Plant Physiology

Schuster, Gadi; Lisitsky, Irena; Klaff, Petra

doi: 10.1104/pp.120.4.937pmid: 10444076

Chloroplast development is characterized by the synthesis and assembly of the photosynthetic complexes of the thylakoid membranes. This maturation process requires the coordinated expression of many nuclear- and chloroplast-encoded genes. As chloroplasts are semiautonomous organelles, they possess their own genome with its inherent transcriptional and translational machinery. However, nuclear-encoded gene products are also necessary for all of the processes occurring in the chloroplast. In this Update we will focus on the control of chloroplast gene expression at the posttranscription stage. Posttranscriptional processes are widely used in controlling the steady-state levels of plastid mRNAs, and are mediated mainly by nuclear-encoded proteins, suggesting a way in which the nucleus can modulate gene expression in the chloroplast. For example, mutants of nuclear-encoded genes affecting the accumulation of specific chloroplast transcripts have been described in maize, Arabidopsis, and the green algae Chlamydomonas reinhardtii(Goldschmidt-Clermont, 1998). mRNAs of higher plant and green algae chloroplasts are transcribed as precursor RNAs that undergo a variety of maturation events, includingcis- and trans-splicing, cleavage of polycistronic messages, processing of 5′ and 3′ ends, and RNA editing (Fig. 1). A general feature of plastid protein-encoding genes is the presence of inverted repeat sequences in the 3′-UTR, which form a stem-loop secondary structure when transcribed to RNA. The 3′ end of mature chloroplast mRNAs is located several nucleotides 3′ to this stem-loop structure (Fig. 1). Contrary to similar structures found in bacterial mRNA, these elements do not function as transcriptional terminators in chloroplasts; instead, they serve as RNA-processing elements capable of stabilizing upstream RNA fragments in vivo and in vitro (Barkan and Stern, 1998). Fig. 1. Open in new tabDownload slide A, RNA metabolism in the chloroplast. Schematic representation of a monocistronic transcription unit in the chloroplast. UTRs are marked with a thin black line; sequences that code for amino acids with a thick black line; and introns with a thick white line. The inverted repeats characterizing the 3′-UTR are symbolized by two arrowheads. A tRNA gene usually presenting 3′ to the inverted repeats is shown as a dashed line. B, The precursor RNA transcribed from this transcription unit undergoes 5′- and 3′-end processing to generate the 5′ and 3′ ends of the mRNA, respectively. C, The 3′ end is located several nucleotides 3′ of the stem-loop structure formed by the inverted-repeats sequence. The introns are removed by splicing, and the tRNA gene is processed by RNaseP. The mRNA is then translated and later degraded. Fig. 1. Open in new tabDownload slide A, RNA metabolism in the chloroplast. Schematic representation of a monocistronic transcription unit in the chloroplast. UTRs are marked with a thin black line; sequences that code for amino acids with a thick black line; and introns with a thick white line. The inverted repeats characterizing the 3′-UTR are symbolized by two arrowheads. A tRNA gene usually presenting 3′ to the inverted repeats is shown as a dashed line. B, The precursor RNA transcribed from this transcription unit undergoes 5′- and 3′-end processing to generate the 5′ and 3′ ends of the mRNA, respectively. C, The 3′ end is located several nucleotides 3′ of the stem-loop structure formed by the inverted-repeats sequence. The introns are removed by splicing, and the tRNA gene is processed by RNaseP. The mRNA is then translated and later degraded. Different aspects of chloroplast mRNA processing and stability have been the subject of recently published reviews (Barkan and Stern, 1998;Drager and Stern, 1998; Goldschmidt-Clermont, 1998; Nickelsen, 1998) and will not be discussed here. We will focus on recent discoveries concerning the molecular mechanism of mRNA polyadenylation and degradation in the chloroplast, and the proteins involved. HOW CAN mRNA DEGRADATION CONTROL GENE EXPRESSION DURING CHLOROPLAST DEVELOPMENT? During leaf development and plastid differentiation, the levels of many plastid mRNAs vary dramatically. The concentration of a specific mRNA is determined by its transcription rate in comparison with its degradation rate. Run-on experiments revealing relative rates of RNA synthesis showed that the different mRNA steady-state levels cannot be related to gene-specific transcriptional activity (Gruissem, 1989). Therefore, changes in the degradation rate of specific mRNAs during chloroplast development occur. To study the degradation of mRNAs in the chloroplast of higher plants in vivo, spinach and barley plants were treated with transcription inhibitors. The rate of decay of chloroplast-encoded mRNAs was determined by quantitative northern analysis using gene-specific probes. The half-life of several chloroplast mRNAs, such as psbA, changed during development from young to mature leaves, whereas that of others, such asrbcL, did not. These results showed that differential mRNA stability contributes to chloroplast mRNA concentrations during leaf development (Klaff and Gruissem, 1991; Kim et al., 1993). Analysis of the transcription, translation, and mRNA levels of 15 plastid genes during barley chloroplast development revealed a dynamic modulation of gene expression and mRNA stability (Rapp et al., 1992). Furthermore, enhanced levels of psbA mRNA in mature barley chloroplasts were due primarily to its selective stabilization. Although data about the RNA elements and the proteins involved in this process are slowly emerging (see below), the precise mechanism by which the stability of a specific chloroplast mRNA changes during plant development and in response to physiological changes (such as light intensity or quality) is still not understood. cis-REGULATORY ELEMENTS CONFERRING STABILITY AND/OR INSTABILITY OF CHLOROPLAST mRNAs A central issue for understanding the regulation of mRNA stability is the identification of cis-regulatory elements. These elements are defined by their ability to transfer properties of a certain mRNA to a reporter gene. In general, this is achieved by constructing chimeric genes consisting of the respective RNA element and a reporter gene, such as the Escherichia coli GUS gene, and transforming these constructs into a recipient organism. Chloroplasts of the green algae C. reinhardtii have been routinely transformed, and, recently, techniques for higher plant chloroplast transformation have become available. Nevertheless, most information on cis-regulatory elements in chloroplasts is still derived from C. reinhardtii. In addition, a wide variety of nuclear mutants affecting mRNA stability have been described for C. reinhardtii and higher plants, providing an additional genetic tool for understanding mRNA-degradation pathways (for review, see Goldschmidt-Clermont, 1998). A well-characterized example of an impaired mRNA metabolism in C. reinhardtii is the nuclear mutation nac2-26, which results in the decreased stability of psbD mRNA. Fusing the 5′-UTR of the psbD mRNA to aadA(aminoglycoside adenyltransferase) as a reporter gene showed a destabilized chimeric transcript in the mutant background and normal accumulation in the wild type, reflecting the properties of thepsbD mRNA. These data indicate that the 5′-UTR of the mRNA includes a determinant for psbD mRNA degradation. Instability of the psbD mRNA in the mutant correlates with a 47-kD protein binding to the psbD leader that is present in wild-type C. reinhardtii chloroplasts but not innac2-26 cells, making this protein a candidate for a gene-specific, nuclear-encoded, trans-acting factor that stabilizes the mRNA (Nickelsen et al., 1994). A more complex mechanism for the regulation of mRNA stability has been proposed for the C. reinhardtii petD mRNA. Extensive mutational analysis and experiments using reporter constructs revealed that sequences within the 5′-UTR are essential for translation and affect RNA stability. In all mutants in which translation was compromised, petD mRNA accumulated to a lower level than in wild-type strains, indicating that mRNA stability is not only regulated by RNA-binding proteins but may also be linked to translatability (Sakamoto et al., 1994). The role of nuclear-encoded factors forpetD mRNA stability was confirmed by F16, a nuclear mutant harboring the mutation mcd1-1, which failed to accumulatepetD mRNA. The analysis of this mutant suggested that the mcd1-1 gene product is involved in protein binding the 5′-UTR to prevent digestion of the mRNA by a 5′ to 3′ exonuclease (Drager et al., 1998). In chloroplasts of higher plants information oncis-regulatory elements for mRNA stability has been obtained from mutational studies and from the recent construction of transplastomic tobacco plants. Similar to the situation in C. reinhardtii, no general scheme emerged, but several parameters contributing to the stability of a certain mRNA were observed. Nuclear mutants in which many chloroplast mRNAs were associated with abnormally few ribosomes showed that the level of rbcL mRNA was reduced 4-fold, indicating that the rbcL mRNA is destabilized as a consequence of its decreased polysome association (Barkan, 1993). For the same mRNA, the analysis of constructs consisting of the 5′-UTR fused to a reporter gene showed that mRNA accumulation in the dark is mediated by this region, whereas a leaderless molecule cannot be detected. That study also provided evidence that rbcLmRNA stability is regulated via the 5′-UTR (Shiina et al., 1998). In addition to the 5′-UTR and its role in mRNA accumulation in chloroplasts, the 3′-UTR is also remarkably important. As described above, most chloroplast mRNAs are flanked by a stem-loop structure 3′ of their coding region that takes part in the mature 3′-end processing (for review, see Barkan and Stern, 1998). In addition, these elements are important for impeding the progress of processive exoribonucleases, which can be shown in vitro using a soluble extract from higher plant chloroplasts, and of synthetic RNA fragments as the substrates (Barkan and Stern, 1998; Drager and Stern, 1998). Partial or complete in vivo deletion of the atpB gene stem-loop in transformed C. reinhardtii chloroplasts led to a dramatic decrease in the accumulation of 3′-end-processed mRNA, whereas the transcription rate of this gene remained unaffected. This result indicated that the stem-loop structure is required for the correct 3′-end processing and mRNA accumulation (for review, see Barkan and Stern, 1998; Drager and Stern, 1998). The stem-loop structure can be replaced in vivo by a stretch of 18 guanosines, which also serves as a barrier for a 3′ to 5′ exonuclease in vitro. The correct 3′-end processing of the mRNA mediated by the structural element is nevertheless essential, since it is required for, or strongly stimulates, its translation in C. reinhardtiichloroplasts (Rott et al., 1998). The results summarized in this section suggest that ciselements responsible for the modulation of mRNA stability of specific genes are mainly localized in the 5′-UTR. On the other hand, RNA structural elements located in the 3′-UTR are required for correct processing and are therefore necessary for mRNA stability. We describe below the search for the molecular mechanism of mRNA degradation and the proteins involved. WHICH ARE THE RNA DEGRADATION ENZYMES IN THE CHLOROPLAST? The in vitro RNA processing and degradation system, in which a synthetic RNA is processed or degraded when incubated with soluble chloroplast extract, was utilized to isolate the enzymes involved. The proteins were fractionated using conventional biochemical separation methods, and the purified fractions were analyzed for activity until one or more polypeptides were present. The search for the RNase involved in the 3′-end processing of chloroplast mRNAs yielded 100RNP (100-kD RNA-binding protein; Hayes et al., 1996). Purified 100RNP has biochemical properties similar to PNPase, one of the two exonucleases discovered to date in bacterial cells. Furthermore, the deduced amino acid sequence of the chloroplast 100RNP cDNA was highly homologous to the bacterial PNPase. Does this result imply that the chloroplast RNA processing and degradation system is similar to recently discovered mechanisms in E. coli? (Nierlich and Murakawa, 1996;Carpousis et al., 1999). Together with the discoveries about the mechanisms of mRNA polyadenylation and degradation, which we will describe later, the answer to this question appears to be yes. Nevertheless, unlike bacteria, plastid mRNA metabolism and its associated enzymes are controlled by the nucleus and may be regulated by light or by the redox state of the chloroplast (Hayes et al., 1996). IS THERE A CHLOROPLAST DEGRADOSOME SIMILAR TO BACTERIA? The E. coli RNA degradosome is a multienzyme complex consisting of the exoribonuclease PNPase, the endonuclease RNase E, a DEAD-box ATP-dependent RNA-helicase, and the enzyme enolase (Carpousis et al., 1999). This high-molecular-mass protein complex is important in RNA processing and mRNA degradation in the bacterial cell, since two of its components, PNPase and RNase E, have been shown to be key elements in these processes. The chloroplast 100RNP/PNPase was isolated in a high-molecular-mass complex of about 600 kD. A 67-kD protein cross-reacting with antibodies prepared against RNase E of E. coli and displaying endoribonuclease activity was copurified with that complex (Hayes et al., 1996). Therefore, it is tempting to suggest that a complex similar to the bacterial degradosome exists in the chloroplast, preserving its ancestral prokaryotic origin (Carpousis et al., 1999). In both bacteria and chloroplasts, it appears that not all of the PNPase population is associated with the degradosome (Carpousis et al., 1994; Lisitsky et al., 1997b). The question of whether different forms of the 100RNP/PNPase are involved in each RNA metabolic activity, such as 3′-end processing and degradation, therefore, remains open. In vitro experiments using synthetic RNAs and purified 100RNP/PNPase have shown much higher enzyme activity on polyadenylated RNA (see below). This selectivity to polyadenylated RNA resulted from the high-affinity binding of the 100RNP/PNPase to poly(A) sequences (Lisitsky et al., 1997b). It is interesting that a similar function and mode of action have recently been reported for the E. coli PNPase and for another exoribonuclease, RNase II (Coburn and Mackie, 1996; Lisitsky and Schuster, 1999). Therefore, identification, isolation, and characterization of the other chloroplast exoribonucleases will determine whether the preference for poly(A)-rich RNAs is only intrinsic to the 100RNP/PNPase, or if it is shared by several or all of the chloroplast exonucleases. In addition, a protein complex composed of several RNA-degradation enzymes was recently identified and isolated in yeast, and was named the exosome (Mitchel et al., 1997;Jacobs et al., 1998). Are the bacterial degradosome, the chloroplast degradosome, and the yeast cytoplasm exosome related to each other functionally and/or evolutionarily? This interesting question is now under intensive study. POLYADENYLATION OF mRNA IN EUKARYOTIC CELLS Posttranscriptional addition of a poly(A) tail to the 3′ end of mRNA was first identified and characterized in eukaryotic cells for viral and nuclear-encoded mRNAs. In these cells the poly(A) tails are formed by the addition of about 250 adenylate residues to a 3′ end generated by endonucleolytic cleavage of the precursor RNA. A complex assembly of proteins is required, along with the activity of poly(A) polymerase. The result of this process is that most of the mRNA molecules are polyadenylated (Wahle and Keller, 1996). What is the function of the poly(A) tail? Many studies have revealed that the long poly(A) tail of eukaryotic nuclear-encoded mRNAs is an important determinant of their maturation and initiation of translation. When referring to maturation, we include the transfer of the precursor RNA from the nucleus to the cytoplasm and the determination of stability. In yeast, deadenylation is a major step in the degradation pathway of nuclear-encoded mRNAs. The poly(A) tail and the protein that bound to it were found to be very important for the initiation of the translation process. How can the poly(A) located in the 3′ end of the RNA molecule control the translation starting at the 5′ end? A model invoking mRNA circularization has been proposed whereby the mRNA 5′ and 3′ ends can interact with each other. In this way, the poly(A)-binding protein that is associated with the poly(A) tail stimulates the binding of the 40S ribosomal subunit to mRNA by associating with the translation initiation factor eIF4G, which also binds to eIF4E and the 5′ cap of mRNA. The circularization of the mRNA in this way is required for the initiation of translation (Sachs et al., 1997). IS mRNA POLYADENYLATED IN PROKARYOTE CELLS? For a long time, polyadenylation was believed to be exclusively associated with eukaryotic mRNAs. Other RNAs, such as rRNAs, tRNAs, and RNAs in prokaryotes, were believed not to be polyadenylated. Most of these RNA molecules do not have a poly(A) tail in their 3′ end. Nevertheless, poly(A) tails have recently been detected in bacteria (Sarkar, 1997). The polyadenylated RNA accounts for only a tiny fraction of the population of the same RNA in the cell. This fraction increases severalfold in mutant bacteria cells that lack exoribonuclease(s) activity. On the other hand, in mutants in which RNA polyadenylation was inhibited due to the lack of poly(A) polymerase enzymes, the half-life of the RNA molecules dramatically increased. What do these results suggest? They imply that, unlike the nucleus and cytoplasm of eukaryotic cells, where the poly(A) tail is important for the stability, maturation, and translation of mRNA and the deadenylation of the long poly(A) tail in part of the mRNA degradation pathway, the addition of poly(A) tails in bacterial mRNAs promotes their degradation. Taken together, polyadenylation of RNA molecules in bacteria cells is a part of the molecular mechanism of RNA degradation in bacteria. Is polyadenylation required for RNA degradation, or is there an additional nonpolyadenylated-dependent degradation pathway in the cell? Is the polyadenylation-dependent degradation pathway specific to certain types of RNA molecules such as mRNA? What is the sequence of events in this RNA-degradation pathway and what is the rate-limiting step? These questions are now under intensive investigation (Blum et al., 1999). IS RNA POLYADENYLATED IN THE CHLOROPLAST? Chloroplasts evolved from free-living prokaryotes that were introduced into eukaryotic cells in an endosymbiotic event(s). Many characteristics of the gene expression machinery of the chloroplast resemble those of the bacteria. However, some characteristics of the gene expression apparatus in the chloroplast are similar to the eukaryotic system. For example, chloroplast genes are usually interrupted with introns that have not been found in bacteria. The question then arose, does RNA polyadenylation occur in the chloroplast? Is it similar to the eukaryotic nuclear-encoded genes, such as bacteria, or is it a unique feature? It is interesting to note that poly(A) RNA was detected in the chloroplast more than 20 years ago (Haff and Bogorad, 1976). Using hybridization experiments with ctDNA and125I-labeled RNA from maize seedlings, it was determined that about 6% of the poly(A)-containing RNA hybridized to ctDNA, and that the chloroplast poly(A) tracts averaged about 45 nucleotides in length. Nevertheless, like the situation in the polyadenylation of bacterial RNA, polyadenylation has long been regarded as a feature of eukaryotic nuclear and viral mRNAs. However, polyadenylation of prokaryotic and organellar mRNAs has recently returned as the focus of research as part of the mRNA degradation mechanism. To detect polyadenylated RNA in the chloroplast, the powerful method of RT-PCR was used. RNA was isolated from purified chloroplasts and oligo(dT)-primed cDNA was synthesized from the polyadenylated RNA molecules. The cDNA corresponding to a specific gene was PCR amplified using a gene-specific primer on one side and a (dT)n tail on the other (Kudla et al., 1996;Lisitsky et al., 1996). Analyzing the nucleotide sequences of the poly(A) tails of chloroplast RNAs revealed several interesting features. First, compared with poly(A) tails in bacteria and yeast, the chloroplast poly(A) tails are very long. Several tails of 270 nucleotides were detected, compared with only 40 to 60 nucleotides in bacteria and yeast. Second, unlike eukaryotic nuclear-encoded and bacterial RNAs, the poly(A) moiety in the chloroplast was not found to be a ribohomopolymer of adenosine residues, but rather was composed of clusters of adenosines mostly bound by guanosines, and, on rare occasions, by cytidines and uridines. A chloroplast poly(A)-rich tail usually contains 70% adenosines, 25% guanosines, and 5% cytidines and uridines, making them purine-rich sequences (Lisitsky et al., 1996). Why are the poly(A) tails heterologous in the chloroplast but not in nuclear-encoded or bacteria RNA? Is there a biological function for this heterogeneity, or does it just reflect less specificity of the chloroplast poly(A) polymerase enzyme? Currently, we do not know the answers to these questions. Other nucleotides in addition to adenosine were recently found in RNA isolated from E. coli cells under stationary growth conditions (Cao and Sarkar, 1997). The third phenomenon is related to the location of the polyadenylation sites in the RNA molecule and will be discussed below. Two possibilities for the formation of polyadenylated sites are truncated transcription termination and cleavage of mature, full-length RNA. Most of the polyadenylation sites that were found by RT-PCR of oligo(dT)-primed chloroplast RNA were localized within the amino acid-coding region of the mRNA. RT-PCR clones of mRNA polyadenylated at the 3′ end were also obtained, but were at least 50 times lower in frequency (Lisitsky et al., 1996). This result indicated that most truncated mRNAs are polyadenylated. How did the truncated RNA molecules undergoing the addition of poly(A)-rich tails originate? The truncation may originate from either early transcription termination or cleavage of a full-length transcript. Two observations suggest that in vivo polyadenylation occurs subsequent to the cleavage of an mRNA as part of the specific degradation pathway. First, five of the polyadenylation sites mapped by RT-PCR in spinachpsbA mRNA perfectly matched endonucleolytic cleavage sites that were mapped by a primer extension in the lysed chloroplast mRNA-degradation system (Fig. 2) (Lisitsky et al., 1996). Since mapping the endonucleolytic cleavage site by primer extension marks the first nucleotide at the 5′ end of the distal cleavage product, and the poly(A) tail is added to the 3′ end of the proximal product, the nucleotide labeled by primer extension was the adjusted nucleotide 3′ to the polyadenylated one (Fig. 2). Second, polyadenylation sites located in the 3′-UTR of thepetD mRNA determined by RT-PCR were mapped to the positions observed as cleavage sites of the purified endoribonuclease p67 on synthetic transcribed RNA corresponding to the petD 3′-UTR (Kudla et al., 1996). These results strongly suggest that most of the polyadenylated mRNAs result from endonucleolytic cleavage of full-length transcripts. Nevertheless, the possibility cannot be ruled out that some of the polyadenylated RNA molecules in the chloroplast are the result of polyadenylation of truncated transcribed molecules. These molecules must immediately enter the degradation pathway, because otherwise they might serve as the templates for the translation of truncated, defective proteins. Full-length polyadenylated transcripts were also detected, albeit in very small amounts relative to the nonpolyadenylated or truncated polyadenylated transcripts. Fig. 2. Open in new tabDownload slide Experimental design used to analyze polyadenylation of endonucleolytically cleaved mRNA. The mRNA 3′- and 5′-UTRs are symbolized by thin lines and the coding region by a thick black line. RT-PCR of purified chloroplast RNA using the primers p1 and p2 revealed clones whereby the poly(A)-rich sequences were added to the last 3′ nucleotide of the proximal cleavage product (N1). The 5′ nucleotide of the distal cleavage product (N2) was determined by primer extension analysis (using the primer p3) of endonucleolytically cleaved RNA in the lysed chloroplast mRNA degradation system (see text). N2 was found to be one nucleotide 3′ to N1. Therefore, polyadenylation occurs at the 3′ end of the proximal endonucleolytic cleavage product. Fig. 2. Open in new tabDownload slide Experimental design used to analyze polyadenylation of endonucleolytically cleaved mRNA. The mRNA 3′- and 5′-UTRs are symbolized by thin lines and the coding region by a thick black line. RT-PCR of purified chloroplast RNA using the primers p1 and p2 revealed clones whereby the poly(A)-rich sequences were added to the last 3′ nucleotide of the proximal cleavage product (N1). The 5′ nucleotide of the distal cleavage product (N2) was determined by primer extension analysis (using the primer p3) of endonucleolytically cleaved RNA in the lysed chloroplast mRNA degradation system (see text). N2 was found to be one nucleotide 3′ to N1. Therefore, polyadenylation occurs at the 3′ end of the proximal endonucleolytic cleavage product. IS POLYADENYLATION REQUIRED FOR CHLOROPLAST mRNA DEGRADATION? The question of whether polyadenylation is required for mRNA degradation, or whether other decay pathways also exist for chloroplast mRNAs, was approached via studies using the lysed-chloroplast system and the polyadenylation inhibitor 3′-dATP (cordycepin triphosphate). Blocking the polyadenylation of RNA inhibits RNA degradation and has similar, if not identical, effects as the direct blocking of the exonucleases by the addition of excess yeast tRNA (Fig.3) (Lisitsky et al., 1997a). In both treatments the full-length mRNA was endonucleolytically cleaved to distinct degradation products that accumulated instead of being exonucleolytically degraded. Therefore, the addition of poly(A)-rich sequences to the endonucleolytic cleavage products of mRNA is required to target these molecules for rapid exonucleolytic degradation in the chloroplast. A system that degrades mRNA without the addition of poly(A)-rich sequences to the endonucleolytic cleavage product either does not exist or was inactive under this experimental system. Fig. 3. Open in new tabDownload slide Similar inhibition of mRNA degradation by blocking polyadenylation and inhibiting the exoribonucleases. Lysed chloroplasts were incubated for 0 or 60 min in the presence or absence (−) of 2 mm 3′-dATP (cordycepin triphosphate) or 0.5 mg/mL yeast tRNA (tRNA), as specified in the figure. RNA was extracted from equal amounts of lysed chloroplasts and analyzed by an RNA gel blot that was hybridized with a psbA-specific probe. The migration of DNA size markers is indicated on the left as the number of nucleotides (nt). Fig. 3. Open in new tabDownload slide Similar inhibition of mRNA degradation by blocking polyadenylation and inhibiting the exoribonucleases. Lysed chloroplasts were incubated for 0 or 60 min in the presence or absence (−) of 2 mm 3′-dATP (cordycepin triphosphate) or 0.5 mg/mL yeast tRNA (tRNA), as specified in the figure. RNA was extracted from equal amounts of lysed chloroplasts and analyzed by an RNA gel blot that was hybridized with a psbA-specific probe. The migration of DNA size markers is indicated on the left as the number of nucleotides (nt). THE BIOCHEMISTRY OF POLYADENYLATION IS ELUCIDATED USING AN IN VITRO SYSTEM An in vitro polyadenylation system was used to elucidate the biochemistry of mRNA polyadenylation and degradation activities, to isolate the proteins involved, and to reconstitute their activities. In vitro-transcribed RNAs corresponding to the chloroplast RNAs could be polyadenylated at their 3′ end using a soluble chloroplast protein extract complemented by the addition of ATP (Lisitsky et al., 1996). In vitro analysis of chloroplast polyadenylation activity revealed specificity to ATP and GTP, reflecting the composition of the poly(A)-rich tails observed by RT-PCR described above. In this respect, it is interesting to note that poly(A)- and poly(G)-polymerase activities were purified from wheat chloroplasts 25 years ago (Burkard and Keller, 1974). Furthermore, in vitro polyadenylation activity is dependent on the substrate structure. Unstructured RNAs were polyadenylated in a highly efficient manner compared with those molecules forming the stem-loop structure characteristic of the mature plastid mRNA 3′ end (Lisitsky et al., 1996). Again, this observation is in agreement with the RT-PCR clones obtained, where polyadenylated RNA molecules at the mature 3′ end (characterized by a stem-loop structure) were found 50 less times than those at the middle of the RNA molecule. CAN THE POLYADENYLATED-DEPENDENT DEGRADATION OF CHLOROPLAST RNA BE MIMICKED IN VITRO? RNA can be synthesized in a test tube using a DNA template of the corresponding nucleotide sequence of the interested gene and an RNA polymerase from a bacteriophage. RNA degradation by chloroplast proteins can be followed by incubating a chloroplast soluble protein extract harboring the ribonucleases, as well as the other RNA-binding proteins and components of this process, together with a radioactive-labeled synthetic RNA. In vitro-transcribed, synthetic polyadenylated RNA was rapidly degraded compared with the same nonpolyadenylated RNA incubated in a soluble chloroplast protein extract. Competition experiments revealed that polyadenylated RNA molecules are more efficient competitors for the degradation machinery than nonpolyadenylated molecules. These results suggest that poly(A)-rich tails play a major role in the rapid degradation of intermediate products of mRNA decay in the chloroplast by targeting the cleavage products for rapid degradation, due to their high affinity to chloroplast exonuclease(s) (Kudla et al., 1996; Lisitsky et al., 1996,1997b). A possible scenario for the situation in the chloroplast is that the relative concentration of the exonucleases is such that they are all occupied in degrading polyadenylated endonucleolytic cleavage products. In this scenario, only polyadenylated RNA molecules will be degraded, as was described above for the experiments using polyadenylation inhibition in the lysed chloroplast system. WHAT IS THE BIOCHEMICAL MECHANISM OF PREFERENTIAL DEGRADATION OF POLYADENYLATED RNA? The 100RNP/PNPase discussed above, similar to the bacterial PNPase, is a processive exoribonuclease binding to the 3′ end, digesting the RNA nucleotide by nucleotide without dissociating from the molecule. As described above, this protein could be obtained as a purified single polypeptide. Therefore, it was possible to determine whether the purified enzyme would retain the preferential degradation activity to polyadenylated RNA observed with the chloroplast protein extract. The other possibility was that other auxiliary proteins are required. In competition experiments using isolated, purified 100RNP/PNPase, the polyadenylated RNA competed with the nonpolyadenylated RNA for the exonuclease, as shown for the soluble protein extract (Lisitsky et al., 1997b). The results implied that competition for polyadenylated RNA is an intrinsic phenomenon of the enzyme as one polypeptide. Therefore, competition for polyadenylated RNA does not depend on the association with the multiprotein complex, the degradosome described above. Is the enzyme's preferred activity to polyadenylated RNA due to the higher binding activity of this protein to a poly(A)-rich sequence or to the faster degradation activity of polyadenylated RNA? Affinity-binding assays of the 100RNP/PNPase to poly(A), as well as to other RNA molecules, displayed higher binding affinity of this protein to poly(A) than to other RNA molecules (Lisitsky et al., 1997b). On the other hand, the degradation rate was similar for all RNA molecules examined. These results suggest that the preferential degradation of polyadenylated RNA in the chloroplast is based on the exoribonuclease 100RNP/PNPase's high binding affinity to the poly(A) sequence. This polypeptide possibly harbors a poly(A) high-affinity binding site in addition to the RNA degradation active site. In addition, the possibility should be emphasized that another, as-yet-unidentified chloroplast exoribonuclease(s) also binds polyadenylated RNA with higher affinity than nonpolyadenylated RNA. Higher in vitro degradation activity of bacterial RNase II to polyadenylated RNA was recently detected (Coburn and Mackie, 1996). Similarly, the E. coli PNPase was recently found to bind polyadenylated RNA with higher affinity than other RNA molecules (Lisitsky and Schuster, 1999). THE MOLECULAR MECHANISM OF mRNA DEGRADATION IN THE CHLOROPLAST Our recent model of the mRNA degradation pathway in the chloroplast is presented in Figure 4. The initial event is endonucleolytic cleavage(s) producing RNA molecules with no stem-loop structure at the 3′ end (Fig. 4B). RNAs ending in a stem-loop structure were poorly polyadenylated in vitro, and RT-PCR clones of poly(A)-rich sequences at the end of the mRNA molecule (characterized by a stem-loop structure) were obtained with much lower frequency than those having the additional site inside the coding region (Lisitsky et al., 1996). Therefore, we suggest that the stem-loop structure characterizing most of the chloroplast mRNAs, and shown to be an effective 3′-end processing signal, also serves as a poor polyadenylation site for preventing exonucleolytic degradation of the functional molecule. Following the endonucleolytic cleavage(s), the proximal fragments were polyadenylated by the addition of poly(A)-rich sequences (Fig. 4C). This stage is inhibited by 3′-dATP (cordycepin) and is required for the continuation of mRNA degradation in the chloroplast. Due to the higher affinity of this enzyme(s) to the poly(A)-rich sequence, the RNAs were rapidly digested only following polyadenylation (Fig. 4D). This last stage can be slowed down by the addition of yeast tRNA (Lisitsky et al., 1997a). Fig. 4. Open in new tabDownload slide A model for the degradation pathway of mRNA in the chloroplast. A, Schematic representation of the psbAmRNA molecule. The white box represents the amino acid-coding region, and the stem-loop structures represent inverted repeats in the 5′- and 3′-UTRs that potentially form stem-loop structures. B, The initial step in the mRNA degradation process is suggested to be endonucleolytically cleaved by an as-yet-unidentified endonuclease(s). The endonuclease is schematically symbolized by scissors. C, A poly(A)-rich tail, which can be up to several hundred nucleotides in length, is then added to the 3′ end of the 5′ endonucleolytically cleaved product. D, The polyadenylated RNA molecule is rapidly degraded by an exonuclease(s), possibly the 100RNP/PNPase. Fig. 4. Open in new tabDownload slide A model for the degradation pathway of mRNA in the chloroplast. A, Schematic representation of the psbAmRNA molecule. The white box represents the amino acid-coding region, and the stem-loop structures represent inverted repeats in the 5′- and 3′-UTRs that potentially form stem-loop structures. B, The initial step in the mRNA degradation process is suggested to be endonucleolytically cleaved by an as-yet-unidentified endonuclease(s). The endonuclease is schematically symbolized by scissors. C, A poly(A)-rich tail, which can be up to several hundred nucleotides in length, is then added to the 3′ end of the 5′ endonucleolytically cleaved product. D, The polyadenylated RNA molecule is rapidly degraded by an exonuclease(s), possibly the 100RNP/PNPase. The model suggests two mechanisms for modulating the half-life of a particular RNA molecule. The first implies that once the mRNA molecule is endonucleolytically cleaved, it is targeted for the degradation process. In this model the rate-limiting step is the initial endonucleolytic cleavage. Once this has occurred, cleaved mRNA will be rapidly polyadenylated and exonucleolytically degraded. In such a mechanism, the nature, specificity, and modulation of activity and/or expression of the endonuclease(s) determine the half-life of a particular mRNA molecule, and the activities of the poly(A) polymerase and exonuclease do not control the rate-limiting step. The second model suggests that other steps in the degradation pathway could limit the degradation rate. It is still unknown if the poly(A) tail length, which can amount to several hundred nucleotides, influences the degradation rate. Furthermore, the guanosine residues characteristic of the poly(A) tails of chloroplast psbA mRNA may be involved in modulating the activity of the respective enzyme. On the other hand, they may simply reflect the specificity of the poly(A) polymerase(s) (Burkard and Keller, 1974; Lisitsky et al., 1996). In vitro experiments in which synthetic transcribed RNA with poly(A) tails of different lengths and different proportions of guanosine residues were incubated with chloroplast protein extract revealed remarkable differences in degradation rates. The significance of these results in relation to the in vivo situation is still unclear and awaits further investigation. So far, our model does not explain how the distal endonucleolytic cleavage product is degraded. One possibility is that many endonucleolytic cleavages along the mRNA occur until the small RNA fragment representing the 3′-UTR is degraded by the polyadenylation-dependent pathway, possibly by demolishing the stem-loop structure, thereby enabling polyadenylation. A low degree of polyadenylation of the mature 3′ end has already been obtained in vivo and in vitro (Lisitsky et al., 1996). Moreover, an additional exonucleolytic degradation pathway in the chloroplast that is independent of polyadenylation may exist. For example, evidence of 5′ to 3′ exonuclease activity was recently obtained in C. reinhardtii chloroplasts (Drager et al., 1998). Whether or not an additional degradation pathway exists, the results of the experiments using the polyadenylation inhibitor 3′-dATP (cordycepin) indicated that the rbcL and psbA mRNAs are exonucleolytically degraded only in the polyadenylation-dependent degradation pathway. To understand the role of this degradation pathway in the developmental regulation of plastid mRNA stability, the degradation pathway and the necessity for polyadenylation in etioplasts and root amyloplasts has to be determined. In these developmental stages the psbA andrbcL RNAs have short half-lives and are rapidly degraded. Over the past few years, our understanding of the chloroplast mRNA degradation pathway has progressed significantly. For certain mRNAs, such as psbA, the different steps of its specific decay have been revealed. The succession of endonucleolytic degradation events is reminiscent of that of the prokaryotic ancestor. This is supported by the observation that one of the enzymes involved has been identified as sharing structural and functional homology with its prokaryotic counterpart. The chloroplast, however, is in part regulated by the nucleus and by external stimuli such as light. Moreover, the longevity of its mRNA better reflects the properties of plants than those of bacteria. Therefore, the chloroplast probably adopted an intermediate position by combining these different features. Following the biochemical pathway of plastid mRNA degradation, research will continue on the regulation of mRNA stability as part of the regulatory network determining leaf development and adaptation to environmental conditions. ACKNOWLEDGMENTS This paper is dedicated to Irena Lisitsky, who passed away on October 21, 1998. We would like to thank the members of the P.K. and G.S. laboratories for numerous helpful comments and suggestions. We would also like to thank Prof. Riesner for his continuous support. Abbreviations: PNPase polynucleotide phosphorylase RT reverse transcription LITERATURE CITED 1 Barkan A Nuclear mutants of maize with defects in chloroplast polysome assembly have altered chloroplast RNA metabolism. Plant Cell 5 1993 389 402 Google Scholar Crossref Search ADS PubMed WorldCat 2 Barkan A, Stern D (1998) Chloroplast mRNA processing: intron splicing and 3′-end metabolism. In J Bailey-Serres, DR Gallie, eds, A Look Beyond Transcription: Mechanisms Determining mRNA Stability and Translation in Plants. American Society of Plant Physiologists, Rockville, MD, pp 162–173 3 Blum E Carpousis AJ Higgins CF Polyadenylation promotes degradation of 3′-structured RNA by the Escherichia coli mRNA degradosome in vitro. J Biol Chem 274 1999 4009 4016 Google Scholar Crossref Search ADS PubMed WorldCat 4 Burkard G Keller EB Poly(A) polymerase and poly(G) polymerase in wheat chloroplasts. Proc Natl Acad Sci USA 71 1974 389 393 Google Scholar Crossref Search ADS PubMed WorldCat 5 Cao G Sarkar N Stationary phase-specific mRNAs in Escherichia coli are adenylated. Biochem Biophys Res Commun 239 1997 46 50 Google Scholar Crossref Search ADS PubMed WorldCat 6 Carpousis AJ Van Houwe G Ehretsmann C Krisch HM Copurification of E. coli RNAase E and PNPase: evidence for a specific association between two enzymes important in RNA processing and degradation. Cell 76 1994 889 900 Google Scholar Crossref Search ADS PubMed WorldCat 7 Carpousis AJ Vanzo NF Raynal LC mRNA degradation, a tale of poly(A) and multiprotein machines. Trends Genet 15 1999 24 28 Google Scholar Crossref Search ADS PubMed WorldCat 8 Coburn GA Mackie GA Differential sensitivities of portions of the mRNA for ribosomal protein S20 to 3′-exonucleases dependent on oligoadenylation and RNA secondary structure. J Biol Chem 271 1996 15776 15781 Google Scholar Crossref Search ADS PubMed WorldCat 9 Drager RG Girard-Bascou J Choquet Y Kindle KL Stern DB In vivo evidence for 5′→3′ exoribonuclease degradation of an unstable chloroplast mRNA. Plant J 13 1998 85 96 Google Scholar Crossref Search ADS PubMed WorldCat 10 Drager RG Stern DB Chloroplast RNA synthesis and processing. Rochaix J-D Goldschmidt-Clermont M Merchant S Molecular Biology of Chlamydomonas: Chloroplasts and Mitochondria, Vol 7. 1998 165 140 Kluwer Academic Publishers Dordrecht, The Netherlands 11 Goldschmidt-Clermont M Coordination of nuclear and chloroplast gene expression in plant cells. Int Rev Cytol 177 1998 115 180 Google Scholar Crossref Search ADS PubMed WorldCat 12 Gruissem W Chloroplast gene expression: how plants turn their plastids on. Cell 56 1989 161 170 Google Scholar Crossref Search ADS PubMed WorldCat 13 Haff LA Bogorad L Poly(adenylic acid)-containing RNA from plastids of maize. Biochemistry 15 1976 4110 4115 Google Scholar Crossref Search ADS PubMed WorldCat 14 Hayes R Kudla J Schuster G Gabay L Maliga P Gruissem W Chloroplast mRNA 3′-end processing by a high molecular weight protein complex is regulated by nuclear-encoded RNA binding proteins. EMBO J 15 1996 1132 1141 Google Scholar Crossref Search ADS PubMed WorldCat 15 Jacobs JS Anderson AR Parker RP The 3′ to 5′ degradation of yeast mRNAs is a general mechanism for mRNA turnover that requires the SKI2 DEVH box protein and 3′ to 5′ exonucleases of the exosome complex. EMBO J 17 1998 1497 1506 Google Scholar PubMed OpenURL Placeholder Text WorldCat 16 Kim M Christopher DA Mullet JE Direct evidence for selective modulation of psbA, rpoA, rbcL, and 16S RNA stability during barley chloroplast development. Plant Mol Biol 22 1993 447 463 Google Scholar Crossref Search ADS PubMed WorldCat 17 Klaff P Gruissem W Changes in chloroplast mRNA stability during leaf development. Plant Cell 3 1991 517 530 Google Scholar Crossref Search ADS PubMed WorldCat 18 Kudla J Hayes R Gruissem W Polyadenylation accelerates degradation of chloroplast mRNA. EMBO J 15 1996 7137 7146 Google Scholar Crossref Search ADS PubMed WorldCat 19 Lisitsky I Klaff P Schuster G Addition of poly(A)-rich sequences to endonucleolytic cleavage sites in the degradation of spinach chloroplast mRNA. Proc Natl Acad Sci USA 93 1996 13398 13403 Google Scholar Crossref Search ADS PubMed WorldCat 20 Lisitsky I Klaff P Schuster G Blocking polyadenylation of mRNA in the chloroplast inhibits its degradation. Plant J 12 1997a 1173 1178 Google Scholar Crossref Search ADS WorldCat 21 Lisitsky I Kotler A Schuster G The mechanism of preferential degradation of polyadenylated RNA in the chloroplast: the exoribonuclease 100RNP/PNPase displays high binding affinity for poly(A) sequence. J Biol Chem 272 1997b 17648 17653 Google Scholar Crossref Search ADS WorldCat 22 Lisitsky I Schuster G Preferential degradation of polyadenylated and polyuridinylated RNAs by bacterial exoribonuclease polynucleotide phosphorylase (PNPase). Eur J Biochem 261 1999 468 474 Google Scholar Crossref Search ADS PubMed WorldCat 23 Mitchel P Petfalski E Shevchenko A Mann M Tollervey D The exosome: a conserved eukaryotic RNA processing complex containing multiple 3′ to 5′ exoribonucleases. Cell 91 1997 457 466 Google Scholar Crossref Search ADS PubMed WorldCat 24 Nickelsen J Van-Dillewijn J Rahire M Rochaix JD Determinants for stability of the chloroplast psbD RNA are located with in its short leader region in Chlamydomonas reinhardtii. EMBO J 13 1994 3182 3191 Google Scholar Crossref Search ADS PubMed WorldCat 25 Nickelsen N Chloroplast RNA stability. Rochaix J-D Goldschmidt-Clermont M Merchant S Molecular Biology of Chlamydomonas: Chloroplasts and Mitochondria, Vol 7. 1998 151 163 Kluwer Academic Publishers Dordrecht, The Netherlands 26 Nierlich DP Murakawa GJ The decay of bacterial messenger RNA. Prog Nucleic Acid Res Mol Biol 52 1996 153 216 Google Scholar Crossref Search ADS PubMed WorldCat 27 Rapp JC Baumgartner BJ Mullet J Quantitative analysis of transcription and RNA levels of 15 barley chloroplast genes: transcription rates and mRNA levels vary over 300-fold; predicted mRNA stabilities vary 30-fold. J Biol Chem 267 1992 21404 21411 Google Scholar Crossref Search ADS PubMed WorldCat 28 Rott R Levy H Drager RG Stern DB Schuster G 3′-Processed mRNA is preferentially translated in Chlamydomonas reinhardtii chloroplasts. Mol Cell Biol 18 1998 4605 4611 Google Scholar Crossref Search ADS PubMed WorldCat 29 Sachs AB Sarnow P Hentze MW Starting at the beginning, middle, and end: translation initiation in eukaryotes. Cell 89 1997 831 838 Google Scholar Crossref Search ADS PubMed WorldCat 30 Sakamoto W Chen X Kindle KL Stern DB Function of the Chlamydomonas reinhardtii petD 5′ untranslated region in regulating the accumulation of subunit IV of the cytochrome b 6/f complex. Plant J 6 1994 503 512 Google Scholar Crossref Search ADS PubMed WorldCat 31 Sarkar N Polyadenylation of mRNA in prokaryotes. Annu Rev Biochem 66 1997 173 197 Google Scholar Crossref Search ADS PubMed WorldCat 32 Shiina T Allison L Maliga P rbcL transcript levels in tobacco plastids are independent of light: reduced dark transcription rates is compensated by increased mRNA stability. Plant Cell 10 1998 1713 1722 Google Scholar Crossref Search ADS PubMed WorldCat 33 Wahle W Keller W The biochemistry of polyadenylation. Trends Biochem Sci 21 1996 247 250 Google Scholar Crossref Search ADS PubMed WorldCat Author notes 1 This work was supported by the Israel Science Foundation (administered by the Israel Academy of Science and Humanities), by a U.S.-Israel Binational Agricultural Research and Development Fund by U.S.-Israel Binational Science Foundation grants to G.S., and by a grant from the Deutsche Forschungsgemeinschaft to P.K. * Corresponding author; e-mail [email protected]; fax 972–4–8225153. Copyright © 1999 American Society of Plant Physiologists This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)

McNeil, Scott D.; Nuccio, Michael L.; Hanson, Andrew D.

doi: 10.1104/pp.120.4.945pmid: 10444077

Osmoprotectants (also termed compatible solutes) occur in all organisms from archaebacteria to higher plants and animals. They are highly soluble compounds that carry no net charge at physiological pH and are nontoxic at high concentrations. Osmoprotectants serve to raise osmotic pressure in the cytoplasm and can also stabilize proteins and membranes when salt levels or temperatures are unfavorable. Osmoprotectants therefore play important roles in the adaptation of cells to various adverse environmental conditions (Yancey, 1994). Chemically, there are three types: betaines and allied compounds, polyols and sugars (e.g. mannitol and trehalose), and amino acids such as Pro. This Updatewill focus on betaines and their biosynthetic pathways. Advances in polyol and Pro research are covered elsewhere (Delauney and Verma, 1993; Stoop et al., 1996). Betaines are amino acid derivatives in which the nitrogen atom is fully methylated, i.e. they are quaternary ammonium compounds. Figure1 shows the structures of the three best-known betaines from plants, Gly betaine, Pro betaine (stachydrine), and β-Ala betaine, as well as choline-O-sulfate and DMSP (Rhodes and Hanson, 1993). The last two compounds are strictly speaking not betaines because DMSP has a tertiary sulfonium in place of a quaternary ammonium group, and choline-O-sulfate has a sulfate ester instead of a carboxyl group—but they are close structural analogs of betaines and have similar chemical and physiological properties. Fig. 1. Open in new tabDownload slide A, Chemical structures of betaines and related compounds that occur in flowering plants and marine algae. B, Concentrations of Gly betaine or DMSP measured in chloroplasts isolated from leaves of salt-tolerant plants grown in nonsaline (black bars) or saline (white bars) conditions. Data are from Robinson and Jones (1986), Génard et al. (1991), and Trossat et al. (1998). Conc., Concentration. Fig. 1. Open in new tabDownload slide A, Chemical structures of betaines and related compounds that occur in flowering plants and marine algae. B, Concentrations of Gly betaine or DMSP measured in chloroplasts isolated from leaves of salt-tolerant plants grown in nonsaline (black bars) or saline (white bars) conditions. Data are from Robinson and Jones (1986), Génard et al. (1991), and Trossat et al. (1998). Conc., Concentration. The compounds in Figure 1 differ in their taxonomic distribution (Blunden and Gordon, 1986; Rhodes and Hanson, 1993). For instance, Gly betaine is widespread among both flowering plants and algae, whereas DMSP is rare in higher plants but common in algae. Certain crop plants such as rice, soybeans, and potatoes lack significant amounts of betaines or any other osmoprotectant. This deficiency is the rationale for recent interest in using metabolic engineering technology to install the synthesis of osmoprotectants in such crops in order to improve their tolerance to drought, salinity, and other stresses. The levels of betaines and other osmoprotectants typically rise during exposure to stresses such as salinity, water deficit, and low temperature because the biosynthetic enzymes are stress induced. Osmoprotectants are largely confined to the cytoplasm (including organelles) and are almost absent from the vacuole, which generally occupies about 90% of the cell volume. For example, the halophyteAtriplex gmelini was found to have 320 mm Gly betaine in the cytoplasm, but only 0.24 mm in the vacuole (Matoh et al., 1987). Isolated chloroplasts of various species have also been shown to contain high concentrations of Gly betaine or DMSP, particularly when isolated from salt-stressed plants; Figure 1B illustrates this point with data from three species. MAGNITUDE AND MECHANISM OF PROTECTIVE EFFECTS The protective properties of betaines were first recognized in experiments in which they were supplied to bacteria whose growth was inhibited by high salt concentrations (Le Rudulier et al., 1984). Typical data for Gly betaine and DMSP are shown in Figure2: In media containing 0.6 mNaCl, the bacteria grow very slowly unless supplied with one of these compounds, which they take up from the medium and accumulate to intracellular levels of >1 m. The physicochemical basis for this striking osmoprotective effect is not fully understood, but there is good evidence that it lies partly in the exclusion of osmoprotectant molecules from the water layer in contact with protein surfaces (Timasheff, 1992). This creates a situation in which native (i.e. folded) protein structures are thermodynamically favored because they present the least possible surface area to the water. Most other solutes such as NaCl or MgSO4 interact directly with protein surfaces and favor unfolding, which leads to denaturation. Osmoprotectants also have cryoprotectant and heat-protectant properties, and exclusion from protein surfaces is probably part of the protective mechanism in these cases as well (Carpenter and Crowe, 1988; Winzor et al., 1992). Fig. 2. Open in new tabDownload slide An illustration of the osmoprotective effect of Gly betaine or DMSP on cells of the bacterium Escherichia coli cultured in a highly saline medium (0.6 mNaCl). Growth in the presence of 1 mm Gly betaine or DMSP was far more rapid than in their absence (control). These compounds accumulated to intracellular levels of >1 m. Bacterial growth was measured by the optical density of the cultures at 420 nm. Data are from Paquet et al. (1994). Fig. 2. Open in new tabDownload slide An illustration of the osmoprotective effect of Gly betaine or DMSP on cells of the bacterium Escherichia coli cultured in a highly saline medium (0.6 mNaCl). Growth in the presence of 1 mm Gly betaine or DMSP was far more rapid than in their absence (control). These compounds accumulated to intracellular levels of >1 m. Bacterial growth was measured by the optical density of the cultures at 420 nm. Data are from Paquet et al. (1994). BIOSYNTHETIC PATHWAYS, ENZYMES, AND GENES Detailed knowledge of biochemical pathways is a prerequisite for metabolic engineering. Biosynthetic routes have now been established for all of the compounds shown in Figure 1. Most of the enzymes participating in these pathways have been identified, and genes for some of them have been cloned. The following is a summary of the current status of knowledge for each compound. Gly Betaine Gly betaine occurs in diverse marine algae and at least 10 flowering plant families, including Chenopodiaceae, Amaranthaceae, Gramineae, Compositae, and Malvaceae (Blunden and Gordon, 1986; Rhodes and Hanson, 1993). Its synthesis has been studied mainly in species of Chenopodiaceae, and to some extent in Amaranthaceae and Gramineae. In these cases, Gly betaine is synthesized via a two-step oxidation of choline (Fig. 3). In Chenopodiaceae and Amaranthaceae, the first step (choline → betaine aldehyde) is mediated by CMO, an unusual Fd-dependent monooxygenase with a Rieske-type iron-sulfur cluster, which has been characterized and cloned. The active enzyme is a dimer or trimer of identical 43-kD subunits and is localized in the chloroplast stroma (Burnet et al., 1995; Rathinasabapathi et al., 1997). Because reduced Fd is generated by photosynthetic electron transport, CMO activity in vivo is strongly light dependent. Both drought and salinity induce CMO expression (Rathinasabapathi et al., 1997; Russell et al., 1998). As CMO has only been found in Chenopodiaceae and Amaranthaceae (Russell et al., 1998), other families may have different choline-oxidizing enzymes that may not be located in plastids. Could these families have choline dehydrogenases or choline oxidases similar to the enzymes known to catalyze choline oxidation in bacteria (Le Rudulier et al., 1984;Hayashi et al., 1997)? Fig. 3. Open in new tabDownload slide Biosynthetic pathways of Gly betaine in Chenopodiaceae and of choline-O-sulfate in Plumbaginaceae. The product of the CMO reaction is the hydrate form of betaine aldehyde. CST, Choline sulfotransferase; PAP, 3′-phosphoadenosine-5′-phosphate; PAPS, 3′-phosphoadenosine-5′-phosphosulfate. Fig. 3. Open in new tabDownload slide Biosynthetic pathways of Gly betaine in Chenopodiaceae and of choline-O-sulfate in Plumbaginaceae. The product of the CMO reaction is the hydrate form of betaine aldehyde. CST, Choline sulfotransferase; PAP, 3′-phosphoadenosine-5′-phosphate; PAPS, 3′-phosphoadenosine-5′-phosphosulfate. The second step of Gly betaine synthesis is catalyzed by BADH (Fig. 3), an NAD-dependent dehydrogenase that has been characterized and cloned from Chenopodiaceae, Amaranthaceae, and Gramineae. BADH is not unique to Gly betaine synthesis; it also attacks DMSP-aldehyde (see below) and seems to be identical to ω-aminoaldehyde dehydrogenase, a ubiquitous enzyme of polyamine metabolism (Trossat et al., 1997). This probably explains why plants with no Gly betaine have some BADH activity (Holmstrom et al., 1994; Rathinasabapathi et al., 1994). BADH is a dimer of identical 54-kD subunits. It is a stromal enzyme in Chenopodiaceae, and, surprisingly, lacks a typical transit peptide. Nonetheless, BADH is efficiently targeted to chloroplasts in transgenic tobacco (Rathinasabapathi et al., 1994). In Gramineae, however, most of the BADH may be peroxisomal (Nakamura et al., 1997), which reinforces the idea that the choline-oxidizing enzyme in this family may not be CMO, as CMO could not function in peroxisomes where there is no reduced Fd. BADH is induced by osmotic stress in both Chenopodiaeae and Gramineae (Rhodes and Hanson, 1993). Choline-O-Sulfate Choline-O-sulfate occurs throughout the Plumbaginaceae (Hanson et al., 1994) and in some marine algae (Blunden and Gordon, 1986). In addition to being an osmoprotectant, it also provides a means of detoxifying sulfate, which is often abundant in saline environments (Hanson et al., 1994). The conversion of choline to choline-O-sulfate is catalyzed by a 3′-phosphoadenosine-5′-phosphosulfate-dependent choline sulfotransferase (Fig. 3) (Rivoal and Hanson, 1994). This sulfotransferase is a soluble enzyme with an extremely high affinity for choline (Km = 5.5 μm), which probably reflects the low cytoplasmic choline levels that prevail in plants (Nuccio et al., 1998). Saline conditions induce sulfotransferase enzyme activity in both leaves and roots (Rivoal and Hanson, 1994). Pro Betaine and β-Ala Betaine β-Ala betaine and Pro betaine are found in some species of Plumbaginaceae; Pro betaine also occurs in Rutaceae, Leguminosae, Compositae, and other families (Rhodes and Hanson, 1993; Hanson et al., 1994; Nolte et al., 1997). Both compounds are also found in marine algae (Blunden and Gordon, 1986) and are known to be formed by methylation of the corresponding amino acid; however, the enzyme(s) involved have yet to be identified (Essery et al., 1962; Hanson et al., 1994). For Pro betaine, in vivo radiolabeling and phytochemical data suggest that the two methylations are catalyzed by different enzymes (Essery et al., 1962; Naidu et al., 1987). DMSP DMSP biosynthesis is important not only in relation to osmoprotection but also because the DMSP produced by marine algae is the precursor of atmospheric dimethylsulfide gas. This gas has a key role in the global sulfur cycle and influences climate (Malin and Kirst, 1997). DMSP is present in diverse marine algae (Blunden and Gordon, 1986; Malin and Kirst, 1997), but has so far been found in only a few flowering plants: Spartinaspp. and sugarcane from the Gramineae and Wollastonia biflora from the Compositae. In all cases DMSP is synthesized from Met. However, conversion of Met to DMSP seems to have evolved independently three times, once in algae and twice in flowering plants, because there are three different pathways (Fig.4). Fig. 4. Open in new tabDownload slide Biosynthetic pathways of DMSP in marine algae and flowering plants. The algal pathway has been found in the green macroalga E. intestinalis and various plankton species. Two related pathways have been found in flowering plants, one in the saltmarsh grass S. alterniflora and the other in the Pacific coastal plant W. biflora. The breakdown of DMSP to give dimethylsulfide (DMS) and acrylate is also shown; this reaction has so far only been demonstrated in marine algae and bacteria. DMSPald, DMSP-aldehyde; D-DMSHB,d-4-dimethylsulfonio-2-hydroxybutyrate; D-MTHB,d-4-methylthio-2-hydroxybutyrate; MAT, Met aminotransferase; MMT, Met S-methyltransferase; MTHBMT, 4-methylthio-2-hydroxybutyrate S-methyltransferase; MTOB, 4-methylthio-2-oxobutyrate; MTOBR, MTOB reductase. Fig. 4. Open in new tabDownload slide Biosynthetic pathways of DMSP in marine algae and flowering plants. The algal pathway has been found in the green macroalga E. intestinalis and various plankton species. Two related pathways have been found in flowering plants, one in the saltmarsh grass S. alterniflora and the other in the Pacific coastal plant W. biflora. The breakdown of DMSP to give dimethylsulfide (DMS) and acrylate is also shown; this reaction has so far only been demonstrated in marine algae and bacteria. DMSPald, DMSP-aldehyde; D-DMSHB,d-4-dimethylsulfonio-2-hydroxybutyrate; D-MTHB,d-4-methylthio-2-hydroxybutyrate; MAT, Met aminotransferase; MMT, Met S-methyltransferase; MTHBMT, 4-methylthio-2-hydroxybutyrate S-methyltransferase; MTOB, 4-methylthio-2-oxobutyrate; MTOBR, MTOB reductase. DMSP Synthesis in Algae The steps involved in DMSP synthesis have been demonstrated in the green macroalga Enteromorpha intestinalis, and appear to be the same in diverse microalgae (Gage et al., 1997). They are: transamination of Met to give the corresponding 2-oxo acid, followed by reduction to a 2-hydroxy acid, then S-methylation and oxidative decarboxylation (Fig. 4). The enzymes catalyzing the first three steps have been demonstrated in vitro (Summers et al., 1998); in vivo 18O2-labeling data indicate that the final step is mediated by an oxygenase (Gage et al., 1997). DMSP Synthesis in Flowering Plants DMSP synthesis in flowering plants has been investigated inW. biflora and S. alterniflora, which have somewhat different pathways. In both species, the first step is conversion of Met to SMM, catalyzed by MetS-methyltransferase, and the last is oxidation of DMSP-aldehyde, catalyzed by BADH (Trossat et al., 1996, 1997; Kocsis et al., 1998). Met S-methyltransferase has been shown to be cytosolic and has been cloned (Trossat et al., 1996; Bourgis et al., 1999). In W. biflora, SMM is apparently converted to DMSP-aldehyde in a single step via an unusual coupled transamination-decarboxylation reaction (Fig. 4) (Rhodes et al., 1997). This conversion is known to occur in the chloroplast (Trossat et al., 1996), but the enzyme(s) involved has not yet been identified. On the other hand, S. alterniflora converts SMM to DMSP-aldehyde in two steps via the intermediate DMSP-amine (Kocsis et al., 1998) (Fig. 4). The enzymes have not yet been isolated, but in vivo radiolabeling evidence points to an SMM-specific decarboxylase plus a specific amine oxidase, dehydrogenase or transaminase (Kocsis et al., 1998). METABOLIC ENGINEERING OF Gly BETAINE SYNTHESIS Gly betaine accumulation has long been a target for engineering stress resistance (Le Rudulier et al., 1984). The idea that introducing the Gly betaine pathway into plants that lack it will enhance their stress tolerance is based both on comparative physiology (Yancey, 1994) and on genetic evidence from a mutation in maize that abolishes Gly betaine synthesis and reduces salt and heat tolerance (Saneoka et al., 1995; Yang et al., 1996). Wide-crossing work on Lophopyrum elongatum and wheat has provided further physiological-genetic evidence that Gly betaine accumulation contributes to salt tolerance (Colmer et al., 1995). Because plants have been found not to catabolize Gly betaine (Rhodes and Hanson, 1993; Nuccio et al., 1998), it has been reasonably assumed that engineering the synthesis of Gly betaine will lead to its accumulation. Several groups have taken the first step toward this goal by expressing choline-oxidizing enzymes from bacteria (Lilius et al., 1996; Hayashi et al., 1997; Alia et al., 1998; Sakamoto et al., 1998) or spinach CMO (Nuccio et al., 1998) in tobacco and other plants that do not contain Gly betaine (“nonaccumulators”). The transgenic plants produced a little Gly betaine and, in some cases, showed small but significant increases in tolerance to various stresses (for review, see Nuccio et al., 1999). These results are encouraging inasmuch as they confirm that Gly betaine synthesis can be engineered. However, the Gly betaine levels obtained to date in transgenic plants (typically about 0.1–1 μmol g−1 fresh weight) are only a small percentage of those in spinach, sugar beet, and other plants that are natural Gly betaine accumulators. The main constraint on Gly betaine production in transgenic plants appears to be the endogenous choline supply, because providing choline exogenously leads to a massive increase in Gly betaine synthesis (Nuccio et al., 1998). It will therefore most likely be necessary to up-regulate the de novo synthesis of choline in order to increase Gly betaine synthesis in nonaccumulators expressing foreign choline-oxidizing enzymes (Nuccio et al., 1998). CONCLUSIONS AND PROSPECTS Characterization and cloning of the enzymes of Gly betaine synthesis has enabled us to start using transgenic plants to understand the role of Gly betaine in stress adaptation. This in turn is helping to define the potential of osmoprotectant engineering in crop improvement. Now that pathways to DMSP have been established, a similar approach could be followed for this compound; the same is true for the other osmoprotectants discussed above. There are two reasons to do this. First, some osmoprotectants may be better than Gly betaine in certain environments, which has important implications for engineering crops. We have already indicated that choline-O-sulfate could be particularly suitable in high-sulfate conditions because its synthesis can detoxify the sulfate anion. DMSP is another example; since it does not require nitrogen to produce, it may be a better choice than Gly betaine for environments that are poor in nitrogen. The second reason to transgenically express enzymes that produce various osmoprotectants is to explore the in vivo control of metabolism, about which we currently know very little. Introducing novel pathways increases the demand for precursors, and quantitative analysis of how this impacts metabolic fluxes, pool sizes, and gene expression can be highly informative in regard to the control architecture of metabolic networks (Bailey, 1991). For instance, because betaines and related compounds require very large amounts of methyl groups relative to primary metabolism, engineering betaine overproduction has the potential to teach us much about how the supply of one-carbon units is regulated. Abbreviations: BADH betaine aldehyde dehydrogenase CMO choline monooxygenase DMSP 3-dimethylsulfoniopropionate DMSP-aldehyde 3-dimethylsulfoniopropionaldehyde DMSP-amine 3-dimethylsulfoniopropylamine SMM S-methylmethionine LITERATURE CITED 1 Alia Hayashi H Sakamoto A Murata N Enhancement of the tolerance of Arabidopsis to high temperatures by genetic engineering of the synthesis of glycinebetaine. Plant J 16 1998 155 161 Google Scholar Crossref Search ADS PubMed WorldCat 2 Bailey JE Toward a science of metabolic engineering. Science 252 1991 1668 1675 Google Scholar Crossref Search ADS PubMed WorldCat 3 Blunden G Gordon SM Betaines and their sulphonio analogues in marine algae. Prog Phycol Res 4 1986 39 80 Google Scholar OpenURL Placeholder Text WorldCat 4 Bourgis F, Roje S, Nuccio ML, Fisher DB, Tarczynski MC, Li C, Herschbach C, Rennenberg H, Pimenta MJ, Shen T-L, Gage D, Hanson AD (1999) S-Methylmethionine has a major role in phloem sulfur transport, and is synthesized by a novel type of methyltransferase. Plant Cell (in press) 5 Burnet M Lafontaine PJ Hanson AD Assay, purification, and partial characterization of choline monooxygenase from spinach. Plant Physiol 108 1995 581 588 Google Scholar Crossref Search ADS PubMed WorldCat 6 Carpenter JF Crowe JH The mechanism of cryoprotection of proteins by solutes. Cryobiology 25 1988 244 255 Google Scholar Crossref Search ADS PubMed WorldCat 7 Colmer TD Epstein E Dvorak J Differential solute regulation in leaf blades of various ages in salt-sensitive wheat and a salt-tolerant wheat × Lophopyrum elongatum (Host) A. Love amphiploid. Plant Physiol 108 1995 1715 1724 Google Scholar Crossref Search ADS PubMed WorldCat 8 Delauney AJ Verma DPS Proline biosynthesis and osmoregulation in plants. Plant J 4 1993 215 223 Google Scholar Crossref Search ADS WorldCat 9 Essery JM McCaldin DJ Marion L The biogenesis of stachydrine. Phytochemistry 1 1962 209 213 Google Scholar Crossref Search ADS WorldCat 10 Gage DA Rhodes D Nolte KD Hicks WA Leustek T Cooper AJL Hanson AD A new route for synthesis of dimethylsulphoniopropionate in marine algae. Nature 387 1997 891 894 Google Scholar Crossref Search ADS PubMed WorldCat 11 Génard H Le Saos J Billard J-P Trémolières A Boucaud J Effect of salinity on lipid composition, glycine betaine content and photosynthetic activity in chloroplasts of Suaeda maritima. Plant Physiol Biochem 29 1991 421 427 Google Scholar OpenURL Placeholder Text WorldCat 12 Hanson AD Rathinasabapathi B Rivoal J Burnet M Dillon MO Gage DA Osmoprotective compounds in the Plumbaginaceae: a natural experiment in metabolic engineering of stress tolerance. Proc Natl Acad Sci USA 91 1994 306 310 Google Scholar Crossref Search ADS PubMed WorldCat 13 Hayashi H Mustardy L Deshnium P Ida M Murata N Transformation of Arabidopsis thaliana with the codA gene for choline oxidase: accumulation of glycinebetaine and enhanced tolerance to salt and cold stress. Plant J 12 1997 133 142 Google Scholar Crossref Search ADS PubMed WorldCat 14 Holmstrom KO Welin B Mandal A Kristiansdottir I Teeri TH Lamark T Strom AR Palva ET Production of the Escherichia coli betaine-aldehyde dehydrogenase, an enzyme required for the synthesis of the osmoprotectant glycine betaine, in transgenic plants. Plant J 6 1994 749 758 Google Scholar Crossref Search ADS PubMed WorldCat 15 Kocsis MG Nolte KD Rhodes D Shen T-L Gage DA Hanson AD Dimethylsulfoniopropionate biosynthesis in Spartina alterniflora: evidence that S-methylmethionine and dimethylsulfoniopropylamine are intermediates. Plant Physiol 117 1998 273 281 Google Scholar Crossref Search ADS PubMed WorldCat 16 Le Rudulier D Strom AR Dandekar AM Smith LT Valentine RC Molecular biology of osmoregulation. Science 224 1984 1064 1068 Google Scholar Crossref Search ADS PubMed WorldCat 17 Lilius G Holmberg N Bülow L Enhanced NaCl stress tolerance in transgenic tobacco expressing bacterial choline dehydrogenase. Bio-Technology 14 1996 177 180 Google Scholar OpenURL Placeholder Text WorldCat 18 Malin G Kirst GO Algal production of dimethyl sulphide and its atmospheric role. J Phycol 33 1997 889 896 Google Scholar Crossref Search ADS WorldCat 19 Matoh T Watanabe J Takahashi E Sodium, potassium, chloride, and betaine concentrations in isolated vacuoles from salt-grown Atriplex gmelini leaves. Plant Physiol 84 1987 173 177 Google Scholar Crossref Search ADS PubMed WorldCat 20 Naidu BP Jones GP Paleg LG Poljakoff-Mayber A Proline analogues in Melaleuca species: response of Melaleuca lanceolata and M. uncinata to water stress and salinity. Aust J Plant Physiol 14 1987 669 677 Google Scholar OpenURL Placeholder Text WorldCat 21 Nakamura T Yokota S Muramoto Y Oguri Y Fukui K Takabe T Expression of a betaine aldehyde dehydrogenase gene in rice, a glycinebetaine nonaccumulator, and possible localization of its protein in peroxisomes. Plant J 11 1997 1115 1120 Google Scholar Crossref Search ADS PubMed WorldCat 22 Nolte KD Hanson AD Gage DA Proline accumulation and methylation to proline betaine in Citrus: implications for genetic engineering of stress resistance. J Am Soc Hortic Sci 122 1997 8 13 Google Scholar Crossref Search ADS WorldCat 23 Nuccio ML, Rhodes D, McNeil SD, Hanson AD (1999) Metabolic engineering of plants for osmotic stress resistance. Curr Opin Plant Biol (in press) 24 Nuccio ML Russell BL Nolte KD Rathinasabapathi B Gage DA Hanson AD The endogenous choline supply limits glycine betaine synthesis in transgenic tobacco expressing choline monooxygenase. Plant J 16 1998 487 498 Google Scholar Crossref Search ADS PubMed WorldCat 25 Paquet L Rathinasabapathi B Saini H Zamir L Gage DA Huang Z-H Hanson AD Accumulation of the compatible solute 3-dimethylsulfoniopropionate in sugarcane and its relatives, but not other gramineous crops. Aust J Plant Physiol 21 1994 37 48 Google Scholar OpenURL Placeholder Text WorldCat 26 Rathinasabapathi B Burnet M Russell BL Gage DA Liao P-C Nye GJ Scott P Golbeck JH Hanson AD Choline monooxygenase, an unusual iron-sulfur enzyme catalyzing the first step of glycine betaine synthesis in plants: prosthetic group characterization and cDNA cloning. Proc Natl Acad Sci USA 94 1997 3454 3458 Google Scholar Crossref Search ADS PubMed WorldCat 27 Rathinasabapathi B McCue KF Gage DA Hanson AD Metabolic engineering of glycine betaine synthesis: plant betaine aldehyde dehydrogenases lacking typical transit peptides are targeted to tobacco chloroplasts where they confer betaine aldehyde resistance. Planta 193 1994 155 162 Google Scholar Crossref Search ADS PubMed WorldCat 28 Rhodes D Gage DA Cooper AJL Hanson AD S-Methylmethionine conversion to dimethylsulfoniopropionate: evidence for an unusual transamination reaction. Plant Physiol 115 1997 1541 1548 Google Scholar Crossref Search ADS PubMed WorldCat 29 Rhodes D Hanson AD Quaternary ammonium and tertiary sulfonium compounds in higher plants. Annu Rev Plant Physiol Plant Mol Biol 44 1993 357 384 Google Scholar Crossref Search ADS WorldCat 30 Rivoal J Hanson AD Choline-O-sulfate biosynthesis in plants: identification and partial characterization of a salinity-inducible choline sulfotransferase from species of Limonium (Plumbaginaceae). Plant Physiol 106 1994 1187 1193 Google Scholar Crossref Search ADS PubMed WorldCat 31 Robinson SP Jones GP Accumulation of glycinebetaine in chloroplasts provides osmotic adjustment during salt stress. Aust J Plant Physiol 13 1986 659 668 Google Scholar OpenURL Placeholder Text WorldCat 32 Russell BL Rathinasabapathi B Hanson AD Osmotic stress induces expression of choline monooxygenase in sugar beet and amaranth. Plant Physiol 116 1998 859 865 Google Scholar Crossref Search ADS PubMed WorldCat 33 Sakamoto A Alia Murata N Metabolic engineering of rice leading to biosynthesis of glycinebetaine and tolerance to salt and cold. Plant Mol Biol 38 1998 1011 1019 Google Scholar Crossref Search ADS PubMed WorldCat 34 Saneoka H Nagasaka C Hahn DT Yang W-J Premachandra GS Joly RJ Rhodes D Salt tolerance of glycinebetaine-deficient and -containing maize lines. Plant Physiol 107 1995 631 638 Google Scholar Crossref Search ADS PubMed WorldCat 35 Stoop JMH Williamson JD Pharr DM Mannitol metabolism in plants: a method for coping with stress. Trends Plant Sci 1 1996 139 144 Google Scholar Crossref Search ADS WorldCat 36 Summers PS Nolte KD Cooper AJL Borgeas H Leustek T Rhodes D Hanson AD Identification and stereospecificity of the first three enzymes of 3-dimethylsulfoniopropionate biosynthesis in a chlorophyte alga. Plant Physiol 116 1998 369 378 Google Scholar Crossref Search ADS WorldCat 37 Timasheff SN A physiochemical basis for selection of osmolytes by nature. Somero GN Osmond CB Bolis CL Water and Life. 1992 70 84 Springer-Verlag Berlin 38 Trossat C Nolte KD Hanson AD Evidence that the pathway of dimethylsulfoniopropionate biosynthesis begins in the cytosol and ends in the chloroplast. Plant Physiol 111 1996 965 973 Google Scholar Crossref Search ADS PubMed WorldCat 39 Trossat C Rathinasabapathi B Hanson AD Transgenically expressed betaine aldehyde dehydrogenase efficiently catalyzes oxidation of dimethylsulfoniopropionaldehyde and ω-aminoaldehydes. Plant Physiol 113 1997 1457 1461 Google Scholar Crossref Search ADS PubMed WorldCat 40 Trossat C Rathinasabapathi B Weretilnyk EA Shen T-L Huang Z-H Gage DA Hanson AD Salinity promotes accumulation of 3-dimethylsulfoniopropionate and its precursor S-methylmethionine in chloroplasts. Plant Physiol 116 1998 165 171 Google Scholar Crossref Search ADS PubMed WorldCat 41 Winzor CL Winzor DJ Paleg LG Jones GP Naidu BP Rationalization of the effects of compatible solutes on protein stability in terms of thermodynamic nonideality. Arch Biochem Biophys 296 1992 102 107 Google Scholar Crossref Search ADS PubMed WorldCat 42 Yancey PH Compatible and counteracting solutes. Strange K Cellular and Molecular Physiology of Cell Volume Regulation. 1994 81 109 CRC Press Boca Raton, FL 43 Yang G Rhodes D Joly RJ Effects of high temperature on membrane stability and chlorophyll fluorescence in glycinebetaine-deficient and glycinebetaine-containing maize lines. Aust J Plant Physiol 23 1996 437 443 Google Scholar OpenURL Placeholder Text WorldCat Author notes 1 This research was supported in part by grants to A.D.H. from the U.S. Department of Agriculture National Research Initiative Competitive Grants Program (no. 98–35100–6149) and the National Science Foundation (nos. IBN–9816075 and IBN–9813999), by an endowment from the C.V. Griffin, Sr., Foundation, and by the Florida Agricultural Experiment Station. This article is journal series no. R–06838. * Corresponding author; e-mail [email protected]; fax 352–392–6479. Copyright © 1999 American Society of Plant Physiologists This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)

Gijzen, Mark; Miller, S. Shea; Kuflu, Kuflom; Buzzell, Richard I.; Miki, Brian L.A.

doi: 10.1104/pp.120.4.951pmid: 10444078

Abstract Soybean (Glycine max[L.] Merr.) hydrophobic protein (HPS) is an abundant seed constituent and a potentially hazardous allergen that causes asthma in persons allergic to soybean dust. By analyzing surface extracts of soybean seeds with sodium dodecyl sulfate-polyacrylamide gel electrophoresis and amino-terminal microsequencing, we determined that large amounts of HPS are deposited on the seed surface. The quantity of HPS present varies among soybean cultivars and is more prevalent on dull-seeded phenotypes. We have also isolated cDNA clones encoding HPS and determined that the preprotein is translated with a membrane-spanning signal sequence and a short hydrophilic domain. Southern analysis indicated that multiple copies of the HPS gene are present in the soybean genome, and that the HPS gene structure is polymorphic among cultivars that differ in seed coat luster. The pattern of HPS gene expression, determined by in situ hybridization and RNA analysis, shows that HPS is synthesized in the endocarp of the inner ovary wall and is deposited on the seed surface during development. This study demonstrates that a seed dust allergen is associated with the seed luster phenotype in soybean and that compositional properties of the seed surface may be altered by manipulating gene expression in the ovary wall. Angiosperm seeds are composite structures that develop from fertilized ovules. The essential components, the embryo, endosperm, and seed coat, each have separate developmental origins and fates in the reproductive cycle. Although these features are common to all angiosperms, seeds from different species follow distinct developmental patterns that produce a vast array of sizes, shapes, colors, textures, and compositions. The development of complex, highly differentiated seed coats is a general feature of legumes and is a characteristic that is often used as an aid for their identification and classification (Corner, 1951;Esau, 1977). At maturity, the seed coat tissues of soybean (Glycine max [L.] Merr.) consist of several cell layers that together constitute 4% to 8% of the seed mass. The color, luster, and permeability of the seed and its resistance to seed-borne diseases are all properties that may be determined by the seed coat and associated tissues. The composition, texture, and nutritional value of food and feed products derived from the seed are also influenced by the presence of the seed coat. For these reasons, we are interested in identifying genes that control seed coat traits. Seeds of Glycine spp. are highly variable in their surface texture and appearance. In many wild species the seed coat is completely obscured by the adherence of endocarp to the seed surface (Wolf and Baker, 1972; Newell and Hymowitz, 1978). Specifically, it is the membranous inner epidermis of the endocarp that detaches from the other tissues of the pericarp, or pod wall, to cover the seed. The presence of adhering endocarp on the seed also occurs in the domesticated soybean and influences the luster of the seed surface. As shown in Table I, many different seed luster phenotypes have been described, including shiny, intermediate, dull, light bloom, bloom, and dense bloom (Bernard and Weiss, 1973; Juvik et al., 1989). Three complementary genes,B1, B2, and B3, have been proposed to control the development of bloom (Woodworth, 1933; Goudong et al., 1987), but there is no genetic model to account for all of the different luster phenotypes observed. For example, most soybean cultivars are described as having either dull or shiny seed coats, yet genetic and biochemical control of this trait remains undetermined. Table I. Seed coat luster phenotypes for soybeana Abbreviation . Phenotype . Description . B Bloom Heavy coating of powdery substance adhering to seed coat Db Dense bloom Heavy bloom Lb Light bloom Slight bloom D Dull Trace amounts of bloom I Intermediate Between dull and shiny S Shiny Absence of bloom Abbreviation . Phenotype . Description . B Bloom Heavy coating of powdery substance adhering to seed coat Db Dense bloom Heavy bloom Lb Light bloom Slight bloom D Dull Trace amounts of bloom I Intermediate Between dull and shiny S Shiny Absence of bloom Soybean seeds show variation in light-reflective and surface properties. F0-a Juvik et al. (1989). Open in new tab Table I. Seed coat luster phenotypes for soybeana Abbreviation . Phenotype . Description . B Bloom Heavy coating of powdery substance adhering to seed coat Db Dense bloom Heavy bloom Lb Light bloom Slight bloom D Dull Trace amounts of bloom I Intermediate Between dull and shiny S Shiny Absence of bloom Abbreviation . Phenotype . Description . B Bloom Heavy coating of powdery substance adhering to seed coat Db Dense bloom Heavy bloom Lb Light bloom Slight bloom D Dull Trace amounts of bloom I Intermediate Between dull and shiny S Shiny Absence of bloom Soybean seeds show variation in light-reflective and surface properties. F0-a Juvik et al. (1989). Open in new tab To begin to resolve these uncertainties, we have compared proteins occurring on the surface of seeds with different luster phenotypes. We show that a previously characterized allergen, HPS, is an abundant seed surface protein associated with dull-seeded phenotypes. We isolated clones encoding cDNA copies of HPS to study the expression and structure of the HPS gene in different seed luster phenotypes. We show that HPS is synthesized in the endocarp and deposited on the seed surface. Furthermore, there is variation in the amount of HPS present among different soybean lines that arises from polymorphic HPS gene structure. Overall, our results suggest a functional role for HPS in influencing the physical properties of the seed surface, and illustrate how seed phenotype and allergenicity are linked. MATERIALS AND METHODS Plant Materials Soybean (Glycine max [L.] Merr.) seed was from the collections at Agriculture and Agri-Food Canada in Harrow and Ottawa, Ontario. Plants were grown in field plots outdoors or in glass-enclosed greenhouses. The Clark isoline L69-4544 is a self-colored (i/i) bloom (B1/B1) genotype, hereafter referred to simply as “Clark B1”. This isoline originated from the U.S. Department of Agriculture Soybean Germplasm Collection and was derived through backcrossing L67-3469 (6) × cv Sooty, where L67-3469 is a self-colored (i/i) Clark mutant. Seed Surface Protein Analysis Seed surface proteins of different soybean cultivars were compared by SDS-PAGE analysis. A single seed was placed in a 2-mL plastic-capped test tube, and surface proteins were extracted by adding 0.5 mL of a buffer-detergent solution containing 10 mm Tris-Cl, pH 7.5, 0.5% (v/v) SDS, and 20 mm DTT and placing the tube in a boiling water bath for 2 min. The contents of the tube were mixed and an aliquot was withdrawn and centrifuged for 5 min at 14,000g. Freshly prepared loading buffer containing 20 mm DTT was added to the sample and proteins were electrophoretically separated on 15% acrylamide gels in the presence of SDS using a modified Laemmli system, as described by Fling and Gregerson (1986). The DTT was omitted from the extraction solution but included in the loading buffer at a range of concentrations to determine its effect on protein migration. Fixation and visualization of the proteins by silver staining followed the method of Blum et al. (1987). Amino-terminal microsequencing of blotted proteins was according to the method of Moos et al. (1988). Isolation of HPS cDNA Clones and DNA Sequencing A seed coat cDNA library was constructed from poly(A+) mRNA isolated from soybean cv Harosoy 63 seeds in the mid to late stage of development (Gijzen, 1997). A sample of the total amplified library was used to subclone inserts from the original vector (Lambda ZAP, Stratagene) into pBK-CMV (Stratagene). Random clones were chosen from this mass excision for plasmid purification and DNA sequencing to establish an expressed sequence tag database of seed coat genes. Automated sequencing of DNA was performed (model 377, Applied Biosystems) using dye-labeled terminators. These DNA sequences were searched for open reading frames encoding HPS by using the BLASTX program at the National Center for Biotechnology Information site (http://www.ncbi.nlm.nih.gov/). DNA and RNA Hybridizations Soybean genomic DNA was isolated from frozen, lyophilized tissue according to the method of Dellaporta et al. (1983). Restriction enzyme digestion of 30 μg of DNA, separation on 0.5% agarose gels, and blotting to nylon membranes followed standard protocols (Sambrook et al., 1989). Digoxigenin-labeled cDNA was prepared and used to probe DNA blots according to the instructions provided by the manufacturer (Boehringer Mannheim). Hybridization was carried out at 65°C for 16 h in 0.25 mNa2HPO4, pH 7.2, 20% (w/v) SDS, 1 mm EDTA, and 0.5% (w/v) blocking reagent (Boehringer Mannheim). Filters were then washed four times for 15 min each at 22°C in high-stringency wash solution (20 mm Na2HPO4, pH 7.2, 1% [w/v] SDS, and 1 mm EDTA), followed by three 15-min washes in the same solution at 68°C. Total RNA was isolated from roots, stems, leaves, flowers, pods, seed coats, and embryos dissected from soybean plants at various stages of development according to published methods (Wang and Vodkin, 1994). Samples of total RNA (10 μg each) were electrophoretically separated in formaldehyde gels and briefly stained with ethidium bromide to ensure equal loading of samples prior to blotting to nylon membranes. Filters were preincubated at 65°C for 4 h in 0.25 mNa2HPO4, pH 7.2, 1% (w/v) BSA, 7% (w/v) SDS, and 1 mm EDTA. The hybridization solution was identical to that used for preincubation, except that 2.5 ng mL−1[32P]cDNA probe was added. Filters were hybridized for 16 h at 65°C, and then washed several times at 68°C and at 22°C. In Situ Hybridization Analysis of HPS Gene Expression Tissue samples were fixed in a solution of 50% ethanol, 5% acetic acid, and 3.7% formaldehyde (all solutions v/v) for 3 h at room temperature, dehydrated in an ethanol series (50%, 60%, 70%, 80%, 90%, 95%, and 100%), and infiltrated with t-butyl alcohol in a stepwise series. Samples were then embedded in paraffin embedding medium (Paraplast, Sigma), placed in blocks, and allowed to harden. Sections of 8 to 10 μm were cut on a rotary microtome and affixed to glass slides. Prior to hybridization, sections were dewaxed in xylene and rehydrated in an ethanol series (100%, 95%, 85%, 70%, 50%, 30%, 15%, and 0% ethanol in distilled, RNase-free water). Sections were then treated with proteinase K and acetylated with acetic anhydride in triethanolamine. Antisense [35S]RNA probes were generated from the HPS cDNA clone. Hybridization methods followed published protocols (Cox and Goldberg, 1988). Sections were hybridized overnight at 42°C, then washed and dehydrated in an ethanol series before application of track emulsion (NTB-2, Kodak). After 1 week at 4°C, slides were developed in (D-19 developer, Kodak), fixed (Kodak), and briefly stained in Toluidine Blue O. They were then dehydrated in an ethanol and xylene series and placed in synthetic mounting medium (Permount, Fisher Scientific). Slides were photographed on slide film (EPL 400, Kodak) using dark-field optics. SEM and Droplet-Surface Analysis Whole, fully mature seeds were mounted to stages with conductive adhesive, sputter coated with gold, and examined using an ISI-DS-130 (International Scientific Instruments, Tokyo, Japan) or a field emission SEM (model S-4500, Hitachi, Tokyo). For the contact angle analysis, seeds were analyzed using a contact angle goniometer (model 100, Ramè-Hart, Mountain Lake, NJ) equipped with a microsyringe attachment. A random sample of four or five individual seeds were measured for each cultivar using water as a probe liquid. To measure static angles, 4 μL of water was deposited on the seed surface. More water was added to the drop to measure the advancing angle. RESULTS HPS Occurs on the Seed Surface To determine the composition of proteins deposited on the soybean seed surface, seeds were washed with a detergent-buffer solution and the extracted peptides were separated by SDS-PAGE. Protein extracts from the seed coat and embryo were also prepared for comparison. These results are shown in Figure 1A. The embryo and seed coat extracts contained many proteins covering a wide range of molecular masses. In contrast, extracts from the seed surface were dominated by a few low-molecular-mass proteins. Initial inconsistencies in the quantity and composition of the surface-extracted proteins was found to result from two main factors: First, oxidation of DTT in the gel loading buffer caused striking changes in the peptides detected by this analysis (Fig. 1B); fresh solutions containing high concentrations of DTT were required to obtain consistent patterns. Second, the amount of protein detected in these extracts varied greatly among different soybean cultivars. Figure 1C shows that the presence of surface protein is correlated with the luster, or light-reflective, properties of the seed surface. Surface extracts from shiny-seeded phenotypes usually contained far less protein than dull-seeded extracts. Moreover, there were large differences in the amount of protein present on the seed surfaces of the two bloom phenotypes examined. To determine the connection between surface protein and seed phenotype, seeds of 80 F2 plants developed from a cross of dull (OX281) and shiny (cv Mukden) parents were scored for luster and the presence of surface protein. This analysis clearly indicated that the presence of surface protein either contributes to the development of dull phenotypes or that corresponding genes controlling seed luster and surface protein are tightly linked in this cross. The genetics are being studied further and will be reported elsewhere. Fig. 1. Open in new tabDownload slide SDS-PAGE analysis of protein extracts from seed tissues and surface. Shown are silver-stained protein gels. Lanes marked “M” indicate protein standards, and their corresponding mass in kilodaltons is provided at left. A, Soluble protein extracts from the embryo, seed coat, and seed surface of a dull phenotype (cv Harosoy 63). Each sample was approximately 1 μg of total protein. B, Seed surface protein extracts of a dull phenotype (cv Harosoy 63) with different concentrations of DTT present in the sample loading buffer, as indicated at the top of each lane. C, Seed surface protein extracts of dull (D), shiny (S), and bloom (B) phenotypes, as indicated at the top of each lane. Fig. 1. Open in new tabDownload slide SDS-PAGE analysis of protein extracts from seed tissues and surface. Shown are silver-stained protein gels. Lanes marked “M” indicate protein standards, and their corresponding mass in kilodaltons is provided at left. A, Soluble protein extracts from the embryo, seed coat, and seed surface of a dull phenotype (cv Harosoy 63). Each sample was approximately 1 μg of total protein. B, Seed surface protein extracts of a dull phenotype (cv Harosoy 63) with different concentrations of DTT present in the sample loading buffer, as indicated at the top of each lane. C, Seed surface protein extracts of dull (D), shiny (S), and bloom (B) phenotypes, as indicated at the top of each lane. Next we wanted to identify the most abundant of these seed surface proteins. Two peptides were purified and subjected to amino-terminal microsequencing, as indicated in Figure 1B. The resulting amino acid sequences were identical and matched existing sequences in the protein database for HPS (Odani et al., 1987; Baud et al., 1993) and soybean dust allergen (Gonzalez et al., 1995). Both peptides had alternative N-terminal residues of Ala or Ile, as has been previously noted for HPS. The different electrophoretic mobilities of the two peptides could not be accounted for from the microsequencing analysis, but may have been due to differences in glycosylation. The HPS Preprotein Contains a Signal Sequence and a Short Hydrophilic Domain To obtain the cDNA transcript of HPS, sequences in a seed coat EST database were searched for reading frames corresponding to the HPS amino acid sequence. Using this strategy, several identical cDNA transcripts that included in their reading frames peptide sequences exactly matching HPS were isolated. A 700-bp transcript that was fully sequenced included 30 bp of 5′ untranslated region, an open reading frame of 119 amino acids, and 313 bp of 3′ untranslated region. The complete deduced amino acid sequence of HPS is shown in Figure2A. The final 80 residues of this sequence correspond to the peptide sequence reported for the HPS (Odani et al., 1987). Thus, the cDNA transcript indicates that HPS is translated with a leader sequence of 39 amino acids that is cleaved during processing. Figure 2B shows that this long leader sequence consists of a hydrophobic membrane-spanning domain and a short hydrophilic domain. This is significant because similar structural features occur in a group of hybrid proteins identified from several plant species and in plant lipid transfer proteins. Fig. 2. Open in new tabDownload slide A, Deduced amino acid sequence of HPS preprotein. Alternate N-terminal residues, as determined by peptide microsequence analysis, are boxed. B, A Kyle-Doolittle hydrophilicity plot of HPS (LASERGENE software, DNASTAR, Madison, WI). In this plot, positive values indicate greater hydrophilic character. Also represented are the three domains of the HPS preprotein and the length of the mature peptide. C, A schematic comparison of HPS domain structure to three other plant proteins. Bold numbers indicate the length in amino acid residues for the domain segments. The pattern of spacing between the eight Cys residues within the hydrophobic domains is also shown below each protein. Sequences for the tobacco N16 polypeptide (accession no.D86629), the maize Pro-rich hydrophobic protein (PRHP) (accession no.X60432), and the Arabidopsis lipid transfer protein 1 (LTP1) (accession no. M80567) were retrieved from GenBank. Fig. 2. Open in new tabDownload slide A, Deduced amino acid sequence of HPS preprotein. Alternate N-terminal residues, as determined by peptide microsequence analysis, are boxed. B, A Kyle-Doolittle hydrophilicity plot of HPS (LASERGENE software, DNASTAR, Madison, WI). In this plot, positive values indicate greater hydrophilic character. Also represented are the three domains of the HPS preprotein and the length of the mature peptide. C, A schematic comparison of HPS domain structure to three other plant proteins. Bold numbers indicate the length in amino acid residues for the domain segments. The pattern of spacing between the eight Cys residues within the hydrophobic domains is also shown below each protein. Sequences for the tobacco N16 polypeptide (accession no.D86629), the maize Pro-rich hydrophobic protein (PRHP) (accession no.X60432), and the Arabidopsis lipid transfer protein 1 (LTP1) (accession no. M80567) were retrieved from GenBank. The hybrid or bimodular proteins are so named because their deduced peptide sequences consist of two discrete domains, one hydrophobic and one hydrophilic. Examples of two of these hybrid proteins and a lipid transfer protein are compared with HPS in Figure 2C. This comparison shows that all of these proteins possess an N-terminal membrane-spanning signal sequence and a 9- to 10-kD hydrophobic domain with eight regularly spaced Cys residues. However, in HPS and the hybrid proteins, the N-terminal signal sequence and the hydrophobic domain are interrupted by a hydrophilic domain. The hydrophilic domains of these proteins are highly variable in their length and in their amino acid sequence and compositions. Different Seed Luster Phenotypes Show Polymorphic HPSGene Structure To compare HPS gene structure in two different seed luster phenotypes that were also different in the amount of HPS present on the seed surfaces, we hybridized genomic DNA blots with probes derived from the HPS cDNA sequence under high-stringency conditions. A typical result from such a Southern analysis is shown in Figure3. Genomic DNA blots from cultivars that accumulated large amounts of HPS on the seed surface produced strong hybridization signals. These intensely hybridizing fragments were not present in genomic DNA from plants with only trace amounts of HPS on the seed surface. However, several fainter signals were also present in DNA blots from both types of plants. These results indicate that sequences related to the HPS cDNA are prevalent in the soybean genome, and that the HPS gene structure is polymorphic among soybean cultivars. Soybean types that accumulate large amounts of HPS on the seed surface possess additional copies of this gene. Fig. 3. Open in new tabDownload slide Restriction fragment length polymorphisms between dull and shiny phenotypes. Genomic DNA from dull (cv Harosoy 63) and shiny (cv Williams 82) soybeans with abundant (+) or trace (−) amounts of HPS on the seed surface was digested with restriction enzymes, electrophoretically separated, blotted, and hybridized to the HPS cDNA probe. The size of hybridizing fragments was estimated by comparison with standards and is shown on the left (in kb). Fig. 3. Open in new tabDownload slide Restriction fragment length polymorphisms between dull and shiny phenotypes. Genomic DNA from dull (cv Harosoy 63) and shiny (cv Williams 82) soybeans with abundant (+) or trace (−) amounts of HPS on the seed surface was digested with restriction enzymes, electrophoretically separated, blotted, and hybridized to the HPS cDNA probe. The size of hybridizing fragments was estimated by comparison with standards and is shown on the left (in kb). High Expression of HPS Occurs in the Pod Endocarp Developmental and tissue-specific expression patterns forHPS were determined by RNA analysis and in situ hybridization. Representative RNA blots probed with HPS cDNA are shown in Figure 4. These results show that HPS is highly expressed in the pod during the mid to late stages of seed development. Hybridization signals were also observed in seed coat RNA samples. No expression was evident in the flower, leaf, embryo, stem, or root. We also compared HPStranscript levels of two different seed luster phenotypes that differ in the amount of HPS present on their seed surfaces. Figure 4B shows that HPS mRNA levels are several times greater in dull-seeded plants that accumulate large amounts of HPS on the seed surface compared with shiny-seeded plants that have only trace amounts of HPS on the seed surface. Faint signals corresponding to low HPS transcript levels were detectable in shiny-seeded phenotypes after prolonged exposure times (not shown). Fig. 4. Open in new tabDownload slide Analysis of HPS gene expression by RNA hybridization. Total RNA was isolated from leaf, flower, pod shells, seed coat, embryo, stem, or root tissue. Equal amounts of RNA (10 μg) were blotted to nylon and probed with HPScDNA. rRNA, visualized by staining with ethidium bromide, is shown as control. A, RNA from tissues at early (E), mid (M), or late (L) stages of development were compared for HPS gene expression. All samples shown are from a dull-seeded phenotype (cv Harosoy 63). B, RNA from pod tissues of dull (cv Harosoy 63)- and shiny (cv Williams 82)-seeded soybeans were compared for HPS gene expression. Fig. 4. Open in new tabDownload slide Analysis of HPS gene expression by RNA hybridization. Total RNA was isolated from leaf, flower, pod shells, seed coat, embryo, stem, or root tissue. Equal amounts of RNA (10 μg) were blotted to nylon and probed with HPScDNA. rRNA, visualized by staining with ethidium bromide, is shown as control. A, RNA from tissues at early (E), mid (M), or late (L) stages of development were compared for HPS gene expression. All samples shown are from a dull-seeded phenotype (cv Harosoy 63). B, RNA from pod tissues of dull (cv Harosoy 63)- and shiny (cv Williams 82)-seeded soybeans were compared for HPS gene expression. Localization of HPS gene expression by in situ hybridization is shown in Figure 5. At 6 DPA, the expression of HPS was limited to the membranous inner layer of the pericarp. By 12 DPA expression was very strong and the inner epidermis was showing signs of becoming detached from the rest of the pericarp (and in places was adhering to the seed surface). Tissue sections from this stage of development also showed strong hybridization signals in the sclerenchyma, indicating thatHPS expression occurs throughout the endocarp. Fig. 5. Open in new tabDownload slide Localization of HPS mRNA transcript by in situ hybridization. Cross-sections of soybean pods containing immature seeds (dull phenotype, HPS [+], cv Maple Presto). Hybridization of 35S-labeled HPS probe to complementary mRNA appears as a bright white signal in these dark-field microscopy images. E, Embryo; Ep, inner epidermal layer of endocarp; Ex, exocarp; F, funiculus; M, mesocarp; SC, seed coat; Sm, sclerenchyma layer of endocarp. Bar = 100 μm. A, Expression at 6 DPA. B and C, Expression at 12 DPA. Fig. 5. Open in new tabDownload slide Localization of HPS mRNA transcript by in situ hybridization. Cross-sections of soybean pods containing immature seeds (dull phenotype, HPS [+], cv Maple Presto). Hybridization of 35S-labeled HPS probe to complementary mRNA appears as a bright white signal in these dark-field microscopy images. E, Embryo; Ep, inner epidermal layer of endocarp; Ex, exocarp; F, funiculus; M, mesocarp; SC, seed coat; Sm, sclerenchyma layer of endocarp. Bar = 100 μm. A, Expression at 6 DPA. B and C, Expression at 12 DPA. Physical Properties of the Seed Surface Are Affected by the Luster Phenotype Figure 6 shows SEM images of the seed surfaces of four soybean cultivars. The four cultivars represent three distinct surface phenotypes: shiny, dull, and bloom. The dull-seeded cv Clark and its bloom isoline Clark B1accumulate large amounts of HPS on their surfaces, whereas bloom cv Sooty and shiny cv Williams 82 have only trace amounts of HPS. SEM analysis showed that the shiny seeded soybeans have a relatively smooth and undulating surface, whereas dull types are uniformly covered with bits of adhering endocarp. Large patches of contiguous membranous endocarp produce a honeycomb-like pattern on the surface of bloom phenotypes, although this tissue appears more fragmented in ClarkB1 than in cv Sooty. Fig. 6. Open in new tabDownload slide SEM micrographs of seed surfaces of shiny, dull, and bloom phenotypes. Four different combinations of phenotype and HPS content (−, trace; +, abundant) are shown at three magnifications. The lowest magnifications (top micrographs) show views of the whole seeds. The large, oval-shaped scar on the seed surface is the hilum, corresponding to the point of detachment of the mature seed from the funiculus. Higher magnifications are focused outside of hilum region. Lengths of scale bars or dashed lines are indicated in micrometers. Lengths across the horizontal field of view for each of the magnifications are: 7.1 mm (top); 1.1 mm (middle); and 0.2 mm (bottom). Fig. 6. Open in new tabDownload slide SEM micrographs of seed surfaces of shiny, dull, and bloom phenotypes. Four different combinations of phenotype and HPS content (−, trace; +, abundant) are shown at three magnifications. The lowest magnifications (top micrographs) show views of the whole seeds. The large, oval-shaped scar on the seed surface is the hilum, corresponding to the point of detachment of the mature seed from the funiculus. Higher magnifications are focused outside of hilum region. Lengths of scale bars or dashed lines are indicated in micrometers. Lengths across the horizontal field of view for each of the magnifications are: 7.1 mm (top); 1.1 mm (middle); and 0.2 mm (bottom). Static and advancing surface-droplet contact angles were also compared for the four soybean cultivars to determine how seed phenotype and HPS may affect surface hydrophobicity. In this analysis, high contact angles were characteristic of hydrophobic surfaces but may have also resulted from differences in surface topography. As shown in Figure7, the highest contact angles were observed for seeds that accumulated large amounts of HPS on the surface. Dull-seeded phenotypes consistently displayed the greatest contact angles, higher than either bloom or shiny phenotypes. Fig. 7. Open in new tabDownload slide Surface droplet contact angles for seeds of shiny, dull, and bloom phenotypes. Four different combinations of phenotype and HPS content (−, trace; +, abundant), corresponding to the four cultivars shown in Figure 6, were compared for surface droplet contact angles. Values are means and se values for four or five independent measurements. Fig. 7. Open in new tabDownload slide Surface droplet contact angles for seeds of shiny, dull, and bloom phenotypes. Four different combinations of phenotype and HPS content (−, trace; +, abundant), corresponding to the four cultivars shown in Figure 6, were compared for surface droplet contact angles. Values are means and se values for four or five independent measurements. DISCUSSION Soybean seeds display a wide variety of phenotypes that differ in coloration, size, shape, luster, and permeability. For example, self-colored (black)-seeded phenotypes differ markedly from the commonly grown yellow seeded types. This trait is in part determined by the I locus, a cluster of chalcone synthase genes that control anthocyanin biosynthesis in the seed coat (Todd and Vodkin, 1996). There is also variation in the composition of proteins from seed coats of different soybean varieties (Lindstrom and Vodkin, 1991; Gijzen et al., 1993), and corresponding genes encoding both structural and soluble seed coat proteins have been isolated (Schmidt et al., 1994; Gijzen, 1997). Despite these examples, there are no clear genetic or biochemical models to account for many of the observed phenotypes. Thus, we undertook a study comparing seed surface protein composition to the luster, or light-reflective, properties of the seed surface. To differentiate proteins that are present in the tissues of the seed coat from those that are deposited on the surface of the seed, we prepared seed surface extracts without dissection or homogenization. This analysis resulted in the identification of HPS as an abundant seed surface protein and provided a link between HPS and dull phenotypes. Whereas HPS has been purified and characterized as a seed constituent and a potent allergen, there have been no studies on the expression, localization, or function of the protein or any description of the corresponding gene. Our initial results raised many questions that could only be addressed by a more extensive investigation of HPS. The association of HPS and seed luster phenotypes was further tested by scoring different soybean cultivars and a segregating F2 population for HPS and luster phenotype. This revealed a strong association between HPS and dull phenotypes in soybean cultivars and in the F2 population. However, the quantity of HPS on the seed surface is not simply dependent upon the amount of adhering endocarp tissue, since the bloom phenotype cv Sooty possessed a heavy coating of endocarp tissue but only trace amounts of surface HPS. The integration of the B1gene from cv Sooty into Clark B1 apparently occurred without loss of the abundant surface HPS present in the recurrent Clark parent. Although cv Sooty and Clark B1 are both described as bloom phenotypes, SEM analysis showed that the endocarp is more fragmented in Clark B1. This fragmentation may result from higher levels of HPS gene expression in Clark B1. The interrelationships among seed luster, adhering endocarp, and HPS are not entirely clear, but the present study did suggest the following. The amount of endocarp tissue adhering to the seed influences the luster of the surface in a quantitative manner. The progression from shiny to intermediate, dull, and bloom phenotypes seems to depend mostly on the amount of adhering endocarp tissue. However, the appearance of the underlying surface and the pattern of attachment of the endocarp may also be important contributing factors. Taken together, the evidence suggests that seed luster is a quantitative trait determined by several loci. Thus, the expression of HPS in the endocarp may be one factor of many that influence how this tissue clings to the seed surface and produces a spectrum of luster phenotypes. It is also possible that HPS does not have any role in the fragmentation or attachment of the endocarp to the seed, but that it is tightly linked to other genes that control this trait. Regardless, DNA and RNA analysis clearly shows that HPS gene structure and transcript levels are very different in plants that accumulate large amounts of HPS on the seed surface than in those that do not. Isolation of cDNA clones encoding HPS has provided the complete sequence of the protein precursor to HPS and confirmed its relationship to a group of hybrid Pro-rich and extensin-like proteins from several other plant species. All of these proteins possess a distinct hydrophobic domain of 80 to 100 amino acids encompassing eight regularly spaced Cys residues. Transcripts encoding hybrid proteins have been isolated from many different plant species under conditions of cold (Castonguay et al., 1994), high salt (Deutch and Winicov, 1995), mechanical stress (Huang et al., 1998), or tissue-specific selection (Josè-Estanyol et al., 1992; Coupe et al., 1993; Yasuda et al., 1997). Ascribing functional roles to these proteins has been difficult and in no case has a protein of this type been associated with a phenotypic character. Plant lipid transfer proteins also show similarity to HPS in size, hydrophobicity, and in the number and spacing of Cys residues in the peptide chain. These proteins are commonly found on leaf surfaces, where they are thought to participate in cuticle biosynthesis and possibly in defense and environmental adaptation (Kader, 1996). Another group of small, Cys-rich, hydrophobic proteins that occur on surfaces are the fungal hydrophobins, a group of secreted proteins that cover hyphae or reproductive structures and influence physical properties of the fungal surface (Wessels, 1997; Kershaw and Talbot, 1998). Thus, a common feature shared by HPS, many lipid transfer proteins, and fungal hydrophobins is surface localization. These compact, hydrophobic, and Cys-rich proteins offer properties that make them attractive for covering surfaces. For example, Sc3p is a hydrophobin fromSchizophyllum commune that self assembles in vitro to form rodlet structures identical to those occurring on the surface of aerial hyphae (Wösten et al., 1994). The capacity of HPS to quickly crystallize out of solution (Odani et al., 1987) and the requirement for high concentrations of DTT to reduce soluble extracts of the protein to monomers demonstrates that HPS also has strong self-associative properties. Results from RNA analysis suggest that HPS is highly expressed in both the pod and seed coat tissues during the mid to late stages of development. However, localization of HPS mRNA by in situ hybridization suggests that HPS expression is tightly restricted to the inner epidermis and sclerenchyma of the pod endocarp. Hybridization signals observed in seed coat RNA blots are likely due to contamination of the seed coat with the membranous inner epidermis of the pericarp, since this tissue sticks to surface of developing seeds and is difficult to completely remove. Thus, we conclude thatHPS is specifically expressed in the endocarp. Proteins expressed in this tissue, or whole sections of the inner epidermis itself, adhere to the seed surface during development and become a component of the seed coat of mature, fully developed soybeans. Odani et al. (1987) estimated the abundance of HPS to be in the range of 200 mg kg−1 whole seed. The presence of such large amounts of protein, restricted entirely to the seed surface, would alter the physical properties of the surface and suggest a structural or defensive function for the protein. Results from contact angle analysis of surface droplets provide correlative evidence that HPS reduces the wettability of seed surfaces. The hydrophobicity and topography of the surface could affect pathogen attachment and penetration or influence the water-absorptive properties of the seed. It is also possible that HPS acts directly as a feeding deterrent or toxin against specific herbivores, pests, or pathogens. More experimentation is required to clarify the functional role of HPS. The demonstration that large amounts of HPS are present on the seed surface is consistent with the localization of the soybean dust allergen to the seed hull fraction (Rodrigo et al., 1990; Swanson et al., 1991), since this allergen was subsequently identified as HPS (Gonzalez et al., 1995). Re-occurring, community-wide outbreaks of asthma in Barcelona and Cartagena (Spain) from 1981 to 1987 were caused by the release of soybean dust through the unloading of seed from container vessels (Antó et al., 1989). These epidemics affected hundreds of individuals and resulted in several deaths (Antó et al., 1993). Soybean dust is also the probable cause of earlier asthma outbreaks in other cities, including New Orleans (Weill et al., 1964), and is listed as a workplace hazard for food industry workers (Pepys, 1986). Our work offers new opportunities for lessening the health hazard of seed dust exposure. For example, phenotypic or genetic screens may be devised to select plants with reduced amounts of HPS on the seed surface. More broadly, results presented here indicate that physical, textural, or compositional properties of the seed surface may be altered by manipulating gene expression in the ovary wall. The accession number for the sequence reported in this article isAF100159. ACKNOWLEDGMENTS We thank Ross Davidson, Mark Biesinger, and Mary Jane Walzak at Surface Science Western for electron microscopy and droplet-surface analysis; Pearl Campbell and Heather Schneider at the Robarts Research Institute for DNA sequencing; Aldona Gaidauskas-Scott and Lu-Ann Bowman for technical assistance; Dorothy Drew for library services; and the Biotechnology Service Centre at the University of Toronto for peptide microsequencing. Abbreviations: HPS hydrophobic protein from soybean SEM scanning electron microscope(y) LITERATURE CITED 1 Antó JM Sunyer J Reed CE Sabrià J Martı́nez F Morrell F Cordina R Rodriguez-Roisin R Rodrigo MJ Roco J Preventing asthma epidemics due to soybeans by dust-control measures. N Engl J Med 329 1993 1760 1763 Google Scholar Crossref Search ADS PubMed WorldCat 2 Antó JM Sunyer J Rodriguez-Roisin R Suarez-Cervera M Vazquez L Community outbreaks of asthma associated with inhalation of soybean dust. N Engl J Med 320 1989 1097 1102 Google Scholar Crossref Search ADS PubMed WorldCat 3 Baud F Pebay-Peyroula E Cohen-Addad C Odani S Lehmann MS Crystal structure of hydrophobic protein from soybean: a member of a new cysteine-rich family. J Mol Biol 231 1993 877 887 Google Scholar Crossref Search ADS PubMed WorldCat 4 Bernard RL, Weiss MG (1973) Qualitative genetics.In BE Caldwell, ed, Soybeans: Improvement, Production, and Uses. American Society of Agronomy, Madison, WI, pp 117–154 5 Blum H Beier H Gross HJ Improved staining of plant proteins, RNA and DNA in polyacrylamide gels. Electrophoresis 8 1987 93 99 Google Scholar Crossref Search ADS WorldCat 6 Castonguay Y Laberge S Nadeau P Vézina LP A cold-induced gene from Medicago sativa encodes a bimodular protein similar to developmentally regulated proteins. Plant Mol Biol 24 1994 799 804 Google Scholar Crossref Search ADS PubMed WorldCat 7 Corner EJH The leguminous seed. Phytomorphology 1 1951 117 150 Google Scholar OpenURL Placeholder Text WorldCat 8 Coupe SA Taylor JE Isaac PG Roberts JA Identification and characterization of a proline-rich mRNA that accumulates during pod development in oilseed rape (Brassica napus L.). Plant Mol Biol 23 1993 1223 1232 Google Scholar Crossref Search ADS PubMed WorldCat 9 Cox KH Goldberg RB Analysis of plant gene expression. Shaw CH Plant Molecular Biology: A Practical Approach. 1988 1 35 IRL Press Oxford 10 Dellaporta SL Wood J Hicks JB A plant DNA minipreparation: version II. Plant Mol Biol Rep 1 1983 19 21 Google Scholar Crossref Search ADS WorldCat 11 Deutch CE Winicov I Post-transcriptional regulation of a salt-inducible alfalfa gene encoding a putative chimeric proline-rich cell wall protein. Plant Mol Biol 27 1995 411 418 Google Scholar Crossref Search ADS PubMed WorldCat 12 Esau K (1977) Anatomy of Seed Plants. Wiley, New York 13 Fling SP Gregerson DS Peptide and protein molecular weight determination by electrophoresis using a high-molarity Tris buffer system without urea. Anal Biochem 155 1986 83 88 Google Scholar Crossref Search ADS PubMed WorldCat 14 Gijzen M A deletion mutation at the ep locus causes low seed coat peroxidase activity in soybean. Plant J 12 1997 991 998 Google Scholar Crossref Search ADS PubMed WorldCat 15 Gijzen M van Huystee R Buzzell R Soybean seed coat peroxidase: a comparison of high-activity and low-activity genotypes. Plant Physiol 103 1993 1061 1066 Google Scholar Crossref Search ADS PubMed WorldCat 16 Gonzalez R Varela J Carreira J Polo F Soybean hydrophobic protein and soybean hull allergy. Lancet 346 1995 48 49 Google Scholar Crossref Search ADS PubMed WorldCat 17 Guodong Z Jinling W Qingxi M Inheritance of bloom on seed coat in soybean. Soybean Genet Newsl 14 1987 91 93 Google Scholar OpenURL Placeholder Text WorldCat 18 Huang YF Jordan WR Wing RA Morgan PW Gene expression induced by physical impedance in maize roots. Plant Mol Biol 37 1998 921 930 Google Scholar Crossref Search ADS PubMed WorldCat 19 Josè-Estanyol M Ruiz-Avila L Puigdomènech P A maize embryo-specific gene encodes a proline-rich and hydrophobic protein. Plant Cell 4 1992 413 423 Google Scholar PubMed OpenURL Placeholder Text WorldCat 20 Juvik GA, Bernard RL, Chang R, Cavins JF (1989) Evaluation of the USDA wild soybean germplasm collection: maturity groups 000 to IV (PI 65.549 to PI 483.464). USDA Technical Bulletin No. 1761 21 Kader JC Lipid transfer proteins in plants. Annu Rev Plant Physiol Plant Mol Biol 47 1996 627 654 Google Scholar Crossref Search ADS PubMed WorldCat 22 Kershaw MJ Talbot NJ Hydrophobins and repellents: proteins with fundamental roles in fungal morphogenesis. Fungal Genet Biol 23 1998 18 33 Google Scholar Crossref Search ADS PubMed WorldCat 23 Lindstrom JT Vodkin LO A soybean cell wall protein is affected by seed color genotype. Plant Cell 3 1991 561 571 Google Scholar PubMed OpenURL Placeholder Text WorldCat 24 Moos M Nguyen NY Liu TY Reproducible high yield sequencing of proteins electrophoretically separated and transferred to an inert support. J Biol Chem 263 1988 6005 6008 Google Scholar Crossref Search ADS PubMed WorldCat 25 Newell CA Hymowitz T Seed coat variation in Glycine Willd. subgenus Glycine (Leguminosae) by SEM. Brittonia 30 1978 76 88 Google Scholar Crossref Search ADS WorldCat 26 Odani S Koide T Ono T Seto Y Tanaka T Soybean hydrophobic protein: isolation, partial characterization and the complete primary structure. Eur J Biochem 162 1987 485 491 Google Scholar Crossref Search ADS PubMed WorldCat 27 Pepys J Occupational allergic lung disease caused by organic agents. J Allergy Clin Immunol 78 1986 1058 1062 Google Scholar Crossref Search ADS PubMed WorldCat 28 Rodrigo MJ Morell F Helm RM Swanson M Greife A Anto JM Sunyer J Reed CE Identification and partial characterization of the soybean-dust allergens in the Barcelona asthma epidemic. J Allergy Clin Immunol 85 1990 778 784 Google Scholar Crossref Search ADS PubMed WorldCat 29 Sambrook J Fritsch EF Maniatis T Molecular Cloning: A Laboratory Manual, Ed 2. 1989 Cold Spring Harbor Laboratory Press Cold Spring Harbor, NY 30 Schmidt JS Lindstrom JT Vodkin LO Genetic length polymorphisms create size variation in proline-rich proteins of the cell wall. Plant J 6 1994 177 186 Google Scholar Crossref Search ADS PubMed WorldCat 31 Swanson MC Li JTC Wentz-Mutha PE Trudeau WL Fernadez-Caldas E Greife A Rodrigo MAJ Morell F Reed CE Source of the aeroallergen of soybean dust: a low molecularmass glycopeptide from the soybean tela. J Allergy Clin Immunol 87 1991 783 788 Google Scholar Crossref Search ADS PubMed WorldCat 32 Todd JJ Vodkin LO Duplications that suppress and deletions that restore expression from a chalcone synthase multigene family. Plant Cell 8 1996 687 699 Google Scholar Crossref Search ADS PubMed WorldCat 33 Wang CS Vodkin LO Extraction of RNA from tissues containing high levels of procyanidins that bind RNA. Plant Mol Biol Rep 12 1994 132 145 Google Scholar Crossref Search ADS WorldCat 34 Weill H Ziskind MM Dicherson RC Derbes VJ Epidemic asthma in New Orleans. JAMA 190 1964 11 14 Google Scholar Crossref Search ADS WorldCat 35 Wessels JGH Hydrophobins: proteins that change the nature of the fungal surface. Adv Microbial Physiol 38 1997 1 45 Google Scholar OpenURL Placeholder Text WorldCat 36 Wolf WJ Baker FL Scanning electron microscopy of soybeans. Cereal Sci Today 17 1972 124 130 Google Scholar OpenURL Placeholder Text WorldCat 37 Woodworth CM Genetics of the soybean. J Am Soc Agron 25 1933 36 51 Google Scholar Crossref Search ADS WorldCat 38 Wösten HAB Ásgeirsdóttir SA Krook JH Drenth JHH Wessels JGH The fungal hydrophobin Sc3p self-assembles at the surface of aerial hyphae as a protein membrane constituting the hydrophobic rodlet layer. Eur J Cell Biol 63 1994 122 129 Google Scholar PubMed OpenURL Placeholder Text WorldCat 39 Yasuda E Ebinuma H Wabiko H A novel glycine-rich/hydrophobic 16 kDa polypeptide gene from tobacco: similarity to proline-rich protein genes and its wound-inducible and developmentally regulated expression. Plant Mol Biol 33 1997 667 678 Google Scholar Crossref Search ADS PubMed WorldCat Author notes 1 This research was supported in part by a grant from the Ontario Soybean Grower's Marketing Board. * Corresponding author; e-mail [email protected]; fax 519–457–3997. Copyright © 1999 American Society of Plant Physiologists This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)

Barrieu, François; Chrispeels, Maarten J.

doi: 10.1104/pp.120.4.961pmid: 10444079

Abstract To further understand how membrane proteins are sorted in the secretory system, we devised a strategy that involves the expression of a membrane-anchored yeast invertase in transgenic plants. The construct consisted of a signal peptide followed by the coding region of yeast invertase and the transmembrane domain and cytoplasmic tail of calnexin. The substitution of a lysine near the C terminus of calnexin with a glutamic acid residue ensured progression through the secretory system rather than retention in or return to the endoplasmic reticulum. In the transformed plants, invertase activity and a 70-kD cross-reacting protein were found in the vacuoles. This yeast invertase had plant-specific complex glycans, indicating that transport to the vacuole was mediated by the Golgi apparatus. The microsomal fraction contained a membrane-anchored 90-kD cross-reacting polypeptide, but was devoid of invertase activity. Our results indicate that this membrane-anchored protein proceeds in the secretory system beyond the point where soluble proteins are sorted for secretion, and is detached from its membrane anchor either just before or just after delivery to the vacuole. The endomembrane system of plant cells consists of a series of compartments and membrane systems, each with unique proteins, and shuttle vesicles that transport proteins and lipids from one compartment to another. The correct delivery of a protein to a particular compartment depends on information within the protein itself (sorting signals) and on transport machinery that interacts with this information. The sorting signals specify retention in a particular compartment or sorting to the appropriate compartment. Considerable progress has been made in recent years in understanding the sorting signals on soluble proteins and the cellular machinery needed for correct delivery of these proteins to various cellular destinations. For example, when a retention or sorting signal is removed, the protein still enters the secretory pathway as long as it has a signal peptide, but is then secreted from the cells. Thus, transport to the vacuole requires positive sorting information (Dorel et al., 1989; for reviews, see Neuhaus and Rogers, 1998; Raikhel and Vitale, 1999). There is considerably less information about the sorting of integral membrane proteins with transmembrane domains. Such proteins may enter the secretory system because they have a cleavable signal peptide (like soluble proteins) followed by a transmembrane domain that acts as a stop-transfer signal, or they may become integrated into the membrane because of the presence of one or more internal transmembrane domains that act as uncleaved signal peptides and stop-transfer signals. For plants, there is essentially no information about the domains or motifs of integral membrane proteins that specifies their targeting or retention. With respect to targeting, we need to understand why certain proteins proceed from the ER (their point of entry) to the plasma membrane, whereas others go to the tonoplast. This is particularly relevant in the case of homologous proteins such as aquaporins, which are found in both membranes. Although amino acid sequence comparisons have revealed differences between the aquaporin homologs (Schaffner, 1998), it is not clear that these domains contain the targeting information. The issue is further complicated by the presence of more than one type of vacuole in plant cells (Paris et al., 1996; Swanson et al., 1998). For soluble proteins, it has been shown that the critical sorting event occurs in the TGN, and that secretion is a default destination, whereas transport to the vacuole requires positive sorting information. In the yeast Saccharomyces cerevisiae, the vacuolar membrane appears to be the default destination of integral membrane proteins (Roberts et al., 1992; Gaynor et al., 1994; Chang and Fink, 1995;Roberg et al., 1997). Our previous attempt to answer the default question for plant integral membrane proteins (Höfte and Chrispeels, 1992) did not yield an unequivocal answer. We found that a reporter protein attached to the sixth transmembrane domain of a tonoplast aquaporin was delivered to the tonoplast, but we could not be sure that this transmembrane domain was devoid of any tonoplast targeting information, because it was derived from a tonoplast protein. A similar experiment was recently carried out by Jiang and Rogers (1998), who used a mutated form of barley pro-aleurain as the reporter protein and the transmembrane domain of the vacuolar sorting receptor BP80 together with the cytoplasmic C-terminal tails of two different TIPs (α-TIP and γ-TIP). These cytoplasmic tails caused targeting to different post-Golgi compartments. In the present study we examine a related question: Can a protein that is normally secreted—in this case yeast invertase—be delivered to the vacuole if it is membrane anchored? We chose yeast invertase because it is known not to have plant vacuolar targeting information. When yeast invertase is equipped with a plant signal peptide, catalytically active protein is secreted in the apoplast (von Schaewen et al., 1990;Dickinson et al., 1991). We chose the transmembrane domain of yeast calnexin because we assumed that this domain would not have information for targeting to the plant tonoplast. Our results show that soluble invertase accumulated in the vacuoles of transformed tobacco (Nicotiana tabacum) plants, suggesting that membrane-anchored invertase was transported to a destination beyond the point where soluble proteins are sorted for secretion, possibly to a PVC or to the vacuole itself, before it was detached from its membrane anchor. MATERIALS AND METHODS Reagents Restriction and DNA modification enzymes were obtained from New England Biolabs. Pfu DNA polymerase was purchased from Stratagene. Unless otherwise stated, all other chemicals were obtained from Sigma. Plasmid Construction The 5′-GACTGGTACCCTAGAGTTTG-3′ and 3′-GATCATATACAAAAGTATAGG-5′ primers were used for PCR amplification of the 3′ region of the plasmid pEG-1-QK (Fig. 1) (Gaynor et al., 1994). The PCR product obtained was digested by KpnI and introduced as an KpnI/SmaI fragment into a plant expression cassette containing the CaMV 35S promoter and the polyadenylation signal of the OCS (octopine synthase) gene. This construct was then digested with KpnI to allow the insertion of theKpnI/KpnI fragment of the PI-3-Inv plasmid (von Schaewen et al., 1990) corresponding to the signal peptide of the proteinase inhibitor II gene from potato (Keil et al., 1986) and the remaining part of the Suc 2 gene (Fig. 1) (Johnson et al., 1987). The constructs were finally cloned into a binary vector (PDE 1001, Plant Genetic Systems, Ghent, Belgium) asEcoRI/HindIII fragments, and directly transformed into Agrobacterium tumefaciens strain C58 AGL-0 (Lazo et al., 1991). Fig. 1. Open in new tabDownload slide Summary of the construction of the plasmid used in this study. Fig. 1. Open in new tabDownload slide Summary of the construction of the plasmid used in this study. Transformation and Regeneration of Transgenic Plants Leaf discs of tobacco (Nicotiana tabacum cv Xanthi) were transformed as described by Voelker et al. (1989). Transformed plants were grown in tissue culture under a 16-h light/8-h dark regime on Murashige and Skoog medium (Murashige and Skoog, 1962) containing 3% (w/w) Suc and 100 μg/mL kanamycin. The kanamycin-resistant plants were transferred to soil (Special Blend, Sun Gro Horticulture, Bellevue, WA) and grown in individual pots in a growth chamber under a 16-h light/8-h dark regime. Leaves were collected for invertase activity analysis and the highest expressors were used for further analysis. Detection of Invertase Activity in Native Polyacrylamide Gels Yeast invertase activity in leaves of transformed tobacco plants was detected using a native invertase activity gel assay (Gabriel and Wang, 1969). Triton X-100 (final concentration 0.1% [v/v]) was added to aliquots of protein extracts before loading on a 10% polyacrylamide gel; the gel and running buffers were 100 mmTris-phosphate, pH 6.7. After running overnight at 40 V and 4°C, gels were incubated in an acidic Suc solution (0.1 m Suc and 0.1m NaOAc, pH 5) for 30 min at 30°C. Following a brief wash in distilled water, gels were developed by incubation in a boiling solution of 0.5 m NaOH containing 0.1% (w/v) 2,3,5-triphenyltetrazolium chloride, giving rise to red bands at positions of invertase activity. Protein Extraction and Preparation of Soluble and Microsomal Fractions Total protein extracts were obtained by homogenizing tobacco leaves (500 mg) in 2 mL of cold extraction buffer (50 mm Tris-phosphate, pH 6.7, 1% [v/v] β-mercaptoethanol, 12% [w/w] Suc, 0.2 mm aminoethylbenzene-sulfonylfluoride [Calbiochem-Novabiochem], 2 μg/mL aprotinin, and 1 μg/mL leupeptin) and collecting the supernatant after centrifugation at 10,000g for 10 min. The homogenate was then fractionated into a soluble and crude microsomal fraction by centrifugation at 100,000g for 1 h through a cushion of extraction buffer containing 16% (w/w) Suc. The upper phase, containing the soluble proteins, was collected and is referred to as the soluble fraction. The microsomal pellet was resuspended in extraction buffer. Immunoprecipitation, SDS-PAGE, and Immunoblotting Immunoprecipitation experiments were carried out as described byFaye and Chrispeels (1989). SDS-PAGE was performed on 15% (w/v) polyacrylamide slab gels according to the method of Laemmli (1970). Proteins separated by SDS-PAGE were transferred to nitrocellulose membranes (Micron Separations, Westborough, MA) according to the method of Faye and Chrispeels (1985). Immunodetection was carried out essentially as described by Laurière et al. (1989), except that the yeast invertase and BiP antisera were diluted 1:1,000 with 0.05% (v/v) Tween 20 in TBS (20 mm Tris-HCl, pH 7.5, and 500 mm NaCl). Isolation of Protoplasts and Vacuoles Protoplast and vacuole isolation from whole tobacco leaf tissue was carried out as described by Dombrowski et al. (1994). The purity and integrity of the vacuoles were monitored microscopically. To compare invertase activity in protoplasts and in the vacuolar fraction, equal α-mannosidase activities, determined according to the method ofVan der Wilden et al. (1980), were subjected to PAGE and stained for invertase activity as described above. RESULTS Construction of a Membrane-Anchored Chimeric Yeast Invertase Gene A chimeric gene that encodes a membrane-anchored invertase that would enter the secretory system so that the path of the invertase protein in the cell could be followed contained the following parts: the signal peptide of the potato PR1 protein fused in frame to the coding sequence of yeast invertase, which was fused in frame to the transmembrane domain, followed by the short C-terminal cytoplasmic tail of yeast calnexin (see Methods and Fig. 1 for details). The derived amino acid sequence of this construct and the accompanying hydropathy plot (Fig. 2) reveal the presence of two hydrophobic domains. We reasoned that the signal peptide would allow the nascent polypeptide to enter the protein into the ER lumen, and that the transmembrane domain of calnexin would act as a stop-transfer sequence, creating a type I membrane protein with a large luminal domain and a short cytoplasmic tail. Fig. 2. Open in new tabDownload slide Amino acid sequence and derived hydropathy plot of the membrane-anchored chimeric yeast invertase. A, Amino acid sequence of the fusion protein. The single-letter amino acid code is used. The signal peptide of the potato PR1 protein is represented in italics. The transmembrane domain of the yeast calnexin Wbp1 is underlined. PutativeN-glycosylation sites are bolded. B, Hydropathy plot of the fusion protein. The plot was generated using a moving window of 11 residues (Kyte and Doolittle, 1982). Fig. 2. Open in new tabDownload slide Amino acid sequence and derived hydropathy plot of the membrane-anchored chimeric yeast invertase. A, Amino acid sequence of the fusion protein. The single-letter amino acid code is used. The signal peptide of the potato PR1 protein is represented in italics. The transmembrane domain of the yeast calnexin Wbp1 is underlined. PutativeN-glycosylation sites are bolded. B, Hydropathy plot of the fusion protein. The plot was generated using a moving window of 11 residues (Kyte and Doolittle, 1982). There is considerable evidence in yeast that KKXX motifs on the cytosolic tails of transmembrane proteins can act as ER retention signals (Gaynor et al., 1994; Letourneur et al., 1994). Although there is as yet (to our knowledge) no good evidence for plant cells regarding the retention function of this motif, we mutated the KKTN C terminus to QKTN by site-directed mutagenesis to eliminate this retention possibility. The construct was fused to the CaMV 35S promoter and introduced into tobacco via A. tumefaciens-mediated transformation. We recovered a dozen transgenic plants and all showed the typical “stress” phenotype previously observed in tobacco plants expressing yeast invertase in their leaves: There were large yellow and brown sectors between the major veins and these sectors turned necrotic as the leaves matured (see also von Schaewen et al., 1990; Dickinson et al., 1991). The subcellular distribution of yeast invertase activity and protein was examined in these transgenic plants. Yeast Invertase Activity Is Found in Vacuoles We used a combination of activity gels and subcellular fractionation to determine the location of the yeast invertase activity within the cells. In the type of gels used here, plant extracts do not give a reaction product, possibly because the invertase is inactivated in the heating step (see also Dickinson et al., 1991), but the thermostable yeast invertase yields the red reaction product of the tetrazolium reaction that uses oxidized Glc as its substrate (seeMethods). Leaves of young plants were homogenized in buffered 12% (w/w) Suc, and the homogenate was fractionated into a soluble and crude microsomal fraction by centrifuging the microsomes through a 16% Suc layer to free them of soluble proteins. Invertase activity gels showed that there was considerable invertase activity in the soluble fraction but no activity in the microsomal fraction (Fig.3A, lanes 3 and 4). In some experiments we found traces of yeast invertase in the microsomal fraction of the transformed plants, but this may have been caused by contamination from the soluble fraction. Fig. 3. Open in new tabDownload slide Detection of invertase activity in transgenic plants. A, Proteins from soluble (S, lanes 1 and 3) and microsomal (M, lanes 2 and 4) fractions of wild-type (WT) and transgenic (INV) leaves of transgenic plants were prepared as described in Methods and assayed for invertase activity after gel electrophoresis. Yeast invertase activity was detected only in the soluble fraction of transgenic plants. No invertase activity was detected in the soluble or microsomal fractions from wild-type plants. B, Detection of invertase activity in protoplasts (lane P) and vacuoles (lane V) of transgenic plants. Vacuoles were isolated from leaf protoplasts of transgenic plants and assayed for α-mannosidase activity. The invertase activity in the vacuole (lane 2) was compared with the invertase activity present in intact protoplasts (lane 1) after loading the same amount of α-mannosidase activity onto each lane. Fig. 3. Open in new tabDownload slide Detection of invertase activity in transgenic plants. A, Proteins from soluble (S, lanes 1 and 3) and microsomal (M, lanes 2 and 4) fractions of wild-type (WT) and transgenic (INV) leaves of transgenic plants were prepared as described in Methods and assayed for invertase activity after gel electrophoresis. Yeast invertase activity was detected only in the soluble fraction of transgenic plants. No invertase activity was detected in the soluble or microsomal fractions from wild-type plants. B, Detection of invertase activity in protoplasts (lane P) and vacuoles (lane V) of transgenic plants. Vacuoles were isolated from leaf protoplasts of transgenic plants and assayed for α-mannosidase activity. The invertase activity in the vacuole (lane 2) was compared with the invertase activity present in intact protoplasts (lane 1) after loading the same amount of α-mannosidase activity onto each lane. The presence of a plant signal peptide on yeast invertase causes the enzyme to be secreted in the apoplast of transgenic plants (von Schaewen et al., 1990; Dickinson et al., 1991). We therefore checked whether the soluble enzyme might represent enzyme extracted from the cell wall during homogenization of the leaves. We prepared extracellular fluid from leaf tissue samples according to the method ofKlement (1965), but found that it contained no yeast invertase (data not shown), whereas isolated protoplasts contained abundant amounts of soluble yeast invertase. To determine if the soluble invertase was located in the vacuoles, we isolated vacuoles by gentle lysis of leaf protoplasts obtained from the transformed plants. These vacuole fractions are generally free of contaminating organelles. Both the protoplast and the vacuole fraction were assayed for the vacuolar marker enzyme α-mannosidase. Lanes 1 and 2 of the gel, shown in Figure 3B, were loaded with aliquots containing equal amounts of α-mannosidase activity. Visualization of the invertase activity in this gel showed that the two lanes contained roughly the same amount of invertase, suggesting that the soluble yeast invertase activity is in the vacuoles. These results lead to the conclusion that an enzyme that would normally be secreted because of the presence of a signal peptide and the lack of vacuolar targeting determinants can be delivered to the vacuole if the protein is synthesized in a membrane-attached form. In the present study, the membrane attachment was apparently disrupted by proteolysis either along the secretory pathway or in the vacuole. In any case, invertase remained membrane attached beyond the point where soluble proteins without vacuolar signals are packaged for secretion. Microsomes Contain a Membrane-Anchored 90-kD Invertase Cross-Reacting Polypeptide The absence of invertase activity from the microsomes was a puzzling finding, because a protein that enters the secretory system would be expected to be found there as well as at its final destination. We used an antiserum to yeast invertase to locate cross-reacting polypeptides in the soluble and microsomal fractions of the transformed plants on an immunoblot. The results (Fig.4) show that the soluble fraction contained a 70-kD species and the microsomal fraction contained a 90-kD species. The 70-kD size is commensurate with glycosylated mature invertase, since the polypeptide itself is 58 kD and there are 10 glycosylation sites. The 90-kD size is commensurate with the glycosylated translation product of the chimeric invertase gene. Fig. 4. Open in new tabDownload slide Immunodetection of yeast invertase in protein extracts from transgenic plants. Proteins from soluble (S, lanes 1 and 3) and microsomal (M, lanes 2 and 4) fractions of wild-type (WT) and transgenic (INV) plants were fractionated by SDS-PAGE, electroblotted onto nitrocellulose membrane, and probed with a yeast invertase antiserum. Molecular standards are shown on the left (in kD). Fig. 4. Open in new tabDownload slide Immunodetection of yeast invertase in protein extracts from transgenic plants. Proteins from soluble (S, lanes 1 and 3) and microsomal (M, lanes 2 and 4) fractions of wild-type (WT) and transgenic (INV) plants were fractionated by SDS-PAGE, electroblotted onto nitrocellulose membrane, and probed with a yeast invertase antiserum. Molecular standards are shown on the left (in kD). The microsomes were subjected to five cycles of freezing and thawing, and the membranes were sedimented again by centrifugation. A comparison of the polypeptides present in the membranes and the supernatant by immunoblotting with yeast invertase and BiP antisera showed that the 90-kD polypeptide was not released by freezing and thawing of the microsomal vesicles, although a substantial portion of the soluble ER-resident BiP was released by this procedure (Fig.5). Similar results were obtained when the vesicles were treated with 0.03% Triton X-100, a concentration of detergent known to release soluble microsomal proteins (Kreibich et al., 1973; Van der Wilden et al., 1980). We interpret these data to mean that the microsomes contain a 90-kD species of invertase that has no catalytic activity. Detachment of the enzyme from its membrane anchor apparently allows the enzyme to be active. Fig. 5. Open in new tabDownload slide Immunodetection of yeast invertase in microsomes of transgenic plants. A microsomal extract prepared from leaves of transgenic plants was subjected to five freeze-thaw cycles. After centrifugation at 100,000g for 1 h, proteins from the supernatant (soluble microsomal proteins, lane 3), the pellet (microsomal membrane proteins, lane 2) and the original total microsome fraction (lane 1) were separated by SDS-PAGE and blotted onto a nitrocellulose membrane. A, Immunodetection of yeast invertase. B, Immunodetection of BiP using a serum against tomato BiP. Molecular standards (in kD) are shown on the right. Fig. 5. Open in new tabDownload slide Immunodetection of yeast invertase in microsomes of transgenic plants. A microsomal extract prepared from leaves of transgenic plants was subjected to five freeze-thaw cycles. After centrifugation at 100,000g for 1 h, proteins from the supernatant (soluble microsomal proteins, lane 3), the pellet (microsomal membrane proteins, lane 2) and the original total microsome fraction (lane 1) were separated by SDS-PAGE and blotted onto a nitrocellulose membrane. A, Immunodetection of yeast invertase. B, Immunodetection of BiP using a serum against tomato BiP. Molecular standards (in kD) are shown on the right. Transport to the Vacuole Is Mediated by the Golgi Apparatus There is some evidence that there are multiple pathways to the vacuole in plant cells, one of which may bypass the Golgi apparatus (for review, see Okita and Rogers, 1996; Beevers and Raikhel, 1998;Robinson et al., 1998). Modification of Asn-linked high-Man glycans by Golgi-located glycosidases and glycosyltransferases is diagnostic of protein transport mediated by the Golgi apparatus. Through the action of these Golgi enzymes these glycans become “complex” with α-1,3 Fuc and β-1,2 Xyl residues. The presence of such residues can be detected with a complex, glycan-specific antiserum (Laurière et al., 1989). To find out if the soluble invertase in the vacuole contained such glycans, we immunoprecipitated invertase in the soluble fraction of the homogenate with anti-invertase antibodies and protein A-Sepharose beads and then used the precipitated polypeptides for an immunoblot with the complex glycan antiserum. The results show that a 70-kD polypeptide precipitated by the invertase serum reacted strongly with the anti-complex glycan serum (Fig. 6). This means that the soluble vacuolar invertase acquired complex glycans on its way to the vacuole, and suggests that vacuolar transport is Golgi mediated. The 60-kD cross-reacting polypeptide seen in lane 2 of Figure6 is also present in lane 1, which contains proteins from the control plants, and therefore does not represent a yeast invertase polypeptide. It is likely that this polypeptide represents a component of the serum used for immunoprecipitation, which reacts with the secondary antibodies used for the immunoblot (see Methods). Fig. 6. Open in new tabDownload slide Immunoblot analysis of proteins from the soluble fractions of wild-type and transgenic plants. Proteins from the soluble fractions of wild-type (WT, lane 1) and transgenic (INV, lane 2) plants were selectively immunoprecipitated using the yeast invertase antiserum, separated by SDS-PAGE, electroblotted onto nitrocellulose membrane, and probed with a plant complex glycan antiserum (Lauriere et al., 1989). Molecular standards are shown on the left (in kD). Fig. 6. Open in new tabDownload slide Immunoblot analysis of proteins from the soluble fractions of wild-type and transgenic plants. Proteins from the soluble fractions of wild-type (WT, lane 1) and transgenic (INV, lane 2) plants were selectively immunoprecipitated using the yeast invertase antiserum, separated by SDS-PAGE, electroblotted onto nitrocellulose membrane, and probed with a plant complex glycan antiserum (Lauriere et al., 1989). Molecular standards are shown on the left (in kD). DISCUSSION The results presented in this paper support the interpretation that a soluble protein that would normally be secreted after it enters the secretory system can be delivered to the vacuole if the protein is synthesized in a membrane-attached configuration. The reason for vacuolar delivery is still unclear, but two interpretations are possible: Either the tonoplast is the default destination of a membrane-attached protein or, after this particular protein is detached from its anchor, it contains sufficient vacuolar targeting information to target it to the vacuole. If the first interpretation is correct then plants resemble the yeast S. cerevisiae, in which the vacuolar membrane is the default destination for membrane proteins, and are unlike mammalian cells in that sorting of lysosomal membrane proteins requires positive sorting information whereas sorting to the plasma membrane does not. Sorting Vacuolar/Lysosomal Membrane Proteins in Mammals, Yeasts, and Plants In mammalian cells, lysosomal membrane proteins can reach their destination by two routes: a direct route from the Golgi, possibly via a prelysosomal or endosomal compartment that requires information in the C-terminal cytoplasmic domain of the protein, or indirectly, after first being transported to the plasma membrane along the default pathway and then being retrieved by virtue of the presence of a specific sorting signal (for review, see Hunziker and Geuze, 1996).Roberts et al. (1992) examined the targeting of two integral membrane dipeptidylaminopeptidases in yeast, DPAP-A and DPAP-B, which reside in the Golgi apparatus and the vacuolar membrane, respectively. They carried out domain swaps with the two proteins and found that all their results were consistent with a model in which proteins are delivered to the vacuolar membrane along a default pathway. All subsequent studies have confirmed this interpretation. In plants, there is little information about the targeting of integral membrane proteins, and there are no studies that attempt to answer this question specifically. In two previous studies (Höfte and Chrispeels, 1992; Jiang and Rogers, 1998), a soluble reporter protein was fused to a membrane anchor derived from α-TIP, and the soluble protein was found in the vacuoles of the transformed plants expressing this construct. The interpretation may be complicated by the recent finding that some plant cells contain at least two types of vacuoles (Paris et al., 1996; Neuhaus and Rogers, 1998; Swanson et al., 1998). In young seedlings that are digesting stored reserves, both the storage parenchyma cells and the meristematic cells contain protein storage vacuoles as well as lytic vacuoles. The tonoplasts of these vacuoles have their own specific integral TIPs: α-TIP in the protein storage vacuoles and γ-TIP in the lytic vacuoles. The presence of different types of vacuoles with their own tonoplast proteins would require specific targeting information for at least one of these. Jiang and Rogers (1998) recently showed differential targeting of a membrane-anchored proaleurain when the C-terminal cytoplasmic tails of these two TIPs were used. However, when expressing α-TIP in tobacco leaves, we found that it was targeted to the tonoplasts of lytic vacuoles (Höfte and Chrispeels, 1992). It is not clear whether α-TIP has specific targeting information for the tonoplast of protein storage vacuoles. However, when these vacuoles are absent, as we presume they are in tobacco leaves, α-TIP goes to the tonoplast of the lytic vacuole. This is consistent with the lytic vacuole as a default destination. Membrane-Anchored Invertase Proceeds Beyond the TGN By equipping yeast invertase with a plant signal peptide derived from potato proteinase inhibitor 2, we ensured the entry of invertase into the secretory system. Indeed, when such a construct is expressed in transgenic tobacco plants the yeast invertase is secreted into the apoplast (von Schaewen et al., 1990; Dickinson et al., 1991). Fusion with a transmembrane domain at the C terminus normally ensures that this transmembrane domain acts as a stop-transfer sequence so that the invertase itself is anchored to the membrane on the luminal side of the cisterna. We chose the transmembrane domain of yeast calnexin on the assumption that this long (32-amino acid) transmembrane domain would not cause retention along the transport path. In yeast and mammalian cells, proteins with short transmembrane domains (16–18 amino acids) are retained in the ER and Golgi, whereas proteins with longer domains are allowed to proceed to the plasma membrane or vacuolar membrane unless they have other retention information (Munro, 1995; Rayner and Pelham, 1997). Changing the C terminus from KKTN to QKTN should have abolished any ER retention information if the same motif that is active in yeast and mammalian cells is also active in plant cells (Jackson et al., 1990; Gaynor et al., 1994). Progression through the secretory system would eventually allow this invertase to accumulate inside the tonoplast or outside the plasma membrane. Detachment of the invertase would result in free invertase in the vacuole or the apoplast. The results presented here (Fig. 3B) show quite clearly that the invertase was all in the vacuoles. The absence of active invertase from the microsomal fraction raises the possibility that membrane-anchored invertase may not be active, because we certainly would expect to find invertase protein in the ER and Golgi fractions. We confirmed that this was indeed the case, and a 90-kD cross-reacting polypeptide was found in the microsomal fraction. A size of 90 kD is consistent with a translation product of 632 amino acids that has a Mr of 72,100 and the presence of eight to 10 small glycans of 1,200 to 2,000Mr depending on the degree of processing (invertase has 10 possible glycosylation sites, see Fig.2A). That this 90-kD form of invertase is indeed membrane anchored was shown by freeze-thawing the microsomes repeatedly or by treating them with 0.03% Triton X-100. Such treatments solubilize soluble ER residents while still allowing the membrane proteins to be sedimented (Kreibich et al., 1973). This treatment solubilizes a considerable amount of BiP, but much of it still sedimented with the permeabilized vesicles (see Fig. 5). We postulate that this BiP is bound to proteins that are not yet completely folded and that it is therefore not readily released from the vesicles. The release of BiP from unfolded or not-yet-assembled proteins in the ER requires ATP (for review, seeGalili et al., 1998). It is entirely possible that some part of this chimeric invertase protein never folds quite “correctly” and that BiP therefore remains attached to it until such time as soluble invertase is released from its membrane anchor. The presence of soluble invertase in the vacuole indicates that invertase remained in the membrane-anchored form until it had been sorted beyond the TGN, where sorting of secreted and vacuolar proteins is thought to take place. Recent evidence obtained with soluble proteins indicates that they are sorted by receptors (Ahmed et al., 1997; Paris et al., 1997) and are postulated to pass through a PVC between the TGN and the vacuole (Conceição et al., 1997;Sanderfoot et al., 1998). Thus, detachment of invertase from its membrane anchor may have occurred in the PVC or in the vacuole, and this uncertainty is shown in the model depicted in Figure7. The presence of proteases in the lytic vacuoles of plant cells is well documented (Butcher et al., 1977;Boller and Kende, 1979; and others subsequently), and these proteases may also be active in the PVC. The fate of the transmembrane domain after invertase detachment is not known, but it may be degraded by the proteolytic system that disposes of incomplete or improperly folded proteins (Pueyo et al., 1995; Pedrazzini et al., 1997). Fig. 7. Open in new tabDownload slide Schematic diagram showing protein sorting in the secretory system. Membrane-anchored invertase reaches a compartment beyond the TGN, such as the PVC or the vacuole itself, before it is detached from the membrane. Whether invertase detachment from its membrane anchor occurs in the PVC or in the vacuole is not known. ○, Invertase. Fig. 7. Open in new tabDownload slide Schematic diagram showing protein sorting in the secretory system. Membrane-anchored invertase reaches a compartment beyond the TGN, such as the PVC or the vacuole itself, before it is detached from the membrane. Whether invertase detachment from its membrane anchor occurs in the PVC or in the vacuole is not known. ○, Invertase. We do not know if the C terminus of the soluble invertase includes amino acids that are not part of the yeast invertase translation product but came instead from the small luminal portion of calnexin used to make the fusion construct. We cannot rule out that invertase was detached earlier in the secretory pathway (in the TGN?) and that these few (putative) amino acids constituted a vacuolar targeting signal on the detached invertase. The absence of soluble invertase from the microsomes argues against this possibility, but the TGN and the PVC may be very small compartments through which proteins pass rapidly. In this case we would not expect a substantial amount of soluble invertase in the microsomal fraction, even if it were detached in the PVC or TGN. An antiserum to the membrane anchor may help resolve this issue, but our attempts to do so were unsuccessful. The scenario described above assumes that the transport of this membrane-anchored protein went through the Golgi apparatus and that sorting involved the TGN and the PVC. Based on ultrastructural evidence, a direct route from the ER to the vacuole has been proposed, at least in developing seeds that make copious quantities of vacuolar proteins (Hara-Nishimura et al., 1998). This direct route differs from the ER-derived protein bodies known to exist in the endosperm of maize and other cereals. In wheat, such protein bodies are thought to enter the vacuole through endocytosis (Levanony et al., 1992). Based on our finding that the soluble vacuolar invertase contains complex glycans (with α-1,3 Fuc and/or β-1,2 Xyl) (Fig. 7), we can conclude that the chimeric protein passed through the Golgi apparatus. The presence of complex glycans on glycoproteins of animal and plant cells is diagnostic of their passage through the Golgi apparatus. Jiang and Rogers (1998) used the same complex, glycan-specific antiserum to conclude that membrane-anchored pro-aleurain accumulated in the membranes of post-Golgi compartments. A Membrane Anchor Provides a Novel Way to Deliver a Protein to the Vacuole Because the vacuole is the largest compartment of the plant cell, it is an ideal compartment for the accumulation of proteins produced in transgenic plants if those proteins are stable in the vacuolar environment. The experiments reported here may represent a novel way to deliver a protein to the vacuole. Until now, delivery to the vacuole could only be ensured by the attachment of a vacuolar sorting signal at the N terminus or the C terminus of a soluble protein that also carries a signal peptide. The vacuolar accumulation of soluble invertase from a membrane-anchored microsomal form, indicates that it may be possible to deliver other enzymes to the vacuole in the same manner. It is apparently not necessary to use the membrane anchor of a tonoplast protein to obtain this result. Our experiments do not exclude the possibility that the length of the transmembrane domain and the characteristics of the amino acids play a role in the targeting of integral membrane proteins in cells that have more than one type of vacuole, and this issue needs to be further explored. It would also be interesting to find out if enzymes that are membrane anchored in this way are generally inactive until they are detached. ACKNOWLEDGMENTS We thank Rebecca Dyer and Bernard de la Cruz for their help in making the initial constructs, Scott Emr (University of California, San Diego) for kindly providing the yeast invertase antiserum, and Nigel Crawford (University of California, San Diego) for the gift of theA. tumefaciens strain. Abbreviations: BiP binding protein PVC pre-vacuolar compartment TGN trans-Golgi network TIP tonoplast intrinsic protein LITERATURE CITED 1 Ahmed SU Bar-Peled M Raikhel NV Cloning and subcellular location of an Arabidopsis receptor-like protein that shares common features with protein-sorting receptors of eukaryotic cells. Plant Physiol 114 1997 325 336 Google Scholar Crossref Search ADS PubMed WorldCat 2 Beevers L Raikhel NV Transport to the vacuole: receptors and trans elements. J Exp Bot 49 1998 1271 1279 Google Scholar Crossref Search ADS WorldCat 3 Boller T Kende H Hydrolytic enzymes in the central vacuole of plant cells. Plant Physiol 63 1979 1123 1132 Google Scholar Crossref Search ADS PubMed WorldCat 4 Butcher H Wagner GJ Siegelman HW Localization of acid hydrolases in protoplasts. Plant Physiol 59 1977 1098 1103 Google Scholar Crossref Search ADS PubMed WorldCat 5 Chang A Fink GR Targeting of the yeast plasma membrane [H+]ATPase: a novel gene AST1 prevents mislocalization of mutant ATPase to the vacuole. J Cell Biol 128 1995 39 49 Google Scholar Crossref Search ADS PubMed WorldCat 6 Conceição ADS Marty-Mazars D Bassham DC Sanderfoot AA Marty F Raikhel NV The syntaxin homolog AtPEP12p resides on a late post-Golgi compartment in plants. Plant Cell 9 1997 571 582 Google Scholar PubMed OpenURL Placeholder Text WorldCat 7 Dickinson C Altabella T Chrispeels MJ Slow-growth phenotype of transgenic tomato expressing apoplastic invertase. Plant Physiol 95 1991 420 425 Google Scholar Crossref Search ADS PubMed WorldCat 8 Dombrowski JE Gomez L Chrispeels MJ Raikhel NV Targeting of proteins to the vacuole. Gelvin SB Schilperoort RA Verma DPS Plant Molecular Biology Manual. 1994 Kluwer Academic Publishers Dordrecht, The Netherlands , section J3, pp 1–29 9 Dorel C Voelker TA Herman EM Chrispeels MJ Transport of proteins to the plant vacuole is not by bulk flow through the secretory system, and requires positive sorting information. J Cell Biol 108 1989 327 337 Google Scholar Crossref Search ADS PubMed WorldCat 10 Faye L Chrispeels MJ Characterization of N-linked oligosaccharides by affinoblotting with concanavalin A-peroxidase and treatment of the blots with glycosidases. Anal Biochem 145 1985 218 224 Google Scholar OpenURL Placeholder Text WorldCat 11 Faye L Chrispeels MJ Apparent inhibition of β-fructosidase secretion by tunicamycin may be explained by breakdown of the unglycosylated protein during secretion. Plant Physiol 89 1989 845 851 Google Scholar Crossref Search ADS PubMed WorldCat 12 Gabriel O Wang S Determination of enzymatic activity in polyacrylamide gels. I. Enzymes catalyzing the conversion of nonreducing substrates to reducing products. Anal Biochem 27 1969 545 554 Google Scholar Crossref Search ADS PubMed WorldCat 13 Galili G Sengupta-Gopalan C Ceriotti A The endoplasmic reticulum of plant cells and its role in protein maturation and biogenesis of oil bodies. Plant Mol Biol 38 1998 1 29 Google Scholar Crossref Search ADS PubMed WorldCat 14 Gaynor EC Heesen ST Graham TR Aebi M Emr SD Signal-mediated retrieval of a membrane protein from the Golgi to the ER in yeast. J Cell Biol 127 1994 653 665 Google Scholar Crossref Search ADS PubMed WorldCat 15 Hara-Nishimura I Shimada T Hatano K Takeuchi Y Nishimura M Transport of storage proteins to protein storage vacuoles is mediated by large precursor-accumulating vesicles. Plant Cell 10 1998 825 836 Google Scholar Crossref Search ADS PubMed WorldCat 16 Höfte H Chrispeels MJ Protein sorting to the vacuolar membrane. Plant Cell 4 1992 995 1004 Google Scholar PubMed OpenURL Placeholder Text WorldCat 17 Hunziker W Geuze HJ Intracellular trafficking of membrane proteins. BioEssays 18 1996 379 389 Google Scholar Crossref Search ADS PubMed WorldCat 18 Jackson MR Nilsson T Peterson PA Identification of a consensus motif for retention of transmembrane proteins in the endoplasmic reticulum. EMBO J 9 1990 3153 3162 Google Scholar Crossref Search ADS PubMed WorldCat 19 Jiang LW Rogers JC Integral membrane protein sorting to vacuoles in plant cells: evidence for two pathways. J Cell Biol 143 1998 1183 1199 Google Scholar Crossref Search ADS PubMed WorldCat 20 Johnson LM Bankaitis VA Emr SD Distinct sequence determinants direct intracellular sorting and modification of a yeast vacuolar protease. Cell 48 1987 875 885 Google Scholar Crossref Search ADS PubMed WorldCat 21 Keil M Sanchez-Serrano J Schell J Willmitzer L Primary structure of a proteinase inhibitor II gene from potato (Solanum tuberosum). Nucleic Acids Res 14 1986 5641 5650 Google Scholar Crossref Search ADS PubMed WorldCat 22 Klement Z Method of obtaining fluid from the intercellular spaces of foliage and the fluid's merit as substrate for phytobacterial pathogens. Phytopathology 55 1965 1033 1034 Google Scholar OpenURL Placeholder Text WorldCat 23 Kreibich G Debey P Sabatini DD Selective release of content from microsomal vesicles without membrane disassembly. J Cell Biol 58 1973 436 462 Google Scholar Crossref Search ADS PubMed WorldCat 24 Kyte J Doolittle RF A simple method for displaying the hydropathic character of a protein. J Mol Biol 157 1982 105 132 Google Scholar Crossref Search ADS PubMed WorldCat 25 Laemmli UK Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227 1970 680 685 Google Scholar Crossref Search ADS PubMed WorldCat 26 Laurière M Laurière C Chrispeels MJ Johnson KD Sturm A Characterization of a xylose-specific antiserum that reacts with the complex asparagine-linked glycans of extracellular and vacuolar glycoproteins. Plant Physiol 90 1989 1182 1188 Google Scholar Crossref Search ADS PubMed WorldCat 27 Lazo GR Stein PA Ludwig RA A DNA transformation-competent Arabidopsis genomic library in Agrobacterium. Biotechnology 9 1991 963 967 Google Scholar Crossref Search ADS PubMed WorldCat 28 Letourneur F Gaynor EC Hennecke S Demolliere C Duden R Emr SD Riezman H Cosson P Coatomer is essential for retrieval of dilysine-tagged proteins to the endoplasmic reticulum. Cell 79 1994 1199 1207 Google Scholar Crossref Search ADS PubMed WorldCat 29 Levanony H Rubin R Altshuler Y Galili G Evidence of a novel route of wheat storage proteins to vacuoles. J Cell Biol 119 1992 1117 1128 Google Scholar Crossref Search ADS PubMed WorldCat 30 Munro S An investigation of the role of transmembrane domains in Golgi protein retention. EMBO J 14 1995 4695 4704 Google Scholar Crossref Search ADS PubMed WorldCat 31 Murashige T Skoog F A revised medium for rapid growth and bioassays with tobacco tissue culture. Physiol Plant 15 1962 473 497 Google Scholar Crossref Search ADS WorldCat 32 Neuhaus JM Rogers JC Sorting of proteins to vacuoles in plant cells. Plant Mol Biol 38 1998 127 144 Google Scholar Crossref Search ADS PubMed WorldCat 33 Okita TW Rogers JC Compartmentation of proteins in the endomembrane system of plant cells. Annu Rev Plant Physiol Plant Mol Biol 47 1996 327 350 Google Scholar Crossref Search ADS PubMed WorldCat 34 Paris N Rogers SW Jiang LW Kirsch T Beevers L Phillips TE Rogers JC Molecular cloning and further characterization of a probable plant vacuolar sorting receptor. Plant Physiol 115 1997 29 39 Google Scholar Crossref Search ADS PubMed WorldCat 35 Paris N Stanley CM Jones RL Rogers JC Plant cells contain two functionally distinct vacuolar compartments. Cell 85 1996 563 572 Google Scholar Crossref Search ADS PubMed WorldCat 36 Pedrazzini E Giovinazzo G Bielli A de Virgilio M Frigerio L Pesca M Faoro F Bollini R Ceriotti A Vitale A Protein quality control along the route to the plant vacuole. Plant Cell 9 1997 1869 1880 Google Scholar PubMed OpenURL Placeholder Text WorldCat 37 Pueyo JJ Chrispeels MJ Herman EM Degradation of transport-competent destabilized phaseolin with a signal for retention in the endoplasmic reticulum occurs in the vacuole. Planta 196 1995 586 596 Google Scholar Crossref Search ADS PubMed WorldCat 38 Raikhel NV Vitale A What do proteins need to reach different vacuoles? Trends Plant Sci 4 1999 149 155 Google Scholar PubMed OpenURL Placeholder Text WorldCat 39 Rayner JC Pelham HRB Transmembrane domain-dependant sorting of proteins to the ER and plasma membrane in yeast. EMBO J 16 1997 1832 1841 Google Scholar Crossref Search ADS PubMed WorldCat 40 Roberg KJ Rowley N Kaiser CA Physiological regulation of membrane protein sorting late in the secretory pathway of Saccharomyces cerevisiae. J Cell Biol 137 1997 1469 1482 Google Scholar Crossref Search ADS PubMed WorldCat 41 Roberts CJ Nothwehr SF Stevens TH Membrane protein sorting in the yeast secretory pathway: evidence that the vacuole may be the default compartment. J Cell Biol 119 1992 69 83 Google Scholar Crossref Search ADS PubMed WorldCat 42 Robinson DG Galili G Herman E Hillmer S Topical aspects of vacuolar protein transport: autophagy and prevacuolar compartments. J Exp Bot 49 1998 1263 1270 Google Scholar Crossref Search ADS WorldCat 43 Sanderfoot AA Ahmed SU Marty-Mazars D Rapoport I Kirchhausen T Marty F Raikhel NV A putative vacuolar cargo receptor partially colocalizes with AtPEP12p on a prevacuolar compartment in Arabidopsis roots. Proc Natl Acad Sci USA 95 1998 9920 9925 Google Scholar Crossref Search ADS PubMed WorldCat 44 Schaffner AR Aquaporin function, structure, and expression: are there more surprises to surface in water relations? Planta 204 1998 131 139 Google Scholar PubMed OpenURL Placeholder Text WorldCat 45 Swanson SJ Bethke PC Jones RL Barley aleurone cells contain two types of vacuoles: characterization of lytic organelles by use of fluorescent probes. Plant Cell 10 1998 685 698 Google Scholar Crossref Search ADS PubMed WorldCat 46 Van der Wilden W Gilkes N Chrispeels MJ The endoplasmic reticulum of mungbean cotyledons: role in the accumulation of hydrolases in protein bodies during seedling growth. Plant Physiol 66 1980 390 394 Google Scholar Crossref Search ADS PubMed WorldCat 47 Voelker TA Herman EM Chrispeels MJ In vitro mutated phytohemagglutinin genes expressed in tobacco seeds: role of glycans in protein targeting and stability. Plant Cell 1 1989 95 104 Google Scholar PubMed OpenURL Placeholder Text WorldCat 48 von Schaewen A Stitt M Schmidt R Sonnewald U Willmitzer L Expression of a yeast-derived invertase in the cell wall of tobacco and Arabidopsis plants leads to accumulation of carbohydrate, inhibition of photosynthesis and strongly influences growth and phenotype of transgenic tobacco plants. EMBO J 9 1990 3033 3044 Google Scholar Crossref Search ADS PubMed WorldCat Author notes 1 This work was supported by a grant from the U.S. Department of Energy, Energy Biosciences Program. * Corresponding author; e-mail [email protected]; fax 619–534–4052. Copyright © 1999 American Society of Plant Physiologists This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)

Sung, Soon-Kee; Yu, Gyung-Hee; An, Gynheung

doi: 10.1104/pp.120.4.969pmid: 10444080

Abstract A MADS-box gene,MdMADS2, was isolated from the apple (Malus × domestica Borkh.) var Fuji and its developmental expression pattern was studied during flower development. MdMADS2 shares a high degree of amino acid sequence identity with the SQUAMOSA subfamily of genes. RNA blot analysis showed that MdMADS2 is transcribed through all stages of flower development, and its transcription was seen in the four floral organs. RNA in situ hybridization revealed that the MdMADS2 mRNA is expressed both in the inflorescence meristem and in the floral meristem. The MdMADS2transcript was detected at all stages of flower development. Protein localization analysis showed that MdMADS2 protein was excluded from the stamen and carpel primordia, in which a considerableMdMADS2 mRNA signal was detected. This indicates that posttanscriptional regulation may be involved in theMdMADS2-mediated control of flower development. Transgenic tobacco expressing the MdMADS2 gene from the cauliflower mosaic virus 35S promoter showed early flowering and shorter bolts, but did not show any homeotic changes in the floral organs. These results suggest that MdMADS2 plays an important role during early stages of flower development. Flower formation in higher plants is a complex process controlled by genetic and environmental factors (Bernier, 1988;Yanofsky, 1995; Amasino, 1996; Levy and Dean, 1998). Much has been learned from genetic and molecular studies of floral meristem and floral organ formation in Arabidopsis and snapdragon. It was found that processes of flower development are controlled by MADS-box genes that encode proteins sharing similarity with transcription factors from yeast and mammals (Schwarz-Sommer et al., 1990). The plant MADS-box genes have a conserved DNA-binding domain called MADS (MCM1, AGAMOUS, DEFICIENS, and SRF) domain and a second conserved domain called K, which is involved in protein to protein interaction (Schwarz-Sommer et al., 1990; Ma et al., 1991; Davies et al., 1996). The majority of plant MADS-box genes that have been characterized function as floral meristem or organ identity genes. The MADS-box genes, such as AP1 (APETALA1) and CAL(CAULIFLOWER) from Arabidopsis (Irish and Sussex, 1990;Mandel et al., 1992; Weigel et al., 1992; Bowman et al., 1993; Kempin et al., 1995) and SQUA (SQUAMOSA) from snapdragon (Huijser et al., 1992), are characterized as floral meristem identity genes. The MADS-box genes, such as AP3(APETALA3), PI (PISTILLATA), andAG (AGAMOUS) from Arabidopsis (Yanofsky et al., 1990; Jack et al., 1992; Goto and Meyerowitz, 1994) and DEF(DEFICIENS), GLO (GLOBOSA), andPLE (PLENA) from snapdragon (Schwarz-Sommer et al., 1990; Sommer et al., 1990; Tröbner et al., 1992; Bradley et al., 1993), are characterized as floral organ identity genes. It became evident that several other MADS-box genes have more subtle functions that are associated with floral meristem and floral organ identity. These genes include AGL2 (Flanagan and Ma, 1994),AGL4 (Savidge et al., 1995) and AGL9 (Mandel and Yanofsky, 1998) from Arabidopsis, SaMADS D from white mustard (Sinapis alba; Bonhomme, 1997), TM5 from tomato (Pnueli et al., 1991, 1994), and FBP2 from petunia (Angenent et al., 1992, 1994). Recently, MADS-box genes have been isolated from woody plants, which include DAL1 to DAL3 from Norway spruce (Tandre et al., 1995, 1998), EAP1 to EAP3 andegm1 to egm3 from eucalyptus (Eucalyptus glabra) (Kyozuka et al., 1997; Southerton et al., 1998),PrMADS1 to PrMADS3 from Monterey pine (Pinus radiata) (Mouradov et al., 1998), andMdMADS1 from Fuji apple (Malus ×domestica Borkh. var Fuji; Sung and An, 1997). Apple is one of the most economically important woody plant species, cultured for its valuable fruits. Fuji apple is the most important and widely cultivated commercial fruit in East Asia. The factors that affect the formation of flowers in apple trees are of particular interest in horticulture, but relatively little attention has been given to the molecular and genetic control of apple flower development. We previously isolated and characterized a MADS-box gene,MdMADS1, from Fuji apple, which was classified as a member of the AGL2 subfamily (Sung and An, 1997). In the present study, we report characteristics of another MADS-box gene from Fuji apple, MdMADS2, which shows high sequence similarities to the SQUA subfamily genes. MATERIALS AND METHODS Plant Materials The apple (Malus × domestica Borkh.) var Fuji was used in this study. Plant samples were provided by the Kyungbuk Provincial Rural Development Administration (Taegu, Korea). Construction of cDNA Library and Isolation ofMdMADS2 Young flower buds at 1 to 2 mm in length were used for construction of a unidirectional cDNA library according to the manufacturer's protocols (Stratagene). The initial plaque forming units were 1.9 × 106 and the average size of the inserts was 1.7 kb. The library was amplified once on agar plates and stored in a 7% DMSO solution at −70°C.MdMADS2 cDNA clone was isolated according to the method ofSung and An (1997). Overlapping subclones were created in a pBluescript SK(−) vector (Stratagene), and nucleotide sequences were determined by the dideoxynucleotide chain termination method (Sanger et al., 1977) using a kit (Sequenase version 2.0, United States Biochemicals). Comparison of the deduced amino acid sequence was performed on GenBank databases and amino acid alignment was performed using the FastDB program (IntelliGenetics, Mountain View, CA). RNA Isolation and RNA Blot Analysis Total RNA was isolated from leaves, immature flower buds, mature flowers before anthesis, and mature, post-anthesis flowers of outdoor-grown trees. Mature flowers before anthesis were dissected into their individual floral organs (sepals, petals, stamens, and carpels), and total RNA was isolated from each. RNA was extracted with a solution containing 4 m guanidine isothiocyanate and further purified by ultracentrifugation in a 5.7 m CsCl solution (Sambrook et al., 1989). For RNA hybridization, 25 μg of total RNA from each sample was separated on an agarose-formamide gel. The gel was blotted onto a Hybond-N+ nylon membrane (Amersham) and hybridized as described previously (Church and Gilbert, 1984) at 60°C for 16 h with the 548-bp DNA fragment between nucleotides 682 and 1,230 of the MdMADS2 cDNA. The blot was washed twice with a solution containing 0.2× sodium chloride/sodium phosphate/EDTA buffer and 0.1% SDS for 10 min at room temperature, followed by two washes with the same solution at 55°C for 10 min each. The blot was exposed to x-ray film with an intensifying screen at −70°C for 3 d. RNA in Situ Hybridization The 321-bp 3′ region (nucleotides 682–1,005) of theMdMADS2 cDNA was cloned into the pBluescript SK(−). This plasmid, pGA1526-3, was used as a template for synthesizing single-strand RNA. To generate an antisense-RNA probe, pGA1526-3 was linearized with BamHI and an approximately 370-bp single-stranded RNA was synthesized by T7 RNA polymerase using a DIG RNA labeling kit (Boehringer Mannheim). To generate a sense strand probe, the pGA1526-3 plasmid was linearized with XhoI, and an approximately 350-bp DIG-labeled sense strand RNA was synthesized using T3 polymerase. Flower buds at different developmental stages were fixed and paraffin embedded according to the method of Dixon et al. (1995). The samples were sectioned 10 μm thick, dewaxed, and rehydrated. The sections were incubated with proteinase K solution (100 mm Tris-HCl, pH 7.5, 50 mm EDTA, and 20 μg mL−1proteinase K) at 37°C for 30 min and then acetylated with 0.25% acetic anhydride in 100 mm triethanolamine, pH 8.0, for 5 min at room temperature. The sections were washed with 2× SSC and dehydrated through an ethanol series. The sections were pre-hybridized in a hybridization solution (50% deionized formamide, 10% dextran sulfate, 300 mm NaCl, 20 mm Tris-HCl, pH 8.0, 10 mm NaPO4, pH 8.0, 10 mm DTT, 5 mm EDTA, 1× Denhardt's solution, 500 μg mL−1 tRNA, 100 μg mL−1 salmon-sperm DNA, and 40 units mL−1 RNase inhibitor) at 45°C for 2 h, and hybridized with 0.8 μg mL−1 RNA probe at 48°C for 16 h. The sections were washed with 3× SSC for 5 min at room temperature and incubated with NTE buffer (500 mmNaCl, 10 mm Tris-HCl, and 1 mm EDTA, pH 7.5) containing 50 μg mL−1 RNase A at 37°C for 30 min. After RNase treatment, sections were washed with NTE buffer three times for 5 min at room temperature and followed by washing with 0.2× SSC at 60°C for 1 h. For detection of hybridization signals, the sections were incubated in a color development solution (10% of polyvinyl alcohol [Mr 70,000–100,000], 100 mm Tris-HCl, pH 9.0, 100 mmNaCl, 5 mm MgCl2, 0.2 mm 5-bromo-4-chloro-3-indolylphosphate, and 0.2 mm nitroblue tetrazolium) at 30°C for 8 to 12 h. The slides were extensively washed with distilled water and dehydrated and mounted with nonaqueous mounting medium using standard protocols. Slides were photographed on a microscope (Labophot-2, Nikon) using bright-field optics. Antibody Preparation To express a truncated form of the MdMADS2 protein (amino acid residues 66–256) lacking the MADS domain, the 802-bpBamHI-XhoI fragment was cloned into the T7 expression vector pRSET C (Invitrogen, Carlsbad, CA). This construct was introduced into the Escherichia coli strain BL21 (pLysS) and protein was induced with 1 mm IPTG. The overexpressed MdMADS2 protein was purified on a Ni-affinity column (Invitrogen) according to the manufacturer's protocol. Isolated protein was further purified by electrophoresis on 12% SDS-PAGE as follows. After electrophoresis, the gel was soaked in cold 100 mm KCl and incubated at 4°C until protein bands were visualized. The protein band corresponding to the truncated MdMADS2 protein was excised from the gel and the protein was extracted with 100 mm KCl. The protein was concentrated with Centriprep columns (Amicon, Beverly, MA) with a cutoff size of 10 kD. Rats were immunized with 40 μg of protein per injection and were boosted a total of five times at 2-week intervals; the fifth boosting was done without adjuvant. Antibodies for the MdMADS3 (accession no.U78949), MdMADS4 (accession no. U78950), and OsMADS1 (accession no. L34271) were also prepared according to the same procedure described above. The antibodies were affinity purified using a western purification procedure (Burke et al., 1982). Cross-reaction between the antibodies was examined by protein blot analysis. One microgram of purified MADS (MdMADS2, MdMADS3, MdMADS4, and OsMADS1) proteins lacking the MADS domain were separated on a 12% SDS-PAGE and blotted onto a Hybond-C nitrocellulose membrane (Amersham). The membrane was incubated with MdMADS2 antibodies (1:5,000 dilution) and immunodetection was carried out using the secondary antibody (peroxidase-labeled affinity purified antibody to rat IgG [H+L]; KPL, Gaithersburg, MD) and a chemiluminescence system (Amersham). The membrane was then exposed to x-ray film for 10 s. Protein Immunolocalization For detection of immunoreactive proteins, the sections were treated with 3% H2O2 for 20 min at room temperature. The slides were washed for 10 min in PBS buffer (10 mm phosphate, pH 7.5, and 0.9% [w/v] NaCl) and incubated in a blocking solution (5% [v/v] normal rabbit serum and 0.5% [w/v] IgG-free BSA [A-9085, Sigma] in PBS buffer) for 2 h at room temperature. The sections were incubated with an avidin/biotin blocking solution (Vector Laboratories, Burlingame, CA) to prevent nonspecific binding of avidin/biotin according to the manufacturer's protocols. The sections were briefly washed with PBS buffer and incubated with primary antibodies (150 μL of the affinity-purified first antibodies were mixed with 10 mL of the blocking solution) for 2 h at room temperature. After incubation with primary antibodies, bound antibodies were detected using a rat IgG ABC (avidin:biotinylated enzyme complex) kit (Vectastain Elite, Vector Laboratories) according to the manufacturer's protocol. Detection of the amplified signal was carried out with a substrate kit (VIP, Vector Laboratories) according to the manufacturer's protocol. The stained sections were washed with distilled water for 5 min, and then dehydrated and mounted in nonaqueous mounting medium using standard protocols. Slides were examined on a microscope (Labophot-2, Nikon) using bright-field optics. Plant Transformation The MdMADS2 cDNA was cloned into the expression vector pGA1530, a binary vector containing the 35S promoter and T7 terminator, and the npt (neomycin phosphotransferase) gene as a selective maker (An et al., 1988). Agrobacterium tumefaciensLBA4404 (Hoekema et al., 1983) was used for transformation of tobacco (Nicotiana tabacum L. cv Petit Havana SR1 and cv Xanthi) plants. Transgenic plants were maintained in greenhouse conditions. Polyamine Determination Polyamines were extracted from the tobacco leaves, dansylated, solvent purified, separated by TLC, and quantified with a fluorometer according to the method of Goren et al. (1982). The level of soluble protein was measured by the method of Bradford (1976) using BSA as a standard. RESULTS Sequence Analysis of the MdMADS2 cDNA The MdMADS2 cDNA clone is 1230 bp long and contains an ORF of 256 amino acid residues with a 233-bp 5′ leader region and a 229-bp 3′ UTR (accession no. U78948). Alignment of the deduced amino acid sequence of MdMADS2 and other MADS proteins is represented in Figure 1. The MdMADS2 protein has the MADS domain located between amino acid residues 2 and 56 and the K domain located between amino acid residues 93 and 158. The highest overall amino acid identity (76%) was shown by the sequence from silver birch BpMADS5 (accession no. X99655). The MADS domain shares an identity of 96%. The MdMADS2 protein shows over 60% overall amino acid identity to TM4 from tomato (Pnueli et al., 1991), PTOM1-1 from potato (Kang and Hannapel, 1995), SLM5 from white campion (Hardenack et al., 1994), AGL8/FRUITFULL from Arabidopsis (Mandel and Yanofsky, 1995b; Gu et al., 1998), SaMADS B from white mustard (Menzel et al., 1996), SQUA from snapdragon (Huijser et al., 1992), AP1 from Arabidopsis (Mandel et al., 1992), and Boi2AP1 from broccoli (Carr and Irish, 1997), which are the SQUA subfamily members (Theissen et al., 1996). In the MADS domain there are more significant similarities (over 90% amino acid identity) between MdMADS2 and the SQUA subfamily members. Fig. 1. Open in new tabDownload slide Comparison of the amino acid sequence of the MdMADS2 protein with other MADS proteins in the SQUA subfamily. A, Sequence alignment of the MADS domain. Shown here are the MADS-box sequences of MdMADS2 (accession no. U78948), silver birch BpMADS5 (accession no. X99655), tomato TM4 (accession no. X60757), potato PTOM1-1 (accession no. U23757), white campion SLM5 (accession no.X80492), Arabidopsis AGL8 (accession no. U33473), white mustard SaMADSB (accession no. U25695), snapdragon SQUA (accession no. X63701), Arabidopsis AP1 (accession no. S35631), and broccoli Bio2AP1 (accession no. U67452). The asterisks represent amino acid residues identical to the corresponding residues in MdMADS2. The numbers at the left represent the positions of the first amino acid residues shown for each sequence. B, Alignment of the K domains. The asterisks and the numbers at the left are as represented in A. C, Alignment of the C-terminal regions. The last 20 amino acid residues are shown. The left numbers are the positions of the first amino acid residues shown for each sequence. The numbers at the right represent the percentage of identical amino acid residues with MdMADS2. Fig. 1. Open in new tabDownload slide Comparison of the amino acid sequence of the MdMADS2 protein with other MADS proteins in the SQUA subfamily. A, Sequence alignment of the MADS domain. Shown here are the MADS-box sequences of MdMADS2 (accession no. U78948), silver birch BpMADS5 (accession no. X99655), tomato TM4 (accession no. X60757), potato PTOM1-1 (accession no. U23757), white campion SLM5 (accession no.X80492), Arabidopsis AGL8 (accession no. U33473), white mustard SaMADSB (accession no. U25695), snapdragon SQUA (accession no. X63701), Arabidopsis AP1 (accession no. S35631), and broccoli Bio2AP1 (accession no. U67452). The asterisks represent amino acid residues identical to the corresponding residues in MdMADS2. The numbers at the left represent the positions of the first amino acid residues shown for each sequence. B, Alignment of the K domains. The asterisks and the numbers at the left are as represented in A. C, Alignment of the C-terminal regions. The last 20 amino acid residues are shown. The left numbers are the positions of the first amino acid residues shown for each sequence. The numbers at the right represent the percentage of identical amino acid residues with MdMADS2. Sequence comparison of the MADS family genes showed that the 3′ portion of the genes is the most divergent. A 548-bp cDNA fragment (nucleotides 682–1230) of the MdMADS2 cDNA clone was used as a probe to determine whether the fragment is gene-specific. Genomic DNA blot analysis revealed that one HindIII, one BamHI, and one PstI fragment hybridized with the probe, indicating that the 3′ region is gene specific (data not shown). To avoid cross-hybridization, therefore, the cDNA fragment containing only the 3′ region of the gene was used as a probe for analysis of the expression pattern of MdMADS2. RNA Blot Analysis Flower bud formation in apple trees occurs in the previous growing season. The morphological differentiation proceeds until dormancy occurs in winter. In spring, differentiation resumes and rapid development takes place prior to anthesis. The mRNA expression pattern of MdMADS2 was examined in flower buds that resumed growth in the spring (March–May). The flower buds were sampled on the basis of the bud length and morphological events, as previously described (Sung and An, 1997). As shown in Figure2, the MdMADS2 mRNA was not detectable in mature leaves (lane 1), whereas the transcript was detected in flower buds at all five stages of flower development (lanes 2–6). The intensity of the RNA band was strongest at stage 1 (lane 2) and receded as the flower developed. In mature flowers the expression was relatively strong in sepals and petals but weak in stamens and carpels (lanes 7–10). Fig. 2. Open in new tabDownload slide A, RNA blot analysis of the MdMADS2transcript in apple flower bud. Lane 1, Mature leaves; lane 2, stage 1 bud (length = 4– 5 mm); lane 3, stage 2 bud (length = 7–8 mm); lane 4, stage 3 bud (petals begin to emerge from sepals); lane 5, mature flowers; lane 6, post-anthesis flowers; lane 7, sepals; lane 8, petals; lane 9, stamens; and lane 10, carpels of mature flower. B, rRNA stained with ethidium bromide showing the equivalence of RNA loading between the lines. Fig. 2. Open in new tabDownload slide A, RNA blot analysis of the MdMADS2transcript in apple flower bud. Lane 1, Mature leaves; lane 2, stage 1 bud (length = 4– 5 mm); lane 3, stage 2 bud (length = 7–8 mm); lane 4, stage 3 bud (petals begin to emerge from sepals); lane 5, mature flowers; lane 6, post-anthesis flowers; lane 7, sepals; lane 8, petals; lane 9, stamens; and lane 10, carpels of mature flower. B, rRNA stained with ethidium bromide showing the equivalence of RNA loading between the lines. RNA in Situ Localization The temporal and spatial expression patterns of theMdMADS2 transcript were determined by RNA in situ hybridization. A DIG-labeled MdMADS2 antisense RNA probe was used. The MADS domain and the K domain were not included to avoid cross-hybridization with other MdMADS genes. Apple has a determinate inflorescence with a terminal flower and a tendency toward dichasial branching (Pratt, 1988). An early stage of apple flower development occurs during the previous growing season. The development of the flower buds at an early stage can be divided into three stages: evocation of the inflorescence meristem (stage 1), differentiation into flower primordia (stage 2), and sequential initiation of sepal, petal, stamen, and carpel primordia in the floral meristem (stage 3). At stage 1, the MdMADS2 transcript is present throughout the inflorescence meristem, the bud procambium, and the adjacent leaf appendages, but is more concentrated in the early floral meristem arising from the inflorescence meristem (Fig.3A). At stage 2, MdMADS2continues to be expressed in all parts of the flower bud, including flower primordia, bracts, and leaf appendages (Fig. 3B). The signal is high in the floral meristem (Fig. 3B). At early stage 3, when the sepal and petal primordia arise and differentiate, the signal is detected at a high level in the region of the floral meristem interior to the sepal and petal primordia, but at a low level in the sepal and petal primordia (Fig. 3C). At late stage 3, when differentiation of the four floral organ primordia becomes apparent, MdMADS2 is expressed weakly throughout the sepal, petal, carpel, and stamen primordia (Fig. 3D). Fig. 3. Open in new tabDownload slide In situ hybridization patterns ofMdMADS2 mRNA in a developing flower bud. Each section was photographed using a blue filter under bright-field optics. The transcript signal is blue. Bar = 300 μm. A, Stage 1 flower bud with the inflorescence meristem. The arrowhead indicates the early floral meristem emerging from the inflorescence meristem. B, Stage 2 flower bud with young flower primordia. C and D, Flower buds at early and late stage 3, respectively. Floral organs arise and differentiate at these stages. E and F, Stage 4 flower bud with fused carpels and mature flower, respectively. G through I, Hybridization with a sense probe of MdMADS2 in flower bud at stage 1 (G), stage 3 (H), and stage 4 (I). a, Anther; b, bract; c, carpel; f, filament; fp, flower primordium; ft, floral tube; g, gynoecium; im, inflorescence meristem; l, leaf appendage; o, ovule; p, petal; pc, procambium; s, style; se, sepal; st, stamen. Fig. 3. Open in new tabDownload slide In situ hybridization patterns ofMdMADS2 mRNA in a developing flower bud. Each section was photographed using a blue filter under bright-field optics. The transcript signal is blue. Bar = 300 μm. A, Stage 1 flower bud with the inflorescence meristem. The arrowhead indicates the early floral meristem emerging from the inflorescence meristem. B, Stage 2 flower bud with young flower primordia. C and D, Flower buds at early and late stage 3, respectively. Floral organs arise and differentiate at these stages. E and F, Stage 4 flower bud with fused carpels and mature flower, respectively. G through I, Hybridization with a sense probe of MdMADS2 in flower bud at stage 1 (G), stage 3 (H), and stage 4 (I). a, Anther; b, bract; c, carpel; f, filament; fp, flower primordium; ft, floral tube; g, gynoecium; im, inflorescence meristem; l, leaf appendage; o, ovule; p, petal; pc, procambium; s, style; se, sepal; st, stamen. In spring the morphological differentiation of stamens and carpels is accelerated, which is the critical stage in the development of apple flower buds (stage 4). At this stage, stamens develop into anthers and filaments, and the individual carpel fuse with each other to form pistils. At stage 4, MdMADS2 RNA is uniformly expressed at low levels throughout flower, such as the perianth (sepals and petals), the reproductive organs (stamens and fused carpels), and the floral tube (Fig. 3E). As floral organs mature, the overall hybridization signal becomes weaker (Fig. 3F). In situ expression ofMdMADS2 is detected in all four floral organs of the mature flower, which we have shown by the RNA blot analysis. Inflorescence meristem hybridized with the MdMADS2 sense probe showed no signal above background (Fig. 3G). We also observed that the sense probe did not show any significant hybridization signal to the sections from the stage of stamen and carpel development and the stage of mature flower development (Fig. 3, H and I). Immunolocalization To study the expression pattern of MdMADS2 at a protein level, polyclonal antibodies were raised against a truncated form of MdMADS2 lacking the conserved MADS domain. The primary antibodies were affinity purified with antigen and used for further studies. Protein blot analysis was conducted to determine whether the antibodies cross-reacted with other MADS proteins, with OsMADS1 from rice (Chung et al., 1994), or with MdMADS3 and MdMADS4 from apple (S.-K. Sung, G.-H. Yu, and G. An, unpublished data). In all cases, the truncated proteins lacking the MADS domain were expressed in E. coli. As shown in Figure4, the antibodies reacted only with MdMADS2 and did not recognize other MADS proteins. We also tested cross-reactivity with other antibodies prepared against the truncated MdMADS3, MdMADS4, and OsMADS1 proteins. These antibodies immunoreacted with their own proteins, but did not recognize the MdMADS2 protein (data not shown). Therefore, the MdMADS2 antibodies generated in this study are specific for detection of MdMADS2. Fig. 4. Open in new tabDownload slide MdMADS2 protein expressed in E. coli and protein gel blots of MdMADS2 and other MADS proteins with affinity-purified anti-MdMADS2 antibodies. A, Coomassie Blue-stained protein gel showing truncated MdMADS2 protein expressed inE. coli. Lane M, Mr marker; lane −, before IPTG induction; lane +, after IPTG induction; lane P, purified MdMADS2 protein. B, Protein gel blots of truncated MADS proteins with affinity-purified anti-MdMADS2 antibodies. Each protein was loaded in two adjacent lanes. Md2, MdMADS2; Md3, MdMADS3; Md4, MdMADS4; Os1, OsMADS1. Fig. 4. Open in new tabDownload slide MdMADS2 protein expressed in E. coli and protein gel blots of MdMADS2 and other MADS proteins with affinity-purified anti-MdMADS2 antibodies. A, Coomassie Blue-stained protein gel showing truncated MdMADS2 protein expressed inE. coli. Lane M, Mr marker; lane −, before IPTG induction; lane +, after IPTG induction; lane P, purified MdMADS2 protein. B, Protein gel blots of truncated MADS proteins with affinity-purified anti-MdMADS2 antibodies. Each protein was loaded in two adjacent lanes. Md2, MdMADS2; Md3, MdMADS3; Md4, MdMADS4; Os1, OsMADS1. It is possible that the transcription pattern of MdMADS2does not reflect its protein level if the gene is regulated at the posttranscriptional level. Therefore, the expression pattern of the MdMADS2 protein was examined by in situ immunolocalization. At stage 1, MdMADS2 protein is seen in the inflorescence meristem, the bud procambium, and the adjacent leaf appendages in the flower bud (Fig.5A), which is consistent with the results of MdMADS2 mRNA in situ hybridization shown in Figure 3A. In the bud with the vegetative apex (leaf bud), the protein signal was not detected in the apical stem meristem, bud procambium, or the adjacent leaf appendages (Fig. 5C). Expression of MdMADS2 continues throughout the young flower primordia and the emerging sepal primordia by stage 2 (Fig. 5E). Fig. 5. Open in new tabDownload slide In situ immunolocalization of the MdMADS2 protein in a developing flower bud. Sections were hybridized with affinity-purified anti-MdMADS2 antibodies. Sections were photographed using a blue filter under bright-field optics. The protein signal is purple. Bar = 100 μm. A, Flower bud with the inflorescence meristem (stage 1). B, Preimmune serum control for A. C, Vegetative bud with the stem meristem. D, Preimmune serum control for C. E, Flower bud at stage 2. F, Preimmune serum control for E. G, Flower bud at stage 3. H, Preimmune serum control for G. b, Bract; c, carpel primordium; im, inflorescence meristem; fp, flower primordium; l, leaf appendage; p, petal primordium; pc, procambium; rm, rib meristem; se, sepal primordium; sm, stem meristem; st, stamen primordium. Fig. 5. Open in new tabDownload slide In situ immunolocalization of the MdMADS2 protein in a developing flower bud. Sections were hybridized with affinity-purified anti-MdMADS2 antibodies. Sections were photographed using a blue filter under bright-field optics. The protein signal is purple. Bar = 100 μm. A, Flower bud with the inflorescence meristem (stage 1). B, Preimmune serum control for A. C, Vegetative bud with the stem meristem. D, Preimmune serum control for C. E, Flower bud at stage 2. F, Preimmune serum control for E. G, Flower bud at stage 3. H, Preimmune serum control for G. b, Bract; c, carpel primordium; im, inflorescence meristem; fp, flower primordium; l, leaf appendage; p, petal primordium; pc, procambium; rm, rib meristem; se, sepal primordium; sm, stem meristem; st, stamen primordium. The MdMADS2 protein was clearly localized to the nucleus (data not shown). When differentiation of the four floral organ primordia became apparent (stage 3), the MdMADS2 protein was highly expressed in developing sepals and petals, but was hardly detectable in the regions of the emerging stamen and carpel primordia (Fig. 5G). Considering the RNA in situ hybridization result as shown in Figure 3, C and D, which shows a significant expression of MdMADS2 mRNA in stamen and carpel primordia, it is likely that posttranscriptional regulation is involved in controlling MdMADS2 expression in the sites of stamen and carpel primordia. During this stage, the rib meristem is very active in the formation of the elongated receptacle, and a higher level of MdMADS2 expression was observed. The protein signal was also detected in bracts and leaf appendages. Only low background was detected in control experiments utilizing preimmune serum as the primary antibody, demonstrating that the signals observed with the MdMADS2 antibody are specific (Fig. 5, B, D, F and H). Ectopic Expression in Tobacco We used a transgenic approach to study the influence of expression of MdMADS2 on the development of the plants. Because it is a long procedure to generate and analyze transgenic apple trees and their flowers, we employed a heterologous tobacco system. TheMdMADS2 cDNA was placed under the control of the CaMV 35S promoter and the chimeric molecule was introduced to the tobacco genome by A. tumefaciens co-cultivation. More than 20 independent transformants were regenerated from two cultivars, SR1 and Xanthi. Five lines that showed a dwarf phenotype and two lines with weak phenotypes were selected from each cultivar for further analysis. The five lines that showed the severe dwarf phenotype had significantly higherMdMADS2 expression levels than those of the weak phenotype plants (Fig. 6). The SR1 line 4 and Xanthi lines 4 and 5 that showed a high level of transgene expression and the severe dwarf phenotype were self-pollinated and phenotypes of offspring were analyzed. The dwarf phenotype was inherited in the next generation as a dominant Mendelian trait and co-segregated with the kanamycin-resistance gene (data not shown). Fig. 6. Open in new tabDownload slide RNA blot analysis of primary transgenic tobacco plants. The numbers indicate independent transgenic tobacco lines of two different cultivars, SR1 (A) and Xanthi (B). Twenty micrograms of total RNA was isolated from mature leaves and hybridized with the32P-labeled MdMADS2 probe. As a control, a probe prepared from the TSC29 cDNA (Gao et al., 1994) was hybridized to the same membrane after washing off theMdMADS2 probe. Fig. 6. Open in new tabDownload slide RNA blot analysis of primary transgenic tobacco plants. The numbers indicate independent transgenic tobacco lines of two different cultivars, SR1 (A) and Xanthi (B). Twenty micrograms of total RNA was isolated from mature leaves and hybridized with the32P-labeled MdMADS2 probe. As a control, a probe prepared from the TSC29 cDNA (Gao et al., 1994) was hybridized to the same membrane after washing off theMdMADS2 probe. The phenotypes of the homozygous progeny are described in TableI. The morphological changes in transgenic plants were not seen until the appearance of several leaves (data not shown). At flowering time, the plant height of the SR1 line 4 was reduced by an average of 30 cm and that of Xanthi line 4 was reduced by 13 cm (Fig. 7). This reduction seemed to coincide with the decrease in node number; the SR1 line 4 and Xanthi line 4 produced one-half to two-thirds the number of nodes produced by wild-type plants (Table I). The node length was similar to that of wild-type plants. We observed that the number of days to flowering time was shortened by 8 to 10 d in transgenic plants. Furthermore, floral development occurred at the axillary buds positioned at lower nodes, which give rise to leaf primordia in wild-type plants. Therefore, it appears that overexpression of theMdMADS2 gene promoted the commitment of indeterminate meristems to flower and at the same time arrested the vegetative growth. Other phenotypic alterations in vegetative parts include smaller and greener leaves. There was no visible alteration in floral organ morphology. When immunolocalization assays were performed MdMADS2 protein accumulated in transgenic plants (data not shown), whereas there was no MdMADS2 protein signal in wild-type plants (data not shown). Table I. Phenotypes of transgenic tobacco plants Transgenic Line . Days to Flowering-a . Height-b . Nodes-c . Contents of Polyamines-d . cm n μg mg−1 protein cv SR1  Wild type 66.4  ± 0.9 85.0  ± 5.7 16.0  ± 1.3 3.7  ± 0.9  Transgenic S4 56.2  ± 1.0 55.3  ± 4.3 8.0  ± 0.7 14.5  ± 1.7 cv Xanthi  Wild type 86.3  ± 3.8 88.0  ± 2.2 39.0  ± 2.1 9.3  ± 2.2  Transgenic X4 78.2  ± 2.6 75.4  ± 5.1 28.0  ± 1.7 12.6  ± 2.9 Transgenic Line . Days to Flowering-a . Height-b . Nodes-c . Contents of Polyamines-d . cm n μg mg−1 protein cv SR1  Wild type 66.4  ± 0.9 85.0  ± 5.7 16.0  ± 1.3 3.7  ± 0.9  Transgenic S4 56.2  ± 1.0 55.3  ± 4.3 8.0  ± 0.7 14.5  ± 1.7 cv Xanthi  Wild type 86.3  ± 3.8 88.0  ± 2.2 39.0  ± 2.1 9.3  ± 2.2  Transgenic X4 78.2  ± 2.6 75.4  ± 5.1 28.0  ± 1.7 12.6  ± 2.9 The homozygous progeny (T2 generation) from the primary transformed line (T1 generation) were germinated in a peat pellet and maintained in greenhouse conditions. Progeny carrying the transgenes were identified by visually scoring T3 seedlings for kanamycin resistance. Wild-type tobacco plants were used as controls. Each value represents the mean ± sd(n = 10). The experiment was repeated three times with similar results. F0-a Days to flowering is defined as the time from seed sowing to the time when the first petals opened. F0-b Height and F0-c number of nodes were measured when fruits were fully developed. F0-d Contents of polyamines (putrescine + spermidine + spermine) were measured in the fifth and sixth leaves of plants. Open in new tab Table I. Phenotypes of transgenic tobacco plants Transgenic Line . Days to Flowering-a . Height-b . Nodes-c . Contents of Polyamines-d . cm n μg mg−1 protein cv SR1  Wild type 66.4  ± 0.9 85.0  ± 5.7 16.0  ± 1.3 3.7  ± 0.9  Transgenic S4 56.2  ± 1.0 55.3  ± 4.3 8.0  ± 0.7 14.5  ± 1.7 cv Xanthi  Wild type 86.3  ± 3.8 88.0  ± 2.2 39.0  ± 2.1 9.3  ± 2.2  Transgenic X4 78.2  ± 2.6 75.4  ± 5.1 28.0  ± 1.7 12.6  ± 2.9 Transgenic Line . Days to Flowering-a . Height-b . Nodes-c . Contents of Polyamines-d . cm n μg mg−1 protein cv SR1  Wild type 66.4  ± 0.9 85.0  ± 5.7 16.0  ± 1.3 3.7  ± 0.9  Transgenic S4 56.2  ± 1.0 55.3  ± 4.3 8.0  ± 0.7 14.5  ± 1.7 cv Xanthi  Wild type 86.3  ± 3.8 88.0  ± 2.2 39.0  ± 2.1 9.3  ± 2.2  Transgenic X4 78.2  ± 2.6 75.4  ± 5.1 28.0  ± 1.7 12.6  ± 2.9 The homozygous progeny (T2 generation) from the primary transformed line (T1 generation) were germinated in a peat pellet and maintained in greenhouse conditions. Progeny carrying the transgenes were identified by visually scoring T3 seedlings for kanamycin resistance. Wild-type tobacco plants were used as controls. Each value represents the mean ± sd(n = 10). The experiment was repeated three times with similar results. F0-a Days to flowering is defined as the time from seed sowing to the time when the first petals opened. F0-b Height and F0-c number of nodes were measured when fruits were fully developed. F0-d Contents of polyamines (putrescine + spermidine + spermine) were measured in the fifth and sixth leaves of plants. Open in new tab Fig. 7. Open in new tabDownload slide Ectopic phenotypes of transgenic tobacco plant. A, Homozygous transgenic line S4 (left two plants) and wild type (right two plants) of the tobacco cv SR1. B, Homozygous transgenic line X4 (left two plants) and wild type (right two plants) of the tobacco cv Xanthi. Fig. 7. Open in new tabDownload slide Ectopic phenotypes of transgenic tobacco plant. A, Homozygous transgenic line S4 (left two plants) and wild type (right two plants) of the tobacco cv SR1. B, Homozygous transgenic line X4 (left two plants) and wild type (right two plants) of the tobacco cv Xanthi. Polyamine is known as a growth substance involved in apple flower bud development. We therefore examined polyamine (putrescine + spermidine + spermine) contents in transgenic plants. Transgenic plants exhibited elevated levels of the polyamine contents in leaves that were 1.5- to 4-fold over those of the wild type (Table I). DISCUSSION The MADS genes are components of complex networks of genes that play an important role in flower development and are found in a variety of plant species. We report the isolation and characterization of a new MADS-box gene from Fuji apple, MdMADS2. A multiple alignment of MdMADS2 with other MADS proteins revealed that MdMADS2 showed high sequence relatedness to various MADS proteins in the SQUA subfamily (Theissen et al., 1996), which includes TM4 from tomato (Pnueli et al., 1991), PTOM1-1 from potato (Kang and Hannapel, 1995), SLM5 from white campion (Hardenack et al., 1994), AGL8/FRUITFULL from Arabidopsis (Mandel and Yanofsky, 1995b; Gu et al., 1998), SaMADS B from white mustard (Menzel et al., 1996), SQUA from snapdragon (Huijser et al., 1992), AP1 from Arabidopsis (Mandel et al., 1992), and Boi2AP1 from broccoli (Carr and Irish, 1997). MdMADS2 shares over 60% overall amino acid identity with the product of these genes and the functional MADS domain shows over 90% amino acid identity. The most fully characterized MADS genes in the SQUAsubfamily are SQUA of snapdragon and AP1 of Arabidopsis, which control floral meristem identity (Huijser et al., 1992; Mandel et al., 1992). SQUA and AP1 are expressed in floral meristems as soon as they form on the flanks of the inflorescence meristem (Huijser et al., 1992; Mandel et al., 1992). In contrast, AGL8/FRUITFULL, which diverged from AP1via gene duplication of Arabidopsis, is expressed in the inflorescence meristem but not in the floral meristem, suggesting that it may be involved in the inflorescence meristem identity (Mandel and Yanofsky, 1995b; Purugganan et al., 1995). A loss-of-function analysis revealed further that the AGL8/FRUITFULL gene is required for cellular differentiation during fruit and leaf development (Gu et al., 1998). Unlike SQUA, AP1, and AGL8/FRUITFULL,MdMASDS2 is expressed in both the inflorescence meristem and floral meristems in apple flower buds. Expression of SLM5, aSQUA homolog from white campion, has also been detected both in the inflorescence and and in the floral meristems (Hardenack et al., 1994). It has been suggested that the expression of SLM5both in inflorescence meristems and floral meristems in white campion may reflect the structure of the inflorescence (Hardenack et al., 1994). White campion has a determinate dichasial inflorescence in which the apical meristem forms a flower flanked by two lateral inflorescence meristems, whereas snapdragon and Arabidopsis have an indeterminate apical inflorescence meristem from which flowers develop on the flanks. Apple has a determinate inflorescence with a terminal flower and a tendency toward dichasial branching (Pratt, 1988). Thus, the determinate inflorescence of apple could result in expression ofMdMADS2 in both the inflorescence and the floral meristems. In subsequent stages, the MdMADS2 transcript is present in all of the floral organs and organ primordia, including the stamen. We also observed MdMADS2 expression in individually dissected stamens and the other floral organs on northern blots. This feature was unexpected given the expression patterns of SQUA,AP1, AGL8/FRUITFULL, and SLM5; these genes are not expressed in stamens at any developmental stage (Mandel et al., 1992; Hardenack et al., 1994; Kempin et al., 1995; Mandel and Yanofsky, 1995b). In contrast to these genes, BoiAP1 from broccoli, a member of the Brassicaceae, is expressed in the stamen primordia (Carr and Irish, 1997). Moreover, expression ofTM4, a SQUA homolog of tomato, is observed in all of the floral organs when plants are grown at low temperature (Lozano et al., 1998). At standard temperatures, transcripts of TM4 are not detected in total RNA from any floral organ (Pnueli et al., 1991). Regulation of homeotic gene expression by hormonal or environmental factors has been suggested in several studies (Estruch et al., 1993;Okamuro et al., 1996; Venglat and Sawhney, 1996). It is generally accepted that the development of apple flower buds is the process most sensitive to hormonal, nutritional, and environmental factors, because induction and development of flower buds occur on the trees carrying a heavy crop of fruit (Buban and Faust, 1982; Pratt, 1988). Therefore, it is possible that expression patterns of MdMADS2 in apple are not only affected by growth pattern, such as the structure of the inflorescence, but are also somehow affected by hormonal, nutritional, and environmental factors. The diversity in the expression patterns of MdMADS2 and other genes in the SQUA subfamily could result from their divergent functions. If the genes are regulated at posttranscriptional level, however, the transcription patterns may not reflect the region of the plant where the protein is present. Therefore, we examined the expression pattern of the MdMADS2 protein. At early developmental stages of the floral meristem, the accumulation patterns of the protein were generally in agreement with those of theMdMADS2 mRNA. However, we observed a discrepancy between the transcript level and the protein level after stamen and carpel differentiation. The MdMADS2 protein was not detectable in the stamen and carpel primordia, where a significant RNA signal was detected. These results indicate the possibility that a posttranscriptional regulatory mechanism may be involved in theMdMADS2 expression in stamen and carpel primordia. Posttranscriptional regulation of plant gene expression is not uncommon. In situ experiments of the petunia fbp1(floral binding protein) gene demonstrated thatfbp1 gene expression is posttranscriptionally regulated in stamen primordia at later stages of development (Canas et al., 1994). Furthermore, posttranscriptional regulation of the AP3 gene has been shown in transgenic Arabidopsis plants ectopically expressingAP3. AP3 mRNA was detected throughout the inflorescence and vegetative tissues, but the AP3 protein was only detected in the second, third, and fourth floral whorls (Jack et al., 1994). Krizek and Meyerowitz (1996) proposed that the posttranscriptional regulation ofAP3 is modulated by the formation of a heterodimer with the PI protein. Although we were unable to test the functions of the appleMdMADS2 by mutation analysis, it is possible to predict the putative functions of the gene by analysis of its expression patterns. The expression patterns of MdMADS2 at the RNA and protein levels suggest that MdMADS2 is active in inflorescence and floral meristems at early developmental stages and its activity is modulated in stamen and carpel primordia by posttranscriptional regulation. Ectopic expression is a useful method with which to analyze the influence of expression of genes that regulate developmental processes. Transgenic tobacco plants ectopically expressing MdMADS2exhibited early flowering and development of auxiliary flowers. There was no visible change in the floral organs. Mandel and Yanofsky (1995a)demonstrated that ectopic expression of AP1 is not only sufficient to convert apical and lateral shoots into flowers, but also to cause an early flowering phenotype in Arabidopsis. Several other MADS-box genes have been shown to promote early flowering when overexpressed (Chung et al., 1994; Kyozuka et al., 1997; Kater et al., 1998; Tandre et al., 1998). In addition to being involved in early flowering, ectopic expression of MdMADS2 is involved in production of smaller and greener leaves and increase in the polyamine content in transgenic plants. It is generally believed that polyamines function as growth substances and promote flower development in higher plants (Smith, 1985). Polyamines are also involved in morphogenesis (Evans and Malmberg, 1989) and bud break in perennial trees (Wang and Faust, 1994). ACKNOWLEDGMENTS We thank Min-Soon Huh and Kyong-Bai Choi for providing apple samples, Mahru An for critical reading of the manuscript, and Jongmin Nam and Dong-Hoon Jeong for sample preparation. Abbreviations: DIG digoxigenin IPTG isopropyl β-d-thio-galactopyranoside LITERATURE CITED 1 Amasino RM Control of flowering time in plants. Curr Opin Genet Dev 6 1996 480 487 Google Scholar Crossref Search ADS PubMed WorldCat 2 An G Ebert PR Mitra A Ha S-B Binary vectors. Gelvin SB Schilperoort R Plant Molecular Biology Manual. 1988 Kluwer Academic Publishers Dordrecht, The Netherlands , section A3, pp 1–19 3 Angenent GC Busscher M Franken J Mol JNM van Tunen AJ Differential expression of two MADS box genes in wild-type and mutant petunia flowers. Plant Cell 4 1992 983 993 Google Scholar PubMed OpenURL Placeholder Text WorldCat 4 Angenent GC Franken J Busscher M Weiss D van Tunen AJ Co-suppression of the petunia homeotic gene fbp2 affects the identity of the generative meristem. Plant J 5 1994 33 44 Google Scholar Crossref Search ADS PubMed WorldCat 5 Bernier G The control of floral evocation and morphogenesis. Annu Rev Plant Physiol Plant Mol Biol 39 1988 175 219 Google Scholar Crossref Search ADS WorldCat 6 Bonhomme F Sommer H Bernier G Jacqmard A Characterization of SaMADS D from Sinapis alba suggests a dual function of the gene: in inflorescence development and floral organogenesis. Plant Mol Biol 34 1997 573 582 Google Scholar Crossref Search ADS PubMed WorldCat 7 Bowman JL Alvarez J Weigel D Meyerowitz EM Smyth DR Control of flower development in Arabidopsis thaliana by APETALA1 and interacting genes. Development 119 1993 721 743 Google Scholar Crossref Search ADS WorldCat 8 Bradford MM A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72 1976 248 254 Google Scholar Crossref Search ADS PubMed WorldCat 9 Bradley D Carpenter R Sommer H Hartley N Coen E Complementary floral homeotic phenotypes result from opposite orientations of a transposon at the plena locus of Antirrhinum. Cell 72 1993 85 95 Google Scholar Crossref Search ADS PubMed WorldCat 10 Buban T Faust M Flower bud induction in apple trees: internal control and differentiation. Hortic Rev 4 1982 174 203 Google Scholar OpenURL Placeholder Text WorldCat 11 Burke B Griffith G Reggio DL Warren G A monoclonal antibody against a 135-K Golgi membrane protein. EMBO J 12 1982 1621 1628 Google Scholar OpenURL Placeholder Text WorldCat 12 Canas LA Busscher M Angenent GC Beltran J-P van Tunen AJ Nuclear localization of the petunia MADS box protein FBP1. Plant J 6 1994 597 604 Google Scholar Crossref Search ADS WorldCat 13 Carr SM Irish V Floral homeotic gene expression defines developmental arrest stages in Brassica oleracea L. vars. botrytis and italica. Planta 201 1997 179 188 Google Scholar Crossref Search ADS PubMed WorldCat 14 Chung Y-Y Kim S-R Finkel D Yanofsky MF An G Early flowering and reduced apical dominance result from ectopic expression of a rice MADS box gene. Plant Mol Biol 26 1994 657 665 Google Scholar Crossref Search ADS PubMed WorldCat 15 Church GM Gilbert W Genomic sequencing. Proc Natl Acad Sci USA 81 1984 1991 1995 Google Scholar Crossref Search ADS PubMed WorldCat 16 Davies B Egea-Cortines M de Andrade Silva E Saedler H Sommer H Multiple interactions amongst floral homeotic MADS box proteins. EMBO J 15 1996 4330 4343 Google Scholar Crossref Search ADS PubMed WorldCat 17 Dixon CD Cutt JR Klessig DF In situ hybridization for the detection of RNA in plant tissues. Maliga P Klessig DF Cashmore AR Gruissem W Varner JE Methods in Plant Molecular Biology. 1995 111 137 Cold Spring Harbor Laboratory Press Cold Spring Harbor, NY 18 Estruch JJ Granell A Hansen G Prinsen E Redig P Van Onckelen H Schwarz-Sommer Z Sommer H Spena A Floral development and expression of floral homeotic genes are influenced by cytokinins. Plant J 4 1993 379 384 Google Scholar Crossref Search ADS PubMed WorldCat 19 Evans PT Malmberg R Do polyamines have role in plant development? Annu Rev Plant Physiol 40 1989 235 269 Google Scholar Crossref Search ADS WorldCat 20 Flanagan CA Ma H Spatially and temporally regulated expression of the MADS-box gene AGL2 in wild-type and mutant Arabidopsis flowers. Plant Mol Biol 26 1994 581 595 Google Scholar Crossref Search ADS PubMed WorldCat 21 Gao J Kim S-R Chung Y-Y Lee JM An G Developmental and environmental regulation of two ribosomal protein genes in tobacco. Plant Mol Biol 25 1994 761 770 Google Scholar Crossref Search ADS PubMed WorldCat 22 Goren R Palavan N Flores H Galston AW Changes in polyamine titer in etiolated pea seedlings following red light treatment. Plant Cell Physiol 23 1982 19 26 Google Scholar Crossref Search ADS WorldCat 23 Goto K Meyerowitz EM Function and regulation of the Arabidopsis floral homeotic gene PISTILLATA. Genes Dev 8 1994 1548 1560 Google Scholar Crossref Search ADS PubMed WorldCat 24 Gu Q Ferrandiz C Yanofsky MF Martienssen R The FRUITFULL MADS-box gene mediates cell differentiation during Arabidopsis fruit development. Development 125 1998 1509 1517 Google Scholar Crossref Search ADS PubMed WorldCat 25 Hardenack S Ye D Saedler H Grant S Comparison of MADS-box gene expression in developing male and female flowers of the dioecious plant white campion. Plant Cell 6 1994 1775 1787 Google Scholar PubMed OpenURL Placeholder Text WorldCat 26 Hoekema A Hirsch PR Hooykaas PJJ Schilperoort RA A binary vector strategy based on separation of vir- and T-region of the Agrobacterium tumefaciences Ti-plasmid. Nature 303 1983 179 181 Google Scholar Crossref Search ADS WorldCat 27 Huijser P Klein J Lonnig W Meijer H Saedler H Sommer H Bracteomania, an inflorescence anomaly, is caused by the loss of function of the MADS-box gene squamosa in Antirrhinum majus. EMBO J 11 1992 1239 1249 Google Scholar Crossref Search ADS PubMed WorldCat 28 Irish VF Sussex IM Function of the apetala-1 gene during Arabidopsis floral development. Plant Cell 2 1990 741 753 Google Scholar PubMed OpenURL Placeholder Text WorldCat 29 Jack T Brockman LL Meyerowitz EM The homeotic gene APETALA3 of Arabidopsis thaliana encodes a MADS box and is expressed in petals and stamens. Cell 68 1992 683 697 Google Scholar Crossref Search ADS PubMed WorldCat 30 Jack T Fox GL Meyerowitz EM Arabidopsis homeotic gene APETALA3 ectopic expression: transcriptional and posttranscriptional regulation determine floral organ identity. Cell 76 1994 703 716 Google Scholar Crossref Search ADS PubMed WorldCat 31 Kang S-G Hannapel DJ Nucleotide sequences of novel potato (Solanum tubersum L.) MADS-box cDNAs and their expression in vegetative organs. Gene 166 1995 329 330 Google Scholar Crossref Search ADS PubMed WorldCat 32 Kater MM Colombo L Franken J Busscher M Masiero S Van Lookeren Campagne MM Angenent GC Multiple AGAMOUS homologs from cucumber and petunia differ in their ability to induce reproductive organ fate. Plant Cell 10 1998 171 182 Google Scholar Crossref Search ADS PubMed WorldCat 33 Kempin SA Savidge B Yanofsky MF Molecular basis of the caulifower phenotype in Arabidopsis. Science 267 1995 522 525 Google Scholar Crossref Search ADS PubMed WorldCat 34 Krizek BA Meyerowitz EM The Arabidopsis homeotic genes APETALA3 and PISTILLATA are sufficient to provide the B class organ identity function. Development 122 1996 11 22 Google Scholar Crossref Search ADS PubMed WorldCat 35 Kyozuka J Harcourt R Peacock WJ Dennis ES Eucalyptus has functional equivalents of the Arabidopsis AP1 gene. Plant Mol Biol 35 1997 573 584 Google Scholar Crossref Search ADS PubMed WorldCat 36 Levy YY Dean C The transition to flowering. Plant Cell 10 1998 1973 1989 Google Scholar Crossref Search ADS PubMed WorldCat 37 Lozano R Angosto T Gomez P Payan C Capel J Huijser P Salinas J Martinez-Zapater JM Tomato flower abnormalities induced by low temperatures are associated with changes of expression of MADS-box genes. Plant Physiol 117 1998 91 100 Google Scholar Crossref Search ADS PubMed WorldCat 38 Ma H Yanofsky MF Meyerowitz EM AGL1-AGL6, an Arabidopsis gene family with similarity to floral homeotic and transcription factor genes. Genes Dev 5 1991 484 495 Google Scholar Crossref Search ADS PubMed WorldCat 39 Mandel MA Gustafson-Brown C Savidge B Yanofsky MF Molecular characterization of the Arabidopsis floral homeotic gene APETALA1. Nature 360 1992 273 277 Google Scholar Crossref Search ADS PubMed WorldCat 40 Mandel MA Yanofsky MF A gene triggering flower formation in Arabidopsis. Nature 377 1995a 522 524 Google Scholar Crossref Search ADS WorldCat 41 Mandel MA Yanofsky MF The Arabidopsis AGL8 MADS box gene is expressed in inflorescence meristem and is negatively regulated by APETALA1. Plant Cell 7 1995b 1763 1771 Google Scholar OpenURL Placeholder Text WorldCat 42 Mandel MA Yanofsky MF The Arabidopsis AGL9 MADS box gene is expressed in young flower primordia. Sex Plant Reprod 11 1998 22 28 Google Scholar Crossref Search ADS WorldCat 43 Menzel G Apel K Melzer S Identification of two MADS box genes that are expressed in the apical meristem of the long-day plant Sinapis alba in transition to flowering. Plant J 9 1996 399 408 Google Scholar Crossref Search ADS PubMed WorldCat 44 Mouradov A Glassick TV Hamdorf BA Murphy LC Marla SS Yang Y Teasdale RD Family of MADS-box genes expressed early in male and female reproductive structures of Monterey pine. Plant Physiol 117 1998 55 61 Google Scholar Crossref Search ADS PubMed WorldCat 45 Okamuro JK den Boer BGW Lotys-Prass C Szeto W Jofuku KD Flowers into shoots: photo and hormonal control of a meristem identity switch in Arabidopsis. Proc Natl Acad Sci USA 93 1996 13831 13836 Google Scholar Crossref Search ADS PubMed WorldCat 46 Pnueli L Abu-Abeid M Zamir D Nacken W Schwarz-Sommer Z Lifschitz E The MADS box gene family in tomato: temporal expression during floral development, conserved secondary structures and homology with homeotic genes from Antirrhinum and Arabidopsis. Plant J 1 1991 255 266 Google Scholar Crossref Search ADS PubMed WorldCat 47 Pnueli L Hareven D Broday L Hurwitz C Lifschitz E The TM5 MADS box gene mediates organ differentiation in three inner whorls of tomato flowers. Plant Cell 6 1994 175 186 Google Scholar Crossref Search ADS PubMed WorldCat 48 Pratt C Apple flower and fruit: morphology and anatomy. Hortic Rev 10 1988 273 307 Google Scholar OpenURL Placeholder Text WorldCat 49 Puruggnanan MD Rounsley SD Schmidt RJ Yanofsky MF Molecular evolution of flower development: diversification of the plant MADS-box regulatory gene family. Genetics 140 1995 345 356 Google Scholar Crossref Search ADS PubMed WorldCat 50 Sambrook J Fritsh EF Maniatis TJ Molecular Cloning: A Laboratory Manual, Ed 2. 1989 Cold Spring Harbor Laboratory Press Cold Spring Harbor, NY 51 Sanger F Nicklen S Coulson AR DNA sequencing with chain terminating inhibitors. Proc Natl Acad Sci USA 74 1977 5463 5467 Google Scholar Crossref Search ADS PubMed WorldCat 52 Savidge B Rounsley SD Yanofsky MF Temporal relationship between the transcription of two Arabidopsis MADS box genes and the floral organ identity genes. Plant Cell 7 1995 721 733 Google Scholar PubMed OpenURL Placeholder Text WorldCat 53 Schwarz-Sommer Zs Huijser P Nacken W Saedler H Sommer H Genetic control of flower development by homeotic genes in Antirrhinum majus. Science 250 1990 931 936 Google Scholar Crossref Search ADS PubMed WorldCat 54 Smith TA Polyamines. Annu Rev Plant Physiol 36 1985 117 143 Google Scholar Crossref Search ADS WorldCat 55 Sommer H Beltran J-P Huijser P Pape H Lönnig W-E Saedler H Schwarz-Sommer Z Deficiens, a homeotic gene involved in the control of flower morphogenesis in Antirrhinum majus: the protein shows homology to transcription factors. EMBO J 9 1990 605 613 Google Scholar Crossref Search ADS PubMed WorldCat 56 Southerton SG Marshall H Mouradov A Teasdale RD Eucalypt MADS-box genes expressed in developing flowers. Plant Physiol 118 1998 365 372 Google Scholar Crossref Search ADS PubMed WorldCat 57 Sung S-K An G Molecular cloning and characterization of a MADS-box cDNA clone of the Fuji apple. Plant Cell Physiol 38 1997 484 489 Google Scholar Crossref Search ADS PubMed WorldCat 58 Tandre K Albert VA Sundas A Engstrom P Conifer homologues to genes that control floral development in angiosperms. Plant Mol Biol 27 1995 69 78 Google Scholar Crossref Search ADS PubMed WorldCat 59 Tandre K Svenson M Svensson ME Engstrom P Conservation of gene structure and activity in the regulation of reproductive organ development of conifers and angiosperms. Plant J 15 1998 615 624 Google Scholar Crossref Search ADS PubMed WorldCat 60 Theissen G Kim JT Saedler H Classification and phylogeny of the MADS-box gene families in the morphological evolution of eukaryotes. J Mol Evol 43 1996 484 516 Google Scholar Crossref Search ADS PubMed WorldCat 61 Tröbner W Ramirez L Motte P Hue I Huijser P Lönnig W-E Seadler H Sommer H Schwarz-Sommer Z GLOBOSA: a homeotic gene which interacts with DEFICIENS in the control of Antirrhinum floral organogenesis. EMBO J 11 1992 4963 4704 Google Scholar Crossref Search ADS WorldCat 62 Venglat SP Sawhney VK Benzylaminopurine induces phenocopies of floral meristem and organ identity mutants in wild-type Arabidopsis plants. Planta 198 1996 480 487 Google Scholar Crossref Search ADS PubMed WorldCat 63 Wang SY Faust M Changes in polyamine content during dormancy in flower buds of “Anna” apple. J Am Soc Hortic Sci 119 1994 70 73 Google Scholar Crossref Search ADS WorldCat 64 Weigel D Alvarez J Smyth DR Yanofsky MF Meyerowitz EM LEAFY controls floral meristem identity in Arabidopsis. Cell 69 1992 843 859 Google Scholar Crossref Search ADS PubMed WorldCat 65 Yanofsky MF Floral meristem to floral organs: genes controlling early events in Arabidopsis flower development. Annu Rev Plant Physiol 46 1995 167 188 Google Scholar Crossref Search ADS WorldCat 66 Yanofsky MF Ma H Bowman JL Drews GN Feldmann KA Meyerowitz EM The protein encoded by the Arabidopsis homeotic gene agamous resembles transcription factors. Nature 346 1990 35 39 Google Scholar Crossref Search ADS PubMed WorldCat Author notes 1 This work was supported in part by a special grant research program of the Ministry of Agriculture and Forestry of Korea to G.A. S.-K.S. was the recipient of postdoctoral research fellowships from the Korea Science and Engineering Foundation and the Korea Research Foundation. * Corresponding author; e-mail [email protected]; fax 82–56–22–79–21–99. Copyright © 1999 American Society of Plant Physiologists This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)

Chao, Wun S.; Gu, Yong-Qiang; Pautot, Véronique; Bray, Elizabeth A.; Walling, Linda L.

doi: 10.1104/pp.120.4.979pmid: 10444081

Abstract LapARNAs, proteins, and activities increased in response to systemin, methyl jasmonate, abscisic acid (ABA), ethylene, water deficit, and salinity in tomato (Lycopersicon esculentum). Salicylic acid inhibited wound-induced increases of LapA RNAs. Experiments using the ABA-deficient flacca mutant indicated that ABA was essential for wound and systemin induction ofLapA, and ABA and systemin acted synergistically to induce LapA gene expression. In contrast,pin2 (proteinase inhibitor 2) was not dependent on exogenous ABA. Whereas both LapA and le4(L. esculentumdehydrin) were up-regulated by increases in ABA, salinity, and water deficit, only LapAwas regulated by octadecanoid pathway signals. Comparison ofLapA expression with that of thePR-1 (pathogenesis-related 1) andGluB (basic β-1,3-glucanase) genes indicated that these PR protein genes were modulated by a systemin-independent jasmonic acid-signaling pathway. These studies showed that at least four signaling pathways were utilized during tomato wound and defense responses. Analysis of the expression of aLapA1:GUS gene in transgenic plants indicated that theLapA1 promoter was active during floral and fruit development and was used during vegetative growth only in response to wounding, Pseudomonas syringae pv tomatoinfection, or wound signals. This comprehensive understanding of the regulation of LapA genes indicated that this regulatory program is distinct from the wound-induced pin2, ABA-responsive le4, and PR protein genes. Plants respond quickly to pathogen and herbivore attacks by activating wound- and defense-response genes (Bowles, 1990; Dixon et al., 1994; Yang et al., 1996). Tomato (Lycopersicon esculentum) wound/defense-response genes are often expressed both locally and systemically (Enkerli et al., 1993; Pautot et al., 1993;Bergey et al., 1996). The signals that mediate systemic responses must be transmitted rapidly throughout the plant and may involve cell-to-cell signaling. Putative systemic signals include ethylene (Ecker and Davis, 1987; O'Donnell et al., 1996), SA (Malamy et al., 1990), ABA (Peña-Cortés et al., 1989), JA (Farmer and Ryan, 1990), and systemin (Pearce et al., 1991), as well as electrical and hydraulic signals (Wildon et al., 1992; Malone et al., 1994; Herde et al., 1996; Stankovic and Davies, 1998). Multiple signal transduction pathways interact to activate or suppress wound- and defense-response genes in the Solanaceae. Wound-response genes, such as the pin (proteinase inhibitor) genes are activated by systemin and octadecanoid pathway products such as JA (Farmer and Ryan, 1990, 1992; Pearce et al., 1991;Peña-Cortés et al., 1995). Systemin acts locally and systemically to induce synthesis of JA, which induces the expression of wound-response genes (Narváez-Vásquez et al., 1995). ABA has also been reported as a local and systemic signal for the induction of pin genes in potato and tomato (Peña-Cortés et al., 1989, 1991, 1995). However, the role of ABA in the induction of wound-response genes in tomato has remained controversial (Schaller and Ryan, 1995; Birkenmeier and Ryan, 1998). Several tomato wound-response genes are negatively regulated by SA, which acts at multiple steps in the octadecanoid signaling pathway (Doherty et al., 1988; Li et al., 1992; Peña-Cortés et al., 1993; Doares et al., 1995; O'Donnell et al., 1996). This contrasts to SA induction of the PR (pathogenesis-related) protein genes (van Kan et al., 1995). The SA and octadecanoid signaling pathways are reciprocally regulated by a wound mitogen-activated protein kinase (Seo et al., 1995) and a small GTP-binding protein (Sano et al., 1994). This cross-talk may aid in separating early responses to wounding that accompany pathogen or pest attack from long-term responses, such asPR gene expression and the development of SAR. The interactions of ethylene with the wound- and SA-signaling pathways are not completely understood. Ethylene is important for the development of necrotic symptoms that accompany pathogen invasion, but is not essential for the development of SAR (Bent et al., 1992;Lawton et al., 1994; Lund et al., 1998). Several PRtranscripts accumulate in response to ethylene or ethephon treatments (Ecker and Davis, 1987; Raz and Fluhr, 1993; van Kan et al., 1995), while ethylene treatments do not induce some wound-response genes (Ryan, 1974; Kernan and Thornberg, 1989). Recent data have suggested that ethylene and JA interact to induce wound-response genes (Xu et al., 1994; O'Donnell et al., 1996). In addition to responding to wound and defense signals, the expression of each wound- and defense-response gene is modulated during development. While some defense-response genes are silent throughout all of vegetative development and are solely induced in response to stress, other defense- and wound-response genes are expressed in specific vegetative or reproductive organs (Peña-Cortés et al., 1991; Titarenko et al., 1997). Many wound- and defense-response genes are expressed only in floral buds (Peña-Cortés et al., 1991) or in a subset of the floral organs (Lotan et al., 1989;Cote et al., 1991; Uknes et al., 1993; Constabel and Brisson, 1995), while other genes are expressed in all mature floral organs (Peña-Cortés et al., 1991). Finally, the developmental programming may be distinct in different species. For example, thepin2 genes of potato are only expressed in tubers and developing floral buds, while the pin2 genes of tomato are expressed in all mature floral organs (Peña-Cortés et al., 1991). In tomato, LapA (Leu aminopeptidase) transcripts, proteins, and activities increase locally and systemically in response to wounding (Pautot et al., 1993; Gu et al., 1996b). Two tomato genes,LapA1 and LapA2, encode the 55-kD subunits of this exopeptidase (Gu et al., 1996a). The Lap genes of potato and Arabidopsis are regulated differently than the tomatoLapA genes, since the Arabidopsis Lap gene is constitutively expressed (Bartling and Nosek, 1994) and potatoLap RNAs do not accumulate systemically after wounding (Hildmann et al., 1992), nor are they detected in response to pathogens (Herbers et al., 1994). Given the fact that LapA transcripts and proteins are abundant after wounding, pathogen invasion, and insect infestation, LAP-A may play an important role in the tomato defense response (Pautot et al., 1993). For this reason, it was important to develop a comprehensive understanding of LapA expression (at the RNA, protein, and activity level) in response to wound/defense-response signals (including ethylene, SA, JA, ABA, and systemin) and during abiotic stress. Comparisons of LapA gene expression patterns relative to patterns of expression for the wound-response genepin2, the ABA-response gene le4, and threePR protein genes (PR-1,PR-4, and GluB [basic β-1,3-glucanase]) in ABA-deficient or ABA-producing lines demonstrated that each gene responded to wound and defense signals in a distinct manner. These data indicated that at least four signaling pathways are required to modulate wound/defense gene expression in tomato plants. Finally, to understand the organ specificity ofLapA responses to wound signals, transgenic tomato and tobacco (Nicotiana tabacum) plants expressing a chimericLapA1:GUS gene were analyzed. MATERIALS AND METHODS Plant Growth, Tissue Harvest, and Storage Tomato (Lycopersicon esculentum cv Peto 238R, cv Ailsa Craig flacca, and cv Ailsa Craig) plants were grown in soil (University of California mix III) in growth chambers with 16-h (30°C)/8-h (20°C) light/dark cycles. Plants were watered daily and supplemented with 14:14:14 fertilizer (Osmocote, Scotts-Sierra, Maysville, OH). Immediately after treatments, leaves were excised, placed directly into liquid nitrogen, and stored at −80°C until use. Treatments with Wound- and Defense-Response Molecules Shoots from 3- to 4-week-old Peto 238R tomato plants were excised 5 cm above the soil for the ABA, MeJA, and ethylene treatments. Shoots from 3-week-old Peto 238R plants were excised below the third leaf from the plant shoot apex for the systemin and SA treatments. The 24-h 10 μm MeJA treatment and control have been described previously (Gu et al., 1996a). For ABA treatments, shoots were placed in flasks filled with 100 μm ABA (pH 6.0) or water (control) for 24 h. For ethylene treatments, excised shoots were placed in flasks filled with water and incubated in airtight glass desiccators containing a ripe banana and apple or without fruit (control) for 24 h. Ethylene levels typically rose to 29 ppm (D.P. Puthoff and L.L. Walling, unpublished results). The systemin treatment was a modification of that of Pearce et al. (1993). Excised shoots were placed in a microfuge tubes with 90 μL of 15 mm sodium phosphate buffer (pH 6.5) or 90 μL of 15 mm sodium phosphate buffer with 1 pmol of systemin. After 10 min, shoots were transferred to a flask filled with water for 12 h. Systemin was kindly provided by Dr. C.A. Ryan (Washington State University, Pullman). For SA treatments, shoots were placed in flasks filled with distilled water (control) or 0.1 to 0.5 mm SA (Sigma) for 24 h. The maximum concentration of SA tolerated by tomato shoots without inducing physical damage was 0.5 mm. Five- to six-week-old flacca plants were excised below the third leaf from the shoot apex. Four-week-old Ailsa Craig plants were at a similar stage in development. Shoots were treated with systemin (above) or wounded (Pautot et al., 1991), and were subsequently placed in water with or without 100 μm ABA and incubated in closed desiccators. Systemin-treated and wounded leaves were collected 12 and 24 h later, respectively. Water Deficit and Salinity Treatments After 3 weeks of growth (described above), water was withheld from Peto 238R plants and leaves were harvested at 4.5 d (at the initial signs of wilting), 5 d, and 5.5 d later. Control plants were watered once per day and leaves were harvested at the same time as the plants experiencing 5.5 d of water deficit. For salinity treatments, Peto 238R plants were watered with 300 mL of 300 mm NaCl, 400 mm NaCl, or water (control) for 3 d, and leaves were harvested. Construction of a LapA1:GUS Fusion Gene A 970-bp HindIII/DdeI fragment from the λLapA1 genomic subclone pLapA1-EH (W.S. Chao, V. Pautot, F.M. Holzer, and L.L. Walling, unpublished data), was end-filled using Klenow enzyme and cloned into the filled-in BamHI site of pBI101 (CLONTECH). Site-directed mutagenesis was used to remove residual vector sequences and restore the integrity of the LapA15′-UTR (Chao, 1996). pLapA1:GUS was transformed intoAgrobacterium tumefaciens (LBA4404 or EHA105), and transformants were confirmed by minilysates (Gelvin and Schilperoort, 1988; Birnboim and Doly, 1979). Tobacco and Tomato Transformation The tomato lines UC82b (Sunseeds Genetics, Hollister, CA) and VF36 (provided by Dr. S. McCormick, U.S. Department of Agriculture/Agricultural Research Service, Albany, CA) plants and tobacco (Nicotiana tabacum cv Xanthi) plants were used in the transformation experiments. LapA1:GUS transgenic plants were regenerated from tomato cotyledons and tobacco leaf discs using a modification of protocols described by Fillatti et al. (1987)and McCormick (1991). Details were described in Chao (1996). Fifteen independent tomato lines and 12 independent tobacco lines were characterized. DNA blots withHindIII/EcoRI-digested genomic DNAs (10 μg/lane) from T0 plants and reconstruction lanes with pLapA1:GUS were used to determine transgene copy number (Walling et al., 1988). The expression of the LapA1:GUS gene in T1 and T0 plants was confirmed by wounding of cotyledon segments and GUS histochemical staining. Transgenic tomato and tobacco plants expressing 35S:GUS (pBI121, CLONTECH) were also made. GUS Activity Assays The expression of the chimeric LapA1:GUS gene was monitored using histochemical and fluorometric assays for GUS activity (Jefferson, 1987). To reduce endogenous GUS activity, 20% methanol (v/v) was added to the assay buffers (Kosugi et al., 1990). Fluorescence was measured using a mini fluorometer (TKO 100, Hoefer Scientific Instruments, San Francisco). Protein concentrations were determined using a bicinchoninic acid protein assay reagent (Pierce). To reduce interference caused by β-mercaptoethanol, samples were preincubated with an equal volume of 0.1 miodoacetamide in 0.1 mm Tris-HCl (pH 8) at 37°C for 20 min (Hill and Straka, 1988). Wounding, MeJA Treatment, and Infection of LapA1:GUS Plants T1 (LapA1:GUS) and UC82b tomato plants with six to eight leaves and T1(LapA1:GUS) and Xanthi tobacco plants with five to seven leaves were used. Leaves of four to six individual plants per transgenic line were wounded (Pautot et al., 1991) or served as controls. Leaves were harvested into liquid nitrogen 24 h later. Intact 7- to 10-d-old seedlings were treated with MeJA by submerging roots in 10 μm MeJA/0.002% ethanol or 0.002% ethanol (control). T1LapA1:GUS and UC82b plants with six to eight leaves were used for the infection studies. Three to four upper leaves were used. Half of the leaflets on a leaf served as the mock-infected control and were gently swabbed with water using cotton-tipped applicators. The remaining leaflets were inoculated with a Pseudomonas syringae pv tomato suspension (3 × 108 cfu/mL) using cotton swabs (Pautot et al., 1991). Leaflets were harvested 24 h later. RNA Blot Analyses RNA blots and washes were performed as described previously (Pautot et al., 1991). Blots were exposed to film (Hyper-MP, Amersham) at −80°C with an intensifying screen (DuPont) for 24 h unless indicated otherwise. Autoradiographic signals were quantitated using a phosphor imager (Molecular Dynamics). Probes were labeled using [α-32P]dCTP by nick translation. Transcript sizes were determined by running an RNA ladder (GIBCO-BRL) in parallel lanes. The pLe4 cDNA was described previously (Cohen et al., 1991; Kahn et al., 1993). The GluB, PR-1, andPR-4 cDNA clones from tomato have been described previously (van Kan et al., 1992, 1995), and were kindly provided by Dr. P.J.G.M. de Wit (Wageningen Agricultural University, Wageningen, Netherlands). The tomato pin2 clone pT2-47 (Graham et al., 1985) and the LapA cDNA clone pDR57 (Pautot et al., 1993) have also been described previously. Total Protein Extraction, Fractionation, and Immunoblot Analyses Total leaf proteins were extracted and fractionated by two-dimensional PAGE as described by Wang et al. (1992). Electro-transfer and immunoblot procedures were described in Gu et al. (1996b). A 1:500 dilution of the LAP-A polyclonal antiserum and the preimmune serum were used (Gu et al., 1996b). Aminopeptidase Activity Assay Native proteins were extracted from leaves of treated and control plants (Gu et al., 1996b). Protein concentrations were determined by a modified Bradford method (Ramagli and Rodriguez, 1985). The assays were performed in triplicate in 96-well microplates with 2 μg of protein and 250 μL of assay solution (1 mml-Leu-p-nitroanilide [Sigma], 50 mm Tris-HCl, pH 8.0, and 0.5 mm MnCl2). After 30 min, the amount of p-nitroaniline generated was measured spectrophotometrically at A405 using a microplate reader (E-Max, Molecular Devices, Menlo Park, CA). In Situ Hybridizations Floral buds (10-mm) were harvested, fixed, and imbedded in methacrylate as described by Kronenberger et al. (1993). Five-millimeter transverse sections of tomato buds were made. Sections were hybridized to a digoxigenin-labeled antisense or senseLapA1 RNAs. Digoxigenin-labeled RNAs were synthesized using T3 or T7 RNA polymerase (GIBCO-BRL) and pBS-LapA1 according to the manufacturer's instructions (Boehringer Mannheim). pBS-LapA1 has a 1.6-kb EcoRI/XbaI fragment from pDR57 inserted into the EcoRI/XbaI sites of pBS-KS+ (Pautot et al., 1993). RESULTS LapA Was Induced by Wound Signals: Systemin, ABA, MeJA, and Ethylene To understand the impact of wound signals on LapA RNA levels, tomato plants were treated with MeJA, systemin, ABA, or ethylene. RNA blots were hybridized with probes for LapA1and genes that respond to one or more of these signals (le4,PR-1, PR-4, and GluB).Le4 encodes a dehydrin-like protein and is induced by exogenous ABA and water deficit (Cohen et al., 1991; Kahn et al., 1993). PR-1, PR-4, andGluB RNAs and proteins accumulate in response to SA or ethephon (an ethylene-releasing compound) (Christ and Mösinger, 1989; van Kan et al., 1995; Tonero et al., 1997).PR-1 and PR-4 encode extracellular proteins. PR-1 has antifungal activity but its mechanism of action is not known (Niderman et al., 1995). PR-4 is similar to Win and hevein proteins; the role of PR-4 in defense has yet to be elucidated (Linthorst et al., 1991). GluB encodes an intracellular, basic β-1,3-glucanase whose activity can hydrolyze pathogen cell walls (van Kan et al., 1992, 1995). LapA was strongly induced by MeJA and systemin (Fig.1A). A 57-fold increase inLapA transcripts occurred in leaves after 24 h of exposure to 10 μm MeJA. LapA RNAs were 2.5-fold more abundant in systemin-treated plants than in MeJA-treated plants. Treatment of shoots with 100 μm ABA increased LapA RNA levels 2-fold, which is similar to the increase measured for the well-characterized ABA- and water-deficit-response gene le4(Cohen et al., 1991). This may be a minimal estimate of LapAinduction in response to ABA, since larger increases in le4transcripts were observed in an excised leaf assay (Cohen et al., 1991). In contrast to LapA, le4 RNA levels did not increase in response to systemin, MeJA, or ethylene treatments. Fig. 1. Open in new tabDownload slide RNA blot analyses of plants treated with wound- and defense-response signal molecules. A, Tomato plants treated with 10 μm MeJA, 29 ppm ethylene (Eth), 100 μm ABA, 1.0 pmol of systemin (Sys), or 0.1, 0.25, or 0.5 mm SA. For each treatment, the corresponding control is shown as a C in parentheses. B, Tomato plants were wounded and incubated in the absence (lane W) or presence of 0.1 mm or 0.3 mm SA (lanes W + SA). Total RNAs were extracted from treated and healthy control (lane C) leaves. The RNA blots were hybridized with 32P-labeled LapA,le4, GluB,PR-4, orPR-1 probes. The RNA sizes are indicated in kb. Data in A and B are from representative experiments. Photographs presented are optimized for visualization of weak autoradiographic signals. Hybridization signals were quantitated using a phosphor imager. Fig. 1. Open in new tabDownload slide RNA blot analyses of plants treated with wound- and defense-response signal molecules. A, Tomato plants treated with 10 μm MeJA, 29 ppm ethylene (Eth), 100 μm ABA, 1.0 pmol of systemin (Sys), or 0.1, 0.25, or 0.5 mm SA. For each treatment, the corresponding control is shown as a C in parentheses. B, Tomato plants were wounded and incubated in the absence (lane W) or presence of 0.1 mm or 0.3 mm SA (lanes W + SA). Total RNAs were extracted from treated and healthy control (lane C) leaves. The RNA blots were hybridized with 32P-labeled LapA,le4, GluB,PR-4, orPR-1 probes. The RNA sizes are indicated in kb. Data in A and B are from representative experiments. Photographs presented are optimized for visualization of weak autoradiographic signals. Hybridization signals were quantitated using a phosphor imager. Relative to the control, there was a small increase in LapARNAs (2-fold) in plants exposed to ethylene (Fig. 1A). After ethylene treatment, PR-1, PR-4, andGluB transcripts increased 14-, 16-, and 3-fold, respectively. The response of the tomato PR-1,PR-4, and GluB genes to pathogens, ethylene, and SA is well established (Christ and Mösinger, 1989;van Kan et al., 1995; Tonero et al., 1997), but less is known about their responses to wound signals. PR-1,PR-4, and GluB RNAs were unchanged after ABA or systemin treatments (Fig. 1A). WhilePR-4 transcripts did not accumulate in response to MeJA, MeJA caused both GluB andPR-1 RNAs to accumulate relative to the control plants. GluB and PR-1 RNA levels were elevated in the MeJA and ethylene controls relative to controls from other treatments (i.e. ABA, systemin, or SA). This may be due to the fact that the MeJA and ethylene treatments were done in a closed environment and a volatile signal may have accumulated to induceGluB and PR-1. It is clear thatLapA, le4, and PR-4transcript levels were not modulated by this additional signal(s). During the course of these studies, we noted that the position of the incision and the age of the seedling used in the excised shoot assay was important (see Methods). When 3-week-old plants were used in this assay (systemin and SA treatments), LapARNAs were detected in controls. This is in contrast to the extremely low to undetectable levels of LapA RNAs in leaves from excised 4-week-old shoots (MeJA, ethylene, and ABA treatments) or from intact plants (Pautot et al., 1993). It is clear that the developmental state must influence the tomato response to shoot excision. Several other studies have indicated that plant age may influence wound signaling (Wolfson and Murdock, 1990; Alarcon and Malone, 1995). LapA RNA Levels Decreased in Response to Exogenous SA Using the excised shoot assay, PR-4transcripts were not detected after any of the SA treatments (Fig. 1A). In contrast, PR-1 and GluB transcripts increased after 0.1 and 0.25 mm SA treatments, respectively. The steady-state levels of PR-1 andGluB RNAs were distinct, suggesting differences in either transcriptional or posttranscriptional regulation. SA inhibited the accumulation of LapA transcripts, since control leaves had higher levels of LapA transcripts than SA-treated leaves (Fig. 1A). When wounding was followed by 0.1 or 0.3 mm SA treatments, LapA transcript levels were significantly reduced relative to wounded plants (Fig. 1B). ABA Was Required for Wound-Induced Activation of LapA To determine if endogenous ABA was required for wound and systemin induction of LapA, the expression of LapA was examined in the ABA-deficient flacca mutant and the ABA-proficient Ailsa Craig lines. Shoots were treated with systemin or were wounded, and subsequently incubated in water or 100 μm ABA (Fig. 2A). High levels of LapA and pin2 transcripts and low levels of le4 RNAs were detected after wounding in cv Ailsa Craig. In flacca plants, the le4 andLapA transcripts were undetectable in healthy or wounded leaves. Low levels of pin2 transcripts were consistently detected in healthy flacca leaves and, after wounding, the levels of pin2 RNAs increased 2-fold (Fig. 2A). Fig. 2. Open in new tabDownload slide RNA blot analysis of LapA gene expression in ABA-deficient (flacca) and control (Ailsa Craig) plants and in response to salinity and water deficit. A, Plants were mechanically wounded (W) or treated with 1 pmol of systemin (Sys), and excised shoots were subsequently incubated in water with or without 100 μm ABA. Ailsa Craig leaves were mechanically wounded (Ailsa, W). Shoots of healthy flacca plants were incubated in water (H) or ABA (H + ABA). Shoots of woundedflacca plants were incubated in water (W) or ABA (W + ABA). Flacca shoots were treated with systemin (Sys), systemin + ABA (Sys +ABA), or incubated with 15 mmphosphate buffer (Sys [C]). The blots were exposed to film for 48 h. B, Tomato plants were treated with 100 μm ABA, 300 mm NaCl, or 400 mm NaCl, or were not watered (water deficit) for 4.5, 5.0, or 5.5 d. Total RNA was extracted from treated and control (C) leaves. The RNA blots were hybridized with 32P-labeled LapA,le4, GluB,PR-4, orPR-1 probes. The autoradiographic signals of each band were quantitated using a phosphor imager. Fig. 2. Open in new tabDownload slide RNA blot analysis of LapA gene expression in ABA-deficient (flacca) and control (Ailsa Craig) plants and in response to salinity and water deficit. A, Plants were mechanically wounded (W) or treated with 1 pmol of systemin (Sys), and excised shoots were subsequently incubated in water with or without 100 μm ABA. Ailsa Craig leaves were mechanically wounded (Ailsa, W). Shoots of healthy flacca plants were incubated in water (H) or ABA (H + ABA). Shoots of woundedflacca plants were incubated in water (W) or ABA (W + ABA). Flacca shoots were treated with systemin (Sys), systemin + ABA (Sys +ABA), or incubated with 15 mmphosphate buffer (Sys [C]). The blots were exposed to film for 48 h. B, Tomato plants were treated with 100 μm ABA, 300 mm NaCl, or 400 mm NaCl, or were not watered (water deficit) for 4.5, 5.0, or 5.5 d. Total RNA was extracted from treated and control (C) leaves. The RNA blots were hybridized with 32P-labeled LapA,le4, GluB,PR-4, orPR-1 probes. The autoradiographic signals of each band were quantitated using a phosphor imager. Treatment of flacca shoots with 100 μm ABA caused LapA RNAs to rise 5-fold (Fig. 2A). When flacca was wounded and ABA treated,LapA RNAs increased 7-fold. ABA supplementation of healthy or wounded flacca plants caused pin2 RNAs to increase only 2- to 3-fold. Le4 RNA levels also increased when healthy flacca leaves were treated with ABA; wounding did not further increase le4 transcript abundance. To examine if ABA has a role in systemin signal transduction,flacca shoots were treated with 15 mmphosphate buffer (control), systemin, or systemin plus 100 μm ABA (Fig. 2A). In buffer-treatedflacca shoots, LapA or le4 RNAs were undetectable and low levels of pin2 RNAs were observed. After systemin treatment, LapA RNA increased 2-fold inflacca leaves. In contrast, a 50-fold induction ofLapA transcripts was detected when flacca shoots were treated with both systemin and ABA. These levels were comparable to LapA levels in wounded cv Ailsa Craig leaves, and suggest that ABA and systemin act synergistically. These data indicated that ABA was critical for maximal accumulation of LapAtranscripts in response to systemin and that pin2 was regulated in a different manner. pin2 RNAs increased 9-fold in flacca leaves in response to systemin and, when applied simultaneously, systemin and ABA increased pin2 transcripts 18-fold. Le4 transcripts did not increase in response to systemin (Figs. 1A and 2A). The level of le4 RNAs in leaves of flacca treated with both systemin and ABA was actually lower than that observed with ABA alone. LapA Is Induced during Water Deficit and Salinity Stress ABA is not only an important signal in the wound response of tomato, but it is also an important component in abiotic stresses such as water deficit and salinity (Bray, 1993; Chandler and Robertson, 1994). Therefore, we measured changes in LapA RNA levels during water deficit. LapA RNA levels increased during water deficit and reached maximal levels in plants stressed for 5 d (Fig. 2B); Le4 served as positive control (Cohen et al., 1991; Kahn et al., 1993). Le4 transcripts were present at higher levels than LapA RNAs and accumulated throughout the entire stress period. By d 5, le4 RNA levels had increased 53-fold. To determine if LapA RNAs accumulated in response to salinity, tomato plants were watered with 300 or 400 mm NaCl for 3 d. LapA RNAs increased 4- to 6-fold in response to salinity treatments (Fig. 2B), whereas a more dramatic increase (22-fold) in le4 RNA levels was observed. None of the PR gene transcripts accumulated in response to water deficit or salinity stress, which is consistent with the observation that exogenous ABA treatments did not inducePR-1, PR-4, or GluB gene expression (Fig. 1A). LAP-A Proteins and Activities Were Elevated after Stress Treatments To determine if there was a coordinate induction ofLapA RNAs and proteins, total proteins were extracted from leaves that were subjected to water deficit or treated with MeJA, systemin, ABA, or NaCl. Immunoblots showed that four classes of LAP-related proteins and one class of non-LAP protein were resolved (Gu et al., 1996b). The 90-kD proteins were not related to LAP, since they were recognized by preimmune serum (Gu et al., 1996b). The 66- and 77-kD LAP-like polypeptides and 55-kD LAP proteins with neutral pIs (LAP-N) were detected in all control and treated tomato leaf samples (Fig. 3). Only the 55-kD LAP-A polypeptides (with acidic pIs) were induced after stress treatments (Fig. 3, B, D, F, H, and J). LAP-A proteins were most abundant in MeJA- and systemin-treated leaves (Fig. 3, B and J), which is consistent with RNA blot analyses (Figs. 1A and 2B). While the levels ofLapA RNA varied in the control plants (Figs. 1A and 2B), the LAP-A protein levels varied only slightly (Fig. 3, A, C, E, G, and I). Fig. 3. Open in new tabDownload slide Immunoblots of proteins that accumulated in response to wound signals and abiotic stress. Total proteins (80 μg) were fractionated by two-dimensional PAGE. The gels were electroblotted onto nitrocellulose and the blots were incubated with a 1:500 dilution of the LAP-A polyclonal antiserum. The pH range for IEF and molecular mass markers (in kD) are indicated. The 55-kD LAP-A proteins had a pI range of 5.6 to 5.9 (Gu et al., 1996b). A, Control plants for MeJA treatment. B, Plants treated with 10 μm MeJA for 12 h. C, Control plants for ABA treatment. D, Plants treated with 100 μm ABA for 12 h. E, Control plants for 5-d water deficit treatment. F, Plants 5 d after water was withheld. G, Control plants for 300 mm NaCl treatment. H, Plants 3 d after 300 mm NaCl treatment. I, Control plants for systemin treatment. J, Plants 12 h after treatment with 1 pmol of systemin. Fig. 3. Open in new tabDownload slide Immunoblots of proteins that accumulated in response to wound signals and abiotic stress. Total proteins (80 μg) were fractionated by two-dimensional PAGE. The gels were electroblotted onto nitrocellulose and the blots were incubated with a 1:500 dilution of the LAP-A polyclonal antiserum. The pH range for IEF and molecular mass markers (in kD) are indicated. The 55-kD LAP-A proteins had a pI range of 5.6 to 5.9 (Gu et al., 1996b). A, Control plants for MeJA treatment. B, Plants treated with 10 μm MeJA for 12 h. C, Control plants for ABA treatment. D, Plants treated with 100 μm ABA for 12 h. E, Control plants for 5-d water deficit treatment. F, Plants 5 d after water was withheld. G, Control plants for 300 mm NaCl treatment. H, Plants 3 d after 300 mm NaCl treatment. I, Control plants for systemin treatment. J, Plants 12 h after treatment with 1 pmol of systemin. To determine if LAP-A protein levels is correlated with LAP activities, aminopeptidase activity assays were performed. Relative aminopeptidase activities increased in leaves of MeJA-, systemin-, ABA-, water-deficit-, and NaCl-treated plants (Fig.4). The large increases in aminopeptidase activities noted in the MeJA- and systemin-treated samples paralleled the large increases in LapA mRNAs and proteins (Fig.1A). Since the activity assays were performed on total soluble leaf protein extracts, a direct correlation between the amount of LAP-A proteins and LAP activities could not be made. The observed changes in aminopeptidase activities may have been due to increases in LAP-A and/or changes in the activities or levels of additional tomato leaf aminopeptidases (Gu et al., 1996b; Walling and Gu, 1996). Fig. 4. Open in new tabDownload slide Relative aminopeptidase activities. Native proteins were extracted from control (C) leaves and from leaves treated with 10 μm MeJA, 100 μm ABA, 5 d of water deficit, 300 mm NaCl, or 1 pmol of systemin (Sys). Aminopeptidase activity was measured in triplicate spectrophotometrically at A405 by the release of p-nitroaniline. Aminopeptidase activity was calculated as the A405 per milligram of protein. Relative aminopeptidase activities and sds are shown; the highest aminopeptidase levels were detected in the systemin-treated leaves (6.2 A405/μg protein); this value was set at 100%. Each analysis was replicated two times. Fig. 4. Open in new tabDownload slide Relative aminopeptidase activities. Native proteins were extracted from control (C) leaves and from leaves treated with 10 μm MeJA, 100 μm ABA, 5 d of water deficit, 300 mm NaCl, or 1 pmol of systemin (Sys). Aminopeptidase activity was measured in triplicate spectrophotometrically at A405 by the release of p-nitroaniline. Aminopeptidase activity was calculated as the A405 per milligram of protein. Relative aminopeptidase activities and sds are shown; the highest aminopeptidase levels were detected in the systemin-treated leaves (6.2 A405/μg protein); this value was set at 100%. Each analysis was replicated two times. There was substantial variation in the aminopeptidase levels detected in the five controls for these studies. This may have been due to the fact that the treatment regimes varied. Four-week-old plants in soil were used for the water deficit and salinity studies and the aminopeptidase activities were similar in their controls. Three-week-old (systemin)- or 4-week-old (ABA and MeJA) excised shoots were used for the other treatments. The impact of seedling age on theLapA RNA levels detected in controls was noted (Fig. 1A). Finally, although the ages of the seedlings in the ABA and the MeJA treatments were the same, the ABA- and MeJA-treated plants were incubated in open and closed environments, respectively. The LapA1 Promoter Was Activated by Wound Signals andP. syringaepv tomato LapA genes are primarily controlled at the transcriptional level (W.S. Chao, V. Pautot, F.M. Holzer, and L.L. Walling, unpublished data). Therefore, a LapA1:GUSfusion was used to investigate LapA1 promoter activity in response to pathogens, to wound signals, and during development. The response of the LapA1 promoter to wounding was characterized using 15 independent LapA1:GUS transgenic tomato lines (Table I). No GUS activity was detected in wounded or nonwounded leaves from UC82b control plants. Basal GUS activity levels in nonwounded leaves from the transgenic tomato lines varied (12–1,191 nmol 4-methylumbelliferone min−1 mg−1 protein). After wounding, GUS activity increased in all LapA1:GUStransgenic tomato lines except the U55 line. Increases in GUS activity levels after wounding was variable and ranged from 1.3-fold (line V13) to 40-fold (line V14). Wound induction of theLapA1 promoter was also noted in the 12 independentLapA1:GUS transgenic tobacco lines characterized (Chao, 1996). In general, wound induction was less dramatic, ranging from 2- to 8-fold; however, one transgenic LapA1:GUS tobacco line exhibited a 72-fold induction (data not shown). Table I. Fluorometric analysis of GUS activity in transgenic tomato lines in response to wounding Transgenic Line . GUS Activity . Ratio (W/H) . Wounded (W) . Healthy (H) . nmol 4-MU min−1mg−1 protein UC82b 0 0 – V13 1,074 853 1.3 V14 3,373 84 40.2 V15 2,257 154 14.7 U17 454 43 10.6 U26 634 144 4.4 U33 1,072 88 12.2 U38 1,639 443 3.7 U48 222 22 10.1 U49 2,075 256 4.2 U55 722 1,191 0.6 U63 170 71 2.4 U69 109 12 9.1 U78 3,750 407 9.2 U83 694 404 1.7 U93 2,467 90 27.4 Transgenic Line . GUS Activity . Ratio (W/H) . Wounded (W) . Healthy (H) . nmol 4-MU min−1mg−1 protein UC82b 0 0 – V13 1,074 853 1.3 V14 3,373 84 40.2 V15 2,257 154 14.7 U17 454 43 10.6 U26 634 144 4.4 U33 1,072 88 12.2 U38 1,639 443 3.7 U48 222 22 10.1 U49 2,075 256 4.2 U55 722 1,191 0.6 U63 170 71 2.4 U69 109 12 9.1 U78 3,750 407 9.2 U83 694 404 1.7 U93 2,467 90 27.4 Fifteen independent LapA1:GUS transgenic lines were analyzed. Transgenic lines are designated to indicate their parentage: UC82b (U) or VFNT (V). All lines had one to two copies of theLapA1:GUS transgene, except U38, which had five copies. GUS activity was measured in leaf extracts. Four to six GUS-positive T1 plants per line were mechanically wounded or served as healthy controls. Leaves were harvested 24 h later and pooled for each treatment. GUS and protein levels were determined as described in “Materials and Methods.” Open in new tab Table I. Fluorometric analysis of GUS activity in transgenic tomato lines in response to wounding Transgenic Line . GUS Activity . Ratio (W/H) . Wounded (W) . Healthy (H) . nmol 4-MU min−1mg−1 protein UC82b 0 0 – V13 1,074 853 1.3 V14 3,373 84 40.2 V15 2,257 154 14.7 U17 454 43 10.6 U26 634 144 4.4 U33 1,072 88 12.2 U38 1,639 443 3.7 U48 222 22 10.1 U49 2,075 256 4.2 U55 722 1,191 0.6 U63 170 71 2.4 U69 109 12 9.1 U78 3,750 407 9.2 U83 694 404 1.7 U93 2,467 90 27.4 Transgenic Line . GUS Activity . Ratio (W/H) . Wounded (W) . Healthy (H) . nmol 4-MU min−1mg−1 protein UC82b 0 0 – V13 1,074 853 1.3 V14 3,373 84 40.2 V15 2,257 154 14.7 U17 454 43 10.6 U26 634 144 4.4 U33 1,072 88 12.2 U38 1,639 443 3.7 U48 222 22 10.1 U49 2,075 256 4.2 U55 722 1,191 0.6 U63 170 71 2.4 U69 109 12 9.1 U78 3,750 407 9.2 U83 694 404 1.7 U93 2,467 90 27.4 Fifteen independent LapA1:GUS transgenic lines were analyzed. Transgenic lines are designated to indicate their parentage: UC82b (U) or VFNT (V). All lines had one to two copies of theLapA1:GUS transgene, except U38, which had five copies. GUS activity was measured in leaf extracts. Four to six GUS-positive T1 plants per line were mechanically wounded or served as healthy controls. Leaves were harvested 24 h later and pooled for each treatment. GUS and protein levels were determined as described in “Materials and Methods.” Open in new tab GUS activity increased significantly in six transgenic tomato lines 24 h after P. syringae pv tomato inoculation (Table II). Nontransformed UC82b leaves had low to undetectable levels of GUS activity in infected and mock-infected leaves, respectively. The levels of GUS activity in mock-infected LapA1:GUS transgenic tomato plants were variable, but were similar to basal levels in untreated leaves (TableI). Line U49 showed the most dramatic increase in GUS activity (174-fold) in response to P. syringae pv tomatoinfection. These data were well correlated with the wounding results (Table I). Table II. Fluorometric analysis of GUS activity in transgenic tomato lines in response to P. syringae pv tomato infection Transgenic Line . GUS Activity . Ratio (I/M) . Infected (I) . Mock-infected (M) . nmol 4-MU min−1 mg−1 protein UC82b 16 0 – V14 5,328 806 6.6 V15 1,274 239 5.3 U38 6,703 192 34.9 U49 12,509 72 173.7 U78 10,509 129 82.9 U93 8,196 66 124.2 Transgenic Line . GUS Activity . Ratio (I/M) . Infected (I) . Mock-infected (M) . nmol 4-MU min−1 mg−1 protein UC82b 16 0 – V14 5,328 806 6.6 V15 1,274 239 5.3 U38 6,703 192 34.9 U49 12,509 72 173.7 U78 10,509 129 82.9 U93 8,196 66 124.2 GUS activity in leaf extracts from six LapA1:GUS transgenic lines was measured 24 h after P. syringae pvtomato inoculation or mock infection. GUS and protein levels were determined as described in Methods. Open in new tab Table II. Fluorometric analysis of GUS activity in transgenic tomato lines in response to P. syringae pv tomato infection Transgenic Line . GUS Activity . Ratio (I/M) . Infected (I) . Mock-infected (M) . nmol 4-MU min−1 mg−1 protein UC82b 16 0 – V14 5,328 806 6.6 V15 1,274 239 5.3 U38 6,703 192 34.9 U49 12,509 72 173.7 U78 10,509 129 82.9 U93 8,196 66 124.2 Transgenic Line . GUS Activity . Ratio (I/M) . Infected (I) . Mock-infected (M) . nmol 4-MU min−1 mg−1 protein UC82b 16 0 – V14 5,328 806 6.6 V15 1,274 239 5.3 U38 6,703 192 34.9 U49 12,509 72 173.7 U78 10,509 129 82.9 U93 8,196 66 124.2 GUS activity in leaf extracts from six LapA1:GUS transgenic lines was measured 24 h after P. syringae pvtomato inoculation or mock infection. GUS and protein levels were determined as described in Methods. Open in new tab Histochemical staining for GUS activity showed that, like the control line UC82b, LapA1:GUS seedlings did not display significant GUS staining of cotyledons, hypocotyls, roots, or primary leaves (Fig.5, B and C). Occasionally, GUS staining was detected at random sites on the LapA1:GUS seedlings, and this was correlated with sites of inadvertent mechanical wounding.35S:GUS seedlings served as a positive control, and uniform GUS staining in the cotyledons, hypocotyls, and roots was detected (Fig. 5C). When treated with MeJA, the LapA1:GUS seedlings showed strong GUS staining in the aerial portions of the transgenic tomato plants: primary leaves (Fig. 5A), cotyledons, and hypocotyls (Fig. 5C). Cotyledons showed the highest level of GUS staining and the most apical portion of the hypocotyl in most seedlings also exhibited strong GUS staining. GUS staining was rarely detected in roots of JA-treated or control plants. Fig. 5. Open in new tabDownload slide LapA1 promoter activity in response to MeJA and during seedling development. A and B, Primary leaf from 10-d-old LapA1:GUS (line U49) (A) and UC82b control seedlings (B) treated with 10 μm MeJA. Bar = 3.1 mm. C, Seven-day-old LapA1:GUS (U49) seedling treated with 10 μm MeJA (1), 7-d-old LapA1:GUS seedling (U49) treated with 0.02% EtOH (2), 7-d-old 35S:GUSseedling (3), and 7-d-old UC82b seedling (4). Bar = 8.6 mm. Fig. 5. Open in new tabDownload slide LapA1 promoter activity in response to MeJA and during seedling development. A and B, Primary leaf from 10-d-old LapA1:GUS (line U49) (A) and UC82b control seedlings (B) treated with 10 μm MeJA. Bar = 3.1 mm. C, Seven-day-old LapA1:GUS (U49) seedling treated with 10 μm MeJA (1), 7-d-old LapA1:GUS seedling (U49) treated with 0.02% EtOH (2), 7-d-old 35S:GUSseedling (3), and 7-d-old UC82b seedling (4). Bar = 8.6 mm. The LAPA1 Promoter Was Active in Flowers and Fruit Examination of the LapA1:GUS transgene expression in both tomato and tobacco indicated that the LapA1 promoter was active in reproductive organs and in developing tomato fruit (Fig.6). Strong GUS staining was consistently observed in the stamens, ovaries, stigma, and styles of 2-mm buds to fully opened flowers (1.6-cm) from LapA1:GUS transgenic tomatoes (Fig. 6, A and B). GUS staining was more uniform in petals and sepals of younger buds (2 mm) than in older buds (0.8 cm and larger) or petals of fully opened flowers (Fig. 6A). This is well correlated with the accumulation of LapA RNAs during tomato flower development (V. Pautot, F.M. Holzer, J. Chaufaux, and L.L. Walling, unpublished data; Milligan and Gasser, 1995). In situ hybridizations with a LapA antisense RNA probe showed thatLapA transcripts were the most abundant in the integument of the ovaries and in the placental region (Fig.7). LapA RNAs were also detected at lower levels throughout other floral organs (V. Pautot, F.M. Holzer, J. Chaufaux, and L.L. Walling, unpublished data), and LAP-A proteins were detected in all floral organs of open flowers (C.J. Tu, F.M. Holzer, and L.L. Walling, data not shown). Similar results were obtained when transgenic LapA1:GUS tobacco lines were examined (Fig. 6, C and D). All LapA1:GUS flowers (0.5–5 cm) had GUS activity in sepals, petals, stamens, pistils, and sepal trichomes. In tobacco, pistils, ovaries, and stigmas exhibited the strongest GUS staining throughout floral development (Fig. 6, C and D). Fig. 6. Open in new tabDownload slide LapA1 promoter activity during flower and fruit development. A, Flowers of LapA1:GUStomato plants (U49) were excised, cut in half, and infiltrated with GUS histochemical substrate. Flowers displayed are 0.3, 0.8, 1.2, 1.6, and 1.6 cm in length (left to right). A 1.5-cm UC82b flower is presented at the right end of the panel. Variation in GUS activity in tomato and tobacco styles was due to the fact that GUS activity was not readily detected unless the style was bisected. Bar = 4.3 mm. B, Eleven-fold enlargement of the 1.2-cm LapA1:GUS tomato flower. Bar = 0.92 mm. S, Sepal; P, petal; O, ovary; Sty, style; Stg, stigma; Stm, stamen. C, Flowers of LapA1:GUStobacco plants (line X2; Chao, 1996) were excised, cut in half, and infiltrated with GUS histochemical substrate. Flowers displayed are 0.5, 0.5, 1.0, 3.0, and 5.0 cm in length (left to right). A 5-cmN. tabacum cv Xanthi flower is present at the right end of the panel. Bar = 6.3 mm. D, Fourteen-fold enlargement of the 1.0-cm LapA1:GUS tobacco flower (line X2). Bar =0.71 mm. S, Sepal; P, petal; O, ovary; Sty, style; Stg, stigma; A, anther; F, filament. E, Top, Fruit from control UC82b; bottom, fruits ofLapA1:GUS (U49) ranging in size from 0.5 mm to 7 cm incubated with GUS histochemical substrate. Bar = 1.25 cm. P, Pericarp; L, locular tissue; S, seed; Pl, placental tissue; V, vascular bundle; C, collumella. Fig. 6. Open in new tabDownload slide LapA1 promoter activity during flower and fruit development. A, Flowers of LapA1:GUStomato plants (U49) were excised, cut in half, and infiltrated with GUS histochemical substrate. Flowers displayed are 0.3, 0.8, 1.2, 1.6, and 1.6 cm in length (left to right). A 1.5-cm UC82b flower is presented at the right end of the panel. Variation in GUS activity in tomato and tobacco styles was due to the fact that GUS activity was not readily detected unless the style was bisected. Bar = 4.3 mm. B, Eleven-fold enlargement of the 1.2-cm LapA1:GUS tomato flower. Bar = 0.92 mm. S, Sepal; P, petal; O, ovary; Sty, style; Stg, stigma; Stm, stamen. C, Flowers of LapA1:GUStobacco plants (line X2; Chao, 1996) were excised, cut in half, and infiltrated with GUS histochemical substrate. Flowers displayed are 0.5, 0.5, 1.0, 3.0, and 5.0 cm in length (left to right). A 5-cmN. tabacum cv Xanthi flower is present at the right end of the panel. Bar = 6.3 mm. D, Fourteen-fold enlargement of the 1.0-cm LapA1:GUS tobacco flower (line X2). Bar =0.71 mm. S, Sepal; P, petal; O, ovary; Sty, style; Stg, stigma; A, anther; F, filament. E, Top, Fruit from control UC82b; bottom, fruits ofLapA1:GUS (U49) ranging in size from 0.5 mm to 7 cm incubated with GUS histochemical substrate. Bar = 1.25 cm. P, Pericarp; L, locular tissue; S, seed; Pl, placental tissue; V, vascular bundle; C, collumella. Fig. 7. Open in new tabDownload slide In situ hybridization of an antisenseLapA RNA with a tomato floral bud. A methacrylate-imbedded transverse section of a 10-mm floral bud of tomato was hybridized with digoxigenin-labeled antisense (A) and sense (B) LapA RNAs. Bar = 88 μm. Fig. 7. Open in new tabDownload slide In situ hybridization of an antisenseLapA RNA with a tomato floral bud. A methacrylate-imbedded transverse section of a 10-mm floral bud of tomato was hybridized with digoxigenin-labeled antisense (A) and sense (B) LapA RNAs. Bar = 88 μm. The LapA1 promoter was active in all stages of tomato fruit development in the different LapA1:GUS transgenic lines (Fig. 6E). The pericarp generally had the highest levels of GUS staining, but staining was also seen in locular tissue, seeds, placental tissue, vascular bundles, and collumella. In most cases, GUS staining was most uniform in the earlier stages of fruit development (data not shown). While all fruit exhibited GUS staining, the degree of GUS staining was variable, and approximately 10% of the fruit had lower levels of GUS activity (data not shown). DISCUSSION Multiple Signal Transduction Pathways Regulate Wound- and Defense-Response Genes The LapA RNAs and proteins accumulate in response to wounding and P. syringae pv tomato infection in tomato (Pautot et al., 1993; Gu et al., 1996b). Therefore, it was important to determine if LapA genes were regulated by the octadecanoid- or SA-dependent defense-response pathways or ifLapA utilized one of the more recently identified JA-independent (Titarenko et al., 1997) or SA-independent signal transduction pathways (Penninckx et al., 1995; Pieterse et al., 1996). To this end, tomato plants were treated with wound/defense signals and assessed for levels of LapA and three PR gene transcripts. These studies indicated that at least four signaling pathways were important for the expression of wound- and defense-response genes in tomato (Fig.8). Similar to pin2,LapA was modulated by MeJA and systemin, signals associated with the octadecanoid-signaling pathway (Fig. 1; Schaller and Ryan, 1995). Consistent with the role of SA in blocking the octadecanoid signaling pathway (Peña-Cortés et al., 1993; Doares et al., 1995; O'Donnell et al., 1996), wound induction of LapA was suppressed by SA. LapA was not strongly induced by exogenous ethylene (Fig. 1). However, if it were similar to pin genes,LapA would require JA for maximal activation by ethylene (O'Donnell et al., 1996). Fig. 8. Open in new tabDownload slide Schematic diagram showing the four independent signal transduction pathways that activate wound and defense genes in tomato. The PR genes (PR-1,PR-4, and GluB) utilize two signaling pathways: a SA-dependent and an ethylene-dependent pathway. PR-1 and GluButilize a third pathway, a systemin-independent, JA-dependent pathway, which may be analogous to the SA-independent pathway used to induce defensins in Arabidopsis (Penninckx et al., 1995). The fourth signaling pathway is the octadecanoid pathway, which in tomato utilizes systemin, ABA, JA, and ethylene to activate the wound-response genes (LapA1, LapA2, and pin2) and SA to down-regulate the pathway. There is evidence for a fifth signaling pathway in Arabidopsis. A JA-independent mechanism for wound-response gene activation was recently described, but tomato genes activated by this pathway have yet to be identified (Titarenko et al., 1997). Fig. 8. Open in new tabDownload slide Schematic diagram showing the four independent signal transduction pathways that activate wound and defense genes in tomato. The PR genes (PR-1,PR-4, and GluB) utilize two signaling pathways: a SA-dependent and an ethylene-dependent pathway. PR-1 and GluButilize a third pathway, a systemin-independent, JA-dependent pathway, which may be analogous to the SA-independent pathway used to induce defensins in Arabidopsis (Penninckx et al., 1995). The fourth signaling pathway is the octadecanoid pathway, which in tomato utilizes systemin, ABA, JA, and ethylene to activate the wound-response genes (LapA1, LapA2, and pin2) and SA to down-regulate the pathway. There is evidence for a fifth signaling pathway in Arabidopsis. A JA-independent mechanism for wound-response gene activation was recently described, but tomato genes activated by this pathway have yet to be identified (Titarenko et al., 1997). PR-4 may be regulated by two signaling pathways that were not utilized by LapA: an ethylene-dependent and a SA-dependent pathway. PR-4 was strongly induced by exogenous ethylene (Fig. 1; van Kan et al., 1995) and therefore utilized an ethylene signaling mechanism distinct from that used byLapA or pin genes. The excised-shoot assay utilized in these studies did not detect increases inPR-4 transcripts in response to 0.5 mm SA. However, excised-leaf assays have demonstrated that SA is an important regulator ofPR-4 gene expression (van Kan et al., 1995). The SA signaling pathway has been elegantly elucidated in Arabidopsis (Dangl et al., 1996; Ryals et al., 1996; Yang et al., 1996), and is assumed to function in a similar manner in the Solanaceae. A more complex circuitry was used to modulate expression of the tomatoPR-1 and GluB genes.PR-1 and GluB RNAs accumulated in response to exogenous SA and ethylene, and similar observations were made by other investigators (Christ and Mösinger, 1989; van Kan et al., 1995; Tonero et al., 1997). The ethylene and SA signal transduction pathways utilized by PR-1,GluB, and PR-4 were probably the same. However, while SA increased the levels of bothPR-1 and GluB RNAs, the steady-state levels of these RNAs were distinct. These data indicate that there are substantial differences in transcriptional and/or posttranscriptional controls that modulate these genes in response to SA. Finally, PR-1 and GluB genes were also regulated by a signaling pathway that was activated by JA but not by systemin. At present it is not known why the JA generated after systemin treatments was insufficient for PR-1 andGluB transcript accumulation. It is possible that the systemin-induced JA was present in a subset of tomato leaf cells that were not competent for PR-1 and GluBgene expression; similar theories regarding JA compartmentalization have been proposed by Harms et al. (1995). Alternatively, systemin may have induced an inhibitor to interfere with JA induction ofPR-1 and GluB expression, which would be consistent with the reciprocal regulation of the oxylipin and SA signal transduction pathways (Seo et al., 1995). At the present time, it is not known how the SA-independent and JA-dependent signaling pathways identified in Arabidopsis or tobacco relate to the pathways being elucidated in tomato (Penninckx et al., 1995; Pieterse et al., 1996; Vidal et al., 1997). However, it is clear from studies in Arabidopsis that some genes (such as defensins) can be induced by exogenous JA and ethylene but not by SA (Penninckx et al., 1995). These data indicate the independence of these signaling mechanisms. The fact that Arabidopsis defensins are not wound induced suggests that they may utilize the JA and ethylene signaling pathways that are similar to those used by the tomato PR-4,PR-1, and GluB and are distinct from JA-dependent or -independent wound responses (Titarenko et al., 1997). ABA Is Essential for Wound Induction of LapA The role of ABA in the modulation of wound-response gene expression in tomato remains controversial. Peña-Cortés et al. (1989, 1991, 1995) have concluded that ABA is essential forpin2 gene expression and acts early in the octadecanoid pathway. On the other hand, other studies have concluded that ABA does not have a primary role in oxylipin signal transduction pathway (Schaller and Ryan, 1995; Birkenmeier and Ryan, 1998). These discrepancies suggested to us that the role of ABA in wound-response gene induction needed to be re-evaluated, since significant differences in plant genotypes and treatments were present in the previous studies. The studies reported here with ABA-producing lines indicated thatPR gene expression was not influenced by exogenous ABA. Our data support the idea that the systemin-independent, JA-responsive mechanism for PR-1 and GluB gene expression is distinct from the JA-signaling mechanisms utilized byLapA and pin2 genes. Examination of LapA, pin2, and le4transcript levels in the ABA-deficient line flacca indicated that there were significant differences in the role of ABA in the regulation of each of these genes. First, pin2 RNAs accumulated in nonwounded and wounded flacca leaves, whereasLapA RNAs and le4 RNAs were undetectable. Second, the impact of exogenous ABA on le4 and LapA gene expression was accentuated in flacca plants relative to ABA-producing plants. Third, ABA was critical for maximal accumulation of LapA transcripts in response to systemin, and ABA and systemin appeared to act synergistically to modulate LapARNA levels. In contrast, pin2 transcript accumulation was not dependent on exogenous ABA for systemin induction. Finally, while ABA promoted le4 transcript accumulation, le4 did not respond to the signals of the octadecanoid pathway. Collectively, these data indicate that although both LapAand pin2 genes utilized the octadecanoid signaling pathway, they responded differentially to ABA. LapA responses to ABA were more similar to those of le4 than to those ofpin2. Wound induction of LapA was either dependent on ABA or required ABA levels that exceeded the residual levels in the flacca line. These data may indicate thatpin2 gene expression was more sensitive to the residual levels of ABA in flacca plants (Neill and Horgan, 1985) than were le4 (Cohen and Bray, 1990) and LapA. Alternatively, the basal pin2 transcript levels detected inflacca plants were reflective of ABA-independent expression. Data from Peña-Cortés and colleagues (1989, 1996) support the idea that pin2 expression in nonwoundedflacca leaves was due to residual ABA levels. Using an excised leaf assay, the sitiens mutant (which accumulates less ABA than flacca) exhibited no increase inpin2 RNAs, while the ABA-producing control showed a marked increase in pin2 mRNAs. This interpretation is also supported by Carrera and Prat (1998), who showed that transgenic tomato plants expressing the mutant abi1 allele from Arabidopsis, which blocks the ABA signal transduction cascade, prevents the accumulation of pin2 and LapA transcripts in response to ABA. Comparison of the data presented here and those fromPeña-Cortés et al. (1989, 1996), Carrera and Prat (1998), and Birkenmeirer and Ryan (1998) showed that the results obtained from excised shoot versus excised leaf assays are different. First, using the excised leaf assay, pin2 mRNAs are not detected in healthy, ABA-proficient plants (Peña-Cortés et al., 1989,1996; Carrera and Prat, 1998). Excised shoot assays routinely detectpin2 transcripts (Birkenmeirer and Ryan, 1998; Fig. 1). Second, exogenous ABA caused larger increases in pin2 RNA levels in excised leaves than in excised shoots. This is consistent with the difference in le4 expression noted in our studies and in previous studies that utilized detached leaf assays (Cohen et al., 1991). LapA Genes Are Induced during Abiotic Stresses That Are Accompanied by ABA Accumulation Endogenous ABA levels increase when plants are exposed to a saline environment (Downton and Loveys, 1981; Walker and Dumbroff, 1981), water deficit (Zeevaart and Creelman, 1988), or low-temperature stress (Chen et al., 1983). Like the ABA- and water-deficit-response genele4 (Cohen and Bray, 1990; Cohen et al., 1991; Kahn et al., 1993), LapA RNAs increased in response to water deficit and increases in salinity. However, differences in the le4 andLapA responses were noted. le4 RNAs increased more dramatically and persisted for a longer period of time than theLapA transcripts. It is also important that while tomatoLapA RNAs increased during water-deficit stress, water deficit did not change potato Lap transcript levels (Hildmann et al., 1992). Several barley JIP(jasmonate-induced protein) genes are also water-deficit, ABA, and JA-induced; however, these genes are not induced by salinity stress, suggesting differences in stress signaling pathways in barley and tomato (Reinbothe et al., 1992). It is not clear if the signal transduction pathways used for expression of LapA genes in response to water deficit and salinity stress were the same and corresponded to the octadecanoid pathway, or if they represent different signal transduction pathways. Clearly,le4, which is not modulated by systemin or MeJA, must utilize a signal transduction pathway distinct from the octadecanoid pathway. However, it is possible that there is cross-talk between the octadecanoid and abiotic-stress signal transduction pathways to coordinate LapA gene responses. Alternatively,LapA RNA induction by abiotic stress might solely utilize the wound-response octadecanoid pathway, since increases in JA have been measured for several abiotic stresses (Creelman and Mullet, 1995). It is also possible that the active oxygen species that accumulate during water deficit (Davies and Mansfield, 1983; Inzé and van Montagu, 1995) might activate the octadecanoid signaling pathway by causing lipid peroxidation and lipoxygenase production (Keppler and Novacky, 1989; Ádám et al., 1989). Rises in ABA could further activate the octadecanoid pathway to ultimately increase JA levels and activate LapA gene expression. The LapA1 Promoter Is Responsive to Wound and Developmental Signals Analysis of transgenic tomato and tobacco plants expressing a chimeric LapA1:GUS transgene demonstrated that theLapA1 promoter sequences that responded to wound and developmental signals were located within the 1st kb of theLapA1 5′-flanking sequences. Similar results have been reported for the expression of the tomato LapA1 gene in potato (Ruı́z-Rivero and Prat, 1998) and the tomatoLapA2 gene in Arabidopsis (A. El Amrani and V. Pautot, unpublished data). The LapA1 promoter was not active in vegetative organs unless tissues were wounded, P. syringae pv tomato infected, or treated with a wound signal such as MeJA. The lack of LapA1 promoter activity in cotyledons after germination indicated that the LAP-A protein does not have a role in the mobilization of storage protein reserves (Walling and Gu, 1996). Aminopeptidases with properties similar to LAP-A have been characterized from kidney bean and barley seeds (Sopanen and Mikola, 1975; Mikkonen, 1992). The data presented here indicate that the kidney bean and barley seed LAPs are likely to be analogs of the constitutively expressed LAP-N of tomato and Arabidopsis LAP (Bartling and Nosek, 1994; Gu et al., 1996b; C.J. Tu and L.L. Walling, unpublished data). Compared with other wound- and defense-response genes (Lotan et al., 1989; Cote et al., 1991; Uknes et al., 1993; Constabel and Brisson, 1995), the LapA1 promoter has a unique developmental specificity. In transgenic tomato, the LapA1 promoter was active in all floral organs, which is similar to the activity seen for tomato pin2 RNA accumulation (Peña-Cortés et al., 1991). However, unlike pin2 genes, LapAgenes were active throughout all of fruit development. Other wound-response genes (i.e. the wound-induced ACC synthase gene and apin1-like gene) are expressed only during the ripening phase of fruit development (Margossian et al., 1988; Li et al., 1992), when endogenous ethylene levels rise. Recently, Ruı́z-Rivero and Prat (1998) reported an analysis of the tomato LapA1 promoter in transgenic potato, and their results differed from the data reported here. They did not detect LapA1 promoter activity in stigmas, styles, or ovaries, although expression in other floral organs was reported. At present, it is not known what signals are responsible for the activation of the LapA1 promoter in tomato flowers and fruit. However, the signaling mechanisms appear to be different in potato and tomato (Peña-Cortés et al., 1991;Ruı́z-Rivero and Prat, 1998). The availability of tomato mutants that impact biosynthesis or perception of ABA, JA, and ethylene may aid in resolving their roles in developmental programming ofLapA gene expression. Plants utilize intricate systems for the expression of defense genes during floral and fruit development. The overlapping patterns of expression of the vast array of defense- and wound-response genes may ensure production of viable seeds. LAP-A proteins may play a defensive (or protective) role in tomato flowers and fruit by protecting gametes from damage by insect or pathogen attack. While the role of protease inhibitors in the control of insect predation is established and highly publicized (Johnson et al., 1989; Xu et al., 1996), a few studies have shown that proteases are important in plant defense. For example, one study showed that a Cys endoprotease confers resistance to maize against fall armyworm (Jiang et al., 1995). It is possible that exopeptidases such as LAP-A or the tomato wound-induced carboxypeptidases may have important roles in plant defense (Pautot et al., 1993; Mehta et al., 1996; Walling and Gu, 1996). In animals, exopeptidases are important in the activation and inactivation of bioactive peptides and regulation of protein half-lives (Taylor, 1996; Varshavsky, 1996; Bradshaw et al., 1998). In a similar manner, LAP-A may serve to modulate levels or activities of regulatory proteins or peptides. Alternatively, LAP-A may facilitate turnover of proteins that are damaged due to reactive oxygen species generated during wounding, or may hydrolyze proteins to supply the pool of amino acids to support the substantial changes in protein synthesis associated with wounding. Current studies using antisense plants and plants overexpressing LapA are in progress. These studies will help to resolve the roles of LAP-As during floral and fruit development and during wounding and defense responses. ACKNOWLEDGMENTS We would like to thank members of the Walling laboratory for helpful discussions; David Puthoff for ethylene measurements; Fran Holzer for aid in figure modifications; Dr. Timothy Close (Department of Botany and Plant Sciences, University of California, Riverside) for reading earlier versions of this manuscript and for the use of his microplate reader; Dr. Richard Whitkus (Department of Botany and Plant Sciences, University of California, Riverside) for the use of his fluorometer; Dr. Jan Oakes (Calgene, Davis, CA) for her training of W.S.C. in tomato transformation; and Dr. Jocelyne Kronenberger (Laboratoire de Biologie Cellulaire, Institut National de la Recherche Agronomique) for her aid with in situ hybridizations. Abbreviations: JA jasmonic acid MeJA methyl jasmonate SA salicylic acid: SAR, systemic acquired resistance LITERATURE CITED 1 Ádám A Farkas T Somlya G Hevesi M Kiraly Z Consequence of O2 − generation during a bacterially induced hypersensitive reaction in tobacco: deterioration of membrane lipids. Physiol Mol Plant Pathol 34 1989 13 26 Google Scholar Crossref Search ADS WorldCat 2 Alarcon JJ Malone M The influence of plant age on wound induction of proteinase inhibitors in tomato. Physiol Plant 95 1995 423 427 Google Scholar Crossref Search ADS WorldCat 3 Bartling D Nosek J Molecular and immunological characterization of leucine aminopeptidase in Arabidopsis thaliana: a new antibody suggests a semi-constitutive regulation of a phylogenetically old enzyme. Plant Sci 9 1994 199 209 Google Scholar OpenURL Placeholder Text WorldCat 4 Bent AF Innes RW Ecker JR Staskawicz BJ Disease development in ethylene-insensitive Arabidopsis thaliana infected with virulent and avirulent Pseudomonas and Xanthomonas pathogens. Mol Plant-Microbe Interact 5 1992 371 378 Google Scholar Crossref Search ADS WorldCat 5 Bergey DR Howe GA Ryan CA Polypeptide signaling for plant defensive genes exhibits analogies to defense signaling in animals. Proc Natl Acad Sci USA 93 1996 12053 12058 Google Scholar Crossref Search ADS PubMed WorldCat 6 Birkenmeier GF Ryan CA Wound signaling in tomato plants: evidence that ABA is not a primary signal for defense gene activation. Plant Physiol 117 1998 687 693 Google Scholar Crossref Search ADS PubMed WorldCat 7 Birnboim HC Doly J A rapid alkaline extraction procedure for screening recombinant plasmid DNA. Nucleic Acids Res 7 1979 1513 1523 Google Scholar Crossref Search ADS PubMed WorldCat 8 Bowles D Defense-related proteins in higher plants. Annu Rev Biochem 59 1990 873 907 Google Scholar Crossref Search ADS PubMed WorldCat 9 Bradshaw RA Brickey WW Walker KW N-terminal processing: the methionine aminopeptidase and N-alpha-acetyl transferase families. Trends Biochem Sci 23 1998 263 267 Google Scholar Crossref Search ADS PubMed WorldCat 10 Bray EA Molecular responses to water deficit. Plant Physiol 103 1993 1035 1040 Google Scholar Crossref Search ADS PubMed WorldCat 11 Carrera E Prat S Expression of the Arabidopsis abi1-1 mutant allele inhibits proteinase inhibitor wound-induction in tomato. Plant J 15 1998 765 771 Google Scholar Crossref Search ADS PubMed WorldCat 12 Chandler PM Robertson M Gene expression regulated by abscisic acid and its relation to stress tolerance. Annu Rev Plant Physiol Plant Mol Biol 45 1994 113 141 Google Scholar Crossref Search ADS WorldCat 13 Chao WS (1996) Characterization of wound-induced leucine aminopeptidase genes (LapA1 and LapA2) in tomato and their expression in response to environmental and developmental cues. PhD Thesis. University of California, Riverside 14 Chen H-H Li PH Brenner ML Involvement of abscisic acid in potato cold acclimation. Plant Physiol 71 1983 362 365 Google Scholar Crossref Search ADS PubMed WorldCat 15 Christ U Mösinger E Pathogenesis-related proteins of tomato 1. Induction by Phytophthora infestans and other biotic and abiotic inducers and correlations with resistance. Physiol Mol Plant Path 35 1989 53 65 Google Scholar Crossref Search ADS WorldCat 16 Cohen A Bray EA Characterization of three mRNAs that accumulate in wilted tomato leaves in response to elevated levels of endogenous abscisic acid. Planta 182 1990 27 33 Google Scholar Crossref Search ADS PubMed WorldCat 17 Cohen A Plant ÁL Moses MS Bray EA Organ-specific and environmentally regulated expression of two abscisic acid-induced genes of tomato. Plant Physiol 97 1991 1367 1374 Google Scholar Crossref Search ADS PubMed WorldCat 18 Constabel CP Brisson N Stigma- and vascular-specific expression of the PR-10a gene of potato: a novel pattern of expression of a pathogenesis-related gene. Mol Plant-Microbe Interact 8 1995 104 113 Google Scholar Crossref Search ADS WorldCat 19 Cote F Cutt JR Asselin A Klessig DF Pathogenesis-related acidic β-1,3-glucanase genes of tobacco are regulated by both stress and developmental signals. Mol Plant-Microbe Interact 4 1991 173 181 Google Scholar Crossref Search ADS PubMed WorldCat 20 Creelman RA Mullet JE Jasmonic acid distribution and action in plants: regulation during development and response to biotic and abiotic stress. Proc Natl Acad Sci USA 92 1995 4114 4119 Google Scholar Crossref Search ADS PubMed WorldCat 21 Dangl JL Dietrich RA Richberg MH Death don't have no mercy: cell death programs in plant-microbe interactions. Plant Cell 8 1996 1793 1807 Google Scholar Crossref Search ADS PubMed WorldCat 22 Davies WJ, Mansfield TA (1983) The role of abscisic acid in drought avoidance. In FT Addicott, ed, Abscisic Acid. Praeger, New York, pp 237–268 23 Dixon RA Harrison MJ Lamb CJ Early events in the activation of plant defense responses. Annu Rev Phytopath 32 1994 479 501 Google Scholar Crossref Search ADS WorldCat 24 Doares SH Narvaez-Vasquez J Conconi A Ryan CA Salicylic acid inhibits synthesis of proteinase inhibitors in tomato leaves induced by systemin and jasmonic acid. Plant Physiol 108 1995 1741 1746 Google Scholar Crossref Search ADS PubMed WorldCat 25 Doherty HM Selvendran RR Bowles DJ The wound response of tomato plants can be inhibited by aspirin and related hydroxy-benzoic acids. Physiol Mol Plant Pathol 33 1988 377 384 Google Scholar Crossref Search ADS WorldCat 26 Downton WJS Loveys BR Abscisic acid content and osmotic relations of salt-stressed grapevine leaves. Aust J Plant Physiol 8 1981 443 453 Google Scholar OpenURL Placeholder Text WorldCat 27 Ecker J Davis RW Plant defense genes are regulated by ethylene. Proc Natl Acad Sci USA 84 1987 5202 5206 Google Scholar Crossref Search ADS PubMed WorldCat 28 Enkerli J Gisi U Mösinger E Systemic acquired resistance to Phytophthora infestans in tomato and the role of pathogenesis-related proteins. Physiol Mol Plant Path 43 1993 161 171 Google Scholar Crossref Search ADS WorldCat 29 Farmer EE Ryan CA Interplant communication: airborne methyl jasmonate induces synthesis of proteinase inhibitors in plant leaves. Proc Natl Acad Sci USA 87 1990 7713 7716 Google Scholar Crossref Search ADS PubMed WorldCat 30 Farmer EE Ryan CA Octadecanoid precursors of jasmonic acid activate the synthesis of wound-inducible proteinase inhibitors. Plant Cell 4 1992 129 134 Google Scholar Crossref Search ADS PubMed WorldCat 31 Fillatti JJ Kiser J Ronald R Comai L Efficient transfer of glyphosate tolerance gene into tomato using a binary Agrobacterium tumefaciens vector. Biotechniques 5 1987 726 730 Google Scholar OpenURL Placeholder Text WorldCat 32 Gelvin SB, Schilperoort RA (1988) Plant Molecular Biology Manual. Kluwer Academic Publishers, Dordrecht, The Netherlands, pp A3–A4 33 Graham JS Pearce G Merryweather J Titani K Ericsson L Ryan CA Wound-induced proteinase inhibitors from tomato leaves: the cDNA-deduced primary sequence of pre-inhibitor II. J Biol Chem 260 1985 6561 6564 Google Scholar Crossref Search ADS PubMed WorldCat 34 Gu Y-Q Chao WS Walling LL Localization and post-translational processing of the wound-induced leucine aminopeptidase proteins of tomato. J Biol Chem 271 1996a 25880 25887 Google Scholar Crossref Search ADS WorldCat 35 Gu Y-Q Pautot V Holzer FM Walling LL A complex array of proteins related to the multimeric leucine aminopeptidase of tomato. Plant Physiol 110 1996b 1257 1266 Google Scholar Crossref Search ADS WorldCat 36 Harms K Atzorn R Brash A Kühn H Wasternack C Willmitzer L Peña-Cortés H Expression of a flax allene oxide synthase cDNA leads to increased endogenous jasmonic acid (JA) levels in transgenic potato but not to a corresponding activation of JA-responding genes. Plant Cell 7 1995 1645 1654 Google Scholar Crossref Search ADS PubMed WorldCat 37 Herbers K Prat S Willmitzer L Functional analysis of a leucine aminopeptidase from Solanum tuberosum L. Planta 194 1994 230 240 Google Scholar Crossref Search ADS PubMed WorldCat 38 Herde O Atzorn R Fisahn J Wasternack C Willmitzer L Peña-Cortés H Localized wounding by heat initiates the accumulation of proteinase inhibitor II in abscisic acid-deficient plant by triggering jasmonic acid biosynthesis. Plant Physiol 112 1996 853 860 Google Scholar Crossref Search ADS PubMed WorldCat 39 Hildmann T Ebneth M Peña-Cortés H Sánchez-Serrano J Willmitzer L Prat S General roles of abscisic and jasmonic acids in gene activation as a result of mechanical wounding. Plant Cell 4 1992 1157 1170 Google Scholar PubMed OpenURL Placeholder Text WorldCat 40 Hill HD Straka JG Protein determination using bicinchoninic acid in the presence of sulfhydryl reagents. Anal Biochem 170 1988 203 208 Google Scholar Crossref Search ADS PubMed WorldCat 41 Inzé D van Montagu MV Oxidative stress in plants. Curr Opin Biotech 6 1995 153 158 Google Scholar Crossref Search ADS WorldCat 42 Jefferson RA Assaying chimeric genes in plants: the GUS gene fusion system. Plant Mol Biol Rep 5 1987 387 405 Google Scholar Crossref Search ADS WorldCat 43 Jiang B Siregar U Willeford KO Luthe DS Williams WP Association of a 33-kilodalton cysteine proteinase found in corn callus with the inhibition of fall armyworm larval growth. Plant Physiol 108 1995 1631 1640 Google Scholar Crossref Search ADS PubMed WorldCat 44 Johnson R Narvrez J An GH Ryan CA Expression of proteinase inhibitor-I and inhibitor-II in transgenic tobacco plants: effects on natural defense against Manduca sexta larvae. Proc Natl Acad Sci USA 86 1989 9871 9875 Google Scholar Crossref Search ADS PubMed WorldCat 45 Kahn TL Fender SE Bray EA O'Connell MA Plant Physiol 103 1993 597 605 Crossref Search ADS PubMed 46 Keppler LD Novacky A The initiation of membrane lipid peroxidation during bacteria-induced hypersensitive reaction. Physiol Mol Plant Pathol 30 1989 233 245 Google Scholar Crossref Search ADS WorldCat 47 Kernan A Thornberg RW Auxin levels regulate the expression of a wound-inducible proteinase inhibitor II-chloramphenicol acetyl transferase gene fusion in vitro and in vivo. Plant Physiol 91 1989 73 78 Google Scholar Crossref Search ADS PubMed WorldCat 48 Kosugi S Ohashi Y Nakajima K Arai Y An improved assay for β-glucuronidase in transformed cells: methanol almost completely suppresses a putative endogenous β-glucuronidase activity. Plant Sci 70 1990 133 140 Google Scholar Crossref Search ADS WorldCat 49 Kronenberger J Desprez T Höfte H Caboche M Traas J A methacrylate embedding procedure developed for immunolocalization in plant tissues is also compatible with in situ hybridization. Cell Biol Int 17 1993 1013 1021 Google Scholar Crossref Search ADS WorldCat 50 Lawton KA Potter SL Uknes S Ryals J Acquired resistance signal transduction in Arabidopsis is ethylene independent. Plant Cell 6 1994 581 588 Google Scholar Crossref Search ADS PubMed WorldCat 51 Li N Parsons BL Liu D Mattoo AK Accumulation of wound-inducible ACC synthase transcript in tomato fruit is inhibited by salicylic acid and polyamines. Plant Mol Biol 18 1992 477 487 Google Scholar Crossref Search ADS PubMed WorldCat 52 Linthorst HJM Danhash N Brederode FT Vankan JAL Dewit P Bol JF Tobacco and tomato PR proteins homologous to Win and pro-hevein lack the hevein domain. Mol Plant-Microbe Interact 4 1991 586 592 Google Scholar PubMed OpenURL Placeholder Text WorldCat 53 Lotan T Ori N Fluhr R Pathogenesis-related proteins are developmentally regulated in tobacco flowers. Plant Cell 1 1989 881 887 Google Scholar PubMed OpenURL Placeholder Text WorldCat 54 Lund ST Stall RE Klee HJ Ethylene regulates the susceptible response to pathogen infection in tomato. Plant Cell 10 1998 371 382 Google Scholar Crossref Search ADS PubMed WorldCat 55 Malamy J Carr JP Klessig DF Raskin I Salicylic acid, a likely endogenous signal in the resistance response of tobacco to viral infection. Science 250 1990 1002 1004 Google Scholar Crossref Search ADS PubMed WorldCat 56 Malone M, Alarcon J-J, Palumbo L (1994) An hydraulic interpretation of rapid, long-distance wound signaling in the tomato. Planta 193: 181–185 57 Margossian LJ Federman AD Giovannoni JJ Fischer RL Ethylene-regulated expression of a tomato fruit ripening gene encoding a proteinase inhibitor I with a glutamic residue at the reactive site. Proc Natl Acad Sci USA 85 1988 8012 8016 Google Scholar Crossref Search ADS PubMed WorldCat 58 McCormick S (1991) Transformation of tomato withAgrobacterium tumefaciens. In Lindsey K, ed, Plant Tissue Culture Manual: Fundamentals and Applications, Vol B6. Kluwer Academic Publishers, Dordrecht, The Netherlands, pp 1–9 59 Mehta RA Warmbardt RD Mattoo AK Tomato fruit carboxypeptidase: properties, induction upon wounding, and immunocytochemical localization. Plant Physiol 110 1996 883 892 Google Scholar Crossref Search ADS PubMed WorldCat 60 Mikkonen A Purification and characterization of leucine aminopeptidase from kidney bean cotyledons. Physiol Plant 84 1992 393 398 Google Scholar Crossref Search ADS WorldCat 61 Milligan SB Gasser CS Nature and regulation of pistil-expressed genes in tomato. Plant Mol Biol 28 1995 691 711 Google Scholar Crossref Search ADS PubMed WorldCat 62 Narváez-Vásquez J Pearce G Orozco-Cárdenas ML Franceschi VR Ryan CA Autoradiographic and biochemical evidence for the systemic translocation of systemin in tomato plants. Planta 195 1995 593 600 Google Scholar Crossref Search ADS WorldCat 63 Neill SJ Horgan R Abscisic acid production and water relations in wilty tomato mutants subjected to water deficiency. J Exp Bot 36 1985 1222 1231 Google Scholar Crossref Search ADS WorldCat 64 Niderman T Genetet I Bruyere T Gees R Stintzi A Legrand M Fritig B Mosinger E Pathogenesis-related PR–1 proteins are antifungal: isolation and characterization of three 14-kilodalton proteins of tomato and of a basic PR-1 of tobacco with inhibitory activity against Phytophthora infestans. Plant Physiol 108 1995 17 27 Google Scholar Crossref Search ADS PubMed WorldCat 65 O'Donnell PJ Calvert C Atzorn R Wasternack C Leyser HMO Bowles DJ Ethylene as a signal mediating the wound response of tomato plants. Science 274 1996 1914 1918 Google Scholar Crossref Search ADS PubMed WorldCat 66 Pautot V Holzer FM Reisch B Walling LL Leucine aminopeptidase: an inducible component of the defense response in Lycopersicon esculentum (tomato). Proc Natl Acad Sci USA 90 1993 9906 9910 Google Scholar Crossref Search ADS PubMed WorldCat 67 Pautot V Holzer FM Walling LL Differential expression of tomato proteinase inhibitor I and II genes during bacterial pathogen invasion and wounding. Mol Plant-Microbe Interact 4 1991 284 292 Google Scholar Crossref Search ADS PubMed WorldCat 68 Pearce G Johnson S Ryan CA Structure-activity of deleted and substituted systemin, an 18-amino acid polypeptide inducer of plant defensive genes. J Biol Chem 268 1993 212 216 Google Scholar Crossref Search ADS PubMed WorldCat 69 Pearce G Strydom D Johnson S Ryan CA A polypeptide from tomato leaves induces wound-inducible proteinase inhibitor proteins. Science 253 1991 895 898 Google Scholar Crossref Search ADS PubMed WorldCat 70 Peña-Cortés H Albrecht T Prat S Weiler EW Willmitzer L Aspirin prevents wound-induced gene expression in tomato leaves by blocking jasmonic acid biosynthesis. Planta 191 1993 123 128 Google Scholar OpenURL Placeholder Text WorldCat 71 Peña-Cortés H Fisahn J Willmitzer L Signals involved in wound-induced proteinase inhibitor II gene expression in tomato and potato plants. Proc Natl Acad Sci USA 92 1995 4106 4113 Google Scholar Crossref Search ADS PubMed WorldCat 72 Peña-Cortés H Prat S Atzorn R Wasternack C Willmitzer L Abscisic acid-deficient plants do not accumulate proteinase inhibitor II following systemin treatment. Planta 198 1996 447 451 Google Scholar Crossref Search ADS WorldCat 73 Peña-Cortés H Sánchez-Serrano JJ Mertens R Willmitzer L Prat S Abscisic acid is involved in the wound-induced expression of the proteinase inhibitor II gene in potato and tomato. Proc Natl Acad Sci USA 86 1989 9851 9855 Google Scholar Crossref Search ADS PubMed WorldCat 74 Peña-Cortés H Willmitzer L Sánchez-Serrano JJ Abscisic acid mediates wound induction but not developmental-specific expression of the proteinase inhibitor II gene family. Plant Cell 3 1991 963 972 Google Scholar Crossref Search ADS PubMed WorldCat 75 Penninckx IAMA Eggermont K Terras FRG Thomma BPHJ De Samblanx GW Buchala A Métraux J-P Manners JM Broekaert WF Pathogen-induced systemic activation of a plant defensin gene in Arabidopsis follows a salicylic acid-independent pathway. Plant Cell 8 1995 2309 2323 Google Scholar OpenURL Placeholder Text WorldCat 76 Pieterse CMJ van Wees SCM Hoffland E van Pelt JA van Loon LC Systemic resistance in Arabidopsis induced by biocontrol bacteria is independent of salicylic acid accumulation and pathogenesis-related gene expression. Plant Cell 8 1996 1225 1237 Google Scholar PubMed OpenURL Placeholder Text WorldCat 77 Ramagli LS Rodriguez LV Quantitation of microgram amounts of protein in two-dimensional polyacrylamide gel electrophoresis sample buffer. Electrophoresis 6 1985 559 563 Google Scholar Crossref Search ADS WorldCat 78 Raz V Fluhr R Ethylene is transduced via protein phosphorylation events in plants. Plant Cell 5 1993 523 530 Google Scholar Crossref Search ADS PubMed WorldCat 79 Reinbothe S Reinbothe C Lehmann J Parthier B Differential accumulation of methyl jasmonate-induced mRNAs in response to abscisic acid and desiccation in barley (Hordeum vulgare). Physiol Plant 86 1992 49 56 Google Scholar Crossref Search ADS WorldCat 80 Ruı́z-Rivero O Prat S A −308 deletion of the tomato LAP promoters is able to direct flower-specific and MeJA-induced expression in transgenic plants. Plant Mol Biol 36 1998 639 648 Google Scholar Crossref Search ADS PubMed WorldCat 81 Ryals J Neuenschwander UH Willits MG Molina A Steiner H-Y Hunt MD Systemic acquired resistance. Plant Cell 8 1996 1809 1819 Google Scholar Crossref Search ADS PubMed WorldCat 82 Ryan CA Assay and biochemical properties of the proteinase inhibitor inducing factor, a wound hormone. Plant Physiol 54 1974 328 332 Google Scholar Crossref Search ADS PubMed WorldCat 83 Sano H Seo S Orudgev E Youssefian S Ishizuka K Ohashi Y Expression of the gene for a small GTP biding protein in transgenic tobacco elevates endogenous cytokinin levels, abnormally induces salicylic acid in response to wounding, and increases resistance to tobacco mosaic virus infection. Proc Natl Acad Sci USA 91 1994 10556 10560 Google Scholar Crossref Search ADS PubMed WorldCat 84 Schaller A Ryan CA Systemin: a polypeptide defense signal in plants. BioEssays 18 1995 27 33 Google Scholar Crossref Search ADS WorldCat 85 Seo S Okamoto M Seto H Ishizuka K Sano H Ohashi Y Tobacco MAP kinase: a possible mediator in wound signal transduction pathways. Science 270 1995 1988 1992 Google Scholar Crossref Search ADS PubMed WorldCat 86 Sopanen T Mikola J Purification and partial characterization of barley leucine aminopeptidase. Plant Physiol 55 1975 809 814 Google Scholar Crossref Search ADS PubMed WorldCat 87 Stankovic B Davies E The wound response in tomato involves rapid growth and electrical responses, systemically up-regulated transcription of proteinase inhibitor and calmodulin and down-regulated translation. Plant Cell Physiol 39 1998 268 274 Google Scholar Crossref Search ADS WorldCat 88 Taylor A Aminopeptidases. 1996 R.G. Landes Georgetown, TX 89 Titarenko E Rojo E Leó J Sánchez-Serrano JJ Jasmonic acid-dependent and -independent signaling pathways control wound-induced gene activation in Arabidopsis thaliana. Plant Physiol 115 1997 817 826 Google Scholar Crossref Search ADS PubMed WorldCat 90 Tornero P Gdea J Conejero V Vera P Two PR1 genes from tomato are differentially regulated and reveal a novel mode of expression for a pathogenesis-related gene during the hypersensitive response and development. Mol Plant-Microbe Interact 10 1997 624 634 Google Scholar Crossref Search ADS PubMed WorldCat 91 Uknes S Dincher S Friedrich L Negrotto D Williams S Thompson-Taylor H Potter S Ward E Ryals J Regulation of pathogenesis-related protein-1a gene expression in tobacco. Plant Cell 5 1993 159 169 Google Scholar PubMed OpenURL Placeholder Text WorldCat 92 van Kan JAL Cozijnsen T Danhash N de Wit PJGM Induction of tomato stress protein RNAs by ethephon, 2,6-dichloroisonicotinic acid and salicylate. Plant Mol Biol 27 1995 1205 1213 Google Scholar Crossref Search ADS PubMed WorldCat 93 van Kan JAL Joosten MHAJ Wagemakers CAM van den Berg-Velthuis GCM de Wit PJGM Differential accumulation of mRNAs encoding extracellular and intracellular PR proteins in tomato induced by virulent and avirulent races of Cladosporium fulvum. Plant Mol Biol 20 1992 513 517 Google Scholar Crossref Search ADS PubMed WorldCat 94 Varshavsky A The N-end rule: functions, mysteries, uses. Proc Natl Acad Sci USA 93 1996 12142 12149 Google Scholar Crossref Search ADS PubMed WorldCat 95 Vidal S de Léon IP Denecke J Palva ET Salicyclic acid and the plant pathogen Erwinia carotovora induce defense genes via antagonistic pathways. Plant J 11 1997 115 123 Google Scholar Crossref Search ADS WorldCat 96 Walker MA Dumbroff EB Effect of salt stress on abscisic acid and cytokinin levels in tomato. Z Pflanzenphysiol 101 1981 461 470 Google Scholar Crossref Search ADS WorldCat 97 Walling LL Chang YC Demmin DS Holzer FM Isolation, characterization and evolutionary relatedness of three members from the soybean multigene family encoding chlorophyll a/b binding proteins. Nucleic Acids Res 16 1988 10477 10492 Google Scholar Crossref Search ADS PubMed WorldCat 98 Walling LL Gu Y-Q Plant aminopeptidases: occurrence function and characterization. Taylor A The Aminopeptidases. 1996 174 219 RG Landes Georgetown, TX 99 Wang C-S Walling LL Eckard KJ Lord EM Patterns of protein accumulation in developing anthers of Lilium longiflorum correlate with histological events. Am J Bot 79 1992 118 127 Google Scholar Crossref Search ADS WorldCat 100 Wildon DC Thain JF Minchin PEH Gubb IR Reilly AJ Skipper YD Doherty HM O'Donnell PJ Bowles DJ Electrical signaling and systemic proteinase inhibitor in the wounded plant. Nature 360 1992 62 64 Google Scholar Crossref Search ADS WorldCat 101 Wolfson JL Murdock LL Growth of Manduca sexta on wounded tomato: role of induced proteinase inhibitors. Entomol Exp Appl 54 1990 257 264 Google Scholar Crossref Search ADS WorldCat 102 Xu D Xue Q McElroy D Mawal Y Hilder VA Wu R Constitutive expression of a cowpea trypsin inhibitor gene, CpTi, in transgenic rice plants confers resistance to two major rice insect pests. Mol Breed 2 1996 167 173 Google Scholar Crossref Search ADS WorldCat 103 Xu Y Chang P-FL Liu D Narasimhan ML Raghothama KG Hasagawa PM Bressan RA Plant defense genes are synergistically induced by ethylene and methyl jasmonate. Plant Cell 6 1994 1077 1085 Google Scholar Crossref Search ADS PubMed WorldCat 104 Yang Y Shah I Klessig DF Signal perception and transduction in plant defense responses. Genes Dev 11 1996 1621 1639 Google Scholar Crossref Search ADS WorldCat 105 Zeevaart JAD Creelman RA Metabolism and physiology of abscisic acid. Annu Rev Plant Physiol Plant Mol Biol 39 1988 439 473 Google Scholar Crossref Search ADS WorldCat Author notes 1 This research was supported by a National Science Foundation grant (no. IBN–9318260) to L.L.W. W.S.C. was partially supported by the National Science Foundation training grant (no. GER–5355042). 2 Present address: Department of Botany, Washington State University, Pullman, WA 99164–4238. 3 Present address: Boyce Thompson Institute, Cornell University, Ithaca, New York 14853–1801. * Corresponding author; e-mail [email protected]; fax 909–787–4437. Copyright © 1999 American Society of Plant Physiologists This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)

Colleoni, Christophe; Dauvillée, David; Mouille, Gregory; Buléon, Alain; Gallant, Daniel; Bouchet, Brigitte; Morell, Matthew; Samuel, Michael; Delrue, Brigitte; d'Hulst, Christophe; Bliard, Christophe; Nuzillard, Jean-Marc; Ball, Steven

Colleoni, Christophe; Dauvillée, David; Mouille, Gregory; Morell, Matthew; Samuel, Michael; Slomiany, Marie-Christine; Liénard, Luc; Wattebled, Fabrice; d'Hulst, Christophe; Ball, Steven

doi: 10.1104/pp.120.4.1005pmid:

Miller, Jennifer D.; Arteca, Richard N.; Pell, Eva J.

doi: 10.1104/pp.120.4.1015pmid: 10444084

Abstract The expression patterns of senescence-related genes were determined during ozone (O3) exposure in Arabidopsis. Rosettes were treated with 0.15 μL L−1 O3 for 6 h d−1 for 14 d. O3-treated leaves began to yellow after 10 d of exposure, whereas yellowing was not apparent in control leaves until d 14. Transcript levels for eight of 12 senescence related genes characterized showed induction by O3. SAG13(senescence-associated gene), SAG21, ERD1(early responsive to dehydration), and BCB (blue copper-binding protein) were induced within 2 to 4 d of O3 treatment; SAG18, SAG20, and ACS6 (ACC synthase) were induced within 4 to 6 d; and CCH (copper chaperone) was induced within 6 to 8 d. In contrast, levels of photosynthetic gene transcripts,rbcS (small subunit of Rubisco) and cab(chlorophyll a/b-binding protein), declined after 6 d. Other markers of natural senescence, SAG12,SAG19, MT1 (metallothionein), andAtgsr2 (glutamine synthetase), did not show enhanced transcript accumulation. When SAG12promoter-GUS (β-glucuronidase) andSAG13 promoter-GUS transgenic plants were treated with O3, GUS activity was induced in SAG13-GUS plants after 2 d but was not detected in SAG12-GUS plants.SAG13 promoter-driven GUS activity was located throughout O3-treated leaves, whereas control leaves generally showed activity along the margins. The acceleration of leaf senescence induced by O3 is a regulated event involving many genes associated with natural senescence. Leaf senescence is the sequence of degradative processes leading to the remobilization of nutrients and eventual leaf death. The senescence process is highly regulated, involving photosynthetic decline, protein degradation, lipid peroxidation, and chlorophyll degradation (Smart, 1994). Total RNA levels decline during senescence as RNase activity increases (Blank and McKeon, 1991). Chloroplasts are one of the earliest sites of catabolism, while mitochondria remain intact until late in the senescence process in order for respiration to continue (Smart, 1994). Plant hormones are involved in regulating the senescence process, with cytokinins delaying senescence, ethylene modulating the timing of senescence, and the other hormones playing less prominent roles (Smart, 1994). Leaf senescence, like other developmental processes, is actively regulated by differential gene expression. Transcript levels for photosynthetic genes such asrbcS (small subunit of Rubisco) and cab(chlorophyll a/b-binding protein) decline (Bate et al., 1991), while other genes become activated (Buchanan-Wollaston, 1997;Weaver et al., 1997). Using differential screening and subtractive hybridization techniques, researchers have identified genes with increased expression during senescence. These genes have been identified in Arabidopsis, oilseed rape, tomato, barley, potato, cucumber, rice, wheat, and maize (for reviews, see Buchanan-Wollaston, 1997; Weaver et al., 1997). Such genes are often referred to as SAGs or senescence-up-regulated genes. Among the identified senescence-induced genes are genes encoding proteases, RNases, Gln synthetase, metallothioneins, protease regulators, ACC oxidase, lipases, glyoxylate cycle enzymes, catalase, endoxyloglucan transferase, pathogenesis-related proteins, ATP sulfurylase, glutathione S-transferase, Cyt P450, and polyubiquitin (Buchanan-Wollaston, 1997; Weaver et al., 1997). Some identified cDNA clones have no obvious senescence-related function and other senescence-induced clones remain unidentified. While the initiation of leaf senescence depends upon the age of the leaf and the reproductive phase of the plant, external factors such as nutrient deficiency, pathogenic attack, drought, light limitation, and temperature can induce premature senescence (Smart, 1994). Researchers have begun to examine the similarities and differences in gene expression during natural senescence, hormone treatment, and stress by measuring the induction of senescence-related genes (Becker and Apel, 1993; Oh et al., 1996; Chung et al., 1997; Park et al., 1998; Weaver et al., 1998). Studies with ABA, ethylene, cytokinin, methyl jasmonate, wounding, dehydration, and dark treatment have shown that these genes are differentially regulated, suggesting that there are multiple signaling pathways leading to their induction (Gan and Amasino, 1997; Park et al., 1998; Weaver et al., 1998). Expression of some senescence-related genes appears to be quite specific to natural senescence, whereas other transcripts are induced by treatments in addition to natural senescence (Weaver et al., 1998). Ozone (O3) is a stress known to induce accelerated foliar senescence in many plant species including potato, radish, alfalfa, wheat, and hybrid poplar (Pell and Pearson, 1983;Reich, 1983; Held et al., 1991; Nie et al., 1993; Pell et al., 1997). O3 exposure accelerates chlorophyll and protein loss and reduces photosynthetic capacity and efficiency in older leaves (Reich, 1983; Held et al., 1991; Nie et al., 1993). Accelerated loss of Rubisco protein is also closely associated with O3-induced senescence (Pell and Pearson, 1983;Nie et al., 1993; Pell et al., 1997). O3 exposure reduced transcript levels for cab, rbcS, andrbcL (large subunit of Rubisco) in potato (Glick et al., 1995) and cab and rbcS in Arabidopsis (Conklin and Last, 1995) and tobacco (Bahl and Kahl, 1995). Accelerated yellowing of older leaves occurred in Arabidopsis plants following exposure to 0.10 to 0.15 μL L−1O3 given continuously for 2 d (Kubo et al., 1995). Exposure to 0.15 μL L−1O3 for 8 d reduced Arabidopsis rosette dry weight by 44% and reduced total chlorophyll, carotenoids, Rubisco activity, and levels of Rubisco large and small subunits (Rao et al., 1995). These results demonstrate that O3 induces changes associated with natural senescence in many species including Arabidopsis. While a decline in message level for photosynthetic genes has been observed during O3-induced accelerated leaf senescence, other molecular changes known to occur during natural senescence have not, to our knowledge, been reported. The main objective of this study was to determine whether O3 exposure regulates the expression of SAGs. The expression pattern of SAG12 and SAG13 was determined by fluorometric quantification of GUS activity in transgenic Arabidopsis carrying either the SAG12 promoter-GUS or the SAG13 promoter-GUS fusion. Expression levels for SAG12 andSAG13 were also characterized by northern analysis. The spatial distribution of SAG13 expression was determined by staining for GUS activity in O3-treated and control transgenic SAG13-GUS plants. The expression patterns of 10 additional senescence-related genes were characterized by northern analysis in relation to the decline in PAG expression. SAG transcript levels were also analyzed following removal of the O3 treatment to determine whether transcript levels remained elevated or returned to control levels. MATERIALS AND METHODS Plant Growth and O3 Exposure Experiments Seeds of Arabidopsis ecotype Lansberg erectatransformed with the SAG12 promoter-GUS fusion orSAG13 promoter-GUS fusion, were provided by S. Gan and Richard Amasino (University of Wisconsin, Madison). SAG12-GUS, SAG13-GUS, and wild-type Lansberg erecta seeds were planted on a commercial soil mix (Redi-earth Plug and Seedling Mix, Scotts-Sierra, Marysville, OH) supplemented with 20:20:20 fertilizer (Peters Professional, Scotts-Sierra) and imbibed overnight at room temperature. Seeds were placed in 4°C for four nights and then transferred to growth chambers to ensure uniform timing of germination. The plants were grown in growth chambers (Environmental Growth Chambers, Chagrin Falls, OH) at 23°C and 60% RH under a 12-h light/dark cycle at 200 μmol m−2s−1. Seedlings germinated within 2 d and were thinned to a single plant per cell pack. Plants were treated with O3 at 15 d post germination, when the fifth leaf, as counted by order of emergence from the meristem (cotyledons were not counted), was 3 to 4 d old. Half of the plants were exposed to 0.15 μL L−1O3 for 6 h d−1 and the other half remained nontreated in another growth chamber. O3 was generated by passing oxygen through an ozonator (OREC V1-0, Ozone Research and Equipment, Phoenix), and O3 concentrations in the growth chamber were monitored continuously with a UV photometric O3analyzer (model 49, Thermo Environmental Instruments, Franklin, MA). In experiments 1 and 2, plants were exposed to O3for 8 and 14 consecutive d, respectively. GUS activity was analyzed every 2 d in the fifth and sixth leaves harvested from four replicate SAG12-GUS and SAG13-GUS transgenic plants per treatment. For staining of GUS activity, the fifth and sixth leaves were collected from three replicate SAG13-GUS transgenic plants per treatment per sampling time. Leaves for GUS staining were harvested on d 3, 6, and 8 in experiment 1, and on d 4, 8, 12, and 14 in experiment 2. For northern analysis, three replicate samples of wild-type plants were collected per treatment every 2 d; each sample consisted of the fifth and sixth leaves pooled from six plants. In addition, one wild-type rosette was collected at each sampling time per treatment in experiment 2. In a third experiment, wild-type plants were exposed to O3 for 10 consecutive d. Two replicate samples of the fifth and sixth leaves pooled from six plants were collected per treatment at the end of the 6-h exposure and 18 h later, on d 6, 8, and 10 of the O3 exposure. The samples were analyzed for SAG transcript levels. GUS Activity Assays For fluorometric quantification of GUS activity, samples were ground in microcentrifuge tubes under liquid nitrogen. Leaf tissue was lysed in 150 to 200 μL of extraction buffer (50 mmsodium phosphate buffer, pH 7.0, 10 mm EDTA, 0.1% [v/v] Triton X-100, 0.1% [w/v]N-lauroylsarcosine, and 10 mmβ-mercaptoethanol) and stored at −80°C for later analysis (Jefferson et al., 1987). Following centrifugation of the crude extract, 50 μL was incubated at 37°C in 500 μL of assay buffer (2 mm 4-methylumbelliferyl β-d-glucuronide in extraction buffer). At 1-h intervals, 100-μL aliquots were removed and the reaction was stopped with 900 μL of 0.2 mNa2CO3. Fluorescence of the methyl umbelliferone product was quantified with a fluorometer (CytoFluor II multi-well plate reader, PE Biosystems). Protein concentrations were measured with the protein-dye-binding assay (Bradford, 1976) using Coomassie Plus protein assay reagent (Pierce) with BSA as a standard. For staining of GUS activity, leaves were vacuum infiltrated with 50 mm sodium phosphate buffer, pH 7.0, 1 mm EDTA, 0.01% (v/v) Triton X-100, and 1 mm5-bromo-4-chloro-3-indolyl β-d-glucuronide (Gold BioTechnology, St. Louis) (Jefferson et al., 1987; Thoma et al., 1996). Leaves were incubated at 37°C until blue staining became evident, 72 h after infiltration. Following staining, leaves were cleared of chlorophyll with 70% (v/v) ethanol. RNA Extraction and Analysis Northern analysis was conducted with the probes listed in TableI. Leaf tissue was ground under liquid nitrogen and total RNA was extracted from 100 mg of tissue (RNeasy, Qiagen, Chatsworth, CA). Total RNA was fractionated in a 1% (w/v) agarose-formaldehyde gel, transferred to a membrane (Hybond-N, Amersham), and fixed to the membrane by baking for 2 h at 80°C. The membranes were prehybridized in 0.5 m sodium phosphate buffer and 7% (w/v) SDS at 65°C for 1 h (Church and Gilbert, 1984). Probes were random-primed labeled with [α-32P]dCTP and unincorporated nucleotides were removed with spin columns (Quick Spin, Boehringer Mannheim). The membranes were hybridized overnight at 65°C. Following hybridization, the membranes were washed at 65°C twice in 40 mm sodium phosphate buffer, 5% SDS, and 1 mm EDTA for 20 min, and twice in 40 mm sodium phosphate buffer, 1% SDS, and 1 mm EDTA for 20 min. Membranes were exposed to film (X-Omat, Kodak) at −80°C with two intensifying screens. Membranes were stripped with boiling 0.1% SDS for rehybridizing with other probes. The final hybridization on each membrane was performed with cDNA for pea rRNA as a loading check (Jorgenson et al., 1982). Table I. SAGs used in the study of O3-induced accelerated leaf senescence Gene . Identity/Similarity . Reference . Time of Induction . d of O3 exposure SAG12 Cys protease Lohman et al. (1994) NI-a Gan and Amasino (1997) SAG13 Short-chain alcohol dehydrogenase Lohman et al. (1994) 2 –4 Weaver et al. (1997) BCB (SAG14) Blue copper-binding protein (membrane) Van Gysel et al. (1993) 2 –4 Lohman et al. (1994) Weaver et al. (1997) ERD1 (SAG15) ClpC-like gene (chloroplast) Kiyosue et al. (1993) 2 –4 Lohman et al. (1994) Weaver et al. (1998) MT1 (SAG17) Metallothionein Zhou and Goldsbrough (1994) NI Lohman et al. (1994) Weaver et al. (1997) SAG18 Novel gene Weaver et al. (1998) 4 –6 SAG19 Unidentified L.M. Weaver and R.M. Amasino (personal communication) NI SAG20 Novel gene Weaver et al. (1998) 4 –6 SAG21 Late embryogenesis-abundant gene Weaver et al. (1998) 2 –4 CCH Copper chaperone Himelblau et al. (1998) 6 –8 Atgsr2 Glutamine synthetase (cytosol) Peterman and Goodman (1991) NI Bernhard and Matile (1994) ACS6 ACC synthase Vahala et al. (1998) 4 –6 Arteca and Arteca (1999) Gene . Identity/Similarity . Reference . Time of Induction . d of O3 exposure SAG12 Cys protease Lohman et al. (1994) NI-a Gan and Amasino (1997) SAG13 Short-chain alcohol dehydrogenase Lohman et al. (1994) 2 –4 Weaver et al. (1997) BCB (SAG14) Blue copper-binding protein (membrane) Van Gysel et al. (1993) 2 –4 Lohman et al. (1994) Weaver et al. (1997) ERD1 (SAG15) ClpC-like gene (chloroplast) Kiyosue et al. (1993) 2 –4 Lohman et al. (1994) Weaver et al. (1998) MT1 (SAG17) Metallothionein Zhou and Goldsbrough (1994) NI Lohman et al. (1994) Weaver et al. (1997) SAG18 Novel gene Weaver et al. (1998) 4 –6 SAG19 Unidentified L.M. Weaver and R.M. Amasino (personal communication) NI SAG20 Novel gene Weaver et al. (1998) 4 –6 SAG21 Late embryogenesis-abundant gene Weaver et al. (1998) 2 –4 CCH Copper chaperone Himelblau et al. (1998) 6 –8 Atgsr2 Glutamine synthetase (cytosol) Peterman and Goodman (1991) NI Bernhard and Matile (1994) ACS6 ACC synthase Vahala et al. (1998) 4 –6 Arteca and Arteca (1999) Selected references include information on clone identification and expression patterns. F0-a NI, Not induced by 14-d O3 exposure. Open in new tab Table I. SAGs used in the study of O3-induced accelerated leaf senescence Gene . Identity/Similarity . Reference . Time of Induction . d of O3 exposure SAG12 Cys protease Lohman et al. (1994) NI-a Gan and Amasino (1997) SAG13 Short-chain alcohol dehydrogenase Lohman et al. (1994) 2 –4 Weaver et al. (1997) BCB (SAG14) Blue copper-binding protein (membrane) Van Gysel et al. (1993) 2 –4 Lohman et al. (1994) Weaver et al. (1997) ERD1 (SAG15) ClpC-like gene (chloroplast) Kiyosue et al. (1993) 2 –4 Lohman et al. (1994) Weaver et al. (1998) MT1 (SAG17) Metallothionein Zhou and Goldsbrough (1994) NI Lohman et al. (1994) Weaver et al. (1997) SAG18 Novel gene Weaver et al. (1998) 4 –6 SAG19 Unidentified L.M. Weaver and R.M. Amasino (personal communication) NI SAG20 Novel gene Weaver et al. (1998) 4 –6 SAG21 Late embryogenesis-abundant gene Weaver et al. (1998) 2 –4 CCH Copper chaperone Himelblau et al. (1998) 6 –8 Atgsr2 Glutamine synthetase (cytosol) Peterman and Goodman (1991) NI Bernhard and Matile (1994) ACS6 ACC synthase Vahala et al. (1998) 4 –6 Arteca and Arteca (1999) Gene . Identity/Similarity . Reference . Time of Induction . d of O3 exposure SAG12 Cys protease Lohman et al. (1994) NI-a Gan and Amasino (1997) SAG13 Short-chain alcohol dehydrogenase Lohman et al. (1994) 2 –4 Weaver et al. (1997) BCB (SAG14) Blue copper-binding protein (membrane) Van Gysel et al. (1993) 2 –4 Lohman et al. (1994) Weaver et al. (1997) ERD1 (SAG15) ClpC-like gene (chloroplast) Kiyosue et al. (1993) 2 –4 Lohman et al. (1994) Weaver et al. (1998) MT1 (SAG17) Metallothionein Zhou and Goldsbrough (1994) NI Lohman et al. (1994) Weaver et al. (1997) SAG18 Novel gene Weaver et al. (1998) 4 –6 SAG19 Unidentified L.M. Weaver and R.M. Amasino (personal communication) NI SAG20 Novel gene Weaver et al. (1998) 4 –6 SAG21 Late embryogenesis-abundant gene Weaver et al. (1998) 2 –4 CCH Copper chaperone Himelblau et al. (1998) 6 –8 Atgsr2 Glutamine synthetase (cytosol) Peterman and Goodman (1991) NI Bernhard and Matile (1994) ACS6 ACC synthase Vahala et al. (1998) 4 –6 Arteca and Arteca (1999) Selected references include information on clone identification and expression patterns. F0-a NI, Not induced by 14-d O3 exposure. Open in new tab RESULTS Arabidopsis plants exhibited downward leaf rolling after 4 d of treatment with 0.15 μL L−1O3. O3 treatment reduced rosette leaf growth and accelerated the yellowing of older leaves. The fifth leaf began to show signs of senescence after 10 d of O3 exposure, whereas control leaves did not begin to show signs of senescence until d 14, the last day of the experiment. In an independent experiment, chlorophyll levels per unit area declined more rapidly in O3-treated leaves (data not shown). These changes in growth and development occurred without any visible signs of hypersensitive-response-like necrosis. Effects of O3 Exposure on SAG12 andSAG13 Expression In experiment 1, the O3 exposure was 8 d in duration, and in experiment 2 the exposure was for 14 d. As the results in experiment 1 were supported in experiment 2, only the more extensive data of the latter experiment are presented here.SAG12 promoter-driven GUS activity was not detected in control or O3-treated plants on any sampling day throughout the 14 d of the experiment (data not shown), while O3 exposure did accelerate the onset ofSAG13 promoter-driven GUS activity (Fig.1). O3-induced,SAG13 promoter-driven GUS activity was first detected on d 2, whereas GUS activity was not detected until d 6 in control leaves. GUS activity gradually increased in O3-treated and control leaves through the remainder of the experiment.SAG13 promoter-driven GUS activity in O3-treated leaves always exceeded the level found in control leaves, except on d 14, when the difference between treatments was no longer detected (Fig. 1). SAG13promoter-driven GUS activity appeared after 2 d in O3-treated leaves, while yellowing did not occur on the fifth leaf until d 10. No SAG12 or SAG13promoter-driven GUS activity was detected in treated or control nontransformed plants. Fig. 1. Open in new tabDownload slide SAG13 promoter-driven GUS activity was induced by O3 treatment. Fifteen-day-old Arabidopsis ecotype Landsberg erecta plants transformed with theSAG13 promoter-GUS fusion were exposed to 0.15 μL L−1 O3 for 6 h d−1for 14 d. The fifth and sixth leaves were harvested from a single plant and GUS activity was measured by fluorometric quantification of 4-methyl umbelliferone (MU). Black bars, O3-Treated leaves; white bars, nontreated leaves. Each bar represents the mean of four samples ± se, except control bars on d 2 and 10, where the mean of three samples was taken. No GUS activity was detected in nontransformed plants (data not shown). Fig. 1. Open in new tabDownload slide SAG13 promoter-driven GUS activity was induced by O3 treatment. Fifteen-day-old Arabidopsis ecotype Landsberg erecta plants transformed with theSAG13 promoter-GUS fusion were exposed to 0.15 μL L−1 O3 for 6 h d−1for 14 d. The fifth and sixth leaves were harvested from a single plant and GUS activity was measured by fluorometric quantification of 4-methyl umbelliferone (MU). Black bars, O3-Treated leaves; white bars, nontreated leaves. Each bar represents the mean of four samples ± se, except control bars on d 2 and 10, where the mean of three samples was taken. No GUS activity was detected in nontransformed plants (data not shown). The localization of SAG13 expression was determined by staining for GUS activity (Fig. 2). The staining pattern was altered spatially and temporally by O3 treatment. GUS staining was diffusely distributed in the interior of O3-treated leaves on d 4, while no staining could be detected in control leaves. By d 8 of the experiment, intense GUS staining was present at the leaf margin and interior of O3-treated leaves. In control leaves, staining was localized to discrete areas along the margins, with some faint and variable staining at the leaf tip. Following d 12 and 14, O3-treated leaves showed intense blue staining throughout the entire leaf, and control leaves began to show stronger staining in the leaf interior as senescence progressed from the margins inward. Fig. 2. Open in new tabDownload slide Photographs showing O3-induced GUS staining in the fifth leaf of transgenic SAG13-GUS plants. Fifteen-day-old SAG13-GUS plants were exposed to 0.15 μL L−1 O3 for 6 h d−1 for 14 d. Leaves were vacuum infiltrated with 1 mm5-bromo-4-chloro-3-indolyl β-d-glucuronide, incubated at 37°C for 72 h, and cleared of chlorophyll with 70% ethanol. Nontreated leaves are shown on the left and O3-treated leaves on the right from samples harvested 4, 8, 12, and 14 d after exposure in A through D, respectively. A similar pattern of expression was found in the sixth leaf (data not shown). The leaves shown are representative of three leaves per treatment per day. Fig. 2. Open in new tabDownload slide Photographs showing O3-induced GUS staining in the fifth leaf of transgenic SAG13-GUS plants. Fifteen-day-old SAG13-GUS plants were exposed to 0.15 μL L−1 O3 for 6 h d−1 for 14 d. Leaves were vacuum infiltrated with 1 mm5-bromo-4-chloro-3-indolyl β-d-glucuronide, incubated at 37°C for 72 h, and cleared of chlorophyll with 70% ethanol. Nontreated leaves are shown on the left and O3-treated leaves on the right from samples harvested 4, 8, 12, and 14 d after exposure in A through D, respectively. A similar pattern of expression was found in the sixth leaf (data not shown). The leaves shown are representative of three leaves per treatment per day. The effect of O3 exposure on SAG12 andSAG13 expression was also determined by northern analysis. Increased abundance of the SAG13 transcript was detected after 2 to 4 d of O3 exposure (Fig.3), whereas the SAG12transcript remained undetectable in O3-treated and nontreated leaves and rosettes on all sampling days (Fig.4). These results support the GUS activity data obtained from SAG12-GUS and SAG13-GUS transgenic leaves (Fig. 1). SAG13 transcript levels gradually increased in O3-treated leaves five and six at later time points and did not appear in control leaves until d 10 to 12 (Fig. 3A).SAG13 transcript levels in entire rosettes did not show this gradual increase in abundance, yet levels did remain elevated in O3-treated rosettes compared with nontreated rosettes (Fig. 3B). The SAG13 transcript was always more abundant in O3-treated leaves than in control leaves (Fig. 3). In contrast, SAG13 promoter-driven GUS activity was similar in O3-treated and control leaves on d 14 (Fig. 1). This discrepancy may be due to the long half-life of GUS, which is approximately 50 h in living mesophyll protoplasts (Jefferson et al., 1987). Fig. 3. Open in new tabDownload slide Induction of senescence-related transcripts in O3-treated Arabidopsis plants. Fifteen-day-old plants were exposed to 0.15 μL L−1 O3 for 6 h d−1 or remained nontreated. Total RNA was extracted and 3 μg of RNA was separated on 1% formaldehyde-agarose gels, transferred to membranes, and hybridized with the radiolabeled probes indicated. A, Each lane contains RNA extracted from the fifth and sixth leaves pooled from six plants. The samples shown are one representative replicate from a total of three. B, Each lane contains RNA extracted from one rosette and only one replicate was analyzed. C, Control, nontreated plants; O3, O3-treated plants. Fig. 3. Open in new tabDownload slide Induction of senescence-related transcripts in O3-treated Arabidopsis plants. Fifteen-day-old plants were exposed to 0.15 μL L−1 O3 for 6 h d−1 or remained nontreated. Total RNA was extracted and 3 μg of RNA was separated on 1% formaldehyde-agarose gels, transferred to membranes, and hybridized with the radiolabeled probes indicated. A, Each lane contains RNA extracted from the fifth and sixth leaves pooled from six plants. The samples shown are one representative replicate from a total of three. B, Each lane contains RNA extracted from one rosette and only one replicate was analyzed. C, Control, nontreated plants; O3, O3-treated plants. Fig. 4. Open in new tabDownload slide SAG12, SAG19,MT1, and Atgsr2 transcript levels were not altered by O3 treatment. Fifteen-day-old Arabidopsis plants were exposed to 0.15 μL L−1 O3 for 6 h d−1 or remained nontreated. Samples were prepared as in Figure 3. Each lane contains RNA extracted from the fifth and sixth leaves pooled from six plants. The samples shown are one representative replicate from a total of three. Sen, RNA sample extracted from yellowing (senescent) leaves older than 30 d; C, control, nontreated plants; O3, O3-treated plants. Fig. 4. Open in new tabDownload slide SAG12, SAG19,MT1, and Atgsr2 transcript levels were not altered by O3 treatment. Fifteen-day-old Arabidopsis plants were exposed to 0.15 μL L−1 O3 for 6 h d−1 or remained nontreated. Samples were prepared as in Figure 3. Each lane contains RNA extracted from the fifth and sixth leaves pooled from six plants. The samples shown are one representative replicate from a total of three. Sen, RNA sample extracted from yellowing (senescent) leaves older than 30 d; C, control, nontreated plants; O3, O3-treated plants. Effects of O3 Exposure on PAG and SAG Expression Transcript levels for the PAGs rbcS and cabshowed a strong reduction in the fifth and sixth leaves after 6 d of O3 exposure (Fig.5A). PAG transcript levels continued to decline gradually throughout the remainder of the experiment. Only a slight decline in PAG mRNA levels was found in control leaves (Fig.5A). The O3-induced decline in PAG expression, as found in the fifth and sixth leaves, was not readily detectable in RNA samples extracted from entire rosettes (Fig. 5B). PAG transcript levels declined with age in both O3-treated and nontreated rosettes. Fig. 5. Open in new tabDownload slide PAG transcript levels declined after treatment with O3. Fifteen-day-old Arabidopsis plants were exposed to 0.15 μL L−1 O3 for 6 h d−1or remained nontreated. Samples were prepared as in Figure 3. A, Each lane contains RNA extracted from the fifth and sixth leaves pooled from six plants. The samples shown are one representative replicate from a total of three. B, Each lane contains RNA extracted from one rosette and only one replicate was analyzed. C, Control, nontreated plants; O3, O3-treated plants. Fig. 5. Open in new tabDownload slide PAG transcript levels declined after treatment with O3. Fifteen-day-old Arabidopsis plants were exposed to 0.15 μL L−1 O3 for 6 h d−1or remained nontreated. Samples were prepared as in Figure 3. A, Each lane contains RNA extracted from the fifth and sixth leaves pooled from six plants. The samples shown are one representative replicate from a total of three. B, Each lane contains RNA extracted from one rosette and only one replicate was analyzed. C, Control, nontreated plants; O3, O3-treated plants. SAG expression levels were determined in three replicate samples, and the range of days given for the time of induction represents the variability within these samples. SAG13, SAG21,BCB (blue copper-binding protein), and ERD1(early responsive to dehydration) were induced in the fifth and sixth leaves between d 2 and 4 of O3 treatment (Fig.3A), prior to any detectable decline in PAG transcript levels.SAG18, SAG20, and ACS6 (ACC synthase) were induced between d 4 and 6 and CCH (copper chaperone) was induced between d 6 and 8 of the O3 treatment in the fifth and sixth leaves (Fig. 3A). Transcripts for all of these genes continued to accumulate throughout the 14 d of exposure. Transcripts for most of these genes were detected in control leaves, but did not appear until later and levels remained below those found in O3-treated samples. The SAG21transcript was detected in the fifth and sixth control leaves on d 6;ERD1 between d 8 and 10; SAG13, SAG18, and CCH between d 10 and 12; and BCB between d 12 and 14 (Fig. 3A). SAG20 and ACS6 did not show any appreciable accumulation in the fifth and sixth control leaves during the experimental period (Fig. 3A). Transcript levels forSAG13 and BCB were greater in O3-treated rosettes compared with nontreated rosettes; however, transcript accumulation throughout the 14 d of exposure, as found for leaves five and six, was not detected in rosettes (Fig. 3B). Similar results were obtained for SAG21and ERD1 transcript levels in O3-treated rosettes and SAG18,SAG20, CCH, and ACS6 transcript levels were more abundant in O3-treated rosettes on only some of the harvest days (data not shown). Not all of the characterized SAGs were induced by O3 treatment. SAG12, SAG19,MT1 (metallothionein), and Atgsr2 (glutamine synthetase) were not induced by O3 treatment during the 8-d exposure in experiment 1 (data not shown). The O3 exposure period in experiment 2 was extended for a total of 14 d to determine if the expression of these SAGs could be induced with a longer O3 treatment.SAG12, SAG19, MT1, andAtgsr2 were not induced to any measurable degree during the 14-d exposure (Fig. 4). MT1 and Atgsr2transcripts were present in all samples and transcript levels gradually increased in abundance equally in O3-treated and control leaves. Slightly greater signals for the MT1 andAtgsr2 transcripts were found in a few of the O3-treated samples, but this response was not consistent. SAG12 transcript was not detected in any sample and SAG19 transcript remained nearly undetectable (Fig. 4). RNA was extracted from partially yellow leaves harvested from nontreated plants older than 30 d post germination and was included on membranes to demonstrate that SAG12 andSAG19 transcripts could be detected (Fig. 4). The ability of O3-treated leaves to recover from the accelerated induction of SAGs was investigated by analyzing transcript levels following removal of O3 on d 6, 8, and 10 of the exposure. The fifth and sixth leaves were harvested from plants immediately following the daily 6-h O3 treatment and from another set of plants allowed to recover from the treatment for an additional 18 h in O3-free air. Transcript levels forSAG13, BCB, ERD1, SAG20, and SAG21 were greater in leaves treated with 6 h of O3 on d 6, 8, and 10 than in nontreated leaves (Fig. 6). Transcript levels for these genes declined following 18 h in O3-free air. On d 6, SAG transcripts were nearly undetectable in control leaves, but were induced in O3-treated leaves. Following 18 h in O3-free air, transcript levels were undetectable in O3-treated samples. On d 8, transcript levels remained nearly undetectable in control leaves, but were induced in O3-treated samples and once again declined in O3-treated samples following the 18-h period. By d 10, SAG transcripts were detected in controls and O3-treated samples. The decline in transcript level 18 h after the removal of O3 was still apparent. Fig. 6. Open in new tabDownload slide SAG13, BCB,ERD1, SAG20, and SAG21transcript levels declined following a recovery period in O3-free air. Fifteen-day-old Arabidopsis plants were exposed to 0.15 μL L−1 O3 for 6 h d−1 or remained nontreated. The fifth and sixth leaves were harvested from six plants immediately following 6 h of exposure to O3 or 18 h after removal of O3. Samples were prepared as in Figure 3. The samples shown are one of two replicates. C, Control, nontreated plants; O3, O3-treated plants. Fig. 6. Open in new tabDownload slide SAG13, BCB,ERD1, SAG20, and SAG21transcript levels declined following a recovery period in O3-free air. Fifteen-day-old Arabidopsis plants were exposed to 0.15 μL L−1 O3 for 6 h d−1 or remained nontreated. The fifth and sixth leaves were harvested from six plants immediately following 6 h of exposure to O3 or 18 h after removal of O3. Samples were prepared as in Figure 3. The samples shown are one of two replicates. C, Control, nontreated plants; O3, O3-treated plants. DISCUSSION In the present study, chronic O3 treatment accelerated the normal rate of foliar senescence in Arabidopsis plants. This response occurred in the absence of the necrosis observed in response to higher O3 concentrations reported previously for other species (Pell et al., 1997). O3-induced leaf yellowing in Arabidopsis was previously observed in older leaves exposed to O3continuously for 2 d (Kubo et al., 1995). Rosette growth was reduced and downward leaf curling was evident within 4 d of O3 exposure, similar to results obtained bySharma and Davis (1994) and Rao et al. (1995). Leaf curling appeared to be an altered growth response and was not the result of dehydration, since the percent dry matter did not vary between treatments in an independent experiment (data not shown). A suite of O3-induced changes in transcript levels were observed, including reductions in levels of PAGs and increased levels of many but not all SAGs measured (Table I; Figs. 3-5). These changes were only expressed in leaves of a discrete developmental age. Hence, observations of O3-induced decline in PAG transcript levels, for example, were observed in the fifth and sixth leaves but were not detected when whole rosettes were analyzed (Fig.5). Similarities and Contrasts to Natural Senescence O3 induces many changes common to natural senescence, but at an accelerated rate: for example, loss of total protein, Rubisco, chlorophyll, and increased leaf abscission (Pell and Pearson, 1983; Reich, 1983; Held et al., 1991; Nie et al., 1993). Diminishing rbcS and cab transcript levels are indicators of declining photosynthetic activity during natural senescence; the observation in this experiment that O3 treatment reduced the level of these transcripts was supported by previous investigations (Bahl and Kahl, 1995; Conklin and Last, 1995; Glick et al., 1995). Transcript levels for two other genes, SDG1 (senescence-down-regulated gene) and SDG2, declined during the O3exposure with an expression pattern similar to rbcS andcab (data not shown). SDG1 and SDG2showed reduced transcript abundance in a differential screen of nonsenescent versus senescent leaves (Lohman et al., 1994). O3 treatment induced the early expression of many molecular markers of senescence, providing additional evidence that changes in gene expression during chronic O3treatment are similar to natural senescence. Two metal-binding proteins, CCH (copper chaperone) and BCB (blue copper-binding protein), are among the genes induced by O3. These genes were previously shown to be induced by acute O3exposure; BCB was induced within a 3-h exposure to 0.30 μL L−1 O3 (Richards et al., 1998) and CCH transcript levels increased by 30% after a 30-min exposure to 0.80 μL L−1O3 (Himelblau et al., 1998). Metal-binding proteins may play an important role in metal remobilization during senescence. O3 treatment also induced transcript accumulation of a protease regulator, ERD1; proteases are involved in protein degradation during natural senescence and may be further required for degradation of oxidized proteins during O3-induced accelerated senescence. Transcript accumulation of other genes, including SAG13,SAG18, SAG20, and SAG21, was also induced by O3 treatment; the function of these genes in senescence remains unclear. While O3induced the buildup of SAG transcripts, it is not known how this translates into accumulation of the protein products. Transcripts for SAG12, SAG19, Atgsr2, and MT1 accumulate during natural senescence, but were not induced by chronic O3 treatment. These genes may lack responsive elements able to recognize O3-induced signaling compounds. Proteases other than the Cys protease SAG12 may have been available for proteolysis and adequate quantities of Gln synthetase, Atgsr2, and metallothionein, MT1, may have been present due to high basal transcript levels. If all senescence-related genes play critical roles in cellular degradation and nutrient remobilization during natural senescence, the lack of these gene products during O3-induced accelerated senescence may reduce the efficiency of nutrient recovery. Specific and perhaps premature induction of gene expression in response to O3 is reminiscent of molecular changes in response to other stresses (Weaver et al., 1997). Genes induced during chronic O3 exposure have also been shown to be induced by darkness, dehydration, and treatment with ethylene or ABA. Dark treatment induced the O3-responsive genes,ERD1, BCB, and SAG20, dehydration induced ERD1, BCB, SAG20, andSAG21, ethylene treatment induced ERD1,BCB, SAG13, SAG20, andSAG21, and ABA treatment induced ERD1 andSAG13 (Kiyosue et al., 1993; Nakashima et al., 1997; Weaver et al., 1998). The overlap in gene expression suggests that O3 treatment, darkness, and dehydration may induce similar signaling molecules. Ethylene and ABA may play a role as signals during O3-induced accelerated leaf senescence, as discussed below. In addition to affecting the timing of induction of some SAGs, O3 also seems to influence the spatial distribution of that induction. SAG13-promoter driven GUS activity first appeared at the leaf margin in control leaves, which resembles the pattern of yellowing found in naturally senescing leaves (Weaver et al., 1998). In contrast, O3 treatment induced SAG13 expression throughout the leaves. This distribution of SAG13 expression probably coincided with regions where O3 entered through open stomata. Potential Signals of Molecular Events Elevated SAG13, SAG20, SAG21,BCB, and ERD1 transcript levels in O3-treated leaves were not sustained following the removal of O3. Daily O3exposures were required to provide a signal to maintain enhanced SAG transcript levels, suggesting that the leaves may retain some ability to recover from exposure to O3. A similar recovery was shown for rbcS and cab transcripts in Arabidopsis following a 24-h O3-free period after treatment with 0.175 μL L−1O3 for 8 h d−1 for 4 d (Conklin and Last, 1995). Since O3 treatment induced premature changes in transcript levels of genes associated with natural senescence, O3 may elicit some of the same signals involved in natural senescence. The common mechanism regulating O3-induced accelerated leaf senescence and natural leaf senescence may involve reactive oxygen species. Oxidative stress has long been associated with senescence (Thompson et al., 1987), and recently this link was shown in the late-flowering (or extended longevity) Arabidopsis mutant gigantea (gi-3), which exhibited enhanced tolerance to methyl viologen-induced oxidative stress (Kurepa et al., 1998). Following stomatal uptake of O3, internal O3concentrations rapidly drop (Laisk et al., 1989) as decomposition products, including reactive oxygen species, are formed. These reactive oxygen species can react with membrane lipids to produce more reactive oxygen intermediates. A second sustained peak of reactive oxygen species was found in the O3-sensitive tobacco cv Bel W3 following O3 exposure, and was not found in the O3-tolerant cv Bel B (Schraudner et al., 1998). An O3-responsive region in the stilbene synthase promoter has been identified (Schubert et al., 1997), and a comparison of this 150-bp region with the SAG13 promoter (S. Gan and R.M. Amasino, personal communication) did not reveal any strong sequence similarity (data not shown). The presence of O3 or reactive oxygen species responsive elements in SAGs is worthy of future investigation. Alternatively, the O3-induced changes in gene expression could have been induced through a secondary signal. Ethylene treatment induces many of the O3-responsive SAGs (Weaver et al., 1998), and plants exposed to high doses of O3 produce large quantities of ethylene (Pell et al., 1997). Ethylene has been shown to regulate the timing of leaf senescence in Arabidopsis (Grbic and Bleecker, 1995). In our experiments, ACS6, one member of the gene family encoding ACC synthase in Arabidopsis, was detected within 4 to 6 d of O3 exposure. This gene is induced by many stresses, including O3, while ACS1,ACS2, ACS4, and ACS5 are not induced by O3 treatment (Vahala et al., 1998; Arteca and Arteca, 1999). At high O3 concentrations, ethylene emission is one of the first responses and is correlated with the degree of lesion formation (Tuomainen et al., 1997). The importance of ethylene in regulating the response to low O3concentrations in the induction of accelerated leaf senescence remains to be determined. We are currently investigating the need for ethylene perception in the induction of this suite of SAGs. Other potential signaling molecules include ABA, salicylic acid, and calcium. ABA is another senescence-promoting hormone, and some of the O3-responsive SAGs are inducible by ABA treatment (Weaver et al., 1998). Salicylic acid and calcium increase during exposure to high O3 concentrations and are involved in the induction of antioxidant gene expression (Yalpani et al., 1994; Sharma and Davis, 1997; Clayton et al., 1999); however, it is not known whether they are involved in the response to chronic O3. In conclusion, chronic O3 treatment induced SAG expression while suppressing PAG expression. An initial pattern of senescence-related gene induction by O3 has emerged. Future experiments should focus on determining which genes are essential for the induction of O3-induced accelerated leaf senescence and what, if any, interdependency exists between these genes. Further investigation will determine the identity of signals required for O3-induced accelerated leaf senescence and elucidate the role of oxidative stress in the progression of natural leaf senescence. ACKNOWLEDGMENTS The authors thank Richard Amasino for the generous gift of the SAG clones and the Arabidopsis Biological Resource Center (Columbus, OH) for the Atgsr2 clone, stock no. CD3-195, donated by T.K. Peterman. We are also grateful to Ed Himelblau and Michael Weaver for helpful discussions and Nan Eckardt and Judy Sinn for critical reading of the manuscript. Abbreviations: PAG photosynthesis-associated gene SAG senescence-associated gene LITERATURE CITED 1 Arteca JM Arteca RN A multi-responsive gene encoding 1-aminocyclopropane-1-carboxylate synthase (ACS6) in mature Arabidopsis leaves. Plant Mol Biol 39 1999 209 219 Google Scholar Crossref Search ADS PubMed WorldCat 2 Bahl A Kahl G Air pollutant stress changes the steady-state transcript levels of three photosynthesis genes. Environ Pollut 88 1995 57 65 Google Scholar Crossref Search ADS PubMed WorldCat 3 Bate NJ Rothstein SJ Thompson JE Expression of nuclear and chloroplast photosynthesis-specific genes during leaf senescence. J Exp Bot 42 1991 801 811 Google Scholar Crossref Search ADS WorldCat 4 Becker W Apel K Differences in gene expression between natural and artificially induced leaf senescence. Planta 189 1993 74 79 Google Scholar Crossref Search ADS WorldCat 5 Bernhard WR Matile P Differential expression of glutamine synthetase genes during the senescence of Arabidopsis thaliana rosette leaves. Plant Sci 98 1994 7 14 Google Scholar Crossref Search ADS WorldCat 6 Blank A McKeon TA Expression of three RNase activities during natural and dark-induced senescence of wheat leaves. Plant Physiol 97 1991 1409 1413 Google Scholar Crossref Search ADS PubMed WorldCat 7 Bradford MM A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72 1976 248 254 Google Scholar Crossref Search ADS PubMed WorldCat 8 Buchanan-Wollaston V The molecular biology of leaf senescence. J Exp Bot 48 1997 181 199 Google Scholar Crossref Search ADS WorldCat 9 Chung BC Lee SY Oh SA Rhew TH Nam HG Lee CH The promoter activity of sen1, a senescence-associated gene of Arabidopsis, is repressed by sugars. J Plant Physiol 151 1997 339 345 Google Scholar Crossref Search ADS WorldCat 10 Church GM Gilbert W Genomic sequencing. Proc Natl Acad Sci USA 81 1984 1991 1995 Google Scholar Crossref Search ADS PubMed WorldCat 11 Clayton H Knight MR Knight H McAinsh MR Hetherington AM Dissection of the ozone-induced calcium signature. Plant J 17 1999 575 579 Google Scholar Crossref Search ADS PubMed WorldCat 12 Conklin PL Last RL Differential accumulation of antioxidant mRNAs in Arabidopsis thaliana exposed to ozone. Plant Physiol 109 1995 203 212 Google Scholar Crossref Search ADS PubMed WorldCat 13 Gan S Amasino RM Making sense of senescence: molecular genetic regulation and manipulation of leaf senescence. Plant Physiol 113 1997 313 319 Google Scholar Crossref Search ADS PubMed WorldCat 14 Glick RE Schlagnhaufer CD Arteca RN Pell EJ Ozone-induced ethylene emission accelerates the loss of ribulose1,5-bisphosphate carboxylase/oxygenase and nuclear-encoded mRNAs in senescing potato leaves. Plant Physiol 109 1995 891 898 Google Scholar Crossref Search ADS PubMed WorldCat 15 Grbic V Bleecker AB Ethylene regulates the timing of leaf senescence in Arabidopsis. Plant J 8 1995 595 602 Google Scholar Crossref Search ADS WorldCat 16 Held AA Mooney HA Gorham JN Acclimation to ozone stress in radish: leaf demography and photosynthesis. New Phytol 118 1991 417 423 Google Scholar Crossref Search ADS WorldCat 17 Himelblau E Mira H Lin SJ Culotta VC Peñarrubia L Amasino RM Identification of a functional homolog of the yeast copper homeostasis gene ATX1 from Arabidopsis. Plant Physiol 117 1998 1227 1234 Google Scholar Crossref Search ADS PubMed WorldCat 18 Jefferson RA Kavanagh TA Bevan MW GUS fusions: β-glucuronidase as a sensitive and versatile gene fusion marker in higher plants. EMBO J 6 1987 3901 3907 Google Scholar Crossref Search ADS PubMed WorldCat 19 Jorgenson RA Cuellar RE Thompson WF Modes and tempos in the evolution of nuclear encoded ribosomal RNA genes in legumes. Carnegie Inst Wash Year Book 81 1982 98 101 Google Scholar OpenURL Placeholder Text WorldCat 20 Kiyosue T Yamaguchi-Shinozaki K Shinozaki K Characterization of cDNA for a dehydration-inducible gene that encodes a CLP A, B-like protein in Arabidopsis thaliana L. Biochem Biophys Res Commun 196 1993 1214 1220 Google Scholar Crossref Search ADS PubMed WorldCat 21 Kubo A Saji H Tanaka K Kondo N Expression of Arabidopsis cytosolic ascorbate peroxidase gene in response to ozone or sulfur dioxide. Plant Mol Biol 29 1995 479 489 Google Scholar Crossref Search ADS PubMed WorldCat 22 Kurepa J Smalle J Van Montagu M Inzé D Oxidative stress tolerance and longevity in Arabidopsis: the late-flowering mutant gigantea is tolerant to paraquat. Plant J 14 1998 759 764 Google Scholar Crossref Search ADS PubMed WorldCat 23 Laisk A Kull O Moldau H Ozone concentration in leaf intercellular air spaces is close to zero. Plant Physiol 90 1989 1163 1167 Google Scholar Crossref Search ADS PubMed WorldCat 24 Lohman KN Gan S John MC Amasino RM Molecular analysis of natural leaf senescence in Arabidopsis thaliana. Physiol Plant 92 1994 322 328 Google Scholar Crossref Search ADS WorldCat 25 Nakashima K Kiyosue T Yamaguchi-Shinozaki K Shinozaki K A nuclear gene, erd1, encoding a chloroplast-targeted Clp protease regulatory subunit homolog is not only induced by water stress but also developmentally up-regulated during senescence in Arabidopsis thaliana. Plant J 12 1997 851 861 Google Scholar Crossref Search ADS PubMed WorldCat 26 Nie GY Tomasevic M Baker NR Effects of ozone on the photosynthetic apparatus and leaf proteins during leaf development in wheat. Plant Cell Environ 16 1993 643 651 Google Scholar Crossref Search ADS WorldCat 27 Oh SA Lee SY Chung IK Lee CH Nam HG A senescence-associated gene of Arabidopsis thaliana is distinctively regulated during natural and artificially induced leaf senescence. Plant Mol Biol 30 1996 739 754 Google Scholar Crossref Search ADS PubMed WorldCat 28 Park JH Oh SA Kim YH Woo HR Nam HG Differential expression of senescence-associated mRNAs during leaf senescence induced by different senescence-inducing factors in Arabidopsis. Plant Mol Biol 37 1998 445 454 Google Scholar Crossref Search ADS PubMed WorldCat 29 Pell EJ Pearson NS Ozone-induced reduction in quantity of ribulose-1,5-bisphosphate carboxylase in alfalfa foliage. Plant Physiol 73 1983 185 187 Google Scholar Crossref Search ADS PubMed WorldCat 30 Pell EJ Schlagnhaufer CD Arteca RN Ozone-induced oxidative stress: mechanisms of action and reaction. Physiol Plant 100 1997 264 273 Google Scholar Crossref Search ADS WorldCat 31 Peterman TK Goodman HM The glutamine synthetase gene family of Arabidopsis thaliana: light-regulation and differential expression in leaves, roots and seeds. Mol Gen Genet 230 1991 145 154 Google Scholar Crossref Search ADS PubMed WorldCat 32 Rao MV Paliyath G Ormrod DP Differential response of photosynthetic pigments, Rubisco activity and Rubisco protein of Arabidopsis thaliana exposed to UVB and ozone. Photochem Photobiol 62 1995 727 735 Google Scholar Crossref Search ADS WorldCat 33 Reich PB Effects of low concentrations of O3 on net photosynthesis, dark respiration, and chlorophyll contents in aging hybrid poplar leaves. Plant Physiol 73 1983 291 296 Google Scholar Crossref Search ADS PubMed WorldCat 34 Richards KD Schott EJ Sharma YK Davis KR Gardner RC Aluminum induces oxidative stress genes in Arabidopsis thaliana. Plant Physiol 116 1998 409 418 Google Scholar Crossref Search ADS PubMed WorldCat 35 Schraudner M Moeder W Wiese C Van Camp W Inzé D Langebartels C Sandermann H Ozone-induced oxidative burst in the ozone biomonitor plant, tobacco Bel W3. Plant J 16 1998 235 245 Google Scholar Crossref Search ADS PubMed WorldCat 36 Schubert R Fischer R Hain R Schreier PH Bahnweg G Ernst D Sandermann H An ozone-responsive region of the grapevine resveratrol synthase promoter differs from the basal pathogen-responsive sequence. Plant Mol Biol 34 1997 417 426 Google Scholar Crossref Search ADS PubMed WorldCat 37 Sharma YK Davis KR Ozone-induced expression of stress-related genes in Arabidopsis thaliana. Plant Physiol 105 1994 1089 1096 Google Scholar Crossref Search ADS PubMed WorldCat 38 Sharma YK Davis KR The effects of ozone on antioxidant responses in plants. Free Radic Biol Med 23 1997 470 488 Google Scholar Crossref Search ADS WorldCat 39 Smart CM Gene expression during leaf senescence. New Phytol 126 1994 419 448 Google Scholar Crossref Search ADS PubMed WorldCat 40 Thoma S Sullivan ML Vierstra RD Members of two gene families encoding ubiquitin-conjugating enzymes, AtUBC1–3 and AtUBC4–6, from Arabidopsis thaliana are differentially expressed. Plant Mol Biol 31 1996 493 505 Google Scholar Crossref Search ADS PubMed WorldCat 41 Thompson JE Legge RL Barber RF The role of free radicals in senescence and wounding. New Phytol 105 1987 317 344 Google Scholar Crossref Search ADS PubMed WorldCat 42 Tuomainen J Betz C Kangasjärvi J Ernst D Yin ZH Langebartels C Sandermann H Ozone induction of ethylene emission in tomato plants: regulation by differential accumulation of transcripts for the biosynthetic enzymes. Plant J 12 1997 1151 1162 Google Scholar Crossref Search ADS WorldCat 43 Vahala J Schlagnhaufer CD Pell EJ Induction of an ACC synthase cDNA by ozone in light-grown Arabidopsis thaliana leaves. Physiol Plant 103 1998 45 50 Google Scholar Crossref Search ADS WorldCat 44 Van Gysel A Van Montagu M Inze D A negatively light-regulated gene from Arabidopsis thaliana encodes a protein showing high similarity to blue copper-binding proteins. Gene 136 1993 79 85 Google Scholar Crossref Search ADS PubMed WorldCat 45 Weaver LM Gan S Quirino B Amasino RM A comparison of the expression patterns of several senescence-associated genes in response to stress and hormone treatment. Plant Mol Biol 37 1998 455 469 Google Scholar Crossref Search ADS PubMed WorldCat 46 Weaver LM, Himelblau E, Amasino RM (1997) Leaf senescence: gene expression and regulation. In JK Setlow, ed, Genetic Engineering, Vol 19. Plenum Press, New York, pp 215–234 47 Yalpani N Enyedi AJ León J Raskin I Ultraviolet light and ozone stimulate accumulation of salicylic acid, pathogenesis-related proteins and virus resistance in tobacco. Planta 193 1994 372 376 Google Scholar Crossref Search ADS WorldCat 48 Zhou J Goldsbrough PB Functional homologs of fungal metallothionein genes from Arabidopsis. Plant Cell 6 1994 875 884 Google Scholar PubMed OpenURL Placeholder Text WorldCat Author notes 1 External funding for this research was provided by the Environmental Protection Agency (grant no. U915212–01–1) and by the U.S. Department of Agriculture (grant no. 93–38420–8742). This research was also supported in part by the Pennsylvania Agricultural Experiment Station and the Environmental Resources Research Institute. It is contribution no. 2064 from the Department of Plant Pathology, The Pennsylvania State University. * Corresponding author; e-mail [email protected]; fax 814–863–7217. Copyright © 1999 American Society of Plant Physiologists This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)

Blázquez, Miguel A.; Weigel, Detlef

doi: 10.1104/pp.120.4.1025pmid: 10444085

Abstract Phytochromes and gibberellins (GAs) coordinately regulate multiple aspects of Arabidopsis development. Phytochrome B (PHYB) promotes seed germination by increasing GA biosynthesis, but inhibits hypocotyl elongation by decreasing the responsiveness to GAs. Later in the life cycle of the plant, PHYB and GAs have opposite effects on flowering. PHYB delays flowering, while GAs promote flowering, particularly under noninductive photoperiods. To learn how PHYB and GAs interact in the control of flowering, we have analyzed the effect of a phyB mutation on flowering time and on the expression of the floral meristem-identity geneLFY(LEAFY). We show that the early flowering caused by phyBcorrelated with an increase in LFY expression, which complements our previous finding that GAs are required for activation of LFY under noninductive photoperiods (M.A. Blázquez, R. Green, O. Nilsson, M.R. Sussman, D. Weigel [1998] Plant Cell 10: 791−800). Since phyB did not change the GA responsiveness of the LFY promoter and suppressed the lack of flowering of severe GA-deficient mutants under short days, we propose that PHYB modulates flowering time at least partially through a GA-independent pathway. Interestingly, the effects of PHYB on flowering do not seem to be mediated by transcriptional up-regulation of genes such as CO (CONSTANS) andFT (Flowering locusT), which are known to mediate the effects of the photoperiod-dependent floral-induction pathway. The control of plant development by light is exerted through the activity of photoreceptors. Among these, phytochromes mediate the responses to red and far-red light (Fankhauser and Chory, 1997). In Arabidopsis, phytochromes are encoded by five different genes,PHYA through PHYE (Sharrock and Quail, 1989;Quail et al., 1995). While each phytochrome seems to have specific roles, there is also considerable overlap in the function of individual phytochromes (Reed et al., 1994; Aukerman et al., 1997; Devlin et al., 1998). For example, PHYA seems to have a primary role in germination and in the regulation of seedling morphogenesis by far-red light (Nagatani et al., 1993; Parks and Quail, 1993; Whitelam et al., 1993). Later in development, PHYA is involved in sensing photoperiod, which is reflected in the insensitivity of phyA mutants to night breaks (Reed et al., 1994). PHYB, on the other hand, is an essential component of the shade-avoidance mechanism, and modulates the expression of genes in response to red light. Null mutations inPHYB cause increased elongation of hypocotyls, leaf petioles, and stems, as well as decreased chlorophyll accumulation and earlier flowering under both long and short photoperiods (Reed et al., 1993). Phytochromes interact with plant hormones of the GA class to regulate certain aspects of plant development (Chory and Li, 1997). The phenotype of mutants defective in GA biosynthesis has confirmed that GAs regulate processes such as seed germination, cell expansion, and flowering, all of which are also under the control of phytochromes. For example, the Arabidopsis spy (spindly) mutant has a slender stature, is pale green, and flowers early, thus resemblingphyB mutants, or wild-type plants treated with GA3 (Jacobsen and Olszewski, 1993). An opposite phenotype is seen in GA-deficient mutants of Arabidopsis that are dark-green dwarves that flower late (Koornneef and van der Veen, 1980). This phenotype is particularly severe in thega1-3 mutant, which is blocked in a very early step of GA biosynthesis, and never flowers under short days (Wilson et al., 1992). Reduced expression of LFY (LEAFY), a floral meristem-identity gene, seems to be a main cause of the flowering defect in ga1-3 mutants, since constitutive expression of LFY from a transgene is sufficient to restore flowering of ga1-3 mutants under short days (Blázquez et al., 1998). There are at least two possible mechanisms by which phytochromes and GAs may interact. The finding of elevated GA levels in thephyB-deficient mutants ein of Brassica rapa and ma3R of sorghum has suggested that PHYB regulates GA biosynthesis in certain species (Rood et al., 1990; Beall et al., 1991). However, the relationship between GA biosynthesis and PHYB activity in sorghum is complex, as GA biosynthesis follows a circadian regulation, and thema3R mutant shows a phase shift in GA accumulation (Lee et al., 1998). An interaction between phytochromes and GA biosynthesis is also supported by the finding that GA-biosynthetic genes in Arabidopsis and spinach are regulated by light. In both species, GA 20-oxidase mRNA levels are higher under long than under short photoperiods (Xu et al., 1995; Wu et al., 1996). It has been suggested that GA 20-oxidase activity is limiting for stem elongation in long days, and that long photoperiods raise the concentration of active GAs above a certain threshold (Talón et al., 1991). Similarly, GA 3β-hydroxylation is promoted by pulses of far-red light in cowpea (Martı́nez-Garcı́a and Garcı́a-Martı́nez, 1992) and by red light in Arabidopsis seeds (Yamaguchi et al., 1998), indicating that the synthesis of active GA species is under phytochrome control. In different situations, however, phytochromes have been shown to affect the responsiveness to GAs, rather than their biosynthesis. Putative phyB mutants of pea and cucumber, lv and lh, as well as thephyB mutant of Arabidopsis have wild-type levels of endogenous GAs but show enhanced hypocotyl elongation in response to exogenous GAs (Weller et al., 1994; López-Juez et al., 1995; Reed et al., 1996). At least in Arabidopsis, PHYB thus appears to control GA-dependent hypocotyl elongation by modulating GA sensitivity as opposed to regulating GA biosynthesis. In this study, we have addressed the question of whether PHYB and GAs interact in the regulation of flowering in a way similar to what is observed for other responses, such as hypocotyl elongation, in Arabidopsis. Using promoter activity of the floral regulatorLFY as an indicator, we show that PHYB and GAs regulateLFY expression independently. This finding is corroborated by the observation that the loss of PHYB function allowed GA-deficient mutants to flower under short days. MATERIALS AND METHODS Plant Material LFY::GUS lines (DW150-304 and 304G1) in the Landsberg erecta background of Arabidopsis have been previously described (Blázquez et al., 1997, 1998; Hempel et al., 1997). Lines 304B5 (phyB-5 LFY::GUS) and 304G1B5 (ga1-3 phyB-5 LFY::GUS) were constructed by crossing line 304G1 (ga1-3 LFY::GUS) to plants carrying thephyB-5 mutations, a null allele in the Landsbergerecta background (Reed et al., 1993). Transgenic plants homozygous for ga1-3 were initially identified by their short stature and dark-green color, and were then confirmed by PCR (Silverstone et al., 1997). The presence of thephyB-5 allele was also monitored by PCR using a dCAPS marker (Neff et al., 1998). Lines homozygous for the transgene were identified by testing F3 progeny. Growth Conditions For experiments on soil, seeds were stratified for 2 to 3 d at 4°C before sowing. Plants were grown at 23°C in long (16 h of light and 8 h of dark) or short days (9 h of light and 15 h of dark) under a mixture of 3:1 cool-white and Gro-Lux fluorescent lights (Osram Sylvania, Danvers, MA). The spectral quality of the light received by the plants under these conditions was determined with a portable spectroradiometer (model LI-1800, LI-COR) and is shown in Figure 1. ga1-3mutants required exogenous GAs to germinate (Koornneef and van der Veen, 1980) and were incubated with 50 μmGA3 (Sigma) during stratification. Seeds were rinsed thoroughly with water before sowing. Vegetatively growing plants were sprayed twice weekly with a solution of 100 μm GA3 and 0.02% (v/v) Tween 20 (Bio-Rad). Fig. 1. Open in new tabDownload slide Light spectrum inside the growth chambers used in this work. The solid black line represents the spectrum provided by a 3:1 mixture of cool-white to fluorescent light, while the dotted line represents the spectrum provided by cool-white light alone. Fig. 1. Open in new tabDownload slide Light spectrum inside the growth chambers used in this work. The solid black line represents the spectrum provided by a 3:1 mixture of cool-white to fluorescent light, while the dotted line represents the spectrum provided by cool-white light alone. The dose-response experiments with GA3 or paclobutrazol (Zeneca Ag Products, Wilmington, DE) were performed with seedlings growing on MS plates (Murashige and Skoog, 1962) without Suc under the light conditions described above. The fluence rate was around 66 μmol m−2 s−1. In experiments with soil-grown plants, paclobutrazol was applied by watering with a 37 mg/L solution. Hypocotyl and GUS Activity Measurements Hypocotyls were measured using a digitized image of 12 to 18 seedlings that had been placed between transparent acetate sheets (Neff and Chory, 1998). The image was analyzed with the public domain NIH Image program (developed at the United States National Institutes of Health and available on the Internet athttp://rsb.info.nih.gov/nih-image). For quantitative measurements of GUS activity using 4-methylumbelliferyl-β-d-glucopyranoside as a substrate, samples of plants grown on soil or MS plates were collected and treated as previously described (Blázquez et al., 1997). RNA Extraction and Analysis Total RNA was extracted with TRIzol reagent as indicated by the manufacturer (GIBCO-BRL). RT-PCR was conducted on 1 μg of total RNA. cDNA synthesis was performed with a reverse transcription kit (Promega). A fragment of the CO (CONSTANS) gene was amplified using oligonucleotides 5′-ACG CCA TCA GCG AGT TCC-3′ and 5′-AAA TGT ATG CGT TAT GGT TAA TGG-3′ as primers (P. Reeves and G. Coupland, personal communication). FT was amplified using 5′-ACT ATA TAG GCA TCA TCA CCG TTC GTT ACT CG-3′ and 5′-ACA ACT GGA ACA ACC TTT GGC AAT G-3′ (J.H. Ahn and D. Weigel, unpublished data). As a control we used oligos 5′-GAT CTT TGC CGG AAA ACA ATT GGA GGA TGG T-3′ and 5′-CGA CTT GTC ATT AGA AAG AAA GAG ATA ACA GG-3′, which amplify two polyubiquitin gene fragments in the Landsbergerecta ecotype (Callis et al., 1995). The amplified fragments were separated on an agarose gel, blotted onto a membrane, and hybridized with radiolabeled CO, FT, andUBQ10 probes. Signal intensities were determined with a phosphor imager (Molecular Dynamics), and the values in the exponential range of amplification were compared. RESULTS Enhanced LFY Up-Regulation in phyB Mutants Mutations at the PHYB locus cause early flowering, especially under short photoperiods (Goto et al., 1991; Reed et al., 1993) (Table I). Since several other mutations that affect flowering time also change the expression level of LFY (Blázquez et al., 1998; Nilsson et al., 1998), we investigated the effect of the phyB-5 null mutation on the activity of the LFY promoter using a fusion of the LFY promoter to the GUS reporter, which faithfully reflects endogenous LFY expression (Blázquez et al., 1997). Plants homozygous for thephyB-5 mutation and aLFY::GUS transgene (304B5) and isogenicPHYB+ plants (DW150-304) were grown on soil under long and short photoperiods, and GUS activity in the apices was determined at different ages during the vegetative phase. As shown in Figure 2, thephyB-5 mutation caused an increase in the expression of LFY::GUS under both photoperiods, although this effect was more pronounced under short days. This result paralleled the acceleration of flowering observed in these plants (Table I), and indicates that PHYB represses LFY expression. Table I. Flowering time of the ga1-3 and phyB-5 lines used in this study Line . Genotype . Long Days . Short Days . Short Days + GA3 . RL . CL . TL . RL . CL . TL . RL . CL . TL . 150-304 Wild type 8.8  ± 0.2 3.1  ± 0.4 11.9  ± 0.3 25.1  ± 0.9 7.5  ± 0.3 32.6  ± 0.8 16.2  ± 0.9 10.0  ± 0.5 26.2  ± 0.8 304B5 phyB-5 7.4  ± 0.5 4.9  ± 0.7 12.1  ± 0.8 9.1  ± 0.6 4.6  ± 0.4 13.7  ± 0.7 7.0  ± 0.3 5.6  ± 0.4 12.6  ± 0.8 304G1 ga1-3 –-a –-a 15.3  ± 0.7 >58 –-b >58 18.0  ± 0.9 8.3  ± 0.8 26.3  ± 0.9 304G1B5 ga1-3 –-a –-a 17  ± 1 –-a –-a 26  ± 1 8.1  ± 0.6 5.2  ± 0.3 13.3  ± 0.7 phyB-5 Line . Genotype . Long Days . Short Days . Short Days + GA3 . RL . CL . TL . RL . CL . TL . RL . CL . TL . 150-304 Wild type 8.8  ± 0.2 3.1  ± 0.4 11.9  ± 0.3 25.1  ± 0.9 7.5  ± 0.3 32.6  ± 0.8 16.2  ± 0.9 10.0  ± 0.5 26.2  ± 0.8 304B5 phyB-5 7.4  ± 0.5 4.9  ± 0.7 12.1  ± 0.8 9.1  ± 0.6 4.6  ± 0.4 13.7  ± 0.7 7.0  ± 0.3 5.6  ± 0.4 12.6  ± 0.8 304G1 ga1-3 –-a –-a 15.3  ± 0.7 >58 –-b >58 18.0  ± 0.9 8.3  ± 0.8 26.3  ± 0.9 304G1B5 ga1-3 –-a –-a 17  ± 1 –-a –-a 26  ± 1 8.1  ± 0.6 5.2  ± 0.3 13.3  ± 0.7 phyB-5 All lines are in the Landsberg erecta background and homozygous for a LFY::GUS transgene. RL, Rosette leaves; CL, cauline leaves; TL, total number of leaves. Values are the means ± 2 se (i.e. with a 95% confidence interval).n ≥ 12 plants. F0-a – indicates that these plants did not bolt in long days and that rosette and cauline leaves could not be distinguished. F0-b – indicates that ga1-3plants did not flower in short days and that rosette and cauline leaves could not be distinguished. Open in new tab Table I. Flowering time of the ga1-3 and phyB-5 lines used in this study Line . Genotype . Long Days . Short Days . Short Days + GA3 . RL . CL . TL . RL . CL . TL . RL . CL . TL . 150-304 Wild type 8.8  ± 0.2 3.1  ± 0.4 11.9  ± 0.3 25.1  ± 0.9 7.5  ± 0.3 32.6  ± 0.8 16.2  ± 0.9 10.0  ± 0.5 26.2  ± 0.8 304B5 phyB-5 7.4  ± 0.5 4.9  ± 0.7 12.1  ± 0.8 9.1  ± 0.6 4.6  ± 0.4 13.7  ± 0.7 7.0  ± 0.3 5.6  ± 0.4 12.6  ± 0.8 304G1 ga1-3 –-a –-a 15.3  ± 0.7 >58 –-b >58 18.0  ± 0.9 8.3  ± 0.8 26.3  ± 0.9 304G1B5 ga1-3 –-a –-a 17  ± 1 –-a –-a 26  ± 1 8.1  ± 0.6 5.2  ± 0.3 13.3  ± 0.7 phyB-5 Line . Genotype . Long Days . Short Days . Short Days + GA3 . RL . CL . TL . RL . CL . TL . RL . CL . TL . 150-304 Wild type 8.8  ± 0.2 3.1  ± 0.4 11.9  ± 0.3 25.1  ± 0.9 7.5  ± 0.3 32.6  ± 0.8 16.2  ± 0.9 10.0  ± 0.5 26.2  ± 0.8 304B5 phyB-5 7.4  ± 0.5 4.9  ± 0.7 12.1  ± 0.8 9.1  ± 0.6 4.6  ± 0.4 13.7  ± 0.7 7.0  ± 0.3 5.6  ± 0.4 12.6  ± 0.8 304G1 ga1-3 –-a –-a 15.3  ± 0.7 >58 –-b >58 18.0  ± 0.9 8.3  ± 0.8 26.3  ± 0.9 304G1B5 ga1-3 –-a –-a 17  ± 1 –-a –-a 26  ± 1 8.1  ± 0.6 5.2  ± 0.3 13.3  ± 0.7 phyB-5 All lines are in the Landsberg erecta background and homozygous for a LFY::GUS transgene. RL, Rosette leaves; CL, cauline leaves; TL, total number of leaves. Values are the means ± 2 se (i.e. with a 95% confidence interval).n ≥ 12 plants. F0-a – indicates that these plants did not bolt in long days and that rosette and cauline leaves could not be distinguished. F0-b – indicates that ga1-3plants did not flower in short days and that rosette and cauline leaves could not be distinguished. Open in new tab Fig. 2. Open in new tabDownload slide LFY::GUS expression during vegetative growth of phyB-5 mutants. Plants homozygous for the LFY::GUS transgene either in a PHYB+ (DW150-304; white symbols) orphyB-5 background (307B5; black symbols) were grown in long days (squares) or short days (circles) until flower buds were visible to the naked eye. Values are expressed as means ± 2 se (n = 12). Time represents days after sowing. Error bars that are not visible are contained within the symbol. MUG, 4-Methylumbelliferyl-β-d-glucopyranoside. Fig. 2. Open in new tabDownload slide LFY::GUS expression during vegetative growth of phyB-5 mutants. Plants homozygous for the LFY::GUS transgene either in a PHYB+ (DW150-304; white symbols) orphyB-5 background (307B5; black symbols) were grown in long days (squares) or short days (circles) until flower buds were visible to the naked eye. Values are expressed as means ± 2 se (n = 12). Time represents days after sowing. Error bars that are not visible are contained within the symbol. MUG, 4-Methylumbelliferyl-β-d-glucopyranoside. Application of GA3 from germination on accelerated flowering of wild-type plants, but not ofphyB-5 mutants under short days (Table I). Although the number of rosette leaves was lower in GA3-treated phyB-5 plants than in untreated plants, the total number of leaves, including cauline leaves, was not significantly different between these populations (Table I). Consistent with the absence of an effect on flowering time, we observed no further increase in LFY::GUSexpression when phyB-5 plants were treated with GA3 (results not shown). Suppression of the Flowering Defect of ga1 Mutants by phyB The increased expression level of LFY::GUS inphyB-5 mutants resembles what is seen upon application of GAs to wild-type plants (Blázquez et al., 1997) or in mutants with enhanced GA-signaling, such as spy(Blázquez et al., 1998). Since several other effects of PHYB signaling, such as hypocotyl elongation and seed germination, appear to be mediated by GAs, we wanted to know whether the early flowering ofphyB mutants depended on the activity of GAs. Therefore, we constructed ga1-3 phyB-5 mutant plants and cultivated them under long and short photoperiods. Under long days, the double mutants were similar in size to ga1-3plants, and much smaller than phyB-5 or wild-type plants, as previously described (Peng and Harberd, 1997). While flowering time of ga1-3 phyB-5 mutants under long days was not different from that ofga1-3 plants, the phyB-5mutation suppressed the flowering defect of ga1-3mutants under short days (Table I; Fig.3). When plants were grown in a mixture of fluorescent and cool-white light, suppression was observed in over 90% of the double mutants after 7 weeks. Among these plants, the number of leaves produced before flowering did not deviate much from the mean, and the value was intermediate between that of wild-type andphyB-5 plants (Table I). Fig. 3. Open in new tabDownload slide Suppression of thega1-3 flowering defect in short days byphyB-5. Representative plants were photographed 50 d after sowing. Arrowhead indicates flowers. Fig. 3. Open in new tabDownload slide Suppression of thega1-3 flowering defect in short days byphyB-5. Representative plants were photographed 50 d after sowing. Arrowhead indicates flowers. Decreased LFY expression appears to be a main cause of the late-flowering phenotype of ga1 mutants under long days, and of the inability of ga1-3 mutants to flower at all under short days (Blázquez et al., 1998). To determine whether phyB suppressed the ga1 mutant flowering defect by restoring more normal levels of LFY promoter activity, we examined LFY::GUS expression inga1-3 phyB-5 double mutants. As shown in Figure 4A,LFY::GUS expression under long days followed a similar pattern in both ga1-3 andga1-3 phyB-5 mutants. The expression levels were very low during the first 15 d of growth, and althoughLFY::GUS expression eventually increased, it never reached the levels seen in wild-type plants. Application of GA3 restored the expression pattern seen in wild-type plants and phyB-5 single mutants. In contrast to long days, LFY::GUS expression remained very low in short-day-grown ga1-3 single mutants during the entire experimental period (Fig. 4B) (Blázquez et al., 1998). Although LFY::GUS expression inga1-3 phyB-5 double mutants increased shortly before flowering occurred, it was not different from thega1-3 single mutant during the first 4 weeks. Upon application of GA3,ga1-3 and ga1-3 phyB-5 mutants flowered as early as wild-type andphyB-5 plants treated with GA3 (Table I), which was paralleled by a similar increase in LFY::GUS expression (Fig. 4B). Fig. 4. Open in new tabDownload slide LFY::GUS expression during vegetative growth of ga1-3 andga1-3 phyB-5 mutants. Plants homozygous for the LFY::GUS transgene were grown in long (A) or short (B) days until flower buds were visible to the naked eye, except in the case of thega1-3 mutant without GA3treatment under short days (□), which had not flowered at the end of the experiment. 304G1 (ga1-3, white symbols) and 304G1B5 (ga1-3 phyB-5, black symbols) were treated with GA3 (circles) or left untreated (squares). Values are expressed as means ± 2 se (n = 12). Time represents days after sowing. Error bars that are not visible are contained within the symbol. MUG, 4-Methylumbelliferyl-β-d-glucopyranoside. Fig. 4. Open in new tabDownload slide LFY::GUS expression during vegetative growth of ga1-3 andga1-3 phyB-5 mutants. Plants homozygous for the LFY::GUS transgene were grown in long (A) or short (B) days until flower buds were visible to the naked eye, except in the case of thega1-3 mutant without GA3treatment under short days (□), which had not flowered at the end of the experiment. 304G1 (ga1-3, white symbols) and 304G1B5 (ga1-3 phyB-5, black symbols) were treated with GA3 (circles) or left untreated (squares). Values are expressed as means ± 2 se (n = 12). Time represents days after sowing. Error bars that are not visible are contained within the symbol. MUG, 4-Methylumbelliferyl-β-d-glucopyranoside. Interaction between PHYB and GAs The suppression by phyB-5 of the flowering defect of ga1-3 mutants, along with the weak up-regulation of LFY::GUS expression in thega1-3 phyB-5 double mutants are compatible with the idea that PHYB modulates flowering andLFY expression independently of GAs. However, thega1-3 mutation does not completely abolish GA biosynthesis, and several GA species are still detectable in thega1-3 mutant (Zeevaart and Talón, 1992; A. Silverstone, and T.-P. Sun, personal communication). In addition, there is the possibility that physiologically relevant levels of exogenous GAs are carried over from the parental generation or from the seed treatment required for germination. Thus, PHYB could act by increasing the low levels of GA biosynthesis or by enhancing the responsiveness toward the small amount of GAs present in ga1-3mutants. An effect of phyB on GA biosynthesis is unlikely, since overall levels of several GA intermediates are unchanged inphyB mutants compared with the wild type (Reed et al., 1996). phyB mutants show enhanced responsiveness to GAs, as monitored by the dose response of hypocotyl elongation (Reed et al., 1996). We have found that continuous watering of plants with paclobutrazol, an inhibitor of the early steps of GA biosynthesis (Rademacher, 1991), did not prevent the flowering of ga1 phyB or phyB mutants under short days, while it abolished flowering of wild-type plants (results not shown). To resolve the question of whether PHYB affects flowering by regulating GA biosynthesis or GA response, we determined whether increased responsiveness to GAs could account for the increasedLFY::GUS expression in phyB mutants.ga1-3 and ga1-3 phyB-5 seedlings carrying aLFY::GUS transgene were grown on plates containing increasing concentrations of GA3. The hypocotyl length of each seedling was determined before measuring LFY::GUS activity. As previously reported (Reed et al., 1996), the hypocotyl of phyB mutants was longer than that ofPHYB+ plants at all GA3 concentrations. More importantly,phyB mutants were also more responsive to exogenous GA3 than wild-type plants (Fig.5A). In contrast, levels of LFY::GUS activity in phyB mutants did not show increased responsiveness to exogenous GA3 over the range of concentrations tested (Fig. 5B). The results were the same under long or short days. Fig. 5. Open in new tabDownload slide Responses of ga1-3and ga1-3 phyB-5 seedlings to exogenous GA3. Seeds of the lines 304G1 (LFY::GUS ga1-3, ○) and 304G1B5 (LFY::GUS ga1-3 phyB-5, •) were sown on MS plates containing the indicated concentrations of GA3. Hypocotyl length (A) and GUS activity (B) were determined 7 d after sowing. Values are expressed as means ± 2 se (n = 12). Time represents days after sowing. Error bars that are not visible are contained within the symbol. MUG, 4-Methylumbelliferyl-β-d-glucopyranoside. Fig. 5. Open in new tabDownload slide Responses of ga1-3and ga1-3 phyB-5 seedlings to exogenous GA3. Seeds of the lines 304G1 (LFY::GUS ga1-3, ○) and 304G1B5 (LFY::GUS ga1-3 phyB-5, •) were sown on MS plates containing the indicated concentrations of GA3. Hypocotyl length (A) and GUS activity (B) were determined 7 d after sowing. Values are expressed as means ± 2 se (n = 12). Time represents days after sowing. Error bars that are not visible are contained within the symbol. MUG, 4-Methylumbelliferyl-β-d-glucopyranoside. To confirm that the responsiveness to exogenous GA3 reflects the behavior toward endogenous GAs, we analyzed both hypocotyl length and LFY::GUS activity in wild-type and phyB-5 seedlings growing on plates with increasing concentrations of paclobutrazol. As expected, paclobutrazol reduced the elongation of wild-type and phyBhypocotyls starting at concentrations as low as 0.03 μm (Fig. 6A) (Reed et al., 1996). At higher concentrations, paclobutrazol inhibited hypocotyl elongation faster in phyB-5 mutants than in wild-type plants. Although paclobutrazol reducedLFY::GUS expression in both wild-type andphyB-5 plants, there was no difference in responsiveness between the two lines (Fig. 6B). Fig. 6. Open in new tabDownload slide Responses of ga1-3and ga1-3 phyB-5 seedlings to the GA-biosynthesis-inhibitor paclobutrazol. Seeds of the lines DW150-304 (LFY::GUS, ○) and 304B5 (LFY::GUS phyB-5, •) were sown on MS plates containing the indicated concentrations of paclobutrazol. Hypocotyl length (A) and GUS activity (B) were determined 7 d after sowing. Values are expressed as means ± 2 se (n = 12). Time represents days after sowing. Error bars that are not visible are contained within the symbol. MUG, 4-Methylumbelliferyl-β-d-glucopyranoside. Fig. 6. Open in new tabDownload slide Responses of ga1-3and ga1-3 phyB-5 seedlings to the GA-biosynthesis-inhibitor paclobutrazol. Seeds of the lines DW150-304 (LFY::GUS, ○) and 304B5 (LFY::GUS phyB-5, •) were sown on MS plates containing the indicated concentrations of paclobutrazol. Hypocotyl length (A) and GUS activity (B) were determined 7 d after sowing. Values are expressed as means ± 2 se (n = 12). Time represents days after sowing. Error bars that are not visible are contained within the symbol. MUG, 4-Methylumbelliferyl-β-d-glucopyranoside. Effect of phyB Mutation on Expression of COand FT GAs act redundantly with the long-day-dependent pathway of floral induction, as GA deficiency has much weaker effects on flowering in long than in short days (Wilson et al., 1992). In addition, double mutants carrying both the ga1-3 mutation and a mutation in CO, an essential element of the long-day pathway, often do not flower at all in long days (Putterill et al., 1995). Since all of our other data pointed to a GA-independent effect of PHYB on flowering, we wanted to determine whether the early flowering of phyB mutants was caused by an increase in the expression of genes known to be involved in the photoperiod-dependent pathway that promotes flowering. The expression of two genes in this pathway has been shown to be limiting for flowering, since their overexpression causes very early flowering under both long and short days. These genes are CO (Simon et al., 1996) andFT (I. Kardailsky and D. Weigel, unpublished data). When we analyzed the expression of CO and FT by RT-PCR, we found that the expression levels of CO andFT did not differ dramatically between wild-type andphyB-5 plants (Fig. 7), suggesting that the early-flowering phenotype of phyBmutants under short days is not caused by overexpression of genes in the long-day pathway. Fig. 7. Open in new tabDownload slide CO and FT RNA expression in phyB mutants. Seedlings of the lines DW150-304 (wild type [WT], white bars) and 304B5 (phyB-5, black bars) were grown on MS plates under short days (SD), and harvested during the 8th h of light on the indicated days. Expression was analyzed by RT-PCR (bottom panel) and the signals quantified and normalized using UBQ expression as a control (top panel, arbitrary units). Fig. 7. Open in new tabDownload slide CO and FT RNA expression in phyB mutants. Seedlings of the lines DW150-304 (wild type [WT], white bars) and 304B5 (phyB-5, black bars) were grown on MS plates under short days (SD), and harvested during the 8th h of light on the indicated days. Expression was analyzed by RT-PCR (bottom panel) and the signals quantified and normalized using UBQ expression as a control (top panel, arbitrary units). DISCUSSION The level of LFY expression is an important determinant for the identity of the primordia that arise at the flanks of the shoot apical meristem during the transition to flowering (Blázquez et al., 1997). This idea has been corroborated by the observation that certain mutations that delay flowering, such as co orgi (gigantea), also reduce the level of LFY expression during the time that the transition to flowering occurs in wild type (Nilsson et al., 1998). In a similar way, the acceleration of flowering caused by mutations such asspy is paralleled by increased LFY expression (Blázquez et al., 1998). Hence, it is not surprising that PHYB, which represses flowering, functions as a negative regulator of theLFY promoter, although it has been difficult to assess whether phyB mutants flower early exclusively because of increased LFY expression, or also because of an increased response to LFY activity (Nilsson et al., 1998). In our growth conditions (Fig. 1), PHYB affected flowering time mainly under short photoperiods. Although the total number of leaves produced was similar under long and short days (Table I), the number of days needed to produce the first flower was higher under short days, as previously reported (Koornneef et al., 1995). Accordingly, the increase of LFY promoter activity caused by the phyBmutation was more clearly observed under short than under long days. It has been previously observed that the contribution of different photoreceptors to the control of flowering time changes with photoperiod (Bagnall et al., 1995). For example, a specific role for the blue/UVA photoreceptor encoded by CRY2 is the promotion of flowering under long days (Guo et al., 1998; Lin et al., 1998). The observation that a phyB mutation has a more pronounced effect on LFY expression under short days suggests an interaction with GAs, since GAs are essential for flowering under noninductive conditions (Martı́nez-Zapater et al., 1994). However, we present evidence that PHYB and GAs regulate LFYexpression through independent pathways, since a phyBmutation did not enhance the response of the LFY promoter to GAs. An independent action was confirmed by the observation thatphyB suppressed the flowering defect of ga1mutants under short days, even when any GA1-independent synthesis of ent-kaurene was inhibited by paclobutrazol treatment. The simplest scenario for this suppression would be activation of the long-day-dependent pathway of flowering. However, the observation that the levels of RNA expression of CO andFT, the two genes believed to act downstream in the long-day pathway (Simon et al., 1996; Koornneef et al., 1998), are not dramatically changed in phyB mutants suggests that this activation would not occur at the transcriptional level. Alternatively, the suppression could take place through the FCA-dependent autonomous pathway (Macknight et al., 1997). An important conclusion from our study and previous studies is that the relationship between PHYB and GAs is complex. For certain responses, such as germination, the GA-deficient ga1 mutant is completely epistatic over phyB. However, most other characteristics of ga1 phyB double-mutant seedlings are intermediate between those observed for the single mutant parents (Peng and Harberd, 1997; this study). Finally, while PHYB regulates hypocotyl elongation by modulating responsiveness to GAs (Reed et al., 1996), PHYB acts independently of GAs in the control of flowering time (this study). That genes do not always interact in the same fashion, even when controlling the same targets, is not uncommon in development. For instance, PHYA and PHYB regulate common responses to light, but they do so differently depending on the particular response. While both phytochromes have an inhibitory effect on hypocotyl elongation and promote seed germination, flowering is repressed by PHYB but promoted by PHYA. It is not easy to imagine the molecular mechanism that integrates the different interactions between GAs and PHYB. Based on the effects of application of different GAs to ryegrass, it has been proposed that certain GA species possess low florigenic activity but promote stem elongation very efficiently and vice versa (Evans et al., 1990). For instance, 3β-hydroxylation is required for the promotion of stem elongation, but not of flowering in this plant (Evans et al., 1994). If these findings reflect the presence of different receptors for the various active GA species, PHYB might regulate receptors specific for GA species involved in hypocotyl and stem elongation but not the ones involved in flowering. ACKNOWLEDGMENTS We thank Ove Nilsson, Michael Neff, and José Miguel Martı́nez-Zapater for discussion and critical reading of the manuscript; Trevor Phan for technical assistance with the hypocotyl length measurements; Ji Hoon Ahn, Paul Reeves, and George Coupland for advice concerning the RT-PCR experiments; and Aron Silverstone for communicating unpublished results. phyB-5 mutant seeds were obtained from Joanne Chory. Abbreviation: RT reverse transcriptase LITERATURE CITED 1 Aukerman MJ Hirschfeld M Wester L Weaver M Clack T Amasino RM Sharrock RA A deletion in the PHYD gene of the Arabidopsis Wassilewskija ecotype defines a role for phytochrome D in red/far-red light sensing. Plant Cell 9 1997 1317 1326 Google Scholar PubMed OpenURL Placeholder Text WorldCat 2 Bagnall DJ King RW Whitelam GC Boylan MT Wagner D Quail PH Flowering responses to altered expression of phytochrome in mutants and transgenic lines of Arabidopsis thaliana (L.) Heynh. Plant Physiol 108 1995 1495 1503 Google Scholar Crossref Search ADS PubMed WorldCat 3 Beall FD Morgan PW Mander LN Miller FR Babb KH Genetic regulation of development in Sorghum bicolor. V. The ma 3 R allele results in gibberellin enrichment. Plant Physiol 95 1991 116 125 Google Scholar Crossref Search ADS PubMed WorldCat 4 Blázquez MA Green R Nilsson O Sussman MR Weigel D Gibberellins promote flowering of Arabidopsis by activating the LEAFY promoter. Plant Cell 10 1998 791 800 Google Scholar Crossref Search ADS PubMed WorldCat 5 Blázquez MA Soowal L Lee I Weigel D LEAFY expression and flower initiation in Arabidopsis. Development 124 1997 3835 3844 Google Scholar Crossref Search ADS PubMed WorldCat 6 Callis J Carpenter T Sun CW Vierstra RD Structure and evolution of genes encoding polyubiquitin and ubiquitin-like proteins in Arabidopsis thaliana ecotype Columbia. Genetics 139 1995 921 939 Google Scholar Crossref Search ADS PubMed WorldCat 7 Chory J Li H Gibberellins, brassinosteroids and light-regulated development. Plant Cell Environ 20 1997 801 806 Google Scholar Crossref Search ADS WorldCat 8 Devlin PF Patel SR Whitelam GC Phytochrome E influences internode elongation and flowering time in Arabidopsis. Plant Cell 10 1998 1479 1487 Google Scholar Crossref Search ADS PubMed WorldCat 9 Evans LT King RW Chu A Mander LN Pharis RP Gibberellin structure and florigenic activity in Lolium temulentum, a long-day plant. Planta 182 1990 97 106 Google Scholar Crossref Search ADS PubMed WorldCat 10 Evans LT King RW Mander LN Pharis RP The relative significance for stem elongation and flowering in Lolium temulentum of 3β-hydroxylation of gibberellins. Planta 192 1994 130 136 Google Scholar OpenURL Placeholder Text WorldCat 11 Fankhauser C Chory J Light control of plant development. Annu Rev Cell Dev Biol 13 1997 203 229 Google Scholar Crossref Search ADS PubMed WorldCat 12 Goto K Kumagai T Koornneef M Flowering responses to light breaks in photomorphogenic mutants of Arabidopsis thaliana, a long-day plant. Physiol Plant 83 1991 209 215 Google Scholar Crossref Search ADS WorldCat 13 Guo H Yang H Mockler TC Lin C Regulation of flowering time by Arabidopsis photoreceptors. Science 279 1998 1360 1363 Google Scholar Crossref Search ADS PubMed WorldCat 14 Hempel FD Weigel D Mandel MA Ditta G Zambryski P Feldman LJ Yanofsky MF Floral determination and expression of floral regulatory genes in Arabidopsis. Development 124 1997 3845 3853 Google Scholar Crossref Search ADS PubMed WorldCat 15 Jacobsen SE Olszewski NE Mutations at the SPINDLY locus of Arabidopsis alter gibberellin signal transduction. Plant Cell 5 1993 887 896 Google Scholar PubMed OpenURL Placeholder Text WorldCat 16 Koornneef M Alonso-Blanco C Peeters AJM Soppe W Genetic control of flowering in Arabidopsis. Annu Rev Plant Physiol Plant Mol Biol 49 1998 345 370 Google Scholar Crossref Search ADS PubMed WorldCat 17 Koornneef M Hanhart C van Loenen-Martinet P Blankestijn de Vries H The effect of daylength on the transition to flowering in phytochrome-deficient, late-flowering and double mutants of Arabidopsis thaliana. Physiol Plant 95 1995 260 266 Google Scholar Crossref Search ADS WorldCat 18 Koornneef M Rolff E Spruit CJP Genetic control of light-inhibited hypocotyl elongation in Arabidopsis thaliana (L.) Heynh. Z Pflanzenphysiol 100 1980 147 160 Google Scholar Crossref Search ADS WorldCat 19 Koornneef M van der Veen JH Induction and analysis of gibberellin sensitive mutants in Arabidopsis thaliana. Theor Appl Genet 58 1980 257 263 Google Scholar Crossref Search ADS PubMed WorldCat 20 Lee IJ Foster KR Morgan PW Photoperiod control of gibberellin levels and flowering in sorghum. Plant Physiol 116 1998 1003 1011 Google Scholar Crossref Search ADS PubMed WorldCat 21 Lin C Yang H Guo H Mockler T Chen J Cashmore AR Enhancement of blue-light sensitivity of Arabidopsis seedlings by a blue light receptor cryptochrome 2. Proc Natl Acad Sci USA 95 1998 2686 2690 Google Scholar Crossref Search ADS PubMed WorldCat 22 López-Juez E Kobayashi M Sakurai A Kamiya Y Kendrick RE Phytochrome, gibberellins, and hypocotyl growth. A study using the cucumber (Cucumis sativus L.) long hypocotyl mutant. Plant Physiol 107 1995 131 140 Google Scholar Crossref Search ADS PubMed WorldCat 23 Macknight R Bancroft I Lister C Page T Love K Schmidt R Westphal L Murphy G Sherson S Cobbett C FCA, a gene controlling flowering time in Arabidopsis, encodes a protein containing RNA-binding domains. Cell 89 1997 737 745 Google Scholar Crossref Search ADS PubMed WorldCat 24 Martı́nez-Garcı́a JF Garcı́a-Martı́nez JL Phytochrome modulation of gibberellin metabolism in cowpea epicotyl elongation. Karssen CM van Loon LC Vreugdenhil D Progress in Plant Growth Regulation. 1992 585 590 Kluwer Academic Publishers Dordrecht, The Netherlands 25 Martı́nez-Zapater JM, Coupland G, Dean C, Koornneef M(1994) The transition to flowering in Arabidopsis. In EM Meyerowitz, CR Somerville, eds, Arabidopsis. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, pp 403–433 26 Murashige T Skoog F A revised medium for rapid growth and bio-assay with tobacco tissue cultures. Physiol Plant 15 1962 473 497 Google Scholar Crossref Search ADS WorldCat 27 Nagatani A Reed JW Chory J Isolation and initial characterization of Arabidopsis mutants that are deficient in phytochrome A. Plant Physiol 102 1993 269 277 Google Scholar Crossref Search ADS PubMed WorldCat 28 Neff MM Chory J Genetic interactions between phytochrome A, phytochrome B and cryptochrome 1 during Arabidopsis development. Plant Physiol 118 1998 27 35 Google Scholar Crossref Search ADS PubMed WorldCat 29 Neff MM Neff JD Chory J Pepper AE dCAPS, a simple technique for the genetic analysis of single nucleotide polymorphisms: experimental applications in Arabidopsis thaliana genetics. Plant J 14 1998 387 392 Google Scholar Crossref Search ADS PubMed WorldCat 30 Nilsson O Lee I Blázquez MA Weigel D Flowering-time genes modulate the response to LEAFY activity. Genetics 150 1998 403 410 Google Scholar Crossref Search ADS PubMed WorldCat 31 Parks BM Quail PH hy8, a new class of Arabidopsis long hypocotyl mutants deficient in functional phytochrome A. Plant Cell 5 1993 39 48 Google Scholar PubMed OpenURL Placeholder Text WorldCat 32 Peng J Harberd NP Gibberellin deficiency and response mutations suppress the stem elongation phenotype of phytochrome-deficient mutants of Arabidopsis. Plant Physiol 113 1997 1051 1058 Google Scholar Crossref Search ADS PubMed WorldCat 33 Putterill J Robson F Lee K Simon R Coupland G The CONSTANS gene of Arabidopsis promotes flowering and encodes a protein showing similarities to zinc finger transcription factors. Cell 80 1995 847 857 Google Scholar Crossref Search ADS PubMed WorldCat 34 Quail PH Boylan MT Parks BM Short TW Xu Y Wagner D Phytochromes: photosensory perception and signal transduction. Science 268 1995 675 680 Google Scholar Crossref Search ADS PubMed WorldCat 35 Rademacher W (1991) Biochemical effects of plant growth retardants. In HW Gausman, ed, Plant Biochemical Regulators. Marcel Dekker, New York, pp 169–200 36 Reed JW Foster KR Morgan PW Chory J Phytochrome B affects responsiveness to gibberellins in Arabidopsis. Plant Physiol 112 1996 337 342 Google Scholar Crossref Search ADS PubMed WorldCat 37 Reed JW Nagatani A Elich TD Fagan M Chory J Phytochrome A and phytochrome B have overlapping but distinct functions in Arabidopsis development. Plant Physiol 104 1994 1139 1149 Google Scholar Crossref Search ADS PubMed WorldCat 38 Reed JW Nagpal P Poole DS Furuya M Chory J Mutations in the gene for the red/far-red light receptor phytochrome B alter cell elongation and physiological responses throughout Arabidopsis development. Plant Cell 5 1993 147 157 Google Scholar PubMed OpenURL Placeholder Text WorldCat 39 Rood SB Williams PH Pearce D Murofushi N Mander LN Pharis RP A mutant gene that increases gibberellin production in Brassica. Plant Physiol 93 1990 1168 1174 Google Scholar Crossref Search ADS PubMed WorldCat 40 Sharrock RA Quail PH Novel phytochrome sequences in Arabidopsis thaliana: structure, evolution, and differential expression of a plant regulatory photoreceptor family. Genes Dev 3 1989 1745 1757 Google Scholar Crossref Search ADS PubMed WorldCat 41 Silverstone AL Mak PY Martinez EC Sun T-P The new RGA locus encodes a negative regulator of gibberellin response in Arabidopsis thaliana. Genetics 146 1997 1087 1099 Google Scholar Crossref Search ADS PubMed WorldCat 42 Simon R Igeño MI Coupland G Activation of floral meristem identity genes in Arabidopsis. Nature 382 1996 59 62 Google Scholar OpenURL Placeholder Text WorldCat 43 Talón M Zeevaart JAD Gage DA Identification of gibberellins in spinach and effects of light and darkness on their levels. Plant Physiol 97 1991 1521 1526 Google Scholar Crossref Search ADS PubMed WorldCat 44 Weller JL Ross JJ Reid JB Gibberellins and phytochrome regulation of stem elongation in pea. Planta 192 1994 489 496 Google Scholar Crossref Search ADS WorldCat 45 Whitelam GC Johnson E Peng J Carol P Anderson ML Cowl JS Harberd NP Phytochrome A null mutants of Arabidopsis display a wild-type phenotype in white light. Plant Cell 5 1993 757 768 Google Scholar PubMed OpenURL Placeholder Text WorldCat 46 Wilson RN Heckman JW Somerville CR Gibberellin is required for flowering in Arabidopsis thaliana under short days. Plant Physiol 100 1992 403 408 Google Scholar Crossref Search ADS PubMed WorldCat 47 Wu KQ Li L Gage DA Zeevaart JAD Molecular cloning and photoperiod-regulated expression of gibberellin 20-oxidase from the long-day plant spinach. Plant Physiol 110 1996 547 554 Google Scholar Crossref Search ADS PubMed WorldCat 48 Xu YL Li L Wu K Peeters AJ Gage DA Zeevaart JA The GA5 locus of Arabidopsis thaliana encodes a multifunctional gibberellin 20-oxidase: molecular cloning and functional expression. Proc Natl Acad Sci USA 92 1995 6640 6644 Google Scholar Crossref Search ADS PubMed WorldCat 49 Yamaguchi S Smith MW Brown RG Kamiya Y Sun T Phytochrome regulation and differential expression of gibberellin 3β-hydroxylase genes in germinating Arabidopsis seeds. Plant Cell 10 1998 2115 2126 Google Scholar PubMed OpenURL Placeholder Text WorldCat 50 Zeevaart JAD, Talón M (1992) Gibberellin mutants inArabidopsis thaliana. In CM Karssen, LC van Loon, D Vreugdenhil, eds, Progress in Plant Growth Regulation. Kluwer Academic Publishers, Dordrecht, The Netherlands, pp 34–42 Author notes 1 This work was supported by grants from the National Science Foundation (no. MCB–9723823 to D.W.) and the Human Frontiers Science Program Organization (no. RG 303/97 to D.W.). M.A.B. received fellowships from the Spanish Ministry of Education and the Human Frontiers Science Program Organization. D.W. is a National Science Foundation Young Investigator and receives support from Agritope (Oregon). * Corresponding author; e-mail [email protected]; fax 858–558–6379. Copyright © 1999 American Society of Plant Physiologists This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)