Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Mutations in the Arabidopsis Nuclear-Encoded Mitochondrial Phage-Type RNA Polymerase Gene RPOTm Led to Defects in Pollen Tube Growth, Female Gametogenesis and Embryogenesis

Mutations in the Arabidopsis Nuclear-Encoded Mitochondrial Phage-Type RNA Polymerase Gene RPOTm... Abstract The mitochondrial genes in Arabidopsis thaliana are transcribed by a small family of nuclear-encoded T3/T7 phage-type RNA polymerases (RPOTs). At least two nuclear-encoded RPOTs (RPOTm and RPOTmp) are located in mitochondria in A. thaliana. Their genetic roles are largely unknown. Here we report the characterization of novel mutations in the A. thaliana RPOTm gene. The mutations did not affect pollen formation, but significantly retarded the growth of the rpoTm mutant pollen tubes and had an impact on the fusion of the polar nuclei in the rpoTm mutant embryo sacs. Moreover, development of the rpoTm/– mutant embryo was arrested at the globular stage. The rpoTm rpoTmp double mutation could enhance the rpoTm mutant phenotype. Expression of RPOTmp under control of the RPOTm promoter could not complement the phenotype of the rpoTm mutations. All these data indicate that RPOTm is important for normal pollen tube growth, female gametogenesis and embryo development, and has distinct genetic and molecular roles in plant development, which cannot be replaced by RPOTmp. Introduction The plant life cycle alternates between a diploid sporophytic phase and a haploid gametophytic phase. The diploid male and female sporophytes produce haploid microspores and megaspores that give rise to male and female gametophytes, respectively. Specifically, in the anther, the diploid pollen mother cell (PMC) undergoes meiosis to generate haploid microspores. Then, each individual microspore undergoes two rounds of mitosis to form a three-celled pollen grain (male gametophyte) that comprises two sperm cells and a vegetative cell. When the mature pollen grain lands on the female stigma, shortly thereafter it germinates and produces a pollen tube that invades the stigmatic tissue and then elongates in the female transmitting tract to deliver the two sperm cells into an embryo sac (McCormick 1993, McCormick 2004). In the ovule, the megaspore mother cell (MMC) undergoes meiosis to give rise to four haploid megaspores, three of which undergo programmed cell death and only one survives and differentiates into a functional megaspore. The functional megaspore then undergoes three rounds of mitotic nuclear division to generate an eight-nucleus coenocytic embryo sac. The coenocytic embryo sac then goes through the fusion of polar nuclei and cytokinesis to form a seven-celled embryo sac (female gametophyte) that consists of four different cell types: three antipodal cells (n), one central cell (2n), one egg cell (n) and two synergid cells (n). Then, the antipodal cells undergo programmed cell death. When a pollen tube enters the embryo sac by penetrating one of the two synergid cells through the filiform apparatus, the synergid cell also undergoes cell death. Subsequently, the two sperm cells are discharged from the pollen tube and migrate to the egg cell and central cell nucleus to achieve double fertilization. Thereafter, embryogenesis takes place (Drews et al. 1998, Yadegari and Drews, 2004, Shi et al. 2005). Several lines of evidences have shown that the mitochondrion is deeply involved in gametophyte development. First, physiological data showed that during pollen formation, the number of mitochondria per cell increased in maize (Zea mays) and tobacco (Nicotiana tabacum) anthers (Warmke and Lee 1978, Huang et al. 1994) and, in the growing pollen tube, the mitochondria exhibited a high motility and formed continuous streaming throughout the cytoplasm (Parton et al. 2003, Yamaoka and Leaver 2008). Secondly, genetic characterization of Arabidopsis mutants defective in nuclear genes encoding mitochondria-associated proteins showed that mitochondria played important roles in gametophyte development (Skinner et al. 2001, Christensen et al. 2002, León et al. 2007, Yamaoka and Leaver 2008). For example, a mutation in EMB2473/MIRO1 encoding a Miro GTPase in mitochondria altered the mitochondrial structure in pollen grains and strongly affected pollen tube growth (Yamaoka and Leaver 2008). Thirdly, mutations in mitochondrial genes may lead to severe defects in respiration and affect male gametophyte development, which could result in cytoplasmic male sterility (CMS) that has been described in about 150 plant species (Conley and Hanson 1995, Hanson and Bentolila 2004, Linke and Börner 2005, Carlsson et al. 2008). Furthermore, several essential mitochondrial gene transcripts have been found to exhibit high abundance in developing maize male gametophytes (Monéger et al. 1992, Wen and Chase 1999). Recently, microarray analysis also showed that most mitochondrial genes were highly expressed in ungerminated dry pollen grains, growing pollen tubes and ovaries in Arabidopsis (Qin et al. 2009). All these data indicate that mitochondria play important roles in plant gametophyte development. The mitochondrion is a semi-autonomous organelle that possesses its own genomes and gene expression systems (Millar et al. 2004). In Arabidopsis, the mitochondrial genome bears 57 genes (Unseld et al. 1997, Binder and Brennicke 2003). As in most other eukaryotic organisms, the mitochondrial genes in Arabidopsis are transcribed by a small family of nuclear-encoded phage-type RNA polymerases. RNA polymerase of T3/T7 type (RPOT) is most similar to the RNA polymerases (RNAPs) of bacteriophage T3, T7 and SP6 (Greenleaf et al. 1986, Kelly et al. 1986, Cermakian et al. 1996, Hedtke et al. 1997). There are three nuclear-encoded RPOTs in Arabidopsis, the first (RPOTm) exists exclusively in the mitochondrion, the second (RPOTp) is found solely in the plastid and the third (RPOTmp) is presumed to be located in both organelles (Hedtke et al. 1997, Hedtke et al. 1999, Hedtke et al. 2000). Previous studies have indicated that all of them are transcriptionally active RNAPs (Kühn et al. 2007). Both RPOTm and RPOTmp are prominently expressed in meristematic and young tissues that have high mitochondrial activity (Emanuel et al. 2006). Thus, there are at least two active phage-type RNAPs that are located in Arabidopsis mitochondria. RPOTm can recognize the mitochondrial promoter accurately in vitro, whereas RPOTmp does not display any significant promoter specificity (Kühn et al. 2007). A recent study indicated that RPOTmp performed gene-specific transcription in Arabidopsis mitochondria (Kühn et al. 2009). Therefore, both of them participate in mitochondrial gene transcription. However, the genetic roles of these mitochondria phage-type RNA polymerases in plants are still poorly understood. We report here the characterization of the novel mutations in the Arabidopsis RPOTm gene. The rpoTm mutations significantly retarded pollen tube growth and disrupted embryogenesis and fusion of polar nuclei in embryo sacs. These data indicate that RPOTm plays important roles in pollen tube growth, female gametogenesis and embryogenesis. The rpoTm rpoTmp double mutation could enhance the rpoTm mutant phenotype. It almost completely disrupted female gametogenesis and severely impaired pollen formation and germination. Furthermore, expression of RPOTmp under the control of the RPOTm promoter could not complement the rpoTm mutant phenotype. These results indicate that the mitochondrial phage-type RNA polymerases are important for gametogenesis, and they have distinct roles in plant development. Results Isolation and genetic analysis of mutant mgp3-1 The first rpoTm mutant allele used in this study was identified in a screen for male gametophyte-defective (mgp) mutants in Arabidopsis thaliana using the gene- and enhancer-trap system (Sundaresan et al. 1995) and was named male gametophyte- defective 3-1 (mgp3-1). The mutation was generated by the insertion of the modified dissociation (Ds) element (Sundaresan et al. 1995). DNA gel blot hybridization showed a single Ds insertion in the genome (Supplementary Fig. S1). The Ds element carries an NPTII gene as a selective genetic marker. It also carries a GUS (β-glucuronidase) reporter gene that could be co-expressed specifically in the mutant pollen grains. Therefore, the mutant pollen grains could be distinguished from wild-type pollen grains by GUS staining (Fig. 1A, B). The mutant exhibited a segregation ratio of roughly 1 (1,240) kanamycin-resistant (KanR) to 1 (1,314) kanamycin-sensitive (KanS) in the progeny of the self-crossed mgp3-1/+ plants. No mgp3-1 homozygous mutant plant was found in the progeny (n = 230). The outcrosses of mgp3-1/+ plants to wild-type plants showed that both the male and female transmission efficiencies of the mgp3-1 mutation were significantly reduced. In particular, the male transmission efficiency was 31% (519/1,689) and the female transmission efficiency was 59% (744/1,264) (Table 1). This result showed that the mutation had a stronger effect on male gametophytic function than on female gametophytic function. Fig. 1 View largeDownload slide The mgp3-1 mutation affected pollen tube growth in vitro. (A, B) The mgp3-1/+ (A) and wild-type (B) pollen grains with or without GUS stain. (C, D) The SEM images of mgp3-1/+ (C) and wild-type (D) pollen grains. (E, F) The Alexander-stained mgp3-1/+ (E) and wild-type (F) pollen grains. (G, H) The DAPI-stained mgp3-1/+ (G) and wild-type (H) pollen grains. (I–K) In vitro germination of the pollen grains from wild-type (I), mgp3-1/+ (J) and MGP3-complemented mgp3-1 homozygous plants (K). (L) Statistical analysis of pollen tube length in wild-type, mgp3-1/+ and two MGP3-complemented mgp3-1 homozygous plants (MGP3-F-1 and MGP3-F-2). The mutant pollen grains were labeled by GUS staining. The red arrowheads indicate the mgp3-1 mutant pollen grains and pollen tubes. Bars = 20 μm in A–H, 50 μm in I–K. Fig. 1 View largeDownload slide The mgp3-1 mutation affected pollen tube growth in vitro. (A, B) The mgp3-1/+ (A) and wild-type (B) pollen grains with or without GUS stain. (C, D) The SEM images of mgp3-1/+ (C) and wild-type (D) pollen grains. (E, F) The Alexander-stained mgp3-1/+ (E) and wild-type (F) pollen grains. (G, H) The DAPI-stained mgp3-1/+ (G) and wild-type (H) pollen grains. (I–K) In vitro germination of the pollen grains from wild-type (I), mgp3-1/+ (J) and MGP3-complemented mgp3-1 homozygous plants (K). (L) Statistical analysis of pollen tube length in wild-type, mgp3-1/+ and two MGP3-complemented mgp3-1 homozygous plants (MGP3-F-1 and MGP3-F-2). The mutant pollen grains were labeled by GUS staining. The red arrowheads indicate the mgp3-1 mutant pollen grains and pollen tubes. Bars = 20 μm in A–H, 50 μm in I–K. Table 1 Genetic analysis of mgp3-1 and mgp3-2 mutants Crosses (female × male)  KanR  KanS  KanR/KanS  TEF (%)  TEM (%)  mgp3-1/+ × mgp3-1/+  1,240  1,314  0.94  NA  NA  mgp3-1/+ × +/+  744  1,264  0.59  59  NA  +/+ × mgp3-1/+  519  1,689  0.31  NA  31  mgp3-2/+ × mgp3-2/+  1,127  1,332  0.85  NA  NA  mgp3-2/+ × +/+  573  994  0.58  58  NA  +/+ × mgp3-2/+  483  2,219  0.22  NA  22  Crosses (female × male)  KanR  KanS  KanR/KanS  TEF (%)  TEM (%)  mgp3-1/+ × mgp3-1/+  1,240  1,314  0.94  NA  NA  mgp3-1/+ × +/+  744  1,264  0.59  59  NA  +/+ × mgp3-1/+  519  1,689  0.31  NA  31  mgp3-2/+ × mgp3-2/+  1,127  1,332  0.85  NA  NA  mgp3-2/+ × +/+  573  994  0.58  58  NA  +/+ × mgp3-2/+  483  2,219  0.22  NA  22  KanR, kanamycin-resistant; KanS, kanamycin-sensitive; NA, not applicable; TE, transmission efficiency; TE = (KanR/KanS) × 100%; TEF, female transmission efficiency; TEM, male transmission efficiency. View Large The mgp3-1 mutation significantly retarded pollen tube growth We first examined the morphology and viability of mature mgp3-1 pollen grains. As shown in Fig. 1, mgp3-1 pollen grains appeared morphologically normal (Fig. 1C, D). Alexander staining showed that mgp3-1 pollen grains contained cytoplasm like the wild type (Fig. 1E, F). 4′,6-Diamidino-2-phenylindole (DAPI) staining indicated that the mutant pollen grains had one vegetative nucleus and two sperm nuclei like the wild type (Fig. 1G, H). These results suggested that the mgp3-1 mutation did not affect the formation of pollen grains. We further investigated mgp3-1 pollen germination in vitro. The pollen grains from wild-type and mgp3-1 heterozygous plants (mgp3-1/+) were germinated for 6 h and labeled by GUS staining (Fig. 1I–K). The pollen grains from mgp3-1 heterozygous plants included two types of pollen grains, the mgp3-1 mutant pollen grains with GUS stain and the wild-type pollen grains without GUS stain. As shown in Fig. 1L, the germination rates of GUS-positive and GUS-negative pollen grains from mgp3-1/+ plants were 73.6% (733/996) and 73.8% (614/832), respectively, similar to that (76.6%, 660/862) of the pollen grains from wild-type plants, indicating that the mutation did not affect pollen germination. The average length of the wild-type pollen tubes and GUS-negative pollen tubes from the mgp3-1/+ plants was roughly 180 μm (Fig. 1I, L). In contrast, the average length of the GUS-positive mutant pollen tubes was about 80 μm (Fig. 1J, L). These findings indicated that mgp3-1 mutation dramatically affected pollen tube growth in vitro. To determine if the mutant pollen tube growth was also affected in vivo, mgp3-1/+ plants were used as pollen donors to cross with wild-type plants. Seeds from the upper and lower halves of the siliques were harvested separately. The seeds were sown on MS agar plates supplied with kanamycin. The KanR segregation rate of the seedlings derived from the seeds of the upper half of the siliques was 41.7 % (171/401), 3.2 times as high as that (12.8%, 63/492) of the seedlings derived from the seeds of the lower half of the siliques, indicating that the growth of mgp3-1 mutant pollen tubes was also significantly affected in vivo. Taken together, we concluded that the mgp3-1 mutation significantly retarded pollen tube growth. The mgp3-1 mutation affected ovule and embryo development To assess the ovule and embryo development processes in the mgp3-1 mutant, the siliques from wild-type and mgp3-1/+ plants were examined at 8 d after self-pollination. In contrast to the wild-type siliques that only had 3% aborted ovules and no abnormal seeds (Fig. 2A, Table 2), the mgp3-1/+ siliques had 8.4% aborted seeds and 17.5% aborted ovules (Fig. 2B, D, E; also see Table 2). To investigate the cause of seed abortion in mgp3-1/+ plants, the seeds from the crosses of mgp3-1/+ and the wild-type were examined at 8 d after pollination. When mgp3-1/+ was used as the pollen donor to pollinate wild-type pistils, the seed setting was normal as in the wild type. When a wild-type plant was used as the pollen donor to pollinate mgp3-1/+ pistils, 20% (185/943) of ovules were found to be unfertilized (Supplementary Fig. S2B), but no aborted seeds were observed. These results indicated that the seed abortion resulted from mgp3-1 homozygous mutation. Table 2 Statistical analysis of seed setting in mgp3-1/+ and mgp3-2/+ siliques by comparison with wild-type plants Plants  Total seeds and ovules  Normal seeds   Aborted seeds   Aborted ovules   n  %  n  %  n  %  Wild-type Ler  1,852  1,790  96.7  0  0  62  3.3  mgp3-1/+  1,996  1,479  74.1  168  8.4  349  17.5  Wild-type Col-0  2,120  2,049  96.6  0  0  71  3.4  mgp3-2/+  2,132  1,515  71.1  223  10.5  394  18.5  Plants  Total seeds and ovules  Normal seeds   Aborted seeds   Aborted ovules   n  %  n  %  n  %  Wild-type Ler  1,852  1,790  96.7  0  0  62  3.3  mgp3-1/+  1,996  1,479  74.1  168  8.4  349  17.5  Wild-type Col-0  2,120  2,049  96.6  0  0  71  3.4  mgp3-2/+  2,132  1,515  71.1  223  10.5  394  18.5  The seeds were counted on the first eight siliques from five independent wild-type or mutant plants. View Large Fig. 2 View largeDownload slide mgp3-1 mutation affected embryo development. (A–C) The siliques from wild-type (A), mgp3-1 heterozygous (B) and MGP3-complemented mgp3-1 homozygous (C) plants. The white arrows indicate the aborted mutant seeds. The white asterisks indicate the aborted ovules. (D) The magnified images of a normal seed (left) and an aborted seed (right). (E) A DIC image of cleared whole-mounted seeds in the mgp3-1/+ siliques. (F–I) DIC images of wild-type embryos at the globular stage (F), heart stage (G), torpedo stage (H) and matured embryo (I). (J–M) DIC images of mgp3-1 mutant embryos at the 2-cell stage (J), 8-cell stage (K) and multiple cell stages (L) and (M). Bars = 1 mm in A–C, 100 μm in D, E, 25 μm in F–M. Fig. 2 View largeDownload slide mgp3-1 mutation affected embryo development. (A–C) The siliques from wild-type (A), mgp3-1 heterozygous (B) and MGP3-complemented mgp3-1 homozygous (C) plants. The white arrows indicate the aborted mutant seeds. The white asterisks indicate the aborted ovules. (D) The magnified images of a normal seed (left) and an aborted seed (right). (E) A DIC image of cleared whole-mounted seeds in the mgp3-1/+ siliques. (F–I) DIC images of wild-type embryos at the globular stage (F), heart stage (G), torpedo stage (H) and matured embryo (I). (J–M) DIC images of mgp3-1 mutant embryos at the 2-cell stage (J), 8-cell stage (K) and multiple cell stages (L) and (M). Bars = 1 mm in A–C, 100 μm in D, E, 25 μm in F–M. The differential interference contrast (DIC) microscopic images of whole-mount clearing ovules showed no significant difference in the seeds at the zygote and single-terminal cell stages between the wild type and the mutant. The defects in embryo development occurred at the early globular stage. In the mgp3-1/+ siliques, 90% (415/468) of embryos were at the 16-cell to 64-cell stages (Fig. 2F), while the others were just at the 2- or 4-terminal cell stage (Fig. 2J). In the older siliques, the normal embryos had entered the early heart stage (Fig. 2G) while the embryos in the aborted seeds were retarded at the early globular stage (Fig. 2K). When wild-type embryos had developed to torpedo and mature stages (Fig. 2H, I), the mutant embryos were still at the globular stage and had an expanded shape (Fig. 2L, M). When the wild-type seeds matured, the aborted seeds shrunk and had a dark yellow color. No (n > 100) mutant embryo could go beyond the globular stage in the aborted seeds that had shrunk. This observation indicated that the mgp3-1 homozygous mutation could disrupt embryogenesis and consequently led to embryo lethality at the globular stage, suggesting that RPOTm is essential for embryogenesis in Arabidopsis. The mutant embryo sacs were defective in fusion of polar nuclei We characterized the mature mutant ovules by confocal laser scanning microscopy (CLSM). All ovules in wild-type siliques at the mature stage contained one secondary central cell nucleus, one egg cell nucleus and two synergid cell nuclei (Fig. 3C). In mgp3-1/+ siliques, however, only 82.2% (268/326) of ovules could develop normally like the wild-type ovule, and 17.8% (58/326) of ovules showed failure of fusion of the two polar nuclei and the three antipodal cells degenerated earlier (Fig. 3G). To determine further if the aborted embryo sacs were caused by defects in fusion of polar nuclei, wild-type pollen grains were pollinated to wild-type and mgp3-1/+ pistils, respectively, and then the ovules were observed at different times after pollination. At 36 h after pollination, the zygotes in wild-type ovules started dividing and the fertilized secondary nucleus had divided into several endosperm nuclei (Fig. 3D). In mgp3-1/+ pistils, 80.9% (251/310) of ovules developed as normally as the wild type and could be fertilized, and 19.1% ovules could not be fertilized, among which 16.5% (51/310) of ovules had two unfused polar nuclei (Fig. 3H) and 2.6% (8/310) of ovules had one central cell nucleus. The egg cell appeared highly incompact and no embryo and endosperm nucleus were observed in these ovules. At 72 h after pollination, the unfused polar nuclei and egg cell nucleus in the abnormal ovules had degenerated. As a result, the ovules stopped growing (Fig. 3I) while wild-type embryos in the same silique had developed to the globular stage (Fig. 3E). These data indicated that the mgp3-1 mutation affected fusion of the polar nuclei and disrupted female gametophyte development. Fig. 3 View largeDownload slide The mgp3-1 mutation affected fusion of polar nuclei. (A–E) CLSM images of wild-type embryo sacs at stage FG5 (A), stage FG6 (B), the terminal stage before pollination (C), 36 h after pollination (D) and 72 h after pollination (E). (F–I) CLSM images of mgp3-1 embryo sacs with unfused polar nuclei at stage FG5 (F), the terminal stage before pollination (G), 36 h after pollination (H) and 72 h after pollination (I). ac, apical cell; AN, antipodal cell nuclei; bc, basal cell; Ch, chalazal end; CN, central cell nucleus; EdN, endosperm nucleus; Emb, embryo; EN, egg cell nucleus; PN, polar nucleus; SN, synergid cell nucleus. Bars = 10 μm. Fig. 3 View largeDownload slide The mgp3-1 mutation affected fusion of polar nuclei. (A–E) CLSM images of wild-type embryo sacs at stage FG5 (A), stage FG6 (B), the terminal stage before pollination (C), 36 h after pollination (D) and 72 h after pollination (E). (F–I) CLSM images of mgp3-1 embryo sacs with unfused polar nuclei at stage FG5 (F), the terminal stage before pollination (G), 36 h after pollination (H) and 72 h after pollination (I). ac, apical cell; AN, antipodal cell nuclei; bc, basal cell; Ch, chalazal end; CN, central cell nucleus; EdN, endosperm nucleus; Emb, embryo; EN, egg cell nucleus; PN, polar nucleus; SN, synergid cell nucleus. Bars = 10 μm. Molecular analysis of mgp3-1 Thermal asymmetric interlaced PCR (TAIL-PCR) (Liu et al. 1995) was applied to obtain the gene sequences flanking the Ds element in the mgp3-1 mutant genome. Sequence analysis of the TAIL-PCR products indicated that the Ds element was inserted in the first exon of At1g68990 (RPOTm) at a site 115 bp downstream of the ATG codon (Fig. 4A). The insertion site was then further confirmed by conventional PCR using the Ds element-specific primer Ds5-2 and the gene-specific primer P2 (Fig. 4B; see also Materials and Methods). The Ds insertion resulted in an 8 bp (ATCTCAGT) duplication of the host sequence at the insertion site (Fig. 4A). Reverse transcription–PCR (RT–PCR) analysis showed a chimeric transcript product of RPOTm mRNA and the Ds sequence (Fig. 4C), leading to loss of the C-terminal domain of RPOTm. Therefore, mgp3-1 is probably a null rpoTm allele. Fig. 4 View largeDownload slide Molecular characterization of the MGP3 gene. (A) A schematic diagram of the MGP3 gene structure, showing the Ds and T-DNA insertion sites in the mgp3-1 and mgp3-2 mutants. (B) Confirmation of the transgenic mutant plants in complementation experiments. (C) The chimeric transcript containing RPOTm mRNA and Ds 3′ end sequences, revealed by RT–PCR. The black boxes indicate exons, and black arrowheads indicate the positions of primers used for genotyping. 1300H, Ds5-2, Ds3-2, LBa1, P1 and P2 are the primers used in the PCR assays. Hyg, hygromycin; Kan, kanamycin; kb, kilobase pairs; MW, molecular weight; WT, wild-type. Fig. 4 View largeDownload slide Molecular characterization of the MGP3 gene. (A) A schematic diagram of the MGP3 gene structure, showing the Ds and T-DNA insertion sites in the mgp3-1 and mgp3-2 mutants. (B) Confirmation of the transgenic mutant plants in complementation experiments. (C) The chimeric transcript containing RPOTm mRNA and Ds 3′ end sequences, revealed by RT–PCR. The black boxes indicate exons, and black arrowheads indicate the positions of primers used for genotyping. 1300H, Ds5-2, Ds3-2, LBa1, P1 and P2 are the primers used in the PCR assays. Hyg, hygromycin; Kan, kanamycin; kb, kilobase pairs; MW, molecular weight; WT, wild-type. A search through the Arabidopsis Biological Resource Center (ABRC) database further identified an rpoTm T-DNA insertion mutant, SALK_005875. The T-DNA was inserted in the first exon of At1g68990 at a site 72 bp downstream ATG codon in this mutant. The insertion site was confirmed by PCR using a T-DNA left border primer (LBa1) and a gene-specific primer (Fig. 4B). The T-DNA also carried a kanamycin resistance marker gene, and SALK_005875 showed a segregation ratio of roughly 1 (1,127) KanR to 1 (1,332) KanS in the progeny from the self-crossed SALK_005875 heterozygous plants. No SALK_005875 homozygous mutant plant was obtained. Outcrosses of SALK_005875 with wild-type Col-0 plants showed that both the male and female transmission efficiencies of the SALK_005875 mutation were reduced in the same manner as mgp3-1/+ (Table 1). The siliques of SALK_005875 heterozygous mutant plants contained 10.5% aborted seeds and 18.5% aborted ovules (Table 2). The embryo development in the aborted seeds was arrested at the globular stage. Furthermore, the pollen tube growth and fusion of polar nuclei were also affected. All these phenotypes were identical to those observed in mgp3-1/+. Therefore, SALK_005875 was considered to be allelic to mgp3-1 and was renamed mgp3-2. To confirm further that the phenotypes of mgp3-1/+ and mgp3-2/+ are caused by disruption of RPOTm, a 7.2 kb genomic DNA fragment of At1g68990, including the 1,147 bp promoter region, coding region and the 840 bp transcription termination region, was introduced into mgp3-1/+ and mgp3-2/+ plants, respectively. In the case of mgp3-1/+, 45 independent transformants were obtained. All these transformants were confirmed by PCR using the pCAMBIA1300-specific primer 1300-H and the RPOTm gene-specific primer P2 (Fig. 4B). The T2 seedlings from these transformants showed a segregation ratio of 2 KanR to 1 KanS (n > 500), indicating that the defective male and female gametophyte transmission efficiencies had been restored by the At1g68990 gene. Furthermore, mgp3-1 homozygous plants harboring the homozygous transgene could be identified in the T2 generation. In these plants, seed setting and pollen tube growth were as normal as in the wild type (Figs. 1K, 2C; see also Supplementary Fig. S2C). In the case of mgp3-2/+, 25 independent transformants were obtained, whose phenotypes were all recovered as in the complemented mgp3-1/+ plants. All these data indicated that the 7.2 kb DNA fragment could complement the phenotype of mgp3-1/+ and mgp3-2/+ plants, and showed that both mgp3-1 and mgp3-2 were loss-of- function mutations in RPOTm. Mitochondrial structure in mgp3-1 pollen grains was slightly affected The embryo lethal phenotype displayed by the mgp3 mutants and results of previous studies strongly support that RPOTm is the principal RNA polymerase that transcribes mitochondrial genes (Baba et al. 2004, Kühn et al. 2007). However, the embryo lethal phenotype of mgp3/– makes it difficult to detect the difference in mitochondrial gene expression between mutant and wild-type plants. We therefore examined whether the mgp3 mutations had an impact on the distribution and number of mitochondria. The result showed no significant change in the distribution and number of mitochondria between wild-type and mgp3-1 mutant pollen grains. We then examined the mitochondrial ultrastructure in mgp3-1/+ and wild-type pollen grains using transmission electron microscopy (TEM). Thirty-five individual wild-type Ler pollen grains (109 images) were examined. Their mitochondria all appeared as normal rod-like shapes (Fig. 5A, C). They contained an electron-dense matrix inside the outer membrane and well-developed inner membrane cristae (Fig. 5A, C). We then examined 41 individual pollen grains (127 images) from mgp3-1/+ plants, among which, theoretically, 50% of pollen grains were wild types and the other 50% were mgp3-1 mutant pollen grains. No significant alteration in mitochondrial morphology was observed in most of these pollen grains (Fig. 5B, D). Only in approximately 10% (4/41) of pollen grains did the mitochondria show a reduction in cristae that appeared with a curved or circular shape (Fig. 5E, G). This type of mitochondrion was never observed in wild-type pollen grains. We further examined mitochondrial structures in the RPOTm-complemented mgp3-1 pollen grains; no defect in mitochondrial structure was found in 117 images taken from 39 individual pollen grains (Fig. 5F, H). These results suggested that the defective mitochondrial structures in the mgp3-1/+ pollen grains were correlated to mgp3-1 mutation. Furthermore, such a type of abnormal mitochondria was similar to those in cultured ρ0 human cells, whose mtDNAs were fully depleted (Gilkerson et al. 2000). Fig. 5 View largeDownload slide TEM images of the mitochondria in wild-type and mgp3-1 pollen grains. (A, C) Mitochondria in wild-type pollen grains. (B, D) Normal mitochondria in most pollen grains from mgp3-1/+ plants. (E, G) Defective mitochondria in some pollen grains from mgp3-1/+ plants. (F, H) Mitochondria in RPOTm-complemented mgp3-1 pollen grains. The white arrows indicate the mitochondrial cristae. Bars = 500 nm. Fig. 5 View largeDownload slide TEM images of the mitochondria in wild-type and mgp3-1 pollen grains. (A, C) Mitochondria in wild-type pollen grains. (B, D) Normal mitochondria in most pollen grains from mgp3-1/+ plants. (E, G) Defective mitochondria in some pollen grains from mgp3-1/+ plants. (F, H) Mitochondria in RPOTm-complemented mgp3-1 pollen grains. The white arrows indicate the mitochondrial cristae. Bars = 500 nm. mgp3-2 rpoTmp-2 double mutation enhanced the mgp3 mutant phenotype The rpoTmp-2 mutant (SALK_132842) was obtained from the ABRC. It was in the Col-0 background, in which the T-DNA was inserted in the third intron of RPOTmp, at a site 1,731 bp downstream of the ATG codon (Supplementary Fig. S3A). The plants homozygous for the T-DNA insertion were identified by PCR (Supplementary Fig. S3B). The RPOTmp transcripts were not detected by RT–PCR in the mutant plants (Supplementary Fig. S3C). The mutant plants also showed a great reduction in plant growth as reported in the known rpoTmp mutants, 833 and 286E07 (Baba et al. 2004) (Supplementary Fig. S3D). These data indicated that rpoTmp-2 was a null rpoTmp mutant. Genetic analysis showed that the rpoTmp-2 mutant gametophytes were functionally normal (Table 3). Table 3 Genetic analysis of the mgp3-2/+; rpoTmp-2/+ double mutant by comparison with mgp3-2/+ and rpoTmp-2/+ single mutants Crosses (female × male)  Progeny   TEF (%)  TEM (%)  WT  rpoTmp-2/+  mgp3-2/+  mgp3-2/+; rpoTmp-2/+  mgp3-2/+ × +/+  209    117    56  NA  +/+ × mgp3-2/+  646    126    NA  20  rpoTmp-2/+ × +/+  130  121      93  NA  +/+ × rpoTmp-2/+  332  325      NA  97  mgp3-2/+; rpoTmp-2/+ × +/+  120  111  63  0  0  NA  +/+ × mgp3-2/+; rpoTmp-2/+  108  92  15  3  NA  3  Crosses (female × male)  Progeny   TEF (%)  TEM (%)  WT  rpoTmp-2/+  mgp3-2/+  mgp3-2/+; rpoTmp-2/+  mgp3-2/+ × +/+  209    117    56  NA  +/+ × mgp3-2/+  646    126    NA  20  rpoTmp-2/+ × +/+  130  121      93  NA  +/+ × rpoTmp-2/+  332  325      NA  97  mgp3-2/+; rpoTmp-2/+ × +/+  120  111  63  0  0  NA  +/+ × mgp3-2/+; rpoTmp-2/+  108  92  15  3  NA  3  NA, not applicable; TE, transmission efficiency. For the mgp3-2/+ single mutant and rpoTmp-2/+ single mutant, the TE = (No. of progeny plants with T-DNA insertion/No. of the progeny plants without T-DNA insertion) × 100%; for the mgp3-2/+; rpoTmp-2/+ double mutant, the TE = (No. of progeny plants with double T-DNA insertions/No. of progeny plants without T-DNA insertion) × 100%. TEF, female transmission efficiency; TEM, male transmission efficiency. View Large We used the RPOTm mutant allele mgp3-2/+ in the Col-0 background to cross with rpoTmp-2/+ and identified the plants heterozygous for both mgp3-2 and rpoTmp-2 mutations (mgp3-2/+; rpoTmp-2/+) among the F1 progeny. In total, 69 out of 462 F2 plants were identified as having short root phenotypes as described in the rpoTmp mutant 833 (Baba et al. 2004). All of them were transferred to soil and then analyzed by PCR. The result showed that all 69 plants were homozygous for the rpoTmp-2 mutation, but none of them possessed the mgp3-2 mutation, suggesting that the mgp3-2 rpoTmp-2 double mutation could not be transmitted to the next generation through gametophytes. To confirm this observation, the mgp3-2/+; rpoTmp-2/+ mutant plants were out-crossed to wild-type Col-0 plants. The karyotyping analysis of the F1 progeny by PCR indicated that the female transmission efficiency of the mgp3-2 rpoTmp-2 double mutation was zero (Table 3) and the male transmission efficiency was 3% (Table 3). These data showed that all the female and nearly all the male gametophytes bearing the mgp3-2 rpoTmp-2 double mutation lost their ability for transmission to next generation. To understand in detail how the mgp3-2 rpoTmp-2 double mutation impaired male gametophyte development, we studied the pollen phenotypes in mgp3-2/+; rpoTmp-2/+ plants by comparison with those in the mgp3-2/+ single mutant, the rpoTmp-2/+ single mutant and wild-type plants. We first studied the pollen morphology at the anthesis stage. In mgp3-2/+; rpoTmp-2/+ plants, 14% (109/769) of pollen grains were aborted (Fig. 6A) before maturation. Such defective pollen grains were not or rarely found in wild-type, mgp3-2/+ single mutant and rpoTmp-2/+ single mutant plants (Fig. 6B–D), indicating that the abortion of pollen grain development was caused by the mgp3-2 rpoTmp-2 double mutation. We also analyzed the viability, mitosis and ultrastructure of the aborted pollen grains, and found that they lacked cellular contents (Fig. 6F, H, J, L). These observations indicated that the mgp3-2 rpoTmp-2 double mutation affected pollen formation. The genetic data mentioned above showed that the rate of aborted pollen grains was significantly lower than the predicted rate (25%) of double mutant pollen grains. They implied that a proportion of the double mutant microspores could develop into morphologically normal pollen grains. Therefore, we examined the in vitro germination of the pollen grains from the double mutant plants (Fig. 7A–D). The pollen germination rate in the mgp3-2/+; rpoTmp-2/+ mutant was calculated and compared with those of the pollen grains from wild-type Col-0, mgp3-2/+ single mutant and rpoTmp-2/+ single mutant plants (Fig. 7E). The germination rate of pollen grains in mgp3-2/+; rpoTmp-2/+ mutants was found to be approximately 69% (597/865), while that in mgp3-2/+ single mutant, rpoTmp-2/+ single mutant and wild-type plants was approximately 85% (Fig. 7E). These observations indicated that the rpoTm rpoTmp double mutation severely impaired pollen formation and germination. Fig. 6 View largeDownload slide The mgp3-2 rpoTmp-2 double mutation affected pollen formation. (A–D) SEM images of pollen grains from mgp3-2/+; rpoTmp-2/+ (A), wild-type Col-0 (B), rpoTmp-2/+ single mutant (C) and mgp3-2/+ single mutant (D) plants. (E, F) Alexander staining images of wild-type (E) and aborted (F) pollen grains. (G, K) Light-field image (G) and DAPI staining image (K) of wild-type matured pollen grains. (H, L) Light-field (H) image and DAPI stain (L) of the pollen grains from a mgp3-2/+; rpoTmp-2/+ plant. (I, J) TEM images of wild-type (I) and aborted (J) mutant pollen grains. The white arrows indicate the defective pollen grains. Bars = 50 μm in A–D and 10 μm in E–L. Fig. 6 View largeDownload slide The mgp3-2 rpoTmp-2 double mutation affected pollen formation. (A–D) SEM images of pollen grains from mgp3-2/+; rpoTmp-2/+ (A), wild-type Col-0 (B), rpoTmp-2/+ single mutant (C) and mgp3-2/+ single mutant (D) plants. (E, F) Alexander staining images of wild-type (E) and aborted (F) pollen grains. (G, K) Light-field image (G) and DAPI staining image (K) of wild-type matured pollen grains. (H, L) Light-field (H) image and DAPI stain (L) of the pollen grains from a mgp3-2/+; rpoTmp-2/+ plant. (I, J) TEM images of wild-type (I) and aborted (J) mutant pollen grains. The white arrows indicate the defective pollen grains. Bars = 50 μm in A–D and 10 μm in E–L. Fig. 7 View largeDownload slide mgp3-2 rpoTmp-2 double mutation affected pollen germination. (A–D) In vitro germination of pollen grains from wild-type Col-0 (A), rpoTmp-2 single mutant (B), mgp3-2/+ single mutant (C) and mgp3-2/+; rpoTmp-2/+ double mutant (D) plants. (E) Statistical analysis of the pollen germination rates in wild-type Col-0, the rpoTmp-2/+ single mutant, the mgp3-2/+ single mutant and the mgp3-2/+; rpoTmp-2/+ double mutant. The white arrows indicate the aborted pollen grains. Bars = 100 μm in A–D. Fig. 7 View largeDownload slide mgp3-2 rpoTmp-2 double mutation affected pollen germination. (A–D) In vitro germination of pollen grains from wild-type Col-0 (A), rpoTmp-2 single mutant (B), mgp3-2/+ single mutant (C) and mgp3-2/+; rpoTmp-2/+ double mutant (D) plants. (E) Statistical analysis of the pollen germination rates in wild-type Col-0, the rpoTmp-2/+ single mutant, the mgp3-2/+ single mutant and the mgp3-2/+; rpoTmp-2/+ double mutant. The white arrows indicate the aborted pollen grains. Bars = 100 μm in A–D. We also observed the female gametophyte development process in mgp3-2/+; rpoTmp-2/+ pistils by comparison with that in mgp3-2/+ single mutant, rpoTmp-2/+ single mutant and wild-type plants. We found that 36.5% embryo sacs aborted in mgp3-2/+; rpoTmp-2/+ mutant siliques, 18.5% in mgp3-2/+ siliques and 4% in rpoTmp-2/+ and wild-type siliques (Fig. 8A–D, Table 4). This phenotype was also observed when wild-type plants were used as pollen donors, indicating that the aborted embryo sacs were caused by the mutations rather than environmental or paternal factors. The proportion of the aborted embryo sacs in mgp3-2/+; rpoTmp-2/+ double mutant siliques was correlated to the genetic analysis data. No embryo was found in any aborted embryo sacs, suggesting that the embryo sacs aborted at least before embryogenesis took place. We then observed the developmental process of the embryo sacs by CLSM (Fig. 8I–N). At the mature stage before fertilization, all embryo sacs in rpoTmp-2/+ plants could develop normally as in the wild-type embryo sacs (Fig. 3C). In contrast, in the mgp3-2/+ siliques, in 21.8% (61/279) of ovules the two polar nuclei failed to fuse. This phenotype was similar to that observed in mgp3-1/+ mutant female gametophytes (Fig. 3G). In mgp3-2/+; rpoTmp-2/+ siliques, 26.7% (69/258) of ovules showed failure of fusion of the polar nuclei like those in the mgp3 embryo sacs (Fig. 3G). The number of aborted embryo sacs bearing unfused polar nuclei was less than the predicted 35%. We then analyzed the embryo sac development process after fertilization. To do this, we pollinated wild-type pollen to the mutant pistils. At 18 h after pollination, the fertilized egg cell and central cell could develop normally in wild-type and rpoTmp-2 embryo sacs (Fig. 8I). In the mgp3-2/+ single mutant and the mgp3-2/+; rpoTmp-2/+ double mutant, the embryo sacs bearing two unfused polar nuclei failed to develop further as observed in mgp3-1 mutant embryo sacs (Fig. 3H). Furthermore, in the mgp3-2/+; rpoTmp-2/+ double mutant, 12% of ovules that had one central cell, one egg cell and two synergid cell nuclei also failed to develop further (Fig. 8L). No embryo and endosperm nucleus were found in the defective ovules, and the two synergid cells appeared intact (Fig. 8L). No defect in pollen tube attraction was observed in the mgp3-2/+; rpoTmp-2/+ pistil by aniline blue staining (Fig. 8E–H). At 72 h after pollination, the defective embryo sacs stopped growing and eventually died (Fig. 8N). These observations suggested that the double mutation impaired not only fusion of polar nuclei, but also synergid cell death. Table 4 Statistical analysis of seed setting in the siliques from mgp3-2/+; rpoTmp-2/+ double mutant plants by comparison with mgp3-2/+, rpoTmp-2/+ and wild-type plants Genotypes of the plants  Total seeds and ovules  Normal seeds   Aborted seeds   Aborted ovules   n  %  n  %  n  %  Wild-type Col-0  2,120  2,049  96.6  0  0  71  3.4  rpoTmp-2/+  1,979  1,898  95.9  0  0  81  4.1  mgp3-2/+  2,132  1,515  71.1  223  10.5  394  18.5  mgp3-2/+; rpoTmp-2/+  1,778  1,091  61.4  39  2.2  648  36.5  Genotypes of the plants  Total seeds and ovules  Normal seeds   Aborted seeds   Aborted ovules   n  %  n  %  n  %  Wild-type Col-0  2,120  2,049  96.6  0  0  71  3.4  rpoTmp-2/+  1,979  1,898  95.9  0  0  81  4.1  mgp3-2/+  2,132  1,515  71.1  223  10.5  394  18.5  mgp3-2/+; rpoTmp-2/+  1,778  1,091  61.4  39  2.2  648  36.5  The seeds were counted on the first eight siliques from five independent wild-type or mutant plants. View Large Fig. 8 View largeDownload slide mgp3-2 rpoTmp-2 double mutation affected embryo sac development. (A–D) Seed setting in the siliques from wild-type Col-0 (A), rpoTmp-2/+ (B), mgp3-2/+ (C) and mgp3-2/+; rpoTmp-2/+ (D) plants. (E, F) Aniline blue-stained siliques from wild-type Col-0 (E) and mgp3-2/+; rpoTmp-2/+ double mutant (F) plants. (G, H) Pollen tubes entered the wild-type embryo sac (G) and aborted embryo sacs in the mgp3-2/+; rpoTmp-2/+ double mutant (H). (I–K) The wild type embryo sacs at 18 (I), 36 (J) and 72 h (K) after pollination. (L–N) The abnormal embryo sacs found in the siliques from mgp3-2/+; rpoTmp-2/+ plants at 18 (L), 36 (M) and 72 h (N) after pollination. The arrows in A–D indicate the aborted seeds. The asterisks in A–F indicate the aborted ovules. Ch, chalazal end; CN, central cell nucleus; Dcn, degenerated central cell nucleus; Dsc, degenerated synergid cell nucleus; EdN, endosperm nucleus; Emb, embryo; EN, egg cell nucleus; F, funiculus; PT, pollen tube; SN, synergid cell nucleus; Zyg, zygote. Bars = 1 mm in A–D, 0.5 mm in E, F and 10 μm in G–N. Fig. 8 View largeDownload slide mgp3-2 rpoTmp-2 double mutation affected embryo sac development. (A–D) Seed setting in the siliques from wild-type Col-0 (A), rpoTmp-2/+ (B), mgp3-2/+ (C) and mgp3-2/+; rpoTmp-2/+ (D) plants. (E, F) Aniline blue-stained siliques from wild-type Col-0 (E) and mgp3-2/+; rpoTmp-2/+ double mutant (F) plants. (G, H) Pollen tubes entered the wild-type embryo sac (G) and aborted embryo sacs in the mgp3-2/+; rpoTmp-2/+ double mutant (H). (I–K) The wild type embryo sacs at 18 (I), 36 (J) and 72 h (K) after pollination. (L–N) The abnormal embryo sacs found in the siliques from mgp3-2/+; rpoTmp-2/+ plants at 18 (L), 36 (M) and 72 h (N) after pollination. The arrows in A–D indicate the aborted seeds. The asterisks in A–F indicate the aborted ovules. Ch, chalazal end; CN, central cell nucleus; Dcn, degenerated central cell nucleus; Dsc, degenerated synergid cell nucleus; EdN, endosperm nucleus; Emb, embryo; EN, egg cell nucleus; F, funiculus; PT, pollen tube; SN, synergid cell nucleus; Zyg, zygote. Bars = 1 mm in A–D, 0.5 mm in E, F and 10 μm in G–N. The PRPOTm::RPOTmp cDNA construct cannot complement the phenotypes of the mgp3 mutants The rpoTm rpoTmp double mutation enhanced the phenotype of the rpoTm mutant, implying that RPOTm and RPOTmp may have partially overlapping roles in plant development. To address this question, we examined whether the phenotypes of mgp3 mutants could be complemented by the expression of the RPOTmp gene under control of the RPOTm promoter. To do this, we constructed a PRPOTm::RPOTmp cDNA construct and introduced it into mgp3-1/+ and mgp3-2/+ plants. More than 20 independent transformants were obtained in each mutation background. All the T2 seedlings from these transformants showed a segregation ratio of roughly 1 KanR to 1 KanS (n > 500), indicating that the construct could not restore the defective male/female gametophyte transmission efficiencies and the embryo lethal phenotypes. Thus, RPOTmp cannot replace RPOTm in gametophyte and embryo development. Discussion RPOTm plays important roles in gametophyte and embryo development Here we have studied the genetic roles of RPOTm in gametogenesis and embryogenesis by characterization of the RPOTm loss-of-function mutations mgp3-1 and mgp3-2 in Arabidopsis. Our results provide new insight into the roles of RPOTm in plant development. First, our results showed that the mutations in RPOTm did not affect pollen formation, but significantly retarded pollen tube growth. Mitochondrial genes encode the proteins, rRNAs and tRNAs that function in the complexes of respiratory chain, heme and cytochrome assembly, and the ribosome (Unseld et al. 1997, Binder and Brennicke 2003). Loss of mitochondrial RNA polymerase could affect the expression of these genes and lead to defects in ATP synthesis. Consequently, it reduces the energy supply. A striking feature of the pollen tube growth is the marked accumulation of secretion vesicles that contain the components for cell wall expansion (Taylor and Hepler 1997, Krichevsky et al. 2007). Vesicular trafficking and cell wall synthesis are the energy- and metabolite-dependent cellular processes. Shortage of energy supply could affect vesicular trafficking and other metabolic processes. As a result, pollen tube growth would be retarded. Moreover, the genetic analysis showed that at least 31% of rpoTm mutant male gametophytes could be successful in fertilization. All these results indicate that rpoTm is not essential for cell survival, but is important for pollen tube growth. Secondly, the mgp3 mutations also affected fusion of polar nuclei during female gametophyte development. Previous studies have shown that the mutations in the genes encoding mitochondrial proteins in a number of mutants, such as nfd1, nfd3, nfd4, nfd5, nfd6 (Portereiko et al. 2006), gfa2 (Christensen et al. 2002) and sdh1 (León et al. 2007), caused defects in fusion of polar nuclei. Our result reported here provides further evidence to demonstrate that the karyogamy defect is a consequence of impairment in basic mitochondrial functions. However, how mitochondria are involved in karyogamy still remains unclear. Thirdly, our results demonstrated that the rpoTm homozygous mutant embryos stopped further development at the globular stage, indicating that RPOTm was critical for embryo development from the globular stage to the heart stage. It is not clear why the mgp3/– embryos were arrested at this stage. A recent study revealed that during the transition from globular to heart stages, the expression of many genes that were involved in energy metabolism was up-regulated, which comprised 21% of the up-regulated genes (Spencer et al. 2007). This indicates that energy metabolism is very active during these stages. Nevertheless, the embryogenesis at these stages involves active cell division, cell differentiation and growth. All these processes need to consume a lot of energy. Because RPOTm is involved in the expression of mitochondrial genes, the mgp3 mutations may drastically affect the expression of mitochondrial genes including those involved in energy metabolism. Thus, the mgp3 mutant embryos stopped further development, possibly due to shortage of energy supply. The distinct genetic roles of mitochondrial phage-type RNA polymerases As discussed above, the mgp3 mutations caused embryo lethality and partially impaired female gametophyte development, but did not affect pollen formation and pollen germination, while the rpoTmp mutations reduced plant size, but did not affect male and female gametophyte and embryo development. The rpoTm rpoTmp double mutation disrupted female gametophyte development completely and more severely affected male gametophyte development. These results indicate that RPOTm is important and sufficient for normal gametophyte and embryo development. At the molecular level, previous studies have indicated that RPOTm was able to initiate transcription from overlapping subsets of mitochondrial promoters accurately, while RPOTmp did not display any significant promoter specificity (Kühn et al. 2007). A recent study showed that RPOTmp performed gene-specific transcription in mitochondria; it transcribed at least rps4, nad2, nad6, cox1 and ccmC genes. Levels of respiratory chain complexes I and IV were down-regulated in the rpoTmp mutants (Kühn et al. 2009). Taken all together, the results suggest that RPOTm is the major RNA polymerase in mitochondria. It can transcribe most, if not all, mitochondrial genes, which is sufficient for normal gametophyte and embryo development. RPOTmp transcribes only a subset of mitochondrial genes, e.g. rps4, nad2, nad6, cox1 and ccmC. It can partially rescue the loss of RPOTm function in the rpoTm mutant gametophyte, but is not sufficient to support normal gametophyte and embryo development. When these two RPOT genes were knocked out, probably most or all mitochondrial genes could not be transcribed and then gametophyte development was disrupted completely. Materials and Methods Plant material and mutant isolation The mgp3-1 mutant is in the Landsberg erecta (Ler) background, while mgp3-2 (SALK_005875) and rpoTmp-2(SALK_132842) are in the Columbia-0 background. The plants were grown in an air-conditioned room at 22°C under 16 h light/8 h dark cycles. Seedling growth in vitro was performed as described previously by Yang et al. (2009). The generation of Ds insertion lines and screening of mutants were performed as described by Sundaresan et al. (1995). The mgp3-2 (SALK_005875) and rpoTmp-2 (SALK_132842) seeds were obtained from the ABRC. The T-DNA insertion sites in mgp3-2 and rpoTmp-2 were determined by PCR using primer pairs LBa1 (5′-TGGTTCACGTAGTGGGCCATCG-3′)/P2 (5′-CATTACGCACAGGCTCAAACC-3′) and LBa1/P4 (5′-CGA CCATTTGAGAGGAG-3′), respectively. The rpoTmp-2 homozygous plant was identified by PCR using primer pair P3 (5′-GAATATGCACATTCTTGG-3′)/P4. Phenotypic characterization The developing seeds/embryos were clarified overnight in Hoyer’s solution (Liu and Meinke 1998) and observed with a Leica DMRA2 microscope equipped with DIC system (Leica, Wetzlar, Germany). Morphological observations of pollen grains by SEM were performed as described by Jiang et al. (2005). DAPI staining of pollen grains was performed as described by Yang et al. (2009), and pollen viability was examined using Alexander staining (Alexander 1969). Pollen germination in vitro was performed as described by Jiang et al. (2005). The germinating pollen grains were counted under a microscope after being incubated at 22–23°C for 6 h. From each culture, at least 300 pollen grains were examined to calculate the average germination rates and pollen tube length for each sample. The female gametophytes were characterized by CLSM as described by Christensen et al. (1997) and Shi et al. (2005) with the following modifications. The pistils were collected in 4% glutaraldehyde (in 12.5 mM cacodylate, pH 6.9), kept under a vacuum for 1–2 h, followed by further fixation overnight at room temperature. Then, the tissues were dehydrated through an ethanol series [10, 30, 50, 70, 90, 95, 100, 100 and 100% (v/v)] with 30 min per step. The dehydrated tissues were clarified in 2 : 1 (v/v) benzyl benzoate : benzyl alcohol for at least 1 h and then observed with a Zeiss LSM510 META laser scanning microscope (Zeiss, Jena, Germany) with a 488 nm argon laser and an LP 530 filter. Transmission electron microscopy The samples were fixed in fixing solution containing 5% (v/v) glutaraldehyde, 0.1 M sodium cacodylate, pH 7.2, post-fixed in 1% osmium tetroxide for 1 h, dehydrated in a series of acetone solutions [25, 50, 75 and 100% (v/v), followed by a resin/ acetone dilution series [25, 50 and 75% (v/v)], and embedded in epoxy resin. After polymerization at 65°C for 18 h, serial cross-sections were prepared and examined with a JEM1230 transmission electron microscope (Hitachi, Tokyo, Japan). Genetic analysis of the mgp3 mutants All crosses and genetic analysis were performed as described previously (Yang et al. 1999, Yang et al. 2003). Molecular cloning of MGP3 and complementation experiments Isolation of the flanking sequences adjacent to the Ds element by TAIL-PCR (Liu et al. 1995) was performed as described previously (Yang et al. 1999, Yang et al. 2003) with mgp3-1/+ genomic DNA and the Ds5′/AD6 primer set (Grossniklaus et al.1998). The insertion site was then confirmed by PCR using primer pairs P2/Ds5-2 (5′-CGTTCCGTTTTCGTTTTTTACC-3′) and P1 (5′-TGGAGGAACATTCTGGG-3′)/Ds3-2 (5′-CCGGTATATCC CGTTTTCG-3′). The full-length MGP3 genomic DNA fragment was amplified by PCR with the primers MGP3-F-P1 (5′-GCAATGAACAG TAGTGTTGG-3′) and MGP3-F-P2 (5′-GCGGATGCGTTCTTG AATGC-3′). The resulting DNA fragment was cloned into a pMD18-T vector (TaKaRa, Dalian, China) for sequencing. For complementation experiments, the full-length MGP3 genomic DNA fragment was subcloned into the pCAMBIA1300 vector (CAMBIA, Canberra, Australia) and introduced into mgp3-1 and mgp3-2 heterozygous plants using the Agrobacterium- mediated infiltration method. The transformants were selected in Murashige and Skoog (MS) medium (Murashige and Skoog 1962) containing 25 mg l−1 of hygromycin (Roche, Mannheim, Germany) and 50 mg l−1 of kanamycin (Sigma, Saint Louis, Missouri, USA). All the transformants were confirmed by PCR using the primer pair 1300-H (5′-TGGCGAAAGGGGGA TGTGCTG-3′)/P2. The transformants homozygous for mgp3-1 or mgp3-2 mutation were selected in T2 generations and used for further analyses. Southern blotting hybridization The DNA preparation and Southern blotting hybridization were as described by Jia et al. (2009) and Yang et al. (1999). Funding This work was supported by the Ministry of Sciences and Technology [973 project number: 2007CB108700]; the Natural Science Foundation of China [project number: 30530060]; the Ministry of Education, PR of China [111 project number B06003]. Acknowledgments We thank Dr. V. Sundaresan, Dr. Wei-Cai Yang and Mrs. Li-Fen Xie for their kind help with the mutant screens. References Alexander MP.  Differential staining of aborted and nonaborted pollen,  Stain Technol. ,  1969, vol.  44 (pg.  117- 122) Google Scholar CrossRef Search ADS PubMed  Baba K,  Schmidt J,  Espinosa-Ruiz A,  Villarejo A,  Shiina T,  Gardeström P, et al.  Organellar gene transcription and early seedling development are affected in the rpoT;2 mutant of Arabidopsis,  Plant J. ,  2004, vol.  38 (pg.  38- 48) Google Scholar CrossRef Search ADS PubMed  Binder S,  Brennicke A.  Gene expression in plant mitochondria: transcriptional and post-transcriptional control,  Philos. Trans. R. Soc. Ser. B: Biol. Sci. ,  2003, vol.  358 (pg.  181- 189) Google Scholar CrossRef Search ADS   Carlsson J,  Leino M,  Sohlberg J,  Sundström JF,  Glimelius K.  Mitochondrial regulation of flower development,  Mitochondrion ,  2008, vol.  8 (pg.  74- 86) Google Scholar CrossRef Search ADS PubMed  Cermakian N,  Ikeda TM,  Cedergren R,  Gray MW.  Sequences homologous to yeast mitochondrial and bacteriophage T3 and T7 RNA polymerases are widespread throughout the eukaryotic lineage,  Nucleic Acids Res. ,  1996, vol.  24 (pg.  648- 654) Google Scholar CrossRef Search ADS PubMed  Christensen CA,  King EJ,  Jordan JR,  Drews GN.  Megagametogenesis in Arabidopsis wild type and the gf mutant,  Sex. Plant Reprod. ,  1997, vol.  10 (pg.  49- 64) Google Scholar CrossRef Search ADS   Christensen CA,  Gorsich SW,  Brown RH,  Jones LG,  Brown J,  Shaw JM, et al.  Mitochondrial GFA2 is required for synergid cell death in Arabidopsis,  Plant Cell ,  2002, vol.  14 (pg.  2215- 2232) Google Scholar CrossRef Search ADS PubMed  Conley CA,  Hanson MR.  How do alterations in plant mitochondrial genomes disrupt pollen development?,  J. Bioenerg. Biomembr. ,  1995, vol.  27 (pg.  447- 457) Google Scholar CrossRef Search ADS PubMed  Drews GN,  Lee D,  Christensen CA.  Genetic analysis of female gametophyte development and function,  Plant Cell ,  1998, vol.  10 (pg.  5- 17) Google Scholar CrossRef Search ADS PubMed  Emanuel C,  von Groll U,  Müller M,  Börner T,  Weihe A.  Development- and tissue-specific expression of the RpoT gene family of Arabidopsis encoding mitochondrial and plastid RNA polymerases,  Planta ,  2006, vol.  223 (pg.  998- 1009) Google Scholar CrossRef Search ADS PubMed  Gilkerson RW,  Margineantu DH,  Capaldi RA,  Selker JML.  Mitochondrial DNA depletion causes morphological changes in the mitochondrial reticulum of cultured human cells,  FEBS Letters ,  2000, vol.  474 (pg.  1- 4) Google Scholar CrossRef Search ADS PubMed  Greenleaf AL,  Kelly JL,  Lehman IR.  Yeast RP041 gene product is required for transcription and maintenance of the mitochondrial genome,  Proc. Natl Acad. Sci. USA ,  1986, vol.  83 (pg.  3391- 3394) Google Scholar CrossRef Search ADS   Grossniklaus U,  Vielle-Calzada JP,  Hoeppner MA,  Gagliano WB.  Maternal control of embryogenesis by MEDEA, a polycomb group gene in Arabidopsis,  Science ,  1998, vol.  280 (pg.  446- 450) Google Scholar CrossRef Search ADS PubMed  Hanson MR,  Bentolila S.  Interactions of mitochondrial and nuclear genes that affect male gametophyte development,  Plant Cell ,  2004, vol.  16 (pg.  S154- S169) Google Scholar CrossRef Search ADS PubMed  Hedtke B,  Börner T,  Weihe A.  Mitochondrial and chloroplast phage-type RNA polymerases in Arabidopsis,  Science ,  1997, vol.  277 (pg.  809- 811) Google Scholar CrossRef Search ADS PubMed  Hedtke B,  Meixner M,  Gillandt S,  Richter E,  Börner T,  Weihe A.  Green fluorescent protein as a marker to investigate targeting of organellar RNA polymerases of higher plants in vivo,  Plant J. ,  1999, vol.  17 (pg.  557- 561) Google Scholar CrossRef Search ADS PubMed  Hedtke B,  Börner T,  Weihe A.  One RNA polymerase serving two genomes,  EMBO Rep. ,  2000, vol.  1 (pg.  435- 440) Google Scholar CrossRef Search ADS PubMed  Huang J,  Struck F,  Matzinger DF,  Levings CS,  III.  Flower-enhanced expression of a nuclear-encoded mitochondrial respiratory protein is associated with changes in mitochondrion number,  Plant Cell ,  1994, vol.  6 (pg.  439- 448) Google Scholar CrossRef Search ADS PubMed  Jia DJ,  Cao X,  Wang W,  Tan XY,  Zhang XQ,  Chen LQ, et al.  GNOM-LIKE 2, encoding an ARF-GEF protein homologous to GNOM and GNL1, is essential for pollen tube growth in Arabidopsis,  J. Integr. Plant Biol. ,  2009, vol.  51 (pg.  762- 773) Google Scholar CrossRef Search ADS PubMed  Jiang L,  Yang SL,  Xie LF,  Puah CS,  Zhang XQ,  Yang WC, et al.  VANGUARD1 encodes a pectin methylesterase that enhances pollen tube growth in the Arabidopsis style and transmitting tract,  Plant Cell ,  2005, vol.  17 (pg.  584- 596) Google Scholar CrossRef Search ADS PubMed  Kelly JL,  Greenleaf AL,  Lehman IR.  Isolation of the nuclear gene encoding a subunit of the yeast mitochondrial RNA polymerase,  J. Biol. Chem. ,  1986, vol.  261 (pg.  10348- 10351) Google Scholar PubMed  Krichevsky A,  Kozlovsky SV,  Tian GW,  Chen MH,  Zaltsman A,  Citovsky V.  How pollen tubes grow,  Dev. Biol. ,  2007, vol.  303 (pg.  405- 420) Google Scholar CrossRef Search ADS PubMed  Kühn K,  Bohne AV,  Liere K,  Weihe A,  Börner T.  Arabidopsis phage-type RNA polymerases: accurate in vitro transcription of organellar genes,  Plant Cell ,  2007, vol.  19 (pg.  959- 971) Google Scholar CrossRef Search ADS PubMed  Kühn K,  Richter U,  Meyer EH,  Delannoy E,  Falcon de Longevialle A,  O’Toole N, et al.  Phage-type RNA polymerase RPOTmp performs gene-specific transcription in mitochondria of Arabidopsis thaliana,  Plant Cell ,  2009, vol.  21 (pg.  2762- 2779) Google Scholar CrossRef Search ADS PubMed  León G,  Holuigue L,  Jordana X.  Mitochondrial complex II is essential for gametophyte development in Arabidopsis,  Plant Physiol. ,  2007, vol.  143 (pg.  1534- 1546) Google Scholar CrossRef Search ADS PubMed  Linke B,  Börner T.  Mitochondrial effects on flower and pollen development,  Mitochondrion ,  2005, vol.  5 (pg.  389- 402) Google Scholar CrossRef Search ADS PubMed  Liu CM,  Meinke DW.  The titan mutants of Arabidopsis are disrupted in mitosis and cell cycle control during seed development,  Plant J. ,  1998, vol.  16 (pg.  21- 31) Google Scholar CrossRef Search ADS PubMed  Liu Y.-G,  Mitsukawa N,  Oosumi T,  Whittier RF.  Efficient isolation and mapping of Arabidopsis thaliana T-DNA insert junctions by thermal asymmetric interlaced PCR,  Plant J. ,  1995, vol.  8 (pg.  457- 463) Google Scholar CrossRef Search ADS PubMed  McCormick S.  Male gametophyte development,  Plant Cell ,  1993, vol.  5 (pg.  1265- 1275) Google Scholar CrossRef Search ADS PubMed  McCormick S.  Control of male gametophyte development,  Plant Cell ,  2004, vol.  16 (pg.  S142- 153) Google Scholar CrossRef Search ADS PubMed  Millar AH,  Day DA,  Whelan J.  Somerville CR,  Meyerowitz EM.  Mitochondrial biogenesis and function in Arabidopsis,  The Arabidopsis Book ,  2004 Rockville, MD American Society of Plant Biologists  doi/10.1199/tab.0105, http://www.aspb .org/publications/arabidopsis/ Monéger F,  Mandaron P,  Niogret MF,  Freyssinet G,  Mache R.  Expression of chloroplast and mitochondrial genes during microsporogenesis in maize,  Plant Physiol. ,  1992, vol.  99 (pg.  396- 400) Google Scholar CrossRef Search ADS PubMed  Murashige T,  Skoog F.  A revised medium for rapid growth and bioassays with tobacco tissue cultures,  Physiol. Plant. ,  1962, vol.  15 (pg.  473- 497) Google Scholar CrossRef Search ADS   Parton RM,  Fischer-Parton S,  Trewavas AJ,  Watahiki MK.  Pollen tubes exhibit regular periodic membrane trafficking events in the absence of apical extension,  J. Cell Sci. ,  2003, vol.  116 (pg.  2707- 2719) Google Scholar CrossRef Search ADS PubMed  Portereiko MF,  Sandaklie-Nikolova L,  Lloyd A,  Dever CA,  Otsuga D,  Drews GN.  NUCLEAR FUSION DEFECTIVE1 encodes the Arabidopsis RPL21M protein and is required for karyogamy during female gametophyte development and fertilization,  Plant Physiol. ,  2006, vol.  141 (pg.  957- 965) Google Scholar CrossRef Search ADS PubMed  Qin Y,  Leydon AR,  Manziello A,  Pandey R,  Mount D,  Denic S, et al.  Penetration of the stigma and style elicits a novel transcriptome in pollen tubes, pointing to genes critical for growth in a pistil,  PLoS Genet. ,  2009, vol.  5 pg.  e1000621  Google Scholar CrossRef Search ADS PubMed  Shi DQ,  Liu J,  Xiang YH,  Ye D,  Sundaresan V,  Yang WC.  SLOW WALKER1, essential for gametogenesis in Arabidopsis, encodes a WD40 protein involved in 18S ribosomal RNA biogenesis,  Plant Cell ,  2005, vol.  17 (pg.  2340- 2354) Google Scholar CrossRef Search ADS PubMed  Skinner DJ,  Baker SC,  Meister RJ,  Broadhvest J,  Schneitz K,  Gasser CS.  The Arabidopsis HUELLENLOS gene, which is essential for normal ovule development, encodes a mitochondrial ribosomal protein,  Plant Cell ,  2001, vol.  13 (pg.  2719- 2730) Google Scholar CrossRef Search ADS PubMed  Spencer MWB,  Casson SA,  Lindsey K.  Transcriptional profiling of the Arabidopsis embryo,  Plant Physiol. ,  2007, vol.  143 (pg.  924- 940) Google Scholar CrossRef Search ADS PubMed  Sundaresan V,  Springer P,  Volpe T,  Haward S,  Jones JDG,  Dean C, et al.  Patterns of gene action in plant development revealed by enhancer trap and gene trap transposable elements,  Genes Dev. ,  1995, vol.  9 (pg.  1797- 1810) Google Scholar CrossRef Search ADS PubMed  Taylor LP,  Hepler PK.  Pollen germination and tube growth,  Annu. Rev. Plant Physiol. Plant Mol. Biol. ,  1997, vol.  48 (pg.  461- 491) Google Scholar CrossRef Search ADS PubMed  Unseld M,  Marienfeld JR,  Brandt P,  Brennicke A.  The mitochondrial genome of Arabidopsis thaliana contains 57 genes in 366,924 nucleotides,  Nat. Genet. ,  1997, vol.  15 (pg.  57- 61) Google Scholar CrossRef Search ADS PubMed  Warmke HE,  Lee SLJ.  Pollen abortion in T cytoplasmic male-sterile corn (Zea mays): a suggested mechanism,  Science ,  1978, vol.  200 (pg.  561- 563) Google Scholar CrossRef Search ADS PubMed  Wen LY,  Chase CD.  Mitochondrial gene expression in developing male gametophytes of male-fertile and S male-sterile maize,  Sex. Plant Reprod. ,  1999, vol.  11 (pg.  323- 330) Google Scholar CrossRef Search ADS   Yadegari R,  Drews GN.  Female gametophyte development,  Plant Cell ,  2004, vol.  16 (pg.  S133- 141) Google Scholar CrossRef Search ADS PubMed  Yamaoka S,  Leaver CJ.  EMB2473/MIRO1, an Arabidopsis Miro GTPase, is required for embryogenesis and influences mitochondrial morphology in pollen,  Plant Cell ,  2008, vol.  20 (pg.  589- 601) Google Scholar CrossRef Search ADS PubMed  Yang KZ,  Xia C,  Liu XL,  Dou XY,  Wang W,  Chen LQ, et al.  A mutation in THERMOSENSITIVE MALE STERILE 1, encoding a heat shock protein with DnaJ and PDI domains, leads to thermosensitive gametophytic male sterility in Arabidopsis,  Plant J. ,  2009, vol.  57 (pg.  870- 882) Google Scholar CrossRef Search ADS PubMed  Yang SL,  Xie LF,  Mao HZ,  Puah CS,  Yang WC,  Jiang L, et al.  Tapetum determinant 1 is required for cell specialization in the Arabidopsis anther,  Plant Cell ,  2003, vol.  15 (pg.  2792- 2804) Google Scholar CrossRef Search ADS PubMed  Yang WC,  Ye D,  Xu J,  Sundaresan V.  The SPOROCYTELESS gene of Arabidopsis is required for initiation of sporogenesis and encodes a novel nuclear protein,  Genes Dev. ,  1999, vol.  13 (pg.  2108- 2117) Google Scholar CrossRef Search ADS PubMed  Abbreviations Abbreviations ABRC Arabidopsis Biological Resource Center CLSM confocal laser scanning microscopy DAPI 4′,6-diamidino-2-phenylindole DIC differential interference contrast microscopy GUS β-glucuronidase Kan kanamycin RPOT RNA polymerase of T3/T7 type RNAP RNA polymerase RT–PCR reverse transcriptase–PCR SEM scanning electron microscopy TAIL-PCR thermal asymmetric interlaced PCR TEM transmission electron microscopy. © The Author 2010. Published by Oxford University Press on behalf of Japanese Society of Plant Physiologists. All rights reserved. For permissions, please email: journals.permissions@oxfordjournals.org http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Plant and Cell Physiology Oxford University Press

Mutations in the Arabidopsis Nuclear-Encoded Mitochondrial Phage-Type RNA Polymerase Gene RPOTm Led to Defects in Pollen Tube Growth, Female Gametogenesis and Embryogenesis

Download PDF
15 pages

Loading next page...
 
/lp/oxford-university-press/mutations-in-the-arabidopsis-nuclear-encoded-mitochondrial-phage-type-IEtiQTREW1

References (60)

Publisher
Oxford University Press
Copyright
© The Author 2010. Published by Oxford University Press on behalf of Japanese Society of Plant Physiologists. All rights reserved. For permissions, please email: journals.permissions@oxfordjournals.org
ISSN
0032-0781
eISSN
1471-9053
DOI
10.1093/pcp/pcq029
pmid
20231244
Publisher site
See Article on Publisher Site

Abstract

Abstract The mitochondrial genes in Arabidopsis thaliana are transcribed by a small family of nuclear-encoded T3/T7 phage-type RNA polymerases (RPOTs). At least two nuclear-encoded RPOTs (RPOTm and RPOTmp) are located in mitochondria in A. thaliana. Their genetic roles are largely unknown. Here we report the characterization of novel mutations in the A. thaliana RPOTm gene. The mutations did not affect pollen formation, but significantly retarded the growth of the rpoTm mutant pollen tubes and had an impact on the fusion of the polar nuclei in the rpoTm mutant embryo sacs. Moreover, development of the rpoTm/– mutant embryo was arrested at the globular stage. The rpoTm rpoTmp double mutation could enhance the rpoTm mutant phenotype. Expression of RPOTmp under control of the RPOTm promoter could not complement the phenotype of the rpoTm mutations. All these data indicate that RPOTm is important for normal pollen tube growth, female gametogenesis and embryo development, and has distinct genetic and molecular roles in plant development, which cannot be replaced by RPOTmp. Introduction The plant life cycle alternates between a diploid sporophytic phase and a haploid gametophytic phase. The diploid male and female sporophytes produce haploid microspores and megaspores that give rise to male and female gametophytes, respectively. Specifically, in the anther, the diploid pollen mother cell (PMC) undergoes meiosis to generate haploid microspores. Then, each individual microspore undergoes two rounds of mitosis to form a three-celled pollen grain (male gametophyte) that comprises two sperm cells and a vegetative cell. When the mature pollen grain lands on the female stigma, shortly thereafter it germinates and produces a pollen tube that invades the stigmatic tissue and then elongates in the female transmitting tract to deliver the two sperm cells into an embryo sac (McCormick 1993, McCormick 2004). In the ovule, the megaspore mother cell (MMC) undergoes meiosis to give rise to four haploid megaspores, three of which undergo programmed cell death and only one survives and differentiates into a functional megaspore. The functional megaspore then undergoes three rounds of mitotic nuclear division to generate an eight-nucleus coenocytic embryo sac. The coenocytic embryo sac then goes through the fusion of polar nuclei and cytokinesis to form a seven-celled embryo sac (female gametophyte) that consists of four different cell types: three antipodal cells (n), one central cell (2n), one egg cell (n) and two synergid cells (n). Then, the antipodal cells undergo programmed cell death. When a pollen tube enters the embryo sac by penetrating one of the two synergid cells through the filiform apparatus, the synergid cell also undergoes cell death. Subsequently, the two sperm cells are discharged from the pollen tube and migrate to the egg cell and central cell nucleus to achieve double fertilization. Thereafter, embryogenesis takes place (Drews et al. 1998, Yadegari and Drews, 2004, Shi et al. 2005). Several lines of evidences have shown that the mitochondrion is deeply involved in gametophyte development. First, physiological data showed that during pollen formation, the number of mitochondria per cell increased in maize (Zea mays) and tobacco (Nicotiana tabacum) anthers (Warmke and Lee 1978, Huang et al. 1994) and, in the growing pollen tube, the mitochondria exhibited a high motility and formed continuous streaming throughout the cytoplasm (Parton et al. 2003, Yamaoka and Leaver 2008). Secondly, genetic characterization of Arabidopsis mutants defective in nuclear genes encoding mitochondria-associated proteins showed that mitochondria played important roles in gametophyte development (Skinner et al. 2001, Christensen et al. 2002, León et al. 2007, Yamaoka and Leaver 2008). For example, a mutation in EMB2473/MIRO1 encoding a Miro GTPase in mitochondria altered the mitochondrial structure in pollen grains and strongly affected pollen tube growth (Yamaoka and Leaver 2008). Thirdly, mutations in mitochondrial genes may lead to severe defects in respiration and affect male gametophyte development, which could result in cytoplasmic male sterility (CMS) that has been described in about 150 plant species (Conley and Hanson 1995, Hanson and Bentolila 2004, Linke and Börner 2005, Carlsson et al. 2008). Furthermore, several essential mitochondrial gene transcripts have been found to exhibit high abundance in developing maize male gametophytes (Monéger et al. 1992, Wen and Chase 1999). Recently, microarray analysis also showed that most mitochondrial genes were highly expressed in ungerminated dry pollen grains, growing pollen tubes and ovaries in Arabidopsis (Qin et al. 2009). All these data indicate that mitochondria play important roles in plant gametophyte development. The mitochondrion is a semi-autonomous organelle that possesses its own genomes and gene expression systems (Millar et al. 2004). In Arabidopsis, the mitochondrial genome bears 57 genes (Unseld et al. 1997, Binder and Brennicke 2003). As in most other eukaryotic organisms, the mitochondrial genes in Arabidopsis are transcribed by a small family of nuclear-encoded phage-type RNA polymerases. RNA polymerase of T3/T7 type (RPOT) is most similar to the RNA polymerases (RNAPs) of bacteriophage T3, T7 and SP6 (Greenleaf et al. 1986, Kelly et al. 1986, Cermakian et al. 1996, Hedtke et al. 1997). There are three nuclear-encoded RPOTs in Arabidopsis, the first (RPOTm) exists exclusively in the mitochondrion, the second (RPOTp) is found solely in the plastid and the third (RPOTmp) is presumed to be located in both organelles (Hedtke et al. 1997, Hedtke et al. 1999, Hedtke et al. 2000). Previous studies have indicated that all of them are transcriptionally active RNAPs (Kühn et al. 2007). Both RPOTm and RPOTmp are prominently expressed in meristematic and young tissues that have high mitochondrial activity (Emanuel et al. 2006). Thus, there are at least two active phage-type RNAPs that are located in Arabidopsis mitochondria. RPOTm can recognize the mitochondrial promoter accurately in vitro, whereas RPOTmp does not display any significant promoter specificity (Kühn et al. 2007). A recent study indicated that RPOTmp performed gene-specific transcription in Arabidopsis mitochondria (Kühn et al. 2009). Therefore, both of them participate in mitochondrial gene transcription. However, the genetic roles of these mitochondria phage-type RNA polymerases in plants are still poorly understood. We report here the characterization of the novel mutations in the Arabidopsis RPOTm gene. The rpoTm mutations significantly retarded pollen tube growth and disrupted embryogenesis and fusion of polar nuclei in embryo sacs. These data indicate that RPOTm plays important roles in pollen tube growth, female gametogenesis and embryogenesis. The rpoTm rpoTmp double mutation could enhance the rpoTm mutant phenotype. It almost completely disrupted female gametogenesis and severely impaired pollen formation and germination. Furthermore, expression of RPOTmp under the control of the RPOTm promoter could not complement the rpoTm mutant phenotype. These results indicate that the mitochondrial phage-type RNA polymerases are important for gametogenesis, and they have distinct roles in plant development. Results Isolation and genetic analysis of mutant mgp3-1 The first rpoTm mutant allele used in this study was identified in a screen for male gametophyte-defective (mgp) mutants in Arabidopsis thaliana using the gene- and enhancer-trap system (Sundaresan et al. 1995) and was named male gametophyte- defective 3-1 (mgp3-1). The mutation was generated by the insertion of the modified dissociation (Ds) element (Sundaresan et al. 1995). DNA gel blot hybridization showed a single Ds insertion in the genome (Supplementary Fig. S1). The Ds element carries an NPTII gene as a selective genetic marker. It also carries a GUS (β-glucuronidase) reporter gene that could be co-expressed specifically in the mutant pollen grains. Therefore, the mutant pollen grains could be distinguished from wild-type pollen grains by GUS staining (Fig. 1A, B). The mutant exhibited a segregation ratio of roughly 1 (1,240) kanamycin-resistant (KanR) to 1 (1,314) kanamycin-sensitive (KanS) in the progeny of the self-crossed mgp3-1/+ plants. No mgp3-1 homozygous mutant plant was found in the progeny (n = 230). The outcrosses of mgp3-1/+ plants to wild-type plants showed that both the male and female transmission efficiencies of the mgp3-1 mutation were significantly reduced. In particular, the male transmission efficiency was 31% (519/1,689) and the female transmission efficiency was 59% (744/1,264) (Table 1). This result showed that the mutation had a stronger effect on male gametophytic function than on female gametophytic function. Fig. 1 View largeDownload slide The mgp3-1 mutation affected pollen tube growth in vitro. (A, B) The mgp3-1/+ (A) and wild-type (B) pollen grains with or without GUS stain. (C, D) The SEM images of mgp3-1/+ (C) and wild-type (D) pollen grains. (E, F) The Alexander-stained mgp3-1/+ (E) and wild-type (F) pollen grains. (G, H) The DAPI-stained mgp3-1/+ (G) and wild-type (H) pollen grains. (I–K) In vitro germination of the pollen grains from wild-type (I), mgp3-1/+ (J) and MGP3-complemented mgp3-1 homozygous plants (K). (L) Statistical analysis of pollen tube length in wild-type, mgp3-1/+ and two MGP3-complemented mgp3-1 homozygous plants (MGP3-F-1 and MGP3-F-2). The mutant pollen grains were labeled by GUS staining. The red arrowheads indicate the mgp3-1 mutant pollen grains and pollen tubes. Bars = 20 μm in A–H, 50 μm in I–K. Fig. 1 View largeDownload slide The mgp3-1 mutation affected pollen tube growth in vitro. (A, B) The mgp3-1/+ (A) and wild-type (B) pollen grains with or without GUS stain. (C, D) The SEM images of mgp3-1/+ (C) and wild-type (D) pollen grains. (E, F) The Alexander-stained mgp3-1/+ (E) and wild-type (F) pollen grains. (G, H) The DAPI-stained mgp3-1/+ (G) and wild-type (H) pollen grains. (I–K) In vitro germination of the pollen grains from wild-type (I), mgp3-1/+ (J) and MGP3-complemented mgp3-1 homozygous plants (K). (L) Statistical analysis of pollen tube length in wild-type, mgp3-1/+ and two MGP3-complemented mgp3-1 homozygous plants (MGP3-F-1 and MGP3-F-2). The mutant pollen grains were labeled by GUS staining. The red arrowheads indicate the mgp3-1 mutant pollen grains and pollen tubes. Bars = 20 μm in A–H, 50 μm in I–K. Table 1 Genetic analysis of mgp3-1 and mgp3-2 mutants Crosses (female × male)  KanR  KanS  KanR/KanS  TEF (%)  TEM (%)  mgp3-1/+ × mgp3-1/+  1,240  1,314  0.94  NA  NA  mgp3-1/+ × +/+  744  1,264  0.59  59  NA  +/+ × mgp3-1/+  519  1,689  0.31  NA  31  mgp3-2/+ × mgp3-2/+  1,127  1,332  0.85  NA  NA  mgp3-2/+ × +/+  573  994  0.58  58  NA  +/+ × mgp3-2/+  483  2,219  0.22  NA  22  Crosses (female × male)  KanR  KanS  KanR/KanS  TEF (%)  TEM (%)  mgp3-1/+ × mgp3-1/+  1,240  1,314  0.94  NA  NA  mgp3-1/+ × +/+  744  1,264  0.59  59  NA  +/+ × mgp3-1/+  519  1,689  0.31  NA  31  mgp3-2/+ × mgp3-2/+  1,127  1,332  0.85  NA  NA  mgp3-2/+ × +/+  573  994  0.58  58  NA  +/+ × mgp3-2/+  483  2,219  0.22  NA  22  KanR, kanamycin-resistant; KanS, kanamycin-sensitive; NA, not applicable; TE, transmission efficiency; TE = (KanR/KanS) × 100%; TEF, female transmission efficiency; TEM, male transmission efficiency. View Large The mgp3-1 mutation significantly retarded pollen tube growth We first examined the morphology and viability of mature mgp3-1 pollen grains. As shown in Fig. 1, mgp3-1 pollen grains appeared morphologically normal (Fig. 1C, D). Alexander staining showed that mgp3-1 pollen grains contained cytoplasm like the wild type (Fig. 1E, F). 4′,6-Diamidino-2-phenylindole (DAPI) staining indicated that the mutant pollen grains had one vegetative nucleus and two sperm nuclei like the wild type (Fig. 1G, H). These results suggested that the mgp3-1 mutation did not affect the formation of pollen grains. We further investigated mgp3-1 pollen germination in vitro. The pollen grains from wild-type and mgp3-1 heterozygous plants (mgp3-1/+) were germinated for 6 h and labeled by GUS staining (Fig. 1I–K). The pollen grains from mgp3-1 heterozygous plants included two types of pollen grains, the mgp3-1 mutant pollen grains with GUS stain and the wild-type pollen grains without GUS stain. As shown in Fig. 1L, the germination rates of GUS-positive and GUS-negative pollen grains from mgp3-1/+ plants were 73.6% (733/996) and 73.8% (614/832), respectively, similar to that (76.6%, 660/862) of the pollen grains from wild-type plants, indicating that the mutation did not affect pollen germination. The average length of the wild-type pollen tubes and GUS-negative pollen tubes from the mgp3-1/+ plants was roughly 180 μm (Fig. 1I, L). In contrast, the average length of the GUS-positive mutant pollen tubes was about 80 μm (Fig. 1J, L). These findings indicated that mgp3-1 mutation dramatically affected pollen tube growth in vitro. To determine if the mutant pollen tube growth was also affected in vivo, mgp3-1/+ plants were used as pollen donors to cross with wild-type plants. Seeds from the upper and lower halves of the siliques were harvested separately. The seeds were sown on MS agar plates supplied with kanamycin. The KanR segregation rate of the seedlings derived from the seeds of the upper half of the siliques was 41.7 % (171/401), 3.2 times as high as that (12.8%, 63/492) of the seedlings derived from the seeds of the lower half of the siliques, indicating that the growth of mgp3-1 mutant pollen tubes was also significantly affected in vivo. Taken together, we concluded that the mgp3-1 mutation significantly retarded pollen tube growth. The mgp3-1 mutation affected ovule and embryo development To assess the ovule and embryo development processes in the mgp3-1 mutant, the siliques from wild-type and mgp3-1/+ plants were examined at 8 d after self-pollination. In contrast to the wild-type siliques that only had 3% aborted ovules and no abnormal seeds (Fig. 2A, Table 2), the mgp3-1/+ siliques had 8.4% aborted seeds and 17.5% aborted ovules (Fig. 2B, D, E; also see Table 2). To investigate the cause of seed abortion in mgp3-1/+ plants, the seeds from the crosses of mgp3-1/+ and the wild-type were examined at 8 d after pollination. When mgp3-1/+ was used as the pollen donor to pollinate wild-type pistils, the seed setting was normal as in the wild type. When a wild-type plant was used as the pollen donor to pollinate mgp3-1/+ pistils, 20% (185/943) of ovules were found to be unfertilized (Supplementary Fig. S2B), but no aborted seeds were observed. These results indicated that the seed abortion resulted from mgp3-1 homozygous mutation. Table 2 Statistical analysis of seed setting in mgp3-1/+ and mgp3-2/+ siliques by comparison with wild-type plants Plants  Total seeds and ovules  Normal seeds   Aborted seeds   Aborted ovules   n  %  n  %  n  %  Wild-type Ler  1,852  1,790  96.7  0  0  62  3.3  mgp3-1/+  1,996  1,479  74.1  168  8.4  349  17.5  Wild-type Col-0  2,120  2,049  96.6  0  0  71  3.4  mgp3-2/+  2,132  1,515  71.1  223  10.5  394  18.5  Plants  Total seeds and ovules  Normal seeds   Aborted seeds   Aborted ovules   n  %  n  %  n  %  Wild-type Ler  1,852  1,790  96.7  0  0  62  3.3  mgp3-1/+  1,996  1,479  74.1  168  8.4  349  17.5  Wild-type Col-0  2,120  2,049  96.6  0  0  71  3.4  mgp3-2/+  2,132  1,515  71.1  223  10.5  394  18.5  The seeds were counted on the first eight siliques from five independent wild-type or mutant plants. View Large Fig. 2 View largeDownload slide mgp3-1 mutation affected embryo development. (A–C) The siliques from wild-type (A), mgp3-1 heterozygous (B) and MGP3-complemented mgp3-1 homozygous (C) plants. The white arrows indicate the aborted mutant seeds. The white asterisks indicate the aborted ovules. (D) The magnified images of a normal seed (left) and an aborted seed (right). (E) A DIC image of cleared whole-mounted seeds in the mgp3-1/+ siliques. (F–I) DIC images of wild-type embryos at the globular stage (F), heart stage (G), torpedo stage (H) and matured embryo (I). (J–M) DIC images of mgp3-1 mutant embryos at the 2-cell stage (J), 8-cell stage (K) and multiple cell stages (L) and (M). Bars = 1 mm in A–C, 100 μm in D, E, 25 μm in F–M. Fig. 2 View largeDownload slide mgp3-1 mutation affected embryo development. (A–C) The siliques from wild-type (A), mgp3-1 heterozygous (B) and MGP3-complemented mgp3-1 homozygous (C) plants. The white arrows indicate the aborted mutant seeds. The white asterisks indicate the aborted ovules. (D) The magnified images of a normal seed (left) and an aborted seed (right). (E) A DIC image of cleared whole-mounted seeds in the mgp3-1/+ siliques. (F–I) DIC images of wild-type embryos at the globular stage (F), heart stage (G), torpedo stage (H) and matured embryo (I). (J–M) DIC images of mgp3-1 mutant embryos at the 2-cell stage (J), 8-cell stage (K) and multiple cell stages (L) and (M). Bars = 1 mm in A–C, 100 μm in D, E, 25 μm in F–M. The differential interference contrast (DIC) microscopic images of whole-mount clearing ovules showed no significant difference in the seeds at the zygote and single-terminal cell stages between the wild type and the mutant. The defects in embryo development occurred at the early globular stage. In the mgp3-1/+ siliques, 90% (415/468) of embryos were at the 16-cell to 64-cell stages (Fig. 2F), while the others were just at the 2- or 4-terminal cell stage (Fig. 2J). In the older siliques, the normal embryos had entered the early heart stage (Fig. 2G) while the embryos in the aborted seeds were retarded at the early globular stage (Fig. 2K). When wild-type embryos had developed to torpedo and mature stages (Fig. 2H, I), the mutant embryos were still at the globular stage and had an expanded shape (Fig. 2L, M). When the wild-type seeds matured, the aborted seeds shrunk and had a dark yellow color. No (n > 100) mutant embryo could go beyond the globular stage in the aborted seeds that had shrunk. This observation indicated that the mgp3-1 homozygous mutation could disrupt embryogenesis and consequently led to embryo lethality at the globular stage, suggesting that RPOTm is essential for embryogenesis in Arabidopsis. The mutant embryo sacs were defective in fusion of polar nuclei We characterized the mature mutant ovules by confocal laser scanning microscopy (CLSM). All ovules in wild-type siliques at the mature stage contained one secondary central cell nucleus, one egg cell nucleus and two synergid cell nuclei (Fig. 3C). In mgp3-1/+ siliques, however, only 82.2% (268/326) of ovules could develop normally like the wild-type ovule, and 17.8% (58/326) of ovules showed failure of fusion of the two polar nuclei and the three antipodal cells degenerated earlier (Fig. 3G). To determine further if the aborted embryo sacs were caused by defects in fusion of polar nuclei, wild-type pollen grains were pollinated to wild-type and mgp3-1/+ pistils, respectively, and then the ovules were observed at different times after pollination. At 36 h after pollination, the zygotes in wild-type ovules started dividing and the fertilized secondary nucleus had divided into several endosperm nuclei (Fig. 3D). In mgp3-1/+ pistils, 80.9% (251/310) of ovules developed as normally as the wild type and could be fertilized, and 19.1% ovules could not be fertilized, among which 16.5% (51/310) of ovules had two unfused polar nuclei (Fig. 3H) and 2.6% (8/310) of ovules had one central cell nucleus. The egg cell appeared highly incompact and no embryo and endosperm nucleus were observed in these ovules. At 72 h after pollination, the unfused polar nuclei and egg cell nucleus in the abnormal ovules had degenerated. As a result, the ovules stopped growing (Fig. 3I) while wild-type embryos in the same silique had developed to the globular stage (Fig. 3E). These data indicated that the mgp3-1 mutation affected fusion of the polar nuclei and disrupted female gametophyte development. Fig. 3 View largeDownload slide The mgp3-1 mutation affected fusion of polar nuclei. (A–E) CLSM images of wild-type embryo sacs at stage FG5 (A), stage FG6 (B), the terminal stage before pollination (C), 36 h after pollination (D) and 72 h after pollination (E). (F–I) CLSM images of mgp3-1 embryo sacs with unfused polar nuclei at stage FG5 (F), the terminal stage before pollination (G), 36 h after pollination (H) and 72 h after pollination (I). ac, apical cell; AN, antipodal cell nuclei; bc, basal cell; Ch, chalazal end; CN, central cell nucleus; EdN, endosperm nucleus; Emb, embryo; EN, egg cell nucleus; PN, polar nucleus; SN, synergid cell nucleus. Bars = 10 μm. Fig. 3 View largeDownload slide The mgp3-1 mutation affected fusion of polar nuclei. (A–E) CLSM images of wild-type embryo sacs at stage FG5 (A), stage FG6 (B), the terminal stage before pollination (C), 36 h after pollination (D) and 72 h after pollination (E). (F–I) CLSM images of mgp3-1 embryo sacs with unfused polar nuclei at stage FG5 (F), the terminal stage before pollination (G), 36 h after pollination (H) and 72 h after pollination (I). ac, apical cell; AN, antipodal cell nuclei; bc, basal cell; Ch, chalazal end; CN, central cell nucleus; EdN, endosperm nucleus; Emb, embryo; EN, egg cell nucleus; PN, polar nucleus; SN, synergid cell nucleus. Bars = 10 μm. Molecular analysis of mgp3-1 Thermal asymmetric interlaced PCR (TAIL-PCR) (Liu et al. 1995) was applied to obtain the gene sequences flanking the Ds element in the mgp3-1 mutant genome. Sequence analysis of the TAIL-PCR products indicated that the Ds element was inserted in the first exon of At1g68990 (RPOTm) at a site 115 bp downstream of the ATG codon (Fig. 4A). The insertion site was then further confirmed by conventional PCR using the Ds element-specific primer Ds5-2 and the gene-specific primer P2 (Fig. 4B; see also Materials and Methods). The Ds insertion resulted in an 8 bp (ATCTCAGT) duplication of the host sequence at the insertion site (Fig. 4A). Reverse transcription–PCR (RT–PCR) analysis showed a chimeric transcript product of RPOTm mRNA and the Ds sequence (Fig. 4C), leading to loss of the C-terminal domain of RPOTm. Therefore, mgp3-1 is probably a null rpoTm allele. Fig. 4 View largeDownload slide Molecular characterization of the MGP3 gene. (A) A schematic diagram of the MGP3 gene structure, showing the Ds and T-DNA insertion sites in the mgp3-1 and mgp3-2 mutants. (B) Confirmation of the transgenic mutant plants in complementation experiments. (C) The chimeric transcript containing RPOTm mRNA and Ds 3′ end sequences, revealed by RT–PCR. The black boxes indicate exons, and black arrowheads indicate the positions of primers used for genotyping. 1300H, Ds5-2, Ds3-2, LBa1, P1 and P2 are the primers used in the PCR assays. Hyg, hygromycin; Kan, kanamycin; kb, kilobase pairs; MW, molecular weight; WT, wild-type. Fig. 4 View largeDownload slide Molecular characterization of the MGP3 gene. (A) A schematic diagram of the MGP3 gene structure, showing the Ds and T-DNA insertion sites in the mgp3-1 and mgp3-2 mutants. (B) Confirmation of the transgenic mutant plants in complementation experiments. (C) The chimeric transcript containing RPOTm mRNA and Ds 3′ end sequences, revealed by RT–PCR. The black boxes indicate exons, and black arrowheads indicate the positions of primers used for genotyping. 1300H, Ds5-2, Ds3-2, LBa1, P1 and P2 are the primers used in the PCR assays. Hyg, hygromycin; Kan, kanamycin; kb, kilobase pairs; MW, molecular weight; WT, wild-type. A search through the Arabidopsis Biological Resource Center (ABRC) database further identified an rpoTm T-DNA insertion mutant, SALK_005875. The T-DNA was inserted in the first exon of At1g68990 at a site 72 bp downstream ATG codon in this mutant. The insertion site was confirmed by PCR using a T-DNA left border primer (LBa1) and a gene-specific primer (Fig. 4B). The T-DNA also carried a kanamycin resistance marker gene, and SALK_005875 showed a segregation ratio of roughly 1 (1,127) KanR to 1 (1,332) KanS in the progeny from the self-crossed SALK_005875 heterozygous plants. No SALK_005875 homozygous mutant plant was obtained. Outcrosses of SALK_005875 with wild-type Col-0 plants showed that both the male and female transmission efficiencies of the SALK_005875 mutation were reduced in the same manner as mgp3-1/+ (Table 1). The siliques of SALK_005875 heterozygous mutant plants contained 10.5% aborted seeds and 18.5% aborted ovules (Table 2). The embryo development in the aborted seeds was arrested at the globular stage. Furthermore, the pollen tube growth and fusion of polar nuclei were also affected. All these phenotypes were identical to those observed in mgp3-1/+. Therefore, SALK_005875 was considered to be allelic to mgp3-1 and was renamed mgp3-2. To confirm further that the phenotypes of mgp3-1/+ and mgp3-2/+ are caused by disruption of RPOTm, a 7.2 kb genomic DNA fragment of At1g68990, including the 1,147 bp promoter region, coding region and the 840 bp transcription termination region, was introduced into mgp3-1/+ and mgp3-2/+ plants, respectively. In the case of mgp3-1/+, 45 independent transformants were obtained. All these transformants were confirmed by PCR using the pCAMBIA1300-specific primer 1300-H and the RPOTm gene-specific primer P2 (Fig. 4B). The T2 seedlings from these transformants showed a segregation ratio of 2 KanR to 1 KanS (n > 500), indicating that the defective male and female gametophyte transmission efficiencies had been restored by the At1g68990 gene. Furthermore, mgp3-1 homozygous plants harboring the homozygous transgene could be identified in the T2 generation. In these plants, seed setting and pollen tube growth were as normal as in the wild type (Figs. 1K, 2C; see also Supplementary Fig. S2C). In the case of mgp3-2/+, 25 independent transformants were obtained, whose phenotypes were all recovered as in the complemented mgp3-1/+ plants. All these data indicated that the 7.2 kb DNA fragment could complement the phenotype of mgp3-1/+ and mgp3-2/+ plants, and showed that both mgp3-1 and mgp3-2 were loss-of- function mutations in RPOTm. Mitochondrial structure in mgp3-1 pollen grains was slightly affected The embryo lethal phenotype displayed by the mgp3 mutants and results of previous studies strongly support that RPOTm is the principal RNA polymerase that transcribes mitochondrial genes (Baba et al. 2004, Kühn et al. 2007). However, the embryo lethal phenotype of mgp3/– makes it difficult to detect the difference in mitochondrial gene expression between mutant and wild-type plants. We therefore examined whether the mgp3 mutations had an impact on the distribution and number of mitochondria. The result showed no significant change in the distribution and number of mitochondria between wild-type and mgp3-1 mutant pollen grains. We then examined the mitochondrial ultrastructure in mgp3-1/+ and wild-type pollen grains using transmission electron microscopy (TEM). Thirty-five individual wild-type Ler pollen grains (109 images) were examined. Their mitochondria all appeared as normal rod-like shapes (Fig. 5A, C). They contained an electron-dense matrix inside the outer membrane and well-developed inner membrane cristae (Fig. 5A, C). We then examined 41 individual pollen grains (127 images) from mgp3-1/+ plants, among which, theoretically, 50% of pollen grains were wild types and the other 50% were mgp3-1 mutant pollen grains. No significant alteration in mitochondrial morphology was observed in most of these pollen grains (Fig. 5B, D). Only in approximately 10% (4/41) of pollen grains did the mitochondria show a reduction in cristae that appeared with a curved or circular shape (Fig. 5E, G). This type of mitochondrion was never observed in wild-type pollen grains. We further examined mitochondrial structures in the RPOTm-complemented mgp3-1 pollen grains; no defect in mitochondrial structure was found in 117 images taken from 39 individual pollen grains (Fig. 5F, H). These results suggested that the defective mitochondrial structures in the mgp3-1/+ pollen grains were correlated to mgp3-1 mutation. Furthermore, such a type of abnormal mitochondria was similar to those in cultured ρ0 human cells, whose mtDNAs were fully depleted (Gilkerson et al. 2000). Fig. 5 View largeDownload slide TEM images of the mitochondria in wild-type and mgp3-1 pollen grains. (A, C) Mitochondria in wild-type pollen grains. (B, D) Normal mitochondria in most pollen grains from mgp3-1/+ plants. (E, G) Defective mitochondria in some pollen grains from mgp3-1/+ plants. (F, H) Mitochondria in RPOTm-complemented mgp3-1 pollen grains. The white arrows indicate the mitochondrial cristae. Bars = 500 nm. Fig. 5 View largeDownload slide TEM images of the mitochondria in wild-type and mgp3-1 pollen grains. (A, C) Mitochondria in wild-type pollen grains. (B, D) Normal mitochondria in most pollen grains from mgp3-1/+ plants. (E, G) Defective mitochondria in some pollen grains from mgp3-1/+ plants. (F, H) Mitochondria in RPOTm-complemented mgp3-1 pollen grains. The white arrows indicate the mitochondrial cristae. Bars = 500 nm. mgp3-2 rpoTmp-2 double mutation enhanced the mgp3 mutant phenotype The rpoTmp-2 mutant (SALK_132842) was obtained from the ABRC. It was in the Col-0 background, in which the T-DNA was inserted in the third intron of RPOTmp, at a site 1,731 bp downstream of the ATG codon (Supplementary Fig. S3A). The plants homozygous for the T-DNA insertion were identified by PCR (Supplementary Fig. S3B). The RPOTmp transcripts were not detected by RT–PCR in the mutant plants (Supplementary Fig. S3C). The mutant plants also showed a great reduction in plant growth as reported in the known rpoTmp mutants, 833 and 286E07 (Baba et al. 2004) (Supplementary Fig. S3D). These data indicated that rpoTmp-2 was a null rpoTmp mutant. Genetic analysis showed that the rpoTmp-2 mutant gametophytes were functionally normal (Table 3). Table 3 Genetic analysis of the mgp3-2/+; rpoTmp-2/+ double mutant by comparison with mgp3-2/+ and rpoTmp-2/+ single mutants Crosses (female × male)  Progeny   TEF (%)  TEM (%)  WT  rpoTmp-2/+  mgp3-2/+  mgp3-2/+; rpoTmp-2/+  mgp3-2/+ × +/+  209    117    56  NA  +/+ × mgp3-2/+  646    126    NA  20  rpoTmp-2/+ × +/+  130  121      93  NA  +/+ × rpoTmp-2/+  332  325      NA  97  mgp3-2/+; rpoTmp-2/+ × +/+  120  111  63  0  0  NA  +/+ × mgp3-2/+; rpoTmp-2/+  108  92  15  3  NA  3  Crosses (female × male)  Progeny   TEF (%)  TEM (%)  WT  rpoTmp-2/+  mgp3-2/+  mgp3-2/+; rpoTmp-2/+  mgp3-2/+ × +/+  209    117    56  NA  +/+ × mgp3-2/+  646    126    NA  20  rpoTmp-2/+ × +/+  130  121      93  NA  +/+ × rpoTmp-2/+  332  325      NA  97  mgp3-2/+; rpoTmp-2/+ × +/+  120  111  63  0  0  NA  +/+ × mgp3-2/+; rpoTmp-2/+  108  92  15  3  NA  3  NA, not applicable; TE, transmission efficiency. For the mgp3-2/+ single mutant and rpoTmp-2/+ single mutant, the TE = (No. of progeny plants with T-DNA insertion/No. of the progeny plants without T-DNA insertion) × 100%; for the mgp3-2/+; rpoTmp-2/+ double mutant, the TE = (No. of progeny plants with double T-DNA insertions/No. of progeny plants without T-DNA insertion) × 100%. TEF, female transmission efficiency; TEM, male transmission efficiency. View Large We used the RPOTm mutant allele mgp3-2/+ in the Col-0 background to cross with rpoTmp-2/+ and identified the plants heterozygous for both mgp3-2 and rpoTmp-2 mutations (mgp3-2/+; rpoTmp-2/+) among the F1 progeny. In total, 69 out of 462 F2 plants were identified as having short root phenotypes as described in the rpoTmp mutant 833 (Baba et al. 2004). All of them were transferred to soil and then analyzed by PCR. The result showed that all 69 plants were homozygous for the rpoTmp-2 mutation, but none of them possessed the mgp3-2 mutation, suggesting that the mgp3-2 rpoTmp-2 double mutation could not be transmitted to the next generation through gametophytes. To confirm this observation, the mgp3-2/+; rpoTmp-2/+ mutant plants were out-crossed to wild-type Col-0 plants. The karyotyping analysis of the F1 progeny by PCR indicated that the female transmission efficiency of the mgp3-2 rpoTmp-2 double mutation was zero (Table 3) and the male transmission efficiency was 3% (Table 3). These data showed that all the female and nearly all the male gametophytes bearing the mgp3-2 rpoTmp-2 double mutation lost their ability for transmission to next generation. To understand in detail how the mgp3-2 rpoTmp-2 double mutation impaired male gametophyte development, we studied the pollen phenotypes in mgp3-2/+; rpoTmp-2/+ plants by comparison with those in the mgp3-2/+ single mutant, the rpoTmp-2/+ single mutant and wild-type plants. We first studied the pollen morphology at the anthesis stage. In mgp3-2/+; rpoTmp-2/+ plants, 14% (109/769) of pollen grains were aborted (Fig. 6A) before maturation. Such defective pollen grains were not or rarely found in wild-type, mgp3-2/+ single mutant and rpoTmp-2/+ single mutant plants (Fig. 6B–D), indicating that the abortion of pollen grain development was caused by the mgp3-2 rpoTmp-2 double mutation. We also analyzed the viability, mitosis and ultrastructure of the aborted pollen grains, and found that they lacked cellular contents (Fig. 6F, H, J, L). These observations indicated that the mgp3-2 rpoTmp-2 double mutation affected pollen formation. The genetic data mentioned above showed that the rate of aborted pollen grains was significantly lower than the predicted rate (25%) of double mutant pollen grains. They implied that a proportion of the double mutant microspores could develop into morphologically normal pollen grains. Therefore, we examined the in vitro germination of the pollen grains from the double mutant plants (Fig. 7A–D). The pollen germination rate in the mgp3-2/+; rpoTmp-2/+ mutant was calculated and compared with those of the pollen grains from wild-type Col-0, mgp3-2/+ single mutant and rpoTmp-2/+ single mutant plants (Fig. 7E). The germination rate of pollen grains in mgp3-2/+; rpoTmp-2/+ mutants was found to be approximately 69% (597/865), while that in mgp3-2/+ single mutant, rpoTmp-2/+ single mutant and wild-type plants was approximately 85% (Fig. 7E). These observations indicated that the rpoTm rpoTmp double mutation severely impaired pollen formation and germination. Fig. 6 View largeDownload slide The mgp3-2 rpoTmp-2 double mutation affected pollen formation. (A–D) SEM images of pollen grains from mgp3-2/+; rpoTmp-2/+ (A), wild-type Col-0 (B), rpoTmp-2/+ single mutant (C) and mgp3-2/+ single mutant (D) plants. (E, F) Alexander staining images of wild-type (E) and aborted (F) pollen grains. (G, K) Light-field image (G) and DAPI staining image (K) of wild-type matured pollen grains. (H, L) Light-field (H) image and DAPI stain (L) of the pollen grains from a mgp3-2/+; rpoTmp-2/+ plant. (I, J) TEM images of wild-type (I) and aborted (J) mutant pollen grains. The white arrows indicate the defective pollen grains. Bars = 50 μm in A–D and 10 μm in E–L. Fig. 6 View largeDownload slide The mgp3-2 rpoTmp-2 double mutation affected pollen formation. (A–D) SEM images of pollen grains from mgp3-2/+; rpoTmp-2/+ (A), wild-type Col-0 (B), rpoTmp-2/+ single mutant (C) and mgp3-2/+ single mutant (D) plants. (E, F) Alexander staining images of wild-type (E) and aborted (F) pollen grains. (G, K) Light-field image (G) and DAPI staining image (K) of wild-type matured pollen grains. (H, L) Light-field (H) image and DAPI stain (L) of the pollen grains from a mgp3-2/+; rpoTmp-2/+ plant. (I, J) TEM images of wild-type (I) and aborted (J) mutant pollen grains. The white arrows indicate the defective pollen grains. Bars = 50 μm in A–D and 10 μm in E–L. Fig. 7 View largeDownload slide mgp3-2 rpoTmp-2 double mutation affected pollen germination. (A–D) In vitro germination of pollen grains from wild-type Col-0 (A), rpoTmp-2 single mutant (B), mgp3-2/+ single mutant (C) and mgp3-2/+; rpoTmp-2/+ double mutant (D) plants. (E) Statistical analysis of the pollen germination rates in wild-type Col-0, the rpoTmp-2/+ single mutant, the mgp3-2/+ single mutant and the mgp3-2/+; rpoTmp-2/+ double mutant. The white arrows indicate the aborted pollen grains. Bars = 100 μm in A–D. Fig. 7 View largeDownload slide mgp3-2 rpoTmp-2 double mutation affected pollen germination. (A–D) In vitro germination of pollen grains from wild-type Col-0 (A), rpoTmp-2 single mutant (B), mgp3-2/+ single mutant (C) and mgp3-2/+; rpoTmp-2/+ double mutant (D) plants. (E) Statistical analysis of the pollen germination rates in wild-type Col-0, the rpoTmp-2/+ single mutant, the mgp3-2/+ single mutant and the mgp3-2/+; rpoTmp-2/+ double mutant. The white arrows indicate the aborted pollen grains. Bars = 100 μm in A–D. We also observed the female gametophyte development process in mgp3-2/+; rpoTmp-2/+ pistils by comparison with that in mgp3-2/+ single mutant, rpoTmp-2/+ single mutant and wild-type plants. We found that 36.5% embryo sacs aborted in mgp3-2/+; rpoTmp-2/+ mutant siliques, 18.5% in mgp3-2/+ siliques and 4% in rpoTmp-2/+ and wild-type siliques (Fig. 8A–D, Table 4). This phenotype was also observed when wild-type plants were used as pollen donors, indicating that the aborted embryo sacs were caused by the mutations rather than environmental or paternal factors. The proportion of the aborted embryo sacs in mgp3-2/+; rpoTmp-2/+ double mutant siliques was correlated to the genetic analysis data. No embryo was found in any aborted embryo sacs, suggesting that the embryo sacs aborted at least before embryogenesis took place. We then observed the developmental process of the embryo sacs by CLSM (Fig. 8I–N). At the mature stage before fertilization, all embryo sacs in rpoTmp-2/+ plants could develop normally as in the wild-type embryo sacs (Fig. 3C). In contrast, in the mgp3-2/+ siliques, in 21.8% (61/279) of ovules the two polar nuclei failed to fuse. This phenotype was similar to that observed in mgp3-1/+ mutant female gametophytes (Fig. 3G). In mgp3-2/+; rpoTmp-2/+ siliques, 26.7% (69/258) of ovules showed failure of fusion of the polar nuclei like those in the mgp3 embryo sacs (Fig. 3G). The number of aborted embryo sacs bearing unfused polar nuclei was less than the predicted 35%. We then analyzed the embryo sac development process after fertilization. To do this, we pollinated wild-type pollen to the mutant pistils. At 18 h after pollination, the fertilized egg cell and central cell could develop normally in wild-type and rpoTmp-2 embryo sacs (Fig. 8I). In the mgp3-2/+ single mutant and the mgp3-2/+; rpoTmp-2/+ double mutant, the embryo sacs bearing two unfused polar nuclei failed to develop further as observed in mgp3-1 mutant embryo sacs (Fig. 3H). Furthermore, in the mgp3-2/+; rpoTmp-2/+ double mutant, 12% of ovules that had one central cell, one egg cell and two synergid cell nuclei also failed to develop further (Fig. 8L). No embryo and endosperm nucleus were found in the defective ovules, and the two synergid cells appeared intact (Fig. 8L). No defect in pollen tube attraction was observed in the mgp3-2/+; rpoTmp-2/+ pistil by aniline blue staining (Fig. 8E–H). At 72 h after pollination, the defective embryo sacs stopped growing and eventually died (Fig. 8N). These observations suggested that the double mutation impaired not only fusion of polar nuclei, but also synergid cell death. Table 4 Statistical analysis of seed setting in the siliques from mgp3-2/+; rpoTmp-2/+ double mutant plants by comparison with mgp3-2/+, rpoTmp-2/+ and wild-type plants Genotypes of the plants  Total seeds and ovules  Normal seeds   Aborted seeds   Aborted ovules   n  %  n  %  n  %  Wild-type Col-0  2,120  2,049  96.6  0  0  71  3.4  rpoTmp-2/+  1,979  1,898  95.9  0  0  81  4.1  mgp3-2/+  2,132  1,515  71.1  223  10.5  394  18.5  mgp3-2/+; rpoTmp-2/+  1,778  1,091  61.4  39  2.2  648  36.5  Genotypes of the plants  Total seeds and ovules  Normal seeds   Aborted seeds   Aborted ovules   n  %  n  %  n  %  Wild-type Col-0  2,120  2,049  96.6  0  0  71  3.4  rpoTmp-2/+  1,979  1,898  95.9  0  0  81  4.1  mgp3-2/+  2,132  1,515  71.1  223  10.5  394  18.5  mgp3-2/+; rpoTmp-2/+  1,778  1,091  61.4  39  2.2  648  36.5  The seeds were counted on the first eight siliques from five independent wild-type or mutant plants. View Large Fig. 8 View largeDownload slide mgp3-2 rpoTmp-2 double mutation affected embryo sac development. (A–D) Seed setting in the siliques from wild-type Col-0 (A), rpoTmp-2/+ (B), mgp3-2/+ (C) and mgp3-2/+; rpoTmp-2/+ (D) plants. (E, F) Aniline blue-stained siliques from wild-type Col-0 (E) and mgp3-2/+; rpoTmp-2/+ double mutant (F) plants. (G, H) Pollen tubes entered the wild-type embryo sac (G) and aborted embryo sacs in the mgp3-2/+; rpoTmp-2/+ double mutant (H). (I–K) The wild type embryo sacs at 18 (I), 36 (J) and 72 h (K) after pollination. (L–N) The abnormal embryo sacs found in the siliques from mgp3-2/+; rpoTmp-2/+ plants at 18 (L), 36 (M) and 72 h (N) after pollination. The arrows in A–D indicate the aborted seeds. The asterisks in A–F indicate the aborted ovules. Ch, chalazal end; CN, central cell nucleus; Dcn, degenerated central cell nucleus; Dsc, degenerated synergid cell nucleus; EdN, endosperm nucleus; Emb, embryo; EN, egg cell nucleus; F, funiculus; PT, pollen tube; SN, synergid cell nucleus; Zyg, zygote. Bars = 1 mm in A–D, 0.5 mm in E, F and 10 μm in G–N. Fig. 8 View largeDownload slide mgp3-2 rpoTmp-2 double mutation affected embryo sac development. (A–D) Seed setting in the siliques from wild-type Col-0 (A), rpoTmp-2/+ (B), mgp3-2/+ (C) and mgp3-2/+; rpoTmp-2/+ (D) plants. (E, F) Aniline blue-stained siliques from wild-type Col-0 (E) and mgp3-2/+; rpoTmp-2/+ double mutant (F) plants. (G, H) Pollen tubes entered the wild-type embryo sac (G) and aborted embryo sacs in the mgp3-2/+; rpoTmp-2/+ double mutant (H). (I–K) The wild type embryo sacs at 18 (I), 36 (J) and 72 h (K) after pollination. (L–N) The abnormal embryo sacs found in the siliques from mgp3-2/+; rpoTmp-2/+ plants at 18 (L), 36 (M) and 72 h (N) after pollination. The arrows in A–D indicate the aborted seeds. The asterisks in A–F indicate the aborted ovules. Ch, chalazal end; CN, central cell nucleus; Dcn, degenerated central cell nucleus; Dsc, degenerated synergid cell nucleus; EdN, endosperm nucleus; Emb, embryo; EN, egg cell nucleus; F, funiculus; PT, pollen tube; SN, synergid cell nucleus; Zyg, zygote. Bars = 1 mm in A–D, 0.5 mm in E, F and 10 μm in G–N. The PRPOTm::RPOTmp cDNA construct cannot complement the phenotypes of the mgp3 mutants The rpoTm rpoTmp double mutation enhanced the phenotype of the rpoTm mutant, implying that RPOTm and RPOTmp may have partially overlapping roles in plant development. To address this question, we examined whether the phenotypes of mgp3 mutants could be complemented by the expression of the RPOTmp gene under control of the RPOTm promoter. To do this, we constructed a PRPOTm::RPOTmp cDNA construct and introduced it into mgp3-1/+ and mgp3-2/+ plants. More than 20 independent transformants were obtained in each mutation background. All the T2 seedlings from these transformants showed a segregation ratio of roughly 1 KanR to 1 KanS (n > 500), indicating that the construct could not restore the defective male/female gametophyte transmission efficiencies and the embryo lethal phenotypes. Thus, RPOTmp cannot replace RPOTm in gametophyte and embryo development. Discussion RPOTm plays important roles in gametophyte and embryo development Here we have studied the genetic roles of RPOTm in gametogenesis and embryogenesis by characterization of the RPOTm loss-of-function mutations mgp3-1 and mgp3-2 in Arabidopsis. Our results provide new insight into the roles of RPOTm in plant development. First, our results showed that the mutations in RPOTm did not affect pollen formation, but significantly retarded pollen tube growth. Mitochondrial genes encode the proteins, rRNAs and tRNAs that function in the complexes of respiratory chain, heme and cytochrome assembly, and the ribosome (Unseld et al. 1997, Binder and Brennicke 2003). Loss of mitochondrial RNA polymerase could affect the expression of these genes and lead to defects in ATP synthesis. Consequently, it reduces the energy supply. A striking feature of the pollen tube growth is the marked accumulation of secretion vesicles that contain the components for cell wall expansion (Taylor and Hepler 1997, Krichevsky et al. 2007). Vesicular trafficking and cell wall synthesis are the energy- and metabolite-dependent cellular processes. Shortage of energy supply could affect vesicular trafficking and other metabolic processes. As a result, pollen tube growth would be retarded. Moreover, the genetic analysis showed that at least 31% of rpoTm mutant male gametophytes could be successful in fertilization. All these results indicate that rpoTm is not essential for cell survival, but is important for pollen tube growth. Secondly, the mgp3 mutations also affected fusion of polar nuclei during female gametophyte development. Previous studies have shown that the mutations in the genes encoding mitochondrial proteins in a number of mutants, such as nfd1, nfd3, nfd4, nfd5, nfd6 (Portereiko et al. 2006), gfa2 (Christensen et al. 2002) and sdh1 (León et al. 2007), caused defects in fusion of polar nuclei. Our result reported here provides further evidence to demonstrate that the karyogamy defect is a consequence of impairment in basic mitochondrial functions. However, how mitochondria are involved in karyogamy still remains unclear. Thirdly, our results demonstrated that the rpoTm homozygous mutant embryos stopped further development at the globular stage, indicating that RPOTm was critical for embryo development from the globular stage to the heart stage. It is not clear why the mgp3/– embryos were arrested at this stage. A recent study revealed that during the transition from globular to heart stages, the expression of many genes that were involved in energy metabolism was up-regulated, which comprised 21% of the up-regulated genes (Spencer et al. 2007). This indicates that energy metabolism is very active during these stages. Nevertheless, the embryogenesis at these stages involves active cell division, cell differentiation and growth. All these processes need to consume a lot of energy. Because RPOTm is involved in the expression of mitochondrial genes, the mgp3 mutations may drastically affect the expression of mitochondrial genes including those involved in energy metabolism. Thus, the mgp3 mutant embryos stopped further development, possibly due to shortage of energy supply. The distinct genetic roles of mitochondrial phage-type RNA polymerases As discussed above, the mgp3 mutations caused embryo lethality and partially impaired female gametophyte development, but did not affect pollen formation and pollen germination, while the rpoTmp mutations reduced plant size, but did not affect male and female gametophyte and embryo development. The rpoTm rpoTmp double mutation disrupted female gametophyte development completely and more severely affected male gametophyte development. These results indicate that RPOTm is important and sufficient for normal gametophyte and embryo development. At the molecular level, previous studies have indicated that RPOTm was able to initiate transcription from overlapping subsets of mitochondrial promoters accurately, while RPOTmp did not display any significant promoter specificity (Kühn et al. 2007). A recent study showed that RPOTmp performed gene-specific transcription in mitochondria; it transcribed at least rps4, nad2, nad6, cox1 and ccmC genes. Levels of respiratory chain complexes I and IV were down-regulated in the rpoTmp mutants (Kühn et al. 2009). Taken all together, the results suggest that RPOTm is the major RNA polymerase in mitochondria. It can transcribe most, if not all, mitochondrial genes, which is sufficient for normal gametophyte and embryo development. RPOTmp transcribes only a subset of mitochondrial genes, e.g. rps4, nad2, nad6, cox1 and ccmC. It can partially rescue the loss of RPOTm function in the rpoTm mutant gametophyte, but is not sufficient to support normal gametophyte and embryo development. When these two RPOT genes were knocked out, probably most or all mitochondrial genes could not be transcribed and then gametophyte development was disrupted completely. Materials and Methods Plant material and mutant isolation The mgp3-1 mutant is in the Landsberg erecta (Ler) background, while mgp3-2 (SALK_005875) and rpoTmp-2(SALK_132842) are in the Columbia-0 background. The plants were grown in an air-conditioned room at 22°C under 16 h light/8 h dark cycles. Seedling growth in vitro was performed as described previously by Yang et al. (2009). The generation of Ds insertion lines and screening of mutants were performed as described by Sundaresan et al. (1995). The mgp3-2 (SALK_005875) and rpoTmp-2 (SALK_132842) seeds were obtained from the ABRC. The T-DNA insertion sites in mgp3-2 and rpoTmp-2 were determined by PCR using primer pairs LBa1 (5′-TGGTTCACGTAGTGGGCCATCG-3′)/P2 (5′-CATTACGCACAGGCTCAAACC-3′) and LBa1/P4 (5′-CGA CCATTTGAGAGGAG-3′), respectively. The rpoTmp-2 homozygous plant was identified by PCR using primer pair P3 (5′-GAATATGCACATTCTTGG-3′)/P4. Phenotypic characterization The developing seeds/embryos were clarified overnight in Hoyer’s solution (Liu and Meinke 1998) and observed with a Leica DMRA2 microscope equipped with DIC system (Leica, Wetzlar, Germany). Morphological observations of pollen grains by SEM were performed as described by Jiang et al. (2005). DAPI staining of pollen grains was performed as described by Yang et al. (2009), and pollen viability was examined using Alexander staining (Alexander 1969). Pollen germination in vitro was performed as described by Jiang et al. (2005). The germinating pollen grains were counted under a microscope after being incubated at 22–23°C for 6 h. From each culture, at least 300 pollen grains were examined to calculate the average germination rates and pollen tube length for each sample. The female gametophytes were characterized by CLSM as described by Christensen et al. (1997) and Shi et al. (2005) with the following modifications. The pistils were collected in 4% glutaraldehyde (in 12.5 mM cacodylate, pH 6.9), kept under a vacuum for 1–2 h, followed by further fixation overnight at room temperature. Then, the tissues were dehydrated through an ethanol series [10, 30, 50, 70, 90, 95, 100, 100 and 100% (v/v)] with 30 min per step. The dehydrated tissues were clarified in 2 : 1 (v/v) benzyl benzoate : benzyl alcohol for at least 1 h and then observed with a Zeiss LSM510 META laser scanning microscope (Zeiss, Jena, Germany) with a 488 nm argon laser and an LP 530 filter. Transmission electron microscopy The samples were fixed in fixing solution containing 5% (v/v) glutaraldehyde, 0.1 M sodium cacodylate, pH 7.2, post-fixed in 1% osmium tetroxide for 1 h, dehydrated in a series of acetone solutions [25, 50, 75 and 100% (v/v), followed by a resin/ acetone dilution series [25, 50 and 75% (v/v)], and embedded in epoxy resin. After polymerization at 65°C for 18 h, serial cross-sections were prepared and examined with a JEM1230 transmission electron microscope (Hitachi, Tokyo, Japan). Genetic analysis of the mgp3 mutants All crosses and genetic analysis were performed as described previously (Yang et al. 1999, Yang et al. 2003). Molecular cloning of MGP3 and complementation experiments Isolation of the flanking sequences adjacent to the Ds element by TAIL-PCR (Liu et al. 1995) was performed as described previously (Yang et al. 1999, Yang et al. 2003) with mgp3-1/+ genomic DNA and the Ds5′/AD6 primer set (Grossniklaus et al.1998). The insertion site was then confirmed by PCR using primer pairs P2/Ds5-2 (5′-CGTTCCGTTTTCGTTTTTTACC-3′) and P1 (5′-TGGAGGAACATTCTGGG-3′)/Ds3-2 (5′-CCGGTATATCC CGTTTTCG-3′). The full-length MGP3 genomic DNA fragment was amplified by PCR with the primers MGP3-F-P1 (5′-GCAATGAACAG TAGTGTTGG-3′) and MGP3-F-P2 (5′-GCGGATGCGTTCTTG AATGC-3′). The resulting DNA fragment was cloned into a pMD18-T vector (TaKaRa, Dalian, China) for sequencing. For complementation experiments, the full-length MGP3 genomic DNA fragment was subcloned into the pCAMBIA1300 vector (CAMBIA, Canberra, Australia) and introduced into mgp3-1 and mgp3-2 heterozygous plants using the Agrobacterium- mediated infiltration method. The transformants were selected in Murashige and Skoog (MS) medium (Murashige and Skoog 1962) containing 25 mg l−1 of hygromycin (Roche, Mannheim, Germany) and 50 mg l−1 of kanamycin (Sigma, Saint Louis, Missouri, USA). All the transformants were confirmed by PCR using the primer pair 1300-H (5′-TGGCGAAAGGGGGA TGTGCTG-3′)/P2. The transformants homozygous for mgp3-1 or mgp3-2 mutation were selected in T2 generations and used for further analyses. Southern blotting hybridization The DNA preparation and Southern blotting hybridization were as described by Jia et al. (2009) and Yang et al. (1999). Funding This work was supported by the Ministry of Sciences and Technology [973 project number: 2007CB108700]; the Natural Science Foundation of China [project number: 30530060]; the Ministry of Education, PR of China [111 project number B06003]. Acknowledgments We thank Dr. V. Sundaresan, Dr. Wei-Cai Yang and Mrs. Li-Fen Xie for their kind help with the mutant screens. References Alexander MP.  Differential staining of aborted and nonaborted pollen,  Stain Technol. ,  1969, vol.  44 (pg.  117- 122) Google Scholar CrossRef Search ADS PubMed  Baba K,  Schmidt J,  Espinosa-Ruiz A,  Villarejo A,  Shiina T,  Gardeström P, et al.  Organellar gene transcription and early seedling development are affected in the rpoT;2 mutant of Arabidopsis,  Plant J. ,  2004, vol.  38 (pg.  38- 48) Google Scholar CrossRef Search ADS PubMed  Binder S,  Brennicke A.  Gene expression in plant mitochondria: transcriptional and post-transcriptional control,  Philos. Trans. R. Soc. Ser. B: Biol. Sci. ,  2003, vol.  358 (pg.  181- 189) Google Scholar CrossRef Search ADS   Carlsson J,  Leino M,  Sohlberg J,  Sundström JF,  Glimelius K.  Mitochondrial regulation of flower development,  Mitochondrion ,  2008, vol.  8 (pg.  74- 86) Google Scholar CrossRef Search ADS PubMed  Cermakian N,  Ikeda TM,  Cedergren R,  Gray MW.  Sequences homologous to yeast mitochondrial and bacteriophage T3 and T7 RNA polymerases are widespread throughout the eukaryotic lineage,  Nucleic Acids Res. ,  1996, vol.  24 (pg.  648- 654) Google Scholar CrossRef Search ADS PubMed  Christensen CA,  King EJ,  Jordan JR,  Drews GN.  Megagametogenesis in Arabidopsis wild type and the gf mutant,  Sex. Plant Reprod. ,  1997, vol.  10 (pg.  49- 64) Google Scholar CrossRef Search ADS   Christensen CA,  Gorsich SW,  Brown RH,  Jones LG,  Brown J,  Shaw JM, et al.  Mitochondrial GFA2 is required for synergid cell death in Arabidopsis,  Plant Cell ,  2002, vol.  14 (pg.  2215- 2232) Google Scholar CrossRef Search ADS PubMed  Conley CA,  Hanson MR.  How do alterations in plant mitochondrial genomes disrupt pollen development?,  J. Bioenerg. Biomembr. ,  1995, vol.  27 (pg.  447- 457) Google Scholar CrossRef Search ADS PubMed  Drews GN,  Lee D,  Christensen CA.  Genetic analysis of female gametophyte development and function,  Plant Cell ,  1998, vol.  10 (pg.  5- 17) Google Scholar CrossRef Search ADS PubMed  Emanuel C,  von Groll U,  Müller M,  Börner T,  Weihe A.  Development- and tissue-specific expression of the RpoT gene family of Arabidopsis encoding mitochondrial and plastid RNA polymerases,  Planta ,  2006, vol.  223 (pg.  998- 1009) Google Scholar CrossRef Search ADS PubMed  Gilkerson RW,  Margineantu DH,  Capaldi RA,  Selker JML.  Mitochondrial DNA depletion causes morphological changes in the mitochondrial reticulum of cultured human cells,  FEBS Letters ,  2000, vol.  474 (pg.  1- 4) Google Scholar CrossRef Search ADS PubMed  Greenleaf AL,  Kelly JL,  Lehman IR.  Yeast RP041 gene product is required for transcription and maintenance of the mitochondrial genome,  Proc. Natl Acad. Sci. USA ,  1986, vol.  83 (pg.  3391- 3394) Google Scholar CrossRef Search ADS   Grossniklaus U,  Vielle-Calzada JP,  Hoeppner MA,  Gagliano WB.  Maternal control of embryogenesis by MEDEA, a polycomb group gene in Arabidopsis,  Science ,  1998, vol.  280 (pg.  446- 450) Google Scholar CrossRef Search ADS PubMed  Hanson MR,  Bentolila S.  Interactions of mitochondrial and nuclear genes that affect male gametophyte development,  Plant Cell ,  2004, vol.  16 (pg.  S154- S169) Google Scholar CrossRef Search ADS PubMed  Hedtke B,  Börner T,  Weihe A.  Mitochondrial and chloroplast phage-type RNA polymerases in Arabidopsis,  Science ,  1997, vol.  277 (pg.  809- 811) Google Scholar CrossRef Search ADS PubMed  Hedtke B,  Meixner M,  Gillandt S,  Richter E,  Börner T,  Weihe A.  Green fluorescent protein as a marker to investigate targeting of organellar RNA polymerases of higher plants in vivo,  Plant J. ,  1999, vol.  17 (pg.  557- 561) Google Scholar CrossRef Search ADS PubMed  Hedtke B,  Börner T,  Weihe A.  One RNA polymerase serving two genomes,  EMBO Rep. ,  2000, vol.  1 (pg.  435- 440) Google Scholar CrossRef Search ADS PubMed  Huang J,  Struck F,  Matzinger DF,  Levings CS,  III.  Flower-enhanced expression of a nuclear-encoded mitochondrial respiratory protein is associated with changes in mitochondrion number,  Plant Cell ,  1994, vol.  6 (pg.  439- 448) Google Scholar CrossRef Search ADS PubMed  Jia DJ,  Cao X,  Wang W,  Tan XY,  Zhang XQ,  Chen LQ, et al.  GNOM-LIKE 2, encoding an ARF-GEF protein homologous to GNOM and GNL1, is essential for pollen tube growth in Arabidopsis,  J. Integr. Plant Biol. ,  2009, vol.  51 (pg.  762- 773) Google Scholar CrossRef Search ADS PubMed  Jiang L,  Yang SL,  Xie LF,  Puah CS,  Zhang XQ,  Yang WC, et al.  VANGUARD1 encodes a pectin methylesterase that enhances pollen tube growth in the Arabidopsis style and transmitting tract,  Plant Cell ,  2005, vol.  17 (pg.  584- 596) Google Scholar CrossRef Search ADS PubMed  Kelly JL,  Greenleaf AL,  Lehman IR.  Isolation of the nuclear gene encoding a subunit of the yeast mitochondrial RNA polymerase,  J. Biol. Chem. ,  1986, vol.  261 (pg.  10348- 10351) Google Scholar PubMed  Krichevsky A,  Kozlovsky SV,  Tian GW,  Chen MH,  Zaltsman A,  Citovsky V.  How pollen tubes grow,  Dev. Biol. ,  2007, vol.  303 (pg.  405- 420) Google Scholar CrossRef Search ADS PubMed  Kühn K,  Bohne AV,  Liere K,  Weihe A,  Börner T.  Arabidopsis phage-type RNA polymerases: accurate in vitro transcription of organellar genes,  Plant Cell ,  2007, vol.  19 (pg.  959- 971) Google Scholar CrossRef Search ADS PubMed  Kühn K,  Richter U,  Meyer EH,  Delannoy E,  Falcon de Longevialle A,  O’Toole N, et al.  Phage-type RNA polymerase RPOTmp performs gene-specific transcription in mitochondria of Arabidopsis thaliana,  Plant Cell ,  2009, vol.  21 (pg.  2762- 2779) Google Scholar CrossRef Search ADS PubMed  León G,  Holuigue L,  Jordana X.  Mitochondrial complex II is essential for gametophyte development in Arabidopsis,  Plant Physiol. ,  2007, vol.  143 (pg.  1534- 1546) Google Scholar CrossRef Search ADS PubMed  Linke B,  Börner T.  Mitochondrial effects on flower and pollen development,  Mitochondrion ,  2005, vol.  5 (pg.  389- 402) Google Scholar CrossRef Search ADS PubMed  Liu CM,  Meinke DW.  The titan mutants of Arabidopsis are disrupted in mitosis and cell cycle control during seed development,  Plant J. ,  1998, vol.  16 (pg.  21- 31) Google Scholar CrossRef Search ADS PubMed  Liu Y.-G,  Mitsukawa N,  Oosumi T,  Whittier RF.  Efficient isolation and mapping of Arabidopsis thaliana T-DNA insert junctions by thermal asymmetric interlaced PCR,  Plant J. ,  1995, vol.  8 (pg.  457- 463) Google Scholar CrossRef Search ADS PubMed  McCormick S.  Male gametophyte development,  Plant Cell ,  1993, vol.  5 (pg.  1265- 1275) Google Scholar CrossRef Search ADS PubMed  McCormick S.  Control of male gametophyte development,  Plant Cell ,  2004, vol.  16 (pg.  S142- 153) Google Scholar CrossRef Search ADS PubMed  Millar AH,  Day DA,  Whelan J.  Somerville CR,  Meyerowitz EM.  Mitochondrial biogenesis and function in Arabidopsis,  The Arabidopsis Book ,  2004 Rockville, MD American Society of Plant Biologists  doi/10.1199/tab.0105, http://www.aspb .org/publications/arabidopsis/ Monéger F,  Mandaron P,  Niogret MF,  Freyssinet G,  Mache R.  Expression of chloroplast and mitochondrial genes during microsporogenesis in maize,  Plant Physiol. ,  1992, vol.  99 (pg.  396- 400) Google Scholar CrossRef Search ADS PubMed  Murashige T,  Skoog F.  A revised medium for rapid growth and bioassays with tobacco tissue cultures,  Physiol. Plant. ,  1962, vol.  15 (pg.  473- 497) Google Scholar CrossRef Search ADS   Parton RM,  Fischer-Parton S,  Trewavas AJ,  Watahiki MK.  Pollen tubes exhibit regular periodic membrane trafficking events in the absence of apical extension,  J. Cell Sci. ,  2003, vol.  116 (pg.  2707- 2719) Google Scholar CrossRef Search ADS PubMed  Portereiko MF,  Sandaklie-Nikolova L,  Lloyd A,  Dever CA,  Otsuga D,  Drews GN.  NUCLEAR FUSION DEFECTIVE1 encodes the Arabidopsis RPL21M protein and is required for karyogamy during female gametophyte development and fertilization,  Plant Physiol. ,  2006, vol.  141 (pg.  957- 965) Google Scholar CrossRef Search ADS PubMed  Qin Y,  Leydon AR,  Manziello A,  Pandey R,  Mount D,  Denic S, et al.  Penetration of the stigma and style elicits a novel transcriptome in pollen tubes, pointing to genes critical for growth in a pistil,  PLoS Genet. ,  2009, vol.  5 pg.  e1000621  Google Scholar CrossRef Search ADS PubMed  Shi DQ,  Liu J,  Xiang YH,  Ye D,  Sundaresan V,  Yang WC.  SLOW WALKER1, essential for gametogenesis in Arabidopsis, encodes a WD40 protein involved in 18S ribosomal RNA biogenesis,  Plant Cell ,  2005, vol.  17 (pg.  2340- 2354) Google Scholar CrossRef Search ADS PubMed  Skinner DJ,  Baker SC,  Meister RJ,  Broadhvest J,  Schneitz K,  Gasser CS.  The Arabidopsis HUELLENLOS gene, which is essential for normal ovule development, encodes a mitochondrial ribosomal protein,  Plant Cell ,  2001, vol.  13 (pg.  2719- 2730) Google Scholar CrossRef Search ADS PubMed  Spencer MWB,  Casson SA,  Lindsey K.  Transcriptional profiling of the Arabidopsis embryo,  Plant Physiol. ,  2007, vol.  143 (pg.  924- 940) Google Scholar CrossRef Search ADS PubMed  Sundaresan V,  Springer P,  Volpe T,  Haward S,  Jones JDG,  Dean C, et al.  Patterns of gene action in plant development revealed by enhancer trap and gene trap transposable elements,  Genes Dev. ,  1995, vol.  9 (pg.  1797- 1810) Google Scholar CrossRef Search ADS PubMed  Taylor LP,  Hepler PK.  Pollen germination and tube growth,  Annu. Rev. Plant Physiol. Plant Mol. Biol. ,  1997, vol.  48 (pg.  461- 491) Google Scholar CrossRef Search ADS PubMed  Unseld M,  Marienfeld JR,  Brandt P,  Brennicke A.  The mitochondrial genome of Arabidopsis thaliana contains 57 genes in 366,924 nucleotides,  Nat. Genet. ,  1997, vol.  15 (pg.  57- 61) Google Scholar CrossRef Search ADS PubMed  Warmke HE,  Lee SLJ.  Pollen abortion in T cytoplasmic male-sterile corn (Zea mays): a suggested mechanism,  Science ,  1978, vol.  200 (pg.  561- 563) Google Scholar CrossRef Search ADS PubMed  Wen LY,  Chase CD.  Mitochondrial gene expression in developing male gametophytes of male-fertile and S male-sterile maize,  Sex. Plant Reprod. ,  1999, vol.  11 (pg.  323- 330) Google Scholar CrossRef Search ADS   Yadegari R,  Drews GN.  Female gametophyte development,  Plant Cell ,  2004, vol.  16 (pg.  S133- 141) Google Scholar CrossRef Search ADS PubMed  Yamaoka S,  Leaver CJ.  EMB2473/MIRO1, an Arabidopsis Miro GTPase, is required for embryogenesis and influences mitochondrial morphology in pollen,  Plant Cell ,  2008, vol.  20 (pg.  589- 601) Google Scholar CrossRef Search ADS PubMed  Yang KZ,  Xia C,  Liu XL,  Dou XY,  Wang W,  Chen LQ, et al.  A mutation in THERMOSENSITIVE MALE STERILE 1, encoding a heat shock protein with DnaJ and PDI domains, leads to thermosensitive gametophytic male sterility in Arabidopsis,  Plant J. ,  2009, vol.  57 (pg.  870- 882) Google Scholar CrossRef Search ADS PubMed  Yang SL,  Xie LF,  Mao HZ,  Puah CS,  Yang WC,  Jiang L, et al.  Tapetum determinant 1 is required for cell specialization in the Arabidopsis anther,  Plant Cell ,  2003, vol.  15 (pg.  2792- 2804) Google Scholar CrossRef Search ADS PubMed  Yang WC,  Ye D,  Xu J,  Sundaresan V.  The SPOROCYTELESS gene of Arabidopsis is required for initiation of sporogenesis and encodes a novel nuclear protein,  Genes Dev. ,  1999, vol.  13 (pg.  2108- 2117) Google Scholar CrossRef Search ADS PubMed  Abbreviations Abbreviations ABRC Arabidopsis Biological Resource Center CLSM confocal laser scanning microscopy DAPI 4′,6-diamidino-2-phenylindole DIC differential interference contrast microscopy GUS β-glucuronidase Kan kanamycin RPOT RNA polymerase of T3/T7 type RNAP RNA polymerase RT–PCR reverse transcriptase–PCR SEM scanning electron microscopy TAIL-PCR thermal asymmetric interlaced PCR TEM transmission electron microscopy. © The Author 2010. Published by Oxford University Press on behalf of Japanese Society of Plant Physiologists. All rights reserved. For permissions, please email: journals.permissions@oxfordjournals.org

Journal

Plant and Cell Physiology – Oxford University Press

Published: Apr 8, 2010

Keywords: Arabidopsis Embryogenesis Gametogenesis Mitochondria Phage-type RNA polymerase Pollen

There are no references for this article.