TY - JOUR
AU - Sze, Heven
AB - Abstract Flowering plant genomes encode multiple cation/H+ exchangers (CHXs) whose functions are largely unknown. AtCHX17, AtCHX18, and AtCHX19 are membrane transporters that modulate K+ and pH homeostasis and are localized in the dynamic endomembrane system. Loss of function reduced seed set, but the particular phase(s) of reproduction affected was not determined. Pollen tube growth and ovule targeting of chx17chx18chx19 mutant pollen appeared normal, but reciprocal cross experiments indicate a largely male defect. Although triple mutant pollen tubes reach ovules of a wild-type pistil and a synergid cell degenerated, half of those ovules were unfertilized or showed fertilization of the egg or central cell, but not both female gametes. Fertility could be partially compromised by impaired pollen tube and/or sperm function as CHX19 and CHX18 are expressed in the pollen tube and sperm cell, respectively. When fertilization was successful in self-pollinated mutants, early embryo formation was retarded compared with embryos from wild-type ovules receiving mutant pollen. Thus CHX17 and CHX18 proteins may promote embryo development possibly through the endosperm where these genes are expressed. The reticulate pattern of the pollen wall was disorganized in triple mutants, indicating perturbation of wall formation during male gametophyte development. As pH and cation homeostasis mediated by AtCHX17 affect membrane trafficking and cargo delivery, these results suggest that male fertility, sperm function, and embryo development are dependent on proper cargo sorting and secretion that remodel cell walls, plasma membranes, and extracellular factors. Cation, endomembrane, male fertility, pollen, proton, transporter, wall formation Introduction Early land plants, such as mosses and ferns, reproduce using motile male gametes that can swim through an aqueous medium to the female gametes. However, in flowering plants, two sperm cells are carried by a male gametophyte (pollen) to the egg cell and the central cell in the absence of an aqueous environment. After a pollen grain lands on a receptive stigma, it germinates and extends a pollen tube inside pistil tissues that surround the female gametophyte. Once inside the embryo sac, the tube ruptures to release two sperm. One fuses with the egg cell and the other with the central cell to complete double fertilization (Dresselhaus and Franklin-Tong, 2013). Subsequent development of the embryo and the endosperm produces a desiccated seed that can remain viable for long periods of time. Although the major steps of reproduction have been known for decades, the molecular and cellular bases of pollen–pistil interactions that culminate in successful double fertilization are not well understood. Recent studies have identified key molecular players in pollen–embryo sac interactions, including pollen tube attractants secreted by synergid (Okuda et al., 2009; Higashiyama and Takeuchi, 2015) and egg cells (Márton et al., 2005); receptor-like kinases at the plasma membrane (PM) of pollen (Boisson-Dernier et al., 2009) and synergid cells (Escobar-Restrepo et al., 2007); and cell surface proteins of the sperm (HAP2/GCS1) (Mori et al., 2006; von Besser et al., 2006). Pollen germination and tip growth are accompanied by fluxes and oscillations of [Ca2+], [K+], or pH (Feijó et al., 2001). Many transporter genes are preferentially expressed in pollen grains and tubes (Bock et al., 2006; Qin et al., 2009), though their biological roles are mostly unknown. ACA9 is a pollen-specific Ca2+ extrusion pump that is localized to the pollen tube PM. T-DNA insertional mutants show reduced pollen tube growth in vivo and aborted fertilization due in part to failed pollen tube rupture (Schiott et al., 2004). Thus disruption of Ca2+ efflux by ACA9 probably caused a defect in Ca2+ signaling, resulting in slowed tube growth, failed tube rupture, and reduced seed set. A single mutant of a cyclic nucleotide-gated ion channel, cngc18, showed male sterility due to a defect in pollen tube growth and inability to enter the transmitting tract (Frietsch et al., 2007). However, the single mutants cngc7 or cngc8 showed no defect, though pollen grains of the double mutants were sterile, possibly due to a tendency to burst before the tube emerged (Tunc-Ozdemir et al., 2013). These experiments suggest that cNMP may stimulate Ca2+ influx or oscillations required for pollen germination and tip growth. Very little is known about transporters that mediate pH and cation homeostasis in pollen–female gametophyte interactions. The diversification of one transporter subfamily, CHXs (cation /H+ exchangers), in monocot and eudicot plants relative to early land plants is striking, especially as ~18 Arabidopsis thaliana CHX genes are expressed in pollen (Sze et al., 2004). The Cation/Proton Antiporter1 (CPA1) subfamily includes mammalian NHE and plant NHXs (Brett et al., 2005; Chanroj et al., 2012). The CPA2 subfamily (including AtCHX) is predominantly conserved in bacteria, archea, fungi, amoebozoa, and plants, but not in metazoa (Chanroj et al., 2012). Basal land plants, such as moss, contain 3–4 copies of CHX genes per genome (Mottaleb et al., 2013), but ~17 and 28 members are predicted in rice and Arabidopsis, respectively. Their biological roles are mostly unexplored. We proposed that the diversification of CHXs in flowering plants is related to reproductive success (Chanroj et al., 2012), and the ability to survive on land. Several studies support the idea that expansion of the CHX gene family facilitated colonization of land. First, CHX21 and CHX23 were necessary for pollen tube targeting to ovules in A. thaliana. Mutant chx21chx23 pollen grains germinate, but tubes fail to turn towards the ovule and thus do not reach the micropyle. Thus, CHX21 and CHX23 could function either in pollen tube perception of female guidance cues or in signal transduction to mediate a change in growth direction (Lu et al., 2011). Secondly, AtCHX20 in guard cells facilitates light-induced stomatal opening (Padmanaban et al., 2007). Localized to reticulate membranes resembling the endoplasmic reticulum (ER), CHX20 could regulate osmoregulation via membrane trafficking needed to increase vacuolar and PM area. Recently, GmSALT3 or GmCHX1, which encodes a homolog of AtCHX20, was shown to confer salt tolerance in certain soybean varieties (Guan et al., 2014; Qi et al., 2014). The best characterized CHXs from A. thaliana show roles in pH and cation homeostasis. CHX17, CHX18, or CHX19 are localized to the pre-vacuolar compartments (PVCs) and the PM in plant cells (Chanroj et al., 2011, 2013). Their expression in an alkaline-sensitive yeast strain conferred tolerance to growth at pH 7.5, suggesting a role in pH homeostasis. CHX17 also mediated K+ transport, as its expression in Escherichia coli strains deficient in K+ uptake pathways restored growth on low K+ medium and mediated 86Rb+ uptake. Furthermore, CHX17 as well as CHX18 and CHX19 restored growth of yeast lacking K+ uptake transporters (Chanroj et al., 2011). These results in yeast are consistent with their role as a K+/H+ antiporter at the endomembrane; though the mode of transport at the PM is less clear (Chanroj et al., 2011; Mottaleb et al., 2013). Structural modeling predicts that AtCHX17 protein has an NhaA-fold architecture, and mutagenesis showed core residues at positions similar to cation/H+ exchangers (Czerny et al., 2016). The roles of CHX17, CHX18, and CHX19 in planta are less clear. CHX17 is expressed in roots, though vegetative growth of seedlings in single, double, or triple mutants (chx17chx18chx19) under various stress conditions was unaltered (Chanroj, 2011). Curiously, quadruple mutants (chx16chx17chx18chx19) were recovered at a lower frequency than expected. Furthermore, ther quadruple mutant pod contained 60% fewer seeds than the wild type (Chanroj et al., 2013). Homozygous triple mutants, chx17chx18chx19, were also recovered at a lower than expected frequency. Here, we determine which phase of reproduction led to reduced seed set in mutants. We demonstrate that chx17chx18chx19 pollen tubes grow, target, and enter ovules; however, many targeted ovules fail to develop into seeds, suggesting a failure to complete fertilization. Our findings underscore a previously unknown role for endomembrane transporters and K+ and pH homeostasis in male fertility, fertilization, and embryo development possibly through remodeling of the cell wall and PM. Materials and methods Plants and genotyping Arabidopsis thaliana, Columbia-0, was grown in Miracle-Gro® potting mix supplemented with 5% Perlite. Plants were grown under a 16 h photoperiod at 150 μE m–2 s–1 illumination, 21 °C, and 60% relative humidity. All chx mutants used (see Supplementary Table S1 at JXB online) are available from ABRC. Genotype was determined by PCR using two gene-specific primers or one gene-specific primer and a T-DNA border primer (Supplementary Table S2). Genomic DNA was isolated from 1–2 rosette leaves of ~5-week-old plants (Edwards et al., 1991). DNA was amplified by Taq DNA polymerase (NEB M0273L) in a three-step PCR program for 40 cycles. PCR products were separated on a 1% agarose gel and visualized using ethidium bromide. Segregation analysis To test male transmission of the mutant allele, pollen from a parent carrying a heterozygous gene was transferred to a pistil carrying the wild-type gene. To test female transmission, a stigma heterozygous for a gene was given pollen harboring the wild-type gene. Parental plants were genotyped for wild-type genomic sequence or T-DNA insertions in CHX17, CHX18, and CHX19 genes. Stage 12 flower buds (Smyth et al., 1990) were emasculated, and pistils were pollinated 2 d later, with anther from stage 13 flowers. Seeds developed on the plant for 10–12 d until pods dried. F1 seeds were planted, and genomic DNA from ~100 plants was extracted for genotype analyses. Analysis of wild-type and chx17/18/19 plants Siliques from the primary bolts of 9-week-old plants were collected starting from the fourth pod below the inflorescence. Pod lengths were measured in ImageJ (NIH) using at least 10 plants per line. Four pods of average length per plant were split open and seeds were scored under a stereomicroscope (Nikon SMZ1000). To detect nuclei, pollen grains from 1–3 stage 13 flowers (Smyth et al., 1990) were dabbed onto a microscope slide, and incubated with 30 µl of DAPI solution for 15 min. DAPI (Life Technologies D3571) solution contained 0.1 M sodium phosphate pH 7, 1 mM EDTA, 0.1% Triton X-100, and 0.4 µg ml–1 DAPI. Pollen was examined by Nikon E600 with a UV filter (excitation/emission at 360/410 nm). To visualize pollen tubes in vivo, pollinated pistils were stained with aniline blue to label callose (Mori et al., 2006). Self-fertilized wild-type and chx17/18/19 pistils were analyzed 1–4 days after pollination (DAP) (Smyth et al., 1990). Pistils were fixed in 75% ethanol/25% acetic acid for 2 h, and then rehydrated in 75, 50, and 30% ethanol, and deionized water. The pistils were incubated in 8 M NaOH overnight, washed once with deionized water, and then incubated for 2 h in 0.1% decolorized aniline blue (Fisher Scientific A-967) in 100 mM K2HPO4 at pH 10. Each pistil was placed on a glass slide and pressure was applied to expose ovules and the transmitting tract. Images were recorded using a Nikon E600 fluorescence microscope with a UV filter (excitation/emission wavelengths at 360 nm/410 nm). Ovule development from self- or manually pollinated pistils was visualized using chloral hydrate. Seed pods were fixed in 90% ethanol/10% acetic acid overnight at room temperature. Pods were washed twice for 30 min in 90% ethanol and then cleared in chloral hydrate/glycerol/water (8:1:2) solution (Berleth and Jurgens, 1993). Each pod was opened and all ovules or developing seeds were mounted in chloral hydrate solution and examined with DIC (differential interference contrast) microscopy. To examine pollen wall architecture, pollen grains from stage 13 flowers were mounted on stubs over double-sided tape. The specimens were then sputter coated with gold–palladium (60%:40%) (Balzers Med 010) and observed using a scanning electron microscope (SU3500, Hitachi) at an accelerating voltage of 3 kV in high vacuum. Monitoring synergid degeneration and gamete fusion To visualize sperm nuclei, HTR10:HTR10:RFP was introduced into the chx17chx18chx19 mutant by genetic crossing. Self-fertilized chx triple mutants with red fluorescent protein (RFP)-labeled sperm were recovered. Oddly, plants showed a range of RFP-labeled pollen, 20–50%, and a few plants with >55% RFP. A triple chx mutant containing 73% RFP-labeled pollen was selfed, and progeny yielded flowers in which 85% of the pollen was RFP positive. Stage 12b or 12c buds (Smyth et al., 1990) were emasculated and hand pollinated 24 h later. Hand pollinations were performed under a dissecting microscope (Zeiss Stemi 2000C) by pollinating emasculated ACT11:MSI1:GFP stigmas with pollen from male donor plants carrying HTR10:HTR10:RFP (Ingouff et al., 2007). Pollinated pistils were analyzed 16 h after manual pollinations. Pistils were harvested and ovary walls were removed as previously described (Johnson and Kost, 2010). Dissected pistil tissues were mounted in 80 mM sorbitol for analysis by confocal laser scanning microscopy (CLSM) using a ×40 objective with DIC capability (Zeiss LSM800 upright microscope). Green fluorescent protein (GFP) expression was imaged using a diode laser at 10 mW with excitation at 488 nm and emission at 509 nm. mRFP (modified RFP) expression was imaged with a diode laser at 10 mW with excitation at 633 nm and emission at 607 nm. Signal intensities were optimized for each fluorophore and then combined in overlay. Final images represent a merge of single planes at varying depths (z stacks). Synergid status was determined based upon the visualization of nuclear ACT11:MSI1:GFP signal as described before (Leydon et al., 2015). Accession numbers CHX17 (At4g23700), CHX18 (At5g41610), and CHX19 (At3g17630) Results chx17chx18chx19 mutants show reduced seed set Vegetative growth and flower development of chx17chx18chx19 plants were similar to those of the wild type (Fig. 1A-i; Supplementary Fig. S1). However, seed pods of chx17chx18chx19 plants were shorter, varying from 8.8 mm to 10.3 mm, compared with 13.0 mm in the wild type (Fig. 1A-ii, B; Supplementary Fig. S2). When siliques were cleared, mutant pods had random gaps instead of two continuous rows of seeds in wild-type pods (Fig. 1A-iii, iv). Progeny from three chx17chx18chx19 plants showed, on average, 11–21 seeds, compared with 48 seeds per pod in the wild type (Fig. 1C). Seed set or pod length of double mutants chx17chx18 or chx17chx19 were similar to those of the wild type (Fig. 1B, C). Thus, loss of function of three CHX genes, CHX17,CHX18, andCHX19, reduced seed set to half that seen in the wild type. Fig. 1. Open in new tabDownload slide Triple chx17chx18chx19 mutant plants showed reduced seed set. (A) Plants and seed pods. (i) Wild-type (WT) Columbia-0 and chx17/18/19 plants have similar vegetative and reproductive growth. (ii) chx17/18/19 mutant pods were shorter than those of the WT. (iii) chx17/18/19 pods contained fewer seeds. Scale bar=1.0 mm. (iv) chx17/18/19 pods contained aborted and developing seeds. At least 44 pods per genotype were analyzed. Scale bar=0.5 mm. (B) Pod lengths of the chx17/18/19 mutant are 21–29% shorter than pods of the WT or chx17/18 or chx17/19 double mutants. Results show the mean, and bars represent the SE (n=234–610 pods). (C) Seed number per pod was reduced 55–73% in the chx17/18/19 mutant relative to the WT or the chx17/18 and chx17/19 double mutants. Four siliques of average length were split open per plant to count seeds. Results show the mean, and bars represent the SE (n=44–126 pods). Male fertility is compromised in chx17chx18chx19 pollen Segregation distortion in the progeny from a selfed mutant carrying heterozygous CHX18+/– in a double chx17chx19 mutant background (Chanroj et al., 2013) indicated a gametophytic defect. Reciprocal crosses were conducted to determine whether transmission of T-DNA-disrupted genes was due to the male or the female gametophyte, or both. As double mutants behaved like the wild type, we first tested segregation of a heterozygous CHX18+/– or CHX19+/– gene in a double mutant background crossed to the same double mutant. For instance, pollen from a double chx17–/–chx19–/– mutant plant carrying a heterozygous CHX18+/– gene was placed onto the same double mutant pistil carrying CHX18+/+ (Table 1). Transmission of chx18-1 and chx19-2 mutant alleles was tested by genotyping the F1 progeny. Parents were genotyped for mutant or wild-type alleles of CHX17, CHX18, and CHX19 by PCR prior to crossing (Fig. 2A, B). The F1 progeny were tested for homozygous wild-type CHX18+/+ or heterozygous CHX18+/–. PCR-amplified products of 1 kb or 0.5 kb were used to distinguish either the wild-type or mutant allele, respectively (Fig. 2C). If a CHX gene has minimal or no effect in male fertility, then progeny containing homozygous wild-type CHX18+/+ would be approximately equal to progeny carrying heterozygous CHX18+/–. Table 1. Only one of three CHX genes is sufficient to restore male fertility Pollen grains from a double mutant chx17–/–chx19–/– parent heterozygous for CHX18+/– were manually transferred to a double mutant pistil carrying wild-type CHX18+/+ (a). Reciprocal crosses were also conducted (b). Similar reciprocal crosses were performed with the double mutant chx17–/–chx18–/– heterozygous for CHX19+/– (c, d) and the double mutant alone. Seeds from these crosses were planted and leaves were collected for DNA extraction and genotyped. The observed (Obs) segregation of F1 progeny is shown below as the number of individuals recovered or percentage (%). . Female: egg or central cell . Male: pollen or sperm . F 1 progeny . F 1 progeny . . CHX . CHX . CHX . Expect . Obs . . 17 . 18 . 19 . 17 . 18 . 19 . 17 . 18 . 19 . (%) . (%) . a – + – – + – –/– +/+ –/– 50 62 (95) – + – – – – –/– +/– –/– 50 3 (5) b – + – – + – –/– +/+ –/– 50 98 (70) – – – – + – –/– +/– –/– 50 41 (30) c – – + – – + –/– –/– +/+ 50 92 (97) – – + – – – –/– –/– +/– 50 3 (3) d – – + – – + –/– –/– +/+ 50 69 (75) – – – – – + –/– –/– +/– 50 23 (25) . Female: egg or central cell . Male: pollen or sperm . F 1 progeny . F 1 progeny . . CHX . CHX . CHX . Expect . Obs . . 17 . 18 . 19 . 17 . 18 . 19 . 17 . 18 . 19 . (%) . (%) . a – + – – + – –/– +/+ –/– 50 62 (95) – + – – – – –/– +/– –/– 50 3 (5) b – + – – + – –/– +/+ –/– 50 98 (70) – – – – + – –/– +/– –/– 50 41 (30) c – – + – – + –/– –/– +/+ 50 92 (97) – – + – – – –/– –/– +/– 50 3 (3) d – – + – – + –/– –/– +/+ 50 69 (75) – – – – – + –/– –/– +/– 50 23 (25) Open in new tab Table 1. Only one of three CHX genes is sufficient to restore male fertility Pollen grains from a double mutant chx17–/–chx19–/– parent heterozygous for CHX18+/– were manually transferred to a double mutant pistil carrying wild-type CHX18+/+ (a). Reciprocal crosses were also conducted (b). Similar reciprocal crosses were performed with the double mutant chx17–/–chx18–/– heterozygous for CHX19+/– (c, d) and the double mutant alone. Seeds from these crosses were planted and leaves were collected for DNA extraction and genotyped. The observed (Obs) segregation of F1 progeny is shown below as the number of individuals recovered or percentage (%). . Female: egg or central cell . Male: pollen or sperm . F 1 progeny . F 1 progeny . . CHX . CHX . CHX . Expect . Obs . . 17 . 18 . 19 . 17 . 18 . 19 . 17 . 18 . 19 . (%) . (%) . a – + – – + – –/– +/+ –/– 50 62 (95) – + – – – – –/– +/– –/– 50 3 (5) b – + – – + – –/– +/+ –/– 50 98 (70) – – – – + – –/– +/– –/– 50 41 (30) c – – + – – + –/– –/– +/+ 50 92 (97) – – + – – – –/– –/– +/– 50 3 (3) d – – + – – + –/– –/– +/+ 50 69 (75) – – – – – + –/– –/– +/– 50 23 (25) . Female: egg or central cell . Male: pollen or sperm . F 1 progeny . F 1 progeny . . CHX . CHX . CHX . Expect . Obs . . 17 . 18 . 19 . 17 . 18 . 19 . 17 . 18 . 19 . (%) . (%) . a – + – – + – –/– +/+ –/– 50 62 (95) – + – – – – –/– +/– –/– 50 3 (5) b – + – – + – –/– +/+ –/– 50 98 (70) – – – – + – –/– +/– –/– 50 41 (30) c – – + – – + –/– –/– +/+ 50 92 (97) – – + – – – –/– –/– +/– 50 3 (3) d – – + – – + –/– –/– +/+ 50 69 (75) – – – – – + –/– –/– +/– 50 23 (25) Open in new tab Fig. 2. Open in new tabDownload slide Gene structures of Arabidopsis CHX17, CHX18, and CHX19 and verification of genotype. (A) CHX17, CHX18, and CHX19 genes and site of T-DNA insertions. Exons and introns are shown as boxes and lines, respectively. Bold arrows indicate right and left border primers (RP and LP) used to amplify the wild-type CHX sequence. Arrows refer to the left border primer (LBb1 or LBb1.3) used to detect the T-DNA insertion. Number 1 refers to the first base of the start codon ATG, and the number at the end of the gene refers to the third base of the stop codon. Base # above the triangle denotes the position of the T-DNA insertion. chx17-4, chx18-1, and chx19-2 correspond to SALK_033417, SALK_001563, and SALK_100047, respectively. Primer sequences are shown in Supplementary Table S2. (B) The genotype of male (chx17–/–chx18+/–chx19–/–) and female (chx17–/–chx18+/+chx19–/–) parents used in a reciprocal cross was verified. Genomic DNA from parents was tested for homozygous or heterozygous CHX17, CHX18, or CHX19 genes by PCR. Lane ‘a’ shows fragments amplified with LP and RP primer pairs, and lane b shows fragments amplified using the T-DNA primer with either LP or RP primers. The wild-type allele produces a PCR-amplified product of ~1 kb, whereas mutant alleles give a PCR product of ~0.5 kb. (C) The number of F1 progeny heterozygous for CHX18 was less than expected. Lanes 1–7 show a sample of F1 plants that were genotyped. Lanes labeled ‘18LP–18RP primer pair’ or ‘18LP–LBb1’ show the PCR-amplified DNA fragment if genomic DNA had a copy of the wild-type CHX18 or a T-DNA-inserted chx18, respectively. Lane ‘S’ is the 100 bp DNA size standard. Lanes WT or ‘Mut’ contained genomic DNA isolated from the wild type or the triple chx17/18/19 mutant. Plants 1 and 3 are heterozygous for CHX18, and plants 2, 4, 5, 6, and 7 are homozygous wild type for CHX18. Transmission of the chx18-1 allele through pollen of the chx17chx19 double mutant background was 5% instead of the 50% expected (Table 1). Similarly, transmission of the chx19-2 allele through pollen in a chx17chx18 mutant background was only 3% instead of 50% (Table 1). In contrast, the average transmission of the chx18-1 allele through the female was 30% instead of 50% (Table 1). Similarly, transmission of the chx19-2 allele was 25% instead of 50% (Table 1). Together, these results indicate that chx17chx18chx19 gametophytes are significantly less fertile than either chx17CHX18chx19 or chx17chx18CHX19 double mutants. Thus one functional CHX out of three is required for either male or female fertility. We conducted additional reciprocal crosses with the wild type in order to define the impact on male and female reproductive function. When pollen from a double chx17chx19 mutant that is heterozygous for CHX18+/– was transferred to a wild-type pistil, the fertility of the triple chx mutant pollen was severely reduced, as only 10% of the F1 progeny carried a triple heterozygous genotype (Table 2), a significant deviation from the expected 50%. In the reciprocal cross using wild-type pollen, female transmission of the triple chx mutant was 38% instead of 50%, consistent with a role for CHX genes in the female gametophyte (Table 2). These data further support our hypothesis that at least one copy of CHX17, CHX18, or CHX19 is required for complete male and female gametophyte function. Since the defect in male reproductive function was more pronounced, we focused on defining the roles of CHX17, CHX18, and CHX19 in the male gametophyte. Table 2. Reciprocal crosses show that male fertility is compromised in the chx17chx18chx19 mutant Pollen grains from a double mutant chx17chx19 parent heterozygous for CHX18 were used to pollinate a wild-type stigma (a). Wild-type pollen grains were used to pollinate a double mutant pistil heterozygous for CHX18+/– (b). F1 seeds from these crosses were planted and leaves were collected for DNA extraction and genotyped. The observed (Obs) segregation of F1 progeny is shown below as number of individuals recovered or percentage (%). Exp. refers to the expected ratio of F1 progeny in %. . Female: egg or central cell . Male: pollen or sperm . F 1 progeny . . . . CHX CHX CHX Exp Obs χ2 17 18 19 17 18 19 17 18 19 (%) (%) a + + + – + – +/– +/+ +/– (50) 101 (90) 72.3 + + + – – – +/– +/– +/– (50) 11 (10) b – + – + + + +/– +/+ +/– (50) 72 (62) 7.3 – – – + + + +/– +/– +/– (50) 43 (38) . Female: egg or central cell . Male: pollen or sperm . F 1 progeny . . . . CHX CHX CHX Exp Obs χ2 17 18 19 17 18 19 17 18 19 (%) (%) a + + + – + – +/– +/+ +/– (50) 101 (90) 72.3 + + + – – – +/– +/– +/– (50) 11 (10) b – + – + + + +/– +/+ +/– (50) 72 (62) 7.3 – – – + + + +/– +/– +/– (50) 43 (38) The number of F1 plants tested was 112–115. χ2 >6.64 (P value of 0.01) indicates that the observed (Obs) results are significantly different from the expected (Exp) Mendelian ratio. Open in new tab Table 2. Reciprocal crosses show that male fertility is compromised in the chx17chx18chx19 mutant Pollen grains from a double mutant chx17chx19 parent heterozygous for CHX18 were used to pollinate a wild-type stigma (a). Wild-type pollen grains were used to pollinate a double mutant pistil heterozygous for CHX18+/– (b). F1 seeds from these crosses were planted and leaves were collected for DNA extraction and genotyped. The observed (Obs) segregation of F1 progeny is shown below as number of individuals recovered or percentage (%). Exp. refers to the expected ratio of F1 progeny in %. . Female: egg or central cell . Male: pollen or sperm . F 1 progeny . . . . CHX CHX CHX Exp Obs χ2 17 18 19 17 18 19 17 18 19 (%) (%) a + + + – + – +/– +/+ +/– (50) 101 (90) 72.3 + + + – – – +/– +/– +/– (50) 11 (10) b – + – + + + +/– +/+ +/– (50) 72 (62) 7.3 – – – + + + +/– +/– +/– (50) 43 (38) . Female: egg or central cell . Male: pollen or sperm . F 1 progeny . . . . CHX CHX CHX Exp Obs χ2 17 18 19 17 18 19 17 18 19 (%) (%) a + + + – + – +/– +/+ +/– (50) 101 (90) 72.3 + + + – – – +/– +/– +/– (50) 11 (10) b – + – + + + +/– +/+ +/– (50) 72 (62) 7.3 – – – + + + +/– +/– +/– (50) 43 (38) The number of F1 plants tested was 112–115. χ2 >6.64 (P value of 0.01) indicates that the observed (Obs) results are significantly different from the expected (Exp) Mendelian ratio. Open in new tab chx17chx18chx19 mutant pollen contains three nuclei though the wall pattern is disorganized The size and morphology of mutant and wild-type pollen grains from stage 13–14 flowers appeared similar under light microscopy. Pollen grains were stained with DAPI which binds dsDNA. Fluorescence microscopy showed that 95% of triple mutant grains contained three DAPI-stained nuclei like the wild type (Fig. 3A). About 5% of mutant grains showed two DAPI-stained nuclei (Fig. 3B). Thus disruption of CHX17, CHX18, and CHX19 genes did not perturb microspore development significantly to produce tricellular pollen. Fig. 3. Open in new tabDownload slide Mutant chx17chx18chx19 pollen grain contains three nuclei and a disorganized wall pattern. (A) Representative images of DAPI-stained pollen grains observed by UV fluorescence (top) or bright-field (bottom) microscopy. Scale bar=10 µm. (B) Nearly all wild-type and chx17/18/19 pollen grains contained three nuclei as visualized by DAPI staining. Total grains scored were 108 and 200 from the wild type and chx17/18/19, respectively. (C) Loss of reticulate wall pattern. Scanning electron micrographs of the wild type (Wt; i, ii) and chx17/18/19 mutant pollen grains (iii, iv). Scale bar=10 (i, iii) or 5 μm (ii, iv). Intriguingly, SEM showed that the outer wall or exine of all chx17chx18chx19 mutant pollen grains was disorganized, in contrast to the highly reticulate pattern of wild-type grains (Fig. 3C). The single chx17 mutant pollen displayed wall sculpture resembling that of wild-type grains (Supplementary Fig. S3). However, the double chx17chx18 mutant grains gave a mixture of wall patterns where ~52% were fully disorganized, 16% were intermediate, and 31% looked normal (Supplementary Fig. S3). Thus all three CHX genes have overlapping roles in pollen wall formation and patterning. chx17chx18chx19 pollen tubes arrive at ovules, but seed development fails Pollen tube growth in the pistil was examined using aniline blue. Callose from vascular tissue of the funiculus is distinct as seen in unpollinated pistils. At 1 DAP, wild-type pollen tubes are visible at the base of the pistil (Supplementary Fig. S4). Furthermore, most ovules (74–82%) had received a pollen tube, observed at higher magnification (not shown). The estimate is probably low as many tubes were not visible due to ovule crowding. Two days after wild-type pollination, 93% of seeds (568 out of 611) had enlarged, though pollen tubes were mostly undetectable, suggesting that successful fertilization caused degeneration of pollen tubes. In wild-type pistils pollinated with the triple mutant, many tubes were visible along the length of the pistil at 1 DAP, though fewer tubes were visible at the base, suggesting that pollen germination, tube growth, or both are suppressed. By 2 DAP, some ovules have increased in size, consistent with early seed development after successful fertilization (Supplementary Fig. S2). However a significant number of ovules (≥50%) remained small and are referred to as ‘undeveloped ovules’. At 3 DAP, developing ovules/seed increased in size further; however, the ‘undeveloped ovules’ from triple mutant pollen had begun to shrivel (Supplementary Fig. S4). At 2–3 DAP, we noticed that aniline blue-stained pollen tubes were less or not visible in developing ovules, confirming that the pollen tube degenerated in cases when fertilization is successful. To distinguish between size differences of undeveloped and developed ovules in pistils pollinated with triple mutant pollen, we analyzed mutants at 2 DAP. We could not see all undeveloped ovules as some are obstructed by the transmitting tract, so we only scored pollen tubes in any undeveloped ovules in full view. For example, we counted 50 pollen tubes inside 132 undeveloped ovules, which included 28–57% pollen reception by undeveloped ovules per pistil. Thus, on average, at least 39% of the undeveloped ovules had received a tube, while a fraction of undeveloped ovules did not receive a tube (Supplementary Fig. S5). In contrast, pistils pollinated with wild-type pollen show that most ovules (86–97%) have increased in size by 2 DAP (Supplementary Fig. S4). Thus mutant pollen grains germinate, extend tubes, and are able to target ovules; however, tube number and tube lengths are reduced relative to the wild type. Even though mutant pollen tube lengths grown in vitro for 6 h were 10–20% shorter and the percentage germination was less (14–32%) than in the wild type (58%) (data not shown), on average most (65–80%) ovules received a tube when pollen grains germinated in vivo on the stigma. We found that growth of mutant pollen tubes in self-pollinated triple mutants is slightly more robust than that manually transferred to wild-type pistils. In pods of self-pollinated wild-type plants, fluorescent-labeled pollen tubes were visible outside each developing ovule/seed (not shown). Mutant chx17chx18chx19 pollen tubes also grew through the female transmitting tissues to the distal end of the pod, as evident from developing seeds (Fig. 4A). Though triple mutant tubes did reach the distal end, many small undeveloped ovules were seen along the entire length of the pod. Upon closer examination, a pod at the right developmental stage (2–3 DAP) showed that 15 out of 21 undeveloped ovules had received a tube (Fig. 4A). Some undeveloped ovules at the distal end of the pod were homogeneously gray and did not receive a tube (Fig. 4A, far right) thus they were unfertilized. Other undeveloped ovules with a visible pollen tube showed bright grainy content (Fig. 4C, right), suggesting abnormal development after tube reception. Out of three pods at a similar developmental stage, we counted 41 pollen tubes entering 65 undeveloped ovules, indicating that at least 63% of undeveloped ovules have received a tube. In three separate experiments using 10 pods each, we detected that pollen tubes had entered 30–63% of the undeveloped ovules in the pistils from the triple mutant (Fig. 4B, C). These results indicated that the pollen–pistil interactions regulating in vivo pollen tube growth and tube guidance of the triple mutant were not perturbed. Mutant tubes are able to sense chemical cues in the pistil and those from the embryo sac and respond by targeting the micropyle. The results indicate that even though pollen tubes reach ovules, there is a failure to complete fertilization. Fig. 4. Open in new tabDownload slide Mutant pollen tubes reached many ovules though half remain undeveloped. Ovules were examined after self-pollination of the triple chx17chx18chx19 mutant. (A) Some ovules are small and do not develop by 3–4 DAP. Arrows point to visible pollen tubes inside undeveloped ovules. (B, C) Magnified images of the same pod show that aniline blue-stained tubes have entered ovules that remain undeveloped. The vascular strand is visible at the chalazal end of each ovule. Images are representative of 6–10 pods. Scale bar=100 µm. (A) is a composite of frames imaged using a ×10 objective lens and (B) and (C) are imaged with a ×20 objective. chx17chx18chx19 pods contain unfertilized ovules and single fertilization events Pods of manually or self-pollinated flowers were treated with chloral hydrate, and every ovule per pod was examined by DIC microscopy 3–4 DAP. Wild-type pistil pollinated with triple mutant chx17chx18chx19 pollen yielded three types of ovules: (i) unfertilized ovules (63%); (ii) single-fertilized ovules containing either an embryo only or a small endosperm only (7%); and (iii) normal seeds that contain both developing embryo and endosperm (30%) (Fig. 5F). Pods from the control wild type yielded nearly all normal seeds (98%), similar to pods obtained from triple mutant pistils pollinated by wild-type pollen (98%) (Fig. 5F). These results are consistent with the reciprocal cross study where male fertility is severely compromised by loss of function of three CHX genes in the male gametophyte. Fig. 5. Open in new tabDownload slide Pistils receiving chx17chx18chx19 mutant pollen produced unfertilized ovules, embryo only, and developing seeds. (A) Wild type (WT) control of developing seed at ~3 DAP. Pods were cleared with chloral hydrate and examined by microscopy. (B–E) Triple mutant ovules ~3 d after self-pollination. (B) Unfertilized ovule shows an egg (arrowhead) and a central cell (red arrowhead). (C) Globular embryo only. (D) Endosperm only. (E) Heart stage embryo. Scale bar=50 µm. (F) Distribution of ovule development per pod after reciprocal or self-pollination. Manual pollinated pistils: WT pistil receiving WT pollen (WT×WT); chx17/18/19 mutant (3chx) pistil receiving WT pollen (WT×3chx), and WT pistil receiving chx17/18/19 mutant pollen (3chx×WT). ‘WT’ self or ‘3chx self’ refers to pods of the self-pollinated WT or chx17/18/19 mutant. The mumber of ovules per pod was based on 10–14 pods (manual crosses) or 3–6 pods (self). ‘Unfert.’ indicates unfertilized ovule (as in B); ‘Emb’ or ‘Endo’ refers to embryo (C) and endosperm (D) only, and ‘Dev’ refers to developing embryo (as in E). (G) Expression of Arabidopsis CHX18 and CHX19 in sperm and pollen based on the ATH1 transcriptome. Heat map showing normalized transcriptome results (http://arabidopsis-heat-tree.org/) of CHX and other characterized genes in microspore (MS), bicellular pollen (BP), tricellular pollen (TP), and mature pollen (MP) (Honys and Twell, 2004), dry pollen grains (Dry), pollen tubes germinated for 0.5 h (0.5 PT), 4 h (4h PT), or in tubes emerging from a cut style after grain germination on a stigma (SIV, semi-in vivo) (Qin et al., 2009), and sperm cells (Borges et al., 2008). Gene expression in female tissues, include stigma and ovary (Schmid et al., 2005), ovule and unpollinated pistil (pistil) (Boavida et al., 2011). Normalized values of expression in sperm are shown on the far right. Feronia (At3g51550) is expressed in synergid cells. HAP2/GCS1 (At4g11720) and GEX2 (At5g49150) are controls for sperm-expressed genes. Pods of self-pollinated triple chx17chx18chx19 mutants also contained three classes of ovules: (i) 33% unfertilized ovules; (ii) 11% single-fertilized ovules; and (iii) 55% normal developing seeds (Fig. 5A–E). The percentage of undeveloped ovules that includes unfertilized and single fertilized ovules ranged from 45% to 70% based on analysis of six pods. Unfertilized ovules were verified by the presence of a central nucleus or egg cell, as shown in Fig. 5B. Products of single fertilization, though less frequent, were consistently observed mostly as a globular embryo only (Fig. 5C). We also noticed developing endosperm without any visible embryo (Fig 5D). These results indicate that only one fertilization event was successful. Ovules with endosperm alone varied in size, and can increase to twice the length of the unfertilized ovule (Fig. 5D). About one-third to half of the ovules per pod developed normally, judging by the development of both an embryo and an endosperm, indicating that a subset of triple mutant pollen was competent to complete double fertilization with triple mutant female gametes (Fig. 5E). chx17chx18chx19 pollen tubes arrive at the female gametophyte but show reduced sperm delivery and gamete interaction defects It is clear that ovules are targeted by chx triple mutant pollen tubes (Fig. 4), yet fail to develop into seeds (Fig. 5). However, given the expression pattern (Fig. 5G) of CHX17, CHX18, and CHX19, this could be due to a defect in interactions between mutant pollen tubes and female gametophytes, or to failed interactions between mutant sperm and female gametes. To resolve the spatial and temporal relationship of pollen tube interactions with the embryo sac and the success of sperm fusion with the female gametes, we used an assay that simultaneously monitors synergid degeneration and the status of sperm cells (Leydon et al., 2015). After a pollen tube enters and contacts the ovule, one of two synergid cells degenerates. Plants carrying ACT11:MSI1:GFP express nuclear GFP in all cells, but the synergid cells express it most strongly (Fig. 6B). As the receptive synergid degenerates, the nuclear GFP deteriorates and loss of the nuclear integrity is especially clear when the GFP signal diffuses throughout the synergid cytoplasm (Fig. 6B). After wild-type ACT11:MSI1:GFP pistils were pollinated with wild-type pollen carrying HTR10:HTR10:RFP, which labels sperm nuclei (Ingouff et al., 2007), 81% of ovules showed either one or no intact synergid nuclei, and diffuse HTR10:RFP signal overlapping the egg cell nucleus and central cell nucleus, indicative of successful pollen tube reception and double fertilization (Fig. 6A–C, J–K). The remaining 19% of ovules had both synergid nuclei intact, and were probably untargeted by pollen tubes (Fig. 6J, K). In contrast, when wild-type ACT11:MSI1:GFP pistils were pollinated by the chx triple mutant, 95.3% of ovules had no visible sperm cells, though 334 out of 443 ovules showed that at least one synergid had undergone degeneration (Fig. 6D–F, J, K). The degeneration of the synergid cell nuclei indicates that mutant pollen tube had reached the synergid cell, consistent with the observation of sperm cells localized to the micropylar region of the ovule (Fig. 6G). Unfused sperm cells were observed in 3.2% (or 14) of the ovules (Fig. 6H, I); these ovules showed synergid degeneration with no sign of plasmogamy; in many cases, the sperm appeared to be deposited incorrectly at the micropylar end of the synergid, rather than at the interface between the egg and the central cell where gamete fusion takes place (Hamamura et al., 2011). Only 1.6% (seven) ovules showed successful double fertilization and had either one or no intact synergids (Fig. 6J, K). Thus the frequency of double fertilization was extremely low in chx triple mutant pollen, though the frequency of synergid degeneration is high. Thus chx triple mutant pollen tubes can target ovules and undergo normal pollen tube reception as evidenced by synergid degeneration. Therefore, reduced male fertility of chx17chx18chx19 triple mutants is due to partial failure in pollen discharge, sperm positioning, impaired sperm–egg interactions, sperm degeneration during pollen tube growth, or a combination of these defects. Fig. 6. Open in new tabDownload slide Live imaging demonstrates that chx17chx18chx19 pollen tubes trigger synergid degeneration, yet sperm frequently fail to undergo fertilization. (A–I) Confocal micrographs of ACT11:MSI1:GFP ovules receiving pollen tubes expressing a sperm-specific HTR10:HTR10:RFP, 16 h after pollination. Scale bars=20 μm. (A–C) Wild-type HTR10:HTR10:RFP sperm cells are visible in ovules with a diffuse signal (A) (arrows) that overlaps with the nuclear GFP signal (B) from the egg and central cell indicating the successful fusion of gametes and formation of the zygote nucleus (zn) and endosperm nucleus (en). The synergid nucleus of the receptive synergid has degenerated (dsn) and the non-receptive synergid nucleus is still intact (sn), indicating that pollen tube reception has taken place normally. (C) Overlapping RFP and GFP signal. (D–I) Ovules receiving chx17chx18chx19, HTR10:HTR10:RFP sperm cells frequently lack positive signal from the HTR10:HTR10:RFP reporter. (D) Sperm cell signal of HTR10:HTR10:RFP is visible in pollen grains that have detached from the stigma into the media (pollen grains, top left), yet not inside ovules. (E) Ovules show evidence of pollen tube reception and synergid degeneration by the degeneration of one or both synergid nuclei (dsn). (F) Overlapping GFP and RFP signals indicate a loss of visible RFP signal, yet the egg (ecn) and central cell nuclear (ccn) GFP signal remains strong. (G) Overlapping RFP and GFP signal from an ovule with two sperm cells (sc) approaching the micropyle, indicating normal pollen tube arrival at the ovule prior to reception. Ovules with one (H) or two (I) degenerated synergid nuclear GFP signals (dsn) and one visible HTR10:HTR10:RFP-positive sperm cell that has failed to undergo fusion with either the egg or the central cell. (J) Quantification of the HTR10:HTR10:RFP sperm signal in all ovules receiving wild-type or chx17,18,19 pollen tubes. (K) Quantification of nuclear ACT11:MSI1:GFP signal in synergids in ovules from (J), 0 SN, zero intact synergid nuclei; 1 SN, one intact synergid nucleus; 2 SN, two intact synergid nuclei. The total number of ovules observed for each category is written above each stacked column, and the number of ovules in each synergid status category is written on each stacked column unit. Delayed embryo development of self-pollinated triple mutants Most embryos within a wild-type silique develop at similar rates, so they reach the late heart stage by 5 DAP (Fig. 7A). Similarly, most of the embryos (80–90%) produced from a wild-type pistil pollinated with triple mutant pollen developed to the late heart stage by 4–5 DAP (Fig. 7G). However, when fertilization was successful in self-pollinated triple mutants, embryos within a silique showed a range of developmental stages (Fig. 7B–E). When some embryos had reached the late torpedo stage, others were at the globular or early to late heart stages (Fig. 7H). Thus, homozygous triple mutant embryos showed delayed development. From six pods, we estimated that ~55% of the embryos had not reached the heart stage, when 30% were at the early to torpedo stage. The fate of embryos with severely delayed development was not followed so it is unclear whether development would abort prematurely, or progress to full maturity. These results suggest that CHX functions are also critical for post-fertilization development of the young sporophyte. As a wild-type pistil pollinated with triple mutant pollen produced embryos that developed similarly to the wild type, our results suggest that at least one copy of CHX17, CHX18, or CHX19 is required for synchronous development of mutant embryos. Fig. 7. Open in new tabDownload slide The self-pollinated chx17chx18chx19 mutant also show delayed embryo development when fertilization is successful. Six days after pollination, pods were fixed and cleared with chloral hydrate. (A) Wild-type (WT) embryos are mostly at the torpedo stage. (B–F) Mutant ovules at various developmental stages from a single pod: (B) globular, (C) early heart, (D) late heart, (E) early torpedo, and (F) torpedo. Scale bars=100 μm. (G) The embryo developed to a similar stage when a wild-type pistil received chx17/18/19 mutant pollen or vice versa. The percentage of embryo stages per pod from (i) WT×WT cross; (ii)WT pistil×mutant pollen; or (iii) mutant pistil×WT pollen. (H) Self-pollinated triple mutants showed delayed embryo development. WT or triple chx mutants were self-pollinated and pods were examined by DIC microscopy. Results are the average of 10 pods per treatment. Discussion We take advantage of a chx17chx18chx19 mutant to dissect the basis of reduced seed set and shed light on the specific reproductive phase(s) affecting male fertility. Reduced male fertility is attributed to multiple steps, including germination, tube growth, pollen discharge, and fertilization defect in sperm. Our studies highlight the importance of endomembrane transporters involved in pH and cation homeostasis. Combined with previous functional studies, these results suggest that CHX transporters influence reproductive development through membrane trafficking by possibly remodeling cell walls and the PM. Male transmission defect and failed fertilization Based on reciprocal crosses and the low transmission of the mutant chx18 or chx19 allele (3–5% instead of 50%) by the male gametophyte to the F1 progeny, we concluded that a male gametophyte with loss of function in CHX17, CHX18, and CHX19 genes was severely impaired. In contrast, transmission of the mutant chx18 or chx19 allele by the female gametophyte was 25–30% instead of 50%, thus a triple mutant female gametophyte was less severely compromised. The basis of the male gametophyte defect was apparently not due to development of the microspore into the mature pollen based on light microscopy. The male gametophyte begins after meiosis of a pollen mother cell in the anther, forming four haploid microspores. Each microspore divides to produce a bicellular pollen containing a vegetative cell and a generative cell. The generative cell then undergoes mitosis to produce two sperms cells, thus giving rise to a tricellular mature pollen. DAPI staining showed that 95% of triple mutant grains contained three nuclei, similar to those of wild-type grains, indicating that male gametogenesis was largely unaffected by loss of CHX17, CHX18, and CHX19 activity. Ultrastructural studies confirmed that the mutant developed like the wild type (not shown); however, SEM revealed a disorganized wall in mutant pollen. Mutant pollen had a ‘spongy’ wall instead of a reticulate pattern, though that defect alone is unlikely to reduce male fertility (Dobritsa et al., 2011). The outer wall of pollen grains protects against dessication, and enhances adhesion and hydration on the stigma before pollen germination. Germination of chx17chx18chx19 pollen was reduced and tube lengths were shorter in vitro; yet the quantity of pollen in vivo ensured that sufficient grains attached to the stigma and many tubes grew into the ovary. Thus pollen functions in the early phases of reproduction were not severely hampered. Aniline blue staining demonstrated that mutant pollen tubes had targeted and entered many ovules, indicating that loss of CHX17, CHX18, and CHX19 function did not affect the ability of tubes to sense cues secreted by the female gametophyte and to target the embryo sac (Higashiyama and Takeuchi, 2015). Pollen tubes were observed in both developing and undeveloped seeds, suggesting that the pollen tube had entered the micropyle in both cases. About 39–65% of undeveloped ovules in each pod had received a pollen tube, indicating that fertilization was unsuccessful. Moreover, mutant pollen tube reception was verified by the degeneration of one or both synergid cells seen in 80% ovules of a wild-type pistil (Fig. 6K). In wild-type pollen, both sperm cells are released simultaneously at tube rupture after synergid cell reception (Hamamura et al., 2011). One sperm fuses with the egg cell to form a zygote that develops into the embryo. The second sperm nucleus unites with the central cell nucleus and undergoes subsequent nuclear division and cellularization to produce the endosperm. Direct analysis of chx triple mutant sperm cells in female gametophytes suggests that many pollen tubes may fail to deliver sperm. The vast majority of ovules pollinated with triple mutant pollen had no sperm cells visible based on RFP fluorescence (Fig. 6K). Failure to express markers of sperm identity such as HTR10:HTR10:RFP (Fig. 6) may indicate that mutant sperm cells are compromised. The lack of visible sperm signal could account for the apparent low frequency of successful fertilization estimated from the imaging results (Fig. 6) compared with 27–45% of developing seeds per pod in pistils receiving chx mutant pollen in vivo (Fig. 4). Undeveloped ovules that received a pollen tube suggest that either tube rupture was defective, or released sperm cells fail to complete fertilization with the female gametes, or both. Single fertilization events were observed, indicating that defective tube rupture is not the major cause. Some ovules develop only a globular embryo with no detectable endosperm tissue, and other ovules enlarge in size due to proliferation of the endosperm nuclei, with no visible embryo. Either of these examples will lead to abortion and failed seed development, as embryogenesis and endosperm development are co-ordinated to produce successful seed development (Berger, 2003). This idea is verified by the presence of unfused sperm cell in ovules pollinated by chx17chx18chx19 mutant pollen (Fig. 6H, I, K). Single fertilization events suggest that one of the two sperm failed to complete fertilization. Thus loss of CHX17,CHX18, andCHX19 genes compromised sperm function in half to two-thirds of the male gametophytes, leading to failed double fertilization. The ability of any one of three CHX genes to rescue male fertility is unexpected as they are differentially expressed in space and time. CHX17 transcript detected in microspore and bicellular pollen (Honys and Twell, 2004) was confirmed by CHX17p::GUS (β-glucuronidase) staining of anthers in stage 9–10 flowers (Bock et al., 2006). CHX19 transcript is detected in developing pollen grains and tubes (Fig. 5G), where the protein was localized to the PM (Supplementary Fig. S6). CHX18 transcript was detected in sperm cells (Fig. 5G) isolated by fluorescence-activated cell sorting (Borges et al., 2008), whereas CHX19 expression is low or ‘absent’. In contrast, CHX18 transcript was low or not detectable in mature pollen (Honys and Twell, 2004) or in the tube (Qin et al., 2009), suggesting that CHX18 transcript in sperm cells is not contaminated by mRNA from the vegetative cell. HAP2/GCS1 or GEX2 are experimentally demonstrated to be sperm-expressed genes (Mori et al., 2006, 2014; von Besser et al., 2006), whereas Feronia is not expressed in pollen and is expressed in synergid cells of the female gametophyte (Huck et al., 2003; Rotman et al., 2003; Escobar-Restrepo et al., 2007). The transcriptome results of reproductive tissues have been verified by these studies (Fig. 5G), indicating that AtCHX18 and AtCHX17 are expressed in sperm cells. At least one of three CHXs is expressed at every stage of male gametophyte development. Apparently one copy of CHX19 is sufficient to restore male fertility of a chx17chx18 double mutant. Thus plants have mechanisms to compensate when one or more homologous genes are rendered non-functional. Yet some chx17chx18chx19 mutants completed double fertilization successfully. This partial success in fertilization by triple mutant sperm could be due to compensation by other sperm-expressed cation/H+ exchangers, such as AtCHX20 (see Fig. 5G) or AtNHX2. AtCHX20 also mediates cation and pH homeostasis in yeast (Padmanaban et al., 2007), though it differs from CHX17, in endomembrane localization and inability to sort cargo, such as aminoglycoside HygB, in yeast (Chanroj et al., 2011). The female gametophyte and post-fertilization events contribute to seed development The basis for transmission of the CHX18+ allele via the female gametophyte (Table 2) awaits further investigation. Both CHX17 and CHX18 transcripts were detected in the central cell (Schmid et al., 2012) (Supplemenatry Fig. S7), and possibly in the synergid and egg cell (Wuest et al., 2010; S. Sprunck, personal communication). We showed that one functional CHX17, CHX18, or CHX19 is also critical for post-fertilization development, as early embryo development contained in one pod was synchronous in a heterozygous chx17+/–chx18+/–chx19+/– sporophyte but was delayed in a homozygous chx17–/–chx18–/–chx19–/– mutant (Fig. 7G, H). We suggest that after fertilization, early embryo development depends on a functional CHX17 or CHX18 in the embryo, endosperm, or both. Several observations support this idea. First, CHX17p::GFP expression is strong at the anterior end of the developing seed (Chanroj et al., 2013) consistent with CHX17 expression at the micropylar endosperm of the early embryo (Belmonte et al., 2013). Secondly, CHX18 is highly expressed in the micropylar and the peripheral endosperm at the pre-globular and globular embryo stage according to the transcriptome of Arabidopsis seed development (Supplementary Fig. S8) (Belmonte et al., 2013). Thirdly, RNA sequencing revealed CHX17 and CHX18 transcripts in a laser-dissected central cell (Supplementary Fig. S7) (Schmid et al., 2012). Union of a sperm cell with the central cell nucleus and subsequent proliferation produces a micropylar endosperm that surrounds the embryo. The role of the micropylar endosperm is unclear. A recent study demonstrated that cysteine-rich peptides, ESF1 (Embryo Surrounding Factor) accumulated in the central cell before fertilization and in the micropylar endosperm after fertilization (Costa et al., 2014). Furthermore, purified ESF has a role in early development of the embryo, possibly through suspensor elongation and auxin distribution. These results suggest that the endosperm surrounding the embryo plays a critical role in producing peptide cues to facilitate development of the suspensor and pre-globular embryo. One model is that CHX17 and CHX18 influence peptide sorting and secretion from the micropylar endosperm to the target cells. Model: CHX17, CHX18, and CHX19 activities affect PM and wall remodeling How would loss of CHX17, CHX18, or CHX19 produce the various defects we observed? The loss in exine patterning of chx17chx18chx19 mutant grains may be indirectly related to decreased male fertility, yet it may provide a clue to compromised wall formation. There are two schools of thought regarding outer wall formation: (i) primexine (precursor of exine) and exine seen on microspores depend on materials synthesized from tapetum (or sporophyte); and (ii) primexine and exine are formed by the combined activities of the microspore and tapetum (Dobritsa et al., 2011). Ultrastructural studies had suggested that primexine originated from the PM of the microspore. For instance, the dex1 mutant is male-sterile, and a tetrad of dex1 microspores surrounded by callose wall shows defects in early primexine formation and later collapsed pollen grains (Paxson-Sowders et al., 2001). DEX1 is expressed in microspores and encodes a predicted integral protein. Curiously, the sterility defect followed Mendelian inheritance, suggesting control by the maternal sporophyte (Paxson-Sowders et al., 2001). Our genetic studies show that loss of male fertility in chx17chx18chx19 mutants is largely a male gametophyte defect and is accompanied by exine deformation. While we cannot eliminate CHX function in tapetal or pollen mother cells, our SEM results (not shown) support the hypothesis that primexine synthesis is initiated from microspores. It is likely that microspores within a tetrad produce a scaffold or template upon which exine materials attach and assemble (Ariizumi and Toriyama, 2011). Thus, any defects in the scaffold could directly affect subsequent exine patterning. Alternatively, a defect in intine (or inner wall) that is formed later by bicellular pollen between the PM and the outer wall might perturb exine patterning. These ideas warrant reconsideration. How could three CHX transporters with nearly identical activities perturb male fertility? To date, our results are consistent with the hypothesis that chx mutants have perturbed membrane trafficking that remodels the PM and the cell wall (Kim and Brandizzi, 2016), and those changes lead to impaired fertility. First, CHX17 as well as CHX18 or CHX19 confer tolerance to alkaline pH in a yeast mutant sensitive to growth at pH 7.5, suggesting a role in pH homeostasis. Secondly, CHX17 supports growth of K+ uptake mutants in yeast and E. coli, suggesting a role in K+ transport and K+ homeostasis. Thirdly, CHX17 could reduce the secretion of vacuolar carboxypeptidase Y in yeast, indicating that it has a role in proper sorting of cargo (Chanroj et al., 2011). As CHX17 has been localized to the PVC and the PM in plant cells (Chanroj et al., 2013), the results infer that proper regulation of the endomembrane pH and cation level plays a role in protein and cargo sorting (Chanroj et al., 2011). This idea is demonstrated convincingly in a det3 H+-pumping V-ATPase mutant that showed increased pH in the trans-Golgi network (TGN), and altered secretion and recycling of cellulose synthase to the PM (Luo et al., 2015). Our results show that chx17chx18chx19 mutant pollen is compromised at multiple phases in reproduction. As many ovules received a pollen tube, we suggest that failed fertilization is mainly due to reduced tube rupture or impaired sperm activity, or both. Loss of function of PM-localized CHX19 could affect tube growth and rupture perhaps through ANX1/ANX2 receptor-like kinases (Boisson-Dernier et al., 2009). Loss of CHX18 and CHX17 function in sperm cells could impair male gamete activity. CHX17 was shown to alter sorting of protein cargo in yeast (Chanroj et al., 2011), thus plasma membrane proteins (e.g. receptors) or secreted factors of the pollen tube or sperm might be misguided in the triple mutant. HAP2 is a sperm cell-expressed PM protein that is crucial for fertilization (Mori et al., 2006; von Besser et al., 2006). HAP2 is a transmembrane protein that is sorted to the PM from the ER after activation by EC1, an egg cell-secreted peptide that promotes fertilization (Sprunck et al., 2012). A homolog in Chlamydomonas is needed for cell–cell fusion, suggesting that the plant HAP2/GCS1 might act as a gamete fusogen (Liu et al., 2008). GEX2 encodes a PM-localized protein in sperm cells. Mutant sperm cells failed to attach to the egg membrane, indicating that GEX2 is required for gamete–gamete attachment, a prerequisite for successful fusion (Mori et al., 2014). Perhaps sperm CHX18 influences the sorting and delivery of proteins, such as GEX2 or HAP2 (GCS1), to promote gamete attachment and fusion. Our findings support a model in which three related cation/H+ exchangers affect a subset of endomembranes in developing pollen, pollen tube, sperm, and central cells. Pollen development, tube sensing, or wall properties might be sufficiently compromised to reduce sperm release, sperm function, or both in chx17chx18chx19 mutants. Our study highlights the critical roles of cation/H+ exchangers in membrane trafficking, cargo sorting, and wall remodeling for successful fertilization. Supplementary data Supplementary data are available at JXB online. Table S1. Single T-DNA insertion mutants used in this study. Table S2. Gene-specific primers and T-DNA border primers used for genotyping plants. Fig. S1. Flower development of the chx17/18/19 mutant and the wild type is similar. Fig. S2. Pod and seed development at 0–9 days after pollination. Fig. S3. Pollen wall architecture is partially disorganized in chx17/18 double mutants but not in the single chx17 mutant. Fig. S4. Time-course of ovule development in wild-type (WT) pistils receiving WT or chx17/18/19 mutant pollen. Fig. S5. Pollen tubes of the chx17/18/19 mutant entered ovules that remain undeveloped. Fig. S6. CHX19–RFP was localized to the PM along the flanks of the pollen tube. Fig. S7. RNA sequencing revealed expression of CHX17 and CHX18 in a laser-dissected central cell. Fig. S8. Expression of CHX17 and CHX18 in the micropylar and peripheral endosperm during early seed development. Author contributions HS conceived the original research plans; HS, AYC, MAJ supervised the experiments; SP, DC, and RS performed most experiments, KL, TKM, ARL, and YZ performed other experiments; SC generated the mutants; SP, DC, ARL, and HS designed the experiments and analyzed the data; HS wrote the article with contributions of all the authors; and AYC and MAJ complemented the writing. Acknowledgements We thank John Harada for the seed transcriptome (seedgenenetwork.net/) and Dr Guoying Wang for analysis of transporter expression. Tony Pham (Eleanor Roosevelt High School) tested in vitro pollen tube growth. Supported in part by US Department of Energy, Division of Chemical Sciences, Geosciences and Biosciences, Office of Basic Energy Sciences (BES DEFG-0207ER15883) to HS, National Science Foundation RCN on Integrative Pollen Biology # 0955910, NSF #1147165 to AYC, and National Science Foundation IOS-1353798 to MAJ. References Ariizumi T Toriyama K. 2011 . Genetic regulation of sporopollenin synthesis and pollen exine development . Annual Review of Plant Biology 62 , 437 – 460 . Google Scholar Crossref Search ADS PubMed WorldCat Belmonte MF Kirkbride RC Stone SLet al. . 2013 . Comprehensive developmental profiles of gene activity in regions and subregions of the Arabidopsis seed . Proceedings of the National Academy of Sciences, USA 110 , E435 – E444 . Google Scholar Crossref Search ADS WorldCat Berger F . 2003 . Endosperm: the crossroad of seed development . Current Opinion in Plant Biology 6 , 42 – 50 . Google Scholar Crossref Search ADS PubMed WorldCat Berleth T Jurgens G. 1993 . The role of the monopteros gene in organizing the basal body region of the Arabidopsis embryo . Development 118 , 575 – 587 . Google Scholar OpenURL Placeholder Text WorldCat Boavida LC Borges F Becker JD Feijó JA. 2011 . Whole genome analysis of gene expression reveals coordinated activation of signaling and metabolic pathways during pollen–pistil interactions in Arabidopsis . Plant Physiology 155 , 2066 – 2080 . Google Scholar Crossref Search ADS PubMed WorldCat Bock KW Honys D Ward JM Padmanaban S Nawrocki EP Hirschi KD Twell D Sze H. 2006 . Integrating membrane transport with male gametophyte development and function through transcriptomics . Plant Physiology 140 , 1151 – 1168 . Google Scholar Crossref Search ADS PubMed WorldCat Boisson-Dernier A Roy S Kritsas K Grobei MA Jaciubek M Schroeder JI Grossniklaus U. 2009 . Disruption of the pollen-expressed FERONIA homologs ANXUR1 and ANXUR2 triggers pollen tube discharge . Development 136 , 3279 – 3288 . Google Scholar Crossref Search ADS PubMed WorldCat Borges F Gomes G Gardner R Moreno N McCormick S Feijó JA Becker JD. 2008 . Comparative transcriptomics of Arabidopsis sperm cells . Plant Physiology 148 , 1168 – 1181 . Google Scholar Crossref Search ADS PubMed WorldCat Brett CL Donowitz M Rao R. 2005 . Evolutionary origins of eukaryotic sodium/proton exchangers . American Journal of Physiology. Cell Physiology 288 , C223 – C239 . Google Scholar Crossref Search ADS PubMed WorldCat Chanroj S . 2011 . Plant-specific K+ transporters with distinct properties and their emerging roles in endomembrane trafficking . PhD Dissertation, University of Maryland . Google Scholar Chanroj S Lu Y Padmanaban S Nanatani K Uozumi N Rao R Sze H. 2011 . Plant-specific cation/H+ exchanger 17 and its homologs are endomembrane K+ transporters with roles in protein sorting . Journal of Biological Chemistry 286 , 33931 – 33941 . Google Scholar Crossref Search ADS PubMed WorldCat Chanroj S Padmanaban S Czerny DD Jauh GY Sze H. 2013 . K+ transporter AtCHX17 with its hydrophilic C tail localizes to membranes of the secretory/endocytic system: role in reproduction and seed set . Molecular Plant 6 , 1226 – 1246 . Google Scholar Crossref Search ADS PubMed WorldCat Chanroj S Wang G Venema K Zhang MW Delwiche CF Sze H. 2012 . Conserved and diversified gene families of monovalent cation/h(+) antiporters from algae to flowering plants . Frontiers in Plant Science 3 , 25 . Google Scholar Crossref Search ADS PubMed WorldCat Costa LM Marshall E Tesfaye Met al. . 2014 . Central cell-derived peptides regulate early embryo patterning in flowering plants . Science 344 , 168 – 172 . Google Scholar Crossref Search ADS PubMed WorldCat Czerny DD Padmanaban S Anishkin A Venema K Riaz Z Sze H. 2016 . Protein architecture and core residues in unwound α-helices provide insights to the transport function of plant AtCHX17 . Biochimica et Biophysica Acta 1858 , 1983 – 1998 . Google Scholar Crossref Search ADS PubMed WorldCat Dobritsa AA Geanconteri A Shrestha Jet al. . 2011 . A large-scale genetic screen in Arabidopsis to identify genes involved in pollen exine production . Plant Physiology 157 , 947 – 970 . Google Scholar Crossref Search ADS PubMed WorldCat Dresselhaus T Franklin-Tong N. 2013 . Male–female crosstalk during pollen germination, tube growth and guidance, and double fertilization . Molecular Plant 6 , 1018 – 1036 . Google Scholar Crossref Search ADS PubMed WorldCat Edwards K Johnstone C Thompson C. 1991 . A simple and rapid method for the preparation of plant genomic DNA for PCR analysis . Nucleic Acids Research 19 , 1349 . Google Scholar Crossref Search ADS PubMed WorldCat Escobar-Restrepo JM Huck N Kessler S Gagliardini V Gheyselinck J Yang WC Grossniklaus U. 2007 . The FERONIA receptor-like kinase mediates male–female interactions during pollen tube reception . Science 317 , 656 – 660 . Google Scholar Crossref Search ADS PubMed WorldCat Feijó JA Sainhas J Holdaway-Clarke T Cordeiro MS Kunkel JG Hepler PK. 2001 . Cellular oscillations and the regulation of growth: the pollen tube paradigm . Bioessays 23 , 86 – 94 . Google Scholar Crossref Search ADS PubMed WorldCat Frietsch S Wang YF Sladek C Poulsen LR Romanowsky SM Schroeder JI Harper JF. 2007 . A cyclic nucleotide-gated channel is essential for polarized tip growth of pollen . Proceedings of the National Academy of Sciences, USA 104 , 14531 – 14536 . Google Scholar Crossref Search ADS WorldCat Guan R Qu Y Guo Yet al. . 2014 . Salinity tolerance in soybean is modulated by natural variation in GmSALT3 . The Plant Journal 80 , 937 – 950 . Google Scholar Crossref Search ADS PubMed WorldCat Hamamura Y Saito C Awai Cet al. . 2011 . Live-cell imaging reveals the dynamics of two sperm cells during double fertilization in Arabidopsis thaliana . Current Biology 21 , 497 – 502 . Google Scholar Crossref Search ADS PubMed WorldCat Higashiyama T Takeuchi H. 2015 . The mechanism and key molecules involved in pollen tube guidance . Annual Review of Plant Biology 66 , 393 – 413 . Google Scholar Crossref Search ADS PubMed WorldCat Honys D Twell D. 2004 . Transcriptome analysis of haploid male gametophyte development in Arabidopsis . Genome Biology 5 , R85 . Google Scholar Crossref Search ADS PubMed WorldCat Huck N Moore JM Federer M Grossniklaus U. 2003 . The Arabidopsis mutant feronia disrupts the female gametophytic control of pollen tube reception . Development 130 , 2149 – 2159 . Google Scholar Crossref Search ADS PubMed WorldCat Ingouff M Hamamura Y Gourgues M Higashiyama T Berger F. 2007 . Distinct dynamics of HISTONE3 variants between the two fertilization products in plants . Current Biology 17 , 1032 – 1037 . Google Scholar Crossref Search ADS PubMed WorldCat Johnson MA Kost B. 2010 . Pollen tube development . Methods in Molecular Biology 655 , 155 – 176 . Google Scholar Crossref Search ADS PubMed WorldCat Kim SJ Brandizzi F. 2016 . The plant secretory pathway for the trafficking of cell wall polysaccharides and glycoproteins . Glycobiology 26 , 940 – 949 . Google Scholar Crossref Search ADS PubMed WorldCat Leydon AR Tsukamoto T Dunatunga D Qin Y Johnson MA Palanivelu R. 2015 . Pollen tube discharge completes the process of synergid degeneration that is initiated by pollen tube–synergid interaction in Arabidopsis . Plant Physiology 169 , 485 – 496 . Google Scholar Crossref Search ADS PubMed WorldCat Liu Y Tewari R Ning Jet al. . 2008 . The conserved plant sterility gene HAP2 functions after attachment of fusogenic membranes in Chlamydomonas and Plasmodium gametes . Genes and Development 22 , 1051 – 1068 . Google Scholar Crossref Search ADS PubMed WorldCat Lu Y Chanroj S Zulkifli L Johnson MA Uozumi N Cheung A Sze H. 2011 . Pollen tubes lacking a pair of K+ transporters fail to target ovules in Arabidopsis . The Plant Cell 23 , 81 – 93 . Google Scholar Crossref Search ADS PubMed WorldCat Luo Y Scholl S Doering Aet al. . 2015 . V-ATPase activity in the TGN/EE is required for exocytosis and recycling in Arabidopsis . Nature Plants 1 , 15094 . Google Scholar Crossref Search ADS PubMed WorldCat Márton ML Cordts S Broadhvest J Dresselhaus T. 2005 . Micropylar pollen tube guidance by egg apparatus 1 of maize . Science 307 , 573 – 576 . Google Scholar Crossref Search ADS PubMed WorldCat Mori T Kuroiwa H Higashiyama T Kuroiwa T. 2006 . GENERATIVE CELL SPECIFIC 1 is essential for angiosperm fertilization . Nature Cell Biology 8 , 64 – 71 . Google Scholar Crossref Search ADS PubMed WorldCat Mori T Igawa T Tamiya G Miyagishima SY Berger F. 2014 . Gamete attachment requires GEX2 for successful fertilization in Arabidopsis . Current Biology 24 , 170 – 175 . Google Scholar Crossref Search ADS PubMed WorldCat Mottaleb SA Rodríguez-Navarro A Haro R. 2013 . Knockouts of Physcomitrella patens CHX1 and CHX2 transporters reveal high complexity of potassium homeostasis . Plant and Cell Physiology 54 , 1455 – 1468 . Google Scholar Crossref Search ADS PubMed WorldCat Okuda S Tsutsui H Shiina Ket al. . 2009 . Defensin-like polypeptide LUREs are pollen tube attractants secreted from synergid cells . Nature 458 , 357 – 361 . Google Scholar Crossref Search ADS PubMed WorldCat Padmanaban S Chanroj S Kwak JM Li X Ward JM Sze H. 2007 . Participation of endomembrane cation/H+ exchanger AtCHX20 in osmoregulation of guard cells . Plant Physiology 144 , 82 – 93 . Google Scholar Crossref Search ADS PubMed WorldCat Paxson-Sowders DM Dodrill CH Owen HA Makaroff CA. 2001 . DEX1, a novel plant protein, is required for exine pattern formation during pollen development in Arabidopsis . Plant Physiology 127 , 1739 – 1749 . Google Scholar Crossref Search ADS PubMed WorldCat Qi X Li MW Xie Met al. . 2014 . Identification of a novel salt tolerance gene in wild soybean by whole-genome sequencing . Nature Communications 5 , 4340 . Google Scholar Crossref Search ADS PubMed WorldCat Qin Y Leydon AR Manziello A Pandey R Mount D Denic S Vasic B Johnson MA Palanivelu R. 2009 . Penetration of the stigma and style elicits a novel transcriptome in pollen tubes, pointing to genes critical for growth in a pistil . PLoS Genetics 5 , e1000621 . Google Scholar Crossref Search ADS PubMed WorldCat Rotman N Rozier F Boavida L Dumas C Berger F Faure JE. 2003 . Female control of male gamete delivery during fertilization in Arabidopsis thaliana . Current Biology 13 , 432 – 436 . Google Scholar Crossref Search ADS PubMed WorldCat Schiott M Romanowsky SM Baekgaard L Jakobsen MK Palmgren MG Harper JF. 2004 . A plant plasma membrane Ca2+ pump is required for normal pollen tube growth and fertilization . Proceedings of the National Academy of Sciences, USA 101 , 9502 – 9507 . Google Scholar Crossref Search ADS WorldCat Schmid M Davison TS Henz SR Pape UJ Demar M Vingron M Schölkopf B Weigel D Lohmann JU. 2005 . A gene expression map of Arabidopsis thaliana development . Nature Genetics 37 , 501 – 506 . Google Scholar Crossref Search ADS PubMed WorldCat Schmid MW Schmidt A Klostermeier UC Barann M Rosenstiel P Grossniklaus U. 2012 . A powerful method for transcriptional profiling of specific cell types in eukaryotes: laser-assisted microdissection and RNA sequencing . PLoS One 7 , e29685 . Google Scholar Crossref Search ADS PubMed WorldCat Smyth DR Bowman JL Meyerowitz EM. 1990 . Early flower development in Arabidopsis . The Plant Cell 2 , 755 – 767 . Google Scholar Crossref Search ADS PubMed WorldCat Sprunck S Rademacher S Vogler F Gheyselinck J Grossniklaus U Dresselhaus T. 2012 . Egg cell-secreted EC1 triggers sperm cell activation during double fertilization . Science 338 , 1093 – 1097 . Google Scholar Crossref Search ADS PubMed WorldCat Sze H Padmanaban S Cellier Fet al. . 2004 . Expression patterns of a novel AtCHX gene family highlight potential roles in osmotic adjustment and K+ homeostasis in pollen development . Plant Physiology 136 , 2532 – 2547 . Google Scholar Crossref Search ADS PubMed WorldCat Tunc-Ozdemir M Rato C Brown Eet al. . 2013 . Cyclic nucleotide gated channels 7 and 8 are essential for male reproductive fertility . PLoS One 8 , e55277 . Google Scholar Crossref Search ADS PubMed WorldCat von Besser K Frank AC Johnson MA Preuss D. 2006 . Arabidopsis HAP2 (GCS1) is a sperm-specific gene required for pollen tube guidance and fertilization . Development 133 , 4761 – 4769 . Google Scholar Crossref Search ADS PubMed WorldCat Wuest SE Vijverberg K Schmidt Aet al. . 2010 . Arabidopsis female gametophyte gene expression map reveals similarities between plant and animal gametes . Current Biology 20 , 506 – 512 . Google Scholar Crossref Search ADS PubMed WorldCat Author notes These authors contributed equally to this work. © The Author 2017. Published by Oxford University Press on behalf of the Society for Experimental Biology. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. © The Author 2017. Published by Oxford University Press on behalf of the Society for Experimental Biology.
TI - Transporters involved in pH and K+ homeostasis affect pollen wall formation, male fertility, and embryo development
JF - Journal of Experimental Botany
DO - 10.1093/jxb/erw483
DA - 2017-06-01
UR - https://www.deepdyve.com/lp/oxford-university-press/transporters-involved-in-ph-and-k-homeostasis-affect-pollen-wall-0vblfmL74R
SP - 3165
EP - 3178
VL - 68
IS - 12
DP - DeepDyve
ER -