Abstract
Arabidopsis thaliana BOR1 was the first boron (B) transporter identified in living systems. There are four AtBOR1-like genes, OsBOR1, 2, 3 and 4, present in the rice genome. We characterized the activity, expression and physiological function of OsBOR4. OsBOR4 is an active efflux transporter of B. Quantitative PCR analysis and OsBOR4 promoter–green fluorescent protein (GFP) fusion revealed that OsBOR4 was both highly and specifically expressed in pollen. We obtained five Tos17 insertion mutants of osbor4. The pollen grains were viable and development of floral organs was normal in the homozygous osbor4 mutants. We observed that in all Tos17 insertion lines tested, the frequency of osbor4 homozygous plants was lower than expected in the progeny of self-fertilized heterozygous plants. These results establish that OsBOR4 is essential for normal reproductive processes. Pollen from osbor4 homozygous plants elongated fewer tubes on wild-type stigmas, and tube elongation of mutant pollen was less efficient compared with the wild-type pollen, suggesting reduced competence of osbor4 mutant pollen. The reduced competence of mutant pollen was further supported by the crosses of independent Tos17-inserted alleles of OsBOR4. Our results suggest that OsBOR4, a boron efflux transporter, is required for normal pollen germination and/or tube elongation.
Introduction
Boron (B) is an essential micronutrient for vegetative and reproductive development. Rhamnogalacturonan II (RG-II), a component of pectin, provides a binding site for B in the cell walls of higher plants (Kobayashi et al. 1996, Matoh et al. 1996, O’Neill et al. 2004). Cross-linked B–RG-II complexes are required for the structural maintenance of cell walls (O’Neill et al. 2004). In Arabidopsis thaliana, B cross-linked RG-II dimers are involved in the normal expansion of rosette leaves (O’Neill et al. 2001). Moreover, B is also essential for pollen germination and pollen tube growth (Cheng and Rerkasem 1993, Huang et al. 2000, Wang et al. 2003, Iwai et al. 2006). Because B has these central roles in plant development, B deficiency in agricultural fields impairs the yield quantity and quality in many crops, including wheat, barley and rice (Rerkasem and Jamjod 1997, Yu and Bell 2002, Wongmo et al. 2004).
Rice and other monocot cereals require less B for their growth compared with dicots. This lower B requirement correlates with lower amounts of pectic compounds in the cell walls of graminaceous plants (Matoh et al. 1996). In rice, B deficiency affects the reproductive growth more strongly than early vegetative growth (Uraguchi and Fujiwara 2011). Limiting B in the growth medium causes a reduction in both the number of panicles and grain fertility (Uraguchi and Fujiwara 2011). Germination rates of pollen obtained from rice grown under B-limited conditions were lower than those of pollen from rice grown under B-sufficient conditions (Lordkaew et al. 2012). These results suggest that B is essential for grain fertility in rice (and in many other plants) by supporting pollen germination and panicle (floral organ) development.
Despite the importance of B in reproductive growth and fertility, the molecular mechanisms of B transport in floral organs are poorly understood, whereas the molecular mechanisms of B acquisition in the roots and xylem loading have been well characterized (Takano et al. 2002, Takano et al. 2006, Takano et al. 2010). In A. thaliana, AtBOR1, an efflux transporter of B, and AtNIP5;1, a boric acid channel, function together for B acquisition in the roots and loading into the xylem (Takano et al. 2002, Takano et al. 2006, Takano et al. 2010). The rice genome contains four genes similar to AtBOR1; OsBOR1, 2, 3 and 4 (Nakagawa et al. 2007). OsBOR1, the closest homolog of AtBOR1, is the B efflux transporter gene for B acquisition in roots and subsequent translocation to the shoots (Nakagawa et al. 2007). However, the functions of the other OsBOR genes have not been described. Considering the essential role of B in reproductive growth and the potential roles of OsBORs for B transport, members of the OsBOR family may be involved in B homeostasis in the floral organs.
We found that OsBOR4 was expressed in a pollen-specific manner. We demonstrated that Tos17-inserted heterozygous osbor4 mutants showed an abnormal segregation ratio in their progeny, and that homozygous mutants showed defects in pollen tube germination and/or elongation. However, these mutants displayed no visible phenotypes with regard to pollen viability, and vegetative and reproductive growth. We concluded that this B transporter plays a role in fertilization, which is known to require adequate B nutrition.
Results
OsBOR4 transcripts accumulated specifically in the anthers
Four AtBOR1-like genes are present in the rice genome (Nakagawa et al. 2007). According to microarray analysis, OsBOR1, 2, 3 and 4 were expressed in different tissue-specific ways and displayed different developmental expression patterns (Supplementary Fig. S1). Among the OsBOR genes, only OsBOR4 was predominantly expressed in plant anthers. This anther-specific expression pattern indicates the possibility that OsBOR4 functions as a B transporter in the floral organs.
To analyze the expression pattern of OsBOR4 further, we examined the accumulation of OsBOR4 mRNA by quantitative reverse transcription–PCR (RT–PCR) (Fig. 1A, B). Levels of the OsBOR4 transcript were low in the leaf blades, leaf sheaths and roots, irrespective of the developmental stage (Fig. 1A), whereas it was roughly 500-fold higher in the anthers (Fig. 1A). Detailed expression analysis of the floral organs revealed that OsBOR4 transcript accumulation in the anthers was several orders of magnitude higher than that in the other floral organs (Fig. 1B). These results establish that OsBOR4 is specifically expressed in anthers.
Spatial and temporal expression analyses of OsBOR4 in rice. (A) Expression patterns of OsBOR4 in a 3-day-old shoot, 1-month-old leaf blade, 1-month-old leaf sheath, 1-month-old root, leaf blade in the reproductive phase and anther. (B) Expression patterns of OsBOR4 in flag leaf and floral organs (lemma, palea, glume, pedicel, ovary and anther). Each value represents the average (mean) of three independent real-time PCR assays. The y-axis is in log scale. The value for the flag leaf was 3.3E-01. Data in (A) and (B) represent means ± SD (n = 3).
To confirm the anther-specific expression pattern of OsBOR4, we examined green fluorescent protein (GFP) fluorescence in anthers under the control of the OsBOR4 promoter. Fluorescence was observed in immature anthers (Fig. 2A). Detailed observation of the anthers revealed that GFP signals were strong in pollen (Fig. 2B), supporting the idea that OsBOR4 is expressed in pollen. No signals were detected in the control plants without the GFP gene under the same conditions.
GFP image of a pre-anthesis anther in the OsBOR4pro-GFP transgenic rice. (A) GFP signal for the OsBOR4 promoter: GFP in the anther is shown in the upper panel. GFP signal from a negative control transgenic plant carrying the OsBOR4 promoter:GUS is shown in the lower panel. (B) GFP signal for the OsBOR4 promoter: GFP in pollen is shown in the upper panel. The negative control is shown in the lower panel. Bars = 1 mm. Transgenic plants carrying a GUS gene were used as a control.
OsBOR4 has a B efflux activity in yeast cells
We cloned OsBOR4 cDNA from mRNA isolated from anthers to establish the functional analysis of OsBOR4. Nucleotide sequence analysis indicated that OsBOR4 is composed of 13 exons, comprising 2,034 nucleotides, and encoding a 677 amino acid polypeptide (Supplementary Fig. S2A). AtBOR1 and OsBOR1 showed efflux transport activity of B in a yeast expression system (Takano et al. 2002, Nakagawa et al. 2007). To examine the B efflux transport activity of OsBOR4, we expressed OsBOR4 in the scbor1Δ strain of Saccharomyces cerevisiae (Takano et al. 2002, Nozawa et al. 2006). The cells were exposed to media containing 0.1 mM boric acid, and the concentrations of B in the cells were determined. Yeast cells expressing AtBOR1 or OsBOR4 accumulated approximately one-third and two-thirds of the B, respectively, compared with the cells carrying the empty vector pYES2 (Fig. 3), suggesting that OsBOR4 is a functional B efflux transporter like AtBOR1.
Boron transport assay in yeast. B concentration in yeast cells carrying the empty pYES2 vector, AtBOR1 cDNA and OsBOR4 cDNA. Data represent means ± SD (n = 4).
Phenotypes of Tos 17 insertion mutant lines of OsBOR4
To examine the physiological function of OsBOR4 in rice, we obtained the retrotransposon Tos17 insertion lines of the OsBOR4 gene from the Rice Genome Resource Center (Miyao et al. 2003). Their stock includes 185 independent Tos17-inserted lines in the OsBOR4 locus. We selected those that had insertions in the exons, and identified five independent lines for which we were able to confirm a homozygous Tos17 insertion at the expected position in OsBOR4 (Supplementary Fig. S2A).
In all five of these Tos17 lines, the insertion lies in the exon, specifically in the first half of the open reading frame. This indicates that the function of OsBOR4 in these lines is likely to be completely disrupted. Moreover, the accumulation of transcript is probably disrupted by the insertion, which may create a premature stop codon and lead to the degradation of OsBOR4 mRNA. OsBOR4 transcript accumulation was examined in the immature anthers of three Tos17 insertion lines. The transcript was not detectable in any of the three lines tested (Supplementary Fig. S2B), suggesting that these Tos17-inserted plants were OsBOR4 knockout mutants. We also examined the effect of Tos17 insertion within OsBOR4 on B accumulation in the leaves. The concentration of B in the leaves of two Tos17 lines was identical to that of the wild-type plants (Supplementary Fig. S2C), suggesting that OsBOR4 is not involved in B transport from roots to shoots.
In the course of isolating the homozygous mutants, we noticed that the frequency of homozygous mutants for Tos17-inserted OsBOR4 was lower than that expected based on Mendelian genetics. To confirm this initial observation, we isolated heterozygous Tos17 lines for all five mutants and analyzed the segregation of the Tos17-inserted OsBOR4 in the selfed progeny. Fertility of the heterozygous mutant plants was comparable with that of the wild-type plants (Fig. 4A, B), although all five osbor4 heterozygous lines showed abnormal segregation rates (Table 1). Fewer osbor4 homozygous mutants were detected in the progeny of osbor4 heterozygous plants than predicted based on the expected ratio of 1 : 2 : 1 (Table 1). In all five lines, the frequency of the appearance of homozygous mutants was much lower than that expected according to Mendelian segregation. These results establish that disruption of OsBOR4 affects the reproductive process.
Phenotypes of osbor4 heterozygous plants. (A) Mature panicles of the wild type (left), ND5083 and ND8074 heterozygous plants. Bar = 5 cm. (B) Percentage of filled spikelets of the wild type, ND5083 and ND8074 heterozygous plants panicles. Data in (B) represent means ± SD (n = 10).
Segregation of progeny derived from self-pollinated heterozygous plants carrying a Tos17 insertion at the OsBOR4 locus
| Tos17 insertion line | Position | bor4/ bor4 | +/bor4 | +/+ | Total | χ2 (1 : 2 : 1) |
|---|---|---|---|---|---|---|
| ND5083 | Exon 1 | 1 | 27 | 32 | 60 | 8.2 × 10−8 |
| NF9912 | Exon 2 | 6 | 49 | 41 | 96 | 2.8 × 10−6 |
| NC7249 | Exon 3 | 7 | 40 | 35 | 82 | 6.9 × 10−5 |
| ND8074 | Exon 4 | 3 | 57 | 40 | 100 | 4.3 × 10−7 |
| NG0030 | Exon 7 | 4 | 35 | 27 | 66 | 2.9 × 10−4 |
| Tos17 insertion line | Position | bor4/ bor4 | +/bor4 | +/+ | Total | χ2 (1 : 2 : 1) |
|---|---|---|---|---|---|---|
| ND5083 | Exon 1 | 1 | 27 | 32 | 60 | 8.2 × 10−8 |
| NF9912 | Exon 2 | 6 | 49 | 41 | 96 | 2.8 × 10−6 |
| NC7249 | Exon 3 | 7 | 40 | 35 | 82 | 6.9 × 10−5 |
| ND8074 | Exon 4 | 3 | 57 | 40 | 100 | 4.3 × 10−7 |
| NG0030 | Exon 7 | 4 | 35 | 27 | 66 | 2.9 × 10−4 |
χ2 values against the expected segregation rate of 1 : 2 : 1 are shown.
Segregation of progeny derived from self-pollinated heterozygous plants carrying a Tos17 insertion at the OsBOR4 locus
| Tos17 insertion line | Position | bor4/ bor4 | +/bor4 | +/+ | Total | χ2 (1 : 2 : 1) |
|---|---|---|---|---|---|---|
| ND5083 | Exon 1 | 1 | 27 | 32 | 60 | 8.2 × 10−8 |
| NF9912 | Exon 2 | 6 | 49 | 41 | 96 | 2.8 × 10−6 |
| NC7249 | Exon 3 | 7 | 40 | 35 | 82 | 6.9 × 10−5 |
| ND8074 | Exon 4 | 3 | 57 | 40 | 100 | 4.3 × 10−7 |
| NG0030 | Exon 7 | 4 | 35 | 27 | 66 | 2.9 × 10−4 |
| Tos17 insertion line | Position | bor4/ bor4 | +/bor4 | +/+ | Total | χ2 (1 : 2 : 1) |
|---|---|---|---|---|---|---|
| ND5083 | Exon 1 | 1 | 27 | 32 | 60 | 8.2 × 10−8 |
| NF9912 | Exon 2 | 6 | 49 | 41 | 96 | 2.8 × 10−6 |
| NC7249 | Exon 3 | 7 | 40 | 35 | 82 | 6.9 × 10−5 |
| ND8074 | Exon 4 | 3 | 57 | 40 | 100 | 4.3 × 10−7 |
| NG0030 | Exon 7 | 4 | 35 | 27 | 66 | 2.9 × 10−4 |
χ2 values against the expected segregation rate of 1 : 2 : 1 are shown.
There are several possibilities for what may cause this phenotype, one being a defect in reproductive development. To examine this further, we observed the phenotypes of osbor4 mutants at the flowering stage (Fig. 5). The shoot length in the grain-filling period was comparable with that of the wild type under normal growth conditions (Fig. 5A). Furthermore, the structure and size of mutant panicles were also comparable with those from wild-type plants (Fig. 5B). We also counted the number of primary and secondary branches and spikelets per panicle and found no significant differences (Fig. 5C, D). Moreover, seed fertilities of osbor4 mutants were similar to those of the wild type (Fig. 5E). These results indicate that the growth and development of floral organs, as well as the fertility of the osbor4 homozygous mutants, were indistinguishable from those of the wild type.
Phenotypes of osbor4 mutants in the reproductive phase. (A) Wild-type (left) and ND8074 (right) mature plants. (B) Mature panicles of the wild type (left), ND5083 and ND8074 plants. Bar = 5 cm. (C) Numbers of primary branches, secondary branches and spikelets per panicle in the wild type and osbor4 mutants. (D) Number of spikelets per panicle in the wild type and osbor4 mutants. (E) Percentage of filled spikelets of the wild type and osbor4 mutant panicles. Data in (C)–(E) represent means ± SD (n = 10).
Next, we compared the size and structure of the flowers in wild-type plants and osbor4 homozygous mutants, which were comparable (Fig. 6A, B). Development of the stamen and pistil was also similar (Fig. 6C, D). We also examined pollen maturation of the osbor4 mutants using iodine–potassium iodide (I2–KI) staining (Fig. 6E). All of the pollen stained strongly blue, suggesting that pollen maturation was not affected by the mutation. The numbers of mature pollen grains in osbor4 mutants were comparable with those in wild-type plants (Fig. 6E). All images of the pollen for mutants and the wild type in Fig. 6E were taken under the same conditions, and the number of pollen grains seen in the figure represents the total number of pollen grains in a particular anther. We conducted fluorescein diacetate (FDA) staining of wild-type and osbor4 mutant (ND8074) pollen grains to verify pollen viability (Supplementary Fig. S3A, B). Both wild-type and ND8074 pollen grains were stained green with FDA (Supplementary Fig. S3A, B). We also stained wild-type and ND8074 pollen grains with DAPI (4′,6-diamidino-2-phenylindole) to examine cell division in osbor4 mutant grains (Supplementary Fig. S3C, D). Both wild-type and ND8074 pollen stained with DAPI showed two bright nuclei in each mature pollen grain (Supplementary Fig. S3C, D). Our findings indicate that the development of floral organs and pollen maturation and viability in the osbor4 mutants are normal.
Phenotypes of floral organs in osbor4 mutants. (A) Flowers of the wild type (left) and NF9912 (right). Le, lemma; Pl, palea. Bar = 5 mm. (B) Width and length of wild-type and NF9912 flowers. Data in (B) represent means ± SD (n = 10). (C) Stamens of the wild type (left) and NF9912 (right). An, anther. Bar = 1 mm. (D) Pistils of the wild type (left) and NF9912 (right). Bar = 5 mm. (E) I2–KI-stained pollen grains of the wild type and osbor4 mutants. Bars = 100 µm.
OsBOR4 functions in germination and/or elongation of the rice pollen tube
The fertility and reproductive development of the osbor4 homozygous mutants appeared normal (Fig. 5). However, as noted above, the number of osbor4 homozygous mutants was very low in the selfed progeny of heterozygous plants (Table 1). These results suggest the possibility that mutation in OsBOR4 does not completely block the reproductive processes of heterozygous plants, but does carry some negative effects. These may result in a disadvantage to the osbor4 mutant pollen compared wiht the wild-type pollen when the heterozygous mutants are self-pollinated. To determine why the osbor4 mutant plants segregate abnormally, we examined the pollen germination/elongation. We examined pollen elongation by staining with aniline blue (Fig. 7A, B). Wild-type plants were pollinated with pollen grains from the wild-type and homozygous plants of ND8074, and 30 min after pollination wild-type stigmas were stained with aniline blue. In Fig. 7A, we show five independent images of pollinated stigmas. When the osbor4 mutant was used as a male parent, the number of elongated pollen tubes 30 min after artificial self-pollination was lower compared with that when the wild type was used as the male parent (Fig. 7A, B), suggesting that OsBOR4 disruption affects pollen elongation and/or germination.
Elongation of the osbor4 mutant pollen tube. (A) Growth of pollen obtained from the wild-type (WT) and osbor4 mutant (ND8074) plants. Pollen grains were stained with aniline blue at 30 min after artificial pollination. Bars = 200 µm. (B) Germination rate of the wild-type and ND8074 pollen tubes. The germination rate was estimated by staining with aniline blue at 30 min after artificial pollination. (C) Segregation of F2 plants derived from self-pollinated F1 plants crossed with osbor4 alleles. Asterisks represent a significant difference from the elongation frequency of wild-type pollen tube (P < 0.01, t-test). Data in (B) represent means ± SD (n = 5).
We conducted further experiments to test this hypothesis. In the progeny of heterozygous plants of the ND5083, ND8074, NC7249 and NF9912 lines, the proportion of the osbor4 allele was in the range of 1–10%, considerably lower than the 25% expected by Mendelian genetics (Table 1). We crossed these independently isolated osbor4 homozygous mutants, and the resultant F1 plants were self-pollinated. If the negative effects caused by the disruption of OsBOR4 on pollen germination/elongation were similar among the mutant lines, then the selfed progeny of the F1 should have segregated the independent osbor4 allele in a Mendelian fashion. As shown in Fig. 7C, the frequency of homozygous lines among the selfed progeny of the F1 plants between homozygous plants of ND5083 and ND8074, and between homozygous plants of NC7249 and NF9912 were within the expected range, supporting our hypothesis of reduced competitiveness of mutant pollen in germination and/or tube elongation. Our results suggest that OsBOR4 is important for pollen tube elongation.
Discussion
OsBOR4 is a pollen-specific gene encoding an efflux type B transporter
Rice and A. thaliana contain four and seven AtBOR1-like genes, respectively (Nakagawa et al. 2007). We previously demonstrated that OsBOR1, the closest homolog of the A. thaliana B efflux transporter AtBOR1 in rice, also encodes an efflux B transporter and plays a major role in root B uptake and xylem loading (Nakagawa et al. 2007). As demonstrated for AtBOR1, yeast cells expressing OsBOR4 show less B accumulation in cells compared with those with an empty vector (Fig. 3), suggesting an efflux transport activity for OsBOR4. On the basis of microarray analysis, only OsBOR4 showed an anther-specific expression pattern in rice BOR genes (Supplementary Fig. S1). This strong expression of OsBOR4 in anthers was confirmed by real-time PCR (Fig. 1), and the analysis of rice plants carrying an OsBOR4 promoter–GFP also suggests that OsBOR4 expression is specifically abundant in pollen (Fig. 2). It was also shown by laser microdissection (LM)–microarray analysis (http://salad.dna.affrc.go.jp/CGViewer/SALADonARRAYs/) that OsBOR4 is specifically expressed at the tricellular stage of pollen maturation (Mihara et al. 2008, Suwabe et al. 2008). OsBOR1 is also strongly expressed in floral organs, including anthers (Supplementary Fig. S1), and in pollen during the maturation process (Mihara et al. 2008, Suwabe et al. 2008). This expression pattern of OsBOR1 indicates that OsBOR1, as well as OsBOR4, may also be involved in B transport in pollen. In fact, osbor1 mutants showed the sterile phenotype under B-deficient conditions (Nakagawa et al. 2007). However, OsBOR1 shows ubiquitous expression (Supplementary Fig. S1), and the major function of OsBOR1 has been demonstrated to be that of B acquisition in roots (Nakagawa et al. 2007). Disruption of OsBOR1 drastically reduced root to shoot B translocation, and thus resulted in reduced B concentration in shoots. This decreased B level in shoots may be an important factor for mutant sterility (Nakagawa et al. 2007), as observed in the wild-type plants, which showed severe sterility under B-limited conditions (Uraguchi and Fujiwara 2011). These results indicate that the effects of B deficiency, and of OsBOR1 disruption in pollen, on fertilization are almost indistinguishable in the osbor1 background. On the other hand, OsBOR4 is specifically expressed in a floral organ and the osbor4 mutant did not show a significant reduction in shoot B accumulation (Supplementary Fig. S2), suggesting that its function in pollen could be detected separately from the effect of B deficiency in shoots. Indeed, the osbor4 single knockout mutant shows abnormality in the fertilization process even with a functional OsBOR1 allele (Table 1, Fig. 7). The clear phenotypes of osbor4 mutants observed in this study support the significance of the function of OsBOR4 in pollen. However, disruption of OsBOR4 did not cause a crucial defect in fertility (Fig. 5). This suggests that other transporters, such as OsBOR1, also play a role in B transport in pollen.
In Arabidopsis, AtNIP7;1, an aquaglyceroporin, transports boric acid, and the expression of AtNIP7;1 has been shown to be anther specific (Li et al. 2011). Although the physiological function of AtNIP7;1 has not been demonstrated, a channel-type B transporter could also contribute to B homeostasis in pollen. Overall, to our knowledge, the present study is the first demonstration of the physiological function of a B transporter expressed in a floral organ.
Disruption of the OsBOR4 gene causes a defect in pollen germination and/or pollen tube elongation
The partial defect of pollen germination and/or pollen tube growth (Fig. 7) supported by an abnormal segregation rate in osbor4 heterozygous plants (Table 1) suggests that OsBOR4 functions in pollen germination and/or pollen tube elongation as a B-efflux transporter. Borate, as well as Ca2+ and H+, in the extracellular medium are important factors for the rigidity of the cell wall and pollen tube elongation (Steer and Steer 1989). Ca2+ oscillation regulated by several Ca2+ pumps and channels is crucial for growing cells in the pollen tube tip (Konrad et al. 2011). In tobacco and Arabidopsis, a proper gradient of cytosolic Ca2+ is essential for maintaining normal tip growth of the pollen tube (Michard et al. 2009). Ca2+ and H+ affect the modification of pectic compounds in the cell wall. De-esterified pectins can be cross-linked by Ca2+ and increase cell wall rigidity, although the esterified form facilitates plasticity and extensibility of the cell wall (Carpita and Gibeaut 1993). Low pH decreases the activity of pectin methyl esterase (PME), and thus H+ regulates cell wall plasticity (Li et al. 2002). These pectin modifications are essential for proper pollen tube growth (Li et al. 1994, Bosch et al. 2005, Jiang et al. 2005).
Cross-linking of RG-II by B is also important for pollen tube growth (Iwai et al. 2006, Delmas et al. 2008) through maintaining the mechanical strength of the pollen tube cell wall. B-mediated cross-linking of RG-II is thought to play an important role in the oscillatory growth of the pollen tube (Holdaway-Clarke et al. 2003). In Lilium formosanum, an optimal B supply in the extracellular medium leads to rapid pollen tube growth in vitro, achieved by a higher oscillation frequency and a faster growth speed (Holdaway-Clarke et al. 2003). These results indicate the significant role of B transporters in vivo in B supply to the apoplast to support efficient pollen tube growth. We demonstrated that OsBOR4 encodes an efflux B transporter (Fig. 3) expressed in the anther and pollen maturation stage (Fig. 1; and Mihara et al. 2008, Suwabe et al. 2008), and that the disruption of OsBOR4 retarded the growth of the pollen tube (Fig. 7). OsBOR4 may supply B to the apoplast of the pollen tube to regulate cell wall rigidity, which may result in the normal growth of the pollen tube. In the osbor4 mutant, lack of this adequate B supply to the cell wall might retard the growth of the pollen tube (Fig. 7). For pollen tube elongation, an adequate spatial and temporal export of B into the extracellular spaces might be essential for normal cross-linking of RG-II. In the osbor4 mutant pollen tubes, lack of this adequate B transport might lead to disruption of efficient growth of the tube, but may not cause the entire fertilization process to be severely defective. This hypothesis is further supported by our results showing that osbor4 heterozygous plants have an irregular segregation rate but that homozygous mutants show no evidence of sterility.
Materials and Methods
Plant materials
Wild-type rice (Oryza sativa L. cv. Nipponbare) and Tos17-inserted mutants of the OsBOR4 gene were used for the experiments. Plants were grown in pots under natural conditions.
Gene expression profiling
Total RNA was extracted from immature anthers of wild-type plants and ND5083, ND8074, and NF9912 using NucleoSpin RNA Plant (Macherey-Nagel). cDNA synthesis was performed using PrimeScript RT Master Mix (TAKARA). The RT–PCR for OsBOR4 was performed using Ex Taq (TAKARA). We used OsUBIQUITIN (OsUBQ) as an internal control. The primers used for quantification were 5′-GAA GGA GGA GGA AAT CGA AC-3′ and 5′-CTT CAC AGA GGT GAT GCT AAG G-3′ for OsUBQ (Os 01g22490), and 5′-GGA TTC AGG ATA TTG GCA CC-3′ and 5′-TTG CCA CCC AGT GCC GTA TA-3′ for OsBOR4 (Os 05g08430).
To determine accumulation of the OsBOR4 transcript, total RNA was isolated from wild-type 3-day-old shoots, 1-month-old leaves, sheath, root, leaf and anthers in the flowering stages. cDNA synthesis was performed using PrimeScript RT Master Mix (TAKARA). We used OsUBQ as an internal control and quantitative PCR was conducted using SYBR Premix Ex Taq II (TAKARA). The primers used for quantification were 5′-CGT GAT CCG TCT ACG GAC G-3′ and 5′-CAG CCT CTC GTC ACG GAA-3′ for OsBOR4. PCR was performed using Thermal Cycler Dice TP800 (TAKARA).
Pollen maturation assay
To evaluate pollen viability, I2–KI staining of pollen grains was performed as described by Chhun et al. (2007). Four to six anthers were removed from a spikelet before flowering and placed on a glass slide. The anthers were crushed into a powder and stained with 10 µl of 1% (v/v) I2 in 3% (v/v) KI, and a 1 µl aliquot was used to detect fertile and infertile pollen using a light microscope (Olympus BX50WI). Pollen grains that were stained black were judged as viable or living pollen, while sterile or dead pollen was stained yellow or light red.
FDA staining
Fluorescein diacetate (FDA) was used to determine the viability of pollen grains. To obtain mature pollen grains, picked mature anthers were crushed and treated with FDA solution (7% sucrose, 100 µg ml−1 FDA) for 5 min and then observed under blue light (wavelength 460–500 nm, Leica M165FC).
DAPI staining
Pollen grains with anthers were fixed in an ethanol : acetic acid (3 : 1) solution for 1 h at room temperature, and rinsed in 70% ethanol and water. Then they were transferred to 10 µg ml−1 DAPI and 0.5% Triton X-100, and observed under UV light (Olympus BX50WI).
Generation of ProOsBOR4-GFP and ProOsBOR4-GUS rice
The putative promoter region of OsBOR4 (from −4,055 to −1 bp of the translation initiation site) was amplified from genomic DNA using the primers 5′-CAC CAC GTA GTT GAA TGT CC-3′ and 5′-CTG AAA TAA CTG AGC TAA TAA AA-3′. The forward primer contained the sequence CACC (underlined) for subsequent TOPO cloning. The resulting product was cloned into a Gateway pENTR/D-TOPO cloning vector (Invitrogen). The obtained plasmid, pSU38, was subcloned into pGWB504 (Nakagawa et al. 2009) and pMDC163 (Curtis and Grossniklaus 2003) using LR clonase (Invitrogen) to obtain pSU43 (ProOsBOR4-GFP) and pSU44 [ProOsBOR4-GUS (β-glucuronidase)], respectively. pSU43 and pSU44 were introduced into the Agrobacterium tumefaciens strains EHA101 (Rifr) and pEHA101 (Kmr), respectively. Rice plants (cv. Nipponbare) were transformed to obtain ProOsBOR4-GFP and ProOsBOR4-GUS transgenic lines according to previously reported methods (Toki et al. 2006).
Vector integration in the genome of T1 generation plants was confirmed by genomic PCR. The primers used to detect the vector were as follows: OsBOR4pro-01F (5′-AAG ATT GAT GAT GCC CAA CC-3′) and sGFP-02R (5′-GTG CTC AGG TAG TGG TTG TC-3′) for ProOsBOR4-GFP, and OsBOR4pro-01F and GUS-01R (5′-CGT GAC ATC GGC TTC AAA TGG-3′) for ProOsBOR4-GUS. GFP fluorescence of pre-anthesis flowers in the PCR-positive lines was observed with a stereomicroscope (Leica MZ16F) and a confocal laser microscope FV-1000 (Olympus).
Boron uptake assay
Plasmid construction for expressing OsBOR4 in yeast was carried out by the Gap-Repair Cloning method (Ma et al. 1987). OsBOR4 cDNA was amplified by PCR using primers OsBOR4-pYES2-f2 (5′-GAC TCA CTA TAG GGA ATA TTA AGC TCA CCA TGA CGG GAA CTG TGA AAG-3′) and OsBOR4-pYES2-r (5′-GGC CGT TAC TAG TGG ATC CGA GCT CTT AGG ATC GTG GAG GAG-3′). The resulting PCR product has 25 bp of homologous pYES2 vector sequences (Invitrogen) at each 5′ and 3′ end to aid in integration by homologous recombination. Yeast strain Y0-1169 (MATa, his3, leu2, met15, ura3, YNL275W::kanMX4) was co-transformed with the PCR product and a HindIII/KpnI-digested linearized pYES2 vector. The resulting clone (pKKF088) was verified by sequencing. As a control plasmid, pTF477 was used for expressing AtBOR1 (Miwa et al. 2006). Three independent colonies were used for the B transport assay as previously described (Miwa et al. 2006). Cells from 40 ml of the mid-log phase cultured in SG medium (Takano et al. 2007) were collected by centrifugation at 3,000×g for 2 min and resuspended in fresh SG medium supplemented with 0.1 mM boric acid. After incubation with vigorous shaking at 30°C for 1 h, the cells were harvested at 3,000×g for 2 min at 4°C and washed twice with 40 ml of ice-cold deionized water. The cells were boiled in 1 ml of deionized water for 30 min. After centrifugation at 10,000×g for 10 min, the supernatant was used for the determination of intracellular B using inductively coupled plasms mass spectrometry (ICP-MS), and the cell pellets were dried and weighed.
Determination of B concentration
Leaf blades of the wild type and the OsBOR4 mutant at the flowering stage were harvested to determine the B concentration. Dried leaf blades were digested in 2 ml of nitric acid. After dilution, the concentration of B in the leaf blade was analyzed by ICP-MS (model SPQ9700; SII Nano-Technology, Seiko).
Pollen germination and elongation
The wild-type and osbor4 homozygous mutant plants were artificially pollinated by hand. After 30 min, the wild-type pistil was removed and stained with aniline blue on a glass slide for 5 min before observation by UV microscopy (Leica M165FC).
Funding
This work was supported by the Ministry of Agriculture, Forestry, and Fisheries of Japan [Genomics for Agricultural Innovation Grant IPG-0005 (to T.F.) and Genomics for Agricultural Innovation Grant RTR-0002 to (Y.N.)); the Japanese Society for the Promotion of Science [grant No. 22-8989 (to S.U.)]; the Ministry of Education, Culture, Sports, Science, and Technology of Japan [Grant-in-Aid for Scientific Research (to T.F.) and Grant-in-Aid for Scientific Research Priority Areas (to T.F.)].
Acknowledgments
We thank T. Nakagawa (Shimane University) for supplying pGWB504 and K. Aizawa for their excellent technical assistance.
Disclosures
The authors have no conflicts of interest to declare.
Abbreviations
- B
boron
- DAPI
4′,6-diamidino-2-phenylindole
- FDA
fluorescein diacetate
- GFP
green fluorescent protein
- GUS
β-glucuronidase
- ICP-MS
inductively coupled plasms mass spectrometry
- RG-II
rhamnogalacturonan
- RT–PCR
reverse transcription–PCR
