A C2H2-type zinc finger protein, SGR5, is involved in early events of gravitropism in Arabidopsis inflorescence stems

Authors


*(fax +81 743 72 5489; email m-tasaka@bs.naist.jp).

Summary

Plants can sense the direction of gravity and change the growth orientation of their organs. To elucidate the molecular mechanisms of gravity perception and the signal transduction of gravitropism, we have characterized a number of shoot gravitropism (sgr) mutants of Arabidopsis. The sgr5-1 mutant shows reduced gravitropism in the inflorescence stem but its root and hypocotyl have normal gravitropism. SGR5 encodes a zinc finger protein with a coiled-coil motif. The SGR5–GFP fusion protein is localized in the nucleus of Arabidopsis protoplasts, suggesting that SGR5 may act as a transcription factor. Analysis of GUS expression under the control of the SGR5 promoter revealed that SGR5 is mainly expressed in the endodermis, the gravity-sensing tissue in inflorescence stems. Furthermore, the observation that endodermis-specific expression of SGR5 using the SCR promoter in the sgr5-1 mutant restores shoot gravitropism indicates that it could function in the gravity-sensing endodermal cell layer. In contrast to other sgr mutants reported previously, almost all amyloplasts in the endodermal cells of the sgr5-1 mutant sedimented in the direction of gravity. Taken together, our results suggest that SGR5 may be involved in an early event in shoot gravitropism such as gravity perception and/or a signaling process subsequent to amyloplast sedimentation as a putative transcription factor in gravity-perceptive cells.

Introduction

Plants utilize the gravitational stimulus as a directional cue for their growth. Gravitropism, directed growth at a defined angle from gravity, is a coordinated response composed of four sequential processes: gravity perception, signal formation in the gravity-perceptive cell, intracellular and intercellular signal transduction, and an asymmetrical cell elongation between the upper and lower sides of the responding organs (Morita and Tasaka, 2004; Tasaka et al., 1999).

With respect to perception of gravitational stimuli, the starch–statolith hypothesis is widely accepted (Caspar and Pickard, 1989; Kiss et al., 1989; Sack, 1997). Starch-filled amyloplasts sediment towards the direction of gravity in specialized cells (statocytes) such as the columella cells in roots or cells in shoot bundle sheaths. Sedimentation of ‘floating’ amyloplasts of the starch synthesis-deficient phosphoglucomutase (pgm) mutant by hyper-gravitational force resulted in hypocotyl curvature (Fitzelle and Kiss, 2001). Moreover, displacement of amyloplasts under a high-gradient magnetic field in vertically growing stems caused organ curvature (Weise et al., 2000). These results suggested that displacement of amyloplasts within statocytes is the initial event of gravity perception and probably triggers downstream events responsible for organ curvature. In Arabidopsis shoots, genetic studies provided clear evidence that the endodermis containing amyloplasts is the gravity-sensing tissue (Fukaki et al., 1998). shoot gravitropism 1 (sgr1)/scarecrow (scr) and sgr7/short-root (shr) mutants that lack normal endodermis in all organs exhibited no gravitropism in their shoots (Di Laurenzio et al., 1996; Helariutta et al., 2000).

The traditional Cholodony–Went hypothesis proposes that asymmetric cell elongation during tropic growth is due to differential auxin distribution by lateral transport of auxin across the organ. Recent studies using Arabidopsis roots have substantiated the hypothesis (Esmon et al., 2005; Friml and Palme, 2002). An auxin transport facilitator PIN3 in the columella cells has been suggested to be involved in lateral auxin distribution in the root cap. Interestingly, PIN3, which is uniformly distributed in columella cells of vertical roots, is relocalized within 2 min to the new lower side of the cell in response to gravi-stimulation (Friml et al., 2002).

Although there is substantial evidence supporting the starch–statolith hypothesis and the Cholodony–Went theory, there is little information on molecular events linking gravity perception with asymmetric lateral auxin transport. To investigate the molecular mechanism of early events of the gravitropic response, we have isolated and analyzed a number of sgr mutants with no or reduced gravitropic responses in shoots (Fukaki et al., 1996a,b; Yamauchi et al., 1997). Light-oriented growth during elongation of the primary inflorescence stem of sgr mutants indicates that the mutants are capable of phototropic growth, suggesting that they retain the ability for organ curvature by differential cell growth (Fukaki et al., 1996a,b; Yamauchi et al., 1997). We have reported that amyloplasts in endodermal cells did not sediment in the direction of gravity in mutants that exhibit no or reduced gravitropism, namely sgr2, sgr3, zig/sgr4and grv2/sgr8 (Morita et al., 2002; Silady et al., 2004; Yano et al., 2003). Molecular cloning of these SGR genes has demonstrated that vacuoles have a novel function for gravity perception in stems, and that membrane traffic between the trans-Golgi network and the vacuoles via endosomes is important for such vacuolar function (Kato et al., 2002; Niihama et al., 2005; Yano et al., 2003). Live cell imaging enabled us to show that functions and membrane dynamics of vacuoles are closely related to amyloplast movement (Saito et al., 2005).

Here, we report a novel type of SGR gene, SGR5, encoding a C2H2-type zinc finger protein that is localized in the nucleus. In contrast to other sgr mutants, an endodermal layer is formed in the inflorescence stem and almost all amyloplasts sediment in the direction of gravity in the sgr5 mutant. Molecular genetic and cellular studies of this mutant suggest that SGR5 may be involved in gravity perception and/or the signaling process subsequent to amyloplast sedimentation as a putative transcription factor in gravity-perceptive cells.

Results

Phenotypic characterization of the sgr5 mutant

When wild-type Arabidopsis (Columbia ecotype) inflorescence stems were gravi-stimulated by horizontal placement, primary shoots bent upwards and became vertical in approximately 90 min at 23°C in the dark (Figure 1a,b,g). In contrast, sgr5-1 and sgr5-3 mutants exhibited weak gravitropic responses and the curvature did not reach 90° even after 3 h (Figure 1d,e,g). The sgr5-2 allele derived from the Landsberg erecta ecotype also showed reduced shoot gravitropism compared with wild-type L er (data not shown). Hypocotyls and roots of sgr5 exhibited a normal gravitropic response (Yamauchi et al., 1997). In addition, both inflorescence stems and hypocotyls of sgr5 showed a positive phototropic response, suggesting that sgr5 is capable of asymmetric growth, as reported previously (Yamauchi et al., 1997).

Figure 1.

 Gravitropic response of the sgr5 mutant.
Gravitropic response of inflorescence stems of 4-week-old plants after 0 min (a,d) or 90 min (b,e) of horizontal gravi-stimulation. (c,f) Lateral shoots of 5-week-old plants. (a–c) Wild-type plants, (d–f) sgr5-3 plants. g, direction of gravity.
(g) Time course of gravitropic response of inflorescence stems: wild-type (open circles), sgr5-1 (closed squares) and sgr5-3 (closed circles). Four-week-old plants were gravi-stimulated by being placed horizontally at 23°C under dim non-directional light. At least eight individuals of each genotype were examined. Bars represent SD.

Wild-type primary inflorescence stems grow upright and their lateral shoots grow upwards, although not upright (Figure 1c). Lateral shoots of the sgr5 mutant tended to grow in a horizontal direction similar to other sgr mutants (Figure 1f) (Fukaki et al., 1996a,b; Yamauchi et al., 1997). Growth arrest at the primary shoot apex was observed at low frequency during production of flowers. The elongation rate of inflorescence stems is, however, comparable to that of wild-type at the earlier growth stage when the gravitropic response is examined, as reported previously (Yamauchi et al., 1997). Thus, the gravitropic response of sgr5 is unlikely to be influenced by the rare growth arrest observed. Other obvious morphological phenotypes were not observed in sgr5.

Cloning of the SGR5 gene

The SGR5 gene was cloned based on its map position. The SGR5 locus mapped at the top region of chromosome 2. By analyzing approximately 500 independent F2 lines, the locus was narrowed to a 25 kb interval in the bacterial artificial chromosome clones F23I14 and F14H20, containing three predicted ORFs. All three ORFs in sgr5-1 genomic DNA were sequenced, and a single G-to-A substitution was found in one ORF At2g01940 (Figure 2a) leading to the single amino acid substitution Asp to Asn at position 159. Plants homozygous for the sgr5-2 and sgr5-3 alleles, which are T-DNA-insertion mutations, were tested by RT-PCR to determine potential SGR5 expression. In both alleles, a slight amount of transcript could be detected using primer sets that amplify the region downstream of the T-DNA insertion. The transcript from sgr5-3 was sequenced and deduced not to encode an in-frame fusion protein of SGR5 (data not shown). Thus, sgr5-3 at least is probably a null mutant.

Figure 2.

 Molecular cloning of the SGR5 gene.
(a) The SGR5 locus was mapped to an approximately 25 kb region between two cleaved-amplified polymorphic sequence markers on chromosome 2. The gene structure of At2g01940 is shown schematically: boxes (open boxes, untranslated region; closed boxes, coding regions) and lines between boxes indicate exons and introns, respectively. sgr5-1 is a single-base-substitution mutant. sgr5-2 and sgr5-3 are T-DNA-insertion mutants.
(b) Complementation analysis. The gravitropic response of wild-type (open circles), sgr5-1 (closed squares), sgr5-1 containing SGR5 genomic region (sgr5-1/gSGR5; closed triangles) and sgr5-1/pSCR::SGR5 (closed diamonds) measured as described in Figure 1(g). At least five individuals of each transgenic line were examined. Bars represent SD.

The 6 kb wild-type genomic fragment containing the At2g01940 ORF and predicted promoter region was cloned and introduced into sgr5-1 plants for a complementation test. Inflorescence stems of the transgenic plants exhibited gravitropism with similar kinetics as the wild-type (Figure 2b), and their lateral shoots grew upwards (data not shown). Taken together, these results lead us to conclude that At2g01940 is the SGR5 gene.

The SGR5 gene encodes a C2H2-type zinc finger protein

At2g01940 was annotated in the Arabidopsis Information Resource (TAIR) as a C2H2-type zinc finger family protein similar to the putative zinc finger protein INDETERMINATE1 (ID1) (Oryzasativa). Sixteen Arabidopsis genes including SGR5 show sequence similarity with maize ID1 (Colasanti et al., 1998) in their zinc finger domains (Figure 3a). In addition to the predicted four zinc finger motifs, these genes have a conserved sequence at the N-terminal region containing an R/K-rich stretch that could be a nuclear localization signal (Figure 3a). Three genes (SGR5, At1g25250 and At1g68130) show high similarity to one another and are less similar to ID1 than the other 13 genes are. Also, the three genes contain a highly conserved domain in addition to the zinc finger domain at their C-terminal region (Figure 3b). Thus, the three genes probably comprise a small sub-family. An amino acid substitution was found at the highly conserved amino acid in the third zinc finger in sgr5-1, suggesting that the third zinc finger is important for the molecular function of SGR5 and probably for that of other family members.

Figure 3.

 Alignment of amino acid sequences among C2H2-type zinc finger proteins homologous to SGR5.
(a) Alignment of the zinc finger domains of SGR5, its Arabidopsis paralogues (At1g25250 and At1g68130) and ID1 of Zea mays (AAC18941). Black lines and a gray line indicate each C2H2-type zinc finger domain and a putative nuclear localization signal, respectively. The open box indicates amino acids at positions +1 to +6 of the α-helix in the third finger (see text). An asterisk indicates the mutation site of sgr5-1.
(b) An unrooted neighbor-joining tree of the regions indicated in (a) of SGR5 homologues, generated using clustal w.
(c) Alignment of the conserved C-terminal region among three Arabidopsis SGR5 homologues. The gray bar indicates a predicted coiled-coil region. Amino acid alignment was carried out with clustal w.

We tested whether the sub-family members including SGR5 are expressed using semi-quantitative RT-PCR analysis. All three genes were detected in almost all organs examined, with significant organ specificity for levels of expression (Figure 5a). For example, the SGR5 transcripts were most abundant in the stem and the flower, while those of other two genes were most abundant in the flower. The results suggest that each sub-family member may have specific function.

Figure 5.

 Expression pattern of SGR5.
(a) Semi-quantitative RT-PCR analysis of SGR5 and its paralogous genes. RNA samples were prepared from roots (lane 1), inflorescence stems (lane 2), hypocotyls from 4-day-old dark-grown etiolated seedlings (lane 3), 7-day-old seedlings grown under white light (lane 4), cauline leaves (lane 5), flowers (lane 6) and buds (lane 7). ACT8 was used as an internal control.
(b–f) pSGR5::GUS expression patterns in wild-type plants: (b) 7-day-old seedling, (c) young lateral root, (d) stamen, (e) pistil, (f) cross-section of the elongation zone of inflorescence stem.
(g) Quantitative RT-PCR analysis of the expression level of SGR5 in sgr1 and sgr7. RNA samples were prepared from inflorescence stems of 5- or 6-week-old plants. Relative expression levels are shown.

Intracellular localization of SGR5 protein

As the SGR5 protein was predicted to have a nuclear localization signal, we sought more direct evidence that SGR5 is a nuclear protein. A GFP gene was fused at the C-terminus of SGR5 cDNA, and expressed under the control of the CaMV 35S promoter in Arabidopsis protoplasts. The GFP fluorescence of SGR5–GFP was observed in the nucleus (Figure 4a,b), while that of GFP alone was observed both in the cytoplasm and nucleus (Figure 4c,d). When the SGR5–GFP protein was expressed under the control of an authentic promoter in Arabidopsis, the GFP fluorescence was localized in one large organelle in each cell of the inflorescence stem endodermis (Figure 4e). pSGR5::SGR5–GFP restored the gravitropic response to sgr5-1, indicating that the SGR5–GFP fusion protein retains the molecular function of SGR5 (Figure 4f). These results indicate that SGR5 is a nuclear-localized protein.

Figure 4.

 Intracellular localization of SGR5 protein.
SGR5–GFP fusion proteins (a,b) or GFP (c,d) were expressed under the control of the CaMV 35S promoter in Arabidopsis protoplasts. (a,c) GFP fluorescence images and (b,d) Nomarski images were taken with an epifluorescence microscope. Scale bars in (a)–(d) = 10 μm.
(e) Longitudinal section of the sgr5/pSGR5::SGR5GFP transgenic plant observed with confocal microscopy. The GFP fluorescence (green), auto-fluorescence of chloroplasts (red), and optical image were merged. Note that amyloplasts in the endodermal cell are dispersed because the specimen was observed on a horizontal stage. co, cortex; en, endodermis; st, stele. Scale bar = 5 μm.
(f) Gravitropic response of the sgr5/pSGR5::SGR5GFP transgenic plant after gravi-stimulation for 90 min. g, direction of gravity.

The expression pattern of SGR5

Based on RT-PCR analysis results, the SGR5 gene appeared to be expressed throughout the whole plant. To analyze the expression pattern of SGR5, we constructed an SGR5 promoter–GUS reporter gene and introduced it into wild-type plants. The GUS expression pattern regulated by the SGR5 promoter is consistent with the expression pattern of authentic mRNA detected by RT-PCR (Figure 5a). GUS activity was detected in seedlings, hypocotyls, roots, floral organs and stems (Figure 5b–e). Strong GUS activity was detected at the shoot apical meristem but not at the root apical meristem (Figure 5c). GUS activity appeared to gradually decrease with ageing in both the shoot and root vasculature. It is noteworthy that the most striking expression was observed in the endodermis of the inflorescence stem, although a weak GUS signal was also found in the vasculature (Figure 5f). Consistently, GFP fluorescence of SGR5–GFP expressed by the SGR5 promoter was observed within the nucleus of endodermal cells, as described above (Figure 4e). In addition, the expression level of SGR5 transcripts significantly decreased in the sgr1/scr and sgr7/shr mutants lacking a normal endodermis (Figure 5g).

The endodermis is the gravity-perceptive tissue of Arabidopsis inflorescence stems. To test whether endodermal expression of SGR5 is sufficient for a gravitropic response, SGR5 was expressed under the control of the endodermis-specific SCARECROW (SCR) gene promoter in the sgr5-1 mutant (Figure 2b). Transgenic plants showed almost a normal gravitropic response, suggesting that SGR5 is involved in an early process in gravitropism that takes place in the gravi-perceptive cells.

Amyloplasts sedimented in the sgr5 mutant

Amyloplasts within the endodermal cells sedimented in the direction of gravity in wild-type plants. The movement of amyloplasts towards a gravitational force is thought to trigger the gravitropic response. In some sgr mutants, amyloplasts are dispersed within the cell and do not sediment in the direction of gravity (Morita et al., 2002; Silady et al., 2004; Yano et al., 2003). Although one or two amyloplasts were occasionally found in the middle part of the cell in sgr5-1, almost all amyloplasts sedimented in the direction of gravity in endodermal cells of sgr5-1. The result suggests that SGR5 is a novel type of gene involved in a process after amyloplast sedimentation but that takes place in the gravi-perceptive cells.

Discussion

Molecular function of the SGR5 protein

SGR5 is a protein similar to the putative C2H2-type zinc finger protein INDETERMINATE 1 (ID1) (Oryza sativa) based on a TAIR gene search. In maize, ID1 regulates a leaf-generated signal required for the transition to flowering (Colasanti et al., 1998). Another similar protein from Solanum tuberosun, PCP1, has also been reported previously (Kuhn and Frommer, 1995). C2H2-type zinc finger domains represent a general DNA-binding motif found in eukaryotic transcription factors that was first described for Xenopus laevis TFIIIA (Miller et al., 1985). The C2H2-type zinc finger domain is involved in a wide range of functions for DNA or RNA binding and protein–protein interactions. A recent in silico study found 176 C2H2-type zinc finger proteins in the Arabidopsis genome and organized them into classes (Englbrecht et al., 2004). It has previously been proposed that the characteristic feature of most plant C2H2 zinc fingers is the QALGGH sequence within the finger structure (Takatsuji, 1999). Well-characterized proteins such as SUPERMAN (Sakai et al., 1995) and proteins closely related to the EPF family of petunia (Takatsuji, 1999) containing the QALGGH motif are classified in one sub-family (C1) of the largest family (named as set C) according to the in silico analysis. SGR5 belongs to a distinct family (named as set A) containing four zinc finger motifs without the QALGGH motif (Englbrecht et al., 2004).

Arabidopsis proteins closely related to maize ID1 and those related to TRANSPARENT TESTA 1 (TT1) (Sagasser et al., 2002), a WIP family protein, also have been categorized as members of set A. In addition to the nuclear localization of SGR5–GFP (Figure 4) and TT1–GFP fusion proteins (Sagasser et al., 2002), the sequence-specific DNA-binding activity of maize ID1 protein (Kozaki et al., 2004) strongly suggests that the family members are transcriptional regulators.

A mutation in sgr5-1 occurs at the highly conserved aspartic acid residue in the third zinc finger (Figure 3a). The residue is in the variable region that constitutes an α-helix following a β-sheet in the third zinc finger. It has been postulated that amino acid residues in helix positions −1, +3 and +6 (+1, the first amino acid of the helix) contribute to DNA recognition (Wolfe et al., 2000). Aspartic acid 159 is at position +3 in the helix of the third zinc finger. A T-DNA-insertion allele, sgr5-3, yields a comparable phenotype to sgr5-1, suggesting that the third zinc finger plays an important role in their molecular function such as DNA recognition. Consistently, biochemical in vitro analysis of maize ID1 protein has demonstrated that the second and third zinc fingers are indispensable for DNA-binding activity, although the first and fourth zinc fingers are not required for activity (Kozaki et al., 2004).

Our BLAST search against the Arabidopsis genome resulted in finding only two genes closely related to SGR5 (At1g68130 and At1g25250) (Figure 3), consistent with the in silico classification indicating that only these three genes constitute a small subgroup in set A (Englbrecht et al., 2004). Two orthologous genes (Q6YSA1, Q67V55) were found in the O. sativa genome. All these genes share a similar domain presumed to be a coiled-coil domain at the C-terminal region in addition to the N-terminal putative NLS and four zinc fingers (Figure 3c). None of the physiological functions of these genes has been reported yet except for SGR5. The significant residual gravitropic response in sgr5-3 might be due to redundant function of these closely related two paralogues. Alternatively, the genetic pathway involved in gravitopism may be branched at an early step of the response, and SGR5 may have a unique function for part of the multi-branched pathway. The differences in expression pattern among three genes might reflect their independent functions. Unfortunately, a T-DNA-insertion line for At1g25250 does not exist, and we could not confirm the insertion for At1g68130. Further study is required to understand the functions of SGR5 paralogues.

Physiological function of SGR5

GUS expression regulated by the SGR5 promoter was observed in the vasculature tissue of various organs (Figure 5). Interestingly, strong and specific GUS expression was observed in the endodermis of inflorescence stems (Figure 5f). Consistently, the expression of SGR5 significantly decreased in inflorescence stems of sgr1 and sgr7 (Figure 5g). The specific expression in the endodermis implies that expression of SGR5 may be partially regulated by SCR/SGR1 in inflorescence stems. Moreover, complementation of the reduced gravitropic response of sgr5-1 by endodermis-specific expression of SGR5 driven by the SCR promoter indicates that SGR5 function within endodermal cells is required for the gravitropic response (Figure 2b). Although GUS staining was detected in the endodermis in hypocotyls, GUS activity was weak and the expression was not specific to the endodermis (Figure 5b and data not shown). In addition, GUS staining was not detected at the root tip, including the root cap and root apical meristem (Figure 5c). These results are consistent with normal gravitropism in the hypocotyls and roots of the sgr5-1 mutant (Yamauchi et al., 1997).

Our previous studies have demonstrated that amyloplasts sediment to the bottom of wild-type endodermal cells in response to a gravitational force (Fukaki et al., 1998). In sgr2, sgr3, zig/sgr4 and grv2/sgr8, amyloplasts appeared to stick to the cell periphery, not only on the bottom side but also on the upper and lateral sides of the cells (Morita et al., 2002; Silady et al., 2004; Yano et al., 2003). The responsible genes in these mutants are closely related to the biogenesis or function of endodermal cell vacuoles. Shoot endodermal cells are mostly filled by a large central vacuole (Morita et al., 2002). Living endodermal cell imaging has demonstrated that displacement of amyloplasts is due to amyloplast movement through transvacuolar strands that transiently appear and disappear in a highly dynamic mode (Saito et al., 2005). In contrast, transvacuolar strands are rarely formed and movement of amyloplasts is severely restricted in zig and sgr2. In sgr5, almost all amyloplasts sedimented in the direction of gravity in endodermal cells (Figure 6). Although a few amyloplasts were occasionally found in the middle part of the endodermal cells in sgr5, the upper cell periphery was devoid of amyloplasts. The status of amyloplasts in sgr5 is clearly distinct from those of other sgr mutants (Morita et al., 2002; Silady et al., 2004; Yano et al., 2003). Thus, we conclude that amyloplasts sediment towards a gravitational force in the endodermal cells of sgr5-1. These results suggest that SGR5 is a putative transcription factor involved in the early events in the gravitropic response, such as gravity perception and/or a signaling process subsequent to amyloplast sedimentation in gravity-perceptive cells.

Figure 6.

 Amyloplast sedimentation in sgr5-1.
Longitudinal sections of inflorescence stems (approximately 3 cm below the apex) of wild-type (a) and sgr5-1 (b). The growth orientation of stems was maintained during fixation. G, direction of gravity; ep, epidermis; en, endodermis. Arrowheads indicate the location of amyloplasts in endodermal cells. Scale bars = 20 μm.

A number of transcriptional regulators have been reported to be involved in shoot gravitropism. SCR/SGR1 and SHR/SGR7 are GRAS (GAI, RGA, SCR) family proteins required for formation and differentiation of the endodermis during development (Di Laurenzio et al., 1996; Helariutta et al., 2000). Several dominant mutants of the AUX/IAA gene family have been identified as agravitropic mutants (Fukaki et al., 2002; Nagpal et al., 2000; Rouse et al., 1998; Tatematsu et al., 2004; Tian and Reed, 1999). A few ARF family genes that regulate transcription in response to auxin coordinately with members of the AUX/IAA family have also been demonstrated to be involved in shoot gravitropism (Harper et al., 2000; Okushima et al., 2005; Stowe-Evans et al., 1998; Watahiki and Yamamoto, 1997). Phytochrome-interacting PIL5 (PIF3-like 5), a bHLH-type transcription regulator, has been demonstrated to be a negative component in phytochrome-mediated inhibition of hypocotyl negative gravitropism (Oh et al., 2004). We report here that SGR5 is a novel type of C2H2 finger putative transcriptional factor expressed within gravity-perceptive cells. Microarray analysis of gravity-responsive genes with root tip samples has indicated that a number of transcriptional regulators including AUX/IAA genes are up-regulated after gravi-stimulation (Kimbrough et al., 2004). We have also found that some AUX/IAA genes are upregulated upon gravi-stimulation in shoots (unpublished results). In contrast, the mRNA expression level of SGR5 is not affected by gravi-stimulation within the limits of our analysis (data not shown). Then, how is SGR5 involved in gravitropism within the gravity-perceptive cells? Possibly, SGR5 may play a role in the transcription of a certain set of genes required for gravitropism constitutively or during developmental processes. Alternatively, the activity of SGR5 could be regulated upon gravi-stimulation leading to transcription of a set of genes required for the gravitropic response. Further analyses such as exploration of target genes of SGR5 will clarify its function in shoot gravitropism.

Experimental procedures

Plant materials and growth conditions

The Columbia (Col) ecotype of Arabidopsis thaliana was used as the wild-type. The screening strategy used to isolate sgr5-1 mutants has been described previously (Yamauchi et al., 1997). Plants were grown on soil under constant white light at 23°C. The sgr5-2 (Landsberg ecotype) is a gene trap line provided by Cold Spring Harbor Laboratory (New York State, USA) and DsG T-DNA is inserted at the first exon of At2g01940. The sgr5-3 (Columbia ecotype, San Diego, CA, USA ) is a T-DNA-insertion line from the SALK Institute and T-DNA is inserted at the first intron.

Gravitropism assay

To examine the gravitropic responses of inflorescence stems, intact plants with primary stems 4–8 cm in length were placed horizontally under non-directional dim light or in the dark at 23°C (Fukaki et al., 1996a,b). The curvature of the stem was measured at indicated times as the angle formed between the growing direction of the apex and the horizontal base line. At least ten individuals of each genotype were examined.

Mapping of SGR5

A Landsberg erecta wild-type plant was crossed to the sgr5-1 mutant plant to generate a mapping population. SGR5 was mapped to the top of chromosome 2. For fine-scale mapping, DNA was prepared from approximately 500 F2 progeny. We generated cleaved-amplified polymorphic sequence markers that can recognize polymorphisms between Col and L er based on information provided by TAIR.

Cloning of SGR5 and plasmid construction

For a complementation analysis, the 6.0 kb genomic DNA fragment of SGR5 including the 2.4 kb putative promoter region and 2.0 kb 3′ downstream region was cloned from a BAC F14H20 DNA into binary vector pBIN19. The resulting constructs were transformed into Agrobacterium tumefaciens strain MP90 and then introduced into sgr5-1 plants. T1 plants (sgr5-1/gSGR5) were selected based on resistance to kanamycin. The presence of the transgene in these plants was confirmed by PCR. Most sgr5-1 T1 plants containing the transgene exhibited normal morphology and gravitropism (Figure 2b).

Total RNA was isolated from inflorescence stems using the RNeasy Plant Mini Kit (Qiagen, Tokyo, Japan). cDNAs were synthesized using SuperScript II reverse transcriptase (Invitrogen, Tokyo, Japan). PCR amplification of SGR5 cDNA was performed with the following primer set: cSGR5-F (5′-AGGATCCATGTTGTCCAACAAGAACACAAA-3′) and cSGR5-R (5′-AGGATCCTTAAAAACCATTTTCCAACTCTC-3′). The PCR products were cloned into pBlueScript II SK (Stratagene, La Jolla, CA, USA). The SGR5 cDNA was inserted downstream of the SCR promoter on the pBI binary vector, then introduced into sgr5-1 plants. T1 plants (sgr5-1/pSCR::SGR5) were selected for resistance to kanamycin.

Expression analysis of the SGR5 and paralogous genes

The 2.4 kb genomic fragment upstream of the SGR5 ORF was used as a promoter region for expression analysis of SGR5. The SGR5 promoter was inserted upstream of the uidA gene in the pBI101 vector (pBI101-pSGR5::GUS). The construct was transformed into wild-type plants by Agrobacterium-mediated transformation. To detect GUS activity in the transgenic plants, tissues were first immersed in 90% ice-cold acetone for 15 min and then incubated in GUS staining solution (100 mm sodium phosphate, pH 7.0, 10 mm EDTA, 5 mm ferricyanide, 5 mm ferrocyanide, 0.1% TritonX-100, 0.52 mg ml−1 5-bromo-4-chloro-3-indolyl-β-d-glucuronic acid) at 37°C. After being cleared in 70% ethanol, tissues were observed for GUS staining with a light microscope. For histological sections, GUS-stained inflorescence stems were dehydrated by a 30–100% ethanol series, then embedded in Technovit 7100 (Heraeus Kulzer, Wehrheim, Germany), and sectioned with a microtome.

The expression of SGR5 paralogues was analyzed by RT-PCR. cDNA was prepared as described above. ACT8 was used as an internal control (Hibara et al., 2003). The PCR conditions were cycles of 0.5 min at 94°C, 0.5 min at 55°C, and 1 min at 72°C. PCR amplifications were repeated for 29 cycles for SGR5 and At1g25250, 27 cycles for At1g68130, and 24 cycles for ACT8. The following primers used for amplification were designed at a specific region for each gene: SGR5f (5′-AACACAAACACATGTTGTGTGG-3′), SGR5r (5′- TGGTTTCACTAGACGGAGTGC-3′), At1g68130f (5′-CAACACCCATCATCACAACCC-3′), At1g68130r (5′- CAATGGAGAGGTCCCCGTTC-3′), At1g25250f (5′- GAGCTGACGCAACCCATAAG-3′), At1g25250r (5′- AAGGGTAAGCAAAAGCTGGTG-3′), ACT8f (5′-ATGAAGATTAAGGTCGTGGCA-3′) and ACT8r (5′-TCCGAGTTTGAAGAGGCTAC-3′).

Quantitative RT-PCR analysis

RNA was extracted with the RNeasy kit (Qiagen, Tokyo, Japan) from inflorescence stem samples of 5- or 6-week-old plants. cDNA was prepared from 2 μg total RNA using Superscript II reverse transcriptase (Invitrogen) according to the manufacturer's instructions. Quantitative RT-PCR was performed on a LightCycler apparatus (Roche Applied Science, Tokyo, Japan) with SYBER Premix EX Taq (Takara, Ohtsu, Japan) according to the manufacturer's instructions. PCR was carried out using the following program: once for 30 sec at 95°C, 40 cycles of 5 sec at 95°C and 20 sec at 60°C, followed by melting curve analysis. Specific primers were designed based on the universal probe library for Arabidopsis https://www.roche-applied-science.com/sis/rtpcr/upl/adc.jsp. Calibration was performed using a cDNA fragment cloned in a plasmid as a template. SGR5 expression levels were normalized to ACT8 expression levels. RT-PCR experiments were performed using three independent RNA samples.

Intracellular localization analysis

We used a modified pBI221 vector (pBI221ΔuidA), pBI221 lacking the uidA gene, for transient expression in cultured cells. GFP (S65T) was fused in-frame at the C-terminus of SGR5 with a triple glycine linker. The SGR5GFP gene (p35S::SGR5GFP) or the GFP gene (p35S::GFP) were inserted into the multi-cloning site of pBI221ΔuidA. Cultured Arabidopsis cells were maintained in modified Murashige and Skoog medium at 23°C in the dark. Plasmids and carrier DNA were introduced into protoplasts that had been generated from suspension cultures. The experiments were performed according to the method described previously (Takeuchi et al., 2000). Fluorescence of GFP was observed with an epifluorescence microscope (Eclipse E800; Nikon, Tokyo, Japan) and images were captured by cooled CCD camera (VB-6010; Keyence, Osaka, Japan).

To express the SGR5–GFP fusion protein in planta, the uidA gene in pBI101- pSGR5::GUS was replaced with the SGR5GFP gene (pBI101- pSGR5::SGR5GFP). The construct was transformed into sgr5-1 plants by the Agrobacterium-mediated transformation procedure. The majority of T1 plants containing the transgene exhibited almost normal gravitropism in the inflorescence stems and their lateral shoots grew upwards.

Confocal microscopy

Inflorescence stem segments were prepared as described previously (Saito et al., 2005). Fluorescence was imaged by confocal laser scanning microscopy (FV1000; Olympus, Tokyo, Japan). GFP fluorescence (green) and chlorophyll autofluorescence (red) were detected with 500–510 and 640–700 nm spectral settings, respectively, with 488 nm excitation.

Histological analysis

For observation of amyloplast sedimentation, stem segments were treated, embedded, and sectioned as described previously (Morita et al., 2002).

Acknowledgements

We thank Seiko Ishihara and Noriko Ueda for technical assistance. The financial support of a Grant-in-Aid for Scientific Research on Priority Areas (Organelle Differentiation as a Strategy for Environmental Adaptation in Higher Plants) from the Ministry of Education, Culture, Sports, Science and Technology of Japan (grant number 16085205) is gratefully acknowledged.

Ancillary