SEARCH

SEARCH BY CITATION

Keywords:

  • gravitropism;
  • WAV3;
  • root;
  • E3 ligase;
  • RING finger protein;
  • Auxin

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental Procedures
  7. Acknowledgements
  8. References
  9. Accession numbers:
  10. Supporting Information

Regulation of the root growth pattern is an important control mechanism during plant growth and propagation. To better understand alterations in root growth direction in response to environmental stimuli, we have characterized an Arabidopsis thaliana mutant, wavy growth 3 (wav3), whose roots show a short-pitch pattern of wavy growth on inclined agar medium. The wav3 mutant shows a greater curvature of root bending in response to gravity, but a smaller curvature in response to light, suggesting that it is a root gravitropism-enhancing mutation. This wav3 phenotype also suggests that enhancement of the gravitropic response in roots strengthens root tip impedance after contact with the agar surface and/or causes an increase in subsequent root bending in response to obstacle-touching stimulus in these mutants. WAV3 encodes a protein with a RING finger domain, and is mainly expressed in root tips. RING-containing proteins often function as an E3 ubiquitin ligase, and the WAV3 protein shows such activity in vitro. There are three genes homologous to WAV3 in the Arabidopsis genome [EMBRYO SAC DEVELOPMENT ARREST 40 (EDA40), WAVH1 and WAVH2 ], and wav3 wavh1 wavh2 triple mutants show marked root gravitropism abnormalities. This genetic study indicates that WAV3 functions positively rather than negatively in root gravitropism, and that enhancement of the gravitropic response in wav3 roots is dependent upon the function of WAVH2 in the absence of WAV3. Hence, our results demonstrate that the WAV3 family of proteins are E3 ligases that are required for root gravitropism in Arabidopsis.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental Procedures
  7. Acknowledgements
  8. References
  9. Accession numbers:
  10. Supporting Information

The roots of higher plants change their growth direction in response to various environmental stimuli, including gravity, light, moisture, nutrients, temperature and physical obstacles. This facilitates environmental adaptation responses to attain the maximum possible growth advantage. Both differential organ growth and helical organ growth affect the growth direction of roots (Migliaccio et al., 2009). Differential growth induces root bending without root tip rotation, and plant tropisms are defined as differential growth responses that re-orient plant organs in response to the direction of physical stimuli (Okada and Shimura, 1994; Esmon et al., 2005). Helical organ growth with root tip rotation, and the consequent rotation of epidermal cell files, induce circumnutation and slanting growth on the surface of vertical agar medium (Hashimoto, 2002).

The wavy growth pattern of Arabidopsis roots observed on inclined agar medium was described by Okada and Shimura (1990) in their study of the mechanisms controlling changes in root growth direction in response to obstacle-touching stimulus. When Arabidopsis seedlings are grown on a hard agar plate that is inclined by 45° to the direction of gravity, the primary roots do not show straight growth as expected, but show a wavy growth pattern, as though the agar surface represents an obstruction to vertical growth. This wavy growth pattern is produced when the root tip grows alternately to the right and to the left of vertical, and at the same time generally shows a periodic reversal of rotation, which is left-handed when the root is moving to the right (when looking at the plant from the shoot apex), and right-handed when the root is moving to the left (Okada and Shimura, 1990; Thompson and Holbrook, 2004). Buer et al. (2003) indicated that, under some conditions, the wavy growth pattern of roots occurs as the result of differential growth without cell file rotation, suggesting that differential growth primarily causes the wavy growth pattern of roots and that root tip rotations modulate root bending. However, these possibilities do not explain how tropic responses generate periodic wavy growth patterns, which remains an open question.

Molecular genetic analyses using Arabidopsis mutants with alterations in the wavy growth pattern have been performed previously to further elucidate the control of thigmomorphogenesis in roots. Defects in tropic responses, including gravitropism, phototropism and hydrotropism, result in straight root growth with fewer waves than expected on an inclined agar surface. Mutants exhibiting this phenomenon include wavy growth 1/phototropin 1 (wav1/phot1), wav5/aux1, wav6/pin2/agr1/eir1, long hypocotyl 5 (hy5) and mizu-kussei 1 (miz1), among others (Okada and Shimura, 1990; Bennett et al., 1996; Oyama et al., 1997; Chen et al., 1998; Luschnig et al., 1998; Müller et al., 1998; Mochizuki et al., 2005; Kobayashi et al., 2007). WAV1/PHOT1 is a blue light photoreceptor kinase, and is essential for root phototropism (Huala et al., 1997; Esmon et al., 2005); WAV5/AUX1 and WAV6/PIN2 are auxin transporters, both of which are essential for root gravitropism; HY5 is a bZIP transcription factor that is involved in root gravitropism; MIZ1 is a structural protein that is essential for root hydrotropism. It has also been reported previously that suppression of helical growth as a result of an abnormality in anisotropic cell expansion in the root hair defective3 (rhd3) mutant causes straight root growth (Yuen et al., 2005), and that enhancement of helical growth in the wav2 mutants causes a short-pitch pattern of wavy growth (Mochizuki et al., 2005).

Taken together, the cumulative evidence to date suggests that formation of the wavy growth pattern of Arabidopsis roots probably involves at least three processes: sensing of a touch stimulus, root bending as a result of differential growth, and modulation of root bending via root tip rotation. Tropic responses, most notably gravitropism, may be required for root tip impedance upon contact with the agar surface and the sensing of a touch stimulus, and also for the subsequent differential growth that forms the wavy growth pattern of roots. In addition, root tip rotations enhance root bending in response to various environmental stimuli, including touch, gravity, light and moisture, as described in our previous study of wav2 mutants (Mochizuki et al., 2005). Hence, genetic analyses of the wavy growth pattern of roots have proven to be very effective in enhancing our understanding of the mechanisms underlying differential growth and helical growth in roots.

The wav3 mutant was identified in a previous study from our laboratory as it showed a short-pitch pattern of wavy growth on inclined agar medium (Okada and Shimura, 1990). In the present report, we provide a detailed characterization of the phenotypes of the wav3 mutant and the functions of the WAV3 protein. Using map-based cloning analysis, we show that WAV3 encodes a protein belonging to the E3 ubiquitin ligase family, the members of which possess an N-terminal C3H2C3-type RING (RING-H2) domain (Stone et al., 2005) and a central von Willebrand factor type A (VWA) domain (Whittaker and Hynes, 2002), both of which are involved in protein–protein interactions. Using an in vitro ubiquitination assay, we confirmed the E3 ubiquitin ligase activity of WAV3. The Arabidopsis genome contains three WAV3 homologs, EMBRYO DEVELOPMENT ARREST 40 (EDA40) (Pagnussat et al., 2005), WAVH1 and WAVH2, and the present genetic analyses indicate that this RING-H2-type E3 ligase family plays a significant role in root gravitropism.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental Procedures
  7. Acknowledgements
  8. References
  9. Accession numbers:
  10. Supporting Information

Characterization of the wav3-1 mutant

The wav3-1 Arabidopsis mutant shows a short-pitch wavy growth pattern in its roots on an inclined agar surface (Figure 1a), as described previously (Okada and Shimura, 1990). We measured the wave tangent angle, half-wave length and growth rate of wild-type (Landsberg erecta, Ler) and wav3-1 roots to further characterize the phenotype of this mutant (Figure 1b–e). The roots of the wav3-1 mutant had shorter half-wave lengths and larger wave tangent angles than those of wild-type seedlings. These results indicate that the wav3 roots show a short-pitch wavy growth pattern, and that greater curvatures of root bending are one of the underlying causes. However, the growth rates of the mutant roots were found to be similar to those of wild-type, indicating that the mutation at the WAV3 locus causes an abnormality in root bending in response to touch stimuli at the root tip, but does not cause severe root growth defects.

image

Figure 1.  Comparison of root wave growth in Arabidopsis wild-type (Ler and Col) and wav3 mutants. (a) Wavy growth phenotypes in the roots of 6-day-old seedlings of wild-type, wav3, wav2 and wav2 wav3 double mutants on inclined agar medium. Scale bar = 3 mm. (b) Schematic view of a waving root showing the wave tangent angle (θ°) and half-wave length (double arrow). (c–e) Quantification of the root wave growth pattern using 6.5-day-old seedlings. Values are means ± SE for half-wave lengths (c), wave tangent angles (d) and growth rates (e) of roots from Ler, wav3-1, wav3-1 expressing wild-type WAV3 (WAV3 in wav3-1), Col and wav3-2 seedlings: 424–511 measurements were taken in each case for (c) and (d), and 91–101 measurements each in (e). The growth rate of the roots was calculated from their elongation lengths during a 3-day period after the agar surface had been inclined.

Download figure to PowerPoint

Although WAV2 is involved in the regulation of stimulus-induced root bending through inhibition of root tip rotation (Mochizuki et al., 2005), the cell file rotations of wav3 roots showing the wavy growth pattern resemble those of wild-type seedlings (data not shown), and the roots of the wav2-1 wav3-1 double mutant show a much shorter-pitch wavy pattern than either wav2-1 or wav3-1 (Figure 1a). Hence, the short-pitch wavy growth pattern of wav3 roots appears to be caused by mechanisms other than an abnormality of root tip rotation.

We next examined whether the roots of the wav3 mutant also show abnormal bending in response to gravity and light. When we examined the gravitropic response by inverting the agar medium, the wav3-1 roots showed a sharper degree of bending and a smaller radius of curvature than the wild-type roots (Figure 2a). When gravi-stimulation was provided to 7-day-old seedlings by rotating the agar medium by 90°, the root curvature in the Ler wild-type plants that had been turned left was greater than that of that plants that were turned right (Figure 2b), due to slanting of the roots on the agar medium to the right. The wav3-1 roots showed greater curvatures than the wild-type roots when plants were turned in either direction, although the values for the wav3-1 mutant and Ler roots for plants turned to the right did not differ significantly (> 0.05; Figure 2b). When we examined the phototropic response caused by unilateral irradiation with blue light at various fluence rates, the wav3-1 roots showed greater curvatures of root bending than wild-type, particularly at high fluence rates (10 or 100 μmol m−2 sec−1; Figure 2c). Previous studies have demonstrated that this mutant shows an enhanced hydrotropic response in roots (Takahashi et al., 2002). Hence, the wav3 roots show enhanced bending in response to touch, gravity and moisture, and suppressive bending in response to light. These results suggest that the WAV3 locus is involved in the mechanisms that regulate changes in the root growth direction in response to various environmental stimuli.

image

Figure 2.  Root gravitropism and phototropism in wild-type and wav3 mutant seedlings. (a) Root gravitropic response when the agar plate had been shifted to the inverted position. The dishes were positioned vertically for the first 4 days and were inverted for the following 2 days. Scale bar = 4 mm. (b) Root gravitropic responses when the agar plates were shifted to the side positions. The dishes were positioned vertically for the first 7 days under a 16 h light/8 h dark cycle, and were switched to the side position for the next 24 h in darkness. The root curvatures (θ°) of Ler, Col, wav3-1 and wav3-2 were then measured (93–111 measurements each). Values are means ± SE of the angle (in degrees). Asterisks indicate significant differences from the corresponding wild-type data (< 0.05, Student’s t test). (c) Root phototropic response. The angles of the root growth direction against gravity (−θ°) of 3-day-old seedlings were measured after 48 h exposure to unilateral blue light at the indicated fluence rates (111–170 measurements each). Light was supplied from either the right or the left side of the agar plate. Values are means ± SE.

Download figure to PowerPoint

Cloning of WAV3

The WAV3 gene was identified using a map-based cloning strategy (Figure 3a), as described in Experimental procedures. wav3-1 mutants transformed by genomic DNA that contained At5g49665 showed almost normal wavy root growth (Figure 1c,d). A T-DNA insertion line within the WAV3 locus of the Col ecotype (SALK_117706/wav3-2), obtained from the Arabidopsis Biological Resource Center (Alonso et al., 2003), also showed a short-wave pitch, enhanced gravitropic responses in both directions, and suppressive phototropic responses in the roots that were similar to those of wav3-1 (Figures 1 and 2). We therefore conclude that At5g49665 is the WAV3 gene.

image

Figure 3.  Map position of the WAV3 gene and structure of the proteins within the WAV3 family. (a) Map-based cloning of WAV3. The mutated locus in wav3-1 was mapped between the RPS4-CT and LFY3 markers on chromosome 5. Black boxes indicate WAV3 and the predicted genes around it. The double arrow shows the region covered by a restriction DNA fragment isolated from the TAC clone K2I5, which was cloned into the pPZP211 binary vector for use in the complementation assay. (b) The genome structure of WAV3. Black rectangles indicate protein-coding regions and white rectangles indicate non-coding regions. Triangles indicate the site of the point mutation in wav3-1 and the T-DNA insertion site in wav3-2. (c) Structures of RING finger proteins with the VWA domain. The number of amino acids in the deduced protein sequences for each gene and the percentage amino acid identity between each sequence and WAV3 are indicated in parentheses. White boxes and shaded ellipses indicate RING finger and VWA domains, respectively. Stars in the VWA domains indicate the MIDAS motif. Shaded boxes indicate conserved regions I, II and III. We defined the upper four genes as the WAV3 family. (d) Multiple sequence alignment of the RING-H2 motifs of the WAV3 family of proteins. The consensus sequence of the RING-C3H2C3-type motif (RING-H2) is indicated.

Download figure to PowerPoint

The WAV3 coding sequence was determined from its full-length cDNA, which was obtained from the RIKEN Bioresource Center (Seki et al., 2002). The start codon of WAV3 was confirmed by RT-PCR, which showed that there are no other start codons in the 5′ untranslated region of this gene. WAV3 contains two exons and encodes a predicted protein of 740 amino acid residues with a molecular mass of 80 kDa (Figure 3b). The wav3-1 gene possesses a nonsense mutation at residue 5 [TGG (Trp) to TGA (stop)]. Using SMART software (Schultz et al., 1998), we predicted that the WAV3 protein possesses a RING finger domain at its N-terminus and a von Willebrand factor type A (VWA) domain in the central region of the protein (Figure 3c). Functional analysis of the RING finger families of Arabidopsis found that eight protein members harbor an N-terminal RING-H2 domain and a central VWA domain (Figure 3c,d) (Whittaker and Hynes, 2002; Stone et al., 2005). A BLAST search further revealed many homologous genes in other plant species, including rice (Oryza sativa), grape (Vitis vinifera) and caster bean (Ricinus communis), but not in non-plant species (data not shown), suggesting that the combination of the RING and VWA domains represents a plant-specific domain architecture (Whittaker and Hynes, 2002). Both domains are involved in protein–protein interactions, and the WAV3 protein thus appears to function as an adapter protein.

WAV3 was also found to be part of a further sub-family containing the genes At4g37890/EMBRYO SAC DEVELOPMENT ARREST 40 (EDA40), At2g22680 (named WAVH1) and At5g65683 (named WAVH2). We defined these four genes as the WAV3 family (Figure 3c). A previous study has reported that EDA40 is a candidate gene for polar nuclei fusion during Arabidopsis embryogenesis, although its molecular function has not yet been determined (Pagnussat et al., 2005). The WAV3 sub-family can be distinguished from other family members in two ways (Figure 3c). First, the metal ion-dependent adhesion site motif (MIDAS motif: DxSxS…T4…D5) that binds to metal ions (Whittaker and Hynes, 2002) is not conserved in the VWA domains of the WAV3 sub-family. Second, only WAV3 sub-family members show three conserved C-terminal motifs (I, II and III) (Figure 3c).

Expression patterns of the WAV3 gene

RNA gel-blot analysis revealed stronger expression of WAV3 in seedlings and the roots of adult plants compared with the stems, leaves and flowers (Figure 4a). This result is consistent with the function of WAV3 in roots. We next performed immunoblotting analysis and found that the expression levels and molecular size of WAV3 in seedlings showing the wavy growth pattern on inclined agar medium do not differ from those in seedlings grown on vertical agar medium (Figure 4b). Hence, the total amounts of WAV3 and the extent of its protein modifications are unchanged in whole seedlings following induction of wavy growth in the roots. WAV3 was not detectable by immunoblotting in either the wav3-1 or wav3-2 mutants, indicating that they are null alleles.

image

Figure 4.  Analysis of the expression pattern of WAV3. (a) Tissue specificity of WAV3 gene transcripts. RNA gel-blot hybridization was performed using 32P-radiolabeled probes for WAV3 and 18S rRNA. (b) Immunoblotting for the WAV3 protein. Arabidopsis seedlings were grown on the surface of vertical agar medium (lane 1) or inclined agar medium (lanes 2–5). Total protein extracts of 7-day-old seedlings of wild-type Ler (lanes 1 and 2), wav3-1 (lane 3), wild-type Col (lane 4) and wav3-2 (lane 5) were analyzed: 10 μg of each protein preparation was separated by 8% SDS–PAGE and immunoblotted using anti-WAV3 antibodies.

Download figure to PowerPoint

E3 ubiquitin ligase activity of WAV3

Previous functional analyses of the RING-H2-type ubiquitin ligase family in Arabidopsis have shown that a member of the WAV3 family, At2g22680/WAVH1, shows E3 ubiquitin ligase activity (Stone et al., 2005). This suggests the possibility that WAV3 also functions as an E3 ligase, and we performed an in vitro ubiquitination assay to test this. We expressed the WAV3 N-terminal region harboring the RING-H2 domain in Escherichia coli as a fusion product with the glutathione S-transferase protein (GST). In the presence of E1, E2 and ubiquitin, purified GST–WAV3 showed ubiquitination activity, but control purified GST was inactive (Figure 5). These findings suggest that WAV3 has E3 ubiquitin ligase activity.

image

Figure 5.  Analysis of the E3 ligase activity of WAV3. A GST-tagged, RING-containing N-terminal domain of WAV3 is capable of mediating protein polyubiquitination in an E2-dependent manner (lane 1). GST–WAV3 and GST proteins were expressed and purified from E. coli and tested for E3 ubiquitin ligase activity with human E1 and UbcH5b E2. Omission of E2 (lane 2) or ubiquitin (lane 4) from the assay, or substitution of GST–WAV3 by GST (lane 3), resulted in loss of protein polyubiquitination. The reactions were analyzed by immunoblotting against a polyubiquitin antibody. The arrowhead indicates the position of the GST–WAV3 protein.

Download figure to PowerPoint

Function of the WAV3 gene family in root gravitropism

To further elucidate the functions of the WAV3 gene family in regulation of the root growth patterns of plants, we analyzed the phenotypes of T-DNA-tagged mutants of the WAV3 family of genes. The SALK T-DNA-tagged lines wavh1-1, wavh1-2, wavh2-1 and wavh2-2 (Col ecotype; Alonso et al., 2003) were obtained from the Arabidopsis Biological Resource Center (Figure 6a), but we could not obtain a null mutant of the EDA40 gene. RT-PCR analyses revealed that transcripts of WAVH1 or WAVH2 were absent from their respective homozygous mutants (Figure 6b). We next generated multiple mutants via crosses between wav3, wavh1 and wavh2, and observed their root growth patterns, including wavy growth on an inclined agar surface, and their gravitropic and phototropic responses (Figure 6c–e). We found no root growth pattern defects in the wavh1 and wavh2 single mutants. The phenotypes of the wav3-2 wavh1-1 double mutant resembled those of the wav3-2 single mutant, which shows a short-pitch wavy growth pattern on an inclined agar surface, a slight enhancement of the gravitropic response (mean angles of 80.0 ± 1.8, 93.7 ± 1.8 and 89.9 ± 1.8° in wild-type, wav3-2 and wav3-2 wavh1-1, respectively), and a slight decrease in root bending in response to unilateral light irradiation (Figure 6c–e). However, the roots of the wav3-2 wavh2-1 double mutants showed a loss of wavy growth patterns on inclined agar medium and an abnormal gravitropic response (Figure 6c,d). The roots of the wavh1-1 wavh2-1 double mutants showed a slight decrease in the gravitropic response (mean angle of 69.5 ± 1.8°) and a slight enhancement of the phototropic response (Figure 6c,d). Further, wavy growth and the gravitropic response were absent and the phototropic response was remarkably enhanced in roots of the wav3-2 wavh1-1 wavh2-1 triple mutants (Figure 6c–e). This root gravitropism defect was confirmed in other triple mutants for the alleles wav3-1, wavh1-2 and wavh2-2 (Figure 6d). These results indicate that the WAV3 gene family plays an essential role in root gravitropism in Arabidopsis.

image

Figure 6.  Genetic analysis of WAV3 gene family functions. (a) Genomic structures of the WAVH1 and WAVH2 genes. Black and white rectangles indicate protein-coding regions and non-coding regions, respectively. The start codon (ATG) of each gene was determined as the site downstream of a nonsense codon (TAA or TAG) in the 5′ region of the longest open reading frame within the cloned cDNA. Triangles indicate the location of the T-DNA insertion sites. (b) RT-PCR analysis. Total RNA isolated from 7-day-old seedlings of wild-type (Col) and mutants were used. The WAVH1 and WAVH2 genes are not expressed in the mutants (arrowhead). The NPH3 gene was used as an internal control (asterisk). (c) Wavy growth phenotypes of the roots of 6-day-old seedlings of wild-type (Col), wavh1, wavh2 and multiple mutants growing on inclined agar medium. (d) Root gravitropism. Four-day-old seedlings grown under continuous white light were rotated 90° (g) and kept under white light conditions for a further 24 h. The frequencies (%) of root growth direction at intervals of 15° are represented by the lengths of the bars (74–94 measurements each). (e) Root phototropic response. The angles of the root growth direction against gravity (−θ°) in 3-day-old seedlings were measured after 48 h exposure to unilateral blue light at 100 μmol m−2 sec−1 (54–70 measurements each). Values are means ± SE. The asterisks indicate a significant differences from wild-type (< 0.05, Student’s t test).

Download figure to PowerPoint

To elucidate the tissue-specific functions of the WAV3 family members, we analyzed their expression patterns in transgenic seedlings using the GUS reporter system (Figure 7). Consistent with the effects of mutations of wav3, wavh1 and wavh2 on root gravitropism, GUS staining was observed in the root tips of transgenic seedlings carrying WAV3p::WAV3-GUS, WAVH1p::WAVH1-GUS and WAVH2p::WAVH2-GUS (Figure 7a–f). Expression of the WAV3–GUS gene driven by its own promoter was found in the root tip excluding the root cap, most notably in the root meristem region (Figure 7d), and in the leaf primordia (Figure 7g). The expression levels of WAVH1–GUS were found to be lower than those of WAV3–GUS and WAVH2–GUS in all transgenic lines examined (31 lines transformed with WAVH1p::WAVH1-GUS). The expression pattern was similar to that of WAV3–GUS, with weak expression in the root meristem region and the leaf primordia (Figure 7e,h). WAVH2–GUS was expressed in the root tip, including the root cap and the root elongation zone, rather than the root meristem region (Figure 7f), and also in the whole aerial part of the seedling, including the cotyledon, leaf primordia and hypocotyl (Figure 7i). These GUS staining results indicate that WAVH1 shares functions in similar tissues and WAVH2 shares functions in different tissues. These findings suggest the possibility that the WAV3 sub-family also plays a role in the aerial parts of seedlings. However, we did not detect any abnormalities in the aerial part of the wav3 wavh1 wavh2 triple mutants, and the role of the WAV3 sub-family in the aerial parts of seedlings remains to be further investigated.

image

Figure 7.  GUS staining patterns in 3-day-old transgenic seedlings. GUS activity (indicated by blue staining) was observed in transgenic plants carrying WAV3p::WAV3-GUS (a,d,g), WAVH1p::WAVH1-GUS (b, e, h) or WAVH2p::WAVH2-GUS (c,f,i). (a–c) Whole seedlings. Scale bars = 2 mm. (d–f) Root tips. Scale bars = 200 μm. (g–i) Shoots. Scale bars = 200 μm.

Download figure to PowerPoint

Effects of wav3 wavh1 wavh2 triple mutations on DR5rev::GFP expression and root elongation

Auxin is known to regulate gravitropism (Philosoph-Hadas et al., 2005). To analyze the auxin responses in the wav3 wavh1 wavh2 triple mutants showing obvious abnormalities of root gravitropism, we first observed the expression patterns of the auxin reporter gene DR5rev::GFP in the triple mutants (Friml et al., 2003). As shown in Figure 8(a), DR5rev::GFP was highly expressed in the root tip of wild-type seedlings, including the quiescent center, columella, lateral root cap and stele. The DR5rev::GFP expression pattern in the wav3-2 wavh1-1 wavh2-1 triple mutants was similar to that in the wild-type, but with lower levels in the stele and higher levels in the lateral root cap (Figure 8a). When we observed the expression patterns under gravi-stimulation for 6 h (sufficient time to promote a visible curvature response in wild-type seedlings), the GFP signal extended into the lower flank of the lateral root cap (the concave side of the curve in roots) in 17 wild-type seedlings of the 20 tested (Figure 8b, upper left panel). This finding indicated that a gravi-stimulus induces unilateral auxin transport in the wild-type. However, the wav3-2 wavh1-1 wavh2-1 seedlings showed irregular root growth directions and reporter gene expression patterns after 6 h of gravi-stimulation. For example, upper right panel in Figure 8(b) shows normal curvature of the root towards the new gravity direction, but the GFP signal extends into the upper flank of the lateral root cap. The lower left panel in Figure 8(b) shows enhanced curvature of the root and asymmetric GFP staining along the lower convex side. The lower right panel shows an upward growing root with GFP signals in both sides of the lateral root cap. When we assessed the effect of IAA on root elongation, the wav3-2 wavh1-1 wavh2-1 triple mutants were found to be slightly resistant to exogenous auxin at 0.1 μm in a root elongation assay in comparison with wild-type seedlings (Figure 8c). Our results thus indicate that the wav3 wavh1 wavh2 triple mutants show some abnormalities in the auxin signaling pathway in roots.

image

Figure 8.  Auxin responses in wild-type seedlings and wav3 wavh1 wavh2 triple mutants. (a) Typical fluorescence images of DR5rev::GFP expression patterns in the roots of 3-day-old seedlings prior to gravi-stimulation. The left and right panels show GFP signals from a wild-type Col and wav3-2 wavh1-1 wavh2-2 mutant seedling, respectively. The direction of gravity (g) is indicated by an arrow. Extension of the GFP signal into the flank of the root tip is indicated by arrowheads. Scale bars = 50 μm. (b) Typical fluorescence images of DR5rev::GFP expression patterns in the roots of 3-day-old seedlings after 6 h of gravi-stimulation. Extension of the GFP signal into the flank of the root tip is indicated by arrowheads. Scale bars = 50 μm. (c) Effects of IAA on root elongation. Plants were grown vertically for 3 days and then transferred to media with and without IAA. The root elongation of wild-type and wav3-2 wavh1-1 wavh2-1 triple mutants was measured after an additional 2 days. The rates of root growth inhibition relative to the controls (0 μm IAA) are shown. Values are means ± SE (= 74–96).

Download figure to PowerPoint

Starch-filled amyloplasts in the central columella cells of the root cap are necessary for a normal gravitropic response in roots (Kiss et al., 1989). When we observed staining patterns using Lugol’s solution, which can visualize starch-filled amyloplasts, the wav3-2 wavh1-1 wavh2-1 and wild-type roots were very similar (Figure S1). Therefore, no abnormality of amyloplast formation was detected in the wav3 wavh1 wavh2 triple mutants.

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental Procedures
  7. Acknowledgements
  8. References
  9. Accession numbers:
  10. Supporting Information

The results of the present genetic study indicate that the WAV3 gene family is essential for the normal gravitropic response of roots, and that at least three members of this family, WAV3, WAVH1 and WAVH2, positively function in that response in a redundant manner. This was evident from our findings of negative gravitropic responses in the roots of the corresponding single and double mutant seedlings, but not in the wav3 wavh1 wavh2 triple mutants (Figure 6c). The wavh1 wavh2 mutant shows a moderate gravitropic response, indicating that the functions of WAVH1 and WAVH2 are almost compensated for by those of WAV3, which therefore plays a positive role in root gravitropism. The wav3 wavh2 mutant shows a slight root gravitropism and wavy growth phenotype, suggesting that WAVH1 induces a weak gravitropic response in the absence of WAV3 and WAVH2. On the other hand, the wav3 and wav3 wavh1 mutants show a short-pitch wavy growth pattern on inclined agar surfaces and also show marginally enhanced gravitropic responses but suppressed phototropic responses. These results suggest that WAVH2 may induce a strong gravitropic response in roots in the absence of WAV3, and that this is why the wav3 single mutants shows an enhanced gravitropic response in their roots. Hence, WAV3, WAVH1 and WAVH2 induce moderate, weak and strong gravitropic responses, respectively, and the function of WAV3 appears to be epistatic to that of WAVH2 in root gravitropism. When we examined the possibility of genetic interactions among these genes at the transcript level, quantitative RT-PCR indicated that each gene mutation had practically no effect on the expression levels of other genes (Figure S2). Thus their genetic interactions are probably caused by some reason other than transcriptional regulation.

The enhanced gravitropic response phenotype in wav3 roots is consistent with the short-pitch wavy growth pattern and the decreased phototropic responses of this mutant. The enhancement of the gravitropic response strengthens root tip impedance upon contact with the agar surface, and this strengthened impedance may enhance root bending in response to touch in wav3 plants. There is also a possibility that the enhancement of the differential growth in the gravitropic response generates the increased wave tangent angles that are characteristic of the wavy growth pattern in roots. Furthermore, because the gravitropic and phototropic responses of roots can counter each other if the growth directions in response to these stimuli differ (Okada and Shimura, 1992, 1994), it is to be expected that an enhanced gravitropic response will cause suppression of the phototropic response in wav3 roots. In this context, suppression of the gravitropic response appears to enhance the phototropic responses in the roots of the wavh1 wavh2 and wav3 wavh1 wavh2 mutants (Figure 6).

However, there are some notable inconsistencies in the phenotypes of wav3 and its related mutants. The relationship between hydrotropism and gravitropism is similar to that between phototropism and gravitropism in roots, and gravitropism-defective mutants often show enhancement of the hydrotropic response, and the hydrotropism mutant nhr1 shows an enhanced gravitropic response (Takahashi et al., 2002; Eapen et al., 2003). These observations suggest that the hydrotropic response will be suppressed in the wav3 mutant because of the enhancement of the gravitropic response in wav3 roots. However, Takahashi et al. (2002) have reported that wav3-1 shows an enhancement of positive hydrotropism in its roots. Our observations also indicate that a defective gravitropic response does not cause an enhanced phototropic response in the wav3 wavh2 mutants (Figure 6d,e). These results therefore suggest that the WAV3 gene family may be involved in signaling pathways that regulate various tropic responses in roots, including gravitropism, phototropism and hydrotropism, and does not simply function in root gravitropism.

Our current results have indicated that the wav3 wavh1 wavh2 triple mutants show some abnormalities in the auxin signaling pathway in roots, and that WAV3 also shows E3 ubiquitin ligase activity, as reported previously for WAVH1 (Stone et al., 2005). These results suggest that the WAV3 family controls root growth patterns, particularly the gravitropic response, via mechanisms involved in auxin transport and/or signaling and the ubiquitin/proteasome pathways. The control of protein turnover of the auxin efflux carrier PIN2 through the ubiquitin/proteasome pathways is significant for the gravitropic response of roots, and stabilization of PIN2 in the wav6-52 mutant causes abnormalities in auxin distribution and gravitropism in Arabidopsis roots (Abas et al., 2006). It would be worthwhile investigating the expression pattern of the PIN2 proteins in the wav3 wavh1 wavh2 triple mutants. Using yeast two-hybrid screens, we identified several WAV3-binding proteins, including all the members of another E3 ligase family of SINA-type RING finger proteins, SINAT1–5 (Figure S3). The human BRCA1–BARD1 and the yeast Hex3–Slx8 ubiquitin ligases function by forming a heterodimer of two RING proteins (Hashizume et al., 2001; Xie et al., 2007), and our result raises the possibility that the WAV3 family of proteins functions in conjunction with SINA family proteins by forming heterodimers. Another possibility is that WAV3 is a target protein that is ubiquitinated and degraded by the SINA family proteins. SINAT5 is an auxin signaling factor that is involved in lateral root formation (Xie et al., 2002), and the WAV3 family may play a role in auxin signaling pathways involving the SINA family. Future analyses of the molecular function of the WAV3 family that explore these possibilities will be important to enhance our understanding of the regulation of root growth patterns in response to various environmental stimuli, including gravity, light and water.

Experimental Procedures

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental Procedures
  7. Acknowledgements
  8. References
  9. Accession numbers:
  10. Supporting Information

Plant growth conditions and microscopy

Arabidopsis thaliana seeds were sterilized in 10% v/v bleach (Haiter; Kao, http://www.kao.com) containing 0.02% v/v Triton X-100 for 10 min. After five washes in sterilized water, the seeds were germinated and grown on square Petri dishes containing half-strength Okada and Shimura (OS) medium (Okada and Shimura, 1990) supplemented with 1.5% w/v agar. After sowing, the dishes were kept in darkness at 4°C for 3–4 days, and then transferred to a growth chamber at 22°C under constant white light (white fluorescent lamp, model FL20S-EXNH; Toshiba, http://www.tlt.co.jp/tlt/index_e.htm). For observations of the wavy growth patterns of the roots, the Petri dishes were positioned vertically for the first 2.5 days of seedling growth, and then inclined backwards at a 45° angle for 3–5 days under white light conditions (3–5 μmol m−2 sec−1). Light-emitting diode (LED) blue light lamps with a maximum wavelength of 470 nm and a 30 nm half bandwidth (LED-mB; Eyela, http://www.eyelaworld.com) were used for the analysis of root phototropism. Seedling images were obtained using either a CCD color camera (VB-7010; Keyence, http://www.keyence.com/), digital microscope (VH8000; Keyence) or digital scanner (ES8500; Epson, http://www.epson.com/). Root wave growth patterns and root curvatures were quantified using the software supplied for the VH8000 digital microscope and also using Canvas 9 software (ACD Systems International Inc., http://www.acdsee.com/).

Map-based cloning of WAV3

Wav3-1 (Ler) was crossed with Col wild-type Arabidopsis, and F2 seedlings were selected by screening for short-pitched wavy root growth on agar plates inclined at 45° to the vertical axis. Genomic DNA from individual F2 seedlings was isolated and amplified using primers specific for various PCR-based cleaved amplified polymorphisms, including RPS4-CT and LFY3 (http://www.arabidopsis.org), and also using simple sequence-length polymorphism markers, which were designed from SNPs stored in the Cereon Arabidopsis Polymorphism Collection (Jander et al., 2002). The seedlings were then assayed for Ler- and Col-specific polymorphisms. The WAV3 locus was mapped to an approximately 133 kb region of chromosome 5 between the Arabidopsis Genome Initiative loci At5g49530 and At5g49780. We sequenced all of the putative genes annotated by the Arabidopsis Information Resource (http://www.arabidopsis.org) in this region of the wav3-1 mutant and compared them against wild-type. We subsequently found only one base change (in At5g49665).

For complementation analysis, a SalI/SacI-digested 6.2 kb fragment of the transformation-competent artificial chromosome K2I5 clone, which includes the At5g49665/WAV3 gene, was cloned into the binary vector pPZP211 (Hajdukiewicz et al., 1994). K2I5 was obtained from the Arabidopsis Biological Resource Center. The resulting plasmid, pPZP-WAV3SALISACI, was used for transformation of Agrobacterium tumefaciens strain ASE, which was used for floral dipping of the wav3-1 plants (Clough and Bent, 1998). Transgenic plants were selected on half-strength OS medium containing 50 μg ml−1 kanamycin. The wavy phenotypes of the roots of the T3 plants were then examined. Several transgenic lines carrying pPZP-WAV3SALISACI showed complementation of the wav3 mutation. The wav3-2 mutant, carrying a T-DNA insertion in At5g49665 in the Col ecotype, was obtained from the Arabidopsis Biological Resource Center. The WAV3 full-length cDNA was obtained from the RIKEN Bioresource Center (http://www.brc.riken.go.jp/lab/epd/Eng/) (Seki et al., 2002). We used SMART software (http://coot.embl-heidelberg.de/SMART) (Schultz et al., 1998) to analyze the structure of the WAV3 protein.

RNA gel-blot analysis

Total RNA was isolated from the roots, stems, rosette leaves and flowers of adult plants and from 5-day-old Ler-seedlings using the RNeasy plant mini kit (Qiagen, http://www.qiagen.com/). Total RNA (30 μg) was loaded into each lane of a denaturing agarose gel, electrophoresed and transferred to a Biodyne B nylon membrane (Pall, http://www.pall.com/). The membrane was then hybridized with a 32P-radioactive random-primed DNA probe prepared using WAV3 cDNA as a template. To normalize for the amount of RNA loaded in each lane, we hybridized the filter to a DNA probe corresponding to 18S rRNA. Signal detection was performed as described previously (Mochizuki et al., 2005).

In vitro ubiquitination assay

To perform an in vitro ubiquitination assay, a WAV3N2 fragment harboring the RING-H2 domain of the WAV3 protein (Figure S3) was prepared by PCR using 5′ and 3′ primers 5′-GAATTCAGTAATCCTTCAACTCCCCGAT-3′ and 5′-GTCGACTTCAGGTATAGTGACAAACC-3′, respectively, incorporating EcoRI and SalI sites, and then inserted into pGEX6T-1 (GE Healthcare, http://www.gehealthcare.com/). This construct and a control vector without WAV3N2 were then introduced into E. coli strain BL21(DE3)pLysS (Novagen, http://www.merck-chemicals.com), and GST-fused WAV3 proteins or control GST were prepared in accordance with the manufacturer’s instructions. Ubiquitination assays were performed using a ubiquitination kit (BIOMOL, http://www.enzolifesciences.com). Briefly, each reaction in a 30 μl final volume contained ubiquitination buffer, 2 mm Mg2+-ATP, 1 mm DTT, 0.6 U inorganic pyrophosphatase, 40 μm ZnSO4, 0.1 nm human recombinant E1, 1.5 μg E2 UbcH5b, 2 μg purified GST–WAV3 or GST, and 2.5 μg ubiquitin. The reactions were incubated at 30°C for 3 h and stopped by addition of SDS–PAGE sample buffer, followed by 10 min incubation at 65°C. Samples were then separated by SDS–PAGE followed by protein gel blotting and immunodetection using FK2 polyubiquitin antibody (MBL, http://www.mblintl.com).

Immunoblotting

Anti-WAV3 antiserum was generated in rabbit using GST-fused WAV3N2 protein as an antigen. Anti-WAV3 antibodies were purified from the resulting antiserum using antigen-blotted polyvinylidene difluoride Hybond-P membranes (GE Healthcare). Immunoblotting was then performed as described previously (Inada et al., 2004).

RT-PCR

Each total RNA preparation was pre-treated with DNase I, and cDNAs were synthesized by reverse transcriptase using a gene-specific primer, in addition to an NPH3 gene primer as an internal control, as described previously (Sakai et al., 2008). PCR was performed using primers 5′-CAACACACGAGTCACGAGG-3′ and 5′-CTAAAATCTAGCGTTTTCGAAACC-3′ for WAVH1, 5′-ATGGTGTTTGGTTGGAGAAAAGC-3′ and 5′-TTAAAATCTGGCGTTTTCCAAGCC-3′ for WAVH2, and 5′-GAAGATGTCTCCATCTTAAGAATCG-3′ and 5′-CCTTGAGATATGAATCAATGGCTC-3′ for NPH3. Equal volumes of PCR products were then separated in 1% w/v agarose gels containing 0.5 mg L−1 ethidium bromide. Gel images were then captured with the FAS-III CCD camera (Nippon Genetics, http://www.nippongenetics.eu).

GUS staining analysis

A blunt-end GATEWAY cassette, reading frame B (RFB) (Invitrogen, http://www.invitrogen.com/), was inserted into the SmaI site of the pBI101.1 binary vector (Clontech, http://www.clontech.com/) to obtain pBI-GW-GUS. To construct the plasmids WAV3p::WAV3-GUS, WAVH1p::WAVH1-GUS and WAVH2p::WAVH2-GUS, in which fusion proteins are expressed under the control of the native promoter, WAV3, WAVH1 or WAVH2 genomic DNA covering the nucleotide regions −2614 to +2299, −2128 to +2049 and −2171 to +2230, respectively (numbers indicate the distances from the ATG initiation codon), were amplified from Arabidopsis genomic DNA by PCR, and then sub-cloned into the RFB sites of pBI-GW-GUS using GATEWAY recombination (Invitrogen). These plasmids were then used to transform Agrobacterium tumefaciens strain C58C1, which was then used for subsequent transformations of wild-type (Col) plants. For GUS expression analysis, 3-day-old T2 transgenic seedlings were immersed in X-Gluc solution containing 0.5 mg ml−1 X-Gluc (5-bromo-4-chloro-3-indolyl-β-glucuronide), 0.5 mm potassium ferricyanide, 0.5 mm potassium ferrocyanide, 100 mm NaPO4 (pH 7.0), 10 mm EDTA and 0.05% Triton X-100, and incubated at 37°C for 24 h.

DR5rev::GFP images

A transgenic line harboring the DR5rev::GFP transgene (Col background) was obtained from the Arabidopsis Biological Resource Center and crossed with the wav3-2 wavh1-1 wavh2-1 triple mutant. For gravi-stimulation, plants were grown vertically for 3 days, and then transferred into half-strength OS agar medium to adjust the root growth direction towards gravity. Then, seedlings were rotated 90° and kept for a further 6 h. GFP fluorescence was visualized using a confocal laser-scanning microscope (TCS SP5; Leica, http://www.leica-microsystems.com).

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental Procedures
  7. Acknowledgements
  8. References
  9. Accession numbers:
  10. Supporting Information

We thank the Arabidopsis Biological Resource Center for providing the K2I5 TAC clone, the MNI5 P1 clone, the SALK T-DNA insertion mutants (wav3-2, wavh1-1, wavh1-2, wavh2-1 and wavh2-2) and a transgenic line harboring the DR5rev::GFP gene, and RIKEN Bioresource Center for providing the WAV3 cDNA. We are grateful to Drs Marco Trujillo (RIKEN Plant Science Center, Yokohama) and Noriyuki Matsuda (RIKEN Genome Science Center, Yokohama) for technical advice on the in vitro ubiquitination assay, and Professor Toshiaki Mitsui (Niigata University, Niigata) for providing the Bio-Rad CFX96 facility. This work was supported by a Grant-in-Aid for Scientific Research on Innovative Areas ‘Plant Environmental Sensing’ (number 23120510) from the Ministry of Education, Culture, Sports, Science and Technology, Japan, to T.S.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental Procedures
  7. Acknowledgements
  8. References
  9. Accession numbers:
  10. Supporting Information

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental Procedures
  7. Acknowledgements
  8. References
  9. Accession numbers:
  10. Supporting Information

Figure S1. Lugol staining of root tips. Starch granules were visualized by Lugol staining. The left panel shows a root tip of a wild-type seedling and the right panel shows a root tip from the wav3-2 wavh1-1 wavh2-1 triple mutant. At least 20 seedlings for each genotype were stained and representative patterns are shown. Bars, 50 μm.

Figure S2. Quantitative real-time RT-PCR analyses of the WAV3 gene family. The graphs show the relative abundance of each transcript in the roots of 5-day-old seedlings. The gene expression levels were normalized with respect to the ACTIN 2 gene, used as an internal control. The values were calculated relative to those of wild type (Col). Data and error bars represent the mean ± SD from the average values of three independent experiments. (a) WAV3 gene expression in wild type, and in wavh1-1 and wavh2-1 mutants. (b) WAVH1 gene expression in wild type, and in wav3-2 and wavh2-1 mutants. (c) WAVH2 gene expression in wild type, and in wav3-2 and wavh1-1 mutants. (d) NPH3 gene expression in wild type, and in wav3-2, wavh1-1 and wavh2-1 mutants. NPH3 gene expression was confirmed to be constant in wild type and mutant seedlings. As well as NPH3, expression levels of WAV1, WAVH1 and WAVH2 transcripts were not significantly different between wild type and each mutant (> 0.05).

Figure S3. Identification of genes encoding proteins that interact with WAV3. (a) Schematic representation of the truncated WAV3 sequences used as bait in both two-hybrid screens and in vitro pulldown assays. The RING-finger domain, the VWA domain and conserved regions I, II, and III of WAV3 are represented by a white box, a shaded ellipse and a shaded box, respectively. The amino acid residues used in each construct are indicated in parentheses. (b) Yeast two-hybrid assay of WAV3-SINAT, WAV3-SPIRAL 2 (SPR2), and WAV3-AT4G17060 interactions. Protein–protein interactions between WAV3N1 and SINAT1 (yeast cell line number 1), WAV3N1 and SINAT4 (number 3), WAV3N1 and SINAT3 (number 5), WAV3N1 and SINAT2 (number 7), WAV3N1 and SINAT5 (number 9), WAV3C and SPR2 (number 12), and WAV3C and AT4G17060 induced the expression of the HIS3 and ADE2 reporter genes. These yeast cell lines had been grown on synthetic medium without the amino acids His and Ade. Combinations using the pGB vector without bait (segments 2, 4, 6, 8, 10, 13, and 15) or with the pGA vector without a target (segments 11 and 16) failed to induce His3 and Ade2 reporter gene expression. (c) In vitro pulldown experiments demonstrating specific interactions between GST-fused WAV3N2 and the HA-tagged SINAT proteins, MBP-fused WAV3C and HA-tagged SPR2, and MBP-fused WAV3C and HA-tagged AT4G17060 proteins. The amino acid residues encoded by clones isolated from the two-hybrid assay and used in the pulldown assay are indicated in parentheses. HA-tagged proteins were resolved in 10% SDS-PAGE gels using in vitro translation products as input size controls. The interactions were detected by immunoblotting with HA antibodies. All of the members of the SINA protein family (SINAT1/At2g41980, SINAT2/At3g58040, SINAT3/At3g61790, SINAT4/At4g27880 and SINAT5/At5g53360: Welsch et al., 2007) harbor a C3HC4-type RING-finger domain at their N-termini. In addition, Xie et al. (2002) have previously reported that SINAT5 has E3 ubiquitin ligase activity. In our current analyses also, we isolated two genes encoding SPR2 (At4g27060: Buschmann et al., 2004; Shoji et al., 2004) and an unknown protein (At4g17060), both of which bind to the C-terminal region of WAV3.

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer-reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

FilenameFormatSizeDescription
TPJ_4870_sm_FigS1-S3.pdf2001KSupporting info item
TPJ_4870_sm_SupportingInfomationlegends.doc59KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.