Report

Secreted Peptide Signals Required for Maintenance of Root Stem Cell Niche in Arabidopsis

See allHide authors and affiliations

Science  27 Aug 2010:
Vol. 329, Issue 5995, pp. 1065-1067
DOI: 10.1126/science.1191132

Abstract

Stem cells are maintained in the niche by intercellular interactions and signaling networks. In this work, we study extracellular signals required for maintenance of the root stem cell niche in higher plants. We identify a family of functionally redundant homologous peptides that are secreted, tyrosine-sulfated, and expressed mainly in the stem cell area and the innermost layer of central columella cells. We name these peptides root meristem growth factors (RGFs). RGFs are required for maintenance of the root stem cell niche and transit amplifying cell proliferation in Arabidopsis. RGF1 defines expression levels and patterns of the stem cell transcription factor PLETHORA, mainly at the posttranscriptional level. The RGFs function independently of the auxin pathway. These peptide signals play a crucial role in postembryonic root development.

Secreted peptides are now recognized as important members of intercellular signals that coordinate and specify cellular functions in plants. Some of the secreted peptide hormones undergo complex posttranslational modifications that are mediated by specific enzymes that recognize particular sequences of multiple target peptides. Because such modifications are generally critical for the functions of individual peptide hormones, the presence of previously unidentified peptide hormones should be revealed through phenotypic analysis of the mutants of posttranslational modification enzymes.

Tyrosine sulfation is a posttranslational modification that has been found in several peptide hormones in both animals and plants (1). This modification is mediated by tyrosylprotein sulfotransferase (TPST), which catalyzes the transfer of sulfate to the phenolic group of tyrosine. Arabidopsis TPST (AtTPST) is a type I transmembrane protein localized in cis-Golgi (2). Because AtTPST is a single-copy gene, phenotypes of its loss-of-function mutant should reflect the deficiency in the biosynthesis of all the functional tyrosine-sulfated peptides found in Arabidopsis.

Of the pleiotropic phenotypes of the tpst-1 mutant, we focused on the severe short-root phenotype characterized by reduction in root meristem size and loss of coordination between cell elongation and expansion in the elongation-differentiation zone (Fig. 1, A to C) (3). Impaired expression of a cell-cycle marker CYCB1;1:GUS (4) indicates that root meristematic activity of tpst-1 is considerably decreased compared with that of the wild type (Fig. 1, D and E, and fig. S1A). We occasionally observed extra quiescent center (QC) cells in tpst-1, as indicated by the expression of WOX5, a QC-specific gene (Fig. 1, F and G, and fig. S1, B to D) (5). In addition, starch granules accumulate in all columella layers including cells at the position of the columella stem cells, indicating that tpst-1 fails to maintain root stem cells (Fig. 1, H and I, and table S1). In contrast, the auxin maximum in tpst-1 visualized by DR5:GUS marker (6) was comparable to that of the wild type (Fig. 1, J and K), suggesting that this short-root phenotype is not directly associated with auxin distribution. Embryogenesis is normal in tpst-1, suggesting that only postembryonic root development is affected by this mutation (fig. S1, E to H).

Fig. 1

Root meristem phenotype of tpst-1. (A) WT and tpst-1 seedlings at 7 days after germination (DAG). (B and C) Confocal images of root meristem of WT and tpst-1 seedlings at 5 DAG. White arrowheads indicate the root meristem boundary. (D and E) CYCB1;1:GUS expression at 5 DAG. (F and G) Whole-mount in situ hybridization of WOX5 mRNA. Asterisks denote WOX5-expressing cells at the QC position. (H and I) Starch granules stained by Lugol. Asterisks indicate the QC position counterstained by in situ hybridization with WOX5. Yellow arrowheads indicate the position of columella stem cells. Insets show close-up views. (J and K) DR5:GUS expression. Scale bars, (A) 5 mm; (B and C) 100 μm; (D to K) 50 μm.

To further understand the defects in tpst-1 plants, we added peptide hormones phytosulfokine (PSK) (7) and PSY1 (8), both of which are tyrosine sulfated and also expressed in roots. These peptides restored cell-elongation activity in the elongation-differentiation zone (figs. S2 and S3, A to D) but did not promote meristematic activity in tpst-1, as evidenced by meristem size (fig. S3, A to E), root length (fig. S3F), and CYCB1;1:GUS expression (fig. S3, G to I). Extra QC cells were still observed in tpst-1 in the presence of PSK and PSY1 (fig. S3, J to L). In addition, starch-granule staining showed no rescue of columella stem cells, indicating that tpst-1 fails to maintain root stem cells, even in the presence of PSK and PSY1 (fig. S3M and table S1).

On the basis of these observations, we speculated that an as-yet undiscovered tyrosine-sulfated peptide(s) regulates root meristematic activity and maintenance of the stem cell niche in Arabidopsis. To identify this peptide signal, we searched the Arabidopsis genome for genes likely to encode the sulfated peptide(s). Hormones PSK and PSY1 are encoded by multiple paralogous genes whose primary translated polypeptides are ~70 to 110 amino acids in length, are relatively poor in cysteine residues (Cys < 6), contain a secretion signal, and contain Asp-Tyr sequences that function as tyrosine-sulfation motifs (8, 9). On the basis of these criteria, we found a candidate polypeptide family with a C-terminal conserved domain containing Asp-Tyr sequences (fig. S4, A and B). We overexpressed one of these peptides, At5g60810, in Arabidopsis and used nano–liquid chromatography–tandem mass spectrometry to analyze the apoplastic peptides derived from the transgenic plants to determine the peptide’s mature structure (fig. S5A). We determined that mature peptide derived from At5g60810 is 13 amino acids long, carries a tyrosine sulfation, and derives from the C-terminal conserved domain of the encoded protein through proteolytic processing (Fig. 2, A and B, and figs. S4A and S5B).

Fig. 2

A tyrosine-sulfated peptide family that maintains root stem cell niche and regulates meristematic activity. (A) Sequence of the mature At5g60810 (RGF1) peptide. (B) Multiple sequence alignments of the conserved 13–amino acid domain of the At5g60810 (RGF1) peptide family. ++, strong acitivity; +, weak activity; –, no activity (see also fig. S9). (C) Whole-mount in situ hybridization of WOX5 mRNA in tpst-1 root meristem at 5 DAG cultured in the presence of At5g60810 (RGF1) peptide. (D) Changes in number of WOX5-expressing cells at the QC position. Data represent mean values ± SD (error bars) (n = 9, ***P < 0.001; paired t test). (E) Starch granules stained by Lugol. The yellow arrowhead indicates the position of columella stem cells. Inset shows a close-up view. (F) Confocal image of root meristem of tpst-1 seedling grown in the presence of At5g60810 (RGF1) peptide. (G) Quantitative analysis of the number of meristematic cortex cells (n = 10). (H) CYCB1;1:GUS expression in tpst-1 root meristem cultured in the presence of At5g60810 (RGF1) peptide. (I) Quantification of CYCB1;1:GUS spots (n = 10). (J) Whole-mount in situ hybridization of RGF1 mRNA. (K) Whole-mount immunostaining with antibodies against RGF1. Scale bars, (C, E, H, J, and K) 50 μm; (F) 100 μm.

To test whether the At5g60810 peptide is responsible for the tpst-1 root phenotype, we cultured tpst-1 seedlings in liquid medium containing the chemically synthesized sulfated form of At5g60810 peptide. At5g60810 peptide suppressed formation of extra QC cells (Fig. 2, C and D) and restored columella stem cells to a level comparable to the wild type (Fig. 2E and table S1), which indicates that At5g60810 peptide is sufficient for maintenance of root stem cell niche. At5g60810 peptide also restored meristematic activity of tpst-1 to ~70% of the wild-type (WT) level, as indicated by root meristem size (Fig. 2, F and G) and CYCB1;1:GUS expression (Fig. 2, H and I). We named this peptide root meristem growth factor 1 (RGF1). RGF1 further increased meristematic activity of tpst-1 to a level comparable to that of the wild type in the presence of PSK and PSY1, suggesting that PSK and PSY1 are required for full activity of RGF1 (fig. S6, A to D). RGF1 fully restored root growth of tpst-1 in the presence of PSK and PSY1 (fig. S6, E to H and table S1).

RGF1 is active at concentrations above 0.1 nM in a dose-dependent manner (fig. S7, A to F and H). Tyrosine sulfation is critical for the function of this peptide (fig. S7, G and H). The earliest detectable increase in meristem size occurred 6 to 12 hours after RGF1 treatment (fig. S8, A to E). Root meristem activity of tpst-1 was also recovered, albeit to varying degrees, by the application of the synthetic sulfated peptides corresponding to the 13–amino acid conserved domain of the other members of this RGF peptide family; the only exception was RGF8, which shows no activity (Fig. 2B and fig. S9, A to M).

In situ hybridization revealed that the expression of RGF1 is limited to the QC and columella stem cells (Fig. 2J and fig. S10J). In addition, two of the homolog peptides that showed strong activity (RGF2 and RGF3) mainly expressed in the innermost layer of central columella cells (fig. S10, B and C). Whole-mount immunostaining with the use of an antibody that specifically recognizes RGF1 and several of its homologs revealed that RGF1 and cross-reactive homologs diffuse into the meristematic region (Fig. 2K and fig. S11, A and B).

To analyze the functions of RGFs, we compared the root phenotype of transgenic seedlings overexpressing RGF1 and WT seedlings treated with RGF1 peptide. Both types of seedlings showed enlarged meristems, indicating that RGF1 peptide treatment phenocopies RGF1 overexpression (fig. S12, A to D). We confirmed that the pskr1-2 pskr2-1 pskrl1-1 triple mutant (8), which lacks PSK receptors and the putative PSY1 receptor, also showed enlarged meristems in the presence of RGF1 (fig. S12, E to G). Plants carrying transferred DNA insertion mutations in the promoter region of RGF1 (rgf1-1) and in the introns of RGF2 (rgf2-1) and RGF3 (rgf3-1) (fig. S12H) showed no RGF1, RGF2, and RGF3 transcripts [quantitative reverse transcription polymerase chain reaction (RT-PCR); fig. S12, I to K]. Immunostaining confirmed the decrease in RGF peptides in the root meristem of the rgf1 rgf2 rgf3 triple mutant (fig. S12L). Although single mutants did not exhibit any obvious phenotypes in roots, the rgf1 rgf2 rgf3 triple mutant showed a short-root phenotype characterized by the decrease in meristematic cell number (Fig. 3, A, B, and D, and fig. S12M). Externally applied RGF1 restored meristem activity in the rgf1 rgf2 rgf3 triple mutant (Fig. 3, B to D). Thus, we concluded that RGFs redundantly regulate root meristem activity.

Fig. 3

Phenotype of rgf1 rgf2 rgf3 triple mutant. (A to C) Confocal images of the root meristem of WT (A), rgf1 rgf2 rgf3 triple mutant (B), and rgf1 rgf2 rgf3 triple mutant treated with RGF1 (C) at 5 DAG. (D) Quantitative analysis of the number of meristematic cortex cells (n = 10). Scale bars in (A) to (C), 100 μm.

To dissect the molecular mechanisms by which RGFs maintain the stem cell niche and regulate meristem activity, we analyzed the response of several root-specific transcription factor mutants to RGF1 by exogenous application of RGF1 peptide. We found that a loss-of-function mutant of PLETHORA (PLT) transcription factors (plt1-4 plt2-2 double mutant) shows little or no response to RGF1 (fig. S13, A to E). PLT genes, specifically expressed in root meristem, encode AP2-domain transcription factors that mediate patterning of the root stem cell niche (10). PLT proteins display gradient distributions with maxima in the stem cell area, and these gradients are essential for maintenance of the root stem cell niche and transit-amplifying cell proliferation (11).

To determine whether the amount of RGF regulates PLT expression, we observed expression of PLT1pro:PLT1-GFP and PLT2pro:PLT2-GFP in WT and tpst-1 backgrounds. In WT seedlings, PLT1–green fluorescent protein (GFP) and PLT2-GFP fusions in WT seedlings show gradients that extend into the region of transit-amplifying cells and, for the PLT2 fusion, as far as the elongation zone (11) (Fig. 4, A and D). In tpst-1 seedlings, both PLT1-GFP and PLT2-GFP signals were reduced, with weak expression in the meristematic zone (Fig. 4, B, E, and G). We confirmed that expression of PLT1-GFP and PLT2-GFP in tpst-1 was recovered when treated with RGF1 for 24 hours, irrespective of the absence or presence of PSK (Fig. 4, C, F, and G, and fig. S13, F to J). PLT2-GFP signal was also reduced in the rgf1 rgf2 rgf3 triple mutant (Fig. 4, H and I). These results indicate that RGFs positively regulate PLT expression levels.

Fig. 4

RGF1 defines expression levels of PLETHORA transcription factors. (A and B) Confocal image of root meristem of WT and tpst-1 seedlings expressing PLT1-GFP at 5 DAG. (C) Root meristem of tpst-1 seedling expressing PLT1-GFP treated with RGF1 for 24 hours. (D and E) Confocal image of root meristem of WT and tpst-1 seedlings expressing PLT2-GFP. (F) Root meristem of tpst-1 seedling expressing PLT2-GFP treated with RGF1 for 24 hours. (G) Quantitative analysis of PLT-GFP signals. BG, background fluorescence (n = 7). (H) Confocal image of root meristem of rgf1 rgf2 rgf3 triple-mutant seedling expressing PLT2-GFP at 5 DAG. (I) Quantitative comparison of PLT2-GFP signal between WT and rgf1 rgf2 rgf3 triple mutant (n = 6). Scale bars in (A) to (F) and (H), 100 μm.

Quantitative RT-PCR analysis revealed that the PLT1 level in tpst-1 is comparable to that of the wild type (fig. S14A). PLT2 transcripts were decreased but not absent in tpst-1 (fig. S14G), suggesting that PLT expression is regulated at both transcriptional and posttranscriptional levels. Consistent with this, in situ hybridization confirmed that most PLT1 and PLT2 transcripts are detected in tpst-1 root meristem, with the highest amounts found in the stem cell region (fig. S14, B to E and H to K). We next treated WT seedlings expressing PLT2-GFP with RGF1 for 24 hours and observed that the PLT2-GFP–expression domain expanded into the basal zone (fig. S13K). Accumulation of PLT2 transcripts, however, is restricted to the stem cell region (fig. S14L), suggesting that expanded expression of PLT2 protein may be due to posttranscriptional regulation, possibly through stabilization of PLT2 protein by RGF1 signaling. Similarly, localization of PLT1 transcripts was also not altered by RGF1 treatment (fig. S14F). When RGF1-treated seedlings expressing PLT2-GFP were further incubated in the absence of RGF1 for 24 hours, gradient expression of PLT2-GFP was restored in the meristematic region (fig. S13L). Thus, RGF acts posttranscriptionally to define PLT expression levels and patterns.

We predict that modulation of local concentration of RGF would alter PLT expression patterns. To confirm this, we placed dextran microbeads containing RGF1 directly on the surface of the meristematic region of WT roots expressing PLT2-GFP (fig. S15A). In the absence of RGF1 beads, PLT2-GFP displayed vertically symmetric distributions, as indicated by pseudocolor intensity images (fig. S15B). In contrast, when RGF1 beads were placed on the meristematic region for 6 hours, PLT2-GFP expression was up-regulated, specifically at the side in contact with the beads, resulting in asymmetric distributions of PLT2-GFP protein (fig. S15, C to E). These results suggest that RGF acts directly on individual cells to define PLT expression levels. RGF expression is not affected by auxin levels (fig. S16).

Thus, RGFs, a redundant family of sulfated peptides, maintain the postembryonic root stem cell niche by defining expression levels and patterns of PLT transcription factors independently of the auxin pathway. The physiological and molecular characteristics of RGFs are such that they might act as morphogens (12). It will be very interesting to unravel how individual cells perceive and integrate the concentration-dependent information provided by RGFs and auxin to generate precise root meristem patterning during root development.

Supporting Online Material

www.sciencemag.org/cgi/content/full/329/5995/1065/DC1

Materials and Methods

Figs. S1 to S16

Table S1

References

References and Notes

  1. Materials and methods are available as supporting material on Science Online.
  2. We thank P. Doerner for providing the CYCB1;1:GUS seeds, T. Guilfoyle for the DR5:GUS seeds, and B. Scheres for the plt1-4 plt2-2 seeds. This research was supported by a Grant-in-Aid for Scientific Research for Priority Areas from the Ministry of Education, Culture, Sports, Science, and Technology (no. 19060010) and a Grant-in-Aid for Creative Scientific Research from the Japan Society for the Promotion of Science (no. 19GS0315).
View Abstract

Navigate This Article