Report

A Genetic Framework for the Control of Cell Division and Differentiation in the Root Meristem

See allHide authors and affiliations

Science  28 Nov 2008:
Vol. 322, Issue 5906, pp. 1380-1384
DOI: 10.1126/science.1164147

Abstract

Plant growth and development are sustained by meristems. Meristem activity is controlled by auxin and cytokinin, two hormones whose interactions in determining a specific developmental output are still poorly understood. By means of a comprehensive genetic and molecular analysis in Arabidopsis, we show that a primary cytokinin-response transcription factor, ARR1, activates the gene SHY2/IAA3 (SHY2), a repressor of auxin signaling that negatively regulates the PIN auxin transport facilitator genes: thereby, cytokinin causes auxin redistribution, prompting cell differentiation. Conversely, auxin mediates degradation of the SHY2 protein, sustaining PIN activities and cell division. Thus, the cell differentiation and division balance necessary for controlling root meristem size and root growth is the result of the interaction between cytokinin and auxin through a simple regulatory circuit converging on the SHY2 gene.

Postembryonic root growth is sustained by the root meristem. Stem cells in the root meristem generate transit-amplifying cells, which undergo additional division in the proximal meristem, and differentiate at the meristem transition zone that encompasses the boundary between dividing and expanding cells in the different cell files (Fig. 1A). For meristem maintenance, the rate of cell differentiation must equal the rate of generation of new cells: How this balance is achieved is a central question in plant development.

Fig. 1.

The SHY2 gene controls root meristem size and is a direct regulatory target of ARR1. (A to C) Five-days-postgermination (dpg) root meristems of wild-type (A), shy2-31 loss-of-function mutant (B), and shy2-2 gain-of-function mutant (C). Root meristem size is expressed as the number of cortex cells in a file extending from the quiescent center (blue arrowheads) to the first elongated cortex cell (black arrowheads). STN indicates stem cell niche, PM, proximal meristem; EDZ, elongation-differentiation zone; and TZ, transition zone. The transition zone is different for each cell type, giving a jagged shape to the boundary between dividing and expanding cells. (D) Root meristem cell number of wild-type, shy2-2, and shy2-31 mutants depicted in (A) to (C), measured over time. (E to G) Expression of the SHY2::GUS construct in wild-type roots (E), in roots treated 3 hours with 5 μM transzeatin (Zt) and stained for 6 hours (F), and in the arr1-3 mutant treated 3 hours with 5 μM Zt and stained for 6 hours (G). (H) Schematic representation of the SHY2 promoter region. Thick and thin lines correspond to coding and non-coding regions, respectively. The bent arrow indicates the transcription start site. Bars with numbers illustrate the DNA fragments used in both ChIP (J) and gel mobility shift analyses (I). (I) Gel mobility shift analysis. Labeled DNA fragments were subjected to gel mobility shift analysis with a fusion protein consisting of GST and ARR1 DNA binding domain. (J) ChIP analysis. Chromatin preparations from ARR1:GFP and wild-type (negative control) roots were coimmunoprecipitated with anti-GFP antibody. An enrichment fold of each fragment was obtained by normalizing the recovery rate for the ARR1:GFP preparation against that for the negative control preparation (24). Error bars indicate SE.

Classic plant tissue culture experiments showed that auxin and cytokinin are key signaling molecules controlling meristems activity because they antagonistically affect, in vitro, shoot and root organogenesis (1, 2). Recently, the in vivo importance of the cytokinin and auxin antagonistic interaction has been proven during Arabidopsis root meristem size determination (3) and for embryonic root stem cell niche specification (4), but the genetic and molecular basis of this interaction remains to be clarified.

The expression domains of the genes encoding the two-component cytokinin signaling pathway, the AHK3 receptor kinase and the ARR1 and ARR12 transcription factors, and the root meristem phenotype of ahk3, arr1, and arr12 mutants demonstrate that cytokinin acts at the transition zone to control cell differentiation rate (3, 5, 6). Furthermore, the expression patterns of cytokinin biosynthesis genes (7) and experiments of tissue- and spatial-specific cytokinin depletion in the root meristem showed that cytokinin specifically acts at the vascular tissue transition zone, where it controls the differentiation rate of all the other root cells by antagonizing a non–cell-autonomous signal that we suggested may be auxin (3). On the other hand, experiments of exogenous application of auxin and the effects of mutations in the PIN auxin efflux facilitators are consistent with a role of auxin in controlling cell division (3, 8). Thus, the size of the root meristem may be established by a balance between the antagonistic effects of cytokinin, which mediates cell differentiation, and auxin, which mediates cell division.

In order to dissect the cytokinin-auxin interaction in the root meristem, we sought to identify the gene(s) immediately responsive to cytokinin signaling. Cytokinin control of root meristem size is mediated by both ARR1 and the ARR12 transcription factors, specifically expressed at the root transition zone and acting, respectively, early and late during meristem development (3). Whereas the root meristems of arr12 mutants were larger than those of wild type but eventually stopped growing, arr1 mutant meristems kept increasing in size over time (3). ARR1 seems therefore to have a critical role in determining root meristem size, and thus we set to verify whether this transcription factor alone would be sufficient to control cell differentiation and root meristem size. To this aim, we checked the root phenotype of plants harboring a glucocorticoid-inducible construct of ARR1 (35S::ARR1ΔDDK:GR) (9). As in the case of exogenous cytokinin applications (3), root meristems of 35S::ARR1ΔDDK:GR plants were significantly reduced after 8 hours of dexamethasone (a glucocorticoid derivative) induction, suggesting that ARR1 is capable alone of controlling root meristem size (fig. S1, A and B).

The same 35S::ARR1ΔDDK:GR construct had been previously used to identify a number of putative direct-target genes of the ARR1 transcription factor (10). Among the 23 putative targets of ARR1 (10), there is notably SHY2, a member of the auxin/indole-3-acetic acid inducible (Aux/IAA) family of genes (11) that act as auxin-response inhibitors by forming heterodimers with the ARF (auxin response factor) transcription factors, thereby preventing activation of auxin-responsive genes (among which are the Aux/IAA genes) by these latter (12, 13). Auxin causes the degradation of the Aux/IAA proteins via the SKIP-CULLIN-FBOX and transport inhibitor response 1 (SCFTIR1) ubiquitin-ligase complex, thus releasing the ARFs from inhibition (13). Degradation of the Aux/IAA proteins and consequent ARF activity depend on auxin concentration: High concentration of auxin causes Aux/IAA degradation, whereas at low auxin concentration these proteins are stable and interact with the ARFs (13).

The presenc eof SHY2 among the putative targets of ARR1 suggests that, in determining root meristem size, cytokinin antagonizes auxin by modulating auxin signaling.

To verify this hypothesis, we analyzed the root-meristem phenotype of shy2 gain- and loss-of-function mutants (14, 15). The shy2-2 gain-of-function mutant allele displayed a shorter root (14, 15) (fig. S1D) and a reduced root meristem size already 2 days after germination (Fig. 1, C and D), which phenocopied the effects of application of cytokinin and of induction of 35S::ARR1ΔDDK:GR. On the contrary, the shy2-31 loss-of-function mutant allele (15) showed an increased number of meristematic cells and kept accumulating cells in the meristem 5 days after germination, vastly exceeding the fixed number of cells of wild-type meristems (Fig. 1, B and D). The increased size of the root meristem correlated with an enhanced rate of root growth and resulted in a longer root (fig. S1D). Because the shy2-31 roots closely resembled arr1 mutant roots, we constructed and analyzed arr1,shy2-2 and arr1,shy2-31 double mutant plants. The root meristem sizes of both arr1,shy2-2 and arr1,shy2-31 double mutants were indistinguishable from that of the arr1 mutant (fig. S1E), supporting the notion that the SHY2 protein may act downstream of the ARR1 transcription factor.

To substantiate the suggested regulatory role of ARR1 on the SHY2 gene, we compared the expression of this gene in wild-type and arr1 mutant roots by using a SHY2::GUS promoter fusion (11, 16). In wild-type root meristems, the activity of this fusion was specifically detected at the vascular tissue transition zone (16) (Fig. 1E and fig. S4J) and was strongly induced by cytokinin treatment (Fig. 1F). This induction of SHY2 promoter activity was matched by a corresponding accumulation of the SHY2 protein, as visualized by Western blot analysis of a transgenic line harboring a Myc-epitope–tagged SHY2 protein (16) (fig. S3A). In contrast, in the arr1 mutant background SHY2::GUS activity was hardly detectable and did not show any increase upon cytokinin treatment (Fig. 1G), confirming that activation of the SHY2 promoter in roots depends on ARR1.

To assess whether SHY2 is necessary and sufficient to mediate the cytokinin control of root meristem size, we analyzed the root phenotype of the shy2-31 loss-of-function mutant harboring the 35S::ARR1ΔDDK:GR construct. The shy2-31, 35S::ARR1ΔDDK:GR roots did not show any reduction in size after 8 hours of dexamethasone induction (fig. S1, A to C), suggesting that SHY2 alone mediates the effect of ARR1 in response to cytokinin to control root meristem size.

The genetic evidence provided establishes a regulatory interaction between ARR1 and SHY2. To confirm the physical interaction between the ARR1 protein and the SHY2 promoter in vivo, we performed chromatin immunoprecipitation (ChIP) assays using roots expressing an ARR1:GFP (green fluorescent protein) fusion protein under the control of the ARR1 promoter. Two DNA fragments in the SHY2 promoter region (fragments 4 and 5 in Fig. 1, H and J) but not those in the promoter of a negative control gene, TUA4 (fig. S2, A and B), were significantly enriched by coimmunoprecipitation with anti-GFP using fragmented chromatin prepared from ARR1:GFP roots as compared with the preparation from wild-type roots. As in the case of the promoter region of ARR6, a direct target-gene of ARR1 (9), the enrichment of DNA fragments was correlated with the binding affinity of the fragments to the ARR1 DNA binding domain in an in vitro gel mobility shift analysis (Fig. 1, I and J, and fig. S2, C and D), supporting the evidence that ARR1 specifically bound to the SHY2 promoter in vivo. The interaction was confirmed to be sequence-specific by gel mobility shift analysis with specific and nonspecific competitors and deoxyribonuclease (DNase) I footprinting analysis (fig. S2, E to G).

These data confirm that SHY2 is a direct regulatory target of the ARR1 transcription factor and provide support for the suggestion that cytokinin antagonizes auxin at the vascular tissue transition zone by modulating auxin signaling and thus determines root meristem size.

The existence of a positive feedback loop between auxin and the PIN auxin-transport facilitators had been established (17), and we had previously reported that cytokinin, acting from the vascular tissue transition zone, requires polar auxin transport to determine the transition zone of all other tissues (3). Thus, we asked whether SHY2 activation by cytokinin, and the resulting reduction of auxin response at the vascular tissue transition zone, would interfere with the activity of the PIN genes in order to control root meristem size.

We first monitored the effect of application of exogenous cytokinin on the expression of the PIN1, PIN3, and PIN7 genes. These particular PIN genes were chosen because they are expressed in the vascular tissue transition zone (17) and because a role for these genes in controlling root meristem size has already been established (3, 7); in fact, the triple pin1,pin3,pin7 mutant displays a short root meristem (7) and is cytokinin-resistant (fig. S3B). Real-time quantitative reverse transcription polymerase chain reaction (qRT-PCR) showed that 4 hours of cytokinin treatment were sufficient to significantly reduce expression of PIN1, PIN3, and PIN7 (Fig. 2J), whereas cytokinin-mediated reduction in root meristem size was first detectable only after 12 hours of exposure to cytokinin (fig. S3B). Notably, as visualized by PIN:GFP translational reporter fusions (8, 18), down-regulation of PIN genes was particularly evident in the vascular tissue transition zone (Fig. 2, B, E, and H).

Fig. 2.

Cytokinin perception at the vascular tissue transition zone controls root meristem size restricting PIN1, PIN3, and PIN7 expression domain. (A, D, and G) Five-dpg wild-type root meristems expressing, respectively, PIN1:GFP, PIN3:GFP, and PIN7:GFP. (B, E, and H) Reduction of PIN expression in cytokinin-treated roots (5 μM Zt, 6 hours) expressing, respectively, PIN1:GFP, PIN3:GFP, and PIN7:GFP. No reduction in root meristem size is still detectable. (C, F, and I) PIN expression in the arr1-3 mutant root meristem expressing, respectively, PIN1:GFP, PIN3:GFP, and PIN7:GFP. In the arr1-3 mutant, the PIN expression domain at the vascular tissue transition zone is expanded (white arrowheads). Blue arrowheads indicate quiescent center cells; white arrowheads indicate cortex transition zone. (J) qRT-PCR confirm PINs gene mRNA down-regulation after 4 hours of cytokinin treatment (5 μM Zt) in wild-type roots, up-regulation in the arr1-3 mutant background, and no down-regulation in the arr1-3 mutant upon cytokinin treatment. Relative expression is normalized to ACTIN, and 0 corresponds to PINs mRNA amount in untreated wild-type roots. Error bars, SD; P < 0.5; Student's t test; n = 3.

To assess whether this cytokinin-dependent down-regulation of the PIN1, PIN3, and PIN7 genes was mediated by the AHK3/ARR1 signaling pathway, we monitored the expression and distribution of PIN mRNAs in the ahk3 and arr1 mutant backgrounds. The expression of the three PIN mRNAs was higher in the ahk3 and arr1 mutants (Fig. 2J and fig. S4K) and the PIN expression domains in the vascular tissue transition zone were expanded as compared with the wild type (Fig. 2, C, F, and I, and fig. S4, B, E, and H). Furthermore, no reduction in root meristem size (fig. S3B) and no down-regulation of PIN1, PIN3, and PIN7 were observed in the ahk3 and arr1 mutant backgrounds in response to cytokinin treatment (Fig. 2J and fig. S4, C, F, I, and K). Thus, perception of cytokinin at the root vascular tissue transition zone, mediated by the AHK3/ARR1 signaling pathway, down-regulates the PIN1, PIN3, and PIN7 genes and restricts their expression domain.

To assess whether this PIN down-regulation depended on SHY2 activation by cytokinin, we monitored the expression and distribution of PIN1, PIN3, and PIN7 mRNAs in the shy2-31 loss-of-function mutant background. PIN::GFP fusion analysis (Fig. 3, B, E, and H) and real-time qRT-PCR (Fig. 3J) showed that, as in the case of the ahk3 and arr1 mutants described above, the expression of PIN1, PIN3, and PIN7 in the shy2-31 mutant was increased and expression domains expanded at the vascular tissue transition zone as compared with the expression of wild type. Furthermore, no down-regulation of the PIN genes and no root meristem reduction were observed in the shy2-31 mutant in response to cytokinin (Fig. 3, C, F, I, and J, and fig. S3B). On the contrary, the expression of PIN1, PIN3, and PIN7 mRNAs was lower in the shy2-2 gain-of-function mutant (fig. S5H).

Fig. 3.

Cytokinin-mediated SHY2 gene induction at the vascular tissue transition zone controls root meristem size restricting the PIN1, PIN3, and PIN7 expression domain. (A, D, and G) Five-dpg wild-type root meristem expressing, respectively, PIN1:GFP, PIN3:GFP, and PIN7:GFP. (B, E, and H) Five-dpg shy2-31 root meristems expressing, respectively, PIN1:GFP, PIN3:GFP, and PIN7:GFP. PIN expression domain at the vascular tissue transition zone is expanded (white arrowheads). (C, F, and I) No change in PIN expression in cytokinin-treated shy2-31 mutant root meristem (5 μM Zt, 12 hours) expressing, respectively, PIN1:GFP, PIN3:GFP, and PIN7:GFP. No changes in root meristem size are detectable. Blue arrowheads indicate quiescent center cells; white arrowheads indicate cortex transition zone. (J) qRT-PCR confirms up-regulation of the PIN gene mRNAs in the shy2-31 mutant and no down-regulation in the shy2-31 mutant upon cytokinin treatment (5 μM Zt, 12 hours). Relative expression is normalized to ACTIN, and 0 corresponds to PIN mRNA levels in untreated wild-type roots. Error bars, SD; P < 0.5; Student's t test; n = 3. (K) qRT-PCR showing PIN down-regulation in root of HS::shy2-6 plants induced at 37 °C for 1 hour. Relative expression is normalized to ACTIN, and 0 corresponds to PIN mRNA amounts in wild-type roots kept for 1 hour at 37°C. Error bars, SD; P < 0.5; Student's t test; n = 3.

To further substantiate the regulatory role of the SHY2 protein on the expression of the PIN genes, we transiently expressed a nondegradable version of the SHY2 protein under a heat-shock promoter (HS::shy2-6) (15) and recorded changes in PIN1, PIN3, and PIN7 gene expression by real-time qRT-PCR. Induction of HS::shy2-6 for 1 hour was sufficient to reduce the expression of PIN1, PIN3, and PIN7 (Fig. 3K), and the consequent reduction of root meristem size was observed after 3 hours (fig. S5, E to G). These data suggest that the activation of SHY2 by cytokinin at the vascular tissue transition zone restricts the expression of PIN1, PIN3, and PIN7 to the vascular tissue transition zone and that this is necessary to position the transition zone of all other tissues of the root.

It had been shown that auxin controls root meristem size, sustaining the activity of the PIN genes as well as cell division (3, 8). Accordingly, exogenous application of auxin to wild-type roots during growth caused an increase in meristem size detectable after 24 hours of exposure to auxin (fig. S5, A, B, and G), whereas 12 hours of auxin treatment were sufficient to induce the expression of PIN1, PIN3, and PIN7 (fig. S5H).

We then asked whether this up-regulation was the result of the auxin-dependent SHY2 protein degradation via the SCFTIR1 ubiquitin-ligase complex (fig. S3A) (13, 19). No up-regulation of the PIN genes and no increase of meristem size (i.e., cell division) upon application of auxin were observed in the gain-of-function mutant shy2-2 in which the interaction of SHY2 with the F-box protein TIR1 is affected (19) (fig. S5, C, D, G, and H).

The cytokinin biosynthetic gene adenosine triphosphate/adenosine diphosphate isopentenyl-transferase 5 (AtIPT5), involved in controlling root meristem size (3) and specifically expressed in the columella and at the vascular tissue transition zone (7) (fig. S6, A to E), is rapidly induced by auxin (7). To establish whether the control of auxin on cytokinin biosynthesis is mediated by SHY2, we analyzed the expression of an IPT5::GUS promoter fusion (7) in the shy2-2 gain-of-function mutant and in the shy2-31 loss-of-function mutant. Although IPT5::GUS activity was higher and more sensitive to auxin in the shy2-31 mutant than in the wild type, auxin-dependent IPT5::GUS activity was totally lost in the shy2-2 gain-of-function mutant (fig. S6, A to E).

We propose a model where cytokinin and auxin antagonistically interact at the vascular tissue transition zone to balance cell differentiation with cell division, thus determining root meristem size, by controlling in opposite ways the abundance of the SHY2 protein. Cytokinin reduces auxin response by activating, via the AHK3/ARR1 two-component signaling pathway, transcription of the SHY2 gene; the SHY2 protein in turn regulates negatively the expression of the PIN genes (fig. S7). The resulting redistribution of auxin leads to cell differentiation of all the other tissues, thus reducing root meristem size. Conversely, auxin controls root meristem growth by directing degradation of the SHY2 protein, thus sustaining the activity of the PIN genes and cell division (fig. S7). The SHY2 protein negatively controls auxin transport on the one hand and cytokinin biosynthesis on the other, thus conferring robustness to the feedback loop we propose.

This model is consistent with the low concentration of auxin at the transition zone observed by other authors (20) and with our previous demonstration that the vascular tissue transition zone is the developmental domain of cytokinin action (3).

It is tempting to hypothesize that in root meristem size determination the PLETHORA (PLT) (21) proteins mediate the effect of cytokinin by responding to changes in polar auxin transport caused by cytokinin. It has been shown that the gradient of auxin in the root meristem creates a gradient of PLT proteins that direct different outputs in response to auxin in different regions of the meristem (22). Changes in PIN proteins activity affect the gradient distribution of auxin and of the PLT proteins (8, 23). Interestingly, we observed that the PLT1 and PLT2 genes are down-regulated by cytokinin (fig. S6F).

Supporting Online Material

www.sciencemag.org/cgi/content/full/322/5906/1380/DC1

Materials and Methods

Figs. S1 to S7

References

References and Notes

View Abstract

Navigate This Article