SEARCH

SEARCH BY CITATION

Keywords:

  • cytokinin;
  • auxin;
  • PIN;
  • two-component signaling;
  • root development

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

The phytohormones cytokinin and auxin regulate a diverse array of plant processes, often acting together to modulate growth and development. Although much has been learned with regard to how each of these hormones act individually, we are just beginning to understand how these signals interact to achieve an integrated response. Previous studies indicated that exogenous cytokinin has an effect on the transcription of several PIN efflux carriers. Here we show that disruption of type-A Arabidopsis response regulators (ARRs), which are negative regulators of cytokinin signalling, alters the levels of PIN proteins and results in increased sensitivity to N-1-naphthylphthalamic acid, an inhibitor of polar auxin transport. Disruption of eight of the 10 type-A ARR genes affects root development by altering the size of the apical meristem. Furthermore, we show that the effect of cytokinin on PIN abundance occurs primarily at the post-transcriptional level. Alterations of PIN levels in the type-A ARR mutants result in changes in the distribution of auxin in root tips as measured by a DR5::GFP reporter, and an altered pattern of cell division and differentiation in the stem cell niche in the root apical meristem. Together, these data indicate that cytokinin, acting through the type-A ARRs, alters the level of several PIN efflux carriers, and thus regulates the distribution of auxin within the root tip.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Cytokinin and auxin affect diverse aspects of plant growth and development, including cell division, and shoot and root initiation and growth. Cytokinins were first identified by their ability to stimulate cell division in cultured plant cells in concert with auxin (Miller et al., 1955, 1956), and subsequently these two phytohormones have been shown to act together in many processes (Su et al., 2011). Recently, a model has emerged for the interaction of cytokinin and auxin in the regulation of root meristem function (Moubayidin et al., 2009).

The cytokinin response pathway is similar to bacterial two-component phosphorelays (To and Kieber, 2008; Argueso et al., 2010; Perilli et al., 2010). In Arabidopsis, cytokinin binds to Arabidopsis histidine kinase (AHK) receptors, activating their ability to transfer a phosphoryl group to the Arabidopsis homologues of histidine phosphotransfer (Hpt) proteins (AHPs). The AHPs transfer the phosphoryl group to Arabidopsis response regulators (ARRs), which include type-A and type-B ARRs. Analyses of single and multiple loss-of-function mutations in the AHK, AHP and type-B ARR genes indicate that they are positive, redundant elements in the cytokinin primary signal transduction pathway (Higuchi et al., 2004; Nishimura et al., 2004; Mason et al., 2005; Hutchison et al., 2006; Riefler et al., 2006; Argyros et al., 2008), and the type-A ARRs act primarily as negative regulators (To et al., 2004). The type-A ARRs are transcriptionally induced by cytokinin via direct activation by the type-B ARRs (D’Agostino et al., 2000; Hwang and Sheen, 2001; Sakai et al., 2001; Mason et al., 2004).

Auxin is transported through plant tissues by the regulated expression and localization of efflux and influx carriers (Blakeslee et al., 2005), the best understood of which are the PIN efflux carriers (Zazímalováet al., 2007). These are encoded by eight partially redundant genes in Arabidopsis. The PIN proteins localize asymmetrically on the plasma membrane to direct polar auxin transport within a tissue. Recent studies demonstrated that exogenous cytokinin alters the transcription of several PIN genes in the Arabidopsis root (Dello Ioio et al., 2008; Pernisováet al., 2009; Ruzicka et al., 2009). Auxin and cytokinin interact to regulate the size of the root apical meristem, with cytokinin acting to increase the rate of cell differentiation by negatively regulating PIN1, PIN3 and PIN7 expression through induction of SHY2/IAA3, which encodes a repressor of auxin signalling (Dello Ioio et al., 2008). Here, we examine the role of the type-A ARRs in this process. We show that disruption of multiple type-A ARRs leads to a decrease in the size of the root apical meristem, consistent with previous studies with exogenous cytokinin. We demonstrate that expression of a subset of PINs is modulated by cytokinin primarily at the post-transcriptional level, and that this leads to a change in the distribution of auxin in the root tip and perturbation in the pattern of cell division.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

An octuple type-A ARR mutant is affected in cytokinin response and root development

The type-A Arabidopsis response regulator (ARR) gene family includes 10 genes grouped in five gene pairs based on phylogenetic analysis (Figure S1a). Previously, we have shown that three gene pairs (ARR3/ARR4, ARR5/ARR6 and ARR8/ARR9) are expressed in the root, and that they act partially redundantly to negatively regulate the response of roots to exogenous cytokinin (To et al., 2004). We further examined the expression pattern of two additional type-A ARR genes, ARR7 and ARR15, using promoter::GFP reporters (Müller and Sheen, 2008). No insertional mutants were available for ARR16 and ARR17 when the studies described in this paper began, and so the remaining two type-A ARRs were not analysed. Both ARR7 and ARR15 showed strong expression in the root tip (Figure S1c), implying that this gene pair may also be involved in regulation of root growth by cytokinin. We introduced T-DNA insertional alleles for these two genes (Figure S1b) into various multiple type-A ARR mutant lines to generate higher-order mutants. Expression analysis by quantitative RT-PCR indicated that arr7 is essentially a null allele, while arr15 is a hypomorphic allele (Figure S1d). Previous studies indicated that the arr7 arr15 double mutant was gametophyte-lethal (Leibfried et al., 2005). Consistent with this, very few lines homozygous for both the arr7 and arr15 mutations were obtained in the background of other type-A ARR loss-of-function mutants. For example, almost no quadruple homozygous mutants were obtained from self-pollinated arr5 arr6 arr7 arr15/arr5 arr6 arr7 ARR15 plants. However, several rare lines were obtained from these selfed plants that were homozygous for all four type-A ARR mutations. These arr5,6,7,15 quadruple mutants obtained were fully fertile, indicating that disruption of arr7 and arr15 results in neither gametophyte- nor embryo-lethality. Similar results were obtained with a higher-order arr3,4,5,6,7,8,9,15 octuple mutant. The segregation studies suggested that the arr7 and arr15 gametophytic lethality resulted from the nature of these insertional alleles. In the heterozygous state, approximately 50% of the pollen from both arr7 and arr15-2 is defective; in the homozygous state, both mutants produce little defective pollen (Figure S1e) and were fully viable and fertile. These results suggest that these mutations most likely contain chromosomal translocations that occurred as a result of the T-DNA insertion events into the ARR7 and ARR15 genes on the top (ARR7) and bottom (ARR15) arm of chromosome 1.

The role of ARR7 and ARR15 in root growth and development was analysed by examining various multiple mutant lines. Neither the arr7 nor the arr15 single mutants showed any substantial effect on cytokinin responsiveness or growth and development (data not shown). The arr5,6,7,15 quadruple mutant was more sensitive to cytokinin in a root growth assay than the arr5,6 double mutant at concentrations of cytokinin ≥ 10 nm. Moreover, inclusion of either the arr7 or arr15 mutation into the sextuple arr3,4,5,6,8,9 mutant further increased the sensitivity to cytokinin (Figure 1a). These results indicate that ARR7 and ARR15 are partially redundant with the other type-A ARR genes with respect to a negative role in the response of roots to cytokinin. In addition to affecting the response to exogenous cytokinin, disruption of ARR7 and/or ARR15 in the background of other type-A ARR mutations strongly affected root elongation in the absence of exogenous cytokinin (Figure 1a), suggesting that these type-A ARRs play a role in regulating primary root growth. Previous studies indicated that exogenous cytokinin altered root growth by decreasing the size of the root apical meristem (Dello Ioio et al., 2007). Consistent with this, the root meristem size in the arr3,4,5,6,7,8,9,15 mutant was significantly smaller than that of the wild-type at 3 days post-germination, and this difference became more pronounced from days 5 to 7 (Figure 1b). Disruption of type-A ARRs also led to reduced root architecture, with the higher-order mutants displaying stronger phenotypes (Figure 1c).

image

Figure 1.  Identification and initial characterization of the arr3,4,5,6,7,8,9,15 octuple type-A arr mutant. (a) Growth of wild-type (WT) or the indicated multiple type-A arr mutant from day 4–9 on MS medium supplemented with the specified concentrations of BA or DMSO (control). Error bars represent SE (> 30). The experiment was repeated at least twice with consistent results. (b) Number of cells in cortex of the root meristem of wild-type and arr3,4,5,6,7,8,9,15 mutants during the first 11 days after germination. Error bars represent SD (> 10). The experiment was repeated at least twice with consistent results. (c) Numbers of 1st-order (light grey) and 2nd-order (dark grey) lateral roots in wild-type and multiple type-A ARR mutants after 12 days of growth on MS medium. Error bars represent SE (> 20). Asterisks indicate values that were significantly different from wild-type at < 0.05 (Student’s t test). The experiment was repeated twice with consistent results. (d) Number of 1st-order lateral roots per mm root length in wild-type and multiple type-A ARR mutants grown on MS medium supplemented with DMSO, 0.1 or 1.0 μm NPA as indicated. Error bars represent SE (= 15). Asterisks indicate values that were significantly different from wild-type (*< 0.1, **< 0.05; Student’s t test).

Download figure to PowerPoint

Disruption of the type-A ARRs alters the level of PIN proteins

To determine whether altered auxin transport contributes to the root phenotype observed in the type-A ARR mutants, we examined the sensitivity of the mutants to the auxin transport inhibitor N-1-naphthylphthalamic acid (NPA) (Figure 1d). Consistent with previous results suggesting a link between PIN auxin efflux carriers and cytokinin, the type-A ARR mutants displayed an increased sensitivity to NPA as measured by the effect on lateral root formation, with the highest-order arr3,4,5,6,7,8,9,15 mutant displaying the most enhanced sensitivity.

The altered sensitivity of the arr3,4,5,6,7,8,9,15 mutant to NPA suggests defects in polar auxin transport, consistent with recent results indicating that exogenous cytokinin down-regulates PIN expression (Dello Ioio et al., 2008; Pernisováet al., 2009; Ruzicka et al., 2009). To test this, transgenes expressing PIN–GFP fusion proteins from their endogenous promoters (PIN1::PIN1-GFP, PIN3::PIN3-GFP, PIN4::PIN4-GFP and PIN7::PIN7-GFP) were introgressed into the arr3,4,5,6,7,8,9,15 mutant, and the levels of PIN expression were analysed. Consistent with the hypersensitivity to NPA, expression of multiple auxin efflux carriers was altered in the mutant (Figure 2). In arr3,4,5,6,7,8,9,15 root tips, the expression of PIN4–GFP was greatly reduced (Figure 2c), and expression of both PIN1–GFP and PIN3–GFP in the stele was slightly reduced compared to the wild-type (Figure 2a,b). The relative PIN fluorescence in these fusion lines was quantified (Figure S2); the levels of PIN1–GFP and PIN3–GFP fluorescence were reduced approximately 20% in the arr3,4,5,6,7,8,9,15 mutant root tips. The expression of PIN7–GFP in the stele was slightly reduced, but its expression in the root cap was increased and expanded (Figure 2d). To confirm these GFP fusion results, we examined the endogenous PIN1 protein level in the root using in situ immunocytochemistry with an anti-PIN1 antibody. Consistent with the analysis of the fusion proteins, the level of endogenous PIN1 protein was also reduced in the root tips of the arr3,4,5,6,7,8,9,15 mutant (Figure 3).

image

Figure 2.  Disruption of type-A ARRs alters the level of PIN–GFP proteins. (a–d) Five- to 6-day-old wild-type (WT) and arr3,4,5,6,7,8,9,15 mutant seedlings expressing PIN1–GFP, PIN3–GFP, PIN4–GFP or PIN7–GFP were treated with 5 μm BA for the specified durations, or with DMSO (control) for 24 h, as indicated, and then imaged by confocal microscopy. Scale bar = 20 μm. The roots were stained with 5 μm propidium iodide to visualize the outlines of the cells.

Download figure to PowerPoint

image

Figure 3.  Disruption of the type-A ARRs reduces the native PIN1 protein levels. Native PIN1 protein (detected using an anti-PIN1-specific antibody) in the root tips of wild-type (WT) and arr3,4,5,6,7,8,9,15 mutants treated with 5 μm BA or DMSO for 24 h as revealed by whole-mount immunohistochemistry (see Experimental procedures).

Download figure to PowerPoint

We next examined the effect of disruption of the type-A ARRs on regulation of PIN expression in response to exogenous cytokinin. The levels of PIN1–GFP and PIN3–GFP responded more strongly to cytokinin in arr3,4,5,6,7,8,9,15 mutant roots, decreasing more rapidly and to a greater extent in response to exogenous cytokinin compared to wild-type roots. Treatment with cytokinin for 14 and 24 h modestly reduced PIN1–GFP expression in wild-type roots, and by 48 h, expression was almost eliminated (Figures 2a and S2a). In contrast, in arr3,4,5,6,7,8,9,15 mutant roots, the level of PIN1–GFP was strongly reduced 14 h after cytokinin treatment and was nearly absent by 24 h. The level of native PIN1 protein showed comparable cytokinin hypersensitivity in arr3,4,5,6,7,8,9,15 mutant roots (Figure 3). Pronounced hypersensitivity to cytokinin in the arr3,4,5,6,7,8,9,15 mutant roots was also observed for expression of PIN3–GFP in the stele, although expression of PIN3–GFP in the columella cells of the root cap appeared almost insensitive to cytokinin (Figures 2b and S2b). PIN4–GFP expression was greatly reduced by cytokinin in wild-type roots, but was nearly undetectable in the arr3,4,5,6,7,8,9,15 mutant roots even in the absence of exogenous cytokinin (Figure 2c). Finally, expression of the PIN7–GFP reporter was slightly elevated in response to cytokinin in wild-type roots, primarily in the stele, but cytokinin had little effect on PIN7–GFP levels in the mutant roots (Figure 2d). Together, these results indicate that disruption of the type-A ARRs alters the basal level of some PINs, and sensitizes the root to the effects of exogenous cytokinin on PIN expression.

PIN function is regulated primarily by a post-transcriptional mechanism

Previous studies have shown that cytokinin alters PIN function through an alteration of PIN transcript levels mediated by the SHY2 gene (Dello Ioio et al., 2008). We therefore examined expression of the PIN and SHY2 genes in wild-type and arr3,4,5,6,7,8,9,15 mutant root tips. RNA was isolated from root tips of approximately 0.5 mm (including the meristem and transition zones), and analysed using the Nanostring nCounter analysis system, which is a highly accurate and sensitive method of mRNA analysis based on hybridization of RNA to fluorescently bar-coded probes (Geiss et al., 2008). Surprisingly, the endogenous transcript levels of the five PIN genes did not show any substantial differences in untreated arr3,4,5,6,7,8,9,15 mutant root tips compared to the wild-type (Figure 4a), with the possible exception of PIN2, which showed an approximately 25% decrease in the mutant (Figure S3). We also analysed expression of the PIN4–GFP transgene using quantitative RT-PCR to confirm the analysis of expression of the endogenous PIN4 gene. Consistent with the nCounter analysis of endogenous PIN4 gene expression, the level of the RNA transcript of the PIN4–GFP transgene was not reduced in the arr3,4,5,6,7,8,9,15 mutant roots (Figure 4b), despite the strong reduction in PIN4–GFP protein levels.

image

Figure 4.  Effects of cytokinin on the level of PIN transcripts in wild-type and arr3,4,5,6,7,8,9,15 mutant root tips. (a) Normalized counts for PIN1, PIN3, PIN4, PIN7 and SHY2 transcripts in 0.5 mm root tips from wild-type (WT) and arr3,4,5,6,7,8,9,15 mutant seedlings treated with DMSO or 5 μm BA for 8 or 24 h, as determined using Nanostring nCounter gene expression analysis. Error bars represent SE from three biological replicates. (b) Relative expression level of the PIN4–GFP transcript in 0.5 mm root tips of wild-type and arr3,4,5,6,7,8,9,15 mutants treated with 5 μm BA or DMSO for 8 or 24 h, as determined by quantitative RT-PCR. Error bars represent SEM from three biological replicates.

Download figure to PowerPoint

We next examined the transcript level of the PIN genes in response to cytokinin. Cytokinin treatment had little or no effect on PIN1 transcript levels in wild-type even after 24 h (Figure 4a). This is distinct from the strong down-regulation of PIN1–GFP fusion protein levels observed in wild-type root tips in response to cytokinin. In the arr3,4,5,6,7,8,9,15 mutant, the transcript level of PIN1 was reduced approximately twofold 8 h after treatment with cytokinin (Figure 4a). The endogenous PIN4 transcript level was increased by cytokinin treatment in wild-type and arr3,4,5,6,7,8,9,15 mutant roots (Figure 4a), and the transgenic PIN4–GFP transcript level did not change significantly in response to cytokinin treatment (Figure 4b). In contrast, PIN4–GFP protein levels were strongly reduced in wild-type roots in response to cytokinin (Figure 2c).

Cytokinin treatment had only modest effects on the level of PIN7 transcripts in wild-type and mutant root tips (Figure 4a). PIN3 transcript levels were not significantly altered by cytokinin in wild-type root tips, but were down-regulated approximately twofold in the mutant. The expression of PIN2, which was not analysed using GFP fusions, was slightly reduced by cytokinin in wild-type roots, and this effect was enhanced in the arr3,4,5,6,7,8,9,15 mutant (Figure S3). Surprisingly, induction of the SHY2 transcript in the arr3,4,5,6,7,8,9,15 mutant was similar to that observed in wild-type roots (Figure 4a), i.e., unlike the response of many other cytokinin-responsive genes, induction of SHY2 was not enhanced in the mutant. Together, these data suggest that down-regulation of PIN1, PIN3 and PIN4 in response to cytokinin in wild-type and mutant roots occurs primarily at the post-transcriptional level, and that the hypersensitive response in the mutant is not the result of increased transcriptional induction of SHY2.

The distribution of the auxin response is altered in arr3,4,5,6,7,8,9,15 mutant roots

PIN4 has been shown to play an important role in generating an auxin sink and maintaining auxin gradients in the root tips (Friml et al., 2002). As PIN4–GFP levels were decreased in the root tips in the arr3,4,5,6,7,8,9,15 mutant, we examined the spatial distribution of the auxin response using the auxin-response reporter DR5::GFP. We introgressed the DR5::GFP transgene into the arr3,4,5,6,7,8,9,15 mutant by multiple back-crosses. The root tips of the arr3,4,5,6,7,8,9,15 mutant showed an altered pattern of DR5::GFP expression, most notably a reduction of DR5::GFP expression in cells of the quiescent centre (QC) (Figures 5a and S2c). This is consistent with previous studies showing that cytokinin treatment reduced DR5 expression in the QC (Ruzicka et al., 2009). Interestingly, although DR5::GFP expression in QC cells was reduced in the arr3,4,5,6,7,8,9,15 mutant, the expression in columella stem cells was somewhat elevated (Figure 5b).

image

Figure 5.  The distribution of auxin is altered in arr3,4,5,6,7,8,9,15 mutant roots. (a,b) Wild-type (WT) and arr3,4,5,6,7,8,9,15 mutant roots expressing an auxin-responsive reporter, DR5::GFP. The yellow arrowheads indicate quiescent centre cells; the blue arrowheads indicate columella stem cells. (a) Root tips from five-day-old untreated seedlings. (b) Five- to 6-day-old wild-type and arr3,4,5,6,7,8,9,15 mutant roots treated with 5 μm BA for the specified time or DMSO for 24 h as a control. Scale bar = 20 μm. The roots were stained with 5 μm propidium iodide to visualize the outlines of the cells, and were visualized by confocal microscopy.

Download figure to PowerPoint

The pattern of cell division in the arr3,4,5,6,7,8,9,15 mutant is altered

The distribution of auxin in the root tip plays an important role in regulating the pattern of cell division and differentiation. PIN4 plays a critical role in this auxin-mediated patterning; the distribution of auxin is altered in pin4 mutant roots, leading to aberrant cell division patterns, especially near the QC (Friml et al., 2002; Blilou et al., 2005). As PIN4 expression is greatly reduced in the arr3,4,5,6,7,8,9,15 mutant, we examined the pattern of cell division in the mutant root. The arr3,4,5,6,7,8,9,15 mutant roots showed cell division in the normally mitotically inactive QC cells (Figure 6a). Furthermore, some of the columella stem cells in the arr3,4,5,6,7,8,9,15 mutant showed signs of premature differentiation into columella cells, as indicated by staining for starch granules (Figure 6b). Approximately 40% of the arr3,4,5,6,7,8,9,15 mutant seedlings showed either QC division or starch-granule staining in columella stem cells, although these two phenotypes did not always appear in the same root.

image

Figure 6.  The pattern of cell division and differentiation is altered in arr3,4,5,6,7,8,9,15 mutant root tips. (a) Five-day-old root tips of wild-type (WT) and arr3,4,5,6,7,8,9,15 mutants visualized by DIC microscopy. Yellow asterisks indicate divided quiescent centre cells. E, endodermis; C, cortex. Note that the quiescent centre cells in arr3,4,5,6,7,8,9,15 mutant roots appear to undergo anticlinal divisions. (b) Lugol staining of wild-type (WT) and arr3,4,5,6,7,8,9,15 mutant roots. The columella stem cells (the tier of cells below the quiescent centre) show premature differentiation as indicated by Lugol staining for starch granules (white asterisk). Red arrowheads indicate quiescent centre cells. (c) Percentage of roots with quiescent centre (QC) divisions or starch-staining columella stem cells (CSC) in wild-type and arr3,4,5,6,7,8,9,15 mutants. At least 50 seedlings were analysed for each phenotype.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Cytokinin has been shown to inhibit auxin transport in the root tip by inducing expression of an auxin signalling repressor SHY2, thus down-regulating the transcription of several PIN auxin efflux carriers (Dello Ioio et al., 2008; Ruzicka et al., 2009). Here, we show that type-A response regulators are involved in the negative regulation of PINs by cytokinin in the root tips, and that this regulation primarily involves post-transcriptional inputs. PIN1 protein levels decreased in wild-type root tips in response to cytokinin, and by 48 h were nearly absent. However, the transcript level of PIN1 was not changed in response to exogenous cytokinin, in contrast to previous reports (Dello Ioio et al., 2008; Ruzicka et al., 2009). This difference may reflect the fact that previous studies used 2 mm root sections (Ruzicka et al., 2009), whereas we analysed only the 0.5 mm tip of the root, which is more specific for the root meristem region, and thus the transcript levels more accurately reflect gene expression changes in the root meristem region. The discrepancy may also reflect differences in the growth or treatment conditions. Alternatively, use of the NanoString method to analyse transcript levels, which does not involve synthesis or amplification of cDNA, may be more accurate than the quantitative RT-PCR method used in the previous studies. Consistent with the effect of cytokinin on PIN1 expression, the level of PIN4–GFP protein level in the root cap was greatly reduced in response to exogenous cytokinin. Previous studies using anti-PIN4 antibodies found high PIN4 protein levels in the root cap region (Blilou et al., 2005; Vieten et al., 2005), slightly different from the pattern observed with PIN4–GFP. However, the transcript levels of endogenous PIN4 and transgenic PIN4–GFP were not reduced in response to exogenous cytokinin (in fact, the endogenous PIN4 level was slightly elevated). This, coupled with the strong decrease in PIN4–GFP protein levels, is consistent with post-transcriptional regulation of PIN4 by cytokinin. Together, these results suggest that cytokinin negatively regulates the abundance of these PIN proteins in the root tip primarily through a post-transcriptional mechanism, possibly by increasing their rate of turnover.

The finding that the arr3,4,5,6,7,8,9,15 mutant has a smaller root meristem suggests that type-A ARRs may be involved in regulation of PIN protein levels by cytokinin. Indeed, we found that the levels of PIN1, PIN3 and PIN4 protein were significantly reduced in the arr3,4,5,6,7,8,9,15 mutant. However, the transcript level of all three PIN genes was not appreciably altered in arr3,4,5,6,7,8,9,15 roots, consistent with a model in which the negative regulation of PINs by cytokinin in the root meristem involves a post-transcriptional mechanism involving the type-A ARRs. One model consistent with the data is that the type-A ARRs negatively regulate turnover of the PIN proteins, and the low level of type-A ARR gene function remaining in the arr3,4,5,6,7,8,9,15 mutant is sensitized to exogenous cytokinin. Alternatively, the type-A ARRs may act indirectly by mediating negative feedback on the down-regulation of PIN proteins by cytokinin. Whichever is the case, the kinetics of SHY2 induction by benzyladenine (BA) treatment are comparable in arr3,4,5,6,7,8,9,15 and wild-type roots, suggesting that the amplified response of the PIN proteins to cytokinin in this mutant does not involve the ARR1–SHY2 pathway.

The PIN auxin efflux carriers have been shown to play important roles in pattern formation by focusing the auxin maximum in the QC and restricting the expression domain of major determinants of the root stem cell niche (Aida et al., 2004; Blilou et al., 2005). In the arr3,4,5,6,7,8,9,15 mutant, the spatial pattern of DR5::GFP expression is altered, with a less pronounced maximum in the QC cells, similar to the pin2 pin3 pin7 and pin3 pin4 pin7 triple mutants (Blilou et al., 2005). This suggests that the type-A ARRs play a positive role in maintaining the auxin maximum in the root tip and stem cell activity through their regulation of PIN protein levels in the root tips. The QC cells are mitotically inactive and play essential roles in maintaining the stem cell fate of surrounding cells. We found that the QC cells in arr3,4,5,6,7,8,9,15 mutant roots undergo mitotic division at a higher frequency and the columella stem cells show signs of differentiation, suggesting that the type-A ARRs are necessary for proper QC function. It is likely that the altered QC function in arr3,4,5,6,7,8,9,15 roots reflects the reduced auxin response maximum in the QC. However, we cannot exclude an additional auxin-independent role for the type-A ARRs in maintaining QC function. Ethylene has been shown to induce the division of QC cells independently of auxin (Ortega-Martinez et al., 2007). As cytokinin promotes ethylene production via the two-component signalling pathway (Hansen et al., 2009), the QC division in arr3,4,5,6,7,8,9,15 mutants may result from increased ethylene production. However, treatment with the ethylene inhibitor 1-methylcyclopropene (MCP) did not inhibit the aberrant QC division in arr3,4,5,6,7,8,9,15 mutants (data not shown), suggesting that it is probably not due to ethylene over-production.

ARR7 and ARR15 have been reported to be essential for early embryo development, in particular for formation of the embryonic root stem cell niche (Müller and Sheen, 2008). Although we observed an effect of disruption of ARR7 and ARR15 on root meristem function, disruption of both genes is clearly not lethal. This may be because the arr15-2 allele was used, and this allele may not be null. However, we have also examined the arr15-1 allele used in previous studies (Müller and Sheen, 2008), and, consistent with the analysis here, double arr7 arr15-1 mutants are also not lethal (data not shown). A second possibility is that the previous study used inducible RNAi to disrupt ARR7 function, and the inconsistency could reflect the difference between inducible down-regulation and genetic disruption. Finally, the embryos in the previous study were analysed in vitro, and this may affect the phenotype of the double mutant. Our results indicate that ARR7 and ARR15 are not essential for embryo development, although they probably play a redundant role with other type-A ARRs in proper root meristem function.

Previously, we have shown that six Arabidopsis type-A response regulators (ARR3, ARR4, ARR5, ARR6, ARR8 and ARR9) are negative regulators of cytokinin signalling (To et al., 2004). Here, we further analysed two additional type-A ARR genes, ARR7 and ARR15. Analysis of various mutant combinations using root cytokinin response assays indicated that both ARR7 and ARR15 are partially functionally redundant with the other type-A ARR genes. Thus, at least eight of the 10 type-A ARRs have overlapping roles as negative regulators of cytokinin signalling. The finding that the arr3,4,5,6,7,8,9,15 mutant has a much shorter root than the sextuple mutants, coupled with the observation that ARR7 and ARR15 are highly expressed in roots, suggests that ARR7 and ARR15, together with the other type-A ARRs, play an important role in mediating the function of endogenous cytokinin in the root. In Arabidopsis, five pairs of type-A ARR genes appear to have arisen as a result of a hypothesized whole-genome duplication event (To et al., 2004), estimated to have occurred approximately 24–40 million years ago (Vision et al., 2000; Blanc et al., 2003). Population genetic theory postulates that duplicate gene pairs will revert to a single copy over a relatively short evolutionary timescale (Lynch and Conery, 2000). Pairs that are retained probably have at least partially diverged in terms of function via neofunctionalization or by evolving distinct spatial patterns of expression. The fact that each member of the five pairs of Arabidopsis type-A ARRs is maintained following the most recent genome duplication events suggests that, despite the functional redundancy of these genes with regard to the response to exogenous cytokinin, they probably also have non-overlapping roles. For example, ARR3 and ARR4 have been linked to the circadian rhythm (Saloméet al., 2005), and their function in this pathway is specifically opposed by ARR8 and ARR9. Likewise, the ARR7 and ARR15 genes have specifically been shown to be differentially regulated by auxin in the root and shoot apical meristem (Müller and Sheen, 2008; Zhao et al., 2010). The effects of cytokinin on the transcription and stabilization of the type-A ARRs also varies in terms of the magnitude and kinetics of induction (D’Agostino et al., 2000; To et al., 2007), suggesting that the ARRs vary in terms of negative feedback regulation on the cytokinin signalling pathway. We postulate that negative feedback regulation of cytokinin signalling is the ancestral role of the type-A ARRs, as all members appear to contribute to this function in Arabidopsis. As more is learned regarding the function of this gene family, unique roles will probably continue to emerge for the individual members.

Experimental procedures

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Plant materials

The Col-0 ecotype was used in this study. DR5::GFP (Blilou et al., 2005), PIN1::PIN1-GFP (Benkováet al., 2003), PIN3::PIN3-GFP (Žádníkováet al., 2010), PIN4::PIN4-GFP (Vieten et al., 2005), PIN7::PIN7-GFP (Blilou et al., 2005), ARR7::GFP and ARR15::GFP (Müller and Sheen, 2008) have been described previously. The arr7 and arr15-2 alleles were identified by PCR screening of pooled DNA from T-DNA collections generated by the Arabidopsis Knockout Facility at the University of Wisconsin-Madison (Sussman et al., 2000) and the Salk Institute (Alonso et al., 2003), respectively. The arr7 allele in the WS ecotype was introgressed into the Col ecotype three times before further analysis. The insertions in arr7 and arr15-2 were confirmed by genomic PCR using gene-specific and T-DNA border primers. The gene-specific PCR primers used were as follows: 5′-GGCGGTTTGCAGACTCACTTACCTGA-3′ and 5′-GACTCTCTCAAACATTGTCTTT-3′ for ARR7, and 5′-CCATTTATTCTCCTCTCATCTC-3′ and 5′-ATCTAATCATCCCCATCTCC-3′ for ARR15. The T-DNA border primer used for arr7 was JL202 (5′-CATTTTATAATAACGCTGCGGACATCTAC-3′), and that for arr15 was JMLB1 (5′-GGCAATCAGCTGTTGCCCGTCTCACTGGTG-3′). The arr7 and arr15-2 alleles were crossed with arr3,4,5,6 and arr3,4,5,6,8,9, and segregants from those crosses were genotyped and selected for further crosses to generate arr5,6,7,15, arr3,4,5,6,7,8,9, arr3,4,5,6,8,9,15 and arr3,4,5,6,7,8,9,15. Genotyping primers for the arr3, arr4, arr5, arr6, arr8 and arr9 alleles are as previously described (To et al., 2004).

Plant growth conditions and phenotypic analysis

Arabidopsis seeds were surface-sterilized, cold-treated at 4°C for 4 days, and grown on 1×MS/1% sucrose vertical plates as previously described (To et al., 2004). Cytokinin root growth assays were performed as previously described (To et al., 2004) with the specified media supplements. Root lengths at days 4 and 9 after germination were marked on the plates. Root elongation between days 4 and 9 was measured using the ImageJ program (http://rsb.info.nih.gov/ij). A dissecting microscope was used to observe and quantify higher-order lateral roots 12 days after germination. For NPA response assays, lateral root primordia were quantified under a dissecting microscope at 9 days after germination and normalized to total root length. Root meristem size was measured as previously described (Dello Ioio et al., 2007), except that Micropore surgical tape (3M, http://solutions.3m.com) was used to seal the plates.

Microscopy and whole-mount immunohistochemistry

For analysis of GFP reporters in roots, intact seedlings were counter-stained in propidium iodide immediately after the cytokinin treatment to visualize cell walls, and analysed using confocal scanning laser microscopy. The GFP fluorescence and propidium iodide staining were visualized and captured using a Zeiss LSM 510 META scanning confocal microscope with a Zeiss Plan Neofluar 20×/0.5 objective (http://www.zeiss.com/). For GFP, a 488 nm line was used for excitation, and an emission range between 505 and 530 nm was used for detection. For propidium iodide, a 543 nm line was used for excitation, and a 560 nm long-pass filter was used for detection. The wild-type and arr3,4,5,6,7,8,9,15 mutant lines harbouring each GFP reporter construct were analysed at the same time using identical microscope settings throughout the treatments. The GFP florescence was quantified from confocal sections that included the QC using ImageJ (http://rsb.info.nih.gov/ij).

Whole-mount immunohistochemistry in roots was performed as described previously (Sauer et al., 2006). The primary anti-AtPIN1 antibody (Friml et al., 2002) was used at 1:330 dilution. The secondary CY3-conjugated anti-rabbit antibody (Jackson ImmunoResearch, http://www.jacksonimmuno.com/) was used at 1:600 dilution. The immunostained roots were visualized using a Zeiss DUO confocal microscope with an EC Plan Neofluar 40×/1.30 oil-immersion objective. A 560 nm line was used for excitation, and an emission range between 583 and 709 nm was used for detection.

For analysis of cell division in the QC, root tips of 5-day-old seedlings were fixed in 3:1 ethanol:acetic acid for 15 min, incubated in 70% ethanol for another 15 min, washed in water for 2 min, cleared in chlorohydrate solution, and observed immediately using a Nikon E800 photomicroscope with a Nikon Plan Apo 100×/1.40 oil-immersion objective (http://www.nikon.com/) using differential interference contrast (DIC) optics. To visualize starch granules in the root tips, 5-day-old seedlings were stained with Lugol’s solution (Sigma, http://www.sigmaaldrich.com/) for 5 min, mounted on slides with chlorohydrate solution (8:2 chlorohydrate:water), and photographed immediately using the same Nikon DIC microscope.

RNA extraction, quantitative RT-PCR and NanoString nCounter gene expression analysis

Seedlings were grown on MS medium and then treated with either DMSO or BA for the indicated times. For RNA extractions, cytokinin treatment was terminated by transferring tissue to RNAlater Solution (Ambion, http://www.ambion.com/). Root samples (the last 0.5 mm of the root tip) were collected under a Leica dissection microscope (http://www.leica.com/), and subjected to total RNA extraction using an RNeasy Plus kit (Qiagen, http://www.qiagen.com/).

cDNA was prepared from the total RNA using SuperScript III reverse transcriptase (Invitrogen, http://www.invitrogen.com/) as described by the manufacturer. Quantitative RT-PCR was performed using SYBR Premix Ex Taq polymerase (TaKaRa, http://www.takara-bio.com/) in a DNA Engine OPTICON 2 (MJ Research, http://www.bio-rad.com). The following primers were used: 5′-AGAGGTTGACGAGCAGATGA-3′ and 5′-ACCAATGAAAGTAGACGCCA-3′ for TUB4, 5′-GAGCAAGGGCGAGGAGCTGTTC-3′ and 5′-TGGTGCAGATGAACTTCAGG-3′ for GFP, 5′-TCTCTTCTTGTAAAGTGACGACTG-3′ and 5′-TCAAATTCACCTTCAAATCCTT-3′ for ARR7 (5′), 5′-AAACCGGTGAAGCTAGCAGA-3′ and 5′-TCGTTTTGAACATGAAGAGTCC-3′ for ARR7 (3′), and 5′-GAGATTGCTTAAGATCTCTGGTTG-3′ and 5′-CAAATCCTTAAGACCAGAAGATCC-3′ for ARR15. At least two biological samples each were analysed with three technical replicates. The relative expression for PIN4–GFP (normalized to β-tubulin as a reference gene and to the wild-type grown on DMSO as a control sample) and standard errors were determined using REST 2009 software (Qiagen).

The Nanostring nCounter gene expression analysis was performed as described previously (Geiss et al., 2008) by the University of North Carolina Genomics and Bioinformatics Core Facilities using 160 ng of total RNA extracted from 0.5 mm root tips. The Nanostring probes targeting PIN1, PIN2, PIN3, PIN4, PIN7 and SHY2, and the control probes for TUB4 (At5g44340), UBQ10 (At4g05320) and APT1 (At1g27450), were designed and synthesized by NanoString Technologies (http://www.nanostring.com). The mRNA counts for PIN1, PIN2, PIN3, PIN4, PIN7 and SHY2 were normalized to the reference genes TUB4, UBQ10 and APT1 as described in the nCounter Gene Expression Assay Manual (http://www.nanostring.com/uploads/Manual_Gene_Expression_Assay.pdf/).

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

We thank Tony Perdue for excellent help with confocal microscopy, Jiri Friml (University of Gent, Department of Plant Systems Biology) for the PIN–GFP lines and the PIN antibody, Bruno Muller (University of Zürich, Institute of Plant Biology) for the ARR7::GFP and ARR15::GFP lines, and Yan Shi and Mike Topal at the University of North Carolina Genomics Core for help with the NanoStrings analysis. This project was supported by National Science Foundation grant IOS-1022053 to J.J.K. and G.E.S.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Figure S1. (a) An unrooted phylogenetic tree of type-A ARRs. Full-length protein sequences of the type-A ARRs were aligned using the Clustal W program (http://workbench.sdsc.edu/). The phylogenetic tree was constructed with 1000 boot-strapping replicates, and presented in TreeView. The bootstrap values are indicated on the tree. Scale bar represents 0.1 amino acid substitution per site. (b) Cartoon of the positions of T-DNA insertions in arr7 and arr15 mutants. (c) The expression patterns of ARR7 and ARR15 in the root tips revealed by promoter::GFP reporter constructs. (d) Relative expression level of ARR7 and ARR15 in 10-day-old seedlings of arr3,4,5,6,7,8,9,15 mutants as determined by quantitative RT-PCR. Error bars represent SE from three replicates. ARR7 (5′) represents the 5′ region of the transcript upstream of the arr7 T-DNA insertion site. ARR7 (3′) represents the 3′ region of the transcript downstream of the arr7 T-DNA insertion site. (e) Upper panel: pollen produced by WT, arr7/ARR7 and arr15/ARR15 plants. Lower panel: the percentage of irregular pollen produced by WT, arr7/ARR7, arr7/arr7, arr15/ARR15 and arr15/arr15.

Figure S2. Quantification of GFP fluorescence in the root tips of wild type and arr3,4,5,6,7,8,9,15 mutants. (a-c) Quantification of the GFP fluorescence in the root tips of wild type and arr3,4,5,6,7,8,9,15 mutants treated with 5 µm BA for the specified time or DMSO for 24 h as a zero control. Mean gray value was determined by ImageJ and defined as the sum of the gray values of all the pixels in the selection divided by the number of total pixels (see Methods) (a) PIN1-GFP in the stele. (b) PIN3-GFP in the stele. (c) DR5::GFP in the QC.

Figure S3. The effects of cytokinin on the level of PIN2 transcripts in wild-type and arr3,4,5,6,7,8,9,15 mutant root tips. The normalized counts of PIN2 transcripts in 0.5 mm root tips from wild-type (WT) and arr3,4,5,6,7,8,9,15 mutant seedlings treated with DMSO or 5 µm BA for 8 or 24 h as determined using the Nanostring n Counter gene expression analysis. Error bars represent SEM from three biological replicates.

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_4668_sm_FigS1.eps20963KSupporting info item
TPJ_4668_sm_FigS2.eps543KSupporting info item
TPJ_4668_sm_FigS3.eps756KSupporting 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.