An Arabidopsis kinase cascade influences auxin-responsive cell expansion

SUMMARY

Mitogen-activated protein kinase (MPK) cascades are conserved mechanisms of signal transduction across eukaryotes. Despite the importance of MPK proteins in signaling events, specific roles for many Arabidopsis MPK proteins remain unknown. Multiple studies have suggested roles for MPK signaling in a variety of auxin-related processes. To identify MPK proteins with roles in auxin response, we screened mpk insertional alleles and identified mpk1-1 as a mutant that displays hypersensitivity in auxin-responsive cell expansion assays. Further, mutants defective in the upstream MAP kinase kinase MKK3 also display hypersensitivity in auxin-responsive cell expansion assays, suggesting that this MPK cascade affects auxin-influenced cell expansion. We further found that MPK1 interacts with and phosphorylates ROP BINDING PROTEIN KINASE 1 (RBK1), a protein kinase that interacts with members of the Rho-like GTPases from Plants (ROP) small GTPase family. Similar to mpk1-1 and mkk3-1 mutants, rbk1 insertional mutants display auxin hypersensitivity, consistent with a possible role for RBK1 downstream of MPK1 in influencing auxin-responsive cell expansion. We found that RBK1 directly phosphorylates ROP4 and ROP6, consistent with the possibility that RBK1 effects on auxin-responsive cell expansion are mediated through phosphorylation-dependent modulation of ROP activity. Our data suggest a MKK3 • MPK1 • RBK1 phosphorylation cascade that may provide a dynamic module for altering cell expansion.

INTRODUCTION

Mitogen-activated protein (MAP) kinase signaling, a conserved mechanism of protein modification via phosphorylation across eukaryotes, plays multiple roles in plant development and defense signaling (reviewed in Xu and Zhang, 2015). MAPK signal transduction consists of reversible phosphorylation through a cascade of ATP-dependent protein kinases. The most upstream kinase in the cascade is typically a Ser/Thr MAP KINASE KINASE KINASE (MAPKKK), which phosphorylates a dual specificity MAP KINASE KINASE (MAPKK), which will in turn phosphorylate a MAP KINASE (MAPK or MPK) at Thr and Tyr residues in its activation loop. The terminal MAPK often phosphorylates downstream targets to change target activity or interaction partners. There are approximately 60 MAPKKKs, 10 MAPKKs, and 20 MAPKs in Arabidopsis thaliana (Ichimura et al., 2002). Combinations of these MAP kinases act to coordinate multiple aspects of plant growth, stress response, and pathogen interactions (reviewed in Xu and Zhang, 2015).

The plant hormone auxin influences cell division, cell expansion, and cell differentiation to affect nearly all aspects of plant growth and development (reviewed in Enders and Strader, 2015). Because auxin potently influences these processes, auxin responses are tightly modulated. Several studies suggest MAP kinase roles in auxin signaling; however, these roles have not been fully elucidated. For example, general MAP kinase activity increases in response to exogenous auxin treatment in Arabidopsis thaliana (Mockaitis and Howell, 2000) and the MAPKK inhibitor PD98059 alters the auxin-responsive gene transcription in Oryza sativa (Zhao et al., 2013), suggesting potential roles for MAPK signaling in mediating auxin responses. Specific components of these cascades implicated in auxin response include Arabidopsis thaliana MKK7, which negatively impacts PINOID protein localization (Dai et al., 2006), the tobacco MAPKKK NPK1, which acts to suppress early auxin-responsive gene transcription (Kovtun et al., 1998), the Arabidopsis thaliana MAPKKK MEKK1 and MPK4, which influence expression of several Aux/IAA genes in response to auxin treatment (Nakagami et al., 2006), and Arabidopsis thaliana MPK12, which negatively impacts auxin signaling through interactions with INDOLE-3-BUTYRIC ACID RESPONSE5 (IBR5) (Lee et al., 2009). Additionally, mutants defective in the dual-specificity MAP kinase phosphatase IBR5 display auxin resistance phenotypes (Monroe-Augustus et al., 2003; Strader et al., 2008a), suggesting that negative modulators of MAPK signaling also play roles in mediating auxin responses.

Here we report that Arabidopsis MPK1 likely negatively impacts auxin-responsive cell expansion. Mutants defective in MPK1 or in its upstream MAP kinase kinase, MKK3, are hypersensitive to the effects of auxin on root and cotyledon cell expansion. This effect may be mediated by the MPK1 target ROP-BINDING KINASE1 (RBK1); rbk1 mutants also display hypersensitivity to auxin in cell expansion assays. We further found that MPK1 phosphorylates RBK1 in vitro, and that RBK1 phosphorylates Rho-like GTPases from Plants4 (ROP4) and ROP6. Thus, MKK3 • MPK1 likely affects auxin-responsive cell expansion by modulating RBK1 and/or ROP activity.

RESULTS

MPK1 influences auxin-responsive cell expansion

To identify MAP kinases involved in auxin response, we screened MAPK T-DNA insertional alleles, obtained from the Arabidopsis Biological Resource Center, for altered auxin sensitivity in root elongation assays. We found that a mutant defective in MPK1 (Figures 1a, 1b, and 1c) displays hypersensitivity to the effects of the auxins indole-3-acetic acid (IAA), 2,4-dichlorophenoxy acetic acid (2,4-D), and 2,4-dichlorophenoxy butyric acid (2,4-DB) on inhibition of root elongation (Figures 1b, 1d, 1e, 1f, and S1). Because the effects of the synthetic auxin 2,4-DB were striking in these assays and in assays from our previous studies (Strader et al., 2008b; Strader and Bartel, 2009), we chose to focus on the effects of this synthetic auxin precursor for many of our assays. Although the mpk1-1 allele accumulates transcript downstream of the T-DNA insert (Figure 1c), it fails to accumulate detectable active MPK1 protein (Ortiz-Masia et al., 2007), and is likely a loss-of-function allele. mpk1-1 phenotypes are rescued by complementation with a construct driving expression of a C-terminal GFP fusion with MPK1 protein behind the MPK1 upstream regulatory region (MPK1:MPK1-GFP; Figures 1b, 1c, 1d, 1e, 1f, and 1h), suggesting that loss of MPK1 activity is causative for mpk1-1 auxin hypersensitivity. In addition, two independent mpk1-1 MPK1:MPK1-GFP complementation lines display an elongated hypocotyl phenotype in the absence of exogenous hormone (Figure 1b) and weak auxin resistance in root elongation assays (Figure 1b and 1d), raising the possibility that either MPK1 is misexpressed in these lines despite being driven behind the MPK1 upstream regulatory region or that the GFP tag leads to a slight mis-function of the fusion protein.

Figure 1. mpk1 displays hypersensitivity to auxin in root elongation and cotyledon expansion assays.

(a) Gene diagram of MPK1 depicting mpk1-1 T-DNA insert and primer locations.

(b) Image of 8-day-old Wt, mpk1, and mpk1 MPK1:MPK1-GFP seedlings grown under continuous white light at 22 °C on medium supplemented with ethanol or 1 μM 2,4-DB. The seedling with the longest root, the seedling with the shortest root, and two representatives of the population mean were imaged. Scale bar: 10 mm.

(c) Relative quantity of MPK1 transcript from of 8-day-old Wt, mpk1, and mpk1 MPK1:MPK1-GFP seedlings grown under continuous white light at 22 °C, determined by qPCR.

(d) Mean normalized primary root lengths (+SE; n ³ 15) of 8-day-old Wt, mpk1, and mpk1 MPK1:MPK1-GFP seedlings grown under continuous white light at 22 °C on medium supplemented with ethanol or 1.5 μM 2,4-DB. Non-normalized root length data is shown in Figure S1a.

(e) Mean normalized primary root lengths (+SE; n ³ 15) of 8-day-old Wt and mpk1 seedlings grown under yellow-filtered light at 22 °C on medium supplemented with ethanol, 60 nM 2,4-D, 1 μM 2,4-DB, or 80 nM IAA. Non-normalized root length data is shown in Figure S1b.

(f) Mean primary root cortex and epidermal cell lengths (+SE; n ³ 559 cells sampled from at least 8 seedlings) of 7-day-old Wt, mpk1, and mpk1 MPK1:MPK1-GFP seedlings grown under continuous white light at 22 °C on medium supplemented with ethanol or 1 μM 2,4-DB. Seedlings were imaged with a Leica Upright microscope using Nomarski optics. Cell lengths were measured using ImageJ.

(g) Mean cotyledon area (+SE; n ³ 20) of 8-d-old Wt, mpk1, and mpk1 MPK1:MPK1-GFP seedlings grown under continuous white light at 22 °C on medium supplemented with ethanol or 1 μM 2,4-DB. Cotyledons were excised, mounted, and imaged with a Leica dissecting microscope. Areas were measured using ImageJ.

(h) Confocal images of cotyledon epidermal pavement cells stained with propidium iodide from 7-day-old Wt, mpk1, and mpk1 MPK1:MPK1-GFP seedlings grown under continuous white light at 22 °C on medium supplemented with ethanol or 0.5 μM 2,4-DB. Scale bar: 100 μm.

Error bars represent SE of the means. Statistically significant differences from the Wt in two-tailed t-tests assuming unequal variance are indicated by an asterisk (P ≤ 0.02).

To determine the cellular effects of auxin in the mpk1-1 mutant, we closely examined the root cells of lines grown in the absence and presence of auxin. We found that epidermal and cortex root cell lengths in the elongation zone of mpk1-1 were significantly shorter than those of wild type in the presence of auxin, whereas mpk1-1 and wild type root cell lengths were indistinguishable under mock conditions (Figure 1f), consistent with the hypersensitivity of mpk1-1 to auxin when examined on a whole-root level (Figure 1d and 1e). Additionally, the mpk1-1 MPK1:MPK1-GFP complementation lines displayed longer root cell lengths compared to wild type when grown in the presence of auxin (Figure 1f), consistent with the mild auxin resistance observed in root elongation assays (Figure 1d).

Growing Arabidopsis seedlings in the presence of auxin results in reduced cotyledon expansion. Similar to the hypersensitivity displayed by mpk1-1 to auxin inhibition in roots, we found that mpk1-1 displays hypersensitivity to the effects of auxin on cotyledon expansion (Figure 1g). Further, the mpk1-1 MPK1:MPK1-GFP complementation lines displayed resistance to the inhibitory effects of auxin on cotyledon expansion (Figure 1g). Because Arabidopsis cotyledons grow primarily by cell expansion (Mansfield and Briarty, 1996), we examined wild type, mpk1-1, and rescue line cotyledon epidermal cell sizes from seedlings grown in the absence and presence of the auxinic compound 2,4-DB. We found that mpk1-1 displays smaller cotyledon cells than wild type when grown in the presence of auxin (Figure 1h).

Because MPK1 appears to repress auxin responsiveness, we hypothesized that it might act downstream of known auxin response factors. To examine the genetic interaction between mpk1 and mutants defective in the auxin receptor TRANSPORT INHIBITOR RESPONSE1 (TIR1), the AUXIN RESPONSE1 (AXR1) component of the RUB-activating enzyme, the AUXIN RESISTANT1 (AUX1) auxin import carrier, and the MAP kinase phosphatase INDOLE-3-BUTYRIC ACID RESPONSE5 (IBR5), we crossed mpk1-1 to tir1-1, axr1-3, aux1-7, and ibr5-1 and examined the auxin responsive root elongation of the resultant double mutants. We found that mpk1-1 failed to suppress the auxin resistant root elongation displayed by tir1-1, axr1-3, aux1-7, or ibr5-1 (Figure S2), suggesting that MPK1 does not act downstream of any of these components of auxin response. Conversely, TIR1, AXR1, AUX1, and IBR5 may genetically act downstream of MPK1, as these double mutants display auxin resistance similar to the resistant parent in each case (Figure S2).

MPK1 is an active kinase

The MAP kinase activation loop acts as a “gate” to allow substrate access to the enzyme active site (reviewed in Adams, 2001). Phosphorylation of key residues on the activation loop results in loop opening, allowing substrate access. In its unphosphorylated form, the MAPK loop remains mostly closed, restricting substrate access. Typically, the upstream MAPKK activates its downstream MAPK by phosphorylating these activation loop residues (Payne et al., 1991). To assess activity of heterologously-expressed MPK1, we identified the predicted phosphorylation sites on the MPK1 activation loop (Figure 2a). We then generated vectors to express wild-type GST-MPK1, GST-MPK1^T191E/Y193E (predicted to be constitutively active) and GST-MPK1^T191A/Y193A (predicted to be kinase dead) in E. coli. Using purified protein, we characterized kinase activity of each of these variants using the generic kinase substrate myelin basic protein (MBP) using in vitro assays with [γ-³²P]ATP. Similar to previous reports (Mizoguchi et al., 1994; Takahashi et al., 2007), we found that GST-MPK1 auto-activated and also displayed kinase activity using MBP as a substrate (Figure 2b). The MPK1 T191E/Y193E amino acid changes mimic phosphorylation of the MPK1 activation loop, likely leading to a constitutively active MPK1 protein (Teige et al., 2004; Asai et al., 2002). Similar to GST-MPK1, GST-MPK1^T191E/Y193E auto-phosphorylated and also phosphorylated MBP (Figure 2b), confirming that the heterologously-expressed wild type MPK1 protein and its phosphomimic variant are active kinases in vitro. Conversely, the GST-MPK1^T191A/Y193A variant carries mutations that prevent activation loop phosphorylation and MPK1 activation and fails to auto-phosphorylate or phosphorylate MBP (Figure 2), indicating that phosphorylation of the activation loop residues may be important for full kinase activity. Thus, MPK1 is an active kinase requiring active loop phosphorylation for full activity.

Figure 2. Heterologously expressed GST-MPK1 is an active kinase.

(a) Diagram of MPK1 protein. Mutations to make the kinase dead and constuitively active MPK1 variants are indicated.

(b) Autoradiogram of GST-MPK1, GST-MPK1^T191E/Y193E (constituitively active), and GST-MPK1^T191A/Y193A (kinase dead) phosphorylation activity using MBP as a substrate. Purified proteins from E. coli were incubated in kinase buffer and 1 μCi of [γ-³²P]ATP in the presence or absence of the general kinase substrate MBP. Reactions were stopped, proteins separated by SDS-PAGE, and activity assessed by autoradiography.

mkk3 is hypersensitive to auxin

The MAP KINASE KINASE MKK3 interacts with MPK1 in yeast two-hybrid assays (Ichimura et al., 1998; Dóczi et al., 2007; Matsuoka et al., 2015). Additionally, active MKK3 phosphorylates and activates MPK1 in vitro (Takahashi et al., 2007), in Arabidopsis protoplasts ( Dóczi et al., 2007; Danquah et al., 2015), and in tobacco leaves (Hwa and Yang, 2008), suggesting that MKK3 is the MAP kinase kinase that regulates MPK1 activity. To determine whether the MKK3 • MPK1 MAPK cascade is involved in auxin response, we examined auxin responsiveness of the mkk3-1 insertional mutant, which displays reduced MKK3 transcript (Dóczi et al., 2007; Lee, 2015). Indeed, mkk3-1 displays phenotypes similar to those of mpk1-1 (Figure 3), consistent with the possibility that MKK3 and MPK1 may act in a single pathway to regulate auxin-responsive cell expansion. The mkk3-1 mutant displays hypersensitivity to the effects of the auxins IAA, 2,4-D, and 2,4-DB on inhibition of root elongation (Figures 3b and 3c). To determine if this phenotype is caused by altered cell expansion, we closely examined root cell lengths of wild type and mkk3-1, both in the presence and absence of auxin. Similar to mpk1-1, root cell lengths of mkk3-1 are significantly shorter than those of the wild-type in response to auxin treatment (Figure 3d). However, in contrast to mpk1, mkk3-1 displays slightly longer root cells than wild type in the absence of auxin (Figure 3d). Additionally, mkk3-1, like mpk1, is hypersensitive to the effects of auxin on the inhibition of cotyledon expansion (Figure 3e). The phenotypes of mkk3-1 further support roles for the MKK3 • MPK1 cascade in negatively regulating auxin-responsive cell expansion. Further, MKK3 activates MPK2 (Takahashi et al., 2007), MPK6 (Takahashi et al., 2007), and MPK8 (Takahashi et al., 2011) in addition to MPK1, and the phenotypes we observe in mkk3-1 that differ from mpk1-1 phenotypes, such as longer starting root cell lengths, may be caused by misregulation of additional MKK3 targets.

Figure 3. mkk3 displays auxin hypersensitivity in root elongation and cotyledon expansion assays.

(a) Gene diagram of MKK3 showing location of the mkk3-1 T-DNA insertion.

(b) Image of 7-day-old Wt and mkk3-1 seedlings grown under continuous white light at 22 °C on medium supplemented with ethanol or 1 μM 2,4-DB. The seedling with the longest root, the seedling with the shortest root, and two representatives of the population mean were imaged. Scale bar: 10 mm.

(c) Mean normalized primary root lengths (+SE; n ³ 20) of 7-day-old Wt and mkk3-1 seedlings grown under yellow-filtered light at 22 °C on medium supplemented with ethanol, 80 nM IAA, 60 nM 2,4-D, or 0.5 μM 2,4-DB. Non-normalized root length data is shown in Figure S1c.

(d) Mean root cortex and epidermal cell lengths (+SE; n ³ 646 cells sampled from at least 10 seedlings) of 7-day-old Wt and mkk3-1 seedlings grown under continuous white light at 22 °C on medium supplemented with ethanol or 1 μM 2,4-DB. Seedlings were imaged with a Leica Upright microscope using Nomarski optics. Cell lengths were measured using ImageJ.

(e) Mean normalized cotyledon area (+SE; n ³ 14) of 8-day old Wt and mkk3-1 seedlings grown under continuous white light at 22 °C on medium supplemented with ethanol or 1.5 μM 2,4-DB. Cotyledons were excised, mounted, and imaged with a Leica dissecting microscope. Cotyledon areas were measured using ImageJ. Non-normalized root length data is shown in Figure S1d.

Error bars represent SE of the means. Statistically significant differences from the Wt in two-tailed t-tests assuming unequal variance are indicated by an asterisk (P ≤ 0.0006).

RBK1 affects auxin-responsive cell expansion

In a protein microarray experiment, MPK1 phosphorylated ROP BINDING PROTEIN KINASE1 (RBK1) (Popescu et al., 2009). To investigate the possible role of RBK1 in the MAPK cascade influencing auxin responses, we examined the auxin responsiveness of two rbk1 insertional alleles (Figure 4a). The rbk1-1 insertion lies upstream of the start codon (Figure 4a) and may lead to a decrease in RBK1 transcript whereas the rbk1-2 insert affects RBK1 gene splicing, resulting in retention of the fifth intron, based on sequencing of the PCR product resultant from amplification with primers flanking this region (Figure 4b). If translated, the resultant rbk1-2 protein would carry a premature stop codon after amino acid 320, suggesting that this allele has a loss of RBK1 function. Similar to mpk1-1 and mkk3-1, these two rbk1 alleles display hypersensitivity to the effects of the auxins IAA, 2,4-D, and 2,4-DB on root elongation inhibition (Figures 4c, 4d, and 4e). Both rbk1 insertional alleles display similar phenotypes, suggesting that the auxin hypersensitive phenotypes are caused by loss of RBK1 function. To ascertain if the rbk1 auxin-responsive inhibition of root elongation phenotypes are caused by cell expansion effects, we closely examined root cell lengths of wild type, rbk1-1, and rbk1-2 seedlings in the absence and presence of auxin. The auxin-responsive inhibition of root elongation phenotypes of the rbk1 alleles are likely due to defects in cell expansion, because root cell lengths are shorter than those of wild type in response to auxin (Figure 4e). Additionally, unlike mpk1 and mkk3, both rbk1 alleles displayed shorter root cell lengths than wild type in the absence of auxin (Figure 4e). Similar to mpk1-1 and mkk3-1, both rbk1-1 and rbk1-2 display hypersensitivity to the effects of auxin inhibition of cotyledon expansion (Figure 4f). Because rbk1 mutants display similar phenotypes to both mkk3-1 and mpk1, and MPK1 has been suggested to phosphorylate RBK1, RBK1 may act in the MKK3 • MPK1 signaling cascade to influence auxin-responsive cell expansion.

Figure 4. rbk1 displays auxin hypersensitivity in root elongation and cotyledon expansion assays.

(a) Gene diagram of RBK1 showing rbk1-1 and rbk1-2 T-DNA insertion and primer locations.

(b) RT-PCR on cDNA synthesized from RNA isolated from 8-day-old Wt, rbk1-1, and rbk1-2 seedlings. Primers used are indicated in gene diagram in 4A. The genomic RBK1 fragment is 376-bp, the alternative-spliced fragment is 280-bp, and the cDNA fragment is 195-bp.

(c) Image of 8-day-old Wt, rbk1-1, and rbk1-2 seedlings grown under continuous white light at 22 °C on medium supplemented with ethanol or 1 μM 2,4-DB. The seedling with the longest root, the seedling with the shortest root, and two representatives of the population mean were imaged. Scale bar: 10 mm.

(d) Mean normalized primary root lengths (+SE; n ³ 15) of 8-day-old Wt, rbk1-1, and rbk1-2 seedlings grown under continuous white light at 22 °C on medium supplemented with ethanol, 1.5 μM 2,4-DB, or 100 nM 2,4-D. Non-normalized root length data is shown in Figure S1e.

(e) Mean primary root lengths (+SE; n ³ 20) of 7-day-old Wt, rbk1-1, rbk1-2 seedlings grown under yellow-filtered light at 22 °C on medium supplemented with ethanol, 80 nM IAA, or 60 nM 2,4-D.

(f) Mean root cortex and epidermal cell lengths (+SE; n ³ 443 cells sampled from at least 13 seedlings) of 7-day-old Wt, rbk1-1, and rbk1-2 seedlings grown under continuous white light at 22 °C on medium supplemented with ethanol or 1 μM 2,4-DB. Seedlings were imaged with a Leica Upright microscope using Nomarski optics. Cell lengths were imaged in ImageJ.

(g) Mean normalized cotyledon area (+SE; n ³ 13) of 8-day-old Wt, rbk1-1, and rbk1-2 seedlings grown under continuous white light on medium supplemented with ethanol or 1.5 μM 2,4-DB. Cotyledons were excised, mounted, and imaged with a Leica dissecting microscope. Areas were measured using ImageJ. Non-normalized root length data is shown in Figure S1f.

Error bars represent SE of the means. Statistically significant differences from the Wt in two-tailed t-tests assuming unequal variance are indicated by an asterisk (P ≤ 0.02).

RBK1 is an active kinase

RBK1 is a predicted Ser/Thr protein kinase and a member of Group VI of the Receptor Like Cytoplasmic Kinase family in Arabidopsis (Shiu et al., 2004). Although RBK1 interacts with ROP4 (Molendijk et al., 2008), no substrates of RBK1 are known, and no rbk1 phenotypes have previously been described. To determine if RBK1 is an active kinase, we tested heterologously-expressed His-RBK1 activity by analyzing its ability to phosphorylate MBP in an in vitro kinase assay with [γ-³²P]ATP. We found that His-RBK1 auto-phosphorylates and also phosphorylates the generic kinase substrate MBP (Figure 5b). Often, mutating specific lysine and aspartate kinase residues disrupts its catalytic function and thus its ability to phosphorylate substrates (reviewed in Adams, 2001). We therefore examined the RBK1 sequence with ScanProsite (de Castro et al., 2006) and identified Lysine 181, Lysine 182, and Aspartic Acid 278 as likely candidate residues for participation in the RBK1 catalytic triad (Figure 5a). To assess the requirement for these residues for catalytic activity, we examined kinase activity of heterologously-expressed His-RBK1^{K181M/K182R/D278A} and found that it fails to auto-phosphorylate or to phosphorylate MBP (Figure 5b). These results demonstrate that RBK1 is an active kinase and that Lysine 181, Lysine 182, and Aspartic Acid 278 are necessary for its kinase activity.

Figure 5. Heterologously expressed His-RBK1 is an active kinase.

(a) Diagram of RBK1 protein. Mutations to create a kinase-dead RBK1 variant are indicated.

(b) Autoradiogram of His-RBK1 and His-RBK1^{K181M/K182R/D278A} kinase activity on MBP. Purified proteins from E. coli were incubated in kinase buffer and 1 μCi of [γ-³²P]ATP in the presence or absence of the general kinase substrate MBP. Reactions were stopped, proteins separated by SDS-PAGE, and activity assessed by autoradiography.

MPK1 interacts with and phosphorylates RBK1

MAPK substrates typically contain a docking site (or D-motif) for MAPK interaction, consisting of [K/R]_1-2-X_2-6-[L/I]-X-[L/I] (Bardwell et al., 2009). Common docking (CD) domains at the C-terminal end of MAPKs interact with these D domains (Bardwell et al., 2009). Eukaryotic Linear Motif analysis (Dinkel et al., 2015) identified a predicted D-domain and multiple predicted phosphorylation sites in the RBK1 protein, consistent with the possibility that RBK1 interacts with a MAP kinase. A protein microarray study identified 148 proteins putatively phosphorylated by MPK1, including RBK1 (Popescu et al., 2009). To verify MPK1 interaction with RBK1, we investigated whether MPK1 and RBK1 could interact in yeast through a yeast two-hybrid assay (Figure 6a). Previous studies (Hannig et al., 1994; Shiozaki and Russell, 1995) have demonstrated that catalytically inactive versions of MAP kinases bind substrates tightly. Indeed, using the kinase dead form of MPK1, we observed an interaction between MPK1^T191A/Y193A and RBK1, confirming the interaction results from the protein microarray study (Popescu et al., 2009). These results suggest that MPK1 and RBK1 may have a direct association.

Figure 6. MPK1 uses RBK1 as a substrate and RBK1 uses ROP proteins as substrates.

(a) Yeast two-hybrid assays were conducted with MPK1^T191A/Y193A (kinase dead) and RBK1 fused to the GAL4-DNA binding domain (DBD) or the GAL4 activation domain (AD). Serial dilutions were spotted on permissive plates (+His) and plates on which growth requires protein interaction (-His + 3-AT) and imaged after five days of growth at 30 °C.

(b) MPK1 uses RBK1 as a substrate. Heterologously-expressed MPK1^T191E/Y193E (constitutively active) and MPK1^T191A/Y193A (kinase dead) proteins were incubated in kinase buffer and 1 μCi of [γ-³²P]ATP with either RBK1 or RBK1^{K181M/K182R/D278A} (kinase dead) proteins. Reactions were stopped, proteins separated by SDS-PAGE, and activity assessed by autoradiography.

(c) Normalized primary root lengths of 8-day-old Wt, rbk1-2, mpk1-1, and mpk1-1 rbk1-2 seedlings grown under continuous yellow light at 22 °C on medium supplemented with ethanol, or the indicated concentration of 2,4-DB. Error bars represent SE of the means

(d) His-RBK1 phosphorylates His-ROP4. Heterologously-expressed RBK1 and RBK1^{K181M/K182R/D278A} (kinase dead) proteins were incubated in kinase buffer and 1 μCi of [γ-³²P]ATP in the presence or absence of His-ROP4. Reactions were stopped, proteins separated by SDS-PAGE, and activity assessed by autoradiography.

(e) His-RBK1 phosphorylates His-ROP6. Heterologously-expressed RBK1 and RBK1^{K181M/K182R/D278A} (kinase dead) proteins were incubated in kinase buffer and 1 μCi of [γ-³²P]ATP in the presence or absence of His-ROP6 or in the presence or absence of the general kinase substrate MBP, as indicated. Reactions were stopped, proteins separated by SDS-PAGE, and activity assessed by autoradiography.

Because mpk1 and rbk1 mutants display similar phenotypes (Figures 1 and 4) and because MPK1 may interact with RBK1, we hypothesized that RBK1 serves as an MPK1 phosphorylation target. We therefore tested GST-MPK1^T191E/Y193E (constitutively active) and GST-MPK1^T191A/Y193A (kinase dead) phosphorylation of heterologously-expressed His-RBK1^{K181M/K182R/D278A} to determine whether RBK1 can serve as an MPK1 substrate (Figure 6b). We found that GST-MPK1^T191E/Y193E (constitutively active), but not GST-MPK1^T191A/Y193A (kinase dead), phosphorylated His-RBK1^{K181M/K182R/D278A} (kinase dead) in a manner dependent on its activation loop (Figure 6b). Conversely, His-RBK1 failed to phosphorylate GST-MPK1^T191A/Y193A and only His-RBK1 autophosphorylation was observed (Figure 6b). This data suggests that MPK1 phosphorylates RBK1, but RBK1 is unable to phosphorylate MPK1^T191A/Y193A, either because MPK1 is not a substrate of RBK1 or because RBK1 specifically phosphorylates the T191 or Y193 residues of MPK1. These data are consistent with RBK1 serving as an MPK1 substrate, and with MPK1 unlikely to be an RBK1 substrate.

Because rbk1 mutants display similar phenotypes to mpk1, and because MPK1 phosphorylates RBK1, we hypothesize that RBK1 may act in the MKK3 • MPK1 signaling cascade to influence auxin-responsive cell expansion. To genetically test whether MPK1 and RBK1 act in a single pathway to influence auxin responsiveness, we examined root elongation responses of mpk1-1 rbk1-2 to the synthetic auxin precursor 2,4-DB and found that the double mutant displayed similar levels of hypersensitivity as either parental line (Figure 6c). Because mpk1 and rbk1 phenotypes are not additive in this assay, MPK1 and RBK1 likely act in the same pathway. Further, our data that MPK1 directly phosphorylates RBK1 suggests that MPK1 acts upstream of RBK1 in this pathway.

RBK1 phosphorylates ROP4 and ROP6

ROP proteins are small GTPases impacting diverse aspects of plant development and responses to the environment. ROP proteins play roles in cell polarity, cell expansion, and tip growth (Nibau et al., 2006). RBK1 was identified as a protein that interacts with the small GTPase ROP4 in yeast two-hybrid, bimolecular fluorescence complementation, and in vitro pull-down assays, independent of the ROP4 activation state (Molendijk et al., 2008); however, whether RBK1 uses ROP proteins as substrates was unknown. We therefore tested heterologously-expressed His-RBK1 kinase activity on His-ROP4 and His-ROP6 (Figures 6d and 6e). Indeed, we found that His-RBK1 phosphorylates both these ROP proteins in a manner dependent on the RBK1 catalytic triad (Figures 6e and 6e), similar to RBK1 activity on the generic kinase substrate MBP (Figures 5b and 6e). Because loss of RBK1 function results in increased auxin responsiveness (Figure 4) and loss of ROP function leads to decreased auxin responsiveness (Xu et al., 2010; Robert et al., 2010; Chen et al., 2012; Huang et al., 2014), we hypothesize that ROP phosphorylation by RBK1 could negatively impact ROP function (Figure 7).

Figure 7. Proposed mechanisms for MAPK/RBK1 influences on ROP activity.

Interactions between MKK3, MPK1, RBK1, and ROP proteins, combined with the altered auxin responsiveness of mutants defective in these signaling components, suggests a pathway for modulating of auxin-responsive cell expansion. In this pathway, MKK3 phosphorylates and activates MPK1, MPK1 phosphorylates and activates RBK1, and RBK1 phosphorylates ROP, likely leading to altered ROP activity. We hypothesize four potential mechanisms of phosphorylation-mediated regulation of ROP activity: 1) Phosphorylated ROP displays altered affinity for RIC effector proteins, leading to downstream effects such as altered cytoskeleton organization; 2) ROP phosphorylated increases ROP affinity for GDIs, leading to decreased ROP localization at the plasma membrane; 3) Phosphorylated ROP displays increased affinity for GAPs, leading to an accumulation of the inactive GDP-bound form of ROP proteins; 4) Phosphorylated ROP displays decreased affinity for ROP GEFs, leading to an accumulation of the inactive GDP-bound form of ROP.

Interestingly, neither His-RBK1 nor GST-MPK1^T191E/Y193E were able to phosphorylate a monomeric version of truncated His-AUXIN RESPONSE FACTOR7 (His-ARF7) (Korasick et al., 2014) in our assay (Figure S3), suggesting these proteins may have roles outside of the major TIR1 auxin response pathway. This result also demonstrates that our in vitro kinase assay is capable of determining the specificity of the kinase activity of GST-MPK1^T191E/Y193E and His-RBK1, in that these proteins do not phosphorylate any substrate they are provided.

Additionally, we investigated auxin-responsive root elongation in the rop2i rop4-1 double mutant (Figure S4), which displays altered auxin-responsive pavement cell morphology (Fu et al., 2005). In these assays, we did not observe any dramatic alteration in auxin responsiveness, although the rop2i rop4-1 seedlings were slightly resistant to the inhibitory effects of 2,4-DB on root elongation. Given the redundancy of ROP activity and the high similarity among ROP family members, this negative data is difficult to interpret. To explore additional physiological significance of the proposed pathway, we examined root gravitropism responses in mpk1-1, mkk3-1, rbk1-2, mpk1-1 rbk1-2, and rop2i rop4-1 seedlings. Each mutant displayed a response similar to wild type (Figure S5), suggesting these genes do not have roles in influencing gravitropic responses, at least in our assay.

DISCUSSION

ROP proteins influence actin cytoskeleton polymerization and microtubule alignment to dynamically impact plant cell growth and formation of the final shape of cells (reviewed in Chen and Friml, 2014 and Lin et al., 2015). ROP2 activation leads to actin cytoskeleton polymerization through the activity of ROP-INTERACTIVE CRIB MOTIF-CONTAINING PROTEIN4 (RIC4) (Fu et al., 2005; Xu et al., 2010). Consistent with its roles in actin polymerization, ROP2 overexpression results in increased root hair elongation and misplaced root hairs, suggesting roles for ROP2 in influencing cell expansion and/or tip growth in these cells (Jones et al., 2002). In addition, ROP6 activation alters the alignment of microtubules, through the action of RIC1 (Fu et al., 2005; Fu et al., 2009; Xu et al., 2010). These findings demonstrate the significance of ROPs in the arrangement of cytoskeleton components and their downstream effects on developmental events.

Several lines of evidence link ROP activity to auxin responses. For example, mutants defective in ROP3 (Huang et al., 2014) or in the ROP effector RIC1 (Choi et al., 2013) display resistance to effects of auxin on inhibition of root elongation. Further, auxin activates tobacco Rac1, a positive regulator of auxin responsive gene expression (Tao et al., 2002). Overexpression of constitutively active Arabidopsis ROP2 results in increased auxin responsive lateral root initiation whereas overexpression of dominant negative ROP2 results in decreased auxin-responsive lateral root initiation (Li et al., 2001), suggesting some ROP proteins may positively influence auxin-responsive cell division. Additionally, ROP9 RNAi lines are hypersensitive to auxin effects on root elongation inhibition and lateral root promotion (Nibau et al., 2013), suggesting some ROP proteins may negatively impact auxin responses. Recently, Li et al. (2017) demonstrated that TARGET OF RAPAMYCIN (TOR) is activated by auxin, and ROP2 is involved in downstream signaling of TOR. Our data suggest MPK1 and RBK1 may act upstream of at least some ROP proteins to influence auxin responses through phosphorylation-mediated regulation of ROP activity (Figure 7).

In human cells, Rho phosphorylation increases interactions with GUANINE NUCLEOTIDE DISSOCIATION INHIBITORS (GDIs) and thus alters Rho subcellular localization, inhibits GTP-binding activity, and increases protein stability (reviewed in Loirand et al., 2006). Likewise, phosphorylation has been proposed as a regulator of plant ROP activity, either by direct phosphorylation of ROP proteins, or indirectly by phosphorylation of the ROP regulators, such as GTPASE ACTIVATING PROTEINS (GAPs), GUANINE EXCHANGE FACTORS (GEFs), and GUANINE NUCLEOTIDE DISSOCIATION INHIBITORS (GDIs) (reviewed in Fehér and Lajkó, 2015). Similar to phosphorylation effects on Rho proteins, ROP protein phosphorylation may influence their activity by altering GTP-binding by ROP, by altering ROP subcellular localization, or by affecting ROP protein interactions (Figure 7). Multiple kinase families have been implicated in phosphorylation of ROP proteins or ROP protein regulators, including AGC-type kinases, calcium-dependent protein kinases, MAP kinases, receptor-like kinases, and other membrane-bound kinases (reviewed in Fehér and Lajkó, 2015). In this study, we identified MKK3 • MPK1 • RBK1 as a potential kinase cascade that terminates in ROP protein phosphorylation. Molendijk et al. (2008) suggested that RBK1 may act downstream of ROP proteins, because they did not observe phosphorylation of ROP proteins by RBK1 in a kinase assay. Our data suggest that RBK1 likely acts upstream of ROPs, as evidenced by His-RBK1 phosphorylation of both ROP4 and ROP6. Because mutants defective in MKK3, MPK1, and RBK1 display hypersensitivity to the effects of auxin (Figures 1, 3, and 4) and mutants defective in ROP4 or ROP6 display resistance to auxin (Xu et al., 2010; Robert et al., 2010; Chen et al., 2012), we hypothesize the RBK1-mediated phosphorylation may act to negatively influence ROP4 and ROP6 activity (Figure 7). Further, additional kinases, including AGC kinases, calcium-dependent protein kinases, MAP kinases, receptor-like kinases, and other membrane-bound kinases, likely also act to modulate ROP activity, either through phosphorylation of ROPs or by phosphorylation of ROP regulators (Figure 7; reviewed in Fehér and Lajkó, 2015). Additional in planta experiments will allow researchers to tease apart roles for each of these kinases on ROP activity.

MPK1 activation by multiple hormones and stresses (Ortiz-Masia et al., 2007; Hwa and Yang, 2008; Danquah et al., 2015) likely reflects a wide range of potential substrates for MPK1. Further, MPK1 effects on cell expansion may be caused by an accumulation of effects on multiple substrates and not limited to RBK1-mediated effects. Similarly, RBK1 transcript is upregulated in response to Phytophthora infestans and Botrytis cinerea infections (Molendijk et al., 2008), suggesting that RBK1, and perhaps ROP phosphorylation, may have multiple roles in plant growth and stress response. Whether downstream responses to these infections are impacted by particular ROP proteins, or perhaps other RBK1substrates, remains unknown.

Roles for the Arabidopsis Group C MAP kinases in Arabidopsis, consisting of MPK1, MPK2, MPK7, and MPK14 (Ichimura et al., 2002), have not been well delineated. These Group C MAP kinases contain a modified Common Docking (CD) domain compared to the fully conserved CD domain in other kinases (Ichimura et al., 2002; Dóczi et al., 2007), making this clade unique among MAP kinases. Kinase activity of MPK1, MPK2, and MPK7 is upregulated in response to multiple stress signals, including wounding, jasmonic acid, abscisic acid, and H₂O₂ (Dóczi et al., 2007; Ortiz-Masia et al., 2007). MPK2 also displays increased kinase activity in response to the synthetic auxin 2,4-D (Mizoguchi et al., 1994). Further, the MAP3Ks MAPKKK17 and MAPKKK18 are activated by ABA and are likely activators of MKK3 (Danquah et al., 2015; Matsuoka et al., 2015), which we have found to be involved in suppressing auxin responsiveness (Figure 3). In addition to MPK1 regulation by MKK3 (Dóczi et al., 2007; Takahashi et al., 2007; Hwa and Yang, 2008; Danquah et al., 2015), other kinases may act to regulate MPK1, either directly or indirectly. The kinase Snrk2, which plays roles in ABA responses (Boudsocq et al., 2007), affects MPK1 and MPK2 phosphorylation (Umezawa et al., 2013), suggesting a potential mechanism for intersections in ABA-auxin signaling. Delineating the complex phosphorylation patterns of these regulators will be important for understanding not only the signal transduction pathways affecting each of these responses, but may also uncover points of convergence amongst them.

Here, we describe identification of MPK1 as negatively influencing auxin-dependent cell expansion. We further identify RBK1 as an MPK1 substrate. RBK1 phosphorylation of ROP proteins may be the molecular basis for MPK1 impacts on auxin-responsive cell expansion. However, we do not know how MPK1-mediated phosphorylation of RBK1 affects its activity, nor do we understand how ROP protein phosphorylation by RBK1 affects ROP activity. Alternatively, MPK1 may indirectly effect auxin-responsive cell expansion by increasing activity of a separate, unidentified pathway which ultimately inhibits auxin-responsive cell expansion. Our mutant phenotypes of MPK1 and RBK1 suggest that MPK1 positively influences RBK1 activity (Figure 7), because both mpk1 and rbk1 mutants display hypersensitivity to the effects of auxin on cell expansion. We hypothesize that RBK1 likely negatively influences ROP proteins, because rbk1 mutants are hypersensitive to auxin (Figure 3), His-RBK1 can directly phosphorylate His-ROP4 and His-ROP6 (Figure 6), and phosphorylation of ROP4 negatively affects its activation (Fodor-Dunai et al., 2011). In this study, we have described a kinase cascade with ROP GTPases as the terminal target, potentially resulting in phosphorylation-mediated ROP regulation. Future work will be needed to address whether MKK3 • MPK1 • RBK1-mediated ROP phosphorylation affects ROP activity to affect auxin effects on cell expansion.

EXPERIMENTAL PROCEDURES

Seed source and genotyping

Insertional alleles disrupted in mpk1-1 (At1g10210; SALK_063847), mkk3-1 (At5g40440; SALK_051970), rbk1-1 (At5g10520; SALK_124882), and rbk1-2 (At5g10520; SALK_043441) were obtained from the Arabidopsis Biological Resource Center. Genotyping primers are listed in Table S1. The tir1-1, axr1-3, aux1-7, and ibr5-1 mutations as previously described (Strader et al., 2008a). PCR amplification using primers MPK1-7 and MPK1-8 results in a 539-bp product in wild type and no product in mpk1-1. PCR amplification using primers MPK1-8 and LB1-SALK results in no product in wild type and a ~550-bp product in mpk1-1. mkk3 was genotyped using primers described in (Yoo et al., 2008). PCR amplification using primers RBK1-14 and RBK1-17 results in a 371-bp product in wild type and no product in rbk1-1. PCR amplification using primers RBK1-17 and LB1-SALK results in no product in wild type and a ~550-bp product in rbk1-1. PCR amplification using primers RBK1-1 and RBK1-7 results in a 444-bp product in wild type and no product in rbk1-2. PCR amplification using primers RBK1-1 and LB1-SALK results in no product in wild type and a ~500-bp product in rbk1-2.

Phenotypic assays

Arabidopsis thaliana mutants were in the Columbia (Col-0) background, which was used as the wild type in all experiments. For phenotypic assays, seeds were surface sterilized (Last and Fink, 1988), stratified for 2 days at 4 °C, and plated on plant nutrient (PN) media (Haughn and Somerville, 1986) supplemented with 0.5% (w/v) sucrose (PNS) and solidified with 0.6% (w/v) agar. Seedlings were grown under continuous illumination at 22 °C.

To examine auxin-responsive root elongation, seeds were plated on PNS supplemented with ethanol (mock) or the indicated auxin and were grown at 22 °C under continuous white light or yellow-filtered light, and measured and imaged after 7 or 8 days of growth, as indicated. To examine the effects of auxin on root cell expansion, seeds were plated on PNS supplemented with ethanol (mock) or the indicated auxin and were grown at 22 °C under continuous white light or yellow-filtered light, imaged after 7 days of growth using a Leica DM6 equipped with Nomarski optics. Cell lengths in the epidermis, cortex, endodermis, and pericycle layer were measured in fully expanded root cells spanning from the differentiation zone to within a few cells from the base of the hypocotyl.

To examine cotyledon expansion, seeds were plated on PNS supplemented with ethanol (mock) or the indicated auxin and were grown at 22 °C under continuous white light. Cotyledons were excised after the indicated number of days of growth, mounted on slides and imaged with a Leica Axioplan dissecting scope. Cotyledon blade areas were measured using NIH Image software.

Confocal microscopy

To examine Arabidopsis pavement cells, cotyledons were excised at the petiole, incubated in 10 μg/mL propidium iodide, mounted in water, and imaged used a 20X objective on a Zeiss LSM510 laser scanning confocal microscope. Samples were excited with the 488-nm laser line from an argon laser, and the resultant fluorescence was split through a 545-nm secondary dichroic beamsplitter. Fluorescence was further filtered through a 565-615-nm band-pass filter.

Vector construction

MPK1:MPK1-GFP construction

The MPK1 genomic region was amplified from Col-0 genomic DNA (Thole et al., 2014) using Pfx Platinum Taq (Life Technologies) with caccMPK1-URR and MPK1g-nostop2. The resultant 2897-bp PCR product contained the MPK1 upstream region and entire MPK1 coding region (from –1693 to 1198, where the A of the ATG start codon is position 1) with the TGA stop codon replaced with a GGA codon encoding for Gly. This PCR product was captured into the pENTR/D-TOPO vector (Life Technologies). The pENTR-DTOPO-MPK1:MPK1nostop vector was linearized by digestion with MluI. The MPK1:MPK1nostop genomic region was recombined into the pMDC107 plasmid (Curtis and Grossniklaus, 2003) using LR Clonase (Life Technologies) to form MPK1:MPK1-GFP, which expresses a C-terminal GFP fusion with MPK1 driven by the region upstream of MPK1.

Protein expression constructs

MPK1 was amplified from the pda02860 cDNA using primers BamHI-MPK1-FOR and EcoRI-MPK1-REV using Pfx Platinum Taq (Life Technologies). The resultant 1125-bp product was captured into the pCR4 vector (Life Technologies). The insert was released using BamHI and EcoRI and ligated into pGEX-4TI to form pGEX-MPK1, which expresses the MPK1 protein with an N-terminal GST tag. RBK1 was amplified from the DKLAT5G10520 cDNA using EcoRI-HisRBK1-F and NotI-HisRBK1-R using Pfx Platinum Taq (Life Technologies). The resultant 1418-bp product was captured into the pCR4 vector (Life Technologies). The insert was released from pCR4-RBK1 using EcoRI and NotI and ligated into pET28a to form pET28a-RBK1, which expresses RBK1 with an N-terminal 6X-His tag. ROP4 was amplified from the U17137 cDNA using primers ROP4-F-His and ROP4-R-His and Pfx Platinum Taq (Life Technologies). The resultant 605-bp product was captured into the pCR4 vector (Life Technologies). The insert was released using EcoRI and NotI and ligated into pET28a to form pET28a-ROP4, which expresses the ROP4 with an N-terminal 6X-His tag. ROP6 was amplified from the S69209 cDNA using primers ROP6-F-His and ROP6-R-His and Pfx Platinum Taq (Life Technologies). The resultant 611-bp product was captured into the pCR4 vector (Life Technologies). The insert was released using EcoRI and NotI and ligated into pET28a, to form pET28a-ROP6, which expresses ROP6 with an N-terminal 6X-His tag.

Yeast two-hybrid constructs

MPK1 cDNA was amplified from pda02860 cDNA using MPK1-SalI and MPK1-NotI using Pfx Platinum Taq (Life Technologies). The resultant 1483-bp product was captured into the pCR4 vector (Life Technologies). The insert was released using SalI and NotI and ligated into the pBI770 and pBI771 yeast two-hybrid vectors (Kohalmi et al., 1998). RBK1 cDNA was amplified with RBK1-SalI and RBK1-NotI using Pfx Platinum Taq (Life Technologies). The resultant 1420-bp was captured into the pCR4 vector (Life Technologies). The insert was digested with SalI and NotI and ligated into the pBI770 and pBI771 vectors (Kohalmi et al., 1998).

Site directed mutagenesis

Using the QuikChange Lightning site-directed mutagenesis kit (Agilent) according to manufacturer’s instructions, the MPK1-T191A/Y193A-FOR and MPK1-T191A/Y193A-REV primers were used to create the MPK1^T191A/Y193A mutations in pGEX-MPK1, pBI770-MPK1, and pBI771-MPK1 vectors. Likewise, the MPK1-T191E/Y193E-FOR and MPK1-T191E/Y193E-REV primers were used to create the MPK1^T191E/Y193E mutations in pGEX-MPK1, pBI770-MPK1, and pBI771-MPK1 vectors. The RBK1-D278A-FOR and RBK1-D278A-REV primers were used to create the RBK1^D278A mutation in pET28a-RBK1 vector. The RBK1-K181M/K182R-FOR and RBK1-K181M/K182R-REV primers were used to create the RBK1^K181M/K182R mutations in pET28a-RBK1^D278A vector.

Plant transformation

The MPK1:MPK1-GFP plasmid was transformed into Agrobacterium tumefaciens strain GV3101 (Koncz and Schell, 1986). This strain was used to transform Col-0 and mutants via the floral dip method (Clough and Bent, 1998). Transformants were selected by selection on 15 μg/mL of hygromycin.

Transcript analysis

RNA from 8 d old seedlings was isolated with TriReagent (Sigma) according to the manufacturer’s suggestions and equal amounts of RNA were treated with DNaseI (NEB). cDNA was synthesized using the Invitrogen Superscript III kit. 2 μL of the 1:10 cDNA, diluted with 10 mM Tris, pH 8.0, was used as the template for qPCR, along with final primer concentrations of 200 nM, and a 0.5X final reaction volume of Bio-Rad iTaq Universal SYBR Green Supermix. A Bio-Rad CFX Connect Real-Time PCR system was used with the following cycle conditions: 5 minute initial denaturation at 95°C, 40 cycles of annealing at 95°C for 10 seconds and extension at 56°C for 40 seconds, followed by a melt curve analysis. Primers for MPK1 transcript were MPK1-7 and MPK1-8qPCR, and MPK1-12 and MPK1-17. Primers for the reference gene TUB4 (At5g44340) were bTUB F2 and bTUB R2 (Argyros et al., 2008). Data was analyzed using the efficiency calibrated method (Pfaffl, 2001). Data represents the average of 3 biological replicates, each themselves the average of 3 technical replicates. Error bars represent standard error.

Yeast two-hybrid

Plasmids were transformed (Gietz and Schiestl, 1995) into the Saccharomyces cerevisiae strain YPB2 (MATa ura3-52 his3-200 ade2-101 lys2-801 trp1-901 leu2-3,112 can^R gal4-542 Gal80-538 LYS2::GALI_UAS-LEU2_TATA-HIS3 URA3:: GAL4_17mers(×3)-CyCl_TATA-lacZ) (Bartel et al., 1993) with the described yeast two-hybrid constructs. Transformants were selected on synthetic complete (SC) growth media lacking leucine and tryptophan (SC-L-W). Individual colonies from the initial transformation were streaked onto SC-L-W plates for secondary selection. Individual colonies from these plates were resuspended in 60 μL dH₂O in a 96-well plate and plated onto either SC-L-W plates or SC plates lacking leucine, tryptophan, and histidine supplemented with 3-amino-1,4,5-triazole (3-AT) using an inoculation frogger. Plates were incubated at 30 °C and photographed after 5 d of growth.

Protein purification

Protein expression constructs were transformed into E. coli Rosetta cells (Invitrogen). Cultures from single colonies were grown to OD₆₀₀=0.5-1, induced with 1 mM IPTG, and grown overnight at 18 °C. Cells were harvested by centrifugation at 4,000 rpm for 12 minutes. Pellets were resuspended in lysis buffer (for GST-tagged proteins: 50 mM Tris, pH 8.0; 50 mM NaCl, 0.15 mM PMSF; for His-tagged proteins: 50 mM Tris, pH 8.0; 20 mM imidazole; 500 mM NaCl, 10% glycerol; 1% Tween) and lysed by sonication. Cell debris was pelleted by centrifugation at 16,000xg for 60 minutes, and supernatant was applied over the appropriate resin column (glutathione agarose (Thermo/Pierce) for GST-tagged proteins or Ni-NTA agarose (Qiagen) for His-tagged proteins). Bound resin was washed with wash buffer (for GST-tagged proteins: 50 mM Tris, pH 8.0; 150 mM NaCl; for His-tagged proteins: 50 mM Tris, pH 8.0; 20 mM imidazole; 500 mM NaCl, 10% glycerol), and protein was eluted with elution buffer (for GST-tagged proteins: 50 mM Tris, pH 8.0; 150 mM NaCl, 10 mM glutathione; for His-tagged proteins: 50 mM Tris, pH 8.0; 250 mM imidazole; 500 mM NaCl, 10% glycerol). Eluates were concentrated by centrifugation with Vivaspin protein concentrators (GE Healthcare). Comassie gels of purified proteins are shown in Figure S6. Protein concentration was approximated using a Bradford assay and approximately 1 μg of each protein used in kinase assays. Myelin basic protein was purchased from Sigma. The His-ARF7 protein was expressed as previously described (Korasick et al., 2014).

Kinase assays

Proteins and substrates were incubated in kinase assay buffer (50 mM Tris, pH 7.5; 50 μM cold ATP; 10 mM MgCl₂; 1 mM DTT), with addition of 1 μCi [γ-³²P]ATP (Perkin Elmer) to each reaction, and incubated at room temperature for 45 minutes before adding 2X NuPAGE buffer to stop the reactions. Myelin basic protein (MBP; Sigma, St. Louis, MO) was used as a substrate in some assays. Reactions were boiled at 100 °C for 10 minutes, and proteins separated by SDS-PAGE. Gels were then dried and exposed to film.

Accession numbers

Sequence data from this article can be found in the Arabidopsis Genome Initiative or GenBank/EMBL databases under accession number At1g10210 (MPK1), At3g62980 (TIR1), At1g05180 (AXR1), At2g38120 (AUX1), At2g04550 (IBR5), At5g40440 (MKK3), At5g10520 (RBK1), At1g75840 (ROP4), and At4g35020 (ROP6), At5g20730 (ARF7).

Supplementary Material

Supp FigS1-6

Figure S1. Primary root lengths corresponding to normalized data.

Figure S2. Genetic interaction of mpk1 with tir1, axr1, aux1, and ibr5.

Figure S3. MPK1 and RBK1 do not phosphorylate the ARF7 PB1 domain.

Figure S4. ROP2 and ROP4 auxin-responsive inhibition of root elongation.

Figure S5. MKK3, MPK1, RBK1, ROP2, and ROP4 do not appear to influence gravitropism responses.

Figure S6. Protein purification.

Supp TableS1. Table S1.

Primers used in this study.

SIGNIFICANCE STATEMENT.

Specific roles for MAP kinase proteins influencing auxin responses have remained elusive. In this work, we identify a potential mechanism influencing auxin-responsive cell expansion.