Receptor kinase module targets PIN-dependent auxin transport during canalization

Abstract

Spontaneously arising channels transporting the phytohormone auxin provide positional cues for self-organizing aspects of plant development such as flexible vasculature regeneration or its patterning during leaf venation. The auxin canalization hypothesis proposes a feed-back between auxin signaling and transport as underlying mechanism but molecular players await discovery. Here we identified part of the machinery that routes auxin transport. The auxin-regulated receptor CAMEL (Canalization-related, Auxin-regulated Malectin-type RLK) together with CANAR (Canalization-related Receptor-like kinase) interact with and phosphorylate PIN auxin transporters. camel and canar mutants are impaired in PIN1 subcellular trafficking and auxin-mediated PIN polarization, which macroscopically manifests as defects in leaf venation and vasculature regeneration after wounding. The CAMEL-CANAR receptor complex is part of the auxin feed-back that coordinates polarization of individual cells during auxin canalization.

Introduction

Plant development flexibly adapts the plant’s architecture and physiology to an ever-changing environment. Much of this adaptive development is characterized by self-organization of patterning processes, such as integration of new organs with the pre-existing vascular network, rise of complex leaf processes, such as integration of new organs with the pre-existing vascular network, rise of complex leaf venation patterns and flexible vasculature regeneration around the wound.

Formation of organized vasculature from originally uniform tissues involves coordinated polarization of individual cells. Canalization hypothesis proposes that the plant hormone auxin acts as a polarizing cue by means of its directional intercellular flow and the feed-back between auxin signaling and transport (1). Auxin transport is mediated by polarly localized PIN auxin transport proteins (2) and thus auxin signaling coordinating the repolarization of PINs in individual cells can generate auxin transport channels demarcating the future position of forming vasculature. The emergence of PIN-expressing auxin channels preceding vasculature formation has been observed in different plant species connecting newly formed organs (3) or lateral shoot branches (4) with pre-existing vasculature network, also during leaf venation (5), in embryogenesis (6) and during regeneration after wounding (7, 8). Similar PIN-expressing auxin channels arise from an artificial local auxin source revealing that auxin is the necessary and sufficient signal for channel formation (4, 8).

It remains enigmatic how such auxin feed-back on subcellular PIN localization leading to coordinated tissue polarization can integrate directional and positional cues. Auxin transcriptionally regulates PIN expression (9) and inhibits PIN endocytic recycling (10), which may explain auxin-mediated PIN repolarization by de novo secretion and by a differential endocytosis rate of PIN proteins from the plasma membrane leading to establishment of polarity (11). A mechanistic model of auxin canalization (12) predicts that PIN polarization away from the auxin source can arise from a combination of intracellular, transcriptional auxin signaling regulating PIN abundance and cell surface auxin signaling regulating PIN internalization rate (10) and thereby stabilizing PINs at the given cell side. This mechanism would sense an auxin gradient throughout the tissue and translate it into tissue polarization. Additionally, a so far elusive short-range signaling mechanism would mediate coordination between individual cells during this process.

Here, we identified a CAMEL-CANAR cell surface receptor complex, acting downstream of the canonical TIR1/AFB-WRKY23 auxin signaling, which is required for the auxin effect on PIN trafficking and polarity in individual cells as well as for coordinated tissue polarization during vascular formation and regeneration.

Results

Identification of potential auxin canalization regulators downstream of WRKY23

To gain more insight into the auxin regulation of PIN polarity, we designed a microarray experiment to find genes downstream of the WRKY DNA-BINDING Protein 23 (WRKY23) transcription factor that regulates auxin-mediated PIN repolarization (13) (Fig. S1A,B). We obtained 14 genes as auxin-inducible, potential targets of WRKY23 (Fig. 1A; Tab. S1). To identify among them regulators of auxin canalization, we isolated or used previously described corresponding loss-of-function mutants. We analyzed the vasculature formation during leaf development as a classical process requiring auxin feedback on PIN-dependent auxin transport (5). Typically, wild type (Wt) cotyledons form a conserved pattern of four loops as in most mutants with exception of mutants in AT5G40780 and AT1G05700, where we observed strong venation defects: (Fig. S1C; AT5G40780/lht1-1: 41% and AT1G05700/camel-1: 45%; where % stands for any type of abnormality deviating from typical Wt pattern).

Fig. 1. CAMEL expression is regulated by WRKY23 and depends on the TIR1/AFB pathway.

(A) Scheme of the microarray experimental setup to identify auxin-regulated genes downstream of the TIR1/AFB-WRKY23 signaling module.

(B) Map of physical interactions between extracellular domains of putative interactors of CANAR from a previous study (16) with CAMEL being one of high confidence, illustrated with the BAR interaction viewer.

(C) Schematic representation of the domain organization of CAMEL and CANAR.

(D and E) RT-qPCR experiments showing that (D) CAMEL expression depends on WRKY23 and (E) auxin-mediated upregulation of CAMEL requires the TIR1/AFB activity.

(F) Luciferase assay in Nicotiana benthamiana: 35S::WRKY23 co-expressed with CAMELpro:LUC and negative control of 35S:WRKY23 or CAMAL:LUC was indicated respectively.

(G and H) FRET-FLIM analysis of transiently expressed 35S::CANAR-GFP and 35S::CAMEL-mCherry in protoplasts. The GFP fluorescence lifetime was calculated as described in the Methods section and the heat map represents the fluorescent lifetime values. A One-Way ANOVA test compared marked sets of data (***, p<0.001; ****, p<0.0001). n denotes the number of scored protoplasts.

Thus, we identified AT5G40780 and AT1G05700 acting downstream of TIR1/AFB-WRKY23 and being required for canalization-based processes such as leaf venation.

Malectin-type LRR Receptor-like kinase CAMEL downstream of auxin signaling

AT1G05700 encodes a previously uncharacterized member of the Leucine-Rich Repeat (LRR) receptorlike kinase (RLK) family from the subfamily I and its extracellular domain consists of a large Malectin-like domain and three LRR repeats (Fig. 1C; Fig. S1D). We named AT1G05700 CAMEL (Canalization-related Auxin-dependent Malectin-like RLK).

qRT-PCR on lines where WRKY23 is either targeted inducibly to the nucleus (35S::WRKY23-GR) or engineered into a transcriptional repressor (35S::WRKY23-SRDX) (14) confirmed that CAMEL mRNA levels increased after activation of WRKY23 and decreased after its repression (Fig. 1D). The JASPAR database of transcription factors (TFs) (15) also predicted WRKY23 among the top candidates binding to a 2000bp CAMEL promoter (Tab. S2). We confirmed an activation of CAMELpro by WRKY23 using the luciferase-based reporter system in Nicotiana benthamiana. A co-expression of 35S::WRKY23 and CAMELpro::LUC led to activation of luciferase activity (Fig. 1F). This supports that CAMEL is a downstream gene of WRKY23.

As shown previously, the TIR1/AFB auxin pathway is required for auxin-mediated PIN polarity re-arrangements and canalization-based development (4, 8) and this goes in part through WRKY23 (13). Consistently with this, CAMEL transcription, similar to WRKY23, is induced by auxin in a time- and dose-dependent manner (Fig. S1E,F) and this auxin effect is not observed in 35S::WRKY23-SRDX line or in mutants defective in transcriptional auxin signaling (HS::axr3-1 and arf7arf19) (Fig. 1D,E). Furthermore, CAMELpro contains 6 auxin responsive elements (Fig. S1G), suggesting additional auxin regulation, possibly directly by ARFs, also supported by fast upregulation of CAMEL by auxin (Fig. S1F).

Thus, CAMEL is transcriptionally regulated by WRKY23 downstream of the TIR1/AFB-ARF signaling module.

CAMEL-CANAR signaling complex at the cell surface

To identify other receptor-like kinases that might interact with CAMEL, we used a map of physical interactions between extracellular domains of Arabidopsis LRR-RLKs (16). A large class VII LRR-RLK was retrieved as a single, high-confidence interactor (Fig. 1B,C; Fig. S1H). Given its association with CAMEL in canalization processes, we named it CANAR (Canalization-related Receptor-like kinase).

To determine whether CAMEL and CANAR interact in vivo as full-length proteins we used three approaches. Firstly, we immunoprecipitated CANAR-GFP using anti-GFP antibodies and performed tandem mass spectrometry to identify proteins co-immunoprecipitated with our bait. We obtained a list with possible interactors including CAMEL among the top 10 interactors (Tab. S3). Next, we tested the CAMEL-CANAR interaction using bimolecular fluorescence complementation (BiFC) in Nicotiana benthamiana leaves. Co-infiltration of CAMEL and CANAR resulted in a positive signal in both combinations (35S::CAMEL-(C)CFP + 35S::CANAR-(N)CFP and vice versa). A negative control 35S::TMK2(C)CFP + 35S::CANAR(N)CFP did not show a signal (Fig. S1I).

To test the CAMEL-CANAR interaction more quantitatively, we used Förster resonance energy transfer combined with fluorescence lifetime imaging microscopy (FRET-FLIM). We co-expressed 35S::CANAR-GFP and 35S::CAMEL-mCherry in Arabidopsis root protoplasts and detected a reduction in the GFP fluorescence life time as compared to 35S::CANAR-GFP expressed alone (Fig. 1G,H). Auxin treatment had no effect on CAMEL-CANAR interaction. Spatial resolution of FRET-FLIM experiments suggested that both proteins are localized and form complexes at the cell surface (Fig.1G).

These observations showed constitutive, auxin-insensitive interaction between CAMEL and CANAR.

CAMEL and CANAR in leaf venation

Next, we tested genetically an involvement of CAMEL-CANAR complex in auxin-mediated canalization. We isolated publicly available T-DNA insertional loss-of-function mutants in both genes: camel-1/-2 and canar-1/-2 (Fig. S2A-D), and generated gain-of-function transgenic lines overexpressing CAMEL and CANAR under the constitutive RPS5A and 35S promoters, respectively (Fig. S2E,F).

The camel-1/-2 mutants showed abnormal vascular patterning with disconnected upper loops, extra branches and extra or missing loops. Two independent overexpression lines also exhibited vasculature defects albeit with lesser frequency (Fig. 2A,B). Also both canar-1/-2 mutant alleles showed similar, even more pronounced vasculature abnormalities whereas 35S::CANAR-GFP exhibited overall less frequent defects (Fig. 2C,D). Double mutant camel-1xcanar-1 exhibited largely rescued venation implying a possible antagonistic action of CAMEL and CANAR (Fig. 2A,C; S2G). CAMEL/CANAR function appears to be rather specific to vasculature formation as no other obvious growth defects were observed (Fig. S2H-P).

Fig. 2. Abnormal patterning and PIN1 polarization during leave venation in camel and canar mutants.

(A and C) Representative images of venation patterning defects in cotyledons associated with camel-1 (A) or canar-1 (C), respectively. Scale bar: 100 μm.

(B and D) Quantification of venation defects in camel and canar mutants (n>75 for each genotype). Scored categories: No phenotype, less loops, higher structure (including extra loops or branches) and upper disconnections.

(E and F) DR5rev::GFP signal distribution and (F) intensity in cotyledons of camel-1/canar-1 mutants. (n>12 for each genotype). Scale bar: 100 μm.

(G and H) Coordinated PIN1 polarity in Col-0 and defective PIN1 polarity in camel-1/canar-1 mutants. Colored boxes in (G) illustrate positions of close-ups in (H). (H) Representative images of PIN1 immunolocalization in firstleaves. A number in left top corner indicates an incidence of observed PIN1 defective polarity to a total amount of analyzed leaves. White arrows show a typical PIN1 polar localization. Red arrows mark a defective PIN1 polarity. Evaluation criteria are described in Material and Method section. Scale bar: 10 μm

Next, we tested if these venation defects are linked to altered auxin distribution and auxin transport. We analyzed the expression of the auxin-responsive reporter DR5rev::GFP, which can be correlated with auxin distribution (17). In both camel-1 and canar-1 mutants DR5 activity was decreased compared to the control (Fig. 2E,F). When analyzing the basipetal (rootward) auxin transport in hypocotyls, we observed that both camel-1 and canar-1 mutants have reduced auxin flow (Fig. S2Q). Given that formation of PIN1-expressing, polarized channels has been linked to vein formation (5) we examined PIN1 polarity in young first leaves by means of anti-PIN1 antibody staining. In Wt leaves, coordinated PIN1 polarity defining auxin-transporting channel was observed with rare PIN1 polarity abnormalities in primary and secondary branches. In contrast, both camel-1 and canar-1 mutants showed higher incidence of PIN1 polarity defects (marked by red arrows) in primary and canar-1 also in secondary branches (Fig. 2G,H) whereas no defects were observed in midvein for any tested genotypes (Fig. S2R).

These observations show that the CAMEL and CANAR mediate vasculature development during leaf venation and coordinate PIN1 polarities during this process.

CAMEL and CANAR in vasculature regeneration after wounding

Another classical example of a canalization-mediated process is vasculature regeneration after wounding when new vasculature is generated circumventing the wound (7, 8).

We interrupted the vasculature in the Arabidopsis thaliana inflorescence stem by a horizontal cut (Fig. 3A) and analyzed GUS expression at 1-7 days after wounding (DAW). Both CAMEL and CANAR as well as their upstream regulator WRKY23 showed promoter activities during the regeneration (Fig. 3B; S3A). Next, we analyzed the extent of vasculature regeneration in loss-of-function mutants and overexpressing lines as visualized by toluidine blue staining (TBO) of regenerated vasculature. In Wt, vasculature was fully developed and both newly regenerated vessel cells (white asterisk) and lignified parenchyma cells (red asterisk) stained in blue were visible. All tested mutants showed defective regeneration caused by inability to form a continuous strand of regenerated cells. RPS5A::CAMEL showed less frequent defects and 35S::CANAR-GFP exhibited even improved regeneration over the Wt (Fig. 3C,D; S3B,C). Similar defects were observed also for flexible formation of auxin transport channels. In PIN1-GFP line, but not in canar-1xPIN1-GFP, PIN1-GFP expressing channel (marked by yellow arrow) circumventing the wound was formed (Fig. S3D). In contrast to leaf venation, camel-1xcanar-1 double mutant showed regeneration defects comparable to individual mutants (Fig. S3B,C).

Fig. 3. Defective vasculature regeneration after wounding in camel, canar and wrky23-1 mutants.

(A) Scheme of the experiment for analyzing vasculature regeneration after wounding.

(B) Expression of CANARpro::GUS, CAMELpro::GUS and WRKY23pro::GUS 4 days after wounding (DAW). The wound site is marked by a white arrowhead. Scale bar: 100 μm.

(C) Wounded stems of canar-1, camel-1, wrky23-1 mutants and RPS5A::CAMEL overexpressing line at 0 and 7 DAW. Stems are stained by toluidine blue to visualize newly regenerated blue-vessel cells (white asterisks) and lignified parenchyma cells (red asterisks). The wound site is marked by a white arrowhead.

(D) Quantification of regeneration defects in (C). n denotes the number of evaluated plants. Scale bar: 100 μm.

To analyze more directly auxin-mediated formation of auxin transport channels, we used external, local auxin application (18). Application of IAA droplet on the stem side below the wound (marked by magenta) led to formation of PIN1-expressing channel connecting with the pre-existing vasculature already after 2 DAW in Wt, an effect not observed in canar-1 background even after 4 DAW (Fig. S3E). Accordingly, the similar, newly formed vascular strands (as revealed by TBO staining) could be observed only in Wt (Fig. S3F).

These results revealed a role for CAMEL and CANAR as well as their upstream regulator WRKY23 in the flexible regeneration of vasculature following wounding.

CAMEL and CANAR in trafficking and polarity of PIN auxin transporters

Since auxin feed-back on PIN polarity is one of the main features of canalization and PIN polar localization is linked to its constitutive endocytic recycling (11, 12) we tested whether CAMEL and CANAR are involved in these processes. PIN endocytic recycling can be indirectly visualized by PIN intracellular aggregation in response to treatment with the trafficking inhibitor Brefeldin A (BFA) (19). Anti-PIN1 immunostaining in roots showed that following BFA treatment, PIN1 intracellular aggregation was reduced in camel-1 and canar-1 mutants (Fig. 4A,B). The same phenomenon was observed for camel-1xPIN2-GFP and canar-1xPIN2-GFP crosses (Fig. S4A,B) indicating a defect in the PIN endocytic recycling.

Fig. 4. Subcellular trafficking and auxin feed-back on PIN polarity is compromised in camel and canar mutants.

(A) Representative confocal images of primary root stele cells after immunostaining PIN1 in Wt, camel-1 and canar-1. Seedlings were BFA-treated (25µM) for 30min. Scale bar: 10 μm.

(B) Quantitative evaluation of (A) shows ratio of total number of BFA bodies/total number of cells per root. n denotes the number of evaluated seedlings (****p<0.0001).

(C) Immunolocalization of PIN1 in endodermis of root meristem after 4h NAA (10µM) treatment. Scale bar: 10 μm.

(D) Quantitative evaluation of (C) shows mean ratio of PIN1 lateral-to-basal signal intensity ratio in endodermal cells. Error bars indicate standard error. The experiment was carried out three times, one representative experiment is presented. A One-Way ANOVA test compared marked datasets (****p<0.0001; n>80 cells corresponding to a minimum of 10 roots per treatment and experiments were imaged using comparable settings).

The auxin effect on PIN polarity can be approximated by the repolarization of PIN1 from the basal to the inner lateral side in the root endodermis cells (8). Anti-PIN1 immunolocalization revealed that following auxin treatment, PIN1 repolarization was reduced in camel and canar mutants (Fig. 4C,D; S4C-E).

These results imply that the CAMEL-CANAR complex not only plays a role in the canalization-related development at the level of organs and tissues but also targets PIN1 in individual cells, regulating its subcellular trafficking and auxin feed-back on PIN polarity.

CAMEL-CANAR receptor complex targets and phosphorylates PIN auxin transporters

To get insight into the mechanism of CAMEL-CANAR action and downstream processes, we immunoprecipitated CAMEL-GFP from seedlings to identify the interactome of CAMEL. Co-immunoprecipitated proteins were pulled-down and analyzed using mass spectrometry (Tab. S4). Among the list of putative interactors multiple PIN proteins were found.

To confirm the interaction with PIN1, we transiently co-expressed 35S::CAMEL-GFP+35S::PIN1-mRFP and 35S::CANAR-GFP+35S::PIN1-mRFP in Arabidopsis root protoplasts. PIN1-mRFP co-immunoprecipitated with both CANAR-GFP and CAMEL-GFP (Fig. 5A). Furthermore, we performed FRET-FLIM in root protoplasts expressing 35S::CAMEL-GFP or 35S::CANAR-GFP. The lifetime of CAMEL-GFP and CANAR-GFP was reduced after co-expression with PIN1^HL-mCherry (HL=hydrophilic loop) further confirming an interaction between CAMEL/CANAR and PIN1 (Fig. 5B,C).

Fig. 5. The CAMEL-CANAR signaling module directly targets PIN1.

(A) Co-immunoprecipitation (Co-IP) assay from Arabidopsis root protoplasts. The protein complex of CAMEL-PIN1 and CANAR-PIN1 was co-immunoprecipitated by anti-GFP beads. Anti-AHA2 was used as loading control. The experiment was carried out three times.

(B and C) FRET-FLIM analysis of transiently expressed 35S::CAMEL-GFP and 35S::CANAR-GFP with 35S::PIN1^HL-mCherry (HL-hydrophilic loop) in protoplasts. The GFP fluorescence lifetime was calculated as described in the Methods section and the heat map indicates fluorescent lifetime values. A One-Way ANOVA test compared the marked data sets (****, p<0.0001). n denotes the number of scored protoplasts.

(D) Autoradiograph of an in vitro kinase phosphorylation assay of PIN1/2/3^HL by CAMEL^CD (CD-cytoplasmic domain). Aliquots of the samples were separated by SDS-PAGE and exposed to autoradiography. The coomassie blue staining was used as loading control and presence of the respective recombinant proteins. The blots shown are representative for three biological replicates.

Since CAMEL and CANAR are expected to act as kinases and PIN phosphorylation is a well-established mode of regulation of PIN activity and polar localization (2), we tested the ability of CAMEL and CANAR to phosphorylate PINs. We therefore performed an in vitro kinase assay by incubating purified PIN1^HL, PIN2^HL or PIN3^HL with purified cytoplasmic kinase domains of CAMEL and CANAR with radiolabeled ATP. We detected phosphorylation of PIN loops by CAMEL with PIN1^HL being the best substrate (Fig. 5D). However, CANAR did not show kinase activity (Fig. S5A). Lack of kinase activity can be explained by losing an aspartic acid from the conserved HRD motif in the catalytic core similarly to other known pseudokinases: BIR2, GHR1, PRK5 (Fig. S5B).

Considering the lack of CANAR kinase activity, constitutive CAMEL-CANAR interaction and complementation of leave vasculature defects of camel-1xcanar-1 double mutant, we hypothesize that CANAR might be a negative regulator of CAMEL. This is further supported by ability of CANAR kinase domains to reduce CAMEL auto-phosphorylation and kinase activity towards PIN1 (Fig. S5C).

To test relevance of CAMEL-mediated PIN1 phosphorylation, we analyzed in vitro kinase reaction using mass spectrometry and identified five mostly conserved putative phosphosites in PIN1^HL (Fig. S5D; 6A-C). These sites seem unique since they are not shared by any previously reported kinase phosphorylating PIN loop such as PID/WAGs, D6PK or MPKs (Fig. S5D) and when mutated, they decreased the ability of CAMEL kinase domain to phosphorylate PIN1^HL (Fig. S6A,B).

Fig. 6. CAMEL-targeted phosphosites in the PIN1 cytoplasmic loop are important for PIN polarity and venation.

(A) Schematic representation of PIN1 in the plasma membrane with marked positions of phosphosites in the cytoplasmic loop targeted by CAMEL.

(B) Phenotypes of PIN1pro::PIN1-GFP^T3AS2A and PIN1pro::PIN1-GFP^T3ES2E. 35 days old.

(C) Subcellular localization of PIN1-GFP, PIN1-GFP^T3AS2A and PIN1-GFP^T3ES2E in root meristem endodermal cells. White arrows mark the predominant subcellular localization.

(D) Representative images of vasculature defects in cotyledons of PIN1pro::PIN1-GFP, PIN1pro::PIN1-GFP^T3ES2E (line13). Scale bar: 100 μm.

(E) Quantification of vasculature defects in PIN1pro::PIN1-GFP, PIN1pro::PIN1-GFP^T3AS2A (lines A, B, F) and PIN1pro::PIN1-GFP^T3ES2E (line 13) (n>68 for each genotype). Scored categories: normal vasculature, less loops, higher structure (including extra loops or branches) and upper disconnections.

(F) Quantitative evaluation of (Fig. S6D) shows the mean lateral-to-basal ratio of PIN1-GFP signal in endodermal cells. Error bars indicate standard errors. The experiment was carried out three times, one representative experiment is presented. A One-Way ANOVA test was performed to compare marked datasets (*p<0.05,**p<0.01,***p<0.001,****p<0.0001; n>40 cells corresponding to a minimum of 8 roots per treatment and experiments were imaged using comparable settings).

We generated phosphodead PIN1-GFP^T3AS2A and phosphomimic PIN1-GFP^T3ES2E constructs by substitution of three threonine and two serine to alanine or glutamic acid, respectively, placed them under control of the native PIN1 promoter and introduced into Wt plants. Positive, GFP-expressing transformants for both constructs showed already in the first generation naked inflorescence stems (7/20 for PIN1-GFP^T3AS2A and 3/18 for PIN1-GFP^T3ES2E) strongly reminiscent of pin1 loss-of-function (Fig. 6B). Other positive plants did not show strong phenotypes and produced seeds allowing analysis in the next generation. Venation in cotyledons of both PIN1-GFP^T3AS2A and PIN1-GFP^T3ES2E lines exhibited increased incidence of vascular abnormalities (Fig. 6D,E; S6C). All positive transformants in the first generation for PIN1^T3AS2A (4/4) and PIN1^T3ES2E (2/2) showing naked inflorescence stems exhibited no vasculature regeneration after wounding characterized by fragmented vessel cells, non-functional parenchyma cell connections or extensive callus formation in the wound (Fig. S6E). To test the role of the identified phosphosites in canalization, we tested auxin effect on mutated PIN1 variants. While PIN1-GFP is localized in roots predominantly basal, both PIN1-GFP^T3AS2A and PIN1-GFP^T3ES2E showed more apolar localization already without any treatments (Fig. 6C; S6D). When immunolocalized with anti-GFP antibody, both PIN1-GFP^T3AS2A and PIN1-GFP^T3ES2E already partially polarized to the inner-lateral side in the mock situation, did not show any further polarity changes following auxin application (Fig. 6F; S6D).

In conclusion, the CAMEL-CANAR complex interacts with PINs and CAMEL is capable of phosphorylating their cytosolic loops. Effects of phosphomimic and phosphodead mutations in the PIN1 loop support the relevance of these phosphorylations for auxin transport and auxin canalization. The stronger defects in lines carrying PIN1 with mutated CAMEL-targeted phosphorylation sites as compared to the camel/canar mutants suggest that these phosphosites are shared by other kinases controlling auxin transport in a different developmental contexts.

Discussion

In this study, we provided mechanistic insights into how auxin controls its own directional cell-to-cell transport and identified molecular components of so-called auxin canalization mechanism underlying flexible and self-organizing formation of auxin channels guiding vasculature formation. Identification of CAMEL-CANAR complex downstream of transcriptional auxin signaling and its direct regulation of PIN-dependent auxin transport provide a potential means how to integrate global auxin signals with a so far hypothetical short range signaling for coordinating cell polarities during plant adaptive development.

One Sentence Summary.

An auxin-responsive receptor-like kinase complex regulates directional auxin transport

Data and materials availability

All data is available in the main text or the supplementary materials. Raw microarray data from this article can be found in the EMBL ArrayExpress repository under accession number: XXXX.