Activation of Self-Incompatibility Signaling in Transgenic Arabidopsis thaliana Is Independent of AP2-Based Clathrin-Mediated Endocytosis

Internalization of plasma membrane (PM)-localized ligand-activated receptor kinases and their trafficking to sorting endosomes have traditionally been viewed as functioning primarily in the down-regulation of receptor signaling, but are now considered to be also essential for signaling by some receptors. A major mechanism for internalization of PM proteins is clathrin-mediated endocytosis (CME). CME is mediated by the Adaptor Protein Complex 2 (AP2), which is involved in interaction of the AP2 μ-adaptin subunit with a tyrosine-based Yxxϕ motif located in the cytoplasmic domain of the cargo protein. In this study, we investigated the role of AP2-mediated CME for signaling by the S-locus receptor kinase (SRK), a protein localized in the PM of stigma epidermal cells, which, together with its pollen coat-localized S-locus cysteine-rich (SCR) ligand, functions in the self-incompatibility (SI) response of the Brassicaceae. Using Arabidopsis thaliana plants that were made self-incompatible by transformation with an A. lyrata-derived SRK/SCR gene pair, we tested the effect on SI of site-directed mutations in each of the two Yxxϕ motifs in SRK and of a CRISPR/Cas9-induced null mutation in the AP2 μ-adaptin gene AP2M. Both in vitro SRK kinase activity and the in planta SI response were abolished by substitution of tyrosine in one of the two Yxxϕ motifs, but were unaffected by elimination of either the second Yxxϕ motif or AP2M function. Thus, AP2-mediated CME is considered to be unnecessary for SRK signaling in the SI response.

Boucrot 2011). Homologs of these components, including the 4 subunits of the AP2 complex, are conserved in A. thaliana, indicating that CME occurs in plants . Indeed, analysis of mutants of the single m-adaptin-encoding gene found in the A. thaliana genome, designated AP2M (At5g46630), has demonstrated a role for this gene in CME of several plasma membrane-localized proteins, including the cellulose synthase CESA6, the auxin-efflux carrier PIN FORMED2 (PIN2), and the brassinosteroid receptor BRASSINOS-TEROOID INSENSITIVE 1 (BRI1) (Bashline et al. 2013;Di Rubbo et al. 2013;Kim et al. 2013).
Consistent with its endocytosis by CME, the BRI1 receptor kinase localizes to the PM and to endosomes and it contains five Yxxu motifs (Geldner et al. 2007;Irani et al. 2012), although a role for these motifs in endocytosis has not been reported. Interestingly, Yxxu motifs are found in several other plant transmembrane receptor-like kinases (RLKs), a few of which have been localized to the PM and to endosomes (Geldner and Robatzek 2008). However, for the majority of these RLKs, neither the mechanism nor the possible requirement of internalization for signaling has been addressed. One of these receptors is the S-locus receptor kinase (SRK), which is a transmembrane protein expressed in the stigma epidermal cells of self-incompatible members of the Brassicaceae (Stein et al. 1991). The SRK gene and the gene for the small pollen coat-localized S-locus cysteine-rich protein (SCR) (Schopfer et al. 1999), which functions as the ligand for SRK (Kachroo et al. 2001;Takayama et al. 2001), are tightly linked and highly polymorphic genes that together constitute the S-locus haplotype (S haplotype) and determine specificity in the self-incompatibility (SI) response (reviewed in Nasrallah and Nasrallah 2014). SRK is only bound and activated by the SCR that is encoded in the same S haplotype. This allele-specific receptor-ligand interaction, which occurs only when stigma epidermal cells and pollen are derived from plants that express the same S haplotype, triggers a downstream signaling cascade in stigma epidermal cells that causes inhibition of pollen germination and pollen tube growth at the surface of these cells .
In the study reported herein, we wished to determine if CME of SRK is required for the SI response. Like other recent structure-function studies of SRK (Boggs et al. 2009a;Yamamoto et al. 2014), this study was facilitated by the availability of a transgenic self-incompatible Arabidopsis thaliana model (Nasrallah et al. 2002(Nasrallah et al. , 2004, in which the ease of transformation allows for efficient in planta testing of mutated SRK proteins generated by site-directed mutagenesis. We previously showed that several SRK/SCR gene pairs, including the SRKb/SCRb gene pair derived from the Sb haplotype of self-incompatible Arabidopsis lyrata, confer an intense and developmentally stable SI response in several accessions of the normally self-fertile A. thaliana, including the C24 accession (Nasrallah et al. 2004;Boggs et al. 2009a,b). This transgenic A. thaliana platform was recently used for live-cell imaging of functional YFP-tagged SRKb proteins in stigma epidermal cells (Rea and Nasrallah 2015). As expected, the full-length SRKb-YFP receptor was observed predominantly at the plasma membrane of epidermal cells in unpollinated stigmas. Notably, this localization of the SRK at the plasma membrane and in juxtaposition to the cell wall, which make its extracellular domain accessible to its pollen-derived SCR ligand, is critical for SRK function. Indeed, loss of SI has been observed when SRK targeting to the plasma membrane is disrupted (Yamamoto et al. 2014;Tantikanjana and Nasrallah. 2015), and when stigmas are treated with 5M NaCl, which causes plasmolysis and retraction of the plasma membrane from the cell wall of epidermal cells (Rea and Nasrallah. 2015;Tantikanjana and Nasrallah. 2015).
In addition to its primary localization at the plasma membrane, the full-length SRKb-YFP receptor was also observed, albeit at much lower levels, in transvacuolar strands and small intracellular vesicle-like structures of unpollinated stigmas (Rea and Nasrallah 2015). Some of these vesicles might be endosomes as inferred from a previous immunolocalization study of the Brassica oleracea SRK3 variant (Ivanov and Gaude 2009). Interestingly, neither of the two SRK localization studies detected a redistribution of SRK in response to self-pollination. Although this result might be related to the resolving power of the visualization methods used, the possibility that SRK is internalized was suggested by internalization of an anti-SRK3 antibody that served as a mimetic for the SCR3 ligand (Ivanov and Gaude 2009). In any case, a functional connection between SRK internalization and SRK signaling in the SI response has not been established.
To address this issue, we focused on the CME pathway as a potential mechanism of SRK internalization, and more specifically on the Yxxu motif-AP2M interaction, which is an essential step in CME. We generated strains that express SRKb mutant proteins in which Yxxu motifs were destroyed by site-directed mutagenesis and we constructed strains carrying an ap2m null mutation generated using the CRISPR/Cas9 gene editing method. Our analysis of these mutant strains strongly suggests that CME is not required for SRK signaling in the SI response.
The transfected cells were collected by centrifugation and resuspended in 200 ml of lysis buffer containing 20 mM sodium phosphate buffer pH 7.4, 150 mM NaCl, 1 mM EDTA pH 8.0, 10%[v/v] glycerol, 0.1%[v/v] Triton X-100, 1 mM PMSF, 1x protease inhibitor cocktail (Cat#P9599, Sigma-Aldrich, St. Louis, MO), and 1x phosphatase inhibitor cocktail (Cat#P0044, Sigma-Aldrich, St. Louis, MO). The resuspended cells were disrupted by vortexing with 0.2 g of glass beads and the mix was centrifuged to remove cell debris and glass beads. The resulting supernatant (180 ml) was transferred to a new tube, to which 300 ml of lysis buffer and 20 ml of a 50%[v/v] suspension of anti-FLAG M2 affinity gel (Sigma-Aldrich, St. Louis, MO #A2220) were added. Following incubation for 3 hr at 4°, the anti-FLAG M2 affinity gel was washed three times with 1 ml of 20 mM sodium phosphate buffer, pH 7.4, containing 150 mM NaCl, 1 mM EDTA pH 8.0, 10%[v/v] glycerol, and 0.1%[v/v] Triton X-100, followed by two washes with 1 ml of 20 mM TrisHCl pH 7.5, 150 mM NaCl, 10 mM MgCl 2 , and 10 mM MnCl 2 . The anti-FLAG M2 affinity gel was resuspended in 20 ml of kinase buffer [20 mM TrisHCl pH 7.5, 150 mM NaCl, 10 mM MgCl 2 , 10 mM MnCl 2 , 1 mM PMSF, 1x protease inhibitor cocktail, and1 · phosphatase inhibitor cocktail], 1 ml of 32 P-gATP was added, and the mix was incubated for 30 min at room temperature. The samples were then washed twice with 1x phosphate buffered saline containing 20 mM EDTA, and the samples were eluted with 20 ml of SDS-PAGE sample buffer (Laemmli 1970) by boiling for 5 min at 95°.
For detection of phosphorylated SRKb proteins, the samples were run on SDS-PAGE gels, the gels were dried, and radioactive signals were detected using a Storm 860 Image Analyzer (Molecular Dynamics, Sunnyvale, CA). For western blot analysis, samples were subjected to SDS-PAGE followed by transfer to Immobilon-P membranes (Millipore, Billerica, MA) as described by Towbin et al. (1979). SRKb-FLAG proteins were detected using a 1:1,000 dilution of monoclonal anti-FLAG antibody (Sigma-Aldrich, St. Louis, MO) as primary antibody and a 1:3,000 dilution of goat anti-mouse-IgG peroxidaselabeled antibody (Sigma-Aldrich, St. Louis, MO) as secondary antibody. Immunodetection was performed using the ECL2 system (Thermo Fisher Scientific, Waltham, MA) and exposure to X-ray film.
CRISPR/Cas9 mutagenesis of the AP2M gene A plasmid carrying AP2M sgRNA and the Cas9 gene was generated as described by Fauser et al. (2014). A 20-bp spacer sequence complementary to a region spanning nucleotides 62-81 in the first exon of AP2M was designed using the CRISPR DESIGN website (http://crispr.mit.edu). Synthetic oligonucleotides AP2M gRNA F (59-ATTGGCACCTACCGAGATGACGTC-39) and AP2M gRNA R (59-AAACGACGTCATCTCGGTAGGTGC-39) were annealed, inserted into the BbsI restriction site of pEn-Chimera and subsequently introduced into pDe-CAS9 using LR Clonase II (Thermo Fisher Scientific, Waltham, MA). The resulting plasmid, designated AP2MsgRNA-Cas9, was introduced into C24 plants as described for SRKb mutant constructs. Seed from Agrobacterium-treated plants were sown on soil and transformants were selected by spraying seedlings with BASTA (Bayer CropScience, Berlin, Germany) at a 1:1,540 dilution. The AP2MsgRNA-Cas9 gene cassette was subsequently introduced into AP2M[AtS1 pro :SRKb-FLAG+SCRb] by crossing, and ap2m[AtS1 pro :SRKb-FLAG+SCRb] mutant plants were identified among the second-and third-generation progenies of the cross.

Pollination assays
The stigmas of buds at developmental stage 13 (Smyth et al. 1990) were manually pollinated with pollen grains from mature flowers under a stereomicroscope. Two hours after pollination, the stigmas were fixed, stained with decolorized aniline blue, and examined by epifluorescence microscopy as described previously (Kho and Bear 1968). Each pollination assay was performed in triplicate. Pollen was derived either from wild-type C24 plants or from AtS1 pro :SRKb-FLAG+SCRb plants. In these assays, results were scored as incompatible when fewer than five pollen tubes are observed per pollinated stigma, as partially incompatible when 6 to 29 pollen tubes are observed per pollinated stigma, and as compatible when more than 30 pollen tubes are observed per pollinated stigma. Images of pollinated stigmas were captured using an Axio Imager M1 microscope fitted with an AxioCam MRm camera or an Axioskop microscope with an AxioCam ERc 5s camera (Carl Zeiss, Oberkochen, Germany).

Data availability
Strains and plasmids are available upon request. The authors affirm that all data necessary for confirming the conclusions of the article are present within the article, figures, and tables. Supplemental material available at Figshare: https://doi.org/10.25387/g3.5970091.

RESULTS AND DISCUSSION
Two potential clathrin-mediated endocytosis Yxxu motifs are conserved in the kinase domain of SRK variants As shown in Figure 1A, the A. lyrata SRKb protein possesses two Yxxu motifs (hereafter designated Yxxu motif 1 and Yxxu motif 2) within its kinase domain, which might potentially function as endocytosis signals for CME of the receptor. To assess the possible conservation of these motifs in SRKs, the Clustal Omega program (Sievers et al. 2011) was used to align the kinase-domain amino-acid sequences of SRKb and SRK variants available in public databases. Seventy-eight SRK variants, including nine from A. lyrata (AlSRKs), 12 from A. halleri (AhSRKs), one from Capsella grandiflora (CgSRK), 25 from Brassica rapa (BrSRKs), and 31 from B. oleracea (BoSRKs) were aligned with SRKb. As shown in Figure 1B and Figure S1, Yxxu motif 1 [YEYL (Y600E601Y602L603) in AlSRKb] is conserved among all SRKs analyzed, while Yxxu motif 2 [YGYM (Y693G694Y695M696) in AlSRKb] is conserved in all SRKs except BoSRK1, in which YGYM is replaced by CGYM. In any case, the conservation of these two Yxxu motifs in the overwhelming majority of SRK variants in various Brassicaceae species suggests that these motifs are important for SRK trafficking or activity.
The Y600 residue in Yxxu motif 1 of AlSRKb is required for kinase activity To gain some clues regarding the potential role of Yxxu motifs in SRK, we assessed their conservation in several A. thaliana RLKs. Figure 2A shows an alignment of the amino-acid sequences of the region that spans Yxxu motif 1 in SRKb with the corresponding regions of several Yxxu-containing A. thaliana RLKs as well as the FLS2 receptor, which contains no Yxxu motifs (Geldner and Robatzek 2008). All RLKs included in the alignment, with the exception of FLS2, contain a Yxxu motif at a location equivalent to that of Yxxu motif 1 in SRKb ( Figure 2A).
In BRI1, this conserved Yxxu motif is YEFM, and the role of its tyrosine residue (Y956), which corresponds to Y600 in SRKb ( Figure  2A), has been elucidated. Structural analysis of BRI1 has shown that Y956 contributes to the formation of the ATP binding site in the BRI1 kinase domain by forming a hydrogen bond with the glutamic acid residue E927 (Bojar et al. 2014), which corresponds to E571 in SRKb (Figure 2A and Figure S1). Additionally, biochemical analysis has shown that a mutation of the Y956 residue in BRI1 eliminates kinase activity (Oh et al. 2009). To determine if the Y600 residue of SRKb is also important for kinase activity, possibly by contributing to ATP-binding site formation, we analyzed the activity of site-directed SRKb mutants that were transiently expressed under control of the CaMV 35S promoter in A. thaliana protoplast cells prepared from rosette leaves. Three SRK proteins carrying a C-terminal 3xFLAG epitope were analyzed: (1) wild-type SRKb-FLAG; (2) SRKb(Y600A)-FLAG, a mutant in which the Yxxu motif was destroyed by replacing theY600 residue with alanine; and (3) a kinase-dead SRKb(K555E)-FLAG protein, in which a glutamic acid residue replaces a lysine residue located in the putative ATP binding site that was previously shown to be required for SRK kinase activity in vitro (Goring and Rothstein 1992;Stein and Nasrallah 1993) and SI function in planta (Tantikanjana et al. 2009). The protoplast-expressed wild-type and mutant SRKb proteins were immunoprecipitated using anti-FLAG antibodies and incubated with 32 P-gATP to analyze in vitro autophosphorylation activity. As shown by the protein blots in Figure 2B, an autophosphorylation signal was detected in wild-type SRKb-FLAG but not in kinase-dead SRKb(K555E) ( Figure 2B, 32 P panel), as expected. Additionally, no autophosphorylation signal was observed in SRKb(Y600A)-FLAG ( Figure 2B, 32 P panel), despite the fact that this mutant protein was synthesized and precipitated at levels equivalent to those of wild-type SRKb-FLAG ( Figure 2B, FLAG panel). These results demonstrate that the Y600 residue is required for SRKb kinase activity.
To confirm the expectation that the kinase defect in SRKb(Y600A) affects the ability of the mutant receptor to activate the SI response in planta, an AtS1 pro :SRKb(Y600A)-FLAG chimeric gene was constructed, in which the promoter of the AtS1 gene, which is highly active specifically in stigma epidermal cells (Dwyer et al. 1994), drives expression of the SRKb(Y600A)-FLAG mutant protein. This gene was introduced  Figure S1. Asterisks, colons, and periods indicate identical amino acids, amino acids having strongly similar properties, and amino acids having weakly similar properties, respectively.
into A. thaliana C24 plants, and a total of 15 transformants were generated. Stigmas of all AtS1 pro :SRKb(Y600A)-FLAG transformants were pollinated with SCRb-expressing pollen (hereafter SCRb pollen) derived from AtS1 pro :SRKb-FLAG+SCRb plants, which express wildtype SRKb-FLAG in stigmas and the SCRb protein in pollen (Yamamoto et al. 2014). In contrast to the strong inhibition of SCRb pollen exhibited by stigmas expressing wild-type SRKb-FLAG, large numbers of elongated pollen tubes were observed in the stigmas of all AtS1 pro :SRKb(Y600A)-FLAG transformants, similar to untransformed C24 plants ( Figure 2C). Thus, the Y600 residue was revealed to be essential for both SRKb kinase activity and its function in SI.

Elimination of Yxxu motif 2 does not disrupt SRKb kinase activity and function
Unlike the Y600 residue, the Y693 residue in Yxxu motif 2, which is located in the activation loop of the SRKb kinase domain is not . The location of the glutamic acid residue (E571 in SRKb) that is part of the ATP-binding site is indicated. Note that all of the receptor kinases shown, except FLS2, contain a Yxxu motif in the same position as Yxxu motif 1 in SRKb. Asterisks, colons, and periods are as in Figure 1B. (B) In vitro kinase assays. The figure shows SDS-PAGE electrophoretic patterns of immunoprecipitated protoplast-expressed and FLAG-tagged wild-type SRKb, the SRKb(Y600A) mutant, and an SRKb(K555R) kinase-dead mutant. Proteins from untransformed plant cells [untransformed] were used as negative control. The radioactively-labeled proteins were detected with a phosphorimager ( 32 P panel), and the same samples subjected to immunoblot analysis with anti-FLAG antibody (IB:FLAG panel) were used loading controls. Note that the SRKb(Y600A) mutant protein lacks kinase activity similar to the kinase-dead SRKb(K555R) protein. (C) Pollination phenotypes of untransformed plants (C24) and transformants expressing wild-type SRKb or SRKb(Y600A) proteins. Stigmas were pollinated with SCRb-expressing pollen. Note that SRKb-expressing stigmas exhibited a robust SI response, manifested by inhibition of SCRb-expressing pollen germination and tube elongation. By contrast, SRKb(Y600A)-expressing stigmas exhibited a highly compatible response manifested by the development of numerous SCRb pollen tubes, similar to the stigmas of untransformed plants. Bar = 100 mm.
conserved among A. thaliana RLKs ( Figure 3A). However, it has been shown that phosphorylation of amino-acid residues within the activation loop of BRI1 kinases enhances both the formation of the substrate binding site and catalytic activity (Bojar et al. 2014). We therefore examined the possibility that the Y693 residue is required for SRKb kinase activity and function. Using the transient expression system and in vitro kinase assay described above, an SRKb(Y693A)-FLAG mutant protein was analyzed along with wild-type SRKb-FLAG and kinasedead SRKb(K555R) proteins. As shown in Figure 3B, SRKb(Y693A) protein exhibited an autophospholylation signal equivalent to that observed for wild-type SRKb-FLAG. This result indicates that the Y693A mutation does not affect the phosphorylation activity of SRKb, at least in its ligand-unbound state.
To assess the function of the SRKb(Y693A)-FLAG in planta, C24 plants were transformed with the AtS1 pro :SRKb(Y693A)-FLAG chimeric gene and 11 independent transformants were generated. In pollination assays with SCRb pollen, the stigmas of five AtS1 pro :SRKb (Y693A)-FLAG transformants were found to exhibit an incompatibility phenotype that was as intense as that exhibited by stigmas expressing wild-type SRKb-FLAG ( Figure 3C). Thus, the Y693 residue and Yxxu motif 2 are not required for SRKb function and SI in planta.
A CRISPR/Cas9-induced null mutation in AP2M does not affect the SI response of SRKb-expressing stigmas Because the Y600 residue of Yxxu motif 1 was found to be essential for SRKb kinase activity, it was not possible to derive any conclusion regarding the role of CME in SRKb signaling. To address this issue, we focused on AP2M, the gene that encodes the Yxxu motif-binding m-adaptin of the AP2 complex (Traub 2009) and that has been shown to Figure 3 The Y693 residue of Yxxu motif 2 is not conserved in A. thaliana receptor kinases and is not essential for SRKb kinase activity and SI. (A) Alignment of amino-acid sequences in the region spanning Yxxu motif 2 in SRKb with the corresponding regions of several A. thaliana RLKs. Except for CLV1, the SRKb Yxxu motif 2 is not conserved in the A. thaliana proteins. Asterisks, colons, and periods are as in Figure 1B. (B) In vitro kinase assay of wild-type SRKb, the SRKb(Y693A) mutant, and the SRKb(K555R) kinase-dead protein. Protoplast-expressed proteins were prepared and assayed by autoradioagraphy ( 32 P panel) or immunoblot analysis (IB:FLAG panel) as described in Figure 2B. Note that SRKb(Y693A) exhibits kinase activity similar to wild-type SRKb. (C) Pollination phenotypes of untransformed plants (C24) and transformants expressing wild-type SRKb or SRKb(Y693A) proteins. Stigmas were pollinated with SCRb-expressing pollen. Note that the stigmas expressing SRKb(Y693A) exhibit a robust SI response similar to the stigmas expressing wild-type SRKb. Bar = 100 mm.
be essential for CME of plasma membrane-localized proteins in A. thaliana (Bashline et al. 2013;Di Rubbo et al. 2013;Kim et al. 2013;Yamaoka et al. 2013). Since a panel of T-DNA insertion lines in the A. thaliana C24 accession is not publicly available, we used the CRISPR/Cas9 gene-editing method to generate an ap2m loss-of-function mutation in C24[AtS1 pro :SRKb-FLAG+SCRb] plants. A single-guide RNA (sgRNA) was designed to introduce a mutation in the first exon of AP2M ( Figure 4A). A plant transformation plasmid carrying the designed sgRNA along with the Cas9 gene was generated essentially as described by Fauser et al. (2014). Because C24[AtS1 pro :SRKb-FLAG+SCRb] plants are highly self-incompatible and set no seed (Yamamoto et al. 2014), the standard floral dipping method of transformation cannot be used with these plants. Consequently, the AP2MsgRNA-CAS9 plasmid was first introduced into untransformed wild-type C24 plants, and the resulting transformants were crossed with a wild-type AP2M[AtS1 pro : SRKb-FLAG+SCRb] plant to generate ap2m[AtS1 pro :SRKb-FLAG+SCRb] homozygous plants.
Mutagenized sgRNA target sites were detected by loss of a BsaHI restriction site within the targeted region ( Figure 4A) using cleaved amplified polymorphic sequence (CAPS) analysis. Genomic DNA was isolated from untransformed C24, AP2M[AtS1 pro :SRKb-FLAG +SCRb], and ap2m[AtS1 pro :SRKb-FLAG+SCRb] plants, and 243-bp AP2M fragments spanning the BsaH1 site were amplified and digested with BsaHI. Figure 4B shows that digestion products of 81 bp and 162 bp were detected in untransformed and AP2M[AtS1 pro :SRKb-FLAG+SCRb] plants ( Figure 4B), as expected. By contrast, the amplification products of ap2m[AtS1 pro :SRKb-FLAG+SCRb] DNA were not digested by BsaHI ( Figure 4B), confirming loss of the BsaHI recognition site. Furthermore, Sanger sequencing of amplification products derived from the ap2m[AtS1 pro :SRKb-FLAG+SCRb] mutant revealed the presence of a frame-shifting cytosine insertion after the 78 th nucleotide of the first exon of AP2M ( Figure 4A). Moreover, amplification of ap2m[AtS1 pro :SRKb-FLAG+SCRb] DNA with SRKb-specific primers confirmed the presence of the AtS1 pro :SRKb-FLAG+SCRb transgenes and amplification with Cas9-specific primers demonstrated the absence of the AP2MsgRNA-CAS9 gene cassette ( Figure 4B).
The phenotypic consequences of the null ap2m mutation were investigated by manual pollination of ap2m[AtS1 pro :SRKb-FLAG +SCRb] stigmas with pollen from untransformed C24 plants or with pollen from an AP2M[AtS1 pro :SRKb-FLAG+SCRb] plant, followed by  Figure 4B) and the red line below the first exon shows the location of the targeted site. In the nucleotide sequences of the region containing the targeted site in wild-type (AP2M) and mutant (ap2m), the sequence of the sgRNA spacer is shown in red characters, the Protospacer Associated Motif (PAM) sequence GGG is underlined, and the BsaHI restriction site is highlighted in bold red characters. The asterisk shows the cytosine insertion in ap2m that abolishes the microscopic monitoring of pollen tube growth. When the stigmas of ap2m[AtS1 pro :SRKb-FLAG+SCRb] plants were pollinated with pollen from untransformed plants (cross-pollination), numerous pollen tubes were observed ( Figure 4C), indicating that the ap2m mutation did not disrupt stigma function. By contrast, when ap2m[AtS1 pro : SRKb-FLAG+SCRb] stigmas were pollinated with SCRb-expressing pollen (self-pollination), an intense SI response identical to that of self-pollinated AP2M[AtS1 pro :SRKb-FLAG+SCRb] stigmas was observed ( Figure 4C). This result indicates that AP2M is not required for the SI response in the transgenic A. thaliana model.

Conclusions
Our attempt to investigate the requirement of CME for SRK signaling in the SI response by eliminating each of the Yxxu motifs in AlSRKb produced mixed results. On the one hand, the observation that the SI response remained intact when Yxxu motif 2 was eliminated clearly demonstrated that this motif is not required for SRK signaling. On the other hand, elimination of Yxxu motif 1 caused by introducing the Y600A mutation resulted in loss of the SI response. This finding cannot be directly related to a defect in CME-mediated receptor trafficking because of the requirement of the Y600 residue for AlSRKb kinase activity. However, the fact that ap2m[AtS1 pro :SRKb-FLAG +SCRb] plants, which harbored a CRISPR/Cas9-induced null mutation in AP2M, retained the ability to reject SCRb-expressing pollen clearly demonstrates that the AP2M-mediated endocytosis pathway is not required for SRKb signaling and for the ability of A. thaliana SRKb-expressing stigmas to mount a robust SI response. This conclusion is consistent with available evidence showing that the SRK receptor signals at the PM.
Our results do not exclude the possibility that SRK is internalized via an alternative clathrin-independent endocytosis (CIE) pathway (Doherty and McMahon 2009;Mayor et al. 2014). A detailed analysis of A. thaliana endocytic pathways concluded that under normal conditions, transmembrane proteins are internalized by CME, while lipid-anchored proteins and lipids are internalized by CIE (Baral et al. 2015). However, CIE has been recently implicated in internalization of BRI1, particularly in its ligand-bound form (Jaillais and Vert 2016;Wang et al. 2015). CIE might similarly function in SRK internalization, possibly as a post-signaling mechanism for attenuating the SI response and desensitizing the stigma epidermal cell to the SCR ligand. Further studies are required to resolve this issue. It should be noted, however, that the SI response differs from other well-studied plant receptor/ligand-mediated processes, such as BRI1 signaling, in which signal termination is essential for desensitizing the cell in readiness for further response to ligand. Rather than being a cell-wide response to ligand, SRK activation and signaling are restricted to the site of contact between a stigma epidermal cell and a "self" pollen grain. Moreover, once on the stigma surface, a pollen grain is not typically dislodged by newly arriving pollen grains (Rea and Nasrallah 2015). These features suggest that internalization for signal attenuation may not play a role in the regulation of SRK signaling.