Abstract
Intercellular communication plays a central role in organogenesis. Tissue morphogenesis in Arabidopsis thaliana requires signaling mediated by a cell surface complex containing the atypical receptor kinase STRUBBELIG (SUB) and the multiple C2 domains and transmembrane region protein QUIRKY (QKY). QKY is required to stabilize SUB at the plasma membrane. However, it is unclear what the in vivo architecture of the QKY/SUB signaling complex is, how it is controlled, and how it relates to the maintenance of SUB at the cell surface. Using a combination of yeast two-hybrid assays and Förster resonance energy transfer (FRET)/fluorescence lifetime imaging microscopy (FLIM) in epidermal cells of seedling roots we find that QKY promotes the formation of SUB homo-oligomers in vivo, a process that appears to involve an interaction between the extracellular domains of SUB. We also show that QKY and SUB physically interact and form a complex at the cell surface in vivo. In addition, the data show that the N-terminal C2A-B region of QKY interacts with the intracellular domain of SUB. They further reveal that this interaction is essential to maintain SUB levels at the cell surface. Finally, we provide evidence that QKY forms homo-multimers in vivo in a SUB-independent manner. We suggest a model in which the physical interaction of QKY with SUB mediates the oligomerization of SUB and attenuates its internalization, thereby maintaining sufficiently high levels of SUB at the cell surface required for the control of tissue morphogenesis.
Introduction
Tissue morphogenesis is a complex process that depends on sophisticated cell-to-cell communication. Receptor kinases (RKs) represent central components of cell-surface receptor complexes involved in extracellular ligand perception in plants (Hohmann et al, 2017). In Arabidopsis, the atypical RK STRUBBELIG (SUB) (Chevalier et al, 2005), also known as SCRAMBLED (SCM) (Kwak et al, 2005), participates in multiple developmental and stress pathways. SUB controls root hair patterning, leaf development, floral organ shape, and ovule development (Chevalier et al, 2005; Kwak et al, 2005; Fulton et al, 2009; Lin et al, 2012). Recently, SUB has also been shown to be required for the stress response elicited by a reduction in cellulose biosynthesis (Chaudhary et al, 2020, 2021). Genetic evidence suggested that SUB functions as a scaffold protein in several cell surface signaling complexes (Vaddepalli et al, 2011; Chaudhary et al, 2021). In support of this notion SUB harbors an ECD with only six LRRs (Vaddepalli et al, 2011) and thus is akin to the ECD of known co-receptors, such as BRI1-ASSOCIATED RECEPTOR KINASE 1/SOMATIC EMBROYGENESIS RECEPTOR-LIKE KINASE 3 (BAK1/SERK3) that carries five LRRs (Li et al, 2002; Nam & Li, 2002). BAK1 and other SERK members are implied in the control of growth, development, and immunity by acting as co-receptors for several different signaling RKs, including BRASSINOSTEROID-INSENSITIVE 1 (BRI1) and FLAGELLIN SENSITIVE 2 (FLS2) (Chinchilla et al, 2007; She et al, 2011; Hothorn et al, 2011; Liu et al, 2020).
A signaling RK with which SUB partners has yet to be identified. QUIRKY (QKY) represents the only known other factor of a SUB signaling complex in vivo. QKY is essential for SUB signaling involved in tissue morphogenesis but plays a minor role in the cell wall stress response (Trehin et al, 2013; Fulton et al, 2009; Song et al, 2019; Chaudhary et al, 2020). Several lines of evidence suggest that QKY and SUB are components of a complex (Trehin et al, 2013; Vaddepalli et al, 2014; Song et al, 2019) and physically interact at plasmodesmata (PD) in epidermal root cells of young Arabidopsis seedlings (Vaddepalli et al, 2014). SUB undergoes constitutive clathrin-dependent endocytosis followed by degradation in the vacuole (Gao et al, 2019; Song et al, 2019) and QKY is also required for the stabilization of SUB at the cell surface (Song et al, 2019; Chaudhary et al, 2021). Apart from its function in SUB-dependent signaling, QKY has been reported to play other roles in growth and development.Genetic evidence indicates that QKY and its maize homolog Carbohydrate Partitioning Defective33 (cpd33) play a role in sugar export from leaves and thus carbohydrate partitioning (Tran et al, 2019). In addition, QKY participates in the control of flowering time (Liu et al, 2019), a process in which SUB does not seem to play a role.
QKY belongs to the family of multiple C2 domains and transmembrane region proteins (MCTPs). MCTPs carry three to four lipid-binding C2 domains in the N-terminal half and multiple transmembrane domains in the C-terminal region (Shin et al, 2005). Representatives of the plant MCTP family were identified in algae and numerous land plant species (Fulton et al, 2009; Liu et al, 2012; Zhu et al, 2020; Liu et al, 2018a; Brault et al, 2019; Tran et al, 2019; Hao et al, 2020; Hu et al, 2021).Several studies indicated multiple members of the MCTP family to be present at PD (Fernandez-Calvino et al, 2011; Liu et al, 2012; Vaddepalli et al, 2014; Tran et al, 2019; Brault et al, 2019). PD are considered plant-specific membrane contact sites (MCS) between the endoplasmic reticulum (ER) and the plasma membrane (PM) (Tilsner et al, 2016). The structure of MCTPs and their presence at PD led to the proposal that some MCTPs function as plasmodesmal membrane tethers (Brault et al, 2019). In addition to QKY, the function of several other plant MCTPs has been described. MCTP1/FT-INTERACTING PROTEIN 1 (FTIP1) controls flowering time in Arabidopsis and rice (Liu et al, 2012; Song et al, 2017), while MCTP3 and MCTP4 affect meristem development (Liu et al, 2018b; Brault et al, 2019), as do GhMCTP7, GhMCTP12, and GhMCTP17 in cotton (Hu et al, 2021).
Many aspects of the QKY/SUB signaling complex in its native cellular environment are unknown. In this study, we provide an in-depth description of the QKY subcellular localization. We further tested the ability of SUB and QKY to form multimers. We mapped the SUB interaction domain of QKY and analyzed its topology in the SUB/QKY complex. Thus, we provide new molecular insights into the QKY-dependent control of SUB complex architecture in vivo.
Results
QKY localizes to the ER and to PD
Experiments involving stable transgenic Arabidopsis plants carrying a functional QKY:EGFP reporter indicated a PD localization for QKY (Vaddepalli et al, 2014). Transient expression studies in Nicotiana benthamiana leaves frequently suggested ER localization of MCTP proteins, but also indicated a number of different subcellular associations, depending on the family member (Liu et al, 2018a; Brault et al, 2019; Liu et al, 2019; Tran et al, 2019; Hu et al, 2021). In particular, they suggested that QKY is localized to the ER, intracellular vesicles, PM, and cytoplasm. We reinvestigated the subcellular localization of QKY in its native context in Arabidopsis cells. First, we transformed the qky-9 mutant, harboring a null allele (Fig. S1), with a reporter carrying an N-terminal translational fusion of mCherry to QKY driven by its endogenous promoter (pQKY::mCherry:QKY) (see Materials and Methods) and analyzed the phenotype of 17 independent transformant lines (Table1). We found that pQKY::mCherry:QKY robustly restored the floral tissue and root hair patterning defects of qky-9 (Table1, Table 2) (Fig. S2).
Second, we analyzed the subcellular localization of the pQKY::mCherry:QKY reporter in six of these lines in live tissues by confocal laser scanning microscopy (CLSM). We observed a modest signal with undistinguishable subcellular signal distribution between all six lines. For further analysis we used line 6 which exhibited average signal intensity. We detected reporter signals in several tissues, including the root epidermis of 6-day seedlings and several stage 13 floral organs (Fig. 1) (floral stages according to (Smyth et al, 1990)). As previously reported, we observed a spotty PD-related pattern along the cell periphery of the root epidermis of seedlings (Fig. 1A-C) (Vaddepalli et al, 2014). In addition, we detected a faint ER-like signal. A similar pattern, though with a somewhat less prominent ER-like and PD-related signal, was observed in the epidermis of mature sepals, petals, and carpels (Fig. 1D-L). An endosomal localization of a p35S::GFP:QKY reporter was observed after transient expression in tobacco leaves in line with its suggested function in the vesicular transport of FLOWERING TIME (FT) in leaf phloem companion cells (Liu et al, 2019). We never observed intracellular vesicle-like signals in the analyzed cells of our Arabidopsis lines with our reporters. Third, we performed a co-localization experiment. As a control, we stained lateral root cap cells of 6-day seedlings expressing Q4, a well-established marker for the ER (Cutler et al, 2000; Cutler & Ehrhardt, 2002), with the membrane marker FM4-64 for 5 minutes, thereby labeling the PM, and investigated the subcellular distribution of Q4 and FM4-64-derived signals. We detected very little if any overlap between two signals in these cells (Fig. 1M-Q). We then crossed Q4 and pQKY::mCherry:QKY lines and analyzed the subcellular distribution of the two reporters in the same cells of F1 6-day seedlings. pQKY::mCherry:QKY localization in lateral root cap cells exhibited a prominent ER-like pattern (Fig. 1Q,R) and we could readily observe co-localization of Q4 and pQKY::mCherry:QKY signals in these cells (Fig. 1R-V). In summary, our data are compatible with the notion that QKY localizes to the ER and PD in Arabidopsis cells.
QKY undergoes SUB-independent homo-oligomerization
Maize MCTP CPD33 was reported to form homodimers in bimolecular fluorescence complementation (BiFC) assays in tobacco leaf cells (Tran et al, 2019). We therefore investigated whether QKY forms homo-multimers. We first performed a yeast two-hybrid (Y2H) experiment in which the region spanning the C2A to C2D domains (C2A-D) served as bait and prey. We observed yeast growth on the selective medium, indicating that this region can interact with itself in this assay (Fig. 2). To map the interaction domain, we generated a series of deletions within the C2A-D domain and tested their ability to interact with the C2A-D domain. The results indicate that the C2A-B domain is necessary for interaction in yeast (Fig. 2).
Next, we tested if QKY can form homo-multimers in its native environment. To this end, we tested the interaction in epidermal root cells of 5-day seedlings using steady-state Förster resonance energy transfer (FRET)/fluorescence lifetime imaging microscopy (FRET/FLIM). We analyzed 5-day seedlings expressing functional pUBQ10::EGFP:QKY or pUBQ10::EGFP:QKY and pUBQ10::mCherry:QKY reporters. When we analyzed the subcellular distribution of the PD-related pQKY::mCherry:QKY signals, we detected stronger signals along the longitudinal cell periphery (Fig. S3A) (Vaddepalli et al, 2014). Therefore, we distinguished between the transverse PM, the longitudinal PM, and the PD located along the longitudinal PM. We found that the mean fluorescence lifetime of pUBQ10::EGFP:QKY was significantly reduced when UMQ was present indicating that pUBQ10::EGFP:QKY forms homo-oligomers in Arabidopsis cells (Fig. 3A). As negative control, we used a line co-expressing the PM-localized p35S::Lti6a:GFP (Lti6a:GFP) (Cutler et al, 2000) and p2×35S::Lti6b:2xmCherry (Lti6b:2xmCherry) (Noack et al, 2022) reporters. We never observed significant mean fluorescence lifetime reductions for Lti6a:GFP in the presence of Lti6b:2xmCherry (Fig. 3B). The combined results indicate that QKY, and possibly other MCTP proteins, undergo homo-oligomerization in vivo.
We then explored whether QKY oligomerization was dependent on SUB. To this end we measured the steady-state fluorescence anisotropy (FA) of pUBQ10::EGFP:QKY in epidermal root cells of 5-day wild-type and sub-1 seedlings. Fluorescence anisotropy is a measure of the rotational freedom of a fluorescent molecule, such as GFP. Upon protein homo-oligomerization of GFP-based fusion proteins homo-FRET can occur resulting in a decrease in fluorescence anisotropy (Bader et al, 2011; Weidtkamp-Peters & Stahl, 2017) (Fig. S4). This approach has been successfully applied in receptor kinase interaction studies involving for example CLAVATA1 or BAK1 (Somssich et al, 2015; Stahl et al, 2013). We did not find any significant differences in the FA values of pUBQ10::EGFP:QKY in wild type compared with sub-1 (Fig. 3C). The combined results support the model that QKY undergoes SUB-independent homo-oligomerization in vivo and that this interaction is mediated by the C2A-B domain.
QKY physically interacts with SUB in vivo
Previous results suggested the physical interaction of QKY and SUB at PD (Vaddepalli et al, 2014). To confirm the QKY/SUB interaction we generated novel lines expressing a newly generated translational fusion of SUB to GFP driven by the endogenous SUB promoter (pSUB::SUB:GFP) and the functional mCherry::QKY fusion driven by the endogenous QKY promoter (pQKY::mCherry:QKY). The pSUB::SUB:GFP reporter is functional as it complements the null allele sub-1 (Fig. S5). Taking advantage of an advanced microscopy set up (see Materials and Methods) we performed FRET/FLIM experiments in root epidermal cells of 5-day seedlings (Fig. 4A-I). As outlined above, we distinguished between the transverse PM, the longitudinal PM, and the PD located along the longitudinal PM. We detected a significant decrease in the mean fluorescent lifetime of pSUB::SUB:GFP in the presence of pQKY::mCherry:QKY. We also observed a polar distribution of the reduction in mean fluorescence lifetime of pSUB::SUB:GFP in the presence of pQKY::mCherry:QKY (Fig. 4I). The reduction in pSUB::SUB:GFP mean fluorescence lifetime was more pronounced at the longitudinal PM than at the transverse PM, consistent with the observed polar distribution of signal intensity of pSUB::SUB:GFP and pQKY::mCherry:QKY. We could not detect an obvious difference in the reduction of pSUB::SUB:GFP mean fluorescence lifetime when we compared the longitudinal PM with the PD-related signals along the longitudinal PM (Fig. 4I). As negative control, we used a line co-expressing pSUB::SUB:GFP and the Lti6b:2xmCherry reporter. We never observed significant mean fluorescence lifetime reductions for pSUB::SUB:GFP in the presence of Lti6b:2xmCherry (Fig. 4I). The results support the notion of a physical interaction between QKY and SUB in vivo. Interestingly, they also suggest that QKY and SUB preferentially interact at the longitudinal edges of epidermal root cells.
The C2A-B domain of QKY is required for interaction with SUB in vivo
Next, we mapped the SUB interaction domain of QKY. We first confirmed previous results (Vaddepalli et al, 2014) that the region comprising the C2A-D domain of QKY can interact with the intracellular domain of SUB in a Y2H assay (Fig. 5A-B). To map the interaction domain in more detail, we then generated a series of deletions within the C2A-D domain of QKY and tested their ability to interact with the intracellular domain of SUB in a Y2H assay. We found that the 187 amino acid linker region (L) separating the C2A and C2B domains was sufficient for interaction in this assay.
Next, we generated qky-9 lines homozygous for different pQKY::mCherry:QKY variants exhibiting sequential deletions up to and including the C2D domain. We observed a PD-like localization pattern for pQKY::mCherry:QKY variants carrying a deletion of the C2A-B region. Interestingly, however, the stronger PD-related signals along the longitudinal cell periphery of pQKY::mCherry:QKY were not observed in the pQKY::mCherry:QKYdelC2A or pQKY::mCherry:QKYdelC2A-B variants (Fig. S3A-I). pQKY::mCherry:QKY variants that carried additional deletions of the C2C and C2D domains exhibited an ER-like pattern (Fig. S3J-O). The results suggest that the C2A-B domain of QKY is not required for PD localization but is involved in the preferential accumulation of QKY along the longitudinal periphery of epidermal root cells, that the C2C-D region mediates PD localization, and that the TMR is required for anchoring of QKY in the ER.
To test if the C2A-B domain was required for physical interaction with SUB in vivo we performed FRET/FLIM experiments. By crossing, we generated F1 seedlings carrying one copy of pSUB::SUB:GFP and one copy of different pQKY::mCherry:QKY variants. Compared with the mean fluorescent lifetime of pSUB::SUB:GFP in combination with pQKY::mCherry:QKY, we found a small increase in pSUB::SUB:GFP mean fluorescence lifetime when pSUB::SUB:GFP was combined with the pQKY::mCherry:QKYdelC2A construct in root epidermal cells of 5-day seedlings (Fig. 5C-E). However, its mean fluorescence lifetime was still significantly reduced compared to the pSUB::SUB:GFP control in the absence of pQKY::mCherry:QKY. When pSUB::SUB:GFP was combined with pQKY::mCherry:QKYdelC2AL and pQKY::mCherry:QKYdelC2A-B reporters, its mean fluorescence lifetime returned to the level of pSUB::SUB:GFP control in the absence of pQKY::mCherry:QKY (Fig. 5C-E). We obtained similar results for the transverse and longitudinal PMs and the longitudinal PD. Unfortunately, mapping the interaction domain of SUB is difficult to do in vivo as SUB variants carrying alterations in its ECD or intracellular domain are retained in the ER where they undergo ER-associated degradation (ERAD) (Vaddepalli et al, 2011; Hüttner et al, 2014).
Physical interaction between QKY and SUB is required to ensure high SUB levels at the cell surface
QKY-mediated stabilization of SUB at the cell surface (Song et al, 2019; Chaudhary et al, 2021) (Fig. 6) counteracts the constitutive endocytosis and degradation of SUB (Gao et al, 2019; Song et al, 2019). It remains unclear, however, if physical interaction between QKY and SUB is required for this process. The pQKY::mCherry:QKY variants carrying progressive N-terminal deletions of the C2 domains failed to complement the qky-9 phenotype (Fig. 6) (Table1, Table 2) (Fig. S6) indicating that all C2 domains are necessary for QKY function. To test whether the C2A-B SUB interaction domain controls the amount of SUB at the cell surface, we investigated the effect of the pQKY::mCherry:QKYdelC2A, pQKY::mCherry:QKYdelC2AL, and pQKY::mCherry:QKYdelC2A-B deletion constructs on the level of pSUB::SUB:GFP signal at the PM of root epidermal cells in qky-9 seedlings. We found that all three constructs were unable to rescue pSUB::SUB:GFP levels (Fig. 6B,C). The results support the notion that the C2A-B domain is essential for maintaining high cell surface SUB levels required for the control of root hair patterning and floral morphogenesis. They also suggest that this process requires the physical interaction between QKY and SUB.
SUB undergoes QKY-dependent homo-oligomerization
There is evidence that plant RKs such as BRI1 can form homo-oligomers (Russinova et al, 2004; Wang et al, 2005, 2015a). One way QKY could affect the architecture of SUB complexes at the cell surface is by controlling the oligomerization of SUB. Therefore, we investigated whether SUB undergoes homo-oligomerization. We first performed Y2H experiments with the ECD or intracellular domains of SUB. Robust yeast growth was observed on the selective medium when the ECD, but not the intracellular domain, was present as bait and prey (Fig. 7A, Fig. S7). The result demonstrates that the ECD of SUB can interact with itself in this assay. In complementary experiments we tested for interaction in root epidermal cells of 5-day seedlings expressing functional pSUB::SUB:EGFP (Vaddepalli et al, 2011) or pSUB::SUB:EGFP and pUBQ10::SUB:mCherry (Chaudhary et al, 2020) reporters. In FRET/FLIM experiments we observed a clear reduction in the mean fluorescence lifetime of pSUB::SUB:EGFP in the presence of pUBQ10::SUB:mCherry indicating that SUB forms homo-oligomers in vivo (Fig. 7B). Next, we explored if SUB oligomerization depended on QKY. To this end we measured the steady-state fluorescence anisotropy (FA) of pSUB::SUB:EGFP in root epidermal cells of 5-day wild-type and qky-9 seedlings. We observed a significant increase in the FA value of pSUB::SUB:EGFP in qky-9 compared with wild type (Fig. 7C). Moreover, when comparing the FA values of pSUB::SUB:GFP in qky-9 or pSUB::SUB:GFP pQKY::mCherry:QKY qky-9, we found a significant decrease in pSUB::SUB:GFP FA value when pQKY::mCherry:QKY was present (Fig. 7D-L) supporting the notion that QKY promotes SUB oligomerization. We hypothesized that higher levels of SUB could lead to stronger SUB oligomerization. To test this assumption, we investigated whether the amount of SUB present at the PM affects the homo-oligomerization of SUB. To this end, we compared the FA of pSUB::SUB:EGFP in wild type and sub-1, a null allele of SUB (Chevalier et al, 2005). We found a modest and statistically nonsignificant increase in pSUB::SUB:EGFP FA in sub-1, suggesting that SUB concentration at the PM has a small, if any, effect on SUB oligomerization (Fig. 7C).
Discussion
The available data support the notion of a modular structure of QKY, in which individual domains or their combinations perform specific functions (Fig. 8A). We showed that the N-terminal C2A-B region is required for the physical interaction of QKY with the RK SUB in vivo. The combination of C2C-D and TMR domains targets QKY to PD and the TMR domain is necessary to anchor QKY to the ER, as is the case for other MCTPs (Fig. S3) (Vaddepalli et al, 2014; Brault et al, 2019; Liu et al, 2018a).
The data presented here and in previous work (Vaddepalli et al, 2014; Liu et al, 2019) reveal that reporters carrying N-terminal fusions of EGFP or mCherry to QKY driven by the endogenous QKY or pUBQ10 promoters can robustly complement the phenotypes of qky-8, qky-9, qky-14, and qky-17 mutants, demonstrating their functionality. Similarly, reporters containing N-terminal fusions of YFP to AtMCTP3 rescued the mctp3 mctp4 phenotype (Brault et al, 2019). The collective data contrast with a recent study reporting that an N-terminal fusion of GFP to QKY driven by the endogenous promoter fails to complement the qky-16 phenotype (Song et al, 2019). We do not know the reason for the discrepancy in the results, but suspect that qky-16 is not a straightforward hypomorph but possibly a dominant-negative allele. There is a discrepancy in the literature about which domain of QKY interacts with SUB. In one study, the C2A-D domain of QKY was reported to interact with the ECD of SUB in a Y2H assay. This result in combination with proteinase K-digestions of protoplasts obtained from transgenic lines harboring tagged QKY and SUB reporters was interpreted to indicate that the C2A-D domain is extracellularly localized (Song et al, 2019). By contrast, we had shown earlier that the C2A-D domain of QKY interacts with the SUB intracellular domain in a Y2H assay (Vaddepalli et al, 2014). Moreover, we failed to detect interaction between the C2A-D domain of QKY and the ECD of SUB in multiple independent Y2H assay repetitions (Fig. S8). Finally, our two independent FRET/FLIM studies strongly suggest that the physical interaction between QKY and SUB in vivo involves the C2A-D domain of QKY and the intracellular domain of SUB, as the fluorescent tags were added to the N-terminus of QKY and the C-terminus of SUB, respectively (Fig. 4, this study) (Vaddepalli et al, 2014). Both studies of the in vivo QKY-SUB interaction were performed in transgenic Arabidopsis seedlings carrying functional reporters and investigated this interaction in root epidermal cells, a cell type that requires QKY and SUB function for its development (Kwak et al, 2005; Fulton et al, 2009). The combined results from our independent lines of experiments strongly suggest that the C2A-D domain of QKY localizes to the cell interior.
Our results provide deeper insight into the molecular architecture of the QKY/SUB receptor complex in vivo. Results from the Y2H analysis suggest that the N-terminal linker region of QKY flanked by the C2A and C2B domains is sufficient for the physical interaction between QKY and SUB. The FRET/FLIM data confirm a central role of the linker in the interaction in vivo. However, they also suggest that the C2A and C2B domains play an additional, albeit minor, role in this interaction, as the stepwise deletion of both domains has small but noticeable effects on the interaction. It remains to be determined whether the two domains represent additional but less important interaction interfaces or whether they form a necessary structural scaffold within QKY that allows the linker to interact with SUB. X-ray crystallography of QKY/SUB complexes will provide the answer to this question.
The available evidence strongly supports the notion that the physical interaction between QKY and SUB in Arabidopsis epidermal root cells occurs in the cytoplasm, with the C2A-B domain of QKY and the intracellular domain of SUB playing a central role in this process. The topology is consistent with the ER-localization of QKY and the PM-localization of SUB (Yadav et al, 2008) if one assumes that physical interaction between QKY and SUB in these cells, and possibly in other cell types, is limited to sites where the cortical ER and the PM are close enough to allow interaction (Fig. 8B). This would be the case in case of PD and is in line with our earlier model, which states that the physical interaction between QKY and SUB is confined to PD (Vaddepalli et al, 2014). However, this model does not explain the overall reduction in SUB at the cell periphery in qky mutants, which includes PD and the sections of the PM between the PD, unless one postulates that the QKY/SUB complex at the PD indirectly maintains SUB concentration at the PM between the PD. We propose a parsimonious model to explain the effect of QKY on SUB levels at the PM. We hypothesize that QKY interacts with SUB not only at PD but also at other sites along the cell periphery where the cortical ER carrying QKY is closely enough associated with the PM to allow physical interaction between QKY and SUB (Fig. 8B). In this model, physical interaction between QKY and SUB would be involved in maintaining a high level of SUB at the PM not only at PD but at numerous sites along the cell periphery.
The molecular mechanism of how QKY stabilizes SUB at the cell surface is poorly understood. Our results indicate that SUB undergoes QKY-dependent homo-oligomerization. Based on these results, we propose the following model. SUB monomers form homo-oligomers in a QKY-dependent manner at the PM and PD in epidermal root cells and possibly in other cell types. SUB homo-oligomers are less amenable to ubiquitination and internalization of SUB, a process which is inhibited by QKY (Song et al, 2019). For example, the ubiquitination site of SUB or its clathrin binding motif may be masked in a QKY/SUB complex. SUB oligomerization could be achieved via two non-mutually exclusive pathways. In one scenario, SUB clusters are initiated by a physical interaction between ECDs and stabilized by a subsequent physical interaction of intracellular domains with the C2A-B domain of QKY. Alternatively, an initial QKY-mediated association of the intracellular domains of the SUB may be followed by stabilization of the clusters through interactions between the ECDs. A number of questions remain. For example, different stoichiometries of the QKY/SUB complex are conceivable (Fig. 8C), and it is unclear which of these occur in vivo. It further remains to be investigated if QKY homo-oligomerization is relevant for QKY-SUB interaction and/or stabilization of SUB at PM. We will address these and other aspects in future work.
Materials and Methods
Plant work and genetics
Arabidopsis thaliana (L.) Heynh. var. Columbia (Col-0) and var. Landsberg (erecta mutant) (Ler) were used as wild-type strains. Plants were grown essentially as described previously (Fulton et al, 2009). Plate-grown seedlings were grown in long-day conditions (16 h light/8 h dark) on half-strength Murashige and Skook (1/2 MS) agar plates supplemented with 1% sucrose. The mutant alleles sub-1 (Ler), qky-9 (Ler) and sub-9 (Col) have been described previously (Chevalier et al, 2005; Fulton et al, 2009; Vaddepalli et al, 2011). The qky-17 (Ler) allele was generated using a CRISPR/Cas9 system in which the egg cell-specific promoter pEC1.2 controls Cas9 expression (Wang et al, 2015b). Two single-guide RNAs (sgRNA), sgRNA1 (5’-ACTCGGATCCTCCGCCGTCG-3’) and sgRNA2 (5’-TTACGACGAGCTCGATATCG-3) were employed. sgRNA1 binds to the region +20 to +39, while sgRNA2 binds to the region +237 to +256 of the QKY coding sequence. The sgRNAs were designed according to the guidelines outlined in (Xie et al, 2014). The qky-17 mutant carries a frameshift mutation at position 36 relative to the QKY start AUG, which was verified by sequencing. The resulting predicted short QKY protein comprises 60 amino acids. The first 12 amino acids correspond to QKY, while amino acids 13-60 represent an aberrant amino acid sequence. The lines carrying pSUB::gSUB:EGFP (SSE, in sub-1, Ler, sub-9), pUBQ10::gSUB:mCherry (USM, in Col), pUBQ10::EGFP:QKY (UEQ, in sub-1, Ler), pUBQ10::mCherry:QKY (UMQ, in Ler) have been previously reported (Chaudhary et al, 2020; Vaddepalli et al, 2011, 2014). The reporter lines Q4, Lti6a:GFP, Lti6b:2xmCherry, TMO7:1xGFP, and TMO7:3xGFP have been described previously (Cutler et al, 2000; Noack et al, 2022; Schlereth et al, 2010). The reporter constructs pGL2::GUS:GFP and pQKY::mCherry:QKY have been described earlier (Gao et al, 2019; Vaddepalli et al, 2014). Wild-type, sub-1 and qky-9 plants were transformed with different constructs using Agrobacterium strain GV3101/pMP90 (Koncz & Schell, 1986) and the floral dip method (Clough & Bent, 1998). Transgenic T1 plants were selected on either kanamycin (50 μg/ml) or hygromycin (20 μg/ml) plates and transferred to soil for further inspection. All the crossed materials used in this study were F1 seeds.
Recombinant DNA work
For DNA work, standard molecular biology techniques were used. PCR fragments used for cloning were obtained using Q5 high-fidelity DNA polymerase (New England Biolabs, Frankfurt, Germany). All PCR-based constructs were sequenced. Primer sequences used in this work are listed in Table S1. The pSUB::gSUB:GFP construct was assembled using the GreenGate system (Lampropoulos et al, 2013). The sequences of pSUB and gSUB were amplified as previously described (Yadav et al, 2008; Vaddepalli et al, 2011). Other sequences, including GFP and the plant resistance modules, were available from the GreenGate vectors. The pSUB::gSUB:GFP was assembled into the intermediate vectors and then combined into the pGGZ0001 destination vector with a standard GreenGate reaction. The pCambia2300-based progressive N-terminal deletions of QKY carrying EGFP driven by the UBQ10 promoter (pUBQ10::mCherry:QKYdel) were described previously (Vaddepalli et al, 2014). To generate the deletion constructs of QKY fused to mCherry driven by the endogenous QKY promoter (pQKY::mCherry:QKYdel), the fragment of pQKY::mCherry was digested with KpnI/SpeI from plasmid pQKY::mCherry:QKY, and subcloned into the KpnI/SpeI digested pUBQ::mCherry:QKYdel. For the generation of pGADT7-SUB carrying ECD or ICD, the fragment of pGBKT7-SUB carrying the ECD or ICD was digested by NdeI/XmaI and subcloned into the NdeI/XmaI-digested pGADT7. Similarly, in order to generate pGBKT7-QKYC2A-D, the fragment of pGADT7-QKYC2A-D was digested by NdeI/XmaI and subcloned into pGBKT7 digested with NdeI/XmaI. In order to generate the pGADT7 carrying the various truncated fragments of QKY, the different truncated fragments of QKY were amplified with the following primers: NdeI/QKY(-TM)_F and QKY(C2A-B)/XmaI_R for the fragment of C2A-B, QKY(C2C-D)/NdeI_F and QKY(-TM)/XmaI_R for the fragment of C2C-D, NdeI/QKY(-TM)_F and C2ALinker_XmaI_R for the fragment of C2AL, NdeI/QKY(-TM)_F and C2A_XmaI_R for the fragment of C2A, C2ALinker_NdeI_F and C2ALinker_XmaI_R for the fragment of L. Fragments were digested by NdeI/XmaI and subcloned into the NdeI/XmaI digested pGADT7. All clones were verified by sequencing. Primers are listed in Table S1.
Yeast two-hybrid assay
The Matchmaker yeast two-hybrid system (Takara Bio Europe, Saint-Germain-en-Laye, France) was employed and experimental procedures followed the manufacturer’s recommendations. The pGBKT7 plasmids containing either the SUB extracellular (ECD) or intracellular (ICD) domain were described previously (Bai et al, 2013). The construct pGADT7-QKYC2A-D is equivalent to a previously described construct pGADT7-QKYΔPRT_C (Vaddepalli et al, 2014). In order to assess possible interactions in yeast, the different combinations of pGBKT7 and pGADT7 plasmids were co-transformed into the yeast strain AH109. Transformants were selected on synthetic complete (SC) medium lacking leucine and tryptophan (-LW) at 30°C for 3 days. To examine yeast-two-hybrid interactions, the transformants were grown on solid SC medium lacking leucine and tryptophan (SC-LW) or leucine, tryptophan and histidine (-LWH). Standard -LWH growth medium was supplemented with 2.5 mM 3-amino-1,2,4-triazole (3-AT) to minimize false positive signals. Yeast were grown for 3 days at 30°C.
Microscopy
Floral organs and siliques were imaged using a Leica SAPO stereo microscope equipped with a digital MC 170 HD camera (Leica Microsystems GmbH, Wetzlar, Germany). Clearing and imaging of ovules was performed as reported previously (Vijayan et al, 2021). Ovule morphology was investigated using an Olympus BX61 upright microscope. For the analysis of root epidermal cell patterning, 6-days-old seedlings were imaged using an Olympus BX61 upright microscope equipped with an XM10 monochrome camera (Olympus Europe, Hamburg, Germany). The number of H-position and N-position cells in at least ten seedlings were scored and the relative ratio of H-position with hair and N-position without hair were deduced. For the quantitative analysis of pGL2::GUS:GFP-expressing cells in the root epiderms, 6-days-old seedlings of different genotypes carrying the pGL2::GUS:GFP reporter were counterstained with 5 μg/ml propidium iodide and examined using a FV3000 confocal laser scanning microscope (Olympus Europe, Hamburg, Germany). A high sensitivity detector (HSD) and a 60x water immersion objective (NA 1.2) were employed. Scan speed was set at 4.0 μs/pixel (image size 1024×1024 pixels), line average at 2, and the digital zoom at 1. GFP was excited using a 488 nm diode laser (2% intensity) and emission was detected at 500 to 540 nm. Propidium iodide was excited using a 561 nm diode laser (1% intensity) and emission detected at 584-653 nm. To assess pQKY::mCherry:QKY (QMQ) subcellular localization, confocal microscopy was performed on root epidermal cells of 6-days-old seedlings using an Olympus FV1000 confocal microscope and on epidermal cells of several stage 13 floral organs using an Olympus FV3000 confocal microscope. High sensitivity detectors (HSDs) and a 60x water immersion objective (NA 1.2) were employed in both microscopes. Scan speed was set at 4.0 μs/pixel (image size 1024×1024 pixels) and line average at 2. The mCherry fluorescence excitation was performed with a 561 nm diode laser (5% intensity) and detected at 584-653 nm. For the colocalization of mCherry:QKY with the ER marker Q4, 6-days-old seedling roots were imaged using an Olympus FV3000 confocal microscope equipped with HSD detectors. In some instances, the seedlings of the ER marker Q4 were counterstained with 4 μM FM4-64 (Molecular Probes) for 5 minutes. A 60x water immersion objective (NA 1.2) was employed. Scan speed was set at 2.0 μs/pixel (image size 1024×1024 pixels), line average at 2, and the digital zoom at 2. The mCherry fluorescence and the FM4-64 stain were excited using a 561 nm diode laser (7% intensity for mCherry and 2% intensity for FM4-64) and emission was detected at 584-653 nm. GFP was excited using a 488 nm diode laser (1% intensity) and emission was detected at 500-540 nm. To observe the subcellular localization of pQKY::mCherry:QKY mutant variants, the root epidermal cells of 6-days-old seedlings were imaged using an Olympus FV3000 confocal microscope equipped with a HSD and a 60x water immersion objective (NA1.2). Scan speed was set at 4.0 μs/pixel (image size 1024×1024 pixels) and the digital zoom at 4. The mCherry fluorescence excitation was performed with a 561 nm diode laser (7% intensity) and detected at 584-653 nm. Images were adjusted for color and contrast using ImageJ/Fiji software (Schindelin et al, 2012). For the quantification of SUB-GFP accumulation at the PM, epidermal cells of root meristems of 6-days-old seedlings in the corresponding plant line were imaged using an Olympus FV3000 microscope equipped with a HSD and a 60x water immersion objective (NA 1.2) at digital zoom 4. Scan speed was at 4.0 μs/pixel (image size 1024×1024 pixels) and line average at 2. GFP was excited using a 488 nm diode laser (1% intensity) and emission was detected at 500 to 540 nm. For the direct comparisons of fluorescence intensities, laser, pinhole, and gain settings of the confocal microscope were kept identical when capturing the images in different QKY mutant backgrounds. The mean gray values of GFP fluorescence signal intensity at the PM were measured in 5-10 cells from each image by creating a region of interest covering the PM area using ImageJ/Fiji.
FRET-FLIM and fluorescence anisotropy measurements
FRET-FLIM was performed using an Olympus FV3000 microscope equipped with a time-correlated single photon counting (TCSPC) device (LSM upgrade kit, PicoQuant, Berlin, Germany). Root epidermal cells of 5-days-old seedlings co-expressing SUB:GFP and mCherry:QKY were imaged with a 60x water immersion objective (NA 1.2) at digital zoom 4. Scan speed was 4.0 μs/pixel (image size 512×512 pixels). The FLIM filter cube DIC560 for GFP/RFP was employed. GFP fluorescence lifetimes were measured with two photon-counting PMA hybrid 40 detectors and a pulsed 485 nm diode laser (LDH-D-C-485, PicoQuant) using a laser pulse rate of 40 MHz. For each image, a minimum of 300 photons per pixel were acquired with a TCSPC resolution of 25.0 ps (image size 512×512 pixels). Results were analyzed using SymPhoTime 64 software 2.7 (PicoQuant) using n-exponential reconvolution and an internally calculated instrument response function (IRF). The lifetime fitting model parameter as 1 was defined (n=1). The analysis results with a correctional factor (χ2) between 0.9 and 1.9 were accepted. Intensity-weighted average lifetime (τ Av Int) of membrane ROIs was taken as the final value for each FLIM image. Steady-state fluorescence anisotropy was performed and analyzed as reported previously (Chaudhary et al, 2021; Chaudhary & Schneitz, 2022). Note that if one has to work with weak reporter expression and therefore around the minimum number of detected photons required for a meaningful result (on average 15 photons per pixel of a ROI detected in the channel belonging to the perpendicular detector of the PicoQuant system (channel 1 in the SymPhoTime 64 software 2.7), maximum scan time of 5 min), the actual calculated anisotropy value is also influenced to some extent by the number of photons collected. In such a scenario, a lower number of photons leads to a low FA value, and a higher number of photons leads to a higher FA value.
Statistical analysis
Statistical analysis was performed with PRISM 9.4.0 software (GraphPad Software, San Diego, CA, USA). All statistical tests and P-values are described in the respective figure legends.
Funding
This work was funded by the German Research Council (DFG) through grant SFB924 (TP A2) to KS. XC was supported by a research fellowship from the Chinese Science Council (CSC).
Supplement
Acknowledgements
We thank members of the Schneitz lab for helpful comments and discussion. We also thank Farhah Assaad (Technical University of Munich), Ziqiang Patrick Li and Emmanuelle Bayer (CNRS-University of Bordeaux) for the Q4, Lti6a:GFP and Lti6b:2xmCherry reporters. We thank Ajeet Chaudhary for help with the yeast two-hybrid method, and acknowledge Rachele Tofanelli and Tejasvinee Mody for their help with ovule imaging. We further thank Ajeet Chaudhary and Sebastian Wolf for suggestions and comments. We also acknowledge support by the Center for Advanced Light Microscopy (CALM) of the TUM School of Life Sciences.