PREreview del Cysteine-rich receptor-like kinases mediate Wall Teichoic Acid perception in Arabidopsis

Pierdzig et al. report the identification of wall teichoic acids (WTAs) from Gram-positive bacteria as microbe-associated molecular patterns (MAMPs) perceived by the Arabidopsis cysteine-rich receptor kinase CRK7. Through an impressive forward genetic screen combined with biochemical fractionation, the authors identify WTAs as the eliciting molecules and demonstrate that CRK7 is required for WTA-triggered responses, including transcriptional reprogramming, priming, and programmed cell death. The study further implicates the related receptor kinase CRK43 as a signaling partner of CRK7 and places EDS1–PAD4–ADR1 helper NLR modules downstream of this pathway. Structural modeling and interaction assays suggest homomeric or higher-order associations of CRK7. Moreover, the authors show ligand-induced conformational changes which lead to the dissociation of a pre-formed CRK7-CRK43 complex, providing a framework for receptor activation. Together, this work identifies the first MAMP perceived by a CRK-family receptor and uncovers a non-canonical, lineage-specific immune signaling pathway linking cell-surface perception of Gram-positive bacteria to intracellular immune hubs. Overall, the study is ambitious, technically sophisticated, and addresses an important gap in our understanding of plant immune receptor diversity.

The story is largely complete, and the breadth of approaches is commendable. We have a few comments and suggestions that we hope will be useful to the authors to improve this high-quality manuscript even further, particularly with regards to the molecular mechanistic work performed towards the end of the story. 

General Comments

-This is a very impressive and ambitious study, particularly the genetic and biochemical fractionation screen leading to the identification of wall teichoic acids (WTAs) from Gram-positive bacteria as MAMPs and CRK7 as their receptor. To our knowledge, this represents the first reported and genetically supported evidence that a MAMP is perceived by a CRK-family receptor, which is an important conceptual advance which we feel could be highlighted more.

-The genetic and biochemical screen used to identify the MAMP and its receptor is outstanding and a major strength of the paper.

-The story is largely complete and technically sophisticated, but the mechanistic/biochemical section toward the end feels comparatively weaker. Some of the experiments in figure 4 would benefit from additional controls, clarification, or toned-down conclusions.

-Several key and exciting results are buried in supplementary figures and could reasonably be promoted to the main figures if there are no space constraints.

-The Discussion feels short and could benefit from deeper conceptual framing and speculation, particularly regarding lineage specificity and potential signaling mechanisms.

Introduction

-The Introduction is very well written and well structured, and does an excellent job of motivating the study and introducing most of the information required to interpret the results.

-The section introducing EDS1, PAD4, ADR1, and helper NLR signaling (currently in results) might be better placed (or at least previewed) here, rather than appearing abruptly in the Results.

Results

Figure 1

-The manuscript discusses programmed cell death (PCD) very early. It would be helpful to show more clearly in Figure 1 that the observed phenotype truly reflects cell death rather than chlorosis or general tissue damage. This distinction is addressed later (Figure 2, with the conductivity assays), but clarifying this earlier would improve the flow of the story.

-For the bacterial infections (panel a and C), it would be useful to specify what type of cell death is being observed. It is not clear whether this is host-programmed (in the case of B. subtilis) or pathogen-driven damage (in the case of Pst DC3000).

Figure 2 and associated supp (S1 to S7)

-Figure 2 panel D. It was nice to see that the 1ug/ml treatment exhibits a similar effect than the treatment with B. subtilis. We also noted  a decrease rather than a plateau when increasing the WTA concentration beyond 1ug/ml, do the authors have any hypotheses/speculation on this observation?

Figure 3 and associated supp (S8 to S12)

-It would be helpful to provide more clarity on what CRK43 is. The manuscript initially introduces CRK43 as an RLK but then classifies it as an RLCK based on the fact that it is truncated and lacks a ligand-binding and transmembrane domain. It would be good to understand how this classification is phylogenetically supported.

-The proposed constitutive interaction followed by ligand-induced dissociation between CRK7 and CRK43 is a very exciting mechanism. One question we had when reading the story is: Why is CRK43 not autoactive in a crk7 knockout background? More speculation or discussion on this point would be welcome. Is ligand induced transphosphorylation of CRK43 by the CRK7 homomers required for CRK43 activation?

-In panel C, there seems to be a higher baseline of PR1 expression in CRK7 complementation lines even in the control treatment, but there is no discussion or speculation of why this could be the case in the text.

-For the WTA-triggered cell death assays in figures S8 and S11, healthy/mock-treated leaves would have been appreciated. In figure S8, the absence of a mock control for the Col-3 gl1 mutant would have helped clarify whether the apparent change in leaf morphology is a symptom of WTA perception or a developmental phenotype.

-Panels D-F would benefit from minor aesthetic changes for improved presentation. We felt that the colors used for Control and flg22 treatments were very similar and slightly hard to distinguish. Moreover, the change of scale in the Y axis for panel F confused some readers.

-Figure S11: The color code of the taxonomic relationship of the infiltrated plant species is not explained in the figure legend. Potentially color coding the presence of CRK43/CRK7 and responsiveness to WTA at the bottom would facilitate interpretation of the figure.

-Figure S12: The lack of responsiveness of Ws-0 and Gd-1 is intriguing, particularly because they cluster together with responsive CRK7 homologs. Digging deeper into the underlying polymorphisms would be very interesting, via an alignment and/or via highlighting them on a predicted CRK7 structure, as was done for the pad4 loss of function mutant. Highlighting these could provide important clues about binding or specificity.

-Generally speaking, a more in depth structural analysis of where the polymorphisms identified in the crk7 loss-of-function mutant or in the natural variants that do not detect WTA are located would have strengthened the paper.

-In general, we feel like the genetic screen was a key strength of the paper which was not discussed in a lot of detail and was largely confined to supplementary figures. We would have liked to see more detailed description of this part.

Figure 4 and associated supp. (S13 to S15)

-In the Results, Co-IP is introduced before FRET-FLIM, but panel A shows FRET-FLIM first. Consider reordering text or panels for consistency.

-In Co-IP experiments: “Total extract” could be renamed “Input”, which is the more widely used term for this kind of experiments.

-A key weakness for both CoIPs shown in Figure 4 is the lack of negative controls, which would strengthen the conclusions. In panel D, the authors could use another unrelated RLK to address whether CRK7 is generally sticky, or perhaps one of the non WTA-responsive CRKs shown previously. For Panel E, perhaps including treatments with this same unrelated CRK or with an unrelated RLCK would have strengthened the conclusions. In general, including the proteins that are being tested for interaction in CoIP separately is not a strong negative control, as this does not address the question of whether the interaction is specific and not due to “sticky” non-specific associations. For panel E, the sudden switch from tagged CRK7 detected with anti-GFP to an untagged version detected with Anti-CRK7 antibody was abrupt, and it was not clear whether these experiments were also done in N. benthamiana with transiently expressed protein or whether this was done in Arabidopsis.

-The conclusion that CRK7 forms a homodimer should be stated more cautiously. The data could also be consistent with higher-order oligomerization. The authors switch to using the term homomer later on, which is a bit more inclusive. Have the authors tried AlphaFold modelling with higher order oligomers or with CRK-CRK heteromers? Do the authors have any explanation for the functional differences between CRK7 and closely related CRKs?

-Related to the point above, mutating the predicted dimerization/oligomerization interface would strongly strengthen this claim and test whether this interaction is required for function. How do you explain that the mutation in a single Cys in the crk7 mutant identified in the screen constitutes a loss-of-function mutant?

-Evidence for a ligand-induced conformational change is currently limited, and the justification feels somewhat speculative.

-Figure S13: The arrow (with the 180º label) does not clearly represent the direction of model rotation. Please specify which AlphaFold version was used in the figure legend or associated results section.Including pLDDT plots, pTM, and iPTM scores in the supplement would be very helpful. Related to this, it might be possible to model a defined WTA fragment (not a long polymer) in complex with the CRK7 ectodomain using a SMILES code. This model, if confident, could be combined with the authors’ already available data on polymorphisms that determine sensitivity to WTA to generate testable hypotheses with regards to the precise WTA-binding interface in CRK7.

-The N. benthamiana experiments are very elegant and convincing. Based on the fact that the system is movable to N. benthamiana, it would be extremely interesting to test: whether cell death is eds1/pad4/adr1-dependent in this system as well. This may give clues as to the potential mechanism via which the CRK7/CRK43 module is communicating with EDS1/PAD4/ADR1. Given that it is unlikely that a TNL is conserved between Arabidopsis and N. benthamiana to mediate this link (even less likely considering that the CRK7/CRK43 module appears to be very lineage specific), is it possible that CRK43 is communicating directly with EDS1/PAD4/ADR1? Also, since transient expression is working reliably in N. benthamiana it would be straightforward to test whether CRK7 or CRK43’s kinase activity is required for induction of cell death.

-Figure S14: The authors show that CRK43 and PAD4 are is required for WTA triggered cell death, but it would have been interesting to test if they are also required for the WTA-triggered priming effect and/or transcriptional reprogramming as well.

Discussion

The Discussion is relatively short and largely reiterates the findings. We would like to see more speculation on why this pathway appears lineage-specific to Brassicaceae. Exploration of why related CRKs cannot recognize WTAs and where specificity may lie/what the binding interface could be based on variation in CRK7/loss-of-function mutants that are insensitive to WTA. More explicit framing of this pathway as a non-canonical PTI / priming response and highlighting that this is the first ligand identified for a CRK-family receptor. Moreover, the proposed working model is compelling, but the graphical abstract might benefit from question marks to distinguish demonstrated steps from speculation, such as the arrow connecting CRK43 to the EDS1/PAD4/SAG101 hub or the depiction of CRK7 as a homodimer.

Final Remarks

Despite the detailed comments outlined above, we would like to emphasize that this is already a very strong and impressive study that represents a major conceptual and technical advance in the field of plant–microbe interactions. The identification of wall teichoic acids as MAMPs and the discovery of CRK7 as their cognate receptor open an exciting new avenue for understanding CRK-mediated immunity and non-canonical PTI signaling. The genetic screen and biochemical dissection are particularly elegant, and the overall story is compelling and timely. We hope that our comments are received in the constructive spirit in which they are intended and help further strengthen an already excellent manuscript. We greatly enjoyed reading and discussing this work and look forward to seeing its continued development and eventual publication.

Conflicto de intereses

Los/as autores/as declaran que no tienen intereses en conflicto.

Uso de Inteligencia Artificial (IA)

The authors declare that they did not use generative AI to come up with new ideas for their review.