PREreview of bioRxiv article “Structure-function analysis of ZAR1 immune receptor reveals key molecular interactions for activity”
This is a review of Baudin, Schreiber, Martin et al. bioRxiv doi:
https://doi.org/10.1101/592824 posted on March 29, 2019. The authors used structural modelling to identify elements required for self-association of the NLR immune receptor ZAR1, specifically its N-terminal CC-domain ZAR1CC. They discovered that the N-terminal α1 helix and EDVID motif in ZAR1CC are important for oligomerization and function of ZAR1. This complements recent findings by Wang et al. (2019) based on cryo-EM structures, highlighting the importance of the α1 helix for the activity of ZAR1 although some differences were noted that could reflect the different experimental set ups (CC domain vs full-length protein) as discussed in the paper.
Summary
In this paper, Baudin, Schreiber, Martin et al. used structural modelling to identify structural elements required for the self-association of the NLR immune receptor ZAR1, specifically its N-terminal CC-domain ZAR1CC. They used monomeric and dimeric forms of CC domain structure templates available at the time to make ZAR1CC models and tried to map the region involved / responsible for oligomerization. They discovered that the N-terminal α1 helix and EDVID motif in ZAR1CC are important for oligomerization and function of ZAR1. This complements recent findings by Wang et al. (2019) \cite{Wang2019,Wang2019a} based on cryo-EM structures, highlighting the importance of α1 helix for the activity of ZAR1 although some differences were noted that could reflect the different experimental set ups (CC domain vs full-length protein) as discussed in the paper. They also identified intramolecular interactions between ZAR1 subdomains, notably NB-ARC with CC domain, that keep ZAR1 in an inactive state. They also demonstrated that loss of function mutants in NB-ARC domain do not play a role in interaction with NB-ARC and suppression of ZAR1CC autoactivity. We believe that this is a nice study that identifies structural elements responsible for CC self-association and is overall consistent with recent structural findings. The set of interaction assays are important to complement the recent findings of ZAR1 cryo-EM structure in inactive and active states. However, in this study, the biological relevance of the NB-ARC/LRR domains in negative regulation of ZAR1 by binding to CC domain is not clear. Detailed comparison between binding affinity of NB-ARC/LRR to CC and the CC-mediated cell death is required for a firm conclusion that intramolecular interactions between ZAR1 subdomains participate in keeping ZAR1 immune complexes inactive.
Main comments
Results section - Model-based identification of ZAR1CC structural determinants for oligomerization They describe two homology models of ZAR1CC based on MLA10 and Rx/Sr33 structures to understand structural dynamics that lead to oligomerization. We feel that some statements in this section are not clear and need further explanation. For example1) Lines 161-163 They should show superimposition of the two models in Fig 1 highlighting α1 and α4 helices to support their statement. Also explain how t1 and t3 region are critical for oligomerisation2) Lines 169-171 They say that ZAR1CCH1 was designed to reduce the probability of favourable transient charge-charge interactions between the two monomers. It is not clear how this version can reduce interactions between two monomers. Most of the residues (78, 85, 89, 91) in Fig S1C seem exposed to the surface and not interacting to the other monomer.3) They should explain why they chose to mutate most of the residues to glutamine and some to alanine.
Results section - Impact of ZAR1CC architecture on its activity Lines 215-219. They report that ZAR1CC variants migrate to slightly different sizes (Fig S2) which could be due to post translational modifications (PTMs). Do they have any hints about potential PTMs? Interestingly mutants in full length ZAR1 don’t migrate at different sizes (Fig2 D). Does PTM occurs only when CC domain is expressed or exposed?In Figure 2C, they should show the leaf pictures, similar to Figure 2A.
Results section - Autoactivity of ZAR1CC-YFP is inhibited by the ZAR1NBARC domain In Figure 3B, they should use CC domain of At5g48620 as a negative control in the co-IP assay, to be consistent with the yeast two-hybrid assay (Figure 3A).In Figure 3B, it is better to show the loading control.In Figure 3C, they should show the leaf pictures, similar to Figure 2A.
Results section - Effect of loss- and gain-of-function mutations on the ZAR1NBARC domain They tried to make autoactive mutants of ZAR1 based on literature. AtZAR1G194E/K195A, AtZAR1D268E and AtZAR1D489V. However, these mutants did not cause HR. This might be explained by the possibility that
Nicotiana benthamiana endogenous RLCKs are not compatible with Arabidopsis ZAR1. Can they test cell death inducing activity by co-expressing AtZAR1G194E/K195A, AtZAR1D268E and AtZAR1D489V mutants with Arabidopsis ZED1 in
N. benthamiana?They could also test the autoactive mutants by agroinfiltration/trangsenics in Arabidopsis. See for example \cite{Wróblewski2018}.In Figure 4C, are there any differences between ZAR1CC-NBD mutants (K195N, V202M, S291N) and ZAR1CC-NBD, ZAR1CC-NBDHD1 mutant (P359L) and ZAR1CC-NBDHD1, ZAR1CC-NBARC mutant (L465F) and ZAR1CC-NBARC, respectively? There seems to be differences between Fig 4C with Fig 3C for the reported loss-of-function mutations.In Figure 4B, it is better to show the loading control.In Figure 4C, they should show the leaf pictures.
Results section - Interaction between the NBARC and LRR domains How does the association between NB-ARC and LRR subdomains contribute to the negative regulation of the CC domain function? Does P816Q mutation affect the activities of full-length ZAR1?In Figure 5, it is better to show the loading control.
Results section - Structure-informed analysis of ZAR1CC intramolecular associations They found that ZAR1CCH1, ZAR1CCH2a, ZAR1CCH2b and ZAR1CCED mutants do not bind ZAR1NBARC and ZAR1LRR. In Fig 1 and 2, the authors showed that the same mutants are reduced in homo-association and lost cell death activity when fused to the C-terminal YFP tag. Please discuss how the mutation sites contribute to the intra- and inter-molecular associations and cell death.
Editorial Comments
Line 139 “recently solved crystal structures of ZAR1” should be “recently solved cryo-EM structures of ZAR1”.Mis labelling in Supp Fig 1B of monomer (1CCα4, α2 is mis-labelled as α1).Typo error in Fig S4. ‘At3g4660’ should be ‘At3g46600’. Typo error in Fig 3 and Fig S5. ‘At3g4660’ should be ‘At3g46600’.
Reviewers
Abbas Maqbool, Hiroaki Adachi and Sophien Kamoun. The Sainsbury Laboratory, Norwich Research Park, University of East Anglia, Norwich, UK.
