Review of: Cross-reactivity of a rice NLR immune receptor to distinct effectors from the blast pathogen leads to partial disease resistanceAuthors: Freya A. Varden, Hiromasa Saitoh, Kae Yoshino, Marina Franceschetti, Sophien Kamoun, Ryohei Terauchi, Mark J. BanfieldDOI: http://dx.doi.org/10.1101/530675This manuscript examines the molecular mechanism of the recognition of the “mismatched” Magnaporthe oryzae (Mo) effector AvrPia by the rice CC-type NLR Pikp. The authors show that transient co-expression of Pikp and AvrPia induces a weak cell death response in N. benthamiana (Nb). In line with these results, a Pikp-containing rice cultivar induces a weak immune response, whereas the allelic Pikm cannot. Next, they show that the Pikp HMA domain (Pikp-HMA) directly interacts with AvrPia in vitro, while Pikm cannot. X-ray crystallography showed that Pikp-HMA indeed directly binds AvrPia in a manner reminiscent of the interaction between RGA5-HMA and AvrCO39. This mode of interaction is in contrast to the previously published AvrPikD Pikp-HMA structure. The authors propose that the N-terminal extension of AvrPikD is responsible for the configuration of the AvrPikD Pikp-HMA heterodimer and could enhance the AvrPia Pikp-HMA interaction. However, addition of this N-terminal extension to AvrPia did not enhance, but instead reduced, AvrPia-Pikp cell death output in Nb. Furthermore, this modified AvrPia version was also not recognised by Pikm, which, like Pikp-AvrPikD, depends on the AvrPikD N-terminal extension.The finding that an integrated domain of an NLR is able to recognise two effectors using two distinct interfaces is a novel finding and will be of interest for the wider NLR biology community. Furthermore, these findings have potential implications for the plant breeding industry. However, we suggest that addressing the issues that we raise below will significantly strengthen the authors’ claims and the impact of their study.Major comments1. The authors use a cell death index to quantify the severity of the cell death response in Nb (Fig. 1 and 5). If the authors wish to show these data we recommend to use stacked bar plots instead of box plots. The use of box plots implies the data has a Gaussian distribution, which clearly is not the case for many, if not all, of the genotypes tested. Moreover, as the index only consists of a scale of 6 possible values, a Gaussian distribution cannot really be made. Furthermore, it is unclear whether the differences depicted are statistically significant. We would recommend to include this, and clearly state the statistical test used in the figure legend. Most importantly, we would advise to complement these data with ion leakage measurements that are known for a quantifiable dose-response relationship. Such physiological experiments will provide a more quantitative readout and strengthen the claim that Pikp induces an intermediate cell death response upon AvrPia perception. This is of particular significance as “partial disease resistance” is one of the important findings of this paper as stressed in the title.   2. The authors show that the rice cultivar K60 containing Pikp induces a partial immune response upon infection with Mo Sasa2 (AvrPia). This claim can be made based on the pictures shown in figure 2. However, a quantification of the lesion size (area and/or lesion length) would be desirable. Furthermore, it would strengthen the authors claim if they would back up these observations with some physiological experiments, e.g. 1) measuring fungal growth by qPCR on fungal genomic DNA, 2) Ion leakage, 3) staining rice leaves for ROS with DAB or H2DCFDA, 4) measuring transcriptional outputs by qPCR on relevant marker genes, such as RbohD. Particularly comparing transcriptional outputs and ion leakage over time in plants infected with Mo AvrPikD and Mo AvrPia would be desirable.  Not only will these data strengthen the author’s conclusions, they would yield insight into the resistance pathway dynamics between Mo AvrPikD- and Mo AvrPia-treated plants. 3. In figures 3 and 4 the authors clearly show that Pikp-HMA is able to directly interact with AvrPia by gel filtration, SPR and x-ray crystallography. Although these data are convincing, one aspect is still missing. Since the interaction experiments are based on in vitro systems, it would strengthen the authors’ claims if they would include in vivo interaction data. Yeast-2-Hybrid or in planta split-luciferase assays are two options the authors could explore. 4. The interaction data between Pikp and AvrPia are based on the interaction of the Pikp-HMA and the effector AvrPia. It would strengthen the authors’ claims if they would confirm their interaction results by performing interaction assays using the full length NLR instead of merely the HMA domain. Yeast-2-Hybrid and split-luciferase would be suitable techniques. 5. Given that the presented AvrPia Pikp-HMA structure as well as the previously reported AvrPikD Pikp-HMA structure are based on HMA domains lacking the adjacent NLR modules, how do the authors know that the recognition of two distinct HMA interfaces occurs and is physiologically relevant in the corresponding full-length NLRs? The authors do not show a variant of Pikp (-HMA) that is unable to interact with AvrPia, even though the authors discuss residues in Pikp-HMA that seem to be responsible for the interaction, e.g. PikpS202 PikpD217, and PikpDS204 (Fig. 4C). The authors could try to evaluate the effect of mutating these three sites and determine whether this leads to a reduced interaction between Pikp (-HMA) and AvrPia. Moreover, if this does reduce binding, the authors could test whether these mutations also affect Pikp outputs in planta, e.g. by co-expressing the Pikp variant with AvrPia in Nb and quantifying signalling outputs by ion leakage.Minor comments 1. Although the manuscript is well written, clear and succinct, in some instances the authors refer to a previous finding, but do not supply the according reference, e.g. Introduction; paragraph 5: “The AvrPikC... Pik NLR.”, and Introduction; paragraph 7: “Despite... on its own.”. Therefore, I would recommend the authors to carefully go through the manuscript and make sure all published data are appropriately referenced.Joram A. Dongus and the Plant Microbe Interactions Journal Club at MPIPZ