PREreview of Graft incompatibility between pepper and tomato can be attributed to genetic incompatibility between diverged immune systems

Preprint review of “Graft incompatibility between pepper and tomato can be attributed to genetic incompatibility between diverged immune systems”

By Daniel Lüdke, Mauricio P. Contreras, Emma Raven, AmirAli Toghani and Andrés Posbeyikian

Summary

This article by Thomas et al. investigates the molecular basis of delayed graft incompatibility between tomato and pepper, which disrupts vascular connections and compromises plant health. The study uses a variety of approaches to study programmed cell death and the accumulation of non-viable tissue in hetero-grafts vs. self-grafts, including viability staining and transcriptomic analysis at multiple post-grafting intervals.

Their data revealed that incompatible grafts maintained non-viable tissue over time, unlike their self-grafted counterparts. RNA sequencing exposed a variety of differentially expressed genes which pointed to a sustained immune response in these incompatible hetero-grafts, particularly noting an upregulation of NLRs possibly involved in hypersensitive responses. Furthermore, a transcriptomic meta-analysis comparing the grafting response to other biotic stresses, like pathogen infections, illustrated a significant overlap with plant parasitism mechanisms. 

The study proposes a model of genetic incompatibility driven by cross-species NLR interactions, leading to defense responses and genotoxic damage—a novel finding in the field of grafting. This research significantly enriches our understanding of grafting incompatibility, which is a severely understudied topic, and offers insights into potential strategies to mitigate these challenges. 

General comments

• The manuscript is concise yet comprehensive, providing a robust framework for future work on overcoming graft incompatibility. All in all, we felt that it was well written and that the figures were very beautifully put together. We have included some suggestions below, in case they are of use to the authors. We believe they may improve clarity/readability and enhance some of the analyses performed. Nonetheless, the manuscript is already of high quality.

• We noted that the supplementary information was not uploaded to bioRxiv, which made interpretation of some of the results and methods complicated. While this information will surely be available in the peer-reviewed version, we would have appreciated if it could have been made available as part of the pre-print as well.

Introduction

• The introduction is generally well written and covers most of the background required to interpret the results. One minor point is that perhaps some sections could be broken into smaller paragraphs for ease of reading. For example, the paragraph that goes from lines 71-100 is a bit long, and could be split into two paragraphs.

• We felt that some of the reviews cited were a bit outdated. It would be good to update the references with more up-to-date reviews whenever possible. Particularly, the review on NLR biology cited in the section regarding NLRs is from 2014. Many important breakthroughs in the NLR field have happened since then, and a multitude of excellent reviews on this topic have been published in the last year.

• Related to this point, the introduction could use more background on NLRs and how programmed cell death in NLRs works, given that this is a major aspect discussed in this paper. There is a lot of literature available on this topic, which used to be more of a black box but is now much better understood. In particular with regards to NLR-mediated cell death in Solanaceae, many of the NLR genes that execute cell death are well known. EDS1, SAG101 and PAD4 are also mentioned as “regulators of HR in Arabidopsis” in the results, but this is not covered in the introduction.

• Hypersensitive response, discussed in line 85, should be introduced as a form of PCD, ideally in lines 66-67 where NLRs are mentioned as triggering PCD. 

Results

Figure 1.

• No comments, great figure! 

• One small comment is that perhaps they could indicate what is a statistically significant difference directly on the figure.

Figure 2.

• Nice side-by side, very clear visual!

• We found it very satisfying that we have exactly panels A-Z.

Figure 3.

• Great figure! However, it has a lot of individually labeled panels. Does every subpanel need its own label (subpanel a through ak)? The nomenclature can be complicated to follow in the text. Perhaps the subpanels could just be subdivided into A for a-r, B for s through aj and C for ak?

• No statistics were performed on data in subpanel ak.

Figure 4.

• An extra line briefly explaining the TUNEL method would be appreciated, just indicating that it measures fragmented DNA, which is a proxy for cell viability. 

• This is a great figure and a well executed experiment. We wondered, however, if given that the conclusions are negative this figure could potentially be supplementary.

Question to the authors, outside of the scope of the study: Instead of growing the calli and applying the exudates ectopically, would it be possible to apply plant exudates directly on graft junctions somehow? For example, by wrapping the graft with the exudate, like people do with hormones. Applying the DAMP exudates from incompatible hetero-grafts to self-grafts which are compatible would be interesting. This might be technically more challenging but it might be more biologically relevant than the callus-based assay.

Figure 5.

• This is a super interesting experiment. We noted that although there’s a lot of downregulated genes identified, the mansucript mostly discusses the upregulated ones. It’s a shame because several NLR or immune-related genes might be in the downregulated gene group. For example,there may be downregulation of immune associated genes to for a successful graft to happen, 

• Related to this, downregulated genes are not mentioned or analysed in any of the RNAseq analyses performed in the manuscript. Why is this? There could be some interesting information hiding in that set of downregulated genes.

Figure 6.

• A main point discussed when reviewing the paper was that given so many NLRs were identified, we would have liked a more thorough description of the NLRs involved. What are the NLRs found? In solanaceae one would focus on CC-NLRs and on the NRC immune receptor network which can make up to 50% of the NLRome of solanaceous species.

• Although SAG101, EDS1 are considered marker genes for immunity, they are not the actual executors of PCD for all pathways. Highlighting helper NLR families would be very interesting.

• What defines the order of the samples in the heatmap. Is it phylogenetic? Is it by gene ID? 

• Organizing the heatmap data by NLR family/category would potentially reveal interesting patterns. For example, organizing the NLRs phylogenetically according to their NB-ARC domain phylogeny/according to their N-terminal executor domain (CC-NLR, TIR-NLR, etc…) would be super useful. Perhaps all the upregulated NLRs belong to a specific phylogenetic clade, for example.

• Classifying NLRs into Sensor and Helper NLRs would be interesting too. Perhaps the cell death is triggered by an instance of sensor-helper incompatibility?

Out of scope for this paper, but perhaps of interest for future research. Many NLRs in tomato and pepper require downstream helper NLRs of the NRC family. There are tomato KO lines lacking several key cell death executing helper NLRs (nrc2/3 and nrc4 KO). It would be interesting to see if these KO lines exhibit altered grafting compatibility/incompatibility phenotypes!

Figure 7.

• Similar to our comment in Figure 5, only upregulated DEGs are looked into, but there wasn’t any analysis of downregulated genes..

Figure 8.

• Similar to our comment in Figure 5, only upregulated DEGs are looked into, but there isn’t any analysis of the downregulated genes.

• Overlap between 3 processes is shown, but there is no mention of their own dataset here.

Discussion

Discussion is well written and provides a nice summary of the findings as well as though-provoking speculation. Potential avenues of exploration are mentioned. We really enjoyed reading it. Our only minor comment is that, similar to our comment on the introduction, it would be nice if the text could be broken into smaller paragraphs (section in lines 596- 641 could be split into smaller paragraphs).

Competing interests

The authors declare that they have no competing interests.