Write a comment

This paper by Margets et al. focuses on the identification and functional characterization of a novel effector protease, Cysteine Protease 1 (CPR1; Hetgly22189), from the Soybean Cyst Nematode (SCN; Heterodera glycines), an obligate endoparasite responsible for significant yield losses in soybean crops. The study aims to deepen our understanding of SCN virulence mechanisms, which are crucial for developing durable resistance strategies in soybean varieties.

The authors employed a combination of gland transcriptome analysis, transient expression in Nicotiana benthamiana, and proximity-based labeling experiments in transgenic soybean roots to identify CPR1 and its interaction with a soybean protein target, branched-chain amino acid aminotransferase (GmBCAT1; Glyma.06G050100). The study reveals that CPR1 localizes to mitochondria and suppresses cell death triggered by the plant immune receptor RPS5. Furthermore, the degradation of GmBCAT1 by CPR1 in a protease-dependent manner underscores the role of CPR1 in promoting SCN virulence.

By silencing CPR1 in SCN, the authors demonstrated reduced nematode penetration, while expression of CPR1 in transgenic soybean roots increased susceptibility to SCN, confirming CPR1's role in virulence. 

The paper highlights the potential of using CPR1 as a candidate for developing an engineered decoy substrate, leveraging the innovative PBS1 decoy system previously developed by the Innes lab. This research paves the way for novel approaches to bioengineer SCN resistance in soybean, which are critical as traditional strategies lose efficacy due to evolving nematode populations.

General comments:

All in all, we thoroughly enjoyed reading the paper, and thought that it was very well written, with high-quality data. We were impressed by the varied methodologies employed in the study. Although the paper is already of very high quality, below we detail some suggestions which we hope are of use to the authors.

Title:

While the general feeling was that the title was good, there was some discussion about whether it would have been worth simplifying it, in an attempt to make it less “wordy” or technical.

For example: “A Soybean Cyst Nematode Effector Targets a Mitochondrial Soybean Branched-Chain Amino Acid Aminotransferase for Degradation to promote virulence”, avoids mentioning gene names while drawing the reader in with a clear message with regards to what the contents of the paper are. Nonetheless, this is ultimately a matter of personal preference.

Introduction:

Overall, we thought that the introduction was well written. A few things that we felt could have been briefly covered:

-Maybe mention a bit more about what is known on the genetic basis of SCN resistance, for example Peking-related resistance?

-What are J2 nematodes? Maybe a sentence on this? Although some people in the group already knew what this was about, some members on the group that were not familiar with nematodes were unsure about what J2 nematodes were and why they are used for several assays. Maybe a sentence on the nematode life-cycle, stating that J2 nematodes are the infective stage?

-Perhaps in the section detailing the RPS5 system, explaining what type of NLR it is? (CC, CC_G10, etc.) 

Results:

Overall the results were of very high quality and the narrative of the paper is clear. Below are a few comments and suggestions that arose while reading. 

Identification of CPR1 - Figure 1

In the first part of the paper, covering Figure 1 and the Identification of CPR1, more details regarding the rationale for focusing on CPR1 vs. all the other candidate effectors would have been nice.

-The authors write “To identify candidate SCN effector proteases, we searched SCN esophageal gland cell transcriptomes for transcripts annotated as proteases.” Could the authors clarify which transcriptomes were used and how this was analysed in M&M?

 -Also, potentially some of the details described in these results potentially could be moved to M&M, such as which signal peptide version was used, etc..

-The authors write “Of the proteases in the SCN gland cell library, CPR1 was the top priority candidate as it was annotated as a cysteine protease and highly conserved in the SCN genomes available (early release data available on SCNBase) (Masonbrink et al. 2019a).” Was CPR1 the only highly conserved protease? Was this the only rationale for following up on CPR1? Could the authors explain in more details how the degree of conservation was established? Potentially a phylogeny could help the reader understand just how conserved CPR1 is across SCN genomes compared to other proteases. Maybe this could be added to Figure 1. Potentially a phylogeny of all of the SCN proteases would be nice to see, with information about their presence or absence in different SCN genomes. One of the key rationales for working on this protease is that it is highly conserved but no data is presented related to this claim.

-Minor point in Figure 1B: could the authors clarify in figure 1 legend what were these two treatments normalised to?

CPR1 suppresses RPS5-mediated cell death in N. benthamiana leaves - Figure 2

-Perhaps beyond the scope of this particular study but have the authors tested whether CPR1 affects cell death mediated by other NLRs other than RPS5? We feel like it is a bit premature to say that CPR1 suppresses ETI. Maybe to be more specific it would be better to say CPR1 suppresses RPS5-mediated cell death as was stated in the heading of this section?

-Minor point: Would have been nice to see UV light imaging for Figure 2B, as was shown for 2A

-A question that arose, how was the accumulation of the RPS5 autoactive mutant assessed given that this triggers very strong cell death at 24 hours? Was this harvested at very early timepoints? Was this done with dexamethasone induction? Please clarify the methodology more precisely in the results or in the figure legend. It would be very useful to us to know what approach was used. Also is there any particular reason WT RPS5 was not used to assess the effects of CPR1 on protein accumulation?

-In general, we feel like it is better to replace the acronym HR with cell death to make plant immunity papers more accessible to people from outside the plant immunity field.

CPR1 localizes to the cell periphery in puncta-like structures in N. benthamiana leaves - Figure 3

-We were curious about the distribution of the MitoTracker Red marker. We would have liked to see a Z-stack maximum projection to make it clearer that the mitotracker red is showing the correct distribution, with puncta both in the periphery but also inside the cell. The single plane images are harder to interpret, in this regard. Because the mitochondrial localisation of CPR1 is based on co-localization with this marker, it would be nice to have more images to clearly see where the mitochondrial marker is localised.

Identification of candidate substrates of CPR1 using biotin-mediated proximity labeling - Figure 4

We found the proximity labelling experiments in soybean hairy roots very elegant and impressive! However, we felt like the rationale for selecting BCAT1 out of all the other hits from the proximity labelling experiment was not super clear in the text. This is perhaps similar to what was mentioned in the comments related to the Identification of CPR1 section,

-“Notably, several of the selected proteins have links to SCN infection and/or plant immunity from prior studies.“ Would be nice to add references accompanying Table S1 for proteins with previously described involvement in SCN infection and plant immunity.

-“Of the proteins identified in the miniTurbo datasets, we prioritized soybean branched-chain amino acid aminotransferase 1 (GmBCAT1) (Glyma.06G050100), as it was absent in the YFP:miniTurbo negative control and all other SCN effector datasets.“ Out of the 12 proteins in Table S1, was this BCAT1 the only protein that filled these conditions? 

-What are the “SCN effector datasets” that are mentioned? Is this something that could be referenced more clearly for the readers to access?

-In Table S1: Gene names for soybean are conventionally written in the format Glyma.06G050100, as was done throughout the paper. Maybe this can be fixed in the table? Also, what database do the annotations come from? How were the genes/proteins annotated?

CPR1 interacts with GmBCAT1 - Figure 4 and Figure 5

-One minor comment: Maybe rephrase the sentence discussing trans-membrane proteins “complicating” the CoIP assay? Although a few of us had done CoIP assays and understood what was meant by this, perhaps it would be good to specify what sort of complications may arise from having a trans-membrane protein in a CoIP assay, for the readers that are not as familiar with this methodology. Are they worried about lack of specificity in the IPs due to membrane association?

-Maybe add Ref for the AvrPphB control used in Figure 4 in the results as well.

-Although it might not be as relevant in this scenario, an uninduced mock control would have potentially been nice to assess whether the system is leaky in the absence of dex induction.

-In Figure 5B, a phylogenetic analysis is used to determine relationships between the soybean BCATs and the previously characterized Arabidopsis ones. We had a number of comments related to this phylogeny. In general, the analysis does not feel like it should be a main figure, as it does not add much to the narrative of the paper. We feel like additional, more comprehensive, phylogenetic analyses are required to define orthology to AtBCAT1 more confidently.

1) The tree needs a proper outgroup.

2) The bootstrap value of the node that links GmBCAT1 to AtBCAT1 is 45, not particularly high. Any chance additional BCAT sequences from more plant species could be added to make this analysis more confident? 

In the discussion, the authors speculate on this by saying “Our phylogenetic analysis revealed GmBCAT1 to be orthologous to Arabidopsis BCAT1 (Fig. 5B), which localizes to mitochondria and contributes to BCAA degradation (Schuster and Binder 2005; Binder 2010).” As mentioned above, we feel like the phylogenetic analyses presented are not enough to determine orthology.

CPR1 contributes to SCN virulence - Figure 6 and Figure 7

We really enjoyed this part. We appreciate the complexity of carrying out virulence assays with nematodes! 

-We had one potentially naive question regarding the different methodologies used in Figs. 6 and 7.

Why were 120 nematodes used for the RNAi assays, whereas 2000 eggs were used for the CPR1 over expression assays? We were wondering if this could be explained further in the text.

-We found it particularly interesting that the catalytic dead CPR1 over expression leads to enhanced susceptibility. This was not covered in the discussion. Could the authors speculate why this could be the case?

Additional comments/questions

These are thoughts or suggestions that arose while reading the paper, that are beyond the scope of this study:

-Experiments connecting BCAT1 to NLR-mediated signaling would have been really nice! Does silencing or overexpression of BCAT1 modulate RPS5/other NLR-mediated cell death or other ETI related readouts?

-In this same train of thought, does BCAT over-expression or silencing affect susceptibility to SCN? 

-An AlphaFold prediction of CPR1 with BCAT1 would be nice. Even AlphaFold models of CPR1 or BCAT1.

These experiments are not required but it would be nice to somehow connect the NLR suppression to the CPR1-BCAT interactions. 

Discussion

Overall the discussion was well written. Our main comment on this section of the paper is that at times it reads a bit like a repetition of the results, without too much speculation. While of course we do not expect any of this speculation to be addressed experimentally in the paper, which is already a very complete and well rounded story, it would have been nice to hear the authors’ views on some open questions that arose from this study. Below are some interesting points where we would have appreciated further comments from the authors in the discussion, for example:

-How generalised is the suppression of NLR-mediated cell death? Is general ETI suppression or just restricted to certain classes of NLRs? Can it suppress CC, CC_G10 and CC_R-triggered cell death indistinctly? Perhaps not relevant for RPS5 engineering but more information on this front could inform deployment of other NLR-based resistance genes in soybean, which could potentially be suppressed by CPR1 as well.

-How do CPR1-BCAT1 interactions link to suppression of cell death? 

-Is BCAT1 required for NLR-mediated cell death?

-What cleavage sequences are being recognised by CPR1? Surely this would be important towards engineering PBS1 cleavage by the SCN effector.

-Why does over expression of the catalytic dead CPR1 still enhance susceptibility? 

-While the authors discuss the feasibility of engineering RPS5-based recognition of CPR1 given that CPR1 suppresses RPS5 signaling, more discussion on this front or alternatives could be nice. Related to this, testing other NLRs could be interesting to see how broad the suppression of cell death is. This would clarify other potential NLR scaffolds that could be used for engineering.

-The paper is initially framed with RPS5 engineering as the main focus. RPS5 engineering is presented as critical for combatting SCN infection in soybean due to the lack of reliable genetic resistance against this pathogen. Maybe coming back to this point a bit more in the discussion would be good.

All in all, we would appreciate more speculation, proposal of models and hypotheses and future work in the discussion. A final figure with a schematic of the proposed working model would be nice as well, although of course not strictly required.

Our conclusion

We congratulate the authors for this excellent paper. Overall, we greatly enjoyed reading it and were impressed by all the different methodologies used. SCN is an extremely important and understudied pathogen, and we commend the authors for tackling this difficult pathosystem. We are sure that the findings from this work will be useful to the field of SCN-soybean interactions and disease resistance engineering. Moreover, further studies of BCAT1 may uncover novel regulators of NLR-mediated signaling/ETI.

Competing interests

The authors declare that they have no competing interests.