The Pik NLR pair accumulates at the plasma membrane as a hetero-oligomeric sensor-helper immune protein complex prior to activation

  1. Hsuan Pai
  2. Mauricio P. Contreras
  3. Jose Salguero Linares
  4. Daniel Lüdke
  5. Andres Posbeyikian
  6. Jiorgos Kourelis
  7. Sophien Kamoun
  8. Clemence Marchal

Reviewed by

Read the full article See related articles

Listed in

Log in to save this article

Abstract

Following the perception of pathogen virulence proteins in plants, nucleotide-binding and leucine-rich repeat immune receptors (NLRs) are activated via a wide range of mechanisms. Singleton NLRs can both perceive effectors and trigger an immune response, whereas other NLRs specialise in either pathogen recognition (sensor NLRs) or activation of the immune response (helper NLRs). Sensor and helper NLRs can function as genetically linked pairs or in unlinked receptor networks. Although growing evidence suggests that NLRs conditionally oligomerise upon activation, our understanding of the resting state of NLRs prior to effector perception remains limited. Here, we investigated the oligomeric state of the genetically linked rice ( Oryza sativa ) sensor Pik-1 and helper Pik-2 NLR pair prior to effector activation when transiently expressed in Nicotiana benthamiana leaves. We show that both wild-type Pikm-1 and engineered Pikm-1 Enhancer sensors associate with Pikm-2 and form ∼1 MDa hetero-complexes in the resting state that accumulate at the plasma membrane. Our findings contribute to the growing evidence that pre-activation mechanisms vary widely across NLRs. This knowledge could be leveraged for disease resistance engineering strategies complementary to approaches focussing solely on effector binding.

One sentence summary

The Pik sensor-helper pair of rice immune receptor proteins forms a hetero-oligomer that accumulates at the plasma membrane in its resting state.

Synopsis

Nucleotide-binding and leucine-rich repeat receptors (NLRs) are key players in plant immune systems that recognise and respond to harmful pathogen effector proteins. NLRs can play specialised roles in either detecting effectors (sensor NLRs) or triggering the immune response (helper NLRs) and might function together as pairs or within larger networks. Although many NLRs are known to oligomerise upon the perception of effectors, much less is understood about their resting state prior to pathogen recognition. In this study, we investigated the resting state of a pair of rice NLRs, Pik-1 and Pik-2, transiently expressed in Nicotiana benthamiana leaves. These two proteins formed a sensor/helper complex at the cell membrane even in the absence of the effector protein. This discovery is significant because it reveals a new aspect of how these immune proteins are organised before activation, contributing to our understanding of the diverse ways plants prepare for immune responses. This finding adds to the growing body of evidence that, although many NLR proteins form complexes to respond to pathogens, their resting states and pre-activation mechanisms can vary widely.

Article activity feed

  1. Review Commons

    Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

    Learn more at Review Commons

    Reply to the reviewers

    We thank the reviewers for their time and their helpful comments and suggestions. In this updated version of the manuscript, we addressed the main concerns regarding autoactivity, carried out ion leakage assays of the tested NLR combinations and added the recommended negative controls. Additionally, we performed new confocal imaging experiments with a plasma membrane marker instead of FDA and carried out additional membrane fractionation replicates to strengthen our finding that the Pik complex accumulates at the plasma membrane. We replaced the previous data that included FDA and the autoactive constructs C-terminally tagged with mNeonGreen with this new experiment (updated Figure 3 and Figure S4). We replaced the previous membrane fractionation replicates with 2 new replicates showing that the H+ATPase does not show differential enrichment between samples. On these new replicates, we observed that Pikm-2 only accumulates in the membrane fraction, similar to that of the H+ ATPase control. These new data are now presented in Figure 3 and Figure S3.

    Please find below a detailed response to each comment.

    Reviewer #1 (Evidence, reproducibility and clarity (Required)):

    In this manuscript, Pai et al show that the NLR pair Pikm-1/Pikm-2 is putatively present as an hetero-oligomer at the plasma membrane in the inactive state. This report could show that the mechanisms of NLR inhibition are diverse and could potentially inform strategies to engineer synthetic NLRs. As such, it could be an interesting study. However, there are major issues in this study.

    Major comments:

    One major comment comes from the use of modified Pik receptors. In this study, the fact that GFP is seen in the Pikm-1 enhancer / Pikm-2 active complex in fig 4 may not represent the normal active state of the NLR pair. It is interesting that such a complex can exist, but the author should clarify figure 5 and precise that the complex represented is not an NLR-effector complex but an artificial NLR-GFP complex. It is claimed that the NLR pair oligomerizes at the plasma membrane in the inactive state. It cannot be ruled out that the pair is partially active when overexpressed. Activity of NLRs does not always result in cell death. In figure 2C, a mild chlorosis phenotype is clearly seen in the mcherry co-expression side. This experiment needs more controls (empty vector, use EV + Pikm-1, EV + Pikm-2) and some other proxy for NLR activation should be used to claim that the complex is inactive (gene expression or else as appropriate).

    Answer: We carried out ion leakage assays in all previously tested combinations and added the suggested combinations of co-expressing Pikm-1 variants and Pikm-2 with Rx as negative controls. Rx provides a superior baseline for non-specific NLR overexpression effects compared to an empty vector control. The updated hypersensitive response assays can be found on Figure 1B, 2C and 4A with the corresponding ion leakage measurements at 24 hpi on Figure S1B and Figure S2.

    This experiment demonstrates that only the functional combinations (Pikm-1 + Pikm-2 + AVR-PikD and Pikm-1Enhancer+ Pikm-2 + GFP) lead to significant increase of ion leakage compared to all other tested combinations, including the Rx controls.

    While we observe mild chlorosis in some overexpression contexts (such as Figure 2C), our quantitative ion leakage data confirm that this does not amount to full cell death activation. We acknowledge the reviewer's suggestion to use gene expression proxies, however, this would require setting up an RNAseq experiment across multiple combinations (Pikm-1 + Rx, Pikm-2 + Rx, Pikm-1 + Pikm-2) that is beyond the scope of this project. Additionally, it has been shown in multiple studies that cell death serves as reliable readouts for Pik pair activation by effectors (https://doi.org/10.1038/s41477-018-0194-x; https://doi.org/10.1371/journal.pone.0238616).

    I am not convinced by the confocal microscopy data presented in figure 3 and by the Co-IP data in figure 1.

    Figure 3: The confocal microscopy presented in figure 3 needs some refining. This experiment may not be suitable to estimate the change in receptor localization as the difference is quantitative. We can always see some signal in the cytoplasm and putatively in the plasma membrane and it is difficult to evaluate what is signal and what is noise. A plasma membrane marker should be used in this experiment to conclude on the localization of the receptors.

    Upper panels: It is claimed that Pikm-1 fully co-localized with fluoresceine when co-expressed with Rx. On the figure, I can see purple signal and blue signal that do not co-localize. Co-localization of Pikm-1 when co-expressed with Rx or Pikm-2 is not clearly different.

    Most importantly, the images were not taken using the same magnification which prevents a clear comparison. Define FDA in the caption.

    Answer: the aim of this experiment was not to measure the change in receptor localization, as this is not the correct setup for such an experiment. Indeed, as the reviewer pointed out, we expect to see signal in both cytosol and plasma membrane in all tested combinations based on the membrane fractionation experiment.

    The aim was to identify in which membrane compartment Pik1 and Pik2 accumulate based on the membrane fractionation assays. The confocal imaging experiment suggests that Pikm-1 and Pikm-2 accumulate is the plasma membrane since they do not localize to any other membrane compartment. We updated the text and changed the order of Figure 2A,B and 2 C,D to emphasize this. Additionally, we conducted a new confocal experiment with N-terminally tagged LTI6b from Arabidopsis as plasma membrane marker.

    Figure 1: Co-IP experiment shows many bands that may or may not be considered as background noise.

    In the first lane, Pikm1-v5 is expressed with Rx-HA. On the v5 blot of the input, we can see a band at approximately 130kDa that is supposedly Pikm1-v5. On the HA blot, we can also see a band at 130kDa that is supposed to be Rx-Ha. However, on the third lane of the v5 input blot, Rx-v5 is much lower. Thus, it is not clear what this 130kDa band is. Size of Pik1 and 2 are said to be around 110kDa. In the IPv5 blot v5 part, lane 1 where Pikm-1-v5 is co-expressed with Rx-HA, we cannot see a clear Pikm-1 band at around 130kDa. Both Pikm1 and 2 bands are supposed to be around 130kDa on this gel, as indicated by the red arrows, but this size is very close to this non-specific band. Overall, there are too many non-specific bands to draw conclusions from this experiment.

    Answer: We re-examined the blots and realigned the IP blots as we noticed a slight difference in migration between the input and IP gels.

    Rx-2xv5 is 4.5 kDa smaller than Rx-6xHA (see below) so a slight difference in band migration is expected. Additionally, the Pikm-1 and Pikm-1Enhancer bands on the IP blot are visible at the top (close to the 130 kDa mark), although the signal was too strong to see a complete band. We added an arrow to clarify this. The space between the lower non-specific bands and the Pik band is wide enough to differentiate these.

    We listed below the weight of each tagged construct for reference:

    Pikm-2-6xHA: 122.04 kDa

    Pikm-1-v5: 129.88 kDa

    Pikm-1Enhancer-v5: 135.28 kDa

    Rx-2xv5 110.51 kDa

    Rx-6xHA 115.01 kDa

    Minor comments:

    I would like to ask the author to be careful about the use of jargon and to always define terms at first use. In the abstract, Pikm is not defined; in the introduction, ID is not defined, ZAR1, Sr35, WAI3 etc... The term Pikm1 "enhancer" is confusing, it seems that this protein is just a hyperactive version of Pikm-1. Please try to be more accessible to students and non-specialists, this field is becoming increasingly opaque for non-specialists.

    Answer: we made the suggested adjustments. Enhancer is the published name of the nanobody that was integrated into the Pik chassis, so we clarified this in the introduction.

    Figure 3:

    A: Labels are not aligned with the lanes.

    Page 7 "To determine whether treatment with the ligand would alter the Pik hetero-complex" this statement is confusing, this is not what you test.

    Answer: we made the suggested adjustments.

    Reviewer #1 (Significance (Required)):

    This study is potentially interesting but the novelty is limited. The interaction of these NLRs was shown before. Here the authors do not show the nature of the complex, its stoichiometry. The presence of effector in the complex is shown using an artificial NLR and GFP that may not represent the normal complex. If the NLRs were forming oligomers and were present at the plasma membrane (and this is still not clearly demonstrated), the question is: How is the complex inhibited? If the authors were to describe this putative inhibition mechanism, this would be highly interesting.

    Answer: while interaction between Pik-1 and Pik-2 was indeed demonstrated, their oligomeric state in the resting state has not been described. Identifying the stoichiometry of the complex and how it is maintained in an inactive state would require extensive structural biology work that proved difficult to conduct and is beyond the scope of this study.

    Reviewer #2 (Evidence, reproducibility and clarity (Required)):

    Summary This manuscript investigates the resting-state organisation of the rice NLR pair Pikm-1/Pikm-2 and reports that these receptors form a ~1 MDa hetero-oligomeric complex that accumulates at the plasma membrane prior to activation. Using a combination of experiments (coIP, Blue Native-PAGE, subcellular fractionation and confocal microscopy) the authors provide evidence that both native and engineered Pikm-1 associate with Pikm-2 in a pre-activation.

    The study addresses an important and timely question in plant immunity on how paired sensor-helper NLRs are organised before activation and offers some conceptual advances relevant to NLR mechanistic. The work is technically ambitious and generally well executed (proper controls, loading samples). However, some data presentation, interpretation and experiments, especially regarding membrane localisation, and proposed functional implications would benefit with further clarification.

    Major comments

    • Clarity and focus of the Introduction: Some paragraphs in the Introduction are very dense and read more like a review article than a research manuscript. While the background is comprehensive, streamlining several sections would improve focus and help guide the reader more directly toward the specific questions addressed in this study.

    We shortened (-300 words) and restructured the introduction for better clarity.

    • Interpretation of intermediate oligomeric status: Intermediate oligomeric bands are observed along the manuscript when Pikm-1 or Pikm-1Enhancer is expressed without Pikm-2. While mentioned, their possible significance is not fully discussed. Would this intermediates be functionally relevant, assembly precursors or artefacts of overexpression? A clearer discussion would help contextualise the ~1 MDa complex.In same lines, the statement "Overall, the biochemical nature of these puncta remains elusive, as they might consist of resistosome foci or clusters of inactive NLR complexes." would benefit from further clarification.

    We added a couple of sentences to the “The Pik pair forms a hetero-oligomer in its resting state” discussion section to elaborate on the intermediate complexes. Regarding the observation of punctate, we do not have the proper system to study their dynamics in the Pik pair since the Pikm-2 constructs C-terminally tagged with fluorescent proteins are autoactive.

    • The organisation of Figure 1C is somewhat confusing. I suggest rearranging the blot so that the Ponceau S staining is placed directly under the corresponding input lanes, as both should originate from the same membrane. The immunoprecipitated (IP v5) samples could then be shown below.

    The suggested adjustments were made.

    Given the central role of membrane localisation in the proposed model, the membrane fractionation data raise concerns. In Figure 3A, the decrease in soluble Pikm-1 upon co-expression with Pikm-2 is not clearly accompanied by a corresponding increase in the membrane fraction, whereas Figure S3 suggests a more pronounced enrichment. Similar inconsistencies appear for Pikm-2. In addition, the plasma membrane ATPase control shows variable enrichment between Figure 3A and Figure S3. Because these experiments underpin key conclusions about pre-assembly and membrane recruitment, I strongly suggest repeating and consolidating these fractionation experiments with clearer quantification.

    We performed two additional replicates of the membrane fractionation experiment that shows consistent accumulation of the ATPase control in the total and membrane fractions. This now allows us to conclude that 1) Pik-2 accumulates in the membrane fraction and 2) Pik-1 is always present in both soluble and membrane fractions, but shows higher accumulation at the membrane in presence of Pik-2 compared to Rx. We added these new replicates to the main Figure 3D and Figure S3 and updated the text accordingly.

    • The reference to Fig. 4"We detected a GFP signal (using anti-FLAG antibodies) at ~1 MDa only when GFP was co-expressed with Pikm-1Enhancer and Pikm-2EE. " is misleading as GFP is also detected in Pikm1 homo-oligomers at 480 kDa in Fig.4B anti FLAG lane 1. Please clarify, explain or acknowledge.

    We clarified the statement to illustrate that a signal was observed at 1 MDa in the FLAG blot only where GFP was co-expressed in the Pik pair and not mCherry.

    The confocal imaging data are not fully convincing as presented. The use of FDA as a cytosolic marker is acceptable, but the fluorescence intensity profiles differ substantially between samples (peaks vs slopes in fg3.B), and scale bars does not appear consistent, text cannot be read. Not all the presented images show the epidermal cell periphery. To strengthen the conclusions, the authors could consider use a plasma membrane marker (FM4-64 or PI) and/or plasmolysis assays to more definitively demonstrate plasma membrane localisation. Intensity plots should span comparable distances across conditions.

    We attempted to image with a FM dye compatible with mCherry but the signal was always leaking into the cytosol, probably due to the fact that most constructs show auto activity. Alternatively, we identified that mTFP1 is not recognized by the nanobody Enhancer and thus Pikm-1Enhancer does not induce strong HR in presence of the N-terminally tagged mTFP1-AtLTI6b plasma membrane marker. We have thus repeated the confocal imaging experiments with this construct (Figure 3A).

    • Several figures contain very small text, making them difficult to read (e.g. Fig. 1A, 1C, plots in Fig. 2B, and Fig. S4). Panel sizes in Fig. S4 are particularly small and should be enlarged.

    We made the suggested adjustments.

    • The authors state the use of Pikm2 EE to abolish downstream cell death but conclusive results from this experiments did not show any major difference in NLR pair association Pikm1Enh and Pikm2EE biochemichaly. A more explicit discussion on alternative activation models (e.g. conformational changes within a pre-assembled complex, changes in membrane topology or altered interactions with additional components) and how to address this issue would strengthen discussion section.

    At this stage, we cannot rule out that Pikm2EE is a nonfunctional version of Pikm2 that cannot be activated. Hence we cannot conclude whether the Pikm-1/ Pikm2EE/GFP complex is truly activated and this is why we did not expand this discussion point. We edited this part of the discussion to better reflect this.

    • OPTIONAL experimental suggestion: it would be informative to examine the dynamics of puncta formation using fluorescently tagged, non-autoactive versions of the NLRs during activation by GFP-HF, rather than relying on autoactive constructs.

    These experiments could help clarify whether puncta formation reflects NLR activation or is instead a secondary outcome of autoactivity.

    We agree that these experiments would help better understand the dynamics of puncta formation. Unfortunately, we haven’t yet identified a suitable Pik-1+ Pik-2 combination that is not autoactive when fused to fluorescent proteins. We are currently exploring indirect imaging techniques to circumvent this issue. However, the system is not yet ready for this study.

    • OPTIONAL: For image clarity I suggest to add Size (kDa) to all WB and BN-PAGE gels.

    We made the suggested adjustments.

    Minor comments (including typos and presentation issues)

    • Figure 3A: lane labels at the top of the blot appear misaligned and should be repositioned. Ponceau blot seems not to be "boxed" as in the other images.
    • Figure 3B: scale bars are illegible; remove them from the images and report magnification or scale clearly in the legend.
    • Labelling in different panels along S4 is hard to understand on Max projection or confocal plane. Please rewrite. Details on laser excitation and acquisition in the legend should be moved to M&Ms, currently missing. Also the number of cells with puncta now indicated as a fraction is not very indicative, I suggest making an additional graph with these quantifications.
    • Figure 5. Please label the plasma membrane.
    • In confocal microscopy (Methods), please indicate the acquisition wavelength ranges used for each fluorophore. Correct the reference from Figure S5 to Figure S4. Language and typographical issues

    We made the suggested adjustments.

    -Legend Fig1. "(kDa) are shown at left" to "on the left".

    -Correct "focussing" to "focusing"

    -"0.45 μM filter" to "0.45 μm filter"

    -"leaf disk were sampled" to "leaf disks were sampled".

    -Improve clarity in "High-order oligomeric resting states were also observed..." → "...have also been observed."

    -Please cross-check Refs 33 and 72.

    We made the suggested adjustments.

    **Referees cross-commenting**

    After reading both reports, I am quite pleased to noticed quite a bit of overlap between Reviewer 1's points and my own. Here is a summary of the main points I would highlight: -NLR constructs: 1) Both reviewers flagged that the GFP-tagged or "enhancer" Pikm-1 constructs might not fully reflect the natural NLR complex. 2) The claim that the hetero-oligomer at the plasma membrane is inactive is not completely supported-yet I acknowledge it is a complex task. Some additional controls or discussion are needed to make this clear (e.g., empty vector, single NLR expression or other activity readouts).

    Thank you for the summary. As mentioned above, we have now added controls to the HR assays and carried out ion leakage assays to show that Pik-1 and Pik-2 co-expressed with Rx or without their cognate effector/ligand are not autoactive. We also added a new confocal imaging experiment with a plasma membrane marker and performed additional replicates of the membrane fraction assays. We hope this additional evidence strengthens the claim that the Pik pair is not autoactive in N. benthamiana and that the Pik complex accumulates at the plasma membrane.

    -Evidence for inactivity1) Mild chlorosis or other subtle phenotypes suggest that overexpression might induce some activity. 2) I agree with Reviewer 1 that more functional evidence would strengthen the claim that the pre-assembled complex is truly inactive.

    See comments above.

    -Membrane localization: 1) The confocal images are hard to interpret quantitatively. Cytosolic signal, inconsistent intensity profiles and different magnifications make it difficult to conclude membrane localization. 2) Using a plasma membrane marker (like FM4-64) or alternatively plasmolysis assays would help, as would standardizing imaging parameters across samples.

    See comments above.

    • Co-IP and BN-PAGE data: 1) Background bands and ambiguous molecular weights make it tricky to confidently interpret oligomeric states-I would understand in some cases finding extra bands for using N.benthamiana transient expression, not so importnant SDS-PAGE in the CoIPs but given the poposed hypothesis oligomeric state for NLRs this should be better addressed, indicated or discussed. 2) Improvement on molecular weight markers, clearer labeling

    Labelling and alignment with molecular weight markers have been refined. Additionally, the ~ 1 MDa band on the blue native is only observed upon coexpression of Pik-1 and Pik-2 and absent in other lanes where the same constructs are used, which demonstrates that this is not an non-specific band.

    -Mechanistic interpretation: 1) The resting-state complex is very interesting and aligns with a recent study on NLR oligomer resting-state (Toth et al, 2026 Nature Plants). Since the data here does not fully explain how the complex is inhibited or regulated, I would also suggest authors to refer to that model in this study in their discussion. 2) Discussion should acknowledge this and consider alternative mechanisms, including conformational changes or other interacting partners.

    We added this suggestion.

    • Figures and clarity: 1) Several figures have small or misaligned text, illegible scale bars or confusing labels (especially Figures 1, 3, S4). 2) Terms should be defined at first use and should clearly explain acquisition parameters and molecular weight labels.

    We have implemented these suggestions.

    Overall: This work is technically solid and conceptually interesting, it aligns with the current direction in the field and proposes a new mechanistic model but it does not fully characterize it. Revisions are needed to clarify some aspects, provide additional evidence at least for inactivity status and subcellular plasma membrane localization. Other changes such as improve figure presentation, readability and format would also be required. Addressing all these points would make the manuscript much clearer, more convincing and ready for publication.

    Reviewer #2 (Significance (Required)):

    This is a strong and conceptually interesting study in a very challenging area of work that provides novel insight into the resting-state organisation of paired NLR immune receptors. The authors also acknowledge the limitations of the study as it is very challenging working biochemically with NLRs in the context of cell death. The experimental approach is elegant and adequate, and the findings will be of broad interest to the plant immunity community. Addressing the issues above, particularly figure clarity, membrane localisation evidence, and mechanistic interpretation will substantially strengthen the manuscript and its impact

  2. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #2

    Evidence, reproducibility and clarity

    Summary

    This manuscript investigates the resting-state organisation of the rice NLR pair Pikm-1/Pikm-2 and reports that these receptors form a ~1 MDa hetero-oligomeric complex that accumulates at the plasma membrane prior to activation. Using a combination of experiments (coIP, Blue Native-PAGE, subcellular fractionation and confocal microscopy) the authors provide evidence that both native and engineered Pikm-1 associate with Pikm-2 in a pre-activation. The study addresses an important and timely question in plant immunity on how paired sensor-helper NLRs are organised before activation and offers some conceptual advances relevant to NLR mechanistic. The work is technically ambitious and generally well executed (proper controls, loading samples). However, some data presentation, interpretation and experiments, especially regarding membrane localisation, and proposed functional implications would benefit with further clarification.

    Major comments

    • Clarity and focus of the Introduction: Some paragraphs in the Introduction are very dense and read more like a review article than a research manuscript. While the background is comprehensive, streamlining several sections would improve focus and help guide the reader more directly toward the specific questions addressed in this study.
    • Interpretation of intermediate oligomeric status: Intermediate oligomeric bands are observed along the manuscript when Pikm-1 or Pikm-1Enhancer is expressed without Pikm-2. While mentioned, their possible significance is not fully discussed. Would this intermediates be functionally relevant, assembly precursors or artefacts of overexpression? A clearer discussion would help contextualise the ~1 MDa complex.In same lines, the statement "Overall, the biochemical nature of these puncta remains elusive, as they might consist of resistosome foci or clusters of inactive NLR complexes." would benefit from further clarification.
    • The organisation of Figure 1C is somewhat confusing. I suggest rearranging the blot so that the Ponceau S staining is placed directly under the corresponding input lanes, as both should originate from the same membrane. The immunoprecipitated (IP v5) samples could then be shown below.
    • Given the central role of membrane localisation in the proposed model, the membrane fractionation data raise concerns. In Figure 3A, the decrease in soluble Pikm-1 upon co-expression with Pikm-2 is not clearly accompanied by a corresponding increase in the membrane fraction, whereas Figure S3 suggests a more pronounced enrichment. Similar inconsistencies appear for Pikm-2. In addition, the plasma membrane ATPase control shows variable enrichment between Figure 3A and Figure S3. Because these experiments underpin key conclusions about pre-assembly and membrane recruitment, I strongly suggest repeating and consolidating these fractionation experiments with clearer quantification.
    • The reference to Fig. 4"We detected a GFP signal (using anti-FLAG antibodies) at ~1 MDa only when GFP was co-expressed with Pikm-1Enhancer and Pikm-2EE. " is misleading as GFP is also detected in Pikm1 homo-oligomers at 480 kDa in Fig.4B anti FLAG lane 1. Please clarify, explain or acknowledge.
    • The confocal imaging data are not fully convincing as presented. The use of FDA as a cytosolic marker is acceptable, but the fluorescence intensity profiles differ substantially between samples (peaks vs slopes in fg3.B), and scale bars does not appear consistent, text cannot be read. Not all the presented images show the epidermal cell periphery. To strengthen the conclusions, the authors could consider use a plasma membrane marker (FM4-64 or PI) and/or plasmolysis assays to more definitively demonstrate plasma membrane localisation. Intensity plots should span comparable distances across conditions.
    • Several figures contain very small text, making them difficult to read (e.g. Fig. 1A, 1C, plots in Fig. 2B, and Fig. S4). Panel sizes in Fig. S4 are particularly small and should be enlarged.
    • The authors state the use of Pikm2 EE to abolish downstream cell death but conclusive results from this experiments did not show any major difference in NLR pair association Pikm1Enh and Pikm2EE biochemichaly. A more explicit discussion on alternative activation models (e.g. conformational changes within a pre-assembled complex, changes in membrane topology or altered interactions with additional components) and how to address this issue would strengthen discussion section.
    • OPTIONAL experimental suggestion: it would be informative to examine the dynamics of puncta formation using fluorescently tagged, non-autoactive versions of the NLRs during activation by GFP-HF, rather than relying on autoactive constructs. These experiments could help clarify whether puncta formation reflects NLR activation or is instead a secondary outcome of autoactivity.
    • OPTIONAL: For image clarity I suggest to add Size (kDa) to all WB and BN-PAGE gels.

    Minor comments (including typos and presentation issues)

    • Figure 3A: lane labels at the top of the blot appear misaligned and should be repositioned. Ponceau blot seems not to be "boxed" as in the other images.
    • Figure 3B: scale bars are illegible; remove them from the images and report magnification or scale clearly in the legend.
    • Labelling in different panels along S4 is hard to understand on Max projection or confocal plane. Please rewrite. Details on laser excitation and acquisition in the legend should be moved to M&Ms, currently missing. Also the number of cells with puncta now indicated as a fraction is not very indicative, I suggest making an additional graph with these quantifications.
    • Figure 5. Please label the plasma membrane.
    • In confocal microscopy (Methods), please indicate the acquisition wavelength ranges used for each fluorophore. Correct the reference from Figure S5 to Figure S4. Language and typographical issues -Legend Fig1. "(kDa) are shown at left" to "on the left". -Correct "focussing" to "focusing" -"0.45 μM filter" to "0.45 μm filter" -"leaf disk were sampled" to "leaf disks were sampled". -Improve clarity in "High-order oligomeric resting states were also observed..." → "...have also been observed." -Please cross-check Refs 33 and 72.

    Referees cross-commenting

    After reading both reports, I am quite pleased to noticed quite a bit of overlap between Reviewer 1's points and my own. Here is a summary of the main points I would highlight:

    • NLR constructs: 1) Both reviewers flagged that the GFP-tagged or "enhancer" Pikm-1 constructs might not fully reflect the natural NLR complex. 2) The claim that the hetero-oligomer at the plasma membrane is inactive is not completely supported-yet I acknowledge it is a complex task. Some additional controls or discussion are needed to make this clear (e.g., empty vector, single NLR expression or other activity readouts).
    • Evidence for inactivity1) Mild chlorosis or other subtle phenotypes suggest that overexpression might induce some activity. 2) I agree with Reviewer 1 that more functional evidence would strengthen the claim that the pre-assembled complex is truly inactive.
    • Membrane localization: 1) The confocal images are hard to interpret quantitatively. Cytosolic signal, inconsistent intensity profiles and different magnifications make it difficult to conclude membrane localization. 2) Using a plasma membrane marker (like FM4-64) or alternatively plasmolysis assays would help, as would standardizing imaging parameters across samples.
    • Co-IP and BN-PAGE data: 1) Background bands and ambiguous molecular weights make it tricky to confidently interpret oligomeric states-I would understand in some cases finding extra bands for using N.benthamiana transient expression, not so importnant SDSPAGE in the CoIPs but given the poposed hypothesis oligomeric state for NLRs this should be better addressed, indicated or discussed. 2) Improvement on molecular weight markers, clearer labeling
    • Mechanistic interpretation: 1) The resting-state complex is very interesting and aligns with a recent study on NLR oligomer resting-state (Toth et al, 2026 Nature Plants). Since the data here does not fully explain how the complex is inhibited or regulated, I would also suggest authors to refer to that model in this study in their discussion. 2) Discussion should acknowledge this and consider alternative mechanisms, including conformational changes or other interacting partners.
    • Figures and clarity: 1) Several figures have small or misaligned text, illegible scale bars or confusing labels (especially Figures 1, 3, S4). 2) Terms should be defined at first use and should clearly explain acquisition parameters and molecular weight labels. Overall: This work is technically solid and conceptually interesting, it aligns with the current direction in the field and proposes a new mechanistic model but it does not fully characterize it. Revisions are needed to clarify some aspects, provide additional evidence at least for inactivity status and subcellular plasma membrane localization. Other changes such as improve figure presentation, readability and format would also be required. Addressing all these points would make the manuscript much clearer, more convincing and ready for publication.

    Significance

    This is a strong and conceptually interesting study in a very challenging area of work that provides novel insight into the resting-state organisation of paired NLR immune receptors. The authors also acknowledge the limitations of the study as it is very challenging working biochemically with NLRs in the context of cell death. The experimental approach is elegant and adequate, and the findings will be of broad interest to the plant immunity community. Addressing the issues above, particularly figure clarity, membrane localisation evidence, and mechanistic interpretation will substantially strengthen the manuscript and its impact.

  3. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #1

    Evidence, reproducibility and clarity

    In this manuscript, Pai et al show that the NLR pair Pikm-1/Pikm-2 is putatively present as an hetero-oligomer at the plasma membrane in the inactive state. This report could show that the mechanisms of NLR inhibition are diverse and could potentially inform strategies to engineer synthetic NLRs. As such, it could be an interesting study. However, there are major issues in this study.

    Major comments:

    One major comment comes from the use of modified Pik receptors. In this study, the fact that GFP is seen in the Pikm-1 enhancer / Pikm-2 active complex in fig 4 may not represent the normal active state of the NLR pair. It is interesting that such a complex can exist, but the author should clarify figure 5 and precise that the complex represented is not an NLR-effector complex but an artificial NLR-GFP complex. It is claimed that the NLR pair oligomerizes at the plasma membrane in the inactive state. It cannot be ruled out that the pair is partially active when overexpressed. Activity of NLRs does not always result in cell death. In figure 2C, a mild chlorosis phenotype is clearly seen in the mcherry co-expression side. This experiment needs more controls (empty vector, use EV + Pikm-1, EV + Pikm-2) and some other proxy for NLR activation should be used to claim that the complex is inactive (gene expression or else as appropriate).

    I am not convinced by the confocal microscopy data presented in figure 3 and by the Co-IP data in figure 1. Figure 3: The confocal microscopy presented in figure 3 needs some refining. This experiment may not be suitable to estimate the change in receptor localization as the difference is quantitative. We can always see some signal in the cytoplasm and putatively in the plasma membrane and it is difficult to evaluate what is signal and what is noise. A plasma membrane marker should be used in this experiment to conclude on the localization of the receptors. Upper panels: It is claimed that Pikm-1 fully co-localized with fluoresceine when co-expressed with Rx. On the figure, I can see purple signal and blue signal that do not co-localize. Co-localization of Pikm-1 when co-expressed with Rx or Pikm-2 is not clearly different. Most importantly, the images were not taken using the same magnification which prevents a clear comparison. Define FDA in the caption. Figure 1: Co-IP experiment shows many bands that may or may not be considered as background noise. In the first lane, Pikm1-v5 is expressed with Rx-HA. On the v5 blot of the input, we can see a band at approximately 130kDa that is supposedly Pikm1-v5. On the HA blot, we can also see a band at 130kDa that is supposed to be Rx-Ha. However, on the third lane of the v5 input blot, Rx-v5 is much lower. Thus, it is not clear what this 130kDa band is. Size of Pik1 and 2 are said to be around 110kDa. In the IPv5 blot v5 part, lane 1 where Pikm-1-v5 is co-expressed with Rx-HA, we cannot see a clear Pikm-1 band at around 130kDa. Both Pikm1 and 2 bands are supposed to be around 130kDa on this gel, as indicated by the red arrows, but this size is very close to this non-specific band. Overall, there are too many non-specific bands to draw conclusions from this experiment.

    Minor comments:

    I would like to ask the author to be careful about the use of jargon and to always define terms at first use. In the abstract, Pikm is not defined; in the introduction, ID is not defined, ZAR1, Sr35, WAI3 etc... The term Pikm1 "enhancer" is confusing, it seems that this protein is just a hyperactive version of Pikm-1. Please try to be more accessible to students and non-specialists, this field is becoming increasingly opaque for non-specialists.

    Figure 3: A: Labels are not aligned with the lanes. Page 7"To determine whether treatment with the ligand would alter the Pik hetero-complex" this statement is confusing, this is not what you test.

    Significance

    This study is potentially interesting but the novelty is limited. The interaction of these NLRs was shown before. Here the authors do not show the nature of the complex, its stoichiometry. The presence of effector in the complex is shown using an artificial NLR and GFP that may not represent the normal complex.

    If the NLRs were forming oligomers and were present at the plasma membrane (and this is still not clearly demonstrated), the question is: How is the complex inhibited? If the authors were to describe this putative inhibition mechanism, this would be highly interesting.

  4. Version published to 10.64898/2025.11.30.691369 on bioRxiv