Global fungal-host interactome mapping identifies host targets of candidalysin

Candidalysin, a cytolytic peptide toxin secreted by the human fungal pathogen Candida albicans, is critical for fungal pathogenesis. Yet, its intracellular targets have not been extensively mapped. Here, we performed a high-throughput enhanced yeast two-hybrid (HT-eY2H) screen to map the interactome of all eight Ece1 peptides with their direct human protein targets and identified a list of potential interacting proteins, some of which were shared between the peptides. CCNH, a regulatory subunit of the CDK-activating kinase (CAK) complex involved in DNA damage repair, was identified as one of the host targets of candidalysin. Mechanistic studies revealed that candidalysin triggers a significantly increased double-strand DNA breaks (DSBs), as evidenced by the formation of γ-H2AX foci and colocalization of CCNH and γ-H2AX. Importantly, candidalysin binds directly to CCNH to activate CAK to inhibit DNA damage repair pathway. Loss of CCNH alleviates DSBs formation under candidalysin treatment. Depletion of candidalysin-encoding gene fails to induce DSBs and stimulates CCNH upregulation in a murine model of oropharyngeal candidiasis. Collectively, our study reveals that a secreted fungal toxin acts to hijack the canonical DNA damage repair pathway by targeting CCNH and to promote fungal infection.

4. The animal data is troubling for a few reasons.The first, is that changes in MN PCE%, H2AX, and CCNH are marginal, and at times the data is not quantitative.Moreover, the ece1Δ/Δ and CLΔ mutants fail to colonize the murine tongue (100-fold less), so it is impossible to make interpretations about the role of CL in driving DNA damage.With significantly less fungus present, it could be due to any number of fungal effectors or host responses.Performing experiments in point 3 above might help clarify this. 5.A scrambled peptide or similar length amphipathic peptide (e.g.antimicrobial peptide) would have been incredibly helpful to rule out non-specific interactions.Due to its capacity to interact with membranes, CL could be associated with mammalian targets due to indirect interactions.Using a peptide with similar amphipathicity may help further delineate true targets from relatively non-specific interactors, especially given that the Ece1 peptides are expressed at high levels.This would also be a useful control for Biacore experiments.

Fig 4D.
There appears to be very few cells in the control panel (should at least be DAPI+), so comparison against CL treatment is unfair.Experiment should be repeated and quantified.7. Fig. 4G.Knockdown experiment results are not very convincing.There appears to be only a very modest increase in H2AX with CL-treatment.Moreover, why is this experiment done in an A549 (lung) cell line as opposed to oral epithelial cells used throughout and relevant to the animal experiments?8. Fig. 4J.Unsure what the horizontal line is depicting.The figure likely needs edited.9.This reviewer is not convinced that protein-protein interactions between CL and CCNH are driving DNA damage.Given its capacity to permeabilize membranes and lyse cells, it is possible that these processes in general are driving DNA damage and subsequent responses.The effects may not depend on specific interactions.The authors should utilize fluorescence microscopy to show that CCNH and CL specifically interact in cells or in vivo using a similar V5-epitope tagged CL as described (PMID: 33471869).
10.More detail regarding strain construction is required.Were these mutants acquired from the PIs who originally described CL or were they made in house?The authors state the strains/primers are listed in Table S2 but I do not see them.
11. Were the peptides expressed with 'K' or 'KR' as the terminal sequence?While they both induce damage similarly, the terminal R is typically cleaved by the Kex1p protease and the majority of peptides produced by Ca are of the 'K' variety.How does this impact interactions or data interpretation?It should at least be discussed.
12. Recently, it was shown that CL can polymerize into loops and insert into membranes (PMID: 36173096).How does this potentially impact its interaction with CCNH?This should be discussed.
Minor comments 1.The word "Candidalysin" should be changed to "candidalysin" throughout to be consistent with current usage and nomenclature for bacterial toxins.

A figure (perhaps new 1B
) showing the sequence of each Ece1 peptide would be useful.
3. Line 80: better to state "including systemic or mucosal candidiasis".Otherwise, you leave out vaginal and cutaneous candidiasis.
7. Line 278-286.I find the description here quite confusing.You may wish to reword for clarity.13.Line 505: tense incorrect 14.Line 527-533: This is written like a protocol instead of a description of the procedure performed.
Reviewer #3 (Remarks to the Author): The paper entitled "Global fungal-host interactome mapping identifies a novel target of Candidalysin for fungal infection" the authors performed an HT-eY2H assay to map the global interactomes of Ece1 peptides, including Candidalysin, with a part of the human ORFeome.They identified Candidalysin peptide target proteins in the host, and functionally characterized Candidalysin's interaction with CCHN, and its mechanism of action of this peptide.The manuscript is well executed, presents novel and interesting data to understand the pathogenesis of C. albicans infection.Only, there are points that they need to improve, which I describe below.In the abstract the authors need to describe that they identified targets for the other Ece1 peptides, and that they also found shared targets between the peptides.In the introduction, please report the sizes in amino acids of the rest of the peptides, they only describe that of Candidalysin.The authors describe that "In this study, a HT-eY2H assay was used to map the global interactomes of Ece1 peptides, including Candidalysin, with the human ORFeome", I think it is important that they mention that it is a part of the human ORFeome, since there are only 13761 human proteins.In results, the authors need to make a figure where they show the amino acid sequences of the 8 peptides of the Eci1 protein (I-VIII).They could identify signatures or boxes conserved between them, or also analyze if there is a correlation between the type of amino acids (for example hydrophobic or hydrophilic, etc) of these Eci1 peptides, and the target proteins that were identified in the two hybrids.All of the above, to analyze if any common region can be identified between the Ece1 peptides, when they identified overlapping human interacting proteins.Another experiment that could be carried out is BiFC in human cells between the CCHN protein and the Candidalysin peptide, to observe in vivo the reconstitution of GFP fluorescence during the interaction of CCHN and the peptide.Regarding the mutant ece1Δ/Δ+ECE1Δ184-279, the authors have to describe which peptides the deletion would cover, I suppose amino acids 184 to 279, and I believe that this region should contain the peptide Ece1-III (Candidalysin), is this correct?Please describe this part in detail.In the discussion section: "Thus, studying Ece1 peptide-host protein-protein interactions (PPIs) was crucial for understanding the mechanisms of fungal infection and the host response, and potentially to develop new strategies for fungal disease treatment and prevention" however, the authors only describe in This section mainly about the interactors of Candidalysin, and the other peptides discuss very briefly; I think that in the results section, they describe in greater detail, what they lack in discussion, for example, which are the human interactors that are particular to each Ece1 peptide, and which ones are overlapping.Also, the possible molecular mechanisms that each Ece1 peptide generates in the human interactome, and its effect on the infection of the fungus.In this sentence: "Interaction networks are especially important as proteins generally act not in isolation but in concert with other proteins.Such interactomes can thus reveal biological pathways and processes impacted by the viral proteome, allowing for the discovery of novel drug targets"; however, I believe they are referring to the fungal proteome, rather than the viral proteome.In the sentence "In addition, we also discovered that RHOA has the potential to be a host interactor as shown by an associated increase in mRNA and protein expression (Fig. S2a and b)", the authors have to describe which peptide(s) of Ece1 was identified as a RHOA interactor.This paragraph "Through interactions of these host proteins with Ece1-II and Ece1-V peptides, we observed the synergistic or antagonistic mechanisms among Ece1 peptides and Candidalysin in host immune-related functions", is not very clear, the authors could improve it, on What synergistic or antagonistic mechanisms are they referring to?Finally, the font size of most of the figures is very small, even some diagrams and graphs, please make it as big as possible.
Reviewer #4 (Remarks to the Author): Authors of the presented manuscript uncovered potential host targets for Candida albicans' cytolytic peptide toxin, Candidalysin.Authors performed rigorous genomic-based studies and evaluated the potentiality of human CCNH through ex vivo and in vivo experimental investigations.Additionally, binding affinity between Candidalysin and its postulated target was assessed through molecular docking approach.The manuscript is valuable in its field as it redeems publication following the address of these suggestions and comments.1. Authors performed homology modelling to obtain the 3D structure of CCNH and Candidalysin for molecular docking analysis.However, CCNH atomic structure is actually deposited within the protein Data Bank database (https://www.rcsb.org/)PDB ID: 1KXU.2. Furthermore, authors should made findings for the constructed structures using several online validation tools such as ERRAT, Verify3D and Rampage (like Z-score and Ramachandran plots) to be available even within the supplementary materials.3. Zoomed 3D image(s) for CCNH-Candidalysin binding interface should be provided to highlight the polar and hydrophobic interactions between the two protein structures.Additionally, analysis for such binding interaction should be performed estimating the bond's angle and/or distances.4. Authors are advised to explore the comparative mm-GBSA binding energies and its contributing energy terms (polar, hydrophobic, solvation, …) using utilizing FastDRH web server (https://cadd.zju.edu.cn/fastdrh/ ) for the CCNH-Candidalysin complex.This would highlight the nature and magnitude of binding the thing that would guide further studies of optimization.5. Performing short time Molecular Dynamic simulation (e.g.50 ns) is advised which in turn would highlight the thermodynamic stability of Candidalysin at the CCNH binding site as well as potential target conformational alterations which would guide future interaction analysis.

Reviewer #1:
In this study, the authors present a comprehensive interactome dataset using an enhanced yeast-2 hybrid system to identify potential human cellular interactors with the Candida toxin protein, Ece1.This is an important piece of workparticularly as it relates to the other Ece1 peptides apart from candidalysin, as they have no currently known function.In general, the work has been carried out to a good standard.However, the authors do not take the work too far with these toxins (which would be of particular value), and there are some important concerns around the work: Answer：We are grateful to the reviewer for the pertinent comments.The manuscript has now been carefully revised and additional data were added in the revised manuscript, and the major revisions were marked with red color.The point-by-point responses to the reviewers' comments are shown as follows.

Q1. Host target for Clys isn't elusiveit's the membrane. Better to rephrase to intracellular targets have not been extensively mapped or some such. (Abstract line 29)
Answer：We appreciate the reviewer's suggestion and have rephrased "host targets" to "intracellular targets" in the revised manuscript.(Line 31 on page 2)

Q2. What do the authors mean by an avirulence factor (line 98)? This doesn't sound correct. Do they mean that it functions to promote immune responses as well as being a virulence factor? This is NOT an avirulence factor, as it relates to the host and host responses.
Answer: We thank the reviewer for this critical suggestion.We agree with the reviewer's comment that "avirulence" may mislead the readers, and therefore we rephrased it to "immunoavoidance", also suggested by another reviewer.(Line 87 on page 4) Q3.Line 103saying that the human target of Clys remains unclear is not correctwe know that it targets the phopholipases in the cell membrane as a pore-forming toxin -in the original Nature paper.Direct human protein targets, however, aren't well knownyou need to re-phrase this sentence and other occurrences in the manuscript.
Answer: We apologize for the lack of clarity and thank the reviewer for raising this crucial point.The sentence was rewritten as "However, the direct human protein targets of candidalysin and the related mechanisms in promoting systemic fungal infection have not been fully defined."Moreover, we also described the known target of candidalysin, the phopholipases in the cell membrane, in the revised Discussion section.

Q5. How do the authors explain the similarities in interactions between Candidalysin and ECE1-II, given the high degree of sequence and structural difference in the two
peptides?Similarly for other shared genes.
Answer：Thanks for the reviewer's valuable consideration.Although the sequences of these 8 Ece1 peptides are different, there exists structural binding similarity for Ece1-II and Ece1-III.Specifically, following the analysis of structural prediction with AlphaFold and molecular docking, we found that Ece1-II and Ece1-III could bind to host proteins such as THPO and SPRED2 with similar hydrophobic interactions, as shown below (Fig. R2).The docking results via HPepDock revealed that both Ece1-III and Ece1-II bind within the hydrophobic groove (yellow) of the THPO protein (PDB id: 8G04) or SPRED2 protein (PDB id: 2JP2).Similarly, this interaction pattern could lead to the relatively large number of share genes between these peptides.9 8 8 7 7 6 6 5 4 4 3 3 2 2   4 4 4 3 3 2 2 2 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 2   Answer：We appreciate the reviewer for raising this crucial point.We have repeated the fungal infection experiment in mice, with the introduction of an additional Clysinfected group.The results of micronucleus tests showed a significant increase in the bone marrow micronucleus rate (MN PCE‰) in mice infected with Clys compared to the control group (Fig. 5c).However, there was no statistically significant difference compared to mice infected with the WT strain, confirming the genotoxicity role of Clys during Candida albicans infection in mice.Furthermore, for the Ece1 null strain infected group, the MN PCE‰ was similar to that of sham infected group.These results indicate that the reduced MN PCE rates observed in mice infected with the ece1Δ/Δ and ece1Δ/Δ+ECE1 Δ184-279 strains are possibly due to the absence of Clys rather than the decreased fungal load.All these modifications are adapted in the revised manuscript.(Lines 326 to 331 on pages 12 to 13) Q13.The data is not sufficiently discussed in the discussionthere is too much emphasis on what other studies have shown, rather than discussing the findings shown here in this context.
Answer：We appreciate the reviewer for raising this crucial point.We have reorganized the discussion part in the revised manuscript as follows.
(1) Summary: The paragraph "Candidalysin plays a vital role in fungal infection…" and "In this study, …" (2) Discussion about candidalysin as a pore-forming toxin and discussion about interaction of host protein targets and other Ece1 peptides: "Previous work has shown that candidalysin damages the cell membrane to promote infection.However, how candidalysin does this remained unclear.Russell, Schaefer et al. showed that candidalysin used a unique pore forming mechanism by creating a loop structure to insert into the membrane, leading to the membrane damage of human cells.It would be interesting to investigate whether the CL mutation (G4W) or inhibitors that stop candidalysin from forming the polymers could induce the DNA damage in the future.
Notably, far less attention has been given to the role of the other seven Ece1 peptides.
In our study, we have also identified the host interactors of these Ece1 peptides.There are many overlapping or intersecting functions between the Ece1 peptides.For example, …" (3) Discussion about the role of candidalysin in DNA damage and tumorigenesis: The paragraph "Consistent with mycotoxins which can induce DNA damage in host, …" and the paragraph "Invasive fungal infections have become a high-risk factor of increasing morbidity and mortality in cancer patients.For example, …" (4) Outlook of further study: "Together, our study not only provide the human interactors of Ece1-peptides including candidalysin which may serve as potential therapeutic targets, but also make it possible in the future to analyze the genome-wide interactome between the CANDIDA ORFeome and the Human ORFeome at the species level."(Lines 347 to 413 on pages 13 to 16) Reviewer #2:

Summary
The objective of this manuscript by Zhang, et al. is to identify the mammalian molecular target(s) and additional function of the peptide toxin candidalysin (CL) secreted by the opportunistic fungus Candida albicans (Ca).In addition, the authors also sought to identify potential interactions of non-CL Ece1 peptides, which have not been ascribed functions to date.A yeast 2-hydbrid (Y2H) approach was used in which individual Ece1 peptides were fused to the GAL4 activation domain (bait) and Gibson cloning used to create human ORFeome fusions to the GAL4 DNA binding domain (prey) to identify potential interacting partners.Bioinformatic analyses revealed a striking amount of unique predicted mammalian interactions amongst individual Ece1 peptides and several that have common shared targets.Ontological analysis suggested that most Ece1 peptides play a role in keratinization, while CL) has predicted roles in signaling olfactory receptors and chromosome organization/cyclin D events in the G1 phase.Given the latter predicted role of CL, the authors then focus on CCNH, a subunit of a CDK-activating kinase important for cell cycle control.Flow cytometry revealed dysregulated cell cycle upon CL treatment and immunohistochemistry showed H2AX phosphorylation (a vital step in DNA damage repair).Expression of DNA repair related genes were up-regulated during short exposure of CL but not during extended exposure.Thus, the authors hypothesized that CL induces DNA damage and by binding to CCNH can suppress the DNA repair process.Surface plasmon resonance of recombinant CCNH with CL revealed a dissociation constant of 17 nM, suggesting strong affinity between these two molecules.Lastly, a mouse model of oropharyngeal candidiasis was performed using WT, ece1Δ/Δ, and ΔCL strains where CCNH, H2AX, and micronucleate polychromatic erythrocyte induction were interpreted to be CLdependent, as was fungal burden.

While this is a very potentially exciting body of work, especially regarding potential interactions with non-CL Ece1 peptides, there a number of technical and theoretical concerns throughout. Please find more detailed comments below.
Answer: We appreciate the reviewer's detailed and helpful comments.The manuscript has now been carefully revised and additional data were added in the revised manuscript, with the major revisions marked with red color.Please take a look at the following point-by-point responses addressing the reviewer's comment.

Q1: While I can appreciate the enormous effort that went into performing the Y2H
screen with each Ece1 peptide, there exists a very high false-positive rate with such an approach.The gold standard to validate findings is to perform co-immunoprecipitation.
The authors should do this with tagged CCNH and Ece1-II, -III, -IV, -V since these were found as interactors in the Y2H screen.Do these other peptides also showing binding by surface plasmon resonance?Answer: We appreciate the reviewer for raising this crucial point.We have conducted co-immunoprecipitation assays to validate the direct interaction between intracellular Ece1 peptides and CCNH.Plasmids encoding 3xFlag-tagged CCNH and Ece1-II to V peptides constructs linked with EGFP were transiently transfected into HEK-293T cells.
The co-immunoprecipitation results demonstrated the interaction between intracellular CCNH and Ece1 peptides (Ece1-II to V). (Supplementary Fig. 5a) Furthermore, we validated the binding of Ece1-II to V peptides to CCNH through surface plasmon resonance assays (via BIAcore).The results revealed a strong binding affinity between these peptides and CCNH, as compared to the negative control antimicrobial peptide (Amp).This finding was consistent with our results from the Y2H screening.Among the peptides, the strongest binding observed is CCNH-Clys (Ece1-III).(Supplementary Fig. 5g &  Q2: In the seminal paper in which CL was described (PMID: 27027296), only Ece1-II, -III, -V, and -VI were found in culture supernatants by LC-MS, suggesting that only these peptides are released extracellularly.In fact, Ece1-III (CL) was found to the greatest extent, and should the others get secreted they are likely a minority.Given these, what is the relevance of looking at interactors of most of the Ece1 peptides?Answer：We appreciate the reviewer for raising our attention to this point.In this seminal paper, the supernatant used for LC-MS detection of peptides was from fungal cell culture, but not co-cultures of fungi with cells.We hypothesized that the secretion profile of peptides might differ from that after co-culturing with host cells.We treated FaDu cells separately with Ece1 peptides and then conducted RNA-seq analysis on them to study the changes in cellular gene expression post-peptide treatment.The results revealed the potential different functions of each peptide during fungal infection (Supplementary Fig. 1).That's why we performed the Y2H screening of the interactors of all the Ece1 peptides.All these modifications have been supplemented in the revised manuscript.(Lines 103 to 112 on page 5) Q3: There are additional concerns about concentration dependence.For example, the authors should show that WT Ca infection drives similar DNA damage repair expression profiles, H2AX induction as presented in Fig. 4 and that an ece1Δ/Δ fails to do so.See additional comment below.
Answer: We thank the reviewer for this crucial point.Firstly, we determined the concentration of Clys by infecting cells with different concentrations of Clys to examine DNA damage.Our findings revealed a concentration-dependent effect of Clys on DNA damage, with 10μM concentration as the most significant upregulation of γ-H2AX.(Supplementary Fig. 5e) Secondly, we performed the infection assay by co-culturing cells with the C. albicans WT (SC5314) and ece1Δ/Δ.Then the expression levels of γ-H2AX and DNA damage repair associated genes (53BP1, ERCC1, ERCC2, ERCC4) were measured.We observed a significant activation of γ-H2AX expression in both WT and ece1Δ/Δ+ECE1 strains, indicating DNA double-strand breaks caused by the strains.
Consistently, we also found the upregulation of DNA damage repair associated genes (53BP1, ERCC1, ERCC2, ERCC4) after WT C. albicans infection, while the ece1Δ/Δ strain failed to do so (Supplementary Fig. 5c).All these modifications have been supplemented in the revised manuscript.(Lines 256 to 257 on page 10) Q4: The animal data is troubling for a few reasons.The first, is that changes in MN PCE%, H2AX, and CCNH are marginal, and at times the data is not quantitative.
Moreover, the ece1Δ/Δ and CLΔ mutants fail to colonize the murine tongue (100-fold less), so it is impossible to make interpretations about the role of CL in driving DNA damage.With significantly less fungus present, it could be due to any number of fungal effectors or host responses.Performing experiments in point 3 above might help clarify this.
Answer: We thank the reviewer for this critical comment.We have repeated the fungal infection experiment in mice, with the introduction of an additional Clys-infected group.
The results of micronucleus tests showed a significant increase in the bone marrow micronucleus rate (MN PCE‰) in mice infected with Clys compared to the control group (Fig. 5c).However, there was no statistically significant difference compared to mice infected with the WT strain, confirming the genotoxicity role of Clys during Candida albicans infection in mice.Furthermore, for the Ece1 null strain infected group, the MN PCE‰ was similar to that of sham infected group.These results indicate that the reduced MN PCE rates observed in mice infected with the ece1Δ/Δ and ece1Δ/Δ+ECE1 Δ184-279 strains are possibly due to the absence of Clys rather than the decreased fungal load.Furthermore, the results of IHC assay also indicated the DNA damage in the tissue caused by Clys and strains capable of secreting Clys (WT and ece1Δ/Δ+ECE1) in mice, along with the relative upregulation of CCNH expression.All these data have been quantified (Fig. 5e).Thus, the aboving results indicated the role of Clys in driving DNA damage.All these modifications are adapted in the revised manuscript.(Fig. 5) Q5: A scrambled peptide or similar length amphipathic peptide (e.g.antimicrobial peptide) would have been incredibly helpful to rule out non-specific interactions.Due to its capacity to interact with membranes, CL could be associated with mammalian targets due to indirect interactions.Using a peptide with similar amphipathicity may help further delineate true targets from relatively non-specific interactors, especially given that the Ece1 peptides are expressed at high levels.This would also be a useful control for Biacore experiments.
Answer：We appreciate the reviewer for this pertinent comment.We followed the reviewer's suggestion and used an antimicrobial peptide (Amp) as a negative control in the BIAcore experiment.The equilibrium dissociation constant of Clys-CCNH binding is KD=1.720×10 - M (pKD=7.76)while that of Amp-CCNH binding is KD=6.26×10 - M (pKD=4.20).This result validated the relatively high affinity binding between Clys and CCNH.(Fig. 4k and Supplementary Fig. 5g &  Answer：We apologize for this confusion.We actually have tried several times to knockout CCNH in FaDu cells (oral epithelial cells used throughout) by using the CRISPR/Cas9 system.However, we observed a significantly low survival rate which prevented the selection of the positive depletion mutant cells (Fig. R4).We hypothesized that this might be due to the essential role of CCNH in FaDu cells, the depletion of which led to cell death.Additionally, as a supplement, we have conducted  Answer: We thank the reviewer for giving this significant suggestion.We have added a figure (Fig. 1a) showing the amino acid sequences of 8 Ece1 peptides (I-VIII).We also conducted the conserved sequence analysis using the online tool of multiple sequence alignment (Clustal Omega, https://www.ebi.ac.uk/Tools/) and we found low homology or similarity between the Ece1 peptides as shown below (Fig. R5).We hypothesized that the similarity in interactors of these peptides may be associated with the similar intermolecular interaction between them.Answer: We appreciate the reviewer for raising this crucial point.We have reorganized the discussion part in the revised manuscript as follows.
(1) Summary: The paragraph "Candidalysin plays a vital role in fungal infection…" and "In this study, …" (2) Discussion about candidalysin as a pore-forming toxin and discussion about interaction of host protein targets and other Ece1 peptides: "Previous work has shown that candidalysin damages the cell membrane to promote infection.However, how candidalysin does this remained unclear.Russell, Schaefer et al. showed that candidalysin used a unique pore forming mechanism by creating a loop structure to insert into the membrane, leading to the membrane damage of human cells.It would be interesting to investigate whether the CL mutation (G4W) or inhibitors that stop candidalysin from forming the polymers could induce the DNA damage in the future.
Notably, far less attention has been given to the role of the other seven Ece1 peptides.
In our study, we have also identified the host interactors of these Ece1 peptides.There are many overlapping or intersecting functions between the Ece1 peptides.For Answer: We apologize for this mistake and we deleted it.(Line 358 on page 14) Q9: In the sentence "In addition, we also discovered that RHOA has the potential to be a host interactor as shown by an associated increase in mRNA and protein expression (Fig. S2a and b)", the authors have to describe which peptide(s) of Ece1 was identified as a RHOA interactor.

8.
Fig 4i: I assume that CL is highlighted in pink?More description in the legend needed.9. Fig. 5e: The statistical markers are placed over the wrong strains.10.Line 386: reference incorrect here.11. Line 397-403.I suggest moving this to the Results.12. Line 413-415: I'm not sure what the authors mean.I do not see NLRP3 in the list of potential interactors.Please clarify.
Were the proteins that interacted with the different individual peptides or groups of peptides from similar functional groups or specific pathways?(line 152) Answer：We thank the reviewer for raising this crucial point.Indeed, proteins that interact with different individual peptides or groups of peptides could be enriched in various pathways, which were further clustered as different functional groups (Fig.R1, Extended data 2, Enrichment Sheet).The representative 20 pathways and corresponding genes were shown in Supplementary Fig.2.For example, the interactors of Ece1-I, II and III could be enriched in the pathway of Cyclin D associated events in G1, which involves the genes such as CCNH, PPP2CA and THPO.Ece1-I, II, IV and V are all associated with Herpes simplex virus 1 infection, enriching genes like ZNF90 and EIF2B2.The detailed information about interactors of different peptides were listed in Extended data 2 (Annotation sheet), related to Fig.3.

Fig. R1
Fig. R1 Upset plot showing 20 pathways and the corresponding interactors which could be enriched in specific or shared pathways.

Fig. R2
Fig. R2 The structural binding similarity of Ece1-II and Ece1-III with THPO and SPRED2.a, the docking results via HPepDock revealed that both Ece1-III and Ece1-II bind within the hydrophobic groove (yellow) of the THPO protein (PDB id: 8G04); b, the binding patterns of Ece1-II and Ece1-III with SPRED2 protein (PDB id: 2JP2).Both Ece1-III and Ece1-II bind within the hydrophobic groove (yellow) of the SPRED2.
Fig. R3The binding efficiency of Ece1-II, IV, V and Amp (antimicrobial peptide) to CCNH protein was evaluated by BIAcore.
h) Q6: Fig 4D.There appears to be very few cells in the control panel (should at least be DAPI+), so comparison against CL treatment is unfair.Experiment should be repeated and quantified.Answer：We thank the reviewer for raising this issue.We have repeated the infection assay with an equivalent number of cells and performed the co-localization immunofluorescence analysis of CCNH and γ-H2AX between the Clys-treated and control groups within the field of view using ImageJ software.After infection with Clys, we observed the increased co-localization of CCNH and γ-H2AX intracellularly.All these modifications are adapted in the revised manuscript.(Fig.4e) Q7: Fig.4G.Knockdown experiment results are not very convincing.There appears to be only a very modest increase in H2AX with CL-treatment.Moreover, why is this experiment done in an A549 (lung) cell line as opposed to oral epithelial cells used throughout and relevant to the animal experiments?

a
Fig.4f, we discovered that inhibition of CCNH led to a suppressed expression of γ-H2AX within the cells, revealing the role of CCNH in mediating Clys-induced

Fig. R5
Fig. R5 Conserved sequence analysis using the online tool of multiple sequence alignment.a, multiple sequence alignment of Ece1-I to VIII peptides.The number right of the sequence represents the length of each peptide; b, the 8 Ece1 peptides were clustered according to the homology score.The results of multiple sequence alignment.and homology analysis showed low similarity between the Ece1 peptides.
Discussion about the role of candidalysin in DNA damage and tumorigenesis: The paragraph "Consistent with mycotoxins which can induce DNA damage in host, …" and the paragraph "Invasive fungal infections have become a high-risk factor of increasing morbidity and mortality in cancer patients.For example, …" (4) Outlook of further study: "Together, our study not only provide the human interactors of Ece1-peptides including candidalysin which may serve as potential therapeutic targets, but also make it possible in the future to analyze the genome-wide interactome between the CANDIDA ORFeome and the Human ORFeome at the species level."(Lines 347 to 413 on pages 13 to 16) Q8: In this sentence: "Interaction networks are especially important as proteins generally act not in isolation but in concert with other proteins.Such interactomes can thus reveal biological pathways and processes impacted by the viral proteome, allowing for the discovery of novel drug targets"; however, I believe they are referring to the fungal proteome, rather than the viral proteome.