NIR light-activated nanocomposites combat biofilm formation and enhance antibacterial efficacy for improved wound healing

Nanoparticle-based therapies are emerging as a pivotal frontier in biomedical research, showing their potential in combating infections and facilitating wound recovery. Herein, selenium-tellurium dopped copper oxide nanoparticles (SeTe-CuO NPs) with dual photodynamic and photothermal properties were synthesized, presenting an efficient strategy for combating bacterial infections. In vitro evaluations revealed robust antibacterial activity of SeTe-CuO NPs, achieving up to 99% eradication of bacteria and significant biofilm inhibition upon near-infrared (NIR) irradiation. Moreover, in vivo studies demonstrated accelerated wound closure upon treatment with NIR-activated SeTe-CuO NPs, demonstrating their efficacy in promoting wound healing. Furthermore, SeTe-CuO NPs exhibited rapid bacterial clearance within wounds, offering a promising solution for wound care. Overall, this versatile platform holds great promise for combating multidrug-resistant bacteria and advancing therapeutic interventions in wound management.

authors are able to validate their 4-protein diagnostic signature in an external, blinded cohort, as a limitation of large amounts of data can be a bias towards significance.Nevertheless, the results are interesting and advance our knowledge of MIS-C.Few minor comments: There have been a handful of papers published on the use of proteomics in MIS-C (https://doi.org/10.1172/JCI151520,https://doi.org/10.1038/s41467-021-27544-6)which the authors may want to include.In the JCI paper by Porritt RA et al, authors utilized proteomics to distinguish severe MIS-C from mild disease and KD, and similarly found upregulation in FcGRIIIa and a reduction in proteins involved with lipid metabolic processes, lipoprotein clearance, and components of the coagulation cascade.
Could the authors provide a reference for the external US validation cohort of MIS-C and controls patients?Thank you for the opportunity to review this interesting study.

Rebuttal
In the submitted manuscript, the authors reported that using mass-spectrometry proteomics and AIassisted unbiased analysis they identified four plasma protein biomarkers (LCP1, FCGR3A, SERPINA3, BCHE) that could highly discriminate multisystem inflammatory syndrome in children (MIS-C) from Kawasaki disease (KD) and severe bacterial and viral infections.They used MIS-C patients meeting CDC case definitions and relevant control groups.I believe the manuscript adds to foundation for the development of a diagnostic test that could assist pediatricians to timely diagnose MIS-C.

Response:
Thank you very much for reviewing this paper!However, I have some questions I hope the authors could address:

Major Comments:
-While the authors used relevant control groups to compare with MIS-C cases, one major limitation of this study is the small sample size, particularly in KD patients, limiting further validation of results observed in this study and this is briefly mentioned in the Discussion.

Response:
We agree and acknowledge that the number of KD patients in this study is very small.Before the initiation of the study, we considered omitting KD patients due to the low number (all identified prior to the pandemic).However, we decided to include them to illustrate the performance of the signature on KD patients, despite the few.
-Sample size calculations are not provided in the methods sections to determine the required sample size for each diagnostic category included in this study.

Response:
We did not perform sample size calculations as the nature of the study was exploratory.However, we found numerous significant changes (also corrected for multiple testing) based on a relatively small number of participants.This support that sample size per see was not problematic but we agree with the reviewer that a larger sample size may have provided additional insight and lower p-values.We have now inserted this in the method section in the revised manuscript.
Change in methods section: -Another major limitation was the lack of validation of the mass spectroscopy results by antibody-based methods that are more amenable to translation into a diagnostic test.This was not mentioned as a limitation, only the fact they were unable to validate the results in an independent cohort.

Response:
Mass-spectrometry was considered a more specific measurement technique than antibody-based methods.This is due to the fact that antibodies may suffer from interference due to their epitope specificity and matrix effects (e.g., plasma moieties).We believe the presented data thereby offer a more specific approach to measure proteins as also generally known as 'gold standard' for protein chemistry.

Major points:
1. Did the authors compare the performance of their signature in MIS-C cases arising from the Alpha vs Delta variant?

Response:
Thank you for this question.We did not compare the performance on the 4-protein signature on MIS-C cases arising from the Alpha vs. Delta variant due to the overall low number in our study.In our cohort, we have previously shown that the clinical phenotype was similar between the two variants (Holm, 2022, JAMA Ped;doi:10.1001/jamapediatrics.2022.2206 2. Given the large proteomic dataset that was generated in this study, did the authors also look for markers of severity in MIS-C cases (as many were severe with shock, requiring ICU and/or inotropes) or only diagnostic markers?

Response:
This is a very relevant question.Thank you.As most of our included patients 21 (78%) presented with shock, we did not look for markers of severity markers.Thus, the 4-protein signature primarily?reflects children with severe MIS-C.
3. Lines 146-149: 'All 12 machine-learning algorithms, except one, had MCC and AUC above 0.77 (Figure 3B).We continued the subsequent analysis with a support vector classification model that, during model training, had an AUC and MCC of 100% and 1, respectively.'It's not clear what was done here in terms of improving the AUC by 23%.These needs explained in more detail and confidence intervals added.

Response
We apologize for not being clear on this part.We evaluated 12 machine-learning algorithms.Eleven algorithms had MCC and AUC between 0.77 and 1, while one (QDA) had MCC and 0.27 and AUC of 0.65.
Figure 3B shows the performance of 12 different algorithms.We did not do anything to improve the algorithms, but chose the best model (SVC), which had MCC and AUC of 1.All 12 algorithms were trained on the same data, so the difference in performance depended on how well each algorithm modelled the data.
The SVC algorithm had fully learned the input data, whereas the GP algorithm had partially learned the data structure, which can be seen by the 23% lower MCC on the same data.Thus, we did not further improve the models.Confidence intervals (standard deviations) are illustrated as errorbars on the barplot (Figure 3B).
We have now further elaborated this in the results section to make it clear for the reader.

Response
We apologize for not identifying this paper in preprint.We have now changed the wording in the manuscript and included this reference in the paper.We have rewritten the discussion section to include this paper.
5. A possible role for 3 of the 4 proteins in the diagnostic signature in MIS-C pathogenesis is not mentioned and needs added to the Discussion.

Response
Thank you for this suggestion.We have added roles for all four proteins in the discussion section.

Minor points:
1. Table 1 provides basic clinical information.This could be improved by including data on oxygen requirements, PICU and inotropes and also clinical features including GI involvement, cardiac, shock and comorbidities for the MIS-C cases and febrile controls.The authors comment on the bacteria and sepsis patients in Table 1 but don't mention those with viral infection.

Response
Thank you for these suggestions.We have now revised Table 1 and provided additional data, including PICU, inotropes, and organ manifestations.In the table and results section, we have added details on patients with viral infections.
2. Line 108: '66 proteins related to therapy with intravenous immunoglobulin were excluded'.How were these proteins identified?This should be referenced.

Response
Immunoglobulin proteins excluded were proteins from the heavy-chains, light-chains, j-chains and variable regions.This is now stated in the manuscript in the method section

Response
We apologize for not being clear on the matter.In the initial proteomics analyses on the pathogenesis, MIS-C was compared to febrile controls, including viral and bacterial infections, as well as KD and severe sepsis.
We excluded KD and severe sepsis in the development of a diagnostic signature in order to use these two groups as validation cohorts.This has now been rewritten in the manuscript to make it clear for the reader.

Result section:
5. Lines 150-153: The function of 2/4 of the proteins in the signature is mentioned here.I would recommend to either include the role for all four or save this for the Discussion.

Response
Thank you for this suggestion.We have deleted the function in the results section and added roles for all four proteins in the discussion section.
6. Lines 207-208: Could this also mean these proteins are highly correlated with other proteins significant between MIS-C and febrile controls?

Response
We agree that a protein signature using different proteins could have been a possibility since a different support vector classification model with different proteins also achieved a high diagnostic accuracy of MIS-C.This supports that several proteins may be used for an MIS-C signature and emphasizes the robustness of proteomics as a diagnostic tool for MIS-C.

Results section:
Table 1 9. Lines 314-315: What about KD cases?Where they also included?They cannot be considered viral or bacterial.

Response
We apologize for not being clear on the matter.As also described above, in the initial proteomics analyses on the pathogenesis, MIS-C was compared to febrile controls, including viral and bacterial infections, as well as KD and septic shock.We excluded KD and septic shock in the development of a diagnostic signature in order to use these two groups as validation cohorts.This has now been rewritten in the manuscript to make it clear for the reader.

Result section:
Reviewer #3 (Remarks to the Author): The authors utilized AI-assisted proteomics to characterize a novel 4-protein signature and 7 protein clusters from children with MISC compared to those with non-MISC febrile illnesses (bacterial, viral, severe sepsis, or KD) enrolled prior to the start of the COVID-19 pandemic.The authors utilized a training and validation internal set of data to develop their model and then validated it using an external cohort from a proteomics database to identify four specific proteins-lymphocyte cytosolic protein 1, FcGRIIIa, alpha-aantichymotrypsin, and butyrylcholinesterase, which have the potential to be utilized for novel diagnostic tests.The study is important in that we still do not have a diagnostic test for MIS-C, nor do we know the exact pathogenesis.There is often overlap in the clinical definition between MIS-C and other hyperinflammatory syndromes such as Kawasaki Disease, Toxic shock syndrome, rheumatologic conditions, or severe COVID-19, making the diagnosis and thus optimal management of MISC challenging.One concern I have that the authors also highlight in their limitations is the choice for controls and lack of an external validation cohort.Routine bacterial and viral infections most often do not meet the clinical criteria for MIS-C, and thus having a biomarker to distinguish MISC from these entities may not be the most clinically useful.For example in Table 1, the viral cohort had a median CRP of 5 whereas those with acute MIS-C had a median CRP of 219.Instead, having a biomarker to distinguish MISC from entities with similar clinical presentations, such as severe COVID-19 or Kawasaki disease (which had a median prediction probability of 55.8% in the validation of the 4-protein signature), would in fact be clinically meaningful, particularly as the treatments vary and are time sensitive.Furthermore, it would certainly strengthen the study and increase generalizability of results if the authors are able to validate their 4-protein diagnostic signature in an external, blinded cohort, as a limitation of large amounts of data can be a bias towards significance.

Response
Thank you very much for reviewing this paper.We agree that it would have improved the paper significantly if we could have validated the 4-protein signature in an external, blinded cohort.However, we were not able to identify any studies, which have performed unbiased mass-spectrometry proteomics or any studies who have investigated the four proteins in our signature.Further, the decline in the incidence of MIS-C has refrained us from validating the accuracy of the 4-protein signature in a prospective Danish MIS-C cohort.
Nevertheless, the results are interesting and advance our knowledge of MIS-C.Few minor comments: There have been a handful of papers published on the use of proteomics in MIS-C (https://doi.org/10.1172/JCI151520,https://doi.org/10.1038/s41467-021-27544-6)which the authors may want to include.In the JCI paper by Porritt RA et al, authors utilized proteomics to distinguish severe MIS-C from mild disease and KD, and similarly found upregulation in FcGRIIIa and a reduction in proteins involved with lipid metabolic processes, lipoprotein clearance, and components of the coagulation cascade.

Response
Thank you for these relevant papers, including the paper finding an upregulation in FcGRIIIa.They are now added to the manuscript in the discussion section Could the authors provide a reference for the external US validation cohort of MIS-C and controls patients?

Response
Yes, we excuse.This reference has now been added to the manuscript REVIEWERS' COMMENTS: Reviewer #1 (Remarks to the Author): I have read the revised manuscript and am satisfied with the responses you provided, the additional text and corrections you have made to the manuscript and to the clinical table.I recommend that this paper can now be accepted.
Reviewer #3 (Remarks to the Author): The authors have made adequate effort at addressing the reviewers concerns and comments within the limits of their cohort.

3.
Lines 110-11: Typo as this related to Figure 2 A/B.The correlation coefficient not stated or a p-value in relation to the 105 significant proteins (Line 112).Responses Thank you for identifying these typos.They have been corrected.The p-values are provided in sTable2.Correlation coefficients between each sample (94 samples x 94 samples) are illustrated in 2B and colored according to their Pearson correlations (from dark red to white corresponding to Pearson correlations from 0.75 to 1) 4. Lines 145-146: What about the KD patients?Only viral and bacterial patients are mentioned.

7.
Lines 223-224: 'We found profound changes in lipid metabolism, previously unreported in MIS-C.'I refer the author to these papers: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8624489/https://www.frontiersin.org/articles/10.3389/fcvm.2022.1033660/fullResponse Thank you for this suggestion.We have added these papers in the discussion section.8. Lines 269-271: How were the bacterial and viral infections defined/ phenotyped?Any deaths in the MIS-C or febrile control groups?Response Details of the controls with bacterial and viral infection are now added to the manuscript.