Altered bioenergetics and mitochondrial dysfunction of monocytes in patients with COVID‐19 pneumonia

Abstract In patients infected by SARS‐CoV‐2 who experience an exaggerated inflammation leading to pneumonia, monocytes likely play a major role but have received poor attention. Thus, we analyzed peripheral blood monocytes from patients with COVID‐19 pneumonia and found that these cells show signs of altered bioenergetics and mitochondrial dysfunction, had a reduced basal and maximal respiration, reduced spare respiratory capacity, and decreased proton leak. Basal extracellular acidification rate was also diminished, suggesting reduced capability to perform aerobic glycolysis. Although COVID‐19 monocytes had a reduced ability to perform oxidative burst, they were still capable of producing TNF and IFN‐γ in vitro. A significantly high amount of monocytes had depolarized mitochondria and abnormal mitochondrial ultrastructure. A redistribution of monocyte subsets, with a significant expansion of intermediate/pro‐inflammatory cells, and high amounts of immature monocytes were found, along with a concomitant compression of classical monocytes, and an increased expression of inhibitory checkpoints like PD‐1/PD‐L1. High plasma levels of several inflammatory cytokines and chemokines, including GM‐CSF, IL‐18, CCL2, CXCL10, and osteopontin, finally confirm the importance of monocytes in COVID‐19 immunopathogenesis.

(Note: With the exception of the correction of typographical or spelling errors that could be a source of ambiguity, letters and reports are not edited. Depending on transfer agreements, referee reports obtained elsewhere may or may not be included in this compilation. Referee reports are anonymous unless the Referee chooses to sign their reports.) In their manuscript "Metabolic exhaustion of monocytes in COVID-19 patients" the authors address COVID-19 associated changes in monocyte subset distribution, monocyte phenotype and metabolism. With their study the authors aim at elucidating mechanisms of hyper-inflammation observed in COVID-19 that are associated to monocytes. As hyper-inflammation is a complication of COVID-19 associated with severe disease and at the same time displays an attractive target for therapeutic intervention on different levels, a detailed understanding of the underlying mechanisms is urgently needed. While this is an important clinical aspect with implications also e.g. for treatment with immune checkpoint inhibitors, only very little is known about modulation of PD-1 and PDL-1 during SARS-CoV-2 infection. Therefore the Topic of the manuscript is of high novelty and potential Impact. In their study, the authors analyzed monocyte metabolic parameters and monocyte subset distribution as well as monocyte expression of PD-1, PDL-1 and TIM-3 in COVID-19 patients and healthy controls. While novelty, medical impact and adequacy of the system are rated high, technical quality is rated medium due to lack of information on the status of the SARS-CoV-2 infection of the patients at the time-point of sampling, i.e. in which phase of COVID-19 they were, and lack of information which of the 15 patients were analyzed for monocyte subsets and marker expression (see specific comments to the authors below).

Referee #2 (Remarks for Author):
Major comments to be addressed during revision: 1) The authors analyzed 15 COVID-19 patients and 12 healthy controls which correspond to the data shown in figure 1. Scatter plots in figure 3 and 4 seem to show fewer data points and information on which patients were analyzed and why not all patients were analyzed is lacking. 2) While longitudinal analyses for more time-points would be optimal but are not always feasible or possible, more information on the patients' diseas and inflammatory status at the time-point of analysis would facilitate integration and interpretation of the results. Relevant information would be whether these patients were altogether in a similar stage of the disease, what time-point from diagnosis/onset of the symptoms/positive test the samples were taken or whether they had already developed a specific adaptive immune/antibody response.
3) The interpretation of the clinical relevance of the presented findings would be greatly increased, if there are correlations to inflammatory parameters. If the authors have analyzed, or have a chance to analyze systemic cytokine levels such as for IL-6 for the time-point of monocyte analysis, this should be shown or done. As is stated also by the authors, it is a limitation that cytokine production by the monocytes themselves could not be analyzed -this is the case especially as monocyte metabolic exhaustion seems somewhat counter-intuitive in hyper-inflammation. Also therefore, any correlation to the disease and inflammatory status of the patients at the time-point of analysis would be very interesting and helpful. 4) According to figure 2, total monocytes and monocyte subsets were analyzed for the expression of PDL-1, TIM-3, CXCR3, CCR2 and CD38. These data are however shown only for PD-1 and PDL-1 and described for TIM-3. The description of the TIM-3 data should include "data not shown" while CXCR3, CCR2 and CD38 analyses need not to be mentioned in the text and shown in the gating strategy in figure 2, if the results of these analyses are not shown, described or discussed later in the manuscript or supplement.
Minor comments to be addressed during revision: 1) Organization of multi-panel figures into A, B etc. would facilitate reading and finding the relevant (sub)-figure.
2) Please check scatter-plots in figure 1 as Max Resp maintenance is listed in the figure legend but Basal ECAR is shown (and not listed in the legend). 3) Are those few patients showing substantially higher values in the bioenergetics profile (figure 1) always the same patients and does this correlate with any other parameter of the disease (along major comments 2) and 3))? 4) In the gating strategy (figure 2) the description is a bit misleading regarding CD15/CD11b gating. The text states "Then, cells positive for HLA-DR and CD15 (i.e., leukocytes that did not include lymphocytes, that are CD15-) were selected" and CD11b is not referred to, while in the corresponding plot the y-axis is labelled CD15/CD11b and there does not seem to be a selection for the CD15/CD11b-positive population but rather all HLA-DR positive cells are gated.
General comments (not necessarily need to be addressed for publication but of general interest): 1) As this manuscript clearly focusses in monocytes, also T cell PD1/PD-L1 expression would be of high interest.
2) It would be interesting to comment or speculate on the relevance of these results on a treatment with check-point inhibitors in cancer patients (during concomitant COVID-19).

Referee #3 (Remarks for Author):
This is a very interesting manuscript which characterizes functional changes in peripheral blood monocytes from patients affected by Covid 19 pneumonia. The Authors identified a redistribution among different monocytes classes accompanied by a significant expansion of intermediate/proinflammatory cells and an increased expression of inhibitory checkpoints, including PD-1/PD-L1. It turns out that the metabolic and functional exhaustion in monocytes from patients might alter their capability to rapidly clear the infection. The manuscript is concise, well written and unveils novel aspects of Covid 19 infection, not characterized before. I suggest that some points are addressed before publication, including experiments, if the limited material from patient samples allows to perform them: Major points 1. In Fig. 1, left panel with representative traces, it is not clear to me if cells from Covid patients contain less mitochondria than control monocytes or if their mitochondria are simply less functional. A difference in number of mitochondria might explain a different basal and maximal respiration. Cardiolipin staining with Nonyl Acridine Orange (NAO) fluorescent mitochondrial dye whose staining is independent of mitochondrial membrane potential, might help to characterize the mechanism causing lower respiration in cells derived from Covid patients. 2. Western blotting of respiratory complex subunits, ATP synthase and its inhibitor protein IF1 should be performed, if lysates from patient cells are available. On one hand OXPHOS detection will provide a possible explanation for the functional results presented in Fig. 1, on the other hand, the IF1 levels would be particularly interesting given the emerging role of this protein in the modulation of ATP synthase in pathological condition and inflammation (García-Aguilar A and Cuezva JM, Minor points 1. I suggest to the Authors to briefly mention in the introduction the connection between monocyte ATP level (bioenergetic performance/glycolysis) and their role in the infection clearance. This would help the less specialized audience to follow the importance of the presented data and understand the relevance of differences in the respiratory behavior of patient cells. 2. In Supplementary Table 1, a range of values in normal condition (healthy donors) for bilirubin, LDH, and other parameters (that are reported in the text) should be indicated on the right in the same table, for clarity. 3. The "oxidative burst" which is stimulated by ionomycin treatment should be better explained in the text. This treatment may cause multiple effects such as Ca2+ uptake in mitochondria, stimulation of the respiratory chain, depolarization of the mitochondrial membranes, ... The Author should better explain which of these effects are in their opinion emerging in their experimental conditions and help the reader to understand the importance of differences between patient and control monocytes.

Dear Dr Carret,
We are sorry to disturb you. A few days ago we submitted a manuscript entitled "Metabolic exhaustion of monocytes in COVID-19 patients" to be considered for publication in EMBO Molecular Medicine.
The manuscript was rejected after the evaluation by three reviewers. The overall evaluation of two of them (referee #2 and #3) was fairly good as novelty, medical impact and adequacy of the system were rated high. The main general weaknesses were the lack of due to lack of information on the status of COVID-19 disease of the patients at the time-point of sampling and the characterization of monocytes by flow cytometry. These comments are actually quite easy to address.
We are writing you because we present a rebuttal, as we do not agree with criticism made by referee #1, who made questionable comments and seems not so competent in the field of flow cytometry. This is quite disappointing, as a relevant part of the manuscript is based on data obtained with this technique.
We are thus sending a file, where you can find our point-to-point answers to the observations of referee #1, together with answers to referees #2 and #3.
We have partially re-thought the structure of the manuscript, which could obviously include the requests of the referee #2 and #3, and also new data such as the unsupervised analysis of flow cytometry data (showing that our gating strategy was not wrong at all, as stated by referee #1), electron micrographs showing ultrastructural alterations in mitochondria, and in vitro stimulation of monocytes (that produce TNF and IFN-gamma, in spite of their clear alterations).
For these reasons, we do hope you can accept our comments about the revision and our rebuttal, and can consider again our work for EMBO Molecular Medicine. We thank you in advance for your time.
Looking forward to hearing from you, Thank you for asking us to reconsider our decision on this article and apologies for only getting back to you today. I decided to seek advice on your article and consulted with our chief editor and an expert in the field.
I am happy to convey that we would like to invit e revision of the paper according to your rebut tal let ter.
Our advisors say "est ablishing a causat ive link is impossible in pat ient s, hence the best would be to consolidat e the immunological "environment " of these monocyt es and furt her approach the heart of the process by using monocyt es purified from alveolar lavages [...]". We would therefore st rongly encourage you to make the st udy st ronger by providing addit ional dat a as much as you possibly can and re-analysis as recommended.
Please keep in mind that the referees' concerns must be fully addressed and that accept ance of the manuscript would ent ail a second round of review.
EMBO Molecular Medicine has a "scooping prot ect ion" policy, whereby similar findings that are published by ot hers during review or revision are not a crit erion for reject ion. Should you decide to submit a revised version, I do ask that you get in touch aft er three mont hs if you have not complet ed it , to updat e us on the st at us.
Please also cont act us as soon as possible if similar work is published elsewhere. If ot her work is published we may not be able to ext end the revision period beyond three mont hs.
Please read below for import ant edit orial format ting and consult our aut hor's guidelines for proper format ting of your revised art icle for EMBO Molecular Medicine.
I look forward to receiving your revised manuscript .
Dear Dr. Carret, Thank you for the message and the information. There one unsolvable problem with the note of your advisors. According to international guidelines, BAL can be done only for clarifying a clinical suspect of COVID pneumonia in a patient with a negative swab. In fact, BAL poses at high risk the operators. In our clinics, only 3 BAL have been done in 3 months, in a population of more than 300 patients. Fortunately, in the last weeks we have seen just a couple of pneumonia. So, studying BAL now is almost impossible for us. Thus, I am asking you if the paper will be considered even without data from BAL. Thank you in advance for your consideration. Kind regards, Andrea Cossarizza and Lara Gibellini.
Dear Drs. Cossarizza and Gibellini, I apologize for the brevity of my letter earlier today. Our advisor indeed mentioned BAL to start with but added that it would be very difficult if not impossible to perform given the circumstances. I'm sorry I should have added this bit to the letter.
So yes, to answer your question, we would still consider the paper even if you don't perform BAL. Looking forward to receiving the revision. We are well aware that even following good rules, setting gates is in any case operator dependent and someone considers this an "art". For these reasons, we applied an unsupervised approach to analyze monocytes subpopulations in CTR and COVID-19 patients. Multiparametric unsupervised method, such as the combination of different algorithms of automatic clustering (FlowSOM), dimension reduction representation (UMAP) and statistical methods for differential discovery analyses in high-dimensional cytometry data (diffcyt) are mostly automatic and operator independent. A new Figure  Moreover, CD14-APC mAb has been titrated (like all the other antibodies) and the separating and saturating titer was always used. Thus, the data reported in the paper has been analyzed according to reviewer's suggestions.
Concerning the use of isotype controls we do not agree with the comments for several reasons, that have been discussed several times in the last 15-20 years, especially in the Purdue List (see http://www.cyto.purdue.edu/hmarchiv/index.htm), where the most renown experts of flow cytometry share basic and most complex notions regarding flow cytometry technique and data analyses (both fluorescent and mass cytometry). The main limitations and uselessness of isotype controls are briefly reported here below: 1) Individual antibody conjugates have various levels of background staining, depending upon their specificity, concentration, degree of aggregation and fluorophore:antibody ratio. Hence the hit-ormiss prospect to find an isotype that truly matches the background staining of a particular test antibody. Isotype controls do not define the true level of background staining and the use of these creates a circular position.
2) Isotype controls do not account for fluorescence spillover from other channels. Only for this, the are in fact useless when several fluorescences are simultaneously used.
3) Much more information is contained in the exhaustive note that was written by Dr. Mario Roederer from NIH (universally recognized as the top expert in this field) in 1998, regarding the use of isotype controls (https://lists.purdue.edu/pipermail/cytometry/1998-April/009860.html).
The appropriate figure is now reported in the paper: the MFI shift between CTR and COVID samples clearly shows that there is an increased expression of PD-L1 in COVID monocytes.
The metabolic activity was evaluated on total monocytes because our primary endpoint was to find metabolic impairment between monocytes from COVID patients and healthy donors. We discovered that monocytes from COVID patients display low oxidative burst if compared to healthy donors. For this reason, in order to ascertain if this difference could be due to a different monocyte distribution, monocyte phenotype was evaluated. We tried to sort intermediate monocyte from COVID patients, but the number of cells obtained was not sufficient to perform further metabolic analyses.  Figure 6 and new data in Figure 3B. The authors claim that the monocytes have less ability to maintain maximal respiration; however, the maximal respiration is never the same, and the loss of looks similar to controls. This claim needs to be modified, or appropriate comparisons supporting this claim should be presented.

Authors' answer:
The maintenance of maximal respiration was analyzed by calculating the area under the curve of the OCR trace from the sixth to eleventh measurement, and is different between CTR and COVID-19, as also the maximal respiration is different. However, in order to strengthen this datum, we added the analysis of the spare respiratory capacity. A scatter plot reporting the quantification of the spare respiratory capacity is now present in Figure 1A. Claiming that classical monocytes "only tended to decrease" is not supported by statistical analysis and therefore the claim needs to reflect that.
Authors' answer: We claimed that "classical monocytes only tended to decrease" as we did not observe a statistical difference between the percentage of classical monocytes in CTR vs COVID-19 patients even if there was a trend and the means were slightly different among the two groups. However, as reported earlier in this point-to-point response, we applied an unsupervised approach (operator-independent approach) to analyze monocytes subpopulations in CTR and COVID-19 patients. According to this analysis classical monocytes are decreased in COVID-19 patients vs CTR.

Authors' answer: The citation was Xia et al., Immune Checkpoint Receptors Tim-3 and PD-1 Regulate
Monocyte and T Lymphocyte Function in Septic Patients (2018). However, due to a rearrangement of the introduction, the sentence was modified.

In their manuscript "Metabolic exhaustion of monocytes in COVID-19 patients" the authors address COVID-19 associated changes in monocyte subset distribution, monocyte phenotype and metabolism. With their study the authors aim at elucidating mechanisms of hyper-inflammation observed in COVID-19 that are associated to monocytes. As hyper-inflammation is a complication of COVID-19 associated with severe disease and at the same time displays an attractive target for therapeutic intervention on different levels, a detailed understanding of the underlying mechanisms is urgently needed. While this is an important clinical aspect with implications also e.g. for treatment with immune checkpoint inhibitors, only very little is known about modulation of PD-1 and PDL-1 during SARS-CoV-2 infection. Therefore the Topic of the manuscript
is of high novelty and potential Impact.

Referee #2 (Remarks for Author):
Major comments to be addressed during revision: 1) The authors analyzed 15 COVID-19 patients and 12 healthy controls which correspond to the data shown in figure 1. Scatter plots in figure 3 and 4 seem to show fewer data points and information on which patients were analyzed and why not all patients were analyzed is lacking.

Authors' answer:
We agree with the reviewer with this comment. As explained for referee #1, our primary endpoint was to investigate the bioenergetics of monocytes isolated from COVID-19 patients and healthy donors. We postulated that this impairment could be due to different monocyte distribution, so in a smaller cohort of patients, we analyzed monocyte phenotype.

Authors' answer:
We thank the review for the comments. As discussed before, we have reanalyzed data applying an unsupervised approach to detect monocyte subpopulations in CTR and COVID-19 patients.
Multiparametric unsupervised method, such as the combination of different algorithms of automatic clustering (FlowSOM), dimension reduction representation (UMAP) and statistical methods for differential discovery analyses in high-dimensional cytometry data (diffcyt) are mostly automatic and operator independent. A new Figure (  Authors' answer: As suggested by the reviewer, we organized the figures with multi-panels.

2) Please check scatter-plots in figure 1 as Max Resp maintenance is listed in the figure legend but Basal ECAR is shown (and not listed in the legend).
Authors' answer: We checked and amended.

Authors' answer:
We apologize with the reviewer. This was a mistake, thus we amended the text as follows "Then, cells positive for HLA-DR, CD15 and CD11b were selected and monocytes expressing CD14 were identified." General comments (not necessarily need to be addressed for publication but of general interest):

1) As this manuscript clearly focusses in monocytes, also T cell PD1/PD-L1 expression would be of high interest.
Authors' answer: We totally agree with the reviewer. Together with monocytes, T cell exhaustion is also very interesting. We recently published that COVID-19 patients' T cell compartment displays several alterations involving naïve, central memory, effector memory and terminally differentiated cells, as well as

regulatory T cells and PD1+CD57+ exhausted T cells (De Biasi et al, Nature Communications, 2020).
2) It would be interesting to comment or speculate on the relevance of these results on a treatment with check-point inhibitors in cancer patients (during concomitant .

Authors' answer:
This suggestion is very important and we thank the reviewer for this. We Nonyl Acridine Orange (NAO) fluorescent mitochondrial dye whose staining is independent of mitochondrial membrane potential, might help to characterize the mechanism causing lower respiration in cells derived from Covid patients.

Authors' answer:
We thank the reviewer for this remark. We quantified mitochondrial mass by using cells mitotracker green, which stains mitochondria regardless of mitochondrial membrane potential. Data have been added in the manuscript and in Supplementary Figure 1A. Information regarding mitochondrial mass were obtained also from transmission electron microscopy and were reported in Figure 1B

Authors' answer:
We are aware that western blotting of ETC complexes and IF1 could complete data regarding OXPHOS, thus improving the quality of the manuscript. We thank the reviewer for this comment.
However, unfortunately, to obtain protein lysates from monocytes from a clinical sample is quite challenging as the number of cells required to have even a few mg of proteins is high. Monocytes are indeed 10-20% of peripheral blood mononuclear cells (PBMCs). We usually obtained twenty mL of peripheral blood from patients and controls. From healthy blood, PBMCs yield ranges between 0.5-3 million cells/mL (10-60 million cells in 20 mL). Most COVID-19 patients has leukopenia, thus they had a severe reduction in the number of leukocytes, including monocytes. Therefore, the number of cells was insufficient to get protein lysates to perform western blotting. ligand Pam3CysSK4, increased OCR and glycolysis, which are essential for activation of host defence mechanisms such as cytokine production and phagocytosis. Functional and metabolic re-programming has been also observed in monocytes during sepsis in a process mediated by hypoxia-inducible factor-1 (HIF1) (Shalova et al, 2015)."  Thank you for the submission of your revised manuscript to EMBO Molecular Medicine. We have now received the enclosed reports from the referees that were asked to re-assess it. As you will see the reviewers are now globally supportive and I am pleased to inform you that we will be able to accept your manuscript pending the following final amendments: ***** Reviewer's comment s ***** Referee #2 (Comment s on Novelt y/Model Syst em for Aut hor):

In Supplementary
The comment s raised in the first round of review have been largely adressed and the manuscript has gained from the revisions and the newly included data and points discussed. The title of the revised manuscript better refers to the data shown. The study remains descriptive, which however should not be a problem at this stage, taking into account limited availability of material and novelty of the topic.
Referee #2 (Remarks for Author): There are minor comments to be addressed: #1: in the text, page 4, correct spelling of E. coli, S. aureus and M. tuberculosis #2: readability of figure 1B would be improved if subpanels a, b, c, d and e were labelled CTR or COVID within the figure. #3 (this is the most important point): from the materials and methods section it does not become clear, which patients within the cohort of healthy controls and COVID-19 patients went into which analyses. It makes a large difference, whether e.g. plasma cytokine levels are from the same patients in which also monocyte subsets were analyzed or whether all of the analyses (plasma cytokines, monocyte subsets, immature monocytes, in vitro cytokine production, bioenergetic profile) were performed in separate sets of patients. This should be specified. #4: please specify in the materials and methods section out of how many patient (and control) samples the 225 mitochondria analysed for each group were derived.

B-Statistics and general methods
the assay(s) and method(s) used to carry out the reported observations and measurements an explicit mention of the biological and chemical entity(ies) that are being measured. an explicit mention of the biological and chemical entity(ies) that are altered/varied/perturbed in a controlled manner. a statement of how many times the experiment shown was independently replicated in the laboratory.
Any descriptions too long for the figure legend should be included in the methods section and/or with the source data.
In the pink boxes below, please ensure that the answers to the following questions are reported in the manuscript itself. Every question should be answered. If the question is not relevant to your research, please write NA (non applicable). We encourage you to include a specific subsection in the methods section for statistics, reagents, animal models and human subjects.

definitions of statistical methods and measures:
a description of the sample collection allowing the reader to understand whether the samples represent technical or biological replicates (including how many animals, litters, cultures, etc.).

The data shown in figures should satisfy the following conditions:
Source Data should be included to report the data underlying graphs. Please follow the guidelines set out in the author ship guidelines on Data Presentation.
Please fill out these boxes ê (Do not worry if you cannot see all your text once you press return) a specification of the experimental system investigated (eg cell line, species name).
We studied all the patients that were recruitable during and after the lock-down period. The number of patients per set of experiments was chosen using the software available on line for clinical and lab studies, according to the probability to reach a statistical significance when compared to controls. graphs include clearly labeled error bars for independent experiments and sample sizes. Unless justified, error bars should not be shown for technical replicates. if n< 5, the individual data points from each experiment should be plotted and any statistical test employed should be justified the exact sample size (n) for each experimental group/condition, given as a number, not a range; Each figure caption should contain the following information, for each panel where they are relevant:

NA
We enrolled 28 patinets with pneumonia caused by COVID-19 who had been admitted to the Infectious Disease Clinics of the University Hospital in Modena, North Italy. Age range 37-89, 69% were males, all caucasians. They were compared with 27 age-and sex-martched healthy donors. Blood samples were collected at the Infetious Disease CLinics, brought to the Immunology Lab (500 meters away) and immediately processed.
There was no randomization, being a study that compares patients with COVID pneumonia vs. age-and sex-matched healthy donors

Data
the data were obtained and processed according to the field's best practice and are presented to reflect the results of the experiments in an accurate and unbiased manner. figure panels include only data points, measurements or observations that can be compared to each other in a scientifically meaningful way.