Received: 8 November 2021 Revised: 20 December 2021 Accepted: 29 December 2021 DOI: 10.1002/ctm2.704 LETTER TO EDITOR Fetuin-A, inter-α-trypsin inhibitor, glutamic acid and ChoE (18:0) are key biomarkers in a panel distinguishing mild from critical coronavirus disease 2019 outcomes Dear Editor, pathophysiology and unveil new predictive biomarkers to The mechanistic pathways leading to immune dysreg- promptly ascertain patient’s outcomes. ulation and complications driven by uncontrolled severe Our COVID-19 study cohort included 273 SARS-CoV- acute respiratory syndrome coronavirus-2 (SARS-CoV-2) 2 infected individuals recruited during the first wave infection remain major challenges.1,2 Hence, a detailed (March–April 2020) in three different hospitals and analysis of the proteome, metabolome and lipidome pro- grouped by the disease severity following the medical file of coronavirus disease 2019 (COVID-19) patients show- inclusion criteria3 in mild, severe or critical (Figure 1A), ing different severity grades might shed light on the disease from whom demographic, preexisting clinical conditions F I G U R E 1 Study design and clinical characterization of the study cohort. (A) Flowchart of the clinical strategy followed to categorize patients of the coronavirus disease-2019 (COVID-19) study cohort. (B) Incidence of comorbidities (a), COVID-19 symptoms (b), medication (c), and oxygen & intensive care (d), grouped by disease severity as mild, severe and critical patients. The size of bars (a, c) and circular (b, d) portions is proportional to the percentage of the corresponding comorbidity, symptom, medication or treatment. While patients with mild disease presented mostly anosmia, were treated with antibiotics and did not require oxygen supply, the incidence of dyspnea was significantly higher in the severe and critical groups, many of the latter requiring corticosteroids, hydroxychloroquine, lopinavir/ritonavir. Low-flow oxygen therapies were mainly necessary for severe patients, some of who required high-flow oxygen administration and, a high proportion of critical patients were intubated and required vasopressor administration or dialysis. Please note that COVID-19-related medication (b) was dispensed after blood sample collection, so that is assumed the subsequent analysis are not biased due to exposure to medication at the time of blood collection This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. © 2022 The Authors. Clinical and Translational Medicine published by John Wiley & Sons Australia, Ltd on behalf of Shanghai Institute of Clinical Bioinformatics Clin. Transl. Med. 2022;12:e704. wileyonlinelibrary.com/journal/ctm2 1 of 7 https://doi.org/10.1002/ctm2.704 2 of 7 LETTER TO EDITOR F I G U R E 2 Serum proteomics profile of coronavirus disease-2019 (COVID-19) study cohort. (A) Heatmap showing significant proteins increasing or decreasing in accordance with disease severity. Columns correspond to the degree of disease severity: mild (left), severe (centre) and critical (right) groups. Mean values for each compound in each coronavirus disease-2019 (COVID-19) group (columns) are colour-coded based on relative abundance, low (red) & high (green). Among the 65 significant proteins, 33 increased and 32 decreased with disease severity. (B) Up-(upper) and down-(bottom) regulated protein networks sorted by gene-name showing a tight interconnection within the up- and down-regulated proteins. (C) Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis sorted by pathway impact and -log10 (p). The interconnected up- and down-regulated genes were enriched in 77 pathways, and the 10 relevant pathways whose impact values were greater than 0.1 (p < .05) were further considered. The bubble diagram shows matched pathways according to the p-values and pathway impact values. The size of bubbles shows pathway impact value and the colour denote the level of significance by means of p-values. Numbers in circles correspond to the significantly enriched pathways ordered from the highest to the lowest pathway impact value. The number of matched proteins, impact value and p-value corresponding to each pathway are indicated on the inserted table. (D) Random forest analysis showing the 15 protein-encoding genes ranked by classification accuracy to distinguish between a mild and critical group of patients. Squares on the right represent the COVID-19 (1 = mild, 2 = severe and 3 = critical) and the colours depict the accuracy power (red and blue for high and low accuracy, respectively) and COVID-19 treatments are summarized in Table S1. COVID-19 study cohort. The proteomics analysis identi- The greatest significant differences were observed between fied 65 proteins with a significant increase or decrease in mild and critically ill patients. These findings indicated abundance according to the disease severity (Figure 2A), that older individuals with comorbidities such as hyperten- which resulted to be highly interconnected (Figure 2B). sion, obesity, diabetes and cardiovascular disorders, mostly Hence, the complement and coagulation cascades were presenting dyspnea (Figure 1B), may be at higher risk of markedly the most significantly enriched pathways related suffering from severe respiratory distress with subsequent to COVID-19 severity (Figure 2C). Other protein-coding oxygen and drug requirements and, eventually, died. Simi- genes such as carboxypeptidases, protease inhibitors, acute larly, the serum biochemical composition analysis revealed phase proteins, extracellular matrix stabilizers and antimi- a well-differentiated blood pattern previously defined for crobial enzymes, were also significantly up-regulated in critically ill patients (Figure S1). severely and critically ill patients. These results showed the In light of the promising results already provided by essential contribution of these proteins in the coagulopa- omic technologies in the search for predictive biomark- thy phenomenon and hyperinflammatory state that sub- ers of COVID-19 severity,4,5 we conducted a nontar- sequently enhances SARS-CoV-2 endocytosis and infectiv- geted multi-omic, including proteomic, metabolomic and ity and promotes secondary bacterial infections, previously lipidomic analyses, in the serum from patients of the described as aggravators of severe and critical COVID-19 LETTER TO EDITOR 3 of 7 F I G U R E 3 Metabolomic and lipidomic profile of coronavirus disease-2019 (COVID-19) patients grouped by disease severity. (A) Heatmap plots significant relative abundance of metabolites & lipids increasing or decreasing in accordance with disease severity. Significant differences (p-values < .05) were determined by ANOVA test followed by post-hoc Bonferroni correction for mean relative abundance between mild, severe and critical COVID-19 groups of patients. Columns correspond to the degree of disease severity: mild (left), severe (centre) and critical (right) groups. Mean values for each compound in each COVID-19 group (columns) are colour-coded based on relative abundance, low (red) & high (green). (B) Metabolomic and lipidomic Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis and related blood diseases in COVID-19 patient cohort. Functional metabolic enrichment pathway of 15 metabolites selected from random forest modelling (Fisher’s exact test. p < .05) (left), and the corresponding enriched blood pathway diseases (right). (C) Random forest modelling of significant metabolites and lipids with the highest classification accuracy. Right-legend indicates the capacity of the compounds to differentiate groups of severity, blue (low) and red (high) cases.6 Proteins with reduced abundance in critically ill showed a significant increase in glucose and glutamic patients with COVID-19 were mostly associated with lipid acid (GA) levels but a reduction in glutamine, citrate transport (apolipoproteins), which dysfunction seems to and uric acid levels, suggesting mitochondrial dysfunc- increase SARS-CoV-2 infectivity in patients with COVID- tion, an enhanced glutaminolysis and a shift from anaer- 19.7 For the first time, fetuin-A (AHSG) and inter-α-trypsin obic to aerobic glycolysis (Warburg effect). Accordingly, inhibitor 3 (ITIH3) were determined as the most accurate D-glutamine and D-glutamate metabolism were the most biomarkers (random forest) of the critical clinical progres- significantly enriched pathways (Figure 3B, left panel), sion of COVID-19 (Figure 2D). and were significantly related to seizures disorders, anoxia, The metabolomic and lipidomic analyses revealed 34 heart failure, diabetes, obesity and inflammatory diseases metabolites and 28 lipids that were significantly increased (Figure 3B, right panel). Lipid levels that increased with or decreased in relation to severity (Figure 3A). Inter- severity were mainly triglycerides (TGs) and diacylglyc- estingly, many of the altered metabolites were amino erols, and those that decreased were predominantly sphin- acids and sugars involved in central carbon metabolism. gomyelins (SMs), cholesteryl esters (ChoEs) and lysophos- In line with previous reports,8,9 critically ill patients phatidylcholines. Lipoproteins rich in TGs may trigger dys- 4 of 7 LETTER TO EDITOR F I G U R E 4 Evaluation of potential biomarkers to be a part of a panel to distinguish coronavirus disease-2019 (COVID-19) outcomes. (A) Heatmap showing the Spearman correlation coefficient of pairwise comparison between demographic, clinical and most enriched proteomic, metabolomic and lipidomic biomolecules determined in the blood of patients with mild (1), severe (2) and critical (3) COVID-19. Spearman matrices are colour-coded (-1:1, orange:purple through white) and correlations with p-values < .05 were considered statistically significant. (B) Receiver operating characteristic (ROC) curves analysis for the predictive power of top selected protein-encoding genes, lipids and metabolites in random forest analysis to differentiate patients with mild from those with a critical illness. (C) ROC curve analysis for the most enriched proteomic, metabolomic and lipidomic biomolecules to differentiate severe from critically-ill patients with COVID-19. (D) Binary logistic regression modelling analysis testing the accuracy of the four selected biomarkers to differentiate mild from critically ill patients with COVID-19 in a randomly selected set of patients function in innate immunity and impair the defence mech- of the variables analyzed, patients with mild illness (Fig- anism against COVID-1910 and a reduced abundance of ure 4A1) significantly differed from those of the severe and SMs and ChoEs may interfere in signal transduction and critical groups (Figure 4A2,3, respectively). In brief, sig- in key immune and cellular processes. Among them, GA nificant correlations were obtained across the omic data, and ChoE (18:0) resulted in the most powerful (random which were more intense between lipidomics than within forest) predictive biomarkers for COVID-19 evolution (Fig- the protein-encoding genes, and nearly negligible through ure 3C), confirmed by the prognosis accuracy determined metabolomics. The predictive power of the selected omics by the receiver operating characteristic (ROC) analysis biomolecules as biomarkers for the severe disease was sub- (Figure S2A–C, respectively). The highest accuracy was sequently demonstrated by the high accuracy, sensitivity attained when combining both compounds in the distinc- and specificity obtained by combining the four molecules tion of mild from critical illnesses (Figure 2SD). To pro- in the ROC analysis (Figure 4B) to effectively distinguish vide insights into the biological pathways related to the critical COVID-19 patients from patients with mild disease pathophysiology of the disease, we study the linkage and (area under the curve [AUC] = 0.994). To precisely pre- co-regulation between the distinct classes of biomolecules dict whether a patient will progress from severe to a life- by integrating the most significant demographical and threatening disease, not only the four but all the top-omic clinical data (Table S1 and Figure S1) and the top omic selected biomarkers need to be integrated into the ROC molecules determined above (Figure S3A,B) in Spearman analysis (AUC = 0.811; Figure 4C). correlation matrix analyses (Figure 4A1–3). Despite all Taking a step further, the inclusion of AHSG, ITIH, three groups showing a similar association pattern for most GA and ChoE (18:0) in a predictive biomarker panel for LETTER TO EDITOR 5 of 7 COVID-19 severity was validated in a randomly selected de Enfermedades Infecciosas-ISCIII [CB21/13/00020], subset of patients. The regression modelling analysis con- Madrid, Spain and Consejeria de Transformacion Eco- firmed the usefulness (classification accuracy >90%) of the nomica, Industria, Conocimiento y Universidades Junta biomarker panel in distinguishing mild to critical COVID- de Andalucía (research Project CV20-85418). Elena 19 outcomes (Figure 4D). Once more, all these findings Yeregui was supported by the Instituto de Salud Carlos III highlighted the complex interactions between certain bio- (ISCIII) under grant agreement “FI20/00118″ through the logical processes and the most serious complications aris- programme “Contratos Predoctorales de Formación en ing from SARS-CoV-2 infections and revealed their poten- Investigación en Salud”. Laia Reverté was supported by the tial as predictive biomarkers of disease severity. Instituto de Salud Carlos III (ISCIII) under grant agree- Limitations are the small sample size to perform sub- ment “CD20/00105″ through the programme “Contratos group analyses and the lack of a non-infected SARS-CoV-2 Sara Borrell”. Francesc Vidal was supported by grants group of subjects. However, this study was conducted in from the Programa de Intensificación de Investigadores a representative symptomatic well-characterized Spanish (INT20/00031)-ISCIII and by “Premi a la Trajectòria cohort to determine predictive biomarkers of COVID-19 Investigadora dels Hospitals de l’ICS 2018″. Anna Rull severity. was supported by a grant from IISPV through the project In conclusion, the multi-omic analysis identified new “2019/IISPV/05″ (Boosting Young Talent), by GeSIDA specific molecules related to complement and coagulation through the “III Premio para Jóvenes Investigadores cascades, platelet activation, cell adhesion, acute inflam- 2019″ and by the Instituto de Salud Carlos III (ISCIII) mation, energy production (Krebs cycle and Warburg under grant agreement “CP19/00146″ through the Miguel effect), amino acid catabolism and lipid transport as fin- Servet Program. Maria José Buzón was supported by gerprints of the acute disease. A novel biomarker panel the Miguel Servet Program (CP17/00179). Ezequiel Ruiz- consisting of AHSG, ITIH3, GA and ChoE (18:0) was pro- Mateos was supported by the Spanish Research Council posed for the accurate differentiation of mild from critical (CSIC). Alicia Gutiérrez-Valencia was supported by the COVID-19 outcomes. Instituto de Salud Carlos III, cofinanced by the European Development Regional Fund (“A way to achieve Europe”), AC K N OW L E D G E M E N T S Subprograma Miguel Servet (grant CP19/00159). This This study would not have been possible without the gen- project was also funded by a donation from the city erous collaboration of all the patients and their fami- Council of Perafort (to Teresa Auguet). lies and medical and nursing staff who have taken part in the project. We want to particularly acknowledge the CONFLICT OF INTEREST collaboration of the Departments of Preventive Medicine The authors declare that they have no conflict of interest. and Epidemiology, Internal Medicine, Critical Care, Emer- gency, Occupational Health, Laboratory Medicine and Laia Reverté1,2,3,# Molecular Biology, and BioBank-IISPV (B.0000853 and Elena Yeregui1,2,# B.0000854) integrated into the Spanish National Biobanks Montserrat Olona1,2,3,4 Platform (PT20/00197) and CERCA Programme (Generali- Alicia Gutiérrez-Valencia5,6 tat de Catalunya) and IISPV for their collaboration. We also Maria José Buzón7 thank Pol Herrero, Maria Guirro and Antoni del Pino from Anna Martí1,2 the Proteomics and Metabolomics facilities of the Cen- Frederic Gómez-Bertomeu1,4 tre for Omic Sciences (COS) Joint Unit of the Universitat Teresa Auguet1,2,4 Rovira i Virgili-Eurecat for their contribution to mass spec- Luis F. López-Cortés5,6 trometry analyses. Joaquin Burgos7 Clara Benavent-Bofill1 F U N D I N G I N F O R M AT I O N Carme Boqué1,4 This work has been developed in the framework of the Graciano García-Pardo1,2,4 COVIDOMICS’ project supported by Direcció General de Ezequiel Ruiz-Mateos5,6 Recerca i Innovació en Salut (DGRIS), Departament de Maria Teresa Mestre1 Salut, Generalitat de Catalunya (PoC-6-17 and PoC1-5). Francesc Vidal1,2,3,4 The research has also been funded by the Programa de Consuelo Viladés1,2,3,4 Suport als Grups de Recerca AGAUR (2017SGR948), the Joaquim Peraire1,2,3,4,## SPANISH AIDS Research Network [RD16/0025/0006, Anna Rull1,2,3,4,## RD16/0025/0007 and RD16/0025/0020]-ISCIII-FEDER COVIDOMICS Study Group* (Spain), the Centro de Investigación Biomédica en Red 6 of 7 LETTER TO EDITOR 1 Hospital Universitari de Tarragona Joan XXIII (HJ23), Laura Herrera-Hidalgo, Karin Neukam, Inmaculada Tarragona, Spain Rivas-Jeremias, Silvia Llaves, Montserrat Dominguez, 2 Institut Investigació Sanitària Pere Virgili (IISPV), Cesar Sotomayor, Cristina Roca-Oporto, Nuria Espinosa, Tarragona, Spain Carmen Infante-Domínguez, Juan Carlos Crespo-Rivas, 3 CIBER Enfermedades Infecciosas (CIBERINFEC), Abraham Saborido, Sara Bachiller, Sonsoles Instituto de Salud Carlos III, Madrid, Spain Salto-Alejandre, Judith Berastegui-Cabrera, Pedro 4 Universitat Rovira i Virgili (URV), Tarragona, Spain Camacho-Martínez, Carmen Infante Domínguez, Marta 5 Unit of Infectious Diseases, Microbiology and Preventive Carretero-Ledesma, Juan Carlos Crespo-Rivas, Eduardo Medicine, Virgen del Rocío University Hospital, Seville, Márquez, José Manuel Lomas, Claudio Bueno, Rosario Spain Amaya, José Antonio Lepe, Jerónimo Pachón, Elisa 6 Institute of Biomedicine of Seville (IBiS), Virgen del Rocío Cordero, Javier Sánchez-Céspedes, Manuela Aguilar University Hospital/CSIC/University of Seville, Seville, Guisado, Almudena Aguilera, Clara Aguilera, Teresa Spain Aldabo-Pallas, Verónica Alfaro-Lara, Cristina Amodeo, 7 Department of Infectious Disease, Hospital Universitari Javier Ampuero, María Dolores Avilés, Maribel Asensio, Vall d’Hebron, Institut de Recerca (VHIR)Universitat Bosco Barón-Franco, Lydia Barrera-Pulido, Rafael Autònoma de Barcelona, Barcelona, Spain Bellido-Alba, Máximo Bernabeu-Wittel, Candela Caballero-Eraso, Macarena Cabrera, Enrique Calderón, Correspondence Jesús Carbajal-Guerrero, Manuela Cid-Cumplido, Yael Francesc Vidal and Anna Rull, Hospital Universitari de Corcia Palomo, Juan Delgado, Antonio Domínguez-Petit, Tarragona Joan XXIII (HJ23), Tarragona, Spain. Alejandro Deniz, Reginal Dusseck-Brutus, Ana Email: fvidalmarsal.hj23.ics@gencat.cat and Escoresca-Ortega, Fátima Espinosa, Michelle Espinoza, anna.rull@iispv.cat Carmen Ferrándiz-Millón, Marta Ferrer, Teresa Ferrer, Ignacio Gallego-Texeira, Rosa Gámez-Mancera, Emilio # These authors contributed equally to this paper, and García, Horacio García-Delgado, Manuel should be considered as primary co-authors. García-Gutiérrez, María Luisa Gascón-Castillo, Aurora ## These authors contributed equally to this paper, and González-Estrada, Demetrio González, Carmen should both be considered as senior co-authors. Gómez-González, Rocío González-León, Carmen * The clinical centres and research groups which contribute Grande-Cabrerizo, Sonia Gutiérrez, Carlos to COVIDOMICS Study Group : Hernández-Quiles, Inmaculada Concepción Hospital Universitari Joan XXIII / IISPV / URV: Jenifer Herrera-Melero, Marta Herrero-Romero, Luis Jara, Carlos Masip, Verónica Alba, Montserrat Vargas, Laia Bertrán, Jiménez-Juan, Silvia Jiménez-Jorge, Mercedes Carmen Aguilar, Miguel López-Dupla, José Antonio Jiménez-Sánchez, Julia Lanseros-Tenllado, Carmina Porras, Sergi Veloso, Ajla Alibalic, David Riesco, Mónica López, Isabel López, Álvaro López Barrios, Rafael Real, Jessica Binetti, Judit Poblet, Mercé Sirisi, Gaspar Luque-Márquez, Daniel Macías-García, Guillermo Dalmau, Vanessa Gázquez, Esther Rodriguez, Antonia Martín-Gutiérrez, Luis Martín-Villén, José Molina, Garcia, Esther Picó, Cristina Gutiérrez, Gemma Aurora Morillo, María Dolores Navarro-Amuedo, Dolores Recio-Comí, Carla Martin-Grau, Teresa Sans, Carlos Nieto-Martín, Francisco Ortega, María Paniagua-García, Chiapella, Pilar Carbajo, Mireia Cramp, Carme Bes, Amelia Peña-Rodríguez, Esther Pérez, Manuel Poyato, Rosalia Bote, Nuria Alba, Blanca Rosich, Cristina Varillas, Julia Praena-Segovia, Rafaela Ríos, Jesús F. Rodríguez, Catherine Cabrejo, Rosaura Reig, Llorenç Mairal, Jesús María Jesús Rodríguez-Hernández, Santiago Esteve Ferrán, Neus Camañes, Angela Cortés and Rafael Rodríguez-Suárez, Ángel Rodríguez-Villodres, Nieves Gracia. Romero Rodríguez, Ricardo Ruiz, Zida Ruiz de Azua, Hospital Universitari Vall d’Hebrón / VHIR / UAB: Celia Salamanca, Sonia Sánchez, Víctor Manuel Meritxell Genescà, Anna Falcó, Cristina Kirkegard, Jordi Sánchez-Montagut, Alejandro Suárez Benjumea and Navarro, Vicente Descalzo, Paula Suanzes, Judith Grau, Javier Toral. Nerea Sanchez, Antonio Astorga and Vicenç Falcó. IBiS/Virgen del Rocío Hospital COVID-19 Working Team: ORCID José Miguel Cisneros, Luis E. López-Cortés, Alberto Alicia Gutiérrez-Valencia https://orcid.org/0000-0003- Pérez-Gómez, Joana Vitallé, Carmen Gasca-Capote, 3445-1574 María Trujillo-Rodriguez, Ana Serna-Gallego, Esperanza Maria José Buzón https://orcid.org/0000-0003-4427- Muñoz-Muela, María de los Reyes Jiménez-Leon, 9413 Mohamed Rafii-El-Idrissi Benhnia, Francisco Ostos, Ezequiel Ruiz-Mateos https://orcid.org/0000-0001-6747- 7813 LETTER TO EDITOR 7 of 7 Francesc Vidal https://orcid.org/0000-0002-6692-6186 6. Zhou F, Yu T, Du R, et al. Clinical course and risk factors for Anna Rull https://orcid.org/0000-0002-8907-7754 mortality of adult inpatients with COVID-19 in Wuhan, China: a retrospective cohort study. Lancet. 2020;395:1054-1062. https: //doi.org/10.1016/S0140-6736(20)30566-3. REFERENCES 7. Wang H, Yuan Z, Pavel MA, et al. The role of high cholesterol in 1. WHO. COVID-19 therapeutic trial synopsis. World Health Orga- age-related COVID19 lethality. BioRxiv. 2020. https://doi.org/10. nization. 2020;1-9 1101/2020.05.09.086249. 2. Domingo P, Mur I, Pomar V, et al. The four horsemen of 8. Wu D, Shu T, Yang X, et al. Plasma metabolomic and lipidomic a viral apocalypse: the pathogenesis of SARS-CoV-2 infection alterations associated with COVID-19. Nat Sci Rev. 2020;7:1157- (COVID-19). EBioMedicine. 2020;58:102887. https://doi.org/10. 1168. https://doi.org/10.1093/NSR/NWAA086. 1016/J.EBIOM.2020.102887. 9. Bharadwaj S, Singh M, Kirtipal N, Kang SG. SARS-CoV-2 and 3. Wang GQ, Zhao L, Wang X, Jiao YM, Wang FS. Diagnosis and Glutamine: SARS-CoV-2 triggered pathogenesis via metabolic treatment protocol for COVID-19 patients (tentative 8th edi- reprograming of glutamine in host cells. Front Mol Biosci. 2021;7. tion). Infect Dis Immun. 2021;1:8-16. https://doi.org/10.1097/01. https://doi.org/10.3389/fmolb.2020.627842. id9.0000733564.21786.b0. 10. McKechnie JL, Blish CA. The innate immune system: fighting 4. Wang X, Xu G, Liu X, et al. Multiomics: unraveling the on the front lines or fanning the flames of COVID-19? Cell Host panoramic landscapes of SARS-CoV-2 infection. Cell Mol Microbe. 2020;27:863-869. Immunol. 2021;18:2313-2324. 5. McArdle A, Washington KE, Chazarin Orgel B, et al. Discov- ery proteomicsfor COVID-19: where we are now. J Proteome S U P P O RT I N G I N F O R M AT I O N Res. 2021;20:4627. https://doi.org/10.1021/ACS.JPROTEOME. Additional supporting information may be found in the 1C00475. online version of the article at the publisher’s website.