Integrated analysis of circulating immune cellular and soluble mediators reveals specific COVID19 signatures at hospital admission with utility for prediction of clinical outcomes

Abstract

6 figures, 6 tables.-- Supplementary material available. ; Coronavirus disease 2019 (COVID19), caused by SARS-CoV-2, is a complex disease, with a variety of clinical manifestations ranging from asymptomatic infection or mild cold-like symptoms to more severe cases requiring hospitalization and critical care. The most severe presentations seem to be related with a delayed, deregulated immune response leading to exacerbated inflammation and organ damage with close similarities to sepsis. ; [Methods]: In order to improve the understanding on the relation between host immune response and disease course, we have studied the differences in the cellular (monocytes, CD8+ T and NK cells) and soluble (cytokines, chemokines and immunoregulatory ligands) immune response in blood between Healthy Donors (HD), COVID19 and a group of patients with non-COVID19 respiratory tract infections (NON-COV-RTI). In addition, the immune response profile has been analyzed in COVID19 patients according to disease severity. ; [Results]: In comparison to HDs and patients with NON-COV-RTI, COVID19 patients show a heterogeneous immune response with the presence of both activated and exhausted CD8+ T and NK cells characterised by the expression of the immune checkpoint LAG3 and the presence of the adaptive NK cell subset. An increased frequency of adaptive NK cells and a reduction of NK cells expressing the activating receptors NKp30 and NKp46 correlated with disease severity. Although both activated and exhausted NK cells expressing LAG3 were increased in moderate/severe cases, unsupervised cell clustering analyses revealed a more complex scenario with single NK cells expressing more than one immune checkpoint (PD1, TIM3 and/or LAG3). A general increased level of inflammatory cytokines and chemokines was found in COVID19 patients, some of which like IL18, IL1RA, IL36B and IL31, IL2, IFNα and TNFα, CXCL10, CCL2 and CCL8 were able to differentiate between COVID19 and NON-COV-RTI and correlated with bad prognosis (IL2, TNFα, IL1RA, CCL2, CXCL10 and CXCL9). Notably, we found that soluble NKG2D ligands from the MIC and ULBPs families were increased in COVID19 compared to NON-COV-RTI and correlated with disease severity. ; [Conclusions:] Our results provide a detailed comprehensive analysis of the presence of activated and exhausted CD8+T, NK and monocyte cell subsets as well as extracellular inflammatory factors beyond cytokines/chemokines, specifically associated to COVID19. Importantly, multivariate analysis including clinical, demographical and immunological experimental variables have allowed us to reveal specific immune signatures to i) differentiate COVID19 from other infections and ii) predict disease severity and the risk of death. ; The authors would like to thank the Biobank of the Aragon Health System integrated in the Spanish National Biobanks Network and the Servicios Científico Técnicos de Citometria de Flujo del CIBA for their collaboration. Work in the JP laboratory is funded by the FEDER (Fondo Europeo de Desarrollo Regional, Gobierno de Aragón, Group B29_17R), Health National Institute Carlos III (COV20-00308), Aragón Government (Fondo COVID-19), Fundación Santander-Universidad de Zaragoza (Programa COVID-19), Agencia Estatal de Investigación (SAF2017-83120-C2-1-R; PID2020-113963RBI00), Fundación Inocente, ASPANOA and Carrera de la Mujer de Monzón. EMG is funded by Agencia Estatal de Investigación (SAF2017-83120-C2-1-R and PID2020-113963RB-I00). IUM and SH are supported by a PhD fellowship from Aragon Government, CP by a PhD fellowship from AECC, LS by a PhD fellowship (FPI) from the Ministry of Science, Innovation and Universities. DDM is supported by a postdoctoral fellowship 'Sara Borrell', and MA is supported by a postdoctoral fellowship 'Juan de la Cierva-incorporacion' from the Ministry of Science, Innovation and Universities. EM and BGT are supported by Rio Hortega contract. JP is supported by the ARAID Foundation. ; Peer reviewed