key: cord-0276335-d81gafpt
authors: Garcia-Gonzalez, P.; Tempio, F.; Fuentes, C.; Merino, C.; Vargas, L.; Simon, V.; Ramirez, M.; Rojas, V.; Tobar, E.; Landskron, G.; Araya, J. P.; Navarrete, M.; Bastias, C.; Tordecilla, R.; Varas, M.; Maturana, P.; Marcoleta, A.; Allende, M.; Naves, R.; Hermoso, M.; Salazar-Onfray, F.; Lopez, M.; Bono, M. R.; Osorio, F.
title: Dysregulated immune responses in COVID-19 patients correlating with disease severity and invasive oxygen requirements
date: 2021-09-17
journal: nan
DOI: 10.1101/2021.09.14.21263541
sha: 804de4c68c7fc3b64983f2fbf15a9f04f349df03
doc_id: 276335
cord_uid: d81gafpt

The prognosis of severe COVID-19 patients has motivated research communities to uncover mechanisms of SARS-CoV-2 pathogenesis also on a regional level. In this work, we aimed to understand the immunological dynamics of severe COVID-19 patients with different degrees of illness, and upon long-term recovery. We analyzed immune cellular subsets and SARS-CoV-2-specific antibody isotypes of 66 COVID-19 patients admitted to the Hospital Clinico Universidad de Chile, which were categorized according to the WHO ten-point clinical progression score. These included 29 moderate patients (score 4-5) and 37 severe patients under either high flow oxygen nasal cannula (18 patients, score 6), or invasive mechanical ventilation (19 patients, score 7- 9), plus 28 convalescent patients and 28 healthy controls. Furthermore, six severe patients that recovered from the disease were longitudinally followed over 300 days. Our data indicate that severe COVID-19 patients display increased frequencies of plasmablasts, activated T cells and SARS-CoV-2-specific antibodies compared to moderate and convalescent patients. Remarkably, within the severe COVID-19 group, patients rapidly progressing into invasive mechanical ventilation show higher frequencies of plasmablasts, monocytes, eosinophils, Th1 cells and SARS-CoV-2-specific IgG than patients under high flow oxygen nasal cannula. These findings demonstrate that severe COVID-19 patients progressing into invasive mechanical ventilation show a distinctive type of immunity. In addition, patients that recover from severe COVID-19 begin to regain normal proportions of immune cells 100 days after hospital discharge and maintain high levels of SARS-CoV-2-specific IgG throughout the study, which is an indicative sign of immunological memory. Thus, this work can provide a useful benchmark for improvement of disease outcomes.

The pandemic of Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2), causing coronavirus disease 2019 (COVID- 19) , has led to millions of infections and deaths worldwide, with numbers continuing to increase (WHO dashboard information: https://covid19.who.int). As with other coronaviruses, SARS-CoV-2 is transmitted primarily via respiratory droplets, and upon infection, the median incubation period takes approximately 5-7 days. Disease outcomes range from asymptomatic and mild to more severe and critical courses with pneumonia, Acute Respiratory Distress Syndrome (ARDS), multiorgan failure, and considerable risk of fatality (1, 2) . The risk of developing severe disease has been associated with advanced age, comorbidities such as metabolic syndrome, lung and heart conditions and other factors such as poverty and social determinants in health (3) (4) (5) (6) (7) .

In this context, symptomatic COVID-19 patients develop clinical manifestations within a 14-days window and can exhibit fever and dry cough, fatigue, anosmia, dyspnea, muscle and joint pain, headache, diarrhea and other symptoms (3, 8, 9) . One factor that appears to be related to disease severity and outcome is the requirement for exogenous oxygen supplementation. Critically ill COVID-19 patients rapidly develop ARDS and require oxygen therapies. These therapies can be grouped into those that deliver high flow oxygen, such as high flow nasal cannula (HFNC), and those that require intubation, such as invasive mechanical ventilation (IMV) (10) (11) (12) . The association between oxygen dependency, disease severity and mortality, is a parameter that the World Health Organization (WHO) considers for clinical classification of patients (13).

During this ongoing and evolving pandemic, rapid collaboration from scientists across the globe has begun to unravel the pathogenesis of COVID-19, leading to the definition of patient immune dynamics, clinical courses, and development of different therapeutic strategies to prevent disease progression. These studies have identified components of innate and adaptive immunity dysregulated in symptomatic adult patients (14) (15) (16) (17) (18) , which include exacerbated inflammation characterized by neutrophilia, lymphopenia, T cell exhaustion and myeloid cell activation as hallmarks of severe outcomes (14, 15, (19) (20) (21) (22) (23) (24) . However, whereas most studies have focused on the early stages of the disease, generally within the time surrounding symptom onset and patient hospitalization, fewer reports have addressed the dynamics between disease evolution and immune recovery spanning more extended periods (over three months) of study. Furthermore, the characterization of immune differences in severe COVID-19 patients with different degrees of illness has not been exhaustively studied.

Moreover, few studies have emerged from the Latin-American region, one of the most severely impacted by the COVID-19 pandemic. In this context, regional studies are essential considering the emergence of local SARS-CoV2 variants which may pose an increased risk to human health, and which may confer heterogeneity in clinical outcomes (7, (25) (26) (27) (28) . Genomic surveillance programs in Chile have determined that, during the last year, the country has been successively impacted by most of the variants of concern (VOCs) defined by the WHO, including the B. were compared with 28 COVID-19 recovered/convalescent adult patients and 28 healthy controls. We report that severe patients who rapidly require invasive mechanical ventilation display an immunological profile typified by increased frequencies of plasmablasts, monocytes, eosinophils and Th1 equivalents, which distinguished this group from more stable severe COVID-19 patients treated with high flow nasal cannula.

Finally, a small group of severe COVID-19 patients that recovered from the disease were longitudinally followed for ten months after having a positive PCR result for SARS-CoV-2.

Our data show that these patients started to regain average proportions of key immune cell components dysregulated at the peak of the pathology including granulocyte composition, plasmablasts, and T cell activation, after three months of hospital discharge. (29) .

Whole blood for flow cytometry analysis was collected in EDTA-coated vacutainers, and all blood samples were processed the same day as collection. Before PBMC isolation, 500ul of blood was aliquoted and stabilized with 25ul of Transfix (Cytomark) and stored All rights reserved. No reuse allowed without permission.

perpetuity.

preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in

The copyright holder for this this version posted September 17, 2021. ; https://doi.org/10.1101/2021.09.14.21263541 doi: medRxiv preprint at 4°C in the dark until staining. Blood and isolated PBMC aliquots (approximately 3x10 6 cells) were stabilized using TransFix® (Cytomark) and stored for 5-7 days at 4°C before cell staining and acquisition for immunophenotyping.

Peripheral blood was diluted 1:1 in RPMI-1640 (Gibco) and layered into a SepMate tube (Stemcell Technologies) pre-loaded with 15ml of Lymphoprep (Stemcell Technologies).

SepMate tubes were centrifuged for 20 minutes at 1200g and 20°C without brake and acceleration. The PBMC layer was collected, washed with RPMI for 15 minutes at 600g and 20°C, and treated with ACK lysis buffer (Thermofisher) for 10 minutes before dilution in RPMI for cell count and sample preparation for storage and staining. Approximately 3x10 6 of isolated PBMC was resuspended in 1mL of RPMI and 5% of Transfix and stored at 4°C in the dark for 5-7 days until staining.

Serum samples were collected after whole blood centrifugation at 400 g for 10 minutes at RT without brake. The undiluted serum was then transferred to 15 ml polypropylene conical tubes, aliquoted, and stored at -80 °C for subsequent analysis.

The ELISA protocol was previously adapted from the group of Kramer (30) . Briefly, 96well ELISA plates (MaxiSorp, Thermo Fisher) were coated overnight at 4°C with 50 µl per well of a 2 µg/ml solution of resuspended SARS-CoV-2 Spike protein (Recombinant SARS-CoV-2 S protein S1, Biolegend) on PBS. Then, the coating solution was removed, and the wells were blocked for one hour at room temperature with 150 µl of 3% skim milk prepared in PBS-0.1% Tween-20 (TPBS). After this period, the blocking solution was removed, and the sera to be tested were added. The serum samples to be tested were heated to 56 ° C for 1 hour before use to reduce the risk of residual virus. 100 µl per well of serial dilutions of the sera prepared in 1% skim milk in 0.1% TPBS were added and incubated for 2 hours at room temperature. The plates were then washed three times with 250 µl per well of 0.1% TPBS. Next, 100 µl per well of HRP-conjugated anti-human IgG (HRP Donkey anti-human IgG Clone: Poly24109) or anti-human IgA (Goat Anti-Human IgA alpha chain (HRP) ABCAM ab97215) secondary antibody diluted 1:10,000 in 0.1% TPB was added and incubated for 1 hour at room temperature. The plates were then washed three times with 250 µl of 0.1% TPBS, after which 50 µl of TMB substrate solution (TMB Substrate Reagent, BD Biosciences) were added per well to reveal the reaction which was then stopped after 10 min by adding 50 µl per well of 1M orthophosphoric acid. Optical density at 450 nm was measured on a Molecular Devices Emax ELISA plate reader. The background was established at a DO equal to 0.10. AUC values and data were analyzed using Prism 9 (GraphPad). Differences between groups were calculated using Kruskal-Wallis test and Dunn's multiple comparison post-test. For direct blood staining, 50µL of the corresponding antibody staining mix was added to 100ul of blood samples and incubated for 30 min in the dark at room temperature.

Samples were then treated with 500µL of 1X EasyLyse Erythrocyte-Lysing Reagent perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in 

Statistical analyses were performed using Prism 8 software (GraphPad). Multiple group comparisons were calculated using non-parametric Kruskal Wallis test with Dunn's multiple comparison post-test. Comparisons between two groups were assessed using Mann-Whitney t-tests. Results are shown as individual data with mean ± SEM for each group.

All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted September 17, 2021. ; https://doi.org/10.1101/2021.09.14.21263541 doi: medRxiv preprint

To better understand the different immune signatures of COVID-19 patients, we devised a 14-parameter flow cytometry strategy to stratify patients in terms of innate immune cell composition, lymphocyte populations, activated/exhausted CD4 + and CD8 + T cells, and B cell/plasmablasts, following a methodology adapted from reported work (14, 16, 19) . This protocol was applied to whole blood (for analysis of granulocytes including neutrophils and eosinophils) or to peripheral blood mononuclear cells (PBMC, for analysis of lymphocytes and monocytes) (Scheme of the study is depicted in Supp. To classify COVID-19 patients according to disease severity, we employed the WHO tenpoint clinical progression scale that provides standard outcome measures for COVID-19 studies (13). According to this scale, our study included 29 patients categorized as 'Hospitalized Moderate disease' (referred to as "M''), which consisted of 5 patients with a score of 4 (hospitalized without oxygen therapy) and 24 patients with a score of 5 (oxygen by mask/nasal prongs). We also recruited 37 patients categorized as 'Hospitalized Severe Disease 'patients (referred to as "S"), consisting of 18 patients with a score of 6 (hospitalized with HFNC), 5 patients with a score of 7 (hospitalized with intubation and mechanical ventilation), 9 patients with score 8 (hospitalized with intubation, mechanical ventilation, PaO2:FiO2 < 150 or vasopressor treatment), and 5 patients with score 9 (hospitalized with intubation, mechanical ventilation, PaO2:FiO2 < 150 and vasopressor dialysis) ( Table 1 ). This study also included a group of 28 convalescent COVID-19 patients who went through SARS-CoV2 infection 118,5 days (IQR: 22,5-33), 46,3%, have moderate hypertension and 15,8% have diabetes. We did not find significant differences in the BMI of moderate and severe patients (Table 1) .

However, among the severe patient group, we found a significantly higher number of individuals with moderate arterial hypertension compared to the moderate patient group. Three moderate and 5 severe COVID-19 patients died during the time of the study. All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted September 17, 2021. ; https://doi.org/10.1101/2021.09.14.21263541 doi: medRxiv preprint Finally, our study includes 28 healthy controls (HC) matched by age and comorbidities. A positive SARS-CoV-2 PCR confirmed all COVID-19 patients before the analysis, and a negative result verified the absence of infection in healthy controls.

We analyzed the immune subset composition of moderate, severe, and convalescent COVID-19 patients and healthy controls to identify significant discriminators of the pathology (gating strategy depicted in Supp Fig 2) . In total blood, analysis of innate immune cells showed a marked presence of polymorphonuclear leukocytes (PMNs) and a clear increase in neutrophils in moderate and severe patients, which segregated them from healthy controls as reported (14, 31) , and also from convalescent patients (Fig 1a) .

Even though the frequencies of eosinophils oscillated within a reduced range, there was a significant reduction of these cells in moderate and severe patients compared to healthy controls and convalescent patients (Fig 1b) . Furthermore, as reported (14), basophil frequency was drastically reduced in moderate and severe patients compared to healthy controls but also compared to convalescent individuals, suggesting that alterations in granulocyte populations occur during the onset of pathology (Fig 1c) . We also observed significant reductions in plasmacytoid dendritic cell (pDCs) frequency in moderate and severe patients compared to healthy controls and convalescent patients (Fig 1d) .

Interestingly, even though monocyte frequency was not altered among COVID-19 patients (Supp Fig 3a, Fig 1e) , the frequency of intermediate monocytes in moderate, severe, and convalescent patients was increased compared to healthy controls (Fig 1e) , suggesting that the pathology re-shapes this myeloid cell repertoire even in patients that have recovered from the disease. Accordingly, non-conventional monocyte frequency was decreased in severe and convalescent patients (Fig 1e) . Along these lines, severe COVID-19 patients downregulated expression of HLA-DR in all monocyte subtypes compared to healthy controls (Fig 1f, Supp Fig 3b) , but convalescent patients showed expression of the molecule that was comparable to that of healthy controls. This data is accompanied by decreased expression of the costimulatory molecule CD86 in monocyte subtypes from severe patients compared to healthy controls (Fig 1g) .

Finally, we measured surface expression levels (through the Mean Fluorescence Intensity, MFI) of PD-L1, an immunoregulatory molecule with reported roles in myeloid cells in COVID-19 (32) . Our data showed that neutrophils from severe patients have All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted September 17, 2021. ; https://doi.org/10.1101/2021.09.14.21263541 doi: medRxiv preprint decreased PD-L1 expression compared to healthy controls and convalescent patients (Fig   1h, Supp Fig 3c) . Altogether, these data indicate that severe COVID-19 patients show a dysregulated innate immune cell balance, evidenced at the level of granulocyte composition and in the repertoire and activation status of monocytes.

Next, we sought to characterize the dynamics of lymphocytes in COVID-19 patients throughout different stages of the pathology. Data depicted in Figure 2a (14), were significantly increased in moderate, severe and convalescent patients compared to healthy controls, which suggests that convalescent patients show long term cellular changes associated to the infection (Supp Fig 4a) . Furthermore, although we did not observe significant changes in the frequency of näive or IgD-CD27-B cells, we found a decrease in B cell memory switch in severe and moderate patients compared to healthy controls and convalescent patients (Supp Fig 4b) . Remarkably, plasmablasts, which are dividing antibody-secreting cells circulating in the blood (33) were detected in COVID-19 patients in a pattern that correlated with disease severity (Fig 2b-c) . As such, these data prompted us to investigate the antibody responses in COVID-19 patients (Fig 2d) . To this end, we performed ELISAs to detect antibodies specific against the Spike (S) protein of SARS-CoV-2. These assays indicated that moderate and severe COVID-19 patients produced detectable levels of S-specific IgM, IgG and IgA antibodies, suggesting that SARS-CoV-2 infection triggers universal production of antibody isotypes against the virus. In addition, convalescent patients expressed high levels of S-specific IgG but not IgM antibodies, agreeing with a temporal resolution of the infection and with the generation of immune memory (Fig 2d) . Altogether, these data indicate that plasmablasts and the presence of antibody isotypes are hallmarks of COVID-19 patients, segregating these groups from patients that have resolved the infection.

Next, we analyzed T cell dynamics in SARS-CoV-2 infected individuals. In agreement with reported work (14, 22, 24) , moderate and severe COVID-19 patients showed decreased T All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted September 17, 2021. ; https://doi.org/10.1101/2021.09.14.21263541 doi: medRxiv preprint cell frequencies in a severity-related manner (Fig 2e) . Interestingly, although convalescent patients restore B cell frequencies to levels comparable to those observed in healthy controls (Fig 2a) , these individuals do not restore T cell populations to normal levels (Fig 2e) , suggesting that lymphopenia recovery occurs at different timings for B cells and T cells. We next determined the activated/exhausted status of CD4 + and CD8 + T cells (Fig 2f-h) and observed that patients with moderate and severe COVID-19 show increased accumulation of activated CD8 + T cells, which distinguished these cohorts from healthy controls and convalescent individuals (Fig 2f-h) . In addition, CD8 + T cells expressing the exhaustion marker PD-1 accumulated in moderate patients and there was a clear, albeit not significant, trend towards higher frequencies in severe COVID-19 patients (Fig 2h) . On the other hand, analysis of activated CD4 + T cells showed a similar tendency, with high frequencies in moderate and severe patients compared to healthy controls (Fig 2i) , whereas CD4 + PD-1 + T cells segregated moderate and severe patients from healthy controls and convalescent individuals (Fig 2j) . Furthermore, using chemokine receptors as correlates of T helper subsets (14) , we found that the frequency of Th1 equivalents (CCR6 + CXCR3 + ) was unaltered between healthy controls, moderate and severe COVID-19 patients, but it was increased compared to convalescent individuals (Fig 2k) . Frequencies of Th17 equivalents (CCR6 + CXCR3 -) remained unaltered among groups (Supp Fig 2c) . Finally, we investigated if the production of granzyme B, a primary cytotoxic mediator, distinguished COVID-19 patients depending on the severity of the disease (Fig 2l-m) . Our data indicate that expression of granzyme B in CD8 + T cells was not different between healthy controls, moderate and severe patients suggesting that this molecule does not discriminate between different degrees of SARS-CoV2 infection.

However, we noticed that convalescent individuals express markedly low molecule levels, possibly reflecting resolution from an effector immune response (Fig 2l-m) . In summary, these data suggest that hospitalized COVID-19 patients show a T cell signature typified by an accumulation of activated CD8 + and CD4 + T cells and CD4 + PD-1 + T cells.

Finally, to obtain a broad overview of immune cell dynamics of COVID-19 patients, we devised an unsupervised flow cytometry analysis of CD45 + cells to cover representative innate and adaptive cells present in the PBMC. Data were visualized using a Uniform Manifold Approximation Projection (UMAP) (Fig 2n) . PBMCs were grouped into populations by FlowSOM automated clustering tool of FlowJo Software for unbiased identification of leukocytes in healthy controls, moderate, severe, and convalescent All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted September 17, 2021. ; https://doi.org/10.1101/2021.09.14.21263541 doi: medRxiv preprint patients (Fig 2n) . This analysis identified 8 main clusters corresponding to CD4 + T cells (cluster 1, CD3 + CD4 + ); CD8 + T cells (cluster 2, CD3 + CD8 + ), B cells (cluster 3, CD3 -CD19 + ), cDCs (cluster 4, CD3 -HLA-DR + CD14 -CD11c + ), monocytes (Cluster 5, CD3 -CD14 + CD16 -/int and cluster 6, CD3 -CD14 + CD16 + ), pDCs (cluster 7, CD3 -HLA-DR + CD14 -CD123 + ) and plasmablasts (cluster 8, CD3 -CD19 + CD138 -CD38 + ) (Fig 2n) . We observed changes in the composition of pDCs, CD8 + T cells and CD4 + T cells, along with the emergence of plasmablasts in moderate and severe COVID-19 patients, a feature no longer observed in convalescent patients. Altogether, this data confirms our previous findings obtained through manual gating analysis (Fig 1d, 2c, 2e) and supports the conclusion that hospitalized COVID-19 patients display marked changes in the composition of pDCs, T cells and plasmablasts.

Our data indicate that severe COVID-19 patients show significant changes in subset composition of innate and adaptive immune cells. These findings prompted us to investigate whether there was a correlation between immune cell frequency and disease severity in this group of patients. To this end, we subclassified severe COVID-19 patients according to the World Health Organization (WHO) ten-point clinical progression scale for COVID-19 patients and analyzed immune parameters associated with each group.

Eighteen severe patients categorized in score 6 and corresponding to hospitalized patients with HFNC were compared to 19 severe patients categorized in scores 7-9, which corresponds to hospitalized patients with intubation and mechanical ventilation with or without vasopressor use. Data in Figure 3 indicates that several immune cell types including B cells, CD4 + and CD8 + T cells, activated and PD-1 + CD4 + and CD8 + T cells, neutrophils, DCs, and basophils did not discriminate between score 6 and score 7-9

patients. In addition, D-dimer concentration in plasma was similarly elevated in both groups, concordant with the clinical severity of the disease (Fig 3a) . However, specific immune cells including plasmablasts, monocytes, eosinophils, and Th1 equivalents, were increased in patients with higher clinical scores (Fig 3b-e) . Remarkably, the concentration of S-specific IgG also differentiated patients under HFNC treatment versus invasive ventilated patients, with higher antibody titers in patients with higher severity rank (Fig 3b) . Conversely, granzyme B expression was reduced in patients with mechanical ventilation compared to HFNC patients (Fig 3d) . Altogether, these data All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted September 17, 2021. ; https://doi.org/10.1101/2021.09.14.21263541 doi: medRxiv preprint indicate that patients that required invasive mechanical ventilation display an exacerbated innate and adaptive immune response typified by increased frequencies of plasmablasts, monocytes, Th1 equivalents, and S-specific IgG, which may provide relevant information regarding the progression of severe COVID-19 pathology.

To better understand how the immune system of severe COVID-19 patients recovers after infection, we carried out a longitudinal analysis of 6 severe patients, which were followed (Fig 4b-c) . These data indicate a reversal of the pan-lymphocytopenia reported for severe COVID-19 patients (19) . Remarkably, plasma levels of S-specific IgG remained elevated 300 days after infection (Fig 4c) , confirming the notion that COVID-19 patients generate a durable immunological memory (34) . In contrast, we also measured the trajectory of S-specific IgA and found that titers of the isotype decline over time (Fig   4c) . Finally, we also observed a constant decrease in the frequencies of activated and PD-1 + CD8 + and CD4 + T cells during R1-R2 time points. Altogether, these data indicate that severe COVID-19 patients that can resolve SARS-CoV-2 infection show signs of restoration of specific immune cell compartments on a trajectory that resembles transition to homeostasis and generation of immunological memory. All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted September 17, 2021. ; https://doi.org/10.1101/2021.09.14.21263541 doi: medRxiv preprint

Although SARS-CoV2 infection generally leads to a mild disease in a large proportion of infected individuals, 5-15% of COVID-19 patients develop a severe pathology that progresses to pneumonia and respiratory failure (35) (36) (37) (38) . In this context, the immune cell landscape of severe COVID-19 patients is reported to be dysregulated (14, 19, 31, 32) , and detailed insights on the cellular dynamics of severe COVID-19 patients are urgently needed to identify potential disease intervention points. In this work, we analyzed immune cell dynamics of moderate, severe, and convalescent COVID-19 patients and further investigated the dynamics of severe patients in terms of oxygen dependence and long-term trajectories of ICU patients during the acute symptomatic stage and after recovery.

Previous studies have shown dysregulation of innate and adaptive immune cell compartments in patients with moderate, severe, and convalescent COVID-19 (14, 19, 22) . As per adaptive immune parameters, the frequencies of plasmablasts and activated CD8 + and CD4 + T cells were increased in severe COVID-19 patients. Additionally, we also found higher frequency of PD-1 + CD8 + and CD4 + T cells, suggestive of potential T cell exhaustion.

As expected, antibody data indicated that moderate and severe COVID-19 patients produce S-specific IgM, IgG, and IgA. However, in contrast to previous observations (19, 20) , our data show that convalescent patients no longer express detectable levels of All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted September 17, 2021. ; https://doi.org/10.1101/2021.09.14.21263541 doi: medRxiv preprint S-specific IgM, which may reflect differences in the time of sample collection between these studies.

Interestingly, we also detected differences related to immune subset composition among severe COVID-19 sub-cohorts, which were associated to the intensity of respiratory support defined as the type of oxygen, severity measured by PaO2:FiO2, and the presence of organ dysfunction defined by the WHO classification. In addition, a clinical debate persists as to why some severe COVID-19 patients get worse rapidly and require IMV while others display a more stable pathological course. In this context, we identified specific immune cell types that distinguished between severe COVID-19 patients under HFNC (displaying a more stable clinical course) versus patients who access rapidly to mechanical ventilation (Scores 6 and 7-9, respectively). These changes include increased frequencies of plasmablasts, monocytes, eosinophils, Th1 equivalents and underrepresentation of granzyme B + producing CD8 + T cells in ventilated severe COVID-19 patients. How exactly these cell types contribute to promote or prevent the pathology remains to be fully elucidated. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted September 17, 2021. ; https://doi.org/10.1101/2021.09.14.21263541 doi: medRxiv preprint https://auspice.cov2.cl/ncov/chile-global-2021-07-02). In Chile, the gamma and lambda variants have dominated the genomic landscape of SARS-CoV-2 in the first half of 2021, are also prevalent in many Latin American countries and remain conspicuously understudied. Further attention should be directed at immunological and clinical features associated with particular variants in local populations to detect potential effects of viral evolution on clinical outcomes. All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted September 17, 2021. ; https://doi.org/10.1101/2021.09.14.21263541 doi: medRxiv preprint data in the study and the corresponding authors had final responsibility for the decision to submit for publication. Mechanical ventilation SpO2/FIO2 <150 (SpO2/FiO2 <200) or vasopressors 9

Severe 9

Mechanical ventilation SpO2/FiO2 <150 and vasopressors, dialysis, or ECMO 5 All rights reserved. No reuse allowed without permission.

perpetuity.

preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in

The copyright holder for this this version posted September 17, 2021. ; perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted September 17, 2021. ; https://doi.org/10.1101/2021.09.14.21263541 doi: medRxiv preprint samples after recovery and clinical discharge (R1 and R2). The first sample (T1) was taken upon ICU admission and positive SARS-CoV-2 PCR, while second (T2) and third (T3) blood samples were taken 5-7 and 9-13 days after, correspondingly. R1 samples for follow-ups were taken once the patient had recovered and been released from the hospital, between 1 and 3 months after ICU admission. R2 samples were taken 6 months after patient recovery and clinical discharge. B. Changes in frequency of innate immune populations during disease evolution and recovery. C-D. Changes of adaptive immune components throughout time. Graphs show individual data for each patient and the mean ± SEM per group connected by a red solid line. Representative histograms for HLA-DR and CD86 expression assessed in the different All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted September 17, 2021. ; https://doi.org/10.1101/2021.09.14.21263541 doi: medRxiv preprint l. m. 

COVID-19: Specific and Non-Specific Clinical Manifestations and Symptoms: The Current State of Knowledge

Risk Factors Associated with Acute Respiratory Distress Syndrome and Death in Patients with Coronavirus Disease

Clinical features of patients infected with 2019 novel coronavirus in Wuhan

Sex-specific clinical characteristics and prognosis of coronavirus disease-19 infection in Wuhan, China: a retrospective study of 168 severe patients

Sex differences in immune responses that underlie COVID-19 disease outcomes

Factors associated with COVID-19-related death using OpenSAFELY

Socioeconomic status determines COVID-19 incidence and related mortality in

The Incubation Period of Coronavirus Disease 2019 (COVID-19) From Publicly Reported Confirmed Cases: Estimation and Application

Incubation period of COVID-19: A systematic review and meta-analysis

A new coronavirus associated with human respiratory disease in China

Risk Factors Associated with Mortality among Patients with COVID-19 in Intensive Care Units in Lombardy, Italy

WHO Working Group on the Clinical Characterisation and Management of COVID-19. A minimal common outcome measure set for COVID-19 clinical research

A dynamic COVID-19 immune signature includes associations with poor prognosis

Longitudinal immune profiling reveals key myeloid signatures associated with COVID-19

Breadth of concomitant immune responses prior to patient recovery: a case report of non-severe COVID-19

COVID-19 and the human innate immune system

Adaptive immunity to SARS-CoV-2 and COVID-19

Comprehensive mapping of immune perturbations associated with severe COVID-19

Imbalanced Host Response to SARS-CoV-2 Drives Development of COVID-19

Neutrophil-to-lymphocyte ratio predicts critical illness patients with 2019 coronavirus disease in the early stage

Deep immune profiling of COVID-19 patients reveals distinct immunotypes with therapeutic implications

A single-cell atlas of the peripheral immune response in patients with severe COVID-19

Lymphopenia is associated with severe coronavirus disease 2019 (COVID-19) infections: A systemic review and meta-analysis

Lessons learnt from easing COVID-19 restrictions: an analysis of countries and regions in Asia Pacific and Europe

Leading Edge Emerging Pandemic Diseases: How We Got to COVID-19

Integrated analyses of single-cell atlases reveal age, gender, and smoking status associations with cell type-specific expression of mediators of SARS-CoV-2 viral entry and highlights inflammatory programs in putative target cells

The trinity of COVID-19: immunity, inflammation and intervention

The REDCap consortium: Building an international community of software platform partners

A serological assay to detect SARS-CoV-2 seroconversion in humans

Longitudinal analyses reveal immunological misfiring in severe COVID-19

Severe COVID-19 Is Marked by a Dysregulated Myeloid Cell Compartment

Factors That Regulate the Generation of Antibody-Secreting Plasma Cells

Immunological memory to SARS-CoV-2 assessed for up to 8 months after infection

Clinical and immunological features of severe and moderate coronavirus disease 2019

COVID-19 severity correlates with airway epithelium-immune cell interactions identified by single-cell analysis

The Emerging Roles of Endoplasmic Reticulum Stress in Balancing Immunity and Tolerance in Health and Diseases: Mechanisms and Opportunities

Characteristics of and Important Lessons from the Coronavirus Disease 2019 (COVID-19) Outbreak in China: Summary of a Report of 72314 Cases from the Chinese Center for Disease Control and Prevention

Longitudinal analysis shows durable and broad immune memory after SARS-CoV-2 infection with persisting antibody responses and memory B and T cells

We thank patients and blood donors who consented to this study and the health team at the Hospital Clínico Universidad de Chile. We also thank members of the laboratory of cell therapy and experimental medicine for establishing appropriated conditions for experimental work in pandemic. We acknowledge the Millennium Institute on Immunology and Immunotherapy (IMII, ICM-ANID, ICN09_016) for access to critical flow cytometry equipment. We express our appreciation to the colleagues that belong to the