key: cord-0809740-j6vyatv1
authors: Teixeira, P. C.; Dorneles, G. P.; de Santana Filho, P. C.; da Silva, I. M.; Schipper, L. d. L.; Postiga, I. A. d. L.; Neves, C. A. M.; Junior, L. C. R.; Peres, A.; de Souto, J. T.; Fonseca, S. G. d.; Eller, S.; Oliveira, T. F.; Rotta, L.; Thompson, C. E.; Romao, P. R. T.
title: Increased LPS levels coexist with systemic inflammation and result in monocyte activation in severe COVID-19 patients.
date: 2021-06-27
journal: nan
DOI: 10.1101/2021.06.24.21259468
sha: d11033a0f5c716489f9ea37bdf61e5acc0adf895
doc_id: 809740
cord_uid: j6vyatv1

This study aimed to evaluate the link between microbial translocation markers and systemic inflammation at the earliest time-point after hospitalization and at the last 72 h of hospitalization in survivors and non-survivors COVID-19 patients. Sixty-six SARS-CoV-2 RT-PCR+ infected patients and nine non-COVID-19 pneumonia controls were admitted in this study. Blood samples were collected at hospital admission (T1) (Controls and COVID-19+ patients) and 0-72 h before hospital discharge (T2, alive or dead) to analyze systemic cytokines and chemokines, LPS concentrations, and soluble CD14 (sCD14) levels. THP-1 human monocytic cell line was incubated with plasma from survivors and non-survivors COVID-19 patients and their phenotype, activation status, TLR4, and chemokine receptors were analyzed by flow cytometry. COVID-19 patients presented higher IL-6, IFN-{gamma}, TNF-, TGF-{beta}1, CCL2/MCP-1, CCL4/MIP-1{beta}, and CCL5/RANTES levels than controls. Moreover, LPS and sCD14 were higher at hospital admission in SARS-CoV-2-infected patients. Non-survivors COVID-19 patients had increased LPS levels concomitant with higher IL-6, TNF-, CCL2/MCP-1, and CCL5/RANTES levels at T2. Increased expression of CD16 and CCR5 were identified in THP-1 cells incubated with the plasma of survivor patients obtained at T2. The incubation of THP-1 with T2 plasma of non-survivors COVID-19 leads to higher TLR4, CCR2, CCR5, CCR7, and CD69 expression. In conclusion, increased microbial translocation during hospitalization coexist with the inflammatory condition of SARS-CoV-2 infection and could lead to higher monocyte activation in non-survivors COVID-19 patients.

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection causes Coronavirus disease 2019 (COVID-19), a disease with diverse clinical manifestations (1) . While most COVID-19 cases are asymptomatic or mild, severe form of COVID-19 can occur with detrimental manifestations such as acute respiratory distress syndrome (ARDS), multi-organ failure, and death (2) . The COVID-19 immunopathology is mainly characterized by a hyperinflammatory state, strong innate cells response, and lymphocyte activation with an exhausted profile (3) . Furthermore, there is a clear association between COVID-19 severity and higher levels of systemic cytokines (4) .

The mechanism underlying the inflammatory state of COVID-19 is not fully understood at this time. Although the respiratory tract is the main site of infection for COVID-19, the pathogenesis of the disease can also involve different organs and systems, such as the heart, kidneys, intestines, vasculature, and liver (3) . In the gastrointestinal (GI) tract, symptoms such as diarrhea and abdominal distention have been reported in several patients (5) . Furthermore, enterocytes in the ileum and colon express the angiotensinconverting enzyme 2 (ACE-2) receptor and may serve as a site for SARS-CoV-2 entrance and predispose to enteric infection (6) . Additionally, RNA viral particles were detected in stool samples, arising the possibility that bacterial translocation and microbial products from the GI tract to the peripheral blood might contribute to the hyperinflammatory state and COVID-19 severity (7) .

Moreover, systemic inflammation caused by lung infection can lead to the disruption of the gut barrier integrity and increase the permeability to gut microbes and microbial products (8) (9) (10) . Interestingly, disruption in gut barriers and microbial translocation is more likely to occur in older individuals and individuals with chronic diseases, such as . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 27, 2021. ; https://doi.org/10.1101/2021.06.24.21259468 doi: medRxiv preprint type 2 diabetes, cardiovascular diseases, and obesity, which contribute to the exacerbation of systemic inflammation by activating innate immune cells, mainly monocytes and neutrophils (8) .

Lipopolysaccharide (LPS) is a product derived from the membrane of gram-negative bacteria that acts as a potent immune-activating stimulus when recognized by innate immune cells expressing toll-like receptor 4 (TLR4) (11) . Then, innate immune cells become activated and produce high amounts of proinflammatory mediators including cytokines, such as interleukin (IL)-6 and tumor necrosis factor-alpha (TNF-) and chemokines (i.e., CCL2/MCP-1 and CCL5/RANTES) (12) . In this sense, peripheral blood LPS is a well-recognized marker of translocation of microorganisms after events that compromise the integrity of the intestinal mucosa (13, 14) . Besides, soluble CD14 (sCD14) is released after cleavage from the membrane form (mCD14) on the surface of monocytes through exposure to LPS stimulation and is considered a marker of microbial translocation (14, 15) .

Emerging evidence indicates that severe SARS-CoV-2 infection induces disruption in the gut barrier contributing to the systemic spread of bacteria and microbial products which affect the host's response to the infection (9, 16, 17) . However, the time course of the presence of microbial product in the peripheral blood of COVID-19 patients during hospitalization remains unknown. Here, we investigated the link between microbial translocation markers and systemic inflammation in the hospital admission and at the discharge time from the hospital in survivors and non-survivors COVID-19 patients.

We hypothesized that microbial translocation markers are associated with systemic inflammation and could contribute to the death of hospitalized COVID-19 patients.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) Clinical and socio-demographic data were collected from the patient's electronic medical records upon admission to the unit. Body mass index was calculated from weight and height data and was classified according to age. Blood samples from 9 ageand sex-matched SARS-CoV-2 RT-PCR negative controls admitted to hospital with pneumonia were also obtained.

Blood samples were collected from the antecubital vein of the patients into 4 mL tubes with EDTA as anti-coagulant. Blood collection were taken at the earliest time-point after hospitalization (T1) and before discharge time from the hospital (T2: 0 to 72 hours before leaving hospital or death). Plasma samples were obtained by centrifugation (2000g, 10 minutes), aliquoted and immediately kept at -80ºC until analysis.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) Plasma levels of IL-6, IL-10, TNF-α, TGF-β1 (all from Invitrogen Life Sciences, USA),   IFN-γ, CCL2/MCP-1, CCL4/MIP-1β, and CCL5/RANTES (all from Peprotech, USA) were analyzed by Enzyme Linked Immunosorbent Assay (ELISA) following the manufacturer's procedure using a microplate reader (EzBiochrom, USA). The intraassay coefficient of variability was < 7.5%. The detection limits of each cytokine were: IL-6, 2 -200 pg/mL; IL-10, 4 -200 pg/mL; IFN-γ, 10 -300 pg/mL; TNF-α, 2 -200 pg/mL; TGF-β1, 2 -500 pg/mL; CCL5/RANTES, 20 -1000 pg/mL; CCL4/MIP-1β, 50 -1500 pg/mL; CCL2/MCP-1, 50 -1000 pg/mL. The soluble form of CD14 was analyzed using Human CD14 ELISA Kit from Invitrogen (USA), with detection limits ranging 8.23 -8000 pg/mL. 

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Table 1 .

Interestingly, COVID-19 non-survivors (n=23, 30.7% of COVID-19 patients) were heavier than controls (body mass, p=0.01; BMI, p=0.01) and COVID-19 survivors (n=43, 57.3% COVID-19 patients) (body mass, p=0.03; BMI, p=0.02), but no . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 27, 2021. ; https://doi.org/10.1101/2021.06.24.21259468 doi: medRxiv preprint difference between groups was found in the hospital length of stay (p>0.05). COVID-19 non-survivors required more oxygen use during hospitalization than COVID-19 survivors (p=0.002), but similar rates of ICU admission were observed (p>0.05). The most reported symptoms by COVID-19 patients were fever (74.2%), dyspnea (62.8%), and cough (58.5%). Furthermore, 21.7% of COVID-19 patients were diagnosed with hypertension, 18.7% with diabetes mellitus 2, 14.7% with some neurologic disease and 12% with some cardiovascular disease. Regarding pharmacological treatment, azithromycin (24.3%) and corticosteroids therapy (23.7%) were the predominant medicine adopted by medical staff during hospitalization. There was no difference between groups regarding symptoms, comorbidities, or pharmacological therapy during hospitalization (p<0.05), unless the prevalence of cardiovascular diseases (COVID-19 survivors, 4.8; COVID-19 non-survivors, 22.7%; p=0.04).

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 27, 2021. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) 

patients COVID-19 patients presented a higher inflammatory profile at hospital admission compared to controls as presented in Table 2 However, only non-survivors had increased IFN-γ (p<0.01) and TNF-α (p<0.05) compared to controls, and higher TNF-α levels (p<0.05) than survivors. Conversely, survivors had higher levels of CCL5/RANTES (p<0.01) and CCL4/MIP-1β (p<0.05) compared to control and non-survivors groups, respectively. Furthermore, diminished TGF-β1 levels were found in both survivors and non-survivors COVID-19 patients compared to controls (p<0.01). Finally, a strong monocytopenia was observed in COVID-19 non-survivors compared to COVID-19 survivors (p=0.01). These data indicate that COVID-19 patients have higher levels of inflammatory cytokines at the hospital admission, and higher TNF-α plasma levels that may be associated with death.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) 

The time course of systemic inflammatory markers and variables related to microbial translocation in 14 survivors and 12 non-survivors COVID-19 patients were evaluated in two times. To this, peripheral blood of patients was collected at hospital admission (T1) and at the discharge time from the hospital (T2: 0 to 72 hours before leaving . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. 

(T2). Data presented as mean ± SD. Two-way ANOVA followed by Bonferroni's post hoc were applied to verify time and group effects (p<0.05). * Indicates p<0.05.

Non-survivor patients presented higher LPS levels in T2 compared to T1 (p<0.05), but sCD14 only tended to increase in T2 (p=0.055). On the other hand, no changes were identified in sCD14 and LPS at the end of hospitalization in COVID-19 survivors.

Furthermore, the delta value (Δ, T2 -T1) of LPS levels positively correlated with Δ CCL2/MCP-1 levels (r = 0.54; p=0.008) (Figure 4 ). Together, these data indicate that the increase of LPS may be associated with death.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 27, 2021. ; https://doi.org/10.1101/2021.06.24.21259468 doi: medRxiv preprint

survivors and non-survivors are represented (C). Two-way ANOVA followed by Bonferroni's post hoc were applied to verify time and group effects. (p<0.05). * Indicates p<0.05. Pearson's Coefficient test were applied to verify the correlation between inflammatory markers and LPS levels (p<0.05). CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 27, 2021. ; https://doi.org/10.1101/2021.06.24.21259468 doi: medRxiv preprint infection and could lead to the higher monocyte activation which may be associated with worsening of COVID-19 outcomes, such as death.

Angiotensin-Converting Enzyme 2 (ACE-2), the most recognized receptor that binds with the Spike protein of SARS-CoV-2 and allows the host infection, is widely expressed in epithelial cells along the gastrointestinal (GI) tract (8, 21) . Thus, SARS-CoV-2 can invade the intestinal tract causing GI dysfunction which disrupt intestinal barriers, increasing the translocation of microbial products (9, 21) . Increased systemic endotoxemia has been previously described in several pathological conditions associated with hospitalization, becoming a recognized marker of critical ill who suffer sepsis (22) (23) (24) . Furthermore, recent studies described disrupted gut barrier integrity and increased circulating bacteriome in moderate and severe COVID-19 patients (9, 16, 17) .

Here, we found increased LPS and sCD14 levels in the plasma of hospitalized COVID-19 patients at admission and during ICU hospitalization. Arunachalam and colleagues (20) found higher levels of bacterial DNA products, as measured by PCR quantification of 16S ribosomal RNA gene product, and LPS levels in severe ICU patients. The increased bacterial products in the peripheral circulation may contribute to the enhanced macrophage activation in infected tissue (i.e. lungs and gastrointestinal tissues) and to the failure to suppress the hyperinflammation during SARS-CoV-2 infection. In fact, elevated markers level of gut leakage was previously associated with inflammasome activation which leads to cardiac injury during the course of hospitalization in COVID-19 patients (17) . Corroborating these data, other studies show an important role for endotoxin and sCD14 in heart failure (25, 26) .

Microbial translocation to the peripheral blood also contributes to an early activation of monocytic cells. Indeed, we observed higher sCD14 in survivors and non-survivors COVID-19 patients. In line with this, Bowman and coworkers (27) found increased . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 27, 2021. ; https://doi.org/10.1101/2021.06.24.21259468 doi: medRxiv preprint sCD14 levels in COVID-19 patients independent of the disease severity degree.

Furthermore, the authors revealed that COVID-19 critical ill patients who recovered or deceased had the highest sCD14 and lipopolysaccharide-binding protein (LBP, another marker of endotoxemia) values compared to the other groups (27) . These data underlined the presence of strong activation of the monocytic lineage on hospital admission due to SARS-CoV-2 infection in conjunct with LPS translocation and inflammatory exacerbation in ICU patients (28) . In according with this study, systemic markers of monocyte activation, mainly sCD14 and sCD163, correlated with inflammatory cytokines in a non-intensive care unit COVID-19 cohort patients (29) .

It was previously postulated that perturbations in gut microbiota and microbial translocation may exacerbate the severity of cytokine storm in COVID-19 (30) .

Moreover, sCD14 tended to be higher at T2 in non-survivors COVID-19 patients. Here, we hypothesized that serum factors, such as LPS and inflammatory cytokines and chemokines, may directly impact the phenotype of monocytes. To address this question, we incubated monocytic THP-1 lineage with the plasma samples of hospitalized . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Severe COVID-19 patients presented an inflammatory phenotype of monocytes associated with early activation and downregulation of HLA-DR (38, 40) . Monocytes upregulate CD69 expression which promotes tissue infiltration and retention (41, 42) .

Increased activation and CD69 expression in monocytes of severe COVID-19 subjects . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 27, 2021. ; https://doi.org/10.1101/2021.06.24.21259468 doi: medRxiv preprint was associated with diapedesis and tissue infiltration, contributing to the monocytopenia observed in the peripheral circulation of diseased patients (43) . In this line, the increased expression of CCR2, CCR5 and CCR7 in THP-1 incubated with T2 plasma of nonsurvivors COVID-19 patients may indicate the migratory stimulation of soluble factors during SARS-CoV-2 infection. Moreover, THP-1 cells also presented higher CCR5 expression after the incubation with plasma of survivors COVID-19 patients obtained at T2. Infiltrating monocytes constitute the majority of leukocytes migrating into the infected lungs, contributing to the severe lung inflammation in COVID-19 (40) . In fact, increased transcription of CCR2 and CCR5 in macrophages present in bronchial alveolar lavage fluid of severe COVID-19 patients (44) . Furthermore, increased CCR2 + monocytes elicit calcium influx, producing higher levels of reactive oxygen radicals and upregulates integrin expression to recruit neutrophils and mast cells, which exacerbates the inflammatory local response and induces tissue damage (45) (46) (47) . The activation of CCR5 + in macrophages is linked to the signal transduction downstream of Gαi/PI3K/AKT and Gαi/MEK/ERK pathways that induce anti-apoptotic activity, maintaining inflammatory exhausted macrophages in the inflamed lung (48) .

Interestingly, COVID-19 patients treated with lerolimab, an antibody used to block CCR5, presented a rapid reduction of plasma IL-6, restoration of the CD4/CD8 ratio, and a significant decrease in SARS-CoV-2 plasma viremia (49) . Antigen uptake by human monocyte derived dendritic cells, during viral infections, increases the expression of CCR7 that mediates the migration of antigen-bearing cells to lymphatic tissue (50) . Thus, the upregulation of CCR7 is crucial for the recruitment of antigen presenting cells to secondary lymphoid organs during viral infections (51) . Due to their pivotal role in the migration of inflammatory monocytes to the infected tissues and lymphoid organs, future experimental studies should be conducted to evaluate the . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted June 27, 2021. ; https://doi.org/10.1101/2021.06.24.21259468 doi: medRxiv preprint

SARS-CoV-2 immunity: review and applications to phase 3 vaccine candidates

Clinical and immunologic features in severe and moderate forms of Coronavirus Disease

Immunology of COVID-19: Current State of the Science

Longitudinal immunological analyses reveal inflammatory misfiring in severe COVID-19 patients

Don't Overlook Digestive Symptoms in Patients With 2019 Novel Coronavirus Disease (COVID-19)

The digestive system is a potential route of 2019-nCov infection: A bioinformatics analysis based on single-cell transcriptomes

Prolonged SARS-CoV-2 RNA detection in anal/rectal swabs and stool specimens in COVID-19 patients after negative conversion in nasopharyngeal RT-PCR test

Alterations in Gut Microbiota of Patients With COVID-19 During Time of Hospitalization

Endotoxemia and circulating bacteriome in severe COVID-19 patients [Internet]. medRxiv. medRxiv; 2020

Enteric involvement of severe acute respiratory syndrome-associated coronavirus infection

Recognition of lipopolysaccharide pattern by TLR4 complexes

Experimental and Molecular Medicine

Bacterial lipopolysaccharide induces HIF-1 activation in human monocytes via p44/42 MAPK and NF-κB

Lipopolysaccharide-induced bacterial translocation is intestine site-specific and associates with intestinal mucosal inflammation

Distinct inflammatory profiles in HIV-infected individuals under antiretroviral therapy using cannabis, cocaine or cannabis plus cocaine

Tlaskalová-Hogenová H. CD14 is expressed and released as soluble CD14 by human intestinal epithelial cells in vitro: Lipopolysaccharide activation of epithelial cells revisited

Severe COVID-19 Is Fueled by Disrupted Gut Barrier Integrity

International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted

Elevated markers of gut leakage and inflammasome activation in COVID-19 patients with cardiac involvement

Quantitative lipopolysaccharide analysis using HPLC/MS/MS and its combination with the limulus amebocyte lysate assay

Systems biological assessment of immunity to mild versus severe COVID-19 infection in humans. Science (80-)

BRIEF COMMUNICATIONS Evidence for Gastrointestinal Infection of SARS-CoV-2

Diagnostic and prognostic implications of endotoxemia in critical illness: Results of the MEDIC study

Endotoxemia and Sepsis: Important Roles for Lipopolysaccharide Binding Protein and CD14 in Endotoxin Signal Transduction

CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity

The level of endotoxemia in sepsis varies in relation to the underlying infection: Impact on final outcome

Invasive assessment of bacterial endotoxin and inflammatory cytokines in patients with acute heart failure

Endotoxin and immune activation in chronic heart failure: A prospective cohort study

Levels of Soluble CD14 and Tumor Necrosis Factor Receptors 1 and 2 May Be Predictive of Death in Severe Coronavirus Disease

Increased sCD163 and sCD14 Plasmatic Levels and Depletion of Peripheral Blood Pro-Inflammatory Monocytes, Myeloid and Plasmacytoid Dendritic Cells in Patients With Severe COVID-19 Pneumonia

Increased Serum Levels of sCD14 and sCD163 Indicate a Preponderant Role for Monocytes

Front Immunol

Could Perturbation of Gut Microbiota Possibly Exacerbate the Severity of COVID-19 via Cytokine Storm? Front Immunol

International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity

An inflammatory cytokine signature predicts COVID-19 severity and survival

The cytokine storm in COVID-19: An overview of the involvement of the chemokine/chemokinereceptor system

Chemokines and chemokine receptors during COVID-19 infection

Characterization of the inflammatory response to severe COVID-19

Plasma Markers of Disrupted Gut Permeability in Severe COVID-19 Patients 2

COVID-19 patients upregulate toll-like receptor 4-mediated inflammatory signaling that mimics bacterial sepsis

Monocyte activation in systemic Covid-19 infection: Assay and rationale

Available from: www.frontiersin.org . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted

COVID-19: immunopathogenesis and Immunotherapeutics

CD69: from activation marker to metabolic gatekeeper

South African Mixed Ancestry individuals. Sci Rep

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

Transcriptomic characteristics of bronchoalveolar lavage fluid and peripheral blood mononuclear cells in COVID-19 patients

Glycosaminoglycan binding and oligomerization are essential for the in vivo activity of certain chemokines

The Role of the CCL2/CCR2 Axis in Mouse Mast Cell Migration In Vitro and In Vivo

International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted

Recombinant human MCP-1/JE induces chemotaxis, calcium flux, and the respiratory burst in human monocytes

CCL5-CCR5 interaction provides antiapoptotic signals for macrophage survival during viral infection

Disruption of the CCL5/RANTES-CCR5 pathway restores immune homeostasis and reduces plasma viral load in critical COVID-19

Low CCR7-mediated migration of human monocyte derived dendritic cells in response to human respiratory syncytial virus and human metapneumovirus

Signaling Pathways, and Adjuvant Functions in Viral Infection and Prevention

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 27, 2021. ; https://doi.org/10.1101/2021.06.24.21259468 doi: medRxiv preprint It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 27, 2021. ; https://doi.org/10.1101/2021.06.24.21259468 doi: medRxiv preprint