key: cord-0995410-95q0cea9
authors: Hoffmann, Andrew D.; Weinberg, Sam E.; Swaminathan, Suchitra; Chaudhuri, Shuvam; Mubarak, Hannah Faisal; Schipma, Matthew J.; Mao, Chengsheng; Wang, Xinkun; El-Shennawy, Lamiaa; Dashzeveg, Nurmaa K.; Wei, Juncheng; Mehl, Paul J.; Shihadah, Laura J.; Wai, Ching Man; Ostiguin, Carolina; Jia, Yuzhi; D’Amico, Paolo; Wang, Neale R.; Luo, Yuan; Demonbreun, Alexis R.; Ison, Michael G.; Liu, Huiping; Fang, Deyu
title: Unique molecular signatures sustained in circulating monocytes and regulatory T cells in Convalescent COVID-19 patients
date: 2022-03-28
journal: bioRxiv
DOI: 10.1101/2022.03.26.485922
sha: d83b05168561782ee76e6165fb27458b69bfc0e6
doc_id: 995410
cord_uid: 95q0cea9

Over two years into the COVID-19 pandemic, the human immune response to SARS-CoV-2 during the active disease phase has been extensively studied. However, the long-term impact after recovery, which is critical to advance our understanding SARS-CoV-2 and COVID-19-associated long-term complications, remains largely unknown. Herein, we characterized multi-omic single-cell profiles of circulating immune cells in the peripheral blood of 100 patients, including covenlesent COVID-19 and sero-negative controls. The reduced frequencies of both short-lived monocytes and long-lived regulatory T (Treg) cells are significantly associated with the patients recovered from severe COVID-19. Consistently, sc-RNA seq analysis reveals seven heterogeneous clusters of monocytes (M0-M6) and ten Treg clusters (T0-T9) featuring distinct molecular signatures and associated with COVID-19 severity. Asymptomatic patients contain the most abundant clusters of monocyte and Treg expressing high CD74 or IFN-responsive genes. In contrast, the patients recovered from a severe disease have shown two dominant inflammatory monocyte clusters with S100 family genes: S100A8 & A9 with high HLA-I whereas S100A4 & A6 with high HLA-II genes, a specific non-classical monocyte cluster with distinct IFITM family genes, and a unique TGF-β high Treg Cluster. The outpatients and seronegative controls share most of the monocyte and Treg clusters patterns with high expression of HLA genes. Surprisingly, while presumably short-ived monocytes appear to have sustained alterations over 4 months, the decreased frequencies of long-lived Tregs (high HLA-DRA and S100A6) in the outpatients restore over the tested convalescent time (>= 4 months). Collectively, our study identifies sustained and dynamically altered monocytes and Treg clusters with distinct molecular signatures after recovery, associated with COVID-19 severity.

The COVID-19 pandemic caused by SARS-CoV-2 has infected more than 298 million people and caused over 5.4 million deaths since its emergence in late 2019 (1) (2) (3) . New, evolving variants of concern and delays in vaccination (over 90% unvaccinated in low-income countries) mean that it will continue to disrupt the global society for extended time in the future. Shortly after COVID-19 became a global pandemic, post-acute health issues were reported (4) . Typical issues included shortness of breath, muscle fatigue, and prolonged loss of smell. Other reported symptoms include lingering headaches and cognitive difficulties, skin rashes and gastrointestinal discomfort.

Combined, these constellation symptoms are now informally referred to as "long COVID" and likely represent a diverse group of post-infectious syndromes (5) (6) (7) . These symptoms have been reported to occur at higher frequency in patients with severe initial infection (8, 9) . In addition to these long-term ailments, COVID has frequently been characterized by high amounts of inflammation. Together, these suggest that long-term changes in the immune system even following recovery from COVID could be at the root of some of these long-term symptoms.

An efficient immune response against invading pathogens including SARS-CoV-2 requires the early activation of innate immunity, a nonspecific but quick frontline response able to control infection (10) (11) (12) (13) (14) . This effective innate immune response plays a critical role to mount the antigenspecific adaptive immunity. The latter contributes to clearing the infection and preventing reinfection by the same pathogen and more importantly, often sustains the memory response ready for future threatens by the same pathogen (15) . Numerous examinations of patient response to COVID-19 at the single-cell level have been performed since the disease became widespread.

Multiple meta-analyses of single-cell RNA-sequencing data sets have defined the several immune dysregulations are involved in COVID-19 pathogenicity including lymphopenia, impaired IFN response, hyperactivation of myeloid cells and dysregulated macrophage and monocyte functions (11, (16) (17) (18) (19) (20) (21) (22) (23) (24) (25) (26) . A common feature of SARS-CoV-2 infection in severe COVID-19 patients is lymphopenia with a drastic reduction in both T and B lymphocytes in the circulating blood. This lymphopenic response is often negatively associated with the viral load of SARS-CoV-2 as well as the disease severity (21, 27) . In addition, Wilk et al (21) performed some of the earliest singlecell RNA sequencing of COVID-19 patient-derived cells, identifying changes in the myeloid compartment as a consequence of more severe disease states. Despite these progresses, critical questions regarding to whether the dysregulated immune cell phenotypes observed during SARS-CoV-2 infection persist in COVID-19 patients after their full recovery, and if yes, how long the fingerprints in gene expression can be maintained, remain to be answered. In order to understand the post-COVID immune response, we utilized the single cell RNA sequencing approach and determined the molecular signatures of monocytes and understudied regulatory T (Treg) cells in the circulating blood from convalescent COVID-19 patients (when they were unvaccinated). Our studies revealed several unique clusters and molecular signatures in both populations sustained during the recovery phase after SARS-CoV-2 infection.

During the convalescent phase following SARS-CoV-2 infection, the immune system consists of two distinct cellular pools, the virus experienced and naïve cells. As a patient recovers, the SARS-CoV-2 experienced pool changes from a mixture of short lived myeloid, lymphoid and tissue resident cells to predominantly long-lived lymphocytes and tissue resident macrophages. Thus, in convalescent patients, the circulating monocytes and granulocytes are predominantly COVID-19 naïve. To assess whether and how SARS-CoV-2 infection maintains a sustained impact to immune function, we characterized the levels of circulating monocytes from convalescent COVID-19 patients. We enrolled 100 convalescent patients at least three weeks after recovery from COVID-19 and analyzed both plasma IgG antibodies specific to SARS-CoV-2 receptor binding domain (RBD) and cellular immunity profiling (Table 1, Figure 1A ). Patients were classified into 5 groups based on severity of disease and RBD-IgG levels (Supplementary Figure 1B) (28, 29) , including sero-negative healthy control group (undetectable RBD-IgG and PCR negative/no PCR), asymptomatic (detectable RBD-IgG but symptom free during SARS-CoV-2 infection), outpatient (mild symptoms, detectable RBD-IgG), and hospitalized groups (non-ICU and ICU subgroups).

As expected, severe COVID symptoms in hospitalized groups (non-ICU and ICU) correlated with higher IgG levels (Supplementary Figure 1A) . Consistantly, plasma capacity to block RBD binding to ACE2-expressing cells was positively correlated with RBD-IgG levels and disease severity (Supplementary Figure 1B) . While we expected that time from disease onset would likewise correlate due to eventual senescence of antibody producing cells, the plasma RBD-IgG levels and capacity to inhibit RBD binding in outpatients gradually decline by four months after their recovery (Supplementary Figure 1C-D) .

Initial flow cytometric analysis of the gated CD45 + peripheral white blood cells (WBCs) identified monocytes based on the smaller cell size and lower granularity than granulocytes, negative for expression of CD3, dim positive for CD4, and positive for CD64, a specific marker for human myeloid cells, particularly macrophages and monocytes ( Figure 1B) . These CD45 + CD3 -CD4 dim CD64 + cells were further sub-divided based on their expression of CD14 and CD16 into: CD14 + CD16classical monocytes, CD14 + CD16 + intermediate monocytes, and CD14 -CD16 + nonclassical monocytes ( Figure 1B) . Interestingly, the frequency of monocyte population showed a dynamic pattern in association with the COVID-19 disease severity. As shown in Figure 1C 

we observed a statistically significant increase in all monocytes and the classical CD14 + monocyte levels relative to total CD45 + cells in recovered COVID-19 patients with less severe disease (asymptomatic and outpatient), which are found at lower levels in patients recovered from the most severe disease (ICU). Therefore, the decline in monocyte populations during the recovery phase appear to be a signature memory for severe disease.

One possible explanation for the long-term changes seen in circulating monocytes in patients recovered from COVID-19 is global alterations to the immune system driven by long-lived adaptive immune cells. CD4 + CD25 + FoxP3 + Tregs are known to modulate broad aspects of the immune response including maintenance of immune tolerance and global immunosuppression, and we hypothesized that changes in the Treg compartment may persist following the recovery from COVID-19. To assess changes in Treg population in convalescent patients, CD45 + CD3 + T lymphocytes were further analyzed by CD4 expression as helper T cell population. Tregs were then defined as CD4 + CD25 + CD127 low populations as reported (30) (Figure 1B) . Importantly, the Treg cell frequency during the convalescent phase was increased in less severe outpatients compared to hospitalized patients ( Figure 1F) , suggesting that SARS-CoV-2 infection may promote an immunosuppressive state in patients with mild disease.

To further investigate the effect of COVID-19 severity on the molecular signature of convalescent monocyte phenotypes, we sorted CD45 + CD3 -CD64 hi CD4 dim intermediate side scatter cells from 50 convalescent patients and sero-negative controls, and subjected to the 10X Genomics single cell sequencing (See Supplementary Table S1 for further demographic information). Clustering analysis identified seven distinct clusters, each of which is specified with two top enriched genes, including five CD14 + classical monocyte clusters and one distinct CD16 + non-classical monocyte cluster M5_ISG15_IFI6, and one dendritic cell cluster M6_ZEB2_CCNL1 showing lower expression of both HLA and S100 family genes (Figure 2A-B) . Further evaluation of the five classical monocyte subsets found five subsets with inflammatory features. Clusters M1_S100A8_S100A9 and cluster M3_S100A6_S100A4 both showed high but distinct S100 genes (cluster 1 with high levels of S100A8 & 9, but cluster 3 with high S100A6 & 4). Also, the Importantly, these monocyte clusters with distinct molecular signatures were found to be associated with COVID-19 severity. The S100 family gene-high cluster M1-S100A8_S100A9 decreased but the immune regulatory receptor CD74-high cluster M2-CD74_HLA-DR increased in the convalescent patients with asymptomatic COVID-19 vs other patient groups. In addition, the cluster M5_ISG15_IFI6 with high in IFN signature genes were enriched in asymptomatic, but not other COVID-19 patients with severe disease (Figure 3A ). Of note, in outpatients the M5_ISG15_IFI6 cluster with IFN response genes was lost during the early (0-1 month) but restored later (2 months or after) disease onset ( Figure recovery ( Figure 3D ). The IFITM proteins have been identified as cell-autonomous proteins that suppress the early stages of viral replication(32), our data suggest that the increase of this nonclassical cluster might be responsible for the severe COVID-19 pathogenesis. Circulating, classical monocytes are known to have a short lifespan of no more than a few days, while non-classical monocytes display longer survival, but still are unlikely to live for more than 1 or 2 weeks (33).

Thus, the monocytes we observed in our analysis likely developed following clearance of SARS-CoV-2 and the restoration of some degree of immune homeostasis in these patients. In that context, our data demonstrated the long-term changes in monocyte phenotypes in patients recovered from COVID-19 is surprising, which suggests that acute SARS-CoV-2 infection induces immune system alteration even after elimination of replicating virus.

Our discovery that the frequency of FoxP3 + Tregs is increased in the circulating blood from recovered COVID-19 outpatients but decreased in hospitalized patients (Figure 1F) , suggesting that dynamic changes in circulating Treg frequency may impact both acute COVID-19 pathogenesis and long-term complications. We sorted the CD45 + CD3 + CD4 + CD25 high CD127 low population, which are known to sufficiently define human Tregs from PBMC (30), from 27 convalescent COVID-19 patients and sero-negative controls (Supplemental Table S3 (Figure 5A & B) . Interestingly, patients recovered from an asymptomatic infection showed a unique expansion of cluster T0_HLA-DR_CD74 relative to any of the severity groups. This cluster demonstrates the highest expression of class II HLA molecules ( Figure 5C, Supplementary Table S4 ), which has previously been suggested to mark a specific highly suppressive Treg subset that primarily inhibits CD4 + helper T responses through contact-mediated suppression (36) . In addition, our integrative analysis of single-cell Treg transcriptomes shows increased Cluster T0-HLA-DR_CD74 and Cluster T3-EEF1_IFITM1, but decreased Cluster T6-TCF7_GAS5, in asymptomatic and hospitalized patients. In contrast, Tregs in Clusters T2-ANXA1 and T7-KLRB1-LMS1 are decreased in all COVID groups (Figure 5A) .

Further analysis showed that both TGF-β1 and IL7R upregulation occurred during the early COVID-19 recovery phase (0-3 month), which gradually returned to normal during the later recovery phase whereas KLF2 increased and sustained over time in the Tregs from COVID-19 patients ( Figure 5C) .

We also noticed a significant decrease in Tregs of cluster T5_HLA-DRA_S100A6, which shows lower levels of FOXP3 in hospitalized patients, which is possibly due to the inflammation-induced downregulation of FOXP3 (Figure 5A) . Similarly, the frequency of effector/activated Tregs with high expression of FOXP3 and STAT1 in cluster T5_HLA-DRA_S100A6 were dramatically reduced during the early recovery phase in COVID-19 outpatients, which returned to a comparable level similar to that in sero-negative controls 4 months after disease recovery. As expected, the intermediate Cluster T0_HLA-DR_CD74, which occupy one of the major Treg populations, did not show significant changes ( Figure 5D ). Collectively, our study identified a dynamic association of Treg frequency and their immune regulatory and tissue injury protective molecular signatures with disease severity identified in convalescent COVID-19 patients.

Our study enrolled 100 convalescent patients with different disease severities including asymptomatic, outpatient, hospitalized and ICU as well as healthy controls. All patients were enrolled at least three weeks after recovery from COVID-19 confirmed by negative in nucleic acid tests and complete disappearance in clinical symptoms. Their history with SARS-CoV-2 infection was all confirmed by the presence of abundant RBD-specific IgG in their circulating blood. We focused on the analysis of the monocytes, an innate immune cell population that are largely shortlived in the circulating blood, and the understudied regulatory T cells for dissecting the experienced immune responses of COVID-19 patients at the single-cell resolution.

Surprisingly, flow cytometry analysis revealed that the frequency of the short-lived monocyte and Treg populations display a sustained and dynamic association with disease severity in convalescent COVID-19 patients. Consistent with their dynamic frequency changes, we discovered several COVID-19-associated immune signatures such as the elevated IFN responsive genes in asymptomatic COVID-19 patients, which appears to be downregulated in patients with severe disease in particular the ICU patients. Many studies through single cell transcriptome analysis of PBMCs obtained during the active phase of the disease, have shown that the impairments in the IFN response is the key mediators of severe disease (13, 31) . Our study here confirms that this dynamic association of IFN response with COVID-19 disease severity is sustained after full recovery. Interestingly, further longitudinal analysis showed that the elevated IFN responsive genes in monocytes could be sustained for more than 4 months of the studied time. In addition to IFN response, the altered HLA expression on myeloid cells during active SARS-CoV-2 infection, which is presumably due to the virus-induced inflammatory cytokines, have been identified (21, 37) . Our study here demonstrated that this HLA high signature is also sustained after COVID-19 recovery. During the convalescent phase following COVID-19 infection, the immune system consists of two distinct pools of cells, COVID-19 experienced, such as memory T and B cells specific to SARS-CoV-2 antigens, and COVID-19 unexperienced cells. It has been well documented that the half-life of classical and non-classical monocytes in both mice and human is estimated at half-to 2.2 days (38) (39) (40) . Therefore, in convalescent patients, the circulating monocytes and granulocytes are presumably unexperienced directly to SARS-CoV-2-specific antigens as well as the associated inflammatory environment during the acute phase of infection.

Therefore, it is reasonable to speculate that the COVID-19-associated molecular signatures are largely inherited from their precursors and/or even the hematopoietic stem cells experienced with SARS-CoV-2 infection. Indeed, a recent study has suggested that the CD34 + hematopoietic stem/progenitor cells were primed toward megakaryopoiesis, accompanied by expanded megakaryocyte-committed progenitors and increased platelet activation by SARS-CoV-2 infection (20) .

Tregs cells are known to modulate broad aspects of both innate and adaptive immune responses to maintain immune tolerance and global immunosuppression and consequently protect the host from autoimmunity (41) . It has been well established that Tregs play a critical role in protecting tissue injury during microbial infections including influenza viruses and SARS-CoV-2 (42, 43) .

Consistent with those observations note, our study here demonstrated that compared to seronegative healthy controls, the Treg cell frequency was increased in COVID-19 patients with milder disease severity and this increase was not observed in hospitalized and ICU patients. As expected, our in-depth single cell transcriptome analysis of the sorted Treg populations identified 10 distinct clusters, confirming that human circulating Treg populations are highly heterogenous regardless with or without microbial infection (44) (45) (46) (47) (48) . Interestingly, a Treg population (cluster T4_KLF2_TGFB1) that was characterized by elevated expression of TGF-b and KLF2, was only identified in some COVID-19 patients in particular the ICU patients, but not in sero-negative healthy controls. Further longitudinal analysis shows that the upregulation of both TGF-b and KLF2 sustained only during the first three months after full recovery, which was not observed in these convalescent COVID-19 patients 4 months after their full recovery. The clinical significance of this TGF-b and KLF2 high population is unclear. Since Treg-specific production of TGF-b is critical for a variety of immunologic roles, the fact that this population is only observed in patients recovering from COVID-19 suggests a possible role in COVID-19 pathogenesis and recovery.

Consistent to our observation, it has been recently observed that SARS-CoV-2 in severe COVID-19 patients induces TGF-β expression in plasma blasts and possibly involved in IgA production for improving the lung mucosal immunity (49) . A recent interesting study observed that Tregs from patients with severe disease produce a substantial amount of interleukin (IL)-6 and IL-18 (50), which was not observed in COVID-19 patients in their recovery phase. It will be interesting to study how Treg frequencies and their COVID-19-associated molecular signatures dynamically change from active infection to recovery phase. Table S5 . Significance values: ns: p>0.05, * p<0.05, ** p<0.01, *** p<0.001, **** p<0.0001 

Transmission, Diagnosis, and Treatment of Coronavirus Disease 2019 (COVID-19): A Review

A pneumonia outbreak associated with a new coronavirus of probable bat origin

A Novel Coronavirus from Patients with Pneumonia in China

Special Expert Group for Control of the Epidemic of C-otCPMA. [Consideration on the strategies during epidemic stage changing from emergency response to continuous prevention and control

How and why patients made Long Covid

Characterizing long COVID in an international cohort: 7 months of symptoms and their impact

Facing up to long COVID

Incidence, cooccurrence, and evolution of long-COVID features: A 6-month retrospective cohort study of 273,618 survivors of COVID-19

Inborn errors of type I IFN immunity in patients with life-threatening COVID-19

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

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

Symptoms and Health Outcomes Among Survivors of COVID-19 Infection 1 Year After Discharge From Hospitals in Wuhan, China

Immune dysregulation and immunopathology induced by SARS-CoV-2 and related coronaviruses -are we our own worst enemy?

Risk of SARS-CoV-2 reinfection after natural infection

COVID-19 and the human innate immune system

Circuits between infected macrophages and T cells in SARS-CoV-2 pneumonia

Imbalance of Regulatory and Cytotoxic SARS-CoV-2-Reactive CD4(+) T Cells in COVID-19

A molecular singlecell lung atlas of lethal COVID-19

Integrated single-cell analysis unveils diverging immune features of COVID-19, influenza, and other community-acquired pneumonia

Single-cell multi-omics analysis of the immune response in COVID-19

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

Systems biological assessment of immunity to mild versus severe COVID-19 infection in humans

Longitudinal analysis reveals that delayed bystander CD8+ T cell activation and early immune pathology distinguish severe COVID-19 from mild disease

Host-Viral Infection Maps Reveal Signatures of Severe COVID-19 Patients

Single-Cell Profiling of the Antigen-Specific Response to BNT162b2 SARS-CoV-2 RNA Vaccine. bioRxiv

Local and systemic responses to SARS-CoV-2 infection in children and adults

Lymphopenia during the COVID-19 infection: What it shows and what can be learned

Patterns and persistence of SARS-CoV-2 IgG antibodies in Chicago to monitor COVID-19 exposure

High seroprevalence for SARS-CoV-2 among household members of essential workers detected using a dried blood spot assay

CD4(+)CD25 (+)CD127 (low/-) T cells: a more specific Treg population in human peripheral blood

Genetic ancestry effects on the response to viral infection are pervasive but cell type specific

IFITM proteins restrict viral membrane hemifusion

Monocytes, macrophages, dendritic cells and neutrophils: an update on lifespan kinetics in health and disease

Treg cells expressing the coinhibitory molecule TIGIT selectively inhibit proinflammatory Th1 and Th17 cell responses

The role of CD4(+)FoxP3(+) regulatory T cells in the immunopathogenesis of COVID-19: implications for treatment

MHC class II expression identifies functionally distinct human regulatory T cells

Distinct developmental pathways from blood monocytes generate human lung macrophage diversity

Increased turnover of T lymphocytes in HIV-1 infection and its reduction by antiretroviral therapy

The fate and lifespan of human monocyte subsets in steady state and systemic inflammation

Circulatory and maturation kinetics of human monocyte subsets in vivo

Imbalance of regulatory T cells in human autoimmune diseases

A Distinct Function of Regulatory T Cells in Tissue Protection

Regulatory T cells in COVID-19

Fibroblast Heterogeneity and Immunosuppressive Environment in Human Breast Cancer

Treg Heterogeneity, Function, and Homeostasis

Metabolic Control of Treg Cell Stability, Plasticity, and Tissue-Specific Heterogeneity

Human FOXP3(+) Regulatory T Cell Heterogeneity and Function in

Treg cell therapy: How cell heterogeneity can make the difference

SARS-CoV-2 in severe COVID-19 induces a TGF-β-dominated chronic immune response that does not target itself

Profound Treg perturbations correlate with COVID-19 severity

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

A highly conserved cryptic epitope in the receptor binding domains of SARS-CoV-2 and SARS-CoV

Integrated analysis of multimodal single-cell data

Figure S1: IgG levels and blocking capacity in convalescent patient plasma

Anti-SARS-CoV-2-spike protein IgG levels in patients, separated by disease severity

Ability of patient-derived plasma to inhibit binding of AlexaFluor-647 tagged SARS-CoV-2-spike receptor binding domain (RBD) to ACE2-expressing HEK-293 cells

IgG levels as in B, but outpatient samples only separated by time since disease

RBD binding capacity as in (C) but outpatients only and separated by time since disease