This is a PDF file of a peer-reviewed paper that has been accepted for publication. Although unedited, the content has been subjected to preliminary formatting. Nature is providing this early version of the typeset paper as a service to our authors and readers. The text and figures will undergo copyediting and a proof review before the paper is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers apply.

Background Thrombocytopenia and thrombosis are prominent in coronavirus disease 2019 (COVID‐19), particularly among critically ill patients; however, the mechanism is unclear. Such critically ill COVID‐19 patients may be suspected of heparin‐induced thrombocytopenia (HIT), given similar clinical features. Objectives We investigated the presence of platelet‐activating anti‐platelet‐factor 4 (PF4)/heparin antibodies in critically ill COVID‐19 patients suspected of HIT. Patients/Methods We tested 10 critically ill COVID‐19 patients suspected of HIT for anti‐PF4/heparin antibodies and functional platelet activation in the serotonin release assay (SRA). Anti‐human CD32 antibody (IV.3) was added to the SRA to confirm FcγRIIA involvement. Additionally, SARS‐CoV‐2 antibodies were measured using an in‐house ELISA. Finally, von Willebrand factor (VWF) antigen and activity were measured along with A Disintegrin And Metalloprotease with ThromboSpondin‐13 Domain (ADAMTS13) activity and the presence of anti‐ADAMTS13 antibodies. Results Heparin‐induced thrombocytopenia was excluded in all samples based on anti‐PF4/heparin antibody and SRA results. Notably, six COVID‐19 patients demonstrated platelet activation by the SRA that was inhibited by FcγRIIA receptor blockade, confirming an immune complex (IC)‐mediated reaction. Platelet activation was independent of heparin but inhibited by both therapeutic and high dose heparin. All six samples were positive for antibodies targeting the receptor binding domain (RBD) or the spike protein of the SARS‐CoV‐2 virus. These samples also featured significantly increased VWF antigen and activity, which was not statistically different from the four COVID‐19 samples without platelet activation. ADAMTS13 activity was not severely reduced, and ADAMTS13 inhibitors were not present, thus ruling out a primary thrombotic microangiopathy. Conclusions Our study identifies platelet‐activating ICs as a novel mechanism that contributes to critically ill COVID‐19.

Many patients produce antibodies but few lead to heparin‐induced thrombocytopenia (HIT). Pathogenic epitopes are difficult to identify as HIT antibodies are polyclonal and polyspecific. KKO binding to platelet factor 4 (PF4) depends on 13 amino acids, three of which are newly observed. Five amino acids in PF4 can help distinguish pathogenic from non‐pathogenic antibodies. Summary BackgroundHeparin‐induced thrombocytopenia (HIT) is an adverse drug reaction that results in thrombocytopenia and, in some patients, thrombotic complications. HIT is mediated by antibodies that bind to complexes of platelet factor 4 (PF4) and heparin. The antigenic epitopes of these anti‐PF4/heparin antibodies have not yet been precisely defined, because of the polyspecific immune response that characterizes HIT. ObjectivesTo identify PF4 amino acids essential for binding pathogenic HIT antibodies. MethodsAlanine scanning mutagenesis was utilized to produce 70 single point mutations of PF4. Each PF4 mutant was used in an enzyme immunoassay (EIA) to test their capacity to bind a platelet‐activating murine monoclonal anti‐PF4/heparin antibody (KKO) and HIT patient sera (n = 9). Results and ConclusionsWe identified 13 amino acids that were essential for binding KKO because they directly affected either the binding site or the antigenic conformation of PF4. We also identified 10 amino acids that were required for the binding of HIT patient sera and five of these amino acids were required for binding both KKO and the HIT patient sera. The 10 amino acids required for binding HIT sera were further tested to differentiate pathogenic HIT antibodies (platelet activating, n = 45) and non‐pathogenic antibodies (EIA‐positive but not platelet activating, n = 28). We identified five mutations of PF4 that were recognized to be essential for binding pathogenic HIT antibodies. Using alanine scanning mutagenesis, we characterized possible binding sites of pathogenic HIT antibodies on PF4.

Coronavirus Disease 2019 (COVID-19) is a global pandemic caused by the novel severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). While detection of SARS-CoV-2 by polymerase chain reaction with reverse transcription (RT-PCR) is currently used to diagnose acute COVID-19 infection, serological assays are needed to study the humoral immune response to SARS-CoV-2. Anti-SARS-CoV-2 immunoglobulin (Ig)G/A/M antibodies against spike (S) protein and its receptor-binding domain (RBD) were characterized in recovered subjects who were RT-PCR-positive (n = 153) and RT-PCR-negative (n = 55) using an enzyme-linked immunosorbent assay (ELISA). These antibodies were also further assessed for their ability to neutralize live SARS-CoV-2 virus. Anti-SARS-CoV-2 antibodies were detected in 90.9% of resolved subjects up to 180 days post-symptom onset. Anti-S protein and anti-RBD IgG titers correlated (r= 0.5157 and r= 0.6010, respectively) with viral neutralization. Of the RT-PCR-positive subjects, 22 (14.3%) did not have anti-SARS-CoV-2 antibodies; and of those, 17 had RT-PCR cycle threshold (Ct) values >27. These high Ct values raise the possibility that these indeterminate results are from individuals who were not infected or had mild infection that failed to elicit an antibody response. This study highlights the importance of serological surveys to determine population-level immunity based on infection numbers as determined by RT-PCR.

Survivors of severe SARS-CoV-2 infections frequently suffer from a range of post-infection sequelae. Whether survivors of mild or asymptomatic infections can expect any long-term health consequences is not yet known. Herein we investigated lasting changes to soluble inflammatory factors and cellular immune phenotype and function in individuals who had recovered from mild SARS-CoV-2 infections (n = 22), compared to those that had recovered from other mild respiratory infections (n = 11). Individuals who had experienced mild SARS-CoV-2 infections had elevated levels of C-reactive protein 1–3 months after symptom onset, and changes in phenotype and function of circulating T-cells that were not apparent in individuals 6–9 months post-symptom onset. Markers of monocyte activation, and expression of adherence and chemokine receptors indicative of altered migratory capacity, were also higher at 1–3 months post-infection in individuals who had mild SARS-CoV-2, but these were no longer elevated by 6–9 months post-infection. Perhaps most surprisingly, significantly more T-cells could be activated by polyclonal stimulation in individuals who had recently experienced a mild SARS-CoV-2, infection compared to individuals with other recent respiratory infections. These data are indicative of prolonged immune activation and systemic inflammation that persists for at least three months after mild or asymptomatic SARS-CoV-2 infections.