Evaluation of the Relationship between Clonal Hematopoiesis and Severe COVID-19 Disease in Rhesus Macaques

Clinical manifestations of infection with the novel SARS-CoV-2 in humans are widely varied, ranging from asymptomatic to COVID-19 respiratory failure and multiorgan damage. Profound inflammation is the hallmark of severe COVID-19 disease, and commonly does not occur until the second week of infection. Although risk factors for this late hyperinflammatory disease have been identified, most notably age and pre-existing co-morbidities, even within high-risk groups the specific factors leading to severe COVID-19 illness remain elusive. Acquired somatic mutations in hematopoietic stem and progenitor cells (HSPCs), termed clonal hematopoiesis (CH), are associated with advanced age, and loss of function (LOF) mutations in certain genes, most commonly DNMT3A and TET2, have been linked to a marked hyperinflammatory phenotype as well as clonal expansion of mutant HSPCs. Given the similar age range of frequent CH and severe COVID-19 disease, the presence of CH could impact the risk of severe COVID-19. Several human cohort studies have suggested this relationship may exist, but results to date are conflicting. Rhesus macaques (RM) have been established as a model for SARS-CoV infection and are being utilized to test therapies and vaccine development, but up to now, macaques have not been reported to develop late hyperinflammatory COVID-19 disease. We have created a robust RM model of CH by introducing LOF TET2 mutations into young adult HSPC via CRISPR/Cas9 followed by autologous transplantation, recapitulating the clonal expansion and hyperinflammatory phenotype. Thus, we hypothesized that macaques with CH could develop severe late COVID-19 disease and be utilized as a model to study disease pathophysiology or test therapeutic approaches. Macaques with either engineered (n=2) or natural CH (n=1) along with age-matched transplanted controls (n=3) were inoculated with SARS-CoV-2 and monitored clinically and via laboratory studies until 12 days post-inoculation (dpi). Macaques normally clear infection and symptoms within 3-5 days of infection. No significant differences in clinical symptoms and blood counts were noted, however, an aged animal with natural DNMT3A CH died on 10 dpi. IL-6 levels were somewhat higher in sera of the CH animals until 12 dpi, and in BAL, mean concentrations of MCP-1, IL-6, IL-8 and MIP-1b were consistently higher in CH macaques compared to controls. Interestingly, we found the median copy number of subgenomic SARS-CoV-2 RNA was higher at every timepoint in the CH group as compared with the control group, in both upper and lower respiratory samples. Lung sections from euthanasia at 10 or 12 dpi showed evidence of mild inflammation in all animals. However, in the immunohistochemical analysis, the viral antigen was detected in the lung tissues of all three animals in the CH group even at the time of autopsy, whereas only one animal of three controls had detectable viral antigen. Although the striking inflammation and serious disease have not been observed, data so far provide evidence of potential pathophysiological differences with or without CH upon SARS-CoV-2 infection. We continue to expand sample size and conduct further analyses to draw a solid conclusion, but we believe this model may be of benefit to understand the relationship between COVID-19 disease and CH.