Interspecies analysis to dissect transcriptomic signatures of humans and hamsters in CVID19

In coronavirus disease 2019 (COVID-19) research, hamster species are particularly valuable to study host-pathogen interactions as they develop moderate (Mesocricetus auratus) or severe (Phodopus roborovskii) disease courses following infection. To identify transcriptional similarities between hamsters and humans in the early phase of SARS-CoV-2 infection, we analyzed blood leukocyte single-cell RNA-sequencing datasets from hamsters together with publicly available datasets of COVID-19 patients. Overall, both Mesocricetus auratus and Phodopus roborovskii displayed highest transcriptional changes 2 days after infection. Genes associated with inflammation showed highest expression levels amongst blood leukocytes of Phodopus roborovskii infected with 1 x 10^5 plaque forming units (pfu) SARS-CoV-2. Infecting Phodopus roborovskii with a lower dose of 1 x 10^4 pfu SARS-CoV-2, induced a less critical clinical outcome with fewer inflammatory genes expressed by classical monocytes. Our analyses further include the training of an autoencoder neural network pipeline capable of mapping temporal disease states of hamsters to human COVID-19 severity ranks for individual cell types. Here, the transcriptional state of most Mesocricetus auratus cell types best matched that of patients with moderate disease. However, non-classical monocytes matched the transcriptional state of patients with severe disease in the early phase after infection. In blood leukocytes of Phodopus roborovskii, neutrophils showed the highest similarity to severe COVID-19 patients, while classical monocytes matched the transcriptional state of patients with moderate disease. Taken together, our findings indicate a stronger involvement of monocytes in SARS-CoV-2-infected Mesocricetus auratus and prominent involvement of neutrophils in SARS-CoV-2-infected Phodopus roborovskii.