Abstract
Dogs are occasionally susceptible to SARS-CoV-2, developing few or no clinical signs. Epidemiological surveillance of SARS-CoV-2 in dogs requires testing to distinguish it from other canine coronaviruses. In the last year, significant advances have been made in the diagnosis of SARS-CoV-2, allowing its surveillance in both human and animal populations. Here, using ELISA and automated western blotting (AWB) assays, we performed a longitudinal study on 809 apparently healthy dogs from different regions of France to investigate anti-SARS-CoV-2 antibodies. There were three main groups: (i) 356 dogs sampled once before the pandemic, (ii) 235 dogs sampled once during the pandemic, and (iii) 218 dogs, including 82 dogs sampled twice (before and during the pandemic), 125 dogs sampled twice during the pandemic and 11 dogs sampled three times (once before and twice during the pandemic). Using ELISA, seroprevalence was significantly higher during the pandemic [5.5% (25/453)] than during the pre-pandemic period [1.1% (5/449)]. Among the 218 dogs sampled twice, at least 8 ELISA-seroconversions were observed. ELISA positive pre-pandemic sera were not confirmed in serial tests by AWB, indicating possible ELISA cross-reactivity, probably with other canine coronaviruses. A significant difference was observed between these two serological tests (Q = 88, p = 0.008). A clear correlation was observed between SARS-CoV-2 seroprevalence in dogs and the incidence of SARS-CoV-2 infection in human population from the same area. AWB could be used as a second line assay to confirm the doubtful and discrepant ELISA results in dogs. Our results confirm the previous experimental models regarding the susceptibility of dogs to SARS-CoV-2, suggesting that viral transmission from and between dogs is weak or absent. However, the new variants with multiple mutations could adapt to dogs; this hypothesis cannot be ruled out in the absence of genomic data on SARS-CoV-2 from dogs.