No abstract

Among risk factors for SARS-CoV-2, ABO(H) blood group antigens have been one of the most recognized predictors of infection. However, the mechanisms whereby ABO(H) antigens influence susceptibility to COVID-19 remain incompletely understood. The receptor binding domain (RBD) of SARS-CoV-2, which facilitates host cell engagement, bears significant similarity to galectins, an ancient family of carbohydrate binding proteins. As ABO(H) blood group antigens are carbohydrates, we compared the glycan binding specificity of the SARS-COV-2 RBD with galectins. Similar to the binding profile of several galectins, the RBDs of SARS-CoV-2, including Delta and Omicron variants, exhibited specificity for blood group A. Not only did each RBD recognize blood group A in a glycan array format, but each SARS-CoV-2 virus likewise displayed a preferential ability to infect blood group A expressing cells. Preincubation of blood group A cells with a blood group binding galectin specifically inhibited the blood group A enhancement of SARS-CoV-2 infection, while similar incubation with a galectin that does not recognize blood group antigens failed to impact SARS-CoV-2 infection. These results demonstrate that SARS-CoV-2 can engage blood group A, providing a direct link between ABO(H) blood group expression and SARS-CoV-2 infection.

Microbial glycan microarrays (MGMs) populated with purified microbial glycans have been used to define the specificity of host immune factors toward microbes in a high throughput manner. However, a limitation of such arrays is that glycan presentation may not fully recapitulate the natural presentation that exists on microbes. This raises the possibility that interactions observed on the array, while often helpful in predicting actual interactions with intact microbes, may not always accurately ascertain the overall affinity of a host immune factor for a given microbe. Using galectin-8 (Gal-8) as a probe, we compared the specificity and overall affinity observed using a MGM populated with glycans harvested from various strains of Streptococcus pneumoniae to an intact microbe microarray (MMA). Our results demonstrate that while similarities in binding specificity between the MGM and MMA are apparent, Gal-8 binding toward the MMA more accurately predicted interactions with strains of S. pneumoniae, including the overall specificity of Gal-8 antimicrobial activity. Taken together, these results not only demonstrate that Gal-8 possesses antimicrobial activity against distinct strains of S. pneumoniae that utilize molecular mimicry, but that microarray platforms populated with intact microbes present an advantageous strategy when exploring host interactions with microbes.