ABO Blood Group Incompatibility Protects Against SARS-CoV-2 Transmission

Abstract

ABO blood groups appear to be associated with the risk of SARS-CoV-2 infection, but the underlying mechanisms and their real importance remain unclear. Two hypotheses have been proposed: ABO compatibility-dependence (neutralization by anti-ABO antibodies) and ABO-dependent intrinsic susceptibility (spike protein attachment to histo-blood group glycans). We tested the first hypothesis through an anonymous questionnaire addressed to hospital staff members. We estimated symptomatic secondary attack rates (SAR) for 333 index cases according to spouse ABO blood group compatibility. Incompatibility was associated with a lower SAR (28% vs. 47%; OR 0.43, 95% CI 0.27-0.69), but no ABO dependence was detected in compatible situations. For the second hypothesis, we detected no binding of recombinant SARS-CoV-2 RBD to blood group-containing glycans. Thus, although no intrinsic differences in susceptibility according to ABO blood type were detected, ABO incompatibility strongly decreased the risk of COVID-19 transmission, suggesting that anti-ABO antibodies contribute to virus neutralization.

Keywords: ABO blood groups; COVID-19; SARS-CoV-2 infection; genetic susceptibility and resistance; incompatibility.

Figure 1

Proposed hypotheses to account for the reported impact of ABO phenotype on the risk of SARS-CoV-2 infection. Under the compatibility-dependence hypothesis (left), protection through virus neutralization is mediated by pre-existing natural anti-ABO antibodies that recognize blood group antigens carried by the virus envelope glycans. Index cases of blood groups A, B or AB excrete virions carrying the A antigen (red spikes), the B antigen (green spikes) or both. Contacts may have anti-A (red) and/or anti-B (green) antibodies able to neutralize the virus carrying the cognate antigen. Protection thus occurs only in situations of ABO incompatibility (I) between the index case and the contact. In the context of compatible encounters (C boxed), no effect of ABO phenotype would be expected. Under the ABO-dependent intrinsic susceptibility hypothesis (right), individuals of blood groups A, B and AB may be intrinsically more susceptible to infection than individuals of blood group O, regardless of the blood group of the person transmitting the virus. Thus, for the same numbers of virions excreted by the index case, the number of contact individuals infected depends on the ABO blood group of the contact. This difference in susceptibility may be due to a direct attachment of the virus spike protein to blood group type glycans (such as the A antigen), facilitating the infection process. Note that although the two hypotheses are not mutually exclusive, their expected consequences are very different. Thus, under the compatibility-dependence hypothesis, transmission rates in populations with a high blood group O frequency should be higher than those in populations in which this blood group is less frequent, because the frequency of compatible encounters is higher. Conversely, under the ABO-dependent intrinsic susceptibility hypothesis, the populations with the highest frequencies of blood group O should benefit from the lower susceptibility conferred by the O blood group.

Figure 2

Effects of ABO blood group on COVID-19 transmission within the cohort of couples. Two-tailed Fisher’s exact test of COVID-19 transmission within couples according to ABO compatibility. The number of individuals in each group is indicated above the bar. Black bars: couples with disease transmission; white bars: couples without disease transmission (A). ABO blood group distribution of the partners of COVID-19 primary cases in situations of ABO compatibility (left panel) and ABO incompatibility (right panel). Black bars: cases with COVID-19 transmission; white bars: cases without COVID-19 transmission. Secondary attack rates are shown above bars for each ABO type; n.a., not applicable because the AB blood type is always compatible (universal recipient; B).

Figure 3

Assay of SARS-CoV-2 RBD binding to blood group A or B antigens. Binding of the RBD-Fc recombinant protein to ACE2 relative to that to the A type 1 hexasaccharide and the H type 1 pentasaccharide (A) and to saliva samples from 19 blood group A, 10 blood group B and 19 blood group O secretors (B) as determined by ELISA. The data shown are the OD values obtained in two independent experiments performed in duplicate. The negative control is the mean OD value obtained in the absence of saliva coating, for three independent plates. Lung tissue sections from a Secretor blood group A and a blood group O donor were incubated with either an anti-A blood group monoclonal antibody or the RBD-Fc fusion protein diluted 1/2 (C). Scale bar: 100 μm. Flow cytometry detection of ACE2 on the surface of ACE2 transfected HEK-293 cells by the RBD-Fc construct. The negative control, in absence of fusion protein is shown by a dashed line, labeling of control HEK-293 cells and of ACE2-expressing HEK-293 cells are shown by red and blue histograms, respectively (D).