Focus on the general concepts, as opposed to understanding every word.Answer the following question for your initial post: How does blood type influence symptom variations in people exposed

viruses

Review

ABO Blood Types and COVID-19: Spurious, Anecdotal, or Truly Important Relationships? A Reasoned Review of Available Data

Jacques Le Pendu 1,*, Adrien Breiman 1,2 , Jézabel Rocher 1, Michel Dion 3 and Nathalie Ruvoën-Clouet 1,4

���������� �������

Citation: Pendu, J.L.; Breiman, A.;

Rocher, J.; Dion, M.; Ruvoën-Clouet,

N. ABO Blood Types and COVID-19:

Spurious, Anecdotal, or Truly

Important Relationships? A Reasoned

Review of Available Data. Viruses

2021, 13, 160. https://doi.org/

10.3390/v13020160

Academic Editor: Shan-Lu Liu

Received: 18 December 2020

Accepted: 19 January 2021

Published: 22 January 2021

Publisher’s Note: MDPI stays neutral

with regard to jurisdictional claims in

published maps and institutional affil-

iations.

Copyright: © 2021 by the authors.

Licensee MDPI, Basel, Switzerland.

This article is an open access article

distributed under the terms and

conditions of the Creative Commons

Attribution (CC BY) license (https://

creativecommons.org/licenses/by/

4.0/).

1 CRCINA, INSERM, Université de Nantes, F-44000 Nantes, France; [email protected] (A.B.); [email protected] (J.R.); [email protected] (N.R.-C.)

2 CHU de Nantes, F-44000 Nantes, France 3 Microbiotes Hosts Antibiotics and Bacterial Resistances (MiHAR), Université de Nantes,

F-44000 Nantes, France; [email protected] 4 Oniris, Ecole Nationale Vétérinaire, Agroalimentaire et de l’Alimentation, F-44307 Nantes, France * Correspondence: [email protected]

Abstract: Since the emergence of COVID-19, many publications have reported associations with ABO blood types. Despite between-study discrepancies, an overall consensus has emerged whereby blood group O appears associated with a lower risk of COVID-19, while non-O blood types appear detrimental. Two major hypotheses may explain these findings: First, natural anti-A and anti-B antibodies could be partially protective against SARS-CoV-2 virions carrying blood group antigens originating from non-O individuals. Second, O individuals are less prone to thrombosis and vascular dysfunction than non-O individuals and therefore could be at a lesser risk in case of severe lung dysfunction. Here, we review the literature on the topic in light of these hypotheses. We find that between-study variation may be explained by differences in study settings and that both mecha- nisms are likely at play. Moreover, as frequencies of ABO phenotypes are highly variable between populations or geographical areas, the ABO coefficient of variation, rather than the frequency of each individual phenotype is expected to determine impact of the ABO system on virus transmission. Accordingly, the ABO coefficient of variation correlates with COVID-19 prevalence. Overall, despite modest apparent risk differences between ABO subtypes, the ABO blood group system might play a major role in the COVID-19 pandemic when considered at the population level.

Keywords: COVID-19; ABO blood groups; natural antibodies; thrombosis; attack rate; susceptibility

1. Introduction

Since the first description of COVID-19 in Wuhan, China, the emergence of the new SARS-CoV-2 coronavirus led to a global public health crisis due to massive morbidity and a burden of mortality that rapidly overwhelmed health systems worldwide. However, the disease impact shows considerable variation between countries and geographic areas for reasons that are poorly understood. Besides differences in the economical, sociological, behavioral, and political response to the pandemic, genetic factors might also play a role in their own right. Thus, soon after the beginning of the pandemic a publication from Wuhan, China, reported a higher risk of infection for people of blood group A, and inversely a lower risk for people of blood group O [1]. Since then, associations with the ABO blood groups have been described in several additional publications from China as well as many other locations from Asia, the Middle East, Europe, and North America [2–36]. Associations between ABO phenotypes were described with either the risk of infection or disease severity, although most studies did not explicitly separate these two aspects. Moreover, the ABO phenotypes that appeared associated with either a higher or a lower risk were not always identical across studies, and several studies failed to uncover any significant association [37–41]. Finally, if there is an effect of ABO blood groups on SARS-CoV-2

Viruses 2021, 13, 160. https://doi.org/10.3390/v13020160 https://www.mdpi.com/journal/viruses

Viruses 2021, 13, 160 2 of 20

infection or the outcome of COVID-19, its impact on the evolution of the pandemic in different regions of the world and its potential importance in terms of patients care remain to be assessed. The purpose of this work is therefore to provide a reasoned review of the literature in order to assess the extent of the potential role and usefulness of the ABO blood group system in the COVID-19 pandemic.

2. The ABO Blood Group System at a Glance

The ABO blood group system was discovered more than a century ago and its un- derstanding allowed the development of blood transfusion. However, the corresponding antigens are expressed on a large number of cell types in addition to erythrocytes [42]. They are of the carbohydrate type, constituting terminal motifs of either N-linked or O-linked chains of glycoproteins as well as of glycolipids. Their synthesis proceeds by addition of monosaccharide units to precursor glycan chains through specific glycosyltransferases. It first requires the synthesis of the histo-blood group H precursor antigen, which is catalyzed by alpha1,2fucosyltransferases that add a fucose in α1,2 linkage to a terminal β-galactose of the subjacent glycan chain. The FUT1 enzymes are responsible for this activity in ery- throblasts, megakaryocytes, vascular endothelial cells, and several other cell types, while it is the FUT2 enzyme that catalyzes synthesis of the H antigen in most epithelial cells such as the upper airways, the digestive tract, and the lower genito-urinary tracts. Once the H antigen is produced, addition in α1,3 linkage of an N-acetylgalactosamine or of a galactose to the same subjacent galactose unit by the A or B blood group enzymes generates the A and B antigens, respectively (Figure 1). The A and B enzymes are coded by distinct alleles of the ABO gene, whereas the O alleles correspond to null alleles unable to generate any active enzyme. These three major types of alleles generate the four major phenotypes O, A, B, and AB [43]. Both the FUT1 and FUT2 genes also present null alleles that lead to a lack of precursor H antigen synthesis in the corresponding cell types and therefore to a lack of A and B blood group antigens expression in these cells [44]. FUT1 null alleles are responsible for a rare red cell phenotype called “Bombay”. Given its rare occurrence, it will not be discussed any further. By contrast, null alleles of the FUT2 gene are common and their frequency varies across populations. These alleles are responsible for the so-called “nonsecretor” phenotype which by contrast with the “secretor” phenotype is characterized by a lack of A, B, and H antigens in many secretions such as saliva and in epithelia. In the Western world, the secretors represent around 80% of the population and nonsecretors, the remaining 20% [44].

In addition to its antigens, the ABO system is characterized by the presence of anti- bodies against the A and B antigens. Thus, blood group O individuals possess anti-A and anti-B antibodies, blood group A individuals possess anti-B antibodies, and blood group B individuals have anti-A antibodies. Only blood group AB individuals are devoid of both anti-A and anti-B antibodies. This system of antigens and their cognate antibodies defines the basic rules of transfusion where blood group O constitutes a universal donor, whereas blood group AB represents a universal receiver [45]. The origin of the natural anti-ABO antibodies is still debated. Nonetheless, it seems that most of these antibodies appear during the first year of life under stimulation of microorganisms either pathogenic or from the microbiota that carry similar antigens [46,47]. Their amounts are highly variable between individuals and some data suggest that they may decrease with improved hygiene conditions [48,49].

Viruses 2021, 13, 160 3 of 20

Figure 1. Description of the major characteristics of the ABO blood group system. Biosynthesis of the A and B antigens starts from a precursor structure constituted by a galactose residue in beta linkage to a subjacent sugar located at the termini of either N- or O-glycans as well as glycolipids. In red blood cells (RBC), vascular endothelial cells (VE) and other cell types such as megakaryocytes that give rise to platelets, addition of a fucose in α1,2 linkage by the FUT1 enzyme gives rise to the H blood group antigen. In most epithelial cells, synthesis of the H antigen is performed by the FUT2 enzyme. Blood group A antigen is then synthesized by the A enzyme coded by A alleles of the ABO gene, while blood group B antigen is synthesized by the B enzyme coded by B alleles. O alleles are unable to generate a functional enzyme; therefore, O/O individuals leave the H antigen unchanged. Relationships between genotypes, phenotypes, antigens, and the corresponding natural antibodies are shown.

The ABO gene and the FUT2 gene, which controls expression of ABH antigens in epithelia, are among the few human genes clearly under frequency-dependent balanced selection, suggesting important roles in interactions with environmental factors [50–55]. Histo-blood group antigens, including ABO blood groups, have previously been implicated in the genetic susceptibility to several infectious diseases, including viral diseases. This has been particularly well documented for human noroviruses and rotaviruses that together are responsible for the majority of gastroenteritis cases worldwide. These non-enveloped RNA viruses attach to the carbohydrate antigens expressed in the gastrointestinal mucosa. They have evolved so that distinct strains recognize preferentially different carbohydrate motifs, resulting in a strain-dependent susceptibility in accordance with the person’s blood type [56]. Rabbit Hemorrhagic Disease Virus (RHDV) is a highly pathogenic rabbit calicivirus related to noroviruses also attaches to blood group antigens expressed in the rabbit respiratory and gut epithelia. The rabbit-RHDV pair allowed the documentation of a natural example of host–pathogen co-evolution involving the recognition of A, B, and H blood groups antigens by the virus [57]. These co-evolving host–pathogen pairs explain, at least partly, the maintenance of the ABO polymorphism. Nevertheless, using mathematical modeling, it has been argued that its maintenance would additionally require the role of the anti-ABO antibodies to be taken into account [58]. It has additionally been argued that viral-mediated selection may explain why South American native populations are exclusively of blood group O [59].

3. Hypotheses Linking ABO Types and COVID-19 and Their Consequences on the Interpretation of Reported Associations

Soon after the first report of an association between ABO phenotypes and the risk of COVID-19, several hypotheses were proposed about the underlying mechanisms [60–63]. Prior to presenting the results of the various studies reporting associations, or the lack of

Viruses 2021, 13, 160 4 of 20

association, we will explain the bases and rationale of these hypotheses as they have distinct consequences that may explain discrepant outcomes stemming from variable designs of the association studies. Because of their crucial importance for blood transfusion ABO blood groups are generally perceived as red cell markers. Their potential role in COVID-19 is therefore not obvious. The mechanisms that have been proposed to account for their implication in the disease can be broadly divided into two major groups; those affecting the risk of infection and of SARS-CoV-2 transmission, on the one hand, and those affecting disease severity, on the other hand.

3.1. The Anti-ABO Antibodies

As described above, histo-blood group antigens, including the ABH antigens, are synthesized by many epithelial cell types, including those of the respiratory and digestive tracts that are known to emit large amounts of viral particles [64–67]. The SARS-CoV-2 main envelope protein, the Spike or S protein harbors many glycosylation sites. It is therefore heavily glycosylated and structural analyses of the recombinant glycoprotein produced in HEK293 cells revealed a large panel of glycans, mostly of the N-glycan type, but also of O- glycans [68–71]. As the enzymatic machinery required for their synthesis is that of the host cell, the exact structures present on virions will depend on the infected cell type. Glycans detected on the recombinant S protein from HEK293 cells or on the virus produced in Vero cells therefore do not fully represent those that are synthesized on viral particles emitted from either respiratory or digestive epithelial cells from infected persons. Most importantly they cannot be decorated by ABH antigens as HEK293 cells and Vero cells do not synthesize these epitopes [72]. To uncover whether the S glycoprotein carries A, B, or H antigens when produced in cells that have the ability to synthesize these glycan motifs, in a recent work, we produced the recombinant S1 domain in CHO cells expressing the FUT2 enzyme and either the A or the B enzymes. This showed that H, A, or B epitopes can be present on the viral S protein in accordance with the glycosyltransferase repertoire of the cells [72]. It is therefore to be expected that authentic infectious virions produced by respiratory epithelial cells also carry the antigens in all individuals of the “secretor” phenotype. Consequently, applying the rules of blood transfusion, SARS-CoV-2 viral particles transmitted in ABO incompatible situations might be neutralized by the anti-A and anti-B antibodies. As blood group O individuals possess both types of antibodies, they might benefit from a better protection than blood group A or B individuals who possess only one of these types of antibodies and even more so than blood group AB people who have none of them. As previously discussed, this kind of protection will only be partial as it can only take place in incompatible ABO situations and likely requires sufficient amounts of the anti-A and anti-B antibodies [60]. Because these are highly variable, it follows that individuals who possess low titers of anti-A or of anti-B antibodies are expected to be at a higher risk of infection than people with high titers. This was indeed recently observed in a study where agglutination scores of patients were compared to those of a control group. Patients’ anti-A and/or anti-B scores were lower than those of non-COVID-19 controls. By contrast, no differences were observed for antibodies directed against similar carbohydrates, including the αGal antigen (see below), which cannot be synthesized by human cells, indicating that the lower scores of anti-ABO antibodies observed in patients did not result from a general decrease of anti-carbohydrate antibodies following infection [72].

Earlier observations additionally support the anti-ABO antibodies hypothesis. Follow- ing an outbreak of SARS-CoV in Hong Kong hospital in 2003, it was observed that blood group O hospital staff members in contact with the initial patient had been largely spared by the disease in comparison with non-O blood group staff members [73]. It could then be showed using an in vitro cellular model that the interaction between the virus S protein and the ACE2 protein, its receptor like that of SARS-CoV-2, could be specifically inhibited by anti-A antibodies when the S protein was produced by cells able to synthesize the blood group A antigen [74]. Additionally, Measles virus grown in cells expressing either the A or B blood group antigens was neutralized by human serum anti-B or anti-A antibodies,

Viruses 2021, 13, 160 5 of 20

respectively, in a complement-dependent manner [75]. Likewise, the αGal xenoantigen, similar to the human B blood group antigen, but not present in humans due to mutations in the GGTA1 gene that encodes a closely-related specific α1,3galactosyltransferase, is present in enveloped virus particles produced by animal cells possessing the enzyme [76]. A large body of evidence indicates that expression of the αGal carbohydrate epitope on viral envelopes leads to the elimination of viruses through anti-carbohydrate natural anti- bodies. Various mechanisms appear to be involved in the phenomenon, including blocking of receptor engagement, complement-dependent neutralization and amplification of the specific immune response through targeting of the opsonized virus particles to antigen presenting cells (reviewed in [77]).

In light of all these data, the involvement of natural anti-ABO antibodies in COVID- 19 infection is a serious possibility that needs careful attention. It should be stressed that if these antibodies play any role, they can only act by preventing infection or by decreasing the viral load. As soon as virus replication has taken place in the new host, newly formed virions will carry autologous glycans that can no longer be recognized by the allogeneic anti-ABO antibodies. Another critical consequence of a potential protection effect of anti-ABO antibodies is that, as protection exclusively takes place in situations of ABO incompatibility, the final number of infected individuals cannot be affected to a great extent, because more and more ABO compatible encounters will take place as the epidemic progresses. Yet, it can still have a major impact on the epidemic by significantly slowing down its progression which will enhance the efficacy of non-pharmaceutical measures of protection, such as social distancing. This was modeled using data from a SARS-CoV 2003 outbreak [74]. The model showed that in case of a strong protection afforded by a sufficient amount of anti-ABO antibodies in ABO incompatible transmission events, the epidemic would be very strongly delayed.

3.2. The ABO Effect on Thrombosis

A large number of studies show associations between ABO blood groups and throm- boembolic diseases as recently reviewed [78]. This has been shown for myocardial in- farction, atherosclerotic vascular disease, venous thromboembolism, and cardiovascular ischemic events. In all instances, people with non-O blood groups proved at a higher risk than O blood group individuals. One of the explanations rests on the observation that plasma levels of coagulation factors, most notably von Willebrand’s factor (vWF), are ~30% higher in non-O blood group individuals than in blood group O. Synthesis of vWF takes place in megakaryocytes and vascular endothelial cells that express ABH antigens. As it is heavily N- and O-glycosylated, it carries the antigens depending on the person’s ABO phenotype. Its clearance, largely glycan-driven and mediated by lectins such as the Ashwell hepatic lectin or CLEC4AM, is reduced in the presence of either the A, the B antigen, or both, leading to higher plasma levels of non-O individuals. In addition to their effect on hemostasis, there is evidence that ABO blood groups also affect vascular function, although the exact underlying mechanisms are not fully elucidated. In this context, it is noteworthy that the levels of vascular adhesion molecules such as the soluble forms of ICAM, P-selectin, and E-selectin correlate with ABO blood groups, with higher levels of these factors detected in blood group A individuals in comparison with blood group O [79]. Regardless of the precise underlying mechanisms, these observations suggest that ABO blood groups modulate leucocyte–endothelial interactions and influence the magnitude of the inflammatory response.

Severe COVID-19 is characterized by an inflammatory state damaging the alveolar– capillary barrier and thereby compromising gas exchange. Intracapillary thrombosis and endothelial dysfunction are essential components of the severe form of the disease. Considering that ABO blood groups modulate both hemostasis and endothelial function, including its interactions with inflammatory cells, this has been suggested as an explanation of the reported associations between COVID-19 and ABO blood types [61,62]. Within the framework of this hypothesis, it appears that ABO blood groups would only influence

Viruses 2021, 13, 160 6 of 20

the outcome of the disease at a late stage when ARDS or severe lung dysfunction has taken place.

3.3. ABO Blood Groups and the Furin Cleavage Site

Cell entry of SARS-CoV-2 involves pre-activation of the S protein by the proprotein convertase furin or furin-like proteases [80]. The furin cleavage site is surrounded by O-glycosylation sites, which represents a unique feature of SARS-CoV-2 among coron- aviruses. As ABH antigens are largely present on O-glycans of epithelial cells [81], the distinct versions of O-glycans thus generated on the virions may impact the ability of furin to cleave the S protein. In a recent publication, Abdelmassih et al. also suggested potential relationships between ABO blood groups and furin [62]. They proposed that furin levels might be reduced in blood type O individuals based on a reported negative relationship between blood type O and furin-related proprotein convertases [82]. In addition, they suggested that furin levels modulated by the ABO phenotypes could play a role in the en- dothelial pathogenicity of SARS-CoV-2. In these conditions, the impact of ABO phenotypes on furin could take place both at the infection level and at the late stage of severe disease.

3.4. ABO Blood Groups and Susceptibility to Other COVID-19-Associated Risk Factors

Besides cardiovascular diseases, a large number of COVID-19 comorbidity factors have been described, some of which could also be associated with ABO blood groups. Although many studies looking for associations between ABO phenotypes and inflamma- tory conditions or autoimmune diseases have been conducted, they have often generated conflicting results. Nonetheless, a recent phenomic study involving a very large number of subjects from independent cohorts replicated some of these associations [83]. Thus, levels of C-reactive protein and alkaline phosphatase appear higher in blood group A individuals in comparison with blood group O, which may indicate a higher inflammatory state in the former group. Blood group O has also been reported to represent a protective factor for Crohn’s disease and ulcerative colitis, type I diabetes, and multiple sclerosis, although replications studies are lacking (reviewed in [62]). ABO blood groups additionally show associations with markers of the general metabolism. PSK9 levels appear higher in non-O blood group people, consistent with their higher levels of total cholesterol, LDL-C, and HDL-C [82,83]. Blood group A also appears associated with a lower forced vital capacity and forced expiratory volume in 1 s [83]. Although the mechanisms behind these various associations are unknown, they might contribute to increase the risk of severe COVID-19 or to worsen the disease.

3.5. ABO Blood Groups and the Microbiota

As bacteria of the microbiota trigger the synthesis of anti-A and anti-B antibodies, composition of the gut microbiota might be critical to explain the large inter-individual variations in levels of these antibodies. The microbiota is also known to play a major role in controlling immunity and inflammation and since recent reports showed that the gut microbiota signature of COVID-19 patients appears distinct from that of healthy controls (reviewed in [84]), it is worth mentioning two studies that related ABO phenotypes to the composition of the gut microbiota. Mäkivuokko et al. observed that the overall microbiota composition of blood group B individuals differed from that of the other blood groups and showed increased levels of Actinobacteria in group A healthy individuals [85]. These bacteria are reportedly increased in Crohn’s disease and ulcerative colitis and may contribute to facilitate the development of the inflammatory state associated with severe COVID-19. Another study reported lower levels of Blautia in blood group A secretors than in non-A secretors [86]. A decreased level of these bacteria had previously been described in several inflammatory, autoimmune conditions, and aging [87], it may also reveal a higher risk of uncontrolled inflammation among blood group A COVID-19 patients than among other ABO blood types.

Viruses 2021, 13, 160 7 of 20

Strikingly, all hypotheses linking ABO blood groups to COVID-19 predict a protective effect of the O blood type in comparison with non-O blood groups. Disentangling their relative importance is therefore not straightforward. Nonetheless, those that concern susceptibility to infection and the likelihood of transmission have different consequences than those predicting an impact on late events associated with the cytokine storm, ARDS and severe lung dysfunction. These differences help in analyzing the reported studies as discussed below.

4. Studies Linking ABO Blood Types to COVID-19 4.1. Case–Control Studies Designed to Observe Associations

At the time of writing, 34 case/control studies reporting associations between ABO phenotypes and the risk of COVID-19 have been reported, while only four studies failed to uncover a significant association (Table 1). Importantly, not all of the reports have been peer-reviewed as yet. As immediately apparent from Table 1, the data show a lower risk for O blood group people and/or inversely a higher risk for people with non-O blood types, most often blood group A. The reported significant odd ratios (ORs) generally revealed a modest influence of the ABO blood groups, although some reports suggested stronger effects. Thus, ORs for blood group O ranged from 0.53 to 0.9 and ORs for non-O blood groups ranged from 1.12 to 3.7. Meta-analyses of the earliest available data have already been conducted, confirming the effect [88–92].

The majority of these studies were hypothesis-driven. Yet, six of them concerned genome-wide association studies (GWAS) that looked for genetic associations without a priori [5,20,23,26,27,37]. Of the latter, only one failed to observe a significant signal at the ABO locus on the long arm of chromosome 9 (9q34) [37]. The fact that the majority of both agnostic GWAS and hypothesis-driven studies found the same significant associations with the ABO blood types indicates that these did not derive from biases due to the a priori search for a link between ABO blood groups and COVID-19.

The various studies also differ in many other aspects, including the number of patients included, the definition of cases and the types of controls to whom patients were compared, as well as the relative ABO frequencies in the diverse studied populations. These different settings can seriously affect outcomes. Thus, in studies where the patients’ group was compared to blood donors, a bias might be introduced as blood group O is over-represented among regular blood donors because group O blood representing the universal donor is more widely demanded. In such studies, there is a risk of finding a false apparent increase in non-O blood types or a decrease of the O type among patients. Nonetheless, other studies using anthropological data of the frequency of ABO phenotypes obtained from large fractions of the local population alleviate this potential bias, as long as the population is sufficiently homogeneous and has not changed in composition since the acquisition of the anthropological data. In this respect, data from Asian and Middle-East countries are quite safe at the local level, with population ancestries being rather homogeneous. This is not so in the USA or European countries. Higher risks of COVID-19 have been reported for some disadvantaged subgroups from these countries [93]. The variations in ABO blood group frequencies between populations of different geographical origins and ancestries represent another important source of potential bias. In order to take this bias into account, in several studies, groups of patients and controls were stratified for ancestry [7,20,21,26,37]. Again, with the exception of one study discussed below [37], they consistently documented a lower risk of COVID-19 for blood group O, although not visible in all ethnic groups as illustrated by the study of Leaf et al. who uncovered a significant ABO effect in the white American population, but not in minorities of African Americans or Latin Americans [7]. This is likely due to the impact of the relative frequencies of ABO phenotypes in populations as discussed below. The effect of population admixture also likely explains why a study from New York, USA, failed to reveal any significant association between ABO blood types and the risk of COVID-19 [38]. In that study, no stratification for ancestry was performed even though the control group was composed of historical non-COVID-19 hospitalized patients.

Viruses 2021, 13, 160 8 of 20

Table 1. Case–control studies reporting data on ABO blood groups and COVID-19.