Proposed O-GalNAc/Gal glycosylation path- ways in blood group O and non-O blood group phenotypes during Plasmodium falciparum in- fections driving evolution

The coevolution of species drives diversity in animals and plants and contributes to natural selection, whereas in host–parasite coevolution, a parasite may complete an incomplete evolutionary/developmental function by utilizing the host cell’s machinery. Analysis of related older data suggests that Plasmodium falciparum (P. falciparum), the pathogen of malaria tropica, cannot survive outside its human host because it is unable to perform the evolutionarily first protein glycosylation of serologically A-like, O-GalNAcα1-Ser/Thr-R, Tn antigen (“T nouvelle”) formation, owing to its inability for synthesizing the amino sugar N-acetyl-d-galactosamine (GalNAc). This parasite breaks the species barrier via hijacking the host's physiological A-like/Tn formation through abundantly expressing serine residues and creating hybrid Alike/Tn structures, which in the human blood group O(H) are attacked by the germline-encoded nonimmune polyreactive immunoglobulin M (IgM), exerting the highly anti-A/B/H-aggressive isoagglutinin activities. These activities physiologically undergo the ABO(H) blood group phenotype formation, occurring on the surfaces of red blood cells (RBC), epithelial and endothelial cells and on plasma proteins by identical glycosylation, performed by the ABO(H)-allelic glycotransferases, phenotypically downregulating the anti-A/B/H-reactive IgM (isoagglutinin) activities in the non-O blood groups. ABO(H) phenotype diversity, this way glycosidically linked and molecularly connected to humoral immunity, becomes exposed to the evolution.

and their analysis of a SARS-CoV-1 outbreak in Hongkong in 2003, demonstrated that blood group O(H) was associated with low risk of infection, while the interaction between the viral S protein and the host cell receptor was inhibited by natural and monoclonal anti-A antibodies in vitro [12]. Moreover, the actual and first statistical study, performed during the current SARS-CoV-2 pandemic, indicated that individuals with blood group A have a significantly higher risk of acquiring SARS-CoV-2 or COVID-19 infection, whereas people with blood group O have a significantly lower risk of infection compared to non-O blood groups [13]. Finally, a hypothesis regarding how SARS-CoV-2 invades the human body via ABO(H) blood group carbohydrate formation, and why blood group A individuals are at higher risk, was proposed by the author [14].
According to a glyco-evasion hypothesis, a pathogen can alter host immune systems by hijacking the glycosylation pathways [15], while evidence shows that parasitic helminths (worms) coevolve with vertebrate immune systems [16].
The pathogen P. falciparum, which is transmitted to human hosts through the anthropophilic mosquito Anopheles gambiae [17], does not have the cellular machinery to complete protein glycosylation independently and hypothetically breaks the species barrier by abundantly expressing serine residues [18] for hijacking host glycosylation, specifically the (serologically A-like) O-GalNAcα1-Ser/Thr-R or "T nouvelle" (Tn) [1] antigen formation process. Although several glycans can be O-linked to Ser/Thr residues, O-GalNAc glycans appear to be especially relevant to mucosal sites and other locations of parasitic invasion [19]. The "bulky" [20] polyfunctional GalNAc molecule is a crucial component of glycosylation, an evolutionarily conserved pathway, which dominates critical functions during metazoan growth and reproduction and may even be the target of P. falciparum in a recently observed inverse relationship between global malaria incidence and cancer mortality [21].

Metazoan protein O-glycosylation and human ABO blood group phenotype formation: A hypothetical basis of P. falciparum infection and malaria tropica disease.
In metazoan evolution, up to 20 distinct genetically undefined polypeptide O-GalNAc transferases catalyze the first addition of GalNAc to a protein [22] and synthesizes the above ( serologically A-like) O-GalNAcα1-Ser/Thr-R Tn [1], which results from the most complex and differentially regulated step in protein glycosylation [ and acts as the complementary protein of the A-like/Tn antigen. The primarily pentameric structure of this mammalian antibody gives rise to speculation regarding an evolutionary relationship with the hexameric structure of the anti-A-reactive Helix pomatia agglutinin [28], a primitive invertebrate defense protein, emerging from the coat proteins of ferti-7 1971 [39], indicated an X chromosome/germline-encoded origin for the murine nonimmune anti-A-reactive IgM, although the complementarity of this antibody protein with syngeneic Alike ovarian glycoproteins and/or glycolipids became evident in subsequent studies [40] [41].
Murine and human nonimmune IgM exhibit a serologically similar reaction with human red blood cells (RBCs) and human anti-A and anti-B isoagglutinin activities have been attributed to anti-A/Tn and anti-B/T antibodies owing to their cross-reactivity with Tn and T structures [42]. During the establishment of the human ABO blood group phenotype these innate isoagglutinin activities, exerted by the nonimmune IgM, are downregulated through phenotypic glycosidic accommodation of plasma proteins in the non-O blood groups A, B, and AB [32] [33].
Because of donor-substrate availability, protein glycosylation is thought to occur intracellularly in the Golgi cisternae, in which nonimmune immunoglobulin production may be adapted to ABO phenotype formation. Accordingly, during immunoglobulin secretion, the poorly glycosylated intracellular IgM molecule becomes loaded with L-fucose and D-galactose [43], and the subsequent expression of fucose and galactose residues by the extracellular IgM might form the basis of phenotypic accommodation. This process may also be performed and/or completed by soluble plasma glycotransferases: blood platelets, for example, have been detected as a rich source of both glycosyltransferases and energy-rich sugar and amino sugars that are released from activated platelets to function in the extracellular space [44] [45] and potentially contribute to the phenotypic accommodation of plasma proteins. A single O-glycosidic enzymatic step may create an A, B, or AB mucin-type cell surface epitope and release a secretory IgM, lacking the corresponding anti-A/Tn or anti-B/T isoagglutinin activities, whereas the blood group O(H) maintains these activities.
The lack of any ABO blood group glycosylation or phenotype formation, as shown by the rare O(h) or Bombay type [46] [47], which originates from consanguinity, is associated with strongly elevated isoagglutinin levels and an unusual anti-H reactivity, acting over a wide range of temperatures, with a thermal amplitude at 37 °C. This overexpression of anti-glycan antibody activity might cause autoimmunological impairment of germ cell maturation or function, responsible for the male infertility of this group [47]. In contrast, the blood group AB constitutes (likely to a result of evolutionary selective diseases) the smallest among the human ABO blood groups, precluding any isoagglutinin formation and representing the other extreme of phenotype diversity. Thus, the O(h) or Bombay type and blood group AB appear to mark two opposite directions of negative (natural) selection and demonstrate how phenotype and isoagglutinin production form evolutionary functional unity, wherein the degree of phenotype diversity and   glutinin activity, implicating secondary adaptive IgG production. In blood group A, the appearance of the ancestral anti-A activity is reduced or excluded by human-specific A-allelic GalNAc glycosylation, independent of classic clonal selection, hypothetically allowing the conversion of synthesized glycoconjugates into phenotype-specific plasma glycoproteins and/or molecular complexes that become subject to internalization [48]. The graphic (with minor changes) was constructed based on Fig. 4 in the manuscript by Arend (2017) [33]. group AB appear to mark two opposite directions of negative (natural) selection and demonstrate how phenotype and isoagglutinin production form evolutionary functional unity, in which the degree of phenotype diversity and innate immunity are inversely proportional. Because statistical data are not available from the Bombay type due to its small population size, the isoagglutinin levels are estimated according to existing reports and the hypothetical serological profile of the classic Bombay type (h/h; se/se) [46], which is characterized by the complete lack of ABO blood group glycosylations.

Why life-threatening malaria is especially prominent in non-O blood groups.
Brooks and Mclennan (1992) [49] argued that P. falciparum, transmitted to humans by Anopheles gambiae, has no alternative vertebrate reservoir hosts. We know that this parasite does not have the cellular machinery to complete protein glycosylation independently because it lacks the ability to synthesize the amino sugar GalNAc and to glycosate O-GalNAc, while intriguingly expressing abundant serine residues. Moreover, considering that malaria tropica infections occur regardless of the ABO blood group but life-threatening disease develops almost exclusively in non-O blood groups, the human glycoproteome must play a critical role in these infections. Indeed, these infections occur through two genetically distinct glycosylation pathways of the amino sugar GalNAc: 1) non-somatic,

blood group-independent, serologically A-like or intermediate Tn-reactive O-GalNAc-glycosylation
and 2) somatic glycosylation and formation of human mucin-type ABO-allelic structures or phenotypes.
As explained earlier, after the transmission of infectious P. falciparum sporozoites, their characteristic and abundantly synthesized serine residues and/or phosphorylated serine residues [18], arising throughout the parasite ' s life -cycle, access the human A-like intermediate Tn-reactive O-Gal-NAc-glycosylations. P. falciparum reaches every tissue, with the liver, bone marrow, and erythrocytic stages being the most critical stages during the life cycle progression. However when syngeneic glycosylation of peptides mainly occurs intracellularly, hybrid glycosylation or hybrid, mucin-type ABO blood group antigen formation may occur extracellularly, enabled by abundant, soluble glycotransferases and energy-rich monosaccharides released from activated platelets [45], which play a dominant role in malaria pathogenesis [50].
The complex invasion mechanism utilized by P. falciparum has been comprehensively explained in a recent study by Belachew (2018) [51]. In humans, this invasion implicates a subsequent molecular step for the access of the parasite to the mucin-type version of the ABO allele-specific phenotype formation, which occurs via both intracellular and extracellular O-linked glycosylation in epithelial and some endothelial tissues. It is noteworthy that expression of A-and B-transferring glycotransferases is independent of the secretor status [52] [53] and that the mucin-type expressions of A, B and (OH) mucin-type glycans are exclusively dependent on the A, B and O(H) genotype [54], regardless of the Lewis type [55]. Glycophorins, which are the intrinsic components of the RBC membrane, provide abundant O-linked [56] [57] blood group A-, B-and Tn-reactive [58] oligosaccharides and appear to be identified as pathogen receptors and merozoite ligands of P. falciparum [59]. These hypothetical interactions are consistent with earlier observations by Barragan et al. (2000) [60], suggesting that soluble blood group antigens can inhibit the binding between the parasite and human RBCs. Appropriately enough, blood group A and B trisaccharide haptens inhibit parasite adhesion to the RBC surface, although notably stronger inhibition can be achieved with the synthetic disaccharide Galβ1-3GalNAcα1-O-Ser/Thr or T antigen coupled to bovine serum albumin [61]. Intriguingly the main rosetting ligand and recognition protein of the parasite, called P. falciparum erythrocyte membrane protein 1 (PfEMP1), appears to even discriminate between blood group A qualities or subgroups [62] and thus demonstrates the role of this protein for infection and blood group-dependent severity of malaria disease. The role of fucosylation in P. falciparum infections remains elusive; O-fucosylation performed by the parasite was recently described [67] and the involvement of host-provided hybrid fucosylations has been suggested by this author in a recent review [2]. Human-specific α1,2 L-fucosylation (FUT1/FUT2) represents the basis of ABO-blood-group phenotype formation on the cell surfaces and plasma proteins, and the O(H) type appears to stand for controlled or physiological self-reactivity  (Fig. 4).
Furthermore, ABO mucin-type generation [69] involves fucosylation (Fig. 3), which protects from autoreactive anti-H activity during phenotypic accommodation of plasma proteins [32]. This mechanism may also enable the hybrid fucosylation between the parasite and blood group O(H) RBCs, which thus have the least contact with the parasite when compared with those of non-O blood groups (Fig. 3). Moreover, without A and B blood group-determining glycotransferases, O(H) can maintain nonimmune IgM and adaptive secondary IgG against syngeneic and hybrid A and B antigens, making it the most protected blood group (Fig. 3). This special immunological position of the O(H) phenotype was hinted at in an early study [70]. It has been discussed recently [2] and again becomes evident: innate IgM, which physiologically regulates the expression of the syngeneic intermediate A-like/Tn antigen in blood group O(H) [32] [33], will attack the formation of hybrid A and B formations, which are based on foreign (parasitic) peptides, and the quality and extent of this interaction should become a topic in future experiments. Blood type A phenotype formation merely maintains nonimmune anti-B-reactive IgM and blood type B maintains nonimmune anti-A/Tn reactivity that might protect against hybrid Tn formation but does not affect the formation of T and B cross-reactive hybrid glycoproteins; moreover, although most glycans are cross-reactive, it is important to note that in the non-O blood groups A and B the levels of these nonimmune IgM activities appear reduced due to A/B cross-reactivity: neutralization of innate nonimmune anti-A also affects the levels of nonimmune anti-B and vice versa. The blood group AB enables the strongest contact with the pathogen, and molecularly precluding any isoagglutinin activity, makes this group the least protected and smallest among the ABO-blood groups (Fig. 3).
Although the formation of hybrid carbohydrates is accomplished by circumventing innate immunity, the adhesion of these structures to parasite proteins ultimately makes non-O blood groups an immunological target. The physiological lack of innate anti-A and anti-B antibodies poses the following immunological dilemma in these blood groups: on one side, it protects these blood groups from self-reactivity, but on the other side, it cannot prevent the formation of autoantigenic targets in subsequent pathogenic steps during malaria infection. This is evident in recent observations [71][72], wherein the induction of autoimmune processes contributes to the development of severe malaria disease, most likely even dominated by autoimmune inflammations.

Figure 4. The adhesion of the parasite to O(H) RBCs is initiated blood group-independently via
fleeting hybrid A-like/Tn formation and is completed by mucin-type fucosylation. This process does not affect anti-A/Tn reactivity of nonimmune IgM, which is implicated in the control of syngeneic and the strong interaction with hybrid A-like/Tn expression. Accordingly, pathogen adhesion to A blood group RBCs occurs through fucosylation-driven A-allelic mucin-type formation. This process precludes adaptive and innate antibody reactivity against both syngeneic and hybrid A-like/Tn antigens due to clonal selection and phenotype-associated plasma glycosylation. In these conditions, nonimmune IgM is an adhesion molecule. This figure expands on Fig. 2 of a previous publication [2].

Conclusions
My previous paper proposed self-destructive biological altruism of the host [2], which enables the survival of the malaria parasite P. falciparum. This phenomenon has now been identified as a defined molecular biological or biochemical event, involving two genetically different GalNAc glycosylation steps: in the first step pathogen transmission occurs regardless of the blood group across species barriers via hybrid syntheses or molecular mimicry and shows another evolutionary function of Tn. The chemical simplicity of this antigen does not stand for antigenic and functional unity Cancer mortality may statistically interfere with malaria rates to some extent. In recent retrospective studies, performed on two different Asian populations, cancer mortality was highest in blood group A [82] [83]. It is not surprising that non-O blood groups are the ideal target of P. falciparum owing to comprehensive presentation of both A-like/Tn and A/B allelic glycosylations, whereas the most ideal target should be the early cancer cell, generating a surplus of genetically undefined A-like O-GalNAc glycans, which might explain the above cited, recently observed inverse relationship between global malaria incidence and cancer mortality [21]. According to this report, cancer mortality was significantly reduced in malaria-endemic regions. This phenomenon awaits validation by studying the survival patterns of human tumors transplanted into non-primate animals with insufficient innate immunity or severe combined immunodeficiency, for example the castaneous mouse (CAST) [84] or SCID mouse [85], infected with P. falciparum. The insight into the interaction mechanisms of this parasite with cancer metabolism might provoke speculations on completely new, future therapeutic strategies for cancer. Intriguingly, a glycosaminoglycan binding malaria protein, targeting human cancer has been discovered by Salanti et al. [86] Although P. falciparum cannot survive in non-primate animals, most likely resulting from innate immunity, in humans, infections occur regardless of the blood group. P. falciparum parasitemia does not differ between non-O and O(H) blood groups in malaria-endemic areas [87], lifethreatening infections, however, are more frequently diagnosed in non-O(H) phenotypes [88] [89], wherein the number of erythrocyte rosette formations reflects the severity of the disease [90] [91].
Clearly, blood group phenotypes determine innate immunity: over the decades several concepts have been developed, in which the pathogen evades host immunity, and depending on the experimental approach and/or the interpretation of existing data, different cellular and molecular aspects of blood group A-favored infection become evident: in a model by Moll et al (2015) [92], the patient's RBC surrounds itself with normal RBC, which makes the pathogen's antigen inaccessible to yet undefined antibody activity and clearance by the immune system, whereas in my concept the pathogen utilizes the physiological absence of the corresponding antibody or phenotype-specific isoagglutinin, which has been neutralized by A, B or AB phenotypic glycosylation. ABO epitopes and innate isoagglutinin activities arise molecularly together during the same enzymatic process and represent a developmental and evolutionary unity. The overexpression of isoagglutinins in the Bombay type and their complete downregulation in blood group AB mark the opposite ends of the spectrum of ABO phenotype diversity in natural selection, whereas the central immunological position of the human blood group O(H) guarantees the maintenance of species and phenotype diversity via a mechanism of immunologically-controlled anti-self-reactivity during growth and reproduction [42]. Consequently, the blood group AB is the least protected and remains the smallest among the ABO blood groups. In contrast, O(H) individuals have the least molecular contact with the pathogen, rarely develop severe disease, and survive this coevolution in an immunological balance with the pathogen as the largest blood group worldwide. While the significance of host -parasite relationships such as host identification of foreignness, host susceptibility to parasites, and antibody responses have been discussed for decades as causes of the present-day world distribution of human ABO blood groups [93] [94], the coevolution of humans and P. falciparum [74] as well as respective virus pandemics most likely contributed to a large extent to this distribution, explainable on both a molecular and immunological basis and demonstrating how augmenting ABO blood group phenotype diversity is associated with decreasing immunity.