PETER J. D'ADAMO 

Copyright 1989, 2001, 2004, 2009 Peter D'Adamo. All rights reserved.

Originally published in The Townsend Letter for Doctors, September 1990  

The first description of a human blood group system was published by Landsteiner in 1900, working to understand the unpredictability of hemolytic reactions resulting from early attempts at transfusion. Using the newly discovered lectins abrin and ricin, recently isolated by Stillmark, he was able to describe classically what has still remained the major blood group of clinical interest. Many other blood grouping systems based either on membrane or sera antigen  antibody interactions have also been discovered. The most clinically relevant of these are the MNS and Lewis antigen systems. It is estimated that there are in excess of 400 blood type antigens now known.

Immunochemistry

There are now set parameters which determine if an antigen possesses blood group activity. Most blood group antigens are carbohydrates extending outward from membrane bound glycolipids and glycoproteins. These membrane glyco conjugates are typically rich in sialic acids whose high degree of hydrophilicity and negative ionic charge results in their projection outward from the cell membrane. Sialic acids do not appear to have a role in ABO blood group specificity, but high or low concentrations may enhance or interfere with the expression of blood group activity.

The ABO and Lewis systems possess a common basic structure and their individual specificity is determined by the sequence and linkage of sugars at the end of the carbohydrate chains. It has been estimated that of the oligosaccharides projecting from the cell surface, 100 per 300,000 bear blood group antigenic determinants.

There are two types of backbone structures: Type I chains, which contain galactose linked 6-(1-3) to N-acetylglucosamine, and Type 2 chains where the linkage is 13-(1-4). Oligosaccharides with these terminal ends do not possess blood group specificity, but can and do cross react immunologically with many bacterial polysaccharides.

The ABO System

These antigens are synthesized from an oligosaccharide intermediate, H substance, which is produced by the presence of the monosaccharide fucose on either Type 1 or 2 chains. Group A or B activity is produced by the addition of a single sugar on the non­reducing end of H chain. The addition of this sugar markedly reduces the reactivity of the H substance. Adding the glycoprotein N-­acetylgalactosamine to the end of the chain results in blood group A antigenicity, whereas with blood group B the terminal carbohydrate and B group antigen is the monosaccharide galactose. There is no O antigen: group O cells contain the H antigen, but the designation group H has been main­tained for historical reasons. The terminal carbohydrate of O(H) antigen is the mono­saccharide fucose.

ABH Secretor system

It was first shown by Lehrs in 1930 that some people do and others do not secrete into their saliva antigens corresponding to their ABO blood group. Sakes found that the abil­ity to secrete behaved as a simple Mendelian function dominant to non-secretion. Group A, B, and AB persons who are secretors secrete the antigens corresponding to their blood groups. Group H persons who secrete the H substance, as do all other secretors; to a somewhat less extent. The secretor gene is identified as Sec to distinguish it from the Ss blood group.

It was long ago discovered that genetic secretors secrete their blood group antigens not only into their saliva but in numerous

The Lewis System

The ABH and Lewis glycoproteins possess a common basic structure and their blood group specificity is determined by the sequence and linkage. There are two Lewis antigens, termed Lea and Leb. The presence of fucose linked to C­4 of N-acetylglucosamine on a Type 1 chain results in Lea activity, but a Type 2 oligosaccharide containing fucose linked to C-3 of N-acetylglucosamine on a Type 2 chain results in very weak Lea activity. The appearance of a second fucose on a type one chain results in the appearance of a new antigenic determinant, Leb, and the loss of most H and Lea antigenicity. A Type 2 difucosyl chain has very weak Leb activity.

The MNS System

The genetics of the system seem to imply that the N antigen is actually a precursor substance and the N gene is an amorph which leaves the N anti­gen unchanged while the M gene of the heterozygote converts part of the N antigen into M, and in the homozygote converts nearly all of the precursor to M. The MN antigens seem to have a direct interaction with membrane bound sialic acids, as M and N specificities seem to be linked to the presence of sialic acid variations. It has been suggested that the antigenic variations may result from specific sialyl transferase activi­ties, which transfer sialic acids to disaccha­rides bearing specific T and Tn specificities that characterize specific cryptic antigens. Mourant considers this blood group of interest only to the geneticist, due to a lack of disease association, however several diverse associations have surfaced including: an association of ankylosing spondylitis with homozygous MM (Sharon) and an association between heterozygous MN and homozygous NN with environmental induced hyperlipidemia (Martin). Cruz et. al. studied the tendency of Easter Islanders to become "hypertensive" upon moving to the mainland and concluded that homozygous NN was significantly more liable to develop hypertension. These are interesting disease associations, yet probably result more from co-dependent alleles than from a direct membrane bound antigen interaction.

Rhesus System

Rh incompatibility is the major cause of hemolytic disease of the newborn, however very few searches have been made for any other kinds of disease associations of the Rh groups. Rh- status is largely a European gene (40%), with its highest frequency among the Basques of Spain, a remarkably homogenous group, originally late paleolithic or early mesolithic inhabitants of the Pyrenee Mountains. The U.S population is approxi­mately 15% Rh-  and 85% Rh+.

Duffy System

The Fya and Fyb antigens were discovered in the 1950's. Sanger dis­covered that a high percentage of African Negroes are of the phenotype Fy (a- b-), which is apparently a third gene termed Fyx which does not react with anti-Fya or anti ­Fyb. In 1975 Miller was able to show that this type is probably specifically resistant to Vivax malaria, to which Africans have long been known to be resistant.

P System

The PI antigen is found in hydatid cyst fluid, and in a considerable variety of worm, both parasitic and free living. It is probably not uncommon that anti-P1 is not infrequently found in the sera of humans in response to worm infestation. The distribu­tion of the p2 allele runs north to south and way from 180 degrees E. The highest fre­quencies reported by Blangero for the P2 allele are Circumpolar people, Oriental, and Pacific Islanders, people with a tradition of fish eating, reindeer and caribou hunting/ herding and aquatic mammal diets. Thus the high P2 allele may result from ensuing helminthiasis.

I/i Antigens

These antigen scarcely qualify as a blood grouping system. Anti-I anti­bodies are common, yet more common are anti-IH antibodies which is distinct from I or H but is presence on cells containing both. Anti-I is a cold agglutinin, so termed because its activity is enhanced at low temperature. This antibody is often seem in the serum of patients with infectious mononucleosis.

Other Polymorphisms

Other human ge­netic variations of clinical interest include: the ability to taste phenylthlocarbamide (associated with certain thyroid susceptibilities), ear wax types (associated with carci­noma of the breast), aryl hydrocarbon hydoxylase (lung cancer), the Group Specific Component globulins (vitamin D transport), protease inhibitors, protease inhib­itors, G6PD and a variety of hemoglobins.

Paleoserology of the ABO Groups

Racial Characteristics

Ottenberg first at­tempted racial classification based on blood groups. However the limited information available at that time made for strange bedfellows (including a "Hunan" group composed of Japanese, Southern Chinese, Hungarians and Rumanian Jews). It was not surprising that anthropologists took one look at the list, shuddered, and said, in effect, "No thanks; I'll take vanilla."

A more recent attempt by Lavory, who was aware that racial classification based merely on ABO groups would in many cases give results which would not fit well with older ideas about race. He also incorporated M and N types and occasional other blood factors to distinguish populations not clearly differentiated by A and B. He distinguished the following races: 1) Europeans (Nordics and Alpines of Europeans of the Near East); 2) Mediterranean; 3) Mongolian (Central Asia and Eurasia); 4) African (Blacks); 5) Indonesian; 6) American Indian; 7) Oce­anic (including Japanese); 8) Australian (a sub variety of Oceanic). Lavory failed how­ever to realize that the characteristics needed to define race, such as morphology or skin color are independent of each other.

Wiener has proposed the following racial classification, based largely on ABO and Rh factors: 1) Caucasoid group (highest inci­dence of Rh-, relatively high incidence of genes for Rh- and A2, moderately high incidence of all other types); 2) Negroid group (highest incidence of RhO, moderate fre­quency of Rh-, high relative incidence of genesA2 and the rare intermediate A and Rh genes); 3) Mongoloid group (virtual absence  of Rh- gene and gene A2). Using MN data, he then further classified the Mongoloid group into an Asiatic group, a Pacific Island and Australian group, and a group including American Indians and Eskimos.

Phylogenetics

Group O is almost universally considered the original blood group, or at least the blood type present as the overwhelming majority in 'ancient' or 'isolated' peoples. This seems to bear out well as the O(H) antigen is the precursor to both A and B.

It has been an often stated assertion that all full-blooded Amerindians are group O, Wyman and Boyd, in ingenious fashion, were able to blood type remains of early prehistoric Amerindian (Basket Maker and aboriginal) remains, finding sufficient group A to cast some doubt on this. Nonetheless, even recent studies on largely intermingled Amerindian populations shows a very (67-80%) predominance of group O, implying that their migration was perhaps earlier that previously thought, and most definitely seems to be a paucity of group B. Additionally, it has been shown that introduction of these genes into the population does result in mutation rates much higher than expected. Studies on Egyptian mummies shows the group B was fairly well distributed (if, as some have asserted, it comes from India) in Egypt at least 5,000 years ago. These serological variations in race apparently can not be explained simply by mutation; selection, migration and ran­dom genetic drift must also be accounted for.

Blood group A seems to have followed significant numbers only after group O, appearing during the lce Age (30,000 to 100,000 years ago), and perhaps representing an adaptation to cold. Its minor subtypes A-intermediate, Ax and A-Bantu seem to be adaptations to parasitic infection.

Group B would seem to have reached appreciable numbers last. Lewontin has conjectured that the complete absence of the gene in Amerindian and Australian aboriginal populations suggested that its origin would have occurred after the rise in sea levels that accompanied the melting of the continental glaciers, 10,000 years ago.

Group AB (a product of A and B parents) seems to be a recent phenomenon also. Studies on prehistoric grave exhumations in Hungary showed a distinct lack of this blood group into the Langobard age (4th to 7th century AD). Populations in Europe also seem to differ serologically from most other populations of the world, except perhaps the African. There !sat present time no other human (or anthropod) group with proportionally more A2 than the average European. This gene has been speculated to perhaps convey some disadvantage, which under more stringent selection in Asia was eliminated. In Paleolithic European man the gene, although perhaps inferior to A1, lingered on.

Natural Selection

There is good evidence for considerable effects of selection on blood type distribution. Although most recent work has centered on malignant or degenerative disease associations between the ABO groups, infection undoubtedly ac­counted for the majority of natural selection in prehistoric populations. This can be ex­plained by the phenomenon of "horror auto­ toxicus": i.e. the body has an inherent aver­sion to producing antibodies to self anti­gens. Thus group AB which produced no antibodies with either A or B specificity would be at selective disadvantage to organ­isms possessing either A or B antigenicity.

Livingstone has pointed out the inherent fault in a simple mutation theory of blood group distribution: Given the time frame, these mutations would have had to occur in humans at a rate four times faster than Drosophila!. However as will be seen, the effect of selection via infectious disease on small populations (with added random genetic shift) does explain the blood group variation in a far shorter time frame.

Infectious Disease Associations and ABO Group

Mourant was the first to hypothesize that the relatively high distribution of A in areas with historically high incidence of plague (Turkey, Greece, Italy) would point to selec­tive disadvantage for group O, proven immunochemically by blood group studies on various Yersinia species. Antigenic similarity exists between the blood groups and a great variety of bacterial, rickettsial and helminthic species, including: typhoid, streptococci (group A), staphylococci (group O), Shigella and Proteus. Blood group A antigen is virtually identical with Pneumoccus polysaccharide antigen, which would suggest an association between group A and this organism, which indeed does exist. The generation of isoag­glutinins via pneumoccocal vaccination was so great (fourfold) that the manufacturers advised doctors not to administer the vac­cine to premenopausal women, fearing hemolytic difficulties in ABO incompatible pregnancies. Urinary tract infections have shown great ABO correlation. Ratner showed that anti-B agglutinogens provided greater protection against UTI than anti-A, and was able to demonstrate a significantly greater incidence of UTI from Pseudo­monas, Kelbsielia and Proteus in group B. Group O which does produce anti-B, but in much lower titers than group A was, of allergic dermatosis and tropical eosinophilia show high group A frequencies. Damian gives many examples of blood group-like antigens in parasitic worms, especially of A and B-like ones such as Ascaris lumbricoides.

Other Blood Types

There is a general paucity of information on blood groups other than ABO. Paciokiewicz showed a general deficiency of Rh+ in mumps, infectious mononucleosis and viral meningitis. It is interestingly to note that viruses tend to attack the non-antigenic types of both the Rh and ABO systems (O and Rh-), respectively. Mourant mentions an association between granulomatous disease and the Kell system, which explains the recognized autosomal dominance noted with this syndrome.

Bacterial sensitization

Several studies imply that the development of isoagglutinins results from cross sensitization between bacterial polysaccharides and immature gut wall of the infant. One study shows a persistent sensitization of infant red cells resulting from E. Coli enteritis.

Other Disease Associations and ABO Groups

Perhaps with the exception of Giardiasis, there have been no significantly new disease associations noted between the ABO groups since my 1981 review article. However the classic studies will be reviewed again here.