Blood Group, P1pk System

A number sign (#) is used with this entry because the P1PK blood group system is determined by the A4GALT gene (607922) on chromosome 22q13.

A distinct, but related antigen, P, which belongs to the globoside (GLOB) system (615021), is defined by activity of the B3GALT3 gene (603094) on chromosome 3q25. B3GALT3 synthesizes the P antigen from the P(k) antigen.

Clinical Features

Different combinations or absences of 2 antigens, P1 and P(k), (see BIOCHEMICAL FEATURES) define 5 different P blood group system phenotypes: P(1), P(2), P(1)(k), P(2)(k), and p. The P(1) red cell phenotype is defined as P1+ and P+ antigens and is the most common phenotype with a prevalence of approximately 75%. The P(2) red cell phenotype is defined as P1- and P+, and has a prevalence of about 25%. Both P(1) and P(2) red cells also express the frequent P antigen (see GLOB, 615021). The rare P(1)(k) red cell phenotype is defined as P1+ and P(k+) and the rare P(2)(k) red cell phenotype is defined as P1- and P(k+). The P(k+) phenotypes result from inactivating mutations in the B3GALT3 gene, and thus do not express the P antigen. The rare p phenotype does not express any of these antigens and is known as 'null' (Marcus et al., 1976; Koda et al., 2002).

Marcus et al. (1976) noted that the P(k) phenotype lacks the P antigen and that the p phenotype lacks both P and P(k) antigens. P(k) red cells contained only traces of globoside and a marked excess of trihexosyl ceramide, whereas p cells lacked both globoside and trihexosyl ceramide and contained an excess of lactosyl ceramide and other complex glycolipids.

Individuals with the P(2), P(k), and p phenotypes have clinically significant antibodies against whichever antigen is lacking. Anti-P1 antibodies can be associated with hemolytic transfusion reactions, and P and P(k)-related antibodies are implicated in hemolytic transfusion reactions, hemolytic disease of the newborn, and spontaneous abortion. Anti-P antibodies are associated with paroxysmal cold hemoglobinuria (Cantin and Lyonnais, 1983; Soderstrom et al., 1985; Spitalnik and Spitalnik, 1995).

NOR Polyagglutination Syndrome

Polyagglutination is the occurrence of red cell agglutination by virtually all human sera, but not by autologous serum or sera from newborns. Harris et al. (1982) reported a 2-generation American family in which the red blood cells from 5 healthy individuals showed polyagglutination in vitro when exposed to almost all ABO compatible normal sera, but not to cord sera. The trait was transmitted in an autosomal dominant pattern of inheritance. This polyagglutination syndrome was designated 'NOR,' since the family was from Norton, Virginia. There was some degree of variability in the agglutination among family members. Agglutination with control human serum samples could be enhanced when the NOR red blood cells were treated with proteolytic enzymes. The anti-NOR antibody was determined to be IgM. NOR-positive red cells did not agglutinate in response to the lectin from D. biflorus. The agglutination was inhibited by hydatid cyst fluid and avian P1 blood group substance, suggesting that NOR may be related to the P1PK blood group system, although the reactivity was not due to human anti-P1. This report delineated an inherited form of polyagglutination syndrome.

Kusnierz-Alejska et al. (1999) reported a Polish family in which 4 individuals showed polyagglutination syndrome consistent with NOR. The trait was ascertained during blood typing, and the trait was transmitted in an autosomal dominant pattern. Treatment of the NOR+ cells with alpha-galactosidase impaired the polyagglutination, whereas papain or neuraminidase enhanced the polyagglutination. Further characterization indicated that the NOR+ red blood cells contained neutral glycolipids with an abnormal oligosaccharide structure, most likely terminated with alpha-linked galactose residues. Duk et al. (2006) determined that the unique NOR-related glycolipids from both the American and Polish families were identical. Monoclonal anti-NOR and the lectin GSL-IB4 (Griffonia simplicifolia lectin IB4) strongly stained NOR glycolipids that migrated identically in NOR samples from both families.

Using mass spectrometry and immunochemical methods to analyze blood from the Polish proband reported by Kusnierz-Alejska et al. (1999), Duk et al. (2001) determined that NOR1, a component of the NOR antigen, was a unique glycosphingolipid. It was a pentaglycosylceramide; an alpha-galactosylated globoside terminated with a novel Gal(alpha)1-4GalNAc sequence. The NOR glycolipids were recognized by human antibodies that were distinct from known anti-Gal(alpha)1-3Gal xenoantibodies. Duk et al. (2007) determined that the NOR2 component was a disaccharide extension of NOR1, with a terminally linked additional Gal(alpha)1-4GalNAc(beta)1-3 unit. They also identified an intermediate glycolipid (NOR-int) that migrated between NOR1 and NOR2 when NOR2 was treated with alpha-galactosidase. NOR-int did not react with anti-NOR antibodies, but did react with GalNAc-specific soybean agglutinin. NOR-int was found to be a relatively abundant component of a neutral glycolipid fraction from NOR erythrocytes, suggesting it was a precursor to NOR2. The results indicated that polyagglutination in NOR individuals is due to unique erythrocyte glycolipids that are synthesized by sequential addition of Gal(alpha)1-4 and GalNAc(beta)1-3 to globoside (Gb4Cer).

Biochemical Features

The human P1PK blood group system consists of 2 main antigens: P1 and P(k). P (see 615021) is a related antigen and is synthesized from P(k). These antigens are synthesized by the sequential addition of monosaccharide residues to ceramide by different glycosyltransferases. The first step in the biosynthesis of all 3 antigens is the glucosylation of ceramide, followed by the addition of beta-galactose to form lactosylceramide (LacCer), a common precursor. After this, the biosynthetic pathways diverge. P(k) results from the addition of alpha-Gal by alpha-1,4-galactosyltransferase (A4GALT). P(k) is the substrate for beta-3-galactosyltransferase-3 (B3GALT3), which adds N-acetylglucosamine to form the P antigen. Thus, P(k) and P are closely related. The biosynthesis of P1 is different, but the final step requires the activity of an alpha-1,4-galactosyltransferase (Hellberg et al., 2002). Iwamura et al. (2003) and Thuresson et al. (2011) determined that the P1 synthase is A4GALT.

Mapping

The possible assignment of P1 to chromosome 22 was first suggested by McAlpine et al. (1978) who found linkage to the NADH-diaphorase locus (DIA1; 613213). Julier et al. (1985) presented family linkage data in support of this assignment: maximum lod of 1.66 at theta 0.03 with the SIS oncogene (190040). Julier et al. (1988) found evidence for linkage of P1 with markers on chromosome 22 and suggested the following order on 22q: IGL--0.10--D22S1--0.20--MB--0.24--(SIS, P1)--ter.

Exclusion Studies

Phillips and Rodey (1975) reported a large family that gave strongly negative lod scores for linkage of HLA and P, which had previously been suggested by cell hybrid studies (Fellous et al., 1971).

Molecular Genetics

In 4 blood samples from patients with the P(k) phenotype, Hellberg et al. (2002) identified 4 homozygous mutations (603094.0001-603094.0004) in the B3GALT3 gene that are predicted to render the enzyme nonfunctional. The absence of enzyme activity results in the P(k) phenotype.

In Japanese and Swedish individuals with the p phenotype, Steffensen et al. (2000), Furukawa et al. (2000), and Koda et al. (2002) identified homozygous mutations in the A4GALT gene (see, e.g., 607922.0001). Absence or decrease in the enzyme activity results in the p phenotype.

In red blood cells with the P(2) phenotype (P1-) of the P1PK blood group system, Thuresson et al. (2011) identified a homozygous 42C-T transition in exon 2a of the A4GALT gene (607922.0007). There was full concordance between genotype and phenotype among 207 samples. All the P(1) phenotypes were homozygous for C or were C/T heterozygotes. There was 1 possible discrepancy, a P(1) phenotype with a T/T genotype, but this could not be followed up. Thuresson et al. (2011) postulated that the variant may exhibit regulatory function.

NOR Polyagglutination Syndrome

In NOR+ individuals from the American and Polish families with NOR polyagglutination syndrome (Harris et al., 1982 and Kusnierz-Alejska et al., 1999) Suchanowska et al. (2012) identified a heterozygous mutation in the A4GALT gene (Q211E; 607922.0008). Transfection of the mutation in teratocarcinoma cells resulted in binding of both the anti-P1 antibody, which recognizes Gal(alpha)1-4Gal, and the anti-NOR antibodies, which recognize Gal(alpha)1-4GalNAc. Expression of the NOR antigen correlated with expression of the P1 antigen. All NOR+ individuals had at least one P1 allele (42C; 607922.0007) The Q211E mutation broadened the acceptor specificity of the enzyme, causing the transferase to acquire the ability to catalyze the synthesis of Gal(alpha)1-4GalNAc present in NOR-related glycolipids, without losing its ability to transfer the Gal residue to the C4 of Gal (Gal(alpha)1-4). In NOR+ erythrocytes, the mutant enzyme transfers alpha-Gal to the GalNAc of Cb4Cer, transforming a small portion of Gb4Cer into the NOR1 glycolipid. This becomes a new substrate for highly active B3GALNT1, resulting in the synthesis of NOR-int. A small portion of NOR-int is then further elongated by mutant A4GALT, giving rise to NOR2.

Genotype/Phenotype Correlations

The ability of bacteria to adhere to epithelial cells of the host is a prerequisite for many bacterial infections. In human urinary tract infections (UTI), there is a high correlation between the ability of bacteria to adhere to the urinary epithelium and their virulence (Svanborg Eden et al. (1976, 1978)). The adhesive capacity is likely to endow the bacteria with higher resistance to mechanical elimination by the flow of urine and thus aid in their ascent to the upper urinary tract and kidney. In uroepithelial cells and red cells, P blood group system antigens are receptor for pyelonephritogenic E. coli (see Korhonen et al., 1982). Svanborg Eden et al. (1983) found that individuals with the P1 phenotype had a higher density of receptor glycolipids in their red cell membrane than did those with the P2 phenotype, and that the P1 phenotype was overrepresented among patients with recurrent pyelonephritis without reflux: 97% as compared to 75% in healthy children. This patient group also showed a higher frequency of 'attaching bacteria.' In cases of recurrent pyelonephritis with reflux, no significant increase in prevalence of P1 or of attaching bacteria was seen. Lomberg et al. (1983) presented evidence that the P1 blood group phenotype and bacteria that attach to glycolipid epithelial cell receptors are especially common in girls with recurrent pyelonephritis if they do not have vesicoureteral reflux. The P1 blood group phenotype is not more common in patients with reflux and recurrent pyelonephritis than in the healthy population. In the nonreflux group, bacteria causing the pyelonephritis were often of the type with adhesions, whereas these were rare in patients with reflux. The presence of reflux appears to compensate for the defect in the capacity of the bacteria to attach.

Sheinfeld et al. (1989) reviewed work on the relationship between P blood group phenotype and recurrent urinary tract infections. In their own studies they could find no peculiarity in the distribution of P phenotypes in a group of 49 white women with histories of recurrent urinary tract infections.

Lichodziejewska-Niemierko et al. (1995) examined the distribution of P antigen, Lewis blood group phenotypes (111100), and secretor (182100) status in 65 patients with E. coli UTI (20 asymptomatic bacteriuria, 20 cystitis with normal radiology, and 25 reflux nephropathy) and 45 healthy controls who had never experienced a UTI episode. The distribution of Lewis blood group antigens was similar in all UTI groups and in the controls. The incidence of nonsecretors in the reflux nephropathy group was similar to that in controls. P1 phenotype was present in 100% of patients with asymptomatic bacteriuria, 80% with cystitis and in controls, and only 44% with reflux nephropathy. Combined P1/nonsecretor phenotype was observed in 45% of patients with asymptomatic bacteriuria, 30% with cystitis, 12% with reflux nephropathy, and 22% of control individuals. P2/secretor phenotype was demonstrated in 44% of patients with reflux nephropathy and in only 11% of controls. The data suggested to the authors that having the P2 blood group protects against asymptomatic colonization of urinary tract, but is associated with the type of infection responsible for scarring in reflux nephropathy. It also appears that being a nonsecretor does not predispose to renal scarring and that the combined P2/secretor phenotype may be linked with susceptibility to reflux nephropathy.

Population Genetics

The phenotype p, originally called Tj(a-), is rare (Race and Sanger, 1975). There is a relatively high frequency of the p phenotype in Vasterbotten County, Sweden (Cedergren, 1973), and the same rare blood type has been found in the Old Order Amish of Holmes County, Ohio, where the Tj(a-) gene was traced through several generations of a Schwartzentruber Amish kindred (Lehmann, 1991). Patients with the p phenotype have anti-P, anti-P1, and anti-P(k) antibodies (collectively called anti-Tj(a)), like anti-A and anti-B in the ABO system.

Data on gene frequencies of allelic variants were tabulated by Roychoudhury and Nei (1988).

History

By immunochemical studies, Naiki and Marcus (1975) identified the P(k) and P blood group antigens as ceramide trihexoside (CTH) and globoside, respectively, and proposed a structure for the P1 antigen. Although the P1 and P(k) determinants had identical terminal disaccharides, CTH did not inhibit anti-P1, and P1 antigens did not react with anti-P(k) serum. Further studies showed no cross-reactions between P, P1, or P(k) antigens. These authors suggested that the P1 and P2 antigens (P and P1 in the newer nomenclature) are not allelic, i.e., that there are at least 2 loci determining the P system blood phenotype.

In a large kindred studied in connection with acrokeratoelastoidosis (101850), Greiner et al. (1983) found a suggestion of linkage of HLA and P (maximum lod of 1.48 at theta 0.27). Similar data were reported by Keats et al. (1979).