Single nucleotide polymorphisms in A4GALT spur extra products of the human Gb3/CD77 synthase and underlie the P1PK blood group system

Radoslaw Kaczmarek; Katarzyna Szymczak-Kulus; Anna Bereźnicka; Krzysztof Mikołajczyk; Maria Duk; Edyta Majorczyk; Anna Krop-Watorek; Elżbieta Klausa; Joanna Skowrońska; Bogumiła Michalewska; Ewa Brojer; Marcin Czerwinski

doi:10.1371/journal.pone.0196627

Abstract

Contrary to the mainstream blood group systems, P1PK continues to puzzle and generate controversies over its molecular background. The P1PK system comprises three glycosphingolipid antigens: P^k, P1 and NOR, all synthesised by a glycosyltransferase called Gb3/CD77 synthase. The P^k antigen is present in most individuals, whereas P1 frequency is lesser and varies regionally, thus underlying two common phenotypes: P₁, if the P1 antigen is present, and P₂, when P1 is absent. Null and NOR phenotypes are extremely rare. To date, several single nucleotide polymorphisms (SNPs) have been proposed to predict the P₁/P₂ status, but it has not been clear how important they are in general and in relation to each other, nor has it been clear how synthesis of NOR affects the P₁ phenotype. Here, we quantitatively analysed the phenotypes and A4GALT transcription in relation to the previously proposed SNPs in a sample of 109 individuals, and addressed potential P1 antigen level confounders, most notably the red cell membrane cholesterol content. While all the SNPs were associated with the P₁/P₂ blood type and rs5751348 was the most reliable, we found large differences in P1 level within groups defined by their genotype and substantial intercohort overlaps, which shows that the P1PK blood group system still eludes full understanding.

Citation: Kaczmarek R, Szymczak-Kulus K, Bereźnicka A, Mikołajczyk K, Duk M, Majorczyk E, et al. (2018) Single nucleotide polymorphisms in A4GALT spur extra products of the human Gb3/CD77 synthase and underlie the P1PK blood group system. PLoS ONE 13(4): e0196627. https://doi.org/10.1371/journal.pone.0196627

Editor: Alberto G. Passi, University of Insubria, ITALY

Received: January 17, 2018; Accepted: April 16, 2018; Published: April 30, 2018

Copyright: © 2018 Kaczmarek et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: All relevant data are within the paper and its Supporting Information files.

Funding: This work was funded by National Science Centre of Poland Opus Project DEC-2014/13/B/NZ6/00227 (Prof. Marcin Czerwinski). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. This publication was supported by the Wroclaw Centre of Biotechnology, programme The Leading National Research Centre (KNOW) for the years 2014-2018.

Competing interests: The authors have declared that no competing interests exist.

Introduction

Despite great strides made to understand the molecular background of human blood groups, the P1PK blood group system continues to puzzle. The difference between P₁ and P₂ (the two common P1PK phenotypes) red blood cells has been known since 1927, when Landsteiner and Levine found that rabbits immunized with human erythrocytes produced antibodies reacting with an antigen then named P and now called P1[1]. Since then, the P blood group system has been renamed P1PK (International Society of Blood Transfusion system 003), and while knowledge about the antigens belonging to that system has grown considerably, its molecular background is still far from being completely elucidated. The P1PK blood group system consists of three glycosphingolipid antigens: P^k (Gb3, CD77), P1 and NOR[2]. The P^k antigen is expressed on RBCs of most individuals (except in the null phenotype, denoted p), whereas P1 varies in different populations: from 30% in Japanese to 80% in Caucasians, to 94% in Blacks, thus underlying two common phenotypes: P₁, if the P1 antigen is present, and P₂, if P1 is absent[3]. The structures of the antigens belonging to the P1PK blood group system and phenotypes linked to these antigens are shown in Fig 1.

Download:

Fig 1. Schematic representation of the three glycosphingolipid antigens and phenotypes of the human P1PK blood group system.

https://doi.org/10.1371/journal.pone.0196627.g001

While it is well-established that the P^k antigen is synthesised by Gb3/CD77 synthase (α1,4-galactosyltransferase, P1/P^k synthase, encoded by A4GALT)[4], P1 has only recently been unequivocally shown to be a product of the same enzyme[5]. In addition, c.631C>G mutation in A4GALT, leading to p.Q211E substitution in Gb3/CD77 synthase (rs397514502), alters the enzyme specificity, rendering it able to synthesise all the three P1PK system antigens: P^k, P1 and NOR[6]. This makes Gb3/CD77 synthase a unique case of a promiscuous glycosyltransferase, radically unmoored from the old ‘one enzyme-one linkage’ tenet[7].

Glycosphingolipids containing Galα1,4Gal unit are targeted by bacterial adhesins, toxins and viruses. These glycosphingolipids may localise in distinct membrane microdomains and act as genuine or decoy receptors. Most prominently, the P^k antigen is the receptor for Shiga toxins and its membrane localisation determines the toxins’ fate upon internalisation. P^k may be masked from the binding proteins and its recognition depends on the structure of its ceramide moiety and membrane microenvironment, including cholesterol content[8]. The P^k antigen was once believed to be a centroblast differentiation marker, which is why it is also known as CD77[9,10]. Unlike P^k, the P1 antigen seems to be restricted to the erythroid lineage. Elsewhere than on RBCs, P1 was shown to be present only on ovarian cancer cells[11]. P1PK blood type may play a role in helminth infections, because these worms may display the Galα1,4Gal unit on their glycoconjugates for molecular mimicry[2,12,13].

The level of P1 antigen is related to the P₁/P₂ status and genotype at the A4GALT locus (P¹P¹, P¹P² or P²P²)[14,15]. The genetic background of theP₁/P₂ difference remained controversial for years, because it does not result from missense mutations and only one gene encoding Gb3/CD77 synthase exists in humans. Therefore, it is generally accepted that the P₁/P₂ status derives from varied A4GALT transcript levels. Several groups proposed different SNPs upstream from the coding region to underlie the P₁/P₂ difference. The SNPs rs5845556 (g.4501_4502insC, NG_007495.1) and rs28910285 (g.4892A/G, NG_007495.1) found by Iwamura et al (2003) were later found not to be correlated with the P₁/P₂ status. More recently, rs8138197[14] (g.7326C/T, NG_007495.1), rs2143918 (g.7837C/G, NG_007495.1), rs2143919 (g.7857T/G, NG_007495.1) and rs5751348[15] (g.8084G/T, NG_007495.1) found downstream of exon 1 of A4GALT were shown to be associated with the P₁/P₂ status (Fig 2). However, in either case, the statistical data presented in support of the identified SNPs were based on limited sample sizes, did not show the data distributions or effect size. Since differences in P1 antigen level may be confounded by a number of factors, such as KLF1 expression level, extra scrutiny is desirable[16–19]. Also, none of the previous studies analysed the level of P1 antigen in NOR-positive RBCs, which warrants investigation, because the NOR antigen is synthesised by the same enzyme. To address the controversy over allelic variations of A4GALT gene expression and P₁/P₂ phenotypic differentiation[14,15], we analysed the effect of four SNPs (rs8138197, rs2143918, rs2143919, rs5751348) previously reported to determine the P₁/P₂ status on A4GALT transcript levels and cellular-scale quantity of the P1 and NOR antigens in a sample of 109 NOR-negative and NOR-positive Polish individuals. Importantly, we present the key results of this study in the form of univariate scatter plots, which faithfully represent distributions of continuous data, unlike still too often used bar plots[20].

Download:

Fig 2. Structure of A4GALT, including the SNPs evaluated in this study.

https://doi.org/10.1371/journal.pone.0196627.g002

Materials and methods

Sample preparation and P₁/P₂ blood typing

The study was approved by the Wrocław Medical University Bioethics Committee, Consent 641/2014, December 14, 2014. Blood (n = 109) from apparently healthy individuals was obtained following an informed consent according to the Declaration of Helsinki. Blood samples were collected on EDTA (for DNA extraction, flow cytometry and glycosphingolipid extraction) or on heparin (for RNA extraction), and washed RBCs were stored in CellStab low-ionic strength preservative solution (DiaMed, Cressier, Switzerland). DNA was extracted using Quick Blood DNA Purification Kit (EURx, Gdansk, Poland). RNA was prepared from the buffy coat using Human Blood RNA Purification Kit (EURx, Gdansk, Poland). The P₁/P₂ phenotype was determined by standard haemagglutination method using the human monoclonal anti-P1 antibody (S1 Table).

Identification of the individual with PP1PK null (p) phenotype

During a routine antibody screening performed by the indirect antiglobulin test (IAT, DiaMed, Cressier, Switzerland) we found that the serum from a Polish blood donor agglutinated all panel RBCs (DiaMed, Cressier, Switzerland). To identify the antibody specificity, we employed an expanded panel of RBCs, including P—RBCs (the null phenotype of GLOB, a blood group system related to P1PK, 2 different samples) and p RBCs (the null phenotype of P1PK, 4 different samples) obtained from Serum Cells and Rare Fluid (SCARF) Exchange. These were used to test the donor’s plasma by the 2-stage enzyme (papain) test (DiaMed, Cressier, Switzerland), while the serum was tested by LISS IAT and 22°C saline tube tests. Very strong reactivity, agglutination or hemolysis, respectively, with all RBCs except P—and p phenotypes was observed. No reaction with the donor’s autologous RBCs was detected. The reactivity pattern indicated the anti-PP1Pk alloantibody specificity. The donor’s phenotype was confirmed by a monoclonal anti-P1 antibody and a panel of sera with different reactivities, including anti-P (1 sample) and anti-PP1Pk (3 different samples) from SCARF Exchange. Neither the monoclonal antibody nor any of the sera reacted with the donor’s RBCs, thus confirming his PP1PK null (p) phenotype.

Analysis of the SNPs

Three DNA segments encompassing the studied SNPs were amplified using polymerase chain reaction (PCR). All primers are listed in S2 Table. PCR was performed using an MJ Mini gradient PCR apparatus (Bio-Rad, Hercules, CA, USA) in 20-μl reaction mixes containing 200 ng of genomic DNA, 0.2 mM dNTPs, Taq buffer with KCl (1:10 dilution), 1.5 mM MgCl₂, 0.2 mM forward and reverse primers, and 1 unit of Taq polymerase (Fermentas, Vilnius, Lithuania). The PCR conditions are shown in S3 Table. The resulting DNA fragments were purified with a gel extraction kit (Gel-Out kit; A&A Biotechnology, Gdynia, Poland) and sequenced using the amplification primers. The genotypes (P^1NORP², P¹P², P^1NORP¹, P¹P¹, and P²P²) were assigned based on the SNP that best correlated with the P₁/P₂ blood typing and the c.631C/G status. Nucleotide differences between hetero- and homozygotes for individual P₁/P₂-related SNPs are shown in S1 Table. The pp genotype of the sole p individual in the cohort was confirmed previously [21].

Quantitative analysis of transcripts

The complementary DNAs (cDNAs) were synthesised using 240 ng of RNA and SuperScript III First-Strand Synthesis kit (Life Technologies, Carlsbad, CA, USA) with oligo(dT) primers. Quantitative polymerase chain reaction (qPCR) was performed on 3 μl of cDNA using the 7500 Fast Real-Time PCR System (Life Technologies), according to the manufacturer’s instructions. A predesigned TaqMan assay targeting exon 2–3 boundary (Hs00213726_m1; Life Technologies) was used. Transcript quantities were normalised to ACTB (β-actin) endogenous control (assay Hs99999903_m1). All samples were run in triplicates. P²P² samples were used as the calibrator. Data were analysed using Sequence Detection software Version 1.3.1 (Life Technologies). The real-time PCR conditions are shown in S4 Table.

Antibodies

The human anti-P1, mouse anti-P1 and goat anti-mouse IgG conjugated with fluorescein isothiocyanate (FITC) antibodies were purchased from Immucor Inc. (Norcross, GA, USA), Ce-Immundiagnostika (Eschelbronn, Germany) and Santa Cruz Biotechnology (Dallas, TX, USA), respectively. The mouse monoclonal anti-NOR antibody, nor118 was obtained in our laboratory before and used as a diluted culture supernatant[22]. The goat anti-human IgM conjugated with FITC, biotinylated anti-mouse antibody and biotinylated anti-human antibody were purchased from Pierce (Rockford, IL, USA).

Flow cytometry

0.5% RBC suspensions were incubated with 100 μl appropriately diluted primary antibodies (human anti-P1 1:400, mouse anti-P1 1:200, anti-NOR 1:20) for 40 min on ice. Then the cells were washed (all washes and dilutions were done with PBS) and incubated with 100 μl (diluted 1:100) FITC-labeled anti-mouse IgM antibody) for 40 min on ice in the dark. The cells were washed and suspended in 750 μl of cold PBS, and analysed by flow cytometry using FACSCalibur (BD Biosciences, Franklin Lakes, NJ, USA). The number of events analysed was 10,000/gated cell population. The results analysis was carried out using Flowing Software (Perttu Terho, University of Turku, Turku, Finland)[23].

Quantitative flow cytometric analysis of the P1 antigen

The Quantum (Bio-Rad, Hercules, CA, USA) bead populations with defined quantities of FITC diluted in PBS were used to plot calibration curves (mean fluorescence intensity versus Molecules of Equivalent Soluble Fluorochrome units). The cells were then analysed by flow cytometry and the antibody binding capacities (ABCs, the number of antibody molecules bound per cell) were calculated by interpolation from the calibration curve as described in the manufacturer’s protocol and based on the fluorophore-to-protein molar ratios of the FITC-antibody conjugates. Negative control results (secondary antibody only) were subtracted from the sample results to obtain specific antibody binding capacities.

Lipid panels

Fasting lipid profiles (mg% total cholesterol, HDL and LDL) were measured by Diagnostyka (Krakow, Poland). None of the individuals in the study were treated with statins.

Extraction and purification of glycosphingolipids

The isolation and fractionation of total glycosphingolipids and the orcinol staining were performed as described previously[24,25]. Lipids were extracted with chloroform/methanol from freeze-dried RBC ghosts. The neutral glycosphingolipids were separated from the phospholipids and gangliosides, purified in peracetylated form, de-O-acetylated, and desalted. Glycosphingolipid samples were solubilised in chloroform/methanol (2:1, v/v), applied to HPTLC plates (Kieselgel 60, Merck, Darmstadt, Germany), and developed with chloroform/ methanol/water (55:45:9, v/v/v). The dried plates were immersed in 0.05% polyisobutylmethacrylate (Aldrich, Steinheim, Germany) in hexane for 1 min, dried, sprayed with TBS (0.05 M Tris buffer, 0.15 M NaCl (pH 7.4)), and blocked in 5% BSA. For antibody assays, the plates were sequentially overlaid with 1) primary antibody diluted in TBS/1% BSA (TBS-BSA) for 1–1.5 h; 2) biotinylated goat anti-mouse Ig antibody conjugated with alkaline phosphatase (Dako, Glostrup, Denmark), diluted 1:1000 with TBS-BSA; 3) ExtrAvidin-alkaline phosphatase conjugate (Sigma-Aldrich, St. Louis, MO, USA) diluted 1:5000 with TBS/ BSA/0.2% Tween20 for 1 h; and 4) the substrate solution (nitro blue tetrazolium/5-bromo-4-chloro-3-indolyl phosphate, Sigma-Aldrich). Other details were as described previously[25,26]. Each HPTLC experiment was repeated three times.

Statistical analysis

The data distributions were tested for normality using Shapiro-Wilk test and inspected visually on Q-Q plots. The Box-Cox transformation was employed to reduce skewness and allow the use of parametric tests[27]. The means for transformed data were backtransformed to obtain weighted means in the original scale. Two-tailed two-sample independent t-test (or Welch’s test in the case of P^1NORP² versus P¹P² and P^1NORP¹ versus P¹P¹ human anti-P1 ABCs, and P^1NORP¹ versus P¹P¹ and P¹P² versus P²P² mouse anti-P1 ABCs, because of unequal variances) was employed for intercohort ABC mean comparisons. The backtransformed means were used to calculate mean differences and their 95% confidence intervals. Hedges’ g was used as the standardised measure of effect size to account for the small sample sizes. The 95% confidence intervals of standardised effect sizes were calculated based on noncentrality parameters of noncentral-t distributions. A Pearson’s correlation test was run to evaluate the relationships between total cholesterol, HDL and LDL levels and anti-P1 ABCs (we chose the P¹P² cohort to avail of the largest sample size). The correlation coefficients were tested for significance using a two-tailed t-test. A χ² test with one degree of freedom and a type I error of 0.05 was run to check if the P¹P¹, P¹P² and P²P² genotypes were in the Hardy-Weinberg equilibrium. All the statistical tests were performed with a type II error of 0.20 and type I error of 0.05 (except for multiple comparisons). The type I error was adjusted for multiple comparisons using the Dunn–Šidák correction to 0.0127 in the case of human and mouse anti-P1 ABCs, to 0.0253 in the case of A4GALT transcript levels, and to 0.0085 in the case of Pearson’s correlation test. All analyses were carried out using the Real Statistics Resource Pack software (Release 5.1, Copyright 2013–2017, Charles Zaiontz, www.real-statistics.com) in Microsoft Office Excel environment (Microsoft Corp, Redmond, WA).

Results

SNPs that best predict the P₁/P₂ blood type

To evaluate the role of single nucleotide polymorphisms in the P₁/P₂ blood group differentiation, we determined SNPs rs8138197 (g.7326C/T, NG_007495.1), rs2143918 (g.7837C/G, NG_007495.1), rs2143919 (g.7857T/G, NG_007495.1) and rs5751348 (g.8084G/T, NG_007495.1) in 109 Polish individuals. The results are shown in S1 Table. There were 84 P1-positive (P₁, including 10 NOR-positive), 1 p (null), and 24 P1-negative (P₂) individuals. The P₁/P₂ status of 100 individuals (91,7% of the total number) could be consistently predicted using all 4 evaluated SNPs. rs2143919 showed the weakest association with the phenotype, as shown before[15]. In two cases, we found that rs8138197 predicted P¹P¹ homozygosity, while the other 3 SNPs suggested P¹P². These two individuals were left out of the further analysis. One P²P² (according to all four SNPs) individual was typed P₁ in the agglutination test. Frequencies of the three genotypes (P¹P¹, P¹P² and P²P²) were in the Hardy-Weinberg equilibrium (p = 0.1823).

Anti-P1 antibody binding capacities of RBCs with different genotypes

Since we used two different (mouse and human) anti-P1 antibodies and one anti-NOR antibody, we performed an HPTLC analysis of neutral glycosphingolipids isolated from RBCs of different phenotypes (Fig 3). We found that the mouse monoclonal anti-P1 antibody (clone 650) binds to both P1 and P^k antigens, while the human anti-P1 monoclonal antibody (clone P3NIL100) binds to the P1 antigen only. The mouse monoclonal anti-NOR antibody (nor118) recognized only the NOR antigen represented by NOR1 and NOR2 glycosphingolipids. NOR1 and NOR2 contain the same terminal disaccharide, recognized by nor118, and emerge in the same pathway [28].

Download:

Fig 3. HPTLC analysis of total neutral glycosphingolipids extracted from red blood cells with different genotypes: P^1NORP¹ (lane 1), P¹P¹ (lane 2), P²P² (lane 3), and pp (lane 4).

The image was uniformly treated with a gamma correction tool (IrfanView 4.38) to improve visibility. Individual panels represent separate silica plates, which were cropped and realigned for clarity. Full-length plates with original gamma settings are presented in S1 Fig.

https://doi.org/10.1371/journal.pone.0196627.g003

In the quantitative flow cytometry (qFCM) analysis using the anti-P1 antibodies, P¹P¹ and P²P² RBCs showed the highest and the lowest antibody binding capacities (ABC), respectively, while pp RBCs showed no binding at all (Fig 4). Scatter plots of ABC for all but the P²P² group were skewed and widely distributed, revealing substantial intercohort overlaps. In the case of mouse anti-P1 antibody (clone 650), the weighted means (obtained by adjusting for skewness) of ABC were 12350, 2597, 11, 5735, 1850 and 0 antibody molecules per RBC for P¹P¹ (n = 13), P¹P² (n = 42), P²P² (n = 18), P^1NORP¹ (n = 5), P^1NORP² (n = 5), and pp (null) (n = 1) genotypes, respectively; in the case of human anti-P1 antibody (clone P3NIL100), the weighted ABC means were 9608, 4389, 139, 6028, 5093 and 0, respectively (Fig 5A).

Download:

Fig 4. Flow cytometric analysis of the P1 and NOR antigens on RBCs with different genotypes.

The fluorescence intensity is in log scale.

https://doi.org/10.1371/journal.pone.0196627.g004

Download:

Fig 5. Scatter plots of RBC anti-P1 and anti-NOR antibody binding capacities (A) and relative A4GALT transcript levels in individuals with different genotypes and 95% confidence intervals on the intercohort mean differences (B).

https://doi.org/10.1371/journal.pone.0196627.g005

In contrast to the results of HPTLC analysis, the flow cytometry experiments suggested that the mouse anti-P1 antibody did not bind the P^k antigen. In fact, RBCs of all the cohorts except P¹P¹ showed higher human anti-P1 ABC than the mouse anti-P1 ABC. This would not have been the case if the mouse antibody bound both P1 and P^k, because P^k is one of the major neutral glycosphingolipids of the RBC, unlike P1, which is a minor component of the RBC membrane, if at all present. Despite its abundance, P^k was shown to be hardly detectable for binding proteins when anchored in the cell membrane, in contrast to the lankier P1[29]. Henceforth, we assumed that in the RBC flow cytometry, both the human and mouse antibody recognize P1 only.

Precision of phenotyping using the four SNPs

The ABC means were compared between cohorts to check whether the effect of the P₁/P₂-differentiating SNPs is significant across the three genotypes (P¹P¹ versus P¹P² and P¹P² versus P²P²; NOR-positive individuals were excluded from this analysis to eliminate the potential confounding effect of the altered enzyme activity). The two anti-P1 antibodies performed consistently insofar that in the case of both reagents the mean differences were statistically significant between P¹P¹ and P¹P², and P¹P² and P²P², with large effect sizes (Table 1). Wide 95% confidence intervals reflected dispersion of the data (Fig 5B). The post-hoc statistical power analysis revealed that in all cases the tests were sufficiently powered, well above the 80% benchmark.

Download:

Table 1. Statistics on the effect of rs5751348 SNP on the P1 antigen level.

https://doi.org/10.1371/journal.pone.0196627.t001

The averaged ratio of transcript levels for P¹P¹, P¹P² and P²P² genotypes was 10:3:1 (Fig 5). The mean differences between P¹P² and P¹P¹ or P²P² were statistically significant and revealed a strong effect of the SNPs on A4GALT transcript level. The tests of these differences were underpowered, because qPCR was carried out on samples from fewer individuals (8 P¹P¹, 21 P¹P² and 10 P²P²) than qFCM, due to insufficient quality of some RNA extracts. However, the results of qFCM using both anti-P1 antibodies and qPCR were consistent across all the three cohorts.

Thus, rs5751348[15] showed the strongest association with anti-P1 antibody binding capacity, although outliers were prominent insofar that ABCs of RBCs with different genotypes overlapped extensively.

Influence of the NOR mutation on the activity of Gb3/CD77 synthase

The c.631C>G missense mutation (rs397514502) was shown to broaden the specificity of Gb3/CD77 synthase, rendering it able to catalyse the synthesis of both Galα1,4Gal (the terminal disaccharide of P1 and P^k antigens) and Galα1,4GalNAc moieties (the terminal unit of NOR antigen), but its influence on enzyme activity in vivo remained unknown. P1, P^k and NOR antigens (represented by NOR1 and NOR2 glycosphingolipids) were all detected in glycosphingolipids isolated from NOR-positive RBCs using the anti-P1 and anti-NOR antibodies (Fig 3).

The anti-P1 ABCs of RBCs from NOR-positive individuals with P¹P¹ genotypes (P^1NORP¹) was reduced by 54% in comparison with NOR-negative P¹P¹ individuals, when measured using the mouse antibody, and by 37%, when measured using the human one. However, comparison of NOR-positive (P^1NORP²) and NOR-negative P¹P² individuals (P¹P²) generated conflicting results with the mouse antibody showing 29% lower ABC of P^1NORP² RBCs, yet the human one showing 16% higher ABC of these RBCs (Fig 5). While the differences between P^1NORP¹ and P¹P¹ cohorts were statistically significant, the differences between P^1NORP² and P¹P² were not (Table 2).

Download:

Table 2. Statistics on the effect of c.631C>G (NOR) mutation on the P1 antigen level.

https://doi.org/10.1371/journal.pone.0196627.t002

The mean anti-NOR ABCs were 4354 and 3707 for P^1NORP¹ and P^1NORP², respectively. The mean difference was not statistically significant.

Plasma cholesterol as a confounder of P1 expression

Since it was postulated that the P1PK antigens localise primarily in lipid rafts, which are enriched in cholesterol, we checked total cholesterol, HDL and LDL plasma concentrations in individuals with known P1PK status. The data are shown in S1 Table. Although we did not find statistically significant correlations between plasma HDL or LDL levels and anti-P1 ABC, we found interesting if counterintuitive trends (Table 3, S2 Fig); the LDL- and HDL-ABC relationships were direct and inverse, respectively, and this pattern was consistent between tests involving the human and mouse anti-P1 ABC (Fig 6 and S2 Fig). The relationship between total cholesterol and anti-P1 ABC was positive. It is important to know that the correlation tests lacked statistical power for the observed effect sizes (the largest observed Pearson’s coefficient was 0.137 for the human anti-P1 ABC and LDL level). To confidently reject (or accept) the null hypothesis (no correlation) for the observed effect sizes, the sample would have to comprise over 410 individuals.

Download:

Fig 6. 95% confidence intervals on Pearson’s correlation between RBC anti-P1 antibody binding capacities, HDL, LDL and total cholesterol.

https://doi.org/10.1371/journal.pone.0196627.g006

Download:

Table 3. Statistics on relationships between RBC anti-P1 antibody binding capacities, HDL, LDL and total cholesterol.

https://doi.org/10.1371/journal.pone.0196627.t003

Discussion

Despite perennial efforts, P1PK remains the most elusive human blood group system. The manifold attempts to obtain a clear picture of how P1PK blood group differentiation ensues have been marred by the unorthodox activity of Gb3/CD77 synthase, the unusual ways in which genetic changes alter this enzyme’s activity, and the intertwined synthesis of the P1PK and other glycosphingolipid blood group antigens. While there is a consensus that A4GALT transcript level in P₁ RBCs is higher than in P₂, presumably because it leads to upregulation of Gb3/CD77 synthase (although it was not directly proven), the molecular background of that phenomenon remains obscure. There have been three reports showing that single nucleotide polymorphisms in the non-coding region of A4GALT may play a role in regulating transcription: Iwamura et al (2003), Thuresson et al (2011) and Lai et al (2014)[30,14,15]. Today, it is generally accepted that SNPs described by Iwamura et al play no role in P₁/P₂ polymorphism, while the SNPs proposed in the two other reports do matter, but so far it has not been completely clear how important they are in general and in relation to each other, because of statistical limitations of those studies. Importantly, the SNPs in question densely intersperse within a relatively short stretch of the gene, which adds to the challenge of studying their effects, because of a potential genetic linkage.

In this study, we set out to scrutinize the postulated effects of previously reported P₁/P₂-differentiating SNPs and map the influence of NOR mutation (c.631C>G in A4GALT) onto the P1 antigen level. Our results provide the strongest evidence yet that the four SNPs previously reported by Thuresson et al. and Lai et al. are related to the P₁ or P₂ phenotype[14,15], of which rs5751348[15] proved to be the most reliable, while the three other SNPs underperformed in terms of their predictive value. However, in contrast to the previous studies, the scatter plots of antibody binding capacities and transcript levels revealed wide and largely overlapping distributions. This shows that the SNPs in question, albeit important, cannot universally predict the phenotype and that the P1 antigen level may be confounded by other factors. One such potential confounder is the transcription factor EKLF (Erythroid Krüppel-Like Factor, encoded by KLF1), not least because rs5751348 is located in an EKLF consensus binding site[16]. The most recent studies implicated the transcription factors EGR1[31] and RUNX1[32] in regulation of A4GALT expression, and so P1 synthesis. The P1 antigen level may also be influenced by competition between Gb3/CD77 synthase and other glycosyltransferases for the same substrate, paragloboside, which is the direct precursor of P1, as well as by activities of enzymes acting upstream in its biosynthetic pathway[12]. Here, we looked at another, rarely addressed, potential confounder: the RBC membrane cholesterol content. It was shown that excess plasma membrane cholesterol may cloak glycosphingolipids, including the P^k antigen, from the binding proteins[8,29,33]. Most of these studies focused only on the P^k antigen’s role as a Shiga toxin receptor, which is perhaps why the cholesterol aspect has not crossed over to immunohaematologic studies on the P1PK system so far. However, P1 is likely to be subject to the same masking mechanisms as P^k, although the longer sugar chain may limit the extent to which it can become stealth. To evaluate cholesterol as a potential confounder of the P1 antigen level, we looked at plasma total cholesterol, HDL and LDL concentrations. The parameters we chose to look at seemed to be good surrogates of the RBC membrane cholesterol content, because RBCs engage in the systemic cholesterol flux and RBC membranes exchange cholesterol with plasma lipoproteins [34–36]. This study lacked statistical power to assert or rule out the small effect of cholesterol level, yet it was interesting to see the tendency of LDL and HDL to be related positively and negatively with anti-P1 antibody binding capacities, respectively. Since HDL transports cholesterol from peripheral tissues to the liver, the reverse being the job of LDL, the inverse relationship of HDL with anti-P1 ABC and the direct relationship of LDL therewith is counterintuitive. However, this pattern was consistent across all comparisons and made more plausible by the direct relationship between total cholesterol and anti-P1 ABC. These surprising trends may suggest that the unique features of the RBC membrane, such as its unusual curvature, alter the way in which cholesterol affects crypticity of RBC antigens or that the P1 antigen membrane topology is different than expected and merit further study.

We also attempted to evaluate the effect of NOR mutation (c.631C>G) on activity of the Gb3/CD77 synthase in vivo and how it relates to our earlier studies ex vivo and in vitro[5–7]. In our study on a transfected cell line[7], activity of the consensus Gb3/CD77 synthase was higher than activity of the p.Q211E mutein, in contrast to our results with the purified recombinant variants of the enzyme. The results of this study suggest that the p.Q211E enzyme’s superpromiscuity comes at the cost of P1 synthesis. The use of quantitative flow cytometry assay allowed us to compare binding levels of anti-P1 and anti-NOR antibodies, regardless of differences in specificity and quantities of the fluorophore conjugated to the secondary antibodies. On that basis, it may be estimated that at the cost of one P1 molecule the p.Q211E mutein produces one NOR molecule. The disparate results of in vivo and in vitro studies are not entirely unexpected, since glycosyltransferases in vivo may enter homo- and heterooligomeric complexes and show altered enzyme activities[37–42]. It was demonstrated that Gb3/CD77 synthase may form such complexes as well, but in this case the functional relevance remains unknown[43,44]. Since all the statistical tests involving NOR-positive cohorts lacked power to pinpoint the effect sizes, these tests were subject to either type II or type M error. This problem could not be worked around by increasing the sample sizes, because of extreme rarity of the NOR-positive phenotype.

Notably, the sole pp (null) individual in the study revealed very low A4GALT transcript level despite his P¹P¹ homozygosity (for all the four SNPs). It was markedly lower than the minimal one in the P¹P¹ cohort and lower than the mean in the P²P² cohort. The genetic background of that individual’s null phenotype was previously described[21]. It derives from a nonsense mutation in one allele and a missense mutation in the other, so the low transcript level may be caused by nonsense-mediated mRNA decay, which is a quality-control mechanism removing mRNAs with premature termination codons[45].

In summary, our results show that the SNPs rs8138197 and rs5751348 proposed by Thuresson et al and Lai et al[14,15], respectively, are strongly associated with the P₁/P₂ status, but rs5751348[15] offers the best predictive value, while rs8138197 underperforms. The growing importance of genetic testing quickly caught up with transfusion medicine and prompted the search for genetic markers that would replace standard haemagglutination assays, at least in complex scenarios, such as foetal testing[46]. In the case of P1PK blood group system, P₁/P₂ blood typing based on the known SNPs requires caution, because it lacks perfect accuracy. In addition, individuals with different genotypes show largely varying levels of the P1 antigen, which suggests that it may be strongly confounded by other (not necessarily genetic) factors. We also show for the first time how synthesis of all the three P1PK system antigens (P^k, P1 and NOR) may alter the P₁ phenotype, which affords a better understanding of the system and how the responsible enzyme, Gb3/CD77 synthase, operates in vivo. The promiscuous activity of this enzyme and its impact on the phenotype stand in stark contrast to how carbohydrate-active enzymes used to be viewed. While the complete picture of the P1PK blood group system seems to be closer than ever, uncertainty remains around indirect factors contributing to the phenotype, so the final chapter in the P1PK story is yet to be written.

Supporting information

S1 Fig. HPTLC analysis of neutral glycosphingolipids extracted from RBCs with different genotypes: P^1NORP¹ (lane 1), P¹P¹ (lane 2), P²P² (lane 3), and pp (lane 4).

The image represents chemical staining and immunoverlays of silica plates from different chromatography runs before cropping and realignment.

https://doi.org/10.1371/journal.pone.0196627.s001

(PDF)

S2 Fig. Scatter plots of relationships between Box-Cox transformed RBC anti-P1 binding capacities, HDL, LDL and total cholesterol.

https://doi.org/10.1371/journal.pone.0196627.s002

(PDF)

S1 Table. Phenotypes, genotypes, RBC antibody binding capacities and lipid profiles of recruited individuals.

The colour-coding is explained beneath the table.

https://doi.org/10.1371/journal.pone.0196627.s003

(PDF)

S2 Table. Sequences of primers used for genotyping.

https://doi.org/10.1371/journal.pone.0196627.s004

(PDF)

S3 Table. PCR conditions used for amplification of A4GALT fragments encompassing the studied SNPs.

https://doi.org/10.1371/journal.pone.0196627.s005

(PDF)

S4 Table. Real-time PCR conditions used for A4GALT gene expression assays.

https://doi.org/10.1371/journal.pone.0196627.s006

(PDF)

Acknowledgments

We thank Teresa Karulek and Jadwiga Braszka from the Regional Center of Transfusion Medicine and Blood Bank (Wrocław, Poland) for their help with blood collection.

This work was funded by National Science Centre of Poland Opus Project DEC-2014/13/B/NZ6/00227.

The study was approved by the Wrocław Medical University Bioethics Committee, Consent 641/2014, December 14, 2014.

This publication was supported by the Wroclaw Centre of Biotechnology, programme The Leading National Research Centre (KNOW) for the years 2014–2018.

References

1. Landsteiner K, Levine P. Further Observations on Individual Differences of Human Blood. Exp Biol Med. SAGE PublicationsSage UK: London, England; 1927;24: 941–942.
- View Article
- Google Scholar
2. Kaczmarek R, Buczkowska A, Mikołajewicz K, Krotkiewski H, Czerwinski M. P1PK, GLOB, and FORS Blood Group Systems and GLOB Collection: Biochemical and Clinical Aspects. Do We Understand It All Yet? Transfus Med Rev. 2014;28: 126–136. pmid:24895151
- View Article
- PubMed/NCBI
- Google Scholar
3. Storry JR, Castilho L, Daniels G, Flegel WA, Garratty G, Francis CL, et al. International Society of Blood Transfusion Working Party on red cell immunogenetics and blood group terminology: Berlin report. Vox Sang. 2011;101: 77–82. pmid:21401621
- View Article
- PubMed/NCBI
- Google Scholar
4. Steffensen R, Carlier K, Wiels J, Levery SB, Stroud M, Cedergren B, et al. Cloning and expression of the histo-blood group P^k UDP-galactose: Galβ1-4Glcβ1-Cer α1,4-galactosyltransferase. Molecular genetic basis of the p phenotype. J Biol Chem. 2000;275: 16723–16729. pmid:10747952
- View Article
- PubMed/NCBI
- Google Scholar
5. Kaczmarek R, Duk M, Szymczak K, Korchagina E, Tyborowska J, Mikolajczyk K, et al. Human Gb3/CD77 synthase reveals specificity toward two or four different acceptors depending on amino acid at position 211, creating P^k, P1 and NOR blood group antigens. Biochem Biophys Res Commun. 2016;470: 168–74. pmid:26773500
- View Article
- PubMed/NCBI
- Google Scholar
6. Suchanowska A, Kaczmarek R, Duk M, Lukasiewicz J, Smolarek D, Majorczyk E, et al. A Single point mutation in the gene encoding Gb3/CD77 synthase causes a rare inherited polyagglutination syndrome. J Biol Chem. 2012;287: 38220–38230. pmid:22965229
- View Article
- PubMed/NCBI
- Google Scholar
7. Kaczmarek R, Mikolajewicz K, Szymczak K, Duk M, Majorczyk E, Krop-Watorek A, et al. Evaluation of an amino acid residue critical for the specificity and activity of human Gb3/CD77 synthase. Glycoconj J. Springer; 2016;33: 963–973. pmid:27538840
- View Article
- PubMed/NCBI
- Google Scholar
8. Lingwood C a., Manis A, Mahfoud R, Khan F, Binnington B, Mylvaganam M. New aspects of the regulation of glycosphingolipid receptor function. Chem Phys Lipids. 2010;163: 27–35. pmid:19781539
- View Article
- PubMed/NCBI
- Google Scholar
9. Högerkorp C-M, Borrebaeck CAK. The human CD77- B cell population represents a heterogeneous subset of cells comprising centroblasts, centrocytes, and plasmablasts, prompting phenotypical revision. J Immunol. 2006;177: 4341–9. pmid:16982868
- View Article
- PubMed/NCBI
- Google Scholar
10. Victora GD, Dominguez-Sola D, Holmes AB, Deroubaix S, Dalla-Favera R, Nussenzweig MC. Identification of human germinal center light and dark zone cells and their relationship to human B-cell lymphomas. Blood. American Society of Hematology; 2012;120: 2240–8. pmid:22740445
- View Article
- PubMed/NCBI
- Google Scholar
11. Jacob F, Hitchins MP, Fedier A, Brennan K, Nixdorf S, Hacker NF, et al. The glycosphingolipid P1 is an ovarian cancer-associated carbohydrate antigen involved in migration. Br J Cancer. Nature Publishing Group; 2014;111: 1634–1645. pmid:25167227
- View Article
- PubMed/NCBI
- Google Scholar
12. Cooling L. Blood Groups in Infection and Host Susceptibility. Clin Microbiol Rev. 2015;28: 801–870. pmid:26085552
- View Article
- PubMed/NCBI
- Google Scholar
13. van Die I, Cummings RD. Glycan gimmickry by parasitic helminths: A strategy for modulating the host immune response? Glycobiology. 2010;20: 2–12. pmid:19748975
- View Article
- PubMed/NCBI
- Google Scholar
14. Thuresson B, Westman JS, Olsson ML. Identification of a novel A4GALT exon reveals the genetic basis of the P₁/P₂ histo-blood groups. Blood. 2011;117: 678–687. pmid:20971946
- View Article
- PubMed/NCBI
- Google Scholar
15. Lai Y-J, Wu W-Y, Yang C-M, Yang L-R, Chu C-C, Chan Y-S, et al. A systematic study of single-nucleotide polymorphisms in the A4GALT gene suggests a molecular genetic basis for the P₁/P₂ blood groups. Transfusion. 2014;54: 3222–3231. pmid:25041587
- View Article
- PubMed/NCBI
- Google Scholar
16. Eernstman J, Veldhuisen B, Heshusius S, Philipsen S, von Lindern M, Borg J, et al. KLF1 regulates P1 expression through transcriptional control of A4GALT. Vox Sang. 2017;112: S06–2.
- View Article
- Google Scholar
17. Daniels G. Lutheran Blood Group System. Human Blood Groups. Wiley-Blackwell; 2013. pp. 259–277.
18. Shaw MA, Leak MR, Daniels GL, Tippett P. The rare Lutheran blood group phenotype Lu(a-b-): a genetic study. Ann Hum Genet. 1984;48: 229–37. Available: http://www.ncbi.nlm.nih.gov/pubmed/6465841 pmid:6465841
- View Article
- PubMed/NCBI
- Google Scholar
19. Rowe GP, Gale SA, Daniels GL, Green CA, Tippett P. A study on Lu-null families in South Wales. Ann Hum Genet. 1992;56: 267–72. Available: http://www.ncbi.nlm.nih.gov/pubmed/1449238 pmid:1449238
- View Article
- PubMed/NCBI
- Google Scholar
20. Weissgerber TL, Milic NM, Winham SJ, Garovic VD. Beyond Bar and Line Graphs: Time for a New Data Presentation Paradigm. PLOS Biol. Public Library of Science; 2015;13: e1002128. pmid:25901488
- View Article
- PubMed/NCBI
- Google Scholar
21. Hellberg Å, Steffensen R, Yahalom V, Sojka BN, Heier HE, Levene C, et al. Additional molecular bases of the clinically important p blood group phenotype. Transfusion. 2003;43: 899–907. pmid:12823750
- View Article
- PubMed/NCBI
- Google Scholar
22. Duk M, Kusnierz-Alejska G, Korchagina EY, Bovin N V., Bochenek S, Lisowska E. Anti-α-galactosyl antibodies recognizing epitopes terminating with α1,4-linked galactose: Human natural and mouse monoclonal anti-NOR and anti-P1 antibodies. Glycobiology. 2005;15: 109–118. pmid:15342552
- View Article
- PubMed/NCBI
- Google Scholar
23. Samani FS, Moore JK, Khosravani P, Ebrahimi M. Features of free software packages in flow cytometry: a comparison between four non-commercial software sources. Cytotechnology. Springer; 2014;66: 555–9. pmid:23839300
- View Article
- PubMed/NCBI
- Google Scholar
24. Kuśnierz-Alejska G, Duk M, Storry JR, Reid ME, Wiecek B, Seyfried H, et al. NOR polyagglutination and Sta glycophorin in one family: relation of NOR polyagglutination to terminal alpha-galactose residues and abnormal glycolipids. Transfusion. 1999;39: 32–8. pmid:9920164
- View Article
- PubMed/NCBI
- Google Scholar
25. Duk M, Reinhold BB, Reinhold VN, Kusnierz-Alejska G, Lisowska E. Structure of a neutral glycosphingolipid recognized by human antibodies in polyagglutinable erythrocytes from the rare NOR phenotype. J Biol Chem. 2001;276: 40574–40582. pmid:11504714
- View Article
- PubMed/NCBI
- Google Scholar
26. Duk M, Singh S, Reinhold VN, Krotkiewski H, Kurowska E, Lisowska E. Structures of unique globoside elongation products present in erythrocytes with a rare NOR phenotype. Glycobiology. 2007;17: 304–312. pmid:17118951
- View Article
- PubMed/NCBI
- Google Scholar
27. Box GEP, Cox DR. An Analysis of Transformations [Internet]. Journal of the Royal Statistical Society. Series B (Methodological). WileyRoyal Statistical Society; 1964. pp. 211–252.
- View Article
- Google Scholar
28. Kaczmarek R, Buczkowska A, Mikołajewicz K, Krotkiewski H, Czerwinski M. P1PK, GLOB, and FORS Blood Group Systems and GLOB Collection: Biochemical and clinical aspects. Do we understand it all yet? Transfus Med Rev. 2014;28. pmid:24895151
- View Article
- PubMed/NCBI
- Google Scholar
29. Mahfoud R, Manis A, Binnington B, Ackerley C, Lingwood C a. A major fraction of glycosphingolipids in model and cellular cholesterol-containing membranes is undetectable by their binding proteins. J Biol Chem. 2010;285: 36049–36059. pmid:20716521
- View Article
- PubMed/NCBI
- Google Scholar
30. Iwamura K, Furukawa K, Uchikawa M, Sojka BN, Kojima Y, Wiels J, et al. The blood group P1 synthase gene is identical to the Gb3/CD77 synthase gene. A clue to the solution of the P₁/P₂/p puzzle. J Biol Chem. 2003;278: 44429–44438. pmid:12888565
- View Article
- PubMed/NCBI
- Google Scholar
31. Yeh C-C, Chang C-J, Twu Y-C, Hung S-T, Tsai Y-J, Liao J-C, et al. The differential expression of the blood group P¹-A4GALT and P²-A4GALT alleles is stimulated by the transcription factor EGR1. Transfusion. 2018; in press.
- View Article
- Google Scholar
32. Westman JS, Stenfelt L, Vidovic K, Möller M, Hellberg Å, Kjellström S, et al. Allele-selective RUNX1 binding regulates P1 blood group status by transcriptional control ofA4GALT. Blood. 2018; blood-2017-08-803080. pmid:29438961
- View Article
- PubMed/NCBI
- Google Scholar
33. Volynsky P, Efremov R, Mikhalev I, Dobrochaeva K, Tuzikov A, Korchagina E, et al. Why human anti-Galα1-4Galβ1-4Glc natural antibodies do not recognize the trisaccharide on erythrocyte membrane? Molecular dynamics and immunochemical investigation. Mol Immunol. 2017;90: 87–97. pmid:28708979
- View Article
- PubMed/NCBI
- Google Scholar
34. Hung KT, Berisha SZ, Ritchey BM, Santore J, Smith JD. Red blood cells play a role in reverse cholesterol transport. Arterioscler Thromb Vasc Biol. NIH Public Access; 2012;32: 1460–5. pmid:22499994
- View Article
- PubMed/NCBI
- Google Scholar
35. Tziakas DN, Kaski JC, Chalikias GK, Romero C, Fredericks S, Tentes IK, et al. Total Cholesterol Content of Erythrocyte Membranes Is Increased in Patients With Acute Coronary Syndrome. J Am Coll Cardiol. 2007;49: 2081–2089. pmid:17531656
- View Article
- PubMed/NCBI
- Google Scholar
36. Lange Y, Molinaro AL, Chauncey TR, Steck TL. On the mechanism of transfer of cholesterol between human erythrocytes and plasma. J Biol Chem. 1983;258: 6920–6. Available: http://www.ncbi.nlm.nih.gov/pubmed/6853509 pmid:6853509
- View Article
- PubMed/NCBI
- Google Scholar
37. Ramakrishnan B, Shah PS, Qasba PK. α-lactalbumin (LA) stimulates milk β-1,4-galactosyltransferase I (β4Gal-T1) to transfer glucose from UDP-glucose to N-acetylglucosamine: Crystal structure of β4Gal-T1·LA complex with UDP-Glc. J Biol Chem. 2001;276: 37665–37671. pmid:11485999
- View Article
- PubMed/NCBI
- Google Scholar
38. Busse M, Feta A, Presto J, Wilén M, Grønning M, Kjellén L, et al. Contribution of EXT1, EXT2, and EXTL3 to Heparan Sulfate Chain Elongation. J Biol Chem. 2007;282: 32802–32810. pmid:17761672
- View Article
- PubMed/NCBI
- Google Scholar
39. Busse-Wicher M, Wicher KB, Kusche-Gullberg M. The extostosin family: Proteins with many functions. Matrix Biol. 2014;35: 25–33. pmid:24128412
- View Article
- PubMed/NCBI
- Google Scholar
40. McCormick C, Duncan G, Goutsos KT, Tufaro F. The putative tumor suppressors EXT1 and EXT2 form a stable complex that accumulates in the Golgi apparatus and catalyzes the synthesis of heparan sulfate. Proc Natl Acad Sci U S A. 2000;97: 668–673. pmid:10639137
- View Article
- PubMed/NCBI
- Google Scholar
41. Zhu G, Jaskiewicz E, Bassi R, Darling DS, Young W.W Jr. β1,4 N-Acetylgalactosaminyltransferase (GM2/GD2/GA2 synthase) forms homodimers in the endoplasmic reticulum: A strategy to test for dimerization of Golgi membrane proteins. Glycobiology. 1997;7: 987–996. pmid:9363441
- View Article
- PubMed/NCBI
- Google Scholar
42. Hassinen A, Pujol FM, Kokkonen N, Pieters C, Kihlström M, Korhonen K, et al. Functional organization of Golgi N- and O-glycosylation pathways involves pH-dependent complex formation that is impaired in cancer cells. J Biol Chem. 2011;286: 38329–38340. pmid:21911486
- View Article
- PubMed/NCBI
- Google Scholar
43. Takematsu H, Yamamoto H, Naito-Matsui Y, Fujinawa R, Tanaka K, Okuno Y, et al. Quantitative transcriptomic profiling of branching in a glycosphingolipid biosynthetic pathway. J Biol Chem. 2011;286: 27214–27224. pmid:21665948
- View Article
- PubMed/NCBI
- Google Scholar
44. Kaczmarek R, Suchanowska A, Szymczak K, Lisowska E, Czerwinski M. Gb3/CD77 synthase (α1,4-galactosyltransferase) and its variant form, NOR-synthase, exist as dimers. FEBS J. WILEY-BLACKWELL, 111 RIVER ST, HOBOKEN 07030–5774, NJ USA; 2013;280: 530–530.
- View Article
- Google Scholar
45. Schweingruber C, Rufener SC, Zünd D, Yamashita A, Mühlemann O. Nonsense-mediated mRNA decay—Mechanisms of substrate mRNA recognition and degradation in mammalian cells. Biochim Biophys Acta—Gene Regul Mech. 2013;1829: 612–623. pmid:23435113
- View Article
- PubMed/NCBI
- Google Scholar
46. Poole J, Daniels G. Blood Group Antibodies and Their Significance in Transfusion Medicine. Transfus Med Rev. 2007;21: 58–71. pmid:17174221
- View Article
- PubMed/NCBI
- Google Scholar

[ref1] 1. Landsteiner K, Levine P. Further Observations on Individual Differences of Human Blood. Exp Biol Med. SAGE PublicationsSage UK: London, England; 1927;24: 941–942.
View Article
Google Scholar

[2] View Article

[3] Google Scholar

[ref2] 2. Kaczmarek R, Buczkowska A, Mikołajewicz K, Krotkiewski H, Czerwinski M. P1PK, GLOB, and FORS Blood Group Systems and GLOB Collection: Biochemical and Clinical Aspects. Do We Understand It All Yet? Transfus Med Rev. 2014;28: 126–136. pmid:24895151
View Article
PubMed/NCBI
Google Scholar

[5] View Article

[6] PubMed/NCBI

[7] Google Scholar

[ref3] 3. Storry JR, Castilho L, Daniels G, Flegel WA, Garratty G, Francis CL, et al. International Society of Blood Transfusion Working Party on red cell immunogenetics and blood group terminology: Berlin report. Vox Sang. 2011;101: 77–82. pmid:21401621
View Article
PubMed/NCBI
Google Scholar

[9] View Article

[10] PubMed/NCBI

[11] Google Scholar

[ref4] 4. Steffensen R, Carlier K, Wiels J, Levery SB, Stroud M, Cedergren B, et al. Cloning and expression of the histo-blood group P^k UDP-galactose: Galβ1-4Glcβ1-Cer α1,4-galactosyltransferase. Molecular genetic basis of the p phenotype. J Biol Chem. 2000;275: 16723–16729. pmid:10747952
View Article
PubMed/NCBI
Google Scholar

[13] View Article

[14] PubMed/NCBI

[15] Google Scholar

[ref5] 5. Kaczmarek R, Duk M, Szymczak K, Korchagina E, Tyborowska J, Mikolajczyk K, et al. Human Gb3/CD77 synthase reveals specificity toward two or four different acceptors depending on amino acid at position 211, creating P^k, P1 and NOR blood group antigens. Biochem Biophys Res Commun. 2016;470: 168–74. pmid:26773500
View Article
PubMed/NCBI
Google Scholar

[17] View Article

[18] PubMed/NCBI

[19] Google Scholar

[ref6] 6. Suchanowska A, Kaczmarek R, Duk M, Lukasiewicz J, Smolarek D, Majorczyk E, et al. A Single point mutation in the gene encoding Gb3/CD77 synthase causes a rare inherited polyagglutination syndrome. J Biol Chem. 2012;287: 38220–38230. pmid:22965229
View Article
PubMed/NCBI
Google Scholar

[21] View Article

[22] PubMed/NCBI

[23] Google Scholar

[ref7] 7. Kaczmarek R, Mikolajewicz K, Szymczak K, Duk M, Majorczyk E, Krop-Watorek A, et al. Evaluation of an amino acid residue critical for the specificity and activity of human Gb3/CD77 synthase. Glycoconj J. Springer; 2016;33: 963–973. pmid:27538840
View Article
PubMed/NCBI
Google Scholar

[25] View Article

[26] PubMed/NCBI

[27] Google Scholar

[ref8] 8. Lingwood C a., Manis A, Mahfoud R, Khan F, Binnington B, Mylvaganam M. New aspects of the regulation of glycosphingolipid receptor function. Chem Phys Lipids. 2010;163: 27–35. pmid:19781539
View Article
PubMed/NCBI
Google Scholar

[29] View Article

[30] PubMed/NCBI

[31] Google Scholar

[ref9] 9. Högerkorp C-M, Borrebaeck CAK. The human CD77- B cell population represents a heterogeneous subset of cells comprising centroblasts, centrocytes, and plasmablasts, prompting phenotypical revision. J Immunol. 2006;177: 4341–9. pmid:16982868
View Article
PubMed/NCBI
Google Scholar

[33] View Article

[34] PubMed/NCBI

[35] Google Scholar

[ref10] 10. Victora GD, Dominguez-Sola D, Holmes AB, Deroubaix S, Dalla-Favera R, Nussenzweig MC. Identification of human germinal center light and dark zone cells and their relationship to human B-cell lymphomas. Blood. American Society of Hematology; 2012;120: 2240–8. pmid:22740445
View Article
PubMed/NCBI
Google Scholar

[37] View Article

[38] PubMed/NCBI

[39] Google Scholar

[ref11] 11. Jacob F, Hitchins MP, Fedier A, Brennan K, Nixdorf S, Hacker NF, et al. The glycosphingolipid P1 is an ovarian cancer-associated carbohydrate antigen involved in migration. Br J Cancer. Nature Publishing Group; 2014;111: 1634–1645. pmid:25167227
View Article
PubMed/NCBI
Google Scholar

[41] View Article

[42] PubMed/NCBI

[43] Google Scholar

[ref12] 12. Cooling L. Blood Groups in Infection and Host Susceptibility. Clin Microbiol Rev. 2015;28: 801–870. pmid:26085552
View Article
PubMed/NCBI
Google Scholar

[45] View Article

[46] PubMed/NCBI

[47] Google Scholar

[ref13] 13. van Die I, Cummings RD. Glycan gimmickry by parasitic helminths: A strategy for modulating the host immune response? Glycobiology. 2010;20: 2–12. pmid:19748975
View Article
PubMed/NCBI
Google Scholar

[49] View Article

[50] PubMed/NCBI

[51] Google Scholar

[ref14] 14. Thuresson B, Westman JS, Olsson ML. Identification of a novel A4GALT exon reveals the genetic basis of the P₁/P₂ histo-blood groups. Blood. 2011;117: 678–687. pmid:20971946
View Article
PubMed/NCBI
Google Scholar

[53] View Article

[54] PubMed/NCBI

[55] Google Scholar

[ref15] 15. Lai Y-J, Wu W-Y, Yang C-M, Yang L-R, Chu C-C, Chan Y-S, et al. A systematic study of single-nucleotide polymorphisms in the A4GALT gene suggests a molecular genetic basis for the P₁/P₂ blood groups. Transfusion. 2014;54: 3222–3231. pmid:25041587
View Article
PubMed/NCBI
Google Scholar

[57] View Article

[58] PubMed/NCBI

[59] Google Scholar

[ref16] 16. Eernstman J, Veldhuisen B, Heshusius S, Philipsen S, von Lindern M, Borg J, et al. KLF1 regulates P1 expression through transcriptional control of A4GALT. Vox Sang. 2017;112: S06–2.
View Article
Google Scholar

[61] View Article

[62] Google Scholar

[ref17] 17. Daniels G. Lutheran Blood Group System. Human Blood Groups. Wiley-Blackwell; 2013. pp. 259–277.

[ref18] 18. Shaw MA, Leak MR, Daniels GL, Tippett P. The rare Lutheran blood group phenotype Lu(a-b-): a genetic study. Ann Hum Genet. 1984;48: 229–37. Available: http://www.ncbi.nlm.nih.gov/pubmed/6465841 pmid:6465841
View Article
PubMed/NCBI
Google Scholar

[65] View Article

[66] PubMed/NCBI

[67] Google Scholar

[ref19] 19. Rowe GP, Gale SA, Daniels GL, Green CA, Tippett P. A study on Lu-null families in South Wales. Ann Hum Genet. 1992;56: 267–72. Available: http://www.ncbi.nlm.nih.gov/pubmed/1449238 pmid:1449238
View Article
PubMed/NCBI
Google Scholar

[69] View Article

[70] PubMed/NCBI

[71] Google Scholar

[ref20] 20. Weissgerber TL, Milic NM, Winham SJ, Garovic VD. Beyond Bar and Line Graphs: Time for a New Data Presentation Paradigm. PLOS Biol. Public Library of Science; 2015;13: e1002128. pmid:25901488
View Article
PubMed/NCBI
Google Scholar

[73] View Article

[74] PubMed/NCBI

[75] Google Scholar

[ref21] 21. Hellberg Å, Steffensen R, Yahalom V, Sojka BN, Heier HE, Levene C, et al. Additional molecular bases of the clinically important p blood group phenotype. Transfusion. 2003;43: 899–907. pmid:12823750
View Article
PubMed/NCBI
Google Scholar

[77] View Article

[78] PubMed/NCBI

[79] Google Scholar

[ref22] 22. Duk M, Kusnierz-Alejska G, Korchagina EY, Bovin N V., Bochenek S, Lisowska E. Anti-α-galactosyl antibodies recognizing epitopes terminating with α1,4-linked galactose: Human natural and mouse monoclonal anti-NOR and anti-P1 antibodies. Glycobiology. 2005;15: 109–118. pmid:15342552
View Article
PubMed/NCBI
Google Scholar

[81] View Article

[82] PubMed/NCBI

[83] Google Scholar

[ref23] 23. Samani FS, Moore JK, Khosravani P, Ebrahimi M. Features of free software packages in flow cytometry: a comparison between four non-commercial software sources. Cytotechnology. Springer; 2014;66: 555–9. pmid:23839300
View Article
PubMed/NCBI
Google Scholar

[85] View Article

[86] PubMed/NCBI

[87] Google Scholar

[ref24] 24. Kuśnierz-Alejska G, Duk M, Storry JR, Reid ME, Wiecek B, Seyfried H, et al. NOR polyagglutination and Sta glycophorin in one family: relation of NOR polyagglutination to terminal alpha-galactose residues and abnormal glycolipids. Transfusion. 1999;39: 32–8. pmid:9920164
View Article
PubMed/NCBI
Google Scholar

[89] View Article

[90] PubMed/NCBI

[91] Google Scholar

[ref25] 25. Duk M, Reinhold BB, Reinhold VN, Kusnierz-Alejska G, Lisowska E. Structure of a neutral glycosphingolipid recognized by human antibodies in polyagglutinable erythrocytes from the rare NOR phenotype. J Biol Chem. 2001;276: 40574–40582. pmid:11504714
View Article
PubMed/NCBI
Google Scholar

[93] View Article

[94] PubMed/NCBI

[95] Google Scholar

[ref26] 26. Duk M, Singh S, Reinhold VN, Krotkiewski H, Kurowska E, Lisowska E. Structures of unique globoside elongation products present in erythrocytes with a rare NOR phenotype. Glycobiology. 2007;17: 304–312. pmid:17118951
View Article
PubMed/NCBI
Google Scholar

[97] View Article

[98] PubMed/NCBI

[99] Google Scholar

[ref27] 27. Box GEP, Cox DR. An Analysis of Transformations [Internet]. Journal of the Royal Statistical Society. Series B (Methodological). WileyRoyal Statistical Society; 1964. pp. 211–252.
View Article
Google Scholar

[101] View Article

[102] Google Scholar

[ref28] 28. Kaczmarek R, Buczkowska A, Mikołajewicz K, Krotkiewski H, Czerwinski M. P1PK, GLOB, and FORS Blood Group Systems and GLOB Collection: Biochemical and clinical aspects. Do we understand it all yet? Transfus Med Rev. 2014;28. pmid:24895151
View Article
PubMed/NCBI
Google Scholar

[104] View Article

[105] PubMed/NCBI

[106] Google Scholar

[ref29] 29. Mahfoud R, Manis A, Binnington B, Ackerley C, Lingwood C a. A major fraction of glycosphingolipids in model and cellular cholesterol-containing membranes is undetectable by their binding proteins. J Biol Chem. 2010;285: 36049–36059. pmid:20716521
View Article
PubMed/NCBI
Google Scholar

[108] View Article

[109] PubMed/NCBI

[110] Google Scholar

[ref30] 30. Iwamura K, Furukawa K, Uchikawa M, Sojka BN, Kojima Y, Wiels J, et al. The blood group P1 synthase gene is identical to the Gb3/CD77 synthase gene. A clue to the solution of the P₁/P₂/p puzzle. J Biol Chem. 2003;278: 44429–44438. pmid:12888565
View Article
PubMed/NCBI
Google Scholar

[112] View Article

[113] PubMed/NCBI

[114] Google Scholar

[ref31] 31. Yeh C-C, Chang C-J, Twu Y-C, Hung S-T, Tsai Y-J, Liao J-C, et al. The differential expression of the blood group P¹-A4GALT and P²-A4GALT alleles is stimulated by the transcription factor EGR1. Transfusion. 2018; in press.
View Article
Google Scholar

[116] View Article

[117] Google Scholar

[ref32] 32. Westman JS, Stenfelt L, Vidovic K, Möller M, Hellberg Å, Kjellström S, et al. Allele-selective RUNX1 binding regulates P1 blood group status by transcriptional control ofA4GALT. Blood. 2018; blood-2017-08-803080. pmid:29438961
View Article
PubMed/NCBI
Google Scholar

[119] View Article

[120] PubMed/NCBI

[121] Google Scholar

[ref33] 33. Volynsky P, Efremov R, Mikhalev I, Dobrochaeva K, Tuzikov A, Korchagina E, et al. Why human anti-Galα1-4Galβ1-4Glc natural antibodies do not recognize the trisaccharide on erythrocyte membrane? Molecular dynamics and immunochemical investigation. Mol Immunol. 2017;90: 87–97. pmid:28708979
View Article
PubMed/NCBI
Google Scholar

[123] View Article

[124] PubMed/NCBI

[125] Google Scholar

[ref34] 34. Hung KT, Berisha SZ, Ritchey BM, Santore J, Smith JD. Red blood cells play a role in reverse cholesterol transport. Arterioscler Thromb Vasc Biol. NIH Public Access; 2012;32: 1460–5. pmid:22499994
View Article
PubMed/NCBI
Google Scholar

[127] View Article

[128] PubMed/NCBI

[129] Google Scholar

[ref35] 35. Tziakas DN, Kaski JC, Chalikias GK, Romero C, Fredericks S, Tentes IK, et al. Total Cholesterol Content of Erythrocyte Membranes Is Increased in Patients With Acute Coronary Syndrome. J Am Coll Cardiol. 2007;49: 2081–2089. pmid:17531656
View Article
PubMed/NCBI
Google Scholar

[131] View Article

[132] PubMed/NCBI

[133] Google Scholar

[ref36] 36. Lange Y, Molinaro AL, Chauncey TR, Steck TL. On the mechanism of transfer of cholesterol between human erythrocytes and plasma. J Biol Chem. 1983;258: 6920–6. Available: http://www.ncbi.nlm.nih.gov/pubmed/6853509 pmid:6853509
View Article
PubMed/NCBI
Google Scholar

[135] View Article

[136] PubMed/NCBI

[137] Google Scholar

[ref37] 37. Ramakrishnan B, Shah PS, Qasba PK. α-lactalbumin (LA) stimulates milk β-1,4-galactosyltransferase I (β4Gal-T1) to transfer glucose from UDP-glucose to N-acetylglucosamine: Crystal structure of β4Gal-T1·LA complex with UDP-Glc. J Biol Chem. 2001;276: 37665–37671. pmid:11485999
View Article
PubMed/NCBI
Google Scholar

[139] View Article

[140] PubMed/NCBI

[141] Google Scholar

[ref38] 38. Busse M, Feta A, Presto J, Wilén M, Grønning M, Kjellén L, et al. Contribution of EXT1, EXT2, and EXTL3 to Heparan Sulfate Chain Elongation. J Biol Chem. 2007;282: 32802–32810. pmid:17761672
View Article
PubMed/NCBI
Google Scholar

[143] View Article

[144] PubMed/NCBI

[145] Google Scholar

[ref39] 39. Busse-Wicher M, Wicher KB, Kusche-Gullberg M. The extostosin family: Proteins with many functions. Matrix Biol. 2014;35: 25–33. pmid:24128412
View Article
PubMed/NCBI
Google Scholar

[147] View Article

[148] PubMed/NCBI

[149] Google Scholar

[ref40] 40. McCormick C, Duncan G, Goutsos KT, Tufaro F. The putative tumor suppressors EXT1 and EXT2 form a stable complex that accumulates in the Golgi apparatus and catalyzes the synthesis of heparan sulfate. Proc Natl Acad Sci U S A. 2000;97: 668–673. pmid:10639137
View Article
PubMed/NCBI
Google Scholar

[151] View Article

[152] PubMed/NCBI

[153] Google Scholar

[ref41] 41. Zhu G, Jaskiewicz E, Bassi R, Darling DS, Young W.W Jr. β1,4 N-Acetylgalactosaminyltransferase (GM2/GD2/GA2 synthase) forms homodimers in the endoplasmic reticulum: A strategy to test for dimerization of Golgi membrane proteins. Glycobiology. 1997;7: 987–996. pmid:9363441
View Article
PubMed/NCBI
Google Scholar

[155] View Article

[156] PubMed/NCBI

[157] Google Scholar

[ref42] 42. Hassinen A, Pujol FM, Kokkonen N, Pieters C, Kihlström M, Korhonen K, et al. Functional organization of Golgi N- and O-glycosylation pathways involves pH-dependent complex formation that is impaired in cancer cells. J Biol Chem. 2011;286: 38329–38340. pmid:21911486
View Article
PubMed/NCBI
Google Scholar

[159] View Article

[160] PubMed/NCBI

[161] Google Scholar

[ref43] 43. Takematsu H, Yamamoto H, Naito-Matsui Y, Fujinawa R, Tanaka K, Okuno Y, et al. Quantitative transcriptomic profiling of branching in a glycosphingolipid biosynthetic pathway. J Biol Chem. 2011;286: 27214–27224. pmid:21665948
View Article
PubMed/NCBI
Google Scholar

[163] View Article

[164] PubMed/NCBI

[165] Google Scholar

[ref44] 44. Kaczmarek R, Suchanowska A, Szymczak K, Lisowska E, Czerwinski M. Gb3/CD77 synthase (α1,4-galactosyltransferase) and its variant form, NOR-synthase, exist as dimers. FEBS J. WILEY-BLACKWELL, 111 RIVER ST, HOBOKEN 07030–5774, NJ USA; 2013;280: 530–530.
View Article
Google Scholar

[167] View Article

[168] Google Scholar

[ref45] 45. Schweingruber C, Rufener SC, Zünd D, Yamashita A, Mühlemann O. Nonsense-mediated mRNA decay—Mechanisms of substrate mRNA recognition and degradation in mammalian cells. Biochim Biophys Acta—Gene Regul Mech. 2013;1829: 612–623. pmid:23435113
View Article
PubMed/NCBI
Google Scholar

[170] View Article

[171] PubMed/NCBI

[172] Google Scholar

[ref46] 46. Poole J, Daniels G. Blood Group Antibodies and Their Significance in Transfusion Medicine. Transfus Med Rev. 2007;21: 58–71. pmid:17174221
View Article
PubMed/NCBI
Google Scholar

[174] View Article

[175] PubMed/NCBI

[176] Google Scholar

Figures

Abstract

Introduction

Materials and methods

Sample preparation and P1/P2 blood typing

Identification of the individual with PP1PK null (p) phenotype

Analysis of the SNPs

Quantitative analysis of transcripts

Antibodies

Flow cytometry

Quantitative flow cytometric analysis of the P1 antigen

Lipid panels

Extraction and purification of glycosphingolipids

Statistical analysis

Results

SNPs that best predict the P1/P2 blood type

Anti-P1 antibody binding capacities of RBCs with different genotypes

Precision of phenotyping using the four SNPs

Influence of the NOR mutation on the activity of Gb3/CD77 synthase

Plasma cholesterol as a confounder of P1 expression

Discussion

Supporting information

S1 Fig. HPTLC analysis of neutral glycosphingolipids extracted from RBCs with different genotypes: P1NORP1 (lane 1), P1P1 (lane 2), P2P2 (lane 3), and pp (lane 4).

S2 Fig. Scatter plots of relationships between Box-Cox transformed RBC anti-P1 binding capacities, HDL, LDL and total cholesterol.

S1 Table. Phenotypes, genotypes, RBC antibody binding capacities and lipid profiles of recruited individuals.

S2 Table. Sequences of primers used for genotyping.

S3 Table. PCR conditions used for amplification of A4GALT fragments encompassing the studied SNPs.

S4 Table. Real-time PCR conditions used for A4GALT gene expression assays.

Acknowledgments

References

Sample preparation and P₁/P₂ blood typing

SNPs that best predict the P₁/P₂ blood type

S1 Fig. HPTLC analysis of neutral glycosphingolipids extracted from RBCs with different genotypes: P^1NORP¹ (lane 1), P¹P¹ (lane 2), P²P² (lane 3), and pp (lane 4).