Characterization of GYP(B-A-B) hybrid glycophorins among Thai blood donors with Mi^a-positive phenotypes

Abstract

Background

GYPA and GYPB genes encode the antigens of the MNS blood group system carried on glycophorin A (GPA) and glycophorin B (GPB), or on a hybrid molecule of GPA and GPB. GP hybrid variants are created through unequal crossing over and gene conversion, typically from the parent genes GYPA and GYPB. In the present study, we characterized the GYP(B-A-B) hybrid variants among Thai blood donors with Mi^a-positive phenotypes using PCR-based coupled to DNA sequencing techniques.

Materials and methods

Altogether, 1,020 samples from Thai blood donors were tested with anti-Mi^a by conventional tube technique (CTT). Polymerase chain reaction with sequence-specific primer (PCR-SSP) was initially used to differentiate normal GYPB, GYP*Vw and groups of GYP*Hut, GYP*Mur, GYP*Hop, GYP*Bun and GYP*HF alleles. Subsequently, GYP(B-A-B) hybrid variants were investigated using DNA sequencing.

Results

Among 1,020 blood donors, 127 (12.45%) were Mi(a+) phenotypes. The comparison Mi^a typing results between CTT and PCR-SSP were concordant. All Mi(a+) samples were positive with only group of GYP*Hut, GYP*Mur, GYP*Hop, GYP*Bun and GYP*HF alleles by PCR-SSP. Regarding the sequencing results, 115/1,020 (11.27%) donors carried the GYP*Mur, of which 111/1,020 (10.88%) were GYP*Mur/GYPB heterozygotes and the other 4/1,020 (0.39%) donors were GYP*Mur/GYP*Mur homozygotes. The remaining 12 donors included different GYP*Bun-like alleles; 11 of them (1.08%) were GYP*Thai/GYPB heterozygotes, and one (0.10%) was GYP*Thai II/GYPB heterozygotes. With 5.83% (119/2,040) of the total hybrid alleles, GYP*Mur was the predominant allele. The GYP*HF, GYP*Bun, GYP*Hop and GYP*Kip alleles were not observed in this study.

Discussion

Regarding the hybrid GP variants, a consensus of observed prevalent GYP*Mur and GYP*Bun-like alleles, respectively, was identified in the Thai population. The introduction of our strategy has allowed us to identify the zygosity for GYP hybrid variants, particularly GYP(B-A-B) hybrid genes, when antisera are unavailable and lacking adequate phenotypic features to determine GP variants.

INTRODUCTION

The MNS system is highly polymorphic; to date, 50 antigens have been described¹, and 36 are low-frequency antigens^1,2. These antigens are carried on glycophorin A (GPA), glycophorin B (GPB), or hybrid GP variants, and most are produced by the hybrid formations resulting from crossing-over or gene conversion events that occurred between the highly homologous GYPA, GYPB and infrequently, GYPE within the GYP gene family. Generally, gene conversion occurs from a mechanism of double-strand break repair in which a short sequence of one gene is copied to the other during gene duplications³. Gene conversion events between GYPA and GYPB genes result in GP hybrid alleles, including GYP(A-B), GYP(B-A-B), GYP(A-B-A) and GYP(B-A) hybrid genes. The “repair mechanism” in the GYP(B-A-B) hybrids modifies the region in GYPB, corresponding to GYPA exon 3, replacing an inactive splice site sequence in GYPB with an active GYPA sequence. As a result, a hybrid protein is produced from a partial GYPB nucleotide sequence encoded by the so-called GYPB pseudoexon³.

Presently, eight hybrid GP variants express the Mi^a (MNS7) antigen. Identification of the Mi^a-positive phenotype, anti-Mi^a could react: GP(A-B-A) hybrids, GP.Vw and GP.Hut; GP(B-A-B) hybrids, GP.Mur, GP.Hop, GP.Bun, GP.HF and GP.Kip; GP(B-A), GP.MOT red blood cells (RBCs)². The high prevalence of Mi^a antigen presented in Southeast Asian populations is inferable to the developed anti-Mi^a by alloimmunization involved in mild and severe cases of hemolytic transfusion reactions (HTRs) and hemolytic disease of the fetus and newborn (HDFN)^4–8. The frequency of Mi^a-positive individuals ranges from 4.7 to 22.3% depending on different regions of Thailand^9–13. Therefore, including the typing results of Mi^a antigen in the reagent red blood cells (RBCs) is essential to detect antibodies among patients and blood donors. A related study revealed that among 94 Thai blood donors possessing the Mi^a-positive phenotype, GP.Mur was the most prevalent variant (88.3%), followed by GP.Bun (11.7%) confirmed by DNA sequencing⁹.

As a result, anti-Mi^a, particularly anti-Mur can be detected among patients and donors. Other rare GP alloantibodies such as anti-Hop and anti-Hil might accidentally be identified among Thai patients with suspected anti-Mi^a due to the incomplete patterns of antibodies reacting with Mi^a-positive panel cells. Additional testing with extra panel cells of other GP phenotypes and testing the patient’s RBCs with specific antiserum is required^14,15. Moreover, a transfusion-dependent Thai patient had anti-E, -c, -Jk^b, -S and a panreactive alloantibody, subsequently identified as anti-JENU. This antibody is specific to a high-frequency antigen on GPB and can be produced by homozygous GP.Mur individuals (also known as JENU−). Only two of random 3,600 red cell units used in the mass screening were compatible with this patient¹⁶.

The exact identification of GP variants by serologic methods alone is challenging due to the cross-reactivity of the epitopes and the identical antigens displayed on several variations. Hence, molecular techniques have been developed to determine the hybrid genes for screening Mi^a-positive donors in Thai populations, such as polymerase chain reaction with sequence-specific primer (PCR-SSP)⁹ and multiplex PCR¹⁰. Interestingly, the real-time PCR melting curve technique has been implemented to differentiate the GYP*Mur allele from other Mi^a-positive phenotypes¹⁷. The GYP*Mur, GYP*Hop and GYP*Bun alleles could be identified colorimetrically using gold nanoparticle oligonucleotide probes¹⁸. However, these techniques are still unable to discriminate between the eight hybrid GP variants expressing the Mi^a antigen. A related study revealed that among 63 of 396 (15.91%) Thai blood donors with Mi^a-positive phenotypes, predicted by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS), exhibited a predominant GYP*Mur allele, followed by novel GYP*Bun-like (GYP*Thai) hybrid alleles¹⁹. Therefore, this study aimed to characterize the GYP(B-A-B) hybrid variants among Thai blood donors with Mi^a-positive phenotypes by PCR-based coupled to DNA sequencing techniques.

MATERIALS AND METHODS

Subjects and serological testing

Blood samples in ethylenediaminetetraacetic acid disodium salt (EDTA) tubes were collected from 1,020 first-time blood donors at the Blood Bank, Thammasat University Hospital, from January to May 2023. The research protocol was approved by the Human Research Ethics Committee of Thammasat University (Science), (HREC-TUSc) Pathumthani, Thailand (COE No. 019/2564). All samples were typed for the Mi^a antigen using a conventional tube technique (CTT) using human monoclonal IgM anti-Mi^a (National Blood Centre, Thai Red Cross Society, Bangkok, Thailand). The Mi(a+) samples were typed for S and s antigens by indirect antiglobulin test (IAT) using human IgG anti-S and anti-s (CE-Immundiagnostika GmbH, Eschelbronn, Germany). To confirm the presence of S and s antigens; the direct antiglobulin test (DAT) was performed to rule out false positive results. The IAT and DAT by CTT were performed according to the manufacturer instructions.

Extracting DNA

Genomic DNA was isolated from the samples using the DNeasy Blood & Tissue Kit (QIAGEN, Valencia, CA, USA) according to the manufacturer instructions, then stored at −20°C before further analysis.

Molecular typing by PCR-SSP

The PCR-SSP was initially performed to confirm the presence or absence of Mi^a, S and s antigens at the molecular level. Specific PCR amplification of GYPB*S and GYPB*s alleles in the GYPB gene was performed. A co-amplification of the human growth hormone (HGH) gene was run as the internal control. The PCR amplification conditions and primers used were similar to those described²⁰. Additionally, two primer sets targeting the GYP(A-B-A) and GYP(B-A-B) hybrid genes were used along with HGH-specific control primers. The fragments of the sets of GYP*Hut, GYP*Mur, GYP*Hop, GYP*Bun and GYP*HF, and the fragment of the other set of GYP*Vw were amplified to identify possibly predicted GP variants in conditions previously described⁹.

Identifying GYP(B-A-B) hybrid alleles by DNA sequencing

A genomic region in GYPB exons 2 and 3, was targeted by PCR amplification to characterize GYP(B-A-B) hybrid alleles. Regarding inserting different portions of the active GYPA sequences, it enabled us to identify the GYP*Mur, GYP*Bun, GYP*Bun-like (GYP*Thai and GYP*Thai II), GYP*Hop, GYP*HF, and GYP*Kip variants using the method of DNA sequencing. The PCR amplification reaction system for the targeted region amplification was as follows: template DNA (50 ng/μL) 3 μL, forward primer (GYPB-2564-F: 5′-CCCTAGCAGATGGAGACACTG-3′) 3 μL, reverse primer (GYPB-2564-R: 5′-CTTTGTCTTTACAATTTCGTGTGAA-3′) 3 μL, 2× PCR reaction mixture (Phusion High-Fidelity PCR Master Mix, New England BioLabs, MA, USA) 20 μL, double distilled water 11 μL, for a total volume of 40 μL at 98°C; initial denaturation for 30 s, 98°C for 10 s and 68°C for 1 min, for 10 cycles, 98°C for 10 s, 68°C for 50 s and 72°C for 1 min 30 s, for 25 cycles, and finally 72°C for 5 min. Thereafter, amplification of PCR products was separated on a 1.5% agarose gel containing SYBR Safe DNA Gel Stain (Invitrogen, Paisley, UK), electrophoresed in 1× TBE buffer at 100 volts, and visualized under a blue light transilluminator. The PCR bands were cut and purified, following the instructions provided with a gel extraction kit (GeneJET Gel Extraction Kit, Thermo Scientific, Waltham, MA, USA) and the eluted fragments were then sequenced by next generation sequencing-based technology on the Illumina Platform (BTSeqTM Services, Celemics Inc., Seoul, Korea). The sequencing results were analyzed and compared with those in GenBank (GenBank reference: GYPA, NG_007470.3; GYPB, NG_007483.2; GYPE, NG_009173.1) with BioEdit software (UNT Computing for Arts & Sciences, Denton, TX, USA).

Statistical analysis

The prevalence of the observed variant alleles was described using descriptive statistics and expressed in numbers and percentages. All statistical analyses were conducted using SPSS, Version 16.0 (SPSS Inc., Chicago, IL, USA).

RESULTS

Serological testing and PCR-SSP

A total of 1,020 blood samples were tested using anti-Mi^a by CTT, 127 (12.45%) donor samples were Mi(a+) phenotypes. According to the manufacturer instructions, anti-Mi^a could recognize and bind to RBCs of the Mi^a-positive phenotypes consisting of GP.Vw, GP.Hut, GP.Mur, GP.Hop, GP.Bun and GP.HF. Moreover, each sample was subjected to PCR-SSP analysis. All Mi(a+) samples were positive with only the set of primers specific for GYP*Hut, GYP*Mur, GYP*Hop, GYP*Bun and GYP*HF. No GYP*Vw allele was identified using PCR-SSP. In contrast, 893 Mi(a−) samples showed negative results for both sets of primers specific for GYP*Hut, GYP*Mur, GYP*Hop, GYP*Bun, GYP*HF and GYP*Vw. In all samples tested, the comparison results between CTT and PCR-SSP were concordant. Of those 127 Mi^a-positive donors, the most frequent phenotype was S−s+ (97.64%) followed by S+s+ (2.36%). The S+s− phenotype was not found. In addition, GYPB*S and GYPB*s allele detections using PCR-SSP were tested and the phenotyping, and genotyping results were 100% concordant (Table I).

Table I.

Frequencies of predicted Mi^a, S and s antigens using PCR-SSP

Phenotype	PCR-SSP		No. (%)	Phenotype	PCR-SSP		No. (%)
	GYP*Vw	GYP*Hut, GYP*Mur, GYP*Hop, GYP*Bun and GYP*HF			GYPB*S	GYPB*s
Mi(a+)	Negative	Positive	127 (12.45)	S+s−	Positive	Negative	0 (0.00)
	Positive	Negative	0 (0.00)	S−s+	Negative	Positive	124 (97.64)
	Positive	Positive	0 (0.00)	S+s+	Positive	Positive	3 (2.36)
Mi(a−)	Negative	Negative	893 (87.55)	Total			127 (100.00)
Total			1,020 (100.00)

Identifying GYP(B-A-B) hybrid alleles by DNA sequencing

The differences in nucleotide positions of GYPA, GYPB, and GYP(B-A-B) hybrids are represented in Figure 1. The GYP(B-A-B) hybrid genes comprise GYP*Mur, GYP*Bun, GYP*Bun-like (GYP*Thai and GYP*Thai II), GYP*Hop, GYP*HF and GYP*Kip variants. Sequencing results have been characterized by the insertion of various regions of the active GYPA sequences. The GYP*HF allele revealed the longest GYPA insert of 161 nucleotides that code for 24 amino acids, whereas the shorter insertion of the GYPA segment presents in the GYP*Thai allele. The GYP*Thai allele has a minimum GYPA-derived sequence insert of 22 nucleotides and a maximum of 53 nucleotides. The GYP*Thai II allele has a minimum insert of 46 nucleotides and a maximum of 117 nucleotides. Moreover, sequencing results revealed an insertion of 118 GYPA-derived nucleotides, encoding nine amino acids in the GYP*Mur allele. The GYP*Bun, GYP*Hop and GYP*Kip alleles have a GYPA insert of 131, 131 and 108 nucleotides, carrying the shortest GPA insert of 7 amino acids. The SNP encoded by exon 4, at c.236T/C (analogous to the GYPB*S/GYPB*s defining GYPB c.143T/C) identifies GYP*Hop (c.236T) from the GYP*Bun and GYP*Kip (c.236C) alleles. The GYP*Bun and GYP*Kip alleles differ by a SNP (rs34498102), c.209A/C (Figure 1).

Figure 1.

Differences in nucleotide positions of GYPA, GYPB, and GYP(B-A-B) hybrids

The blue and green highlights demonstrate the GYPA nucleotides inserted into GYPB pseudoexon 3 sequences. Minimum (light green highlight) and maximum (dark green highlight) sequences are identified between the breakpoint boundaries in intron 3. Red letter indicates a SNP (rs34498102) of GYP*Kip allele. The SNP encoded by exon 4 at c.236T/C (analogous to the GYPB*S/GYPB*s defining GYPB c.143T/C) identifies GYP*Hop (c.236T) from the GYP*Bun and GYP*Kip (c.236C).

DNA sequencing analysis results are demonstrated in Table II. All 893 Mi(a−) samples were normal GYPB homozygote. Of the other 127 Mi(a+) GYP(B-A-B) hybrid samples, 115/1,020 (11.27%) donors carried the GYP*Mur, of which 111/1,020 (10.88%) were GYP*Mur/GYPB heterozygotes and the other 4/1,020 (0.39%) donors were GYP*Mur/GYP*Mur homozygotes. The remaining 12 donors included different GYP*Bun-like alleles; 11 of them (1.08%) were GYP*Thai/GYPB heterozygotes, and one (0.10%) was GYP*Thai II/GYPB heterozygotes. The GYP*Mur allele was the predominant hybrid allele and was observed in 119/2,040 (5.83%) alleles. The GYP*HF, GYP*Bun, GYP*Hop and GYP*Kip alleles were not observed in this study.

Table II.

Distribution of GYP(B-A-B) hybrid alleles in the Thai population

Genotype	Genotype frequency No.=1,020 (%)	Allele	Allele frequency No.=2,040 (%)
GYPB/GYPB	893 (87.55)	GYPB	1,909 (93.58)
GYP*Mur/GYPB	111 (10.88)	GYP*Mur	119 (5.83)
GYP*Mur/GYP*Mur	4 (0.39)
GYP*Thai/GYPB	11 (1.08)	GYP*Thai	11 (0.54)
GYP*Thai II/GYPB	1 (0.10)	GYP*Thai II	1 (0.05)

Estimating the risk for Mi^a alloimmunization among Thai and other populations

The estimated risk of Mi^a alloimmunization was obtained by multiplying the probability of being predicted by the Mi^a-negative phenotype frequency by the predicted Mi^a-positive phenotype frequency²⁰. The frequencies of predicted Mi^a phenotypes in the Thai population were compared with related studies among Thai and other Asian populations. From low to high sorting, the Indonesian population²¹ indicated the lowest risk of Mi^a alloimmunization, followed by Taiwanese²¹, southern Thai¹³, Vietnamese²¹, southern Han Chinese²², Filipino²¹, central Thai⁹, Chinese (Guangzhou)²³, Thai (this study) and northern Thai¹⁰ populations, respectively (Table III).

Table III.

Estimation of the risk for Mi^a alloimmunization among Thai and other Asian populations

Populationref	No.	Predicted Mia phenotype frequency		Risk of Mia alloimmunization
		Negative	Positive
Thai (this study)	1,020	0.875	0.125	0.109
Northern Thai 10	300	0.777	0.223	0.173
Southern Thai 13	400	0.953	0.047	0.045
Central Thai 9	1,041	0.910	0.090	0.082
Vietnamese 21	160	0.938	0.062	0.058
Taiwanese 21	167	0.958	0.042	0.040
Indonesian 21	285	0.982	0.018	0.018
Filipino 21	262	0.924	0.076	0.070
Southern Han Chinese 22	3,104	0.935	0.065	0.061
Chinese (Guangzhou) 23	528	0.903	0.097	0.088

DISCUSSION

Low frequency antigens of the MNS blood group system, especially the Mi^a antigen, are clinically relevant in Southeast Asian populations owing to the high prevalence of Mi^a alloimmunization and HTRs and HDFN cases^4–8. Routinely, testing RBCs among patients and donors with anti-Mi^a is convenient and can identify Mi^a-positive and Mi^a-negative individuals. However, serological testing limitations include unavailable specific antiserum to determine eight hybrid GP variants showing Mi^a-positive phenotypes. Among Thai patients with suspected anti-Mi^a, the inconclusive patterns of agglutination reactions with Mi^a-positive panel cells required extra cells with known GP subclasses to determine antibody specificities such as anti-Hop and anti-Hil^14,15. Further testing using molecular techniques are beneficial in clarifying the frequencies of GP subclasses in populations^{17–19,22–24}.

In this study, we initially screened GP variants using human monoclonal anti-Mi^a to distinguish between Mi^a-positive and Mi^a-negative phenotypes, and 127 of 1,020 blood samples were Mi^a-positive representing a frequency of 12.45%, which is consistent with related studies in central Thai blood donor populations^9,25. PCR-based coupled to DNA sequencing techniques were subsequently performed to characterize the GYP*(B-A-B) hybrid GP variants among blood donors with Mi^a-positive phenotypes. The PCR-SSP testing results in 1,020 samples agreed with the serological testing results. Moreover, the GYP*Vw allele was not found using PCR-SSP, similar to related studies in Thai populations^9,19. The occurrence of GP.Vw was reported at 1.43% in Southern Switzerland and 0.057% in Caucasian populations²⁶.

Among the 127 Mi^a-positive donors from the initial screening using the serological test and PCR-SSP, all samples were further sequenced to determine the GYP(B-A-B) hybrids. The GYP*Mur hybrid allele was the most prevalent (5.83%), followed by GYP*Thai (0.59%) consisting of GYP*Thai and GYP*Thai II. A DNA sequencing technique employed to authenticate all 893 Mi^a-negative donors identified normal GYPB homozygotes. The GYP*HF, GYP*Bun, GYP*Hop and GYP*Kip alleles were not found in this study. The related studies of GP. subclasses of Mi^a-positive Thai blood donors revealed that GP.Mur was the most common (93.0%) and the other was GP.Bun (7.0%)^25,27. Thereafter, the other study determined GP. subclasses of Mi^a-positive donors using DNA sequencing confirming that GP.Mur and GP.Bun were observed in 88.0 and 12.0%, respectively⁹. A recent study reported the GYP*Mur and GYP*Bun-like –GYP*Thai and GYP*Thai II–alleles were identified in 88.9 and 11.1%, respectively¹⁹. All the above mentioned studies describe multiple concordance observations in which the high prevalence of GP.Mur and GP.Bun was classified among Thais. In this study a similarity in predicted GP.Mur and GP.Bun frequencies was 90.6% (115/127) and 9.4% (12/127), in rank. In addition, we also identified GYP*Bun-like alleles which was consistent with the study reported by Jongruamklang and colleagues¹⁹. However, the GYP*Bun-like alleles with a shorter insertion of noncoding GYPA sequences spanning intron 3 (Figure 1) remain sufficient to predict the expression of the GP.Bun hybrid protein.

The serologic and PCR-SSP testing results among 127 Mi^a-positive donors confirmed that the S−s+ phenotype was the most common (97.6%), and the other three S+s+ individuals (2.4%) were GYP*Mur/GYPB. A related study reported by Jongruamklang and colleagues revealed that 51 of GYP*Mur/GYPB individuals (100%) had S−s+ phenotype¹⁹. Regarding the qualitatively and semiquantitatively altered s antigen in GP. hybrids, the study demonstrated that the s antigen carried by GP.Mur or GP.Bun was not detected by IgM monoclonal anti-s (P3BER), but the use of two IgG anti-s showed improved reactivity. We have also detected the s antigen carried by these hybrids without any discrepancies in typing results obtained by serological and molecular approaches, supporting the recent report that IgG anti-s (polyclonal) used in our study could discriminate the s phenotypes¹⁹. In addition to the report of alloanti-JENU, a Thai patient carrying JENU−, GYP*Mur/GYP*Mur homozygote lacked normal GPB and had been alloimmunized by exposure to RBCs from donors expressing GP hybrid heterozygotes (Mi^a-positive) or normal GPB (Mi^a-negative)¹⁶. In this study, 115 donors possessed GYP*Mur, of which four samples were homozygous GYP*Mur/GYP*Mur. The probability of finding alloanti-JENU in JENU− individuals may be prone to observe in Thai patient populations, especially in chronic transfusion therapy. These patients could produce alloantibodies with multiple antibody specificities and had trouble searching for compatible blood units by underlying unidentified alloantibodies (perhaps anti-JENU). Therefore, our genotyping strategy using PCR-based coupled to DNA sequencing techniques plays a role in identifying JENU− patients and donors, as well as providing additional reagent RBCs to determine hidden unidentified anti-JENU. Moreover, the estimated risk of Mi^a alloimmunization of central Thais was compared with two Thai populations and another population^{9,10,13,21–23}. For the highest risk of Mi^a alloimmunization in northern Thais (0.173), detecting anti-JENU proneness increased in this population. Similar expectations of anti-JENU production may be observed in Chinese populations.

The advantages of using PCR-based coupled to DNA sequencing techniques are first differentiation of Mi^a-positive and Mi^a-negative samples. Second, the GYP*Vw, GYP(A-B-A) variant can be excluded from GYP*Hut, GYP*Mur, GYP*Hop, GYP*Bun and GYP*HF. Third, the discrimination of those alleles in GYP(B-A-B) variants can be identified by DNA sequencing, leading to determining the zygosity for GYP hybrid variants. In addition, determining GP.Mur homozygote individuals is beneficial in identifying the antibodies related to anti-Mi^a: (1) weak anti-Mi^a using reagent panel cells from GP.Mur homozygotes^28,29, (2) anti-U produced by U+ and GP.Mur homozygous individuals³⁰ and (3) anti-JENU triggered by GP.Mur homozygous individuals¹⁶.

The limitations of this study using PCR-based coupled DNA sequencing techniques included GYP(A-B-A) variant with Mi^a-positive hybrid GP, GYP*Hut allele could not be differentiated; however, the distribution of the GP.Hut was rare (0.04%) in the Thai population, as reported by Chandanayingyong and colleagues³¹. This allele warrants further studies to determine DNA sequencing in larger population samples. Even though the Mi^a-positive phenotype whose GYP*MOT allele, a GYP(B-A) hybrid gene could not be determined, it has been recently discoverd only in one Japanese donor^2,32.

CONCLUSIONS

We have characterized the GYP(B-A-B) variant genes in Mi^a-positive Thai donors by PCR-based coupled with DNA sequencing techniques. For GP variant alleles, a consensus of observed prevalent GYP*Mur and GYP*Bun-like alleles, respectively, was identified in the Thai population. The introduction of our strategy has allowed us to identify the zygosity for GYP hybrid variants, particularly GYP(B-A-B) hybrid genes, when antisera are unavailable and lacking adequate phenotypic features to determine GP variants.