Discovered by Karl Landsteiner in 1900, the ABO blood type is the most important blood type for transfusions [1]. For safe blood transfusions, cell type and serum type test results of the patient should match. If the results of the tests are inconsistent, we call it ABO discrepancy. It is essential to resolve ABO discrepancies in clinical laboratories for safe blood transfusion [2]. Weak ABO subgroups are the main cause of ABO discrepancy. Various genotyping methods can be used to confirm the ABO blood group. Recently, next-generation sequencing has been used for blood grouping. However, this technique is expensive [3]. Here, we describe a weak B phenotype harboring B101/O04-variant alleles observed in a Korean family, and suggest a practical algorithm to work-up ABO subgroups without using whole-genome sequencing. The Institutional Review Board of Chonnam National University Hwasun Hospital (CNUHH), Hwasun, Korea (CNUHH-2022-096) approved this study.

A healthy 24-year-old woman visited the Chonnam National University Hospital to confirm her blood type. We carried out serological ABO typing for her and her family. ABO cell type test was performed using the tube method with anti-A, anti-B, anti-A1, and anti-H agents (Shinyang Diagnostics, Siheung, Korea). Serum type test was performed using the tube method with A1 and B cells (Shinyang Diagnostics). The proband, her father, and her half-sister presented a weak B phenotype: trace positivity (+/–) for anti-B antibodies in cell typing and strong positivity (4+) for A1 cells in serum typing (Fig. 1A). The mother of the proband presented an A phenotype. To clarify the genetic cause of the weak B phenotype, we performed a stepwise molecular work-up. For sequencing, we extracted DNA from the EDTA sample tube with QIA symphony DSP DNA mini kit (Qiagen, Hilden, Germany). PCR products were purified using QIAquick PCR purification kit (Qiagen) according to the manufacturer’s instructions.

ABO genotyping of exons 6 and 7 was performed first. Then, the other regions were genotyped sequentially (Fig. 1B). We performed PCR amplification using the primers ABOe6F and ABOe7R, and sequenced the PCR products using the primers ABOe6R, AB-Oe7F, ABOe7SF1, and ABOe7SF2. PCR amplification protocol and the cycling conditions described by Kim et al. were followed [4]. In exons 6 and 7 of the ABO gene, we found a common B101 allele in the proband, her father, and half-sister. A rare single nucleotide variant (c.579T>C) known to be associated with the O04 allele was observed in the proband and her mother (Fig. 1C).

Then, we performed PCR amplification of ABO exon 2 to coding region of exon 7 using primers ABOe27longF and ABOe27longR, and sequenced ABO exon 2 to intron 6. PCR amplification protocol and the cycling conditions described by Huh et al. [5] were followed. Further sequencing from exon 2 to intron 6 showed an insertion variant in intron 4 (c.203+1622_1623insC), which was confirmed to have originated from the O allele using allele-specific sequencing using a specific primer for ABOc.261GspR (5´-CAA TGG GAG CCA GCC AAG GGG TCA-3´). A base ‘C’, two steps away from the 3´ end, combines with a base ‘G’ at the locus c.261, which would bind specifically to the B allele. As a result, no insertion variant was found in this sequence using the specific primer. Based on these results, we made a tentative determination of the O04-variant allele (Table 1).

Then, we performed PCR amplification, including exon 1, adjacent ABO promoter site, and CBF/NF-Y enhancer site using primers ABOenhe1longF and ABOenhe1longR and sequenced each region. We did not find any remarkable variants for these regions [5, 6]. We followed the PCR amplification protocol and the cycling conditions described by Huh et al. [5].

Next, we focused on intron 1. Long PCR using the primers ABO+4419S and ABO+11078AS was performed as described by Sano et al. [7]. PCR amplification was performed in a final reaction volume of 50 °L, containing 25 °L of KOD One™ PCR Master Mix (TOYO BO, Osaka, Japan), 1.0 °M of each primer, and 3 °L of genomic DNA. The PCR amplification was carried out in Veriti 96-well Thermal Cycler (Thermo Fisher Scientific, Waltham, MA, USA). The cycling conditions were: initial denaturation at 98°C for 180 seconds, 35 cycles of 98°C for 10 seconds, and 68°C for 60 seconds, followed by a final elongation at 68°C for 180 seconds. The PCR product was electrophoresed on 0.7% agarose gel. It revealed no deletion of the +5.8-kb site, including the erythroid cell-specific regulatory element in intron 1. Sequencing using the primers ABO+5.8kbseqF and ABO+5.8kbseqR showed no remarkable variants on this site. We also amplified the 9.9 kb region that has CBF/NF-Y enhancer site, the binding site of the ABO+4419S, and the GATA motif in the +5.8-kb site. The heterozygosity of the 41st nucleotide in ABO minisatellite in the PCR product was maintained, confirming that the binding site of the PCR primer and the GATA motif was not deleted.

Lastly, we focused on the +22.6-kb enhancer region within the 3´ UTR region. PCR amplification was performed using the primers ABO+22.6kbpcrF and ABO+22.6kbpcrR, and the PCR product was sequenced using the primers ABO+22.6kbseqF and ABO+22.6kbpcrR as described by Sano et al. [8]. PCR amplification was performed in a final reaction volume of 50 °L, containing 25 °L of 2x Phusion Flash High-Fidelity PCR Master Mix (Thermo Fisher Scientific), 1.0 °M of each primer, and 3 °L of genomic DNA. The PCR amplification was carried out in Veriti 96-well Thermal Cycler (Thermo Fisher Scientific). The cycling conditions were: initial denaturation at 98°C for 180 seconds, 35 cycles of 98°C for 15 seconds, and 72°C for 60 seconds, followed by a final elongation at 72°C for 180 seconds. The PCR product was electrophoresed on 1.0% agarose gel. We did not find any deletion or nucleotide variants, or large deletion in the amplified sequences. During sequencing, the nucleotide sequence of GenBank accession no. NG_006669.2 was used as a reference. Sequencher software (Gene Codes, Ann Arbor, MI, USA) was used for sequence analysis. We described primer sequences and their targets in the ABO gene used in this study (Table 2).

Whole-genome sequencing may be ideal for determining ABO genotypes [3], but it is not readily available in routine practice yet. Here, we used a stepwise procedure to identify genetic variations in the ABO gene based on previously reported causes of a weak phenotype [6-9]. We made a tentative determination of the B101/O04-variant, which harbored c.203+1622_1623insC in addition to c.579T>C in this case. The O04 allele is rarely observed, alth-ough a case of Ael02/O04 presenting as a typical O phenotype and a chimera with A102/B101/O04 have been reported in Korea [10]. A recent Chinese report suggested the potential effect of the O04 allele associated with the expression of weak A antigen in a case of B101/O04 with a weak AB phenotype [11]. However, the O04-variant allele was unlikely to be the main cause of the weak B phenotype in our case, since the father and half-sister had the common O allele, rather than the O04-variant allele, but still expressed a weak B phenotype. Regrettably, we could not clarify the cause of the weak B phenotype as it is difficult to find the exact cause of the ABO mismatch based on general PCR techniques. However, we found a new O04-variant allele that has both c.579T>C and c.203+1622_1623insC. In conclusion, we described a stepwise algorithm to work-up ABO subgroups, which can be used instead of whole-genome sequencing in a real-world laboratory.