Comprehensive Molecular Analysis of Serologically D-Negative and Weak/Partial D Phenotype in Thai Blood Donors

Antigen expression in the Rh system, which is the most complex and polymorphic blood group system, is driven by two homologous RH genes: RHD and RHCE [1]. The most clinically relevant antigen in this system, antigen D, is encoded by the former gene. Anti-D alloimmunization can occur in D-negative patients when exposed to D-positive red blood cells and may be responsible for hemolytic transfusion reaction and hemolytic disease of the fetus and newborn with severe clinical outcomes [2].

Variations in D phenotype expression in Caucasians, Black-Africans, and East-Asians, as well as the molecular genetics of RH genes, have been extensively studied and described, and the global distribution of variant RH alleles is well known in these populations [3-6]. In Western and East-Asian countries, molecular typing; i.e., genotyping, in addition to serological study, has become routine practice to assess the risk of alloimmunization [7, 8] and is therefore a critical step in the whole management of transfusion and pregnancy at risk.

A key point in molecular RHD typing is allele distribution, which is known to vary largely between ethnicities. The knowledge about the molecular background of the RHD gene on a regional/national basis is necessary to implement an efficient, dedicated genotyping strategy. For example, Caucasian-specific tests for rou tine diagnostics have been developed and implement ed to identify preferentially the RHD*weak D type 1, RHD*weak D type 2, and RHD*weak D type 3 alleles [9-12], which have long been known to be the most common variant alleles in serologic weak D individuals of Caucasian origin [3]. This latter strategy would be definitely inefficient for typing people of African or Asian origin. In Korea, Seo et al. [13] designed and implemented an effective diagnostic strategy for the detection of RHD variants in serologically D– Korean donors, including the typical Asian DEL allele characterized by the splice site variation c.1227G>A [14], which is the most prevalent DEL allele in the East-Asian population [6, 15, 16]. Extensive molecular data are available in Caucasians, Black Africans, East-Asians, and Indians since more recently. Conversely, populations of other origins, such as Thais, have been barely documented. It is worth mentioning that preliminary studies have been conducted to investigate the Rh blood group system in the Thai population. As early as in 2002, a large-scale study reported that 99.69, 0.30, and 0.01% of the Thai blood donors are D+, D–, and weak D, respectively [17]. Later, another investigation in samples with serological weak D expression showed that partial DVI, which is clinically relevant, is the most common phenotype (62/81, 76.5%) in this subset of samples [18]. In terms of molecular analysis, the Asian DEL allele was found in 18.9% of the serologically D– samples [19], and a recent study also reported preliminary RH genotyping data in a subset of nonselected donors from North and Central Thailand [20]. Overall, few studies have been carried out in Thailand, and the molecular determinants of Rh D variability, which is currently unknown in this population, remain to be described in detail at the genetic level. Here, we report for the first time a comprehensive study investigating the molecular basis of those variant D samples in the Thai population.

Blood samples from unrelated donors from Bangkok and suburban provinces were collected and provided by the National Blood Centre (NBC), Thai Red Cross Society (Bangkok, Thailand). Blood samples were analyzed routinely with automate analyzer PK7300 Microplate Agglutination Systems (Beckman Coulter) with monoclonal anti-D (clones LDM2 and RUM-1, NBC). Samples with inconclusive or negative results in D typing were retyped with two commercial blended IgM/IgG monoclonal anti-D reagents: IgM clone LDM2 and IgG clone LHM77 (NBC), IgM clone RUM-1 and IgG clone MS-26 (Plasmatec, Bridport, UK). Using conventional tube test at room temperature, 37°C, and indirect antiglobulin test (IAT), 450 donor samples were classified either as weak/partial D or D negative. Weak/partial D samples showed either agglutination reaction ≤2+ or IAT positive. Rh CcEe typing was carried out in all donor samples by using in-house anti-c and -E (NBC), and commercial anti-C and -e (DiaClon; DiaMed GmbH, Bio-Rad Laboratories) monoclonal IgM antibodies.

IAT-negative samples were subsequently screened for DEL by the adsorption-elution technique with monoclonal anti-D reagent (NBC). Briefly, 200 μL of donor red blood cells were washed with normal saline solution (0.9%) and mixed with equal volume of anti-D reagent. The mixture was incubated at 37°C for 1 h and then washed thoroughly with normal saline solution. The eluate was prepared by the in-house 0.1 M glycine-HCl buffer/ 10% EDTA (V/V: 4: 1) technique as previously described [21]. Eluate and last-washed supernatant were used for IAT (conventional tube test) against D-positive and D-negative control cells (NBC).

Genomic DNA was extracted from whole blood by using commercial kits (BioFact Genomic DNA Prep Kit; BioFact, Celbridge, Ireland; PureDireX Genomic DNA Isolation Kit; Bio-Helix, Keelung City, Taiwan; or PureLink Genomic DNA Mini Kit; Thermo Fisher Scientific, Bangkok, Thailand). DNA concentration and purity (A₂₆₀/A₂₈₀) were measured with a NanoDrop^TM 2000/2000c Spectrophotometer (Thermo Fisher Scientific).

Analysis by quantitative multiplex polymerase chain reaction (PCR) of short fluorescent fragments (QMPSF) was carried out to investigate potential exon copy number variations/rearrangements in the RHD gene by using the F9 and HFE genes for positive amplification control and copy number normalization in conditions previously described [22]. Basically, QMPSF is a simple multiplex, fluorescent PCR-based method, which aim is here to specifically identify RHD exon markers to quantitatively assess exon copy number variations, which are frequently encountered at the RHD locus, and to easily identify variant alleles, including hybrid D-CE genes. Respective fragment length and peak areas were analyzed by GeneMapper v4.0 software (Applied Biosystems, Villebon-sur-Yvette, France). Exon 3 duplication allele was confirmed and screened in weak D samples by a previously reported PCR-based method [23]. Finally, when necessary, all ten RHD exons were amplified and directly sequenced, and sequencing data were analyzed by Sequencher v5.0 software (Gene Codes Corporation, Ann Arbor, MI, USA) [11]. Genotypes were deduced from both QMPSF and sequencing data on the basis of the most likely combination of variant RHD alleles.

Minigene splicing assay was carried out to investigate the effect of an intronic variation on splicing as previously described [11, 23]. Briefly, both wild-type and mutated regions of RHD exon 8 and flanking intronic domains were PCR amplified. PCR products were inserted into the minigene to generate recombinant vectors that were subsequently transfected into HEK293 cells (ATCC Number CRL-1573). Cells were harvested 48 h after transfection. Total RNA was extracted and purified by the RNeasy Mini Kit (Qiagen, Courtaboeuf, France) and served as a template for reverse-transcription (RT)-PCR with the QIAGEN OneStep RT-PCR Kit (Qiagen). Finally, RT-PCR products were visualized on a 2% agarose gel containing ethidium bromide and directly sequenced after gel extraction.

In the 450 Thai blood donors, 129 and 321 samples were classified as weak/partial D and D negative by routine testing, respectively. Subsequent test by adsorption-elution in the 321 latter samples actually identified 121 DEL samples, while the other 200 samples were D negative. The RHD gene was then genotyped in all samples to identify the respective molecular bases of the variant D phenotype.

Samples typed as D negative by serological analysis (n = 200) were first investigated by RHD QMPSF, which has proven useful for genotyping such samples. As commonly observed, whole deletion of the RHD gene at the ho mozygous state is the most frequent genotype (Table 1; 150/200, 75.0%) notably found in all 100 ccee samples, followed by hybrid genes in 26 samples. RHD gene exons were subsequently sequenced in the remaining 24 samples. Interestingly, the Asian DEL allele; i.e., RHD(c.1227G>A), was found at the hemizygous state in 21 samples, while the other three samples carry single-nucleotide variations (Table 1).

All 121 serological DEL samples were first screened in RHD exon 9. The Asian DEL allele was identified in 108 samples (Table 2, 89.2%). QMPSF analysis and direct sequencing resolved twelve samples carrying negative RHD alleles, and are thus supposed to be D negative, while a single sample was found to carry the RHD*weak-D type 15 allele (Table 2).

The 129 weak/partial D samples were investigated by RHD QMPSF and/or direct sequencing. This analysis showed a large heterogeneity in terms of genotypes in the Thai population (Table 3). Indeed, as many as 52 allele combinations, including 33 observed only once, were found in those samples. The most common weak/partial D allele is partial DVI type 3 (or RHD*06.03; allele frequency = 0.120). Interestingly, the most frequent weak D allele (RHD*weak D type 150) recently reported in the Indian population, which involves a duplication of a ∼12 kilobase-region including RHD exon 3, was observed in 8 samples (6.2%). Eight novel alleles (Table 4) were found in 16 donors, suggesting that the Thai population is heterogeneous in terms of RHD molecu lar genetics. Of those, both RHD(R10Q,V174M) and RHD(L110P,F223V) involve two variations already reported separately. The other six novel alleles involve single missense variations.

We next paid attention to a variant in intron 8, c.1153+6T>C (or IVS8+6T>C), which was previously reported (GenBank Accession No.: JX114750). To get insight into the functional defect induced by the intronic variant, we took advantage of our previously reported minigene assay, which has proven useful for the characterization of variants affecting splicing [11, 23]. Our test clearly demonstrated that c.1153+6T>C significantly alters normal splicing by promoting RHD exon 8 exclusion from the mature transcript (Fig. 1A). A “normal” full-length transcript including exon 8 was though identified by sequencing analysis (Fig. 1B). Overall, this assay suggests that c.1153+6T>C is a splicing variant affecting D-antigen expression quantitatively, but not qualitatively, and therefore is not of clinical importance.

The nature and distribution of variant RHD alleles are fundamental data for the Public Healthcare system to implement efficient and cost-effective genotyping tests and to guide population-specific transfusion practice and obstetrical management [24]. In order to get insights into the molecular determinants of variant D expression, we sought to investigate the RHD gene in the Thai population, which has remained globally unexplored so far. For the first time, we report a comprehensive overview of the variant RHD alleles in this population.

As expected, whole RHD gene deletion at the homozygous state is the most common D-negative genotype in serological D-negative Thai blood donors (Table 1, 75.0%). When considering only the subset of C/E+ samples (n = 100), frequency of this genotype is in the range of what has been reported before in other Asian populations (36.4–63.1%) [15, 16, 25, 26]. Interestingly, as many as 21 samples out of 200 were found to carry the Asian DEL allele. This result suggests that either another variation in RHD, which has not been identified yet, impairs RhD protein expression, or reagents and/or conditions for identifying DEL samples are not optimal, thus yielding false-negative results. In this more likely latter hypothesis, DEL screening methodology should be reviewed urgently in a large-scale study including genotyped samples. Unexpectedly, one sample found to be D negative serologically was genotyped as RHD*05.04. This finding may be due to either the absence of some D epitopes recognized by the antibodies used in this study as previously reported [27], leading to a “false” apparent D-negative phenotype, or another unidentified deleterious variation in the RHD gene that impairs D antigen expression.

As also expected, RHD(c.1227G>A) is the most frequent DEL allele found in ∼90% of the samples (Table 2). Based on the genotyping data, in this group of serologic DEL samples, the percentage of false-positive results is significant (Table 2; 12/121, 9.9%), confirming that the methodology for DEL sample identification should be revised.

Finally, as many as 36 variant RHD alleles were found in 129 serological weak/partial D samples, highlighting the large variability of the RHD gene in the Thai population. Eight novel alleles were found, thus contributing to extend the molecular catalogue of RHD polymorphism. It also suggests that many other novel alleles are very likely to be found in future studies in the Thai population. The most frequent allele is RHD*06.03, which was also reported to be the most frequent in D variant donors from South China (28/62, 45.2%) [28]. Very interestingly, the second most common variant allele is RHD*01W.150, which was recently reported in the Indian population [29]. Taken together, besides the population-specific pattern observed in Thais, those data highlight the mixed nature of this population and, in parallel with the geographical context, provide data to bridge the molecular gap between South China and India. Replication studies in other regions of Thailand will definitely be of interest to investigate potential region-specific patterns of variant RHD allele distribution.

It is important to pay attention to the last six genotypes reported in Table 3, including a total of 28 samples. Indeed, an apparent wild-type RHD allele was found in 23/129 samples (17.8%), while they were found to be weak D serologically. It is interesting to compare these data with those reported recently in the Indian population, where 37/223 serologically weak D samples (16.6%) were found to carry at least one apparent wild-type RHD gene copy [29]. On the basis of these observations and by considering that replication studies in both populations yield comparable results, we suggest that more complete RHD gene investigation, including yet unexplored regions of regulation and introns, should be carried out to further identify potential functional variations altering D antigen expression. Discrepancy in the genotype-phenotype correlation in the remaining five samples (i.e., RHD*01EL.01/RHD*01N.01 [n = 1]; RHD*01EL.01/RHD*01N.03 [n = 1], and RHD*01N.01/RHD*01N.01 [n = 3]) remains to be elucidated by additional serological testing in the same samples first.

Overall, we report the molecular basis of variant D expression in Thai blood donors for the first time. Beyond the fundamental interest of this study in terms of population genetics and molecular epidemiology of the Rh blood group system, the next step will be the design and implementation of a population-specific screening test based on our data to help clinicians in their current practice to further improve blood transfusion safety.

The authors are grateful to the Thai blood donors who contributed their blood samples for this study.

The project was approved by Mahidol University Central Institutional Review Board (MU-CIRB 2018/104.1805) and Research Ethics Committee, NBC, Thai Red Cross Society (COA No. NBC 13/2018). All participants provided their informed consent.

The authors declare no conflicts of interest.

This study was supported by the Faculty of Medical Technology, Mahidol University, Bangkok, Thailand; the Etablissement Français du Sang (EFS) – Bretagne, and the Institut National de la Santé et de la Recherche Médicale (INSERM), France.

Pornlada Nuchnoi and Yann Fichou contributed equally to this work.