The low prevalence Rh antigen STEM (RH49) is encoded by two different RHCE*ce818T alleles that are often in cis to RHD*DOL

Abstract

Background

STEM (RH49) is a low prevalence antigen in the Rh blood group system. A scarcity of anti-STEM has precluded extensive study of this antigen. We report that two alleles with a RHCE*ce818C>T change encode a partial e, and a hr^S−, hr^B+, STEM+ phenotype and that both alleles are frequently in cis to RHD*DOL1 or RHD*DOL2.

Materials and methods

Blood samples were from donors and patients in our collections. Hemagglutination, DNA and RNA testing was performed by standard techniques.

Results

Fourteen STEM+ samples were heterozygous RHCE*ce818C/T: six had RHCE*ceBI and eight had a novel allele, RHCE*ceSM. Eleven were heterozygous for RHD*DOL1 or RHD*DOL2. Eleven samples, previously typed STEM−, had RHCE*ce818C/C (consensus nucleotide). RBCs from informative STEM+ samples were e+/− hr^S− hr^B+. One person who was heterozygous RHCE*ceBI and RHCE*cE had an anti-e-like antibody in her plasma, and one person, who was hemizygous for RHD*DOL2 had anti-D in her plasma.

Conclusions

We show that two alleles with a RHCE*ce818C>T change (RHCE*ceBI and RHCE*ceSM) encode a hr^S− hr^B+ STEM+ phenotype. In addition, both alleles are frequently in cis to RHD*DOL1 or RHD*DOL2 and RHCE*ceBI encodes a partial e antigen. In the small cohort of samples tested, RHD*DOL invariably traveled with RHCE*ce818T. Our study also confirmed the presumption that RHD*DOL2, like RHD*DOL1, encodes a partial D antigen and the low prevalence antigen DAK.

Introduction

STEM (RH49) is a serologically defined low prevalence antigen that was shown, by family studies, to be in the Rh blood group system.¹ A scarcity of monospecific anti-STEM has precluded extensive study of this antigen. An article by Marais and coworkers published in 1993¹ is the sole original report in the literature regarding STEM and anti-STEM. These authors reported that STEM has a variable expression, which is an inherited characteristic, and is associated with an altered e antigen. They also found that RBCs from approximately 65% of hr^S− and 30% of hr^B− South African donors typed STEM+ and that anti-STEM induced mild hemolytic disease of the fetus and newborn (HDFN).¹

RHD and RHCE are well studied and numerous variants, and the phenotypes they encode, have been described.^2-4 The high degree of diversity is a consequence of the homology between RHD and RHCE, their opposite orientation and close proximity on chromosome 1p, which favor the exchange of nucleotides. Some of the variant alleles encode so-called partial antigens, which are identified when an antigen-positive person makes the corresponding alloantibody. Several variant RHD and RHCE alleles have been shown, especially in African Americans, to be inherited en bloc. Examples, represented by the transfusion medicine-friendly allele names that have been proposed by the ISBT Working Party on Red Cell Immunogenetics and Blood Group Terminology (see ISBT-web.org), are RHD*DAR-RHCE*ceAR, RHD*DIIIa-RHCE*ceS, and RHD*DAU-RHCE*ceMO.⁵ Variant phenotypes can express more than one low prevalence antigen, e.g., RHD*DIIIa-RHCE*ceS encodes V, VS, and DAK. Similarly, a low prevalence antigen can be encoded by more than one allele, e.g., DAK is expressed on RBCs with the DIIIa, R^N, or DOL phenotypes that are encoded by RHD*DIIIa, RHCE*CeRN, RHD*DOL1, and RHD*DOL2, respectively.⁶

We report here an expansion of our initial findings^7-9 that two alleles with a RHCE*ce818C>T change encode the e+/− hr^S− STEM+ phenotype. These alleles are (i) the previously reported RHCE*ceBI¹⁰ (IBST provisional name RHCE*01.08, or RHCE*ce.08), which was previously not known to encode a novel low prevalence antigen, and (ii) the novel allele, RHCE*ceSM, (ISBT provisional name RHCE*01.09, or RHCE*ce.09). Both alleles are frequently in cis to RHD*DOL1 (ISBT provisional name RHD*12.01) or RHD*DOL2 (ISBT provisional name RHD*12.02). Furthermore, we confirm that RHCE*ceBI encodes a partial e antigen, and that, like RHD*DOL1, RHD*DOL2 encodes a partial D antigen and the DAK antigen.

Material and Methods

Samples and hemagglutination testing

Blood samples were freshly collected samples or were recovered from storage in liquid nitrogen. They were either referred for investigation, found in DNA screening tests, or received from numerous colleagues over many years. Reagent RBCs and antibodies were from our collections. Hemagglutination was performed by methods that were optimal for the antibody being used. STEM typings were determined in our laboratories.

Analyses of DNA and RNA

RNA was isolated from the reticulocytes by standard methods using TriZol and PureLink RNA Mini Kit (Invitrogen, Carlsbad, CA). Reverse transcription was carried out with gene-specific RHD and RHCE primers and Superscript III according to manufacturer’s instructions (Superscript III first-strand synthesis SuperMix, Invitrogen), and polymerase chain reaction (PCR) amplification was carried out for 35 cycles with primers to amplify exons 1 to 4 and exons 5 to 10 in RHD and RHCE as described previously.¹¹ RT-PCR products were checked for appropriate size and purity on agarose gels, purified using PCR product clean-up reagent according to manufacturer’s instructions (ExoSAP-IT, USB Corporation, Cleveland, OH), and directly sequenced by GENEWIZ, Inc. (South Plainfield, NJ). Sequences were aligned, and protein sequence comparisons were performed with Sequencher v4.9 (GeneCodes, Ann Arbor, MI).

For those blood samples where RNA extraction failed, and for confirmatory testing, genomic DNA was isolated using the QIAamp DNA Blood Mini Kit (QIAGEN, Inc. Valenica, CA). Samples were tested for RHCE*ceBI and RHCE*ceSM by analyzing RHCE*ce818C/T and RHCE*ce1132C/G on PCR amplicons using RHCE exon-specific primers as previously described,¹² followed by sequencing by GENEWIZ, Inc. To screen for RHD*DOL1, and RHD*DOL2, genomic DNA was amplified using specific primers flanking RHD exon 4 and exon 8 as previously described,¹² and products were submitted for sequencing to GENEWIZ, Inc to determine the presence of RHD*DOL1 (nt 509T>C in exon 4) and RHD*DOL2 (nt 509T>C in exon 4 and nt 1132C>G in exon 8). The RHD*410T>C nucleotide change in exon 3 that would discriminate RHD*DOL3, was not analyzed.

RHD zygosity analyses were performed to detect the presence or absence of the hybrid box as previously reported.¹³

Results

Two alleles encode STEM and are linked to RHD*DOL

RHD and RHCE cDNA, prepared from samples with a historic STEM+ or STEM− type, were sequenced. We found that two RHCE alleles (RHCE*ceBI or RHCE*ceSM) and two RHD alleles (RHD*DOL1 or RHD*DOL2) were present in the STEM+ samples. We then tested samples from people of African origin for the presence of these alleles using a combination of PCR-based assays (see Material and methods section).

Taken together, a total of fourteen STEM+ samples were tested: all were heterozygous for RHCE*ce818C/T: six (including the original index case, Stemper) were RHCE*ceBI [ce48C (16Cys), 712G (238Val), 818T (273Val), 1132G (378Val)]¹⁰ and eight had a new allele, which we name RHCE*ceSM [ce 48C (16Cys), 712G (238Val), 818T (273Val); so named from the first and last letters of ‘STEM’]. Although RHCE*ceBI has been reported¹⁰, it previously had not been shown to encode the STEM antigen. Of the six samples with RHCE*ceBI, three were heterozygous for RHD*DOL1 [509C (170Thr), 667G (223Val)] and three were heterozygous for RHD*DOL2 [509C (170Thr), 667G (223Val), 1132G (378Val)]. Of the eight samples with RHCE*ceSM, five were heterozygous for RHD*DOL1, and three did not have either a RHD*DOL1 or a RHD*DOL2 allele (see Table 1).

Table 1.

Compiled hemagglutination results and allele assignments on STEM+ and STEM− samples

Name	D	C	E	c	e	hrS	hrB	V/VS†	STEM		DAK	RHD*	RHCE*
									SA	NY
RHCE*ceBI and STEM+
Stemper	+	0	+	+	+/0	0	+	0	+	+	+S	DOL2 D	ceBI cE
MA72-09	+	0	+	+	+/0	0	+	NT	NT	+	+S	DOL1 DFV	ceBI cE
RFA-221	+	0	0	+	+	+W	+	0	NT	+	+S	DOL2 DAU-0	ceBI ce 48C
MA170-09	+/0	0	0	+	NT	NT	NT	NT	NT	+	NT	DOL2 Hemizygous presumed	ceBI ce 48C
J Allen	+	0	0	+	+	NT	NT	0	NT	+	+	DOL1 D	ceBI ceJAL
MA372-09 RFLP only	+	NT	0	+	+	NT	NT	+	NT	+	+	DOL1 Hemizygous presumed	ceBI ceS
RH*CEceBI and STEM NT
MG303-00 RFLP only	+/0	0	0	+	+	NT	NT	0	NT	NT	+	DOL1 Hemizygous	ceBI ce
RFA-075	+	NT	NT	NT	NT	NT	NT	NT	NT	NT	NT	DOL1 DIIIa-CE(4-6)-D	ceBI ceS
PC73 RFLP only	+	NT	NT	NT	NT	NT	NT	NT	NT	NT	NT	DOL1 D	ce818C/T
LC donor RFLP only	+	NT	NT	NT	NT	NT	NT	NT	NT	NT	NT	DOL1 D	ce818C/T
RHCE*ceSM and STEM+
Delia RFLP only	+	0	0	+	+	0	NT	0	+	+	+	DOL1 DAU-0	ceSM ceJAL
Keys	+	0	0	+	+	0	+	0	+	NT	+	DOL1 D	ceSM ceJAL
Hector	+	0	0	+	+	NT	NT	NT	+	NT	+	DOL1 DAU-0	ceSM ceMO
John	+W	0	0	+	+	NT	NT	+	+	NT	+	DOL1 Weak D type 4.2.2	ceSM ceAR
Jack	+w	0	0	+	+	0	+	+	+	+	+	DOL1 Weak D type 4.2.2	ceSM ceAR
KHV RFLP only	+	0	0	+	+	0	+	NT	+	NT	+	DIIIa D	ceSM ceEK
Dennis	+	0	0	+	+	0	+	+/0	+W	+	+W	Weak D type 4.2.2 D186T (novel)	ceSM ceAR
Sybil	+	0	0	+	+	0	+	+/0	+W	+	+W	Weak D type 4.2.2 D186T (novel)	ceSM ceAR
STEM+ but not RHCE*ceBI or RHCE*ceSM
Roman	+	0	0	+	+	+/0	0	+V−	^ +W	+W	+S	DIIIa DIIIa	ceS ce 48C, 733G
Alson	+	0	0	+	+	NT	NT	V−	+W	NT	+	DIIIa, 150C (novel) DIIIa, 150C (novel)	ceS ceS
STEM− and not RHCE*ceBI or RHCE*ceSM
MA127-08	+	+W	0	0	+W	+	0	0	NT	0	+	D D	CeRN CeRN
MA315-09	+	0	0	+	+	NT	0	NT	0	NT	+	DIIIa DIIIa	ceS ceS
MA156-08	+	0	0	+	+	+W	0	+	0	NT	+	DIIIa DIIIa	ceS ceS
E. Dlamini	0	+	0	+	+	+	0	+	0^^	0	0	DIIIa–CE[4-7]–D DIIIa–CE[4-7]–D	ceS ceS
M. Shaba	0	+	0	+	+	+	0	+	0	NT	0	DIIIa-CE[4-7]-D DIIIa-CE[4-7]-D	ceS ceS
MA314-09	0	+	0	+	+	+	0	+	0	0	0	DIIIa-CE[4-7]-D DIIIa-CE[4-7]-D	ceS ceS
Marcus	+w	0	0	+	+	+	0	+	0	0	+W	DIIIb (DIII type 7) DIIIb (DIII type 7)	ceS ceS
Norma	+W	0	0	+	+	0	+	+	0^^	NT	NT	Weak D type 4.2.2 Weak D type 4.2.2	ceAR ceAR
MA318-09 RFLP only	+	+	+	+	+	0	+	+	0	0	NT	Not analyzed	Not ceBI or ceSM
MA149-09	+	0	0	+	+	NT	+	+	NT	0	+W	Weak partial D type 4.0 Homo or hemizygous	ce 48C, 733G Not ceBI or ceSM Hybrid under investigation
CO 06-11	0	+	0	+	+	NT	NT	+	0^^	NT	0	Not DOL1 or DOL2 Hybrid under investigation	ceS ce

^†= positive with a reagent containing anti-V and anti-VS

^{^}= not available for repeat testing

^{^^}originally thought to be STEM+

RFLP = only PCR RFLP assays were performed. Full sequencing was done on the other samples.

NT = not tested

Four additional samples, that could not be tested with anti-STEM (because the RBCs did not recover from cryopreservation and fresh samples were not available), were heterozygous for RHCE*ceBI and heterozygous for RHD*DOL1 (Table 1). Thus, of the 18 samples with a RHCE*ce818C>T change, 12 had RHD*DOL1, three (including the original index case) had RHD*DOL2, and three did not have a RHD*DOL allele (Tables 1 and 2).

Table 2.

Summary of STEM testing interpretations (no homozygous RHCE*ce818T or RHD*509C, 667G [RHD*DOL] were found) in a total of 31 samples

RHCE*ceBI (n = 10)

6 samples STEM+

6 had RHD*DOL

4 samples STEM not tested

4 had RHD*DOL

RHCE*ceSM (n =8)

8 samples STEM+

5 had RHD*DOL

3 had other RHD alleles

RHCE*ce818C and RHD*509T, 667T (n = 13)

2 samples STEM+W

2 had RHD*DIIIa

11 samples STEM−

See Table 1 for alleles

Two samples from two South African donors (code names, Roman and Alson) were historically typed as STEM+^w. However, these samples did not have RHCE*ce818T or RHD*509C, 667G. The reason for this apparent discrepancy could not be determined because our concerted efforts to obtain fresh samples for extensive testing have been unsuccessful and other samples with this unique genotype were not available. Both RBC samples were non reactive by the direct antiglobulin test. As the STEM typing of both samples was considerably weaker than that of samples with RHCE*ce818T, it is highly probable that the agglutination was caused by an additional, but unidentified, antibody in the test plasma and was not due to the anti-STEM. A paucity of plasma containing anti-STEM precluded us from being able to perform absorption and elution studies.

As few samples tested for STEM, either positive or negative, are available, we have included our testing of eleven STEM− samples, which included hr^S− and hr^B− samples; all had the consensus nucleotide, i.e., RHCE*ce818C/C. The RHD and RHCE alleles associated with these STEM− samples are given in Table 1. Attempts to perform RHD zygosity testing confirmed that discrepant results are frequently obtained in the presence of an RHD*DOL allele and, thus the results are not given.^14-16

STEM+ and variant e antigen

Hemagglutination results given in Table 1 are a compilation of those obtained from historic records and those performed during this study. Accessibility of RBCs, the condition of the RBC sample, or lack of anti-STEM prevented us from being able to test all samples.

RBCs from eight informative STEM+ samples [with a hr^S− haplotype (RHCE*cE, RHCE*ceJAL, RHCE*ceAR, RHCE*ceEK, or RHCE*ceMO) in trans to the RHCE*ceBI or RHCE*ceSM)] were tested by hemagglutination and found to be hr^S−. In contrast, RBCs from the only two informative samples with a hr^B− haplotype (RHCE*cE) in trans to RHCE*ceBI were hr^B+. RBCs from the original index case (Stemper) and from another RHCE*ceBI/RHCE*cE sample (MA72-09) carried a single dose of the variant e antigen, they were agglutinated by four commercial anti-e and five single clone monoclonal anti-e (MS16, MS21, MS69, HIRO41, HIRO43), but were not agglutinated by three other monoclonal anti-e (MS19, MS62, MS63), indicating that RHCE*ce818T encodes an altered e antigen. One person (MA72-09) with RHCE*ceBI/RHCE*cE had anti-e-like antibodies in his plasma, providing the first evidence that RHCE*ceBI encodes a partial e antigen.

RHD*DOL encodes a partial D

One person (MA170-09) with RHD*DOL2 (presumed to be hemizygous) had anti-D in her plasma. Given the similarity in D epitope expression between DOL1 and DOL2, the presumption can be made that DOL2 is also a partial D but the presence of anti-D in this patient provides the first actual evidence that this allele encodes a partial D antigen. One person (MG303-00) with RHD*DOL1 had anti-D in her plasma, confirming that this allele encodes a partial D antigen.

DAK typing

All of 14 STEM+ samples tested with anti-DAK (Akom, Riz, Jep), were positive. Twelve gave strong reactions: nine had RHD*DOL1, two had RHD*DOL2, one had RHD/RHD*DIIIa. Two gave weak reactions: both had the RHD*weak D type 4.2.2/RHD*D186T haplotype¹⁷ (Table 1). RBCs with DOL or DIIIa phenotypes have been previously shown to express DAK.⁶ The DAK+ result on RBCs from two samples with RHD*DOL2 (Stemper and RFA-NYBC221) showed that this allele, like RHD*DOL1, encodes DAK. The unexpected reactivity of anti-DAK, albeit weak, with RBCs from the two samples with RHD*weak D type 4.2.2/RHD*186T is likely due to the 186G>T change that is predicted to encode Phe instead of Leu at amino acid residue 62, because this change is present on RBCs with the DIIIa phenotype. However, more samples need to be tested to confirm this hypothesis. The RHD*D186T allele is novel and we are reporting it elsewhere.¹⁷

Discussion

We have identified two RHCE alleles, both with a change of nucleotide 818C>T in RHCE*ce, that encode the STEM antigen. One allele (RHCE*ceBI) has been previously reported but was not known to be associated with STEM,¹⁰ and the other is a novel allele (RHCE*ceSM ). Of the 18 samples with RHCE*ce818C/T, 12 had RHD*DOL1 and three had RHD*DOL2 (hereafter collectively called RHD*DOL). Thus, in this small cohort of 18 samples, RHD*DOL travelled with one of the two RHCE*ce818T variants in 15 cases. Three samples with RHCE*ce818C/T but not RHD*DOL were STEM+ and provided evidence that STEM is encoded by the RHCE*ce818T variant alleles and not by RHD*DOL. Two of these samples had both RHD*weak D type 4.2.2 and the novel allele, RHD*D186T.¹⁷ The third sample had RHD/RHD*DIIIa.

In a separate, previously reported, study, we tested 705 samples that were not serologically typed for STEM and found four had RHCE*818C/T.⁹ Of these 705 samples, 70 samples were from patients with sickle cell disease and 220 samples were from blood donors who self-identified as being of African descent. Two of the 220 donors were heterozygous for RHCE*818C/T and 1132G (RHCE*ceBI); both samples had RHD/RHD*DOL. In these 290 samples, no other RHD*DOL were found. This shows the alleles encoding STEM are not as rare as previously thought, an impression due in part, no doubt, to the dearth of anti-STEM. Predicting the STEM type of samples by DNA analyses provides a means to have appropriate RBCs available to aid in the identification of anti-STEM, to provide matched blood products to immunized patients, and to overcome the problem of the scarcity of anti-STEM. Furthermore, testing plasma from patients who have been transfused with blood predicted to be STEM+ could reveal a source of anti-STEM and/or anti-DAK.

Our serologic tests indicate that RHCE*818T does not produce hr^s and the plasma of one STEM+ person contained an anti-e-like antibody. The exact specificity of this antibody remains to be determined. Other hr^S− RBCs exist, namely those with ceAR, ceMO, and ceEK phenotypes, but there is no evidence that these RBCs are STEM+. The anti-hr^S made by hr^S− STEM− people may not be mutually compatible and are likely to be incompatible with hr^S− STEM+ RBCs, and vice versa. Indeed, Peyrard, et al. briefly reported that the antibody made by a RHCE*ceMO/RHCE*ceMO person reacted strongly (4+) with RBCs from a RHCE*ceBI/RHCE*ceBI person.¹⁸ Neither ceBI nor ceSM express hr^S but, based on two informative samples, do express hr^B. Thus, it is clear why the original study¹ found that approximately 65% of hr^S− samples were STEM+ but it is unclear why it found that approximately 30% of hr^B− samples were STEM+.

Depending on alleles carried on the in trans chromosome, patients with the RHD*DOLRHCE*ce818T haplotypes can make anti-D, anti-e-like/-hr^S, or anti-Rh18¹⁰ and their RBCs carry the low prevalence antigens STEM and DAK. The original JAL+ proband (J. Allen) has RHD*D/RHD*DOL1 and RHCE*ceJAL/RHCE*ceBI and his RBCs express JAL, STEM, DAK, and extremely weak V and VS.¹⁹ In our relatively small study, two other samples (Delia and Keys) harbored the RHCE*ceJAL allele associated with expression of JAL. Our results highlight that many RBCs express multiple low prevalence antigens and this fact should be taken into account when identifying antibodies. From our results, most RBCs expressing STEM also express DAK and a high proportion express V and VS. It is possible that some antibodies have been misidentified and retesting them with a panel of RBC samples from people whose allele makeup has been determined will help to fine-tune the specificity(ies) within a given plasma. Having appropriate RBCs to test would resolve the conundrum of exactly which proteins express DAK.

In summary, our study revealed several new findings: RHCE*ceBI and RHCE*ceSM encode STEM; on limited testing, RHCE*ceBI encodes a slightly stronger expression of STEM than does RHCE*ceSM. This is consistent with the previously reported variable expression¹ and provides evidence that STEM is encoded by these variant RHCE alleles and not the RHD*DOL. RHCE*ceBI and RHCE*ceSM, which encode a partial e and lack of hr^S are often, although not invariably, in cis to RHD*DOL1 or RHD*DOL2. In this study and our previously reported study⁹ in all samples with RHD*DOL, RHCE*ceBI or RHCE*ceSM were found in cis.

The variant RhD encoded by RHD*DOL2, is a partial D antigen. As anti-STEM is in short supply, typing for this antigen by hemagglutination is virtually impossible; however, DNA-based assays provide a useful tool to predict the presence of STEM. Knowledge of the molecular basis of various Rh antigens should help to not only establish which antigenic determinants are encoded by which alleles but also to accurately establish the specificity of an antibody, or antibodies, in a plasma. In table 1, when possible, the donor code name is used so that others who have these samples in their frozen library can use the knowledge of which alleles are present to aid selection of appropriate RBCs for inclusion on a panel used for antibody identification purposes.