KEL*02 alleles with alterations in and around exon 8 in individuals with apparent KEL:1,−2 phenotypes

Background and Objectives Antibodies to antigens in the Kell blood group system, especially anti‐KEL1, are involved in both haemolytic disease of the newborn and foetus and haemolytic transfusion reactions. Correct typing results are important and discrepancies between serologic and genetic typing must be resolved. Here, we describe the investigation of three healthy individuals who were initially phenotyped as KEL:1,?2. Materials and Methods Antigen typing was performed by standard serological techniques and by flow cytometric analysis. The KEL*01/02 polymorphism was tested by an allele‐discrimination TaqMan assay as well as by PCR with allele‐specific primers and PCR–RFLP. DNA sequencing of the KEL coding region was also performed. Results Two KEL*02N alleles with mutated splice sites around exon 8 were identified: intron 7 ?1g>c (novel) and intron 8 +1g>t (previously reported in one case of K₀). In the third sample, a missense mutation in exon 8, 787G>A (novel) predicting Gly263Arg, was detected on a KEL*02 allele and associated with dramatically weakened KEL2 antigen expression. Conclusion Resolution of discrepant phenotype/genotype results identified silencing mutations in or around exon 8. A combination of molecular and serologic methods has the potential to improve the quality of test results and was required to ensure both the accurate KEL2 antigen status and KEL*01 zygosity of these individuals.     

Point mutations in KEL exon 8 determine a high-incidence (RAZ) and a low-incidence (KEL25, VLAN) antigen of the Kell blood group system

BACKGROUND AND OBJECTIVES: The molecular basis of two Kell blood group antigens, RAZ (provisionally KEL27) and VLAN (KEL25), were determined. MATERIALS AND METHODS: The DNA sequences of the open reading frames and the flanking intron regions of the 19 KEL exons from RAZ and VLAN probands were compared with that of common KEL. Genotyping assays were designed to confirm and detect RAZ and VLAN phenotypes. RESULTS: A homozygous G865A mutation, encoding lysine instead of glutamic acid at amino acid position 249 of Kell protein, defines the RAZ phenotype, while a heterozygous G863A mutation in KEL, encoding an arginine to glutamine substitution at amino acid 248, characterizes the VLAN phenotype. CONCLUSION: Point mutations G865A and G863A, in adjacent codons of KEL exon 8, which cause amino acid substitutions, characterize the RAZ and VLAN Kell blood group phenotypes.     

A combination of the effects of rare genotypes at the XK and KEL blood group loci results in absence of Kell system antigens from the red blood cells

The 22 antigens of the Kell blood group system are located on a red blood cell (RBC) membrane glycoprotein that shows sequence homology with a family of metalloendopeptidases. Expression of the Kell system antigens is partially governed by XK, an X-linked gene that encodes the Kx protein; absence of Kx results in reduced Kell antigen expression. Almost total absence of Kell antigens from the RBCs of a German man with no symptoms of neuroacanthocytosis could not be due to the Kell- null phenotype, Ko, because his RBCs had very weak expression of Kx antigen and his three children were Kp(a + b+). Kell antigens were normal on the RBCs of his son but weak on those of his two daughters. An Nla III restriction fragment-length polymorphism within the KEL gene showed the Kpa/Kpa genotype in the propositus. Sequencing of his XK gene showed a single base change within the donor splice consensus sequence of intron 2. A BsaAl restriction fragment-length polymorphism showed the mutation in both of his daughters but not in his son. The extreme depression of the Kell antigens of the propositus must be due to a combination of effects, ie, homozygosity for Kpa and deficiency of Kx protein, each of which is capable of causing some degree of weakening of Kell antigens.     

Phenotype prediction by DNA-based typing of clinically significant blood group systems in Jordanian blood donors

BACKGROUND AND OBJECTIVES: During the past 10 years several DNA-typing methods have been developed to complement routine serological typing for determination of polymorphisms in the ABO, RH, KEL, JK and FY blood group genes. However, the molecular basis of blood groups can differ widely between ethnic groups. The purpose of this study was to evaluate selected DNA-based methods for phenotype prediction in a population not previously investigated. MATERIALS AND METHODS: Blood samples from a random sample of Jordanian blood donors were collected and red cells isolated from these blood samples were phenotyped for common ABO (n = 150) and KEL/FY/JK (n = 90) antigens. RHD-negative and -positive donors were selected for RH typing (n = 120 and 30, respectively). DNA was prepared and blood group genotyping performed according to selected methods in current use. Discordant samples required further investigation by extended serology and DNA sequencing. RESULTS: The degree of concordance between phenotype and genotype was high, but some exceptions were noted. Two of 14 A2/A2B samples lacked all mutations associated with known A2 alleles of the ABO system. RH typing revealed four samples with the c(cyt48) marker, causing false-positive RHC typing. A single D-negative sample was positive for D-specific exon 10 markers. The RHD pseudogene was not found in the 150 donors tested. Nine samples revealed discrepancies that were associated with unknown silent or weakly expressing Fyb-like alleles. CONCLUSIONS: With the exception of the FY system, we conclude that the molecular background of the clinically important blood group antigens studied here is similar to that reported for Caucasoids.     

A New Low-Incidence Antigen in the Kell Blood Group System: VLAN (KEL25)

A multilaboratory investigation has identified a new low-incidence antigen ‘VLAN’ on the red cells of a blood donor. The VLAN antigen is destroyed by 2-aminoethylisothiouronium bromide treatment of the donor's red cells suggesting an association with the Kell system. Monoclonal antibody-specific immobilization of erythrocyte antigen analysis with anti-VLAN and with several mouse monoclonal antibodies directed at epitopes on the Kell glycoprotein gave positive results, indicating that the VLAN antigen is located on the Kell glycoprotein. The VLAN red blood cells have the common Kell phenotype: KEL:–1,2,–3,4,5,–6,7,–10,11,12,13,14,–17,18,19,–21,22,–23,–24. Additional serologic data indicate that the VLAN antigen is not part of any other ISBT blood group system, collection or series. A family study showed that the VLAN antigen is inherited since the red cells of two sisters and one niece of the propositus are also VLAN+. The ISBT Working Party on Terminology for Red Cell Surface Antigens has assigned VLAN to the Kell blood group system as KEL25 (number for computer listings 006025).     

Molecular basis of the Kell (K1) phenotype

K1 (K, Kell) is a strong immunogen; its antibodies can cause severe reactions if incompatible blood is transfused and may cause hemolytic disease of the newborn in sensitized mothers. K1 is a member of the Kell blood group system, which is complex, containing over 20 different antigens. Some of the antigens are organized in allelic pairs of high and low prevalence whereas others are independently expressed. K1, which is present in 9% of the population, is antithetical to the high- prevalence K2 (k) antigen. We have determined the molecular basis of the K1/K2 polymorphism by sequencing the 19 exons of the Kell gene (KEL) of a K1/K1 person. Polymerase chain reaction was performed on genomic DNA isolated from peripheral blood and the amplified products were either directly sequenced or subcloned and sequenced. Comparisons of K1/K1 and K2/K2 DNA showed a C to T base substitution in exon 6 that predicts a threonine to methionine change at amino acid residue 193. This amino acid substitution occurs at a consensus N-glycosylation site (Asn. X. Thr) and probably prevents N-glycosylation, leading to a change in phenotype. The C to T substitution creates a Bsm I restriction enzyme site, which was tested in 42 different samples to confirm that this base change identifies the K1/K1 genotype. This test differentiates genotypes, K1/K1, K2/K2, and the K1/K2 heterozygote and should prove useful in the prenatal diagnosis of K1-related hemolytic disease of the newborn.     

Sequence-specific primers for MNS blood group genotyping

Background

Various techniques of genotyping the MNSs blood group have been described, but none of them enables the complete detection of all MNS antigens.

Materials and Methods.

Blood samples were obtained from blood donors. Primers were created using the published DNA sequences for glycophorins A and B. Genotyping was performed using polymerase chain reaction sequence-specific primers (PCR-SSP).

Results

A total of seven primers were found to specifically amplify the most common MNS antigens. The use of these primers has enabled us to correctly genotype all blood samples tested so far (n=116).

Discussion.

Specifically created primers enable genotyping of the MNS antigens in a single PCR-SSP run. The method is reliable, easy to perform, and can be used in routine practice.     

HPA-1 typing by PCR amplification with sequence-specific primers (PCR-SSP): a rapid and simple technique

Summary. A DNA-based method was developed to genotype donors for the human platelet antigens HPA-la and -1b. Sequence-specific primers (SSP) were used in the polymerase chain reaction (PCR) which allowed the HPA-la/la. -lb/lb and -la/lb genotypes to be determined by PCR alone, no second analytical stage was required. 10 donors were tested by PCR-SSP and the results were concordant with serological phenotyping and independent DNA analysis.     

Identification of a defect in the intracellular trafficking of a Kell blood group variant.

Blood group polymorphisms have been used as tools to study the architecture of the red blood cell (RBC) membrane. Some blood group variants have reduced antigen expression at the cell surface. Understanding the underlying mechanism for this reduced expression can potentially provide structural information and help to elucidate protein trafficking pathways of membrane proteins. The Kp(a+) phenotype is a variant in the Kell blood group system that is associated with a single amino acid substitution (R281W) in the Kell glycoprotein and serologically associated with a weakened expression of other Kell system antigens by an unknown mechanism. We found by immunoblotting of RBCs that the weakening of Kell antigens in this variant is due to a reduced amount of total Kell glycoprotein at the cell surface rather than to the inaccessibility of the antigens to Kell antibodies. Using a heterologous expression system, we demonstrate that the Kpa mutation causes retention of most of the Kell glycoprotein in a pre-Golgi compartment due to differential processing, thereby suggesting aberrant transport of the Kell protein to the cell surface. Furthermore, we demonstrated that single nucleotide substitutions into the coding region of the common KEL allele, as predicted by the molecular genotyping studies, was sufficient to encode three clinically significant low incidence antigens. We found that two low incidence antigens can be expressed on a single Kell protein, thus showing that the historical failure to detect such a variant is not due to structural constraints in the Kell protein. These studies demonstrate the power of studying the molecular mechanisms of blood group variants for elucidating the intracellular transport pathways of membrane proteins and the requirements for cell surface expression.     

Heterozygosity for two novel null alleles of the KEL gene causes the Kell-null phenotype in a Japanese woman

The Kell-null (Ko) phenotype is rare and it does not express the Kell antigens on erythrocyte membranes. Recently, several distinct missense and nonsense base substitutions in the coding region and the donor splice site of intron 3 were identified in the KEL gene in individuals with the Ko phenotype. We analysed both genomic DNA and cDNA sequences of the KEL gene in a Japanese woman with the Ko phenotype. She was found to be heterozygous for two novel null KEL alleles. One allele contained an A to G substitution in intron 5 that changes the 3'-splice site of intron 5 from AAG to AGG, resulting in a reading frameshift by a single guanine insertion in KEL mRNA, and the other allele contained a single G to A substitution in exon 12 (codon 459) creating a termination codon. Neither mutation was found in 114 randomly selected Japanese individuals. The results suggested that the Ko blood group phenotype might be owing to several distinct non-functional alleles without any prevalent allele.     

A Novel Common Kell Antigen, TOU, and Its Spatial Relationship to Other Kell Antigens

A 47-year-old native American (TOU) was admitted to hospital for hip surgery. His serum agglutinated all red blood cells (RBCs) tested except Ko and DTT-treated RBCs and was weakly reactive with RBCs known to have a weak expression of Kell antigens, namely Kmod, McLeod, Kp(a+b-) (KEL:3,-4) and K:-13 (KEL:-13) phenotypes. RBCs from three siblings, a son and a daughter were incompatible with TOU's antibody. TOU's RBCs had the common Kell phenotype: K- k+ Kp(a-b+c-) Ku+ Js(a-b+) Ul(a-) K:11, -17 K:14,-24 K:12,13,18,19,22,-23 (KEL:-1,2,-3,4,5,-6,7,-10,11,12,13,14,-17,18,19,-21,22,-23,-24). Since TOU's RBCs were not agglutinated by an unidentified Kell-related antibody (IAN), tests were performed to show that TOU and IAN were mutually compatible. IAN is a Latino female hospitalised for a hysterectomy. The TOU antigen was shown to be located on the Kell glycoprotein by a monoclonal antibody immobilisation of erythrocyte antigen (MAIEA) assay. The unique pattern of reactivity obtained with TOU and IAN antibodies using this assay indicated the TOU epitope to be in an area remote from other Kell antigens, namely K, k, Kp^a, Kp^b, Kp^c, Ku, Js^a, Js^b, Ul^a, K11, K12, K13, K14, Wk^a, K18, K19, K22 and K24 (KEL1, KEL2, KEL3, KEL4, KEL5, KEL6, KEL7, KEL11, KEL12, KEL13, KEL14, KEL17, KEL18, KEL19, KEL21, KEL22 and KEL24) but close to the low-incidence antigen K23 (KEL23). Investigation of antibodies to previously identified antigens on the Kell glycoprotein by MAIEA using the mouse monoclonal antibodies BRIC18, BRIC68, BRIC107 and BRIC203 has identified six patterns of reactivity and has provided evidence for Kp^c being located in the same region as Kp^a and Kp^b.     

Distribution of Diego blood group alleles and identification of four novel mutations on exon 19 of SLC4A1 gene in the Chinese Han population by polymerase chain reaction sequence–based typing

Background The Diego blood group system plays an important role in transfusion medicine. Genotyping of DI1 and DI2 alleles is helpful for the investigation into haemolytic disease of the newborn (HDN) and for the development of rare blood group databases. Here, we set up a polymerase chain reaction sequence–based typing (PCR‐SBT) method for genotyping of Diego blood group alleles. Study Design and Methods Specific primers for exon 19 of the solute carrier family 4, anion exchanger, member1 (SLC4A1) gene were designed, and our PCR‐SBT method was established and optimized for Diego genotyping. A total of 1053 samples from the Chinese Han population and the family members of a rare proband with DI1/DI1 genotype were investigated by the PCR‐SBT method. An allele‐specific primer PCR (PCR‐ASP) was used to verify the reliability of the PCR‐SBT method. Results The frequencies of DI1 and DI2 alleles in the Chinese Han population were 0·0247 and 0·9753, respectively. Six new single nucleotide polymorphisms (SNPs) were found in the sequenced regions of the SLC4A1 gene, and four novel SNPs located in the exon 19, in which one SNP could cause an amino acid alteration of Ala858Ser on erythrocyte anion exchanger protein 1. The genotypes for Diego blood group were identical among 41 selected samples with PCR‐ASP and PCR‐SBT. Conclusion The PCR‐SBT method can be used in Diego genotyping as a substitute of serological technique when the antisera is lacking and was suitable for screening large numbers of donors in rare blood group databases.     

High-throughput multiplex PCR genotyping for 35 red blood cell antigens in blood donors

Background and Objectives One to two per cent of patients in need of red cell transfusion carry irregular antibodies to red blood cell (RBC) antigens and have to be supplied with specially selected blood units. To be able to respond to those requests, blood centres have to screen a significant number of donors for a variety of antigens serologically, which is a costly and through the shortage of reagents, also limited procedure. To make this procedure more efficient, the Austrian Red Cross has developed a genotyping assay as an alternative approach for high throughput RBC typing. Materials and Methods A multiplex polymerase chain reaction (PCR) assay was designed for typing 35 RBC antigens in six reaction mixes. The assay includes both common as well as high‐frequency‐alleles: MNS1, MNS2, MNS3 and MNS4; LU1, LU2, LU8 and LU14; KEL1, KEL2, KEL3, KEL4, KEL6, KEL7, KEL11, KEL17 and KEL21; FY1, FY2, FYB^WK and FY0 (FYB^ES); JK1 and JK2; DI1, DI2, DI3 and DI4; YT1 and YT2; DO1 and DO2; CO1 and CO2; IN1 and IN2. The assay was validated using 370 selected serologically typed samples. Subsequently 6000 individuals were screened to identify high frequency antigen (HFA)‐negative donors and to facilitate the search for compatible blood for alloimmunized patients. Results All controls showed complete concordance for the tested markers. The screening of 6000 donors revealed 57 new HFA‐negative donors and the blood group database was extended by approximately 210 000 results. Conclusion The study shows that in practice, this high‐throughput genotyping assay is feasible, fast and provides reliable results. Compared to serological testing, this molecular approach is also very cost‐efficient.     

Unmasking of Kx antigen by reduction of disulphide bonds on normal and McLeod red cells

We have investigated the effect of dithiothreitol (DTT) treatment of human red cells upon the Kx blood group antigen. At low concentrations of DTT (less than or equal to 2 microM) there is enhancement of the Kx antigen concomitant with the complete denaturation of the Jsa and Jsb antigens of the Kell blood system. This unmasking of the Kx antigenic site is near maximal using 2 microM DTT. At this concentration of DTT, only the Jsa and Jsb antigens are completely denatured; all other Kell system antigens tested (K, k, Kpb, Ku) are essentially unaffected. These results argue against the Kx antigen serving strictly as a carbohydrate precursor substance involved in a sequential biosynthetic pathway of Kell blood group antigens. Also, McLeod red cells, after treatment with DTT, were found to contain Kx antigen, although in much lower density than normal red cells, indicating that, although not a typical carbohydrate precursor substance, Kx may, nevertheless, be essential for the serological expression of Kell related antigens. It is hypothesized that the Kx structure and the Kell blood group antigen structure are two separate subunits associated in a quaternary conformation involving at least one interchain S-S bond. Our results should allow for a clearer understanding of the relationship between the serological expression of the Kx antigen and the serologically observed reactivity of the Kell blood group antigens of individuals having normal, Ko and McLeod phenotypes.     

Introduction of a real-time-based blood-group genotyping approach

Background and Objective  Genotyping may be applied for rare blood group polymorphisms in a high-throughput mode as well for the molecular determination of blood groups due to unclear serological results.
Material and Methods  We developed and validated a DNA typing method for the determination of KEL1/2, JK1/2, FY1/2, FY0, MNS1/2, MNS3/4, DO1/2, CO1/2 and LU1/2 alleles using a melting curve analysis downstream from a fully automated DNA extraction. All assays were validated in terms of specificity, sensitivity, assay variability and robustness. The usability was proven by a batch of 200 blood samples with partially known phenotype.
Results  Assays for all blood groups were within the range of specificity (100%), assay variability and robustness (coefficient of variance < 3%). Genotypes of 200 random platelet donors were fully consistent with existing phenotype data. The obtained genotype distribution is in complete concordance with existing data for the European population underlined by a complete absence of CO2 homozygous donors and the FY0 allele among the cohort.
Conclusion  We introduce an approach for blood group genotyping of particular samples or gene loci in glass capillary format and for medium-throughput analysis in 96/384-well format. The advantages of this real-time polymerase chain reaction method are its automation potential, the flexibility regarding hardware and the rapid cycling time.     

The Kell blood group system: Kell and XK membrane proteins

Two membrane proteins express the antigens that comprise the Kell blood group system. A single antigen, Kx, is carried on XK, a 440-amino acid protein that spans the membrane 10 times, and more than 20 antigens reside on Kell, a 93-kd, type II glycoprotein. XK and Kell are linked, close to the membrane surface, by a single disulfide bond between Kell cysteine 72 and XK cysteine 347. Although primarily expressed in erythroid tissues, Kell and XK are also present in many other tissues. The polymorphic forms of Kell are due to single base mutations that encode different amino acids. Some Kell antigens are highly immunogenic and may cause strong reactions if mismatched blood is transfused and severe fetal anemia in sensitized mothers. Antibodies to KEL1 may suppress erythropoiesis at the progenitor level, leading to fetal anemia. The cellular functions of Kell/XK are complex. Absence of XK, the McLeod phenotype, is associated with acanthocytic red blood cells (RBCs), and with late-onset forms of muscular dystrophy and nerve abnormalities. Kell, by homology, is a member of the neprilysin (M13) family of membrane zinc endopeptidases and it preferentially activates endothelin-3 by specific cleavage of the Trp21-Ile22 bond of big endothelin-3.     

A1,2BO1,2 genotyping by multiplexed allele-specific PCR

The ABO blood group is the most clinically important human alloantigen system in transfusion medicine. The system involves three antigens A, B and H. H antigen is converted to either A or B by the activity of α1 → 3- n -acetyl-galactosaminyl transferase (A transferase) or α1 → 3 galactosyl transferase (B transferase). The O phenotype is the result of an inactive glycosyltransferase, which is unable to glycosylate the H antigen.
The immunological properties of the ABO system were identified at the turn of the century; however, the genetic basis of the ABO system has only recently been characterized. This has enabled the development of a number of molecular ABO typing methods. Described here is a two-reaction multiplex allele-specific PCR (ASPCR) genotyping assay for the A¹, A², B, O¹ and O² subtypes. 11 different allele-specific oligonucleotide primers were selected to detect the presence or absence of the O¹ associated G →  (−) deletion at base 261, the O² associated G → A substitution at base 802, the B associated G → A substitution at base 803, and finally the A² associated C → (−) deletion at base 1059.
A total of 122 peripheral blood samples were genotyped and serologically forward and reverse typed. A concordance rate of 98.4% (120/122 samples) was observed between the actual genotype and the serologically-based predicted genotype. These results indicate that this assay provides a rapid, accurate, and simple method for A^1,2BO^1,2 genotyping that serves as a useful supplement to standard serological ABO typing.     

Transfusion support for a patient with McLeod phenotype without chronic granulomatous disease and with antibodies to Kx and Km

Background and Objectives   Kell antigens are encoded by the KEL gene on the long arm of chromosome 7. Kx antigen is encoded by the XK gene on the short arm of the X chromosome. Kell and Kx proteins in the red cell membrane are covalently linked by a disulphide bond. The McLeod phenotype is characterized by weakened expression of antigens in the Kell blood group system, absence of Km and Kx antigens, and acanthocytosis. It has an X-linked mode of inheritance with transmission through carrier females. Some males with the McLeod syndrome also have chronic granulomatous disease (CGD). It is generally believed that patients with non-CGD McLeod may develop anti-Km but not anti-Kx, but that those with CGD McLeod can develop both anti-Km and anti-Kx.
Materials and Methods   We present serological data, DNA genotyping and gene sequencing, monocyte monolayer assay and neutrophil oxidative burst test from a patient with the McLeod phenotype without clinical evidence of CGD.
Results   We report here the second example of a patient with non-CGD McLeod who developed anti-Kx in addition to anti-Km. Sequencing of our patient's XK gene confirmed the presence of a mutation resulting in a premature stop codon and lack of Kx protein in the red cell membrane, which is consistent with the diagnosis of McLeod syndrome. Neutrophil oxidative burst test was normal, indicating that our patient did not have CGD. The challenge of providing 10 compatible blood units for multiple surgeries was met.
Conclusion   The second case of a rare entity, a patient with non-CGD McLeod who developed anti-Kx and anti-Km, was managed successfully with a combination of autologous donations and procurement of compatible units from national and international sources.     

HLA typing strategies in a cord blood bank

BACKGROUND AND OBJECTIVES: Although widely used, serological typing of HLA loci does not always produce uequivocal results. This may be particularly the case for cord blood since samples may be of small volume and poor quality, and contaminated. DESIGN AND METHODS: We typed 220 cord blood units (CBU) for HLA class I antigens using the serological technique. For those samples giving doubtful results we repeated the HLA typing by polymerase chain reaction with sequence specific primers (PCR-SSP). RESULTS: Results were satisfactory for 181 samples (82.3%). For the remaining 39 (17.7%) we had a doubtful antigen assignment for A locus in 9/39 cases (23.1%) and for B locus in 22/39 cases (56.4%). Eight of the 39 samples (20.5%) could not be analyzed by serology due to the high mortality of the cell suspension. Using PCR-SSP we obtained clear definition of class I antigens in all cases. All CBU were typed for HLA class II alleles by PCR-SSP with clear results in 100% of cases. INTERPRETATION AND CONCLUSIONS: In our experience, PCR-SSP can resolve the limitations of serology but, at the moment, it cannot substitute the latter in routine practice. The best strategy, in cord blood typing, is to perform both serological and molecular typing in order to obtain an accurate and clear result.     

Identification of RHCE and KEL alleles in large cohorts of Afro-Caribbean and Comorian donors by multiplex SNaPshot and fragment assays: a transfusion support for sickle cell disease patients

To lower the alloimmunization risk following transfusion in blacks, we developed two genotyping assays for large-scale screening of Comorian and Afro-Caribbean donors. One was a multiplex SNaPshot assay designed to identify ce(s) (340), ceMO/AR/EK/BI/SM, ce(s) , ce(s) (1006) and KEL*6/*7 alleles. The other was a multiplex fragment assay designed to detect RHD, RHDψ and RHCE*C and 455A>C transversion consistent with (C)ce(s) Type 1 and DIII Type5 ce(s) . Variant RHCE*ce alleles or RH haplotypes were detected in 58·69% of Comorians and 41·23% of Afro-Caribbeans. The ce(s) allele, (C)ce(s) Type 1, and DIII Type 5 ce(s) haplotypes were identified respectively in 39·13%, 14·67% and 4·88% of Comorians and 32·23%, 5·28% and 1·76% of Afro-Caribbeans. Genotypes consistent with partial D, C, c and/or e antigen expression were observed in 26·08% of Comorians and 14·69% of Afro-Caribbeans. No homozygous genotype corresponding to the RH:-18, -34, and -46 phenotypes were found. However, over 50% of genotypes produced low-prevalence antigens at risk for negative recipients, i.e., V, VS, JAL, and/or KEL6. One new variant RHCE*ce(s) (712) allele was identified. This is the first determination of variant RHCE and KEL allele frequencies. Results indicate the most suitable targets for molecular assay screening to optimize use of compatible blood units and lower immunization risk.