Three novel Er blood group system alleles and insights from protein modeling.

BACKGROUND: The Er blood group system was recently shown to be defined by PIEZO1. The system consists of high prevalence antigens Era, Er3, ERSA, and ERAMA; and low prevalence antigen Erb. Era/Erb are antithetical with Er(a-b+) defined by the ER*B allele [c.7180G>A p.(Gly2394Ser)]. A nonsense variant c.5289C>G p.(Tyr1763*) is associated with a predicted Ernull phenotype, and a missense variant c.7174G>A p.(Glu2392Lys) in close proximity to p.2394 causes loss of both Era and Erb expression. STUDY DESIGN AND METHODS: We investigated PIEZO1 in four Er(a-) individuals who presented with anti-Era. Whole genome sequencing (WGS) and Sanger sequencing were performed. The location and structural differences of predicted protein changes were visualized using the predicted 3-D structure of Piezo1 created using AlphaFold2. RESULTS: One individual was homozygous for the reported ER*B. A second had a novel heterozygous nonsense variant c.3331C>T p.(Gln1111*), but a second allelic variant was not found. In the remaining two individuals, two different heterozygous novel missense variants, c.7184C>T p.(Ala2395Val) or c.7195G>A p.(Gly2399Ser), were in trans to the reported c.7180G>A variant, ER*B. AlphaFold2 protein modeling showed that each of the missense variants is predicted to encode an altered structural conformation near Era and Erb. CONCLUSIONS: Investigation of archived samples resulted in the identification of three novel PIEZO1 alleles including a predicted Ernull and two missense variants. Structural modeling suggests that the missense changes potentially alter Era/Erb epitope expression with p.2399Ser resulting in a small increase in the negative electrostatic potential.

Two new RHD alleles with deletions spanning multiple exons.

BACKGROUND: The most common large-deletion RHD allele (RHD*01N.01) includes the entire coding sequence, intervening regions and untranslated regions. The rest of large-deletion RHD alleles reported to-date consist of single-exon deletions, such as RHD*01N.67 which includes exon 1. MATERIALS AND METHODS: Samples from two donors with RhD-negative serology yielded unclear or inconclusive results when subject to confirmatory testing on RHD genotyping arrays. To determine their RHD genotypes, genomic DNA was analyzed with a combination of allele-specific PCR, long-range PCR, Sanger sequencing, and next-generation sequencing assays. RESULTS: Allele-specific PCR failed to detect products for RHD exons 1 to 3 in one sample and RHD exons 1 to 5 in the other. A quantitative next-generation sequencing assay confirmed deletion of exons 1 to 3 and 1 to 5 respectively, and detected the absence of an RHD gene in trans in both samples. Long-range PCR and Sanger sequencing enabled identification of the breakpoints for both alleles. Both deletions start within the 5' Rhesus box (upstream of the identity region for the 1-to-3 deletion, downstream of it for the 1-to-5 deletion), and end within introns. CONCLUSIONS: Resolution of unclear or inconclusive results from targeted genotyping arrays often leads to the discovery of new alleles. The 5' Rhesus box may be a hot spot for genetic recombination events, such as the large deletions described in this report.

RH genotyping by nonspecific quantitative next-generation sequencing.

BACKGROUND: Conventional sequencing uses gene-specific primers to determine the location of RH variants and permits a qualitative assessment of zygosity. Whole-genome and whole-exome sequencing determine the genetic location of variants and enable a quantitative assessment of zygosity. Nonspecific sequencing uses RH-consensus primers to detect variants and sequencing-read ratios to quantify their copy number. STUDY DESIGN AND METHODS: Two hundred seventy eight samples with diverse genotypes were analyzed by next-generation sequencing with RH- consensus primers. Custom-developed data analysis software was used to detect individual variants and infer the RH genotype. The method was evaluated for its quantitative nature, its ability to discriminate similar genotypes, its accuracy to detect variants, and its accuracy to assign them to RHD or RHCE. RESULTS: As a measure of balanced amplification of RHD and RHCE sequences, observed ratio medians deviate from expected ratios by 3% or less of the ratio range. As a measure of discriminatory power, contiguous RHCE / [RHD + RHCE] ratio averages are separated by 4 or more standard deviations of the mean. Variants are detected with a sensitivity and specificity greater than 99%, and variants at consensus positions are correctly assigned to RHD vs RHCE with a sensitivity greater than 72% and a specificity greater than 99%. The method is successful in the identification of genotypes with large conversion events and in the detection of copy number variation. CONCLUSION: Nonspecific sequencing of homologous gene sets combines detection and quantification of genetic variation in a single assay. Evidence is provided for the quantitative nature of the method, its sensitivity and specificity, and its ability to identify complex RH genotypes.

Concordance of two polymerase chain reaction-based blood group genotyping platforms for patients with sickle cell disease.

CONCLUSIONS: In recent years, polymerase chain reaction-based genotyping platforms, which provide a predicted phenotype, have increased in both patient and high-throughput donor testing, especially in situations where serologic methods or reagents are limited. This study looks at the concordance rate between two platforms commercially available in the United States when used for testing samples from patients with sickle cell disease (SCD), a group particularly vulnerable to alloimmunization. DNA extracted from samples from 138 patients with SCD was tested by human erythrocyte antigen (HEA) BeadChip (Immucor, Norcross, GA) and by ID CORE XT (Progenika-Grifols, Barcelona, Spain). Predicted phenotype results were compared, and a concordance rate was calculated. Discrepancies were resolved by Sanger sequencing. All testing was done under an institutional review board-approved protocol. A concordance rate of 99.9 percent was obtained. Sanger sequencing was performed on four samples with discrepancies in the Rh blood group system. Three samples had a similar allelic variant detected by ID CORE XT. Two of the three discrepant samples were correctly identified as V+w, VS- by ID CORE XT but not by HEA BeadChip. The third sample, predicted to have a phenotype of V+, VS+ by sequencing, was called correctly by HEA BeadChip but not by ID CORE XT, which had predicted V+w, VS-. The fourth discrepancy was identified in a sample that ID CORE XT accurately identified as RHCE*ce[712G] and predicted a partial c phenotype. This result was confirmed by Sanger sequencing, whereas HEA BeadChip found no variants and predicted a c+ phenotype. The high concordance rate of the two methods, along with the known limitations of serology, warrant further discussion regarding the practice of serologic confirmation of extended phenotypes. Clinical significance of the identified discrepancies remains to be determined.

Effect on gene expression of three allelic variants in GATA motifs of ABO, RHD, and RHCE regulatory elements.

BACKGROUND: Only a few genetic variants have been reported in regulatory elements of blood group genes. Most of them affect GATA motifs, binding sites for the GATA-1 transcription factor. STUDY DESIGN AND METHODS: Samples from two patients and one donor with unusual or discrepant serology for ABO, RhD, and RhCE antigens were analyzed by DNA sequencing. Analyzed regions included the coding sequence and portions of regulatory elements. The effect of some variants on gene expression was evaluated in reporter gene assays. RESULTS: Three new alleles were identified. Their key variants are located in the ABO Intron 1 enhancer, the RHD proximal promoter, and the RHCE proximal promoter. IVS1 + 5859A was found in an African American patient with a group O forward type and a group B reverse type. 5'UTR-115C was the only RHD variant sequence found in a mixed-race black and Caucasian prenatal patient showing mixed-field agglutination with anti-D. 5'UTR-83T was found in several black donors and patients in the context of the genetically related RHCE*ceBI and RHCE*ceSM alleles. Reporter assays of promoter constructs including 5'UTR-115C or 5'UTR-83T showed a significant reduction in RH gene expression. CONCLUSION: Three new alleles in the ABO, RHD, and RHCE genes consist of single-nucleotide changes within GATA motifs, emphasizing the key role of GATA transcription factors in the expression of blood group genes.

Red blood cell phenotype prevalence in blood donors who self-identify as Hispanic.

CONCLUSIONS: Molecular genotyping platforms provide a quick, high-throughput method for identifying red blood cell units for patients on extended phenotype-matching protocols, such as those with sickle cell disease or thalassemia. Most of the antigen prevalence data reported are for non-Hispanic populations. Therefore, this study sought to determine the phenotype prevalence in a single blood center's Hispanic population and to compare those results with previously reported rates in non-Hispanic donor populations. We performed a retrospective review of all serologic and molecular typing from donors who self-reported as Hispanic. The phenotype prevalence was reported and compared with rates from other racial/ethnic groups. A total of 1127 donors who selfidentified as Hispanic were screened by serologic methods for Rh and Kell antigens, and 326 were subsequently selected for molecular typing. The most prevalent probable Rh phenotypes were R1r (26.6%), R1R2 (21.5%), and R1R1 (20.7%); rr was found in 7.8 percent of donors tested. The percentage of K+ donors in this population was 2.8 percent. The most prevalent Duffy phenotypes were Fy(a+b+) (35.9%), Fy(a+b-) (35.6%), and Fy(a-b+) (27%). Of the donors studied, 15.3 percent had an FY GATA mutation. Only 1.5 percent of the donors were Fy(a-b-). The Jk(a+b+) phenotype was found in nearly half of the population. M+N+S+s+ was the most prevalent MNS phenotype from that group, constituting 22.4 percent. A total of 95.7 percent of the donors were Lu(a-b+), and Di(a-b+) was observed in 94.4 percent. The most prevalent Dombrock phenotype was Do(a+b+), constituting 46.9 percent, followed closely by Do(a-b+) at 40.5 percent. Hispanic donor antigen prevalence is distinctly different from other racial/ethnic groups and should be considered when attempting to find extended matched units for these patients.

Identification of new KLF1 and LU alleles during the resolution of Lutheran typing discrepancies.

BACKGROUND: The Lu(b) antigen is expressed on red blood cells (RBCs) of the majority of individuals in all populations. Its absence in transfused patients may lead to alloantibody production and mild-to-moderate transfusion reactions, and in pregnancies to mild hemolytic disease of the fetus and newborn. This report describes the results of discrepancy resolution between apparent LU*A/LU*B or LU*B/LU*B genotypes and apparent Lu(b-) or Lu(b+ weak) phenotypes in one Asian and 10 Caucasian blood donors. STUDY DESIGN AND METHODS: Whole blood samples were analyzed by molecular methods to resolve discrepancies between Lu(b-) phenotypes detected by serology and Lu(b+) phenotypes predicted by genotyping. RBC agglutination assays were performed with commercial and patient antisera by tube or gel column methods. Genotyping was performed on commercial arrays that target the LU*A/LU*B polymorphism at Position c.230. The discrepancies were resolved by sequencing of genomic DNA and in some cases by sequencing of cloned DNA fragments. RESULTS: Eleven new alleles with coding sequence variants were identified, seven in the KLF1 gene, which are presumed to act dominantly to silence LU expression, and four in the LU gene itself. The alleles are KLF1*114delC, KLF1*298T, KLF1*304C,484insC, KLF1*304C,1000del2, KLF1*621G, KLF1*948delC, KLF1*1040A,1045delT, LU*B(559T,711T,714T), LU*B(611A,638T), LU*B(1049del2ins3), and LU*B(1306T,1340T,1671T,1742T). CONCLUSION: Besides confirming common phenotypes and detecting rare antigen-negative phenotypes, the use of molecular methods in blood donor typing can prompt the identification of new alleles through discrepancy resolution.

Identification of six new RHCE variant alleles in individuals of diverse racial origin.

BACKGROUND: The introduction of molecular methods into routine blood typing is prompting the identification of new blood group alleles. Discrepancies between the results of genotyping and serology or chance events uncovered during genotyping prompted additional investigations, which revealed six new RHCE variant alleles. STUDY DESIGN AND METHODS: Samples from eight blood donors, two patients (one prenatal), and a patient's relative, all of diverse racial origin, were analyzed by standard serology methods, targeted genotyping arrays, DNA sequencing, and allele-specific polymerase chain reaction. RESULTS: Six new RHCE alleles were identified, namely, RHCE*cE84A, RHCE*ce202G, RHCE*ce307T, RHCE*Ce377G, RHCE*ce697G,712G,733G,744C, and RHCE*Ce733G. CONCLUSION: While implementation of new assays in commercial genotyping platforms to detect the polymorphisms reported here may not be justified given their apparent rarity, software interpretative algorithms may benefit from the identification of new alleles for a more accurate determination of genotypes and prediction of phenotypes.

FY*A silencing by the GATA-motif variant FY*A(-69C) in a Caucasian family.

BACKGROUND: The c.1-67C variant polymorphism in a GATA motif of the FY promoter is known to result in erythroid-specific FY silencing, that is, in Fy(a-) and Fy(b-) phenotypes. A Caucasian donor presented with the very rare Fy(a-b-) phenotype and was further investigated. STUDY DESIGN AND METHODS: Genomic DNA was analyzed by sequencing to identify the cause of the Fy(a-b-) phenotype. Samples were collected from some of his relatives to establish a correlation between the serology and genotyping results. Red blood cells were analyzed by gel column agglutination and flow cytometry. Genomic DNA was analyzed on genotyping microarrays, by DNA sequencing and by allele-specific PCR. RESULTS: In the donor, a single-nucleotide polymorphism T>C within the GATA motif was found at Position c.1-69 of the FY promoter and shown to occur in the FY*A allele. His genotype was found to be FY*A(-69C), FY*BW.01. In six FY*A/FY*B heterozygous members of the family, a perfect correlation was found between the presence vs. absence of the FY*A(-69C) variant allele and a Fy(a-) vs. Fy(a+) phenotype. CONCLUSION: The location of the c.1-69C polymorphism in a GATA motif whose disruption is known to result in a Fy null phenotype, together with the perfect correlation between the presence of the FY*A(-69C) allele and the Fy(a-) phenotype support a cause-effect relationship between the two.

Weak D type 67 in four related Canadian blood donors.

Correct donor D typing is critical to prevent recipient alloimmunization. No method can detect all variants, and the immunogenicity of many variants is unknown. Routine ABO and D serologic typings are performed in our laboratory by automated microplate testing. Until 2011, routine confirmation of D- status of first-time donors was performed by the manual tube indirect antiglobulin test (IAT); this was replaced by automated solid-phase testing including weak D testing by IAT. Selected donors are investigated by other methods. We describe four weak D type 67 (RHD*01W.67) donors whose samples tested as D- by automated microplate and manual methods but were later determined to be D+ by automated solid-phase and RHD gene analysis. Solid-phase serologic and molecular typing results of all four donors were identical. It was identified that the donors are of English-Irish descent; two are brothers and the others are cousins. Transfusion of blood from one of these donors likely resulted in alloimmunization to D in one of three recipients tested since no other documented exposures were identified. Lookback studies determined that two other D- recipients were not alloimmunized.

Strategy for managing maternal variant RHD alleles in Rhesus D negative obstetric populations during fetal RHD genotyping.

OBJECTIVES: Fetal RHD screening programs that aim to reduce unnecessary antenatal anti-D prophylaxis are being introduced into clinical practice. Strategies to manage women serologically typed as Rhesus D negative who have maternal RHD variants are needed. This study describes maternal RHD allelic variants detected in nonselected and alloimmunised Rhesus D negative obstetric populations and explores a mathematical approach to identify these variants. METHODS: Fetal RHD status was defined by testing cell-free fetal DNA in maternal plasma. Women at risk of an RHD variant were identified by selection for C and E haplotypes or by recognition of low polymerase chain reaction cycle threshold on fetal RHD typing. Maternal RHD alleles were defined by SNP profiling or sequencing. RESULTS: The prevalence of RHD variants in nonselected and alloimmunised groups was 1% (6/603) and 5.5% (6/110), respectively (p < 0.001). An inverse association between RHD cycle threshold values and gestational age was described by a linear model (p < 0.001). Standard residual values with a Z score threshold of -3.00 would have detected all maternal variants with one (1/713) false positive. CONCLUSIONS: The prevalence of maternal RHD variants was significantly higher in alloimmunised cases. The causative mechanism for this needs further investigation. Mathematical modeling simplifies the detection of maternal RHD variants.

New RHCE variant alleles encoding the D- - phenotype.

BACKGROUND: Variant alleles that do not produce RhCE antigens are rare. Consequently, they pose a challenge to transfusion when found in alloimmunized patients and make blood units valuable when found in donors. STUDY DESIGN AND METHODS: Five index cases and their relatives were studied by both serologic and molecular techniques. Genomic DNA was subjected to microarray genotyping, sequencing, exon scanning, and/or copy number determination assays to identify the RHCE allele(s) responsible for their D+ C- c- E- e- (D- -) phenotype. RESULTS: The five apparent D- - phenotypes were confirmed by molecular methods. Three of them contained unreported RHCE-null alleles, namely, RHCE*Ce-D(3-9)-Ce, RHCE*Ce87_93insT, and RHCE*cE221A. CONCLUSION: Molecular analysis of D- - phenotypes allows the identification of new RHCE-null variants. Conversely, detection of described RHCE-null variants facilitates confirmation of D- - phenotypes in patients and donors, helping improve transfusion safety.