Red blood cell antigen phenotype by DNA analysis

Ghazala Hashmi

INTRODUCTION

The association of the majority of red blood cell(RBC) antigens with single-nucleotide polymor-phisms (SNPs)1 provides the basis for determin-ing blood group antigen expression at the

deoxyribonucleic acid (DNA) level. Molecular immunohe-matology, by invoking DNA analysis, not only reduces theneed for increasingly rare serologic reagents, but alsopermits the reliable determination of a phenotype in situ-ations that are difficult to resolve by serologic methods,especially when the available antibody reagents are onlyweakly reactive. These applications of DNA analysisinclude the determination of antigens with weak oraltered expression, the analysis of patients in a variety ofconditions including, e.g., chemotherapy or transfusiontherapy, and the identification of fetuses at risk forhemolytic disease of the newborn.2 In the context ofimmunohematology, DNA analysis has been used previ-ously to screen for allele combinations by amplification ofselected mutations and polymorphisms by sequence-specific primers and to analyze specific alleles by allele-specific polymerase chain reaction (PCR), restrictionfragment length polymorphism analysis, or real-timePCR.2 Although these methods have led to advances in theidentification of alleles, they are not ideal for a large-scaleclinical application given their complexity, labor-intensiveformat, and low throughput. Most recently, when imple-mented in an array format, high-throughput DNA analysiswas shown to permit the rapid determination of entiresets of designated SNPs associated with platelet andminor blood group antigens.3-6 The BeadChip format

(BioArray Solutions, Warren, NJ) presented here adopts abead assay format, long established in diagnostic applica-tions,7,8 to optimize assay performance and combines itwith a novel method of bead array assembly to producerandom encoded microparticle arrays that have beenshown to permit rapid customization and to providereliable performance in the clinical setting. Thus, theBeadChip format has been applied to nucleic acid as wellas to protein analysis for a variety of applications, includ-ing immunohematology, the analysis of both donor andpatient samples, transplantation medicine, geneticdisease carrier screening, including the application toneonatal screening, and autoantibody profiling (http://www.BioArrayS.com).

This short review, after a brief description of key tech-nical aspects of the BeadChip format and assay deliverysystem, gives a summary of conclusions from two recentclinical studies, one focused on the analysis of a large anddiverse set of blood donors and the other focused on theanalysis of patient samples typically seen in a metropoli-tan transfusion service. The results of these studiessupport the conclusion that BeadChip DNA analysispermits the determination of an extended human eryth-rocyte antigen (HEA) phenotype (Table 1) for a diversegroup of donors and for patients, in various clinicalconditions.

ABBREVIATIONS: HEA = human erythrocyte antigen;

PB = peripheral blood; SCT = sickle cell trait; SNP(s) = single-

nucleotide polymorphism(s).

From BioArray Solutions Ltd, Warren, New Jersey.

Address reprint requests to: Ghazala Hashmi, PhD, BioArray

Solutions Ltd, 35 Technology Drive, Warren, NJ 07059; e-mail:

ghashmi@bioarrays.com.

BioArray Solutions Ltd, Warren, NJ, uses “BeadChip” as its

trademark.

doi: 10.1111/j.1537-2995.2007.01312.x

TRANSFUSION 2007;47:60S-63S.

TABLE 1. Composition of HEA BeadChip formatBlood factor Analyte Polymorphism

Rh Cc 203A>G, Int 2Ee 676G>C

Kell K/k 698T>CKidd Jka/Jkb 838G>ADuffy Fya/Fyb FY125G>ADuffy-GATA Silencing FY FY-33T>C

Fyx (Fy[b+w] ) 265C>TMNS GYPA (M/N) 60C>T

GYPB (S/s) 143T>CGPB Silencing 230C>T, +5 g>t

Lutheran Lua/Lub 230A>GDiego Dia/Dib DI2561T>CColton Coa/Cob CO134C>TDombrock Doa/Dob DO793A>G

Jo(a+)/Jo(a–) DO350C>THy+/Hy– DO323G>T

Landsteiner-Wiener LWa/LWb LW308A>GScianna Sc1/Sc2 SC169G>AHemoglobin S HbS 173A>T

60S TRANSFUSION Volume 47, July 2007 Supplement

THE BEADCHIP FORMAT

Making BeadChip platformBeadChip manufacturing entails threefundamental operations, namely: pro-ducing collections (“libraries”) ofencoded microarticles (“beads”) func-tionalized with allele-specific oligo-nucleotides (or other capture moieties);pooling aliquots of bead suspensionsand assembling arrays by forming andimmobilizing monolayers of beads indesignated areas of a patterned siliconwafer (or wafer section); singulating thewafer section into individual chips; andplacing sets of chips into a desired con-figuration, e.g., a 1 ¥ 8 configuration, oran 8 ¥ 12 configuration on a carrier(Fig. 1).

Using BeadChip platformThe BeadChip assay protocol, as previ-ously described,6 includes a multiplexedPCR reaction, post-PCR processing togenerate single-stranded targets, andon-chip allele identification by way ofenzyme-mediated probe elongation.That is, allele-specific probes contain-ing variable 3′ termini matching eitherthe normal or a variant allele act asnested primers for the simultaneouselongation of matching probes by aDNA polymerase. Each bead producesan assay signal reflecting the incorpora-tion of fluorescently labeled dNTPs into the elongationproducts displayed on that bead. Using an automatedarray imaging system provided by BioArray Solutions,assay signal intensity patterns are recorded from sets ofchips and passed to software for automated analysis. Thisanalysis produced a reference table for genotype-to-phenotype conversion, showing the combinations ofSNPs at one or more variable sites probed by the BeadChippanel, to antigen expression state (Table 2). The signalintensities produced by paired probes are converted intodiscrimination ratios, and these in turn are converted tophenotypes by reference to look-up tables, thereby elimi-nating manual transcription. The assay signal intensitiesproduced for each SNP by the corresponding pair of elon-gation probes provides the basis for allele discriminationas described previously.6 The BeadChip system thusincludes kits providing complete assay protocols, and aproprietary assay delivery system providing automated“snapshot” array imaging and web-hosted data manage-ment (http://www.BioArrayS.com).

DNA ANALYSIS OF A DIVERSE GROUP OFNEW YORK DONORS

In collaboration with Dr. Marion Reid and colleagues atthe New York Blood Center, BeadChip HEA analysis wasapplied to determine the HEA phenotype for 2355 donorsfrom four principal ethnic backgrounds.9 DNA wasextracted from white blood cells in whole blood of donorsof self-identified ethnicity and partially known minorantigen phenotype. The initial portion of the study wasperformed with a panel of 18 pairs of allele-specificprobes, each pair directed to a specific polymorphic site;this set was later expanded, for an additional 369 donors,to include probes directed to mutations known to silenceGYPBS, namely GYPB 230 C>T and intron 5 +5 g>t.10 Hap-lotype analysis was invoked to predict the expression ofantigens controlled by alleles comprising multiple SNPs,such as those in the MNS and Duffy blood group sys-tems.11,12 Reference tables were designed to facilitategenotype-to-phenotype conversion and to estimate

Fig. 1. Illustration of BeadChip system showing the bead assembly protocol on 6�

silicon wafer, array imaging system, and decoding and assay images. Also shown is

an example of HEA data analysis charts generated automatically by BeadChip analy-

sis software using web-based software (wHEA; BioArray Solutions, Warren, NJ),

showing 28 antigen states derived from DNA analysis, linked to a BeadChip

carrier-ID and sample-ID, stored in a database.

TABLE 2. Example of genotype-to-phenotype conversion tablesG FKEL(698T>C) FJK(838G>A) FDI(2561T>C) FCO(134C>T)

AA K+k– Jk(a+b–) Di(a+b–) Co(a+b–)AB K+k+ Jk(a+b+) Di(a+b+) Co(a+b+)BB K–k+ Jk(a–b+) Di(a–b+) Co(a–b+)G FLW(308A>G) FLU(230A>G) FSC(169G>A) FHGBS(173A>T)AA LW(a+b–) Lu(a+b–) Sc:1,2 SCT–AB LW(a+b+) Lu(a+b+) Sc:1,2 SCT+BB LW(a–b+) Lu(a–b+) Sc:-1,2 HgbSS

* SCT = sickle cell trait.

RBC ANTIGEN PHENOTYPE BY DNA ANALYSIS

Volume 47, July 2007 Supplement TRANSFUSION 61S

antigen prevalence. The analysis of SNPs associated with24 antigens of 10 blood group systems was compared withhistorical serologic phenotypes when available.

BeadChip analysis produced results that were con-cordant with the serologic determination of antigen nega-tivity for 4510 antigens. Eight discordant results were dueto silencing of GYPBS and 16 were likely errors in recordingserologic results or data entry. The analysis produced newantigen-negative typings that had not been serologicallydefined, identified 21 rare donors, and determined allelefrequencies and corresponding antigen prevalences forfour ethnicities (Table 3) that are comparable to previ-ously described studies.13

DNA ANALYSIS OF A DIVERSE GROUPOF PATIENTS14

In collaboration with Drs Yoon Choo and Carolyn Whitsettof Mount Sinai School of Medicine, New York, NY, theclinical utility of the BeadChip HEA analysis for patients incomplex clinical situations was investigated.

It has been well recognized that serologic phenotyp-ing of blood group antigens is unreliable or invalid whenpatients have had a recent transfusion or their direct anti-globulin test is positive for immunoglobulin G. In con-trast, DNA analysis is unaffected by these conditions. Thisis illustrated in Table 4 for a patient in sickle cell crisiswho received 6 units of compatible, but not phenotype-matched, blood. Samples selected for DNA analysis were:pre- and posttransfusion peripheral blood (PB) samplesfrom 20 patients who received up to 16 units of leukore-duced (n = 16) or up to 12 units of nonleukoreduced

(n = 4) RBCs within a few days; and PB samples from 33patients with severe leucopenia, and (n = 60) abnormalchemistry from liver or kidney dysfunction. Genomic DNAwas extracted from 200 uL of PB, amplified, and analyzedto determine alleles of the blood group systems. In all 20paired transfusion samples, identical genotypes were con-firmed after transfusions. The yields of genomic DNAextracted from 200 uL of leukopenic PB were sufficient foranalysis in all 33 cases. All genomic DNA samples (n = 60)derived from PB samples from patients with abnormalchemistries, which included 22 with renal insufficiency, 34with liver dysfunction, 2 with hemolysis, and 2 with hyper-gammaglobulinemia, were successfully amplified. Insummary, blood group genotyping was reliably achievedin several clinical situations tested: patients who receivedtransfusions of multiple RBC units, patients with severeleukopenia, and patients with abnormal serum chemistry.

DISCUSSION

As with human leukocyte antigen molecular typing todetermine the expression of human leukocyte antigen andpermit the selection of compatible prospective marrowand organ donors for a given recipient, the determinationof HEA phenotype invokes the analysis of alleles govern-ing the expression of antigens expressed, in this case, onRBCs. It was the objective of the large-scale study fordonors, and the study of patients in complex clinical situ-ations, to assess the viability of DNA analysis as a rapidand reliable method for determining RBC antigens. Thestudies, as summarized here, support the conclusion thatBeadChip DNA analysis, including automated array

TABLE 3. Genotype frequencies (%) identified in various ethnic groupsEthnicity Genotype FYA/FYB GATA FY-265 GPBS GPA K1/K2 JKA/JKB LUA/LUB HbS173

African American AA 4 12 99 7 41 0 54 0 95AB 19 34 1 24 32 2 37 6 5BB 77 55 0 69 28 98 9 94 0

Caucasian AA 18 96 98 13 34 0 30 0 100AB 43 2 2 40 44 5 44 5 0BB 39 1 0 46 22 95 26 95 0

Hispanic AA 15 68 99 13 39 0 36 1 100AB 45 26 1 32 45 2 42 3 0BB 40 6 0 55 17 98 22 97 0

Asian AA 69 98 100 8 27 0 22 0 100AB 25 2 0 12 57 0 53 2 0BB 6 0 0 80 16 100 25 98 0

TABLE 4. Phenotype determined by serotyping and DNA analysis before and after transfusion. A 15-year-oldAfrican American female patient with sickle cell disease and pain crisis received a total of 6 leukoreduced

RBC unitsK k Fya Fyb Jka Jkb M N S s

Pretransfusion Serotype – + – – + – + + – +Genotype

Posttransfusion Serotype – + + + + + + + +W +Genotype – + – – + – + + – +

HASHMI

62S TRANSFUSION Volume 47, July 2007 Supplement

imaging and computerized data management, is suitablefor these applications.

The ability to test a large number of donors simulta-neously for several antigens, in combination with com-puterized analysis and interpretation of data, makesit feasible to maintain a diverse inventory of well-characterized donors. For patients, the ability to reliablydetermine antigen expression corresponding to the phe-notype, regardless of clinical condition and therapeuticregimen, holds the potential to substantially reduce delays(for patients having multiple antibodies) incurred underthe current “Type & Screen” paradigm in procuring com-patible blood units for transfusion.15 As with the currentparadigm, compatibility is to be confirmed by an anti-globulin crossmatch. With the inclusion of polymor-phisms to identify RhCcEe haplotypes and correspondingantigen expression, the current HEA panel permits thesimultaneous analysis of 28 antigens in 11 blood groupsystems. The routine determination of an extended HEAphenotype by DNA analysis of donors and recipients willmake it possible to consider a new paradigm for the selec-tion of donors for given recipients on the basis of therespective antigen repertoires.

ACKNOWLEDGMENTS

The author is indebted to several colleagues at the New York

Blood Center, Mount Sinai School of Medicine, NY and BioArray

Solutions, NJ for their valuable collaboration, helpful discussion,

and technical help throughout the development of the HEA Bead-

Chip panel.

REFERENCES

1. Reid ME, Lomas-Francis C. Blood group antigen factsbook,

2nd ed. San Diego, CA: Academic Press, 2004.

2. Reid ME. Applications of DNA-based assays in blood group

antigen and antibody identification. Transfusion 2003;43:

1748-57.

3. Beiboer SH, Wieringa-Jelsma T, Maaskant-van Wijk PA,

et al. Rapid genotyping of blood group antigens by multi-

plex polymerase chain reaction and DNA microarray

hybridization. Transfusion 2005;45:667-79.

4. Bugert P, McBride S, Smith G, et al. Microarray-based

genotyping for blood groups: comparison of gene array

and 5′-nuclease assay techniques with human platelet

antigen as a model. Transfusion 2005;45:654-9.

5. Denomme GA, Van Oene M. High-throughput multiplex

single-nucleotide polymorphism analysis for red cell

and platelet antigen genotypes. Transfusion 2005;45:

660-6.

6. Hashmi G, Shariff T, Seul M, et al. A flexible array format

for large-scale, rapid blood group DNA typing. Transfusion

2005;45:680-8.

7. Nustad K, Johansen L, Ugelstad J, Ellingsen T, Berge A.

Hydrophilic monodisperse particles as solid-phase mate-

rial in immunoassays: comparison of shell-and-core par-

ticles with compact particles. Eur Surg Res 1984;16(Suppl.

2):80-7.

8. Taylor J, Brilley D, Nguyen Q, et al. Flow cytometric plat-

form for high-throughput single nucleotide polymorphism

analysis. Biotechniques 2001;30:661-9.

9. Hashmi G, Shariff T, Zhang Y, et al. Determination of 24

minor red blood cell antigens for more than 2000 blood

donors by high-throughput DNA analysis. Transfusion

2007;47:736-47.

10. Storry JR, Reid ME, Fetics S, Huang C-H. Mutations in

GYPB exon 5 drive the S-s-U+var phenotype in persons of

African descent: implications for transfusion. Transfusion

2003;43:1738-47.

11. Pogo AO, Chaudhuri A. The Duffy protein: a malarial and

chemokine receptor. Semin Hematol 2000;37:122-9.

12. Race RR, Sanger R. Blood groups in man, 6th ed. Oxford:

Blackwell Scientific Publications, 1975.

13. Mourant AE, Kopec AC, Domaniewska-Sobczak K. Distri-

bution of the human blood groups and other polymor-

phisms, Vol. 1, 2nd ed. London: Oxford University Press,

1976.

14. Choo Y, Enriquez E, Shariff T, Hashmi G, Whitsett C. Blood

group genotyping of selected patient populations by array

analysis. (Abstract # S82-040E), AABB, 2006.

15. Friedberg RC, Jones BA, Walsh MK. Type and screen

completion for scheduled surgical procedures. A College of

American Pathologists Q-Probe study of 8941 Type and

Screen tests in 108 institutions. Arch Pathol Lab Med

2003;127:533-40.

RBC ANTIGEN PHENOTYPE BY DNA ANALYSIS

Volume 47, July 2007 Supplement TRANSFUSION 63S