Eur J Haematol 2000: 65: 128±131 Copyright # Munksgaard 2000 Printed in UK. All rights reserved EUROPEAN JOURNAL OF HAEMATOLOGY ISSN 0902-4441 Gene frequencies of the HPA-3 and HPA-5 platelet antigen alleles among the Amerindians

Covas DT, BõÂscaro TA, Nasciutti DC, Guerreiro JF, Santos SEB, Zago D. T. Covas1,2,T.A.BõÂscaro2, MA. Gene frequencies of the HPA-3 and HPA-5 platelet antigen alleles D. C. Nasciutti1,2, J. F. Guerreiro3, among the Amerindians. S. E. B. Santos3, M. A. Zago1,2 Eur J Haematol 2000: 65: 128±131. # Munksgaard 2000. 1Department of Clinical Medicine and the 2Regional Blood Center of RibeiraÄo Preto, University of S. Paulo, Abstract: Platelet antigens are of importance in several clinical situations and 3Human Genetics Laboratory, University of Para, and in population genetics. Data are scarce on allele frequencies in ethnic groups other than whites, Asians and African Americans. The frequencies of the alleles of the systems HPA-3 and HPA-5 were determined using the allele-speci®c restriction enzyme for ®ve South- American Amerindian populations and compared with those obtained for Afro-Americans, Japanese and whites from Brazil. The frequency of the HPA-3a allele among the Amerindians as a group did not differ from the values obtained for the other populations. However, differences were observed among the Amerindians, varying from 0.27 to 0.75, the highest Key words: platelet antigens; HPA-3; HPA-5; frequency thus far observed for a population of Asian origin. Only the Amerindian populations; Brazilian populations HPA-5a allele was found among 130 Amerindian chromosomes. The Correspondence: Dimas T. Covas, Hemocentro de determination of gene frequencies of the HPA systems in different RibeiraÄo Preto, 14051±140 RibeiraÄo Preto, Brazil populations allows inference of gene ¯ows and genetic constitution of e-mail: [email protected] populations and the estimation of the risk of platelet-speci®c alloimmunization. Accepted for publication 16 March 2000

Platelet-speci®c antigens are of importance in frequencies have been extensively reported for neonatal alloimmune thrombocytopenia (NATP), whites and Asians, but few data have been post-transfusion purpura (PTP), refractoriness to published for other populations. We recently platelet transfusion therapy and population genetics reported the frequencies of HPA-1 and HPA-2 (1, 2). The 13 diallelic antigens (HPA 1 to 13) thus among six tribes of Amerindians from the Brazilian far reported are characterized by an antigen-speci®c Amazon (14), and in the present study we single base substitution leading to a single amino determined the distributions of the HPA-3 and acid difference in the corresponding glycoprotein HPA-5 alleles in Amerindians by allele-speci®c (3±8). The HPA-3 antigen system is located on restriction enzyme analysis and compared their glycoprotein IIb (GPIIb). The polymorphism frequencies with other selected populations. results from a C-T base substitution at position 2622 of the GPIIb gene which determines the presence of isoleucine (HPA-3a) or serine (HPA-3b) Material and methods at position 843 of the mature protein (9). In the Populations HPA-5 system an A-G substitution at position 1648 The Amerindian sample consisted of 65 individuals of the glycoprotein Ia (VLA-2a) gene determines belonging to ®ve tribes from the Brazilian Amazon: the presence of glutamic acid at position 505 of the 10 (Karib language), 14 Kayapo (Je lan- mature protein in the HPA-5a allele or of lysine in guage), 12 (Tupy-Guarani language), the HPA-5b allele (10, 11). The HPA-5 and HPA-3 17 -Apalai (Karib language) and 12 systems are the second most common causes of Yanomama. The individuals studied were appar- NATP and PTP, respectively (12, 13). HPA allele ently unrelated. Additional genetic characterization

128 HPA-3 and HPA-5 allele distribution in Amerindians of these groups has been published elsewhere (15). under UV light after electrophoresis on 2% agarose Three additional populations were also studied: 30 gel and ethidium bromide staining. whites (mostly of western and southern European ancestry), 30 blacks (excluding mulattos) and 30 Statistical analysis Japanese. Only individuals who reported the 2 absence of any racial admixture for their four Pearson's x test was used to compare sets of data, grandparents were included. with a signi®cance level of 5%.

Results and discussion DNA analysis The results are summarized in Table 1. In addition to Genomic DNA was obtained from leukocytes. The our previous report on the allele frequencies for DNA segments that contain the polymorphic HPA-1 and HPA-2 genes (14), a single Amerindian sequences of the HPA-3 and HPA-5 alleles were population has been studied thus far for the HPA-3 ampli®ed by the polymerase chain reaction (PCR). and HPA-5 genes (17). The set of primers described by Unkelbach et al. was In the present study, the frequency of the HPA- used (16). The HPA-3 primers are: forward 5k CTC 3 alleles among the Amerindians as a group did AAG GTA AGA GCT GGG TGG AAG AAA not differ from the values obtained for Brazilian GAC3k; reverse 5k CTC ACT ACG AGA ACT GGA Afro-Americans (HPA-3a, 0.65; HPA-3b, 0.35), TCC TGA AGC CTC 3k. The HPA-5 primers are: whites (HPA-3a, 0.50; HPA-3b, 0.50) or Japanese forward 5k GTG ACC TAA AGA AAG AGG 3k; and their descendants who live in Brazil (HPA-3a, reverse 5k CTC TCA TGG AAA ATG GCA G 3k . 0.55; HPA-3b, 0.45) (p>0.05). The frequencies The PCR for HPA-3 was carried out with 0.4 mgof found for these populations were similar to those genomic DNA with 20 pmol of speci®c primers and reported for European, North American and 2.0 U of Taq polymerase in a total volume of 50 mL. Asian populations (18±20). However, differences For the HPA-5 system, PCR was carried out with 0.1 were observed when comparing the different mg of DNA with 25 pmol of speci®c primers and Amerindian populations: the HPA-3a allele fre- 2.0 U of Taq polymerase in a total volume of 50 mL. quencies varied from 0.27 in the Wayampi to 0.75 The allele-speci®c restriction enzyme analyses were the Wayana-Apalai, the highest thus far reported performed using the enzymes FokI (4 IU) for HPA-3 for populations of Asian origin, and can be and MnlI (4 IU) for HPA-5 under conditions explained by limited sample size or the isolation recommended by the manufacturer (New England and genetic drift occurring in small groups. A BioLabs, Beverly, MA). Digests were analyzed similar heterogeneity of allele distribution among

Table 1. Frequencies of HPA-3 and HPA-5 alleles in Amerindians and other Brazilian populations

Gene frequency Genotype frequency (%) Populations HPA-3a HPA-3b a+ bx a+ b+ ax b+

Amerindians (126) 0.57 0.43 34.9 44.4 20.6 Arara (20) 0.70 0.30 50.0 40.0 10.0 Kayapo (28) 0.61 0.39 35.7 50.0 14.3 Wayampi (22) 0.27 0.73 9.1 36.4 54.5 Wayana-Apalai (32) 0.75 0.25 56.3 37.4 6.3 Yanomama (24) 0.46 0.54 16.7 58.3 25.0 Blacks (60) 0.65 0.35 43.3 43.3 13.4 Japanese (58) 0.55 0.45 34.5 41.4 24.1 Whites (56) 0.50 0.50 25.0 50.0 25.0 HPA-5a HPA-5b a+ bx a+ b+ ax b+ Amerindians (130) 1.00 0.00 100.0 0.0 0.0 Arara (20) 1.00 0.00 100.0 0.0 0.0 Kayapo (28) 1.00 0.00 100.0 0.0 0.0 Wayampi (24) 1.00 0.00 100.0 0.0 0.0 Wayana-Apalai (34) 1.00 0.00 100.0 0.0 0.0 Yanomama (24) 1.00 0.00 100.0 0.0 0.0 Blacks (60) 0.80 0.20 63.3 33.3 3.4 Japanese (58) 0.97 0.03 93.1 6.9 0.0 Whites (56) 0.97 0.03 93.3 6.7 0.0

The number of chromosomes analyzed for each population is given in parenthesis.

129 Covas et al. different Amerindian populations which results in 6. SIMSEK S, FOLMAN C, vAN DER SCHOOT CE, vON DEM BORNE AE. a high interpopulation diversity has been observed The Arg633His substitution responsible for the private platelet antigen Gro(a) unraveled by SSCP analysis and also for VNTRs and a-globin gene haplotypes direct sequencing. Br J Haematol 1997;97:330±335. (21). This ®nding explains why population studies 7. NORIS P, SIMSEK S, DE BRUIJNE-ADMIRAAL LG, et al. Max(a), a on Amerindians based on one or two populations new low-frequency platelet-speci®c antigen localized on may be completely misleading, even when large glycoprotein IIb, is associated with neonatal alloimmune numbers of individuals are studied. Thus, thrombocytopenia. Blood 1995;86:1019±1026. although Castro et al. found a gene frequency 8. SANTOSO S, KIEFEL V. Human platelet-speci®c alloantigens: update. Vox Sang 1998;74 (suppl. 2):249±253. of HPA-3b of only 0.243, our expanded data on 9. LYMAN S, ASTER RH, VISENTIN GP, NEWMAN PJ. Polymorph- ®ve tribes demonstrate a mean gene frequency of ism of human platelet membrane glycoprotein IIb associated 0.43 (range 0.25±0.73). Thus, in spite of the with Baka/Bakb alloantigen system. Blood 1990;75:2343± occasional differences in allele distribution of 2348. individual tribes when compared with other 10. SIMSEK S, GALLARDO D, RIBERA A, vOM DEM BORNE AEGK. The human platelet alloantigens, HPA-5(a+,bx) and Asian-derived populations, the HPA-3 allele HPA-5(ax,b+), are associated with a Glu505/Lys505 poly- frequencies of Amerindians as a group did not morphism of glycoprotein Ia (the a2 subunit of VLA-2). Br differ from those obtained in Japanese, Chinese J Haematol 1994;86:671±674. and Koreans (19, 20, 22, 23). The results obtained 11. SANTOSO S, KALB R, KROLL H, et al. A point mutation leads to for HPA-5 demonstrate that the HPA-5b allele is an unpaired cysteine residue and a molecular weight poly- morphism of a functional platelet beta 3 integrin subunit. The absent in the ®ve Amerindian populations studied Sra alloantigen system of GPIIIa. J Biol Chem 1994;18: by us and the single population studied by Castro 8439±8444. (17), and probably is absent or extremely rare in 12. BOEHLEN F, KAPLAN C, DE MOERLOOSE P. Severe neonatal the Amerindian populations of . alloimmune thrombocytopenia due to anti-HPA-3a. Vox The frequencies of the HPA-5 alleles observed Sang 1998;74:201±204. 13. MCCRAE KR, HERMAN JH. Post-transfusion purpura: two in the Afro-American, Japanese and white popu- unusual cases and a literature review. Am J Haematol 1996; lations from Brazil did not differ from those 52:205±211. observed in similar populations of North 14. COVAS DT, DELGADO M, ZEITUNE MM, GUERREIRO JF, SANTOS America, Asia and Europe (18, 22, 24±26). SEB, ZAGO MA. Gene frequencies of the HPA-1 and HPA-2 The determination of gene frequencies of the HPA platelet antigen alleles among the Amerindians. Vox Sang systems in different populations is of anthropologi- 1997;73:182±184. 15. ZAGO MA, SILVA WA Jr, TAVELLA MH, SANTOS SEB, cal and genetic importance, since it allows inference GUERREIRO JF, FIGUEIREDO MS. Interpopulational and of gene ¯ows and genetic constitution of popula- intrapopulational genetic diversity of Amerindians as tions. In addition, this kind of study allows the revealed by six variable number of tandem repeats. Hum estimation of the risk of platelet-speci®c alloimmu- Hered 1996;46:274±289. nization, NATP and PTP in selected populations. 16. UNKELBACH K, KALB R, SANTOSO S, KROLL H, MUELLER- ECKHARDT C, KIEFEL V. Genomic RFLP typing of human platelet alloantigens Zw(PlA), Ko, Bak and Br (HPA-1,2,3,5). Br J Haematol 1995;89:169±176. Acknowledgements 17. CASTRO V, ORIGA AF, ANNICHINO-BIZZACCHI JM, et al. Frequencies of platelet-speci®c alloantigen system 1±5 in This work was partly supported by CNPq and FUNDHERP three ethnic groups in Brazil. Eur J Immunogenet 1996; (Brazil). 26:355±360. 18. LEGLER TJ, KOÈ HLER M,MAYR WR, PANZER S,OHTO H,FISCHER GF. Genotyping of the human platelet antigen systems 1 through 5 by multiplex polymerase chain reaction and References ligation-based typing. Transfusion 1996;36:426±431. 1. NEWMAN PJ, MCFARLAND JG, ASTER RH. Alloimmune throm- 19. KIM HO, JIN Y, KICKLER TS, BLAKEMORE K, KWON OH, BRAY bocytopenias. In: LOSCALZO J, SCHAFER AI (eds), Thrombosis PF. Gene frequencies of the ®ve major human platelet and Hemorrhage. Boston: Blackwell Scienti®c Publications, antigens in African American, white, and Korean popula- 1994:529±543. tions. Transfusion 1995;35:863±867. 2. NEWMAN PJ, VALENTIN N. Human platelet alloantigens: recent 20. TANAKA S, TANIUE A, NAGAO N, et al. Simultaneous DNA ®ndings, new perspectives. Thromb Haemost 1995;74:234± typing of human platelet antigens 2, 3 and 4 by allele-speci®c 239. PCR method. Vox Sang 1995;68:225±230. 3. vON DEM BORNE AEG, DEÂ CARY F. ICSH/ISBT working party 21. ZAGO MA, SANTOS EJM, CLEGG JB, et al. a-Globin gene on platelet serology. Nomenclature of platelet-speci®c anti- haplotypes in South American Indians. Hum Biol 1995; gens. Vox Sang 1990;58:176. 67:535±546. 4. PANZER S, KAPLAN C. Report on the 1997 International 22. SEO DH, PARK SS, KIM DW, FURIHATA K, UENO I, HAN KS. Society of Blood Transfusion workshop for genotyping of Gene frequencies of eight human platelet-speci®c antigens in platelet alloantigens. Transf Med 1998;8:125±128. Koreans. Transf Med 1998;8:129±132. 5. PEYRUCHAUD O, BOURRE F, MOREL-KOPP MC, et al. HPA- 23. CHANG YW, MYTILINEOS J, OPEL G, HAWKINS BR. Distribution 10W(b)La(a): genetic determination of a new platelet-speci®c of human platelet antigens in a Chinese population. Tissue alloantigen on glycoprotein IIIa and its expression in COS-7 Antigens 1998;51:391±393. cells. Blood 1997;89:2422±2428. 24. REINER AP, ARAMAKI KM, TERAMURA G, GAUR L. Analysis of

130 HPA-3 and HPA-5 allele distribution in Amerindians

platelet glycoprotein Ia (a2 integrin) allele frequencies in homoduplex formation assay). Br J Haematol 1996;95: three north American populations reveals genetic association 198±203. between nucleotide 807C/T and amino acid 505 Glu/Lys 26. HOU M, ROSENGREN-KOGAN L, FORSBERG B, et al. Genotyping (HPA-5) dimorphisms. Thromb Haemost 1998;80:449±456. of the platelet-speci®c alloantigen HPA-5 (Bra/Brb) using 25. FUJIWARA K, ISA K, OKA T, et al. Large-scale DNA typing for polymerase chain reaction with sequence-speci®c primers human platelet alloantigens by PCR-PHFA (preferential (PCR-SSP). Eur J Haematol 1996;57:208±213.

131