The human GPIa/Ila complex, also known as integrin a2Bfl, serves as a major receptor for collagen in platelets and other cell types. In addition to its role in platelet adhesion to extracellular matrix, GPIa/IIa is also known to bear the clinically important Br' and Br' alloantigenic determinants, which can result in antibody-mediated platelet destruction. Immunochemical studies showed that the Br antigenic epitopes reside solely on the GP Ia subunit and do not depend on sialic acid residues. To define the polymorphism responsible for the Br alloantigen system platelet RNA PCR technique, was used to amplify GPIa mRNA transcripts. Nucleotide sequence analysis of the amplified platelet GPIa cDNA from Bra/a and Brb/b individuals revealed a single A * G polymorphism at base 1648. MnlI RFLP analysis of cDNA from serologically determined individuals confirmed that this polymorphism segregates with Br phenotype. This single base change results in a substitution of Lys (AAG) in Bra to Glu (GAG) in Brb at amino acid residue 505. In spite of the reversal in charge at this position, however, we found no difference in the ability of Bra and Br' homozygous platelets to adhere to collagens types I, III, or V, nor did antiBra or anti-Br' alloantibodies interfere with platelet adhesion to any of these fibrillar collagens. The identification of the nucleotide substitution that defines the Bra/Brb alloantigen system will now permit both pre-and postnatal diagnosis for Br phenotype. (J. Clin. Invest. 1993. 92:2427-2432

The diallelic human platelet alloantigen systems 1-5 have been found to result from single base pair substitutions in the encoding genes of platelet membrane glycoproteins IIIa, Ib, IIb and Ia. This is the basis of DNA methods for determination of platelet alloantigens. In this study, 98 blood donors were typed in the HPA-1, 2, 3 systems and, for the first time, in the HPA-5 system. Serologically obtained data (MAIPA and platelet agglutination) were compared with results from analysis of restriction fragment length polymorphisms (RFLP). Discordances were found in the HPA-2 and 3 systems and can be ascribed to false typing results in both the serological and genomic methods. In the HPA-1, 2 and 5 systems, all samples were typed correctly with RFLP analysis. Serologically, two donors were falsely typed positive with anti-HPA-2b in platelet agglutination and one donor with anti-HPA-3a in MAIPA assay. In the HPA-3 system, another four donors were misinterpreted to be HPA-3a negative in RFLP analysis. Possible technical problems in PCR-RFLP-typing are discussed and another strategy of HPA-1 typing using the restriction enzyme Scr FI is evaluated.

SummaryAlloimmunization against the human platelet alloantigen system Br (HPA-5) is the second most common cause of neonatal alloimmune thrombocytopenia (NAIT) in Caucasian populations. We have recently shown that a single base polymorphism at position 1648 on platelet mRNA coding for GPIa results in an aminoacid substitution at position 505 on the mature GPIa which is associated with the two serological defined Br phenotypes.Since DNA-typing of platelet alloantigens offers possibilities for useful clinical applications, we designed genomic DNA-based restriction fragment length polymorphism (RFLP) typing for Br alloantigens. To establish this technique we analyzed the genomic organization of GPIa adjacent to the polymorphic base. Using the polymerase chain reaction (PCR) of blood cell DNA we have identified two introns (approximately 1.7 and 1.9 kb) flanking a 144 bp coding sequence of the GPIa gene encompassing the polymorphic base 1648. Based on the in- tron sequence, a PCR primer was constructed to amplify a 274 bp fragment which was used for allele-specific RFLP to determine the Br genotypes. The results of RFLP analysis using Mnll endonuclease obtained from 15 donors (2 Br37*, 2 Br^ and 11 Brb/b) correlate perfectly with serological typing by monoclonal antibody-specific immobilization of platelet antigens (MAIPA) assay.

In order to determine whether platelets contain specific messenger RNA encoding for HLA class I molecules, polymerase chain reaction (PCR) was performed with RNA from different platelet donors. Two amplified 300 bp and 279 bp cDNA fragments were obtained which encompassed sequences from 321 to 620 and from 795 to 1073. The 300 bp fragment encodes exon 2 and exon 3, the 279 bp encodes a portion of exon 4, exon 5, exon 6 and a portion of exon 7. A 300 bp nested PCR product from one donor, that encoded for the highly polymorphic region alpha 2, was cloned and sequenced. The resulting nucleotide sequences fitted to the expected sequence for HLA B*3801 of this donor. Sequence analysis of the 279 bp PCR product demonstrated the presence of exon 5 encoding for the 117 bp transmembrane domain. In addition, de novo protein biosynthesis was studied by radioimmunoprecipitation of HLA class I molecules from 35S-methionine metabolically labelled platelet lysates with a monoclonal antibody (mab) w6/32 specific for a monomorphic epitope on the heavy chain of HLA class I antigens. Analysis of the immunoprecipitates on SDS polyacrylamide gel electrophoresis showed a specific band with apparent molecular weight (M(r)) of 44 kD corresponding to integral membrane HLA protein. On the basis of these results, we conclude that platelets contain specific messenger RNA encoding for HLA class I molecules and have the capability to synthesize the integral HLA membrane protein.

Presuming that the mutation is not a singular phenomenon and also occurs with the PlA1 haplotype, it could lead to false interpretations of restriction analysis with Msp I. To exclude that possibility, Nci I is preferred for restriction fragment length polymorphism typing in the PlA system.