EpHLA

Os Professores da Universidade Federal do Piauí, Dr. Luiz Cláudio Demes da Mata Sousa, Dr. Adalberto Socorro da Silva e Dra. Semiramis Jamil Hadad do Monte (coordenadora do Laboratório de Imunogenética e Biologia Molecular da Universidade Federal do Piauí) juntamente com seus colaboradores publicaram este mês no periódico Transplant immunology, o artigo intitulado:

EpHLA: An innovative and user-friendly software automating the HLAMatchmaker algorithm for antibody analysis.

Este software automatiza o uso do algortimo HLAmatchmaker, desenvolvido por René Duquesnoy, diminuindo o tempo necessário para a determinação de mismatchs aceitáveis (com base no conceito de eplets), de 3 horas para 5 minutos!

No artigo, relatamos dois casos interessantes cujo entendimento foi facilitado pelo emprego do software. A seguir colocamos os casos na íntegra, mas convidamos o leitor para uma leitura do artigo completo, que está disponível on line para open acess.

Caso 1.

A 55-year-old man with CDC assay negative received a kidney transplant from his mother. Eight years after transplantation he lost the kidney by chronic rejection. A serum screen of this recipient using single class I and II allele SPA Luminex panels (Labscreen; OneLambda, Canoga Park, CA) revealed the presence of anti-class II donor specific antibodies (anti-DRB5*02:02) as well as non-donor specific antibodies .

We asked why the recipient developed antibodies against antigens to which he was never exposed. In order to solve this problem we used the EpHLA software. A closer view of results in the EpHLA's Histocompatibility Map report showed that all HLA molecules to which the recipient developed antibodies share eplets with DRB5*02:02 from the immunizer. Interestingly, the DRB5*02:02 molecule has three potentially immunogenic eplets: 6C, 71QAA, 108T3. We noted that while 6C eplet appears only on DRB5*02:02 molecule, the 71QAA eplet is shared by molecules of serum group DR15 (DRB1*15:01, DRB1*15:02, DRB1*15:03) and the 108T3 eplet is shared by DRB5*01:01.

Thus, we were able to identify the epitopes targeted by the recipient's HLA antibodies using the EpHLA software, and the alleles DRB5*02:02, DRB5*01:01, DRB1*15:01, DRB1*15:02, DRB1*15:03 are the unacceptable mismatches for this case; they are associated to a 28% cPRA.

Caso 2.

A 35-year-old female in chronic hemodialysis, enrolled in the related renal transplantation program with two potential donors (brothers). The donors were typed as identical HLA each other and distinct as regards the recipient. The result of the T and B

CDC assays were positive to both donors. Four years later, the CDC assays performed with the same donors were negative and flow cytometry crossmatches were positive for T and B cells.

The SPA results using current serum showed preformed antibodies directed to a myriad of different class I and II HLA antigens (cPRA=91%). The possible immunization events were blood transfusions and three gestations from the same husband, whose HLA typing is shown in Table 2. A closer examination of the SPA results revealed:

(i) specific antibodies against the husband's HLA mismatches, including allele A*02:01 (MFI =8,994), and (ii) antibodies against potential donors' HLA antigens, including allele A*68:02 (MFI =12,353). Because A*02:01 and A*68:02 alleles belong to the same CREG, we reasoned that such alleles could share the same eplet targeted by the recipient's HLA antibodies.

We tested our hypothesis using the EpHLA software analyzing recipient versus immunizer, and then versus potential donors. We found that the recipient HLA antibodies recognize the pair of eplets 142MT +145KHA. Such eplets are shared by the HLA allele A*02:01 from the immunizer and allele A*68:02 from potential donors. As showed, the immunizer and potential donors share the same beta subunit in the HLA DQ molecule (DQB1*03:01), however combined to different alpha subunits. Such beta subunit presents an only potentially immunogenic eplet: 45EV. Nevertheless, as the MFI value of the HLA heterodimer DQA1*02:01–DQB1*03:01 of the potential donors is 921, the immunological risk is low for class II HLA. Contrariwise, we were able to detect a strong reactivity against A*68:02, representing a high immunological risk for antibody-mediated rejection.

As there is no agreement upon current CDC assay with the flow cytometry crossmatch and SPA results, we believe that the recipient has a mixture of antibodies with a prevalence of non-fixing complement isotypes, or the titles of the fixing complement antibodies present in the current serum were not enough to activate the classic pathway of the complement system. Thus, the potential donors' allele A*68:02 is considered one of the unacceptable mismatches for this recipient. As the calculated PRA was 91%, this case exemplifies the importance of using the Acceptable Mismatch approach.

Adalberto S. da Silva