A major focus of the lab is determining novel characteristics of HLA class I and II that functionally distinguish the specific allotypes, beyond peptide binding preference alone. As far as we know, our lab is unique in this regard. This work requires consideration of all common HLA alleles/allotypes in order to be able to assign quantitative values to each individual in disease cohorts when determining whether the trait has an effect in the pathogenesis of different types of diseases, and whether the effects are opposing depending on disease type. HLA is loaded with peptide by a multiprotein complex within the ER. One key protein of this complex is tapasin, which mediates the peptide editing process. A limited set of HLA allotypes have been shown to exhibit differential dependence on tapasin to form a stable peptide complex. Tapasin dependence is likely to influence the peptide repertoire and HLA-peptide complex stability, and thus, the quality of immune responses. The exact molecular features of HLA class I alleles that define levels of tapasin dependence are unknown and there is no prediction tool that can characterize tapasin dependence based on the HLA class I structure. Just one amino acid difference can dramatically change tapasin dependency, as in the case of HLA-B4402 and HLA-B4405, which differ only at position 116. We experimentally determined tapasin dependence for all HLA class I alleles with frequencies greater than 0.5% in either white or black US populations, including 27 HLA-A, 42 HLA-B and 24 HLA-C alleles. We expressed each allotype with an N-terminal FLAG-tag in the tapasin negative lymphoblastoid cell line 721.220 using lentiviral constructs. In addition, the same set was expressed in 721.220 cells reconstituted with tapasin. Expression levels were measured by flow cytometry using anti-FLAG antibody. The ratio of expression levels in tapasin positive vs tapasin negative cells was defined as the tapasin dependency value. We are now able to assign a tapasin dependency score to any individual based on his/her HLA class I genotype. This score can be applied to various tests for association with human disease outcomes. Polymorphism within TAPBP and other genes involved in peptide processing/loading may influence antigen presentation by HLA class I molecules. We have found that variation in the UTRs of TAPBP marks differential expression levels of the gene as measured by qPCR in people of African ancestry, whereas these positions are fixed in Europeans. The 5'UTR SNP rs111686073C/G and the 3'UTR SNP rs73410010A/G were identified to significantly and independently associate with mRNA levels of TAPBP in two South African cohorts (p0.001 for each SNP). The 5'UTR SNP has been verified as a direct modulator of tapasin expression via luciferase assays in which the promoter region of tapasin was cloned into a luciferase expression vector. The G variant induced greater expression of luciferase relative to the C variant. Elecrophoretic mobility shift assays (EMSAs) using HeLa extract indicate that AP-2a, a transcription factor predicted to have a binding site overlapping the SNP, binds both the G and C variants of the 5'UTR SNP, but the G variant appears to bind AP-2a more strongly than does the C variant. We will study the effect of the 3'UTR SNP on expression via luciferase assays as well. We plan to test for genetic associations of these SNPs with disease outcome in various cohorts and any potential interactions these variants may have with tapasin dependence of HLA class I allotypes.
Showing the most recent 10 out of 20 publications
