Background/Purpose:

Human leukocyte antigen (HLA) genes code for proteins that allow the immune system to distinguish self from foreign cells. These genes have important indications in rheumatology, as well as other autoimmune disorders, transplantation, cancer, and pharmacotoxicity. The gold standards for HLA-typing are Sequence Specific Oligonucleotide Probe and Sequence Specific Primed PCR typing (PCR-SSOP and PCR-SSP), both of which are manually-intensive and costly. Multiplexing samples with next-generation sequencing (NGS) could easily reduce costs. Unfortunately, HLA genes are the most variable in the human genome and also have high sequence similarity, each of which pose considerable challenges to new technologies. However, nascent assays and bioinformatic methods are emerging in support of more cost-effective platforms for finding new biomarkers of inflammatory disease and for typing patients in the clinic. Here we apply new and traditional methods to investigate the HLA B27 allele, which has known links to a large family of inflammatory diseases.

Methods:

We obtained 90 samples from RA patients and compared the HLA allele calls on three platforms: 1) GenDx’s NGSengine for the NGSGo kit for Illumina’s MiSeq, 2) SNP2HLA [1] for both the Affymetrix Axiom and the Illumina Omni 5 microarrays, and 3) the Luminex PCR process. We stratified the sample populations by principal components to identify 47 samples of European ancestry to call alleles using array genotypes against a European reference panel. We measured HLA B27-positive status in those 47 samples to check for enrichment in both the NGS-based and the array-based calls.

Results:

Alleles were called and compared to the Luminex calls for two-digit and four-digit precision, and with the exception of DRB1, NGS-based allele calls were 97-100% concordant. However, two-digit precision suffices for testing B27 status, which will be positive for all 4-digit precision alleles ranging from HLA-B*27:01-HLA-B*27:117. The microarray approach measured 19% of the RA subjects as B27-positive, and the NGS technique measured 12%. This is enriched, compared with 3.65 (+/- 1.7)% observed in typical European populations [2]. The less accurate array-based technique found the association and the NGS technique confirmed it. Notably, the NGS-based technique can assay the same number of loci as the high-throughput gold standard option but at close to half the cost and with fewer ambiguities. Some NGS-based options have already been submitted for CLIA-validated clinical use.  Alternatively, the array-based techniques are offered at a cost that is an order of magnitude lower than the gold standard and can be used on legacy datasets, presenting a lower cost option for association discovery.

Conclusion:

Existing, legacy microarray data can guide exploration and hypotheses at reduced costs, and should be confirmed with gold standard or CLIA-validated NGS technologies, the latter of which poses a viable candidate for reducing costs, time and effort in a clinical setting.

References:

1. Xiaoming et al. PLoS One 2013; 8(6):e64683
2. Gonzalez-Galarza et al. NAR 2015; 28, D784-8

« Back to 2015 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/new-technologies-for-typing-self-identifying-antigens-encoded-by-hla-genes-provide-cost-effective-alternatives-to-identifying-b27-allele-status-in-rheumatoid-arthritis-subjects/