High-throughput multiplex HLA genotyping by next-generation sequencing using multi-locus individual tagging.

Philip K Ehrenberg,Nelson L Michael,Jerome H Kim,Karen M Baldwin,Rasmi Thomas,Merlin L Robb,Richard Apps,Aviva Geretz,Victoria R Polonis

doi:10.1186/1471-2164-15-864

Abstract

BackgroundUnambiguous human leukocyte antigen (HLA) typing is important in transplant matching and disease association studies. High-resolution HLA typing that is not restricted to the peptide-binding region can decrease HLA allele ambiguities. Cost and technology constraints have hampered high-throughput and efficient high resolution unambiguous HLA typing. We have developed a method for HLA genotyping that preserves the very high-resolution that can be obtained by next-generation sequencing (NGS) but also achieves substantially increased efficiency. Unambiguous HLA-A, B, C and DRB1 genotypes can be determined for 96 individuals in a single run of the Illumina MiSeq.ResultsLong-range amplification of full-length HLA genes from four loci was performed in separate polymerase chain reactions (PCR) using primers and PCR conditions that were optimized to reduce co-amplification of other HLA loci. Amplicons from the four HLA loci of each individual were then pooled and subjected to enzymatic library generation. All four loci of an individual were then tagged with one unique index combination. This multi-locus individual tagging (MIT) method combined with NGS enabled the four loci of 96 individuals to be analyzed in a single 500 cycle sequencing paired-end run of the Illumina-MiSeq. The MIT-NGS method generated sequence reads from the four loci were then discriminated using commercially available NGS HLA typing software. Comparison of the MIT-NGS with Sanger sequence-based HLA typing methods showed that all the ambiguities and discordances between the two methods were due to the accuracy of the MIT-NGS method.ConclusionsThe MIT-NGS method enabled accurate, robust and cost effective simultaneous analyses of four HLA loci per sample and produced 6 or 8-digit high-resolution unambiguous phased HLA typing data from 96 individuals in a single NGS run.Electronic supplementary materialThe online version of this article (doi:10.1186/1471-2164-15-864) contains supplementary material, which is available to authorized users.

Highlights

Unambiguous human leukocyte antigen (HLA) typing is important in transplant matching and disease association studies
Since multiple HLA loci genotypes are typically required from each individual we developed a multiplex individual tagging approach to enable HLA-A, B, C, and DRB1 to be simultaneously typed in each of the 96 homozygous or heterozygous individuals per sequencing run
Peripheral blood mononuclear cells (PBMCs) from 81 healthy donors of mainly Asian, Caucasian and African ancestry enrolled in the U.S Military HIV Research Program (MHRP) RV144 and RV229 study protocols were included in this study

Summary

Introduction

Unambiguous human leukocyte antigen (HLA) typing is important in transplant matching and disease association studies. The defining feature of HLA genes is their extreme polymorphism (http://www.ebi.ac.uk/ ipd/imgt/hla/stats.html). These genes encode surface expressed glycoproteins with a peptide-binding region (PBR) that presents peptides to autologous T lymphocytes and NK cells during the immune surveillance process. The excluded regions include the other exons, enhancer or promoter regions, introns, and the untranslated regions (UTR) Variation in these regions may be functionally important and can influence HLA transcription rate, gene expression, splice site variation, mRNA stability, and protein structural integrity [4,5,6]. Importantly, genotyping when limited to the sequences in PBR may give rise to HLA types that are not phase-defined or are ambiguous

Methods

Results

Discussion

Conclusion