Search

The recombinant, live, attenuated, tetravalent dengue vaccine CYD-TDV has shown efficacy against all four dengue serotypes. In this exploratory study (CYD59, NCT02827162), we evaluated potential associations of host human leukocyte antigen (HLA) alleles with dengue antibody responses, CYD-TDV vaccine efficacy, and virologically-confirmed dengue (VCD) cases. Children 4-11 years old, who previously completed a phase 2b efficacy study of CYD-TDV in a single center in Thailand, were included in the study. Genotyping of HLA class I and II loci was performed by next-generation sequencing from DNA obtained from 335 saliva samples. Dengue neutralizing antibody titers (NAb) were assessed as a correlate of risk and protection. Regression analyses were used to assess associations between HLA alleles and NAb responses, vaccine efficacy, and dengue outcomes. Month 13 NAb log geometric mean titers (GMTs) were associated with decreased risk of VCD. In the vaccine group, HLA-DRB1*11 was significantly associated with higher NAb log GMT levels (beta: 0.76; p = 0.002, q = 0.13). Additionally, in the absence of vaccination, HLA associations were observed between the presence of DPB1*03:01 and increased NAb log GMT levels (beta: 1.24; p = 0.005, q = 0.17), and between DPB1*05:01 and reduced NAb log GMT levels (beta: -1.1; p = 0.001, q = 0.07). This study suggests associations of HLA alleles with NAb titers in the context of dengue outcomes. This study was registered with clinicaltrials.gov: NCT02827162.

The human leukocyte antigen (HLA) genes are highly variable and are known to play an important role in disease outcomes, including infectious diseases. Prior knowledge of HLA polymorphisms in a population usually forms the basis for an effective case-control study design. As a prelude to future disease association analyses, we report HLA class I and II diversity in 334 unrelated donors from a Dengue vaccine efficacy trial conducted in Thailand. Long-range PCR amplification of six HLA loci was performed on DNA extracted from saliva samples. HLA-A, -B, -C, -DPB1, -DQB1 and -DRB1 were genotyped using a next-generation sequencing method presented at the 17th International HLA and Immunogenetics Workshop. In total, we identified 201 HLA alleles, including 35 HLA-A, 57 HLA-B, 28 HLA-C, 24 HLA-DPB1, 21 HLA-DQB1 and 36 HLA-DRB1 alleles. Very common HLA alleles with frequencies greater than 10 percent were A∗11:01:01, A∗33:03:01, A∗24:02:01, B∗46:01:01, C∗07:02:01, C∗01:02:01, C∗08:01:01, DPB1∗05:01:01, DPB1∗13:01:01, DPB1∗04:01:01, DPB1∗02:01:02, DQB1∗03:01:01, DQB1∗05:02:01, DQB1∗03:03:02, DRB1∗12:02:01, DRB1∗09:01:02, and DRB1∗15:02:01. A novel HLA allele, B∗15:450, had a non-synonymous substitution and occurred in more than one donor. Population-based full-length NGS HLA typing is more conclusive and provides a sound foundation for exploring disease association in a given population.

Aim A tetravalent, live attenuated dengue vaccine demonstrated efficacy, safety and immunogenicity in several clinical trials in Asia and Latin America. Efficacy differed based on infecting serotypes, presence of pre-existing dengue neutralizing antibody (NAb) titers and age. HLA class II molecules expressed on antigen presenting cells mediate CD4+ T cell stimulation of antibody production by B cells involved in vaccine-induced responses. We hypothesized that the differences in observed vaccine efficacy could be due to variation in NAb immune responses in conjunction with host HLA class II genes. Methods Samples were available from a subset of subjects that took part in the first tetravalent dengue vaccine efficacy trial conducted in Thailand. DNA was extracted from 335 saliva samples and HLA genotyping was performed using next-generation sequencing (NGS) of full-length genes. A panel of ancestry informative markers (AIMs) was genotyped to assess population stratification. Serotype-specific NAb titers were measured by plaque-reduction neutralization test 28 days after last injection. The association of NAb titers and HLA class II on dengue infection was tested by logistic regression. Linear regression was used to test association of HLA class II alleles with NAb levels after accounting for sex, age, and serotype as covariates. A minimal false discovery rate to account for multiple comparisons, with a two-sided p-value Results NGS identified 197 HLA class I and II alleles in the Thai Dengue vaccine trial. AIMs analysis did not identify population stratification comparing cases and controls. Magnitude of NAb levels post vaccination was significantly higher in the presence of HLA-DRB1*11 (p = 0.002, q = 0.08). HLA-DPB1*03:01 and *05:01 presence correlated with pre-existing NAb titers in the placebos (p = 0.005, q = 0.09; p = 0.001, q = 0.04). We did not observe a direct effect of HLA on dengue infection in either the placebo or treatment arm. Conclusions These findings suggest that specific HLA class II alleles modulate protective NAb titers in dengue infection. This is an exploratory study to identify signals to replicate in other dengue vaccine clinical studies. Understanding this HLA class II mechanism will enable improved vaccine design.

Aim RNA sequencing (RNA-Seq) is a powerful next-generation sequencing method for discovering, profiling and quantifying RNA transcripts. RNA-Seq can also identify splice junctions and single nucleotide polymorphisms (SNP) in coding exons. In this study, we tested if RNA-Seq reads provide sufficient coverage and depth for high-resolution HLA genotyping. Methods RNA-Seq data was available from 24 individuals from an acutely infected HIV cohort. We used two methods to extract HLA reads from the RNA-Seq data: 1) FR-HIT recruited RNA-Seq reads to the IMGT coding sequence reference database; and 2) the SNP-tolerant aligner GSNAP aligned the reads to a constructed GMAP database consisting of cDNA from HLA genes. The resulting HLA-specific FASTQ reads that were generated were analyzed by commercial HLA typing software. Results We obtained 100% sequence coverage and a median sequence read depth of 331, which was sufficient to genotype 15 HLA loci from each of the donors. HLA genotypes for 8 classical HLA class I and II loci matched genotypes generated from the Sanger sequence-based typing method, as well as targeted next-generation sequencing of genomic DNA. We identified a novel HLA-DQB1 allele in the Thai population with a frequency greater than 5% that was identical to the 05:01:01 allele in all exons except exon 1, where it differed by one SNP at position 81. The DQB1 * 05:01:01 allele, like several other DQB1 alleles, has exon 5 spliced out. This deletion however, cannot be confirmed from the reads that are obtained by sequencing genomic DNA. RNA-Seq confirmed that the novel allele does not express exon 5, as the RNA-Seq data showed a complete absence of exon 5 reads compared to reads from targeted genomic sequences. Conclusions We have successfully genotyped 15 HLA loci and confirmed a splice variant in a novel allele using RNA-Seq. This method highlights the importance of mining RNA-Seq data for high-resolution HLA genotyping. Additional analysis methods may be able to identify allele-specific expression of HLA alleles from RNA-Seq data.

Aim The Multi-locus Individual Tagging – Next Generation Sequencing (MIT-NGS) method was previously shown to efficiently genotype four HLA loci (A, B, C, and DRB1) from 96 donors while maintaining phase, and resolving ambiguities associated with traditional sequence based HLA typing (Ehrenberg et al., BMC Genomics, 2014). Here we expanded MIT-NGS to six loci to include DPB1 and DQB1, and evaluated HLA genotypes for all loci against those obtained from both the HLA sequence-based typing (SBT) method and the four loci MIT-NGS method. Methods Full-length genes of HLA-A, B, C, DPB1, DQB1 and DRB1 loci from 96 donors of Caucasian, African and Asian ethnicities were amplified using long-range PCR. Individual donor pools, each consisting of equimolar inputs corresponding to the target size of the six HLA loci amplicons were enzymatically fragmented and uniquely indexed. This 6-plex approach enabled sequencing of 576 loci from 96 individuals in a single Illumina MiSeq run. Locus specific sequence reads were demultiplexed, aligned, and genotyped using commercially available software and an in-house pipeline. Results There were no observed discordances between the 6-plex MIT-NGS method and the combined SBT or 4-plex NGS methods for all 1152 alleles analyzed in this study. Furthermore, this method was able to resolve 20 DQBI and 22 DPB1 genotypes that were ambiguous with SBT. These alleles included DPB1 ∗ 104:01,DPB1 ∗ 105:01, DPB1 ∗ 107:01, DPB1 ∗ 131:01, DPB1 ∗ 135:01, DPB1 ∗ 296:01, DPB1 ∗ 414:01, DQB1 ∗ 02:02, DQB1 ∗ 03:19 and DQB1 ∗ 06:01:03. All of these ambiguous changes affected the protein coding sequence. Conclusions We have developed a six-locus high resolution HLA typing method to accurately and efficiently genotype all classical HLA class I and II loci while maintaining phase, and resolving ambiguities associated with conventional sequence based HLA typing.