21-OR

Abstract

Aim The clonal property of next generation sequencing systems allows for the quantitative analysis of mixed samples by simply counting sequences corresponding to the component alleles. Previously, we developed a method using 454 amplicon sequencing and Conexio software for high resolution and high throughput genotyping of the HLA class I and class II loci. This system was used to analyze the blood of a SCID child and estimate the proportion of maternal cells by counting HLA-C allelic sequence reads corresponding to the non-transmitted maternal allele. Here we report the development of a system that allows us to quantify HLA allelic mixtures in plasma. Methods DNA from plasma or contrived mixtures of cell lines was amplified using primers that targeted a short region (∼150 bp to amplify small DNA fragments in plasma) of HLA DPB1 or DQB1 exon 2. Amplicons were further amplified by emulsion PCR and sequenced on a 454 GS FLX or GS Junior. Sequences were examined using modified Conexio Assign ATF 454 software. This software allowed for rapid digital analysis of each DPB1 or DQB1 allele. We also analyzed mixtures using the Illumina MiSeq. Results Using the modified Conexio software, minority HLA alleles in a mixture were readily identified and separated from background “noise” i.e., sequences generated by PCR or sequencing errors. In mixtures of two heterozygous cell lines, the minority HLA alleles could be detected at 0.5% with 1 ng (∼140 diploid genomes) DNA input. In 3/3 plasma samples from kidney transplant recipients undergoing acute or chronic rejection, donor HLA alleles were detected at an average of 0.6% of the total of donor plus recipient alleles. Conclusions We have developed a system that allows the sensitive and precise analysis of mixtures using either 454 or Illumina MiSeq sequencing technology. This method could be useful in numerous clinical applications; we have demonstrated the detection of donor HLA alleles in the plasma of kidney transplant recipients undergoing rejection. Holcomb: Roche: Employee. Rastrou: Roche: Employee. Hoglund: Roche: Employee. Goodridge: Conexio Genomics: Employee. Erlich: Roche: Employee.