Voltar
Artigo Acesso aberto Revisado por pares
BIBTEX RIS

Using Nanopore Whole-Transcriptome Sequencing for Human Leukocyte Antigen Genotyping and Correlating Donor Human Leukocyte Antigen Expression with Flow Cytometric Crossmatch Results

2019; Elsevier BV; Volume: 22; Issue: 1 Linguagem: Inglês

10.1016/j.jmoldx.2019.09.005

ISSN

1943-7811

Autores

Maureen C. Montgomery, Chang Liu, Rosanne Petraroia, Eric T. Weimer,

Tópico(s)

Cytomegalovirus and herpesvirus research

Resumo

Transplant centers are increasingly using virtual crossmatching to evaluate recipient and donor compatibility. However, the current state of virtual crossmatching fails to incorporate donor human leukocyte antigen (HLA) expression in the assessment, despite numerous studies that have demonstrated the impact of donor HLA expression on physical crossmatch outcomes. Whole-transcriptome sequencing (RNA-Seq) for HLA enables simultaneous determination of HLA genotyping and relative HLA expression. Ultimately the RNA-Seq needs to be faster to be incorporated into the virtual crossmatching process. However, to demonstrate feasibility, the utility of the MinION sequencer (Oxford Nanopore Technologies, Oxford, UK) was evaluated in combination with RNA-Seq to generate HLA genotypes and to determine HLA class I expression. Although HLA class I expression varied among individuals, the pattern of HLA expression remained relatively consistent (HLA-B > HLA-A = HLA-C). HLA-A and -C had similar expression profiles. The impact of donor HLA expression was evaluated using serum samples containing a single donor-specific antibody (DSA). By making DSA consistent, donor HLA expression variability could be assessed. With consistent DSA mean fluorescence intensity, there was a direct relationship between the donor HLA expression to which the DSA is against and flow cytometric crossmatch median channel shifts. Transplant centers are increasingly using virtual crossmatching to evaluate recipient and donor compatibility. However, the current state of virtual crossmatching fails to incorporate donor human leukocyte antigen (HLA) expression in the assessment, despite numerous studies that have demonstrated the impact of donor HLA expression on physical crossmatch outcomes. Whole-transcriptome sequencing (RNA-Seq) for HLA enables simultaneous determination of HLA genotyping and relative HLA expression. Ultimately the RNA-Seq needs to be faster to be incorporated into the virtual crossmatching process. However, to demonstrate feasibility, the utility of the MinION sequencer (Oxford Nanopore Technologies, Oxford, UK) was evaluated in combination with RNA-Seq to generate HLA genotypes and to determine HLA class I expression. Although HLA class I expression varied among individuals, the pattern of HLA expression remained relatively consistent (HLA-B > HLA-A = HLA-C). HLA-A and -C had similar expression profiles. The impact of donor HLA expression was evaluated using serum samples containing a single donor-specific antibody (DSA). By making DSA consistent, donor HLA expression variability could be assessed. With consistent DSA mean fluorescence intensity, there was a direct relationship between the donor HLA expression to which the DSA is against and flow cytometric crossmatch median channel shifts. The human leukocyte antigen (HLA) region has been studied extensively due to its association with various diseases and its importance in transplant compatibility. Within transplantation, HLA is a significant source of allograft rejection, and HLA compatibility is routinely assessed before kidney allograft allocation. The HLA class I genes HLA-A, HLA-B, and HLA-C are expressed on all nucleated cells. Most of the studies to date have focused on the substantial allelic variation in the HLA genes for hematopoietic cell transplantation. However, few studies have focused on HLA expression in solid organ transplantation, in which potential variations in HLA expression may affect initial recipient and donor compatibility. In solid organ transplantation, HLA expression influences crossmatch sensitivity and outcomes.1Badders J.L. Jones J.A. Jeresano M.E. Schillinger K.P. Jackson A.M. Variable HLA expression on deceased donor lymphocytes: not all crossmatches are created equal.Hum Immunol. 2015; 76: 795-800Crossref PubMed Scopus (30) Google Scholar Currently, donor HLA expression is determined using only flow cytometry as part of crossmatch quality assurance. However, determining donor HLA expression before initiating crossmatch provides additional data that may enhance virtual crossmatch assessments. Donor HLA genotyping is required before the initiation of a United Network for Organ Sharing match run and organ donation. Currently, all clinical HLA genotypes are determined using DNA. However, DNA cannot estimate expression. In contrast, RNA can provide both HLA genotype and relative HLA expression. In addition, using RNA enables the detection of splice site variants and null transcripts.2Gerritsen K.E. Voorter C.E. Joannes M.O. Senden-Gijsbers B.L. Agis F. Wieten L. Tilanus M.G. RNA and protein expression of HLA-A*23:19Q.Hum Immunol. 2015; 76: 286-291Crossref PubMed Scopus (9) Google Scholar, 3Balas A. Garcia-Sanchez F. Vicario J.L. RNA processing and protein expression of HLA-B*07:44N.HLA. 2017; 89: 230-234Crossref PubMed Scopus (5) Google Scholar, 4Reinders J. Rozemuller E.H. Otten H.G. Houben A.J. Dormoy A. Mulder A. van den Tweel J.G. Petersen E.J. Tilanus M.G. Identification of HLA-A*0111N: a synonymous substitution, introducing an alternative splice site in exon 3, silenced the expression of an HLA-A allele.Hum Immunol. 2005; 66: 912-920Crossref PubMed Scopus (25) Google Scholar, 5Laforet M. Froelich N. Parissiadis A. Bausinger H. Pfeiffer B. Tongio M.M. An intronic mutation responsible for a low level of expression of an HLA-A*24 allele.Tissue Antigens. 1997; 50: 340-346Crossref PubMed Scopus (52) Google Scholar, 6Mine H. Ishikawa Y. Hojo M. Imashuku S. Tokunaga K. Juji T. Saji H. Yokoyama S. An HLA-B null allele (B*1526N) with a stop codon in exon 3 generated by a point mutation.Tissue Antigens. 1997; 50: 351-354Crossref PubMed Scopus (42) Google Scholar Numerous reports have indicated accurate HLA genotyping from whole-transcriptome sequencing (RNA-Seq) with the use of Illumina technology (San Diego, CA).7Kim H.J. Pourmand N. HLA typing from RNA-seq data using hierarchical read weighting [corrected].PLoS One. 2013; 8: e67885Crossref PubMed Scopus (37) Google Scholar, 8Boegel S. Lower M. Schafer M. Bukur T. de Graaf J. Boisguerin V. Tureci O. Diken M. Castle J.C. Sahin U. HLA typing from RNA-seq sequence reads.Genome Med. 2012; 4: 102Crossref PubMed Scopus (149) Google Scholar, 9Boegel S. Löwer M. Bukur T. Sorn P. Castle J.C. Sahin U. HLA and proteasome expression body map.BMC Med Genomics. 2018; 11: 36Crossref PubMed Scopus (59) Google Scholar, 10Buchkovich M.L. Brown C.C. Robasky K. Chai S. Westfall S. Vincent B.G. Weimer E.T. Powers J.G. HLAProfiler utilizes k-mer profiles to improve HLA calling accuracy for rare and common alleles in RNA-seq data.Genome Med. 2017; 9: 86Crossref PubMed Scopus (24) Google Scholar To our knowledge, no studies have evaluated HLA expression from RNA-Seq using nanopore technology. A combination of RNA-Seq and nanopore technology may enable rapid HLA genotyping and the determination of expression. HLA expression has been measured by a variety of technologies, including real-time quantitative PCR11Cajander S. Tina E. Backman A. Magnuson A. Stralin K. Soderquist B. Kallman J. Quantitative real-time polymerase chain reaction measurement of HLA-DRA gene expression in whole blood is highly reproducible and shows changes that reflect dynamic shifts in monocyte surface HLA-DR expression during the course of sepsis.PLoS One. 2016; 11: e0154690Crossref PubMed Scopus (20) Google Scholar, 12McCutcheon J.A. Gumperz J. Smith K.D. Lutz C.T. Parham P. Low HLA-C expression at cell surfaces correlates with increased turnover of heavy chain mRNA.J Exp Med. 1995; 181: 2085-2095Crossref PubMed Scopus (158) Google Scholar, 13Mine K.L. Tedesco-Silva H. Mourão T.B. Campos E.F. Salzedas L.A. Aguiar B. Felipe C.R. Medina-Pestana J.O. Gerbase-DeLima M. Heightened expression of HLA-DQB1 and HLA-DQB2 in pre-implantation biopsies predicts poor late kidney graft function.Hum Immunol. 2018; 79: 594-601Crossref PubMed Scopus (10) Google Scholar, 14Petersdorf E.W. Malkki M. O'HUigin C. Carrington M. Gooley T. Haagenson M.D. Horowitz M.M. Spellman S.R. Wang T. Stevenson P. High HLA-DP expression and graft-versus-host disease.N Engl J Med. 2015; 373: 599-609Crossref PubMed Scopus (189) Google Scholar and RNA-Seq.7Kim H.J. Pourmand N. HLA typing from RNA-seq data using hierarchical read weighting [corrected].PLoS One. 2013; 8: e67885Crossref PubMed Scopus (37) Google Scholar, 8Boegel S. Lower M. Schafer M. Bukur T. de Graaf J. Boisguerin V. Tureci O. Diken M. Castle J.C. Sahin U. HLA typing from RNA-seq sequence reads.Genome Med. 2012; 4: 102Crossref PubMed Scopus (149) Google Scholar, 9Boegel S. Löwer M. Bukur T. Sorn P. Castle J.C. Sahin U. HLA and proteasome expression body map.BMC Med Genomics. 2018; 11: 36Crossref PubMed Scopus (59) Google Scholar, 10Buchkovich M.L. Brown C.C. Robasky K. Chai S. Westfall S. Vincent B.G. Weimer E.T. Powers J.G. HLAProfiler utilizes k-mer profiles to improve HLA calling accuracy for rare and common alleles in RNA-seq data.Genome Med. 2017; 9: 86Crossref PubMed Scopus (24) Google Scholar,15Gerritsen K.E. Olieslagers T.I. Groeneweg M. Voorter C.E. Tilanus M.G. An improved and validated RNA HLA class I SBT approach for obtaining full length coding sequences.Tissue Antigens. 2014; 84: 450-458Crossref PubMed Scopus (6) Google Scholar, 16Greene J.M. Wiseman R.W. Lank S.M. Bimber B.N. Karl J.A. Burwitz B.J. Lhost J.J. Hawkins O.E. Kunstman K.J. Broman K.W. Wolinsky S.M. Hildebrand W.H. O'Connor D.H. Differential MHC class I expression in distinct leukocyte subsets.BMC Immunol. 2011; 12: 39Crossref PubMed Scopus (32) Google Scholar, 17Lang K. Surendranath V. Quenzel P. Schofl G. Schmidt A.H. Lange V. Full-length HLA class I genotyping with the MinION nanopore sequencer.Methods Mol Biol. 2018; 1802: 155-162Crossref PubMed Scopus (15) Google Scholar, 18Liu C. Yang X. Duffy B. Mohanakumar T. Mitra R.D. Zody M.C. Pfeifer J.D. ATHLATES: accurate typing of human leukocyte antigen through exome sequencing.Nucleic Acids Res. 2013; 41: e142Crossref PubMed Scopus (82) Google Scholar, 19Szolek A. Schubert B. Mohr C. Sturm M. Feldhahn M. Kohlbacher O. OptiType: precision HLA typing from next-generation sequencing data.Bioinformatics. 2014; 30: 3310-3316Crossref PubMed Scopus (352) Google Scholar Although real-time quantitative PCR is faster and less expensive compared with RNA-Seq, the resolution of HLA genotyping is lower compared with the high-resolution HLA genotyping generated by RNA-Seq.7Kim H.J. Pourmand N. HLA typing from RNA-seq data using hierarchical read weighting [corrected].PLoS One. 2013; 8: e67885Crossref PubMed Scopus (37) Google Scholar,8Boegel S. Lower M. Schafer M. Bukur T. de Graaf J. Boisguerin V. Tureci O. Diken M. Castle J.C. Sahin U. HLA typing from RNA-seq sequence reads.Genome Med. 2012; 4: 102Crossref PubMed Scopus (149) Google Scholar,10Buchkovich M.L. Brown C.C. Robasky K. Chai S. Westfall S. Vincent B.G. Weimer E.T. Powers J.G. HLAProfiler utilizes k-mer profiles to improve HLA calling accuracy for rare and common alleles in RNA-seq data.Genome Med. 2017; 9: 86Crossref PubMed Scopus (24) Google Scholar,17Lang K. Surendranath V. Quenzel P. Schofl G. Schmidt A.H. Lange V. Full-length HLA class I genotyping with the MinION nanopore sequencer.Methods Mol Biol. 2018; 1802: 155-162Crossref PubMed Scopus (15) Google Scholar,18Liu C. Yang X. Duffy B. Mohanakumar T. Mitra R.D. Zody M.C. Pfeifer J.D. ATHLATES: accurate typing of human leukocyte antigen through exome sequencing.Nucleic Acids Res. 2013; 41: e142Crossref PubMed Scopus (82) Google Scholar,20Lank S.M. Wiseman R.W. Dudley D.M. O'Connor D.H. A novel single cDNA amplicon pyrosequencing method for high-throughput, cost-effective sequence-based HLA class I genotyping.Hum Immunol. 2010; 71: 1011-1017Crossref PubMed Scopus (42) Google Scholar, 21Ton K.N.T. Cree S.L. Gronert-Sum S.J. Merriman T.R. Stamp L.K. Kennedy M.A. Multiplexed nanopore sequencing of HLA-B locus in Maori and Pacific Island samples.Front Genet. 2018; 9: 152Crossref PubMed Scopus (11) Google Scholar, 22Laver T.W. Caswell R.C. Moore K.A. Poschmann J. Johnson M.B. Owens M.M. Ellard S. Paszkiewicz K.H. Weedon M.N. Pitfalls of haplotype phasing from amplicon-based long-read sequencing.Sci Rep. 2016; 6: 21746Crossref PubMed Scopus (44) Google Scholar Of importance, several reports have demonstrated a strong correlation between mRNA and cDNA sequencing–derived expression levels with the cell-surface expression for HLA molecules.23Kulkarni S. Ramsuran V. Rucevic M. Singh S. Lied A. Kulkarni V. O'HUigin C. Le Gall S. Carrington M. Posttranscriptional regulation of HLA-A protein expression by alternative polyadenylation signals involving the RNA-binding protein syncrip.J Immunol. 2017; 199: 3892-3899Crossref PubMed Scopus (18) Google Scholar, 24Kulkarni S. Savan R. Qi Y. Gao X. Yuki Y. Bass S.E. Martin M.P. Hunt P. Deeks S.G. Telenti A. Pereyra F. Goldstein D. Wolinsky S. Walker B. Young H.A. Carrington M. Differential microRNA regulation of HLA-C expression and its association with HIV control.Nature. 2011; 472: 495-498Crossref PubMed Scopus (277) Google Scholar, 25Kaur G. Gras S. Mobbs J.I. Vivian J.P. Cortes A. Barber T. Kuttikkatte S.B. Jensen L.T. Attfield K.E. Dendrou C.A. Carrington M. McVean G. Purcell A.W. Rossjohn J. Fugger L. Structural and regulatory diversity shape HLA-C protein expression levels.Nat Commun. 2017; 8: 15924Crossref PubMed Scopus (66) Google Scholar, 26Revenfeld A.L. Steffensen R. Pugholm L.H. Jorgensen M.M. Stensballe A. Varming K. Presence of HLA-DR molecules and HLA-DRB1 mRNA in circulating CD4(+) T cells.Scand J Immunol. 2016; 84: 211-221Crossref PubMed Scopus (10) Google Scholar, 27Hauber I. Gulle H. Wolf H.M. Maris M. Eggenbauer H. Eibl M.M. Molecular characterization of major histocompatibility complex class II gene expression and demonstration of antigen-specific T cell response indicate a new phenotype in class II-deficient patients.J Exp Med. 1995; 181: 1411-1423Crossref PubMed Scopus (38) Google Scholar, 28Rene C. Lozano C. Villalba M. Eliaou J.F. 5' and 3' untranslated regions contribute to the differential expression of specific HLA-A alleles.Eur J Immunol. 2015; 45: 3454-3463Crossref PubMed Scopus (22) Google Scholar, 29Dellgren C. Nehlin J.O. Barington T. Cell surface expression level variation between two common Human Leukocyte Antigen alleles, HLA-A2 and HLA-B8, is dependent on the structure of the C terminal part of the alpha 2 and the alpha 3 domains.PLoS One. 2015; 10: e0135385Crossref PubMed Scopus (12) Google Scholar In this study, we evaluated the utility of RNA-Seq using nanopore technology in HLA genotyping and in the determination of HLA expression. RNA-Seq yielded different data output but with enough HLA-specific reads to generate accurate HLA genotyping. Relative HLA expression varied significantly among healthy individuals, but with a consistent expression pattern of HLA-B > HLA-A = HLA-C. The importance of HLA expression and particularly locus-specific expression in flow cytometric crossmatches (FCXM) was determined. Peripheral blood was drawn into acid citrate dextrose tubes, and total lymphocytes were isolated using the EasySep Direct Human Total Lymphocyte isolation kit (StemCell Technologies, Vancouver, BC, Canada). Briefly, acid citrate dextrose tubes were spun for 10 minutes at 2113 × g and the buffy coat was removed. Equal parts of EasySep isolation cocktail and RapidSpheres (Stemcell Technologies, Vancouver, BC, Canada) were added to the buffy coat and incubated at room temperature for 5 minutes. EasySep buffer was added, and the sample was placed on a magnet for 10 minutes. The supernatant was removed and retained, as the lymphocytes were negatively selected to reduce gene up-regulation during isolation and were in the suspension. The same volume of RapidSpheres was added to the suspension a second time and mixed well. The sample was incubated at room temperature for 5 minutes, followed by 5 minutes on the magnet. Supernatant was removed and retained again, once more placed on the magnet to remove any residual RapidSpheres. Lymphocytes were expected to have been pure and ready for either mRNA isolation or flow crossmatching. Up to 1 × 107 lymphocytes were pelleted, and mRNA was isolated with the GenElute Direct mRNA MiniPrep Kit (Sigma-Aldrich, St. Louis, MO). Cells were lysed and then bound to oligo (dT) beads to isolate mRNA. The suspension was spun down to pellet the beads, and a series of washes was performed through a filtration column. mRNA was eluted off of the oligo (dT) beads, and the mRNA was either used or stored at −70°C. Isolated mRNA was quantified using a QuBit 2.0 fluorometer RNA HS Assay (Thermo Fisher Scientific, Carlsbad, CA). To increase the downstream MinION sequencing output, one sample (cDNA014) was further purified using the ReliaPrep RNA clean-up and concentration system (Promega, Madison, WI). Purification was done by combining mRNA, membrane binding solution, and 100% isopropanol. The sample was run through a filtration column and eluted with nuclease-free water. Quantification with the QuBit RNA HS Assay was performed. The SQK-PCS108 cDNA-PCR Sequencing Kit (Oxford Nanopore Technologies, Oxford, UK) was used according to the manufacturer's instructions. Briefly, mRNA was denatured at 65°C for 5 minutes. SuperScript IV Reverse Transcriptase (Thermo Fisher Scientific) and buffer were used in reverse transcription, which included a strand-switching component to create double-stranded cDNA. A short PCR was performed to add on barcodes and rapid-attachment primers. Rapid-sequencing adapters were added with a short incubation at room temperature. Samples were quantified with the QuBit dsDNA HS kit before loading on MinION. Sequencing was performed on MinION (Oxford Nanopore Technologies), using standard flow cells FLO-MIN106D R9.4.1 and MinKNOW software version 3.1.13. Samples were sequenced for a mean duration of 18 hours (range, 15.5 to 22 hours). Fast5 files from the MinION software were used for downstream analysis. Base-calling and demultiplexing were performed on all samples, except cDNA014, with Albacore software version 2.3.4 (Oxford Nanopore Technologies). Because of the timing of the experiments, Albacore was no longer available for cDNA014; thus it was base-called and demultiplexed with Guppy software version 2.3.1 (Oxford Nanopore Technologies). Fast5 files were processed for quality and binned into folders indicating pass or fail. The passed files were base-called, and if a sample was barcoded, it was demultiplexed according to the barcodes, and an output fastq file was created. Adapter trimming was performed using Porechop software version 0.2.3 (https://github.com/rrwick/Porechop, last accessed January 5, 2019), with sequences of middle adapters discarded. Nanopore sequencing reads were mapped to reference sequences of exons 2 and 3 of the HLA-A, HLA-B, and HLA-C genes from IMGT software version 3.26.0 using minimap2.30Li H. Minimap2: pairwise alignment for nucleotide sequences.Bioinformatics. 2018; 34: 3094-3100Crossref PubMed Scopus (3053) Google Scholar Locus-specific reads were extracted and subsequently analyzed using the Athlon version 1.0 pipeline to determine the genotype at the key-exon level, as reported previously.31Liu C. Xiao F. Hoisington-Lopez J. Lang K. Quenzel P. Duffy B. Mitra R.D. Accurate typing of human leukocyte antigen class I genes by Oxford nanopore sequencing.J Mol Diagn. 2018; 20: 428-435Abstract Full Text Full Text PDF PubMed Scopus (20) Google Scholar Postprocessed fastq files were analyzed in CLC Genomics Workbench software version 12.0 (Qiagen Bioinformatics, Redwood City, CA) using the standard RNA-Seq analysis workflow. The reference sequence was hg38, and genes were annotated with transcripts and aligned using standard settings. HLA expression data were normalized between donors using the RNA-Seq analysis, in transcripts per million (TPM). FCXM was performed according to the Halifax protocol.32Liwski R.S. Greenshields A.L. Conrad D.M. Murphey C. Bray R.A. Neumann J. Gebel H.M. Rapid optimized flow cytometric crossmatch (FCXM) assays: the Halifax and Halifaster protocols.Hum Immunol. 2018; 79: 28-38Crossref PubMed Scopus (23) Google Scholar Briefly, pronase-treated lymphocytes at a concentration of 0.25 × 106 were combined with 25 μL of patient serum or controls in a well of a 96-well U-bottom plate and incubated at room temperature for 20 minutes. Cells were pelleted by centrifugation at 1352 × g for 1 minute, and a series of three washes was performed. Cells were then incubated with fluorescein isothiocyanate, CD3 PerCP, CD19 PE, and viability allophycocyanin stains at room temperature for 10 minutes in the dark. Cells were pelleted, washed two more times, and resuspended in 1% paraformaldehyde. Samples were acquired on FACCanto II using the high-throughput sampler (BD Biosciences, Franklin Lakes, NJ) and FACSDiva software version 8.0.2. One-way analysis of variance using Geisser-Greenhouse correction, followed by the Tukey multiple-comparisons test, was performed using Prism software version 8.2.0 (GraphPad, San Diego, CA). Correlation values were determined using linear regression analysis. P values of ≤0.05 were considered significant. Peripheral blood mononuclear cells were collected from 12 healthy donors, and total lymphocytes were negatively selected and used for mRNA extraction. Sequencing libraries were prepared using standard nanopore protocols. Standard MinION flow cells were used and sequenced for a mean duration of 18 hours (Figure 1). Since total lymphocytes were used for mRNA isolation, only HLA class I loci were evaluated. HLA class II loci were underrepresented in total lymphocytes due to the abundance of T cells (Supplemental Figure S1). The absolute number of reads achieved from each sequencing run is shown in Figure 2A. The range of reads was 510,733 to 3,761,507 (Figure 2A), with a mean ± SEM of 84.7% ± 1.6% of reads mapping to hg38 (Figure 2B). To determine whether RNA-Seq using nanopore technology would support HLA class I genotyping, the absolute number of mapped reads specific for HLA was determined. HLA-B had the most mapped reads, with a mean ± SEM of 2217 ± 429, followed by HLA-A, with 1432 ± 246, then HLA-C, with 1328 ± 280 (Figure 2C). No correlation was observed between the numbers of RNA-Seq reads and HLA locus–specific reads (Figure 2D).Figure 2Overall nanopore whole-transcriptome sequencing (RNA-Seq) data and human leukocyte antigen (HLA) class I–specific reads. A: Total number of reads achieved per sample after MinION sequencing (Oxford Nanopore Technologies, Oxford, UK). B: Percentage of the total reads that mapped to hg38 transcripts using CLC Genomics Workbench (Qiagen Bioinformatics, Redwood City, CA). C: Individual samples for HLA class I specific read count. D: HLA locus–specific reads versus total RNA-Seq reads. Each point represents an individual donor.View Large Image Figure ViewerDownload Hi-res image Download (PPT) Given the number of HLA-specific reads achieved in Figure 2, using nanopore RNA-Seq to generate HLA genotypes may be possible. To evaluate HLA genotyping, nanopore RNA-Seq data were compared with those from existing high-resolution HLA class I genotypes. The reference method was DNA-based next-generation sequencing using an Illumina platform.33Weimer E.T. Montgomery M. Petraroia R. Crawford J. Schmitz J.L. Performance characteristics and validation of next-generation sequencing for human leucocyte antigen typing.J Mol Diagn. 2016; 18: 668-675Abstract Full Text Full Text PDF PubMed Scopus (53) Google Scholar,34Montgomery M.C. Petraroia R. Weimer E.T. Buccal swab genomic DNA fragmentation predicts likelihood of successful HLA genotyping by next-generation sequencing.Hum Immunol. 2017; 78: 634-641Crossref PubMed Scopus (9) Google Scholar Previously, Athlon has been successfully used with DNA amplicons and nanopore data for the genotyping of HLA class I genes accurately.31Liu C. Xiao F. Hoisington-Lopez J. Lang K. Quenzel P. Duffy B. Mitra R.D. Accurate typing of human leukocyte antigen class I genes by Oxford nanopore sequencing.J Mol Diagn. 2018; 20: 428-435Abstract Full Text Full Text PDF PubMed Scopus (20) Google Scholar HLA class II genotyping analysis is being developed. The nanopore RNA-Seq data were blindly analyzed (C.L.) using Athlon for HLA class I genotyping. Using a maximum of 200 random HLA-specific reads, 24 of 24 (100%) HLA-A and HLA-B genotypes were correctly determined to two-fields (Table 1). However, only 20 of 24 HLA-C genotypes (83.3%) were correctly determined to two-fields. Of interest, all of the missed genotypes were HLA-C*07, which has been established as the lowest-expressed HLA-C allele group.35Apps R. Meng Z. Del Prete G.Q. Lifson J.D. Zhou M. Carrington M. Relative expression levels of the HLA class-I proteins in normal and HIV-infected cells.J Immunol. 2015; 194: 3594-3600Crossref PubMed Scopus (104) Google Scholar,36Apps R. Qi Y. Carlson J. Chen H. Gao X. Thomas R. Yuki Y. Del Prete G. Goulder P. Brumme Z. Brumme C. John M. Mallal S. Nelson G. Bosch R. Heckerman D. Stein J. A Soderberg K. Anthony Moody M. Carrington M. Influence of HLA-C expression level on HIV control.Science. 2013; 340: 87-91Crossref PubMed Scopus (292) Google ScholarTable 1Accuracy of RNA-Seq HLA Class I Genotyping Compared with Reference HLA Class I GenotypingHLA-A*HLA-A*HLA-B*HLA-B*HLA-C*HLA-C*Sample IDReferenceRNA-SeqReferenceRNA-SeqReferenceRNA-SeqcDNA102:01:0102:01:01G35:01:0135:01:01 G03:04:0103:04:01 G11:01:0111:01:5140:01:0240:01:01 G04:01:0104:01:67cDNA202:01:0102:01:01G27:05:0227:05:0601:02:0101:02:01 G11:01:0111:01:2955:01:0155:01:01 G03:03:0103:03:01 GcDNA302:01:0102:01:01G35:01:0135:01:01 G04:01:0104:01:1123:01:0123:01:01G49:01:0149:01:01 G07:01:0107:01:16cDNA502:01:0102:01:01G07:02:0107:02:01 G01:02:0101:02:0503:01:0103:01:2315:01:0115:01:0307:01:0107:02:30cDNA601:01:0101:01:01G07:02:0107:02:01 G07:01:0107:01:01 G02:01:0102:01:01G08:01:0108:01:01 G07:02:0103:296:XXcDNA802:01:0102:01:01G07:02:0107:02:0506:02:0106:02:01 G24:02:0124:02:4157:01:0157:01:01 G07:02:01XcDNA902:01:0102:01:01G08:01:0108:01:01 G04:01:0104:01:6703:01:0103:01:0235:01:0135:01:01 G07:01:0107:01:32cDNA1003:0103:01:01G44:0244:02:01 G05:0105:01:01 G11:0111:01:01G51:0151:01:01 G15:0215:02:01 GcDNA1102:01:0102:01:0244:2944:29:XX05:01:0105:01:01 G24:02:0124:02:01G51:01:0151:01:01 G15:02:0115:02:01 GcDNA1201:01:0101:01:1814:02:0114:02:01 G07:01:0205:01:0702:01:0102:01:01 G15:17:0115:17:01 G08:02:0108:02:01 GcDNA1302:01:0102:01:01 G14:01:0114:01:0103:04:0103:04:3830:04:0130:04:01 G40:01:0240:01:01 G08:02:0108:02:01 GcDNA1403:01:0103:01:01 G07:02:0107:02:01 G07:02:0107:02:0203:01:0103:01:01 G07:02:0107:02:01 G07:02:0103:296:XXInaccurate HLA genotyping from RNA sequencing (-seq) data is noted in bold. Open table in a new tab Inaccurate HLA genotyping from RNA sequencing (-seq) data is noted in bold. The second function of RNA use in the HLA laboratory is the determination of relative HLA expression. To determine relative HLA expression, RNA-Seq data were mapped against genes in hg38 (as in Materials and Methods), and the expression of HLA loci was calculated using TPM for normalization across experiments since the objective was comparing within genes. (TPM normalizes for the differences in the composition of the transcripts in the denominator, rather than dividing by the number of reads in the library, and is considered more comparable between samples of different origins and composition.37Conesa A. Madrigal P. Tarazona S. Gomez-Cabrero D. Cervera A. McPherson A. Szcześniak M.W. Gaffney D.J. Elo L.L. Zhang X. Mortazavi A. A survey of best practices for RNA-seq data analysis.Genome Biol. 2016; 17: 13Crossref PubMed Scopus (1328) Google Scholar) In 11 of 12 individuals (91.6%), HLA-B expression was higher than those of HLA-A and HLA-C (Figure 3A). In 7 of 12 cases (58.3%), HLA-A expression was higher than that of HLA-C. There were samples in which this pattern was not observed. For example, cDNA5 had higher HLA-A expression compared with HLA-B, and cDNA9 had higher HLA-C compared with HLA-A. Across all samples, the mean ± SEM HLA-A, -B, and -C expression values were 1052 ± 83 TPM, 1440 ±160 TPM, and 1017 ± 133 TPM, respectively. HLA-B expression was significantly higher compared with those of HLA-A and -C (P = 0.0064 and P = 0.0003, respectively). However, the expression of HLA-A and -C were similar (P = 0.9413). To assess the reproducibility of HLA expression and to demonstrate how HLA expression varies over time, three individual HLA expression values were measured at four different time points separated by at least 2 days. For cDNA3, mean ± SEM HLA-A, -B, and -C expression values were 1060 ± 99 TPM, 1185 ± 114 TPM, and 896 ± 89 TPM, respectively. The day-to-day variations in HLA-A, -B, and -C were 18.7%, 19.3%, and 20.0%, respectively (Figure 3B). Sample cDNA6 had mean HLA-A, -B, -C HLA class I expression values of 945 TPM, 1108 TPM, and 659 TPM, respectively. In addition, cDNA6 had greater variations among samples compared with cDNA3 or cDNA9; cDNA6 varied from 18.5% (HLA-A) to 41.7% (HLA-C). Sample cDNA9 was in the minority of samples in which HLA-C expression (mean, 1184 TPM) was greater than HLA-A expression (mean, 1159 TPM). Previous studies have demonstrated that surface HLA expression was correlated with FCXM outcomes.1Badders J.L. Jones J.A. Jeresano M.E. Schillinger K.P. Jackson A.M. Variable HLA expression on deceased donor lymphocytes: not all crossmatches are created equal.Hum Immunol. 2015; 76: 795-800Crossref PubMed Scopus (30) Google Scholar However, this is done at the HLA class level. Since

Referência(s)