Human Immunology 75 (2014) 1239–1243

Contents lists available at ScienceDirect

www.ashi-hla.org

journal homepage: www.elsevier.com/locate/humimm

HLA-C locus allelic dropout in Sanger sequence-based typing due to intronic single nucleotide polymorphism Christopher Cheng a, Zahra Mehdizadeh Kashi a, Russell Martin a,⇑, Gillian Woodruff a, David Dinauer b, Tina Agostini b a b

Kashi Clinical Laboratories, Inc., 10101 SW Barbur Blvd., Ste. 200, Portland, OR 97219, USA Life Technologies, Brown Deer, Wisconsin, 9099 N Deerbrook Trail, Brown Deer, WI 53223, USA

a r t i c l e

i n f o

Article history: Received 7 January 2014 Accepted 27 September 2014 Available online 12 October 2014 Keywords: HLA-C C*17:01:01:02 C*06:02:01:01 C*17:01:01:03 C*06:02:01:03

a b s t r a c t We report a novel HLA-C allele that was identified during routine HLA typing using sequence-based methods. The patient was initially typed as a C*06:02, 06:04 with two nucleotide mismatches in exon 3, (C to T and T to G changes) which would have resulted in a non-synonymous mutation of a leucine residue being replaced with tryptophan. Further resolution of the patient’s type by using sequencespecific primers (SSP) revealed that the companion allele to C*06:02 was a novel C*17:01. Confirmation of the existence of the new allele was performed across multiple platforms: Sanger sequencing, SSP, and Next Generation Sequencing (NGS) on the original sample and allele-specific clones for the entire HLA-C locus. The investigation revealed a single nucleotide mismatch within the Sanger sequencing primer binding site in intron 3. The mutation caused the initial C*17 dropout in exons 2 and 3. Further analysis of the Sanger and NGS data revealed that the C*17 had two additional unique positions in introns 2 and 7. The companion C*06:02 allele also possessed a novel position at intron 3. On August 31, 2013, the WHO nomenclature committee officially named the novel C*17:01 allele sequence as C*17:01:01:03 and the novel C*06:02 allele sequence as C*06:02:01:03. Ó 2014 American Society for Histocompatibility and Immunogenetics. Published by Elsevier Inc. All rights reserved.

1. Introduction The Human Leukocyte Antigen (HLA), located on the short arm of chromosome 6 [1], is the most polymorphic set of genes in the human genome [2] and plays a large role in acquired immune responses. HLA genes code for cell-surface molecules that allow for recognition and presentation of a multitude of antigens [3] and vary widely between non-related individuals. There are over 2300 reported HLA-C locus alleles to date in the IMGT/HLA Database [4]. HLA typing, or determination of HLA alleles held by a particular individual, is very important in the successful transplantation of solid organs and bone marrow [5]. Matching a donor and recipient’s HLA type can increase a patient’s probability of survival [6] by reducing the chance of the recipient’s immune system

Abbreviations: CDS, Coding Sequence; SBT, Sequence Based Typing; dNTP, Deoxynucleotide Triphosphate; PCR, Polymerase Chain Reaction; HLA, Human Leukocyte Antigen; SSP, Sequence Specific Primer; LB, Luria Broth; WHO, World Health Organization; SNP, Single Nucleotide Polymorphism; NGS, Next Generation Sequencing; GvHD, Graft vs. Host Disease. ⇑ Corresponding author. Fax: +1 503 206 6939. E-mail address: [email protected] (R. Martin).

recognizing transplanted material as a foreign substance [5]. Having an allele library for HLA alleles that is as complete as possible is necessary to increase the likelihood of finding a donor with an HLA type that matches a patient’s type as closely as possible. Often this work is performed with sequence based typing (SBT), which is a powerful approach that can reveal the presence of novel alleles. SBT involves using Sanger sequencing to obtain the sequences of allele-defining portions of HLA genes (particularly exons 2 and 3 for class I genes and exon 2 for class II genes). We present a case where the initial SBT analysis detected a novel allele in a Caucasian patient but could not fully characterize it. Under the wrong circumstances, a mistyping of this nature could lead to graft rejection or serious Graft vs. Host Disease (GvHD) for the patient. This case study illustrates the need to utilize other molecular platforms to facilitate identification of novel alleles. 2. Materials and methods 2.1. Sample DNA extraction An EDTA blood specimen was submitted for testing as a potential healthy sibling donor for a patient diagnosed with acute

http://dx.doi.org/10.1016/j.humimm.2014.09.016 0198-8859/Ó 2014 American Society for Histocompatibility and Immunogenetics. Published by Elsevier Inc. All rights reserved.

1240

C. Cheng et al. / Human Immunology 75 (2014) 1239–1243

Fig. 1. Schematic representation of HLA-C gene, as well as two pairs of primer binding locations. Genomic amplification of sample 2454 with these 2 sets of alternative SeCore HLA-C SBT amplification primers was performed to determine if the location of sample mutation was in the 50 untranslated (forward) or intron 3 (reverse) amplification primer binding sites.

myelogenous leukemia. Genomic DNA was isolated from whole blood with a semi-automated M48 DNA extractor (Qiagen, Germantown, MD). DNA purity and concentration were evaluated using a NanoDropÒ ND-1000 spectrophotometer (ThermoFisher Inc., Fremont, CA).

from an SSP kit that targeted C*06, C*17, or both alleles and that also bind within exons 3 or 4 to increase sequence resolution. The final nucleotide sequence was compared to a database of known HLA alleles using BLAST. 2.5. TOPO-TA cloning

2.2. Sequence based typing for HLA-C Extracted genomic DNA was amplified with a SeCoreÒ C Locus Sequencing Kit (Life Technologies, Brown Deer, WI) according to the manufacturer’s instructions and sequenced with an ABI 3100 Genetic Analyzer (Life Technologies, Foster City, CA). HLA-C locus alleles were typed with uTYPEÒ HLA SBT Analysis Software v5.0 (Life Technologies. Brown Deer, WI) and Assign v3.5 (Conexio, San Francisco, CA) software. The kit creates two amplicons: one of which includes exons 2 and 3, and one of which includes exon 4. These three exons are then sequenced using one sequencing primer pair per exon. 2.3. HLA-C SSP Resolution of inconclusive typing by SBT for the specimen necessitated the use of high resolution SSP to determine the occurrence of a novel allele variant or PCR dropout. Extracted DNA was run with a C Locus Hi Res SSP UniTrayÒ (Life Technologies, Brown Deer, WI) according to the manufacturer’s instructions, and was analyzed using UniMatchÒ v5.0 software (Life Technologies, Brown Deer, WI). 2.4. HLA-C SBT with alternate primer mixes Sample aliquot was submitted to SBT and SSP kit manufacturer (Life Technologies, Brown Deer, WI) for verification of kit quality and determination of sample-specific issue. The sample was retested alongside control samples (UCLA524 C*04:01:01:01 + 17:01:01:01 and IHW9433 C*07:02:01:01 + 17:01:01:01) on ABI 3730 platform (Life Technologies, Foster City, CA) to demonstrate that the event was sample-specific. Two alternate amplification mixes were utilized to characterize the region corresponding to intron 3. The alternate amplification mixes were made from a combination of standard SeCoreÒ C Locus SBT Kit primers and C locus-specific development primers. (Life Technologies, Brown Deer, WI). One mix amplified C locus exons 1 through 6 (Fig. 1) in a single fragment. The second mix amplified C locus intron 1 through intron 3. Life Technologies sequenced over the primer binding site to determine the position and type of any potential base changes present in the sample. Secondary confirmation of these results was performed by sequencing with primers

Once the presence of a mutation was confirmed, the PCR product from the confirmatory sequencing was used to segregate and clone the alleles to show a relationship between the mutation and the initial typing results. Agar plates were made containing Luria Broth (LB), agar, and 100 ng/lL ampicillin and stored at 4 °C. A TOPO-TA CloningÒ Kit for Sequencing (Life Technologies, Carlsbad, CA) was used for the cloning process according to the manufacturer’s instructions. 2.6. Colony screening Colonies were prescreened for the presence of target C*06 or C*17 alleles using C*06- and C*17-specific SSP AmbiSolv mixes (Life Technologies, Brown Deer, WI), which are PCR primer pairs that only amplify specific alleles; this amplification is then visualized using agarose gel electrophoresis. All selected plasmids were sequenced for HLA-C at exons 2, 3, and 4 according to protocol with a plasmid concentration of 100 ng/lL. Screening with SBT confirmed that selected colonies were homozygous for either C*06 or C*17 alleles. Life Technologies sequenced exon 1 through intron 6 of the selected clones using a 3500xL (Life Technologies, Foster City, CA) with run module StdSeq50_POP7 to collect the maximum number of bases. Sequencing reactions were set up according to the BigDyeÒ Terminator v1.1 Cycle Sequencing Kit protocol, using SeCoreÒ cycle sequencing conditions. Files were loaded into uTYPEÒ software for analysis. Base pair errors due to the cloning process were identified by comparing the original specimen sequence with 2 group-specific colonies. Initial review focused on the intron 3 region in order to pinpoint the location of the SNP and associate it with the HLA-C*17 allele type. Review was expanded to encompass exon 1 through intron 6. Data was compared to C*06:02:01:01 and C*17:01:01:02 reference sequences as these alleles were the closest match. 2.7. NGS sample sequencing Sample DNA was amplified using long range PCR amplification. The design of the HLA-C locus-specific primers and the protocol for library creation and PGM sequencing are based on the previous publication [9], except as follows: 1.0 lL (4 pmol/lL) of each

C. Cheng et al. / Human Immunology 75 (2014) 1239–1243

1241

Fig. 2. The figure shows the novel positions of sample 2454 associated with the C*06 and C*17 alleles when compared to the IMGT/HLA Database’s C*06:02:01:01 and C*17:01:01:02 [4]. Each novel polymorphism was concordant between Sanger (bottom row) and NGS (top row) results. NGS Position Read Statistics Legend: Read Depth = number of individual sequence reads obtained for a given position; A, C, G, T, D, I = percent and number of sequence reads containing adenine, cytosine, guanine, thymine, a deletion, or an insertion at the given position.

primer mix was used for amplification, and the long-range PCR reactions were performed using Veriti™ Dx 96-well Thermal Cycler (Life Technologies, Carlsbad, CA). The NGS results for the entire C locus gene from the PGM were loaded into a custom plug-in software for comparison to the C*06:02:01:01 and C*17:01:01:02 alignments as these alleles were the closest sample match. The C locus 30 untranslated region was reviewed in IGV 2.0 software. The 50 untranslated region and introns 6 through the 30 untranslated region were Sanger sequenced using the NGS amplicon to cover regions not previously covered by the colony screening in Section 2.6. 3. Results 3.1. Initial sample characterization A conclusive HLA C locus typing result for sample 2454 was not obtained using the SeCore SBT HLA typing kit due to an allele drop out in the exon 2 and 3 amplicon. The dropped allele was confirmed to be a C*17:01 with companion allele C*06:02 by SSP Unitray. Independent laboratory testing also confirmed the high resolution HLA-C typing C*06:02 + C*17:01. Results on a new DNA extract matched that of the original DNA extraction results. The SBT kit manufacturer, Life Technologies, confirmed the C*17 dropout in the exon 2, 3 amplicon and identified the dropout to be sample-specific. 3.2. HLA-C SBT with alternate amplification and sequencing primers Alternative HLA-C amplification mixes (Life Technologies, Brown Deer, WI) indicated a mutation within the primer binding region of intron 3. SBT typing results with the alternative mix were C*06:02:01:01/02:01:02/03:02/55 + 17:01:01:01/01:01:02/02/03, similar to what was observed in previous SSP testing. Sequencing of the 30 exon 2, 3 amplification primer binding site in intron 3 utilizing standard sequencing primers and Group

Specific Sequencing Primers (GSSPs) consistently resulted in a sequence that differed from all currently known alleles in the IMGT/HLA Database [4]. Sample 2454’s mutation was identified as a C peak at position 1246 of intron 3, specific only to sample 2454 and absent in C*17 control samples (data not shown). Both the intron 3 alignment with the IMGT/HLA Database and a BLAST analysis of the sequence showed no perfect matches with any known HLA alleles [4]. 3.3. Sequencing of HLA-C locus Isolation of C*17 and C*06 homozygous clones was necessary to establish a relationship between the novel intron 3 sequence and the C*17 allele. Cloning produced 14 purified colonies of which 5 were C*06 positive and the remainder C*17 positive as determined by SSP and SBT screening. Use of SBT allowed for a sequence comparison between the two target alleles. Alignment of the chosen C*17-positive clone sequences against the IMGT/HLA Database showed a C mismatch at position 1246 of intron 3, confirming results from previous sequencing analyses and a T mismatch at position 684 of intron 2 [4]. Sequencing of the original sample 2454 with NGS confirmed the C mismatch at position 1246, T mismatch at position 684, and also identified a T mismatch at position 2902. The T mismatch at position 2902 was confirmed by Sanger sequencing of intron 7 using standard sequencing primers and GSSPs (Fig. 2). Alignment of the chosen C*06-positive clone sequences against the IMGT/HLA Database showed a T mismatch at position 1522 of intron 3, indicative of a C*06:46N allele; however, the C*06:46N was ruled out by a C at position 742 of exon 4. The C*06:46N allele was the only allele with a ‘‘T’’ at position 1522. Sequencing of sample 2454 with NGS confirmed the only nucleotide differentiating the C*06 allele from C*06:02:01:01 was located at position 1522 of intron 3 (Fig. 2). Overall the Sanger sequencing and NGS data from the original specimen and C*17 colonies established novel intron 2, 3 and 7

1242

C. Cheng et al. / Human Immunology 75 (2014) 1239–1243

Fig. 3. An alignment of C*06:02:01:01 and C*06:04 [4], illustrating the nucleotide differences between the two alleles at codon 156 as well as the lack of sequence data available for C*06:04 in exon 4.

sequence unique to the C*17:01:01:03 allele for sample 2454. The C*17:01:01:01 ambiguity was ruled out by the intron 6 sequence; while introns 2, 3, and 7 ruled out the C*17:01:01:02 ambiguity C*17:02 and C*17:03 ambiguities were ruled out by exon 1 sequences. The final typing for sample 2454 as assigned by the WHO nomenclature committee is a C*17:01:01:03 + 06:02:01:03 with novel positions in introns 2, 3, and 7.

4. Discussion The Caucasian patient in this study was initially typed as a C*06:02, 06:04 with two nucleotide mismatches in exon 3, (C to T and T to G changes) which would have resulted in a non-synonymous mutation of a leucine residue being replaced with tryptophan. Had this been entered into the patient’s record, there would have been a much stronger likelihood of GvHD due to immunologic recognition of the patient’s C*17 allele by the graft. However, there were several clues that an allele dropout might be occurring. First, while the sequences in exons 2 and 3 appeared to be homozygous, the sequence in exon 4 was clearly heterozygous. Combined with the knowledge that exons 2 and 3 are located on a different amplicon than exon 4, this raises suspicion that one of the two amplicons may not be producing an accurate result. Furthermore, C*06:04 is not sequenced in exon 4, explaining how the

initial typing could result in relatively few mismatches with an entire antigen missing from most of the sequence, and the two mismatches that it contains in exon 3 are the very same that distinguish C*06:02 and C*06:04 (Fig. 3), further indicating that the data may be incomplete. In this study, the high resolution capabilities of HLA SBT were highlighted. What was believed to be an allele dropout due to a faulty test kit was actually a novel C*17:01:01:02 allele variant (since named C*17:01:01:03) with intronic mutations. The initial sequencing results hinted at the possibility of an allele dropout due to the presence of allelic heterozygosity in exon 4 but not in exons 2 and 3. Another high resolution typing method was required to clarify the disagreement between exon 4 and exons 2 and 3. Use of SSP confirmed that an allele dropout had occurred and identified it as a C*17. Once the missing allele was identified, SBT made it possible to run an in-depth analysis of the full length HLA-C sequence to identify the cause of the allele dropout. The novel C*17:01:01:03 allele identified in this study possessed a single nucleotide mismatch in intron 3. This mutation caused the allele dropout during the initial SBT run by being located in the binding site for the 30 exon 2, 3 primer, thus explaining why exons 2 and 3 did not exhibit allele heterozygosity as exon 4 did. While identifying the C*17-associated mutation, further analysis of the SBT and NGS data revealed the presence of mutations in introns 2 and 7, as well as a novel companion C*06 allele.

C. Cheng et al. / Human Immunology 75 (2014) 1239–1243

This case emphasized the need for multiplatform analysis when a conclusive HLA type cannot be determined by a single methodology. In most cases, SBT has been shown to provide higher resolution HLA typing [7]. In this case study, SBT was initially only able to show the possibility of an allele dropout, not identify it. However, SSP identified the dropout as a C*17 allele. The use of different HLA testing platforms should be a standard protocol in order to ensure that accurate haplotypes are produced, particularly when the initial typing yields suspicious data such as that produced in this case. This practice is useful, as well, for ruling out allele dropouts due to mutations caused by hematologic malignancy. The need for other high resolution platforms, however, does not diminish the resolution power of SBT, but rather reinforces it. In this case, the intronic mutations would not have been revealed without SBT and NGS coverage of the full length sequence. In fact, SBT and NGS were able to identify novel sequence not only for the target C*17 allele, but unexpectedly for the companion C*06 allele as well. The SBT platform makes it possible to identify novel alleles. When coupled with full length sequencing, SBT is able to provide more ambiguity resolution with the additional sequencing data. This point was evident during the study as C*17:02 and C*17:03 ambiguities were ruled out by exon 1 sequences, which typically are not used in HLA class I typing. The SBT platform allows for a broader analysis of sequencing data, which was a necessity in this study as the mutation did not lie within the SeCore HLA-C kit’s target regions (exons 2, 3, and 4). Sequencing across the entire length of an HLA locus should be encouraged as it would increase the amount of sequence data and hence aid the discovery of novel allele variants [8]. Widespread implementation of NGS or development of new HLA SBT kits that allow for analysis of the entire locus would promote the adoption of a new standard in HLA laboratories. As technology and testing materials both evolve and drop in

1243

price, sequencing methods will continue to provide high quality HLA typing at a reasonable cost. Acknowledgments We would like to thank Dr. Tsiporah Shore of Cornell Medical Center Presbyterian, New York, New York, for all of her support and medical advice. References [1] Reinders J, Rozemuller E, Otten H, Houben A, Dormoy A, Mulder A, et al. 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;66:912–20. [2] Gill KK, Leligdowicz A, Luo M, Bielawny T, Brunham R, Plummer F. Identification of four novel HLA-A alleles from an East African population by high-resolution sequence-based typing. Hum Immunol 2006;67:833–88. [3] Scheltinga SA, Johnston-Dow LA, White CB, van der Zwan AW, Bakema JE, Rozemuller EH, et al. A generic sequencing based typing approach for the identification of HLA-A diversity. Hum Immunol 1997;57:120–8. [4] Robinson J, Halliwell JA, McWilliam H, Lopez R, Parham P, Marsh SGE. The IMGT/HLA database. Nucleic Acids Res 2013;41:1222–7. [5] Mullighan CG, Petersdorf EW. Genomic polymorphism and allogeneic hematopoietic transplantation outcome. Biol Blood Marrow Transplant 2006;12:19–27. [6] Tiercy J-M, Bujan-Lose M, Chapuis B, Gratwohl A, Gmur J, Seger R, et al. Bone marrow transplantation with unrelated donors: what is the probability of identifying an HLA-A/B/Cw/DRB1/B3/B5/DQB1-matched donor? Bone marrow Transplant 2000;26:437–41. [7] Adams SD, Barracchini KC, Simonis TB, Stroncek D, Marincola FM. High throughput HLA SBT utilizing the ABI Prism 3700 DNA Analyzer. Tumori 2001;87:40–3. [8] Deng Z, Wang D, Xu Y, Gao S, Zhou H, Yu Q, et al. HLA-C polymorphisms and PCR dropout in exons 2 and 3 of the C*0706 allele in sequence-based typing for unrelated Chinese marrow donors. Hum Immunol 2010;71:577–81. [9] Shiina T, Suzuki S, Ozaki Y, Taira H, Kikkawa E, Shigenari A, et al. Super high resolution for single molecule-sequence-based typing of classical HLA loci at the 8-digit level using next generation sequencers. Tissue Antigens 2012;80(4):305–16. http://dx.doi.org/10.1111/j.1399-0039.2012.01941.x. Epub 2012 Aug 4.