HLA-A, -B, -C typing by next generation sequencing in a sample of Turkish population
M. Dorak (Liverpool, GB)
HLA typing was performed on 767 unrelated subjects as per the EFI-accredited Alp Şen Tissue Typing and Genetic Research Laboratory in Istanbul University, Istanbul Faculty of Medicine using Illumina's MiSeq Sequencing System. The subjects were genotyped for the volunteer bone marrow donor registry. HLA-A, -B, -C typing on next-generation sequencing data was achieved using Omixon Holotype HLA(TM) assay and Omixon HLA Twin(TM) software. Population genetic analyses were done on Arlequin v.18.104.22.168. Genotype frequencies at all three classical HLA class I loci were in Hardy-Weinberg equilibrium (P >= 0.25). Also, when assessed by the inbreeding coefficient (FIS), observed and expected heterozygosity did not differ at any loci (P >= 0.40). Ewens-Watterson tests of selective neutrality did not indicate any statistically significant selection (P = 0.99). The numbers of alleles detected in each locus were 55, 104, 51 for HLA-A, -B, -C loci, respectively. The most common three alleles (and their frequencies) at each locus were A*02:01:01 (0.201), A*24:02:01 (0.145), A*01:01:01 (0.111); B*51:01:01 (0.114), B*35:01:01 (0.081), B*18:01:01 (0.060); and C*04:01:01 (0.175), C*12:03:01 (0.114), C*07:01:01 (0.100). The most common B-C haplotypes (and their frequency / D' as LD measure) were: B*35:01:01 - C*04:01:01 (f = 0.072; D' = 0.864), B*49:01:01 - C*07:01:01 (f =0.043; D' = 0.967), B*38:01:01 - C*12:03:01 (f = 0.040; D' = 0.965). These B-C haplotypes most frequently had the following HLA-A alleles as part of three-locus haplotypes: A*11:01:01, A*23:01:01, and A*26:01:01, respectively. However, the most common three-locus haplotype was none of these, but A*03:01:01 - B*07:02:01 - C*07:02:01 (f=0.018). Of the 11 three-locus haplotypes with more than 0.01 frequency, three were the B-C haplotype B*35:01:01 - C*04:01:01 with different alleles at HLA-A: A*11:01:01 - A*24:02:01 - A*03:01:01, suggesting the presence of a recombinational hot-spot between HLA-A and -C on this particular B-C haplotype. The addition of these high-resolution HLA class I types in a sample of Turkish population should fill a gap in global databases.
Common and rare HLA-allele groups and HLA-haplotypes in the Moscow Russian donors of hematopoietic stem cells
E. Khamaganova (Moscow, RU)
The aim of our study was to analyze frequencies of HLA-A, -B, -C, -DRB1, and -DQB1 allele groups haplotypes in a representative group of Moscow Russian donors of hematopoietic stem cells. The study included 1507 unrelated donors who according to their self-assessments were the Russians by ethnicity. The donors were recruited and HLA-typed in the registry of National Research Center for Hematology (Moscow). The HLA-A, -B, -C, -DRB1, -DQB1 typing (low or intermediate resolution) was performed using Luminex 200 system (Luminex, TX, USA) with Lifecodes HLA SSO typing kits (Immucor, CT, USA). Statistical analysis was performed using the ARLQUIN version 3.5 software [http://cmpg.unibe.ch/software/arlequin35]. We calculated maximum-likelihood estimates for HLA-A,-B,-C,-DRB1 and -DQB1 allele group frequencies. Five-locus haplotype frequencies were estimated by expectation–maximization (EM) algorithm. Hardy–Weinberg exact tests were performed for each of the HLA loci. There were no deviations from HWE in all investigated HLA loci. We found 16 allele groups with frequencies >0.001 in HLA-A, 29 allele groups with frequencies >0.001 in HLA-B, 13 allele groups with frequencies >0.001 in HLA-C, 13 allele groups in HLA-DRB1 and five allele groups in DQB1, all defined allele groups of HLA class II had frequencies >0.001. 1284 different five-locus HLA-haplotypes were calculated. 201 of HLA-haplotypes had frequency >0.001. Seven of HLA-haplotypes had frequency >0.01: A*01-B*08-C*07-DR*03-DQB*02 (0.0415), A*03-B*07-C*07-DR*15-DQB*06 (0.0265), A*03-B*35-C*04-DR*01-DQB*05 (0.0262), A*25-B*18-C*12-DR*15-DQB*06 (0.0176), A*02-B*13-C*06-DR*07-DQB*02 (0.0163), A*02-B*18-C*07-DR*11-DQB*03 (0.0153), A*02-B*07-C*07-DR*15-DQB*06 (0.0139). Distribution of HLA-allele groups and HLA-A-B-C-DRB1-DQB1 haplotypes in the Russian population of Moscow is similar to their distribution in the Russians from other regions of our state.
Differential HLA-C expression in Caucasians: Useful data from direct HLA-protein measurements in 188 healthy German donors
C. Tsamadou (Ulm, DE)
Differential HLA-C expression, although already investigated in African Americans (Apps et al. Science 2013), has not been accordingly researched in Caucasians. Furthermore, the possibility that HLA-C expression heterogeneity could reflect an overall HLA class I expression regulation has not been yet experimentally addressed. This work aims at elucidating both issues through actual protein expression data retrieved from healthy German blood donors.
188 buffy-coats provided from the Ulm Blood Donor Service were used for leucocyte collection in order to determine by flow cytometry the protein expression levels of HLA-C and HLA class I molecules on lymphocytes as previously described. To this end a DT-9 and an anti-HLA-ABC Antibody (Ab) were used. Median Intensity Fluorescence (MFI) coefficients were calculated for each C allotype through implementation of a linear regression model, as previously described. According to our findings HLA-C*03, C*07 and C*08 (MFI: 408, 450, 481) were ranked as low-expressed C antigens in contrast to C*12, C*14 and C*01, which were notably higher expressed (MFI: 1261, 1280, 1322). In addition, our data on overall HLA class I expression levels indicated no direct correlation between HLA-C and HLA class I expression on the cell surface. First, the divergence range of expression levels for HLA class I molecules was markedly narrower (MFIs: 450-758 vs 408-1322) and secondly, low expressed C allotypes were found linked with high HLA class I expression levels (e.g. C*08 758 for Cl-1 vs 481 for C). Our HLA-C expression results match significantly those of Apps et al., at least for low and high -expressed HLA-C antigens, despite the different ethnicity of subjects included in the studies. Moreover, of high interest is our finding regarding the lack of association between HLA-C and HLA class I expression, which in turn points to an HLA-C rather than an HLA class I specific expression-regulating factor accounting for this marked variability in HLA-C expression levels. Further research is required before final conclusions can be drawn.
Tri-allelic pattern at the TPOX locus: Back to the literature about one case
X. Lafarge (Bordeaux, FR)
Pre-graft short tandem repeat profile determination for a patient from Mayotte displayed a peculiar pattern, using PP16 Promega kits, suggesting a contamination. A tri-allelic pattern could be observed for the TPOX locus (8, 10, and 11 AATG repeat motifs). Three of four parental HLA haplotypes were from African origin, the fourth was from Malaysian origin. The 4 patient’s grandparents were from Madagascar and had Swahili as a native language, a typical Bantu idiom. Those tri allelic patterns for TPOX are not contaminants and are described, despite their rarity, present in 2% in the black sub-Saharan populations. In 90%, this third allele (10) is due to an insertion in the X chromosome. It would have occurred in the Bantu tribes before the beginning of their geographical expansion from the Cameroon-Nigeria region to south east Africa then Madagascar and the Comorian islands. The detection of this profile in America would be in line with the migration of African populations in the direction to America linked to the slave trade. In Brazil, where diverse populations were largely mixed, this pattern is also observed in individuals with Caucasoid phenotypes.
Indigenous populations from Madagascar are from Polynesian origin (in particular Indonesian) and speak an Austronesian language despite strong African immigration. Parental HLA haplotypes and grand-parents native language, indicate that they are of African origin, despite a possible race mixing with the indigenous population. This ethnic origin, probably explains the observed tri-allelic pattern in this patient. This genetic particularity must be known in the chimerism post graft follow up for these patients.
Association of the CYP21A2 gene p. V282L mutation with HLA alleles and haplotypes in the Croatian population
Z. Grubic (Zagreb, HR)
The CYP21A2 mutations that are in linkage disequilibrium with particular HLA-A, -B, -DRB1 alleles/haplotypes, cause deficiency of the 21-hydroxylase enzyme (21-OHD) and account for the majority of congenital adrenal hyperplasia (CAH) cases. The aim of this study was to investigate those associations linked to the non-classical (NC) form of CAH among Croatians. The study included parents of patients with the NC form of CAH, positive for the p.V282L mutation (n=55) and cadaveric donor samples (n=231). All subjects were HLA-A, -B, and -DRB1 typed and tested for the presence of the p.V282L mutation. Among parents of patients, 92.73% of subjects were positive for the B*14:02 allele and almost half of them carried the HLA-A*33:01-B*14:02-DRB1*01:02 haplotype. Among cadaveric samples 77 out of 96 subjects positive for the B*14:02 allele had the p.V282L mutation. Among them, 37 were positive for the HLA-A*33:01-B*14:02-DRB1*01:02 haplotype, 23 had the HLA-A*33:01-B*14:02-DRB1*03:01 haplotype, eight had the B*14:02-DRB1*01:02 combination and five were carrying the HLA-A*68:02-B*14:02-DRB1*13:03 haplotype. Four of these subjects were positive only for B*14:02 allele. HLA-B*14:02 was the only single allele which association with the p.V282L mutation reached statistically significant P value (RR=12.00; P=0.0024). Haplotypes B*14:02-DRB1*01:02 (P<0.001) and HLA-A*68:02-B*14:02-DRB1*13:03 (P<0.001) as well as HLA-A*33:01-B*14:02-DRB1*01:02 and HLA-A*33:01-B*14:02-DRB1*03:01 showed high relative risks (RR=45.00, RR=41.63 and RR=36.96, respectively). Our data support the previously documented association of the HLA-A*33:01-B*14:02-DRB1*01:02 haplotype with the p.V282L mutation, but also point out a high frequency of the p.V282L mutation among Croatians with HLA-A*33:01-B*14:02-DRB1*03:01 and HLA-A*68:02-B*14:02-DRB1*13:03 haplotypes.
The first population of the Canary Islands is not only North African: presence of European Atlantic genes and ancient Iberian Language scripts
A. Arnaiz-Villena (Madrid, ES)
Canary Islands (Spain) First Inhabitants (“Guanches”) where postulated to come only from North Africa. Our studies show that not only North Africans but also Iberian/Atlantic Europeans (and possibly others) must have been first Canarians. Debate whether North Africans or Iberians were the first “Guanches” is artificial since Iberian Peninsula-North African gene flow in ancient times was abundant and Iberians share a great part of genetic profile with North Africans. New genetic (HLA) and linguistic data shown in the present paper, is supported by diverse early anthropological and “Guanches” mummies characters which confirm existence of at least two “Guanches” types and a correct interpretation of R1b Y chromosome high frequency in Atlantic Europe (Ireland, British Isles, North Spain, Basque Coast and Portugal), and also, is present in Canary Islands (10%). In the present paper HLA genes and presence of abundant old Iberian language scripts (which show an easy translation proposal by using Basque) in Fuerteventura and also in Lanzarote and the El Hierro Islands suggest that a present day dogma of a hypothetically North African single origin should be changed. Both Atlantic/Europeans and North Africans define origin of Canary Islands first inhabitants.
Identification of genomic full-length sequence of HLA-E in Chinese individuals and two novel HLA-E alleles
Y. Xu (Shenzhen, CN)
HLA-E is one of non-classical HLA class I genes. At present, the study of polymorphism analysis in China is mainly aimed at the variation in exon 3 of HLA-E, which determines HLA-E*01:01 or E*01:03. However, study of identification of the full-length HLA-E and its novel alleles is rarely reported. We aimed to establish a method of identification of HLA-E genomic full-length sequences, and identify novel alleles in healthy blood donors in Shenzhen, China. We extracted DNA from peripheral blood of individuals and designed amplification and sequencing primers in conserved regions according to the sequences of HLA-E published on the IPD-IMGT/HLA Database. We used a high-fidelity polymerase to amplify genomic full-length of HLA-E followed by sequencing, assembling, confirming and typing. During this research, we successfully established a method for amplifying genomic full-length sequence and sequence-based typing. Two novel HLA-E alleles were detected and named by the WHO HLA Nomenclature committee as HLA-E*01:01:01:06 and HLA-E*01:01:01:07. Compared with the closet allele HLA-E*01:01:01:01, HLA-E*01:01:01:06 has one nucleotide change at position -26G->T in the 5'-promoter region, and HLA-E*01:01:01:07 has one nucleotide change at position 3345T->C in the 3'-UTR. HLA-E*01:01:01:06 appears to be a common allele among Chinese. The polymorphism data of genomic full-length HLA-E in Chinese individuals need to be further investigated.
Characterisation and functional implications of the two new HLA-G alleles found in Amerindian (HLA-G*01:08:02) and Caribbean (HLA-G*01:20) populations
A. Arnaiz-Villena (Madrid, ES)
HLA-G polymorphism has been found to be relatively low in all world populations. In the present paper two new HLA-G alleles are described in ancient American natives. A new HLA-G allele (HLA-G*01:08:02) from an Ecuadorian Amerindian individual (male) showed four codon changes with respect to HLA-G*01:01:01. Silent changes in the alpha-1 domain (residue 57, Pro, CCG>CCA) and alpha-2 domain (residue 93, His, CAC>CAT and residue 100, Gly, GGC>GGT) and one non-synonymous change in the alpha-3 domain (residue 219 changed from Arg to Trp). This alpha-3 change may dramatically alter HLA-G interactions with beta-2 microglobulin, CD8, ILT-2 and ILT-4 ligands present in subsets of T, B, NK, monocytes, macrophages and dendritic cells. Another HLA-G new molecule (HLA-G*01:20) was found in a woman from Hispaniola Island, Dominican Republic (Sto Domingo): it presented a silent change in the alpha-2 domain residue 107, Gly, GGA>GGT and non-silent change at residue 178, Met>Thr (with respect to HLA-G*01:01:01) which is close to class I molecule/clonotypic T cell receptor interaction sites. Functional implications of these findings are discussed.
The diversity in the associations between the HLA-DRB4*01:03:01:02N allele and HLA-DRB1 alleles
M. Maskalan (Zagreb, HR)
It is well known that the DRB1 gene is expressed for each HLA haplotype and depending on the DRB1 alleles present, additional DRB genes can be expressed. In the haplotypes with DRB1*04, DRB1*07, and DRB1*09 alleles the second active DRB gene is DRB4. This gene has now 57 known alleles and among them five are nonexpressed: DRB4*01:03:01:02N, DRB4*01:16N, DRB4*01:38N, DRB4*02:01N and DRB4*03:01N. Here we describe HLA-DRB1~DRB4 haplotypes in a group of 242 cadaveric samples positive for one of the DR53 alleles in Croatia. One hundred and eleven samples were DRB1*07:01 positive, 126 were DRB1*04 positive, while 5 samples were DRB1*09:01 positive. HLA-A, -B, -C, -DRB1 and -DQB1 typing was performed by PCR-SSP low resolution method, while HLA-DRB4 and HLA-DRB1*04 typing was performed by PCR-SSP at allelic level. Three different DRB4 alleles were observed among DRB1*04 samples; DRB4*01:02, DRB4*01:03 and DRB4*01:03:01:02N with frequencies of 2.38%, 91.27%, and 6.35%, respectively. Of the eight times DRB4*01:03:01:02N was observed among DRB1*04 samples, the associated allele was DRB1*04:02 (7/26; 26.92%). This is statistically significant in comparison to DRB1*04:01 samples (P=0.0019) and DRB1*04:04 samples (P=0.0113), while comparison with other subgroups of DRB1*04 samples did not reach statistical significance. Among six different DRB4 alleles in the group of DRB1*07:01~DQB1*02:02 haplotypes, the most frequent were DRB4*01:03 (49.09%) and DRB4*01:01:01:01 (41.18%) while alleles DRB4*01:08, DRB4*03:01N, DRB4*01:05 and DRB4*01:02 were detected once or twice. The DRB1*07:01~DQB1*03:03 haplotype, in 98.21% of cases, carried the DRB4*01:03:01:02N allele. This study supports necessity of DRB4 sub-typing to correctly identify the presence/absence of DRB4 null alleles especially in the cases when DRB1*04:02 allele is present. This may have implication in unrelated HSCT program as well as for HLA-disease association studies.
HLA haplotype frequencies based on family analysis in Serbia
Z. Andric (RS)
The knowledge of HLA haplotype distribution in a population is a useful tool for donor search process. The aims of this study were to analyse real HLA haplotype frequencies and to compare them with estimated haplotypes of Serbian bone marrow donors. We analyzed 714 haplotypes of 357 samples collected between 2014 and 2016 belonging to hematological patient families. HLA typing was performed by Luminex SSO and SSP. HLA haplotype frequencies were obtained by direct counting. We ranked the haplotypes by the decreasing order of their frequencies and compared them with previously estimated haplotypes of 1999 individuals from the Serbian bone marrow donor registry. A total of 378 different HLA-A-B-DRB1 haplotypes, 143 repetitive and 235 unique, were found in the family study. Only 6 haplotypes were found with frequency higher than 1%. The most frequent haplotypes were A*01-B*08-DRB1*03 (3.6%), A*02-B*18-DRB1*11 (2.8%) and A*33-B*14-DRB1*01 (1.7%). These haplotypes are also the most frequent among estimated haplotypes of Serbian donors (5.9%, 1.9% and 1.4%, respectively). The most frequent alleles were A*02 (28.8%), A*24 (13.2%), A*03 (12.6%) and A*01 (12.5%), the same as in the registry donors (29.5%, 11.1%, 11.3% and 14.3% respectively). HLA-B*35 (14.7%) and HLA-B*51 (14.1%), following with B*18 (10.8%) were the most frequent in the family group as well as in the donor group (13.1%, 12.8% and 9.9% respectively). The most frequent HLA-DRB1 in both groups were DRB1*11 (20.31% in the family group and 16.89% in donors), following with DRB1*01 (10.5%) and DRB1*13 (10.1%). We confirmed distribution of the most common allele group in the Serbian population. Determining real haplotypes without the use of algorithms could help in searching donor for haematological patients and provide useful information for extended haplotype profiles of the Serbian population. A high number of unique haplotypes argue in favor of increasing the number of donors in Serbian registry.
Major histocompatibility complex-DMB allelic diversity in Old and New World non-human primates: intra-species pattern of evolution
A. Arnaiz-Villena (Madrid, ES)
New MHC-DMB complete cDNA sequences have been obtained in individuals belonging to the following primate species/families: Hylobates lar, Papio hamadryas, Macaca mulatta, Macaca fascicularis, Cercopithecus aethiops and Saguinus oedipus. Exonic allelism has been recorded all along the DM molecule domains and analyses of the critical residues in the conformation of the MHC-DR peptide-binding site were done; it was found that an evolutionary pressure over the putative peptide-binding region of the DMB molecule favours synonymous changes. These results are in contrast with those found in the MHC class I and class II genes, where non-synonymous DNA base substitutions are favoured The immunoreceptor inhibition motif Tyr230-X231-X232-Leu233 (ITIM) is invariably present in all extant studied primates since 40 million years ago. It confirms the important function for this molecule, directing DR molecules towards the endosomal/lysosomal HLA class II compartment and sending inhibitory signals to cells in order to stop synthesis of unnecessary MHC-DR molecules. Some Macaca individual's DMB molecules (appearing on Earth more than ten million years ago) do bear both short (without ITIM) and long cytoplasmic tails (with ITIM), similarly to what has been found in human individuals. These differences may have important functional implications. Obtained MHC-DMB allele phylogenetic trees suggest an intraspecies evolution, since alleles of the same species cluster together, as it occurs in other MHC related genes (Bf, C4d, HLA class I and class II). Other MHC class I and class II molecules' phylogenetic trees show a trans-species pattern of evolution. Also, these trees show that gorilla and human are genetically closer than human and chimpanzee (like C4d phylogenetic trees).
Identification of two distinct haplotype structures elucidated by polymorphism and gene expression analyses in dog leukocyte antigen class I and II genes: DLA-88, DLA-12, DLA-64 and DLA-DRB1
J. Miyamae (Fujisawa, Kanagawa, JP)
The current information on the polymorphism variation and haplotype structure of the domestic dog leukocyte antigen (DLA) genes is limited in comparison to other experimental animals. In this paper, to better elucidate the degree and types of polymorphisms and genetic differences. For DLA-88, DLA-12, DLA-64 and DLA-DRB1, we genotyped four families of 38 beagles and another 404 unrelated dogs representing 49 breeds by RT-PCR based Sanger sequencing. We identified 76 alleles for DLA-88, 21 for DLA-12, seven for DLA-64 and 47 for DLA-DRB1, of which 44, 20, seven and six were newly described, respectively. Phylogenetic analysis supported that the DLA-88, DLA-12 and DLA-64 alleles were independently generated after the original divergence of the DLA-79 alleles. Our genotyping data suggested that there were two distinct two-gene DLA-class I (DLA-88a and DLA-12/88b) haplotype structures, tentatively named DLA-88a - DLA-12 and DLA-88a - DLA-88b. Of them the haplotype frequency of DLA-88a - DLA-88b occupied 31.9% of the unrelated dogs. Quantitative real-time PCR analysis showed that the gene expression levels of DLA-88b and DLA-88a were the same, and that the gene expression level of DLA-12 was significantly lower. In addition, haplotype frequency estimations revealed 143 different DLA haplotypes (88a-12/88b-64-DRB1) overall, and 37 different DLA haplotypes in homozygous dogs for 29 breeds and mongrels. Further studies and large scale genetic screening of dogs would help to better identify and define various DLA haplotypes. This DLA polymorphism information and genetic differences among the dog breeds could be used as a standard internal control of the MHC genetic background for the benefit of biomedical research into regeneration medicine using the most common DLA haplotypes as models for human MHC related diseases.