Associations of alleles atloci located throughout the HLA complex

Abstracts

5

1.1. HLA Diversityof Alleles and Haplotypes(I) 001

RECOMBINATIONHOTSPOTS IN THE HLAREGIONOF CHROMOSOME6 Barmada M. Michael, Bias W. B., Delaney N.L., SchmeckpeperB. J. Johns Hopkins University, Baltimore, MD, USA To localize potential recombination"hotspots"in the HLA region of chromosome6p, we have analyzed 28caucasianfamilies in which recombinant individuals were identified by serologic(lILA-A, B, C, DR, and DQJ and/or protein (F13AI, BF, GLOl) typing. In additionto the five microsatellite markers which were required for the 12th InternationalHistocompatibility Workshop(D6SI05, D6S265, TNFa, DQCAR, D6S291), we analyzed four more microsatellitemarkers (FI3AOI, D6S89, D6S285, D6S46I). Statisticalanalysis of our data revealed 1706heterozygous couplings (of 2772 possible) giving a 62% average heterozygosity over all markers. From 99 childrenanalyzed, 84 recomhinationevents were identified. Recombinationfractions and Kosambi map distances werecalculatedfor each interval andcomparedto values found in the Genome Database (GDB) usingcbi-squaredanalysis. Physical distances between micro satellite markers were not available, therefore no precise comparison betweeophysical and genetic distances could be made. However, comparing the calculatedrecombinationfractions to the estimatedmap distances between loci (from GDB) yieldedestimatedmeasuresof the recombination et al., Am. J. Hum. Genet., 56:1350intensity in eacb marker interval (Cullen 1358, 1995). These results suggest that threerecombination"hotspots"may be located in this region: onepreviously described "hotspot"between DQ and D6S291 (in the area of the TAP genes) (Cullenet al., ibid.) with recombination intensity in this regionreaching3 times thatexpected given the genetic distances reported; one betweenD6SI05 and HLA-A, withrecombinationintensity in this region reaching6 times that expected; and the other between D6S265 and HLAC, with recombinationintensity in this regionreaching 12 times that expected. When sex-specific differences were analyzed, it was noted that the majority of recombinationsin this region are attributable to female meiotic events.

003

COMPREHENSIVESEQUENCE ANALYSIS OF HLA-DQAl AND -DQBI ALLELES AND CHARACTERIZATION OF DRBI-QAP.DQA1.DQBl HAPLOTYPES Shin'ichiroYasunaga I), Akinori Kimura1) 2), TakehikoSasazuki I) I) Departmentof Genetics, Medical Instituteof Bioregulation,Kyushu University, Fukuoka,Japan 2) Departmentof Tissue Phisiology, Division of Adult Diseases, Medical Reserch Institute,Tokyo Medical and Dental University, Tokyo, Japan It is well known that both a chain and 13 chain of HLA-DQ are highly polymorphic. However the polymorphisms outside the hypervariableregion were notfully examinedso far. To furtherclarify the polymorphisms in DQ genes, we determinedthe nucleotidesequences of full length cDNA, spanningfrom the leadersequence to the stop codon, from 15 DQAI alleles and 15 DQBl alleles. We identifiedseveralnew DQ alleles which had identical exon 2 sequence and were differentin other exons. On the basis of the sequence analyses, a comprehensivePCRbased oligotyping system for DQAI gene was established. We then characterizedDRBI-QAP(DQAl promoterl-DQAI-DQBIhaplotypesof B-Iymphoblastoidcell lines homozygous for HLA and healthy unrelated Japaneseand Norwegianpopulations. It was revealed that DQAI alleles, which were identical in exon 2 but differentin other exons, showed close linkagedisequilibriumwith diferentcharacteristicDRB1, QAP and DQBl alleles. These results suggest that DR-DQ haplotypeshave been generated in the earlystage of molecularevolution.

005

SEQUENCING OF A NEW HLA·DR4 ALLELE WITH AN UNUSUAL RESIDUE AT POSITION 88 THAT DOES NOT AFFECT T CELL ALLO RECOGNITION. GebuhrerLucette, Adami Nadia",Freidel Anne-Catherine.Javaux Francoise, JeannetMichel", BetuelHerve. Tiercy Jean-Marie. * Histocompatibility Laboratory. Blood Transfusion Center - Lyon - France. Transplantation Immunolgy Unit".Hilpital Cantonal - Geneva - Switzerland.

It is rathercommon to encounternew HLA alleles identified by unorthodox performedwith sequence specific patterns observed during low resolution typing oligonucleotideprobes (SSOP). One of the best examples is locus DRB I, where allelic subtypesare characterizedby a combination of a limited numberof residues located in threehypervariableregions of exon 2. HLA-DR oligotyping

analysis of a female caucasoId bone marrow donor of the Lyon registry led to the identificationof an individual that typedas ORBI· 11, DRB3· 02. DRB4· 01, DQBI* 0301-0302. This donor washowevertyped by serology as DRII DR4, DR52 DR53, DQ7 DQ8. PCR-SSP typing for DR4 subtype revealed an amplification pattern typical for DRBI· 0404. After PCR amplification of genomic DNA and sequencing the entire cxon 2, a new DRB t allele was identified: DRBl· 04var that is identical to DRBl· 0404. except for one nucleotide at codon 88resultingin a Ser ==> Arg exchange. This mutation had preventedamplificationwith the DR generic primers. The donor inherited this new DRB I· 04var from her mother who presented the same molecular characteristics.The complete SSOP typing of theimplicated haplotype is ; A*6801l, Cw·0704, B·4402, DRBI· 04var, DRB4· 0103, DQA· 0301, DQBI· 0302, DPBl* 0601. Cellulartyping by three HTCs-DRB1·0404/DWI4 from the 9th Workshop showed that this DRBI· 04var typed exactly like aDWl4 cell (the S.R.R. being 40%, 34%.36%,respectively). This suggests that residue 88 docs not affect T cell recogrtition possibly due to itslocation at the far end of the alpha helix flanking the peptide binding groove.

002

MAPPINGRECOMBINATIONSITES IN THE HLA CLASS ITREGION.

Michael Cullen', 1ane1le Noble', Hemy Erlich', Karen Thorpe', Stephan Beck', 10hn Trowsdale',andMary Carrington'

'SAIC Frederick, NCI-FCRDC,Frederick, MD, 'Imperial Cancer ResearchFUIId,London, England, 'Roche MolecularS Y'-, Alameda, CA

been useful in Studies of linka&e disequilibriumacross the HLA class IT region have

predictingwhererecombination is moseHkelyto occur. SIronJ associations exist within the IOOkbregion from DQBIto DRSI, suggesting a low frequency of recombinationin this

region. Conversely,a lackof associationbetween alleles of TAPIandTAPZ(aboutISkb) bas beenobserved, suggesting that recombination o ccurs here wilh relative frequency. Analysis of familial-
004

COMPLETEPRIMARY STRUCTUREOF A NEW DQBl*06 ALLELE

C. Vilches, ·J.M. Garcia-Pacheco,C. de Pablo, ·S. Puente, M. Kreisler. H. Puerta de Hierro & 'Carlos III, Madrid, ·H. do Meixoeiro, Vigo, Spain. Molecularcloning has been greatlyfacilitated by PCR, but most full-lengthclass II sequences were obtainedbefore the PCR era. Nowadays it has almost become a rule to amplify and clone only the second exon of class II genes (often not even to completion) from genomic DNA. While this procedureis rapidand simple, partialsequences can be misleadingfor nomenclatureand typing since sequence polymorphismof uncertainphysiological significance exists outside that region. We have designed a RT-PCR method to amplify the whole coding region of DQB 1 with primersrecognizing its 5'- and 3 '-ut regions. With this approach, we have characterized a new DQBl*06 allele (06-GB), detected by oligotyping in 2 Bubi individuals (Bioko I.), in association with DRBI*1301 and DQAl*0103. The new allele shows greatest similarity with DQB1*0609: DQB1* 06-G8 0609 06051 0604 0603 0602

9

TAC

13 27 ATG GTA

c--T-

--G

30 TAC

48 57 CGG GTT

c-c-- --c --c

-A-A-

70 GGG A-A-A--

86 GGG

-c-c-

87 TAC

-T-T-

91 130 CTG CAG *** •• * T-T--

-G-G-

006

ASSOCIATIONSOF ALLELESAT LOCILOCATEDTHROUGHOUTTHElILA COMPLEX CarringtonM J, Martin MI, Newell

W,

Beck 52, Klitz W3

'BCDP, SAIC Frederick, NCI-FCRDC, Frederick, Maryland, 'Imperial Cancer Research Fund, London, England, 'University of California, Berkeley, California. Frequencies of recombination within regions of the HLA complex have been determined previously using 59 Centre d'Etude du Polymorphisme Humain (CEPH) pedigrees. Here, these families were usedto identify independent haplotypes composed of 20 markers spaced across the complex. Pairwise analysis of the 20 markers indicated segments of the complex varying in significance and strength of association. The degree of statistical codependence between alleles at one locus and those at a second locus D QBl, which is not very (mutual information) was highest between DRB! and surprising given their relatively closeproximity (about 80kb) and high variability. However, the lowest mutualinformation value was between the furthest variants of TAPI and TAP2«50kb), reflecting in part the limited variability of these two "alleles" (2 in both cases), but, perhaps, also reflecting a relatively high level of recombination between these markers. Normalizationof mutual informationvalues by individual locus variabilities indicated four main regions of highly associated loci among the 20 typed: 0) the telomeric end of the class II region including markers DRBl, DQAl, and DQBl, (ii) the telomeric end of the class I regionincluding markers near HLA-A, HLA-F, and MOG (iii) four closely-spaced markers located within and immediately telorneric to TAPl, and (iv) lWO markers located within intron 2 of TAP2. All four regions contained multiple polymorphisms in close proximity, explaining values of high with greater association, whereas other markers typed were more evenly spaced distances to their adjacent markers. Relative recombinationrates derived from measured will be discussed. association and known physical distances between markers

PH 50198-8859(96)00072-9