CA2035813A1 - Human leukocyte antigen typing - Google Patents

Abstract

ABSTRACT

A method of human HLA-DR and/or Dw allotype matching is provided. The process comprises effecting polymerase chain reaction (PCR) amplification of a HLA-DRB gene exon 2 nucleotide sequence of a first sample of DNA. The DNA fragments resulting from this amplification are then separated according to size and the length polymorphism of these separated fragments is then determined. Thereafter, the determined length polymorphism is compared with the length polymorphism of DNA fragments resulting from PCR amplification of the HLA-DRB gene exon 2 nucleotide sequence of a second sample of DNA.

Description

HUMAN LEU~OCYTE ANTIGEN TYPING

The present invention relates to matching human leu~ocyte antigens.
The human leukocyte antigen (HLA) class Il genes of the human major histocompatibility complex ~MHC) encode cell surface glycoproteins with a fundamental role in the immune response: the presentation of antiqenic peptides to ~ helper cells. The recognition by T helper cells of foreign antigen in association with self-M~C class II molecules triggers a cascade of immunological responses resulting in the activation of both cytotoxic T cells and B cells to effect killing of antigen-presenting cells and induction of an antibody response, respectively.
At least six HLA class II loci have been defined, three of which (HLA-DR, DQ and DP) are known to express functional products. Pairs of A (formerly alpha) and B
(formerly beta) genes within these three loci encode heterodimeric protein products which are multi-allelic and alloreactive. In addition, combination~ of epitopes on DR
and/or DQ molecules are recognised by alloreactive T cells;
This reactivity has been used to define "DW" type~`by cellular a~says based upon the mixed lymphocyte reaction (MLR).
~ s a corollary to DR and DQ alloreactivity, it has been demonstrated that matching of donor and recipient HLA-DR and DQ alleles prior to allogeneic transplantation 2~3~8~ ~

has an important influence on allograft survival.
~herefore, HLA-DR and DQ matching is ~ow generally under-taken as a clinical prerequisite for renal and bone marrow transplantation.
Methods for the identification of alloreactive epitopes have until recently been confined to serological and cellular typing. Serological typing of DR and DQ is well established and employs antisera generated as a result of humoral recponses to DR and DQ alloantigens. However, serological typing is frequently problematic, due to the availability and crossreactivity of alloantisera and because live cells are required.
Extensive molecular genetic analysis of the HLA class II region has recently enabled the definition of HLA-DR, DQ
and DP alleles at the DNA level. It is now recognised that many polymorphisms detected at the DNA level could not previously be defined serologically. However, the polymorphisms revealed by DNA probes are slowly being confirmed as functional since new alloantisera are being reported which can now define serological "splits". It therefore appears that the polymorphisms detected ~at the'DNA
level might well reflect functional epitopes. Therefore, DNA typing is becoming more widely used as an ad~unct, or alternative, to serological tests.
To date, the most extensively employed DNA typing method for the identification of these alleles has been restriction fragment length polymorphism (RFLP) analysis.

Thi6 well e6tabli6hed method for HLA cla6s II DNA typing suffer6 from a number of inherent drawback6. Thu6, RFLP
typing i6 too time-con6uming for clinical u6e prior to cadaveric renal' tran6plantation, and for thi6 rea60n it i6 be6t 6uited to live donor tran6plantation or retro6pective 6tudie6. Furthermore, RFLP does not generally detect polymorphi6m within the exon6 which encode functionally 6ignificant HLA cla66 II epitope6, but relie6 upon the 6trong linkage between allele-6pecific nucleotide 6eguence6 within the6e exon6 and restriction endonuclease recognition site di6tribution within 6urrounding, generally noncoding, DNA.
The inten6ive effort by many groups to sequence HLA
cla6s II allele6 ha6 revealed that the ma~ority of alloreactive epitope~ of DRa, DQB, DQA and DPB protein product6 are confined to the membrane-di6tal domain, encoded by the 6econd exon of the re6pective gene6. The flanking 6equence6 of the6e exons are both locu6-6pecific and highly con6erved between all~le6, and a6 6uch they are amenable to enzymatic amplification u6ing the polymera6e chain reaction ~PCR) technique (Saiki et al, Science 230, 1350, 1~85;
Scharf et al, Human Immunol. 23, 143, l9B8). Thi6 technique ha6 expedited the acqui6ition of nucleotide 6equence data for virtually all of the allele6, and'ln turn h~s permittea the con6truction of allele-6peciflc oligonucleotide (ASO) probe6 wh~ch are able to detect allele-6pecific nucleotide 6equence microheterogeneity. A6 a re6ult, PCR-ASO typing methods have been developed. These rely on the generation by PCR amplification of sufficient target DNA to permit typing by AS0 probe6 using ~lot- or dot-blot hybridization analysi6. Thus, PCR-AS0 typing has been u6ed to develop improved procedure6 for HLA-DR typing, HLA-DQ typing and HLA-DP typing.
The ma~or methodological drawback of these ~ystem6 i6 that the complexity of the technique is directly related to the number of alleles under investigation. thu~, at least one AS0 probe is employed per allele, and therefore one membrane containing immobilized target DNAS is required for each AS0 probe used. An alterantive 6trategy has been developed by Scharf et al (1988), whereby immobilized AS0 probe6 are hybridized to enzymatically labelled amplified target DNAs: in this manner, a single membrane containing all of the requisite AS0 probes may be used to type each amplified DNA.
We have now devi6ed a new and 6imple method for HLA-DR/Dw allotype matching which employs PCR ampllfication of ~LA-DRB gene 6econd exon 6equence6 and 6ubsequent product analy6i6 by electrophoretic ~eparation. Rapid 6eparation r can be achieved in nondenaturing polyacrylamide minigels.
In contra~t to currently available DNA typing technology, there are no requirements for po6t-ampllfication 6ample proce6slng 6uch as target DNA denaturation/neutrallzation, immobilization on 601id 6upport membrane6, hybridization with radioi60tope- or enzyme-labelled AS0 probes or development of hybridi6ation signals.
Accordingly, the present invention provide6 a method of human HLA-DR and/or Dw allotype match$ng, which method compri6e6:
(i) effecting PCR amplification of a HLA-DRB gene exon 2 nucleotide 6equence of a fir6t sample of DNA;
(ii) separat$ng according to 6ize the DNA fragment6 re~ulting from the ~aid amplification;
(iii) determining the length polymorphi6m of the thus ~eparated DNA fragment6; and (iv) comparing the thu6-determined length polymorphi6m with the length polymorphi6m of DNA fragment6 resulting from PCR amplification of the 6aid HLA-DRB gene exon 2 nucleotide 6equence of a second sample of DNA.
The method i8 founded on PCR amplification of HLA-DRB
gene exon 2 seguences with defined PCR primer6 to generate character$stic allele-~pecific PCR products (PCR
fingerprint6). It differ~ from other e6tabli6hed DNA typing aethod6 6ince there i~ no requirement for po6t-amplification enzyae dige~tion, chemical treatments, DNA immobilization, target-probe hybridization6, or for the u6e of m~ltiple combin~tion6 of PCR oligonucleotide primers. Both HLA-DR
~nd Dw PCR fingerprint6 can easily be identified in DR/Dw homozygou6 or heterozygous individuals withln ~ hour6.
Hence, the pre~ent method of direct ViBUal compari~on between PCR fingerprint6 of panel6 of individual6 can be adapted for HLA-DR/Dw allotype matching for example in the h ~3 ~ 3 selection of HLA-DR/Dw-matched living related or unrelated volunteer donors for bone marrow transplantation.
When the DNA of HLA-DR/Dw heterozygou6 ind$vidual6 was studied, the ob6erved PCR fingerprint6 were 6imilar, but not identical, to the patterns expected by the 6imple addition of two corresponding HLA-DR/Dw homozygous cells.
However, if the DNAs from two HLA-DR/Dw homozygous cells are premixed before PCR amplification, the pattern observed in the corre6ponding heterozygote i6 reproduced. Therefore the presently described method may be developed for general HLA-DR/Dw allotyping. The DR and/or Dw specificities of heterozygous individuals can therefore be analysed.
The pre6ent method is carried out on a fir6t 6ample of DNA. A sample of DNA is obtained from an individ~al or any object who~e HLA-DR and/or Dw allotypes it i6 wi~hed to study. Individual include6 a foetus. HLA DNA can be extracted from all nucleated celle. Typically, HLA DNA i6 obtained from peripheral blood cell6 for convenience.
Foetal HLA DNA can be obtained from placental cells or amniotic fluid. Other sources of DNA include hair follicles, mummified bodies, etc.
The DNA is i601ated under conditions which preclude degradation. Cell~ are digested with a protease under 6uch condition6 that there is likely to be little or no ~NAase ~ctivity. The digest i5 extracted with ~ DNA 801vent. The extracted DNA may be purified by, for example, dialy6is or chromatogr~phy. Suitable DNA isolation technigues are ,~ ~ 3 ~ ~? ~ ~

described by Kan et al in N. Eng. J. Med. 297, 1080-1084, 1977 and Nature 251, 392-393, 1974 and by ~an and Dozy, Proc. Natl. Acad. Sci. USA 75, 5631-5635, 1978.
Exon 2 nucleotide sequence6 of the HLA-DRB gene of the ~ample DNA are then amplified by PCR. The flanking 6equence6 of exon 2 Are highly con6erved between DRs allele6. To the HLA DNA are added two oligonucleotide primers for annealing to complementary sequences at either end of exon 2, a heat-6table DNA polymera6e 6uch a6 Taq$, dATP, dCTP, dGTP and dTTP. The DNA i6 denatured, the oligonucleotide primerC anneal to their complementary sequence~ with the 3' ends pointing towards each other and the DNA polymera6e results in extension of the annealed primers and amplification of the segment of DNA defined by the 5' ends of the primers.
The cycle of DNA denaturation, primer annealing and 6ynthe6is of the DNA 6egment defined by the 5' ends of the primer6 i6 repeated as many time6 as i6 nece~6ary to amplify the HBA-DRB DNA unti} sufficient i6 available for 6tep (ii) of the pre6ent method. Amplification may proceed for from 20 to 40 cycles, for example from 25 to 35 cycles.;
Any appropriate oligonucleotide primer6 may be employed. The primer6 may be suitable for amplificatlon of multiple HLA-DR and/or Dw alleles or for ampllfication of 6pecific 6uch alleles. Preferred primer6 for multi-allelic amplification are:

~ ~3 3 ~

GH46 (left~ : CCGGATCC~TCGTGTCCCCA~ACCACG
G~SO (right) : CTCCCCAACCCCGTAGTTGTGTCTGCA

For 6pecific amplification of HLA-DRw8 and -DRW5 (w12), GHSO may be u6ed a6 the rigbt primer and, for the left primer:

PL8/~2: TTCTTGGAGTACTCTACGGG

The primer6 may be labelled for facilit-tlng analy6i6 in 6tep (iii) of the present ~ethod. The primer~ can be labelled with a directly detectable tag, for example a radionuclide 6uch a~ 3 2 p, 3 5 S, l~C or 1 2 5 I, a fluore~cent compound 6uch a6 fluorescein or rhodamine derivatives, an enzyme ~uch a6 a peroxida6e or alkal~ne pho6phata6e) avidin or biotin. The two pr$mer6 ~ay have the ~ame or different labelr.
The fragment6 of amplified DNA, i.e. the product of amplification of the exon 2 nucleotide cequence of the ~A-DRB gene of the 6ampl~ DNA a6 defin~d by the PCR
primer~, are then ~eparated according to cize. Thi6 may bè
achieved by electrophore~ic or by high pre6~ure llqu~d chromatography. The ~eparation i~ effected on a ~ubstrate.
For electrophor~ , thi6 typically 1~ a g-l wh~ch doe~ not denatur- the DNA, ~uch as polyacrylamide gel.
She amplified DNA i6 6eparated on the gel according to the ~ize of each fragment. Electrophore6i6 i6 conducted 203~3 g under conditions which effect a de6ired degree of resolution of fragments. A deqree of resolution that 6eparates fragment6 that differ in 6ize by as little a6 about 10 bp is u6ually sufficient. Size marker~ may al60 be run on the gel to permit e6timation of the 6ize of fragment6.
The size di~tribution, i.e. the re601ution pattern, of the amplified DNA fragment6 will be allele-6pecific.
Thi~ resolution pattern or PCR fingerprint can next be vi6uali6ed. Where a PCR primer has bcen labelled, thi6 label may be revealed. A substrate carrying the ~eparated labelled DNA fragments is contacted with a reagent which detects the presence of the label. Where the PCR primers were not labelled, the substrate bearing the PCR fingerprint may be contacted with ethidium bromide and the DNA fragments vi~ualised under ultraviolet light.
The length polymorphi6m of the DNA fragments is thus determined. This i6 compared with the length polymorphi6m of DNA fragments resulting from PCR amplificat$on of the HLA-DRB exon 2 nucleotide sequence of a second ~ample of DNA. In this way, corre6pondence between the HLA-DR and/or Dw allotypes of the DNA sample6 may be obtained. ;It may be a6certained whether the length polymorphi6m of the DNA
fragment6 obtained, and therefore whether the allotypes of tho two 6ample~ are the L~me or not. If the allotype oE tbe ~econd DNA sample i6 known, the ~llotype of the fir6t DNA
cample may be identified if the allotypes are the same.
~ he length polymorphism in respect of the 6econd ~ 3~

rample of DNA may be obtained using the same conditions as are employed to obtain the length pol~morphi~m in respect of the fir~t DNA ~ample. Typically the same primer~ are used.
PCR amplification need not necessarily be for the same number of cycles or under identical reaction conditions, though. Similarly, 6eparation of the resulting DNA
fragments need not be carried out in an identical fa~hion provided it is possible to a66ess the relative correspondence of the length polymorphi6ms of the DNA
fragments resulting from amplification of each DNA sample.
The 6econd sample of DNA may be analysed according to the pre6ent method simultaneously with or at a different time to analysis of the first DNA 6ample. Indeed, a multiplicity of DNA 6amples may be analysed. Typically the or each ~ample is a 6elected 6ample. Sample6 from ~elected individuals can be analy6ed. The length polymorphism determined for each sample may be held in a computer.
A computer databa6e may therefore be generated containing the ~ize distribution patterns for different sample6, for different D~ and/or Dw ~pecificities or, indeed, for all DR and Dw specificitie6. The different molecular 6izes of the PCR generated fragments for 6amples and/or for the different DR and/or Dw specificitie6 determined for known ~amples could be`keyed into A compute~.
Any unknown DR and/or Dw type8 could therefore be determined ("typed") by comparing the size6 of the fragments generated after PCR of DNA of the ~nknown sample with the reference '` - ll -database. In step (iv) of the pre6ent method, therefore, the length polymorphism in re6pect of the fir6t sample of DNA may be compared with a length polymorphi6m held on a computer databi6e in respect of a 6econd 6ample of DNA.
The PCR fingerprint may be compared with another PCR
fingerprint to determine whether the individuals, whose DNA
ha6 been tested to obtain the two fingerprint6, have matching HLA-DR and Dw allotype6. The present method can therefore be applied to DNA 6amples from two or more individuals. Alternatively, a PCR fingerprint may be compared with a 6tandard fingerprint previously obtained.
Corre6pondence between fingerprints can therefore be determined.
~ he present method can be u6ed to determine whether a donor of a transplant or transfusion and a receipient or proposed recipient of the transplant or transfusion have matching HLA-DR and Dw allotypes. PCR fingerprint6 from the donor and the recipient or propo6ed recipient can be compared. The transplant may be a ti66ue transplant 6uch as a heart, lung, liver or kidney transplant or a bone marrow transplant. The transfu6ion may be a blood tran~f~6ion.
HLA-DR/Dw matching of living related or unrelated donors for allogenic transplantation may therefore be achieved.
Altornatively, the present method ca~ be used in determining paternlty of an individual. By comparing PCR
fingerprint6 obtained for the individual, the individual's mother and the suspected father of the individual, thi6 ! - 12 -determination can be made. Also, the present method can be used in determining whether an indiviaual is susceptible to or has a di6ease associated with HLA-DR and/or Dw allotypes.
Such diseases are reviewed in Immunol. Rev. 70, 1-218, 1983.
The invention also provides a test kit, which kit comprises:
(a) two oligonucleotide primer~ suitable for use in PCR and capable of annealing to complementary sequence6 at respective ends of exon 2 of a H~A-DRB gene; and (b) a control DNA and/or control PCR amplification product.
The primers may be labelled as above. The control DNA and control PCR amplification products are also typically labelled. The kit may further comprise one or more of the following:
- a heat-stable DNA polymerase such as TaqI;
- dATP, dCTP, dGTP and dTTP; and - A database comprising the length polymorphi6ms of DNA
fragments generated by PCR amplification of 6elected DNA
samples.
The length polymorphisms stored in the data~ba6e may be those of DNA samples of known DR and/or Dw specificity.
The database may therefore comprise the PCR fingerprints for known DR and/or Dw allotypes, indeed for all such allotype~
ava~lable. The length polymorphl6ms as reported in Flgures 2, 3 and 5 for DR or Dw allotypes or as reported in Figure for the possible combinations of these allotypes may 2 0 ~ 3 therefore be provided in the database.
The following Example illustrates the invention. In the accompanying drawings:
Figure 1 shows sample temperature during PCR
amplification. One PCR cycle is 6hown. The duration6 of cycl- ~eqment6 (a) to (f) are detailed in the Example.
Figure 2 6hows PCR fingerprints of HLA-DR/Dw homozygous B-lymphobla6toid cell line6 (BLCL6). HLA-DRB
gene exon 2 6equences were amplified using GH46 plu6 GH50 PCR primers. M1 and M2, molecular 6ize markers (M1, 8stEII
digest of bacteriophage lambda; M2, MspI digest of pBR322.
Molecular 6ize shown in base pair6 (bp)). Cell6 6hown are:
Lane 1, BAF (ECACC 87033001); Lane 2, CI (9w0201); Lane 3, BGE (9w1201); Lane 4, RML (10w9016); Lane 5, CAA-O (ECACC
85051626); Lane 6, QBL (ECACC 85022807); Lane 7, BOB-2; Lane 8, TS-10 (9w1005); Lane 9, SSTO (9w1303); Lane 10, LS-40 (9w1403); Lane 11, HAS-15 (9w9902); Lane 12, RAG; Lane 13, J-SIN (T29639: from E. Bidwell); Lane 14, HBS (9w0601); Lane 15, EMJ (9w0606); Lane 16, JTED (9w0603)~ Lane 17, HAG
(9w1802); Lane 18, BRU (9w0901); Lane 19, PIT (9w0704); Lane 20, BH13 (9w1901); Lane 21, KIJ (9w1104); Lane 22,~BAE t ~9w0807); Lane 23, LUY (9w0805); Lane 24, DRB (9w and 10w:
Ninth and Tenth International Histocompatibility Work6hop reference number8; ECACC: European Collection of Anlmal Cell Culture6 reference number). The HL~-DR and Dw allotypes are ~hown: workshop (w) prefixes have been omitted for the 6ake of clarity. Where appropriate, RFLP-defined splits are also 2 ~3~5~ 3 chown .
Figure 3 show6 PCR fingerprints of HLA-DRwl0-positive heterozygous cells. Cells ghown are: Lane 1, CI (9w0201);
Lane 2, C164; Lane 3, BOB-2; Lane 4, C1103; Lane 5, ~RO
(9w0905); Lane 6, R287; Lane 7, BUP (9w0702); Lane 8, R295.
The DR serologic 6pecificities are shown. M, molecular size marker6 (MspI digest of pBR322). Primers used and abbreviations as for Figure 2.
Figure 4 show6 PCR fingerprints of HLA-DR/Dw heterozygous cells. For each of the five groups of cell6 shown, PCR fingerprints of homozygous and heterozygous cells are compared. M, DNAs from two homozygous cells (0.5 ~g of each DNA) mixed before PCR amplification. H, DNA from HLA-DR/Dw heterozygous individuals. Cells shown are: Lane 1, AL (9w0101); Lane 2, BOB-2; Lane 3, AL plus BOB-2; Lane 4, C84; Lane S, CI (9w0201); Lane 6, BOB-2; Lane 7, CI plus BOB-2; Lane 8, C650; Lane 9; CAA-O (ECACC 85051626); Lane 10, BOB-2; Lane 11, CAA-O plus BOB-2; Lane 12, C508; Lane 13, ~RO (9w0505); Lane 14, BOB-2; Lane lS, KRO plus BoB-2;
Lane 16, CS9S; Lane 17, BUP (9w0702); Lane 18, BOB-2; L~ne 19, BUP plu8 BOB-2; Lane 20, C667. The HLA-DR 8ër~1Ogic ~pecificitie6 are shown. Primers used and molecular size marker6 M1 and M2 as for Figure 2~
Figure S ohow~ PCR fingerprlnts of HLA-DRw8-positive cells. Combination8 of PCR primers used were: a, GH46 plus GH50; b, PL8/12 plus GH50; c, GH46 plu6 PL8/12 plus GH50.
Allele-6pecific amplificition of DRw8 and DRw12 i6 shown ~3.~8~

(lanes 2 and 4 respectively): no amplification of other DR
haplotypes was observed (not shown). Cells 6hswn are: Lanes 1 and 2, BAE (9w0807); Lanes 3 and 4, J-SIM (T29639); Lane 5, BOB-2; Lane6 6 and 7, C810; Lane 8, RAG; Lanes 9 and 10, C246; Lane 11, JTED (9w0603); Lanes 12 and 13, C372. The HLA-DR seroloqic specificities are shown. Molecular size markers Ml and M2 as for Figure 2.
Pigure 6 shows the results of regeneration of primary PCR product from 6atellite DNA sequences. (a) Reamplification of primary PCR product: Lane 1, genomic DNA
from the DRl-homozygou6 BLCL, BAF (ECACC 87033001) amplified with GH46 plus GH50 primers; Lanes 2 and 3, purified primary (271 bp) product from lane 1, reamplified without primers ~lane 2) and with GH46 plus GH50 primers (lane 3). (b) Reamplification of satellite DNA ~equences: Lane 4, genomic DNA from the DR3(wl7)-homozygous BLCL, CAA-O (ECACC
85051626) amplified with GH46 plus GH50 primers 6howing primary (271 bp) product (P) and upper (U) and lower (L) satellite DNA sequences; L~nes 5 and 6, purified U and L
satellite DNA, respectively, reamplified with GH46 plu~ GH50 primers. (c) Satellite DNA 6equences are not gene;rated from HLA-DRB cDNA: Lane 7, genomic DNA from the HLA-DR9-homozygous BLCL, D~B, amplified with GH46 plus GH50 primer8; Lane 8, pla8mid DNA (2~g) from a DR9 DRB cDNA clohe ampl$fied with GH46 plus GH50 primers, Lane 9, a8 lane 8 but not amplified. Molecular size markers Ml and M2 as for F$gure 2.

203~3 Figure 7 s~ows the P~R ~ing~rprints o~ Example 2 EXAM.~m~

Ce~ls Well characterized ~LC~s, homozygou~ for DLA-~r/Dw alleles, were obtained ~rom or~ginating laboratories partloipiat~ng ~n t~e N~nth ~nd ~enth International Histocompatibillty Workshopss or where appropriat~, from the Europcnn Co lection of Animal Cell Cultures, Porton Do~n, GB.
~LA-DR~Dw het~rozygous c~ wer~ char~ct~riz-d by, and obtained frem, ~rs. E. Bidwell (United ~ingdo~ Transpl~nt Servi¢e, Bristol, GB). ~L~-DR ~nd DW allotypes were coh~irmed by RFLP typing o~ ~aqI-digested genomic CNA, using short DRB, DQB and DQA c~NA p~obes, as previously describod ~Bidwell, Immunol. ~oday 9, lB-23, 1988~ Bidw~
Tran~plantation 45, 640-646, 1988).

~eaction mixtur~e tlOO~l) contained l~g (for BLC~8) or 2~g ~for ~ DR~w heterozygous c-ll#~ o~ genomic DN~, prepared a3 de w rib-d tBidwell 9~ ~1, 1988), lOm~ Trlo-~Cl, pH 8.3, 1.5~M ~qC~2~ 0.01~ ~w/v) gel4tin, 0.2~ ach o~ dATP, dCTP, dGTP and dTTP, ana 1~ each o~ H~-DRB g-n~ seoond eXon ~' ~le~t) ~nd 3' (rlqht) ollgonucleoti~e prlmer~. ~he primers us~d wer- (a) for multi-allellc ampllfication, G~46 (le~t, nucleotld~ soquence 5'CCGGATCCTTCGTGTCCCCAGACCACG3') (S¢har~

203~ 3 ' et al, ~uman. Immunol. 22, 61-69, 1988J plu~ GH50 ~right, nucleotide sequence 5~CTCCCCAACCCCGTA~GTGTCTGCA3~) (Scharf et al, Human Immunol. 22, 61-69, 1988); (b) for ~pecific amplification of ~LA-DRw8 and DRw5(w12), PL8/12 (left, nucleotide 8eguence 5'TTCTTGGAGTACTCTACGGG3') plus GH50 (right). For combined ~mplification, 1~M eaCh Of GH46, PL8/12 and GHSO were u~ed. Mixture6 were heated at 94C for 5 ~in6, placed on $ce for 2 mins, and 2.5 unit6 of Taq polymera~e (Perkin Elmer Cetus, Norwalk, CTO6859, U.S.A.) werc added. DNA amplification wa6 performed u~ing a programmable cyclic reactor (Ericomp ~nc., San Diego, CA92121, U.S.A.). Amplification cycle parameters were: 60 6ecs at 9qC (Fig l(d)), 90 ~ecs cooling ramp (Pig l(e)), 120 ~ec~ at 65C (Fig l~f)), 70 ~ec~ heating ra~p (Fig l(a)), 120 6ecs at 72C (Fig l(b)) and 160 ~ec6 heating ramp (Fig l(c)). After 35 cycles of amplification, the 72C
heating 6tep wa6 extended by 10 min~.

Pol~yacrylamide gel electroDhore~i6 Aliquot6 of PC~-ampllfied DNA (10~1) were 8ub~ected to electrophore6i~ for 95 min at 200 volt6 $n nonde~a;turinq polyacrylamide minigel~ (12% w/v acylamide:
N,N'-bisaerylamide ~29:1), in 1 x TBE ~89mM Tr$~-boratc, 89mM borle acid, 2mM EDTA pH 8.0)) uslng a proprietary vertical minigel eleetrophore6is system tB$o-Rad Laboratorie~, Richmond, CA94804, U.S.A.). The dimen~ion~ of the formed gel were 82mm~w) x 72mm(h) x 0.75mm~d), 203~3 containing 15 sample slots of dimensions 3mm(w) x lOmm(h) x 0.75mm(d). After electrohoresi6, gels were immersed for 30 mins in 1 x TBE containing ethidium bromide (0.5~g/ml) and DNA was visualized using a 302nm ultraviolet transilluminator. For reamplification experiments, DNA was exc$sed from the polyacrylamide gels and purified by isotachophoresis and éthanol precipitation. Aliquots were reamplified using the same condition6 described above.

RESULTS

PCR fingerprints of BLA-DR/Dw homozyqous cells ~o establish an easy and reproducible molecular method for the determination of HLA-DR and Dw allotype6, we empLoyed the PCR amplification technique to amplify allele-specific ~equences. When the HLA-DRB gene exon 2-specific GH46 and GH50 primer6 were used to amplify DNA from a 6eries of HLA-DR/Dw homozygous LLCLs, allele-specific pattern6 of polymorphic PCR product~ (PCR fingerprint~) could be established that were associated with individual HLA-DR
~erotypes (Fig 2). Moreover, split6 of HLA-DR spe;cificities defined by both alloreactive ~ cells (Dw allotypes) and by DNA-RFLP typing produced further allele-specific PCR
~ingerprint6. For ~ach HLA-DR/Dw speclficity, groups of phenotypically identical BLCLs were examined: each group revealed identical PCR finqerprints (not shown). The PCR
fingerprints con6isted of a primary product (271bp as ~3~
t -- 19 --predicted from a eeries of HLA-DRB gene exon 2 nucleotide sequences) together with multiple higher molecular weight products (hereafter designated satellite DNAs). With the exception of HLA-DRw8 homozygous BLCLs ~ee below), allele-specific multiple satellite DNAs ranging from 290bp to 1000bp were observed. One cluster of satellite DNAS
~650bp - 1000bp) appeared to correlate with the supertypic HLA-DrwS3 serologic specificity, associated with HLA-DR4, DR7 and DR9 ~Fig 2, lanes 7-11, 19-21 and 24).

PCR fingerprints of HLA-DR/Dw heterozyqous cells Experiments were conducted to compare the PCR fingerprints of HLA-DR/Dw homozygous and heterozygous cells. These revealed that the PCR fingerprints of HLA-DR/Dw heterozygous cells could not be entirely predicted from the simple summation of patterns observed in the corresponding HLA-DR/Dw homozygous cells ~Figs 4 and 5~. However, the PCR
fingerprints of the former were exactly reproducible by premixing DNAs from the corresponding HLA-DR/Dw homozygous cells before PCR amplificiation. Fig 4 shows five such representative experiments. Premixing DNAs thus permits the construction of predetermined combinations of HLA-DR/Dw specificities for use as reference DNAs.
Si;~ce no HLA-DRw10 homozygous cells h~ve b~en described, we examined this specificity by analysis of PCR
fingerprints of HLA-DR210 heterozygous individuals. In each case studied, unique PCR fingerprints were observed, though 2 ~ 3 .~ 8 1 ~

no consensus HLA-DRwlo-specific ~atellite DNA band W8S
observed (Fig 3).
The PCR fingerprints of HLA-DRw8 homozygous BLCLs revealed a primary (271bp) product, but no detectable atellite DNAs (Pig 2, lanes 22 and 23) and thu6 represented a potential difficulty in identification of the HLA-DRwB
alleie in HLA-DR/Dw heterozygous lndividual6. We therefore constructed a 5' ~left) PCR primer for the purpose of allele specific amplification. This primer, PLQ/12 (~ee Materials and Methods) is complementary only to HLA-DRw8 and gene exon 2 5'-sequence specificity of PL8/12 on an extended panel (not shown) of yg us and heterozygous cell b and DR25(wl2)-positive cells revealed a primary amPlifiCatin product of 242bp as predicted from the primer annealing location (Fig 5, lanes 2 and 4). ~owever, neither the miSsion of the PL8/12 ~mplification with GH46 plus GHS0 primers affected resulting PCR fingerprints. These revealed, as for all other HLA-DR/Dw heterozygou8 cells examined, unique additional satellite DNA bands not observed in the relevant h~mozygou#
cells (Fig 5, lanes 5-13). This permitted the ~dentification of the HLA-DRw8 specificity in HLA-Dr/Dw without recour6e to ~llele primers.

2 ~

Regeneration of primary PCR products from purified satellite DNA seguences To investigate the possible origin of the satellite DNA
sequences obser~ed, primary (271bp) product bands and ~atellite DNA bands were excised from polyacrylamide gels, purified, and reamplified as described above.
Reamplification of the primary product band from the HLA-DRl homo~ygous cell line BAF failed to regenerate the satellite DNA seguences observed after amplification of genomic DNA
~Fig 6(a)), indic.~ting that these satellites do not arise from HLA-DRB gene exon 2 sequences in isolation from flanking DNA. Reamplification of purified satellite DNA
bands from the HLA-DR3(w17) homozygous cell line CAA-O
showed that, for each band reamplified, the lower molecular weight bands observed in genomic DWA PCR fingerprints were regenerated. Thus, primary product (P) and lower satellite (L) bands were regenerated from the upper satellite (U) band; and the P band was regenerated from the L band (Fig 6(b)). This further indicates that satellite DNA bands might consist of a nested set of seguences, each containing part or all of the P sequence.
To test whether satellite DNA sequences could be generated from a full length HLA-DR~ gene cDNA clone, amplification product6 of genomic DNA and cDNA from DR9 haplotypes were compared. only the P sequence was generated from the cDNA clone (Fig 6(c)), indicating again that ~pecific flanking sequences may be required to generate satellite DNAs.

~ ~ 3 P~

P~R ~ingerprints consi~t of ~ principal (leading~ DNA band ~nd a 6erles of satellite ~trailing) DN~ bandY. ~h~
principal band repre~ent~ ho~odupl~xed alA-DRB qene ~econa exon PC~ produots. The number of indivi~ual ho~odup~exes pre~ent in this band corr~spond~ to the num~er o~ dif~eren~
DRB genes ~hioh contaln amplifiablc ~econd QXOn ~equence~
~nd whlch are present on both haplotypes of an individual.
Satellite DNA bands are heteroduplexes formed by het-rologous as~ociation o~ dif~er-nt HIA-~Rs gene produ¢ts du~ing the l~st annealing ~age of the PCR. The~e hetero~uplexe~ contain ~ariabl~ regions o~ nuol~oti~e ~equence mi~atch~, d~pending upon t~e combination of DXB
alleles ~ithin both haplotyp-s. Regions o~ mi~match oon~er on such heteroduplsxe~ a mole¢ular con~or~ation different ~rom th- proqenitor homoduplexe~. 5hi8 permits the resolution of homo~uplex~s and h-teroauplex~s by nond~naturing polyaoryl~mido g-l elect~ophorQsis. The xtent of nualeotid- ~equence mi~matches bet~een allele~ is haplotyp- ~pecific: thus, t~e nu~bsr ~nd gel ~ob~l$tiet Or h-terodupleXos i- haplotyp~ ~p-oi~lc. ThlR permit~ th~
di~crimin~t1on b-t~--n h4~10types u~in~ PCR ~ln~pr~nting.
For HLA-~R/Dw het-rozygotes, add1tional hetoroduplexQs are ob~erved w~en compar.ed to t~ ~up-rimpoDod PCR ~ingerprinto Or the ~orre~pondlng ~omozygou~ typing c~lls, 6~nc~ tran~
~ociatlon g~nerate~ n~w h~tercduplex conrormations.
, ' 2 ~ 3 ~

Thi6 technique wa~ developed a~ a ~urther application of P~R
fingerprintlng. It permito ~a) r~ wlut~on bQtween similar PCR fingerprints ~hown by cells of dif~er~nt ~LA-DRJ~w ~llotyp-~, and (b) the assignment o~ HLA-DR/Dw allotypes.

1. ~esoluti~n between cimilar PCR ~inq~r~rints Coll~ of differ~nt HIA-~RJDw allotype~ ~ay occasionally ~amonstrate ~imilar PCR ~ingerprint-. Idontity ox differQnce in allotype may be ~onfi~ed by DNA
crossmatching, def$nea as:
(1) the mlxing equiv~lent ~mount~ o~ genomic DNAs ~ro~ th~
two aell~ prior ~o PCR a~plification, o~ (2) mixing ~quivalent amount~ of po~ta~pli~lcation PCR products, denat~r~ng th- mixture, for example at 94-C for 1 minute, allowi~g r-~nne~ling ~o ocour by incubatlon, ~or example at 65-C (2 minute~), and subsequent prim-r xt~n~ion, for example at 72'~ ~9 minutes). ~at~hin~ o~ X~A-DR/Dw allotype3 o~ the coll~ i~ indlcated when no di~ference i6 ob~erved between PCa fingerprints derived from indlvldu~l cell DNA~ and ~rom the ~xture. Di~erenc- botween HLA-DR~Dw allotype- o~ the cqlls 1~ lndlaated whon on~ or mor~ new DNA ~atelllte band~ teroduplexe~), not pre~ont in PCR fing~rprints of ~ith~r Or tho individual aell DNA~, are observed in the PCR fingerprint of the miXtUre of cell DNAs (th~ DNA cros6matchJ.

2.
A~ignment of H~-DR/Dw ~llotypes m~y be ~Ghieve~ by a serie~ of DNA cros~match~s, between a Test DNA o~ unknown ~LA-DR/Dw type, and a panel of homozy~ous typing cells (HTCs) of known ind~vidual ~C DNA/Te~t DNA crossmatch and that of the Te6t DNA ("a negative DNA cros~matchn) indicates tnat t~ ~est DNA contai~ the haplotype represented by the HTC DNA. Di6parity of PCR fin~erprints between an individual ~TC DNA/T~st DNA croesmatch and that o~ ~he ~st DNA t~ ~po~itiv~ DNA cro~match") indicates that th~ T~t DNA doe~ not contain ~he haplotyp~ representQd by the ~TC
D~A. An HLA-DR~Dw neterozygou~ ~es~ DNA will givo negat~ve ~A oros~matches with two diSforent ~TC D~As.

3. HLA-DR/Dw Cro~matchinq Xit ~n H~A-DR~Dw Cros~matohing ~it co~priP,Q~ an oligonucleotide prim-r ot for ~pli~ying HLA-DRB gone ~econd xon ~Qguenoe6~ ~2) bu~orB ~or th~ poly~-r~ chain re~t$on ~F~CRI ~pli$1catlon and ~or gel load$ny; ~nd ~3) A
6et of geno~io ~NA6 from ~ ~et o~ oharactQri~ed HLA-DR/Dw homozygou~ typ~ng cell~. Instruotion- for por~or~ing g~nomic D~A i~olst~on, ~uri~ioation snd a~ay, PCR

.

2 0 ~

amplificatlon, DNA cros~match~ng, and analys$s of results, are a'so typically provided. DNA cro~smatahing permits the typin~ of HLA-DR/~w allotypes by ~imple ~i~ual co~pari~on o~
PCR f~ngerprint~ created either by ~1) the mixed amplification of test D~A With a panel of separat~
homozyqo~s typing cell (HTC) DNAs, or by (2) post PCR-am~lification ~ixing of test and H~C PCR products, rollowed ~y denaturation and reannealing.
~a~
A~thoug~ PC~ ~ingerprint~ appear to be uniq~e for each comblnation of DR haplotypes so far examined, tho~e fro~
cert~in unrelatad DR haplotype~ ~ay d~onstrat- ~arked 5imilaritie8. ~or exampl~, the PCR fing~rprints o~ HLA
DR2(w~5~-Dw2 and D~w6(W13)-Dw18 homozygous cells ara very similar ~F~g ~a, lane~ 1 and 2 respectively). In order to Qas~ly discriminate between suah speci~icltie~, we devised exper~mental ~pLking o~ PCR mixtures be~ore ampli~ication with genomic D~A ~rom a DRw8-homozygous typing cell. Since the D~w8 haplotype contains only one ampli~l~ble DRB gene ~co~d ~xon, wh~ch di~f-r~ ln sequence ~rom all other al~les~ m~x-d Amplification o~ genomic DNA ~rom a DRw8-homozygouo oell w1th that ~rom othe~ DR ~p-cifiai~ie~
resulted in the format~on o~ neW heteroduplexeo. The aonfo~a~ons of these new h~teroduplexe~ ~llow-d discri~ination between oth-rwise si~ilar PCR ~inqerprlnts.
~hus in Fig 7a, spiking o~ D~2~w15)-Dw2 And DRw6(W13)-DW18 genomic DNA~ w~th DRw8 genomiC DNA be~ora ampli~lcation gave ris~ to ne~ and a~ily dlstingui-habl- PCR ~ingerpr$nt~
~lane~ 3 and 4, r-~pectlv-ly). ~hls t-ohn~quo m~y al~o b-applied with eqy~l er~ect ¢no~ shown) by ~lxlng re~peotiv~
PCR produets a~ter a~pllfication, danaturlng, and allowing mixtures to reanneal.

Claims (11)

1. A method of human HLA-DR and/or Dw allotype matching, which method comprises:
(i) effecting polymerase chain reaction (PCR) amplification of a HLA-DRB gene exon 2 nucleotide sequence of a first sample of DNA;
(ii) separating according to size the DNA fragments resulting from the said amplification;
(iii) determining the length polymorphism of the thus separated DNA fragments; and (iv) comparing the thus-determined length polymorphism with the length polymorphism of DNA fragments resulting from PCR amplification of the said HLA-DRB gene exon 2 nucleotide sequence of a second sample of DNA

2. A method according to claim 1, wherein step (i) is effected with two oligonucleotide primers each carrying a radionuclide, fluorescent, enzyme, avidin or biotin label

3. A method according to claim 1 or 2, wherein two oligonucleotide primers having the following nucleotide sequences are employed in step (i):
; and .

4. A method according to any one of the preceding claims, wherein PCR amplification is effected in step (i) for from 25 to 35 cycles.

5. A method according to any one of the preceding claims, wherein step (ii) is effected by electrophoresis or by high pressure liquid chromatography.

6. A method according to claim 5, wherein step (ii) is effected by electrophoresis of the DNA fragments on a non-DNA-denaturing polyacrylamide gel.

7. A method according to any one of the preceding claims, wherein step (i) to (iii) are carried out on a sample of DNA from a donor of a transplant or transfusion and a sample of DNA from a recipient or proposed recipient of the transplant or transfusion and the thus-determined length polymorphisms are compared in step (iv) to ascertain whether the donor and recipient have matching HLA-DR and allotypes.

8. A test kit suitable for use in a method of human HLA-DR and/or DW allotype matching at defined in claim 1, which kit comprises:
(a) two oligonucleotide primers capable of annealing to complementary sequences at respective ends of exon 2 of a HLA-DRB gene; and (b) a control DNA and/or control PCR amplification products.

9. A kit according to claim 7, wherein the primers each carry a radionuclide, fluoresent, enzyme, avidin or biotin label.

10. A kit according to claim 8 or 9, which further comprises a heat-stable DNA polymerase and/or dATP, dCTP, dGTP and dTTP.

11. A kit according to claim 8, comprising:
(1) an oligonucleotide primer set for amplifying HLA-DRB gene second exon sequences;
(2) buffers for PCR amplification and for gel loading, and (3) a set of genomic DNAs from a set of characterised HLA-DR/Dw homozygous typing cells.