HLA-DRBl*Olsubtyping by allele-specific PCR amplification: A sensitive, specific and rapid tecfhque 0. Olerup, H. Zetterquist. HLA-DRBl*OI subtyping by allele-specific PCR amplification: A sensitive, specific and rapid technique. Tissue Antigens 1991: 37: 197-204.

Abstract: The two DR1-associated cellular specificities Dwl and Dw20, as well as DRBr’ (Dw’BON’), cannot be unequivocally assigned by serological typing or restriction fragment length polymorphism (RFLP) analysis. We have developed and compared two polymerase chain reactionbased (PCR) typing methods for distinguishing these DRBl alleles; allelespecific amplification of DRBI*OI alleles followed by an agarose gel electrophoresis detection step and group-specific DRBl*Ol amplification followed by hybridization with sequence-specificoligonucleotide probes. The two typing strategies gave completely concordant results in the 33 DRBl *Ol-positive and the 46 DRBl*Ol-negativeindividuals and cell lines studied. No false-negative or false-positive typing results were obtained. All possible heterozygous combinations of the DRB1*0101-0103 alleles could be distinguished by both typing methods. DRBPOI subtyping by allele-specificPCR amplification was performed in less than 3 hours, including PCR amplification, detection and interpretation steps. The technique will be a valuable complement to DR typing by serology and RFLP analysis. Allele-specificDRBl amplifications or group-specific amplifications followed by directed allele-specific amplifications of DRBI alleles, typing based on the absence or presence of amplified products, may well prove to be the technical innovation that will firmly establish PCR-based I DR typing in routine clinical tissue typing.

Introduction Most expressed HLA loci exhibit a remarkable degree of allelic polymorphism, which derives from sequence differences predominantly localized to discrete hypervariable regions of the amino-terminal domain of the molecule. Allelic HLA variability determines the immune response phenotype of an individual by influencing the self-adjustment of the T-cell repertoire during thymic maturation and the presentation of antigenic peptides. In the elucidation of HLA polymorphism, serological, cellular and biochemical typing methods have been at least partially replaced by indirect [restriction fragment length polymorphism (RFLP) analysis] and direct [hybridizing DNA amplified by the polymerase chain reaction with sequence-specific oligonucleotide probes (PCR-SSO) and sequencing] techniques to detect genetic variability. Precise definition of HLA alleles is of importance in population studies and in HLA and disease association studies, as well as in clinical transplantation.

Olle Olerup and Henrik Zetterquist Center for BioTechnology, Karolinska Institute, NOVUM, Huddinge and Department of Clinical Immunology, Karolinska Institute at Huddinge Hospital, Huddinge, Sweden.

Key words: allele-specific amplification - group specific amplification - histocompatibility testing - HLA-DR antigens - polymerase chain reaction - restriction fragment length polymorphism Received 20 December 1990, revised, accepted for publication 26 February 1991

The DRBr’ (Dw’BON’) specificity was initially characterized serologically as DQw 1 -positive and negative with all known D R alloantisera (1, 2). DR’Br’ has subsequently also been found in association with DQw7 (3). DR1 and DRBr’ are indistinguishable by RFLP analysis (4), but may be separated by group-specific PCR-RFLP typing (5). In most cases, serological typing combined with RFLP analysis will enable both identification of DR’Br’ and discrimination between D R l and DRBr’. However, good serological D R typings are in many instances not possible to perform due to, e.g., low number or low quality of B cells or decreased expression of HLA class I1 antigens. Most DRI-positive Caucasians are Dwl-positive, whereas Dwl and Dw20 are represented almost equally in DRl-positive black Americans (6). Dwl and Dw20 cannot be separated by TaqI RFLP analysis (7). Thus, a genomic typing technique unequivocally identifying the three DRBl*OI alleles would be desirable. The second exon of DRBl*OIOI (Dwl), 197

Olerup & Zetterquist

DRB1*0102 (Dw20) and DRBI*0103 (Dr’Br’) genes do not contain allele-specific sequence motifs (6, 8, 9). Consequently, PCR-SSO typing in which all DRB alleles are amplified may be difficult to interpret, especially in heterozygotes (10). The aim of the present study was to develop and compare two PCR-based typing strategies for distinguishing DRBI*0101-0103; allele-specific amplification of DRBl*OI alleles visualized by ethidium bromidestained agarose minigel electrophoresis and groupspecific DRBl*OI amplification followed by hybridization with SSOs. Material and Methods Study subjects

Ten DR1-positive Northern Europeans (determined by serological typing and TuqI RFLP analysis), 8 DRBr’-positive Englishmen (determined by TuqI RFLP analysis and PCR-SSO typing) and 12 West Africans with the DRl,DQw5-associated Tug1 DRB, DQA and DQB RFLP patterns were included in the study. (One individual was DRl/ DRBr‘ heterozygous.) Twenty-three DR1-negative Swedes representing DR specificities DR2-wl7 (determined by serological typing and/or TuqI RFLP analysis) and 2 DRl-negative, DRwl8-positive (determined by serological typing and TuqI RFLP analysis) West Africans were used as amplification and specificity controls. Cell lines

One Dwl (DRBI*OIOI, SA), one Dw20 (DRB1*0102, MZ070782) and one Dw’BON’ (DRBI *0103, Rain-Ter) homozygous cell line were included in the study. In addition, 21 DRBl*OInegative homozygous cell lines of the 10th International Histocompatibility Workshop were used as amplification and specificity controls. These cell lines represented the following DRBI alleles: 0301, 0302, 0401-0406, 0701, 0702, 0801-0803, 0901, 1101, 1201, 1301, 1302, 1401, 1402, 1501, 1502, 1601 and 1602.

HLA-DR serological typing HLA-DR antigens were determined by the microlymphocytotoxicity technique using locally and commercially bbtained antisera. Nine of the DR1positive persons, all DRBr’-positive Englishmen, 3 of the West Africans and 18 of the DRl-negative Swedes were typed by serology. Serological typings of the DRBr’-positive individuals were performed at the United Kingdom Transplant Service in Bristo1 and the Anthony Nolan Research Centre in Great Britain. 198

Southern blot analysis

DNA was extracted from peripheral blood leukocytes by phenol/chloroform extraction of proteinase-K treated nuclei in microscale. Restriction enzyme digests, agarose gel electrophoresis, capillary blotting of DNA fragments onto nylon membranes, labeling of purified probe inserts, stringency washes and autoradiography were performed according to standard techniques with minor modifications described in ref. (1 l). Taq I DRB, DQA and DQB RFLP typing

Allelic TuqI DRB, DQA and DQB restriction fragment patterns were analyzed as previously described (7, 12, 13). To avoid local RFLP nomenclature, the serologically defined specificities associated with the different allelic Tug1 RFLP patterns are given in the text (7, 12, 13). All individuals included in the study were typed by TuqI RFLP analysis. Amplification primers

The 22mer 5’TTG TGG CAG CTT AAG TTT GAA T3‘(5’A), matching a unique shared sequence motif (codons 8-15) of DRBI*OI specificities, was used as 5’-primer for allele- and group-specific amplifications of the second exon of DRBI*OI alleles. For allele-specific DRBI *01 amplification four 3‘primers were used: the 20mer ”ACT GTG AAG CTC TCA CCA AC3’(3’A), complementary to the sense strands of codons 85-91 of the DRBI*OIOI and DRBI*0103 alleles; the 19mer 5’TGT GAA GCT CTC CAC AGC C3’(3’B), complementary to codons 84-90 of the DRBl*0102 allele; the 17mer ”GGC CCG CCT CTG CTC CA3’ (3’C), complementary to codons 68-73 of the DRBI*OIOl and DRBl*0102 alleles and the 17mer 5’CGG CCC GCT CGT CTT CC3’ (3’D), complementary to codons 68-74 of the DRB1*0103 allele (Fig. 1). The melting temperatures (T,) (4 x number of GC bp+2 x number of AT bp) for primers 5’A and 3’A-D were 60°C. For group-specific DRBI*OI amplification the 19mer ”TCG CCG CTG CAC TGT GAA G3‘ (3’E), complementary to codons 88-94 of almost all published DRBl alleles (14), was used as 3’-primer. Allele-specific amplification of DRBl’U1 alleles

For allele-specific DRBl*OI amplifications the 5’primer 5’A and four 3’-primers, 3‘A-3‘D7were used. Each DNA sample was amplified in four tubes all containing the same 5‘-primer, but different 3’primers. The PCR reaction mixture consisted of

HLA-DRBl*Ol subtyping

Figure I. Comparison of the nucleotide sequences of the first domain exon of DRBI*0101-0103 alleles (5, 7, 8), codons 1-20 and codons 66-92. The localizations of primers 5‘A and YA-3’D are indicated by shadowed areas. Primers 3‘A and 3’B generate PCR products of 251 bp, whereas the amplified products obtained by primers 3’C and 3’D are 195 bp.

0.1-0.3 jig genomic DNA (diluted to a concentration of 0.2-0.3 pg/jil), PCR buffer [50 mM KCl, 1.5 mM MgCl,, 10 mM Tris-C1 pH 8.3 and 0.01% (w/v) gelatin], 200 pM each of dATP, dCTP, dGTP and dTTP, 1 pM of each of the primers and 1 unit of AmpliTaq (Perkin-Elmer Cetus Instruments) in a total volume of 25 pl, covered with 50 pl mineral oil. Thirty amplification cycles were carried out in a DNA Thermal Cycler (Perkin-Elmer Cetus Instruments). Each cycle consisted of denaturation at 94°C for 45 seconds, annealing at 67°C (7°C above TJ for 45 s and extension at 72°C for 45 s. A negative amplification control was included in each experiment’. DRBl*OI allele-specific amplification of DNA from all DRBl*Ol-positive and negative individuals and cell lines of the study was performed. Visualization of allele-specific amplification by agarose minigel electrophoresis

Five microliter aliquots of PCR reaction mixtures were loaded on 1% (w/v) ME agarose minigels (SeaKem, FMC BioProducts) stained with ethidium bromide. Gels were run for 15 min at 20 V/cm in 0.5xTBE buffer (89 mM Tris borate/89 mM boric acid/2 mM EDTA, pH 8.0). Gels were examined under UV illumination and documented by photography.

’ Subsequent to the completion of this study, primers were designed to be used as internal positive controls in each PCR reaction. Primers ’TGC CAA GTG GAG CAC CCA A” (complementary to codons 173-179 in the 3’ end of exon 3) and ’GCA TCT TGC TCT GTG CAG AT” (complementary to codons 193-200 in the 5’ end of exon 4) amplify the conserved third intron of DRBI genes giving rise to a 796 bp fragment. The T, is the same for the control primers and the allele-specific primers 5’A and 3’A-YE, i.e. 60°C. The concentrations of the control primers in the amplification mixture are kept 10- to 15fold lower than the concentrations of the allele- and groupspecific primers.

Group-specific amplification of DRB7*07 alleles

For group-specific DRBl*Ol amplifications primers 5’A and 3’E were used. The PCR amplifications were carried out as described above. The specificity of DRBI*OZ group-specific PCR amplification was determined by agarose gel electrophoresis and by hybridizing dot blotted PCR products with a DRB cDNA probe. DNA from all individuals and cell lines were amplified by DRBl*Ol group-specific amplification. Hybridization with oligonucleotide probes

Two microliters of the group-specific PCR reaction mixtures were manually dot blotted onto nylon membranes (Biodyne A, Pall) according to the manufacturer’s protocol. After prehybridization for 30-60 min at 42°C in 10 x Denhardt’s solution (0.02 YOpolyvinylpyrrolidone/O.02YO Ficoll/O .02% bovine serum albumine)/6 x SSC (3 M NaCL/0.3 N sodium citrate, pH 7.0)/0.5% SDS a 32P-3’end labeled oligonucleotide probe was added. Filters were hybridized for 1 h at 42°C with the labeled 3‘primers, 3’A-3‘D7 now used as oligonucleotide probes. One 5-min wash at room temperature in 6 x SSC/O.l% SDS was followed by one 20-min wash at 56°C in 6 x SSC/0.1?4 SDS. Membranes were autoradiographed (X-omat AR5, Kodak) with intensifying screens for 2-24 h. Results Sensitivity

DNA from all the 30 individuals carrying the DR1, DQw5-associated Tag1 DRB, DQA and DQB RFLP patterns and the DRBZ*OIOI-, DRBZ*OI02and DRBI*0103-positive cell lines were amplified by allele-specific DRBI *01 amplification. As 199

Olerup & Zetterquist shown in Fig. 2, the three DRBZ*OZ specificities gave rise to allele-specific amplification patterns: DRBI*OZOl was amplified by primers 3'A and 3'C (Fig. 2 A); DRBZ*OZ02 by primers 3'B and 3'C (Fig. 2 B) and DRBZ*0103 by primers 3'A and 3'D (Fig. 2 C). All 10 DR1-positive Northern Europeans were DRBZ*OZOZ-positive. DNA from the 8 DR'Br'positive (DRBZ*0103) individuals was specifically amplified by primers 3'A and 3'D. Eighty-three percent of the 12 DR1-positive West Africans carried the DRBI*0102 allele, only 2 were DRBI*OZOZ-positive (Table 1). Identical typing results were obtained by allelespecific amplification, visualized by agarose gel electrophoresis, and group-specific amplification, detected by hybridization with SSOs 3'A-3'D, in all the 33 DRBl*OZ-positive individuals and cell lines of the study (Table 1).

A

3 ' - p r i mer

None of the DNAs from the 25 DR1-negative individuals representing the serological specificities DR2-wl8 or from the 21 DRBZ*OZ-negativehomozygous cell lines was amplified by allele-specific DRBl*OZ amplification, i.e. no PCR products were observed on ethidium bromide-stained agarose gels (Table 1). Furthermore, samples from the 46 DRBZ*OZnegative persons and cell lines were not amplified by DRBI*OZ group-specific amplification, detected both by agarose gel electrophoresis and by hybridizing the dot blotted PCR products with a DRB cDNA probe and SSOs 3'A-3'D (Table 1). identification of DRB7'07 heterozygotes

The three heterozygous combinations of DRBI*OZ01-0Z03 alleles gave rise to distinct ampli-

B

A

B

C

D

C

A

D R B 7 '010 1

B

C

D

B

C

D

O R B 1 ' 0 10 110 702

B

C

D

O R 8 1 'Old3

E

A

A

D R B 7 '0702

D

3'-pr i m e r

Specificity

F

A

B

C

D

DRB1'0107/0103

A

B

C

D

D R B 1 '0 1 0 2 / 0 103

Figure 2. PCR-products obtained by HLA-DRBf*01 allele-specific amplification were size-separated by agarose gel electrophoresis and visualized by staining with ethidium bromide. A, DRBI*OfOf homozygous cell line; B, DRBf*Of02 homozygous cell Line; C, DRB1*0103 homozygous cell line; D, premixed DNAs from DRBI*OIOf and DRBf*0102 homozygous cell lines; E, DRBf*OfOl/ Of03 heterozygous individual; F, premixed DNAs from DRBf'Of02 and D R B f *0103 homozygous cell lines. Below each lane, the 3'-primer used is indicated. Sizes of PCR products are given in bp.

200

HLA-DRBl*Ol subtyping Table 1. Results of DRB1'01 subtyping by allele-specific PCR-amplification followed by an agarose gel detection step compared to DRBl O f subtyping by group-specific PCR-amplification followed by hybridization with SSOs in the investigated 55 individuals and 24 homozygous cell lines

+

group-specific PCR

allelespecific PCR gel Individuals and cell lines OR1-positive Northern Europeans (n= 10) DR'Bt-positiveEnglishmen' (n= 8) OR1-positive West Africans (n= 12)

ORB1'0101

DRB1'0102

10 1 2

0

0 10

1

-

ORBi'0103

ORBl '01 01

ORBi'0102

10 1 2

0

0 10

1

-

DRB7'0707-positive cell line DRB7 0702-positive cell line DRBl'0103positive cell line

-

-

-

-

-

DR1-negative individuals (n= 25) DFISI'OI-negative cell lines (n= 21)

0 0

0 0

0 0

0

-

1

+ SSOs ORBl '01 03

1

0

'One individuals was OR1IDR'Br' heterozygous.

fication patterns (Fig. 2D-F), which were easily distinguishable from each other as well as from the homozygous patterns (Fig. 2 A-C). A DRB1*0101/ 0103 heterozygous person was included in the study (Fig. 2 E). When determining the amplification patterns of the other two DRBI*OI heterozygous combinations, equal amounts of DNAs from DRBI*OI-positive homozygous cell lines were premixed and amplified together (Fig. 2 D and F). The 3 different DRBI*OI heterozygotes were also unequivocally identified by hybridizing dot blotted PCR products, obtained by group-specific DRB1*01 amplifications, with SSOs 3'A-3'D. Optimization of DRBl '07 allele-specific amplification

Raising the annealing temperature of the PCR cycle above 67"C, shortening the annealing or extension steps to 30 s or less, or performing less than 30 cycles of DNA amplification resulted in less efficient allele-specific amplification. The primer 3'D was the first to fail to amplify when the amplification conditions were more demanding. Lowering the annealing temperature of the PCR cycle to 66°C or below sometimes gave rise to unspecific amplification of DRBI*OI-negative DNA samples, most often with the GC-rich primer 3'C, as detected by agarose gel electrophoresis. Increasing the number of PCR cycles to 35 or more resulted in unspecific amplification of several DRBZ *OI-positive and -negative samples. When allele-specific DRBZ *01 amplification is used as a complement to DR typing by serology or RFLP analysis the annealing temperature might be lowered to 65°C. This will increase the yield of the PCR reaction, increasing the strength of the ethidium bromide-stained allele-specific bands. The slightly decreased amplification specificity will not be disadvantageous in this situation. However,

when allele-specific DRB1*01 amplification is performed on its own, the annealing temperature should be kept at 67°C to ensure optimal typing specificity. Mixing two 3' primers - giving rise to PCR products of different lengths - in the same reaction mixture was not successful when the two combined primers should amplify the same allele. For example, amplification of DRBl*OIOI-positive samples with primers 3'A and 3'C in the same tube should result in two PCR products of 251 bp and 195 bp, respectively (Fig. 1 and Fig. 2 A). However, only the longer fragment was obtained consistently, making this approach to reduce the number of PCR reactions/sample impossible. Such reduction might be made feasible by redesigning the primers, by varying the relative concentrations of primers, or by changing the composition of the PCR reaction or the profile of the PCR cycle. These modifications have not been tested systematically. Time required for allele-specific DRB7*07 amplification and detection

Typing for DRBl*OI alleles was performed in less than 3 h: setting up the amplification reactions 15 min; allele-specific PCR amplification - 2 h; agarose gel electrophoresis including loading of samples, documentation and interpretation of results - 30 min. Discussion

In the present study primers were designed and the profile of the PCR cycle was adjusted to obtain highly sensitive and specific DRBI*OI allele-specific and group-specific amplifications. The two amplification strategies were tested on DNAs from 79 individuals and cell lines and were found to be applicable for extremely discriminatory DRBl*OI 201

Olerup & Zetterquist subtyping. The methods gave completely concordant results with no false-negative or false-positive typing responses. All possible DRBl*0101-0103 heterozygotes could be distinguished. Allele-specific DRBl*Ol amplification followed by a simple agarose minigel detection step is the method of choice as it requires very little post-amplification processing of samples, compared to group-specific DRBl*Ol amplification followed by hybridization with SSOs or separation of D R l and DR’Br’ by selective PCR-RFLP typing (5). All 10 Northern Europeans in the study who carried the DR17DQw5-associated TaqI DRB, DQA and DQB RFLP patterns were found to be DRBl*OlOI-positive. This indicates that the DR’Br’ specificity may be more rare in Scandinavia than in Western Europe where gene frequencies of DRBr’ of 1 to 2.3% have been observed (2, 9). In black Americans the DRBl*OlOl and DRBl*0102 alleles are approximately equally distributed in DRl-positive individuals (6). Only 17% of the investigated DR1-positive West Africans carried the DRBl*0101 specificity, which is consistent with a less pronounced admixture of Caucasian genes in West Africans compared to black North Americans. Until the last few years, HLA class I1 allelic polymorphism has been identified with serological, cellular and biochemical techniques. However, DNA typing methods are becoming more widely used as an adjunct or alternative to serological typing. RFLP analysis of HLA class I1 genes, especially DR and DQ, has been developed into a powerful and accurate typing technique, with a close correlation between specific RFLPs and expressed allelic variability (7, 12, 13, 15). This wellestablished technique suffers from two major inherent drawbacks: RFLP typing is too time-consuming for clinical use before cadaveric transplantations and it does not directly identify alloreactive epitopes, as most RFLPs are localized to noncoding regions. Most allelic DR, DQ and DP polymorphism is confined to the membrane-distal domain, encoded by the second exon of the respective gene. The flanking sequences of these exons are both locusspecific and highly conserved between alleles. Thus, they are amenable to enzymatic in vitro amplification by the PCR technique (16, 17). The PCR product may be used for genomic typing, directly identifying expressed genetic variabilility. The major drawback of this typing approach is the large number of alleles of several of the class I1 loci and also that alleles share sequence motifs/epitopes in a variety of complex patterns. Thus, many class I1 alleles do not contain unique epitopes, but rather unique combinations of sequence motifs, which 202

may cause difficulties especially when heterozygotes are investigated. The complexity of PCRbased typing may be reduced by amplification of groups of alleles, e.g. DR1-, DR4- and DRw52associated specificities (18-20) or, as shown here, by allele-specific amplification. The PCR product may be analyzed in several ways. (i) Hybridizing membrane-bound amplified DNA with radioisotope- or enzyme-labeled SSOs. (ii) Hybridizing immobilized SSOs with labeled target PCR product - ’reverse dot blot’. This approach has so far only been successfully employed for the DQAl locus (21), but it will most likely also be possible to apply it to other class I1 genes. (iii) In a recently described technique (PCRRFLP), the PCR product is cleaved with multiple restriction enzymes and allele-specific RFLPs may be detected by a simple electrophoresis step (22). The PCR-RFLP method will, in its present form, fail to unequivocally identify many HLA heterozygotes (23). However, the technique might be improved by, e.g., the use of additional restriction enzymes and group-specific amplifications (5, 23, 24). (iv) An interesting method has been described for HLA-DR matching by analysis of PCR product polymorphism (PCR “fingerprints”). In this technique allelic patterns of PCR satellite bands, representing heteroduplexes, are detected by nondenaturing polyacrylamide gel electrophoresis (25). (v) The post-amplification sample processing may be reduced to a minimum by allele-specific amplification, described here for the DRBl*Ol alleles and by Tonai et al. for 23 DRB alleles (26). Allele-specific PCR amplification is based on the principle that a completely matched primer will be more efficiently used in the PCR reaction than a primer with one or several mismatches. Single base pair mismatches should be placed in the 3‘ end of primers as Taq polymerase lacks 3/45’ proofreading exonuclease activity (27). Allele-specific PCR amplification, distinguishing two alleles - ‘wild type’ and ‘mutant’ - has previously been described for the diagnosis of sickle-cell anemia (28, 29), alantitrypsin deficiency (30) and cystic fibrosis (3l). The absence or presence of amplified product has so far mostly been detected by agarose gel electrophoresis (28, 30, 31). By labeling primers at their 5’ end with a fluorescent group, amplification can be detected by UV illumination after removal of unextended primers (29). Wu et al. have suggested a dual labeling system in which one of the oligonucleotide primers in a primer pair is tagged at the 5’ end with a fluorescent group and the other with biotin (28). The amplified biotin-conjugated PCR product may be purified on a streptavidin-agarose column (28) or perhaps more easily by streptavidin-coated magnetic beads. The allele-specific

HLA-DRBl*Ol subtyping primer pair(s) and internal control primers (Footnote 1) may be conjugated with different fluorescent groups. However, it might be difficult to distinguish primer-dimer formation and unspecific amplification from specific amplification with a dual-labeling system. We have so far only used an agarose gel electrophoresis detection step, by which the amplified product is directly visualized and can be discriminated from primer-dimer formation and unspecific amplification. In conclusion, the present study shows that DRBl*Ol subtyping may be accurately and rapidly performed by allele-specific amplification. The technique will be a valuable adjunct to serological DR typing and RFLP analysis. We have very promising preliminary data using this typing strategy for other alleles and groups of alleles2,as well as other HLA class I1 loci. The technique is highly discriminatory and very rapid. Combined with fast DNA preparation methods, genomic typing by allele-specific amplification may be performed in less than 4 h, i.e. approximately the same time as DR typing by serology. The time may be shortened even further by the recently available improved thermal cyclers. The method is technically very simple, is not labor-intensive and is amenable to automation. The typing results are extremely easy to interpret. Furthermore, the technique is cheaper than typing by PCR-SSO and, most important, the cost is almost independent of the number of samples analyzed simultaneously. We believe that allele-specific amplifications of HLA class I1 genes or group-specific amplifications followed by a second step of directed allele-specific amplifications of class I1 genes, typing based on the absence or presence of amplified products requiring minimal post-amplification processing of samples, is the new concept that is going to establish PCR-based genomic typing in routine clinical practice, including donor-recipient matching in cadaveric transplantations. Acknowledgments

We thank Dr Jeffrey L. Bidwell for the generous gift of DNA from 5 DR’Br’-positive individuals. This study was supported by grants from the Swedish Medical Research Council (Projects No 00793, 08890 and 09514). We have recently designed nine 5’ and thirteen 3’ primers which, when used in 16 separate PCR reactions (with identical PCR cycle parameters) per individual, will identify DRBI genetic variability corresponding to the serological specificities DRl-DRwl8 (to be published). This set of primers will also allow the separation of DR7,DQw9 and DR9,DQw9, which is not possible by TaqI DRB-DQA-DQB RFLP analysis. Eight 3‘ primers for identification of the DRCassociated DRBI*04010408 alleles have also been designed.

References 1. Laundy GJ, Middleton D, Klouda PT. A new DQw 1 associated antigen: HLA-DRBr’. Tissue Antigens 1985: 2 6 210-211. 2. Cambom-Thomsen A, Thomsen M, Abbal M, Sommer E, Calot M, Ohayon E. A new HLA-D specificity associated with DR blank: D-BON. Tissue Antigens 1986: 27: 256-261. 3. Middleton D, Savage DA, Bidwell JL, Cullen C, Martin J. Association of HLA-DRBr’ with HLA-DQw3. Tissue Antigens 1988: 31: 79-82. 4. Bidwell JL, Jarrold EA, Laundy GJ, Klouda PT, Bradley BA. Molecular genetics of HLA-DRBr’: Allogenotypes of DRl and DRBr’ are indistinguishable. Tissue Antigens 1986: 27: 99-101. 5. Bignon JD, Cesbron A, Cheneau ML, Bonneville F, Herry P, Muller JY. Selective PCR-RFLP method to distinguish HLA-DRl from HLA-DR-“BR (Dw Bon) alleles: Applications in clinical histocompatibilitytesting. Tissue Antigens 1990: 3 6 171-173. 6. Hurley CK, Ziff BL, Silver J, Gregersen PK, Hartman R, Johnson AH. Polymorphism of the HLA-DRl haplotype in American blacks. Identification of a DR1 P-chain determinant recognition in the mixed lymphocyte reaction. J Immunol 1988: 140: 40194023. 7. Bidwell JL, Bidwell EA, Bradley BA. HLA class I1 genes: typing by DNA analysis. In: Contreras M, ed. Blood Transfusion: The Impact of New Technologies. London: Ballitre, 1990: 355-384. 8. Bell JI, Denney D, Foster L, Belt T, Todd JA, McDevitt HO. Allelic variation in the DR subregion of the human major histocompatibilitycomplex. Proc Natl Acud Sci USA 1987: 84: 6234-8. 9. Bidwell JL, Bidwell EA, Sansom DM, Klouda PT, Bradley BA. The origin of HLA-DRBr’: Exon 2 nucleotide sequence implicates possible gene conversion of DRl by DR4-Dwl0, DR5 or DRw6-Dw18. Human Imrnunol 1989: 2 6 191-197. 10. Bidwell EA, Bidwell JL, Klouda PT. Discrimination between HLA-DRl and DRBr using sequence-specificoligonucleotide probes. Abstract. 3rd European Histocompatibility Conference, Strasbourg, France, March 21-23, 1990. 11. Olerup 0, Hillert J, Fredrikson S, et al. Primarily chronic progressive and relapsing/remittingmultiple sclerosis: Two immunogenetically distinct disease entities. Proc Nut1 Acad Sci USA 1989: 86: 7113-7117. 12. Carlsson B, Wallin J, Bohme J, Moller E. HLA-DR-DQ haplotypes defined by restriction fragment length analysis: Correlation to serology. Human Immunol1987: 20: 95-113. 13. Bidwell JL, Bidwell EA, Savage DA, Middleton D, Klouda PT, Bradley BA. A DNA-RFLP typing system which positively identifies serologically well-defined and ill-defined HLA-DR and -DQ alleles, including DRwlO. Transplantation 1988: 45: 640-646. 14. Marsh SGE, Bodmer JG. HLA-DRB nucleotide sequences, 1990. Zmmunogenetics 1990: 31: 141-144. 15. Simons MJ, Wheeler R, Cohen D, Lalouel J-M, Dupont B. Restriction fragment length polymorphism of HLA genes: Summary of the Tenth International Workshop Southern blot analysis. In: Dupont B, ed. Immunobiology of HLA, Vol. I. New York, Springer-Verlag, 1989: 959-1023. 16. Saiki RK, Bugawan TL, Horn GT, Mullis KB, Erlich HA. Analysis of enzymatically amplified P-globin and HLADQa DNA with allele-specificoligonucleotide probes. Nulure 1986: 324: 163-166. 17. Saiki RK, Gelfand DH, Stoffel S, et al. Primer-directed enzymatic amplification of DNA with a thermostable DNA polymerase. Science 1988: 238: 487-489. 18. Lanchbury JSS, Hall MA, Welsh KI, Panayi GS. Sequence

203

Olerup & Zetterquist analysis of HLA-DR4Bl subtypes: Additional first domain variability is detected by oligonucleotide hybridization and nucleotide sequencing. Human Immunoll990: 27: 136-144. 19. Gao X, Fernandez-Vina M, Shumway W, Stastny P. DNA typing for class I1 HLA antigens with allele-specific or group-specific amplification. I. Typing for subsets of HLADR4. Human Imrnunoll990: 27: 40-50. 20. Fernandez-Vina M, Shumway W, Stastny P. DNA typing for class I1 HLA antigens with allele-specific or groupspecific amplification. 11. Typing for alleles of the DRw52associated group. Human Imrnunol 1990: 28: 51-64. 21. Saiki RK, Walsh PS, Levenson CH, Erlich HA. Genetic analysis of amplified DNA with immobilized sequencespecific oligonucleotide probes. Proc Natl Acad Sci USA 1989: 86: 623M234. 22. Uryu N, Maeda M, Ota M, Tsuji K, Inoko H. A simple and rapid method for HLA-DRB and -DQB typing by digestion of PCR-amplified DNA with allele specific restriction endonucleases. Tissue Antigens 1990: 35: 20-3 1. 23. Olerup 0. HLA class I1 typing by digestion of PCR-amplified DNA with allele-specificrestriction endonucleases will fail to unequivocally identify the genotypes of many homozygous and heterozygous individuals. Tissue Antigens 1990: 36: 83-87. 24. Inoko H. PCR-RFLP methods holds great promise for complete HLA class I1 genotyping. Tissue Antigens 1990: 36 88-92. 25. Bidwell JL, Hui KM. Human HLA-DR/Dw allotype matching by analysis of HLA-DRB gene PCR product polymorphism (PCR “fingerprints”). Technique - a Journal of Methods in Cell and Molecular Biology 1990: 2: 93-100.

26. Tonai RJ, Geer L, Wang Y, Park MS, Terasaki PI. Oligonucleotide typing of HLA DR by allele-specificPCR amplification. Abstract, 16th Annual Meeting of the American Society for Histocompatibility and Immunogenetics, Los Angeles, California, USA, November 33-7, 1990. 27. Fugger L, Morling N, Ryder LP, 0dum N, Svejgaard A. Technical aspects of typing for HLA-DP alleles using allelespecific DNA in vitro amplification and sequence-specific oligonucleotide probes. J Irnmunol Methodr 1990: 129 175-185. 28. Wu DY, Ugozzoli L, Pal BK, Wallace RB. Allele-specific enzymatic amplification of P-globin genomic DNA for diagnosis of sickle cell anemia. Proc Natl Acad Sci USA 1989: 86 2757-2760. 29. Chehab FF, Kan YW. Detection of sickle cell anaemia mutation by colour DNA amplification. Lancet 1990 335: 15-17. 30. Newton CR, Graham A, Heptinstall LE, et al. Analysis of any point mutation in DNA. The amplification refractory mutation system (ARMS). Nucleic Acidr Res 1989: 17: 2503-2516. 31. Ballabio A, Gibbs RA, Caskey CT. PCR test for cystic fibrosis deletion. Nature 1990: 343: 220. Address: OIIe Olerup, M.D. Center for BioTechnology Karolinska Institute NOWM S-141 52 Huddinge Sweden