HLA-DPB1 allele mismatches between unrelated HLA-A,B,C,DR (generic) DQA1identical unrelated individuals with unreactive MLC M, Salazar, I. Yunis, S. M. Alosco, M. Chopek, E. J. Yunis. HLA-DPB1 allele mismatches between unrelatedPHLA-A,B,C,DR (generic) DQA1identical unrelated individuals with unreactive MLC. Tissue Antigens 1992: 39: 203-208. Abstract: We have used a PCR-RFLP method with one generic amplification of HLA-DPB1 second exon and 6 endonucleases to differentiate the 19 HLA-DPBl alleles and 171 heterozygous combinations. The set of primers used in our studies produced fragment sizes different from those published before (1). The HLA-DPBl alleles in Caucasians showed a higher frequency of DPB1*0401 and DPB1*0402, when compared to a small group of Colombians who showed a higher frequency of DPB1*0402 and DPB1*0201. We found three HLA-DPB1 alleles associated with two HLA haplotypes that result from non-random association of alleles: DPB1*0401 with HLA-A26, B38, DR4, DQA1*0301 and DPB1*0101 and DPB1*0401 with HLA-A1, B8, DR3, DQA1*0501. We also report that 70% of combinations between HLA (generic A,B,C,DR) and DQA1identical MLC-unreactive cell mixtures showed HLA-DPBl mismatches, suggesting that HLA-DPBl differences are not important in MLC reacI tivity.

Introduction

HLA-DPB class I1 antigens were first described in 1980 (2), but due to intrinsic difficulties in its typing, low level of expression and lack of serological reagents, its role in autoimmunity and alloreactivity could not be clearly determined. With the development of the PCR technique, the field of study of the DPBl locus was open. Soon it was evident that HLA-DPB1 locus was much more polymorphic than was determined by PLT typing: 20 alleles versus only 6 specificities.Now several reports indicate that the HLA-DPB1 alleles can be associated with autoimmune diseases (3,4), and the generation of GvHD in bone marrow transplantation is influenced by HLA-DPB1 differences (5). In 1984, it was reported that HLA-DPBI differences were abIe to induce an allogeneic response in mixed lymphocyteculture (6). In 1990, it was reported that the predictive value between DP compatibility / incompatibility as defined by RFLP analysis and reactivity in MLC in HLA, A,B,C,DR and DQ-compatible pairs was higher than 99% (7). Recent reports showed that some HLA-DPB1 incompatibilities as determined by oligonucleotide typing did not induce alloreactiv-

M. Salazarl2, I. Yunis3, S. M. Alosco3, M. ChopekS and E. J. Yunis’wza ‘Dana Farber Cancer Institute, Boston, ’Center for Blood Research, Boston. and ’American Red Cross, Dedham, Mass. U.S.A.

Key words: HLA-DPB1 - MHC - MU: RFLP

- PCR-

Received 7 November, accejtea for publication 10 December 1991

ity (8,9). In this study, we typed HLA-DPB1 alleles in 16 unrelated HLA-identical (generic) individuals whose MLC were unreactive, and found that 70% of MLC combinations presented HLA-DPB 1 mismatches. Our studies also suggested that there is association of HLA-DPB1 alleles with two HLA high delta haplotypes. DPB1*0101 with HLA Al, B8, DR3 DQA1*0501 and the HLA A26, B38, DR4 DQAl*O301 with DPB1*0401. Material and Methods DNA samples

DNA from unrelated individuals was from the American Red Cross DNA repository, all samples having been typed by serology and RFLP for DR and DQw specificities and with PCR-RFLP method for DQAl alleles (10). DNA from homozygous typing cells from the Tenth Workshop panel were used as controls (1 1). DNA extraction

DNA was extracted from EBV-transformed cell lines, or from 10 ml whole blood collected on EDTA by the salting-out procedure (12). 203

Salazar et al. Table 1. HLA-DPB AFLPs allele patterns Dde I

BStU I

2 1 5 9 8 5 7 8 1 7

I l l I l l I l l I l l

3 1 1 3 9 4 a 4 4

EcoN I

Fok I

3 2 2 1 1 3 8 2 7 1 5 a 4 8 4 0 4

3 2 2 1 3 8 2 1 5 5 a 1 5 3 7 6

I I I I I I I I I

I

I

I

l

I I

I

I

l I I

I I

Sau 961

3 2 2 1 1 3 8 0 4 3 7 6 5 8 1 4 7 4 8 9 7

I I

I

I I

Rsa I

I I I

I

I

I I I I I

I I I I I

I I I I I

I I

I I I I I I I I I I

I I I

I I

I I

I

I

I

I

I I 1 I

I

I

I

I

I

I

I

I

I

I I I I

I I

I I

I I

I I

I I I I I I I I

I

I

I

I

I

# samples

3 2 2 1 2 8 2 0 6 3 6 6 2 4 4 9

I I

I 1

HLA-DPB1 Allele

I

I I I I I I I I I I I I I

I I

I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I 1 I I I I I I I I I I I I I I I

'0801 '0901

1

'1 301

2

'1 601 '1 701

1

'1 901

1

'0401+'1101 *0401+'1501 *0401+'1301 '0101+'1101 *0101+'0401 '0201+'0601 '0301+'0201 '0201+*1401 '0301 +'1401 '0301 +'1701 '0402+'0201 '0402 +'0801 '0402+'1701 '0402+*0301 '0402+'1901 '0501 +'a402 '0201 +'1301 *0201+'1501 '0201 +'0401 '0402 +'1301 '0401 +'lo01 '0401 +'1701 *0401+'0601 '0401+'0301 *0401+'0402 I '0101 +*a201 '01 01 +'1701 '0101 +'a301 '0101 +'1401 I '0101 +'a402 I . '0401+'0501 '0501 +'1301 -0101 +'0501

1

1

1 1 1 1 16 1 5 1

1 1 5 1

1 2 1 2 2 1 14 1 2 4 2 9

15 2 1 2 1

a 2 2 1

Pattern of bands for the 19 HlA-DPBl alleles and heterozygous combinations found in the population studied. The patterns found corresponded exactly to the predicted patterns for those combinations. The sizes of the fragments obtained with each enzyme are indicated in base pairs (bp) at the top of the columns.

The presence of a particular band is noted as 1.

204

HLA-DPBI allele mismatches PCR amplification

'

Genomic DNA (1-2 pg) was amplified by the PCR method (13). DNA samples were amplified in 100 pl of reaction containing 0.125 mM dNTPs, 50 mM KCl, 1.5 mM of MgC12, 10 mM Tris-HC1 pH 8.0, 1 pM of each primer, 2.5 u of AmpliTaq polymerase (Cetus) and 0.01% (w/v) gelatin. The HLA-DPB1 second exon was amplified using the primers described before (3). DNA samples were subjected to 30 PCR cycles in a programmable Thermal Cycler (PE Cetus) using a two steps profile: Denaturation at 96°C for 30 9 and annealing and extension at 72" for 1.5 min. A negative control was always included to detect any contamination: 5-7 pl of amplified product were loaded in a 2% NuSieve, 2% agarose gel to determine band size and amplification-efficiency. Endonuclease digestions and gel electrophoresis

Amplified DNA was divided in 15 p1 aliquots, dispensed on polypropylene 96-well microtiter plates and digested for 3 h with 5 units of BstU I, Dde I, EcoN I, Fok I, Rsa I and Sau96 I (Biolabs MA) at 37°C for a11 of them except for BstU I which was incubated at 60°C. The digestion buffer supplied by the vendor was used at a final concentration of lx. Further correction of the buffer prior

Table 2. HIA-OPE1 allele frequencies in North American Caucasians (N.A.C.) and Colombian Mestizos (C.M) ~~

~

C.M

N.A.C. N=136 OPE1 Alleles DPB1'0101 DPB1'0201 DPBl'O202 DPB1'0301 OPE1'0401 DPB1'0402 DPB1'0501 OPE1'0601 DPBl'O801 OPB1'0901 DPB1'1001 DPB1'1101 DPB1'1301 DPB1'1401 OPE1'1 501 OPE1'1 601 DPBl'1 701 DPBl'1801 DPB1'1901

Total

N-21

#

%

#

%

32

11.8 10.3 0 4.8 47 12.5 4 1.1 0.4 0 1.1 0.4 2.2

3 9

7.1 21.4 0 14.3 16.7 23.8 0 0

28

-

13 128 34 11 3 1

-

3 1 6 2

2

-

7

-

1 272

6 7 10

0 0

-2

0.7

1

0.7 0 2.6 0 0.4

4

100

N=Total number of individuals tested

-

4.8 0 0 2.4 0 0 9.5 0

0 42

100

to the addition of the enzyme was unnecessary. Double distilled water was added to complete the final volume of 20 p1. The heterozygous combinations HLA-DPB1 *0501/0201 was always included as a control for enzyme activity. Digested DNAs were then run on a 4% (2%NuSieve (FMC Bioproducts, ME) 2% LE (Bethesda Research Laboratories, ME)) agarose submarine gel for 3 h 0.5 x TBE, pH 8.0 containing 0.5 pg/ml of ethidium bromide. Resulting patterns were analyzed by measuring band sizes comparing band migration to size markers and recording those that were present in each of the digests. Mixed lymphocyte cultures (ME)

MLC experiments were carried out among HLAA, B, C, DR, DQ-identical unrelated individuals using 3 controls as responders on a checkboard design. Results were calculated as counts per minute (cpm) of incorporated tritium [3H]. Relative response values (RR) were calculated by comparing the median (100%) cpm of lymphocytes from 3-5 unrelated controls (14) with the cpm obtained from the experimental groups. Values of RR below 5% were considered unreactive.

Results HLA-DPBl typing of homozygous and heterozygous individuals by AFLP

Using one generic amplification of the second exon of HLA-DPBI and 6 endonucleases as described before (1) we were able to assign the 19 HLADPBl alleles recognized by the World Health Organization nomenclature committee. As shown in Table I , all the homozygous patterns and the heterozygous patterns could be identified in the cells typed. We theoretically predicted that all 171 heterozygous combinations were different with the exception of *0801/*0202 and *0201I1901 which gave identical patterns with the six enzymes used, but could be distinguished using another enzyme Ban I1 which gives a specific pattern with *0202. We typed 170 unrelated individuals from the ARC repository and the Dana Farber repository, and identified only 52 out of 190 possible homozygous and heterozygous combinations, due to the low frequency of most of the HLA-DPB1 alleles. HLA-DPBl allele frequencies in North American Caucasians (N.A.C) and Colombian Mestizos (C.M)

Table 2 shows the frequencies of HLA-DPB1 alleles in a population of N.A.C and C.M. In the N.A.C population our results were similar to the frequencies reported by other studies (15). HLA205

Salazar et al. DPB1*0401 was by far the most frequent allele with a frequency of 47%. We couldn't find any examples of HLA-DPBI "0901 , DPB1*0202, DPB1*1601 and DPB*1801 alleles. In the Colombian population, although the number of samples was only 20, we found that the most frequent allele was *0402 instead of *0401, and also found an increase in the frequency of *1701 as compared to the N.A.C. We did not find examples of DPBl alleles *0202, *0501, *0601, *0801, *0901, *1101, *1301, *1501, *1601, *1801 and *1901 due to their low frequency and the small size of the population studied.

match and 8 (53.3Y0)showed two incompatibilities at HLA-DPB1 . HLA-DPB1 alleles association with high delta HLA haplotypes

Table 5 shows the HLA-DPBI association with 5 known high delta haplotypes. We found that the A26, B38, DR4 haplotype was associated with HLA-DPB1*0401 allele in a11 the cases studied (6/ 6). The high delta haplotype Al, B8, DR3, which was reported to be associated with HLADPB1*0101 (6), was found to be also associated with both *0101 and *0401 with the same frequency 10/23 (45%). We did not find specific HLADPBl alleles associated with the other 3 high delta haplotypes studied.

HLA-DPB1 allele typing in a group of individuals with nonreactive MLC

We performed HLA-DPBl allele typing in a group of unrelated HLA (generic)-identical individuals who had an MLC negative as determined by a RR < 5%. We studied 16 individuals and 21 MLC reactions which were unreactive. Table 3 shows the HLA-A,-B,-C and -DR generic typings and HLADQAl and HLA-DPBI allele typings for these individuals. As shown in Table 4, 6 out of 21 (29?h) combinations were HLA-DPB 1-matched, while 15/21 were HLA-DPBl-mismatched (71%). In the group of mismatched mixtures, 7 (46.70/) had one mis-

Discussion

PCR-RFLP typing of HLA-DPB1 alleles has been shown to be a good alternative to the use of oligotyping. Although the first report (16) failed to identify most of the alleles due to a lack of known sequences at that time, several other reports have demonstrated that HLA-DPBI typing of AFLP is easy to perform and can identify all HLA-DPBl alleles. The method we used included a set of primers already known (3) and the enzymes described by others (l), but the combined set of

Table 3. HIA-DPB1 allele typings of unrelated HIA A, 6, C DR (generic) OQAl-identical individuals tested Serological typing

#

Allele typing

Sample

A

B

C

DR

DQAl

DPBl

1

f 2

2 2

7.~50 7,w50

w6 w6

wl5.7 w15J

'01 01/0201 '0101/0201

'0401 '0401 1'0402

2

3 4

11,30 11,30

13,35 13,35

w4,w6

w11,7 wl1,7

'0501 . '0501

3

5 6

26,31 26.31

38 38

4

7 8

l,w33 1.~33

5

9 10

24,30 24,30

6

11 12 13 14

1 1 1 1

15 16

12 1.2

MLC group

7

'04011'1701 '0401/'0402

4 4

'0301 '0301

'0401 '0401

3,7 3J

'0501/'0201 '0501 1'0201

'0201 '0201 1'0401

13.38 13,38

w6,7 w6,7

'01 011'0201 '01 011'0201

'0401/*1301 '0402/'1301

8 8 8

3 3 3 3

'0501 '0501 '0501

'0401 '0301 /*0401 '0201 1'1 401 '0401 PO1 01

'0501 '0501

'01 01/'0402

14,w57 14,w57

w6 w6

a

8,w60

8.~60

w3 w3

3,4 3,4

'0501

HLA typing of patients and prospective unrelated bone marrow donors with unreactive MLC reactions. HLA-A,-B,G and -OR were assigned by serology. HLA-DQA1 and HLA-DPBl alleles were defined by PCR-RFLP method.

206

'

'0301 1'0401

HLA-DPB1 allele mismatches Table 4. HLA-OPE1 allele mismatch in unreactive MLC mixtures

MLC group

Responder

Stimulator

HLA-DPB1 Matching

1

040110401 0401/0402

040110402 040110401

mismatch match

2

040111701

040110402

mismatch

3

0401/0401 040110401

0401/040i 0401/0401

match match

4

020110201 020110401

020110401 020110401

mismatch match

5

040111301 0402/1301

0402/1301 0401I1301

mismatch mismatch

6

040110401 0301/a401 0401/0401 020111401 040i/o40i 0401/0101 030110401 020111401 040110101 040i/aioi

030110401 0401/0401 0201/1401 040110401 04oi/ai 01 0401/0401 020111401 0301/0401 030111401 0201M 401

mismatch match mismatch mismatch mismatch match mismatch mismatch mismatch mismatch

a10110402 0301/040i

mismatch mismatch

7

A

0301/0401 010110402

Unreactive MLCs were considered as a RR c 5%. MU: groups corresponded to those described in Table 3. In group 2 the MLC was positive in one

direction.

primers and enzymes used had not been published before. In previous studies using the same primers and different enzymes it was not possible to distinguish DPB 1*0201 from DPB 1* 1601. Now, using Sau96 I instead of BstN I it was possible to solve this ambiguity. We have schematically shown that 6 enzymes can distinguish 19 HLA-DPB1 alleles including the heterozygous combinations with the exception of DPB1*0801/*0202 from *0201/*1901 " combinations. The latter, when found, can be differentiated using the enzyme Ban 11; therefore, the use of PCR-RFLP is a viable alternative to SSO typing especially in HLA-DPB1 wherein the pattern of polymorphism requires may probes which are not allele-specific, making the analysis of the results tedious and time-consuming. The HLA-DPBl allele frequencies in the Cauca< I '

Table 5. HLA-DPBl allele association with high delta HLA haplotypes Haplotypes

DPBl Allele

Frequency

~~

HLA-A1, 68, DR3, DQA1'0501

-0101

*ow HLA-A26, 838, DR4, DQA1'0301

'0401

10 Out of 23 43.50% 10 Out of 23 43.50% 6 out of 6 100%

High delta haplotypes: HL4 A2, 8w62, OR% Al, Bw57, DR7; and A24, 87, DR2 did not show significant association with any specific HLA-DPB1 allele.

sian population found by us were different from published studies (17): for example, 29.3% of DPB1*0201 and 34.3% of DPB1*0402 as compared to ours of 10.3% for *0201 and 12.5% for *0402. The frequencies reported by us were similar to those of others (15). The differences may be due to the fact that one study was not calculated correctly (17). Also, we found that although the HLA-DPB1 locus is very polymorphic based on sequences, it has a limited polymorphism when studying populations (5 alleles account for 86% of the total number in Caucasians). The limited polymorphism in some populations will allow us to design different HLA-DPB1 typing strategies with emphasis on identifying common alleles, for example *0401, the most frequent allele in Caucasians, can be identified using only two enzymes (BstU I+EcoN I). Only four enzymes (Ban 11, Bstu I, EcoN I, Rsa I) can identify all the alleles except 0301 from 0601; 0801 from 1901; 0901 from 1701 and 1101 from 1301, alleles that with the exception of *0301 have a frequency of less than 3% in that population. Our results of HLA-DPB1 typing in patients and unrelated bone marrow donors who are MLCnegative demonstrated that HLA-DPB1 allele incompatibilities in the MLC-negative group were as high as 70%. It was reported before (8) that differences between DPw4 and DPa (DPB*O401 and *0402) were not able to induce MLC proliferation. and that this mismatch accounted for 88% of HLA-DPB1 incompatibilities in MLC-unreactive mixtures. In our data, only 4/15 (26.70/0) YO of HLA-DPBl mismatches were found in mixtures between *0401 and *0402 individuals. In addition, mismatches between these two alleles were able to induce proliferation in otherwise HLA-identical (generic-typed) individuals (data not shown). Analysis of our data could not explain why some HLA-DPB1 alleles failed to produce any stimulation in some cases while in others they did stimulate. These findings may be due to differences at other Class I1 MHC loci. Nevertheless, our results confirmed another published analysis (9) in which it was reported that a number of unreactive MLC mixtures have HLA-DPB 1 disparities. Three HLA-DPBl alleles were found associated with HLA haplotypes known to result from nonrandom association of HLA alleles: DPBl*O101 and DPB1*0401 with HLA, Al, B8, DR3 DQA1*0501; however, we found 10/23 Al, B8, DR3 DQA1*0501 haplotypes associated with DPB 1*040 1, an allele which as a frequency of 47% in the population; and DPB1*0401 with HLA A26, B38, DR4, DQA1*0301. These results are consistent with previous findings of association of D h l with Al, B8, DR3 and DPw4 with A26, B38, DR4 207

Salazar et al. (6). Taken together, our findings will be important for the analysis of the of HLA-DPBl alleles in disease association since Al, B8, DR3 has been found associated with HLA-DPBl*OlOl in Celiac disease (18), a finding that suggest that linkage disequilibrium in Celiac disease extendt into the DP region to include the HLA-DPBl*OlOl allele. Acknowledgements

’

This work was supported by a grant from the American Red Cross and National Institutes of Health grants CA20531 and HL-29583, USA. References Jw,Easteal S. HLA-DP typing by amplified fragment lengh polymorphism (AFLPs). Immunogenetics 1990: 32: 56-59. Shaw S, Johnson AH, Shearer GM. Evidence for a new segregant series of B cell antigens that are encoded in the HLA-D region and that stimulate secondary allogeneic proliferative and cytotoxic responses. J Exp Med 1980: 152: 565 Begovich AB, Bugawan TL, Nepon SB, Klitz W, Nepom GT, Erlich HA. A specific HLA-DPB allele is associated with pauciarticular juvenile rheumatoid arthritis but not adult rheumatoid arthritis. Proc Natl Acad Sci USA 1989: 86: 9489-93. Bugawan TI, Angelini G, Larrick L, Auricchio S, Ferrara GB, Erlich HA. A combination of a particular HLA-DPP allele and an HLA-DQ heterodimer confers susceptibility to coeliac disease. Nature 1989: 339: 470-3. Odum N, Platz P, Jakobsen BK, et al. HLA-DP and bone marrow transplantation: DP-incompatibility and severe acute graft-versus-host disease. Tissue Antigens 1987: 30: 2 13-6. Matsui Y, Alosco SM, Awdeh Z, et al. Linkage disequilibrium of HLA-SBI with the HLA-A1, B8, DR3, SCOl and of HLA-SB4 with the HLA-A26, Bw38, DwlO, DR4, SC21 extended haplotypes. Immunogenetics 1984 20: 623-3 1. Olerup 0, Moller E, Persson U. HLA-DP incompatibilities induce significant proliferation in primary mixed lymphocyte cultures in HLA-A, -B, -DR and -DQ compatible individuals: Implications for allogeneic bone marrow transplantation. Tissue Antigens 1990: 36.194-202.

1. Dekker

2.

3.

4.

5.

6.

7.

208

8. Al-Daccak R, Loiseau P, Miramont P, Rabian C, Raffoow C, Colombani J. Evaluation of HLA-Class I1 identity between unrelated individuals by serological typing, DNARFLP method and mixed lymphocyte reaction. Hum Immuno1 1990: 29 189-201. 9. Tennijtelen A, Erlich HA, Braun LA, et al. Oligonucleotide typing is a perfect tool to identify antigens stimulatory in the Mixed Lymphocyte Culture. Hum Immunol 1991: 31: 241-5. 10. Yunis I, Salazar M, Alosco S, Gomez N, Yunis EJ. HLADQAl and MLC among HLA (generic) identical unrelated individuals. Tissue Antigens (in press). 1 1. Yang SY, Milford E, Hammerling V, Dupont B. Description of the reference panel of B-lymphoblastoid cell lines for factors of the HLA system: The B cell line panel designated for the 10th International Histocompatibility Workshop. In: Dupont B, ed. Immunobiology of HLA, vol 1. Histocompatibility testing, 1987. Berlin: Springer, 1988: 11-19. 12. Miller SA, Dykes DD, Polesky HF. A simple salting out procedure for extracting DNA from human nucleated cells. Nucleic Aciak Research 1988: 1 6 1215. 13. Mullis KB, Faloona F. Specific synthesis of DNA in vitro via a polymerase-catalyzedchain reaction. Methoah Enzym01 1987: 155: 335-50. 14. Dupont B, Hansen JA, Yunis El. Human mixed-lymphocyte culture reaction: Genetics, specificity and biological implications. Adv Immunol 1976: 23: 107-202. 15. Al-Daccak R, Wang FQ, Theophille D, Lethielleux P, Colombani J, Loisea P. Gene polymorphism of HLA-DPBI and DPAl Loci in Caucasoid population: Frequencies and DPBI-DPAl Associations. Hum Immunoll991: 31: 277-85. 16. Maeda M, Uryu N, Murayama N, et al. A simple and rapid method for HLA-DP genotyping by digestion of PCRamplified DNA with allele specific restriction endonucleases. Hum Immunol 1990: 27: 11 1-21. 17. Fernandez Vina M.Morales E, Stastny P. DNA typing for Class I1 HLA antigens with allele-specific or group-specific amplification. Hum Immunol 1991: 3 0 60-68. 18. Hall MA, Lanchbury JS, Bolsover WJ, Welsh KI, Ciclitira PJ. Celiac Disease is associated with an extended HLADR3 haplotype which includes HLA-DPwl. Hum Immunol 1990: 27: 220-8. Address: Dr. Marcela Salaiar. M.D. Division of Immunogenetics Dana Farber Cancer Institute 44 Binney Street Boston, MA 02115 U.S.A.