TAP2 polymorphisms in Australian multiple sclerosis patients

Journal of Neuroimmunology ELSEVIER

Journal

of Neuroimmunology

59 (1995) 113-121

TAP2 polymorphisms in Australian multiple sclerosis patients Bruce H. Bennetts a,*, Suzy M. Teutsch a, Robert N.S. Heard b, Heather Dunckley ‘, Graeme J. Stewart a a Department of Immunology, Westmead Hospital, Westmead, Sydney, NS W 2145, Australia b Department of Medicine, University of Sydney, Sydney, NS W 2006, Australia ’ NS W Red Cross Tissue Typing Laboratory, Sydney, NS W 2000, Australia Received

8 June 1994; revised 9 February

1995; accepted

9 February

1995

Abstract Polymorphism of the TAP2 gene locus, situated approximately 150 kb centromeric to the MHC class II loci HLA-DR, DQ was examined in 100 Australian patients with relapsing/remitting multiple sclerosis (MS), in 100 random controls and in 37 selected HLA-DRBl* 1501-positive controls. The results were correlated with HLA class I and class II phenotypes. TAP2 encodes a protein involved in the transport and presentation of antigenic peptides by MHC class I molecules and hence is a candidate locus for a putative MS susceptibility gene either through functional interactions with class I alleles or as an explanation, via linkage disequilibrium (LD), for the known association between MS and the alleles DRBl* 1501, DQAl * 0102, DQBl* 0602. Strong LD was found between the allele TAP2 * 01 and DRBl* 1501 in both the MS and control populations. The MS-associated haplotype can therefore be extended to DRBl* 1501, DQAl* 0102, DQBl* 0602, TAP2 * 01, and the putative gene locus could reside on the centromeric side of DQ. TAP2 typing, however, could not explain the DRBl* 1501, DQAl* 0102, DQBl* 0602-negative patients in whom, interestingly, the frequency of TAP2 * 01 was decreased compared to controls. The results of this study exclude TAP2 as a locus for a necessary MS/MHC gene but indicate that an MS gene carried by the DRBl* 1501, DQAl * 0102, DQBl* 0602 haplotype could reside centromeric of DQ. Keywords:

Linkage disequilibrium;

Multiple sclerosis; TAP2

1. Introduction

Multiple sclerosis (MS) is associated with the major histocompatibility complex (MHC), haplotype DR2 (Dw2), DQw6 (Olerup and Hillert, 1991) in Caucasian populations in several countries throughout the world. This association has been further defined at the genetic level, mapping to the class II alleles DRBl * 1501, DQAl* 0102, DQBl * 0602. However, the DRBl* 1501, DQAl* 0102 and DQBl* 0602 alleles are found in 54% of Australian (G.J. Stewart, B.H. Bennetts, S.M. Teutsch, M. Castle and R.N.S. Heard, submitted for publication) and 72% of Norwegian MS patients (Spurkland et al., 1991) compared with 17% of the -* Corresponding +61 (2) 891 3889

author.

Elsevier Science B.V. SSDI 0165-5728(95)00033-X

Phone

+ 61 (2) 633 6790 or 633 6791; Fax

Australian and 33% of the Norwegian control populations. An adequate explanation for the incomplete association between MS and these DR and DQ alleles in Caucasian populations, and the lack of association in other ethnic groups (Marrosu et al., 1988) has not yet been found. On the one hand, there may be no need for a necessary MS/MHC susceptibility gene to account for these incomplete associations. The lack of HLA haplotype sharing in siblings concordant for MS can be interpreted as evidence against a necessary gene within the MHC region (Ebers et al., 1982). However, two possibilities consistent with a necessary MS/MHC region gene hypothesis have been considered. First, that a critical MHC sequence or residue occurs on a MHC class II structural gene on several haplotypes including DRBl* 1501, DQAl * 0102, DBQBl * 0602. Second, that the susceptibility gene is separate from, but in linkage disequilibrium with, the currently de-

114

B.H. Bennetts et al. /Journal

of Neuroimmunology

59 (1995) 113-121

conserved compared with the rat alleles, which can vary by up to 25 amino acids. There have been to date a number of published studies investigating TAP associations with autoimmune diseases. Whilst a primary association with a TAP2 allele was not detected in studies of rheumatoid arthritis (Wordsworth et al., 19931, coeliac disease (Colonna et al., 1992; Powis et al., 1993b; Tighe et al., 19941, or juvenile rheumatoid arthritis (Ploski et al., 19941, Caillat-Zucman et al. (1993) reported a dominant protective effect associated with the TAP2 * 0201 allele in patients with insulin-dependent diabetes mellitus (IDDM). This protection was independent of, and additive to, the protection conferred by the DRBl* 02, DQBl* 0602 haplotype and antagonistic to the predisposing effect of the DRBl* 03, DQBl* 0201 and DRBl * 04, DQBl * 0302 haplotypes. Susceptibility appeared to be conferred by homozygosity for the TAP2 * 0101 allele. There are now several published studies investigating associations between MS and TAP genes. These studies looked at Swedish (Liblau et al., 19931, English (Kellar-Wood et al., 1994), Norwegian (Spurkland et al., 1994) and Northern Irish (Middleton et al., 1994) populations. None of these reports have identified any significant association between MS and TAP2 alleles. We have chosen to investigate whether there is an association between TAP2 polymorphisms and MS in an Australian population. In looking for associations, we stratified our subjects according to whether they were positive or negative for a number of HLA alleles, including DRBl* 1501, DQAl* 0102 and DQBl* 0602. We have used the primers designed by Powis et al.

tected MHC class II marker alleles. Each is the subject of ongoing study. The discovery of a group of polymorphic non-MHC genes, which map to the MHC class II region of chromosome 6 (Deverson et al., 1990; Monaco et al., 1990; Spies et al., 1990; Trowsdale et al., 1990) has made them candidates for MS susceptibility genes (Fig. 1). Moreover, these genes play a role in antigen processing and presentation by MHC class I molecules. Two of these genes (TAP1 and TAP2) encode subunits of a heterodimeric transporter protein (Spies et al., 19921, the function of which is to transport antigenic peptides, generated in the cytoplasm, across the endoplasmic reticulum (ER) membrane into the ER where they can associate with class I molecules. Polymorphism of TAP2 exists, with the most significant allelic variation being due to an additional stop codon resulting in a short protein of 669 amino acids (TAP2 short) compared with 686 amino acids (TAP2 long). These major alleles have been designated TAP2 * 01 and TAP2 * 02, respectively, by the WHO nomenclature committee (Bodmer et al., 1992). Further polymorphism has been described in both TAP2 short (TAP2A, C, D and E) and TAP2 long (TAP2B and F) proteins due to single amino acid substitutions at four polymorphic sites (Caillat-Zucman et al., 1993; Powis et al., 1993a). Polymorphisms in the rat TAP2 gene have been found to alter the spectrum of peptides bound by the MHC class I allele RT1.A” and its subsequent recognition by allogeneic T cells (Powis et al., 1992a). However, differences in the functional properties of TAP alleles in humans have not yet been reported. This may be due to the human alleles being relatively

HLA-H

DNA DOB DP

\

-\

TNF HLA-B HLA-C

DQA

HLA-A

HLA-E

HLA-G HLA-F

I

0

1000 <-CLASS

HLA-X

II-

2000 VLASSIII-

<

3000

4000 kb

CLASS I V

Fig. 1. Physical map of the human major histocompatibility complex showing the more important genes in each of the class II, class III and class I regions. (a) denotes non-MHC genes lying within the Class II region.

B. H. Bennetts et al. /Journal

of Neuroimmunology 59 (1995) 113-121

115

2.4. PCR

a) 2044 TAP2*0 1

AAG Cl-f

GCC

CAG CTC TAG GAG ‘SAC ___

TAP2*02

C..

___

b) 682 TAP2*01

KLAQL*

TAP2*02

-

Q

E

o

Fig. 2. (a) The polymorphism distinguishing the short and long allelic form of TAPZ. The polymorphic BstOI is underlined in the TAP:! * 02 sequence. The sequence begins at residue 2044 (Powis et al., 1992b). (b) The corresponding amino acid sequence, which begins at residue 682.

(1992b) to create a PCR-RFLP strategy to type genomic DNA for the TAP2 *01 and the TAP2 *02 alleles.

2. Materials and methods 2.1. MS patients and control subjects

A group of 100 Australian MS patients were selected from Royal Prince Alfred and Westmead Hospitals, Sydney, by established criteria (Poser et al., 1983). All subjects had the relapsing/remitting form of the disease. A normal control panel of 100 subjects was gathered from staff at Westmead Hospital, Sydney and unrelated partners of MS subjects. An additional 37 DRBl* 1501-positive controls were supplied by the NSW Red Cross Tissue Typing Laboratory, Sydney to improve the comparison of the DRBl* 1501 individuals. The ethnic origin of patients and controls was similar and predominantly Caucasian.

TAP2-specific primers described by Powis et al. (1992b) were used. Genomic DNA (0.1-1.0 pg) was amplified in a 20-~1 reaction containing 100 ng of TAPZspecific primers, 10 mM Tris (pH 9 at 25” C), 50 mM KCl, 0.01% gelatin (w/v), 0.1% Triton-X 100, 200 FM dNTPs, 4 mM MgCl, and 1.5 U Taq polymerase. Amplifications were carried out in a capillary thermocycler (Corbett Research) using the following temperature profile: 1 cycle of 95” C for 3 min, 57” C for 15 s, 72” C for 15 s; 4 cycles of 94” C for 15 s, 57” C for 15 s, 72” C for 15 s; 35 cycles of 94” C for 3 s, 57” C for 3 s, 72” C for 15 s. 2.5. PCR-RFLP analysis The 270-bp PCR product was digested with the restriction enzyme Bst 01 (Promega), according to the manufacturer’s directions. This enzyme recognized the polymorphism encoding glutamine at amino acid residue 687 (Gin-687), distinguishing the TAP2 short (Stop-687) and long alleles (Gin-687). This polymorphic restriction site is shown in Fig. 2. Digestion products were electrophoresed on a 3-40% gradient polyacrylamide gel (Gradipore), and stained with ethidium bromide. The fragment sizes for this digestion are shown in Table 1. Resolution of the 34-, 33-, 32- and 27-bp fragments required the use of gradient gels. Using these techniques we have classified all our subjects into 2 groups, those containing the short alleles (TAP2 *01) and those containing the long allele (TAP2 *02). The BstOI digestion patterns showing TAP2*01, TAP2”02 homozygotes and a TAP2*01, TAP2 *02 heterozygote are seen in Fig. 3. The genotypes of some of these subjects were confirmed by digestion of the amplified fragments with BfaI (Ronningen et al., 1993). 2.6. Statistical analysis

2.2. HLA typing DQAl and DQBl genotyping was performed (Westmead) using a PCR-RFLP technique (S.M. Teutsch, B.H. Bennetts, M. Castle, M. Hibbins, R.N.S. Heard and G.J. Stewart, submitted for publication). HLA-A, -B and -DR typing was performed by the NSW Red Cross Tissue Typing Laboratory, Sydney. DR2-positive subjects were also genotyped by PCR-RFLP (Ota et al., 1992). 2.3. DNA extraction DNA was extracted from whole blood and from frozen white blood cells by a rapid salting out method (Lahiri and Nurnberger, 1991).

Associations between TAP2 allele groupings and MS were examined by chi-squared analysis and P-val-

Table 1 Fragment sizes (bp) produced TAP2 PCR product

from

BstOI digestion

TAP2 short allelic forms

TAP2 long allelic forms

64 60 34

64 _

32 _ 9

34 33 32 27 9

of the 270-bp

116

B.H. Bennetts et al. /Journal

of Neuroimmunology

59 (1995) 113-121

67

26

digest patterns of TAP2 PCR fragments electrophoresed restriction Fig. 3. BstOI and a TAP2 * 01,’ * 02 heterozygote. TAP2 ’’ 01,’ * 01 and TAP2 * 02,’ * 02 homozygotes,

ues were corrected by multiplying by the number of comparisons (Bonferoni inequality method). In this study 60 comparisons were made. Where the overall distribution of TAP2 genotypes was significant, odds ratios were calculated for each genotype (Woolf’s formula). The statistical difference of the odds ratios (OR) from unity was determined by using the chisquared test with one degree of freedom. Linkage disequilibrium between TAP2 and HLA alleles was examined by calculating delta values (Svejgaard et al., 1979).

3. Results 3.1. TAP2 associations in MS and control subjects The gene, genotype and phenotype frequencies of the TAP2 *01 and TAP2 *02 alleles are shown in Table 2 where they are correlated with MHC phenotype. There was no significant difference between the gene frequencies of all MS patients and controls. The distribution of the genotypes in the control population fitted the Hardy-Weinberg equilibrium. However, the

on

a 3-40%

gr ,adient

polyacrylamide

gel

Jing

distribution in the patients was distorted with the number of heterozygotes lower than expected (22% observed vs. 36% expected). There were no significant differences in TAP2 associations between MS patients and controls in DRBl* 1501-positive individuals. There was an increase in the TAP2 * 01 gene frequency in the DRBl* 1501-positive patients and a trend towards having more of the TAP2 * Ol/ * 01 homozygous genotype. Due to the strong linkage disequilibrium between the DRBl* 1501, DQAl * 0102 and DQBl * 0602 alleles, the associations with TAP2 observed for DQAl * 0102- and DQBl * 0602-positive individuals were very similar to those of the DRBl* 1501-positive individuals. There were no significant differences in TAP2 associations between the DRBl* 1501-negative MS patients and the controls. However, there was an increase in the TAP2 * 02,’ * 02 homozygous genotype in the patients. In the patients the distribution of the genotypes was distorted with fewer heterozygotes than expected (25% observed vs. 46% expected), the distribution of genotypes in the control group was close to the expected distribution. An association between MS and DQAl alleles en-

B.H. Bennetts et al. /Journal

of Neuroimmunology 59 (1995) 113-121

coding glutamine at amino acid residue 34 (DQA1*0102, *0103, *0401, “0501 and *0601) has been proposed (Spurkland et al., 1991). Similarly, an association between MS and DQBl alleles (DQBl* 0602, * 0603, * 0604, *0302, *0303 and * 0201)

Table 2 Analysis of TAP2

allele distribution

TAP2 allele

in MS patients

Phenotype

1501-positive

Genotype

‘01

‘02

‘01

* 02

*01/

0.7900 0.7500

0.2100 0.2500

90 93

32 43

0.9035 0.8426

0.0965 0.1574

100 100

0.6395 0.7289

0.3605 0.2711

0.8810 0.8182

0.1190 0.1818

DQAl* 0102.negative MS n = 37 0.6351 0.3649 Controls n = 67 0.7164 0.2836 DQBl* 0602.positive MS n = 54 0.9259 0.0741 Controls n = 18 0.8611 0.1389 DQBl* OCOZ-negative MS n = 46 0.6304 0.3696 Controls n = 82 0.7256 0.2744 DRBl * 1501 /DQAl * 0102 / DQBl * 0602.positive MS n = 54 0.9259 0.0741 Controls n = 17 0.8529 0.1471 DRBI * lSOl/ DQAl * 0102 /DQBl* 0602-negative MS n = 46 0.6304 0.3696 Controls n = 83 0.7289 0.2711 DQAl* OlOl-positive MS n = 15 0.5333 0.4667 Controls n =31 0.6290 0.3710 DQAl* OlOl-negative MS n = 85 0.8353 0.1647 Controls n = 69 0.8043 0.1957 Gln3Cpositive MS n = 89 0.8202 0.1798 Controls n = 77 0.7857 0.2143 Leu26-positive MS n = 90 0.8278 0.1722 Controls n = 76 0.7500 0.2500 HLA-A3-positive MS n = 39 0.8077 0.1923 Controls n = 24 0.6667 0.3333 HLA-A3-negative MS n = 61 0.7787 0.2213 Controls n = 76 0.7763 0.2237 HLA-B7-positive MS n = 36 0.8472 0.1528 Controls n = 17 0.7059 0.2941 HIA-B7-negative MS n = 64 0.7578 0.2422 Controls n = 83 0.7590 0.2410 a 17 DRBl*

encoding leucine at amino acid residue 26 has been reported (Haegert and Francis, 1992). We examined associations between these alleles and those of TAP2 to establish whether or not they were stronger than the association between the alleles of the MS susceptibility

and controls

Gene frequency

All individuals MS n = 100 Controls n = 100 DRBl* 1501-positive MS n = 57 Controls n=54” DRBl* 1501~negative MS n = 43 Controls n = 83 DQAl* 0102.positive MS n = 63 Controls n = 33

controls

(from

100 random

117

frequency

* 02

*02j*o2

P”,.

68 57

22 36

10 7

0.09

19 31

81 69

19 31

0 0

0.14

77 92

49 46

51 54

26 37

23 8

0.05

98 97

22 33

78 67

21 30

2 3

0.49

76 91

49 48

51 52

24 39

24 9

0.07

100 100

15 28

85 72

15 28

0 0

0.22

78 91

52 46

48 54

30 38

22 9

0.10

100 100

15 29

85 71

15 29

0 0

0.18

78 92

52 46

48 54

30 37

22 8

0.10

73 87

67 61

33 39

40 48

27 13

0.51

93 96

26 35

74 65

19 30

7 4

0.22

92 96

28 39

72 61

20 35

8 4

0.08

93 92

28 42

72 58

21 34

7 8

0.14

87 88

26 54

74 46

13 42

13 13

0.03

92 95

36 39

64 61

28 34

8 5

0.63

92 88

22 47

78 53

14 35

8 12

0.16

89 94

38 42

63 58

27 36

11 6

0.33

individuals)

+ 37 additional

DRBl*

*01

1501-positive

blood donors.

B.H. Bennetts et al. /Journal

118

of Neuroimmunology 59 (1995) 113-121

haplotype (DRBl * 1501, DQAl* 0102, DQBl* 0602) and TAP2 alone. There was no significant difference in the TAP2 associations between the patients and controls having DQ alleles encoding these putative critical residues. As TAP2 is involved in the MHC class I antigen presentation, the potential exists for functional interactions between TAP2 alleles and MHC class I alleles. The class I antigens, HLA-A3 and HLA-B7 are part of the extended MS susceptibility haplotype (Stewart and Kirk, 1983). There were no significant TAP2 associations between HLA-A3- or HLA-B7-positive or negative MS patients compared with corresponding controls when corrected for multiple comparisons. 3.2. HLA-DRBl*1501-, DQAl positive vs. negative MS patients

* 0102-,

DQBl*

3.3. Linkage alleles

HLA

DRl DRl DRBl* DRBl* DR3 DR4 DR7 DQAl’ DQAl* DQAl* DQAl* DQAl DQAl DQBl DQBl DQBl* DQBl* DQBl DQBl DQBl

TAP2

1501 1501

0101 0101 0102 0102 * 0201 * 0301 * 0303 * 0303 0501 0501 * 0602 * 0602 * 0603

“01 ‘02 *01 ‘02 *02 ‘02 ‘01 ‘01 *02 ‘01 ‘02 “01 *02 “01 *02 ‘01 ‘02 ‘01 *02 *02

between

TAP2 allelic groupings

0602-

and selected

Controls

(N = 100)

++

+-

-+

--

A (x 104)

20 15 54 17 10 3 28 27 18 32 11 27 4 7 7 20 15 18 5 3

3 8 0 37 16 19 5 4 13 1 22 5 19 3 3 2 7 0 13 0

73 26 76 38 31 38 65 66 23 61 30 66 37 86 34 73 26 75 36 38

4 51 7 45 43 40 2 3 46 6 37 2 40 4 56 5 52 7 46 59

- 322 401 501 b -291 b -50 - 459 -751 - 466 402 284 - 205 - 768 -416 -510 197 - 101 427 250 -173 116

between

TAP2 and HLA

The associations between TAP2 and HLA alleles were further investigated by determining delta values for both controls and MS patients. The significant values in one or both subject groups are shown in Table 3. There was significant positive linkage disequilibrium in the MS patient group between the class II alleles of the MS susceptibility haplotype (DRBl *1501, DQAl* 0102, DQBl* 0602) and the TAP2 *01 allele, and negative linkage disequilibrium with the TAP2 *02 allele. This association was not seen as strongly in the control group, with significant positive linkage disequilibrium only occurring between DRBl* 1501 and the TAP2 *01 allele. Fourteen individuals (9 patients and 5 controls) were homozygous for DRBl* 1501, DQAl* 0102, DQBl* 0602; all 14 were also homozygous for TAP2 *01. In addition, no TAP2 *02 homozygotes were found amongst 57 patients and 54 controls positive for this haplotype (Table 2). Strong linkage disequilibrium was also observed between the DRl, DQAl* 0101, DQBl* 0501 alleles (which form a known haplotype) and the TAP2* 02 allele in both subject groups. Significant values were also noted for both groups between this TAP2 allele and the DR7 and DQAl* 0201 alleles, which also form a known haplotype.

Not analysed in Table 2, there were highly significant differences with TAP2 between the haplotypepositive and negative patients with respect to the overall distribution of TAP2 genotypes (P, = 0.0031, the TAP2 “Ol/TAP2 ‘01 genotype (85% vs. 48%, OR = 6.25, PC = 0.009), gene frequency (TAP2 * 01-0.9259 vs. 0.6304, PC = 0.00002) and the TAP2*02 phenotype (15% vs. 52%, OR = 0.16, PC = 0.009). None of these differences were found at a significant level amongst the haplotype-positive and negative controls.

Table 3 Linkage disequilibrium

disequilibrium

a HLA class II alleles in control MS patients P,,

< 0.01 < 0.05


< < < <

0.025 0.01 0.005 0.05

< 0.005

< 0.05

subjects

and MS patients

(N = 100) ++

+-

-+

--

A (x 104)

PnE *

10 9 57 11 3 7 15 11 10 62 14 15 7 4 2 11 11 54 8 8

4 5 0 46 20 18 5 4 5 1 49 5 20 1 3 4 4 0 46 7

80 23 33 21 29 25 75 79 22 28 18 75 25 86 30 79 21 36 24 24

6 63 10 22 48 50 5 6 63 9 19 5 48 9 6.5 6 64 10 22 61

- 483 290 1089 -717 - 308 -70 -592 -466 355 1076 - 657 -592 -117 -82 -34 - 466 397 1018 - 902 208

< < < < <

0.025 0.01 0.05 0.005 0.05

< < < < < <

0.025 0.025 0.005 0.0005 0.01 0.025

< < < <

0.025 0.0005 0.0005 0.0005

a Selected on the basis of (i) an MS-associated allele and/or (ii) significant linkage disequilibrium population. b Calculation based on 17 DRBl* 1501-positive controls (from 100 random individuals) + 37 additional * P,, = not corrected.

detected

in either

DRBl * 1501-positive

the MS or control blood donors.

B.H. Bennetts et al. /Journal

of Neuroimmunology 59 (1995) 113-121

4. Discussion The results of this study provide two significant pieces of information for the ongoing quest of finding a full explanation for the association between MS and alleles of the MHC. This quest involves examination of two competing hypotheses to explain the incomplete DRBl* 1501 (DR2) association: (i) that DRBl* 1501, or something close to it, is a direct susceptibility factor which is not necessary in all patients or (ii) DRBl* 1501 is in linkage disequilibrium with an allele of a linked locus, this allele being necessary and therefore in both DRBl* 1501-positive and DRBl* 1501-negative patients (Hedge, 1993). The results of this study show firstly that a single TAP2 allele is not common to all patients, that is, an association with a TAP2 allele does not explain the HLA DRBl* 1501-negative patient and the TAP2 locus is therefore not a candidate for a necessary MS gene. This finding is in keeping with one other recent study that has included DRZnegative MS patients (Spurkland et al., 1994). Our results could be expanded by subtyping the short and long alleles; however, to do so would not affect this conclusion which can be reached using the diallelic polymorphism at amino acid 687 that separates TAP2 * 01 from TAP2 * 02. Second, the results indicate that the MS susceptibility haplotype can be extended to include the TAP2 locus. It is likely that almost all, if not all, DRBl* 1501-bearing haplotypes also carry the TAP2 short mutation: linkage disequilibrium by phenotype analysis is strong (none of the 57 patients or 54 controls positive for DRBl* 1501 were TAP2 * 02 homozygous). All 28 DRBl* 1501, DQAl* 0102, DQBl* 0602 haplotypes identified in 14 individuals homozygous for these alleles (9 patients and 5 controls) also carried TAP2 * 01. In a previous study of linkage disequilibrium between TAP2 and HLA alleles (Powis et al., 1993a) only four homozygous typing lines containing complete genotyping for DRBl* 1501, DQAl* 0102, DQBl* 0602 were examined. Kellar-Wood et al. (1994) reported that there was no linkage disequilibrium between DR15 and TAP2 in a group of 15 DR15 homozygous individuals, although they did not show data to support this conclusion. Thus, within the DRBl* 1501, DQAl* 0102, DQBl* 0602, TAP2 *01 haplotype, this finding provides the first evidence that the putative MS susceptibility gene could reside centromeric of DQB (Fig. 1). Previously, the only polymorphic locus studied on the centromeric side of DQA2 (Olerup and Hillert, 1991) was DP which shows only weak linkage disequilibrium between DQBl * 0602 and a DP allele (Dekker et al., 1993). On the telomeric side of these loci, a recently reported study indicated that approximately 90% of DRBl* 1501, DQAl* 0102, DQBl * 0602 haplotypes carry a particular haplotype of

119

the MHC class III C4A and cytochrome P21 loci, and 50% carry the HLA-B locus allele B7 (Hillert and Olerup, 1993). This led the authors to conclude that susceptibility to MS is coded within or close to the HLA-DR-DQ subregion. On current data, that conclusion can be extended to the HLA-DR-DQ/TAP2 subregion at least to the point of the location of the additional stop codon responsible for the short/long TAP2 polymorphism. Further subtyping of TAP2*01 into the TAP2A, C, D and E alleles (Powis et al., 1993a; Powis et al., 1993b) would refine the description of the MS associated haplotype but would not alter this conclusion. Liblau et al. (1993) demonstrated that the DR2 associated TAP2 short allele appears to be TAP2A (TAP2 * 0101) but the data was not presented in a manner to allow testing of linkage disequilibrium (LD) with the haplotype DRBl* 1501, DQAl* 0102, DQBl* 0602. Our data for all individuals are similar to other recent studies where no significant associations with TAP2 alleles were found (Kellar-Wood et al., 1994; Middleton et al., 1994; Spurkland et al., 1994). Szafer et al. (1994) have reported similar findings; however, they did not publish actual frequency data. TAP2 associations were examined in DRBl* 1501 (DR2)-positive individuals in Swedish (Liblau et al., 19931, English (Kellar-Wood et al., 1994) and Norwegian (Spurkland et al., 1994) MS populations. As in our study, these reports found no significant associations. Interestingly, the three studies, like our own, showed a lack of TAP2 * 02 homozygotes in DRBl* 1501 (DR2)positive individuals. No TAP2 * 02 homozygotes were observed in the DR2 Swedish (Liblau et al., 1993) and Norwegian (Spurkland et al., 1994) MS patients, and only one such individual was observed in DR15 English patients (Kellar-Wood et al., 1994). These results, in conjunction with our own study, indicate a strong association between the TAP2 * 01 allele and DRBl* 1501. Liblau et al. (1993) observed that three DR2 controls were homozygous for TAP2 * 02. However, eight of their 60 DR2 controls did not have DQBl* 0602, DQAl’OlO2 and thus may not have had the DRBl’ 1501 allele. We have observed a TAP2 * 02/ TAP2 * 02 genotype in a DRBl* 1601 (DR2) control individual. Only Spurkland et al. (19941, in a Norwegian MS population, examined individuals who were negative for DRBl* 1501 (DR2) and DQBl* 0602 alleles. Although no significant differences were found between the patient and control groups in their study, as in our own, interestingly there was a trend towards an increase in TAP2 * 02 homozygotes in the patient group in both studies. The difference between the DRBl * 1501-positive patients compared with the DRBl * 1501-negative patients is intriguing. It is a result of the DRBl* 1501

120

B.H. Bennetts et al. /Journal

of Neuroimmunology 59 (1995) 113-121

patients having a slight increase in the TAP2 * 01 allele compared with their corresponding controls and the DRBl* 1501-negative patients having a slight decrease in the TAP2 * 01 alleles compared with their corresponding controls. This suggests that the MHC haplotypes of the DRBl* 1501-positive patients are quite different to those of the DRBl* 1501-negative patients. Thus different MHC susceptibility factors may need to be searched for in the two MS populations. A detailed correlation of TAP2 typing with each of the reported DQ/MS-associated alleles, which was not conducted in any of the other studies, did not add to the above conclusions. Of interest, we and others have noted a reduced frequency of the DRl/DQAl* 0101 haplotype (Allen et al., 1994; G.J. Stewart, B.H. Bennetts, S.M. Teutsch, M. Castle and R.N.S. Heard, submitted for publication). We have found this haplotype to be reduced in DRBl* 1501 patients but not in DRBl* 1501-negative patients. This potentially protective haplotype showed significant linkage disequilibrium with the TAP2 * 02 allele in both the MS and normal populations in our study. This linkage disequilibrium was also noted by Caillat-Zucman et al. (1993) and Ploski et al. (1994). Thus, the DRl, DQAl* 0101, TAP2 *02 haplotype may be protective in the presence of the MS susceptibility haplotype DRBl* 1501, DQBl* 0602, DQAl* 0102, TAP2 * 01. Whilst a direct effect of a TAP2 allele on susceptibility to MS is unlikely, this may not be so for other MHC-associated autoimmune diseases. Caillat-Zucman et al. (1993) concluded that the TAP2 * 0201 allele had a dominant protective effect in IDDM which could neutralise the susceptibility effect of the IDDM associated phenotypes, DR3 and/or DR4. They also found that the TAP2 *OlOl-blank genotype (most likely homozygous) conferred predisposition to IDDM, suggesting that the TAP2 * 0101 allele had a recessive effect. Overall, the results to date are too few for definite conclusions to be drawn about the role of TAP2 in susceptibility to autoimmune disease. They indicate, however, that the MHC-encoded genes involved in MHC class I antigen processing and presentation appear to be worthy of detailed study in several autoimmune disorders, in varying countries and ethnic groups.

Acknowledgements This study was supported by grants from the National Multiple Sclerosis Society of Australia. We wish to thank the NSW Red Cross Tissue Typing Laboratory, Sydney for HLA typing and the provision of additional DRBl* 1501-positive controls. Assistance in the recruitment of MS patients was given by Prof. James McLeod and Claudia Charlton (Royal Prince Alfred Hospital, Sydney). We also wish to thank Mark

Hibbins, Marc Buhler and Marianne Castle (Westmead Hospital, Sydney) for control samples, raising EBVtransformed cell lines on patient samples, and DQBl genotyping, respectively.

References Allen, M., Sandberg-Wollheim, M., Sjogren, K., Erlich, H.A., Petterson, U. and Gyllensten, U. (1994) Association of susceptibility to multiple sclerosis in Sweden with HLA class II DRBl and DQBl alleles. Hum. Immunol. 39, 41-48. Bodmer, J.G., Marsh, S.G.E., Albert, E.D., Bodmer, W.F., DuPont, B., Erlich, H.A., Mach, B., Mayr, W.R., Parham, P., Sasazuki, T., Schreuder, G.M.Th., Strominger, J.L., Svejgaard, A. and Terasaki, P.I. (1992) Nomenclature for factors of the HLA system, 1991. Hum. Immunol. 34, 4-18. Caillat-Zucman, S., Bertin, E., Timsit, J., Boitard, C., Assan, R. and Bach, J.-F. (19931 Protection from insulin-dependent diabetes mellitus is linked to a peptide transporter gene. Eur. J. Immunol. 23, 1784-1788. Colonna, M., Bresnahan, M., Bahram, S., Strominger, J.L. and Spies, T. (1992) Allelic variants of the human putative peptide transporter involved in antigen processing. Proc. Nat]. Acad. Sci. USA 89, 3932-3936. Dekker, J.W., Easteal, S., Jakobsen, I.B., Gao, X., Stewart, G.J., Buhler, M.M., Hawkins, B.R., Higgins, D.A., Higgins, D.A., Yu, Y.L. and Serjeantsen, SW. (1993) HLA-DPBl alleles correlate with risk for multiple sclerosis in caucasoid and Cantonese patients lacking the high-risk DQBl* 0602 allele. Tissue Ant. 41, 31-36. Deverson, E.V., Gow, I.R., Coadwell, W.J., Monaco, J.J., Butcher, G.W. and Howard, J.C. (1990) MHC class II region encoding proteins related to the multidrug resistance family of trans-membrane transporters. Nature 348, 738-741. Ebers, G.C., Paty, D.W., Stiller, C.R., Nelson, R.F., Seland, T.P. and Larsen, B. (1982) HLA-typing in multiple sclerosis sibling pairs. Lancet ii, 88-90. Haegert, D.G. and Francis, G.S. (1992) Contribution of a single DQb chain residue to multiple sclerosis in French Canadians. Hum. Immunol. 34, 85-90. Hillert, J. and Olerup, 0. (1993) Multiple sclerosis is associated with genes within or close to the HLA-DR-DQ subregion on a normal DRl$DQ6,Dw2 haplotype. Neurology 43, 163-168. Hodge, SE. (1993) Linkage analysis versus association analysis: distinguishing between two models that explain disease-marker associations. Am. J. Hum. Genet. 53, 367-384. Kellar-Wood, H.F., Powis, S.H., Gray, J. and Compston, D.A.S. (1994) MHC-encoded TAP1 and TAP2 dimorphisms in multiple sclerosis. Tissue Ant. 43, 129-132. Lahiri, D.K. and Nurnberger, J.I. Jr. (1991) A rapid and non-enzymatic method for the preparation of HMW DNA for RFLP studies. Nucleic Acids Res. 1991, 4491. Liblau, R., van Endert, P.M., Sandberg-Wollheim, M., Patel, S.D., Lopez, M.T., Land, S., Fugger, L. and McDevitt, H.O. (1993) Antigen processing gene polymorphisms in HLA-DR2 multiple sclerosis. Neurology 43, 1192-1197. Marrosu, M.G., Muntoni, F., Murru, M.R., Spinicci, G., Pischedda, M.P., Goddi, F., Cossu, P. and Pirastu, M. (1988) Sardinian multiple sclerosis is associated with HLA-DR4: a serologic and molecular analysis. Neurology 38, 174991753. Middleton, D., Megaw, G., Cullen, C., Hawkins, S., Darke, C. and Savage, D.A. (1994) TAP1 and TAP2 polymorphism in multiple sclerosis patients. Hum. Immunol. 40, 131-134.

B.H. Bennetts et al. /Journal

of Neuroimmunology 59 (1995) 113-121

Monaco, J.J., Cho, S. and Attaya, M. (1990) Transporter protein genes in the murine MHC: possible implications for antigen processing. Science 250, 1723-1726. Olerup, 0. and Hillert, J. (1991) HLA class II-associated genetic susceptibility in multiple sclerosis: a critical evaluation. Tissue Ant. 38, I-15. Ota, M., Seki, T., Fukushima, H., Tsuji, K. and Inoko, H. (1992) HLA-DRBl genotyping by a modified PCR-RFLP method combined with group specific primers. Tissue Ant. 39, 187-202. Ploski, R., Undlien, D.E., Vinje, O., Forre, Q., Thorsby, E. and Ronningen, K.S. (1994) Polymorphism of human major histocompatibility complex-encoded transporter associated with antigen processing (TAP) genes and susceptibility to juvenile rheumatoid arthritis. Hum. Immunol. 39, 54-60. Poser, C.M., Paty, D.W., Scheinberg, L., McDonald, WI., Davis, F.A., Ebers, G.C.. Johnson, K.P., Sibley, W.A., Silberberg, D.H. and Tourtellotte, W.W. (19831 New diagnostic criteria for multiple sclerosis: guidelines for research protocols. Ann. Neural. 13, 227-231. Powis, S.J., Deverson, E.V., Coadwell, W.J., Ciruela, A., Huskinson, N.S., Smith, H., Butcher, G.W. and Howard, J.C. (1992a) Effect of polymorphism of an MHC-linked transporter on the peptides assembled in a class I molecule. Nature 357, 211-215. Powis, S.H., Mockridge, I., Kelly, A., Kerr, L., Glynne, R., Gileadi, U., Beck, S. and Trowsdale, J. (1992b) Polymorphisms in a second ABC transporter gene located within the class II region of the human major histocompatibility complex. Proc. Natl. Acad. Sci. USA 89, 1463-1467. Powis, S.H., Tonks, S., Mockridge, I., Kelly, A.P., Bodmer, J.G. and Trowsdale, J. (1993al Alleles and haplotypes of the MHC-encoded ABC transporters TAP1 and TAP2. Immunogenetics 37, 373-380. Powis, S.H., Rosenberg, W.M.C., Hall, M., Mockridge, I., Tonks, S., Ivinson, A., Ciclitira, P.J., Jewell, D.P., Lanchbury, J.S., Bell, J.I. and Trowsdale, J. (1993b) TAP1 and TAP2 polymorphism in coeliac disease. Immunogenetics 38, 345-350. Ronningen, K.S., Undlien, D.E., Ploski, R., Maouni, N., Konrad, R.J., Jenson, E., Hornes, E., Reijonen, H., Colonna, M., Monos, D.S., Strominger, J.L. and Thorsby, E. (1993) Linkage disequilibrium between TAP2 variants and HLA class II alleles; no primary

121

association between TAP2 variants and insulin-dependent diabetes mellitus. Eur. J. Immunol. 23, 1050-1056. Spies, T., Bresnahan, M., Bahram, S., Arnold, D., Blank, M.E., Pious, D. and De Mars, R. (1990) A gene in the major histocompatibility complex class II region controlling the class I antigen presentation pathway. Nature 348, 744-749. Spies, T., Cernundolo, V., Colonna, M., Cresswell, P., Townsend, A. and De Mars, R. (1992) Presentation of endogenous viral antigen dependent on putative peptide transporter heterodimer. Nature 355, 644-649. Spurkland, A., Ronningen, K.S., Vandvik, B., Thorsby, E. and Vartdal, F. (1991) HLA DQAl and HLA DQBl genes may jointly determine susceptibility to develop multiple sclerosis. Hum. Immunol. 30, 69-75. Spurkland, A., Knutsen, I., Undlien, D.E. and Vartdal, F. (1994) No association of multiple sclerosis to alleles at the TAP2 locus. Hum. Immunol. 39, 299-301. Stewart, G.J. and Kirk, R.L. (1983) The genetics of multiple sclerosis: the HLA system and other genetic markers. In: J.F. Hallpike, C.W.M. Adams and W.W. Tourtellotte fEds.1, Multiple Sclerosis. Chapman and Hall, London, pp. 97-128. Svejgaard, A., Hauge, M., Jersild, C., Platz, P., Ryder, L.P., StaubNielsen, L. and Thomsen, M. (1979) The HLA System. Monogr. Hum. Genet. 7. Szafer, F., Oksenberg. J.R. and Steinman, L. (1994) New allelic polymorphisms in TAP genes. Immunogenetics 39, 374. Tighe, M.R., Hall, M.A., Cardi, E., Ashkenazi, A., Lanchbury, J.S. and Ciclitira, P.J. (1994) Associations between alleles of the major histocompatibility complex-encoded ABC transporter gene TAP2, HLA class II alleles, and celiac disease susceptibility. Hum. Immunol. 39, 9-16. Trowsdale, J., Hanson, I., Mockridge, I., Beck, S., Townsend, A. and Kelly, A. (1990) Sequences encoded in the class II region of the MHC related to the ABC superfamily of transporters. Nature 348, 741-744. Wordsworth, B.P., Pile, K.D., Gibson, K., Burney, R.O., Mockridge, I. and Powis, S.H. (1993) Analysis of the MHC-encoded transporters TAP1 and TAP2 in rheumatoid arthritis: linkage with DR4 accounts for the association with a minor TAP2 allele. Tissue Ant. 42, 153-155.