Kidney International, Vol. 41 (1992), pp. 1 75—182
Analysis of HLA class II genes in Japanese patients with idiopathic membranous glomerulonephritis SATORU OGAHARA, SETSUYA NAIT0, KASUMI ABE, ISAO MICHINAGA, and KIKuo ARAKAWA Second Department of Internal Medicine, Fukuoka University School of Medicine, Fukuoka, Japan
Analysis of HLA class II genes in Japanese patients with idiopathic membranous glomerulonephritis. Thirty unrelated Japanese adult patients with idiopathic membranous glomerulonephritis (1MG) and 50 unrelated Japanese normal controls were examined for HLA class II genes. Both the 3.0 kb DQB and the 6.2 kb DQA restriction fragments were simultaneously identified in 86.7% of the patients. These were
different between Japanese and Caucasians, specifically HLADR2-DQwl and DR3-DQw2 [7, 81, respectively.
length polymorphism (RFLP) of DQB and DQA of the patients coincided with that of homozygous cell lines with Dw2. The DRB probe showed no significant differences in the hybridization patterns between the patient group and HLA-matched controls. These findings indicate that the specific HLA-DQ polymorphism correlates primarily with
probes using the PCR technique [9]. We compared the published amino acid sequences of second exon of HLA class II genes related with Japanese and Caucasian 1MG.
In the present study, we attempted to ascertain the HLA
regions much more specifically associated with 1MG by examining the restriction fragment length polymorphisms (RFLP) of significantly different from the 14% frequency of the fragments in HLA class II genes. We also tried to determine the DQA area controls (relative risk = 39.9; P = 1.7 x l0°). The haplotype of HLA class II genes associated with 1MG using the polymerDR2-DQw1 associated with 1MG seemed to be Dw2 which was essentially different from the haplotype common to the normal Japanese ase chain reaction (PCR) method, since Saiki et al established it population, Dw12, This was because most of the restriction fragment for HLA DQA typing with sequence-specific oligonucleotide
Methods
Japanese 1MG.
Patients, controls and cell lines Specific HLA class II alloantigens, HLA-DR2 and -DQw1, have been found in Japanese patients with idiopathic membranous glomerulonephritis (1MG) about eight times more frequently than in the normal Japanese population [1—3]. Since HLA-DR2 and DQw 1 were very closely linked to each other [4], the association of HLA-DR2 and -DQw1 with 1MG was considered to be due to that of a haplotype of HLA-DR2-DQwl with 1MG, However, the association clarified by serologically
typed HLA was too rough to define the "host factors for susceptibility" to 1MG.
HLA-DR and -DQ regions were subdivided according to specificities which were characterized by mixed lymphocyte culture. Recent advances in DNA technology have made it possible to define almost HLA-D clusters [5] and HLA class II gene sequences [6], and depict some DNA sequences of HLA
Thirty unrelated Japanese patients were diagnosed as having 1MG clinically and by renal biopsy, and then were determined of their HLA phenotype. At the time of renal biopsy, their mean age was 56.0 12.1 years. Ten patients were male and 20 were female. Fifty normal persons were chosen as random unrelated Japanese controls and 24 individuals with HLA-DR2 were used as HLA-DR-matched controls. Epstein-Barr virus-transformed B-lymphoblastoid cell lines which were homozygous for HLADR2 were used to produce contrast electrophoretic pictures. The D02089l5, WT8, SHCU and MGAR cell lines as Dw2, the E4181324 cell as Dwl2, the KASO1I as Dw21 cell line, and the RML as Dw22 were obtained from the Tenth International HLA Workshop core cell lines. Serological HLA typing
HLA-A, -B, -C, -DR and -DQ typings were done by the
class II genes being responsible for the susceptibility of diseases standard microdroplet lymphocytotoxicity test, using antisera like insulin-dependent diabetes mellitus (IDDM) and rheuma- from the Tenth International HLA Workshop, Japanese retoid arthritis (RA) from the serological association of HLA with gional antisera and the local antisera comparable to them.
those diseases. The haplotype of HLA-DR2-DQw1 has also Restriction fragment length polymorphism analysis (RFLP) been known to be one of the most common among Japanese, The analysis procedure was performed according to the and is able to be divided into at least four distinct HLA-D standardized Southern blot workshop technique established at clusters. Furthermore, the haplotype associated with 1MG is
the Tenth International HLA Workshop [10]. In brief, genomic DNA samples were obtained by phenolchloroform extraction of sodium dodecyl sulfate (SDS) and by lysing cells with protein-
Received for publication February 7, 1990 and in revised form August 12, 1991 Accepted for publication August 19, 1991
ase K from peripheral blood leukocytes of patients, controls and the cell lines. The probes used for hybridization were 0.78
© 1992 by the International Society of Nephrology
provided from the Tenth International HLA Workshop [5]. The
kb DRB, 1.2 kb DQB and the 1.6 kb DQA cDNA clone, 175
176
Ogahara et a!: HLA-class 11 genes in 1MG
A 10
DOA1 * 0101
0102 0103
40
50
60
80
70
Q
0201 0301
Y
0401
Y Y
0501
30
20
DHVASCGVNLYQFYGPSGQYTHEFDGDEEFYVDLERKETAWRWPEFSKFGGFDPQGALRNMAVAKHNLNIMIKRYNSTAATN S S
Y
S S S
0601
--n-
-
F F
K
V-KL—L—HRLR ————F——T—I——L S
Q Q Q
F
L———S
V—QL—L—RR—RR————F——T—I——L
V———S
G———-V—CL—VLRQ—R —-——F——T—I——L G-———V—CL—VLRQ—R ————F——T—I——T G————V-CL-VLRQ—R ————F——T—I——T
SL———S L———S L———S
GH27
GH26
B
DQA1 Specificity
5' SSo
Sequence
3'
1 CACGTAGAACTCCTCATC 2 CACGTAGAACTGCTCATC
3 CTCCTTCCTCTCCAGGTC 4 GTCTCCTTCTTCTCCAGG 5 CAGAGGCAACTTCCAGAC
6 ATCTTCTAAATCTGCGGA 7 TCAGACTGTTCAAGTTAT
8 CGTAGAACTGCTCGTCTC 9 TCATGGGTGTACTGGCCA 10 TCATGGGTGAACTGGCCA
aa
0101
01020103 0201 03010401 0501 0601
32-37 * 32-37 38-43 * 39-44 45-50 51-56
* *
*
* *
* * *
*
71-7 6
31-36 23-28 * 23-28
*
* *
*
*
*
*
*
*
Note, aa, amino acid position; , positive spot
Fig. 1. A. Comparison of amino acid sequences of DQA. The PCR primers (GH26, GH27) and their relationship to the DQA genes by arrows. B. The sequence of allele-specific oligonucteotide probes and their specificity.
restriction endonuclease, Taq I (Toyobo Co. Ltd., Tokyo) were used to digest the genomic DNA. Ten g of the endonuclease-
digested DNAs were subjected to electrophoresis on 0.7% agarose gel for 40 hours at 35 volts in 40 mvt Tris-HC1, 20 mM glacial acetic acid and 2 mM EDTA (TAE buffer) [5]. The gel was soaked in 0.15 M HCI for ten minutes, and then soaked in
0.4 N NaOH for 30 minutes. The gel was transferred onto a nylon membrane (Gene Screen Plus, NEN, Massachusetts, USA) by the method of Southern [11]. DNAs were transferred overnight with 0.4 N NaOH solution as transfer buffer, The probes were labeled with [a-32P] dCTP by the multiprime labeling method [121. The membranes were prehybridized overnight in prehybndization mix (50% deionized formamide, 0.1% Denhardt solution [13], 5x SSPE (30x SSPE is 5.4 M NaC1, 0.3 M NaH2PO4 and 30 mst EDTA), 0.1% SDS, 5% Dextran sulfate
Amp1ficarion of genoinic DNA
The second exon of the DQA gene was amplified from genomic DNA with the use of PCR. The PCR primers used were GH26 and GH27 as described [14] (Fig. lA). DNAs were prepared from a part of the RFLP analysis samples. One g of
genomic DNA was amplified by PCR. Two units of Taq polymerase (Stratagene, CA) were added to a 100 1.d volume of polymerase buffer containing 10 mrvi Tris-HC1 pH 8.4, 50 mM KCI, 2.5 mr't MgCl2 and 100 g/ml gelatin, overlaid with mineral
oil. The reaction was adjusted to 200 M for each of the four deoxynucleoside triphosphates and 0.2 /LM for each of the PCR primer pairs, amplified by using a programmable thermal cycler (Perkin Elmer-Cetus Co., Norwalk, Connecticut, USA). Each cycle consisted of a 30 second denaturation step at 95°C, a 55°C annealing step, and a two minute extension step at 72°C except for initial denaturation for five minutes at 95°C, and a final seven
and 100 tg/m1 Herring sperm denatured DNA) at 42°C with agitation. The membranes were hybridized in prehybridization minute extension step at 72°C. The PCR products were conmix with labeled probes for 48 hours at 42°C. The membranes firmed by the electrophoresis in a 1.5% agarose gel. were washed with 2 x SSPE twice for five minutes at room temperature, and with 2x SSPE and 0.5% SDS once for 15 Dot typing of PCR DQA products minutes at 65°C, and then with 0.5x SSPE for 15 minutes at 65°C. The membranes were subsequently dried and exposed to Ten sequence-specific oligonucleotide (SSO) probes were an X-OMAT-AR film (Kodak, Rochester, New York, USA) designed and supplied by the Tenth Japan HLA Workshop (Fig. from overnight to seven days at —80°C. IB). Thirty to 50 ng of specific PCR products was spotted onto
177
Ogahara et alt HLA-class 11 genes in 1MG
Table 1. HLA-A, B, DR, DQ type, RFLP of DQB and DQAI genes and subregion of DQAI of patients with 1MG
RFLP HLANo.
A
Patient
I
M. MOR.
2
T.YAM.
2 24
3 4 5
M. MIZ.
2
6
T. ISH.
11
M. FUJ. A. HAS.
24 24 2 2 26 2 24
B 11
60
31
7
24 24 26 31
39
11
54
24
13
7 8 9
H. MAE.
10 11
K. KIM. Y. ARA.
12
A. HIG.
11
26
13
N. MAS.
31
14 15
T. KUS. T. TOK.
24 24 24
16
S. MIT.
11
17
H. YOS. T. TAN.
31
18 19 20 21
22 23 24 25 26 27 28 29 30
K. FUJ. M. HIR.
24
26 26 24 26 26
S. OOT.
2 2 24 24 26 2
K. 00K.
24
26 33 33
S. MIY.
T. 00K. K. YOS. 5. 00K.
U. KUM. T. TIK.
T. TAK. H. SAl.
M. NAK. K. KUM.
a , existent;
—,
2
24
DQB
31 11
2 2 2
31
11
24
31 31
51 7 51
62 35 7 61 54 52 51 35 35 52 39 48 7 35 35 7
DR 61 51 56 60 35
52
2
8
1
2 2 2 2 2 2
1
1
8 8
1
1
1
60
2
54
2 2
52 62 62 35
DQ
1
1
4
1
12
1
1
1
60
13
1
7 62 35 54
51
4 9 11
7 7 4
59 59
— —
5
12
1
+ + + + + + + + + + + + + + + + + + + + + + +
--
13
4 4
+ + + + + + + + +
61
51
DQA1
1
44 44
55 55 60
7.0 kb
1
39 35
48 39
DOT
6.2 kb
4 9 4
2 2 2 2 2 2 2 2 2 2 2 2 2 2
60
4 7
DQA1
3.0 kb
1 1
3 3
1
8 8 8
1
4-
1 1
3
1
1
1
8
1
1
1
3
1
1
12
4
1
+ --
1
9 14
+ + + + +
7 3
+ + + + + + + + — — —
—
+ +
— —
0102 0102 0102 0102 0102 0102 0102 0102 0102 0102 0102 0102 0102 0102 0102 0102 0102 0102 0102 0102 0102 0102
— —
— —
0101 0401
— —
—
0401
—
0301
--
—
+ +
0102 0102 0102 0102
+ — — — — — — —
+ — — —
+ — — — — — —
0103 0101
0103 0103 0101
0103 0301 0401 0301 0301 0301 0103
0301 0103 0301 0101 0101 0103 0101 0401 0301 0301
not existent
nylon membranes, which were presoaked in soaking solution [0.15 M NaC1, 10 m sodium phosphate, 1 mrsi EDTA (pH 7.4)]
for 15 minutes, and then dried at 60°C. Membranes were rewetted with 0.4 N NaOH for one minute for denaturation, and then soaked in a soaking solution for five minutes, and exposed to UV light to fix the PCR products for 30 seconds. SSO probes were phosphorylated at their 5'- termini with [y-32P] ATP and
by convention were expressed as relative risks. When one element of the equation was zero, the relative risk was calculated with the formula of Haldane [17]: [(a +
112)(d + 1/2)]
[(b +
1/2)(c +
l/2)f
Results T4 polynucleotide kinase (TOYOBO, Osaka, Japan). Membranes were prehybridized in a hybridization solution (50 mM Table 1 lists HLA-A, B, DR and DQ antigens typed serologTris-HCI pH 8.0), 3M tetramethylammonium chloride [151, 2 ically, existence of the specific DQB and DQA RFLP, and DQA mM EDTA (pH 8.0), 5x Denhardt's solution [13], 0.1% SDS subtypes by dot typing of the 30 Japanese patients with 1MG. and 100 j.tg/ml denaturated Herring sperm DNA) for one hour at Hybridization with a DQB probe of genomic DNA digested 54°C. Hybridization with 32P-labeled SSO probes was accom- with Taq I showed two constants (3.4 and 4.5 kb) and at least plished by the addition of a 2 pmol SSO probe in 5 ml of the two polymorphic bands (3.0 and 5.3 kb). Hybridization with a hybridization solution for one hour at 54°C. The membranes DQA probe of genomic DNA digested with Taq I showed at were rinsed twice in 0.3 M NaCI, 20 m sodium phosphate, 2 least five polymorphic DQA1 bands (2.8, 4.6, 5.5, 6.2 and 7.0 mM EDTA, and 0.1% SDS for 10 minutes at room temperature, kb) and two polymorphic DQA2 bands (2.3 and 2.0 kb). These washed in 50 m Tris-HC1 (pH 8.0), 3 M tetramethylammonium hybridization patterns were consistent with those described by chloride, 2 mt EDTA and 0.1% SDS for 10 minutes at room Marcadet et a! [181 and Bontrop et a! [19]. temperature and then for 30 minutes at 58°C, and exposed for Table 2 presents the antigen phenotypic frequencies of patients and controls with HLA-DR2, -DRw53 and -DQw1, and autoradiography with an intensifying screen for one hour. the frequencies of patients and controls with the Taq I-digested Statistical analysis 3.0 kb DQB and 6.2 kb DQA restriction fragments. Both of Statistical analysis was performed by the chi-square method these fragments were simultaneously identified in 86.7% (26 of and Fisher's exact test. P values were multiplied by the number 30) of patients. This percentage was significantly different from of antigens and fragments tested (corrected P) [16]. Odds ratios the 14% (7 of 50) frequency of the fragments between patients were calculated with the use of the formula (a x d)/(b X c) and and controls identified in controls. The relative risk of these
178
ogahara ci al: HLA-class II genes in 1MG
Table 2. Comparison of frequencies of HLA serological class II typing and fragments of DQB and DQA RFLP
Serology DR2 DRwS3 DQw1
RFLP DQB 3.0 kb DQA 6.2 kb
Patients = 30)
Controls
(N = 50)
Relative risk
80.0% 36.7% 90.0% 86.7% 86.7%
30.0% 60.0% 58.0% 14.0% 14.0%
9.3 0.4 6.5 39.9 39.9
P 1.5 X 0.04 2.5 1.7
x iO x io'°
1.7 x 10—0
Corrected P value
1.8 x iO NS
0.013
1.7 x iO—
1.7 x lO
NS, not significant.
LC)
A
cc CD
I0
'-C)
c'.1
B cc
00
ccw
w
Cl)
kb
5.3 4.5
kb
_ 7.0
iI:I!i!:• •II.e.— a
S
6.2
- 4.4
3.4 3.0
eeeØee
0.9
Dw 2222221212 Dw2 222221212 DQB
Taq I
OQA
Fig. 2. Electrophoretic patterns of Taq Idigested DNAs from homozygous cell lines hybridized with (A) DQB and (B) DQA probes.
fragments were much higher than those of DR2 or DQwI (39.9 lines with Dw2 and Dw2 1. The 4.4 kb fragment was found only vs. 9.3 or 6.5), and the statistical differences were much more in Dw22. The 7.0 kb fragment was found only in Dwl2. significant in the fragments than serologically typed antigens, Twenty-four Japanese adult patients with 1MG and a haploDR2 and/or DQw1 (1.7 x 10_b vs. 1.5 x l0 or 2.5 x 10). type of HLA-DR2-DQw 1 (DR2-patients) and twenty-four JapFigure 2A shows that electrophoresis of genomic DNA of anese normal controls with the same haplotype (HLA-matched each homozygous cell lines digested with Taq I produced four controls) were examined. Figure 3 presents the electrophoretic restriction fragment patterns when hybridized with the DQB patterns of Taq I-digested DNAs from a portion of the patients probe. The 4.5 and 3.4 kb fragments were observed in all cell and HLA-matched controls hybridized with DQB probes. All lines. The 3.0 kb fragment was observed in all cell lines with DR2-patients had the 3.0 kb fragment. This fragment was found Dw2, but was not found in the cell lines with Dwl2, 21 and 22. in only three out of 24 HLA-matched controls (RR 301; P = The 0.9 kb fragment was found in only the cell lines with Dw22. 1.0 x 10). The 5.3 kb fragment was found in only the cell line with Dw21. Figure 4 illustrates the electrophoretic patterns of Taq I-diThe cell line with Dw12 did not have 3.0, 0.9 or 5.3 kb
gested DNAs from a portion of the DR2-patients and HLAmatched controls hybridized with DQA probes. The 6.2 kb fragment was also found in all DR2-patients. In contrast, this easily definable. Figure 2B shows that the electrophoresis of fragment was found in only 12.5% of the HLA-matched con301; P = 1.0 x l0—). The 7.0 kb fragment was genomic DNA of each homozygous cell lines digested with Taq trols (RR fragments. Since the number of fragments hybridized with DQA was less than that of DQB, the polymorphisms of DQA fragments were
I produced three restriction fragment patterns when hybridized found in 25% of the DR2-patients, but was found in 88% of the with the DQA probe. The 6.2 kb fragment was found in all cell HLA-matched controls (RR = 0.04; P = 3.7 x 10—s).
179
Ogahara et al: HLA-class II genes in 1MG B Normal
A 1MG
kb!flK Fig. 3. Electrophoretic patterns of Taq I-
es
digested DNAs from a portion of (A) patients with 1MG and (B) normal controls hybridized with DQB probes.
DQB/Taq I
A 1MG
B Normal
._aM_eu
7.0—.-
-
-
kb
—e
—.7.0 6.2
5.5 •' 4.6 2.8 — —
-a S
SS — DQA/Taq I
S
13.8—
S
Fig. 4. Electrophoretic patterns of Taq 1digested DNAs from a portion of (A) patients with 1MG and (B) normal controls hybridized with DQA probes.
B Normal
A 1MG
5.8 —
— 2.8
Nd.
a
as ••1
—a
—'-S
S —
j
I
'a.
S. 1.4——.
-1'
U DRB/Taq I
Figure 5 illustrates the electrophoretic patterns of Taq I-digested DNAs from a portion of the DR2-patients and HLAmatched controls hybridized with DRB probes. The DRB probe gave similar electrophoretic patterns between the patient group
and the controls. The 1.4 kb fragment was found in all DR2 patients and HLA-matched controls. The DQA subregion of all patients with 1MG was determined by DQA dot typing. Twenty-six patients were assigned to have
DQA1*0102, six patients, DQA1*010l, seven patients, DQA1 *0103, nine patients, DQA 1*0301, and four patients,
Fig. 5. Electrophoretic patterns of Taq Idigested DNAs from a portion of (A) patients with 1MG and (B) normal controls hybridized with DRB probes.
DQA 1*0401. However, none of the patients were assigned to have DQA1*0201, DQA1*050l or DQA1*0601 (Table 1). Discussion
The present study suggests the existence of a strong association between Japanese 1MG and some specific fragments of RFLP in DQ. The allele was represented by the 3.0 kb fragment
of DQB RFLP and the 6.2 kb of DQA RFLP. Since these fragments were thought to correspond to the HLA-Dw2 haplo-
type when they were compared to the patterns of RFLP of
180
Ogahara et a!: HLA-class II genes in 1MG
HTCs with DR2 related D type, Dw2, Dw12, Dw2l and Dw22,
86.7% of patients and 12.5% of the normal controls were
DQB. Todd, Bell and McDevitt reported that the specific amino
acid residue of DQ molecules could confer the resistance to suggested to have Dw2. Most of the controls with HLA-DR2 IDDM [28], and Gregersen, Silver and Winchester reported that probably had Dw12 as seen in a report on the normal Japanese specific amino acid residues of DR molecules could confer the susceptibility to RA [29]. Therefore, it may be possible to point population [201. On the other hand, the patterns of DRB RFLP digested with out common amino acid residues of class II molecules which Taq I were very similar between patients with DR2 and HLA- showed a strong association with 1MG in Japanese and also in matched controls. RFLP of DRB was clarified by using a Caucasians. The published amino acid sequence of the second variety of restriction enzymes such as BamH I, Bgl II, EcoR I, exon at the DQA1, DQB1, DRB1 DRB3, and DRB5 genes EcoR V, Hinc II, Hind III, Kpn I, Msp I, Pst I, Pvu II, Sst I and coding HLA-DR2, DR3, DQw1 and DQw2 antigens were comTaq I at the Tenth HLA International HLA Workshop [211. pared with each other [30]. There was no existence of a However, there were few fragments distinguishing between common specific sequence in DQA1*0102 and DQA1*0401 Dw2 and Dw12. genes, in DQB*0602 and DQB1*0201 genes, and in DRB1*1501 The present study revealed that patients with 1MG commonly and DRB 1*0301 genes, respectively. The DRB 1*0102 gene have a relatively less frequent DR2-positive haplotype assigned indicated a strong linkage disequilibrium with the DRB5*0l01 to Dw2 in the Japanese population in contrast to the Bw52-DR2 gene, and the DRB1*0301 gene with the DRB3*0101 gene, assigned to Dw12 more frequently found in the same population respectively. When the amino acid sequence of the DRB5*010l [22]. Two of six patients without DR2 also had both the DQB gene was compared with that of the DRB3*0101 gene, both 3.0 kb and DQA 6.2 kb fragments. These patients (patients No. genes were found to have leucine at position 38. It was known 25 and 26) were typed as HLA-DRw13. Although DRw13 was that the DRB 1 * 1201 genes also had leucine at the same position. less frequent in patients because of the much more frequent Almost all other DRB3 or B5 genes have valine or alanine at occurrence of DR2, the possibility exists that DRw13 might be position 38. Residue 38 of the first domain of the DRB gene was associated with the susceptibility to 1MG. This is because included in the second hypervariable region (residues 25-38). DRwl3 has a linkage disequilibrium with DQw1, and the According to the predicted structure of class II molecules [311, frequency of DQw1 was higher than that of DR2 in the patients. this region is located in the base of the cleft formed by the Patients with DQA1*0102 simultaneously exhibited both frag- extracellular domain of class II molecules. If allelic polymorments of 3.0 kb DQB and 6.2 kb DQA, while patients with phisms in the second hypervariable region determine the funcDQA1*0103 had fragments of 7.0 kb DQA except for patients tional effects on the antigen binding and on the T cell recogniNo. 1 and 22. It is known that DQA1*0102 was in strong linkage tion, these polymorphisms might confer the susceptibility to disequilibrium with DR2(Dw2)-DQw6, DR2(Dw21)-DQw5 and 1MG. Since these sequences were from the literature, it is DRw13(Dw19)-DQw6, whereas the DQA1*0l03 was in strong essential to sequence the class II genes from Japanese and linkage disequilibrium with DRwl3 (Dw18)-DQw6, DR2 Caucasian patients with 1MG if common factors exist between (Dw12)-DQw6 and DRw8-DQwI [23]. These results thus sup- HLA-DR2 (Dw2) in Japanese and HLA-DR3 in Caucasians. port the fact that patients with DQA1*0102 had DR2 (Dw2) or Also important is whether or not these genes may have a DRw13 (Dw19). DQA dot typing clearly confirmed that some linkage disequilibrium with common susceptible genes of 1MG. Some diseases have been reported to be associated with the patients with the serologically homozygous gene of DQ (No. 6, 14 and 18) were able to be assigned as a heterozygote. HLA-Dw2 haplotype. Honda et al first reported that all JapaIn Caucasians, there are HLA-B genes in the strong linkage nese patients with narcolepsy had HLA-DR2 [32]. Inoko et al disequilibrium with HLA-DR3 forming two haplotypes, that is, demonstrated the 100% presence of the DQB EcoR I 2.4 kb, HLA-B8-DR3 and HLA-B18-DR3. Sacks et al reported that BamH I 2.9 kb, and Pst 112 kb fragments in narcolepsy [33]. B8-linked DR3 is different from B 18-linked DR3 from the Matsuki et al described the differences between Dw2 and Dw12 analysis of the DRB genes by RFLP, biochemical studies of DR using Taq I-generated DQB RFLP and reported that all patients molecules by two-dimensional gel electrophoresis and func- with narcolepsy had Dw2 [34]. Unlike 1MG, Caucasian patients tional studies of mixed lymphocyte reactions [24]. Caucasian with narcolepsy also have Dw2. Jacobson et al reported that the patients with 1MG have been reported to be associated with a RFLP proffle of HLA DR2 (Dw2) is found in Caucasian patients haplotype of HLA B8-DR3 [25—27]. It is interesting to note that with multiple sclerosis (MS) [35]. In contrast, Naito, Tabira and a HLA-DR2 subtype in Japanese and that of HLA-DR3 in Kuroiwa reported that there was no significant association with Caucasians were associated with susceptibility to 1MG. HLA- HLA in Japanese patients with MS [36]. Although it is unclear B8, B18 or DR3 are very rare in Japanese. Thus, no disease whether or not racial heterogeneities exist among these disassociated with these antigens has been reported in Japanese. eases, 1MG might be a good model for revealing the host factors The possibility may exist that the HLA-DR2 (Dw2)-DQw1 of susceptibility to diseases. haplotype in Japanese and HLA-B8-DR3 in Caucasians possess In conclusion, there may be a definite association between a a similar unknown gene susceptible to 1MG in linkage disequi- specific DQ gene and Japanese patients with 1MG. Two hypothlibrium specific to races. This would mean that the common eses are proposed. One is the existence of some specific HLA host factors to 1MG might be existent in HLA-DR2 (Dw2) in class II genes actually in the part of the pathogenesis of 1MG. Japanese and HLA-B8-DR3 in Caucasians. The other is the existence of unknown disease-susceptible Since the DRA gene is invariant in structure except for the genes in a linkage disequilibrium with HLA class II genes. cytoplasmic portion, allelic polymorphism in the DR subregion There might be two ways to clarify these two possibilities. One is found only in the DRB gene and most of the allelic variability is to ascertain the whale DNA sequences of the HLA class II is localized to the first domain of the DRB gene. Unlike the DR area in Japanese and Caucasian patients with 1MG and compare gene, polymorphism in the DQ gene is found both in DQA and them. The other is to find some common gene sequences
Ogahara et al: HLA-class II genes in 1MG
between the patients with 1MG with associated HLA haplotype and rare ones without the associated HLA haplotype. Acknowledgments This study was supported in part by the Grant from the Intractable Disease Division, Ministry of Health and Welfare, and by the grant-inaid for scientific research (No. 61304038) from the Ministry of Educa-
181
A method for oligonucleotide screening of highly complex gene
libraries. Proc NatI Acad Sci USA 82:1585—1588, 1985 16. SVEJGAARD SS, RYDER LP: Association between HLA and disease, in HLA and Disease, edited by DAU5SET J, SVEJGAARD A, Copenhagen, Munksgaard, 1977, pp. 46—71 17. S0ME5 GW, O'BRIEN KF: Odds ratio estimators, in Encyclopedia of Statistical Sciences (vol. 6), edited by KOTZ 5, JOHNSON NL, New York, A Wiley-Interscience Publication, 1985, pp. 407—410
tion, Science, and Culture of Japan. We are indebted to Dr. Akinori Kimura, Department of Genetics, Medical Institute of Bioregulation, Kyushu University, for his valuable comments and suggestions. The technical assistance of Miss Yumi Tukazaki, Miss Takako Nakamizo and Miss Kazumi Senda are gratefully acknowledged.
18. MARCADET A, MARTELL M, COHEN D, BIGNON JD, KOBAYASHI K, CARPENTER CB, WALFORD RL: RFLP standardization report
Reprint requests to Dr. Setsuya Naito, Second Department of 7-45-1,
I in Immunobiology of HLA, Histocompatibility Testing 1987, edited by DUPONT B, New York, Springer-Verlag, 1989, pp.
1. Hiici Y, KOBAYASHI 1, KASHIWAGI N: Strong association of HLA-DR2 and MT1 with idiopathic nephropathy in Japan. Kidney
20. SASAZUKI T, KANEOKA H, Tsuii K, NosE Y, KASHIWAGI N, KANEKO T: HLA-D in Japan, in Histocompatibility Testing 1980, edited by TEB.ASAKI P1, Los Angeles, UCLA Tissue Typing Laboratory, 1980, pp. 274—276 21. SIM0Ns MJ, WHEELER R, COHEN D, LALOUEL JM, DUPONT B: Restriction Fragment Length Polymorphism of HLA Genes: Summary of the Tenth Internal Workshop on Southern Blot Analysis, in Immunobiology of HLA, Histocompatibility Testing 1987, edited
Internal Medicine, Fukuoka University, School of Medicine, Nanakuma, Jonan-ku, Fukuoka 814-01, Japan.
References
mt 25:953—967, 1984 2. TOMURA S, KASHIWABARA H, SI-usHiDo H, TSUCHIDA H, SAKU-
iw S, MIYMIMA T, Tsui K, TAKEUCHI J: Strong association of idiopathic membranous nephropathy with HLA-DR2 and MTI in Japanese. Nephron 36:242—245, 1984
3. N.iio 5, KOHARA M, ARAKAWA K: Association of class II antigens of HLA with primary glomerulopathies. Nephron 45:111— 114, 1987 4. BAUR MP, NEUGEBAUER M, ALBERT ED: Reference Table of
Two-Locus Haplotype Frequencies for All MHC Marker Loci, in Histocompatibility Testing 1984, edited by ALBERT ED, BAUR MP,
DQ betalTaq I, in Immunobiology of HLA, Histocompatibility Testing 1987, edited by DUPONT B, New York, Springer-Verlag, for
1989, pp. 620—622 19. BONTROP RE, CARPENTER CB, WALFORD R, SEKIGUCHI S, BIGNON JD, COHEN D: RFLP standardization report for DQ AlphalTaq 815—816
by DUPONT B, New York, Springer-Verlag, 1989, pp. 959—1023
22. KOHARA M: Association between HLA and circulating immune complex in chronic glomerulonephritis. Jap J Nephrol 28:25—35, 1986 23. BUGAWAN TL, SAIKI RK, LEVENSON CH, WATSON RM, ERLICH
HA: The use of nonradioactive oligonucleotide probes to analyze enzymatically amplified DNA for prenatal diagnosis and forensic HLA typing. Bio-technol 6:943—947, 1988
MAYR WR, Berlin, Springer-Verlag, 1984, pp. 677—755 5. MARCADET A, DUPON B, COHEN D: Organization and Design of the
24. SACKS SH, BUSHELL A, RUST NA, KARAGIANNIS JA, JEWELL DP,
Southern Blot Component of the Tenth Histocompatibility Workshop, in Immunobiology of HLA, vol. 1, Histocompatibility Testing 1987, edited by DUPONT B, New York, Springer-Verlag, 1989, pp.
biological subtypes of the haplotype HLA-DR3 in patients with Celiac disease or idiopathic membranous nephropathy. Hum Im-
560—566
6. BODMER JG, MARSH SE, ALBERT ED, BODMER WF, DUPONT B,
ERLICH HA, MACH B, MAYR WR, PARHAM P, SASAZUKI T, SCHREUDER GM, STROMINGER JL, SVEJGAARD A, TEasAKI P1:
Nomenclature for factors of the HLA system, 1990. Tissue Antigens 37:97—104, 1991
7. KLOUDA PT, MANOS J, ACHESON EJ, DYER PA, GOLDBY FS, HARRIS R, LAWLER W, MALLICK NP: Strong association between idiopathic membranous nephropathy and HLA-DRw3. Lancet 13: 770—771, 1979
LEDINGHAM JG, WOOD KJ, MCMICHAEL AJ: Functional and munol 20:175—187, 1987 25. MULLER GA, MULLER C, LIEBAU G, KOMPF J, Isn H, WERNET
P: Strong association of idiopathic membranous nephropathy (IMN) with HLA-DR3 and MT-2 without involvement of HLA-B18 and no association to BfF1. Tissue Antigen 17:332—337, 1981 26. BERTHOUX FC, LAURENT B, LE-PETIT J-C, GENIN C, BROUTIN F, TOURAINE F, HA5SAN AA, CHAMPAILLER A: Immunogenetics and
immunopathology of human primary membranous glomerulonephritis, HLA-A, B, DR antigens; functional activity of splenic macrophage Fc receptors and peripheral blood T-lymphocyte subpopulations. Gun Nephrol 22:15—20, 1984
8. LE PETIT JC, LAURENT B, BERTHOUX FC: HLA-DR3 and idio-
27. PAPIHA SS, PAREEK SK, RODGER RS, MORLEY AR, WILKINSON R,
pathic membranous nephritis (IMN) association. Tissue Antigen
ROBERTS DF, KERR DN: HLA-A, B, DR and Bf allotypes in patients with idiopathic membranous nephropathy (IMN). Kidney
20:227—228, 1982
9. S.iu RK, BUGAWAN TL, HORN GT, MULLIS KB, ERLICH HA: Analysis of enzymatically amplified f-globin and HLA-DQa DNA with allele-specific oligonucleotide probes. Nature 324:163—166, 1986
Int 31:130—134, 1987 28. TODD JA, BELL JI, MCDEVITT HO: HLA-DQI3 gene contributes to susceptibility and resistance to insulin-dependent diabetes mellitus,
Nature
329:599—604,
1987
10. MARCADET A, O'CONNELL P, COHEN D: Standardization Southern
29. GREGERSEN PK, SILVER J, WINCHESTER RJ: The chared epitope
Blot Workshop Technique, in Immunobiology of HLA, Histocompatibility Testing 1987, edited by DUPONT B, New York, Springer-
hypothesis: An approach to understanding the molecular genetics of susceptibility to rheumatoid arthritis. Arthr Rheum 30:1205—
Verlag, 1989, pp. 553—560
11. SOUTHERN EM: Detection of specific sequences among DNA fragments separated by gel electrophoresis. J Mol Biol 98:503—517, 1975
12. FEINBERG AP, VOGELSTEIN BA: A technique for radiolabeling DNA restriction endonuclease fragments to high specific activity. Anal Biochem 132:6—13, 1983
13. DENHARDT DT: A membrane-filter technique for the detection of complementary DNA. Biochem Biophys Res Commun 23:641, 1966 14. SCHARF SJ, HORN GT, ERLICH HA: Direct cloning and sequence analysis of enzymatically amplified genomic sequences. Science 233:1076—1078, 1986
15. Woor WI, GITSCHIER J, LASKEY LA, LAWN RM: Base composition-independent hybridization in tetramethylammonium chloride:
1213, 1987 30. GREGERSEN PK, TODD JA, ERLICH HA, LONG E, SERVENIUS B, CH0I E, KAO HT, LEE JS: First domain sequence diversity of DR
and DQ subregion alleles, in Immunobiology of HLA, Histocompatibility Testing 1987, edited by DUPONT B, New York, SpringerVerlag, 1989, pp. 1027—1031 31. BROWN JH, JARDETZKY T, SAPER MA, SAMRA0UI B, BJORKMAN PJ, WILEY DC: A hypothetical model of the foreign antigen binding site of class II histocompatibility molecules, Nature 332:845—850, 1988
32. HONDA Y, Dol Y, JUJI T, SATAKE M: Narcolepsy and HLA: Positive DR2 as a prerequisite for the development of narcolepsy. Folia Psychiatr Neurol Jpn 38:360, 1984 33. INOKO H, ANDO A, TSUJI K, MATSUKI K, JIJJI T, HONDA Y:
182
Ogahara et al: HLA -class II genes in 1MG
HLA-DQ/3 chain DNA restriction fragments can differentiate between healthy and narcoleptic individuals with HLA-DR2. Immunogenetics 23:126—128, 1986
34. MATSUKI K, MAEDA H, Jun 1, INOKO H, ANDO A, Tsusi K, HONDA Y: Taq I-generated HLA-DQa polymorphism in Japanese patients with narcolepsy. Immunogenetics 27:87—90, 1988 35. JAKOBSON S, SORRENTINO R, NEPOM GT, MCFARLIN DE,
STROMINGER JL: DNA restriction fragment length polymorphism
of HLA-DR2: Correlation with HLA-DR2-associated function. J Neuroimmuno! 12:195—203, 1986
36. NAITO S, TABIRA T, KuRoiwA Y: HLA studies of multiple sclerosis in Japan, in Multiple Sclerosis East and West, edited by KUROIWA Y, KURLAND LT, Kyushu University Press, Fukuoka, Japan, 1982, pp. 215—222
Analysis of HLA class II genes in Japanese patients with idiopathic membranous glomerulonephritis SATORU OGAHARA, SETSUYA NAIT0, KASUMI ABE, ISAO MICHINAGA, and KIKuo ARAKAWA Second Department of Internal Medicine, Fukuoka University School of Medicine, Fukuoka, Japan
Analysis of HLA class II genes in Japanese patients with idiopathic membranous glomerulonephritis. Thirty unrelated Japanese adult patients with idiopathic membranous glomerulonephritis (1MG) and 50 unrelated Japanese normal controls were examined for HLA class II genes. Both the 3.0 kb DQB and the 6.2 kb DQA restriction fragments were simultaneously identified in 86.7% of the patients. These were
different between Japanese and Caucasians, specifically HLADR2-DQwl and DR3-DQw2 [7, 81, respectively.
length polymorphism (RFLP) of DQB and DQA of the patients coincided with that of homozygous cell lines with Dw2. The DRB probe showed no significant differences in the hybridization patterns between the patient group and HLA-matched controls. These findings indicate that the specific HLA-DQ polymorphism correlates primarily with
probes using the PCR technique [9]. We compared the published amino acid sequences of second exon of HLA class II genes related with Japanese and Caucasian 1MG.
In the present study, we attempted to ascertain the HLA
regions much more specifically associated with 1MG by examining the restriction fragment length polymorphisms (RFLP) of significantly different from the 14% frequency of the fragments in HLA class II genes. We also tried to determine the DQA area controls (relative risk = 39.9; P = 1.7 x l0°). The haplotype of HLA class II genes associated with 1MG using the polymerDR2-DQw1 associated with 1MG seemed to be Dw2 which was essentially different from the haplotype common to the normal Japanese ase chain reaction (PCR) method, since Saiki et al established it population, Dw12, This was because most of the restriction fragment for HLA DQA typing with sequence-specific oligonucleotide
Methods
Japanese 1MG.
Patients, controls and cell lines Specific HLA class II alloantigens, HLA-DR2 and -DQw1, have been found in Japanese patients with idiopathic membranous glomerulonephritis (1MG) about eight times more frequently than in the normal Japanese population [1—3]. Since HLA-DR2 and DQw 1 were very closely linked to each other [4], the association of HLA-DR2 and -DQw1 with 1MG was considered to be due to that of a haplotype of HLA-DR2-DQwl with 1MG, However, the association clarified by serologically
typed HLA was too rough to define the "host factors for susceptibility" to 1MG.
HLA-DR and -DQ regions were subdivided according to specificities which were characterized by mixed lymphocyte culture. Recent advances in DNA technology have made it possible to define almost HLA-D clusters [5] and HLA class II gene sequences [6], and depict some DNA sequences of HLA
Thirty unrelated Japanese patients were diagnosed as having 1MG clinically and by renal biopsy, and then were determined of their HLA phenotype. At the time of renal biopsy, their mean age was 56.0 12.1 years. Ten patients were male and 20 were female. Fifty normal persons were chosen as random unrelated Japanese controls and 24 individuals with HLA-DR2 were used as HLA-DR-matched controls. Epstein-Barr virus-transformed B-lymphoblastoid cell lines which were homozygous for HLADR2 were used to produce contrast electrophoretic pictures. The D02089l5, WT8, SHCU and MGAR cell lines as Dw2, the E4181324 cell as Dwl2, the KASO1I as Dw21 cell line, and the RML as Dw22 were obtained from the Tenth International HLA Workshop core cell lines. Serological HLA typing
HLA-A, -B, -C, -DR and -DQ typings were done by the
class II genes being responsible for the susceptibility of diseases standard microdroplet lymphocytotoxicity test, using antisera like insulin-dependent diabetes mellitus (IDDM) and rheuma- from the Tenth International HLA Workshop, Japanese retoid arthritis (RA) from the serological association of HLA with gional antisera and the local antisera comparable to them.
those diseases. The haplotype of HLA-DR2-DQw1 has also Restriction fragment length polymorphism analysis (RFLP) been known to be one of the most common among Japanese, The analysis procedure was performed according to the and is able to be divided into at least four distinct HLA-D standardized Southern blot workshop technique established at clusters. Furthermore, the haplotype associated with 1MG is
the Tenth International HLA Workshop [10]. In brief, genomic DNA samples were obtained by phenolchloroform extraction of sodium dodecyl sulfate (SDS) and by lysing cells with protein-
Received for publication February 7, 1990 and in revised form August 12, 1991 Accepted for publication August 19, 1991
ase K from peripheral blood leukocytes of patients, controls and the cell lines. The probes used for hybridization were 0.78
© 1992 by the International Society of Nephrology
provided from the Tenth International HLA Workshop [5]. The
kb DRB, 1.2 kb DQB and the 1.6 kb DQA cDNA clone, 175
176
Ogahara et a!: HLA-class 11 genes in 1MG
A 10
DOA1 * 0101
0102 0103
40
50
60
80
70
Q
0201 0301
Y
0401
Y Y
0501
30
20
DHVASCGVNLYQFYGPSGQYTHEFDGDEEFYVDLERKETAWRWPEFSKFGGFDPQGALRNMAVAKHNLNIMIKRYNSTAATN S S
Y
S S S
0601
--n-
-
F F
K
V-KL—L—HRLR ————F——T—I——L S
Q Q Q
F
L———S
V—QL—L—RR—RR————F——T—I——L
V———S
G———-V—CL—VLRQ—R —-——F——T—I——L G-———V—CL—VLRQ—R ————F——T—I——T G————V-CL-VLRQ—R ————F——T—I——T
SL———S L———S L———S
GH27
GH26
B
DQA1 Specificity
5' SSo
Sequence
3'
1 CACGTAGAACTCCTCATC 2 CACGTAGAACTGCTCATC
3 CTCCTTCCTCTCCAGGTC 4 GTCTCCTTCTTCTCCAGG 5 CAGAGGCAACTTCCAGAC
6 ATCTTCTAAATCTGCGGA 7 TCAGACTGTTCAAGTTAT
8 CGTAGAACTGCTCGTCTC 9 TCATGGGTGTACTGGCCA 10 TCATGGGTGAACTGGCCA
aa
0101
01020103 0201 03010401 0501 0601
32-37 * 32-37 38-43 * 39-44 45-50 51-56
* *
*
* *
* * *
*
71-7 6
31-36 23-28 * 23-28
*
* *
*
*
*
*
*
*
Note, aa, amino acid position; , positive spot
Fig. 1. A. Comparison of amino acid sequences of DQA. The PCR primers (GH26, GH27) and their relationship to the DQA genes by arrows. B. The sequence of allele-specific oligonucteotide probes and their specificity.
restriction endonuclease, Taq I (Toyobo Co. Ltd., Tokyo) were used to digest the genomic DNA. Ten g of the endonuclease-
digested DNAs were subjected to electrophoresis on 0.7% agarose gel for 40 hours at 35 volts in 40 mvt Tris-HC1, 20 mM glacial acetic acid and 2 mM EDTA (TAE buffer) [5]. The gel was soaked in 0.15 M HCI for ten minutes, and then soaked in
0.4 N NaOH for 30 minutes. The gel was transferred onto a nylon membrane (Gene Screen Plus, NEN, Massachusetts, USA) by the method of Southern [11]. DNAs were transferred overnight with 0.4 N NaOH solution as transfer buffer, The probes were labeled with [a-32P] dCTP by the multiprime labeling method [121. The membranes were prehybridized overnight in prehybndization mix (50% deionized formamide, 0.1% Denhardt solution [13], 5x SSPE (30x SSPE is 5.4 M NaC1, 0.3 M NaH2PO4 and 30 mst EDTA), 0.1% SDS, 5% Dextran sulfate
Amp1ficarion of genoinic DNA
The second exon of the DQA gene was amplified from genomic DNA with the use of PCR. The PCR primers used were GH26 and GH27 as described [14] (Fig. lA). DNAs were prepared from a part of the RFLP analysis samples. One g of
genomic DNA was amplified by PCR. Two units of Taq polymerase (Stratagene, CA) were added to a 100 1.d volume of polymerase buffer containing 10 mrvi Tris-HC1 pH 8.4, 50 mM KCI, 2.5 mr't MgCl2 and 100 g/ml gelatin, overlaid with mineral
oil. The reaction was adjusted to 200 M for each of the four deoxynucleoside triphosphates and 0.2 /LM for each of the PCR primer pairs, amplified by using a programmable thermal cycler (Perkin Elmer-Cetus Co., Norwalk, Connecticut, USA). Each cycle consisted of a 30 second denaturation step at 95°C, a 55°C annealing step, and a two minute extension step at 72°C except for initial denaturation for five minutes at 95°C, and a final seven
and 100 tg/m1 Herring sperm denatured DNA) at 42°C with agitation. The membranes were hybridized in prehybridization minute extension step at 72°C. The PCR products were conmix with labeled probes for 48 hours at 42°C. The membranes firmed by the electrophoresis in a 1.5% agarose gel. were washed with 2 x SSPE twice for five minutes at room temperature, and with 2x SSPE and 0.5% SDS once for 15 Dot typing of PCR DQA products minutes at 65°C, and then with 0.5x SSPE for 15 minutes at 65°C. The membranes were subsequently dried and exposed to Ten sequence-specific oligonucleotide (SSO) probes were an X-OMAT-AR film (Kodak, Rochester, New York, USA) designed and supplied by the Tenth Japan HLA Workshop (Fig. from overnight to seven days at —80°C. IB). Thirty to 50 ng of specific PCR products was spotted onto
177
Ogahara et alt HLA-class 11 genes in 1MG
Table 1. HLA-A, B, DR, DQ type, RFLP of DQB and DQAI genes and subregion of DQAI of patients with 1MG
RFLP HLANo.
A
Patient
I
M. MOR.
2
T.YAM.
2 24
3 4 5
M. MIZ.
2
6
T. ISH.
11
M. FUJ. A. HAS.
24 24 2 2 26 2 24
B 11
60
31
7
24 24 26 31
39
11
54
24
13
7 8 9
H. MAE.
10 11
K. KIM. Y. ARA.
12
A. HIG.
11
26
13
N. MAS.
31
14 15
T. KUS. T. TOK.
24 24 24
16
S. MIT.
11
17
H. YOS. T. TAN.
31
18 19 20 21
22 23 24 25 26 27 28 29 30
K. FUJ. M. HIR.
24
26 26 24 26 26
S. OOT.
2 2 24 24 26 2
K. 00K.
24
26 33 33
S. MIY.
T. 00K. K. YOS. 5. 00K.
U. KUM. T. TIK.
T. TAK. H. SAl.
M. NAK. K. KUM.
a , existent;
—,
2
24
DQB
31 11
2 2 2
31
11
24
31 31
51 7 51
62 35 7 61 54 52 51 35 35 52 39 48 7 35 35 7
DR 61 51 56 60 35
52
2
8
1
2 2 2 2 2 2
1
1
8 8
1
1
1
60
2
54
2 2
52 62 62 35
DQ
1
1
4
1
12
1
1
1
60
13
1
7 62 35 54
51
4 9 11
7 7 4
59 59
— —
5
12
1
+ + + + + + + + + + + + + + + + + + + + + + +
--
13
4 4
+ + + + + + + + +
61
51
DQA1
1
44 44
55 55 60
7.0 kb
1
39 35
48 39
DOT
6.2 kb
4 9 4
2 2 2 2 2 2 2 2 2 2 2 2 2 2
60
4 7
DQA1
3.0 kb
1 1
3 3
1
8 8 8
1
4-
1 1
3
1
1
1
8
1
1
1
3
1
1
12
4
1
+ --
1
9 14
+ + + + +
7 3
+ + + + + + + + — — —
—
+ +
— —
0102 0102 0102 0102 0102 0102 0102 0102 0102 0102 0102 0102 0102 0102 0102 0102 0102 0102 0102 0102 0102 0102
— —
— —
0101 0401
— —
—
0401
—
0301
--
—
+ +
0102 0102 0102 0102
+ — — — — — — —
+ — — —
+ — — — — — —
0103 0101
0103 0103 0101
0103 0301 0401 0301 0301 0301 0103
0301 0103 0301 0101 0101 0103 0101 0401 0301 0301
not existent
nylon membranes, which were presoaked in soaking solution [0.15 M NaC1, 10 m sodium phosphate, 1 mrsi EDTA (pH 7.4)]
for 15 minutes, and then dried at 60°C. Membranes were rewetted with 0.4 N NaOH for one minute for denaturation, and then soaked in a soaking solution for five minutes, and exposed to UV light to fix the PCR products for 30 seconds. SSO probes were phosphorylated at their 5'- termini with [y-32P] ATP and
by convention were expressed as relative risks. When one element of the equation was zero, the relative risk was calculated with the formula of Haldane [17]: [(a +
112)(d + 1/2)]
[(b +
1/2)(c +
l/2)f
Results T4 polynucleotide kinase (TOYOBO, Osaka, Japan). Membranes were prehybridized in a hybridization solution (50 mM Table 1 lists HLA-A, B, DR and DQ antigens typed serologTris-HCI pH 8.0), 3M tetramethylammonium chloride [151, 2 ically, existence of the specific DQB and DQA RFLP, and DQA mM EDTA (pH 8.0), 5x Denhardt's solution [13], 0.1% SDS subtypes by dot typing of the 30 Japanese patients with 1MG. and 100 j.tg/ml denaturated Herring sperm DNA) for one hour at Hybridization with a DQB probe of genomic DNA digested 54°C. Hybridization with 32P-labeled SSO probes was accom- with Taq I showed two constants (3.4 and 4.5 kb) and at least plished by the addition of a 2 pmol SSO probe in 5 ml of the two polymorphic bands (3.0 and 5.3 kb). Hybridization with a hybridization solution for one hour at 54°C. The membranes DQA probe of genomic DNA digested with Taq I showed at were rinsed twice in 0.3 M NaCI, 20 m sodium phosphate, 2 least five polymorphic DQA1 bands (2.8, 4.6, 5.5, 6.2 and 7.0 mM EDTA, and 0.1% SDS for 10 minutes at room temperature, kb) and two polymorphic DQA2 bands (2.3 and 2.0 kb). These washed in 50 m Tris-HC1 (pH 8.0), 3 M tetramethylammonium hybridization patterns were consistent with those described by chloride, 2 mt EDTA and 0.1% SDS for 10 minutes at room Marcadet et a! [181 and Bontrop et a! [19]. temperature and then for 30 minutes at 58°C, and exposed for Table 2 presents the antigen phenotypic frequencies of patients and controls with HLA-DR2, -DRw53 and -DQw1, and autoradiography with an intensifying screen for one hour. the frequencies of patients and controls with the Taq I-digested Statistical analysis 3.0 kb DQB and 6.2 kb DQA restriction fragments. Both of Statistical analysis was performed by the chi-square method these fragments were simultaneously identified in 86.7% (26 of and Fisher's exact test. P values were multiplied by the number 30) of patients. This percentage was significantly different from of antigens and fragments tested (corrected P) [16]. Odds ratios the 14% (7 of 50) frequency of the fragments between patients were calculated with the use of the formula (a x d)/(b X c) and and controls identified in controls. The relative risk of these
178
ogahara ci al: HLA-class II genes in 1MG
Table 2. Comparison of frequencies of HLA serological class II typing and fragments of DQB and DQA RFLP
Serology DR2 DRwS3 DQw1
RFLP DQB 3.0 kb DQA 6.2 kb
Patients = 30)
Controls
(N = 50)
Relative risk
80.0% 36.7% 90.0% 86.7% 86.7%
30.0% 60.0% 58.0% 14.0% 14.0%
9.3 0.4 6.5 39.9 39.9
P 1.5 X 0.04 2.5 1.7
x iO x io'°
1.7 x 10—0
Corrected P value
1.8 x iO NS
0.013
1.7 x iO—
1.7 x lO
NS, not significant.
LC)
A
cc CD
I0
'-C)
c'.1
B cc
00
ccw
w
Cl)
kb
5.3 4.5
kb
_ 7.0
iI:I!i!:• •II.e.— a
S
6.2
- 4.4
3.4 3.0
eeeØee
0.9
Dw 2222221212 Dw2 222221212 DQB
Taq I
OQA
Fig. 2. Electrophoretic patterns of Taq Idigested DNAs from homozygous cell lines hybridized with (A) DQB and (B) DQA probes.
fragments were much higher than those of DR2 or DQwI (39.9 lines with Dw2 and Dw2 1. The 4.4 kb fragment was found only vs. 9.3 or 6.5), and the statistical differences were much more in Dw22. The 7.0 kb fragment was found only in Dwl2. significant in the fragments than serologically typed antigens, Twenty-four Japanese adult patients with 1MG and a haploDR2 and/or DQw1 (1.7 x 10_b vs. 1.5 x l0 or 2.5 x 10). type of HLA-DR2-DQw 1 (DR2-patients) and twenty-four JapFigure 2A shows that electrophoresis of genomic DNA of anese normal controls with the same haplotype (HLA-matched each homozygous cell lines digested with Taq I produced four controls) were examined. Figure 3 presents the electrophoretic restriction fragment patterns when hybridized with the DQB patterns of Taq I-digested DNAs from a portion of the patients probe. The 4.5 and 3.4 kb fragments were observed in all cell and HLA-matched controls hybridized with DQB probes. All lines. The 3.0 kb fragment was observed in all cell lines with DR2-patients had the 3.0 kb fragment. This fragment was found Dw2, but was not found in the cell lines with Dwl2, 21 and 22. in only three out of 24 HLA-matched controls (RR 301; P = The 0.9 kb fragment was found in only the cell lines with Dw22. 1.0 x 10). The 5.3 kb fragment was found in only the cell line with Dw21. Figure 4 illustrates the electrophoretic patterns of Taq I-diThe cell line with Dw12 did not have 3.0, 0.9 or 5.3 kb
gested DNAs from a portion of the DR2-patients and HLAmatched controls hybridized with DQA probes. The 6.2 kb fragment was also found in all DR2-patients. In contrast, this easily definable. Figure 2B shows that the electrophoresis of fragment was found in only 12.5% of the HLA-matched con301; P = 1.0 x l0—). The 7.0 kb fragment was genomic DNA of each homozygous cell lines digested with Taq trols (RR fragments. Since the number of fragments hybridized with DQA was less than that of DQB, the polymorphisms of DQA fragments were
I produced three restriction fragment patterns when hybridized found in 25% of the DR2-patients, but was found in 88% of the with the DQA probe. The 6.2 kb fragment was found in all cell HLA-matched controls (RR = 0.04; P = 3.7 x 10—s).
179
Ogahara et al: HLA-class II genes in 1MG B Normal
A 1MG
kb!flK Fig. 3. Electrophoretic patterns of Taq I-
es
digested DNAs from a portion of (A) patients with 1MG and (B) normal controls hybridized with DQB probes.
DQB/Taq I
A 1MG
B Normal
._aM_eu
7.0—.-
-
-
kb
—e
—.7.0 6.2
5.5 •' 4.6 2.8 — —
-a S
SS — DQA/Taq I
S
13.8—
S
Fig. 4. Electrophoretic patterns of Taq 1digested DNAs from a portion of (A) patients with 1MG and (B) normal controls hybridized with DQA probes.
B Normal
A 1MG
5.8 —
— 2.8
Nd.
a
as ••1
—a
—'-S
S —
j
I
'a.
S. 1.4——.
-1'
U DRB/Taq I
Figure 5 illustrates the electrophoretic patterns of Taq I-digested DNAs from a portion of the DR2-patients and HLAmatched controls hybridized with DRB probes. The DRB probe gave similar electrophoretic patterns between the patient group
and the controls. The 1.4 kb fragment was found in all DR2 patients and HLA-matched controls. The DQA subregion of all patients with 1MG was determined by DQA dot typing. Twenty-six patients were assigned to have
DQA1*0102, six patients, DQA1*010l, seven patients, DQA1 *0103, nine patients, DQA 1*0301, and four patients,
Fig. 5. Electrophoretic patterns of Taq Idigested DNAs from a portion of (A) patients with 1MG and (B) normal controls hybridized with DRB probes.
DQA 1*0401. However, none of the patients were assigned to have DQA1*0201, DQA1*050l or DQA1*0601 (Table 1). Discussion
The present study suggests the existence of a strong association between Japanese 1MG and some specific fragments of RFLP in DQ. The allele was represented by the 3.0 kb fragment
of DQB RFLP and the 6.2 kb of DQA RFLP. Since these fragments were thought to correspond to the HLA-Dw2 haplo-
type when they were compared to the patterns of RFLP of
180
Ogahara et a!: HLA-class II genes in 1MG
HTCs with DR2 related D type, Dw2, Dw12, Dw2l and Dw22,
86.7% of patients and 12.5% of the normal controls were
DQB. Todd, Bell and McDevitt reported that the specific amino
acid residue of DQ molecules could confer the resistance to suggested to have Dw2. Most of the controls with HLA-DR2 IDDM [28], and Gregersen, Silver and Winchester reported that probably had Dw12 as seen in a report on the normal Japanese specific amino acid residues of DR molecules could confer the susceptibility to RA [29]. Therefore, it may be possible to point population [201. On the other hand, the patterns of DRB RFLP digested with out common amino acid residues of class II molecules which Taq I were very similar between patients with DR2 and HLA- showed a strong association with 1MG in Japanese and also in matched controls. RFLP of DRB was clarified by using a Caucasians. The published amino acid sequence of the second variety of restriction enzymes such as BamH I, Bgl II, EcoR I, exon at the DQA1, DQB1, DRB1 DRB3, and DRB5 genes EcoR V, Hinc II, Hind III, Kpn I, Msp I, Pst I, Pvu II, Sst I and coding HLA-DR2, DR3, DQw1 and DQw2 antigens were comTaq I at the Tenth HLA International HLA Workshop [211. pared with each other [30]. There was no existence of a However, there were few fragments distinguishing between common specific sequence in DQA1*0102 and DQA1*0401 Dw2 and Dw12. genes, in DQB*0602 and DQB1*0201 genes, and in DRB1*1501 The present study revealed that patients with 1MG commonly and DRB 1*0301 genes, respectively. The DRB 1*0102 gene have a relatively less frequent DR2-positive haplotype assigned indicated a strong linkage disequilibrium with the DRB5*0l01 to Dw2 in the Japanese population in contrast to the Bw52-DR2 gene, and the DRB1*0301 gene with the DRB3*0101 gene, assigned to Dw12 more frequently found in the same population respectively. When the amino acid sequence of the DRB5*010l [22]. Two of six patients without DR2 also had both the DQB gene was compared with that of the DRB3*0101 gene, both 3.0 kb and DQA 6.2 kb fragments. These patients (patients No. genes were found to have leucine at position 38. It was known 25 and 26) were typed as HLA-DRw13. Although DRw13 was that the DRB 1 * 1201 genes also had leucine at the same position. less frequent in patients because of the much more frequent Almost all other DRB3 or B5 genes have valine or alanine at occurrence of DR2, the possibility exists that DRw13 might be position 38. Residue 38 of the first domain of the DRB gene was associated with the susceptibility to 1MG. This is because included in the second hypervariable region (residues 25-38). DRwl3 has a linkage disequilibrium with DQw1, and the According to the predicted structure of class II molecules [311, frequency of DQw1 was higher than that of DR2 in the patients. this region is located in the base of the cleft formed by the Patients with DQA1*0102 simultaneously exhibited both frag- extracellular domain of class II molecules. If allelic polymorments of 3.0 kb DQB and 6.2 kb DQA, while patients with phisms in the second hypervariable region determine the funcDQA1*0103 had fragments of 7.0 kb DQA except for patients tional effects on the antigen binding and on the T cell recogniNo. 1 and 22. It is known that DQA1*0102 was in strong linkage tion, these polymorphisms might confer the susceptibility to disequilibrium with DR2(Dw2)-DQw6, DR2(Dw21)-DQw5 and 1MG. Since these sequences were from the literature, it is DRw13(Dw19)-DQw6, whereas the DQA1*0l03 was in strong essential to sequence the class II genes from Japanese and linkage disequilibrium with DRwl3 (Dw18)-DQw6, DR2 Caucasian patients with 1MG if common factors exist between (Dw12)-DQw6 and DRw8-DQwI [23]. These results thus sup- HLA-DR2 (Dw2) in Japanese and HLA-DR3 in Caucasians. port the fact that patients with DQA1*0102 had DR2 (Dw2) or Also important is whether or not these genes may have a DRw13 (Dw19). DQA dot typing clearly confirmed that some linkage disequilibrium with common susceptible genes of 1MG. Some diseases have been reported to be associated with the patients with the serologically homozygous gene of DQ (No. 6, 14 and 18) were able to be assigned as a heterozygote. HLA-Dw2 haplotype. Honda et al first reported that all JapaIn Caucasians, there are HLA-B genes in the strong linkage nese patients with narcolepsy had HLA-DR2 [32]. Inoko et al disequilibrium with HLA-DR3 forming two haplotypes, that is, demonstrated the 100% presence of the DQB EcoR I 2.4 kb, HLA-B8-DR3 and HLA-B18-DR3. Sacks et al reported that BamH I 2.9 kb, and Pst 112 kb fragments in narcolepsy [33]. B8-linked DR3 is different from B 18-linked DR3 from the Matsuki et al described the differences between Dw2 and Dw12 analysis of the DRB genes by RFLP, biochemical studies of DR using Taq I-generated DQB RFLP and reported that all patients molecules by two-dimensional gel electrophoresis and func- with narcolepsy had Dw2 [34]. Unlike 1MG, Caucasian patients tional studies of mixed lymphocyte reactions [24]. Caucasian with narcolepsy also have Dw2. Jacobson et al reported that the patients with 1MG have been reported to be associated with a RFLP proffle of HLA DR2 (Dw2) is found in Caucasian patients haplotype of HLA B8-DR3 [25—27]. It is interesting to note that with multiple sclerosis (MS) [35]. In contrast, Naito, Tabira and a HLA-DR2 subtype in Japanese and that of HLA-DR3 in Kuroiwa reported that there was no significant association with Caucasians were associated with susceptibility to 1MG. HLA- HLA in Japanese patients with MS [36]. Although it is unclear B8, B18 or DR3 are very rare in Japanese. Thus, no disease whether or not racial heterogeneities exist among these disassociated with these antigens has been reported in Japanese. eases, 1MG might be a good model for revealing the host factors The possibility may exist that the HLA-DR2 (Dw2)-DQw1 of susceptibility to diseases. haplotype in Japanese and HLA-B8-DR3 in Caucasians possess In conclusion, there may be a definite association between a a similar unknown gene susceptible to 1MG in linkage disequi- specific DQ gene and Japanese patients with 1MG. Two hypothlibrium specific to races. This would mean that the common eses are proposed. One is the existence of some specific HLA host factors to 1MG might be existent in HLA-DR2 (Dw2) in class II genes actually in the part of the pathogenesis of 1MG. Japanese and HLA-B8-DR3 in Caucasians. The other is the existence of unknown disease-susceptible Since the DRA gene is invariant in structure except for the genes in a linkage disequilibrium with HLA class II genes. cytoplasmic portion, allelic polymorphism in the DR subregion There might be two ways to clarify these two possibilities. One is found only in the DRB gene and most of the allelic variability is to ascertain the whale DNA sequences of the HLA class II is localized to the first domain of the DRB gene. Unlike the DR area in Japanese and Caucasian patients with 1MG and compare gene, polymorphism in the DQ gene is found both in DQA and them. The other is to find some common gene sequences
Ogahara et al: HLA-class II genes in 1MG
between the patients with 1MG with associated HLA haplotype and rare ones without the associated HLA haplotype. Acknowledgments This study was supported in part by the Grant from the Intractable Disease Division, Ministry of Health and Welfare, and by the grant-inaid for scientific research (No. 61304038) from the Ministry of Educa-
181
A method for oligonucleotide screening of highly complex gene
libraries. Proc NatI Acad Sci USA 82:1585—1588, 1985 16. SVEJGAARD SS, RYDER LP: Association between HLA and disease, in HLA and Disease, edited by DAU5SET J, SVEJGAARD A, Copenhagen, Munksgaard, 1977, pp. 46—71 17. S0ME5 GW, O'BRIEN KF: Odds ratio estimators, in Encyclopedia of Statistical Sciences (vol. 6), edited by KOTZ 5, JOHNSON NL, New York, A Wiley-Interscience Publication, 1985, pp. 407—410
tion, Science, and Culture of Japan. We are indebted to Dr. Akinori Kimura, Department of Genetics, Medical Institute of Bioregulation, Kyushu University, for his valuable comments and suggestions. The technical assistance of Miss Yumi Tukazaki, Miss Takako Nakamizo and Miss Kazumi Senda are gratefully acknowledged.
18. MARCADET A, MARTELL M, COHEN D, BIGNON JD, KOBAYASHI K, CARPENTER CB, WALFORD RL: RFLP standardization report
Reprint requests to Dr. Setsuya Naito, Second Department of 7-45-1,
I in Immunobiology of HLA, Histocompatibility Testing 1987, edited by DUPONT B, New York, Springer-Verlag, 1989, pp.
1. Hiici Y, KOBAYASHI 1, KASHIWAGI N: Strong association of HLA-DR2 and MT1 with idiopathic nephropathy in Japan. Kidney
20. SASAZUKI T, KANEOKA H, Tsuii K, NosE Y, KASHIWAGI N, KANEKO T: HLA-D in Japan, in Histocompatibility Testing 1980, edited by TEB.ASAKI P1, Los Angeles, UCLA Tissue Typing Laboratory, 1980, pp. 274—276 21. SIM0Ns MJ, WHEELER R, COHEN D, LALOUEL JM, DUPONT B: Restriction Fragment Length Polymorphism of HLA Genes: Summary of the Tenth Internal Workshop on Southern Blot Analysis, in Immunobiology of HLA, Histocompatibility Testing 1987, edited
Internal Medicine, Fukuoka University, School of Medicine, Nanakuma, Jonan-ku, Fukuoka 814-01, Japan.
References
mt 25:953—967, 1984 2. TOMURA S, KASHIWABARA H, SI-usHiDo H, TSUCHIDA H, SAKU-
iw S, MIYMIMA T, Tsui K, TAKEUCHI J: Strong association of idiopathic membranous nephropathy with HLA-DR2 and MTI in Japanese. Nephron 36:242—245, 1984
3. N.iio 5, KOHARA M, ARAKAWA K: Association of class II antigens of HLA with primary glomerulopathies. Nephron 45:111— 114, 1987 4. BAUR MP, NEUGEBAUER M, ALBERT ED: Reference Table of
Two-Locus Haplotype Frequencies for All MHC Marker Loci, in Histocompatibility Testing 1984, edited by ALBERT ED, BAUR MP,
DQ betalTaq I, in Immunobiology of HLA, Histocompatibility Testing 1987, edited by DUPONT B, New York, Springer-Verlag, for
1989, pp. 620—622 19. BONTROP RE, CARPENTER CB, WALFORD R, SEKIGUCHI S, BIGNON JD, COHEN D: RFLP standardization report for DQ AlphalTaq 815—816
by DUPONT B, New York, Springer-Verlag, 1989, pp. 959—1023
22. KOHARA M: Association between HLA and circulating immune complex in chronic glomerulonephritis. Jap J Nephrol 28:25—35, 1986 23. BUGAWAN TL, SAIKI RK, LEVENSON CH, WATSON RM, ERLICH
HA: The use of nonradioactive oligonucleotide probes to analyze enzymatically amplified DNA for prenatal diagnosis and forensic HLA typing. Bio-technol 6:943—947, 1988
MAYR WR, Berlin, Springer-Verlag, 1984, pp. 677—755 5. MARCADET A, DUPON B, COHEN D: Organization and Design of the
24. SACKS SH, BUSHELL A, RUST NA, KARAGIANNIS JA, JEWELL DP,
Southern Blot Component of the Tenth Histocompatibility Workshop, in Immunobiology of HLA, vol. 1, Histocompatibility Testing 1987, edited by DUPONT B, New York, Springer-Verlag, 1989, pp.
biological subtypes of the haplotype HLA-DR3 in patients with Celiac disease or idiopathic membranous nephropathy. Hum Im-
560—566
6. BODMER JG, MARSH SE, ALBERT ED, BODMER WF, DUPONT B,
ERLICH HA, MACH B, MAYR WR, PARHAM P, SASAZUKI T, SCHREUDER GM, STROMINGER JL, SVEJGAARD A, TEasAKI P1:
Nomenclature for factors of the HLA system, 1990. Tissue Antigens 37:97—104, 1991
7. KLOUDA PT, MANOS J, ACHESON EJ, DYER PA, GOLDBY FS, HARRIS R, LAWLER W, MALLICK NP: Strong association between idiopathic membranous nephropathy and HLA-DRw3. Lancet 13: 770—771, 1979
LEDINGHAM JG, WOOD KJ, MCMICHAEL AJ: Functional and munol 20:175—187, 1987 25. MULLER GA, MULLER C, LIEBAU G, KOMPF J, Isn H, WERNET
P: Strong association of idiopathic membranous nephropathy (IMN) with HLA-DR3 and MT-2 without involvement of HLA-B18 and no association to BfF1. Tissue Antigen 17:332—337, 1981 26. BERTHOUX FC, LAURENT B, LE-PETIT J-C, GENIN C, BROUTIN F, TOURAINE F, HA5SAN AA, CHAMPAILLER A: Immunogenetics and
immunopathology of human primary membranous glomerulonephritis, HLA-A, B, DR antigens; functional activity of splenic macrophage Fc receptors and peripheral blood T-lymphocyte subpopulations. Gun Nephrol 22:15—20, 1984
8. LE PETIT JC, LAURENT B, BERTHOUX FC: HLA-DR3 and idio-
27. PAPIHA SS, PAREEK SK, RODGER RS, MORLEY AR, WILKINSON R,
pathic membranous nephritis (IMN) association. Tissue Antigen
ROBERTS DF, KERR DN: HLA-A, B, DR and Bf allotypes in patients with idiopathic membranous nephropathy (IMN). Kidney
20:227—228, 1982
9. S.iu RK, BUGAWAN TL, HORN GT, MULLIS KB, ERLICH HA: Analysis of enzymatically amplified f-globin and HLA-DQa DNA with allele-specific oligonucleotide probes. Nature 324:163—166, 1986
Int 31:130—134, 1987 28. TODD JA, BELL JI, MCDEVITT HO: HLA-DQI3 gene contributes to susceptibility and resistance to insulin-dependent diabetes mellitus,
Nature
329:599—604,
1987
10. MARCADET A, O'CONNELL P, COHEN D: Standardization Southern
29. GREGERSEN PK, SILVER J, WINCHESTER RJ: The chared epitope
Blot Workshop Technique, in Immunobiology of HLA, Histocompatibility Testing 1987, edited by DUPONT B, New York, Springer-
hypothesis: An approach to understanding the molecular genetics of susceptibility to rheumatoid arthritis. Arthr Rheum 30:1205—
Verlag, 1989, pp. 553—560
11. SOUTHERN EM: Detection of specific sequences among DNA fragments separated by gel electrophoresis. J Mol Biol 98:503—517, 1975
12. FEINBERG AP, VOGELSTEIN BA: A technique for radiolabeling DNA restriction endonuclease fragments to high specific activity. Anal Biochem 132:6—13, 1983
13. DENHARDT DT: A membrane-filter technique for the detection of complementary DNA. Biochem Biophys Res Commun 23:641, 1966 14. SCHARF SJ, HORN GT, ERLICH HA: Direct cloning and sequence analysis of enzymatically amplified genomic sequences. Science 233:1076—1078, 1986
15. Woor WI, GITSCHIER J, LASKEY LA, LAWN RM: Base composition-independent hybridization in tetramethylammonium chloride:
1213, 1987 30. GREGERSEN PK, TODD JA, ERLICH HA, LONG E, SERVENIUS B, CH0I E, KAO HT, LEE JS: First domain sequence diversity of DR
and DQ subregion alleles, in Immunobiology of HLA, Histocompatibility Testing 1987, edited by DUPONT B, New York, SpringerVerlag, 1989, pp. 1027—1031 31. BROWN JH, JARDETZKY T, SAPER MA, SAMRA0UI B, BJORKMAN PJ, WILEY DC: A hypothetical model of the foreign antigen binding site of class II histocompatibility molecules, Nature 332:845—850, 1988
32. HONDA Y, Dol Y, JUJI T, SATAKE M: Narcolepsy and HLA: Positive DR2 as a prerequisite for the development of narcolepsy. Folia Psychiatr Neurol Jpn 38:360, 1984 33. INOKO H, ANDO A, TSUJI K, MATSUKI K, JIJJI T, HONDA Y:
182
Ogahara et al: HLA -class II genes in 1MG
HLA-DQ/3 chain DNA restriction fragments can differentiate between healthy and narcoleptic individuals with HLA-DR2. Immunogenetics 23:126—128, 1986
34. MATSUKI K, MAEDA H, Jun 1, INOKO H, ANDO A, Tsusi K, HONDA Y: Taq I-generated HLA-DQa polymorphism in Japanese patients with narcolepsy. Immunogenetics 27:87—90, 1988 35. JAKOBSON S, SORRENTINO R, NEPOM GT, MCFARLIN DE,
STROMINGER JL: DNA restriction fragment length polymorphism
of HLA-DR2: Correlation with HLA-DR2-associated function. J Neuroimmuno! 12:195—203, 1986
36. NAITO S, TABIRA T, KuRoiwA Y: HLA studies of multiple sclerosis in Japan, in Multiple Sclerosis East and West, edited by KUROIWA Y, KURLAND LT, Kyushu University Press, Fukuoka, Japan, 1982, pp. 215—222
