Distribution of HLA class II alleles and haplotypes in insulin-dependent moroccan diabetics

ELSEVIER

Distribution of HLA Class II Alleles and Haplotypes in Insulin-Dependent Moroccan Diabetics Hassan Izaabel, Henri-Jean Garchon, Geneviitve Beaurain, Mohamed Biga, Omar Akhayat, Jean-Fraqois Bach, and Sophie Caillat-Zucman ABSTRACT: HLA class II polymorphism in Moroccan IDDM patients has not been investigated so far. In this study, HLA-DRBl, -DQAl, and -DQBl allele and haplotype frequencies were analyzed in 125 unrelated Moroccan IDDM patients and 93 unrelated healthy controls, all originating from the Souss region and mostly of Berber origin. Some common features with other Caucasian groups were observed, in particular, a predisposing effect of the DRB1*03-DQA1*0501-DQBl*O201 and DRB 1*04-DQA1*0301-DQB1*0302 alleles or allelic combinations. The Moroccan IDDM group also presented with more specific characteristics. Among DRB 1*04 subtypes, DRB1*0405 was associated with susceptibility to

and DRBl”O406 with protection from the disease. The haplotype and the relative predispositional effect (RPE) analyses indicated that the DRB 1*OS-DQA 1*040 lDQB1*0402

haplotype

was also associated

tibility to IDDM. Interestingly, the DRB1*09DQA1*0301-DQB1*0201 haplotype, completely absent from the control group and very rare in North African populations, was observed in 7.2% of the Moroccan diabetics. Conversely, the DRB1*07-DQA1*0201DQB1*0201 and DRB1*15-DQA1*0102-DQB1*0602 haplotypes were associated with protection from IDDM. Finally, we observed an age-dependent genetic heterogeneity of IDDM, the frequencies of predisposing alleles being higher and those of protective alleles lower in childhood- than in adult-onset diabetics. Our data on Moroccan diabetics, together with data on European and Northern Mediterranean patients, suggest a gradient of various HLA class II predisposing and protective markers that link these populations. Human Immunology 49, 137-143

(1996)

with suscep-

INTRODUCTION Insulin-dependent diabetes mellitus (IDDM), one of the most common chronic metabolic disorders in childhood 111, is an autoimmune disease with multigene dependence 121. Family and population studies have shown that the main susceptibility locus for the disease is located within the major histocompatibility complex (MHC) class II region 12, 31. However, characterization of IDDM-predisposing determinants has been complicated by the strong linkage disequilibrium that may

From INSERM U25 and Service d’lmmunologie Clinique, Hopital Ne&er, Paris, France (H.I., H-J.G., G.B., J-F.B., S.C-Z.); H@ital Hassan II. Agadir, Maroc (M. B.); Laboratoire de Biologic Cellulaire et Mokulaire, Universite’IBNOU-ZOHR, Agadiv, Maroc (H.I., O.A.): and Association Claude Bernard, Paris, France (G.B.). Address reprint requests to: Sophie Caillat-Zucman, MD, INSERM U25 and Service d’lmmunologieClinique, Hopital Neck, 161 rue de &&es. 75015, Paris, France. Received October3 1, 1995: accepted March 8, 1996. Human Immunology 49, 137-143 (1996) 0 American Society for Histocompatibility

occur between the numerous HLA and non-HLA genes mapping within the MHC, making it difficult to distinguish primary from secondary disease associations. The HLA-DR3 and -DR4 specificities have been shown to be strongly associated with the disease in Caucasoids [2}. Studies in other ethnic groups have indicated a role for DR7 in Negroids 141, DR4 and DR9 in Japanese [5, 61 and DR3 and DR9 in Chinese 171. Attention has also been paid to loci in linkage disequilibrium with the DRBl locus, in particular to the DQ loci 17-10). Thus in Caucasoids, the DQA1*0301-DQB1*0302 and DQA1*0501-DQBl”0201 allelic combinations, respectively expressed together with DRB 1*04 and DRB 1*03, confer a strong predisposition to the disease. One valuable approach to overcoming problems resulting from linkage disequilibrium has been to examine the pattern of HLA associations in distinct ethnic groups presenting with different class II allelic and haplotypic 0198~8859/96l$15.00

and Immunogenetics,

1996

PII SO198-8859(96)00104-8

H. Izaabel et al.

frequencies

[IO-12).

mined

similar

by

If IDDM genetic

is assumed

to be deter-

mechanisms

in

all

ethnic

groups, recurrent associations of markers with disease should be seen across the groups, in spite of variations in linkage have

disequilibria.

In this

investigated

the

antigens

in predisposition

113-161.

S o f ar,

IDDM analyzed

few

the

IDDM area,

Africa

individual

tion

to the

deviation RPE.

overall above

vious

round

izing

accordingly

from

the

93 healthy

RPE,

maining

from

The

one

with

that

had

both

was repeated the allele

patients

the This

alleles

no significant

for their

to be the one

process

the

process them

deviation

the

next

in the pre-

and controls

sequential

largest

the greatest

to find

distribution

ef-

contribu-

detected

expected

and removing

overall

ex-

in controls.

predisposing

reviewed

x2 value.

to that

observed

the greatest

were

by excluding

alleles.

associated

with

alleles

was held

The

was compared

of frequencies

the allele

fect,

in

in a group

with

To identify

in patients

basis

we have

are available study,

originating

on the

largest

In the present patients

distribution

of

of HLA alleles

by comparison

II

incidence

the HLA class II gene polymorphism

geographic

studies class

in North

concerning

population.

of 125 Moroccan same

very

of HLA

to IDDM

no data

or the distribution

the Moroccan

regard,

participation

allelic pected

and normalof the

re-

of identifying

was continued

until

was observed.

controls. RESULTS SUBJECTS

AND

METHODS

Szlbjects. One hundred can IDDM

and twenty-five

patients

(Agadir)

were studied.

patients

were

dent

ketosis

on insulin

disease

onset

21.6%

>20 years,

patients

was

were

population

from

the

More

80%

were Berbers.

13.1

and were

recruited

years). their

or familial

the

history

~20

families.

age at

years

and

of the

125

The

healthy

same

area

All

The

distribution

types

among

93 healthy

of DRBl,

control

Berber and

in-

without

125 unrelated

controls

is shown

Distrihtion presented

of DRBl a strong

(68.3%

nomic

for HLA-DRBl,

DNA

was performed

md -DQBl

as previously

ampiified

by the polymerase

basis

of the

locus-specific blotted

flanking

onto nylon

following labeled tection

primers.

membranes

hybridization

Mannheim

< 0.02).

Analysis

, and DQB 1 genes

was

showed

(PCR)

and alleles

using

DNA

was

digoxigenin-1

1-ddUTP-

decreased

IDDM

the presence of copy cording when

or the absence

number.

Odds

to Woolfs necessary.

of each allele,

ratios

formula

(OR}

with

were

Haldane’s

The level of significance

use of chi-squared

(x2) analysis

with

scored

for

estimated

ac-

was assessed

by

2 x 2 contingency

tables. Fisher’s test was used where appropriate. The P value was considered significant if ~0.05, and the correction

of Bonferonni

for multiple

tests

was applied

by

multiplying P by the number of alleles compared. The relative predisposing effect (RPE) of the different alleles was performed Payami

according

to the

et al. (181 and Kockum

method

developed

OR

than

by

et al. 1191. The overall

ho-

patients

(OR

tests

= 15.3,

significance

suscep-

(39.2%

vs 23.7%,

P

subtypes

(Table

2)

of DRB1*0405

= 0.06).

had

= 4.21)

(2%

in diabetics more

vs 5.4%,

no synergistic

absent

in 7.2%

P < O.Ol), but after correction

hetin

OR

effect.

completely

was present

a vs

often

=

Surfrom

of IDDM

the difference did for the number of

performed. were nega-

tively associated to IDDM and conferred protection = 0.28 and 0.27 respectively). The DRBl*l5 which

and

The DRB1*03/04 (21.6%

allele,

in

(61.2%

OR

observed

in controls

population,

frequency. with

frequent

in controls

was

the DRB1*09

control

gene

The DRB 1* 13 and DRB 1* 11 phenotypes

modification

= 4.41).

was more

of DRBl*O406

P < O.OOOS), but

4.84,

not reach

independently

OR

associated

in controls,

combination

patients

DRB1*03

frequency

vs 27.3%

in controls,

erozygous

the were

the disease

of DRB1*03/03

of DRB1*04

frequency

prisingly,

and controls

phenotype

with

of controls,

frequency

strongly

than

an increased

in diabetics 27.3%

were identified

DRBl*O3

in Caucasians,

the

Biochemicals).

Statistical analysis, Patients

to IDDM

{17]. In brief,

sequence-specific oligonucleotides (SSO) and dewith a chemoluminescence substrate (AMPPD,

Boeringher

tibility

diabetics

Amplified

with

of ge-

Moroccan

reaction

1.

association

calculated usually

loci

chain

phenoand the

mozygotes in IDDM patients relative to controls (23.2% vs 5.4%) was in fact not different from that expected on

and -DQBl

described

exon of the DRB 1, DQAl

second

typing. Typing

-DQAl,

DQBl patients

in Table

vs 33.3%

increased

The DRB1*04, -DQAl,

and

IDDM

alleles. The positive

of patients

The apparently

the

of diabetes.

HLA-DRBI,

DQAl,

the

depen-

Mean

Twelve

of 93 unrelated from

region

absolutely

(78.4%

l-31 with

Souss

of diagnosis.

years

range

originating

personal

than

the time

consisted

dividuals

Moroc-

originating prone

from

unrelated

has

Caucasians, in controls

been

shown

to protect

was less frequent (4.8% vs 13.9%,

from

in Moroccan P i 0.03).

the

(OR allele,

disease

diabetics

in than

DQA 1 and DQB 1 allelic distribution. The DQA 1*050 1 and 0301 phenotype frequencies were both increased in IDDM

patients

(OR

= 2.74

and

2.3 3, respectively)

HLA Genetics

TABLE

1

of IDDM

Distribution of DRBl, DQAl, and DQBl alleles among IDDM patients and healthy controls IDDM patients (n = 125) (%)

DRBl” 01

15 16

4.8 68.8 39.2 19.2 14.4 7.2 0.0 5.6 0.0 6.4 0.8 4.8 0.8

DQAl* 0101 0102 0103 0201 0301 040 1 0501 060 1

5.6 7.2 2.4 18.4 44.8 17.6 72.0 0.0

03 04 07 08 09 10 11 12 13 14

DQBl* 0201 0301 0302 0303 040 1 0402 0501 0502 0503 060 1 0602 0603 0604 0605

139

in Moroccans

80.8 10.4 37.6 0.0 0.0 12.8 5.6 2.4 0.8 0.8 2.4 0.8 1.6 0.0

TABLE

Controls (n = 93) (%)

15.1 33.3 23.7 35.5 6.5 0.0 6.5 18.3 2.2 19.4 2.2 13.9 4.3

21.5 28.0 3.2 35.5 25.8 7.5 48.4 1.1

55.9 24.7 17.2 1.1 1.1 5.4 22.6 4.3 3.2 0.0 13.9 2.2 4.3 6.5

OR

0.28 4.41 2.08 0.43 2.44 15.25 0.05 0.27 0.15 0.28 037 0.31 0.18

0.22 0.20 0.74 0.41 2.33 2.62 2.74 0.25

3.50 0.35 2.90 0.25 0.25 2.58 0.20 0.55 0.24 2.25 0.15 0.37 0.55 0.05

2

Distribution of DR4 subtypes among IDDM patients and healthy controls IDDM patients (n = 49) (%)

Controls (n = 22) (%o)

OR

6.1 22.4 6.1 2.0 61.2 2.0 0.0

0.0 13.6 18.2 0.0

3.39 1.83 0.29

P

PC

<0.0001

DRBl* 040 1 0402 0403 0404 0405 0406 Others

27.3 27.3 13.6

1.39 4.21 0.06 0.06

0.01 0.03

<0.05 = 5.3).


Two alleles were significantly less frequent in diabetics than in controls, DQB1*0602 (2.4% vs 14%, OR = 0.15) and DQB1*0501 (5.6% vs 22.60/o, OR = 0.2).


RPE analysir. The above analysis allowed in a first step the identification of alleles most strongly associated to IDDM. However, to detect associations of lesser strength, a RPE analysis was performed 118, 191. At the DRB 1 locus, allelic frequencies that deviated the most in IDDM patients from values expected on the basis of control frequencies were the following: DRB1*03 (n expected = 41.7, overall x2 = 151.9, P < O.OOOl), DRB1*04 (n expected = 23.4, P < O.OOOl), DRB1*08 (n expected = 5.2, P < O.OOOl), and the group containing DRB 1*09 and DRBl * 12 (n expected = 1.5, overall x2 = 84.4, P < 0.001). These two alleles were combined into one group for statistical purpose, because the absence of DRB 1*09 in controls prevented the analysis of this allele separately E181. However, with regard to DRB 1* 12 frequencies in controls and patients, the predisposing effect of the DRB1*09-12 group was likely related to DRB1*09 alone, as already suggested above by the comparison of DRB 1*09 phenotype frequencies. For the DQAl locus, the sequence of allele removal was DQA1*0401, 0301, and 0501, and for the DQBl locus the sequence was DQB1*0302, 0201, and 0402. No DRB 1, DQAl, or DQBl protective allele other than the ones already shown was revealed by RPE analysis.

<0.05 <0.05


co.oo1

whereas the DQAl*OlOl, 0102, and 0201 frequencies were decreased (OR = 0.22, OR = 0.2, and OR = 0.41, respectively). The DQB1*0201 phenotype was the most frequently expressed in the IDDM patient group (80.8% vs 55.9% in controls, OR = 3.5), but the DQBl*O302 frequency was also significantly increased (37.6% in IDDM patients vs 17.2% in controls, OR = 2.9). The DQB 1*020 110302 heterozygous combination conferred a high risk (23.2% in diabetics vs 5.4% in controls, OR

Haplotype analysts. DR-DQ haplotypes were assigned in 12 families of IDDM probands. Most of the observed allelic combinations between DRB 1, DQAl , and DQB 1 loci confirmed the known linkage disequilibria in Caucasians. According to the observed haplotypes within families, DR-DQ combinations were deduced in all IDDM patients.

140

H. Izaabel et al.

HLA

heterogeneity according to the age at disease onset. Since our diabetic population contained both childhoodand adult-onset IDDM patients, the DRBl distribution was analyzed separately in these two groups (Table 4). Interestingly, the frequencies of IDDM-associated alleles decreased as the age at IDDM onset increased, no significant difference persisting in older patients except for DRBl*O3 and to a lesser extent for DRB1*09. Conversely, in the childhood-onset group, the DRBl*l5, DRB 1* 13, and DRB 1*07 alleles conferred a protective effect that disappeared in the adult-onset group. This was particularly clear in the case of the DRB 1* 15 allele, which was never observed among diabetic children while its frequency in adult-onset diabetics was strictly similar to that in controls.

The DRB1*03-DQA1*0501-DQB1*0201 haplotype was associated with IDDM predisposition (43.2% vs 16.7% in controls, OR = 3.8) as was the DRB1*04DQA1*0301-DQB1*0302 haplotype (20.4% vs 9.7%, OR = 2.4) (Table 3). However, the heterozygous haplotypic combination did not particularly increase the IDDM risk (OR = 5.84, P < 0.001). The IDDMpredisposing DRBl*O405 allele was always expressed with the DQB1*0302 allele in patients and controls. The DRB1*0406 protective allele, observed in only one IDDM patient in combination with the DQB1*0401 allele, was associated in controls with DQB1*0301, 0302, or 0401. The DRB1*08-DQA1*0401-DQB1*0402 haplotype frequency was enriched in IDDM patients (6.4% in patients vs 1% in controls, OR = 6.3) as was the DRB1*09-DQA1*0301-DQB1*0201 haplotype frequency (3.6% in IDDM patients and 0% in controls, OR = 14.7). However, for these two haplotypes, the differences did not reach significance when the P value was corrected for the number of comparisons. The DRB1*07-DQA1*0201-DQB1*0201 haplotype frequency was significantly reduced in patients (9.6% vs 19.9% in controls, OR = 0.42, PC < 0.01) as was the DRB1*15-DQA1*0102-DQB1*0602 haplotype frequency (1.2% vs 5.9%, OR = 0.19, P < 0.05). The following haplotypes were also less frequent in IDDM patients relative to controls but the differences were not significant after correction: DRB 1* 1302DQA1*0102-DQB1*0605 (0% in patients vs 3.2% in controls), DRBl*ll-DQA1*0501-DQB1*0301 (2.4% vs 8%). To determine the strongest association for alleles expressed on the same haplotype, we used the two-by-four table analysis described by Svejgaard and Ryder [20]. On the DR3 and DR4 haplotypes, no particular allele merged. On the DR7 haplotype, both DRBl*O7 and DQA1*0201 alleles contributed significantly to protection from the disease (p < 0.01). TABLE

3

Distribution and healthy

of HLA DRBl-DQ controls

haplotypes

IDDM patients (2n = 250)

Controls (2~ = 186)

DISCUSSION Association of predisposition to IDDM with particular HLA alleles has long been known. However, the strength and the nature of associations vary among the populations analyzed. Our analysis of HLA class II distribution in the Moroccan population confirms as a whole results previously observed in other Caucasoid populations, but underlines several characteristics that distinguish Moroccan diabetics from other Caucasian diabetics, including North African IDDM patients. Although the DRB1*03 allele conferred the highest risk, RPE and haplotype analyses provided evidence for the participation of other alleles or haplotypes in susceptibility to IDDM. The DRB1*04 allele was not associated to strong IDDM predisposition unless it was expressed together with the DQB1*0302 allele. Even in this case, the risk was lower than that conferred by the DRB1*03-DQB1*0201 combination. At variance with other Caucasian groups, DRBl*O3-DQB 1*0201DQA1*0501/DRB1*04-DQB1*0302-DQA1*0301 heterozygosity was not dramatically associated with IDDM. Several groups have reported a differential paramong

IDDM

DRBl*

-DQAl*

-DQBl*

(%)

(%)

OR

P

PC

0301 04 09 08

-0501 -0301 -0301 -0401

-0201 -0302 -0201 -0402

43.2 20.4 3.6 6.4

16.7 9.7 0 1.1

3.80 2.40 14.70 6.30

0.0001 0.001 0.01 0.01


07 1501 11 13

-0201 -0102 -0501 -0102

-0201 -0602 -0301 -0605

9.6 1.2 2.4 0

19.9 5.9 8.1 3.2

0.42 0.19 0.28 0.05

0.005 0.05 0.01 0.01


141

HLA Genetics of IDDM in Moroccans

TABLE

4

Distribution

of DRBl

alleles among

childhood

Patients < 15 years

IDDM patients

Patients 15-30 years

Controls (n = 96) (%‘o)

n = 81 (%I

OR

PC

DRl DR3 DR4 DR7 DR8 DR9

15.05 33.31 23.66 35.48 6.45 0.00

4.94 67.90 44.44 16.05 17.28 8.64

0.29 4.23 2.58 0.35 3.03 18.83

a h h

DRlO DRll DR12 DR13 DR14 DR15 DRl6

6.45 18.28 2.15 19.35 2.15 13.98 4.10

0.00 4.94 0.00 2.47 1.23 0.00 1.23

0.08 0.23 0.22 0.11 0.57 0.04 0.28

DRB 1

and adult-onset

r

n = 37 (%)

OR

2.70 70.27 32.43 24.32 10.81 2.70

0.16 4.73 1.55 0.58 1.76 7.68

0.00 5.41 0.00 10.81 0.00 13.51 0.00

0.16 0.26 0.49 0.51 0.49 0.96 0.27

Patients >30 years

PC

(

n=? (%)

OR

PC

14.29 57.14 28.57 28.57 0.00 14.29

0.94 2.67 1.29 0.73 0.90 43.15

h

0.00 14.29 0.00 14.29 0.00 14.29 0.00

0.90 0.75 2.44 0.69 2.44 1.03 1.33

@PC < 0.0001. b PC <

0.05.

cPC i 0.001.

ticipation of the DQB 1*0302-associated DR4 subtypes in IDDM predisposition. In North Europeans, DRB1*0401 is associated with an increased risk {2]. A predisposing effect of DRB1*0402 or 0405 and a protective effect of DRB1*0403, 0404, or 0408 have been described in French and American Caucasians, Sardinians, and Mexican Americans 117, 2 l-261, whereas DRB l*O4O6 was associated with protection in Orientals [27]. By contrast, no DR4 subtype association has been found in Spaniards who represent a mixture of northern Europeans and paleo-North Africans {28-30). All these controversial results are likely due to differences in the DR4-subtype distributions between control populations. In Algerians, another Maghrebian population, the DRB 1*0405 frequency was significantly increased in IDDM patients [14]. In our Moroccan patients, the same increase of the DRB1*0405 frequency was indeed observed, together with a specific decrease of DRB l*O406. These alleles differ only by three aminoacid substitutions at positions 57, 74, and 86 of the DRB chain. Recent functional and crystallographic studies indicate that polymorphism at these positions is likely to influence the immune response by modifying the peptide binding affinity or disrupting a critical contact with the helper T cell receptor 131). Interestingly, DRB1*09 was present in 7.2% of Moroccan diabetics but completely absent from controls. This allele is apparently very rare in North Africans. It is absent from the Algerian and Egyptian groups and from the Moroccan Jewish community living in Israel [ 14, 16, 32-341. DRB 1*09 is usually neutral in Caucasian populations but is strongly associated to IDDM in Chinese,

Japanese, and Blacks l-4, 7, 10, 111. In our population, DRBl*O9 was expressed with the DQAl*O301 and DQB1*0201 alleles, as shown by family analysis. The same haplotypic combination is observed in Blacks (47, suggesting a genetic relationship between these two populations. It is possible that the predisposing effect of the DR9 haplotype is due to the DQa/B heterodimer encoded by the DQA*O301 and DQB1*0201 alleles in the cis-position. In the case of DR3/DR4 Caucasian IDDM patients, this molecule is encoded in the transposition. Several protective alleles or allelic combinations were evidenced in our patient group. The major participation of the DQ locus in protection from IDDM was confirmed by the effect of DQA1*0102 and DQB1*0602 alleles, in linkage disequilibrium with the neutral DRBl*l5 allele. The DRB1*07-DQA1*0201-DQB1*0201 haplotype was also strongly protective, as already described in Spaniards [29]. DRB1*07 is usually neutral in Caucasians [2}. It is positively associated to IDDM in Blacks [lo] but in the DRB1*07-DQA1*0301-DQB1*0201 haplotypic combination. Although a protective role of the DQAl*O201 allele remains possible, DRB1*07 could be directly involved in protection from IDDM, as suggested by the stratification analysis. Since DQB1*0201 is also present on the IDDM-associated DR3 haplotype where it confers the most significant risk (PC < O.OOOl), its participation in protection from the disease seems to be excluded. However, the recent discrimination of two DQB1*02 alleles (DQB1*0201 and DQB 1*0202) on the basis of a mutation within the third exon [35] brings new insights into this discussion. The

142

H. Izaabel et al.

DQB 1*0202 allele, present on the DRB 1*07 haplotype, might be associated with protection from IDDM, whereas the DQB1*0201 allele, expressed on the DRB1*03 haplotype, would confer susceptibility to the disease. Our typing methods did not allow differentiation between these two alleles, but experiments in progress in our laboratory should help to confirm this hypothesis. Finally, when discriminating IDDM patients according to the age at disease onset, the frequencies of IDDMpredisposing alleles were higher, whereas those of protective alleles were lower in children than in adult diabetics. These observations are in the same line as previous reports suggesting an age-dependent genetic heterogeneity of IDDM 117, 36, 371. Childhood- and adult-onset IDDM could be distinctly influenced by genetic factors. The presence of strong predisposing alleles such as DR3 and/or DR4 would trigger the disease early in life, while in their absence, other genetic or acquired factors would contribute to a delayed onset of the disease. In summary, while our results share some features with IDDM association studies in Caucasians, they reveal other peculiarities described only in other races, such as in Black and Asian populations. Our assumption is that our Moroccan patients, mostly of Berber origin, present with a mixture of HLA class II IDDM-predisposing factors.

ACKNOWLEDGMENTS

We thank Dr. M. Ouchrif and Dr. A. N’Bou for their help in obtaining donor blood samples and P. Przednowed, E. Audran, and I. Texier for their technical assistance. This work was supported in part by “Action integree inter-universitaire France-Marocaine” (No. 95/926).

REFERENCES Rossini AA, Mordes JP, Like AA: Immunology of insulindependent diabetes mellitus. Annu Rev Immunol 3:289, 1985. Thomson G, Robinson WP, Kuhner MK, Joe S, MacDonald MJ, Gottschall JL, Barbosa J, Rich S, Bertrams J, Baur MP, Partanen J, Tait BD, Schber E, Mayer R, Ludvigson J, Lindblom B, Farid NR, Thopson C, Deschamps I: Genetic heterogeneity, modes of inheritance, and risk estimates for a joint study of Caucasians with insulindependent diabetes mellitus. Am J Hum Genet 43:799, 1988. Owerbach D, Lernmark A, Platz PP, Ryder LP, Peterson PA: HLA-D region beta-chain DNA endonuclease fragments differ between HLA-DR identical healthy and insulin-dependent diabetic individuals. Nature 303:815, 1983. Mijovic CH, Jenkins

D, Jacobs KH, Penny MA, Fletcher

JA, Barnett AH: HLA-DQAl and -DQBl alleles associated with genetic susceptibility to IDDM in black population. Diabetes 40:748, 1991. 5

Aparicio JMR, Wakisaka A, Takada A, Matsuura N, Aizawa M: HLA-DQ system and insulin-dependent diabetes mellitus in Japanese: Does it contribute to the development of IDDM as it does in Caucasians? Immunogenetics 28:240, 1988.

6

Kida K, Mimura G, Kobayashi T, Nakamura K, Sonoda S, Inouye H, Tsuji K: Immunogenetic heterogeneity in type 1 (insulin-dependent) diabetes among Japanese-HLA antigens and organ specific autoantibodies. Diabetologia 32: 34, 1989.

R, 7. Ju L-Y, Sun YP, Semana G, Gu XF, Krishnamoorty Fauchet R, Charron D: Aspartic acid at position 57 of the HLA-DQB chain in insulin-dependent diabetes mellitus: an association with one DRw9-DQw9 subtype in Chinese population. Tissue Antigens 37:218, 1991. 8. Baisch JM, Weeks T, Giles R, Hoover M, Stastny P, Capra DJ: Analysis of HLA-DQ genotypes and susceptibility in insulin-dependent diabetes mellitus. N Engl J Med 322:1836, 1990. 9. Todd JA, Bell JI, McDevitt HO: HLA-DQB gene contributes to susceptibility and resistance to insulindependent diabetes mellitus. Nature 329:599, 1987. 10. Fletcher JA, Mijovic C, Jenkins D, AH: Trans-racial studies implicate nent of genetic susceptibility dependent) diabetes. Diabetologia

Bradwell AR, Barnett HLA-DQ as a compoto type I (insulin31:8764, 1988.

11. Todd JA, Mijovic C, Fletcher J, Jenkins D, Bradwell AR, Barnett AH: Identification of susceptibility loci for insulin-dependent diabetes mellitus by trans-racial gene mapping. Nature 338:587, 1989. 12. Jenkins D, Mijovic C, Fletcher J, Jacobs KH, Bradwell AR, Barnett AH: Identification of susceptibility loci for type 1 (insulin-dependent) diabetes by trans-racial gene mapping. Diabetologia 33:387, 1990. 13. Mercier P, Vallo JJ, Vialettes B, Vague PH: HLA-A, DR antigens and insulin-dependent diabetes mellitus Algerians. Tissue Antigens 25:20, 1985. S, 14. Djoulah S, Khalil I, Beressi JP, Benhamamouch saoud K, Deschamps I, Degos L, Hors J: The DRBl*0405 haplotype is most strongly associated IDDM in Algerians. Em J Immunogenet 19:381,

B, in

BensHLAwith 1992.

S, 15. Jenhani F, Badri R, Gorgi Y, Ayed K, Chammaki Boukhris R: Polymorphisme HLA, A, B, C, DR, C4, Bf chez les diabetiques insulinodependants dans la population Tunisienne. Ann Biol Clin 47:23, 1989. 16. Gaber SA, Mazola G, Berrino M, Canale L, Cornaglia M, Ghali I, Curtoni ES, Amoroso A: Human leukocyte antigen class II polymorphisms and genetic susceptibility of IDDM in Egyptian children. Diabetes Care 17:1341, 1994. S, Garchon H-J, Timsit J, Assan R, 17. Caillat-Zucman Boitard C, Djilali-Saiah I, Bougneres P, Bach J-F: Age-

HLA Genetics

dependent dependent

of IDDM

143

in Moroccans

HLA genetic heterogeneity of type 1 insulindiabetes mellitus. J Clin Invest 90:2242, 1992.

18. Payami H, Joe S, Farid NR, Stenszky V, Chan SH, Yeo PPB, Cheah JS, Thomson G: Relative predispositional effects (RPEs) of marker alleles with disease: HLA-DR alleles and Graves disease. Am J Hum Genet 45:541, 1989. R, Holmberg E, Michelsen B, 19. Kockum I, Wassmuth Lernmark A: HLA-DQ primarily confers protection and HLA-DR susceptibility in type I (insulin-dependent) diabetes studied in population-based affected families and controls. Am J Hum Genet 53:150, 1993. 20. Svejgaard A, Ryder LP: HLA and disease associations: detecting the strongest association. Tissue Antigens 43: 18, 1994. 21. Sheehy MJ, Scharf SJ, Rowe JR, Neme De Gimenez MH, Meske LM, Erlich HA, Nepom BS: Diabetessusceptibility haplotype is best defined by a combination of HLA-DR and -DQ alleles. J Clin Invest 83:830, 1989. 22. Erlich HA, Zeidler A, Chang J, Shaw S, Raffel LJ, Klicz W, Beshkov Y, Costin R, Pressman S, Bugawan TL, Rotter Jr: HLA class II alleles and susceptibility and resistance to insulin dependent diabetes mellitus in MexicanAmerican families. Nature Genet 3:358, 1993.

23. Cucca F, Lampis R, Frau F, Macis D, Angius E, Masile P, Chessa M, Frongia P, Silvetti M, Cao A, De Virgiliis S, Congia M: The distribution of DR4 haplotypes in Sardinia suggests a primary association of type I diabetes with DRBl and DQBl loci. Hum Immunol 43:301, 1995. 24. Erlich HA, Bugawan TL, Scharf S, Nepom GT, Tait BD, Griffith RL: HLA-DQ beta sequence polymorphism and genetic susceptibility to IDDM. Diabetes 39:96, 1990. 25. Rowe JR, Mickelson EM, Hansen JA, MacDonald MJ, Allen CI, Gabbay KH, Yunis EJ, Sheehy MJ: T-cell defined DR4 subtypes as markers for type 1 diabetes. Hum Immunol 22:51, 1988. 26. Bach FH, Rich SS, Barbosa J, Segall M: Insulin-dependent diabetes associated HLA-DR region encoded determinants. Hum Immunol 12:59, 1985. 27. Awata T, Kuzuya T, Matsuda A, Iwamoto Y, Kanazawa Y: Genetic analysis of HLA class II alleles and susceptibility to type 1 (insulin-dependent) diabetes mellitus in Japanese subjects. Diabetologia 35:419, 1992. 28. Morales

P, Martinez-Las0

J, Martin-Villa

JM, Core11 A,

Vicario JL, Varela P, Perez-Aciego P, Arnaiz-Villena A: High frequency of the HLA-DRB 1*0405-(Dwl5)-DQw8 haplotype in Spaniards and its relationship to diabetes susceptibility. Hum Immunol 35:170, 1991. JM, 29. Vicario JL, Martinez-Las0 J, Core11 A, Martin-Villa Morales P, Lledo G, Segudaro OG, de Juan D, ArnaizVillena A: Comparison between HLA-DRB and DQ DNA sequences and classic serological markers as type 1 (insulin-dependent) diabetes mellitus predictive risk markers in the Spanish population. Diabetologia 35:475, 1992. N, Gomez30. Martinez-Las0 J, de Juan D, Martinez-Quikes Casado E, Cuadrado E, Arnaiz-Villena A: The contribution of the HLA-A, -B, -C and -DR, -DQ DNA typing to the study of Spaniards and Basque origins. Tissue Antigens 45:237, 1995. 31. Hammer J, Gallazzi F, Bono E, Karr RW, Guenot J, Valsasnini P, Nagy A, Sinigaglia F: Peptide binding specificity of HLA-DR4 molecules: correlation with rheumatoid arthritis association. J Exp Med 181:1847, 1995. S, 32. Djoulah S, Sanchez-Mazas A, Khalil I, Benhamamouch Degos L, Deschamps I, Hors J: HLA-DRBl, DQAl and DQBl DNA polymorphisms in healthy Algerian and genetic relationships with other populations. Tissue Antigens 43:102, 1994. A, Benmamer D, Alvarez M, Diaz-Campos 33. Arnaiz-Villena N, Varela P, Gomez-Casado E, Martinez-Las0 J: HLA allele and haplotype frequencies in Algerians: relatedness to Spaniards and Basques. Hum Immunol 43:259, 1995. A, Freidmann A, Witt CS, Eisenberg S, 34. Roitberg-Tambur Soskolne A, Shapira L, Sela MN, Battat S, Safirman C, Sherman L, Cohen I, Braubar C: HLA polymorphism in Moroccan Jewry. Hum Immunol 40:6I, 1994. 35. Hall MA, Lanchbury JS, Lee JS, Welsh KI, Ciclitira PJ. HLA-DQ2 second-domain polymorphisms may explain increased trans-associated risk in Celiac disease and dermatitis herpetiformis. Hum Immunol 38:284, 1993. 36. Karjalainen J, Salmela P, Ilonen J, Surcel H-M, Knip M: A comparison of childhood and adult type I diabetes mellitus. N Engl J Med 320:881, 1989. 37. Kumar D, Gemayel NS, Deapen D, Kapadia D, Yamashit PH, Lee M, Dwyer JH, Roy-Burman P, Bray GA, Mack TM: North-American twins with IDDM: genetic, etiological and clinical significance of disease concordance according to age, zygosity, and the interval after diagnosis in first twin. Diabetes 42:1351, 1993.