| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Studies |
Department of Immunology, King’s College School of Medicine and Dentistry (S.A., M.P.); Liver Unit, King’s College Hospital (J.A.U., P.D.), and the Division of Endocrinology, Middlesex Hospital (G.S.C.), London, United Kingdom SE5 9PJ
Address all correspondence and requests for reprints to: Dr. Mark Peakman, Department of Immunology, King’s College School of Medicine and Dentistry, Bessemer Road, London, United Kingdom SE5 9PJ, UK. E-mail: mark.peakman{at}kcl.ac.uk
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
One of the major autoantibody specificities described in POF is steroid cell autoantibody, which reacts with the adrenal cortex, ovary, and testis on immunofluorescence (5). We recently identified a major target of steroid cell autoantibody as the steroid cell enzyme 3ß-hydroxy-steroid dehydrogenase (3ßHSD); autoantibodies against this were found in 21% of patients with idiopathic POF (6). We proposed that anti-3ßHSD autoantibodies could be a marker of true autoimmune POF and thus useful in targeting immune therapies to those patients most likely to respond. In the present study, we seek further evidence that 3ßHSD autoantibodies define a subgroup of POF with autoimmune pathogenesis by examining whether there is an association between autoantibody-positive POF patients and human leukocyte antigen (HLA) genotypes. Many well characterized, organ-specific, autoimmune diseases are associated with HLA class II genes (7). The gene products of the HLA class II region on chromosome 6p21.3 include those encoding HLA-DR, -DQ, and -DP molecules (8). These cell surface glycoproteins present peptide antigens to CD4+ helper T lymphocytes and thus have a key role in immune responses (9). The HLA-DR, -DQ, and -DP genes exhibit a high degree of polymorphism, with many different allelic forms giving rise to HLA molecules with distinctive peptide-binding properties (10). It has been proposed that HLA molecules can thus determine whether disease-inducing peptide antigens are presented, providing a rational basis for their involvement in autoimmune disorders (11). Against this background, we have examined HLA-DRB1 and DQB1 genotypes in a cohort of patients with idiopathic POF, including a subgroup of patients positive for 3ßHSD autoantibodies, as markers of autoimmunity.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
One hundred and eighteen North European caucasoid patients presenting to the Reproductive Endocrine Clinic at Middlesex Hospital (London, UK) with idiopathic POF (median age of onset, 26 yr; range, 11–39 yr) were studied. POF was defined as hypergonadotropic amenorrhea, with serum LH and FSH levels greater than 10 IU/L on two occasions and no menstruation for at least 6 months. Patients with POF secondary to Turner’s syndrome, chemotherapy, pelvic surgery, pelvic irradiation, galactosemia, and 46,XY gonadal dysgenesis were excluded, as previously described (2).
Seventeen patients presented with primary amenorrhea, and 101 presented with secondary amenorrhea. Evidence of Addison’ disease was sought on clinical examination and measurement of serum electrolytes, and formal Synacthen tests were performed when clinically indicated, but no new diagnosis of Addison’s disease was made, and none of the patients had the disease at entry into the study. Four patients had autoimmune thyroid disease, of whom 3 had hypothyroidism treated with T4, and 1 had Graves’ disease.
Autoantibody screening
Autoantibodies to 3ßHSD were measured using a Western blot assay as described previously (6). Steroid cell autoantibodies were determined using a conventional indirect immunofluorescence technique performed on sections of monkey adrenal gland, testis, and ovary as previously described (6). Antithyroglobulin and antithyroid microsomal autoantibodies were measured by gel agglutination as previously described (12).
HLA typing
A total of 32 HLA class II DRB1 and DQB1 alleles or groups of alleles were determined by PCR amplification and PCR-SSOP (sequence-specific oligonucleotide probing), including 18 DRB1 and 14 DQB1 as previously described (13). Two locus haplotypes, based on patterns of linkage previously described in caucasoid individuals, were constructed for all of the subjects (14, 15, 16). The control group comprised 134 Northern European Caucasian subjects recruited from the same geographical area as the patients with POF. In the absence of family data it was not possible to determine true haplotypes; thus, in this analysis only common preselected haplotypes were determined. These represented known HLA associations with the more common HLA alleles and the best known linkage patterns.
Statistical analysis
The distribution of all of the DRB1 and
DQB1 alleles in patients and controls were compared using
2 analysis. P values were corrected
(Pc) for multiple testing following the recommendations of
Svejgaard and Ryder (17), applying a correction factor of 32
(i.e. the total number of different alleles compared).
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Of the 118 patients with POF recruited into the study, DNA samples were available from all, and serum was available from 97 patients.
Initially, we compared the frequencies of HLA-DR and -DQ genotypes in
the POF patients and control subjects. The only difference of note
among the HLA-DR genotypes was an increase in the
DRB1*1101/3 genotype in POF patients (2 =
8.0; P < 0.0047). Increased frequencies of the
DQB1*0603 and -0301 genotypes were also seen
(2 = 4.2 and 5.03, respectively; P =
0.041 and P = 0.025, respectively), and a reduced
frequency of the DQB1*0302 genotype was found
(2 = 4.4; P = 0.036). None of these
trends was significant when corrected for multiple testing
[Pc = P x the number of alleles
(n = 32) tested], consistent with previous reports (18, 19).
However, when patients were divided according to the presence or
absence of 3ßHSD autoantibody (Tables 1
and 2), it became clear that these trends
were predominantly related to the presence of the autoantibody.
Twenty-one of 97 patients were positive for autoantibodies to 3ßHSD
(21.6%), a prevalence identical to that previously observed (6). POF
patients with 3ßHSD autoantibodies had increased frequencies of the
DQB1*0603 (25%) and 0301 (42.9%) genotypes
compared with control subjects (6.0% and 33.6%, respectively;
2 = 7.5 and 4.3; P = 0.006 and
P = 0.037, respectively). In our study population of
North European caucasoids, the DQB1*0301 gene is most
commonly found on an extended haplotype, which includes
DRB1*04 alleles. The increase in the DQB1*0301
genotype was thus associated with a significantly higher frequency of
the DRB1*04/DQB1*0301 haplotype, which was present in 9
patients (21.4% of possible haplotypes) with 3ßHSD autoantibodies
but in only 20 control subjects (7.5%; 2 = 8.35;
P = 0.0039). None of the probability values described
above was significant after correction for multiple testing.
|
|
Despite this, the increase in DQB1*0603 and 0301 genotypes in 3ßHSD autoantibody-positive patients, arising in tandem with the nonsignificant trend for a reduced frequency of the DQB1*0302 genotype, is of interest when the amino acid sequences of the respective DQß chains are considered (20).
Within the regions of polymorphism in the DQB1 locus,
DQB1*0603 and -0301 encode identical residues at
positions 14, 38, 57, 71, 74–75, 77, and 116. The only site among
these at which they differ from DQB1*0302 is position 57.
The difference in the proteins encoded is that position 57 on the
ß-chain of the DQ molecule is occupied by an aspartate (Asp) in the
0603 and 0301 alleles, whereas the same position
has an alanine in the 0302 allele. For this reason, we then
examined the frequency of DQß-Asp57-encoding genes,
namely DQB1*0301, -0303, -0402,
-0503, and -0601–3 (DQB1*0203,
-0401, -0306, -0607, 0611,
and -0614 also encode DQß-Asp57, but are rare
and were not represented in our populations; Table 3). The frequency of
DQß-Asp57-encoding genes in 3ßHSD autoantibody-positive
patients was 27 of 42 haplotypes compared with 109 of 268 haplotypes in
control subjects (2 = 8.2; P = 0.004),
and 9 of 21 (42.8%) patients were homozygous compared with 17 of 134
(12.7%) controls (2 = 11.8; P =
0.0006). None of these probability values was significant after
correction for multiple analyses.
|
2 = 9.5; P = 0.002; Pc =
0.067]. Thyroid and steroid cell autoantibodies
The frequency of autoantibodies to thyroglobulin and thyroid
microsomes was similar to that described previously (6). Steroid cell
antibodies were detected by immunofluorescence in four patients. All
had the tissue distribution typical of 3ßHSD autoantibodies (staining
of zonae fasciculata and glomerulosa only), but not of
21
-hydroxylase autoantibodies (staining of all three zonae) (6).
Antiovarian autoantibodies were present in one patient. None of the
control subjects was positive for these autoantibodies.
The inclusion of these as markers as markers of autoimmunity in POF patients did not enhance the significance of any of the HLA associations described above.
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The HLA class II molecules, DR and DQ, are heterodimers, each composed
of - and ß-chains of approximately 29–34 kDa (21). The
heterodimers form complexes with short peptides (13–24 amino acids)
that are derived from the enzymatic processing of exogenous antigens or
autoantigens (22). Peptides bind within a specialized binding groove
formed by two -helixes and a ß-pleated sheet (23). The HLA
molecule/peptide complex is then expressed on the surface of
antigen-presenting cells for presentation to antigen-specific receptors
on CD4+ T helper cells, one of the pivotal cells in the
induction and maintenance of immune and autoimmune responses (24).
Between individuals, the HLA class II genes display considerable
genetic polymorphism, which results in variability at specific amino
acid residues within the antigen binding groove (21). This has a
profound effect on peptide binding, which, in turn, influences thymic
selection of the T cell repertoire and peripheral T cell activation
(25), although it is not known precisely how these events might lead to
autoimmunity. Our preliminary identification of an HLA association with
a subgroup of POF patients is thus an important step in identifying the
pathogenic mechanisms that lead to ovarian damage in these cases. The
finding requires confirmation in larger series of patients from our own
and other centers before it can lead to future work focused on the
identification of peptides from 3ßHSD that are processed and
presented by antigen-presenting cells bearing HLA-DQ molecules with
Asp57 ß-chains.
Many well characterized immune-mediated diseases have strong HLA
disease associations, particularly with HLA-DRB1 and
HLA-DQB1 genes. These include insulin-dependent diabetes
mellitus (DQB1*0302) (26), coeliac disease
(DQB1*0201) (27), bullous pemphigoid (DQB1*0301)
(28), and autoimmune hepatitis (DRB1 genes encoding lysine
at position 71 of the DR ß-chain) (29). Although the crystallographic
structure of HLA-DQ has yet to be solved, comparisons with HLA-DR
suggest that position 57 occupies a key site at the boundary of the
groove (23). Aspartate at position 57 allows the formation of a salt
bridge with a conserved arginine at position 76 on the -chain,
whereas non-Asp residues are unable to generate this interaction. This
single amino acid difference has been shown to have profound effects on
peptide binding (30). Thus, it is clear that minimal differences in
residues within these key peptide-binding regions can have profound
effects on the immune response.
Previous studies have also attempted to define HLA associations in POF. One study on a small number of patients (n = 19) reported an association with HLA-DR3 (31), found in 58% of patients (vs. 23% of control subjects), whereas another identified a weak association with HLA-DR4 (18), and another showed no association (19). Each study was based on serological determination of HLA-DR, and none included subgroup analysis according to the presence of autoantibody status. Serology is less accurate than the DNA-based technique used in the present study and does not discriminate between most HLA-DQB1 alleles.
It is of interest that the high frequency of HLA-DRB1*1101/3 genes that we identified in our patients was associated with those POF patients who did not have evidence of 3ßHSD autoantibodies, particularly in light of the fact that DQB1*0301 is almost always found on the same haplotype. This finding will need to be confirmed in other series, but it is tempting to speculate that patients with the DRB1*11 genotype may constitute another subgroup of patients with autoimmune POF in whom the target autoantigen is distinct from 3ßHSD.
Received September 30, 1998.
Revised December 14, 1998.
Accepted December 16, 1998.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
ß heterodimers in IDDM
susceptibility. Diabetes. 41:378–384.[Abstract]
This article has been cited by other articles:
 |
|
 |
|
  D. Goswami and G. S. Conway Premature ovarian failure Hum. Reprod. Update, July 1, 2005; 11(4): 391 - 410. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Forges, P. Monnier-Barbarino, G.C. Faure, and M.C. Bene Autoimmunity and antigenic targets in ovarian pathology Hum. Reprod. Update, March 1, 2004; 10(2): 163 - 175. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Arif, R. Varela-Calvino, G. S. Conway, and M. Peakman 3{beta} Hydroxysteroid Dehydrogenase Autoantibodies in Patients with Idiopathic Premature Ovarian Failure Target N- and C-Terminal Epitopes J. Clin. Endocrinol. Metab., December 1, 2001; 86(12): 5892 - 5897. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Reimand, P. Peterson, H. Hyöty, R. Uibo, I. Cooke, A. P. Weetman, and K. J. E. Krohn 3{beta}-Hydroxysteroid Dehydrogenase Autoantibodies Are Rare in Premature Ovarian Failure J. Clin. Endocrinol. Metab., June 1, 2000; 85(6): 2324 - 2326. [Abstract] [Full Text]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
