| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Other Original Articles |
Department of Immunology (S.A., R.V.C., M.P.), Guy’s, King’s and St. Thomas’ School of Medicine, London, United Kingdom SE5 9NU; and the Division of Endocrinology (G.S.C.), Middlesex Hospital, London, United Kingdom W1N 8AA
Address all correspondence and requests for reprints to: Dr. Mark Peakman, Department Immunology, Guy’s, King’s & St. Thomas’ School of Medicine, Rayne Institute, 123 Coldharbour Lane, London, United Kingdom SE5 9NU. E-mail: mark.peakman{at}kcl.ac.uk
Abstract
The steroid cell enzyme 3ß hydroxysteroid dehydrogenase (3ß HSD) has been identified as a target of steroid cell autoantibodies, and autoantibodies to this enzyme are present in patients with premature ovarian failure and patients with autoimmune polyendocrine syndrome 1.
The aim of the present study was to develop a radioligand binding assay for 3ß HSD autoantibodies and to exploit this to examine regions of the molecule targeted by autoantibodies. We generated a construct of 3ß HSD coupled to a luciferase fusion partner in order to maximize the yield of 35S-radiolabeled protein. Labeled 3ß HSD was then immunoprecipitated and the autoantibodies quantified by phosphoimaging. Autoantibodies to 3ß HSD were detected in 12 of 100 (12%) idiopathic premature ovarian failure patients and 0 of 103 (0%) healthy age-matched controls (P < 0.0001). Three overlapping fragments of 3ß HSD cDNA were cloned downstream of luciferase to examine autoantibody binding sites. Two of nine sera with 3ß HSD autoantibodies (22%) displayed reactivity to the N terminus of 3ß HSD, and seven (77%) showed reactivity to the C terminal; no sera reacted with the middle region. Our study demonstrates a markedly enhanced disease specificity of autoantibodies to 3ß HSD detected using this novel assay and shows that distinct regions of the molecule are targeted.
PREMATURE OVARIAN FAILURE (POF) is a syndrome that is clinically defined as the loss of ovarian function before the age of 40 yr (1). POF is a heterogeneous disorder with a multifactorial pathogenesis; chromosomal abnormalities (2), genetic disorders of metabolism (3), iatrogenic effects (4), and infection (5) may all contribute to its etiology. Despite this, in many women, ovarian failure is not readily explained, and an autoimmune basis has been proposed to account for these idiopathic cases (6). The evidence supporting the autoimmune nature of idiopathic POF is based on histopathological studies that have demonstrated lymphocytic infiltration of the ovaries (6), the presence of autoantibodies to target autoantigens in the ovary (7), and an association between the development of POF and other autoimmune diseases (8). Steroid cell autoantibodies (SCA) directed against targets in steroid cells of the adrenal cortex, the theca interna of the ovary, the Leydig cells of the testis, and placental trophoblasts have been described in association with POF, with reported prevalences ranging from 1% (9) to 52% (10).
We have previously identified the steroid cell enzyme 3ß hydroxysteroid dehydrogenase (3ß HSD) as an autoantigen in idiopathic POF (11). Furthermore, patients with POF and autoantibodies to 3ß HSD appear to have a preponderance of human leukocyte antigen (HLA) class II genotypes encoding an aspartate residue at position 57 of the DQß chain (HLA-DQß Asp57) (12). This demonstration of an HLA class II association with 3ß HSD autoantibody positivity reinforces the view that some cases of idiopathic POF are autoimmune in nature, because such associations are typical of other autoimmune diseases (13, 14, 15).
In our original study, 3ß HSD autoantibodies were measured using an immunoblot assay. This assay format has several limitations. First, the antigen is denatured, so that only linear epitopes are available for autoantibody binding. Analytical sensitivity and specificity may be reduced when conformational epitopes are lost. Second, the assay is qualitative rather than quantitative. The arbitrary nature of determining positivity in immunoblots could account for the high prevalence of 3ß HSD autoantibodies (5%) that we observed in our control population (11).
The purpose of the present study, therefore, was to measure autoantibodies by immunoprecipitation of radiolabeled native recombinant 3ß HSD generated in an in vitro transcription/translation system to allow detection of conformation-dependent autoantibodies. To increase the yield of labeled recombinant protein, the enzyme was coupled to luciferase as a fusion partner. In addition, we took the opportunity to adapt this assay format to include an analysis of the regions of 3ß HSD that constitute the main autoantibody binding sites.
Subjects and Methods
Subjects
One hundred consecutive patients presenting to the Reproductive Endocrine Clinic at the Middlesex Hospital (London, UK) with idiopathic POF (median age of onset, 26 yr; range, 11–39 yr), on whom blood samples were available, were studied. Of these, 78 were of Caucasian origin, 14 Asian, 5 Black, 2 Chinese, and 1 Middle Eastern. POF was defined as hypergonadotrophic amenorrhea, with serum LH and FSH levels greater than 20 IU/liter on two occasions and no menstruation for at least 6 months. In our laboratory, reference LH/FSH levels for women aged 11–29 yr are 10 IU/liter (the upper limit of the normal range) and for women older than 30 yr, 15 IU/liter. Polycystic ovary syndrome was excluded by clinical and ultrasound examination in all cases. Patients with POF secondary to Turner’s syndrome, chemotherapy, pelvic surgery, pelvic irradiation, galactosemia, and 46,XY gonadal dysgenesis were excluded as described (12).
Ten patients presented with primary amenorrhea, and 90 presented with secondary amenorrhea. Evidence of Addison’s disease was sought on clinical examination and measurement of serum electrolytes. Formal Synacthen tests were performed when clinically indicated, but no new diagnosis of Addison’s disease was made, and none of the patients had Addison’s disease at entry into the study. Four patients had autoimmune thyroid disease; of these, three had hypothyroidism treated with thyroxine, and one had Graves’ disease.
Sera were also obtained from 103 healthy female blood donors as controls (median age, 27 yr; age range, 18–43 yr).
Cloning of 3ß HSD constructs
The vectors, restriction enzymes, and DNA polymerases used in this study were from Promega Corp. (Southampton, UK). The cDNA for human 3ß HSD [obtained from Dr. V. Luu-The, Department of Molecular Endocrinology, Laval University Research Centre (CHUL), Sainte Foy, Canada] was amplified using Pfu polymerase by the PCR with the following upper and lower strand primers: 5' GCTCTAGAGCCG CCACCATGGGCTGGAGCTGCCTTGTG 3', and 5' GCTCTAGAGCTCACTGA GTCTTGGACTTCAGG 3', respectively. These primers provide XbaI restriction sites and a KOZAK consensus sequence (CCACCATG) to enable more efficient translation (16). The amplified product was subsequently digested with XbaI and cloned into the pSP64 Poly A vector. Automated sequencing with appropriate primers was used to verify the sequence of these constructs.
Constructing luciferase fusion proteins
To introduce relevant restriction enzyme cleavage sites, human 3ß HSD cDNA was amplified using Pfu polymerase by PCR with the following upper and lower strand primers: 5' GATCCTCATAAAGGCCATGGGCT GGAGCTGCCTTGTT 3' and 5' CCGCTCGAGTTAACCCACATGCACATCTC TGTC 3', respectively. Then, the cDNA was cloned into the TOPO-TA cloning vector (Invitrogen, Groningen, The Netherlands) and digested with EcoN1 and XhoI. The 3ß HSD insert was cloned into the pGEM-luc vector that had been digested with EcoN1 and XhoI. This introduces 3ß HSD in-frame into the luciferase gene at codon 541, resulting in the loss of the terminal 9 codons of luciferase. The construct was designated pGEM-luc-3ß HSD, and automated DNA sequencing was used to confirm authenticity.
Epitope mapping studies
In addition to inserting the full-length 3ß HSD into pGEM-luc, three fragments overlapping by 50 bp were generated and cloned into the same site and designated pGEM-luc-3ß HSD1–450 (numbers refer to base pairs and represent amino acids 1–150 of 3ß HSD), pGEM-luc-3ß HSD399–849 (representing amino acids 133–283 of 3ß HSD), and pGEM-luc-3ß HSD798–1116 (representing amino acids 266–372). The fragments were generated by PCR amplification using Pfu polymerase of the original 3ß HSD cDNA using the following primer pairs: 5' GATCCTCATAAAGGCCATGGGCTGGAGCTGCCTTGT 3' and 5' CC GCTCGAGTTAGTTTTCCAGAGGCTCTTCTTC 3'; 5' GATCCTCATAAA GGCCTCCTACAAGGAAATCATCCA 3' and 5' CCGCTCGAGTTAGGAATC AAGGCGGAGGCCG 3';
5' GATCCTCATAAAGGCCAACCTTAATTACACCC TGAGC 3' and 5' CCGCTCGAGTTAACCCACATGCACATCTCTGTC 3'. All three constructs were verified by automated sequencing.
In vitro transcription and translation of 3ß HSD constructs
All of the constructs were translated and radiolabeled in the rabbit reticulocyte-coupled transcription/translation system (Promega Corp.). The proteins were translated according to manufacturer’s instructions and labeled with either 35S-methionine or L-[4,5-3H] leucine (Amersham Pharmacia Biotech, Buckinghamshire, UK) in the case of the 3ß HSD cDNA cloned in pSP64, and only 35S-methionine for the pGEM-luc constructs. The vector pGEM-luc was also translated to generate 35S-methionine radiolabeled control protein. The translation products were analyzed by SDS-PAGE and autoradiography.
Radioimmunoprecipitation assay
For each sample, 17,000–20,000 cpm of the translation mixture was brought to a final volume of 20 µl in immunoprecipitation (IP) buffer [10 mM HEPES (pH 7.4), 150 mM NaCl, 0.5 mM methionine, 10 mM benzamidine, 0.01% BSA, and 0.5% Triton X-114) and incubated with serum at a final dilution of 1:5 overnight at 4 C. Ten microliters of Protein-A Sepharose (Amersham Pharmacia Biotech) (stock suspension washed in IP buffer twice and resuspended in 0.5 ml IP buffer per milliliter of Protein-A Sepharose) slurry was added to each sample and incubated at 4 C for 1 h with shaking. The immune complexes were washed five times in 0.5 ml IP buffer, followed by a final wash in 0.5 ml sterile water, and analyzed by SDS-PAGE and phosphoimaging. Positive controls comprising rabbit antisera to luciferase (Europa Bioproducts, Cambridgeshire, UK) and 3ß HSD (a gift from Dr. Luu-The) were included in each assay. In addition, a single SCA positive serum and a negative control serum were included in all assays.
Radioactivity associated with immunoprecipitates was analyzed by liquid scintillation counting. However, differences in counts per minute between positive and negative sera were not considered sufficient (data not shown) to be discriminative for screening large numbers of sera. For this reason, a more sensitive approach was developed, in which precipitation products were separated by 10% SDS-PAGE and gels were examined by phosphoimaging. 3ß HSD and luciferase antisera were included to indicate positive bands, and then test bands were quantified digitally in relation to the negative control serum.
Quantification of immunoprecipitation products by phosphoimaging
SDS-PAGE gels were analyzed using a phosphoimager (Fujifilm FLA2000). Briefly, the SDS-PAGE gels were exposed to imaging screens of type 20*40 (Fujifilm, Bedford, UK) for a minimum of 48 h. Then, the screens were scanned in the phosphoimager at a pixel size of 100 µm and examined using an FLA-2000 reader. The Aida 2.00 program was used to analyze the data by measuring the intensity of bands using a region determination tool. The intensity of the band was taken as the region determination reading minus the background reading obtained from an adjacent part of the gel. Results were expressed as a ratio BItest/BIneg [arbitrary units (AU)] in which BItest is the band intensity for the test serum and BIneg is the intensity of the band obtained with the negative control serum. Positivity was assigned by determining the ratios for control samples, and any sample giving a ratio greater than the 100th centile of the controls (i.e. ratio > 2.38 AU) was considered positive. Inter- and intra-assay variability of the assay was determined by using the positive and negative control sera.
A similar assay was used for analyzing reactivity to 3ß HSD N-terminus, middle, and C-terminus fragments.
HLA analysis
HLA genotyping and frequency analysis were carried out as described (12). Of the 100 patients tested in the present study, HLA data were available on 85.
Statistical analysis
The difference in prevalence of autoantibodies between patients
and controls was compared by Fisher’s exact test. The distribution of
HLA-DQß-Asp57 encoding genotypes was compared
in the patients and controls using 
2 analyses
as described previously (12).
Results
Yields of radiolabeled 3ß HSD
Table 1
shows the average labeling
efficiency obtained using each construct. Despite the presence of a
KOZAK sequence upstream of the 3ß HSD cDNA, the yields of
35S-methionine-labeled protein from
transcription/translation of the pSP64 construct were poor, resulting
in insufficient counts per minute to perform the immunoprecipitation
assays. In general, labeling is considered sufficient for
immunoprecipitation when counts are at least 5–10 times higher than
the negative control reaction. Labeling the protein with
L-[4,5-3H] leucine in these
constructs also proved to be inefficient.
|
), indicating that antigenic
epitopes on both proteins remained accessible for antibodies.
|
Translated and radiolabeled luciferase-3ß HSD fusion proteins were used to detect autoantibodies by immunoprecipitation. As stated in Subjects and Methods, radioactivity associated with immunoprecipitates detected by liquid scintillation counting was insufficiently sensitive to detect differences between positive and negative sera. For this reason, a more sensitive approach was developed in which precipitation products separated by 10% SDS-PAGE were compared using phosphoimaging and test bands quantified digitally in relation to a negative control serum.
Twelve patients (12%) with idiopathic POF were positive for
autoantibodies to 3ß HSD in this assay. None of the 103 control sera
exhibited any reactivity (Fig. 2), and
the difference in frequency of autoantibodies between patients and
controls was significant (P < 0.0001). To establish
that reactivity was due solely to the 3ß HSD portion of the fusion
protein and not due to heterophilic antibodies reacting with
luciferase, all of the positive samples were tested by
immunoprecipitation for reactivity to the luciferase fusion partner
alone and all tested negative. The interassay variation for this assay
was calculated as 5% by repeat analysis (30 separate determinations)
of the BItest/BIneg (AU)
ratio using the same SCA positive and negative control sera. The
intra-assay variation was calculated as 5.6% on the basis of testing a
single serum six times in the same assay.
|
Previously, using an immunoblot assay, we reported autoantibodies to 3ß HSD in 25 of 118 (21%) patients with idiopathic POF and 5% of healthy controls. Of these 25 positive patients, serum was available on 20 for testing in the immunoprecipitation assay. Fourteen of the 20 sera positive by immunoblot were negative by immunoprecipitation, whereas the remaining 6 were positive in both assays.
HLA associations in patients positive for autoantibodies to 3ßHSD detected by immunoprecipitation
The HLA system represents a group of highly polymorphic genes on chromosome 6 that are involved in immune function. Associations between autoimmune diseases, autoantibody positivity, and subtypes of HLA (e.g. the association between HLA-DQ8, type 1 diabetes mellitus, and glutamic acid decarboxylase autoantibodies) are used to support the contention that a disease is autoimmune. In our previous study, HLA-DQB genes encoding an aspartate at position 57 were associated with POF. We therefore examined the frequency of these genotypes in relation to autoantibodies detected using our novel assay.
The frequency of HLA-DQß-Asp57 encoding
genotypes was analyzed in the 12 patients that were positive for
autoantibodies to 3ß HSD by immunoprecipitation. In this cohort, HLA
data were available on 11 of the 12 patients; 10 of these 11 patients
(91%) were positive for a genotype encoding
HLA-DQß-Asp57; furthermore, 2 of the 10 (20%)
were homozygous for HLA-DQß-Asp57 encoding
genotypes. Previously, using an immunoblot assay, 18 of 21 (85%)
patients positive for autoantibodies were positive for
HLA-DQß-Asp57 encoding genotypes
(12). In the control population, 92 of 134 (68%) were
positive for HLA-DQß-Asp57 encoding genotypes,
and of these, 13% were homozygous. By 2
analyses, the frequency of HLA-DQß-Asp57
genotypes was not significantly different between patients and
controls, possibly reflecting the small number of autoantibody-positive
cases.
As in our previous study (12), no association was observed between autoantibody positivity and a particular HLA- DRB1 genotype.
Clinical characteristics of patients positive for autoantibodies to 3ßHSD
Eleven of the 12 patients identified as positive for autoantibodies to 3ß HSD presented with secondary amenorrhea and 1 with primary amenorrhea. Thyroid and adrenal autoantibody data were available on 11 of these 12 patients. Four patients had thyroid autoantibodies (one thyroid microsomal, one thyroglobulin, and two both thyroid microsomal and thyroglobulin), and one of these presented with clinical hypothyroidism.
Epitope mapping
An identical method to that described above was used to detect autoantibodies to each of the three fragments of 3ß HSD fused to luciferase. Sera were available on 9 of the 12 patients positive for autoantibodies to 3ß HSD. Each 3ß HSD fragment fused to luciferase was effectively immunoprecipitated by polyclonal antisera to 3ß HSD.
Figure 3 illustrates the
reactivity to each of the three fragments of 3ß HSD using sera from
patients with idiopathic POF identified as positive for autoantibodies
to 3ß HSD in the present study. None of the sera exhibited any
reactivity to fragment 2 (generated using the construct pGEM-luc-3ß
HSD399–849). Two sera (22%), displayed
reactivity to fragment 1 (generated with pGEM-luc-3ß
HSD1–450). Seven patients (77%) reacted to
fragment 3 (generated with pGEM-luc-3ß
HSD798–1116).
|
Discussion
This study was aimed at detecting autoantibodies to 3ß HSD in the sera of patients with POF using a quantifiable and sensitive radioimmunoprecipitation assay. In addition, a further goal was to determine whether reactivity was focused selectively on particular regions of the molecule.
Initially, we attempted to label the 3ß HSD protein using 35S-methionine and L-[4,5-3H] leucine and a widely used plasmid for in vitro transcription/translation, pSP64. However, neither of these methods produced sufficient quantities of labeled proteins. For this reason, the 3ß HSD cDNA was fused downstream of the luciferase gene in a similar vector, giving highly labeled fusion protein recognized by both antisera to 3ß HSD and anti-luciferase. Luciferase was selected as a fusion partner because it is highly efficiently translated and labeled in the in vitro transcription/translation system. Fusion of luciferase with the antigen of choice thus allows optimal labeling efficiency for proteins that otherwise label poorly when used alone. This is not the first study to use luciferase-tagged autoantigens in immunoprecipitation. For example, in Graves’ disease, in which autoantibodies to TSH receptor are typically present at low levels and difficult to detect by conventional antibody technology, use of a luciferase-tagged TSH receptor in an immunoprecipitation format showed good correlation with autoantibody binding that was detected using conventional biochemical assays (17). Others have produced radiolabeled 3ß HSD using 35S-cysteine (18). However, this approach was not suitable for the present study, given that one of our aims was to identify epitopic regions of 3ß HSD preferentially bound by patient sera. The lack of cysteine residues in C-terminal region (amino acids 260–372) would have precluded us from labeling with 35S-cysteine. In contrast, by producing a highly efficiently labeled fusion partner (luciferase), we were able to detect autoantibodies to any region of 3ß HSD, regardless of its amino acid constituents.
The luciferase-3ß HSD fusion protein immunoprecipitation assay that we have developed detected 3ß HSD autoantibodies in 12% of patients with idiopathic POF and in none of the healthy women we tested. These results suggest that the immunoprecipitation assay has greater disease specificity than the immunoblot assay we used previously, in which 5% of control sera were positive. Clearly, there is also a loss of sensitivity, as might be expected, and therefore fewer patients are positive in the new format assay. Our data comparing results on sera tested in the present and in our previous study (12) suggest that concordance between the immunoprecipitation and immunoblot assays is not particularly high, as has been reported by others (19). The most likely explanation for the low concordance between immunoprecipitation and immunoblot is the nature of the antigen in the two formats. In the immunoblot assay, 3ß HSD is denatured by boiling, exposure to chemical denaturants, and binding to nitrocellulose. This destroys the secondary and tertiary structure of the protein, and such antigens are referred to as linear. Binding of autoantibodies to linear antigens is dependent upon their ability to bind sites generated by the primary amino acid sequence alone (so-called linear epitopes). In contrast, in the immunoprecipitation format the antigen is in its native conformation and retains secondary and tertiary conformation. Many types of autoantibodies are dependent upon the availability of conformational structures to bind (conformational epitopes) and do not bind linearized antigen. Our results suggest that in the case of detection of 3ß HSD autoantibodies, autoantibodies binding linear epitopes, detected by the immunoblot assay, are less disease-specific than those binding conformational epitopes and detected by our novel immunoprecipitation assay.
A recent study describes a lower prevalence of autoantibodies to 3ß HSD in patients with POF (18) than we detect. This discrepancy is most likely to result from methodological differences. In particular, the system for detection of binding in our study was a phosphoimager that is more sensitive than conventional autoradiography as used by Reimand et al. (18). In addition, there are differences in the patient cohorts in the two studies. For example, in the present study the patients are tested at a much younger age. Interestingly, Reimand et al. (18) report the presence of antibodies to 3ß HSD in 20% of autoimmune polyendocrine syndrome 1 patients, but it is not specified whether any of these patients displayed gonadal failure. The 15 patients in this particular cohort were derived from an initial cohort of 45 autoimmune polyendocrine syndrome 1 patients of whom 15 presented with gonadal failure (20).
The patients identified as positive for autoantibodies to 3ß HSD in the present study displayed a strong association with HLA-DQß-Asp57 encoding genotypes, which were found in more than 90% of cases, compared with 68% in the control population. Although these results are not statistically significant, possibly as a result of the small numbers, they are consistent with our previous study. Such associations between autoantibody positivity and certain HLA-DQ types have been observed in other autoimmune diseases such as type 1 diabetes mellitus, in which autoantibodies to glutamic acid decarboxylase are associated with DQ2/8 or DQ2/X genotypes (21). Taken together, these findings support the contention that approximately 12% of idiopathic POF patients have an autoimmune pathogenesis to their disease. At present, however, it is not clear what role autoimmunity to 3ß HSD has in the disease pathogenesis. Thus far, studies have concentrated on characterization of autoantibodies to this enzyme, but it is unlikely that the autoantibodies themselves cause tissue damage, because 3ß HSD is also present in tissues (e.g. adrenal cortex) not damaged in POF patients. In the present study, we have used the Type II 3ß HSD cDNA as a source of antigen, which is predominantly expressed in adrenal and ovarian tissue (22). It is likely that the autoantibodies we detect also cross-react with the extra-adrenal, Type I form of the enzyme, because they share 93% amino acid homology. In keeping with other organ-specific autoimmune diseases characterized by autoantibodies to enzymes (23), dysfunction of enzymatic processes is rarely seen.
According to the data in this study, the major autoantibody epitopes of 3ß HSD lie in the regions bounded by amino acids 1–150 and 266–372, with no reactivity observed against region 133–283 of the molecule. All patients positive for 3ß HSD autoantibodies showed reactivity with at least the N- or C-terminal regions. It is possible that other epitopes exist but are discontinuous and dependent upon the molecule being intact. Others have inferred that the immunodominant region of 3ß HSD is near the N terminal (18), and the present study confirms that an epitope lies within amino acids 1–150. Although each of our smaller 3ß HSD constructs was immunoprecipitated by 3ß HSD-specific antiserum, we cannot exclude the possibility that fusion with luciferase interferes with autoantibody binding, and this could account for the low reactivity to region 1–150 of the molecule and the lack of reactivity to the 133–283 region.
In conclusion, autoantibodies recognizing 3ß HSD can be detected in the sera of patients with POF by immunoprecipitation of a luciferase fusion protein, and reactivity appears to be directed mainly against the N- and C-terminal regions of the molecule. 3ß HSD autoantibodies are associated with distinctive HLA-DQ genotypes. This study highlights the need for improved assays to promote concordance between studies so that the diagnostic and prognostic value of autoantibodies to 3ß HSD can be explored.
Acknowledgments
Footnotes
M.P. is supported by Diabetes UK.
Abbreviations: 3ß HSD, 3ß Hydroxysteroid dehydrogenase; AU, arbitrary units; HLA, human leukocyte antigen; IP, immunoprecipitation; POF, premature ovarian failure; SCA, steroid cell autoantibodies.
Received March 8, 2001.
Accepted September 11, 2001.
References
This article has been cited by other articles:
 |
|
 |
|
  J. Simard, M.-L. Ricketts, S. Gingras, P. Soucy, F. A. Feltus, and M. H. Melner Molecular Biology of the 3{beta}-Hydroxysteroid Dehydrogenase/{Delta}5-{Delta}4 Isomerase Gene Family Endocr. Rev., June 1, 2005; 26(4): 525 - 582. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
), fragment 2 (3ß HSD amino
acids 133–283) (
), and fragment 3, (3ß HSD amino acids 266–372)
(
). The data are depicted as binding ratios (AU) relative to control
serum as described. The cut-off for positivity is the 100th centile of
control samples.