| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Divisions of Reproductive and Developmental Sciences (A.W.H., S.v.d.D., A.E.K., M.M., P.L., H.O.D.C., W.C.D.), Pathway Medicine (S.B., P.G.), and Pathology (A.R.W.), The University of Edinburgh, Edinburgh EH16 4SB, United Kingdom; and Centre for Protein and Peptides (H.L.), Oxford Brookes University, Oxford OX3 0BP, United Kingdom
Address all correspondence and requests for reprints to: Dr. Andrew W. Horne, Obstetrics and Gynaecology, University of Edinburgh, Simpson Centre for Reproductive Health, Royal Infirmary of Edinburgh, 51 Little France Crescent, Edinburgh EH16 4SB, United Kingdom. E-mail: andrew.horne{at}ed.ac.uk.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Objective: Our objective was to investigate differential gene expression in decidualized endometrium of ectopic pregnancy.
Design: Tissue and serum analysis informed by microarray study was performed.
Setting: The study was performed at a large United Kingdom teaching hospital.
Patients or Other Participants: Women undergoing surgical termination of pregnancy (n = 8), evacuation of uterus for miscarriage (n = 6), and surgery for tubal ectopic pregnancy (n = 11) were included in the study. Endometrium was collected from normally cycling women undergoing hysterectomy.
Interventions: Decidualized endometrium was subjected to microarray analysis, morphological assessment, and immunohistochemistry. Endometrial stromal fibroblasts were cultured in the presence of decidualizing stimuli.
Main Outcome Measures: Differential expression of potentially secreted molecules was calculated.
Results: Inhibin/activin β-B expression was lower in decidualized endometrium from ectopic pregnancies when compared with that of ongoing pregnancies (P < 0.01) or miscarriages (P < 0.01). The localization of the β-B subunit was more marked in decidualized than nondecidualized stroma. Decidualization of stromal fibroblasts in vitro was associated with increased β-B expression (P < 0.05). Endometrial stroma of ectopic pregnancies was less decidualized morphologically (P < 0.05), with lower prolactin (P < 0.01) and IGF binding protein-1 (P < 0.005) expression. Serum activin B was lower in ectopic pregnancies (P < 0.005) than in intrauterine pregnancies, whereas there was no difference in progesterone concentrations.
Conclusions: Despite similar concentrations of progesterone, the endometrium of ectopic pregnancies is less decidualized than intrauterine pregnancies. Expression of the β-B subunit is related to decidualization and can be detected in the circulation as activin B. Serum activin B concentrations are lower in ectopic pregnancy.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Early diagnosis is essential for optimal conservation of fallopian tubal integrity to facilitate less invasive medical management (4) and prevent potentially life-threatening abdominal bleeding (7). Early accurate diagnosis would reduce the number of visits the patient has to make before treatment and avoid unnecessary laparoscopies, with their inherent dangers (8), in which the diagnosis is not clear. Unfortunately, a reliable blood test that can differentiate a tubal from an intrauterine pregnancy is not yet available.
A candidate-based approach for markers of tubal ectopic pregnancy has been unsuccessful in revealing markers that could be used as a single blood test for tubal ectopic pregnancy (5, 9, 10, 11, 12). We believed that the increasing availability of high-density oligonucleotide microarray technology would facilitate the identification of novel genes and signaling pathways that are important for ectopic implantation.
Rather than focus on the ectopic pregnancy tissue, or the underlying fallopian tube, we investigated the endometrial reaction to intrauterine pregnancy. Endometrial differentiation and uterine decidualization occur in early pregnancy regardless of whether the pregnancy is intrauterine or ectopic (13). We hypothesized that trophoblast invasion influences uterine decidual secretory function, and that decidualized endometrium from ectopic pregnancies would differ from decidualized endometrium of intrauterine pregnancies due to the absence of direct physical contact between the trophoblast and decidualized endometrium. We believed that secreted decidual products would be excellent candidate markers for tubal ectopic pregnancy.
This study aimed to investigate gene expression in the decidualized endometrium of tubal ectopic pregnancies using microarray analysis and compare it with the expression in the decidualized endometrium of intrauterine pregnancies that were either ongoing or miscarrying. Here, we report the identification, expression, localization, and secretion of one gene that was reduced in the decidualized endometrium of tubal ectopic pregnancies: the inhibin/activin βB subunit.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Ethical approval for this study was obtained from Lothian Research Ethics Committee, and informed written consent was obtained from all patients before sample collection. First-trimester decidualized endometrium was obtained from women (age 18–45 yr) undergoing surgical termination of pregnancy (TOP) (n = 8, group 1, mean gestation 58.7 d), surgical management of embryonic missed miscarriage (n = 6, group 2, mean gestation 57.7 d), and surgical management of tubal pregnancy (n = 11, group 3, mean gestation 58.1 d). None of the women undergoing surgical management of tubal ectopic pregnancy presented acutely with hemodynamic shock, and all required serial serum β-HCG and ultrasound monitoring before diagnosis. The ranges of serum β-HCG concentrations for each group were as follows: group 1, 42,892–162,895 IU/liter; group 2, 4,259–32,526 IU/liter; and group 3, 203–15,956 IU/liter. The decidualized endometrium and trophoblast were obtained by suction curettage from groups 1 and 2. The decidualized endometrium was obtained by suction endometrial biopsy (Pipelle; Eurosurgical Ltd., Cranleigh, UK) from group 3. The decidualized endometrium was isolated from the trophoblast macroscopically.
The decidualized endometrium was: 1) immersed in RNAlater (Ambion, Inc., Austin, TX) at 4 C overnight then flash frozen at –70 C; and 2) fixed in 10% neutral buffered formalin overnight at 4 C, stored in 70% ethanol, and wax embedded for staining with hematoxylin and eosin and immunohistochemistry. The presence of trophoblast was excluded morphologically and by using immunohistochemical staining for cytokeratin as described previously (14). Endometrial samples were obtained at hysterectomy for benign gynecological conditions from women (18–45 yr) who had regular menstrual cycles (28–35 d) and were not pregnant (15). Endometrial samples were collected at room temperature, and transported to the laboratory in RPMI 1040 (Life Technologies, Inc., Gaithersburg, MD) and processed for endometrial stromal cell isolation as described below (endometrial stromal cell isolation and culture).
RNA extraction
Total RNA was extracted from the decidualized endometrium as detailed in the manufacturer’s protocol (QIAGEN Ltd., West Sussex, UK). The concentration and quality of the extracted RNA were assessed using an Agilent bioanalyzer (Agilent Technologies, Inc., Palo Alto, CA). All samples were standardized for quality control and assigned an RNA integrity number. RNA samples were considered to be of good quality when a mean RNA integrity number value of 7.5 was obtained (16).
Microarray analysis
In brief, 4 µg total RNA from each sample was used to generate double-stranded cDNA using the One-cycle cDNA Synthesis Kit (Affymetrix, UK Ltd., High Wycombe, UK), followed by purification with the GeneChip Sample Cleanup Module (Affymetrix, UK Ltd.). The double-stranded cDNA was used as a template for the in vitro transcription using the GeneChip IVT Labeling Kit (Affymetrix, UK Ltd.) yielding biotin-labeled cRNA. After cleanup and quantification by spectrophotometric analysis, the purified biotinylated target cRNA was then fragmented into short sequences.
The hybridization cocktail consisted of 15 µg fragmented biotin-labeled cRNA spiked with eukaryotic hybridization control. Thereafter, 80 µl of the hybridization cocktail was hybridized to the test chips to check the cRNA integrity and assess the system veracity. After that, the Human HG-U133 Plus 2.0 microarrays (Affymetrix, UK Ltd.) were directly loaded with 200 µl hybridization cocktail solution and then placed in Genechip Hybridization Oven 640 (Affymetrix, UK Ltd.) rotating at 60 rpm at 45 C for 16 h.
After hybridization, the arrays were washed on a Genechip Fluidics Station 450 (Affymetrix, UK Ltd.) and scanned using a Genechip Scanner 3000 (Affymetrix, UK Ltd.) according to the manufacturer’s protocol. To ensure quality and consistency of the sample labeling process and array hybridizations, control information from all 26 arrays was collated and reviewed before the data analysis, and all were found to be consistent with Affymetrix, UK Ltd. recommendations.
Expression was calculated using the robust multiarray average (RMA) algorithm (17) implemented in the BioConductor (http://www.bioconductor.org) extensions to the R statistical programming environment (18). RMA generates a background-corrected and quantile-normalized measure of expression (19) on the log 2 scale of measurement. To investigate further the performance of the array hybridizations, scatter plots comparing each array with every other were produced in BioConductor extensions to the R statistical programming environment. These plots confirmed a linear distribution between arrays and showed a dynamic uninterrupted range of expression values from low to high signal values. Box and whisker visualizations also confirmed that the data had comparable distributions and were of sufficient quality for further analysis. Data and protocols are available to download from the publicly accessible Minimum Information About a Microarray Experiment compliant database "GPX" (www.gti.ed.ac.uk/GPX, accession no. GPX-000067.1).
After data normalization by RMA, the data were filtered to remove invariant transcripts that would contribute to multiple testing errors in the subsequent statistical analysis, using an arbitrary signal intensity threshold value of 64. After the explorative analysis, a rigorous statistical approach was exploited to identify a subset of differentially expressed genes. In brief, genes differentially expressed between the intrauterine and ectopic pregnancy samples were identified using a t test comparison, followed by multiple test correction using the Benjamini-Hochberg false discovery detection method (20). Gene lists were then created using a fold-change threshold of more than two and a corrected P value of less than 0.05.
Quantitative RT-PCR
After DNAse treatment, using RQ1 DNase (Promega, Southampton, UK), the RNA was reverse transcribed into cDNA using random hexamers (Applied Biosystems, Foster City, CA). TaqMan quantitative RT-PCR was then used to measure gene levels using Applied Biosystems’ prevalidated "assay-on-demand" specific primers and probes (Eurogentec, Southampton, UK) for prolactin and IGF binding protein (IGFBP)-1, as well as the inhibin βA subunit (M13436), inhibin βB subunit (M13437), and inhibin 
-subunit (M13144) (21), and levels were related to a ribosomal 18S internal control (Applied Biosystems).
All samples were performed in duplicate, and a relative comparison was made to an appropriate control tissue cDNA. Using the 2-
Ct method, mRNA expression results were normalized against 18S and expressed as fold change compared with controls (TOP sample group).
Immunohistochemistry
Briefly, immunolocalization of inhibin/activin βB subunit and -inhibin was preformed using antibodies developed and kindly provided by Professor N. Groome, Oxford Brookes University, Oxford, United Kingdom, and Professor Alan McNeilly, Medical Research Council Human Reproductive Sciences Unit, Edinburgh, United Kingdom, respectively. To perform the experiment in a controlled and easily repeatable manner, the Bond-X automated immunostaining machine (Vision Biosystems, Newcastle, UK) was used. Five-micron paraffin sections of uterine decidua were cut, dewaxed, rehydrated, and subjected to pressure-cooked antigen retrieval in 0.01 mol/L citrate buffer (pH 6.0) before being place on the Bond-X machine. This method on the Bond-X automated machine uses a specific polymer high-contrast program.
Slides were peroxidase blocked for 5 min, incubated for 2 h with the primary antibodies [46A/F βB mouse monoclonal antibody for inhibin βB subunit (22) and ASMR150 for -inhibin, rabbit polyclonal antibody (23)], diluted 1:500 in the diluent supplied, and then incubated with the post-primary reagent for 15 min. To confirm antibody specificity, control sections were incubated with diluent alone or with nonspecific immunoglobulins, diluted to the same concentration as primary antibody in supplied diluent. Sections were then incubated with the polymer reagent for 15 min to increase sensitivity of detection before 3'3-diaminobenzidine detection for 10 min. Sections were counterstained in hematoxylin for 5 min. Slides were then removed from the machine, and dehydrated and mounted using Pertex (Cell Path, Hemel Hempstead, UK).
Endometrial stromal cell isolation and culture
Endometrial tissue was subjected to collagenase/DNAase (Sigma-Aldrich, St. Louis, MO) digestion for 2 h at 37 C. After digestion, the stroma was dispersed, whereas the epithelial structures remained mostly intact. Human endometrial stromal cells were separated on a size basis, as previously described (24). Endometrial stromal cells were plated in six-well standard culture plates at a density of 2.4 x 105 per well in RPMI 1040 with 2% fetal calf serum, gentamycin, penicillin, and streptomycin as described previously (25). The cells were treated with medroxyprogesterone acetate (10–6 M), estradiol (10–7 M), and 8-bromo cAMP (0.1 mg/ml) for 24, 48, 72, 96, and 120 h. The cells were harvested at each time point for RNA extraction.
Serum assays
Serum progesterone concentrations were measured using a standard RIA (15). Activin B concentrations were measured using the activin B ELISA that incorporates the use of monoclonal antibody 46A/F (as both capture and detection antibody) (22) with a sodium dodecyl sulfate and heat pretreatment of samples. Serum was tested without dilution, and the assay had a lower detection limit of 19 pg/ml.
Analysis of sections
The samples of decidualized endometrium were classified blindly by a subspecialist gynecological pathologist according to the degree of endometrial change. Secretory transformation of the glandular epithelium (absent/early, high, and exhausted) and stromal differentiation (nondecidualized, pre-decidual, confluent decidual change) were scored using the classification of Zimmermann et al. (26). Samples with exhausted glandular transformation with confluent stromal decidualization were classed overall as decidualized (+), those with absent/early secretory transformation and nondecidual stroma were classified as decidualized (–), and other samples were classified as decidualized (±).
Statistical analysis (for all data outside microarray analysis)
Where the data were normally distributed, three groups were analyzed by ANOVA with Bonferroni pairwise comparisons. Two groups were analyzed using a t test. Where the data were not normally distributed, Kruskal-Wallis testing with Dunn’s pairwise comparisons was used to compare three groups. The 
2 test was used to compare proportions, and Spearman (nonparametric) or Pearson (parametric) coefficients were used to assess linear correlations. The statistical test used is given in the text and figure legend, and significance was at the P < 0.05 level.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
A comparison of gene expression in the decidualized endometrium from women with intrauterine (n = 14) and tubal pregnancies (n = 11) revealed that 669 genes were differentially expressed (FC > ±2; P < 0.05). One notable gene highlighted by this analysis was the inhibin/activin βB subunit. This was down-regulated in the decidualized endometrium from women with tubal ectopic pregnancies with a fold reduction of 2.34. Quantitive RT-PCR confirmed reduced expression in the decidualized endometrium from tubal pregnancies when compared with miscarriage (P < 0.01) and TOP (P < 0.01) groups (Fig. 1A
).
|
To be secreted from the cell, the βB subunit has to form a dimer. Because it can combine with itself to form activin B, the inhibin -subunit to form inhibin B, and the activin βA subunit to form activin AB, the expression of the and βA subunits was, therefore, also investigated. Expression of the βA subunit did not follow the same pattern (Fig. 1B
), and there were no significant differences between groups. Although the changes in expression of the inhibin -subunit between groups was less marked (Fig. 1C), it was lower in ectopic pregnancy decidualized endometrium than in that of miscarriage (P < 0.05).
Localization of inhibin and βB subunits in decidualized endometrium
To investigate further the inhibin and βB subunits in the decidualized endometrium, they were localized using immunohistochemistry. Specific immunostaining for both these proteins could be detected in the glandular and stromal areas of the decidualized endometrium (Fig. 2, A–D). However, there was variable staining in the decidual stroma between patients that was particularly marked for the βB subunit (Fig. 2, E–H). When the endometrium was less decidualized and had a more proliferative phenotype (Fig. 2E), the staining was noted to be glandular. When the endometrium had a more secretory phenotype (Fig. 2F), the staining was mainly glandular with light stromal involvement. However, in fully decidualized samples (Fig. 2G), there was marked stromal immunostaining.
|
and β-subunits during decidualization
Because there was an indication that the localization of the βB subunit in the stromal compartment of early pregnancy endometrium changed with the level of decidualization, human endometrial fibroblasts were examined in vitro. Exposure to decidualizing stimuli in vitro increased the expression of the inhibin/activin βB subunit (P < 0.05) in a time-dependent manner (Fig. 3A). There were no changes with regards to the expression of the inhibin βA subunit (Fig. 3C), although there was a significant but much less marked increase in the -subunit (P < 0.05) (Fig. 3B) during stromal fibroblast decidualization in vitro.
|
Because βB expression seems to be related to stromal decidualization, and was lower in the decidualized endometrium of ectopic pregnancies, the morphology of the glandular and stromal compartments in each tissue sample was assessed in hematoxylin and eosin-stained sections by an expert in endometrial pathology blinded to tissue identity. The glandular compartment (P < 0.05) showed a lesser degree of secretory transformation, and the stromal compartment (P < 0.05) showed less decidualization when compared with intrauterine pregnancies (Fig. 4A). Furthermore, the overall level of endometrial decidualization was less in ectopics than TOP or miscarriage (P < 0.05) (Fig. 4B). When the samples were stratified based on the overall degree of morphological decidualization (Fig. 4C), there was also a clear correlation with inhibin βB subunit expression (P = 0.0005).
|
Because these investigations had suggested that the endometrium from ectopic pregnancy showed less advanced decidualization, the expression of genes known to be involved in decidualization was also studied in the tissue samples. IGFBP1 (Fig. 5A) expression was reduced in the decidua of ectopic pregnancies when compared with that of miscarriages (P < 0.01) and TOPs (P < 0.05). Prolactin (Fig. 5B) expression was also reduced in ectopic pregnancy decidua when compared with that of TOP (P < 0.05).
|
To determine whether changes in the decidual expression of the inhibin/activin βB subunit could be detected systemically, serum activin B concentrations were assayed. Activin B concentrations were different between the groups analyzed (Fig. 6A) (P = 0.001), like progesterone (P < 0.05) (Fig. 6D). However, activin B was lower in the serum of women with ectopic pregnancies (P < 0.01) than in viable intrauterine pregnancies (Fig. 6A), unlike serum progesterone concentrations (Fig. 6E). In addition, there was a clear correlation between serum activin B (P = 0.005) and the degree of decidualization (Fig. 6C) that was not seen when progesterone was analyzed (Fig. 6F).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
It is clear that locally produced growth factors have major roles in endometrial development (27). Indeed, the TGF-β superfamily members are abundantly expressed in the endometrium. Inhibin/activin , βA, and βB subunits have previously been expressed in the endometrium across the menstrual cycle and in the decidua of early pregnancy (28). It was felt that their function was to facilitate remodeling and tissue repair after menstruation. However, because their expression is increased in early pregnancy, it is likely they have a role in the endometrial response to pregnancy (29).
There is no doubt that activin itself can affect endometrial function. Activin receptors are present on endometrial stromal cells (30), and their expression increases during decidualization (31). Indeed, treatment of endometrial stromal cells with activin A promoted decidualization and the expression of decidual markers such as prolactin (32), and treatment with follistatin significantly retarded the decidual response (32, 33). It has also been shown that activin A and inhibin A differentially regulate the expression of stromal matrix metalloproteinases, and this suggests a role for activin signaling in the remodeling associated with implantation and fetomaternal interaction (31, 34).
As well as having a role promoting decidualization, activin subunit expression itself increases during decidualization (28, 35). The regulation of this is not clear, but in vivo there seems to be a dialogue with the implanting trophoblast. During blastocyst attachment, the localization of inhibin/activin βA expression within the decidua changes, suggesting paracrine regulation (36). Therefore, it is not surprising that activins may function as markers of endometrial function and may be differentially regulated in intrauterine pregnancy.
Although most previous reports relate to the βA subunit expression and suggest a key role of activin A in decidualization, it was the βB subunit that was differentially expressed in this study. Although βB expression in decidualized endometrium has been reported previously (28), the role, regulation, and effects of activin B have not yet been assessed. Certainly, both endometrial βB subunit expression and serum activin B are highly correlated with the degree of endometrial decidualization in this study. It is not known whether it is influenced directly by neighboring trophoblast cells.
In a tubal ectopic pregnancy, HCG levels tend to be lower than in an intrauterine pregnancy. However, in this study serum progesterone, as a measure of luteal function and, therefore, serial HCG action, was no different between tubal and intrauterine pregnancy. That said, it is not known whether HCG has direct effects on endometrial decidualization and whether this is altered when there is no direct physical interaction between the trophoblast and endometrium in an ectopic pregnancy. Indeed, we hypothesized that in intrauterine pregnancy, the trophoblast would locally influence decidual function, and this interaction would be absent in ectopic pregnancy such that decidual markers would be useful for the diagnosis of tubal ectopic pregnancy. Previously assessed markers have focused on the ectopic implantation [serum creatine kinase (15)], corpus luteum function as a marker of hCG dynamics [serum progesterone (37)], or a marker of both [serum vascular endothelial growth factor (38)]. These tests have not yet proven to be of particular clinical use in the diagnosis of ectopic pregnancy.
The concept of activins and inhibins as markers of ectopic pregnancy is not new. In an early study, inhibin A and the free circulating pro-C inhibin were investigated and found not to discriminate tubal ectopic pregnancy (11). A recent study has suggested that activin A itself might function as a marker of ectopic pregnancy (39). Serum activin A was lower in ectopic pregnancies. Because both the ectopic pregnancy and the corpus luteum express βA subunits rather than βB subunits (40, 41), the source of this activin A is not clear. It seems likely that the source of activin B is the decidua, and whether this can function as an adjuvant marker for decidualization and tubal ectopic pregnancy should be assessed in prospective studies.
The demonstration of inhibin/activin βB underexpression in the decidualized endometrium of women with tubal ectopic pregnancies, associated with lower serum activin B concentrations, is clinically important. The role and regulation of activin B in decidualization should be addressed to dissect further the roles of trophoblast and progesterone in this process. In addition, the role of decidual assessment or activin B measurement in the diagnosis of ectopic pregnancy should be investigated further. We believe that further potential biomarkers of tubal pregnancy could be discovered by focusing on secreted proteins associated with uterine decidualization.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Disclosure Statement: The authors have nothing to declare.
First Published Online April 1, 2008
Abbreviations: hCG, Human chorionic gonadotropin; IGFBP, IGF binding protein; RMA, robust multiarray average; TOP, termination of pregnancy.
Received January 22, 2008.
Accepted March 20, 2008.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  C.J. Wedderburn, P. Warner, B. Graham, W.C. Duncan, H.O.D. Critchley, and A.W. Horne Economic evaluation of diagnosing and excluding ectopic pregnancy Hum. Reprod., November 20, 2009; (2009) dep397v1. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. B. Torres, P. Florio, L. Galleri, F. M. Reis, L. E. Borges, and F. Petraglia Activin A, Activin Receptor Type II, Nodal, and Cripto mRNA Are Expressed by Eutopic and Ectopic Endometrium in Women With Ovarian Endometriosis Reproductive Sciences, August 1, 2009; 16(8): 727 - 733. [Abstract] [PDF]
|
|
|
|
 |
|
 J. Cartwright, W C. Duncan, H. O D Critchley, and A. W Horne Serum biomarkers of tubal ectopic pregnancy: current candidates and future possibilities Reproduction, July 1, 2009; 138(1): 9 - 22. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A.W. Horne, W.C. Duncan, A.E. King, S. Burgess, P.C. Lourenco, P. Cornes, P. Ghazal, A.R. Williams, L. Udby, and H.O.D. Critchley Endometrial cysteine-rich secretory protein 3 is inhibited by human chorionic gonadotrophin, and is increased in the decidua of tubal ectopic pregnancy Mol. Hum. Reprod., May 1, 2009; 15(5): 287 - 294. [Abstract] [Full Text] [PDF]
|
|
Read all eLetters
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
