| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Paediatric Endocrinology Section (G.B., D.I.I., R.S., M.B.R.), University Children’s Hospital, 72076 Tuebingen, Germany; Department of Medical Genetics (A.D.), University of Tuebingen, 72076 Tuebingen, Germany; Paediatric Endocrinology (M.W.), University Children’s Hospital, 89075 Ulm, Germany; and Institute of Biochemistry II (M.S.), University Hospital, 07740 Jena, Germany
Address all correspondence and requests for reprints to: P. D. Dr. Gerhard Binder, Pediatric Endocrinology Section, University-Children’s Hospital, Hoppe-Seyler-Str.1, 72076 Tuebingen, Germany. E-mail gdbinder{at}med.uni-tuebingen.de.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
We studied a family in which seven affected males over three generations had inherited prepubertal gynecomastia in an autosomal dominant manner. The proband developed gynecomastia at 11.5 yr, entered puberty at 12.5 yr, but was incompletely virilized at 19 yr. A similar development was observed in his affected stepbrother and one first-degree cousin. All three boys had acceleration of prepubertal growth and bone age. The older two had a diminished pubertal growth spurt and precocious growth arrest, but their final heights were within the range of their target height. In addition, the maternal grandfather and three maternal uncles were affected, who all had been mastectomized. The mother of the proband had normal age at menarche and no macromastia. Estrone levels of the proband and the other affected boys were elevated, 17ß-estradiol levels were high-normal, and testosterone levels were low. Hormonal analyses of the affected adults, who had all fathered children, revealed pathologically low serum testosterone levels but normal to high-normal levels of estradiol and estrone. The mother of the proband had elevated estrone levels. Treatment of the proband was more effective with anastrozole than with testolactone and increased the initially reduced testes volume to normal size, promoted virilization, and normalized serum estrone and testosterone levels.
Neither preadipocytes from breast fat tissue of the affected stepbrother nor peripheral lymphocytes of the affected boys exhibited increased aromatase activity in culture. Therefore, these cells can be excluded from being the source of estrone excess. In addition, serum of the proband and his stepbrother did not contain factors promoting aromatase activity as assayed using preadipocytes from control individuals.
A repeat polymorphism of the p450 aromatase gene cosegregated with the disease phenotype in the family, making a mutation of the p450 aromatase gene likely. Single-strand conformational polymorphism analysis of the known alternative untranslated exons and all coding exons of the p450 aromatase gene did not indicate any mutation. In addition, fluorescent in situ hybridization analysis using four probes covering the promotor region did not reveal the presence of any major inversion at this locus.
In conclusion, preadipocytes and blood cells were excluded as the cell source of increased aromatization. Fluorescent in situ hybridization and single-strand conformational polymorphism analyses did not reveal any mutation of the p450 aromatase gene, but an intragenic polymorphic marker cosegregated with the disease phenotype. Excess of serum estrone in the presence of normal 17ß-estradiol levels may be the only indicative serum parameter of this mild manifestation of aromatase excess syndrome, which includes prepubertal gynecomastia and moderate hypogonadism in men but not necessarily short stature. In women, this mode of aromatase excess may remain clinically inapparent.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Here we describe the clinical, hormonal, biochemical, and genetic characteristics of a unique family with seven affected men over three generations and the therapeutic effects of aromatase inhibitors. Our data indicate that the aromatase excess syndrome is clinically and genetically heterogeneous. Isolated serum estrone excess may be the only detectable alteration when initial endocrine evaluation is performed.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This study was approved by the Ethics Committee of the Medical Faculty, University Tuebingen, and informed consent was given by patients and their parents. The proband is a member of a large family in which five males, his younger stepbrother, three of his maternal uncles, and his maternal grandfather, had been mastectomized because of severe prepubertal gynecomastia. In addition, one first-degree cousin developed prepubertal gynecomastia at the age of 8.9 yr but has not undergone surgery yet. The family pedigree is shown in Fig. 1
. His mother (II,6), who presumably transmitted the disturbance to both of her sons, had severe proportionate short stature (145.7 cm) but reportedly normal age at thelarche and menarche and no macromastia. The three affected uncles (II,2; II,3; II,4) were married, and each had fathered several children, the oldest boy of whom was also affected (III,11). During this study, clinical examination of the uncles, who all had short stature (165–170 cm), was not performed, but their blood samples were analyzed.
|
, indicating early growth arrest at the age of 14.3 yr. He exhibited a bilateral recurrence of gynecomastia after reductive mammoplasty, which prompted a second presentation at the age of 19 yr. At that time his final height was low (162.0 cm) but corresponded well with his target height (165.0 cm). His virilization was incomplete with no facial hair at all, narrow shoulders, small testes (12 ml each), and genital status G4 according to Tanner. He was lean with a body mass index of 19 kg/m2. Karyotyping was normal (46,XY). For the promotion of his virilization, a therapeutic trial with the aromatase inhibitor testolactone (5 mg/kg·d) was started. During this trial, testes volume increased from 12 to 15 ml, and the penis grew in length and diameter. Testosterone serum levels rose to near normal levels, but estrone serum levels remained elevated. Later, at the age of 20 yr, the new-generation aromatase inhibitor anastrozole (1 mg/d) was given instead of testolactone, which caused a continuous normalization of serum androgen and estrogen hormone levels (Fig. 3).
|
|
). Like his brother, he started puberty early. At the age of 15 yr, his bone age was 18 yr, and his near-final height was 176.2 cm (target height 169.4 cm). Small testes (each 10 ml), no facial hair, and genital stage G4 were in correspondence with the insufficient virilization, mirroring the clinical manifestations of the proband. One first-degree cousin (III;11) developed prepubertal gynecomastia at the age of 8.5 yr. He was prepubertal (genital stage G1, pubes stage PH1) at presentation (age 9.0 yr) with a testes volume of 2 ml and had unilateral breast stage B2 according to Tanner with a bilateral diameter of 2 cm. After 6 months, he developed bilateral thelarche. His height (140.4 cm at the age of 9 yr) was +1.40 SD score above his target height, and his weight was normal for height (32.4 kg). The bone age was accelerated by 3.5 yr, which was suggestive of hormonal activity long before thelarche had started. Because of the limited final height prognosis (<160 cm, target height 171 cm), he was started on anastrozole, 1 mg daily, which caused a fast regression of breast volume.
The stepsister (III;9) of the proband was accidentally diagnosed with late-onset adrenal hyperplasia during this study. She had short stature (151.1 cm) after early menarche (10.9 yr), signs of virilization (moustache, acne, pubes stage PH6), and small mature breasts. Her serum testosterone and 17-hydroxyprogesterone levels were elevated. The genetic analysis of her CYP21B gene identified a missense mutation (Val281Leu) on one allele and a deletion on the other. The missense mutation was inherited from her father (II;7), whereas the deletion was inherited from the mother. Both the proband and his stepbrother were heterozygous for a CYP21B mutation.
Hormone measurements
All blood samples were drawn early in the morning. Basal hormonal analysis was performed with blood from the proband (III,8), his stepbrother (III,10), his cousin (III,11), his mother (II;6), his stepfather (II,7), the three affected uncles (II,2; II,3; II,4), the one nonaffected uncle (II;5), and the affected grandfather (I,1). The proband’s biological father was not available for this study. Functional endocrine testing was performed using LHRH administration in the proband, his stepbrother, and the affected cousin, using additional human chorionic gonadotropin administration only in the proband.
Total testosterone levels were measured by solid-phase 125J RIA in unextracted serum (Coat-A-Count total testosterone, Diagnostic Products Co., Los Angeles, CA). Cross-reactivity to natural steroids was much less than 1%, to 5
-dihydro testosterone 3%. High cross-reactivity to testolactone (aromatase inhibitor therapeutically used in this study) was determined by comparison with mass spectrometry measurements, which indicated that testosterone levels were overestimated by 40% in the presence of the testolactone medication. The interassay coefficient was 6.4–6.7%, and the intraassay coefficient was 6.0%. Normal testosterone reference values of young male adults were determined to be above 350 ng/dl in this assay.
17ß-Estradiol (E2) levels were determined by RIA (Estradiol MAIA, Adaltis Co., Casalecchio di Reno, Italy). The detection limit of this assay was 5 pg/ml. Cross-reactivity was much less than 0.1% to all natural sex steroids, 0.47% to estriol, and 1.77% to estrone (E1). The intraassay coefficient was 2.25%, and the interassay coefficient was between 2.9 and 3.5% at 12.8 pg/ml according to the manufacturer. E2 serum levels of young male adults are less than 50 pg/ml using this assay.
E1 serum levels were measured by RIA [Diagnostic Systems Laboratories (DSL), Sinsheim, Germany]. Cross-reactivity to E2 was 1.25%, the intraassay coefficient was 5.6%, and the interassay coefficient was 11.1% at 90–100 pg/ml according to the manufacturer. Normal E1 levels for young male adults are less than 60 pg/ml using this assay.
Androstenedione serum levels were determined by active androstenedione RIA (DSL). Cross-reactivity to natural androgens was less than 0.33%. However, we observed an extremely high cross-reactivity to testolactone. Comparison with mass spectrometry results indicated an overestimation of the androstenedione levels in the presence of testolactone medication by more than 2000%. Therefore, the high androstenedione levels measured by the assay during testolactone therapy were an artifact, which has already been reported by others (7). The intraassay coefficient was 5.6%, and the interassay coefficient 9.8% at 60–70 ng/dl according to the manufacturer.
The LH and FSH serum levels were measured by immunometric assay (Immulite LH/FSH, DPC). Both assay antibodies were highly specific to LH and FSH. The LH intraassay coefficient was 4.8%, and the interassay coefficient was 8.1–11.6%; the FSH intraassay coefficient was 5.4%, and the interassay coefficient was 8.1%, according to the manufacturer.
Determination of the aromatization of 3H-radiolabeled testosterone by preadipocytes
Preadipocytes from breast fat tissue removed during bilateral reductive mammoplasty in the affected stepbrother were isolated as previously described (8). These preadipocytes were cultured in 24-well plates in medium M199 supplemented with 10% fetal calf serum. When confluence was reached, the cells were washed and cultured serum free for 2 d to reduce the aromatase activity to basal levels. Then the cells were incubated for 24 h with or without 1 µM cortisol in the presence or absence of 10% human serum. During the last 6 h of the incubation, 150 nM [1ß,2ß-3H]testosterone (specific activity 1491.1 Gbq/mmol) was added. Its aromatization caused the release of 3H-H2O, which was measured after extraction as previously described (9). Values were adjusted to the total protein content of the preadipocytes in each culture well studied.
Measurement of the aromatase-inducing activity in the serum
Normal human preadipocytes from three consecutive donors were cultured and incubated as described above. Serum of normal controls, the proband, and his stepbrother were added to the cells in concentrations of 0, 0.1, 1, and 10% during the induction period of the experiments. Aromatization was measured as described above.
Determination of the aromatization of 14C-radiolabeled androstenedione by blood cells
Whole blood from the proband and his stepbrother or the extensively washed cells derived from it were diluted to the 3-fold volume and plated in a total volume of 3 ml/well in 6-well plates. After 6 h cells were incubated with 249 nM [4-14C]androstenedione (specific activity 1983.2 MBq/mmol) for various times (1–5 d). After that, steroids were extracted and analyzed by two-dimensional thin-layer chromatography as described previously (10). Products were quantified on a PhosphorImager system (Molecular Dynamics, Sunnyvale, CA) and in addition visually analyzed after exposure to x-ray film for 28 d.
Tetranucleotide repeat polymorphism
PCR was performed using 50 ng genomic DNA as a template and 25 pmol of each oligonucleotide primer. Oligonucleotide primers (MWG-BIOTECH, Ebersberg, Germany), which were labeled with a fluorescent dye, were chosen as described by Polymeropoulos et al. (11). The reaction conditions were denaturation at 95 C for 1 min, annealing at 55 C for 1 min, and extension at 72 C for 1 min, and the number of cycles was 27. Separation and visualization of the fluorescent products were performed in the LI-COR automatic sequencing apparatus 4200 (LI-COR Biosciences GmbH, Bad Homburg, Germany) under denaturing conditions.
Single-strand conformational polymorphism (SSCP) analysis
SSCP analysis was performed for detection of point mutations of the p450 aromatase gene. The nine coding exons were amplified by PCR as described previously (12). The untranslated alternative exons 1.1, 1.2, 1.3, 1.4, 1.5, 1.6, and 1.F were amplified as described (13, 14, 15, 16, 17, 18). One microliter of PCR product was mixed with 3 µl formamide, heated for 3 min at 95 C, and immediately put in ice water. Single-stranded DNA was separated by PAGE (12.5%) under nondenaturing conditions using PhastSystem-technology (Amersham Biosciences, Uppsala, Sweden). Automatic silver staining was used for detection of the single stranded fragments.
Fluorescent in situ hybridization (FISH) analysis
Chromosome preparations were made from phytohemagglutinin-stimulated peripheral blood lymphocytes. FISH was performed using the following bacterial artificial chromosomes (BACs): RP11–105D1, RP11–522G20, RP11–313P18, and RP11–108K3 (ResGen, Invitrogen Corp., Paisley, UK). The unique probes containing sequences that spanned and flanked the p450 aromatase gene locus were selected from the tile path clones at the ensemble human genome database (www.ensemble.org/homo_sapiens/cytoview) (see Fig. 6A). Bacteria were cultured in Luria Bertani (LB) medium supplied with chloramphenicol, 12.5 µg/ml. BACs were isolated using the EndoFree plasmid maxi kit (Qiagen GmbH, Hilden, Germany). The sequences were amplified using degenerated oligonucleotide primed PCR with 35 cycles of denaturation at 94 C, annealing at 40 C, and extension at 72 C (19). Probes were directly labeled by standard nick translation procedures with SpectrumGreen deoxyuridine 5-triphosphate or SpectrumOrange deoxyuridine 5-triphosphate (Vysis, Abbott Laboratories, Abbott, IL). FISH was carried out according to the method reported by Pinkel et al. (20) with minor modifications. Chromosomes were counterstained using 4',6'-diamino-2-phenylindole and mounted in antifading solution (Vector Laboratories, Burlingame, CA). Images were captured on a Axioplan II microscope (Zeiss, Göttingen, Germany) and the ISIS digital FISH imaging system (MetaSystems, Altlussheim, Germany).
|
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
). Estradiol was not increased in absolute terms, but the testosterone to estradiol ratio was pathologically low due to the low testosterone levels. This hormonal situation sufficiently explained the gynecomastia and the incomplete virilization of the proband. The low basal and stimulated gonadotropin levels suggested a central disturbance of the gonadal axis (Table 1B) and rendered a primary synthesis defect of testosterone unlikely. This was confirmed by the normal increase of testosterone after administration of ß-human chorionic gonadotropin, which was accompanied by a similarly steep increase of estradiol (Table 1C) (21, 22, 23).
|
|
|
and Table 1). Surprisingly and in contrast to the affected adults, his E1 level was mildly elevated; this increase may still be in the upper range of the normal variability.
|
The combination of prepubertal gynecomastia, acceleration of prepubertal growth rate and bone maturation, incomplete virilization, the hormonal findings, and the presumably dominant inheritance of the disorder are consistent with the so-called aromatase excess syndrome caused by dysregulated and elevated aromatization of androgens to estrogens in various tissues (4).
This hypothesis was confirmed in vivo by the therapeutic intervention that blocked aromatase activity: a low-dose therapy with a first-generation aromatase inhibitor, testolactone, caused an increase in serum testosterone levels to near normal values and normalization of testes size of the proband, whereas serum E1 levels stayed elevated (Fig. 3). Therefore, a trial with the potent new aromatase inhibitor anastrozole (1 mg/d) was started, which induced a strong rise of the serum testosterone level and a robust decrease of the serum estrone level in the proband, inducing for the first time a normal endocrine homeostasis, which led to a sufficient promotion of virilization (facial hair and penile growth). After discontinuation of this therapy, E1 went up to the highest levels ever measured and testosterone was pathologically low again (Fig. 3). The still prepubertal affected cousin of the proband showed a decrease of breast size and normalization of E1 levels 3 months after start of therapy with anastrozole (1 mg/d).
For tracing the source of estrogen synthesis, we examined the in vitro aromatization activity of blood cells of the proband and the affected stepbrother and preadipocytes of the affected stepbrother. The aromatase activity of the preadipocytes, which was quantitatively determined by the release of 3H-H2O from [1ß,2ß-3H]testosterone, did not differ from control cells under basal and cortisol-stimulated conditions in the absence of serum (Fig. 4, A–D). This observation excluded constitutively active aromatase in the preadipocytes or the presence of autocrine or paracrine inducers, which may act via promotor II (under basal conditions) or promotor 1.4 (in the presence of cortisol).
|
, A and C). The inducing effects of sera tended to be more pronounced in normal human preadipocytes, this difference being significant in cortisol-stimulated cells (Fig. 4, C and D). Therefore, an increased susceptibility to aromatase induction was excluded for the preadipocytes from the affected stepbrother.
In normal human preadipocytes, normal serum and sera of both patients (proband and stepbrother) acted in an identical manner. That was also seen in dose-response studies (0.1–10% serum used, data not shown). However, both patients’ sera led to a significantly attenuated aromatase induction in preadipocytes from the affected stepbrother (Fig. 4C). These observations excluded the presence of elevated concentrations of aromatase inducers in the serum of the affected.
In summary of the preadipocyte studies, our in vitro data did not provide evidence for the preadipocyte, which is the main aromatizing cell of the fat tissue to be the source of the E1 excess. In addition, there was no evidence for an increased amount of circulating aromatase inducers in the patients’ sera. Moreover, preadipocytes from the affected stepbrother exhibited a blunted response to the aromatase induction by serum, which was even more pronounced when the patients’ sera were used for induction.
Incubation of blood cells from both the proband and his affected stepbrother with [4-14C]androstenedione for 1–5 d did not result in a detectable E1 or E2 synthesis in vitro as analyzed by two-dimensional thin-layer chromatography (data not shown). In addition, there was no difference in formation of testosterone from androstenedione by blood cells in this assay indicating normal 17ß-HSD activity [means (SD) of total steroids; proband: 27% (11) vs. normal control: 29% (11)]. Thus, blood cells did not contribute to the altered steroid profile.
These in vitro observations did not rule out a mutation of the p450 aromatase gene, especially within the promotor region of the gene, because the aromatase overexpression could be present in a tissue that was not available for our biochemical studies. Therefore, we analyzed a p450 aromatase-linked repeat polymorphism and detected the same allele in all affected males and the mother of the proband (Fig. 5). SSCP screening for mutation of all coding exons and all 5' alternative untranslated exons of the p450 aromatase gene was performed but did not reveal any conformational polymorphism indicative of a gene mutation (data not shown). Finally, FISH analysis of the p450 aromatase gene locus, using four BAC probes covering about 0.8 Mb of the p450 aromatase gene locus, did not reveal a major structural aberration, which has been recently described in three males affected by the aromatase excess syndrome (5) (Fig. 6, A–F).
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
As a consequence of prepubertal growth acceleration, pubertal growth spurt was lacking in the two older affected boys in this study. However, in contrast to data from previous studies in which bone age acceleration was always more advanced than in this study, final height of our adolescents was not compromised by the estrogen excess. Therefore, short adult stature does not seem to be a necessary sign of aromatase excess syndrome.
Estrogen excess was accompanied by diminished FSH and LH serum levels, insufficient pubertal growth of the testes, and diminished testosterone serum levels. This central hypogonadism, which was a manifestation of the negative feedback of the increased estrogens on the pituitary-gonadal axis (26), disappeared when pharmacological aromatase inhibition was provided. Interestingly, peripheral quantitative computed tomography of the radius revealed normal bone density in the proband and his stepbrother, suggesting that this hypogonadism in the presence of estrogen excess did not cause osteoporosis.
Notably, the aromatase excess was not associated with infertility in the affected men. Similar observations have been reported by others (1, 3, 4). This may partly be explained by the fact that serum E1 levels of the affected adult men were not elevated, indicating a trend of normalization of the estrogen excess with age. These changes may be explained by the natural decrease of serum dehydroepiandrosterone, which is the main prohormone for peripheral estrogen synthesis in men (27). However, the normal sperm count in the adolescent proband may also indicate that in general moderate estrogen excess and moderate testosterone deficiency is compatible with normal spermatogenesis and fertility.
Because E1 is a weak estrogen, the degree of gynecomastia and bone age acceleration observed is somewhat surprising. However, serum levels of E1 and E2 may incompletely reflect the hormone levels at the site of peripheral action because peripheral tissues like skin, mammary gland, or bone cells provide 17ß-hydroxysteroid dehydrogenase activity to effectively convert E1 to E2 locally. The major location for peripheral E1 formation is the adipose tissue. Interestingly, our biological analysis excluded preadipocytes of breast fat as well as peripheral blood lymphocytes from being the source of the estrogen excess observed. Moreover, the preadipocytes from one affected subject exhibited a blunted response to aromatase induction by serum in comparison with control cells. Also, no serum factors excessively inducing aromatase were detectable in the affected subjects when studied by the preadipocyte assay. Previous in vitro studies on hereditary prepubertal gynecomastia provided evidence of excessive aromatase activity in cultured skin fibroblasts (4, 5) and Epstein-Barr-virus-transformed peripheral lymphocytes (4) but not in peripheral blood lymphocytes (4) as in this study. Adipocytes and their precursor cells had not been examined in detail before. Skin fibroblasts were not available for analysis in this study. Taken together, our experimental data suggest that the source of the estrogen excess was located outside the (breast) adipose tissue and blood and was not driven by serological factors.
We found cosegregation of a known polymorphism at the p450 aromatase gene with the disease phenotype. However, SSCP analysis of the coding region and the various 5' untranslated exon 1 did not reveal any mutation. FISH analysis using four BAC probes covering about 0.8 Mb of the p450 aromatase gene locus was not indicative of an inversion, which has recently been described as the molecular basis of the aromatase excess syndrome in two affected families (5). The affected men in these families, however, presented with an earlier manifestation of aromatase excess and much higher serum E2 levels than our proband and his affected relatives. Nevertheless, smaller inversions within the promotor region not detectable by the used FISH probes or point mutations at the same locus, which escaped our mutational screening method, cannot be entirely excluded.
In conclusion, when our data are compared with previous reports, the hormonal, biochemical, and genetic basis of the aromatase excess syndrome is demonstrated to be heterogeneous. Notably, the disorder cannot be ruled out by isolated determination of E2: excess of serum E1 is the only indicative parameter of the milder manifestation of this syndrome. Cosegregation of a marker and the disease made the presence of a mutation at the p450 aromatase gene likely, which must be different from the previously reported gain-of-function mutations caused by major inversions within the promotor region. Treatment with anastrozole provides a measure for normalization of the disturbed hormonal homeostasis and promotion of virilization.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
First Published Online October 13, 2004
Abbreviations: BAC, Bacterial artificial chromosome; E1, estrone; E2, 17ß-estradiol; FISH, fluorescent in situ hybridization; SSCP, single-strand conformational polymorphism.
Received August 5, 2004.
Accepted September 23, 2004.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
-hydroxyprogesterone, 
4-androstenedione, estrone, and 17ß-estradiol: evidence for the action of human chorionic gonadotropin on intermediate enzymes implicated in steroid biosynthesis. J Clin Endocrinol Metab 49:284–291This article has been cited by other articles:
 |
|
 |
|
  I. Czajka-Oraniec, W. Zgliczynski, A. Kurylowicz, M. Mikula, and J. Ostrowski Association between gynecomastia and aromatase (CYP19) polymorphisms Eur. J. Endocrinol., May 1, 2008; 158(5): 721 - 727. [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 |
