| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Clinical Biochemistry, Bispebjerg Hospital, University of Copenhagen, DK-2400 Copenhagen NV, Denmark
Address all correspondence and requests for reprints to: Jan Fahrenkrug, Professor, M.D., D.Med.Sci., Department of Clinical Biochemistry, Bispebjerg Hospital, DK-2400 Copenhagen NV, Denmark. E-mail: bbhjanf{at}inet.uni2.dk.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Circadian output from the SCN plays a major role in the regulation of female reproductive rhythms (9), and in the rat neuronal projections from the SCN to the GnRH neurons have been demonstrated (10). Furthermore, estradiol acts permissive to prime GnRH and LH release (9, 11). Female rats are spontaneous ovulators, and under standard laboratory conditions with a 12-h light, 12-h dark cycle, the estrous cycle lasts for 4 d. One of the hallmarks of the cycle is the midcyclic gonadotropin peak, which occurs on the afternoon of proestrus with subsequent ovulation approximately 12 h later. Changes in the duration of the light/dark periods cause altered timing of the gonadotropin peak and estrous cycle (12), and destruction of the SCN as well as constant light lead to a state of constant estrus and a change from spontaneous to induced ovulation (13, 14, 15). Furthermore, female clock mutant mice, which carry a 51-amino acid deletion in the transcriptional activation domain of the CLOCK protein gene, display disrupted estrous cycles and inability to produce a gonadotropin peak (16).
Information on the expression of clock genes in mammalian gonads is limited to the testes (17, 18, 19, 20, 21), and both arrhythmic and oscillating expressions have been reported. Most studies, however, have shown that the pattern of expression is developmental rather than circadian, confined to specific stages of spermatogenesis (17, 18, 19). Occurrence of clock genes in the mammalian ovary has not yet been reported, but noncyclic expression of Per and cytoplasmic localization of PER have been reported in the Drosophila ovary (22, 23, 24).
In the present study, we examined the spatiotemporal expression of the two core clock genes, Per1 and Per2, in the rat ovary and report that they oscillate with a period of approximately 24 h under 12-h light, 12-h dark conditions and in continuous darkness. Cyclic changes in Per mRNAs were followed by a rise and fall in PER proteins with a lag of several hours. In situ hybridization histochemistry showed that Per mRNAs were localized to the steroid-producing cells in the ovaries, and by double immunocytochemistry the dynamic intracellular movements of PER1 and PER2 proteins were visualized.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Experimental design
To examine the temporal expression of the Per genes in the rat ovaries during the 4-d estrous cycle, we first examined the mRNA expression during the 2 d of proestrus and estrus. Marked changes in morphology and hormonal output occur during these 2 d in which the primary estradiol-producing preovulatory follicles mature and rupture to release the oocyte. Subsequently the cells luteinize and become progesterone-producing corpora lutea. Thirty-three rats were entrained to 12-h light, 12-h dark cycle [Zeitgeber time (ZT; time-point of exogenous scheduled entraining cycle) 0 = lights on and ZT 12 = lights off] and followed with daily cellular profile analysis of vaginal smears for at least three consecutive 4-d cycles. Three animals were killed by decapitation every fourth hour corresponding to the following ZTs during the 2 d of proestrus and estrus: ZT 22, ZT 2, ZT 6, ZT 10, ZT 14, ZT 18, ZT 22, ZT 2, ZT 6, ZT 10, and ZT 14. The ovaries were rapidly removed and frozen on dry ice. One ovary from each animal was used for quantification of Per1 and Per2 mRNA by real-time RT-PCR.
The expression of the Per genes in the ovaries during proestrus and estrus was obviously found to be diurnal, and subsequently we examined the daily expression of the two genes in a population of rats representing a mixture of all 4 d of the cycle. Forty-five rats were kept in a 12-h light, 12-h dark cycle. The estrous cycle of each animal was determined by cellular profile analysis in vaginal smears. Five animals were killed at each of the following time points: ZT 0, ZT 2, ZT 6, ZT 10, ZT 12, ZT 14, ZT 16, ZT 18, and ZT 20. It was ensured that each time point represented a mixture of stages in the estrous cycle. The ovaries were rapidly removed, and one ovary from each animal was used for quantification of Per1 and Per2 mRNA by real-time RT-PCR, whereas the other ovary was used for in situ hybridization histochemistry or immunocytochemistry (n = 3 and n = 2, respectively, at each time point). Ovaries for real-time RT-PCR and in situ hybridization histochemistry were rapidly frozen on dry ice, whereas ovaries for immunocytochemistry were immersion fixed in Stefanini’s fixative overnight, cryoprotected in 30% sucrose, and frozen.
Finally we examined the expression of the Per genes during constant environmental conditions. Forty-five rats were entrained to a 12-h light,12-h dark cycle for at least 14 d after which the light was turned off. On the second cycle after the transfer into continuous darkness, five animals were killed at each of the following time points: ZT 0, ZT 2, ZT 6, ZT 10, ZT 12, ZT 14, ZT 16, ZT 18, and ZT 20. The ovaries were rapidly removed and treated as described above.
RNA preparation, cDNA synthesis, and real-time PCR analysis
Total ovary RNA was prepared by the guanidinium thiocyanate-phenol-chloroform extraction method (25). cDNA from 1 µg RNA was synthesized using TaqMan reverse transcription reagents (Applied Biosystems, Foster City, CA) in a total volume of 20 µl. After the cDNA synthesis, the samples were diluted 5-fold yielding cDNA synthesized from 10 ng total RNA per microliter. For the standard curves; a pool of ovary RNA was used to make two large batches of cDNA used for the Per1 and ß2-microglobulin (B2m) and Per2 and B2m standard curves, respectively. Five serial 5-fold dilutions were made and frozen in aliquots. The most concentrated sample held cDNA from 50 ng total RNA per microliter and the least concentrated cDNA from 80 pg total RNA per microliter. Doublets of 2.5 µl of each standard were assayed in each run; the highest standard was arbitrarily set to 12,500 and the lowest to 20.
Real-time PCR was performed using an ABI7000 instrument and Taqman chemistry (Applied Biosystems). The primers and TaqMan probes were designed using Primer Express software (Applied Biosystems). The probes for the Per1, Per2, and B2m assays all had a 3'-TAMRA-quencher, the Per1 (6-FAM-CAG CAA GAG TAC AAA CTC ACA GAG CCC ATC C-TAMRA) and Per2 (6-FAM-AGC CCC AGC AAG TGA TCG AGG ACT AAG-TAMRA) probes had a 5'-6-FAM-reporter, whereas the B2m-probe (VIC-TCC CCA AAT TCA AGT GTA CTC TCG CCA TC-TAMRA) used as internal control had a VIC reporter. It was verified that the amount of B2m mRNA used as internal control did not vary as function of time (data not shown). The probes used in all three assays were spanning exon boundaries, thus excluding detection of genomic DNA. Verification for this was done by running samples of 25 ng rat genomic DNA. The amount of DNA was chosen to equal the amount of RNA input in each of the experimental cDNA samples analyzed. No amplification of genomic DNA was seen using any of the assays. All samples, standards, and the nontemplate-negative controls were made in duplicate in a final volume of 25 µl. The PCRs were done using TaqMan universal PCR master mix containing AmpErase7UNG (Applied Biosystems). In the Per1 assay, 150 nM probe, 300 nM forward primer (AGC TCT GCT GGA GAC CAC TGA), and 400 nM reverse primer (CAC TCA GGA GAC TAT AGG CAA TGG A) were used. In the Per2 assay, 200 nM of probe, forward (GCA GGC TCA CTG CCA GAA CT) and reverse primer (CAA GAT GAT TCT ATT CCA GAA GCA TT) were used, and in the B2m assay, 150 nM probe and 300 nM forward (CGT GCT TGC CAT TCA GAA AA) and reverse primer (GAA GTT GGG CTT CCC ATT CTC) were used.
The ABI prism 7000 SDS software program (Applied Biosystems) was used to calculate the concentrations (in arbitrary units) of Per1 or Per2 and B2m mRNA. The amount of Per1 or Per2 mRNA in each sample was normalized with the amount of B2m mRNA from the sample obtained on the same plate.
In situ hybridization histochemistry
Frozen ovaries taken at time points at which Per mRNAs as revealed by real-time RT-PCR were low (Per1: ZT 2; Per2: ZT 6) and high (Per1: ZT 14; Per2: ZT 18) were cut in 12-µm-thick sections. In situ hybridization histochemistry was performed as previously described using 33P-labeled Per1 and Per2 cRNA probes (26). After hybridization, washing, RNase treatment, and a final wash, the radioactively labeled ovary sections were dried, emulsion dipped (Amersham Biosciences, Little Chalfont, UK), and exposed for 7–14 d before being developed. Sections were counterstained with cresyl violet acetate. For control purposes hybridizations were performed in parallel using antisense and sense probes on consecutive sections. Images were obtained via a DC 200 camera (Leica, Wetzlar, Germany) using Leica DC 200 software.
Immunohistochemistry
Two cDNA fragments, one encoding the N-terminal part of mouse Per1 (404 amino acids) and another encoding a long C-terminal part of rat PER2 (836 amino acids) were generated by RT-PCR and subcloned in the vector pCRT7/NT-TOPO (Invitrogen, Taastrup, Denmark). The two recombinant proteins were expressed in Escherichia coli BL21 (DE3) pLysS and extracted, solubilized, and purified according to our previously described procedure (27).
Five rabbits were immunized with 50 µg of the PER1 protein and another five rabbits were immunized with 50 µg of the PER2 protein with intervals of 10 d. Serum from each bleeding was tested by immunoblotting for reaction with the purified recombinant proteins. In addition, Western blot analysis and immunostaining were performed on rat suprachiasmatic nucleus and ovary. Serum from rabbit no. 298 (PER1) and 015S (PER2) was selected. In Western blots these antisera showed distinct bands of the expected molecular size (
150 kDa), and the reactivity was removed by absorption of the anti-PER1 and anti-PER2 antisera with immunization material.
Double immunohistochemistry for visualization of PER1 and PER2 was performed as described in detail previously using the protocol on two primary antibodies from the same species (28, 29). In brief, 12-µm sections were incubated overnight with the PER2 antiserum (diluted 1:80,000) followed on the second day by biotinylated donkey antirabbit antiserum (Jackson ImmunoResearch Laboratories, Inc., West Grove, PA) and visualized by biotinylated tyramide (tyramide system amplification, DuPont NEN Life Science Products, Boston, MA) and streptavidin conjugated Texas Red (Amersham Bioscience). After washing and blocking in 1% H2O2 for 10 min, the sections were incubated overnight with anti-PER1 antiserum (diluted 1:5000). On the third day, the anti-PER1 antibody was visualized using a horseradish peroxidase-conjugated goat antirabbit antiserum (from kit no. 12, code no. T20922, diluted 1:800; Molecular Probes, Eugene, OR) followed by incubation with Alexa-Tyramide 488 (diluted 1:200; Molecular Probes). Some slides were counterstained with 4', 6'-diamidino-2-phenylindole for visualization of cell nuclei. Preabsorption with the respective antigens and/or elimination of the primary antiserum abolished specific staining of the respective antigen (see Fig. 6
, I and J).
|
Statistical analysis
Levels of Per mRNAs were presented as means ± SEM. Diurnal and circadian changes in Per1 and Per2 mRNAs were analyzed using the methods for cosinor-rhythmometry as described by Nelson et al. (30). The data were thus fitted to a combined cosine and sine function: Per = M + k1COS(2
t/24) + k2SIN(2t/24). Substituting COS(2t/24) = X and SIN(2t/24) = Z gives the expression: Per = M + k1X + k2Z. The model fit was then tested using the GLM procedure in the SAS statistical software package (31). P < 0.05 was considered statistically significant.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). Statistical analysis showed that both Per1 mRNA (P = 0.0003) and Per2 mRNA (P < 0.0001) changed significantly as a function of a 24-h-cycle. To substantiate that the oscillations in Per mRNAs were diurnal throughout the ovarian cycle, we performed a 24-h experiment in another group of female rats in which each time point was represented by ovaries from mixed stages of the 4-d estrous cycle. In this experiment, ovarian Per mRNAs also displayed a significant (Per1: P = 0.02; Per2: P = 0.0001) diurnal rhythm of expression (Fig. 2) similar to that observed during proestrus and estrus. Thus, the Per1 mRNA level was low at ZT 0–2 and peaked at ZT 12–14, whereas the Per2 mRNA level from a trough at ZT 5–6 gradually increased, reaching a peak at ZT 17–18, after which it declined again.
|
|
, A and C) and Per2 mRNA (Fig. 3, E and G) were weak at ZT 2 and ZT 6, respectively. At ZT 14 (Per1) and ZT 18 (Per2), the corpora lutea were strongly labeled, and labeling was also observed in the granulosa cell compartments and the theca cell compartments of growing follicles at preantral, antral, and preovulatory stages as well as in cells of the interstitial glandular tissue (Figs. 3, B, D, F, and H, and 4). All these cell types are steroid producing. No signal was detected in the oocytes. Sense probes for Per1 and Per2 mRNA produced no signal at the times of peak expression (Figs. 3, D and H, insets, and 4).
|
|
).
|
). At ZT 6 weak PER1 immunostaining was detected in the cytoplasm of a few cells (Fig. 6A). The cytoplasmic PER1 immunostaining increased during the following hours, and at ZT 12 most cells showed PER1 immunostaining in their cytoplasm (Fig. 6C). Subsequently PER1 was transferred to the nuclei of the cells, and at ZT 16 cytoplasmic staining was less intense, whereas nuclear PER1 immunoreactivity became apparent. From ZT 16 to ZT 0, PER1 was hardly visible in the cytoplasm, whereas intense nuclear immunostaining was observed (Fig. 6, E and G).
The pattern for PER2 immunostaining was similar to that of PER1. The changes in PER2 protein levels and localization were as the mRNA, however, delayed by approximately 4 h in relation to PER1. Thus, at ZT 12 PER2 protein level was low and hardly detectable in the cytoplasm (Fig. 6D). Until ZT 16 the cytoplasmic staining increased (Fig. 6F), and from ZT 18 to ZT 2, a nuclear localization of Per2 was evident (Fig. 6H). At ZT 6 the nuclear Per2 immunostaining was faint (Fig. 6B), and it disappeared completely at ZT 10.
The displaced time course of the changes in intracellular localization of the two PER proteins is clearly illustrated in the high-power merged images (Fig. 7), e.g. at ZT 16 at which Per2 is still in the cytoplasm and Per1 has a nuclear localization.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Previously we (26) reported that Per1 mRNA in the rat SCN showed maximum expression at midday (ZT 6) and minimum expression at night. Per2 mRNA in the SCN was low during early day and late night but peaked at the transition time between day and night (ZT 12). Thus, the rhythms of Per1 and Per2 gene expression in the ovary were phase delayed by approximately 6 h relative to the SCN. In accordance, a delay of 3–9 h has been reported for most peripheral circadian clocks (4). It has been demonstrated that humoral or neural outputs act as functional connections between central and peripheral clocks (37). It remains to be clarified whether the ovarian clock is under the influence of the brain’s biological clock located in the SCN. The Per genes and their gene products were shown to be expressed in steroidogenic cells, which are all endowed with one or both of the gonadotropin receptors. Consequently a putative regulator of ovarian clock gene rhythms could be the gonadotropins and in particular the midcyclic gonadotropin peak. Preliminary experiments in our laboratory using immature prepubertal rats injected with equine chorionic gonadotropin and human chorionic gonadotropin support this notion (Gräs, S., B. Georg, and J. Fahrenkrug, unpublished data).
The contemporary understanding of how rhythmicity is generated is that CLOCK/BMAL1 heterodimers initiate transcription of per and cry genes (38). PER1 and PER2 associate with CRY1 and CRY2 to form heterodimer complexes, also including Casein kinase. Casein kinase I
and I
phosphorylate the PERs and thereby affect both nuclear translocation and destabilization of the PER proteins by ubiquitination (39, 40). After nuclear translocation the PERs inhibit CLOCK/BMAL induction of the per and cry genes. In the mouse it has been shown that PER2 also promotes Bmal1 transcription through mechanism not yet fully understood to provide a positive drive to the system (41). Rev-Erb-
, itself a target for positive induction by CLOCK/BMAL1, inhibits induction of Bmal1 (42).
In the present study, we were able to examine the temporal intracellular movements of PER protein expression by immunohistochemistry using newly developed antisera. From a just detectable level, the PER proteins accumulated within the cytoplasm of the ovarian steroidogenic cells for several hours after which the proteins were transferred to the nuclei. The rise and fall in PER protein followed the cycle of mRNA expression with a lag of several hours. The nuclear accumulation of PER1 and PER2 protein coincided with declining levels for Per1 and Per2 mRNAs, which is consistent with the delayed feedback model of the oscillator, and the simultaneous PER2 mediated increase in the transcription of Bmal1. Our protein findings are somewhat different from those previously reported in the SCN (43). Thus, in the SCN both PER proteins accumulate in the nuclei of the cells at the same time and peak at ZT12, despite that Per1 mRNA and Per2 mRNA reach maximal values at different time points. Furthermore, immunohistochemical studies on the SCN have been unable to show specific cytoplasmic localization of PER immunoreactivity. In the ovary we found that the phase for the cyclic PER2 protein expression was delayed by approximately 4 h relative to PER1, which accords with the rise and falls of their respective mRNAs. Furthermore, we found that the nuclear accumulation of the PER proteins occurred later in the cycle than reported for SCN cells. The discrepancies could be tissue specific or could reflect differences in methodology. Our immunocytochemical data, however, clearly demonstrate the changes in intracellular localization and staining intensities of the two PER proteins and that the protein cycles of PER1 and PER2 were asynchronous.
Clock genes are also expressed in the male gonadal tissue of rodents, but the reported findings are unusual and somewhat conflicting. It appears that Per2 is not expressed in the testis (17, 18, 19, 20). Zylka et al. (20) initially demonstrated that Per1 expression oscillates in mouse testis, but later studies (17, 18, 19) found the levels in the testis to be constant over a 24-h period. It was shown that the expression of both Per1 and clock mRNA and protein is restricted to specific developmental stages of spermatogenesis (17, 19). Species differences may exist because rhythmic expression of Per1 has been reported in hamster testis (21).
Information on the circadian regulation of the female hypothalamo-pituitary-gonadal axis is accumulating (9, 44), but at present the role of an ovarian circadian clock is purely speculative. In addition to the ovarian sex hormones, a large number of other hormones and local regulatory substances are produced to ensure optimal coordination between the different aspects of the hypothalamo-pituitary-ovarian axis. The concept of daily rhythms in ovarian gene expression is new, and consequently the temporal aspect has only occasionally been examined. However, diurnal rhythms have been reported for some of the genes involved in ovarian steroid production during puberty and the menstrual cycle in humans (12, 45, 46). Diurnal rhythms have also been reported for the FSH-regulatory peptides, inhibin-A, inhibin-B, and follistatin (47).
Knockout mice for many of the clock genes, including Per1 and Per2, exist (4, 37). In female clock-mutant mice, estrous cyclicity and maintenance of pregnancy are disrupted, but the changes are subtle and ovulation takes place despite absence of the LH peak (16, 48). No major abnormalities in fertility have been reported for the other clock gene mutants, but systematic studies remain to be conducted. Furthermore, it is generally believed that the roles of the central and peripheral clocks are to adjust the organism in the best possible way to different environmental changes, and photic and feeding inputs are known to be important for entrainment of the systems (12, 37). Normally, experimental animals are kept under very well-defined environmental conditions. It is possible that almost normal fertility may exist under such conditions, but it is likely that fertility (and other systems) could be more severely disrupted if the animals were subjected to irregular environmental changes.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Disclosure statement: The authors have nothing to disclose.
First Published Online May 4, 2006
Abbreviations: B2m, ß2-Microglobulin; BMAL1, brain and muscle ARNT-like protein 1; SCN, suprachiasmatic nucleus; ZT, Zeitgeber time (time-point of exogenous scheduled entraining cycle).
Received March 8, 2006.
Accepted for publication April 24, 2006.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
. Mol Cell Biol 20:4888–4899 (CKI) and CKI in cultured cells. Mol Cell Biol 22:1693–1703 controls circadian transcription within the positive limb of the mammalian circadian oscillator. Cell 110:251–260[CrossRef][Medline]This article has been cited by other articles:
 |
|
 |
|
  W. P. Williams III, E. M. Gibson, C. Wang, S. Tjho, N. Khattar, G. E. Bentley, K. Tsutsui, and L. J. Kriegsfeld Proximate mechanisms driving circadian control of neuroendocrine function: Lessons from the young and old Integr. Comp. Biol., November 1, 2009; 49(5): 519 - 537. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Yoshikawa, M. Sellix, P. Pezuk, and M. Menaker Timing of the Ovarian Circadian Clock Is Regulated by Gonadotropins Endocrinology, September 1, 2009; 150(9): 4338 - 4347. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Bebas, C. P. Goodall, M. Majewska, A. Neumann, J. M. Giebultowicz, and P. E. Chappell Circadian clock and output genes are rhythmically expressed in extratesticular ducts and accessory organs of mice FASEB J, February 1, 2009; 23(2): 523 - 533. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Teboul, F. Guillaumond, A. Grechez-Cassiau, and F. Delaunay Minireview: The Nuclear Hormone Receptor Family Round the Clock Mol. Endocrinol., December 1, 2008; 22(12): 2573 - 2582. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. J. Nakamura, M. T. Sellix, M. Menaker, and G. D. Block Estrogen directly modulates circadian rhythms of PER2 expression in the uterus Am J Physiol Endocrinol Metab, November 1, 2008; 295(5): E1025 - E1031. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. Pilorz and S. Steinlechner Low reproductive success in Per1 and Per2 mutant mouse females due to accelerated ageing? Reproduction, April 1, 2008; 135(4): 559 - 568. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Nakao, S. Yasuo, A. Nishimura, T. Yamamura, T. Watanabe, T. Anraku, T. Okano, Y. Fukada, P. J. Sharp, S. Ebihara, et al. Circadian Clock Gene Regulation of Steroidogenic Acute Regulatory Protein Gene Expression in Preovulatory Ovarian Follicles Endocrinology, July 1, 2007; 148(7): 3031 - 3038. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P.-J. He, M. Hirata, N. Yamauchi, S. Hashimoto, and M.-a. Hattori The disruption of circadian clockwork in differentiating cells from rat reproductive tissues as identified by in vitro real-time monitoring system J. Endocrinol., June 1, 2007; 193(3): 413 - 420. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E Menkhorst, N Ezard, and L Selwood Induction of ovulation and natural oestrous cycling in the Stripe-faced Dunnart, Sminthopsis macroura Reproduction, February 1, 2007; 133(2): 495 - 502. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Gery and H. P. Koeffler The Role of Circadian Regulation in Cancer Cold Spring Harb Symp Quant Biol, January 1, 2007; 72(0): 459 - 464. [Abstract] [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 |
