| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Medicine (Y.M.T.), Townsville Hospital, Townsville, Queensland 4810; Department of Diabetes and Endocrinology (J.C., J.B.P.), Princess Alexandra Hospital, Woolloongabba, Queensland 4102; and School of Molecular and Microbial Biosciences (A.H.F., H.-C.M., N.L., L.B.H., A.D.C.), University of Sydney, Sydney, New South Wales 2006, Australia
Address all correspondence and requests for reprints to: A/Prof. Arthur D. Conigrave, Department of Biochemistry (G08), University of Sydney, NSW 2006, Australia. E-mail: a.conigrave{at}mmb.usyd.edu.au.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MaterialsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MaterialsResultsDiscussionReferences |
|---|
The human CaR is a transmembrane G-protein-coupled receptor with a 612-amino-acid extracellular domain, a transmembrane region composed of seven transmembrane helices and a 216-amino-acid C-terminal intracellular region (for review, see Ref. 5). The N-terminal extracellular domain is composed of two major subdomains: a bilobed, putative ligand-binding region known as the Venus Fly-Trap (VFT) domain composed of amino acid residues 36–513; and a cysteine-rich (Cys-rich) domain, residues 514–612, that is required for signal transduction (7).
More than thirty inactivating mutations causing familial hypocalciuric hypercalcemia and, very rarely, neonatal severe hyperparathyroidism have been described (reviews, Refs. 4 and 6). The majority of these natural mutations lie in the VFT domain, and relatively few have been found in the Cys-rich domain (6). In addition, more than twenty activating mutations of the receptor causing ADH have been described. Approximately 50% of the reported activating mutations lie within the proximal half of the VFT domain (i.e. before residue 256), and the remainder lie in the transmembrane region and tail (reviews, Refs. 4 and 6). With the exception of F612S, which lies at the interface between the extracellular region and first transmembrane helix (3), no activating mutations have been previously identified in the Cys-rich domain; and this region, with the exception of the conserved cysteines, is largely tolerant of amino-acid substitutions (7). In the current work, we have identified a second activating mutation (E604K) in the C-terminal region of the Cys-rich domain in a kindred with ADH. The C-terminal region of the Cys-rich domain may mediate tonic inhibition of the receptor.
The extracellular Ca2+-CaR complex activates several intracellular signaling pathways, including PI-phospholipase C (PI-PLC), MAPK, phospholipase D, and phospholipase A2 (reviews, Refs. 5 and 8). In addition, the CaR inhibits adenylyl cyclase (for review, see Ref. 5). In general, cellular analysis of CaR mutations in HEK-293 cells has focused on extracellular Ca2+-dependent activation of PI-PLC assayed either by the release of inositol phosphates or mobilization of intracellular Ca2+. It would be expected that an activating mutation of the extracellular head would affect not only PI-PLC/Ca2+ mobilization but also other signaling pathways. In the current study, we assessed the impact of E604K on CaR-stimulated MAPK activity as well as Ca2+ mobilization.
|
Subjects and Materials |
|---|
|
TopAbstractIntroductionSubjects and MaterialsResultsDiscussionReferences |
|---|
A 33-yr-old man (the proband II.3; Fig. 1
) was found to be hypocalcemic on family screening. A plasma total calcium level was 1.71 mM (normal range, 2.10–2.55 mM), and serum ionized Ca2+ was 0.77 mM (normal range, 1.20–1.40 mM). The proband had no history of seizures but had infrequent muscle cramps of long standing and paraesthesiae on repetitive fine finger movements. His serum PTH level (9 ng/liter; normal range, 10–65 ng/liter) was just below the lower limit of normal, and serum Mg2+ concentration was also low (0.66 mM; normal range, 0.74–1.03 mM). Before treatment, a 24-h urinary calcium excretion was 0.7 mmol/d (normal range, 2.5–7.5 mmol/d), and the urinary calcium to creatinine ratio was just above the lower limit of normal (0.04 mmol/mmol; normal range, 0.03–0.40 mmol/mmol). However, urinary calcium excretion (9.4 mmol/d) and the urinary calcium to creatinine ratio (0.46 mmol/mmol) were elevated after the introduction of calcium and calcitriol supplements.
|
) had suffered seizures shortly after birth, with low serum calcium (serum ionized Ca2+, 0.88 mM) and low serum intact PTH (0.4 pM; normal range, 1–7 pM) levels at diagnosis. He also exhibited hypercalciuria on calcitriol. The proband’s younger brother (II.4) had previously experienced unexplained childhood seizures and was also found to be hypocalcemic. A daughter (III.3) of II.4 was recently diagnosed, at the age of 8 months, with hypocalcemia (total plasma calcium concentration, 1.79 mM) in the context of nonfebrile seizures and a suppressed level of serum intact PTH (0.3 pM). A review of the proband’s other relatives led to the identification of asymptomatic hypocalcemia in his mother (I.1) and an older brother (II.2). Neither of these individuals had experienced childhood seizures. II.2 had an affected son (III.1) who also suffered from hypocalcemic seizures (serum ionized Ca2+, 0.89 mM) and a suppressed level of serum intact PTH (0.5 pM) at diagnosis in childhood. Serum inorganic phosphate levels were normal in all affected individuals, with the exception of the infant girl (III.3), whose level was 2.5 mM (pediatric normal range, 1.1–1.9 mM). The proband’s oldest brother (II.1) and sister (II.5) were asymptomatic, with normal serum calcium and intact PTH levels.
On the basis of the inheritance pattern, the variable severity of the clinical phenotype, and calcitriol-induced hypercalciuria, a provisional diagnosis of ADH was made. DNA samples from the proband and several other family members, including the unaffected older brother and younger sister, were submitted for mutational analysis of the CaR. Blood for genetic studies could not be obtained from the proband’s father or his affected nephew (III.1) and niece (III.3).
Materials and methods
PCR product amplification and sequencing of the CaR. Genomic DNA was obtained from peripheral blood leukocytes as described previously (9). Initially, all amino-acid-encoding exons (2, 3, 4, 5, 6, 7) of the CaR were sequenced in the proband, leading to the identification of a mutation G2182A (E604K) in exon 7. Subsequent DNA sequencing in other individuals was restricted to exon 7. The PCR primers used were described previously (10). A 50-µl standard PCR reaction mixture, including 20 pmol forward and reverse primers, 1.5 mM MgCl2, 2.5 mM deoxynucleotide triphosphate, and 2.5 U Amplitaq Gold (Roche Diagnostica, Castle Hill, NSW, Australia), was used. PCR was performed on a Corbett Research FTS-thermal cycler using the following protocol: initial denaturation of 10 min at 94 C; then 35 cycles of 94 C, 60 C, and 72 C for 1, 1, and 1.5 min, respectively; and a final extension of 5 min at 72 C. PCR products were purified using a High Pure PCR product purification kit (Roche) according to the manufacturer’s instructions. DNA sequencing was performed using Big Dye Terminator Mix (PE Applied Biosystems, Scoresby, Victoria, Australia) as per the manufacturer’s recommendations and run on a ABI PRISM 377 at the Australian Genome Research Facility (Queensland, Australia). The PCR screening primers were also used to sequence the PCR products, with the addition of the following four internal primers for exons 7.1 and 7.2: 5'-GGATCTTGGCCTCAAACACC-3'and 5'-CTTCTCCAGCTCCCTGTTCTTC-3' for exon 7.1, and 5'-GATGACCTTCTGCTTGCATCTG-3' and 5'-TAACCCAGCAAGAGCAGCAG-3' for exon 7.2. Sequencing was performed for both strands, and the mutation was subsequently confirmed using an independently derived PCR product.
Restriction digest analysis. The mutation G2182A eliminates a predicted Taq1 site in the human CaR DNA; and, in the affected exon 7.1 PCR fragment (881 bp), a single Taq1 restriction site (yielding fragments of 123 bp and 758 bp in the wild-type DNA) was lost. Restriction digest analysis on the 881-bp exon 7 fragment was performed using Taq1 (New England Biolabs, Inc., Beverly, MA; Genesearch, Queensland, Australia) according to the manufacturer’s instructions.
Site-directed mutagenesis. pcDNA3.1 containing the wild-type human CaR (GenBank accession no. U20759) was the generous gift from Dr. Mei Bai and Prof. Edward Brown (Endocrine-Hypertension Division, Brigham and Women’s Hospital, Boston, MA). Site-directed mutagenesis of this construct, to yield G2182A (E604K), was performed using the Stratagene(La Jolla, CA) Quick-change kit according to the manufacturer’s instructions. The sequences of the mutagenic oligonucleotides were: forward, 5'P-CATTGCCAAGGAGATCAAGTTTCTGTCGTGGACGG; and reverse, 5'P-CCGTCCACGACAGAAACTTGATCTCCTTGGCAATG (the position of the mutant base is indicated in bold italics). Successful mutation of the wild-type human CaR sequence to E604K was confirmed by DNA sequencing (Sydney University Prince Alfred Macromolecular Analysis Centre, Camperdown, NSW, Australia).
Cell culture and transfection. HEK-293 cells were cultured in DMEM containing 10% fetal bovine serum. When the cells reached 70–95% confluence, they were transfected with the human wild-type or E604K mutant CaR expression constructs (in equal amounts) and reporter plasmids using LipofectAMINE Plus or LipofectAMINE 2000 according to the manufacturer’s instructions (Invitrogen, Mt. Waverly, Victoria, Australia).
Transfer of apoaequorin gene into pcDNA3.1. The vector pMAQ2, containing apoaequorin cDNA, was obtained from Molecular Probes, Inc. (Eugene, OR). The cDNA was excised by digestion with XbaI and PstI and transferred, via pBluescript (Stratagene), into NheI-KpnI cut pcDNA3.1 (+) (Invitrogen), generating pcAEQ for expression in mammalian cells.
Aequorin luminescence assay for changes in cytoplasmic free Ca2+ concentration. All aequorin assays were performed 2 d after transfection. The protocol developed was based on the method of Ungrin et al. (11). Transfected HEK-293 cells in 96-well plates were charged for 1 h at 37 C in Ca2+-free physiological saline solution (125 mM NaCl, 4 mM KCl, 20 mM HEPES, 1 g/liter D-glucose, 1 mM NaH2PO4, and 1 mM MgCl2) containing 5 µM coelenterazine (cp form, Molecular Probes, Inc.) to form holoaequorin. The charging solution was then replaced with physiological saline solution, and the plates were taken to a Victor 2 Multilabel counter (Wallac, Inc., Turku, Finland, and Perkin-Elmer Corp., Sydney, NSW, Australia) for luminescence measurements. Transfected cells were exposed to various extracellular Ca2+ concentrations to generate dose-response curves; and then ionomycin, to identify a residual CaR-independent response. Wells were tested individually, injecting increasing volumes of CaCl2 (stock concentrations, 2–50 mM) in physiological saline using peristaltic pump-1. Light emission was recorded (at 0.5-sec intervals for 30 sec), after which, ionomycin was injected from pump-2 (final concentration, 2 µM). The light emission was then recorded for an additional 20 sec (recorded as forty 0.5-sec integrations). The luminescence responses obtained at various extracellular Ca2+ (Ca2+o) concentrations were expressed with respect to the maximum response obtained with maximal Ca2+ plus ionomycin.
MAPK assay. HEK-293 cells that had reached 70–80% confluence were transfected with the Path Detect MAPK trans-reporting system (Stratagene) according to the manufacturer’s instructions. Transfection efficiency was monitored routinely by cotransfection with a ß-galactosidase reporter construct. After 24 h, the wells were exposed to various concentrations of Ca2+ (0.5–15 mM), and the cells were cultured for a further 24 h before lysis and recovery of the luciferase-containing supernatant. Twenty-microliter aliquots of supernatant were then transferred to the wells of a 96-well plate, and the plates were read in a Wallac, Inc. multilabel counter using luminescence mode, before and after injection of 50 µl D-luciferin-ATP reagent, according to the manufacturer’s instructions (Stratagene). Supernatants were also routinely assayed for ß-galactosidase as described by Sambrook et al. (12). No extracellular Ca2+-dependent luciferase expression was observed when the CaR genes or either of the MAPK-luciferase reporter genes were omitted from the transfections. For each paired experiment, the data were converted to fractional responses of the response obtained in HEK-293 cells transfected with the wild-type CaR and then exposed to 15 mM Ca2+.
Intact cell enzyme-linked immunoassay to determine CaR surface expression. The cell surface expression of wild-type and the E604K mutant CaR was measured using untransfected and plasmid-only transfected cells as controls as described in (13). Briefly, after 48 h of transfection in 6-well plates, HEK-293 cells (American Type Culture Collection, Manassas, VA) were incubated in DMEM for 1 h at 4 C with 1 µg/ml ADD monoclonal antibody against the CaR peptide residues 214–235 (14), the kind gift from Dr. Edward Nemeth (NPS Pharmaceuticals, Inc., Salt Lake City, UT). The cells were then detached from the plates by washing with ice-cold PBS and incubated with a 1:500 dilution of a peroxidase-conjugated sheep antimouse IgG (Chemicon, Boronia, Victoria, Australia). After additional washing and incubation with the peroxidase substrate, the cells were removed by centrifugation, and the supernatants were transferred to a 96-well microtiter plate. To stop the reaction, sulfuric acid was added to a final concentration of 0.7 M, and the absorbances were read at 490 nm. Each experimental data point was the mean of triplicate determinations. As previously described, the assay exhibited a nonzero background, even in the absence of the CaR (13).
Data analysis. The Ca2+ mobilization and MAPK data were expressed as concentration-response curves and fitted to the following form of the Hill equation: R = b + (a-b)·Cn/(en+Cn), where R = fractional response, a = maximum fractional response, b = basal fractional response, C = extracellular Ca2+ concentration (in mM), e = EC50 (the concentration of Ca2+ that induced a half-maximal response), and n = Hill coefficient. Estimates of the curve-fitting parameters and their SEs were obtained using MacCurveFit 1.5 for Macintosh. Differences between the EC50 values and maximum responses obtained from concentration-response curves for wild-type and E604K mutant forms of the CaR were tested for statistical significance using the F-test as described in (15). Other data are routinely expressed as means ± SEM (number of experiments).
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MaterialsResultsDiscussionReferences |
|---|
Initial DNA sequence analysis of the proband identified a heterozygous single-base mutation, G2182A (GenBank accession no. U20759) in exon 7. This mutation results in a nonconservative amino acid substitution E604K. The amino acid position 604 lies at the C terminus of the Cys-rich domain of the extracellular head. Thus, the mutation lies 8–9 residues to the N-terminal side of the interface between the extracellular head and first transmembrane helix. Family screening demonstrated that the affected phenotype segregated with the E604K mutation. This mutation was found in one allele from all five affected family members that were screened. However, the mutation was not present in either CaR allele from the proband’s unaffected older brother (II.1) or his unaffected younger sister (II.5; Fig. 1).
Taq1 restriction digest analysis was also performed on the exon 7.1 PCR product that included the mutation site (Fig. 2). A predicted 2-fragment pattern (758 bp and 123 bp) was observed using DNA from the two unaffected family members (lanes B and C; Fig. 2) together with 50 normal controls (not shown). However, all affected family members tested (lanes D–H) exhibited a complex three-band Taq1 restriction digest pattern (881 bp, 758 bp, and 123 bp). This pattern was consistent with failure of digestion of the mutant allele together with normal digestion of the wild-type allele (yielding the 758-bp and 123-bp fragments) within the heterozygous DNA samples.
|
Control aequorin experiments were performed using HEK-293 cells that had been transiently transfected with known activating mutations, including L125P (16) and C131S (17), or inactivating mutations, including S147A (18) and E297K (19). The estimates of the maximal responses and EC50 values obtained were comparable with those reported previously (Fig. 3 and Table 1).
|
|
). Cells that had been transfected with the apoaequorin gene were charged with coelenterazine-cp (5 µM) for 1 h at 37 C before stimulation. As expected, both the wild-type CaR and E604K mutant exhibited extracellular Ca2+-dependent activation of Ca2+ mobilization in the aequorin-luminescence assay, and E604K exhibited a left-shift in the Ca2+-dependent concentration curve. For intracellular Ca2+ mobilization, the EC50 values obtained from 10 paired experiments were, respectively: wild-type, 4.4 mM; E604K, 3.6 mM (F[1,76] = 17.8; P < 0.01). There was also a significant reduction in the maximum response (Fig. 4; F[1,76] = 226; P < 0.01).
|
HEK-293 cells were transiently transfected with the wild-type CaR or E604K mutant and the MAPK-luciferase detection system (Fig. 5) and then exposed to various extracellular concentrations of Ca2+. The data obtained for 0.5–12 mM Ca2+ were then expressed as a fraction of the response obtained for the wild-type CaR at 15 mM Ca2+ for curve-fitting analysis. The E604K mutant exhibited a left-shift in the Ca2+-dependent concentration curve. For MAPK activation, the EC50 values for extracellular Ca2+ obtained from four paired experiments were, respectively: wild-type, 6.6 mM; E604K, 3.9 mM (F[1,27] = 50.1; P < 0.01). The maximum levels of MAPK activity for the E604K mutant (1.18 ± 0.11) and the wild-type control (1.12 ± 0.22) were comparable; however, there was a small, positive baseline shift in the E604K mutant (F[1,27] = 5.9; P < 0.05) even at the lowest Ca2+ concentrations tested (Fig. 5). Similar data were obtained in two further experiments using separate preparations of wild-type and E604K DNA.
|
The effect of the E604K mutation on surface expression was assessed in four experiments. Transient transfection with the wild-type CaR yielded an ELISA signal that was 3- to 4-fold higher than the untransfected or plasmid-only controls, comparable with that described previously (13). Relative to the wild-type CaR, E604K surface expression was 1.04 ± 0.16 (n = 4). These data suggest that the effect of the E604K missense mutation on surface expression is minimal or nonexistent.
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MaterialsResultsDiscussionReferences |
|---|
) cosegregated with the affected phenotype, substituting a lysine for a glutamate in the distal extracellular head at amino acid position 604. Furthermore, functional analysis of extracellular Ca2+-activated Ca2+ mobilization (Fig. 4) and MAPK (Fig. 5) in HEK-293 cells provided support for the conclusion that the E604K mutation enhances the extracellular Ca2+ sensitivity of the CaR in vivo and is responsible for ADH in this kindred. Finally, the possibility that E604K is a common polymorphism of the CaR gene was excluded by Taq1 restriction digest analysis of exon 7.1 PCR products generated using DNA from 50 normal individuals. Of the 24 previously reported activating mutations of the CaR, 14 are located in the extracellular domain, and 13 of these are located in the proximal extracellular head, inside the VFT domain (for review, see Ref. 6). This domain shares homology not only with metabotropic glutamate receptors but also with nutrient-sensing bacterial periplasmic binding proteins (18). Similarly, the majority of inactivating mutations associated with familial hypocalciuric hypercalcemia have been identified in the VFT domain (6). This region has been implicated in ligand binding and receptor dimerization by analysis of mutants and chimeric receptors (17, 20).
The E604K-activating mutation reported in this study, unusually, lies in the Cys-rich domain of the extracellular head, which seems to be required for the coupling of ligand-binding to the activation of intracellular signaling pathways (7). Another activating mutation, F612S, lies close-by, at the interface between the Cys-rich domain and the first transmembrane helix (3, 21). The Cys-rich domain has been implicated previously in the transmission of the activating signal from the ligand-bound VFT domain to the transmembrane region (7). The finding that Ca2+ mobilization and MAPK were both affected by the E604K mutation supports the view that the Cys-rich domain is not, however, involved in the selection of specific signaling pathways. Additional work on other signaling pathways, including phospholipase A2 and adenylyl cyclase, and taking advantage of specific pharmacological agents, such as the PI-PLC inhibitor U73122 and pertussis toxin, which inhibit distinct components of CaR-dependent MAPK (22), are required to fully clarify this issue.
Surface expression of the CaR is increased by some activating mutations, e.g. a large in-frame deletion of the CaR tail (23), but not others, e.g. F612S (21). The result for E604K is similar to that for the nearby mutation F612S (although the assay used in the current study could not have reliably detected a modest increase in surface expression, e.g. of the order of 5–10%). It seems probable, therefore, that the activating effect of E604K, like that of F612S, arises primarily from a sensitizing effect on signal transmission, leading us to speculate that the C terminus of the Cys-rich domain normally suppresses receptor activation at low extracellular Ca2+ concentrations.
The E604K mutation was associated with marked hypocalcemia in all affected individuals (ionized Ca2+, 0.7–0.9 mM; normal range, 1.2–1.4 mM). However, there was considerable variation in the clinical histories of affected individuals in this family, particularly with respect to the occurrence of seizures in childhood. Variation in clinical presentation in ADH, ranging from asymptomatic hypocalcemia to childhood hypocalcemic seizures, seems to be typical for the disorder and is currently unexplained (for review, see Ref. 4).
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Abbreviations: ADH, Autosomal dominant hypocalcemia; CaR, calcium-sensing receptor; Cys-rich, cysteine-rich; PI-PLC, phosphoinositide-specific phospholipase C; VFT, Venus Fly-Trap.
Received January 23, 2002.
Accepted October 23, 2002.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MaterialsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  S. Pidasheva, M. Grant, L. Canaff, O. Ercan, U. Kumar, and G. N. Hendy Calcium-sensing receptor dimerizes in the endoplasmic reticulum: biochemical and biophysical characterization of CASR mutants retained intracellularly Hum. Mol. Genet., July 15, 2006; 15(14): 2200 - 2209. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. D. Mittelman, G. N. Hendy, R. A. Fefferman, L. Canaff, I. Mosesova, D. E. C. Cole, L. Burkett, and M. E. Geffner A Hypocalcemic Child with a Novel Activating Mutation of the Calcium-Sensing Receptor Gene: Successful Treatment with Recombinant Human Parathyroid Hormone J. Clin. Endocrinol. Metab., July 1, 2006; 91(7): 2474 - 2479. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H.-C. Mun, E. L. Culverston, A. H. Franks, C. A. Collyer, R. J. Clifton-Bligh, and A. D. Conigrave A Double Mutation in the Extracellular Ca2+-sensing Receptor's Venus Flytrap Domain That Selectively Disables L-Amino Acid Sensing J. Biol. Chem., August 12, 2005; 280(32): 29067 - 29072. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Wystrychowski, S. Pidasheva, L. Canaff, J. Chudek, F. Kokot, A. Wiecek, and G. N. Hendy Functional Characterization of Calcium-Sensing Receptor Codon 227 Mutations Presenting as Either Familial (Benign) Hypocalciuric Hypercalcemia or Neonatal Hyperparathyroidism J. Clin. Endocrinol. Metab., February 1, 2005; 90(2): 864 - 870. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H.-C. Mun, A. H. Franks, E. L. Culverston, K. Krapcho, E. F. Nemeth, and A. D. Conigrave The Venus Fly Trap Domain of the Extracellular Ca2+-sensing Receptor Is Required for L-Amino Acid Sensing J. Biol. Chem., December 10, 2004; 279(50): 51739 - 51744. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. N. Hendy, C. Minutti, L. Canaff, S. Pidasheva, B. Yang, Z. Nouhi, D. Zimmerman, C. Wei, and D. E. C. Cole Recurrent Familial Hypocalcemia Due to Germline Mosaicism for an Activating Mutation of the Calcium-Sensing Receptor Gene J. Clin. Endocrinol. Metab., August 1, 2003; 88(8): 3674 - 3681. [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 |
