| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Studies |
Departments of Medicine (S.P., M.F., R.E.W., N.H.S., S.R.), Pediatrics (S.R.); J. P. Kennedy, Jr., Mental Retardation Research Center (S.R.); and Clinical Endocrinology Laboratory (N.H.S., S.R.), University of Chicago, Chicago, Illinois 60637-1470; and Center for Adolescent Medicine and Foundation Growth Puberty Adolescence (U.E.), Zurich, Switzerland
Address all correspondence and requests for reprints to: Dr. Samuel Refetoff, University of Chicago, 5841 South Maryland Avenue, Chicago, Illinois 60637. E-mail: refetoff{at}medicine.bsd.uchicago.edu
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Although earlier studies based on electrophoretic, chemical, and immunological properties have traced the TT4 abnormality to a modified human serum albumin (HSA) molecule (2, 3, 9, 10, 11), the precise defect was identified only in 1994, independently by Petersen et al. (12) and by Sunthornthepvarakul et al. (13). A missense mutation in the HSA gene that results in the replacement of the normal arginine 218 with a histidine (R218H), produces a HSA with 10- to 15-fold higher affinity for T4 than the wild-type (WT) molecule, and a 5-fold increase in affinity for T3 (13, 14). The identical mutation has now been identified in 22 unrelated Caucasian families, mostly of Portuguese or Hispanic ancestry (13, 14) (Pannain, S., and S. Refetoff, personal observation).
The first report of FDH in an Asian family was published in 1997 by Wada et al. (15). These researchers identified a Japanese kindred with a phenotype characterized by extremely high serum TT4 levels due to a missense mutation in the same codon that replaced the normal arginine 218 with a proline (R218P). T4 binding studies suggested that the variant HSA molecule has an 83-fold increase in affinity for T4. This finding appeared to explain the 14- to 27-fold increase in the mean normal serum TT4 concentrations in subjects harboring the R218P HSA compared to the only 1.5- to 2.5-fold increase in subjects harboring HSA R218H. Both variant HSAs have a lesser increase in affinity for T3, although only subjects with HSA R218P have consistently high serum total T3 (TT3) levels (3, 15).
A third mutation in the HSA gene, causing alterations in serum iodothyronine concentrations, was identified in a Thai kindred in 1998 by Sunthornthepvarakul et al. (16). This missense mutation replaces the normal leucine 66 with a proline (L66P). This variant HSA differed from those of the other two mutations in that the molecule has a higher affinity for T3 than for T4. The 40-fold higher affinity constant for T3 but only 1.5-fold increase in the binding affinity for T4 are responsible for the predominant increase in serum TT3 concentration. The term FDH-T3 was suggested to describe this phenotype.
In this communication we report findings in a second family with HSA R218P and the first Caucasian family with no Asian ancestry. Serum TT4 concentrations were, on the average, 14-fold above the normal mean value and similar to those reported by Wada et al. (15). This very high TT4 concentration was not artifactual, as it could be demonstrated in serum extracts. Unexpectedly, the affinity constant for T4 was identical to that of HSA R218H when determined by Scatchard analysis at ligand saturation. Despite the apparent identical T4 binding affinities of HSA R218P and HSA R218H, the relative amount of dialyzable T4 at equivalent TT4 concentrations was lower in HSA R218P than in HSA R218H.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The probands (III-1 and III-2; see Fig. 1
) were 3.8 and 2.4 yr of age when
referred to one of us (U.E.) because of short stature and
hyperactivity. Physical examination was unremarkable, except for height
of 96.5 cm (-2.1SD) for III-1 and 84.0 cm (-2.2
SD) for III-2, bird-like facies, and hyperactivity.
Although their thyroid glands were not enlarged, the diagnosis of
resistance to thyroid hormone was considered. Indeed, serum
TT3 and FT4 concentrations
were high, and TSH values were within the normal range. However, the
interpretation was complicated by the inconsistency of
TT3 and FT4 results
reported for III-1 from two different laboratories as 2.25 and 8.3
nmol/L for TT3 (normal ranges, 0.8–2.46 and
1.7–3.7 nmol/L) and 43.3 and more than 85 pmol/L for
FT4 (normal ranges, 10.3–29.7 and 9.0–24.5
pmol/L), with corresponding TSH values of 3.6 and 4.2 mU/L. The
FT4 and TSH concentrations in serum from III-2
were 35.8 pmol/L (normal range, 10.3–29.7 pmol/L) and 1.1 mU/L,
respectively. In a follow-up study at 4 yr of age, the bone age of
III-1 was 3.25 yr. At ages 6.8 and 4.7 yr, the children were found to
have attention deficit disorder, with hyperactivity and apparently
reduced intelligence.
|
Blood samples were obtained from both children (III-1 and III-2), their parents (II-1 and II-2), and their maternal grandparents (I-1 and I-2). The family, of Swiss origin and without Asian ancestry, denied consanguinity. The grandfather has type I diabetes mellitus with complications including retinopathy, neuropathy, and nephropathy and had had two myocardial infarctions. He was taking seven different medications in addition to insulin.
Tests of thyroid function
Serum TT4, TT3, TrT3, and thyroglobulin (TG) concentrations were measured by RIAs. Serum samples from affected individuals with very high total T4 concentrations were diluted 10- and 20-fold with a zero T4 standard or with a T4-deficient serum obtained from an athyreotic patient after withdrawal of L-T4 replacement therapy for the routine evaluation of thyroid cancer. TSH was measured by a third generation chemiluminescence immunoassay (Nichols Institute, Inc./Corning, Inc., San Juan Capistrano, CA)
Because of the unusually high concentration of T4 in serum of the affected subjects, cross-reactivity of this iodothyronine in the T3 and rT3 RIAs was determined by measurement of authentic, high pressure liquid chromatography-purified T4 in the same assays. The cross-reactivity in both assay was 0.03%.
To establish the authenticity of the high T4 concentrations in the serum of affected subjects, measurements were carried out at different dilutions by two assays, a RIA (DPC double antibody, Diagnostic Products, Los Angeles, CA) and an electrochemical immunoradiometric assay using the automated system Elecsys 2010 (Roche Molecular Biochemicals, Indianapolis, IN). To exclude the possibility of interference in the T4 assays by serum proteins, the latter were removed by precipitation with 2 vol ethanol. The ethanol extract was evaporated to dryness, the residue was reconstituted in T4-deficient serum, and T4 was measured at the same dilutions as in the unextracted serum. Results were corrected for the recovery of T4 in the extract determined by the addition of a tracer (10,000 cpm/mL serum), [125I]T4 (SA, 969 Ci/mmol; New England Nuclear Life Science Products, Boston, MA), to serum before extraction. The T4 recovery in extracts ranged from 68–71% in serum from the affected subjects as well as control serum, with normal and high T4 concentrations due to HSA R218H or T4-binding globulin (TBG) excess.
The FT4 concentration was measured by equilibrium dialysis at 37 C, using a kit purchased from Nichols Institute, Inc./Corning, Inc. The dialysate buffer was designed to approximate the composition of a protein-free ultrafiltrate of normal human serum and was buffered at pH 7.35 with 4-(2-hydroxyethyl)-1-piperazine ethane sulfonate, as described in detail previously (17). To determine the effect of the high TT4 concentration on FT4 in serum of the affected subjects relative to that in normal serum and serum of subjects with the common form of FDH (HSA R218H variant), the TT4 concentration in such sera was adjusted to 1,500 nmol/L, which is the level found in family members with HSA R218P. FT4 was also determined in these sera after the addition of higher amounts of T4, to total concentrations of 3,200, 6,400, 32,000, 128,700, and 320,000 nmol/L. The latter is a full saturating concentration for HSAs. The dialysates were diluted appropriately, from 0- to 10,000-fold, to keep the concentration of T4 in the measurable region of the standard curve.
Tests to assess excessive T4 binding to HSA
The proportion of T4 bound to HSA was determined by addition of approximately 10,000 cpm [125I] T4 (New England Nuclear Life Science Products) to 1 µL serum diluted in 50 µL phosphate-buffered saline, pH 7.4, containing 2 mmol/L ethylenediamine tetraacetate. After allowing 1-h equilibration at room temperature, 30 µL goat anti-HSA (DiaSorin, Inc., Stillwater, MN) were added. Six hours later, the precipitate was separated by overlaying the suspension on 200 µL 1 mol/L sucrose dissolved in phosphate-buffered saline, placed in a 400-µL microfuge, and centrifuged for 3 min at 10,000 x g. [125I]T4 associated with the precipitated HSA was determined by counting the content of the microfuge tip, cut off with a razor blade after rapid freezing its contents.
Isoelectric focusing (IEF) was performed on Phastgel (Pharmacia Biotech-LKB, Piscataway, NJ), pH 4–6.5, overlaid for 30 min with 8-fold diluted ampholine (Pharmalyte), pH 4.0–4.9. Undiluted serum (0.1 µL) equilibrated with 1000 cpm [125I]T4 was applied on the Phastgel and was subjected to 400 volt-hours electrophoresis on PhastSystem. At the end of the run, the gel was dried, and the locations of [125I]T4-bound proteins were determined by autoradiography. Sera from unaffected individuals were equilibrated with 1.6 µmol/L T4 (Sigma, St. Louis, MO) in addition to the [125I]T4 to saturate the T4-binding sites on TBG and transthyretin, increasing the proportion of the tracer bound to the HSA.
Measurement of the affinities (Ka) of HSAs for T3 and T4
The method for determination of iodothyronine binding to serum proteins has been described in detail (18). Briefly, whole serum was diluted 10- or 40-fold with 75 mmol/L barbital buffer (pH 8.6) to prevent T4 binding to transthyretin as well as in 0.2 mol/L glycine-0.12 mol/L sodium acetate buffer (pH 8.6). [125I]T4 and [125I]T3 (SA, 2200 Ci/mmol; New England Nuclear Life Science Products) binding to the high affinity, low capacity sites on TBG was abolished by saturation with unlabeled T4 and T3 (Sigma), respectively, at 200-fold the maximal binding capacity, which was the lowest concentration of these iodothyronines (50 µmol/L) used in the analysis of binding to HSA. Ka values were derived by the method of Scatchard from the slopes of the best-fit lines, and the coefficients of correlation (r) were calculated.
Serum samples from two normal individuals, two subjects with FDH due to HSA R218H but belonging to different families, and a subject with FDH-T3 caused by HSA L66P were analyzed in parallel using the same reagents and methods.
DNA extraction and sequencing
DNA was extracted from white blood cells. The DNA of the proband’s mother (II-1) served as a template for the amplification of exon 7 of the HSA gene by PCR using the following oligonucleotide primer sequences: 5'- TCTGTATGTCCATTTTGAATTTTC-3' (sense) and 5'-GATACCAAACGCATCCATTCTA-3' (antisense). Reactions were carried out in a 100-µL volume containing 100 ng DNA, 100 pmol of each primer, 200 µmol/L of each deoxy-NTP, 1.5 mmol/L MgCl2, 50 mmol/L KCl, 10 mmol/L Tris-HCl (pH 9.0), 0.1% Triton X-100, and 2 U Taq polymerase. Initial denaturation was at 94 C for 5 min, followed by 35 cycles of 94 C for 1 min, 55 C for 1 min, and 72 C for 1 min and a final extension step at 72 C for 1 min. The PCR products were then subjected to sequencing using the Applied Biosystems, Inc. d-rhodamine terminator cycle DNA sequencing kit (Perkin-Elmer Corp., Foster City, CA) at the following conditions: initial denaturing cycle at 94 C for 4 min, followed by 25 cycles of 96 C for 10 s, 50 C for 5 s, and 60 C for 4 min. The reaction products were purified and analyzed on an Applied Biosystems, Inc. Prism 377 DNA sequencer (Perkin-Elmer Corp.).
Confirmation of the HSA mutation and genotyping
To confirm the presence of the HSA mutation R218P, we performed the endonuclease digestion described by Wada et al. (15). A mismatched sense oligonucleotide primer that creates a restriction site for AvaII in the presence of the mutation (C as the second nucleotide of codon 218) was used to amplify a 153-bp fragment from unaffected and affected individuals of the family. Subsequent digestion with AvaII (New England Biolabs, Inc., Beverly, CA) would reduce the length of the DNA fragment from the affected individuals harboring the mutant codon 218 (CCC) from 153 to 122 and 31 bp, but would leave the amplified DNA fragment of unaffected individuals (CGC) intact. The products of digestion were submitted to electrophoresis on 10% acrylamide gel, and bands were visualized with ethidium bromide under UV light.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Results of thyroid function tests of individuals belonging to the
FDH-30 family are presented in Fig. 1
. The probands (III-1 and III-2)
and their mother (II-1) have predominately high serum
TT4 and TrT3 concentrations
with a lesser increase in TT3. Serum
TT4 values ranged from 1493–1544 nmol/L, on the
average, 14-fold the normal mean value. TrT3
levels ranged from 2.40–2.72 nmol/L, and TT3
from 3.28–4.79 nmol/L, on the average, 7- and 2-fold the respective
mean normal values. The contribution of T4 by
cross-reactivity in the corresponding assays was 0.45 nmol/L, or 29%
and 12% of the TT3 and
TrT3 values, respectively. Serum TSH and TG
concentrations were normal, and there were no detectable TG or thyroid
peroxidase antibodies. The serum TT4 level of the
maternal grandfather (I-1) was also remarkably high at 695 nmol/L,
albeit it was half the level in the other three affected family
members. This was attributed in part to concomitant hypothyroidism
(TSH, 6.5 mU/L) as well as his illness and possibly the effect of some
of the drugs he was taking. Test values of the probands’ father (II-2)
and maternal grandmother (I-2) were in the normal range, except for a
slight increase in TrT3 in the grandmother.
The proportion of T4 bound to HSA was markedly increased in all affected individuals, including I-1, ranging from 34–44% compared to 3.4% and 2.0% for the two unaffected family members. IEF showed that TBG was fully saturated by the endogenous T4, and as in FDH due to R218H (3), the labeled T4 was associated with a band of more acidic isoelectric point than normal HSA (data not shown).
Identification of the HSA gene mutation
Direct sequencing of exon 7 of the HSA gene of subject II-1 revealed a G to C transversion in the second nucleotide of codon 218 in one of the two alleles (result not shown). This nucleotide substitution predicts the replacement of the arginine (CGC) at codon 218 of the WT HSA with a proline (CCC).
The creation of a restriction site for AvaII in the presence
of the mutant C in codon 218 was used to confirm the mutation in
subject II-1 and to genotype all members of the family. As shown in
Fig. 2, the four members of the family
that expressed the FDH phenotype were heterozygous for R218P, whereas
the two subjects with normal serum levels of all three
iodothyronines were homozygous for the WT R218 HSA.
|
Serum samples from subjects II-1 and III-1 were used to measure
the T4 and T3 binding
affinities of HSA R218P, which were compared, in the same assays, to
those in sera from normal unrelated individuals as well as individuals
harboring the HSA variants R218H and L66P. The results are shown in
Fig. 3 and summarized in Table 1. Surprisingly, measurement of the
T4 binding affinity of HSA R218P did not reveal
the expected difference with that of HSA R218H given the 8-fold higher
TT4 concentration in serum. As shown in Fig. 3, the Ka values for both mutant HSAs were similar,
5.4 x 106
mol/L-1 for HSA R218P and
5.2 x 106
mol/L-1 for HSA R218H.
Mean Ka values from three separate determinations
were 5.4 x 106
mol/L-1 for HSA R218P,
6.4 x 106
mol/L-1 for HSA R218H, and
5.1 x 105
mol/L-1 for the WT HSA
R218. Corresponding values measured by the same method but using
glycine acetate, rather than barbital, buffer were 4.6 x
106
mol/L-1 for HSA R218P,
5.8 x 106
mol/L-1 for HSA R218H, and
4.1 x 105
mol/L-1 for WT HSA
R218.
|
|
, the
Ka of 4.7 x 104 for
HSA R218P was only slightly higher than that for WT HSA of 3.8 x
104 despite a 2.3-fold higher serum
TT3 concentration. A much higher
Ka, 1.1 x 106, was
found in the subject with FDH-T3, who had a serum
TT3 concentration 3.3-fold that of subjects with
the WT HSA. This Ka value is similar to that
reported previously of 1.5 x 106 (16). The amounts of the variant HSAs in serum were not significantly different and represented about half the total HSA as determined from the intercepts of the regression lines at the abscissa. The total capacity of HSA was also determined with saturating concentrations of T4 by immunoprecipitation of the added [125I]T4 tracer with anti-HSA serum. In samples with WT HSA, HSA R218H, and HSA R218P, the HSA-bound T4 levels were 53%, 48%, and 56%, respectively.
Authentication of the T4 values in serum of subjects with HSA R218P and determination of the effect of T4 concentration on the FT4 in WT and variant HSAs
Because of the marked difference in serum TT4 concentrations in subjects with HSA R218P (14-fold the mean normal) compared to those with R218H (2-fold the mean normal) in the absence of significant differences in the measured Ka values for T4, the validity of the T4 measurement was investigated. To eliminate possible interference from the mutant HSA in the T4 assay, as observed previously for HSA L66P (16), serum proteins were precipitated, and T4 was extracted and measured in two different assays as described in Materials and Methods. The concentrations of T4 before and after extraction and reconstitution were, respectively, 95 and 95 nmol/L for serum with normal HSA, 166 and 174 nmol/L for serum with high T4 due to HSA R218H, and 1570 and 1441 nmol/L for serum from subject II-1 with HSA R218P. Measurements made by RIA and those made by the electrochemical immunoradiometric assay (see Materials and Methods) were similar.
The FT4 concentrations at different levels of
T4 added to serum samples containing the mutant
HSAs was determined by direct measurement after equilibrium dialysis at
37 C. At the T4 concentration of 1,500 nmol/L,
found in sera from subjects with HSA R218P, the
FT4 levels were 2,008 and 1,931 pmol/L in two
sera with WT HSA, 412 and 451 pmol/L in two sera with HSA R218H, and 40
pmol/L in serum with HSA R218P. Addition of 6,400 nmol/L
TT4 magnified the excess
FT4 concentration for the WT HSA compared to that
for HSA R218P to 130-fold (Fig. 4A). The
11-fold difference between the FT4 of HSA R218H
and R218P at a TT4 level of 1500 nmol/L narrowed
at the higher amount of added TT4 and was
completely obliterated at the HSA saturating concentration of 320,000
nmol T4/L (Fig. 4B).
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Sequencing exon 7 of the HSA gene showed that the mutation of codon 218 was identical to that identified by Wada et al. (15) in a Japanese family. Additionally, the phenotypes were virtually identical in terms of the magnitudes of total T4, rT3, and T3 elevations. In fact, the magnitude of the TT4 elevation was the highest observed under any physiological or pathological condition (21). The small stature, delayed bone age, hyperactivity, reduced intelligence, and facial features in the probands of family FDH-30 could not be explained on the basis of the HSA R218P mutation, and no similar clinical findings have been reported in the affected members of the Japanese family or in the other 32 kindreds with FDH due to HSA R218H that we have evaluated.
The mutation that involves the second nucleotide of codon 218 (CGC) is the same guanine that is substituted by an adenine (CAC) in the common type of FDH (R218H) found in many Caucasian families (13, 14) (Pannain, S., and S. Refetoff, personal observation). However, in contrast to this typical hot spot mutation, the mutation described herein that produces HSA R218P is a G to C transversion (CCC), rather than a G to A transition. The latter occurs with higher frequency due to methylation followed by spontaneous deamination that takes place in CpG dinucleotide hot spots (22). Even if not a hot spot mutation, its occurrence does not appear to be limited to or more prevalent in the Japanese population, although a distant common ancestor to the Swiss and Japanese families cannot be excluded.
Based on the remarkable elevation of the serum TT4 concentration, it was anticipated that the affinity of the mutant HSA R218P would be higher than that of HSA R218H by at least 1 order of magnitude. As a matter of fact, it can be estimated that to maintain a normal free T4 of 19 pmol/L in the presence of 1,500 nmol/L TT4 and 300 µmol/L (half the total HSA concentration) of the mutant HSA R218P, the latter should have a Ka of 1.1 x 108 mol/L-1, or 16-fold that observed in subjects with HSA R218H. It was, therefore, surprising that no significant differences were detected in the Ka for T4. This result was obtained in three independent determinations, using serum samples from two affected members of the family, and in assays performed along with WT and R218H HSAs. Furthermore, our Ka values for the latter two HSAs were similar to those previously reported from our laboratory as well as those of Bhagavan and using three different methods. Closer examination of the results reported by Wada (15) showed that the 80-fold increase in the Ka for T4 in serum from the Japanese subject with HSA R218P was mainly due to the low Ka obtained for the WT HSA. The value of 1.1 x 105 mol/L-1 is 4- to 5-fold lower than that reported by other laboratories (13, 14, 16). In contrast, the absolute Ka value of 9.1 x 106 mol/L-1 for HSA R218P is only 2-fold higher that that reported herein, again showing minimal difference in the apparent affinity for T4 between the two mutant HSAs.
The inconsistency of TT4 and Ka values for the HSA R218P compared to HSA R218H was investigated further by challenging the hormone and affinity measurements. The possibility of an error in TT4 determination was excluded by measurement of T4 after removal of the subjects’ serum proteins and extraction of the iodothyronine with ethanol. The concentration of T4 after reconstitution of the ethanol extract was consistent with the original measurement in serum. In addition, the concentration of dialyzable T4 was determined in the presence of increasing amounts of TT4. At a serum level of TT4 equivalent to that found in the subjects with HSA R218P, the absolute FT4 concentrations were, on the average, 40, 432, and 1970 pmol/L for sera expressing HSA R218P, R218H, and R218 (WT), respectively. Thus, as expected on the basis of clinical observations and in agreement with our calculations, the affinity of HSA R218P for T4 must be higher than that of HSA R218H to produce an 11-fold difference in FT4 at the same concentration of TT4. This difference was obliterated at the saturating concentrations of TT4 used for the determination of Ka values by the method of Scatchard.
The reason why mutations of arginine 218 in HSA increase the affinity of the molecule for T4 is open to speculation. Previous studies from Bhagavan’s laboratory used site-directed mutagenesis of HSA to study T4 binding of the WT and R218H HSAs by means of tryptophan 214 fluorescence quenching (14). Their results suggested that a single T4-binding site in subdomain 2A of HSA is responsible for the increased affinity of HSA R218H for the ligand. The large guanidino group of arginine in the WT HSA interacts unfavorably with the amino group of T4. The replacement with a smaller side-chain of histidine in HSA R218H or alanine increased the T4 binding affinity by more than 1 order of magnitude (23). The same reasoning could explain the effect of the replacement of arginine with a proline.
After completion of the work presented herein and during the time this manuscript was being reviewed, Petersen and Bhagavan (24, 25) published their studies using recombinant HSA R218P synthesized in yeast. The Ka values for T4 binding to R218P, R218H, and WT HSA, determined by fluorescence spectroscopy, were respectively, 3.4 x 105, 4.0 x 106 and 6.9 x 105 mol/L-1 (corresponding to Kd values of 2.64 x 10-7, 2.49 x 10-7, and 1.44 x 10-6 mmol/L). These Ka values are remarkably similar to those presented in the current report, determined by saturation using whole serum and anion exchange resin. As a matter of fact, Ka values of HSA R218P and HSA R218H are within 15% of each other, as measured by fluorescent spectroscopy, and within 20% of each other, as determined by the resin method. These differences are not significant and cannot explain the 7- to 10-fold differences in serum total T4 in subjects expressing the two HSA variants.
Among the different hypotheses to explain this paradox, Petrersen et al. (24) favor interference with the measurement of TT4 in serum resulting in spuriously high values and the presence of serum components that increase the affinity for T4 of R318P, but not R218H HSA. In this communication we have excluded the former by measurement of T4 after its extraction from serum, and the latter by determining the affinities of the variant HSAs in the presence of the subjects’ serum.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received November 15, 1999.
Revised April 13, 2000.
Accepted April 21, 2000.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
