| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Division of Pediatric Endocrinology (F.G.R., J.F., W.G.S., N.K.), Department of Pediatrics, and Department of Pediatric Cardiology (M.V.), University Hospital Schleswig-Holstein, Campus Kiel, 24105 Kiel, Germany; Department of Pediatrics (L.d.S., A.T.), University of Torino, 10126 Torino, Italy; Division of Pediatric Endocrinology (S.E.), Regina Margherita Children’s Hospital, 10100 Torino, Italy; Pediatric Endocrinology (B.K.), University Children’s Hospital, University of Ulm, 89075 Ulm, Germany; Screening Laboratory Hannover (M.P.), 30952 Ronnenberg-Benthe, Germany; Institute of Biochemistry (J.G.), Christian-Albrechts-Universität zu Kiel, 24098 Kiel, Germany; and Department of Physiology (G.F.-T.), Dartmouth Medical School, Lebanon, New Hampshire 03756
Address all correspondence and requests for reprints to: Felix G. Riepe, M.D., Division of Pediatric Endocrinology, Department of Pediatrics, University Hospital Schleswig-Holstein, Schwanenweg 20, D-24105 Kiel, Germany. E-mail: friepe{at}pediatrics.uni-kiel.de.
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Objective: Our objective was to reveal the cause of renal salt loss in six PHA1 patients and analyze the mutants’ functional impact on MR function.
Design: Our study included the following: clinical and hormonal characterization of the patients’ phenotype, analysis of the NR3C2 gene, determination of receptor affinities to aldosterone and the transcriptional activation abilities of the MR mutants, investigation of subcellular translocation using fluorescence-labeled MR, and studying changes in mutant receptor conformation with proteolysis experiments and three-dimensional modeling.
Results: Six heterozygous NR3C2 mutations were detected. One frameshift mutation (c.1131dupT) has been reported previously. The second frameshift mutation (c.2871dupC), which has only recently been reported by our group, showed no aldosterone binding and no transactivation because of a major change in receptor conformation. Two novel nonsense mutations generate a truncated receptor protein. Two missense mutations differently affect MR function. S818L was reported recently without complete in vitro data. S818L does not bind aldosterone or activate transcription or translocate into the nucleus. A major displacement of several residues involved in aldosterone binding was PHA1 causing. The novel E972G mutation showed a significantly lower ligand-binding affinity and only 9% of wild-type transcriptional activity caused by major changes in receptor conformation.
Conclusions: Our data on six mutations extend the spectrum of PHA1-causing NR3C2 gene mutations. Studying naturally occurring mutants helps to clarify their pathogenicity and to identify crucial residues for MR structure and function.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Pseudohypoaldosteronism type 1 (PHA1) is a rare inherited disease characterized by neonatal salt loss resistant to mineralocorticoid treatment first described by Cheek and Perry in 1958 (5). Two forms of PHA1 are distinguished at the clinical and molecular level (6, 7). The autosomal recessive form (OMIM no. 264350) is caused by mutations of the amiloride-sensitive epithelial sodium channel (ENaC) subunit genes SCNN1A, SCNN1B, and SCNN1G (8). The autosomal dominant PHA1 (OMIM no. 600983) is caused by inactivating mutations in the human MR gene (NR3C2 or hMR) (9). The disease expression of the two genetically different PHA1 forms varies dramatically. Autosomal recessive PHA1 presents in the neonatal period with hyponatremia caused by multiorgan salt loss, including kidneys, colon, and sweat and salivary glands. Hyponatremia and hyperkalemia are combined with elevated plasma renin and aldosterone concentrations. Children suffering from this PHA1 form often show lower respiratory tract diseases because of altered sodium-dependent liquid absorption (10). Autosomal recessive PHA1 usually persists into adulthood and shows no improvement over time (6). Autosomal dominant PHA1 is characterized by an isolated renal resistance to aldosterone, leading to renal salt loss, hyponatremia, hyperkalemia, metabolic acidosis, failure to thrive, and elevated plasma renin and aldosterone concentrations in infancy. Patients can be treated with oral salt supplementation. Particularly the autosomal dominant form of PHA1 typically shows a gradual clinical improvement during childhood, allowing the cessation of sodium supplementation.
Here we studied the underlying genetic cause for PHA1 disease expression found in six unrelated families and identified two novel nonsense mutations, two missense mutations, and two single-nucleotide duplication mutations in the NR3C2 gene. The disease-causing nature of the missense mutations and the single-nucleotide duplication mutation described previously by our group were studied in vitro and interpreted in silico using the three-dimensional structure of the MR protein.
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Seven patients presenting with isolated renal salt loss from six families in Italy and Germany were studied. All patients manifested in early infancy (from 2 wk to 2 months) with poor weight gain, failure to thrive, dehydration, or vomiting. The diagnosis of PHA1 was established by the confirmation of hyponatremia, hyperkalemia, elevated plasma renin activity or direct renin concentrations, and high plasma aldosterone levels. All patients were treated with oral sodium supplementation. With the exception of family F, none of the family histories showed any record of prematurity, increased salt craving, episodes of vomiting, or electrolyte disturbances. An uncle in the maternal line of family F was reported to have suffered from feeding problems in the first months of life, but no additional clinical data were available. Family E and the index patient have been described in detail previously (11). The clinical, biochemical, and hormonal data of the PHA1 families are shown in Table 1
.
|
Blood samples for molecular genetic analysis were taken with informed consent. Genomic DNA was extracted from peripheral blood leukocytes, and the MR gene (NR3C2) was amplified using 19 primer pairs and the PCR conditions described previously (9). NR3C2 sequencing using an automated fluorescence sequencer (ABI Prism 310 Genetic Analyzer; Perkin-Elmer, Wellesley, MA) included all translated exons (2, 3, 4, 5, 6, 7, 8, 9) of the NR3C2 gene and the exon/intron boundaries. The NR3C2 cDNA (GenBank NM_000901) was used as template for analysis, numbering the A of the ATG translation initiation codon with +1. The mutations were designated in accordance with the recommendations of the Nomenclature Working Group (12). The previously published three-dimensional MR crystal structure (PDB code 2AA2) was used for analysis of the structural localization of the missense mutations (13, 14, 15). The structural representations were generated with the Ribbons program (16) on a Silicon Graphics Indigo 2 workstation (Silicon Graphics GmbH, Grasbrunn, Germany).
Construction of plasmids
The human full-length NR3C2 cDNA was amplified from the pRS-NR3C2 construct, kindly provided by R. M. Evans (Salk Institute, San Diego, CA), by PCR using the upstream primer 5'-AACAGGTACCCGGCGAGAGA-3' and the downstream primer 5'-TTTCATCTCGAGGAACAGGAA-3', thereby introducing a KpnI and an XhoI restriction site (2). The PCR product was cloned into the pGEM-T Easy vector (Promega, Madison, WI), and the clones were sequenced to check the integrity of the insert. The NR3C2 cDNA was subsequently cloned into the KpnI/XhoI site of a pcDNA3.1 expression vector (Invitrogen, Carlsbad, CA).
The mutagenesis was performed on the pcDNA3.1-NR3C2, using the QuikChange XL site-directed mutagenesis kit (Stratagene, La Jolla, CA) with the following sense primers and their corresponding antisense primers (not displayed): 2453C>T, 5'-TCA TTT GCC TTG AGC TGG AGA T tG TAC AAA CAT ACG AAC AGC-3'; 2871–2872insC, 5'-GCG CTG AAG GTA GAG TTC CCC cGC AAT GCT GGT GGA GAT CAT-3'; 2915A>G, 5'-AGC TGC CCA AGG TGG gGT CGG GGA ACG CCA AGC-3' (positions of nucleotide substitutions are given in lowercase, boldface letters). The introduction of the mutations was verified by sequencing the entire construct. The transfer into a new pcDNA3.1 vector was performed as described above. Full-length wild-type NR3C2 cDNA and mutant NR3C2 cDNAs were also inserted into the HindIII/ApaI site of the pECFP-C2 and pEYFP-C2 plasmids (Clontech, Palo Alto, CA), containing sequences encoding fluorescence-enhanced jellyfish fluorescent protein (cyan fluorescent protein and yellow fluorescent protein) (17).
In vitro expression and transactivation
Approximately 2.5 x 105 COS-7 cells were grown in DMEM supplemented with glutamine and 10% charcoal-stripped fetal calf serum in 24-well plates and transiently transfected 24 h after seeding using lipofectamine (Invitrogen). Cells were transfected with 0.3 µg pcDNA3.1-NR3C2 or pcDNA3.1-NR3C2 mutant in the presence of 0.3 µg of a TAT3-luciferase reporter construct (TAT3-luc) (18) or 0.3 µg of a mouse mammary tumor virus (MMTV)-luciferase reporter construct (MMTV-luc). Cotransfection of 0.1 µg pRL-TK (Promega) coding for renilla luciferase was performed to normalize for transfection efficiencies. In case of cotransfections of different pcDNA3.1-NR3C2 plasmids, the amount of total DNA was kept equal with empty pcDNA3.1 vector. Two days after transfection, aldosterone (Sigma Chemical Co., St. Louis, MO) was added for 24 h in different concentrations ranging from 10–12 to 10–6 mol/liter. Cells were lysed in reporter lysis buffer (Promega), followed by the measurement of firefly and renilla luciferase activity with the dual luciferase assay (Promega). Protein concentrations were estimated by the Bradford method. Results were standardized for transfection efficiency and protein content and expressed as the ratio of firefly luciferase activity over renilla luciferase activity over protein content. The EC50 was calculated using GraphPad Prism software version 4.0 (GraphPad Software, San Diego, CA).
The 1471.1 cells have been described previously (19, 20). These cells were derived from the mouse adenocarcinoma line C127 and contain multiple copies of the stably replicating MMTV-chloramphenicol acetyl transferase (CAT) reporter. The 1471.1 cells were transfected with 10 µg of a pcIL2-receptor plasmid (cytomegalovirus promoter driving the IL-2 receptor gene) (21), 2.5 µg MMTV-luc, and 2.5 µg pcDNA3.1-NR3C2 or the respective pcDNA3.1-NR3C2 mutant by electroporation with a BTX600 electroporator (BTX Inc., San Diego, CA) at 960-µF capacitance, 129-ohm resistance, and 240–280 V in chilled cuvettes. Cells were harvested after 6 h of recovery, and magnetic affinity cell sorting was used to separate transfected from nontransfected cells as described (21). Sorted cells were treated for 12 h with different concentrations of aldosterone (Sigma). The 1471.1 cells were lysed in reporter lysis buffer and assayed for CAT activity, luciferase activity, and protein content. CAT activity was measured using the Molecular Probes fast CAT assay (Invitrogen) by a thin-layer chromatography method. Visualization of the acetylated products was performed with a FluoroImager (Molecular Dynamics, Sunnyvale, CA). Analysis was performed using the ImageQUANT software (Amersham Biosystems, Piscataway, NJ). Luciferase assay was carried out following standard procedures (Promega). Protein content was measured as described above.
Protein studies
MR was transcribed and translated in vitro using the TNT Quick Coupled Transcription-Translation system (Promega) following the manufacturer’s protocol. Plasmids coding for wild-type or mutant MR were used as templates for transcription with T7 polymerase, followed by translation with either cold methionine for the binding studies or with [35S]methionine (1000 Ci/mmol) for SDS-PAGE analysis.
Aldosterone binding characteristics were evaluated as previously described (22). In vitro translated MR protein lysates were diluted 2-fold in ice-cold TEGWD buffer containing 20 mmol/liter Tris-HCl (pH 7.4), 1 mmol/liter EDTA, 1 mmol/liter dithiothreitol, 20 mmol/liter sodium tungstate, and 10% glycerol. Aliquots of 25 µl were incubated with [3H]aldosterone in increasing concentrations (0.25–8 nM) in the absence and presence of an excess of unlabeled aldosterone for 4 h at 4 C to determine specific and nonspecific binding (n = 3). Bound and unbound steroids were separated by the dextran-coated charcoal method (25 µl lysate and 1000 µl 0.5% dextran-coated charcoal in TEGWD buffer, stirred for 10 min at 4 C). The radioactivity of the bound fractions was counted in an LSC TRI-CARB 2300 TR liquid scintillation spectrometer (Packard, Dreieich, Germany) using an external standard. The dissociation constant at equilibrium (KD) of the wild-type MR and the mutants was calculated using GraphPad Prism 4.0 software (GraphPad Software).
For the limited proteolytic digestion assay, 35S-labeled MR proteins were incubated with or without 10–6 mol/liter aldosterone for 10 min at 20 C, and 20 µg/ml trypsin for the unbound MR proteins and 200 µg/ml trypsin for the bound MR proteins was added to 35S-labeled MR proteins for various times. Aliquots of the digestion product were removed and mixed with protein-loading buffer, boiled for 5 min, and subjected to electrophoresis. After electrophoresis, the SDS-polyacrylamide gels were dried and autoradiographed at –80 C overnight.
Intracellular localization and trafficking
Rabbit cortical collecting duct cells were maintained as previously described (23). Approximately 7500 cells were grown in monolayers on small glass coverslips in medium containing charcoal-stripped fetal calf serum. Transient transfections were carried out using lipofectamine PLUS with 20 ng of the YFP-NR3C2 plasmid per coverslip or 40 ng of the CFP-NR3C2 plasmid per coverslip. To study the intracellular localization of the receptor, aldosterone was added for 1 h in a concentration of 10–9 mol/liter on the day after transfection. Cells on coverslips were fixed after hormone treatment with paraformaldehyde and counterstained with 4,6-diamidino-2-phenylindole after permeabilization following standard procedures. For the intracellular trafficking studies, the coverslips with the living cells were placed into a heated chamber (37 C) under a confocal microscope. The chamber was perfused with culture medium supplemented with 10–9 M aldosterone (17). Fluorescence was detected sequentially in a population of 10 cells in three repeated experiments. The confocal Zeiss LSM 510 Meta laser scanning microscope (Carl Zeiss, Oberkochen, Germany) was used for fluorescence detection in both settings. ECFP was excited with the 458-nm line and EYFP was excited with the 514-nm line from an argon laser. Images were captured using a x63 plan apo objective (1.4 NA). Image processing and fluorescence analysis were performed with the software ImageJ 1.36 (http://rsb.info.nih.gov/ij/).
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Six different heterozygous mutations were detected in the family members (Fig. 1). The index patient in family A carries a duplication of thymidine 1131 (c.1131dupT) in exon 2. This leads to a frameshift resulting in a putative premature stop codon at amino acid position 378 (E378X). The mutation was inherited from the father. The mother showed a wild-type sequence at this position. The nonsense mutation c.2017C>T in exon 5 was found in the index case of family B and her mother. The mutation changes arginine 673 to a stop codon (R673X) deleting the ligand-binding domain (LBD). Another nonsense mutation c.2024C>T in exon 5 was detected in family C in the index case only. This mutation changes serine 675 to a premature stop codon (S675X) leading to an absent LBD. Two different missense mutations were detected in families D and F. The missense mutation c.2453C>T in exon 6 was detected in two affected siblings in family D and their father. The nucleotide change results in a substitution of leucine for serine at amino acid position 818 (S818L). The mutation c.2915A>G in exon 9 was detected in the index patient in family F and his mother. This leads to a change of glycine for glutamic acid at position 972 (E972G). The mutational analysis in family E that revealed the 2871dupC mutation exclusively in the index patient had been performed previously and is therefore not displayed in Fig. 1 (11). The location of the mutations within the NR3C2 gene is shown in Fig. 2. In the three-dimensional structure of the LBD of MR, the residue S818 is located in helix 5 and the residue E972 is located at the end of helix 12 (Fig. 3).
|
|
|
Transactivation on an MMTV and TAT3 reporter by wild-type MR was induced through aldosterone in a dose-dependent way, reaching a plateau at 10–8 mol/liter aldosterone (Fig. 4). The ED50 value was calculated with 6.5 x 10–11 ± 1.4 x 10–11 mol/liter and 7.6 x 10–11 ± 3.9 x 10–12 mol/liter for MMTV and TAT3, respectively. The mutants S818L-MR and 2871dupC-MR showed no measurable transactivation with either the MMTV or TAT3 reporter constructs. The aldosterone-induced luciferase activity of E972G-MR showed a shift of the dose-response curve to higher aldosterone concentrations with an ED50 value of 7.2 x 10–9 ± 4.5 x 10–10 mol/liter and 5.4 x 10–9 ± 1.9 x 10–9 mol/liter for MMTV and TAT3, respectively. The maximal luciferase activity of E972G-MR at 10–6 mol/liter aldosterone compared with wild type was 76% for MMTV and 80% for TAT3. To test whether the reduced aldosterone-induced transactivation of the E972G-MR mutant is caused by a disturbed chromosome/DNA accessibility, a transactivation assay was performed in 1471.1 cells. The luciferase activity originating from a cytoplasmic MMTV reporter was compared with the CAT activity resulting from a chromosomally integrated MMTV reporter. The relative transactivation for wild-type MR was the same for MMTV-luc and MMTV-CAT. The same was found for E972G-MR, although the transactivation ability in both systems was only 24 and 21% of the wild type, respectively. Cotransfection experiments of the wild-type MR were performed together with the mutated proteins to rule out a dominant negative effect of the mutations. Using equimolar concentrations of each MR construct, the transactivation induction at 10–10 mol/liter aldosterone was similar for S818L-MR and 2871dupC-MR cotransfected with wild-type MR (Fig. 4). Wild-type MR in cotransfection with E972G-MR induced a significantly higher luciferase activity at 10–10 mol/liter aldosterone compared with wild-type MR alone (60.4 ± 1.1-fold induction vs. 70.3 ± 1.8-fold induction; P < 0.05).
|
Protein translation of the mutant S818L-MR, 2871dupC-MR, and E972G-MR was studied in vitro by SDS-PAGE analysis. All MR mutants showed similar in vitro translation levels resulting in an MR protein of an expected size of approximately 100 kDa for S818L-MR and E972G-MR and approximately 110 kDa for 2871dupC-MR (data not shown).
The ability of wild-type and mutated MR proteins to bind aldosterone was studied in vitro. The mutants S818L-MR and 2871dupC-MR showed a complete absence of aldosterone binding, whereas the mutant E972G-MR showed a detectable but significantly lower aldosterone-binding ability. The binding characteristics of mutant E972G-MR were further defined by the analysis of the dissociation constant KD by Scatchard analysis. Compared with the KD of the wild-type MR protein (3.16 ± 0.20 nM), the KD of the mutant E972G-MR was 4-fold lower (11.32 ± 4.10 nM) under the experimental conditions.
A limited proteolysis assay was performed to analyze whether the differences in ligand binding of the mutant MR proteins were caused by a changed receptor conformation. Without aldosterone binding, two major fragments of 30 and 40 kDa were generated from the wild-type and the S818L and E972G receptors (Fig. 5A). Limited proteolysis of the 2871dupC mutant resulted in a 30-kDa fragment and a slightly longer band of approximately 42 kDa. The intensity of the 40-kDa fragment was higher after 5 min of digestion in the wild-type, S818L, and E972G mutant receptors compared with the 42-kDa fragment from the 2871dupC mutant. In addition, the 42-kDa fragment of the 2871dupC protein had almost completely vanished after 15 min and was absent after 30 min of digestion, whereas the 40-kDa fragment was clearly visible after 15 min in the wild-type, S818L, and E972G receptors. The 2871dupC mutation is thus more sensitive to trypsin digestion in the unbound conformation. Binding of aldosterone induced a change in the wild-type receptor conformation, identifiable by an increase in the resistance of the 40-kDa fragment to trypsin proteolysis (Fig. 5B). Because the S818L and 2871dupC mutant receptors are unable to bind aldosterone, the 40- or 42-kDa fragment was completely digested. The aldosterone-bound E972G mutant 40-kDa fragment was clearly visible, but it was not as resistant to trypsin digestion as the wild-type receptor showing a much weaker band after 30 min and a completely digested fragment after 60 min.
|
Initially, the transcriptional activity of the YFP-NR3C2 and CFP-NR3C2 fusion MR proteins was determined. COS7 cells were transfected with the respective fluorescence protein (FP)-NR3C2 plasmid together with the responsive MMTV-luc reporter. The transcriptional activity of all FP-NR3C2 plasmids was consistently equal to the wild-type MR protein (data not shown). The cellular distribution of the fusion proteins in the presence and absence of hormones was examined. In the absence of aldosterone, the majority of transfected cells exhibited both cytoplasmic and nuclear fluorescence irrespective of whether FP-wild-type-MR or FP-mutant-MR constructs were transfected. No FP-MR was found in the nucleoli. After treatment with aldosterone FP-wild-type-MR and FP-E972G-MR were completely localized to the nucleus in 100% of the transfected cells. The aldosterone-bound FP-MR concentrated in clusters within the nucleus. The FP-2871dupC-MR and FP-S818L-MR constructs did not translocate into the nucleus after aldosterone incubation. No clustering was detectable for these two mutations, either in the cytoplasm or in the nucleus (data not shown). In addition, the cellular distribution of the cotransfected fusion proteins in the presence of hormones was examined to analyze the potential interaction between the wild-type and mutant receptors. CFP-wild-type-MR was cotransfected in equimolar amounts with the three YFP-mutant-MR constructs (Fig. 6). After incubation with aldosterone, the CFP-wild-type-MR translocated into the nucleus and concentrated in clusters. Similar to the single transfection experiments, the YFP-E972G-MR protein translocated into the nucleus forming clusters at the same localization as the wild-type receptor. After aldosterone treatment and cotransfection with wild-type receptor, the YFP-2871dupC-MR and YFP-S818L-MR constructs showed neither translocation into the nucleus nor the regular cluster formation. The wild-type MR translocation and cluster formation was not influenced by these cotransfected mutants. The intracellular trafficking of the wild-type receptor and the E972G mutant was studied in living cells by sequential fluorescence detection. The addition of 10–9 M aldosterone resulted in a complete translocation of the wild-type receptor into the nucleus within 60 min. No significant differences were recorded for the translocation of the mutant E972G receptor (P = 0.162) in three independent experiments (data not shown).
|
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
The novel nonsense mutation R673X was found in the index patient in family B and in his mother. In contrast to the index patient who suffered from renal salt loss, the mother never presented with any symptoms. It is possible that a mild PHA1 phenotype in the mother was missed during infancy because plasma aldosterone and renin activity are elevated to date, but she had no recollection of treatment or hospitalization. This case is thus another example for the wide phenotypic heterogeneity in autosomal dominant PHA1 (11, 24). The second novel nonsense mutation S675X was detected only in the index patient of family C so PHA1 is classified as sporadic in this family. Alternatively, the reported paternity is incorrect. Aldosterone concentrations were still elevated at the age of 15 months and sodium supplementation was still necessary. Both novel nonsense mutations are located at the start of exon 5 and introduce a premature stop codon. This generates a truncated MR protein that lacks the complete LBD. Each truncated MR mutant studied to date had no transcriptional activity (25, 26).
The mutation c.1131dupT was previously described by Sartorato et al. (27) (named insT1354) in two sisters from an Italian kindred. The family studied in this study was not obviously related to the Italian kindred described above. As with the patient described previously, our male index patient carrying the c.1131dupT mutation presented with the typical signs of PHA1 in infancy. It has been shown previously that the mutant receptor is not able to activate transcription from an MMTV promoter (27). Despite these in vitro data, sodium replacement therapy was stopped at the age of 3 yr in our patient, and no electrolyte disturbances occurred. This clinical course exemplifies the age dependency of inactivating MR mutations with regard to the clinical phenotype. As with the family reported by Sartorato et al. (27), the affected father of our patient shows no clinical signs of PHA1. The nonsense mutation c2871dupC was described previously by our group (11) where the NR3C2 gene mutation was detected in a patient of Spanish origin. Segregation analysis revealed that the mutation occurred sporadically or that the paternity was different. The female patient showed the typical signs of autosomal-dominant PHA1 with a hyponatremia of 129 mmol/liter and failure to thrive. The 2871dupC mutant showed no transactivation from an MMTV or TAT3 promoter and no dominant negative effect on the wild-type transactivation ability. Furthermore, no nuclear translocation of the 2871dupC mutant and no interaction with the wild-type receptor translocation in coexpression experiments were observed. However, the in vitro protein translation was similar to the wild-type protein, but no binding of the physiological ligand aldosterone could be detected. The duplication mutation creates an MR nonsense protein elongated by 28 residues. The change of the amino acid sequence starts at residue A958 and changes the H12 helix completely (A958R, M959N, L960A,... X1013). Our experimental data indicate that this changes the receptor conformation dramatically, because the mutant is more sensitive to proteolysis in the unbound conformation. Helix H12 harbors the second ligand-dependent activation function (AF-2), which plays a major role in receptor activity (28). Helix H12 is repositioned after binding of an agonist on the ligand-binding pocket and thereby contributes to an interface for the interaction with coactivators. This mechanism is most probably disturbed in the c2871dupC mutant. We therefore postulate that the change of the receptor conformation caused by the severe sequence alteration in helix H12 is responsible for the absent aldosterone binding and the lack of in vitro transactivation. Because no RNA was available from the patient, it is not clear whether the mutant mRNA is available for in vivo protein translation or whether it is degraded because of an absent or altered 3'-end formation signal (29) and haploinsufficiency is responsible for the patient’s phenotype.
The missense mutation S818L was detected in two infants suffering from a renal form of PHA1 and in their healthy father. The segregation reflects an autosomal dominant trait. The affected siblings had a comparable phenotype, whereas no evidence for a renal salt loss was found in the father. This indicates that a normalization of biological and hormonal parameters can occur with age with the S818L mutation. Only recently was a sporadic case of PHA1 carrying the S818L mutation reported (30). The described phenotype of the index case was quite similar to that in our patients. The authors investigated only the transactivation ability of the mutant receptor and the underlying cause of the inactivation of the S818L-MR has therefore not been fully studied and clarified. In this study, we were able to show a comparable S818L-MR protein translation. The mutant receptor was nearly free of aldosterone binding. This explains the absence of transactivation abilities and the lack of nuclear translocation. The S818L-MR showed no major change in receptor conformation in the unbound state. S818L-MR exhibited no dominant negative effect on the wild-type receptor, as was shown in transactivation and translocation assays. The residue S818 is conserved in the MR LBD of different species as well as in the human progesterone (PR), glucocorticoid (GR), and androgen (AR) receptors (Fig. 2
). This reflects the importance of the S818 residue for proper receptor function. No mutations are reported in the corresponding residues of human PR, GR, or AR. The residue S818 is located in helix H5 of the MR protein (Fig. 3) (13). Helix H5 forms a charge clamp pocket together with helices H3, H4, and H12 (AF-2) necessary for the docking of coactivators. As shown in the three-dimensional structure, S818 is not oriented toward the molecular surface, and it is therefore unlikely that this residue is part of the coactivator binding site. Crystallization experiments of the AR also revealed that the corresponding residue S753 is not in contact with the various binding motifs of the nuclear receptor coactivators (31). In addition, helix H5, together with helices H4, H8, and H9, build the middle layer of the MR protein structure creating an interior cavity for the ligand binding (13). Although the residue S818 is in close vicinity, it has an orientation that allows for no direct interaction with the ligand. However, S818 is necessary to stabilize helix H5 and the ß-sheet 1 via hydrogen bonds to Y828. The position of Y828 in the ß-sheet 1 is of elementary importance for keeping the residue R817 in correct orientation. The MR residue R817 is involved in an extensive hydrogen-bond network locking the A-ring of aldosterone in the ligand-binding pocket (13). In addition, the residue F829, which is also part of the hydrogen-bond network keeping the ligand in place, is anchored in the ß-sheet 1. Hence we hypothesize that the substitution of the bulky apolar leucine for serine at position 818 is displacing the ß-sheet 1, which severely disturbs the interaction of R817 and F829 with the ligand aldosterone. Therefore, ligand binding is impossible for the S818L receptor mutant.
The novel missense mutation E972G was detected in an Italian patient and his mother. Although no events of salt loss were recollected in the mother’s history, she presented with a slightly elevated plasma aldosterone concentration in combination with normal plasma renin activity. A maternal uncle is reported to have suffered from feeding problems and is likely to have been a carrier of the E972G mutation. The ligand-binding ability of E972G-MR was 4-fold lower than that of the wild-type receptor. This is most probably caused by a significant change in the receptor conformation, as was shown in the proteolysis experiments. The residue E972 is situated in the C-terminal part of helix H12 (13). As with other nuclear receptors, MR shows a unique, unconserved amino acid sequence in the C-terminal region (Fig. 2). The C-terminal region, including helix H12 and the following ß-strand, seems to play an important role in receptor activation by stabilizing the active conformation (25, 32). It has been reported that the deletion of the last few residues of the MR and the GR induce changes in hormone binding specificity and a reduction of the receptor-mediated transactivation of target genes (25, 33). In MR, the residue E972 is involved in a hydrogen-bond network with R947 anchoring helix H12 to H10. Through this, the hydrophilic residues protect the hydrophobic internal layer formed by the two helices involved in ligand binding. We speculate that the exchange of glycine for glutamic acid at residue 972 would open up the hydrophobic core and displace helix H10. This would disturb the contact of the residue T945 with aldosterone and explain the decreased ligand-binding ability. The E972G-MR protein displayed significantly lower transactivation abilities. We showed that this loss of transactivation ability is not caused by a deterioration of DNA accessibility, because the mutant protein can equally activate transactivation from a cytoplasmic or a nuclear reporter gene sequence. The transcellular trafficking and nuclear localization was found to be comparable to the wild-type receptor. The duration for a complete nuclear translocation was extended in our in vitro setting compared with that in the literature, which can be explained by the use of different cell lines and MR constructs (17). No dominant negative effect was detected in coexpression experiments; on the contrary, because of the remaining transactivation abilities of E972G, a significantly higher gene transcription was noticed. In conclusion, the impairment of transactivation can be explained by the disturbed ligand-binding ability but may also reflect an additional inactivation of the receptor activation function mediated by the AF-2 region located in helix 12 (34).
An increasing number of NR3C2 mutations have been detected in patients with autosomal dominant and sporadic PHA1. The in vitro and structure-function analyses of these mutations are necessary to clarify their pathogenicity and to classify the patient’s disease as autosomal dominant PHA1. Although these results are obtained in vitro, the analyses are helpful for the clinician because they provide the opportunity to counsel the parents and to withdraw sodium supplementation with a low risk of a relapse of salt loss. In addition, it enables the identification of functionally and structurally important MR protein residues and may open up the possibility of encountering novel factors and pathways regulating sodium homeostasis.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Disclosure statement: F.G.R., J.F., L.S., S.E., A.T., B.K., M.P., M.V., J.G., W.G.S., and N.K. have nothing to declare. G.F.-T. received lecture fees from Pfizer.
First Published Online September 5, 2006
Abbreviations: AR, Androgen receptor; CAT, chloramphenicol acetyl transferase; FP, fluorescence protein; GR, glucocorticoid receptor; LBD, ligand-binding domain; MMTV, mouse mammary tumor virus; MR, mineralocorticoid receptor; PHA1, pseudohypoaldosteronism type 1; PR, progesterone receptor.
Received May 30, 2006.
Accepted August 29, 2006.
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  M. D. Tobin, M. Tomaszewski, P. S. Braund, C. Hajat, S. M. Raleigh, T. M. Palmer, M. Caulfield, P. R. Burton, and N. J. Samani Common Variants in Genes Underlying Monogenic Hypertension and Hypotension and Blood Pressure in the General Population Hypertension, June 1, 2008; 51(6): 1658 - 1664. [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 |
