| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Other Original Articles |
Biophysics Laboratory (T.M.), Department of Pathology (E.M.), Biostatistics Unit of Department of Hematology (J.G.), and Department of Endocrine Surgery (J.-L.K.), Jean Bernard Hospital, Groupe de Recherche en Endocrinologie Experimentale et Clinique, 86021 Poitiers Cedex, France
Address all correspondence and requests for reprints to: Thierry Métayé, Laboratoire de Biophysique, Hôpital Jean Bernard, BP 577, 86021 Poitiers Cedex, France. E-mail: . t.metaye{at}chu-poitiers.fr
Abstract
Most of the TSH effects on the proliferation and differentiation of thyroid cells are mediated by cAMP via an adenylyl cyclase-activating Gs protein. TSH receptor responsiveness in cell cultures, is regulated by G protein-coupled receptor kinase (GRK) 2 and 5. To determine whether an alteration in activity and expression of GRKs might be associated with variable levels of TSH receptor desensitization in vivo, we studied human thyroid tissues including 21 normal tissues and 18 differentiated carcinomas. GRK activity was assayed by rhodopsin phosphorylation, and GRK protein and mRNA expressions assessed by immunoblotting and real-time quantitative RT-PCR, respectively. GRK2 and GRK5 were found as the predominant isoforms in the human thyroid. GRK5 protein expression was significantly decreased in differentiated thyroid carcinoma (P < 0.02) and paralleled a decrease in GRK mRNA expression (P < 0.02). In contrast, no difference in protein and mRNA levels of GRK2 were observed between normal and cancerous thyroid tissues. Although GRK2 protein levels correlated with GRK activities, we demonstrated a significant increase in GRK activity in differentiated thyroid carcinoma (P < 0.02). Less TSH receptor desensitization occurred in differentiated carcinoma than in normal thyroid tissue, as judged by TSH-stimulated cAMP response in human thyroid cells in primary culture. In conclusion, this study indicates that GRK2 activity and GRK5 expression have opposite regulations in cancer cells. Furthermore, the decrease in GRK5 expression may underlie the reduction in homologous desensitization of the TSH receptor in differentiated thyroid carcinoma, contributing to explain the increased cAMP levels in these tumors.
THE TSH RECEPTOR, which belongs to the family of seven-transmembrane domain receptors, is a major determinant of thyroid function. Most of the TSH effects on the proliferation and differentiation of thyroid cells, are mediated by cAMP via an adenylyl cyclase-activating Gs protein (1). The mitogenic effect of cAMP on human thyrocytes has been well established in vitro and received convincing arguments in vivo. Indeed, anti-TSH receptor autoantibodies and point mutations of the TSH receptor or the G
s protein genes, activate the cAMP cascade and cause hyperproliferative diseases including Graves’ disease, toxic thyroid nodule, and carcinoma (2, 3, 4, 5, 6). These observations emphasize the importance of understanding the molecular mechanisms of cAMP synthesis but also those involved in the control of intracellular cAMP accumulation.
Prolonged stimulation with TSH results in desensitization of the TSH receptor in vitro (7, 8) and in vivo (9), as reflected by a decreased cAMP response to subsequent TSH stimulation. The molecular mechanisms of the homologous desensitization (TSH dependent) have been lightened by the prototype models of ß2-adrenergic receptor (10) and rhodopsin (11). Exposure of G protein-coupled receptors (GPCRs) to an agonist results in the uncoupling of the receptor from its heterotrimeric G protein-coupled, a process that starts by GPCR phosphorylation. A unique class of serine/threonine kinases so called the G protein-coupled receptor kinases (GRKs), phosphorylates the agonist-occupied form of receptor. Based on sequence and functional similarities, the GRK family has been divided in three subfamilies: 1/rhodopsin kinase (GRK1), 2/the ß-adrenergic receptor kinases (GRK2 and GRK3), 3/the GRK4, GRK5, and GRK6 subfamily (12). The phosphorylation of GPCRs by GRKs induces its binding to an arrestin protein, which leads to both termination of some aspects of cAMP signaling as well as to the initiation of other pathways, including activation of c-Src and members of MAPK cascades (13).
The presence of GRK2 and GRK5 was first demonstrated in a rat thyroid cell line (FRTL-5), but conflicting results exist regarding the predominant GRK isoforms expressed in these cells (14, 15). TSH-induced cAMP production was substantially decreased in FRTL-5 cells overexpressing GRK2, GRK5, or GRK6. Conversely, treatment of cells with either an oligonucleotide antisense to the GRK5 mRNA (15) or a GRK2 dominant negative mutant (16), resulted in higher TSH- induced cAMP production. Taken together, these studies indicated a key role for GRK2 and GRK5 in the regulation of TSH receptor signaling.
Differentiated thyroid carcinoma (DTC) of follicular cell origin possesses TSH receptors and a functional coupling to adenylyl cyclase (17). The basal cAMP concentration in tissues of these carcinomas has been found significantly greater than in normal thyroid tissue (NTT) (17, 18). Furthermore, the homologous desensitization process, as observed in thyroid slices (19) or in neoplastic thyroid cell lines (20), was weaker in DTC compared with normal tissues. Based on these previous studies, we wondered whether in DTC, decline in the desensitization process, could be associated with a less efficient GRK mechanism. To address this question, we assessed the activity and expression of GRKs in thyroid tissues. We demonstrate an increased GRK activity and a decreased GRK5 expression in DTC compared with normal human thyroid tissue.
Materials and Methods
Subject protocol
All samples from human thyroid tissues, including 21 normal tissues and 18 primary DTCs, were obtained at operation and stored in liquid nitrogen or used immediately for cell culture. Tissue samples were randomly selected after histological examination and classified according to the World Health Organization recommendations (21). This study was approved by the Poitiers Hospital Ethics Committee. NTTs were adjacent to 6 benign nodules, 4 toxic adenomas, and 11 DTCs. The mean age of normal subjects was 51.1 yr (SD, 15.2 yr). Among the 18 DTCs, there were 4 follicular and 14 papillary thyroid carcinomas. Follicular thyroid carcinomas occurred in 2 males and 2 females with a mean age of 56.5 yr (SD, 28.5 yr). The mean tumor size of follicular carcinomas was 4.4 cm (SD, 2.7 cm). Papillary thyroid carcinomas occurred in 3 males and 11 females with a mean age of 51.6 yr (SD, 12.5 yr). The mean tumor size of papillary carcinomas was 2.6 cm (SD, 1.6 cm). Seven subjects exhibited positive lymph nodes and one had distant metastases preoperatively. Patients with thyroid carcinomas were biologically euthyroid, and 99mTc scintigraphies showed cold lesions. All patients were without treatment before operation. Finally, 6 patients had paired samples of both normal and cancerous thyroid tissues.
Sample preparation and GRK partial purification
Human thyroid tissues were cut into small pieces with a razor blade and mechanically pulverized in an impact grinder at -180 C (Freezer Mill, SPEX Industries, Edison, NJ). The powders were then homogenized at 4 C in three times its volume of HEPES buffer (20 mM HEPES; 2 mM EDTA; 250 mM NaCl; 1 mM dithiothreitol; 0.02% Triton X-100; 0.1 mM phenylmethylsulfonyl fluoride; 20 µg/ml leupeptin; 20 µg/ml aprotinin; and 0.1 mg/ml benzamidine, pH 7.2) using an Ultra-Turrax homogenizer (Janke and Kunkel, Staufen, Germany). The homogenates were incubated under agitation for 1 h at 4 C and centrifuged at 140,000 x g. The supernatants, which contain GRKs, were further purified by adding a 50% suspension (vol/vol) of SP-Sepharose (Amersham Pharmacia Biotech, Uppsala, Sweden), and the mixtures were gently agitated for 1 h at 4 C. The SP-Sepharose binds GRKs but not other kinases, such as PKC (22), which phosphorylates rhodopsin. Almost 100% of cytosolic GRKs binds SP-Sepharose as measured by Western blotting for GRK2. The resin was pelleted, washed three times, and the bound GRKs were eluted with three 700-µl volumes of HEPES buffer containing 600 mM NaCl. The eluates were collected, diluted with NaCl-free buffer, and concentrated using Centricon-30 (Amicon, Beverly, MA) concentrators. Typically, 80–90% of SP-Sepharose-bound GRKs were recovered in the concentrates. The final extracts were stored at -70 C and used for Western blot and phosphorylation assays.
Positive GRK controls were obtained from the cytosolic fraction of transfected COS-7 cells with expression vectors for GRK2, GRK3, GRK5, and GRK6 from human origin.
Protein concentrations were determined by the method of Bradford (23) with a Bio-Rad Laboratories, Inc. (Hercules, CA) protein assay reagent, using BSA as standard.
GRK enzymatic activity assay
GRK enzymatic activity was assayed using light-dependent phosphorylation of rhodopsin (24). Rod outer segment (ROS) membranes were prepared from dark-adapted bovine retinas and then treated with 5 M urea to deactivate the endogenous rhodopsin kinase (25). GRK-dependent phosphorylation was measured by incubating 8 µg of protein extracts with a reaction mixture containing 250 pmol rhodopsin, 50 µM [
-32P]ATP (2 Ci/mmol, Amersham Pharmacia Biotech, Buckinghamshire, UK), 20 mM Tris-HCl, 2 mM EDTA and 5 mM MgCl2 at pH 7.5 in a final volume of 40 µl. The reactions were carried out at 32 C for 40 min in the presence or absence of light. The incubations were terminated by the addition (1 ml) of an ice-cold solution of 100 mM sodium phosphate pH 7.0 containing 5 mM EDTA at 4 C, followed by centrifugation at 20,000 x g for 5 min. The ROS membrane pellets were resuspended in 20 µl of SDS sample buffer (62.5 mM Tris-HCl, pH 6.8, containing 1% SDS, 10% glycerol, 2.5% ß-mercaptoethanol and 0.005% bromophenol blue) and electrophoresed on 10% SDS-polyacrylamide gel. Rhodopsin bands identified by Coomassie blue staining, were exposed for autoradiography and counted for 32P radioactivity. Results, obtained by the difference between phosphorylations in the presence and absence of light, were expressed as pmol of phosphate incorporated/min/mg protein. The enzymatic reaction was linear over 40-min period for an activity no greater than 50 fmol of phosphate incorporated/min. All assays were performed with the same ROS membrane preparation and results were confirmed in at least two separate experiments, tested in duplicate.
Determination of GRK protein expression
GRK protein expression was determined by electrophoresis and immunoblotting. SDS/PAGE was performed by the method of Laemmli (26), with a 10% separating gel. After electrophoresis, proteins were electrotransferred to a polyvinylidene difluoride (PVDF) membrane (Bio-Rad Laboratories, Inc.) using a Bio-Rad Laboratories Mini Trans-Blot apparatus (27). Unreacted sites on the PVDF membranes were blocked overnight with a solution of 5% nonfat dry milk in Tris-buffered saline (20 mM Tris-HCl, pH 7.6, containing 137 mM NaCl). The PVDF membranes were then incubated for 1 h at room temperature with anti-GRK antibodies. GRK2 protein expression was determined using a 4 x 10-4 dilution of a mouse monoclonal antibody, raised against a C-terminal peptide (amino acids 467–688) of human GRK3 (Upstate Biotechnology, Inc., Lake Placid, NY) and which recognizes both GRK2 and GRK3. GRK3 protein expression was determined using a 3.5 x 10-3 dilution of a rabbit polyclonal antibody, raised against a C-terminal peptide (amino acids 675–688) of bovine GRK3 (Santa Cruz Biotechnology, Inc., Santa Cruz, CA). GRK5 protein expression was determined using a 2 x 10-3 dilution of a rabbit polyclonal antibody, raised against a C-terminal peptide (amino acids 571–590) of human GRK5 (Santa Cruz Biotechnology, Inc., Santa Cruz, CA). GRK6 protein expression was determined using a 2.5 x 10-3 dilution of a rabbit polyclonal antibody, raised against a C-terminal peptide (amino acids 525–544) of human GRK6 (Santa Cruz Biotechnology, Inc.). After several washings with Tris-buffered saline supplemented with 0.06% Tween 20, the membranes were incubated for another hour in a 5 x 10-4 dilution of peroxidase-conjugated second antibody. Immunoreactive bands were visualized with a commercial electrochemiluminescence system (Amersham Pharmacia Biotech, Buckinghamshire, UK) using Hyperfilm-electrochemiluminescence as described by the manufacturer. Films were optically scanned with model GS 300 densitometer (Hoefer Scientific, San Francisco, CA). The peak areas were analyzed with a GS 365W program, version 2.22 from Hoefer. Results from different blots were normalized using the same positive GRK control and expressed in percentage relative to NTT values. All results were confirmed in at least two separate experiments.
Determination of GRK mRNA expression
GRK2 and GRK5 mRNA expressions were determined by real-time quantitative RT-PCR. Human thyroid tissues, preserved in liquid nitrogen, were cut into small pieces with a razor blade and mechanically pulverized in an impact grinder at -180 C (Freezer mill, SPEX Industries, Edison, NJ). Total RNA was extracted from 100 mg powder, using the RNeasy Mini kit (QIAGEN, Hilden, Germany), according to the manufacturer’s instructions. Deoxyribonuclease digestion was performed during total RNA purification with the ribonuclease-free deoxyribonuclease set (QIAGEN). The total RNA was collected in 50 µl ribonuclease-free water, quantified spectrophotometrically and frozen in aliquots at -70 C.
Oligonucleotide primers were designed to be intron spanning and selected in the 3' region of GRK2 and GRK5 genes, using the computer program Primer Express (PE Applied Biosystems, Foster City, CA). The primer sequences for GRK2 were: sense, 5'-gcagctgggccatgagg-3'; antisense, 5'-tgccactgggtcaggaagg-3'. The primer sequences for GRK5 were: sense, 5'-gaccacacagacgacgacttc-3'; antisense, 5'-cgttcagctccttaaagcattc-3'. The primer sequences for ß-actin were: sense, 5'-tccctggagaagagctacg-3'; antisense, 5'-gtagtttcgtggatgccaca-3'.
Total RNA (5 µg) was reverse transcribed to cDNA in a 25-µl-volume reaction containing 50 mM Tris-HCl (pH 8.3), 75 mM KCl, 3 mM MgCl2, 10 mM dithiothreitol, 400 U Moloney murine leukemia virus reverse transcriptase (Life Technologies, Inc. Ltd., Paisley, UK), 40 U ribonuclease inhibitor, 1 mM deoxy (d) ATP/TTP/CTP/GTP, and 4 µg random hexamers. The ribonuclease inhibitor, dATP/TTP/CTP/GTP and random hexamers were purchased from Amersham Pharmacia Biotech (Uppsala, Sweden). The samples were incubated at 37 C for 1 h. The cDNAs were then diluted (1/4) in ribonuclease-free water and the reverse transcriptase was inactivated at 99 C for 2 min. All RNA samples included in this study were treated with a common reverse transcriptase mixture. Quantitative PCR reaction was carried out in 96 sample tubes/assay, using a qPCR Mastermix kit for Sybr Green I (Eurogentec, Seraing, Belgium) and performed on the ABI Prism 7700 Sequence Detection System (PE Applied Biosystems). A master mixture was prepared according to the manufacturer’s instructions with 1x reaction buffer containing 5 mM MgCl2, 200 µM dATP/CTP/GTP/UTP, 0.625 U of Hotgoldstar DNA polymerase, 1/60000 dilution of Sybr Green I, and sense and antisense primers. The primer concentrations were, 100 nM and 300 nM for GRK2 and 300 nM and 50 nM for GRK5 in sense and antisense orientations, respectively. The primer concentrations were 900 nM for ß-actin. The thermal cycling conditions comprised an initial step at 50 C for 2 min, then a denaturation step at 95 C for 10 min and 45 cycles of 2-step PCR including 15 s at 95 C and 1 min at 60 C. To normalize for differences on the amount and quality of total RNA added to the reaction, amplification of ß-actin RNA was performed as an endogenous control. The samples were assayed in triplicate in two separate experiments and results were expressed as mean ± SEM of 6 different values. Each PCR run included the 5 points of the calibration curve (serially diluted human normal thyroid cDNA), a no-template control, and 18 unknown patient cDNAs. Agarose gel electrophorese of PCR products showed a unique band at the expected size. Furthermore, direct sequencing of PCR products certified the specificity of PCR reactions.
Cell culture, desensitization, and cAMP measurement
Human thyroid tissues were obtained aseptically from patients undergoing thyroid surgery, usually for uninodular or multinodular goitre. Subsequent steps were performed in a laminar-flow chamber as previously described by Roger et al. (28), but with modifications. Thyroid tissues were subjected to enzymatic digestion in Ca2+- and Mg2+-free HBSS, containing 1 g/liter dispase (0.5 U/mg, Roche Molecular Biochemicals, Mannheim, Germany) and 0.1 g/liter collagenase (217 U/mg, Worthington Biochemical Corp., Freehold, NJ). Isolated cells and follicles were washed 3 times in DMEM (Life Technologies, Inc. Ltd., Paisley, UK), containing 25 mM HEPES, 0.6 g/liter bovine albumin and 5% FCS, then centrifuged at 600 x g. The resulting pellet was resuspended in culture medium (DMEM/Ham’s F-12 supplemented with 5% FCS), and the thyroid cells were cultured in 96-well plates for 24 h in a humidified atmosphere of 5% CO2 at 37 C. After this period, the thyroid cells were rinsed with culture medium and incubated (first incubation) for 2 h with or without 1 mU/ml TSH (bovine TSH, Sigma, St. Louis, MO). The cells were then washed twice and incubated for 30 min in culture medium (second incubation) to ensure removal of TSH. Finally, the thyroid cells were tested for responsiveness to 10 mU/ml TSH during a 20-min cAMP stimulation assay (third incubation). This assay was performed in 0.2 ml of culture medium containing 0.5 mM 3-isobutyl-1-methylxanthine (IBMX). In other experiments, the thyroid cells were incubated for the indicated periods at 37 C in culture medium supplemented with 10 mU/ml TSH and 0.5 mM IBMX. The incubations were ended by rapidly discarding the medium and adding 0.25 ml of 12% trichloroacetic acid at 4 C. Following centrifugation, trichloroacetic acid was removed with water-saturated diethylether and the supernatants were evaporated in a vacuum oven. The dried extracts were then dissolved in acetate buffer and stored at -70 C until cAMP was assayed by means of RIA (Amersham Pharmacia Biotech, Buckinghamshire, UK). Desensitization was calculated as the percent decrease of the cAMP response in the third incubation; % desensitization: 100x [1 - (cAMP level in desensitized cells - basal cAMP in desensitized cells/cAMP level in undesensitized cells - basal cAMP in undesensitized cells)].
Statistical methods
GRK2 and GRK5 mRNA expressions were analyzed using paired differences of threshold cycle values with ß-actin mRNA expression as endogenous control. GRK activity, protein expression and mRNA expression were compared between NTT and DTC groups using Wilcoxon rank-sum test. An association between GRK activity and GRK2 immunodetection was sought using the Spearman rank correlation coefficient. Data of intracellular cAMP accumulation and desensitization were compared using t test. A value of P < 0.05 was considered as a minimum level of significance.
Results
GRK activity in normal and cancerous thyroid tissues
To assess GRK activity levels in DTC and NTT, sample tissues were treated with 250 mM NaCl to extract soluble and membrane bound GRKs. Then, GRKs were partially purified by ion exchange chromatography on SP-Sepharose gel to eliminate thyroglobulin, which varies in different thyroid tissues, reducing the accuracy of results expressed per mg protein (29). ROS were used as specific substrate, resulting in phosphorylation of a 38-kDa band that was consistent with the labeling of rhodopsin (Fig. 1A
). The light-dependent phosphorylation of ROS was completely inhibited by the addition of heparin and was not significantly affected by 1 µM protein kinase A inhibitor, PKI or 0.1 µM staurosporine, a PKC inhibitor (data not shown). These biochemical features suggest that our experimental conditions were suitable for measuring GRK activity. Light exposure resulted in 1.9- to 12.2-fold (median, 4.4) increase in rhodopsin phosphorylation. GRK activities in samples from DTC (5.05 ± 1.11 pmol/min/mg protein, n = 8) were significantly increased (P < 0.02, Fig. 1, B and C) compared with those in samples from NTT (2.16 ± 0.18 pmol/min/mg protein, n = 9). In DTC group, 7 patients had papillary carcinoma and 1 had follicular carcinoma. Because of the small follicular carcinoma number, we could not compare these two subgroups. Three patients had paired samples and GRK activities were, 3.33, 3.81, and 10.89 pmol/min/mg protein in DTC, and 2.49, 2.96, and 1.94 pmol/min/mg protein in their adjacent normal tissues, respectively.
|
Before studying GRK protein expression levels in NTT and DTC, we wanted to clarify what GRK isoforms were physiologically expressed in human thyroid tissue. GRK proteins were partially purified from a NTT and analyzed by Western blotting. Using monoclonal antibody raised against GRK2/3, a 79-kDa protein, as determined by the mobility of standard markers and comigrating with recombinant GRK2 and GRK3 proteins, was observed in NTT (Fig. 2, GRK2/GRK3). Because GRK3 protein level was not detected in NTT (Fig. 2, GRK3), even upon much longer exposure of the gel (data not shown), this suggest that GRK2 is the main ß-adrenergic receptor kinase expressed in human NTT. Western blotting with a polyclonal rabbit antibody raised against GRK5, showed an immunoreactive band in NTT (Fig. 2, GRK5), comigrating with recombinant GRK5 protein at an apparent mass of 68 kDa. No GRK6 protein was detected in NTT (Fig. 2, GRK6). Taken together, these results show that GRK2 and GRK5 proteins are the most expressed GRK isoenzymes in normal human thyroid tissue.
|
To determine whether the increase observed in GRK- mediated phosphorylation of ROS in samples from DTC was associated with an increase in immunodetectable GRK2 protein, we performed Western blotting to compare its relative expression in DTC and NTT (Fig. 3). All samples assayed for GRK activity, were also tested for their GRK2 content by immunoblotting, with the exception of 3 NTTs. In DTC group, 9 patients had papillary carcinoma and 1 had follicular carcinoma. A 79-kDa immunoreactive band, comigrating with recombinant GRK2 protein, was present in NTT and DTC (Fig. 3A). Densitometric analysis of immunoblots (Fig. 3B), revealed no difference in relative expression of GRK2 protein between NTTs (100 ± 13%, n = 9) and DTCs (126.4 ± 20.0%, n = 10). Two patients had paired samples showing relative expressions of GRK2 protein in papillary carcinomas of 111.9% and 158.0% compared with 100% expression in their adjacent normal tissues. A positive correlation was observed between GRK2 protein expression and GRK activity (r = 0.82, P < 0.001), confirming that the ROS phosphorylation assay reflected mainly GRK2 activity (30).
|
To determine whether the increase in GRK activity in DTC might be associated with an increase in the expression of other GRKs, we assessed GRK5 protein by Western blotting to compare its relative level in DTC and NTT (Fig. 4). All samples tested for GRK2 protein by immunoblotting were also evaluated for GRK5 protein expression, with the exception of one NTT. The 68-kDa immunoreactive band, comigrating with recombinant GRK5 protein, was present in NTT but was hardly detected in DTC (Fig. 4A). Densitometric analysis of immunoblots (Fig. 4B) confirmed the existence of a significant difference (P < 0.02) in relative expressions of GRK5 protein between NTTs (100 ± 24%, n = 8) and DTCs (27.3 ± 8.1%, n = 10). In our patient group, 5 NTTs had the highest values of GRK5 expression, whereas 4 DTCs had the lowest one. A paired sample of both normal and cancerous thyroid tissues had relative expressions of GRK5 protein of 100% and 72.5%, respectively.
|
To determine whether mRNA expression levels were associated with variations observed in protein expression, we measured GRK2 and GRK5 mRNA levels by real-time quantitative RT-PCR in 8 NTTs and 10 DTCs (Fig. 5, A and B). Standard curves for each of the targets were constructed with the same cDNA obtained from a NTT and the slope used to validate the efficiency of real-time PCR methods. The coefficients of variations on threshold cycle values were calculated for each parameter and were less than 1%. The same thyroid samples were assayed for both GRK2 and GRK5 mRNA expression levels. One patient had a follicular carcinoma, the others carcinomas were of papillary type. Our results indicated a more abundant GRK5 mRNA expression in NTT samples (mean: 100%, range: 53.3–187.5%) than in DTC samples (mean: 42.4%, range: 17.9–100.4%, P < 0.02) and no difference between tissues for GRK2 mRNA expression levels (mean: 100%, range: 57.5–173.9% for NTT, and mean: 132.5%, range: 58.6–299.8% for DTC). Some patients such as Gu, Al, and Bn (Fig. 5) had both low GRK2 and GRK5 mRNA expressions. Taken together, these results fall in line with those obtained for GRK protein expressions.
|
To investigate whether alterations in GRK5 expression and GRK activity were associated with a default in TSH-mediated response, we assessed the kinetics of cAMP synthesis and TSH receptor desensitization in primary cell cultures of 7 NTTs and 5 DTCs. In a first series of experiments, thyroid cells from a papillary carcinoma and its adjacent normal tissue, were incubated with TSH for up to 4 h (Fig. 6A). Basal levels of cAMP were significantly higher in papillary carcinoma cells (0.89 ± 0.04 pmol/well) than in normal thyroid cells (0.74 ± 0.03 pmol/well, P < 0.05). Maximum intracellular cAMP concentrations were observed 30 min after TSH stimulation in normal and cancer cells. Then, intracellular cAMP concentrations declined after 1-h TSH stimulation and the levels were approximately stable until 4 h for each cell type, with values significantly higher in cancer cells than in normal cells. The extent of desensitization was more pronounced in normal thyroid cells. For example, cAMP levels after 4-h TSH stimulation reached 1.8 times the basal value in normal cells and 3.1 times the basal value in cancer cells. In a second series of experiments, we evaluated more accurately the percentages of desensitization after 20-min and 2-h pretreatments of thyroid cells with TSH (Fig. 6B). In DTC cells, the percentages of TSH receptor desensitization were 8.4 ± 4.4% and 56.0 ± 7.3% after 20-min and 2-h TSH preincubation, respectively. These values were significantly lower (P < 0.01) than those obtained with normal thyroid cells, 42.5 ± 6.4% and 82.7 ± 2.2% after 20-min and 2-h TSH pretreatment, respectively.
|
Three significant features of GRKs in NTT and DTC, emerge from the present study. First, increased GRK activity was found in DTC compared with that in NTT. Second, GRK5 protein expression decreased significantly in DTC whereas GRK2 protein levels appeared to be unchanged in the same tumors (this result was confirmed at the mRNA level). Third, GRK3 and GRK6 protein were not detected in human thyroid tissue as previously reported in rat thyroid cells (15, 16). The very low amount of GRKs in human thyroid gland explained the requirement for large tissue samples, and the small number of patients with neoplastic sample and their surrounding normal tissue tested in this study. To validate our results, identical differences, compared with the entire group, were found between the normal and neoplastic tissue in 5 out of 6 measurements using paired samples. The absence of TSH effect on paired samples from the same patient is corroborated by the absence of correlation observed between GRK2 or GRK5 protein expressions and patients’ TSH levels assayed before operation (data not shown). However, in rat thyroid cells, Nagayama et al. have shown that TSH reduced mRNA and protein levels of GRK5 (31). This suggests that in DTC, GRK5 levels may be controlled by both a TSH-dependent mechanism as well as a TSH-independent mechanism. The signaling pathway by which TSH acts to down-regulate GRK5 is mediated through the adenylyl cyclase-cAMP system (31).
Because rhodopsin is a better substrate for GRK2 than for GRK5 (30), the positive correlation observed between GRK activity and GRK2 protein expression confirms that ROS phosphorylation is most likely due to GRK2. Therefore, the dissociation observed between GRK2 amount and enzymatic activity might be explained by the hypothesis of posttranslational modification of the protein itself. For instance, activity and protein expression of PKC have been shown to be elevated in thyroid carcinomas (32). Moreover, phosphorylation of GRK2 by PKC activates its translocation to the membrane and rhodopsin phosphorylation (33). Accordingly, PKC-induced GRK2 phosphorylation may explain the increased GRK activity observed in DTC without requiring changes in protein and mRNA expressions.
Ours results, in primary culture of human thyroid cells, showed that desensitization level of the TSH receptor in DTC was significantly decreased compared with that in NTT. This phenomenon was accompanied by higher intracellular cAMP concentrations observed in thyroid cancer cells after 4-h incubation with TSH. These results fall in line with those obtained in thyroid slices and in human thyroid cancer cell lines, using different cell culture conditions (19, 20). The reduction of TSH receptor homologous desensitization is consistent with the observed decrease in GRK5 level, leading to the suggestion that GRK5 is involved in the regulation process of TSH-stimulated cAMP response in human DTC. Evidence for functional importance of GRK5 in TSH receptor signaling was initially demonstrated in rat thyroid cells in culture by transfection experiments using sense and antisense plasmids (15). To date, GRK/ß-arrestin machinery, TSH receptor internalization and TSH receptor down-regulation have been shown to regulate cAMP synthesis in thyroid cells. It was suggested that GRK2 desensitized the rapid (20 to 30 min) TSH-stimulated cAMP response, whereas GRK5 desensitized the receptor-mediated response after 2- to 24-h exposure to TSH (15, 16). A maximum 30% of the TSH receptor was internalized after 15-min TSH treatment, followed by its recycling to the cell surface (34). Furthermore, prolonged exposure to TSH decreased TSH receptor mRNA expression in dog thyroid and FRTL-5 cells (35), but this effect was very weak in human thyrocytes (36).
We show here for the first time an association between GRK5 expression and cAMP regulation process in human DTC. GRK5 levels were selectively altered because no change in protein and mRNA expressions of GRK2 was detected. Moreover, in FRTL-5 cell line transfected with GRK5 plasmid (15), basal as well as agonist-stimulated adenylyl cyclase activities were significantly reduced, suggesting that a decrease in GRK5 level may facilitate basal cAMP accumulation in DTC. An increase in ß-arrestin 2 expression was observed in toxic adenomas compared with surrounding normal tissues (37), demonstrating that ß-arrestin 2 rather than ß-arrestin 1, is predominantly regulated with GRKs in the control of TSH receptor signaling.
Generation of the oncogenic phenotype is the consequence of somatic chromosome mutations or other genetic alterations that cause constitutive activation or a loss of function of proteins. GRK5 was mapped to the 10q24 region. Importantly, the translocation of the 10q24 region has been specifically observed in thyroid adenoma (38) but not in DTC, suggesting an alteration of some genes in this region that have not yet been identified. Therefore, there is a possibility for the GRK5 gene to be involved in certain thyroid tumors.
Treatment of thyroid cells with phorbol ester (PKC activator) has been reported to mimic TSH-induced desensitization (39) and also to increase GRK activity with no change in total GRK2 protein level (33). These data were not observed in DTC. In contrast, TSH-induced desensitization was decreased, and GRK activity was increased in these tissues. Therefore, in DTC, GRK5 seems to be the predominant regulator of the cAMP pathway activated by the TSH receptor.
In summary, the present studies have shown that GRK2 and GRK5 are the predominant GRK isoenzymes expressed in human thyroid tissue. GRK3 and GRK6 proteins were not detected. Opposite regulations of GRK2 activity and GRK5 expression were observed in cancer cells. Furthermore, the decrease in GRK5 expression may underlie the defect in homologous desensitization of the TSH receptor in differentiated thyroid carcinoma, contributing to explain the increased cAMP levels in these tumors.
Acknowledgments
We thank Drs. J. L. Benovic (Department of Microbiology and Immunology, Thomas Jefferson University, Philadelphia, PA), L. Iacovelli (Department of Molecular Pharmacology and Pathology, Consorzio Mario Negri Sud, Santa Maria Imbaro, Italy), and F. Boulay (Département de Biologie Moléculaire et Structurale/Commissariat à l’Energie Atomique, Grenoble, France) for kindly providing the expression plasmids for GRKs. We also thank Dr. N. Bennett (Laboratoire de Biophysique Moléculaire et Cellulaire, Centre National de la Recherche Scientifique/Commissariat à l’Energie Atomique, Grenoble, France) for the generous gift of rod outer segment membranes, and Dr. S. Patri (Laboratoire de Génétique Cellulaire et Moléculaire du Pr Kitzis, Poitiers, France) for helpful assistance and discussion in the determination of mRNA expression. We are grateful to D. Acun and V. Cassin for their technical assistance, and to Prof. C. J. Larsen for careful reading of the manuscript.
Footnotes
Abbreviations: dATP, Deoxy ATP; DTC, differentiated thyroid carcinoma; GPCRs, G protein-coupled receptors; GRK, G protein-coupled receptor kinase; IBMX, 3-isobutyl-1-methylxanthine; NTT, normal thyroid tissue; PVDF, polyvinylidene difluoride; ROS, rod outer segment.
Received October 29, 2001.
Accepted March 9, 2002.
References
1B-adrenergic, and A1 adenosine receptor-mediated responses. Mol Pharmacol 56:316–324This article has been cited by other articles:
 |
|
 |
|
  T. Metaye, P. Levillain, J.-L. Kraimps, and R. Perdrisot Immunohistochemical detection, regulation and antiproliferative function of G-protein-coupled receptor kinase 2 in thyroid carcinomas J. Endocrinol., July 1, 2008; 198(1): 101 - 110. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H.-Y. Zhang, H.-Q. Wang, H.-M. Liu, Y. Guan, and Z.-X. Du Regulation of tumor necrosis factor-related apoptosis-inducing ligand-induced apoptosis by DJ-1 in thyroid cancer cells Endocr. Relat. Cancer, June 1, 2008; 15(2): 535 - 544. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. Giroux, J. Iovanna, and J.-C. Dagorn Probing the human kinome for kinases involved in pancreatic cancer cell survival and gemcitabine resistance FASEB J, October 1, 2006; 20(12): 1982 - 1991. [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 |
38 kDa) and inhibition of GRK activity by heparin (20 µg/ml) in a NTT. B, Rhodopsin phosphorylation by samples from a DTC and its adjacent NTT. C, Mean distribution of GRK activity in partially purified fractions of 9 NTTs and 8 DTCs. The data represent the mean ± SEM.
