| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
-Subunit in Human Embryonic Kidney-293 CellsDepartment of Pharmacology (I.C., D.R., T.C.), Istituto Superiore di Sanità, 00161 Rome, Italy; Department of Clinical Biochemistry (H.L.), Medical University of Innsbruck, and Austrian Academy of Sciences (C.Z., P.B.), Institute for Biomedical Aging Research, A-6020 Innsbruck, Austria
Address all correspondence and requests for reprints to: Tommaso Costa, Dipartimento del Farmaco, Istituto Superiore di Sanità, Viale Regina Elena 299, 00161, Rome, Italy. E-mail: tomcosta{at}iss.it; or Peter Berger, Ph.D., Institute for Biomedical Aging Research, Austrian Academy of Sciences, Rennweg 10, A-6020 Innsbruck, Austria. E-mail: peter.berger{at}oeaw.ac.at.
| Abstract |
|---|
|
|
|---|
-subunit of human glycoprotein hormones (GPH
). Induction of GPH
transcription in response to cAMP elevations resulted in a dramatic increase (600-fold) of protein secretion as shown by RT-PCR and a highly specific time-resolved immunofluorometric assay. Cloning and sequencing of the GPH
cDNA and mass spectrometric analysis of HPLC-purified GPH
derived from serum-free HEK293-β2-adrenergic receptor-stimulated cells verified the nature of the molecule. Enzymatic deglycosylation with subsequent Western blots revealed that this was a large hyperglycosylated form of GPH
that had not been associated with a β-subunit previously. This uncombined variant is known to be either cosecreted with GPHs from the pituitary, the placenta, and a variety of tumors or secreted without GPHs from APUD cells and rare tumors. Moreover, it is similar to GPH
found at high concentrations in seminal plasma. As shown by a panel of endogenous or transfected G protein-coupled receptors in HEK293 cells, the expression of large GPH
was controlled by Gs- and Gq- but not Gi-dependent receptors and mediated via cAMP and Ca++ release. This suggests that Gq- or Gs-coupled receptors other than the classical GnRH receptor may play a role in the regulation of nonpituitary, nonplacental GPH
secretion under physiological and pathological conditions. | Introduction |
|---|
|
|
|---|
- AND β-subunits of the heterodimeric glycoprotein hormones that control reproduction (chorionic gonadotropin, FSH, and LH) and thyroid function (TSH) are members of the larger family of cysteine knot growth factors (1). Neither the single common
- nor four types of β-subunits are considered to be hormonally active in their uncombined free form because the full activation of the corresponding glycoprotein hormone receptors requires heterodimeric hormonal forms (2).
However, biological effects triggered by free subunits, but not likely mediated by glycoprotein hormone receptors, have been reported. The glycoprotein hormone
-subunit (GPH
) was found to be biologically active on prolactin releasing cells of pituitary and presumably endometrium (3, 4), and was described as the active human chorionic gonadotropin (hCG) molecular species in stimulating human decidualization. Minor effects observed with heterodimeric hCG were assigned to uncombined GPH
, generated by dissociation of hormone subunits catalyzed via a yet-unknown mechanism in endometrial cells (5, 6). Recent findings indicate that the local production of GPH
in prostatic tissue inhibits the growth of stromal cells. Thus, decreased expression of GPH
with age might be a favoring factor for the development of hyperplasia (7).
Physiologically significant amounts of free GPH
are eutopically released by the pituitary and trophoblastic placental cells during pregnancy or ectopically by neuroendocrine cells. Unlike in serum in which holo-GPHs are in excess over free GPH
, in human seminal plasma very high concentrations of GPH
are found in 1000-fold excess over free hCGβ and 10,000-fold over the holohormone hCG (8). In testicular and other cancers, secretion of GPH
alone, or in excess over hCGβ subunits and holo-hCG, may occur (9, 10). GPH
was also observed frequently in transformed cell lines derived from neoplastic tissue, such as bronchogenic tumors (11), or HeLa cell clones (12, 13) in which it might exert autoparacrine growth-promoting effects (14).
Regulation of GPH
subunit expression and secretion in the pituitary was investigated in cellular model systems from human, rat, and mouse. The major stimuli for GPH
production and secretion in response to GnRH1 are the activation of the cAMP pathway with consequent phosphorylation of cAMP-responsive element sites and the intracellular release of Ca++ from storage pools with subsequent activation of the MAPK pathway (see Refs. 15, 16 for reviews). However, non-GnRH-regulated ectopic production of GPH
was not investigated to the same extent of detail. This secretion is of relevance because human seminal plasma contains excessive amounts of GPH
that originates from the APUD and stromal cells from the prostate. APUD cells are also a source of GPH
in the gastrointestinal tract and a variety of testicular (10) and gastroenteropathic tumors (17).
Whereas investigating the pattern of gene expression induced by agonist stimulation in a HEK293 cell line stably transfected with β2-adrenergic receptors, we found that human GPH
was the gene with the highest level of induction. Here we show that HEK293 cells express and release uncombined hyperglycosylated GPH
. This expression, unlike in many previously described transformed cell lines, is strictly dependent on activation of intracellular signaling pathways. In fact, GPH
transcription and release in HEK293 cells is barely detectable under basal conditions but reaches high levels only on activation of cAMP and Ca2+-dependent signals. Moreover, using HEK293 cells transiently or stably transfected with receptors selective for each of the three G protein family members, Gs, Gq, and Gi, we show that both Gs and Gq but not Gi proteins mediate increases in GPH
transcription and secretion.
| Materials and Methods |
|---|
|
|
|---|
Array analysis
Atlas human cDNA arrays version 1.2 (1176 cDNA tags x 3) were purchased from CLONTECH (Palo Alto, CA). 32P-labeled cDNA was synthesized according to the manufacturers instructions from poly-A RNA enriched by streptavidin conjugated magnetic beads using [32P]dATP (6000 Ci/mmol; Amersham, Aylesbury, UK) and the components supplied with the array. The hybridization of the radiolabeled probe to the cDNA array filters and the subsequent washes were performed as outlined by CLONTECH. The dried membranes were measured in a Phosphor imager (Cyclone; PerkinElmer, Palo Alto, CA) and analyzed using CLONTECH software (AtlasImage).
Stimulation of GPH
expression and secretion in HEK293 cells
Cells were plated in 6- or 24-well plates and cultured until close to confluence before replacing the medium with 1 ml FCS-free DMEM. Twelve hours later, isobutyl-methyl-xanthine (IBMX) 100 µM, rolipram 10 ng/ml, 8-bromoadenosine-cAMP (8-Br-cAMP) 1 µM, cholera toxin 0.5 µg/ml, forskolin 1 µM, calcium ionophore A2318 10 µM, isoproterenol 1 µM, and 5-(N-ethyl-carboxamido) adenosine (NECA) 100 µM (all from Sigma, St. Louis, MO) were for 24 h, or, in kinetics experiments, for variable time intervals as indicated. Subsequently the medium was collected and stored frozen for the immunofluorometric assay (IFMA) assay of released GPH
, whereas the cell monolayers were treated with Trizol reagent for RNA extraction and RT-PCR analysis. To determine the relative concentration of secreted and intracellular GPH
immunoreactivity, the cell monolayers were washed thrice in PBS, scraped into 0.25 ml ice-cold Tris/HCl [50 mM (pH 7.4)], and sonicated. After centrifugation (40,000 x g, 30 min), the supernatants were stored frozen at –80 C until analyzed by IFMA.
Analysis of mRNA expression by RT-PCR
Total RNA was extracted from cells monolayers by directly lysing cells with Trizol reagent (Invitrogen). About 2 µg were reverse transcribed using SuperScript III reverse transcriptase and an oligo(dT)12–18 as primer (Invitrogen). Aliquots, corresponding to one twentieth of total cDNAs were amplified through 25 cycles of PCR, using Platinum Taq DNA polymerase (Invitrogen), according to the manufacturers instructions. The pairs of sequence-specific oligonucleotides used for amplifications were: 5' primer, 5'-CCA TGG ATT ACT ACA GAA AAT ATG CAG C-3'; 3' primer, 5'-GCA CGC CGT GTG GTT CTC CAC TTT GGA AA-3' for GPH
mRNA; 5' primer, 5'-TCC GTG CCT CCA AGA TGA CAA A-3'; and 3' primer, 5'-CAG AGA AGA GCC TGT CTT CAG TC-3' for mRNA amplification of the housekeeping gene coding for the ribosomal protein S26 used for normalize the expression of GPH
. Synthetic oligonucleotides were from MWG Biotech (Ebersberg, Germany).
Molecular cloning and sequencing of GPH
cDNA in HEK293 cells
The full-length cDNA encoding the GPH
protein was amplified from HEK293 reverse transcribed RNA using the following pair of sequence-specific oligonucleotides: 5'primer, 5'-GCC CTG AAC ACA TCC TGC AAA A-3' and 3'primer, 5'-GCA GTC ATC AAG ACA GCA CTT G-3'. The amplified fragment was purified by phenol/chloroform extraction, cloned into pCRII vector (Invitrogen), and sequenced by MWG Biotech.
Time-resolved IFMAs (TR-IFMA) for hCG
and hCGβ
Coating, blocking, incubation, and washing procedures of the IFMAs for uncombined, i.e. free hCG
and hCGβ, respectively, were performed as described previously (9). Samples were diluted in IFMA buffer [50 mM Tris-HCl (pH 7.75); 0.9% NaCl, 5 g/liter BSA, 0.1 g/liter Tween 40, and 20 mM diethylenetriaminepenta acid; Sigma-Aldrich, Milwaukee, WI] and run in duplicate. The used monoclonal antibodies (mAbs) against hCG
, and hCGβ served as reference reagents in the international TD-7 Workshop on antibodies to hCG and hCG-related molecules, and their characteristics have been described elsewhere (19). Coating mAbs were coded INN(sbruck)-hCG-72 recognizing selectively free hCG
(hCG
assay) and INN-hCG-68 reacting with hCGβ but not with holo-hCG (hCGβ assay). The detection mAbs INN-hFSH-158, a pan
-mAb, for the hCG
assay and INN-hCG-22, a pan β-mAb, for the hCGβ-assay, were labeled with isothiocyanatophenylene triamintetraacetic acid-europium (Wallac, Turku, Finland) according to the manufacturers recommendations. Time-resolved fluorescence was measured for a second in a fluorometer (Victor2; Wallac). Hormone standards (hCG
first I.R.P. 75/569, hCGβ first I.R.P. 75/551) to determine sensitivity and specificity of both IFMAs were kindly provided by the National Institute for Biological Standards and Control (South Mimms, UK). Inter- and intraassay variations of either IFMA were less than 10% over the entire assay ranges.
GPH
purification via reversed phase-HPLC (RP-HPLC)
GPH
was purified from 80 ml of supernatant from HEK293 stably transfected with β2AR and stimulated with isoproterenol (1 µM). The supernatant was 4-fold concentrated by centrifugation with Ultrafree (Biomax 5 kDa cutoff, 15 ml; UFV2BCC10; Millipore, Anderson, MA) according to the manufacturers recommendations, dialyzed against 0.01 M NaHCO3 overnight at 4 C (Slide-A-Lyzer, 3.5-kDa cutoff; Pierce, Rockford, IL), frozen (–70 C), lyophilized, resolved in 5 ml aqua bidest, dialyzed against 10 mM Na-phosphate buffer (pH 3.5; 4 C, overnight; Slide-A-Lyzer, 7-kDa cut-off; Pierce), and purified by RP-HPLC.
The RP-HPLC equipment used consisted of a 127 Solvent Module and a model 166 UV-visible-region detector (Beckman Instruments, Fullerton, CA). The effluent was monitored at 214 nm. Separation was performed on a Nucleosil C4 column (250 mm x 8 mm inner diameter; 5 µm particle pore size; 30 nm pore size; end capped; Seibersdorf, Austria). Samples of 1 ml were injected onto the column and chromatographed within 60 min at a constant flow of 1.5 ml/min with a two-step acetonitrile gradient starting at solvent A-solvent B (70:20) (solvent A: water containing 0.1% trifluoroacetic acid; solvent B: 70% acetonitrile and 0.1% trifluoroacetic acid). Next, the concentration of solvent B was increased from 20 to 70% in 50 min and from 70 to 100% in 10 min. Fractions were collected and, after adding 100 µl 1% NH3, lyophilized and stored at –20 C. Recovery of GPH
at each purification step was monitored by IFMA. The 80-ml starting material contained 188 µg; this was concentrated to 66.6 µg per 0.4 ml after RP-HPLC purification.
Mass-spectrometric (MS) analysis
RP-HPLC purified GPH
(25 µl equivalent to 3.8 µg) from HEK293 stably transfected with β2AR stimulated with isoproterenol (1 µM) (see above) was reduced by 10 mM dithiothreitol in 100 mM NH4HCO3 (pH 8.3) (30 min at 56 C) and alkylated by addition of 25 of 55 mM iodoacetamide in 100 mM NH4HCO3. After 20 min incubation at room temperature in the dark, the sample was digested with
-chymotrypsin [EC 3.4.21.1; Sigma type I-S, 1:100 (wt/wt)] for 4 h at 37 C. The digest was subjected to nanospray-MS. Protein digest was analyzed using a LCQ ion trap instrument (ThermoFinnigan, San Jose, CA) equipped with a nanospray interface. The nanospray voltage was set at 1.6 kV, and the heated capillary was held at 200 C. MS/MS spectra were searched against a human database using SEQUEST (LCQ BioWorks; ThermoFinnigan).
Digestion with glycosidases
GPH
purified by HPLC as described above from supernatant of HEK293 stably transfected with β2AR and specifically stimulated with isoproterenol (10 µM), and for comparative purposes the frozen carrier-free concentrate (FC 862) of the new World Health Organization (WHO)-adopted first International Reference Preparation for Immunoassay of hCG
(first IRR hCG
99/720) were deglycosylated according to the manufacturers recommendations (New England BioLabs, Frankfurt/Main, Germany). N-linked carbohydrate antennae were digested either with peptide N-glycanase PNGase F [peptide-N4-(N-acetyl-β-glucosaminyl) asparagine amidase] (EC no. 3.5.1.52) or with endoglycosidase H when the extent of trimming of the carbohydrate moiety was determined.
In brief, 22 µl sample equivalent to 1 µg GPH
or hCG
were incubated with 2 µl 10 x denaturing buffer (5% sodium dodecyl sulfate, 10% β-mercaptoethanol) for 10 min (100 C) and put on ice. Then 2.2 µl 0.5 M sodium phosphate buffer (pH 7.5), 2.2 µl Nonidet P-40, and 1 µl enzyme (PNGase F, 500 U) were added and incubated for 2 h at 37 C.
For digestion with endoglycosidase H, 1 µg GPH
or hCG
was denatured as described above, 2.2 µl sodium citrate (0.5 M, pH 5.5), and 500 U (1 µl) endoglycosidase H added. Deglycosylation was carried out for 2 h at 37 C.
Western blot
Approximately 500 µg each of PNGase-F- or endoglycosidase H-deglycosylated, and native GPH
and hCG
, were diluted with sample buffer to 25 µl and loaded on precast polyacrylamide gradient gels (4–20% Tris-glycine gels; Cambrex BioScience Rockland, ME) and separated by SDS-PAGE (SDS-PAGE; Mighty Small II; Hofer Scientific Instruments, San Francisco, CA; 150 V, 75 min). Proteins were electrophoretically transferred (3.5 h, 400 mA) to polyvinyl difluoride-membranes (Immunoblot 0.2 µm; Bio-Rad, Hercules, CA) and blocked with 5% (wt/vol) skimmed milk powder in PBS (45 min, room temperature). A mixture of three mAbs directed against three distinct epitopes on hCG
(codes INN-hCG-72, -hFSH-132, and -158) (20) were diluted in 5% (wt/vol) milk powder in PBS/0.1% Tween 20 to approximately 5 µg/ml each and incubated overnight at 4 C. After extensive washing with 5% milk powder/PBS/0.1% Tween 20, membranes were incubated for 1 h with goat antimouse IgG horseradish peroxidase (W 4021; Promega, Mannheim, Germany) diluted 1:2500 in 5% (wt/vol) milk powder in PBS. Chemiluminescent substrate conversion (Super Signal West Dura; Pierce) was detected by Hyperfilm ECL (Amersham) using exposure times between 10 sec and 45 min.
Signaling pathways involved in GPH
transcription and secretion
The cDNA clones for the receptors indicated below were obtained from the UMR cDNA Resource Centre (www.cdna.org).
Cells plated in 24 wells were transiently transfected with cDNAs coding for different G protein-coupled receptors (GPCRs) [muscarinic M3, histamine H1, histamine H2, opioid type-µ (MOP), β2-adrenergic, and neurokinin-1]. To assess the signaling activity of transiently transfected GPCRs, cells were switched 24 h after transfection to serum-free medium containing [3H]myo-inositol (1 µCi/ml) (NEN Life Science Products, PerkinElmer, Shelton, CT) and incubated for an additional 24 h. After removal of the culture media and washing, monolayers were incubated for 30 min at 20 C in PBS containing LiCl (1 mM) and the agonists active on the transfected receptors. After agonist removal, reactions were arrested by adding 0.5 ml ice-cold HCl 0.1 N. Inositol phosphate accumulation and cAMP levels were measured in the same extracts as described previously (21).
| Results |
|---|
|
|
|---|
in HEK cells
expression.
-subunit of the heterodimeric glycoprotein hormones family was up-regulated 500-fold as a result of agonist treatment (Fig. 1A
in this cell line had not been previously reported in the literature.
|
was not the result of the fortuitous selection of a particular HEK293 phenotype during the isolation of the β2AR-expressing clone, total RNA was prepared from parental untransfected HEK293 cells and analyzed by RT-PCR, using primers specific for the human GPH
gene sequence. Basal GPH
expression was extremely low under nonstimulated conditions but strongly induced in the presence of forskolin (which directly activates adenylyl cyclase bypassing receptors). In addition, despite the much lower level of endogenous β2AR in the wild-type cell line (
25 fmol/mg), compared with the transfected clone (32 pmol/mg), a smaller extent of isoproterenol-stimulated GPH
expression was also detectable in the parental cell line (Fig. 1B
gene is an intrinsic property of HEK293 cells in general.
To investigate whether signaling-dependent activation of GPH
gene expression might be a widespread phenomenon, we examined three additional human transformed cell lines, including HeLa cells and A341 epidermoid carcinoma, both of which express β-adrenergic receptors (22, 23), and the SH-EP line, a transdifferetiated epithelial phenotype from human neuroblastoma, known to express protease-activated receptor 1 (24). Hela cells expressed high levels of GPH
mRNA, but unlike in HEK293, the transcription was constitutive, as the levels were virtually identical in basal and isoproterenol or forskolin stimulated conditions (Fig. 1B
). No detectable mRNA was observed in A341 cells, whereas a very low extent of expression slightly stimulated by thrombin was observed in SH-EP cells (Fig. 1B
). Thus, the peculiar characteristic of HEK293 cells is their ability to express high levels of GPH
only in response to activation of intracellular signals.
cAMP-mediated synthesis and secretion of GPH
To verify that the remarkable induction of GPH
transcription in HEK293 cells was matched by a corresponding enhancement of protein synthesis and release, we compared mRNA induction and protein production of GPH
in both HEK293-β2AR and parental untransfected HEK293 cells. As shown in Fig. 2A
, not only stimulated levels (lanes 2 and 4), but also basal mRNA expression (lanes 1 and 3) were greater in HEK293-β2AR than in the parental cells. This indicates that the higher degree of receptor expression in the transfected clone can generate enough constitutive signaling to activate the expression of the gene. Using TR-IFMA, we found that protein synthesis and release followed a quite similar pattern. Immunoreactive GPH
was stimulated 113-fold over basal in wild-type cells exposed to forskolin and 457-fold in the β2AR-expressing clone exposed to isoproterenol. Basal protein levels were also 4-fold greater in β2AR-expressing cells than in wild type (Fig. 2B
). Thus, the stimulation of GPH
mRNA in HEK293 cells is paralleled by a very similar enhanced production of
-subunit protein.
|
and only traces of hCGβ.
Immunoreactive GPH
was found in both culture medium and cell extracts. In the parental untransfected cells, the fraction of total
-subunit in the medium was only 16% under basal conditions but increased to 87% after stimulation. In the receptor-expressing cells, 61% of the protein was released in the medium even in the absence of agonist, in line with the greater level of constitutive signaling noted by PCR analysis, but in the presence of agonist, the fraction of released GPH
was similarly enhanced to 83% (Fig. 2B
). Such shifts in released vs. intracellular ratios induced by stimulation suggest that the protein is actively secreted from the cells after the enhanced transcription triggered by the elevation of cAMP levels.
Coupling of GPH
expression and secretion
To contrast the temporal evolutions of agonist-induced mRNA transcription with that of protein release, we performed RT-PCR assays and TR-IFMA determinations on cell extracts and media of HEK-β2AR cells collected at different time intervals after stimulation by 1 µM isoproterenol. As expected, the induction of mRNA transcription was detectable earlier than release. The transcription of mRNA started within the first 30 min of agonist stimulation, reached high levels after about 1 h, and continued to increase more slowly during the remaining period (Fig. 3A
). The release of protein in the medium took a longer time lag to start (2.5 h) but then displayed a time course consistent with the kinetics of mRNA accumulation. The rate was nearly log linear with time during the first 5 h and then slowed down without reaching any apparent plateau (Fig. 3B
).
|
production in receptor-expressing cells. Both cellular and released
-subunit accumulation were enhanced with very similar concentration-response relationships by the agonist (Fig. 4A
was also increased by isoproterenol in concentration-dependent fashion, but the apparent EC50 for the maximal enhancement of this ratio (17 nM) was in very good agreement with that for cAMP stimulation (Fig. 4A
|
: identification by cDNA cloning, MS/MS, and glycosylation pattern
was purified by RP-HPLC from a serum-free supernatant of cells in which maximal accumulation of the released protein was achieved by exposure to 10 µM isoproterenol for 48 h. From 80 ml of starting material approximately 188 µg of purified GPH
were obtained. An aliquot (3.8 µg) was reduced, alkylated, digested with
-chymotrypsin, and analyzed by MS/MS. Verification that this specifically induced protein in fact is GPH
was achieved by coverage of nearly half of the amino acid sequence of the hCG
apoprotein by the analyzed peptides (Table 1
|
expressed by HEK293 cells may bear spontaneous mutations or polymorphisms, the full-length cDNA encoding the secreted protein was cloned and sequenced. As shown in Table 2
|
that has been secreted as uncombined protein differs in glycosylation pattern from hCG
dissociated from pregnancy-derived holo-hCG. It contains additional oligosaccharide branching and core fucosylation, which result in a slightly elevated molecular mass (change in apparent molecular size =
2 kDa). The comparison of GPH
purified from HEK293 with dissociated hCG
(WHO first International Reference Preparation, 99/720, FC 862) (Fig. 5
. Accordingly, purified GPH
from HEK293 cells was resistant to endoglycosidase H deglycosylation (Fig. 5
, which was partially digested under the same conditions (Fig. 5
on agonist stimulation.
|
expression
expression in HEK293 cells. Steady-state mRNA levels and protein release were measured simultaneously in cells treated with a variety of cAMP raising agents or a calcium ionophore (summarized in Table 3
was strictly regulated by changes of intracellular cAMP concentrations (Fig. 6
expression and release. The cAMP phosphodiesterase inhibitor IBMX also stimulated GPH
accumulation but to a smaller extent and only significantly in the β2AR-expressing clone. This may reflect tonic stimulation of cAMP signaling exerted by the overexpressed receptor in the transfected line. The calcium ionophore A-23187 produced a marked stimulation of GPH
accumulation (Fig. 6
expression.
|
|
transcription, we used a strategy based on receptor-mediated activation of the endogenous G proteins in HEK293 cells. A number of diverse GPCRs capable of interacting with Gs, Gq, or both or Gi (Table 3
expression. In addition to Gs-coupled receptors, GPCRs that primarily interact with Gq (such as the muscarinic M3, the histamine H1 receptors) or both G protein types (neurokinin-1 receptors), also strongly induced GPH
expression (Fig. 7
mRNA. Thus, both Gs and Gq, through cAMP and Ca2+, respectively, regulate the expression of the common gonadotropin
-subunit in HEK293 cells.
|
expression in HEK 293 cells may exhibit functional antagonism or rather exert their effects in an additive/synergic fashion. To this end, we transfected HEK293 cells with the Gq-coupled H1-histamine receptor and examined the consequence of the concurrent stimulation of GPH
transcription via Ca2+ and cAMP signaling. As shown in Fig. 8
|
| Discussion |
|---|
|
|
|---|
-chain for glycoprotein hormones in response to the activation of the cAMP and Ca2+ signaling pathways via stimulation of endogenous and transfected GPCRs other than classical GnRH receptors.
Purification of secreted GPH
from stimulated HEK293-β2AR cells followed by tryptic digestion and MS/MS identification of the peptide fragments showed that the sequence of the released protein unequivocally corresponded to GPH
with minor impurities present in the final preparation, mainly consisting of residual FCS proteins. Molecular cloning of the expressed mRNA confirmed the sequence of the entire GPH
precursor. It is worth noting that the HEK293
-subunit does not carry the single amino acid substitution (A56E) previously described in a GPH
ectopically secreted by a human carcinoma. This mutation was deemed responsible for the abnormal glycosylation and the consequent high molecular size of the protein (27).
Earlier work suggested that uncombined GPH
contains an additional oligosaccharide O linked to T39 (T43 in bovine GPH
), which might be responsible for the shift in molecular weight and the poor ability to dimerize with free β-subunits (28, 29). However, subsequent studies reported little or no extent of O-glycosylation in several free
-producing systems (30, 31, 32) and show that the N-linked oligosaccharides are primarily responsible for the increase in molecular size of the free
-subunit and its reduced ability to form
β-heterodimers (33, 34). Thus, it is conceivable that the GPH
secreted by HEK293 cells is similar the N-hyperglycosylated form that is frequently produced in many eu- and ectopic sources (30, 31, 32, 33, 35, 36, 37, 38).
Western blot analysis indicated that the apparent molecular weight of the released protein is larger than that of the first WHO reference preparation of hCG
, which was dissociated from pregnancy hCG (39). Large uncombined GPH
has been described to be resistant to the digestion by endoglycosidase H (40), which cleaves the chitobiose core of hybrid and high mannose N-linked carbohydrate antennae. Purified HEK293 protein is resistant to deglycosylation with endoglycosidase H, but after PNGaseF digestion, its apparent molecular weight is identical with dissociated hCG
.
Thus, by gene array screening, RT-PCR, molecular cloning, nanospray MS analysis, SDS-PAGE molecular size, and sensitivity to deglycosylation, we unambiguously show that the major molecule induced by intracellular messengers in HEK293 cells is a genuine large uncombined form of GPH
that is heavily glycosylated.
The ectopic overproduction of GPH
in a transformed cell line as we observed in HEK293 is not a new finding. Early reports showed that bronchogenic tumor cell lines can produce unbalanced amounts of hCG subunits, with some clones expressing virtually only uncombined hCG
(11). Later, selected clones of HeLa cells (a continuous line derived from a cervical carcinoma) were found to prevalently express
-subunits in cell culture (12, 41).
However, the second-messenger control of GPH
transcription and release in neoplastic cells has not been investigated in detail. Modest increases of hCG subunits transcription induced by PKA and protein kinase C (PKC) activators were described in choriocarcinoma cells (42), and more recently release of free CG
subunit in response to PKA and PKC activation was reported in a hepatoma cell line (43). In HeLa cell lines variable levels of constitutive
-subunit expression were reported, but in most cases the expression could be enhanced only by treatment with nonspecific activators of gene expression, such as sodium butyrate or inhibitors of DNA synthesis (44, 45). As shown here in a side-to-side comparison, GPH
transcription in HeLa cells is constitutive and cannot be regulated by cAMP signaling.
Thus, unlike other transformed cell lines, the peculiar feature of HEK293 cells GPH
expression is the tight regulation in response to physiological signals. The expression was very low or undetectable under basal conditions but rapidly induced by stimulation of cAMP and Ca2+ signaling pathways, leading to the accumulation of microgram per milliliter concentrations of GPH
in the serum-free medium of the treated cells.
To investigate the signal specificity of GPH
induction in HEK293 cells, we used two different approaches. In the first, we used a panel of stimulators or inhibitors involved in cAMP signaling and a calcium ionophore. The results clearly indicated that GPH
mRNA expression and protein secretion of GPH
can be stimulated by intracellular elevation of cAMP and Ca2+ ions.
It is known that the GnRH receptor, the major physiological regulator of gonadotropin expression and release, can interact in various cell types with at least three different types of G proteins, Gq/11, Gs, and Gi (46, 47, 48). Therefore, in the second approach, we used a number of GPCRs that show a more restricted G protein specificity than the GnRH receptor as transfectable tools to investigate the G protein subtypes capable of mediating transcriptional control of GPH
in HEK293 cells. Our data clearly show that both Gq/11 and Gs, but not Gi, can regulate the expression of large GPH
.
This dual Ca2+ and cAMP responsiveness of GPH
expression and release in HEK293 is in accord with the mechanisms of gene regulation that occur in cells in which GPH
is physiologically secreted, such as pituitary and placental cells. Distinct cAMP and PKC response sequences have been identified in the promoter region of the human GPH
gene and explain the dual control by these two signal transduction pathways (49).
Consistent with that previous work on promoter regulation, we found that there is a significant level of additivity between Ca2+- and cAMP-induced transcription, which agrees with the observation that promoter activity was enhanced in synergistic fashion by Ca2+ and cAMP stimuli (42, 49). This further confirms the physiological nature of the regulation of GPH
expression in HEK293 cells and indicates that this cell line may provide a better model for investigations on the mechanism of signal regulated secretion of free glycoprotein hormone
-subunit in an ectopic system.
APUD cells of several organs secrete uncombined GPH
. Neuroendocrine cells of the prostate might be responsible for the highly concentrated levels of large free GPH
found in the seminal plasma (8), but very little is known about the signaling mechanisms that control secretion of this hyperglycosylated form of GPH
Therefore, HEK293 cells may constitute a useful experimental model for unraveling the contribution of different signaling pathways to the transcriptional and secretory regulation of this protein in such tissues.
Taking advantage of the monoclonal line expressing large amounts of recombinant β2AR, we also performed a series of experiments on cAMP-mediated induction of GPH
production. They showed a close relationship between enhancement of cAMP levels and extent of
-subunit induction, particularly when comparing the effect of isoproterenol in the receptor-expressing clone and the parental line. In wild-type HEK293 cells with very low levels of endogenous β-adrenoceptors, the enhancement of cAMP levels induced by isoproterenol is barely detectable (1.5- to 2-fold). In contrast, in the β2AR-transfected clone isoproterenol produces a 500- to 1000-fold enhancement of intracellular cAMP. The release of GPH
at 24 h induced by isoproterenol in the two cell lines was consistent with such differences in second-messenger responses but also appeared to magnify minor changes in intracellular nucleotide concentrations because the stimulation of
-subunit release via endogenous β-AR in the nontransfected cells was significant (10-fold). There was also a 3-fold difference in basal GPH
release between the receptor clone and the parental line, indicating that such readout is capable of detecting small changes in steady-state basal nucleotide levels caused by the constitutive signaling activity of the overexpressed receptors. GPH
release in HEK293 may thus be used as a highly sensitive assay for cAMP-mediated signaling, particularly when combined with the ultrasensitive TR-IFMA procedure used here. Such an assay might also be one of the most sensitive available to date for the detection of mutational-induced changes in the constitutive activity of Gs-coupled GPCRs.
Despite the agreement between cAMP mobilization and GPH
release, the EC50 for isoproterenol-mediated enhancement of GPH
production was approximately 2 orders of magnitude higher than the EC50 for stimulation of intracellular cAMP. However, such discrepancy disappeared when release was expressed as change of ratios between extracellular and intracellular amounts of GPH
. This indicates that the initial agonist-induced elevation of cAMP levels is more directly linked to the shift between intra- and extracellular pools of GPH
, rather than to the overall accumulation of protein over a 24-h period.
It may be speculated that the discrepancy in agonist EC50 between second-messenger response and GPH
synthesis might be related to the biphasic kinetics of expression noted in this study. GPH
mRNA increased rapidly within the first hour after agonist addition and then more slowly during the remaining 24 h. GPH
production displayed a similar biphasic course, although delayed in onset. We suspect that only the rapid phase may be directly triggered by the initial receptor-induced elevation of cAMP, whereas the slower and more sustained phase of accumulation might be the result of secondary changes that are independent of receptor activation. If so, the discrepancy in concentration-response relationships between signaling and release may be explained, particularly if release is measured as protein accumulation at later end points when the slower phase is predominant.
In conclusion, we have shown that HEK293 cells express and secrete large amounts of a hyperglycosylated form of human GPH
on stimulation of endogenous and transfected β2AR and other GPCRs in a manner that is stringently regulated at the transcriptional level by the cAMP and calcium signaling pathways. GPH
expression in this readily available and highly transfectable cell line can be an experimental model for the investigation of glycoprotein hormone biosynthesis and regulation and could also be exploited as an endogenous reporter gene assay for studies on orphan GPCRs activity.
Moreover, the finding that several receptors coupled with either Gs and Gq, but not Gi proteins, can specifically control the transcription and secretion of large GPH
is important. It suggests that yet-to-be-identified Gq- and Gs-coupled GPCR types potentially expressed in APUD cells or ectopically producing tumors might have a role in orchestrating the control of nonpituitary, nonplacental regulation of free GPH
, both under physiological and pathophysiological conditions.
| Acknowledgments |
|---|
99/720. | Footnotes |
|---|
Abbreviations: β2AR, β2-Adrenergic receptor; 8-Br-cAMP, 8-bromoadenosine-cAMP; FC, carrier-free concentrate; FCS, fetal calf serum; GPCR, G protein-coupled receptor; GPH
,
-subunit of human glycoprotein hormone; hCG, human chorionic gonadotropin; IBMX, isobutyl-methyl-xanthine; IFMA, immunofluorometric assay; mAb, monoclonal antibody; MOP, opioid type-µ; MS, mass spectrometric; PKA, protein kinase A; PKC, protein kinase C; RP-HPLC, reversed phase-HPLC; TR-IFMA, time-resolved IFMA.
Disclosure Statement: The authors of this paper have nothing to declare.
Received June 11, 2007.
Accepted for publication November 27, 2007.
| References |
|---|
|
|
|---|
subunit. Science 226:566–568
molecules from pregnancy stimulate secretion of prolactin from human decidual cells: a novel function for free
in pregnancy. Endocrinology 129:2257–2259
-subunit functions synergistically with progesterone to stimulate differentiation of cultured human endometrial stromal cells to decidualized cells: a novel role for free
-subunit in reproduction. Endocrinology 137:1332–1339[Abstract]
-subunit from human chorionic gonadotropin, which can synergize with progesterone to induce decidualization. J Clin Endocrinol Metab 83:1575–1578
production. J Clin Endocrinol Metab 87:5297–5303
-subunit, free β-subunit of human chorionic gonadotropin (hCG), and intact hCG in sera of healthy individuals and testicular cancer patients. Clin Chem 38:370–376
subunit of glycoprotein tropic hormones. Nature 260:530–532[CrossRef][Medline]
subunit of human chorionic gonadotropin. J Cell Biol 108:2423–2434
-Subunit and human chorionic gonadotropin-β immunoreactivity in patients with malignant endocrine gastroenteropancreatic tumours. Eur J Clin Invest 24:131–136[Medline]
fusion proteins. The receptor of one covalent complex interacts with the
-subunit of another. J Biol Chem 278:15778–15788
-subunit of human chorionic gonadotropin. Nature 281:351–356[CrossRef][Medline]
subunit of a human carcinoma choriogonadotropin. J Biol Chem 261:10475–10477
subunit of the glycoprotein hormones from bovine pituitaries and identification of its O-linked oligosaccharide. J Biol Chem 258:240–244
-subunit and luteinizing hormone (LH) in the presence of LH-releasing hormone. Endocrinology 117:1699–1706
subunit. Differential processing of free and combined forms in human trophoblast and transfected mouse cells. J Biol Chem 262:14197–14203
-subunit and
-subunit dissociated from intact human chorionic gonadotropin, despite differences in sialic acid contents. Endocrinology 121:1215–1220
interfere with its ability to combine with human chorionic gonadotropin-β subunit. J Biol Chem 265:21951–21956
-subunit does not limit the assembly of chorionic gonadotropin
β dimer in human malignant and nonmalignant trophoblast cells. Endocrinology 124:1602–1612
-subunit from cultured choriocarcinoma (JEG) cells: comparison of the subunit secreted free with that prepared from secreted human chorionic gonadotropin. Endocrinology 105:581–587
-subunit and free β-subunit combination in the extraembryonic coelomic fluid of early pregnancy. Endocrinology 136:903–910[Abstract]
subunits of human glycoprotein hormones. J Clin Invest 56:1043–1052[Medline]
-subunit of human choriogonadotropin (hCG
) and the free
molecules produced in pregnancy: most free
and some combined hCG
molecules are fucosylated. Endocrinology 126:2788–2799
subunit secretion by a human lung cancer cell line (ChaGo) transplanted into nude mice. J Natl Cancer Inst 59:1285–1289[Medline]
-subunit gene in the human hepatoma cell line NPLC. Endocrinology 134:582–587
-subunit-like material secreted by HeLa cells. Biochemistry 27:6474–6487[CrossRef][Medline]
-subunit gene. J Biol Chem 268:3903–3910
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |