Role of cleavage and shedding in human thyrotropin receptor function and trafficking Myle ` ne Quellari 1 , Agne ` s Desroches 1 , Isabelle Beau 1 , Emmanuelle Beaudeux 1 and Micheline Misrahi 1,2 1 INSERM E120, Re ´ cepteurs, Signalisations et Physiopathologie Thyroı ¨ dienne et de la Reproduction, and 2 Laboratoire d’Hormonologie et Biologie Mole ´ culaire, Ho ˆ pital Bice ˆ tre, IFR Bice ˆ tre, Le Kremlin Bice ˆ tre, France The thyrotropin receptor (TSHR) undergoes a cleavage at the cell membrane, leading to a heterodimer, comprising an a extracellular and a b-transmembrane and intracellular subunits, held together by disulfide bonds. Moreover, part of the a-subunit of the receptor is shed from thyroid and transfected L cells. To understand the role of cleavage and shedding, we constructed deletion mutants starting, respectively, at the most N-terminal (S314), and C-terminal (L378) cleavage sites previously mapped, corresponding to free b1orb2-subunits without further modification of receptor structure. Functional studies performed in COS-7 cells showed that both mutants display an increased basal activation of the cAMP pathway when compared with the wild-type receptor. By contrast, deletion of almost the entire extracellular domain of the receptor (TM409 mutant) totally impairs receptor function, thus confirming a role of the juxtamembrane extracellular region in receptor function. The b1 mutant receptor exhibited an increased internalizat- ion when compared with the hormone-activated holo- receptor. Furthermore, no recycling was observed in the case of the b1 mutant receptor. These observations strongly argue for a different conformation between the receptor activated by cleavage and shedding on the one hand, and the receptor activated by the ligand on the other hand. Cleavage and shedding of a receptor already activated by a transmem- brane activating mutation M453T further increase its activity, showing that the extracellular domain still exerts a negative effect in the M453T holoreceptor. An increased internalization of the M453T receptor was observed when compared with the wild-type receptor, which was increased further in the corresponding truncated b1-M453T receptor. Thus cleavage and shedding yield TSHR activation but also increase internalization of the free b-subunits of the receptor, the latter mechanism limiting simultaneously excessive receptor signaling. The combined effects may be responsible for the limited basal constitutive activation of the cAMP pathway that is detected for the TSHR. Keywords: thyrotropin receptor; cleavage; shedding; con- stitutive activity; traffic. The thyrotropin receptor (TSHR) plays a key role in thyroid growth and function [1–3]. This receptor is also the target of stimulating or blocking autoantibodies in patients with autoimmune diseases [4,5]. The TSHR belongs to a particular subgroup of G protein-coupled receptors, inclu- ding the FSH and LH receptors [6]. They are characterized by the presence of a seven transmembrane domain and a large extracellular domain involved in high affinity hormone binding. These three receptors are mainly coupled to G s , leading to the activation of the adenylate cyclase pathway. However, unlike the gonadotropin receptors, the TSHR transduces a signal via adenylate cyclase even in the absence of ligand, thus having a weak constitutive activity [7]. The TSHR undergoes a unique post-translational mat- uration among G protein-coupled receptors. In human thyroid membranes, intramolecular cleavage occurring at the cell surface generates two subunits: an approxi- mately 53 kDa a extracellular subunit, and a wide approxi- mately 33–42 kDa b-transmembrane and intracellular subunit (called A- and B-subunits, according to the terminology of Rees-Smith et al. [4,8]), held together by disulfide bridges [9]. A similar maturation is also observed in an L cell line stably transfected with the human TSHR cDNA [10]. In addition, the extracellular domain of the receptor is shed from the cell surface of thyroid and transfected L cells [11]. The shedding is an enzymatically catalyzed process, where the disulfide bonds reduction is performed by cell-surface associated protein disulfide iso- merase [12]. Other studies suggested that cleavage was required for the formation of TSHR dimers and higher order complexes [13]. PDI activity may also account for disulfide bonding of TSHR oligomers. Immunopurification and microsequencing of the b-sub- units in thyroid membranes and transfected L cells led to the identification of the cleavage sites of the TSHR. In fact, multiple cleavage sites exist [14,15]. They are unrelated and located in a specific extracellular region of the receptor (E3) that displays no homology with the LH and FSH receptors [15–17]. In transfected L cells and in thyroid membranes, the Correspondence to M. Misrahi, Inserm E120, Baˆ timent Gregory Pincus, Hoˆ pital Biceˆ tre, 94275, Le Kremlin Biceˆ tre, France. Fax: + 33 1 45213822, Tel.: + 33 1 45212746/49591828, E-mail: misrahi@kb.inserm.fr Abbreviations: ABTS, 2,2¢-azine-di(ethylbenzthiazoline sulfonate); bTSH, bovine TSH; FSH, follicle stimulating hormone; GPCR, G protein-coupled receptor; LH, luteinizing hormone; TSH, thyrotropin hormone; TSHR, thyroid stimulating hormone receptor. (Received 25 March 2003, revised 7 June 2003, accepted 12 June 2003) Eur. J. Biochem. 270, 3486–3497 (2003) Ó FEBS 2003 doi:10.1046/j.1432-1033.2003.03718.x most N-terminal site mapped is upstream Ser314 and the most C-terminal site detected is upstream Leu378 [15]. The cleavage reaction is sequential and leads to the processive digestion of the specific extracellular region of the TSHR. The enzyme involved in the maturation of the TSHR shares similarities with the ADAM (A Disintegrin and Metallo- protease) family of metalloproteases [12,18]. The cleavage and shedding of the human TSHR may be of physiological importance because the precise quantifica- tion of each subunit in thyroid membranes demonstrated an approximately 2.5 to 3-fold excess of b-overa-subunits [9]. This observation led us to postulate that the a-subunit might be shed from cell membranes and released into the extracellular space or bloodstream. Soluble forms of the TSHR have been described in human thyroid homogenates [19], or proposed in human blood [20,21]. Accumulation of the a-subunit has also been detected in the extracellular matrix [22]. There have been several difficulties in understanding the functional role of receptor cleavage and shedding. Some mutations or small deletions of individual or groups of amino acids did not prevent receptor cleavage, due to the multiplicity of the cleavage sites [23–26]. In addition, the TSHR is expressed in transfected cells in multiple processed and unprocessed forms [10,17]. Monomeric precursors also accumulate in transfected cells [10]. Furthermore, the construction of mutant receptors deleted from various parts of the extracellular domain led to contradictory results: in a first study, one mutant was found to be nonfunctional [27]. More recently, other mutants were shown to display a constitutive activity [28,29]. The latter result led to the proposition of an inhibitory role of the ectodomain on the transmembrane domain [29]. However, because of the lack of adequate immunological tools to trace the receptor, these results were obtained with modified tagged mutant receptors. Moreover, the intracellular traffic of the TSHR has been described [30], but the trafficking of the remaining b-subunits is still unknown. Such trafficking influences the number of molecules present at the cell surface. To understand the role of cleavage and shedding, we constructed deletion mutants, starting at the most N- or C-terminal cleavage sites that we had previously mapped in the divergent E3 region of the receptor [15]. We studied the function and trafficking of such truncated receptors, mimicking cleaved and shed receptors. More- over, we re-evaluated the function of a mutant receptor lacking almost the entire extracellular domain of the TSHR, including the highly conserved region close to the membrane [28]. We also studied a TSHR already constitutively activated by a transmembrane point mutation [31], to know whether cleavage and shedding would modify its function. Experimental procedures Materials DMEM, L -glutamine and gentamycin were from Gibco BRL (Invitrogen Corporation, Paisley, UK); fetal bovine serum was from Biochrom. Bovine thyrotropin hormone (bTSH; 2 IUÆmg )1 ), 3-iso-butyl-1-methylxanthine (IBMX), gelatin, BSA (bovine serum albumin; fraction V) and monensin were obtained from Sigma. 2,2¢-Azine-di(ethyl- benzthiazoline sulfonate) (ABTS) was obtained from Per- bio. Superfect transfection reagent was from Qiagen. [ 125 I]Streptavidin, cAMP-RIA assay kit, peroxidase-conju- gated sheep anti-(mouse IgG) Ig and [ 125 I]Streptavidin (specific activity 20–50 lCiÆlg )1 ) were from Amersham Pharmacia. Alexa488-labeled anti-mouse IgG was from Molecular Probes (Netherlands). Anti-TSHR monoclonal Igs Monoclonal Igs T3-365 and R5T-34 have been described previously [9,15]. Those Igs were raised against fragments of the TSHR expressed in Escherichia coli. The T3-365 Ig is raised against an epitope localized in the intracellular domain of the TSHR [9], and the R5T-34 Ig recognizes an epitope localized between amino acids 357 and 369 [15]. Expression vectors encoding deletion mutants of TSHR The vector encoding the human wild-type TSH receptor cDNA [pSG5-hTSHR] has previously been described [10]. Polymerase chain reaction was used to generate a deletion from position +65 to position +940 (+1 being the first base of the initiation codon), yielding a deletion of amino acids 22–313 in the TSHR, to obtain the b1mutant receptor. The N-terminus of this mutant receptor starts at Ser314, after cleavage of the signal peptide. The four following oligonucleotides, O1 to O4, were used in a polymerase-chain reaction to generate a 348 base-pair fragment (position ) 97–1126, +1 being the first base of the initiation codon). O1: TGG GCA ACG TGC TGG TTA T (position +973); O2: GTC CCT GGA CCC GCC TAG ACA CTT ACG GAA CTT ATC GG (position +1118); O3: CAG GGA CCT GGG CGG ATC TGT GAA TGC CTT GAA TAG CC (position +2010); O4: CTCGAGTTTTTGGGGGGTCCTTC(position +2175). This fragment was digested with EcoRI (position )26) and SacI (position +1105), purified and cloned into the pSG5-hTSHR vector previously digested with the same enzymes, yielding the pSG5-b1 expression vector. To construct the expression vector encoding the b2 mutant receptor, starting at residue Leu378, an oligonucleo- tide encoding amino acids 7–21 and 378–380 was cloned into the pSG5-b1 expression vector, digested with PstI (position +2301) and HindIII (position +1136). In the same way, for generating the expression vector encoding the TM409 mutant receptor, starting at residue Glu409, an oligonucleotide encoding amino acids 7–21 and 409–412 was cloned into the pSG5-b1 expression vector, digested with PstIandBbsI (position +1232). The same strategy was used to construct the deletion mutants containing the Met453Thr mutation, starting from the pSG5-hTSHR-M453T vector, previously described [31]. All the constructs were verified by double-strand DNA sequencing. Cell culture and transfection COS-7 cells were maintained in DMEM with L -glutamine supplemented with 10% fetal bovine serum and 8 lgÆmL )1 Ó FEBS 2003 Role of cleavage and shedding in TSHR function (Eur. J. Biochem. 270) 3487 gentamycin at 37 °C in a humidified 5% CO 2 atmosphere. L cells stably expressing the wild-type and mutant receptors were maintained in DMEM supplemented with 10% fetal bovine serum, L -glutamine and 200 lgÆmL )1 G418, as described [10]. For transient transfection, COS-7 cells were seeded in six-well plates and grown overnight in DMEM supple- mented with 10% fetal bovine serum. They were trans- fected with the Superfect reagent according to the manufacturer’s instructions. Two days after the transfec- tion, cells were assayed for cAMP determination or R5T-34 monoclonal Ig binding experiments. Transfection efficiencies were verified by immunocytochemistry, using the T3-365 monoclonal Ig [9], directed against an intracel- lular epitope of the receptor. Western blot analysis COS-7 cells were grown on 10-mm dishes and transfected as described above. Forty-eight hours later, membrane extracts were prepared as described [15]. The extracts were loaded on a 10% SDS/PAGE and Western blot analysis were performed as described [15] using the monoclonal Ig T3-365 (5 lgÆmL )1 ) directed against the intracellular domain of the TSHR. This Ig recognizes all mutant receptors, including TM409, devoid of almost all extracel- lular sequences. Cellular ELISA COS-7 cells were plated on six-well plates and transfected with the expression vectors encoding the wild-type, b1, b2 and TM409 receptors as described above. Two days after transfection, the cells were fixed for 15 min in 3% paraformaldehyde in NaCl/P i . After washing, the alde- hyde groups were quenched with 50 m M NH 4 Cl in NaCl/ P i for 20 min. After 1 h saturation and permeabilization with NaCl/P i , 1% BSA, 0,1% saponin, the cells were incubated for 2 h with the T3-365 monoclonal Ig (5 lgÆmL )1 in NaCl/P i , 1% BSA, 0,01% saponin). The cells were then washed with NaCl/P i ,1%BSA,0,1% Tween20 and incubated for 1 h with peroxidase-conju- gated sheep anti-(mouse IgG) Ig (dilution 1 : 1000). After washing with NaCl/P i ,400lL ABTS was added into each well and incubated for 15 min under dark. Absorbances were read at 450 nm [28]. cAMP assay COS-7 cells transfected with the expression vectors encoding the wild-type or the truncated receptors were washed twice with DMEM medium containing 20 m M Hepes, pH 7.4, and gelatin 1 mgÆmL )1 at 37 °C. Each dish was then incubated for 1 h at 37 °Cwiththesame medium containing IBMX (0.5 m M ) and 0–100 IUÆL )1 of bTSH. The incubation was stopped by aspiration of the medium and addition of 400 lLÆwell )1 of 1 M perchloric acid. The cells debris were then collected by centrifugation at 15 000 g for 5 min at 4 °C. The resulting supernatants were neutralized with 0.72 M KOH and 0.6 M KHCO 3 . cAMP accumulation was measured by radioimmunoassay [22]. Immunofluorescence and confocal microscopy Indirect immunofluorescence was performed on COS-7 cells expressing the wild-type TSHR and the TM409 mutant receptors, grown on glass culture chambers (Nalge Nunc International), as described [9]. The cells were fixed for 15 min in 3% paraformaldehyde in NaCl/ P i . After washing, the aldehyde groups were quenched with 50 m M NH 4 Cl in NaCl/P i for20min.After1hof saturation and permeabilization with NaCl/P i ,1%BSA, 0.1% saponin, cells were incubated for 2 h with the monoclonal Ig T3-365 [9] (5 lgÆmL )1 in NaCl/P i ,1% BSA, 0.01% saponin). The cells were then washed with NaCl/P i , 1% BSA, 0.1% Tween20 and incubated for 1 h with a 1 : 400 dilution of Alexa488-labeled antimouse IgG. After washing, the cells were mounted with Fluor- escent Mounting Medium. They were examined with a Zeiss LSM-510 confocal scanning laser microscope equipped with a 25 mW Argon laser, using a Plan Apochromat 63 · objective (NA 1.40, oil immersion). Green fluorescence was observed with long pass 505 nm emission filter, under 488 nm laser illumination. The pinhole is set at 1.0 Airy unit. Stacks of images were collected every 0.4 lm along the z-axis. Projections of z median optical slices were projected for each receptor. Moreover, Differential Interference Contrast (DIC, or Nomarski) was used to visualize individual cells [32]. Quantification of receptor–Ig complexes at the cell surface COS-7 cells were seeded in six-well plates and transfected as described above. Two days after, cells were washed with DMEM supplemented with 0.1% BSA and 20 m M Hepes buffer, pH 7.4 (incubation medium) and incubated at room temperature for 15 min, then cooled for 10 min at 4 °Cin the same cold incubation medium. Cells were then incuba- ted with 0.25 mL of incubation medium containing 2 lgÆmL )1 of biotinylated R5T-34 monoclonal Ig as described [30] at 4 °C for 30 min. Next, after removal of the unbound Ig, cells were incubated with 0.4 mL of cold incubation medium containing [ 125 I]Streptavidin at 2ngÆmL )1 at 4 °C for 30 min. After four washes with cold NaCl/P i , cells were trypsinized, collected, and the cell- associated radioactivity, corresponding to receptor-Ig com- plexes present at the cell surface, was measured using a c-counter [30]. Experiments were performed at least three times in duplicate. Normalization of cAMP accumulation to cell-surface expression Basal cAMP accumulation was normalized to cell surface expression for wild-type, b1, M453T and b1-M453T receptors. For that purpose, the receptor-dependent cAMP accumulation (in nmolÆL )1 ) was divided by the radioactivity measured (in c.p.m.), corresponding to cell surface receptor-biotinylated Ig complexes, revealed by [ 125 I]Streptavidin: (cAMP in receptor-transfected cells – cAMP in control pSG5-transfected cells)/(binding of receptor-transfected cells – binding of the control pSG5- transfected cells). The values (relative constitutive activities) 3488 M. Quellari et al. (Eur. J. Biochem. 270) Ó FEBS 2003 were then normalized to the value of the specific consti- tutive activity of the wild-type TSHR, arbitrarily set to 1 [29,33]. Internalization and recycling of receptor-Ig complexes L cell lines stably expressing the wild-type and the b1, the M453T and the b1-M453T receptors were obtained using the calcium phosphate precipitation method and main- tained as described [10]. Internalization of receptor-Ig complexes was measured as described [30]. Briefly, cells were seeded in six-well plates, and grown overnight. Cells were washed with incubation medium (DMEM supple- mented with 0.1% BSA and 20 m M Hepes buffer, pH 7.4) and incubated at room temperature for 15 min, then cooled 10 min at 4 °C in cold incubation medium. They were then incubated with 0.25 mL of incubation medium containing 2 lgÆmL )1 of biotinylated R5T-34 monoclonal Ig [30] at 4 °C for 30 min. In some cases, monensin 40 l M was added to the incubation medium. Next, after removal of the unbound Ig, cells were washed three times with cold incubation medium, once with prewarmed incubation medium and incubated at 37 °C for various periods of times. Cells were incubated with 0.4 mL of cold incuba- tion medium containing [ 125 I]Streptavidin at 2 ngÆmL )1 at 4 °C for 30 min. After four washes with cold NaCl/P i , cells were trypsinized, collected, and the cell-associated radioactivity was measured using a c-counter [30]. Experiments were performed at least three times in duplicate. Statistics Statistical significance was assessed by the Mann–Whitney non-parametric test. Results are expressed as means ± SD. Results Generation of truncated mutants corresponding either to b-subunits of the TSHR, or to a receptor deleted of almost the entire extracellular domain To understand the role of receptor cleavage and shedding, we constructed two deletion mutants, corresponding either to the longest b1 (starting at Ser314) or to the shortest b2 (starting at Leu378) subunit of the TSHR (Fig. 1A). Fig. 1. Schematic representation of the trun- cated mutant receptors. (A) Top panel, human TSHR. The seven transmembrane segments (TM) are shown in gray and the signal peptide is colored in black. The percentage of identity of the different extracellular regions of the TSHR to the corresponding regions of the human LH receptor are indicated above. The E3 region (residues 289–385) is the most divergent region and E5 the most conserved one (residues 403–416). The black arrow indicates the localization of the constitutive mutation Met453Thr (M453T) in the second transmembrane segment of the TSHR [31]. The truncated mutant receptors b1, b2, TM409 starting, respectively, at Ser314, Leu378 and Glu409 are schematized. Note that the N-terminus of the b1andb2mutant receptors originate in the divergent E3 region, whereas the TM409 mutant receptor is deleted of both E3 and most of the E5 regions. I, intracellular domain. (B) Lower panel: West- ern blot analysis of the receptor mutants expressedinCOS-7cells.COS-7cellswere transiently transfected with cDNA encoding the truncated b1, b2, or TM409 mutant receptors. Forty-eight hours after transfection, total cell membrane extracts were prepared (see Experimental procedures) and run on a 10% polyacrylamide gel. Western blot analy- ses were performed using the T3-365 mono- clonal Ig, which recognizes receptor endodomain. Molecular mass standards, in kilodaltons (kDa), are indicated on the left. Lane 1, b1 receptor; lane 2, b2 receptor; lane 3, TM409 receptor. Ó FEBS 2003 Role of cleavage and shedding in TSHR function (Eur. J. Biochem. 270) 3489 We studied also a TM409 mutant receptor (starting at residue Glu409), deleted of 98% of the extracellular region of the TSHR (Fig. 1A). It lacks the E5 region highly conserved in gonadotropin and TSH receptors (approxi- mately 85% of homology with the corresponding part of the LH receptor). This conserved region is present in the b1and in the b2 mutant receptors, which contain, respectively, 102 and 38 amino acids of the extracellular domain of the receptor (Fig. 1A). Western blot analysis of receptors expressed in COS-7 cells were performed using a monoclonal Ig directed against the intracellular domain of the receptor. Figure 1B shows that the truncated receptors are expressed as single mole- cular weights species, with respective apparent molecular weights of approximately 52, 40 and 36 kDa. Effect of deletions corresponding to cleavage and shedding comparative to a deletion of almost the whole extracellular domain of the receptor on its basal activity COS-7 cells transfected with expression vectors encoding either the wild-type or the truncated receptors were incubated with various concentrations of bTSH (0– 10 IUÆL )1 ), and the basal or hormone-induced cAMP levels were measured (Fig. 2A). Transfection efficiencies were verified by immunocytochemistry, using the T3-365 mono- clonal Ig. In addition, total receptor expression was verified using a cellular ELISA ([28] and see Methods) on per- meabilized cells. Figure 2B shows a similar total expression of the wild-type and the truncated b1, b2 and TM409 receptors. As shown in Fig. 2A, the b1andb2 mutant receptors displayed a similar approximately 2.5-fold higher total basal (not normalized) accumulation of cAMP (P <0.01)when compared with the wild-type receptor. This experiment was repeated at least three times with similar results. In contrast, the basal cAMP levels detected in transfected cells expressing the TM409 mutant receptor were signifi- cantly lower when compared with cells expressing the wild- type receptor (P < 0.05) and not significantly different from the values detected in cells transfected with the pSG5 vector alone. Thus, deletion of almost the entire ectodomain of the TSHR in the TM409 mutant receptor suppressed the Fig. 2. cAMP accumulation in COS-7 cells expressing the wild-type or the truncated mutant receptors. (A) COS-7 cells were transiently transfected either with the vector alone (pSG5) or with expression vectors encoding the wild-type (WT) or the truncated b1, b2, and TM409 receptors. Later (48 h), they were incubated (black bars) or not (white bars). One hour, 10 IUÆL )1 of bTSH, and cAMP accumulation was measured. The data presented are expressed as raw values (intracellular cAMP accumulation in nmolÆL )1 ), and represent the mean ± SD of triplicate wells from a representative experiment of three independent experiments. The following P-values were calculated for total basal cAMP accumulation, when compared with the wild- type receptor: b1 receptor, P <0.01;b2 receptor, P <0.05;TM409 receptor, P < 0.05. (B) Cellular ELISA was performed to compare total cellular expression of the different constructs. Briefly, COS-7 cells were transiently transfected either with the vector alone (pSG5) or the wild-type, b1, b2, and TM 409 mutant receptors. Forty-eight hours later, they were fixed, permeabilized and incubated with the T3-365 monoclonal Ig that recognizes an intracellular epitope of the TSHR. After incubation with a peroxidase-conjugated sheep anti-(mouse IgG), then with ABTS, OD was read at 450 nm. (C) Dose–response to TSH of the wild-type, b1, b2 and TM409 receptors. COS-7 cells were transiently transfected either with the pSG5 vector alone (·)orwith expression vectors encoding the wild-type (WT, j) or the truncated b1 (h), b2(r), and TM409 (s) receptors. Forty-eight hours later, they were incubated for 1 h with 0–100 IUÆL )1 of bTSH, and cAMP accumulation was measured. The data presented are expressed as raw values (intracellular cAMP accumulation in nmolÆL )1 ), and represent the mean ± SD of triplicate wells from a representative experiment. 3490 M. Quellari et al. (Eur. J. Biochem. 270) Ó FEBS 2003 basal constitutive activity detected for the wild-type receptor. In all cases, incubation of cells transfected with expression vectors encoding the truncated b1, b2orTM409mutants with bTSH (0.1, 1, 10 or 100 IUÆL )1 ), did not enhance the accumulation of cAMP (Fig. 2C), contrary to the wild-type receptor. Cell-surface expression of the TM409 mutant receptor As the TM409 mutant exhibited a complete loss in receptor function, we also verified the cell surface expression of this truncated receptor, comparatively to the wild-type receptor. For that purpose, we performed indirect immunofluores- cence using a T3-365 monoclonal Ig directed against the intracellular domain of the receptor. Cell surface expression was observed using confocal microscopy and the Nomarski differential interference contrast. As shown in Fig. 3A (top panel), in COS-7 cells expressing the wild-type receptor, staining is observed at the cell surface at the leading edge of lamellipodia, as previously described [22]. An intracellular staining of the cells is also observed, probably correspond- ing to the precursor protein that has been shown to accumulate in the endoplasmic reticulum [10]. For cells expressing the TM409 receptor, staining was also observed at the leading edge of lamellipodia, showing that receptor molecules are indeed present at the cell surface (Fig. 3A, lower panel). However, receptors also accumulate inside the cell, showing that some TM409 receptor molecules are also trapped intracellularly. DIC (Fig. 3B) analysis merged with the fluorescence (Fig. 3C) allowed confirmation of cell surface expression of thewild-typeandtheTM409receptors. Specific basal cAMP-stimulating activity of the wild-type and the b1 mutant receptors To precisely quantify the enhancement in receptor activity, we compared the cell surface expression of the wild-type and the b1 mutant receptors. For this purpose, we used in both cases the same biotinylated R5T-34 anti-TSHR monoclonal Ig, as described [30]. Indeed, this Ig has previously been used to study the intracellular traffic of the wild-type TSHR, by monitoring the disappearance of cell surface associated receptor-Ig complexes. This Ig recognizes an epitope localized between amino acids 357 and 369 [15], in the E3 specific region of the receptor. It does not interfere with receptor function and trafficking ([30], and data not shown). Measurement of the R5T-34 monoclonal Ig binding is thus a convenient tool with which to trace and quantify the cell surface expression of the receptors. To quantify the cell surface expression of the b1 mutant receptor, comparatively to the wild-type receptor, COS-7 cells were transfected with each expression vector and incubated at 4 °Cwiththe Fig. 3. Confocal microscopy of COS-7 cells expressing either the wild-type or the TM409 receptors. Transfected cells were fixed, permeabilized, incubated with the T3-365 monoclonal Ig that recognizes an intracellular epitope of the TSHR, and indirect immunofluorescence was performed (see Experimental procedures). Top panel, wild-type TSHR (WT); lower panel, TM409 receptor. (A) Confocal microscopy study; (B) Nomarski optics was used to study cell morphology; (C) fluorescence image was overlaid on Nomarski image to generate merged image. Ó FEBS 2003 Role of cleavage and shedding in TSHR function (Eur. J. Biochem. 270) 3491 biotinylated R5T-34 anti-TSHR Ig during 30 min. After washing the cells, the concentration of receptor-Ig com- plexes at the cell surface was quantified by measuring the binding of [ 125 I]Streptavidin [30]. No Ig binding was observed in cells transfected with the vector alone (Fig. 4B). As shown in Fig. 4B, quantification of cell surface expression of COS-7 cells expressing the b1 mutant receptor revealed an approximate threefold decrease in cell surface expression when compared with the wild-type receptor. Therefore, the increase in the specific constitutive activity of the b1 receptor, after normalization to cell surface expres- sion, is at least approximately eightfold (mean of three independent experiments, including the data presented in Fig. 2A, see also Fig. 4C). Increased basal internalization of the b1 mutant receptor A decreased cell surface expression of the b1mutant receptor was observed when compared with the wild-type receptor (Fig. 4B). We wondered whether shedding of the ectodomain would modify the trafficking of the receptor. Therefore, we studied the comparative internalization of the wild-type and of the b1 receptors, using the R5T-34 anti- TSHR monoclonal Ig, as previously described [30]. L cell lines expressing the wild-type or the b1mutant receptors were incubated at 4 °C with the Ig in the presence or in the absence of bTSH. After washing the unbound Ig, the cells were incubated for different periods of times (0–60 min) at 37 °C. The concentration of the biotinylated Ig remaining on the cell surface was then quantified by measuring the binding of [ 125 I]Streptavidin. In the absence of bTSH, a very weak internalization of the TSHR is detected (approximately 10%) ([30] and data not shown). In the presence of bTSH, approximately 30% of the wild-type receptor molecules are internalized after 15 min ([30] and Fig. 5). As previously described [30], the majority (approximately 90%) of receptor molecules were recycled back to the cell surface after 30 min. Incubation with monensin confirmed that the recovery of cell surface receptors was the result of receptor recycling. For the b1 mutant receptor, there was a marked constitutive internalization of the receptor. Indeed approxi- mately 45% of the receptor molecules were internalized after 15 min, in the absence of hormone. Furthermore, no recycling of the b1 mutant receptor was detected even after incubation for 60 min at 37 °C. Addition of bTSH or monensin did not modify the intracellular traffic of the b1 mutant receptor (data not shown). The traffic of the b1 mutant receptor was similar to the traffic observed for the wild-type TSHR in the presence of bTSH and monensin (Fig. 5). Effect of cleavage and shedding on a constitutively activated receptor Natural point mutations of the TSHR have been described in familial hyperthyroidism or toxic adenomas [7,34,35]. These mutations are mainly located in the transmembrane domain of the receptor and lead to a constitutive activation of the receptor. We evaluated the functional consequences of receptor shedding in a mutant harboring a constitutive natural Fig. 4. Specific constitutive activity of the wild-type and the truncated b1 receptors. (A) Total basal cAMP accumulation was measured in COS-7 cells transfected with the control pSG5 expression vector or vectors encoding the wild-type (WT) and the b1 mutant receptors. The data are expressed as raw values (intracellular cAMP accumulation in nmolÆL )1 )andrepresentthemean±SDoftriplicatewellsfroma representative experiment of three independent experiments. (B) Quantification of receptor-bound Ig complexes was performed using R5T-34 monoclonal Ig as described [30]. Briefly, transfected cells were incubated at 4 °C with the biotinylated R5T-34 Ig. The receptor-Ig complexes present at the cell surface were quantified using [ 125 I]Streptavidin. The data, expressed in percentage of cell surface expression, represent the mean ± SD of triplicate wells from a rep- resentative experiment of three independent experiments. (C) Relative specific constitutive activity of the wild-type TSHR (WT) (arbitrarily setto1)andoftheb1 mutant receptor: normalization of cAMP accumulation to cell-surface expression was performed (see Experi- mental procedures). Presented data are the mean ± SD of three independent experiments. 3492 M. Quellari et al. (Eur. J. Biochem. 270) Ó FEBS 2003 transmembrane mutation M453T (M453T-TSHR) [31]. To discover whether the deletion of the ectodomain would cause an additional increase in receptor function, we constructed two supplementary mutant receptors b1- M453T and b2-M453T, corresponding, respectively, to shed M453T-TSHR after cleavage at the most N- or C-terminal sites (Fig. 1). We also constructed a TM409- M453T mutant deleted of almost the entire extracellular domain of the receptor. To study comparative receptor activity, COS-7 cells were transfected with expression vectors encoding the wild-type or the mutant receptors. As shown in Fig. 6, the M453T mutant receptor exhibited an approximately sevenfold increase in basal cAMP accumulation, when compared with the wild-type receptor, as described previously [31]. COS-7 cells expressing the truncated b1-M453T and b2- M453T mutant receptors displayed a similar approximately twofold higher total basal accumulation of cAMP (P < 0.01 and P <0.05 for b1-M453T and b2-M453T, respectively) when compared with the wild-type receptor (Fig. 6). However COS-7 cells expressing the truncated TM409-M453T exhibited cAMP accumulation that did not differ significantly from cells transfected with the pSG5 vector alone. In all cases, incubation of COS-7 cells expressing the truncated b1, b2 or TM409-M453T mutant receptors with bTSH (10 IUÆL )1 ) did not enhance the accumulation of cAMP (Fig. 6). When the cell surface expression of the wild-type, the M453T and the b1-M453T truncated receptors was studied using the R5T-34 anti-TSHR monoclonal Ig, a diminished expression of the M453T receptor at the cell membrane of approximately 1.7-fold when compared with the wild-type receptor was observed (Fig. 7B). Its normalized constitutive activity is thus increased at least approximately eightfold (mean of three independent experiments), when compared with the wild-type receptor. Likewise, the expression of the truncated b1-M453T mutant receptor was strongly reduced by 7.7-fold (Fig. 7B), and thus the specific constitutive activity is increased by at least approximately fourfold, when compared with the M453T receptor (Fig. 7C). Increased basal internalization of the M453T receptor, further enhanced by deletion of its ectodomain We also studied the trafficking of the constitutive M453T receptor, which also exhibited a diminished cell surface expression. Therefore, we established L -cell lines expressing either the M453T holoreceptor or the truncated b1-M453T mutant receptor. As shown in Fig. 8, in the absence of hormone, the constitutive M453T receptor exhibited an increased basal internalization when compared with the wild-type receptor, with approximately 40% of receptor molecules being internalized after 20 min at 37 °C, and approximately 50% after 60 min. Addition of bTSH did not significantly enhance the internalization of the M453T mutant receptor (data not shown). However, no recycling was detected in the absence or in the presence of TSH, even after longer time Fig. 6. cAMP accumulation in COS-7 cells expressing a constitutively activated M453T receptor and the corresponding truncated M453T mutant receptors. COS-7 cells were transiently transfected either with the vector alone (pSG5) or with expression vectors encoding the wild- type (WT), the M453T, and the truncated b1-M453T, b2-M453T, TM409-M453T mutant receptors. Forty-eight hours later, they were incubated (black bars) or not (white bars) 1 h with 10 IUÆL )1 of bTSH, and cAMP accumulation was measured. The data presented are expressed as raw values (intracellular cAMP accumulation in nmolÆL )1 )andrepresentthemean±SDoftriplicatewellsfroma representative experiment of three independent experiments. The fol- lowing P-values were calculated for basal cAMP accumulation, when compared with the wild-type receptor: M453T receptor, P <0.01;b1- M453T receptor, P <0.01;b2-M453T receptor, P < 0.05; TM409- M453T receptor, P < 0.05]. Fig. 5. Internalization of the wild-type and the truncated b1 mutant receptors. L cells stably transfected with expression vectors encoding either the wild-type or the truncated b1(j) receptors were incubated with biotinylated R5T-34 monoclonal Ig. For cells expressing the wild- type receptor, bTSH (10 IUÆL )1 )(s) or bTSH + monensin (40 l M ) (m) were also added to the incubation medium. After removal of unbound Ig, cells were incubated for the indicated times at 37 °C. Surface-bound Ig was quantified as described [30] by measuring the binding of [ 125 I]Streptavidin. Specific binding at each point was nor- malized with reference to specific binding before the incubation at 37 °C (0-min point) to derive the fraction of initial total receptor-Ig complexes remaining at the cell surface. Bars, SD of duplicate points. The assay shown is representative of an experiment repeated at least three times. Ó FEBS 2003 Role of cleavage and shedding in TSHR function (Eur. J. Biochem. 270) 3493 periods up to 120 min. In keeping with this observation, incubation with monensin did not modify the trafficking of the receptor (Fig. 8). Deletion of the ectodomain of the receptor (b1-M453T mutant receptor) led to a supplementary increase in the basal internalization, approximately 50% of receptor mole- cules being internalized after 15 min, and approximately 70% after 60 min (Fig. 8). No receptor recycling was observed. bTSH and monensin had no effect on the intracellular traffic of the mutant receptor (data not shown). Discussion In this study, we focused on the role of TSHR cleavage and shedding, by constructing mutant receptors corresponding to the longest (b1) and shortest (b2) b-subunits that we had previously mapped in thyroid and L cells [15]. Altogether, our results show that mutants corresponding to cleaved and shed receptors (at the most N- or C-terminal sites) display an increased constitutive activation of the corresponding b-subunits of the TSHR. By contrast, a mutant receptor lacking almost the entire extracellular domain of the TSHR (TM409 mutant) including the highly conserved region close to the membrane, exhibited a complete loss in receptor function. Thus we extend previous studies which support a role for the extracellular domain in the stabilization of an inactive form of the unliganded receptor [29]. Indeed, only cleavage in the E3 domain of the receptor, followed by receptor shedding, can enhance receptor activity. Recently, Chen et al. reported that TSHR cleavage, by itself, was insufficient to enhance ligand-independent constitutive activity [36]. The complete loss of the extracellular domain yields a complete loss in receptor function. This observation is in agreement with previous data, which highlighted a role for the conserved E5 region, close to the transmembrane domain, in TSH or LH receptor function [37–39]. It was proposed that this invariant sequence in the glycoprotein hormone receptors is required for proper folding, trafficking and ligand-mediated signaling but not for ligand binding [37,40]. The complete loss in the constitutive activity of the Fig. 8. Internalization of the wild-type, M453T and truncated b1-M453T mutant receptors. L cells stably transfected with expression vectors encoding either the wild-type (s), the M453T receptor (j)or the b1-M453T (d) receptors were incubated with biotinylated R5T-34 monoclonal Ig. For cells expressing the M453T receptor, monensin (40 l M ) was also added to the incubation medium (m). After removal of unbound Ig, cells were incubated for the indicated times at 37 °C. Surface-bound Ig was quantified as described ([30] and see Fig. 7) by measuring the binding of [ 125 I]Streptavidin. Bars, SD of duplicate points. The assay shown is representative of an experiment repeated at least three times. Fig. 7. Specific constitutive activity of the wild-type, M453T and b1-M453T receptors. (A) Total basal cAMP accumulation was meas- ured in COS-7 cells transfected with the control pSG5 expression vector or vectors encoding the wild-type (WT), M453T and b1-M453T receptors. The data are expressed as raw values (intracellular cAMP accumulation in nmolÆL )1 )andrepresentthemean±SDoftriplicate wells from a representative experiment of three independent experi- ments. (B) Quantification of receptor-bound Ig complexes was per- formed using R5T-34 monoclonal Ig as described [30], and see Fig. 4. The data, expressed in percentage of cell surface expression, represent the mean ± SD of triplicate wells from a representative experiment of three independent experiments. (C) Relative specific constitutive activity of wild-type (WT) (arbitrarily set to 1), M453T and b1-M453T receptors: normalization of cAMP production to cell-surface expres- sion was performed (see Experimental procedures). Presented data are the mean ± SD of three independent experiments. 3494 M. Quellari et al. (Eur. J. Biochem. 270) Ó FEBS 2003 M453T receptor, when this region is deleted (TM409- M453T receptor), also strongly supports a role for this juxtamembrane region in receptor function and cell surface targeting. We could detect the presence of the TM409 truncated receptor at the cell surface by confocal micros- copy. However some mutant receptor molecules also accumulate inside the cell. We could not reproduce the results obtained by Zhang et al. [28] with a mutant receptor E409, similar to our TM409 mutant, but also including a hemagglutinin (HA) tag sequence. This group detected a basal activity three times higher (not normalized) when compared with the wild-type TSHR. Vlaeminck-Guillem et al. [29] have also constructed truncated mutants devoid of almost the entire extracellular domain of the receptor, and all including a rhodopsine tag localized at the N-terminus to improve and quantify cell surface expression. However, the addition of this tag led to a marked increase in the apparent molecular weight of the TSHR (approximately 40 kDa), due to the introduction of a supplementary glycosylation site [29]. Modification of the structure of the receptor might be responsible for the increased activity detected for such truncated receptors. The same group has found that a truncated mutant containing five amino acids of the extracellular domain and devoid of any tag was not functional [27], while a mutant containing four amino acids of the ectodomain and including a rhodopsine tag was found constitutively activated [29]. In agreement with our results, previous mutations or deletions within the E5 highly conserved region yielded a loss in receptor activity [37,38]. Deletion of receptor ectodomain yielded a diminished expression of the b1 receptor at the cell surface. We thus studied b1 receptor trafficking. While the unliganded wild-type TSHR exhibited a very limited basal internal- ization [30], deletion of the ectodomain in the b1 receptor led to a marked increased basal internalization of the receptor. Thus, the extracellular domain of the TSHR negatively modulates receptor internalization probably through a conformational change transmitted to the b-subunit of the receptor. Deletion of the ectodomain, or addition of TSH, relieves constrained conformations and increases internalization. As the b1 mutant is constitu- tively activated, this observation suggests a link between the conformations necessary for receptor activation and for internalization. A similar situation has been described for other G protein-coupled receptors [41]. However, no recycling was observed for the b1 receptor. Its intracel- lular traffic is very similar to the one of the wild-type receptor activated by TSH, but in the presence of monensin, which inhibits the recycling. This observation strongly argues for a different conformation between the receptor activated by cleavage and shedding on the one hand, and the receptor activated by the ligand on the other hand. This may be due to different post-transla- tional maturation of the receptors, leading to different conformations of the sequence(s) implicated in the recycling of the receptor. The TSHR can also be activated by natural constitutive point mutations found mainly in the transmembrane domain of the receptor [7,34,35]. The current hypothesis maintains that constitutively activated receptors release the conformational constraints of the GPCR inactive state that normally keep the ligand-free receptor silent. Therefore, we wondered whether shedding of the ectodomain of receptors already activated by such a mutation, M453T [31], would lead to a supplementary activation of the corresponding b-subunits. Accordingly, cleavage and shedding of the M453T mutant receptor further increase its activity. Our results show that the extracellular domain still exerted a negative effect on the receptor already constitutively activated by a transmembrane point mutation. This inhibition may be released in part by the hormone, as TSH still up-regulates to some extent the activity of the M453T receptor, or by receptor shedding. Deletion of the ectodomain of other constitutively active mutants (muta- tion D633A located in the sixth transmembrane segment, or A623I in the third intracellular loop of the receptor) did not yield a supplementary increase in receptor activity [29]. Some constitutively activated GPCRs have been shown to be constitutively internalized [41–43]. Therefore, we studied the intracellular trafficking of receptors constitu- tively activated by a transmembrane point mutation. Accordingly, the activated M453T receptor exhibited a marked enhanced constitutive internalization when com- pared with the wild-type receptor. The underlying molecular mechanism that drives the rapid internalization of a constitutive receptor is not understood, but a link between the active conformation and the conformation necessary for receptor internalization has been already proposed [41]. For the LH receptor, the increased internalization has been proposed to be linked to receptor phosphorylation and/or interaction with arrestins, or due to an easier clustering of the activated receptors in coated pits [44]. It has to be noted that in the case of the M453T mutant receptor, contrary to the wild-type receptor, no recycling was observed. This was confirmed by the use of monensin, which did not modify the trafficking of this mutated receptor. These observations strongly suggest a difference in the conformation of the M453T receptor when compared with the hormone-activated receptor. Deletion of the ectodomain of the M453T receptor, mimicking receptor shedding, led to a supplementary increase in receptor internalization, indicating probably a supplementary change in receptor conformation. It has been proposed that each function of the receptor (G protein coupling, internalization, recycling) is not triggered by only one well-defined conformation, but by a continuum of independent conformations [41,45]. In conclusion, cleavage and shedding yield TSHR activation but also increase receptor downregulation through an increased internalization of the b-subunits of the receptor, the latter mechanism limiting simulta- neously excessive receptor signaling. The combined effects may be responsible for the limited basal constitutive activation of the cAMP pathway that is detected for the TSHR. Further studies are necessary to discover whether the shedding may be regulated [46], which might be a novel way to modulate receptor activity. Ó FEBS 2003 Role of cleavage and shedding in TSHR function (Eur. J. Biochem. 270) 3495 [...]... Role of cleavage and shedding in TSHR function (Eur J Biochem 270) 3497 32 Lee, A., Chow, D., Haus, B., Tseng, W., Evans, D., Fleiszig, S., Chandy, G & Machen, T (1999) Airway epithelial tight junctions and binding and cytotoxicity of Pseudomonas aeruginosa Am J Physiol 277, 204–217 33 Nishi, S., Nakabayashi, K., Kobilka, B & Hsueh, A.J.W (2002) The ectodomain of the luteinizing hormone receptor interacts... Targeted restoration of cleavage in a noncleaving thyrotropin receptor demonstrates that cleavage is insufficient to enhance ligand-independent activity Endocrinology 144, 1324– 1330 37 de Roux, N., Misrahi, M., Brauner, R., Houang, M., Carel, J.C., Boumedienne, A., Toublanc, J.E & Milgrom, E (1996) Four families with loss of function mutations of the thyrotropin receptor J Clin Endocrinol Metab 81, 4229–4235... Devauchelle, G & Milgrom, E (1994) Processing of the precursors of the human thyroid-stimulating hormone receptor in various eukaryotic cells (human thyrocytes, transfected L cells and baculovirus-infected insect cells) Eur J Biochem 222, 711–719 11 Couet, J., Sar, S., Jolivet, A., Hai, M.-T.V., Milgrom, E & Misrahi, M (1996) Shedding of human thyrotropin receptor ectodomain J Biol Chem 271, 4545–4552 12 Couet,... constitutive activity of the transmembrane domain of the human TSH receptor: implications for hormone receptor interaction and antagonist design Endocrinology 141, 3514–3517 Vlaeminck-Guillem, V., Ho, S.C., Rodien, P., Vassart, G & Costagliola, S (2002) Activation of the cAMP pathway by the TSH receptor involves switching of the ectodomain from a tethered inverse agonist to an agonist Mol Endocrinol 16, 736–746... Cell surface protein disulfide-isomerase is involved in the shedding of human thyrotropin receptor ectodomain Biochemistry 35, 14800–14805 13 Latif, R., Graves, P & Davies, T.F (2001) Oligomerization of the human thyrotropin receptor: fluorescent protein-tagged hTSHR reveals post-translational complexes J Biol Chem 276, 45217– 45224 14 Chazenbalk, G.D., Tanaka, K., Nagayama, Y., Kakinuma, A., Jaume,... processing of the natural human thyrotropin receptor: demonstration of more than two cleavage sites J Clin Endocrinol Metab 84, 2177–2181 Tanaka, K., Chazenbalk, G.D., McLachlan, S.M & Rapoport, B (1999) Subunit structure of thyrotropin receptors expressed on the cell surface J Biol Chem 274, 33979–33984 Blobel, C.P (2000) Remarkable roles of proteolysis on and beyond the cell surface Curr Opin Cell... presence of the unique, 50 amino acid insertion J Biol Chem 273, 1959– 1963 Tanaka, K., Chazenbalk, G.D., McLachlan, S.M & Rapoport, B (1999) Thyrotropin receptor cleavage at site 1 involves two discontinuous segments at each end of the unique 50-amino acid insertion J Biol Chem 274, 2093–2096 Chazenbalk, G.D., Tanaka, K., McLachlan, S.M & Rapoport, B (1999) On the functional importance of thyrotropin receptor. .. exoloop 2 to constrain the transmembrane region J Biol Chem 277, 3958–3964 34 Wonerow, P., Neumann, S., Gudermann, T & Paschke, R (2001) Thyrotropin receptor mutations as a tool to understand thyrotropin receptor action J Mol Med 79, 707–721 35 Farid, N.R., Kascur, V & Balazs, C (2000) The human thyrotropin receptor is highly mutable: a review of gain -of- function mutations Eur J Endocrinol 143, 25–30 36... An N-linked glycosylation motif from the noncleaving luteinizing hormone receptor substituted for the homologous region (Gly367 to Glu369) of the thyrotropin receptor prevents cleavage at its second, downstream site J Biol Chem 272, 28296–28300 Tanaka, K., Chazenbalk, G.D., McLachlan, S.M & Rapoport, B (1998) Thyrotropin receptor cleavage at site 1 does not involve a specific amino acid motif but instead... J & Puett, D (1999) Characterization of a region of the lutropin receptor extracellular domain near transmembrane helix 1 that is important in ligandmediated signaling Endocrinology 140, 1775–1782 39 Szkudlinski, M.W., Fremont, V., Ronin, C & Weintraub, B.D (2002) Thyroid-stimulating hormone and thyroid-stimulating 40 41 42 43 44 45 46 hormone receptor structure -function relationships Physiol Rev 82, . Role of cleavage and shedding in human thyrotropin receptor function and trafficking Myle ` ne Quellari 1 , Agne ` s. activated by cleavage and shedding on the one hand, and the receptor activated by the ligand on the other hand. Cleavage and shedding of a receptor already