| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Institut National de la Santé et de la Recherche Médicale (INSERM), E120 (I.B., M.-T.G.P., A.D., J.-P.T., M.M.), and Laboratoire d’Hormonologie et Biologie Moléculaire Hôpital de Bicêtre, Institut Fédératif de Recherche 93 (M.M.), 94275 Le Kremlin Bicêtre, France; Unité Mixte de Recherche (UMR) 8125 Centre National de la Recherche Scientifique (CNRS), Institut Gustave Roussy-PR1 (P.D.), 94805 Villejuif Cedex, France; CNRS, UMR 6014, Université de Rouen (E.C.), 76821 Mont Saint Aignan, France; and INSERM, Unité 413, Université de Rouen (J.L., H.V.), 76821 Mont Saint Aignan, France
Address all correspondence and requests for reprints to: Dr. Micheline Misrahi, Institut National de la Santé et de la Recherche Médicale E120, Bâtiment Grégory Pincus, 80 rue du Général Leclerc, 94275 Le Kremlin Bicêtre, France.E-mail: micheline.misrahi{at}bct.ap-hop-paris.fr.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Although many G protein-coupled receptors have been shown to display a polarized expression in epithelial cells (3, 4, 5), the signals involved in their polarized sorting remain largely unknown. The presence of multiple cytoplasmic, transmembrane and extracellular domains in these receptors has been proposed to result in distinct types of hierarchical and competing signals (6).
Gonadotropin [LH (LHR) and FSH (FSHR)] and TSH (TSHR) receptors belong to the large family of G protein-coupled receptors (7, 8). They possess the characteristic seven transmembrane spanning domains, but form a specific subgroup characterized by the presence of a large extracellular domain made up of repeated leucine-rich motifs and are heavily glycosylated. This domain is responsible for high affinity hormone binding. The three receptors are principally coupled to Gs, leading to the hormone-induced activation of adenylate cyclase.
FSHR and LHR have a central role in reproduction through control of the synthesis and secretion of sexual steroids and their role in gamete maturation (7). The TSHR has a key role in the control of thyroid growth and function (8). Natural germinal mutations of these receptors have been described in human hypothyroidism and infertility (8, 9).
A particularity of the TSHR and FSHR is their polarized expression in follicular thyroid and Sertoli cells, respectively (8). In contrast, the LHR is present all over the cell surface of thecal, granulosa, and luteal cells in the ovary and of Leydig cells in the testis (8). This nonpolarized distribution is due to the fact that the LHR is not physiologically expressed in polarized cells.
We have previously expressed all three receptors in Madin-Darby canine kidney (MDCK) epithelial cells (10). This allowed us to reproduce the physiological basolateral targeting of the TSHR and FSHR in these cells. However, we have also shown that the LHR has a polarized basolateral distribution in polarized MDCK cells, suggesting that all three receptors contain a basolateral sorting signal in their structure.
We have initially characterized the basolateral sorting signal of the FSHR (11). It has an unusual localization, occurring in the distal part of the intracellular domain of the receptor at a distance from the membrane. It was the first basolateral sorting signal identified for a G protein-coupled receptor and more generally for a hormone receptor (11).
Few studies have been performed on other G protein-coupled receptors (6, 12, 13), and the basolateral sorting signals of the expanded family of G protein-coupled receptors remain largely unknown. The availability of three related G protein-coupled receptors belonging to the same subgroup provides an advantageous model with which to comparatively analyze the different structural determinants of their polarized targeting. This may enable identification of specific features of basolateral signals for related G protein-coupled receptors.
Another particularity of this receptor subgroup is the existence of soluble forms, deleted of their transmembrane domain by alternative splicing. For the LHR, variant proteins are expressed physiologically in the testis and ovary (7). A soluble form of the TSHR has been detected in human blood (8), and it has been proposed that it was encoded by variant truncated mRNAs. This would necessitate their secretion from the basolateral side of polarized cells. However, the polarized secretion of these soluble variants has not been studied previously. The extracellular domain of those receptors is heavily glycosylated, and N-linked sugars have been reported to correspond to an apical sorting signal in MDCK cells (14). Thus, we have also studied the polarized expression of a truncated TSHR corresponding to its ectodomain.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
). The truncated receptors were stably expressed in MDCK cells, and their distribution was analyzed by confocal microscopy (Fig. 1B) using a monoclonal antibody directed against the extracellular domain of the TSHR. In cells expressing the wild-type receptor, a strong labeling of the membrane was observed on basolateral sections with a reticular pattern characteristic of the basolateral localization. The vertical sections xz of the cells confirmed this basolateral localization, especially with a strong lateral staining. In cells expressing a deleted version of THSR (TSHR
708–764, deleted of the major part of the intracellular domain of the receptor), a strong labeling of the cell surface was detected only on apical sections. The vertical section shows that the labeling is restricted at the top of the cells, confirming apical distribution of the truncated receptor. This suggested that basolateral sorting information was located in the distal part of the intracellular domain of the TSHR.
|
731–764, we observed an apical distribution similar to that of the first mutant, TSHR708–764. In contrast, the mutant TSHR747–764 had mainly a basolateral polarized expression similar to that of the wild-type receptor (Fig. 1B
). Thus, the sequence located between residues 731 and 746 is very likely to contain basolateral sorting information.
To quantify the receptor molecules present at each membrane domain, we performed cell surface immunoprecipitation experiments on polarized monolayers of MDCK cells. After metabolic labeling of the cells, a polyclonal antibody directed against the extracellular domain of the TSHR was added in either the apical or the basolateral compartment of cells (10). Receptor-antibody complexes were then purified and analyzed by gel electrophoresis (Fig. 1C). For the wild-type receptor, approximately 95% of the receptor molecules were immunoprecipitated from the basolateral cell surface. A similar proportion was detected at the basolateral membrane for the TSHR747–764 mutant. However, in the case of the MDCK cells expressing TSHR708–764 and TSHR731–764, only 10–20% of receptor molecules were immunoprecipitated from the basolateral membrane domain; the majority of the receptor molecules were expressed at the apical cell surface.
We verified that the basolateral sorting information mapped in the C-terminal part of the TSHR were also functional in thyroid cells. We thus studied the polarized expression of the wild-type TSHR and two truncated 708–764 and 731–764 receptor mutants in the polarized rat cell line FRT (Fig. 1D). The wild-type receptor has a basolateral localization in these cells, whereas the localization is reversed with the two truncated receptors. We can thus conclude that the sequence located downstream of residue 731 is also implicated in the basolateral localization of the TSHR in thyroid cells.
The Sequence Located between Residues 731–746 Contains an Autonomous Basolateral Sorting Signal Transferable to a Heterologous Protein
We have shown that deletion of a sequence corresponding to residues 731–746 reverses the basolateral localization of the TSHR. Thus, this sequence is very likely to contain a basolateral sorting signal. However, this sequence might have been only part of the signal and might not be sufficient to target the receptor to the basolateral membrane domain. This sequence might also stabilize the receptor at the basolateral cell surface.
To show that the mapped sequence corresponds to an autonomous and independent basolateral sorting signal, transferable to another protein, a chimeric protein was designed with the neurotropin receptor (NTR; Fig. 2A). This receptor presents an apical distribution in MDCK cells. We used a truncated p75 NTR with a deletion after the fifth amino acid of its cytoplasmic tail (NTRt). This construct also presents an apical distribution in MDCK cells (15). The sequence encoding amino acids 729–746 of the TSHR was fused to the cytoplasmic tail of the NTRt (see Materials and Methods).
|
). Although approximately 90% of the p75 NTRt was expressed at the apical membrane domain, only about 10% of the chimeric receptor molecules were detected at the apical cell surface. We can thus conclude that amino acids 731–746 of the TSHR encode an effective basolateral sorting signal. This signal is independent and transferable to a heterologous protein.
The Basolateral Sorting Signal of the TSH Receptor Is Not Dependent on a Tyrosine or a Hydrophobic Doublet
Both TSHR and FSHR basolateral sorting signals are located in the distal part of the intracellular domain of the receptors that share weak primary sequence homology. A tyrosine and a leucine within the signal were shown to be important for the function of the FSHR basolateral sorting signal. A tyrosine and a hydrophobic doublet are also found in the TSHR basolateral targeting signal. To establish whether mutation of these residues alters the function of the autonomous basolateral sorting signal, we also generated chimeric receptors in which the tyrosine and the leucine-isoleucine doublet were mutated into alanines. As shown in Fig. 2B, mutation of these residues did not alter the basolateral targeting of the hybrid receptor. The same mutations were introduced into the sequence of the complete TSHR without altering the expression of the TSHR at the basolateral membrane (data not shown).
The Basolateral Sorting Signal of the TSHR Is Independent of the Endocytosis Signal and Is Not Involved in Gs Protein Coupling
Many basolateral sorting signals have been shown to coincide or overlap with endocytosis signals (2, 16). Therefore, we studied the internalization of [125I]bovine TSH ([125I]bTSH) in MDCK cells expressing either the wild-type or the mutant receptor deleted for the basolateral sorting signal. As shown in Fig. 2C, no alteration of receptor internalization could be detected in the mutant receptor compared with the wild-type receptor.
We then examined the possible involvement of this region of the TSHR in Gs protein coupling. Truncated TSHR708–764, -731–764, and -747–764 were transfected in COS-7 cells. Basal and maximal hormone-induced adenylate cyclase stimulations were similar to those observed with the wild-type receptor (Fig. 2D). Thus, the basolateral sorting signal of the TSHR is independent of its endocytosis signal, and the corresponding sequence is not implicated in Gs coupling.
The C-Terminal Region of the LHR Also Contains a Basolateral Sorting Signal
Despite the lack of sequence homology, the basolateral sorting signals of the FSHR and TSHR are both located in the distal part of the intracellular domain of each receptor. The question was thus raised as to whether the third receptor belonging to this subgroup, the LHR, also contains a basolateral sorting signal in this divergent distal region. Thus, we constructed three receptor mutants deleted from various parts of the intracellular domain of the LHR (Fig. 3A). The distribution of these receptors expressed in MDCK cells was analyzed by confocal microscopy (Fig. 3B). The first mutant LHR651–696, which is deleted for the entire region divergent between gonadotropin receptors and TSHR, exhibited apical localization. The second mutant LHR666–696 also exhibited apical localization, whereas the last mutant, LHR690–696, displayed a basolateral distribution similar to that of the wild-type receptor.
|
), confirming the reverse polarization of the truncated receptors LHR651–696 and -666–696. In contrast, 85–90% of the receptor molecules were detected at the basolateral membrane domain for the wild-type receptor and for the LHR690–696 mutant. Thus, we can conclude that the distal part of the intracellular domain of the LHR is very likely to contain a basolateral targeting signal located between residues 666 and 690.
In the case of the FSHR and TSHR, we have shown that a sequence of 14 or 15 residues, respectively, is sufficient to allow basolateral targeting. To delineate more precisely the region of the LHR implicated in the basolateral targeting, we grafted 18 residues of the receptor (residues 672–689) to the cytoplasmic tail of the truncated p75 NTRt (Fig. 4A). Cell surface immunoprecipitation (Fig. 4B) showed that these 18 residues are sufficient to redirect the chimeric receptor p75 NTRt/LHR672–689 to the basolateral surface of MDCK cells. We can conclude that, in the distal part of the intracellular domain of the LHR, residues 672–689 encode an effective basolateral sorting signal.
|
651–696 were found to be similar (data not shown), as previously described for a similar rat LHR mutant (17). In the same way, no difference in Gs coupling could be detected between the wild-type and the truncated 651–696 and 666–696 receptors deleted of the basolateral sorting signal (Fig. 4C). Thus, the basolateral sorting signal of the LHR is also independent of endocytosis and Gs coupling. Comparison of this signal with those of FSHR and TSHR shows that they share no primary sequence homology. A tyrosine is found within the signal in the porcine receptor. However, this tyrosine is not conserved in LHR cloned in other mammalian species, especially in the human LHR and is thus not expected to play an important functional role.
Circular Dichroism Study of TSHR, LHR, and FSHR Basolateral Sorting Signals
TSHR, LHR, and FSHR basolateral sorting signals share an unusual localization in the distal part of their intracellular domain at a marked distance from the membrane. However, their primary sequences differ. Despite this discrepancy in primary structures, these three basolateral signals may share a common secondary structure.
We determined the secondary structures of the sorting signals of TSHR, LHR, and FSHR using circular dichroism (CD) spectra analysis of synthetic peptides encompassing the sorting signals (Table 1). The CD phenomenon is very sensitive to the secondary structure of the polypeptides. The major absorption in the far UV region comes from peptide bonds (chromophores). For example, an 
-helical pattern is characterized by typical signals at 192, 208, and 222 nm corresponding to the three electronic transitions in the peptide bond (strong 
*
, small *//, and n*, respectively), whereas the random coil structure is characterized by only two signals at 195 and 212 nm (* and n* respectively). As previously described, the easiest method to extract secondary structure content from CD data is to assume that a spectrum is a linear combination of CD spectra of each type of secondary structure (e.g. pure -helix, pure ß-strand, etc.) adjusted in the light of its abundance in the polypeptide conformation (18, 19). From this information it is possible to estimate qualitatively for a polypeptide the prevalent secondary structure or quantitatively the percentage of each secondary structure using deconvolution calculation (see Materials and Methods). Spectra were recorded in various solutions including sodium dodecyl sulfate (SDS), which mimics the membrane bilayer environment (20), and trifluoroethanol (TFE) and methanol (MeOH), which make it possible to evaluate the propensity of peptides to adopt a helical conformation (21) (Fig. 5). The data revealed that the TSHR-derived peptide produces a strong signal at 192, 208, and 222 nm (the so-called Cotton effect) in all solvents, indicating that the peptide adopts a stable helical structure. The LHR-derived peptide produced a weaker Cotton effect, even in TFE, which is a helix promoter, and the CD spectra indicated that the peptide exhibits a tendency to form both helix and ß-sheet structures. It was noticed that the CD spectra of TSHR and LHR in SDS exhibited solvent flattering. This effect, which is probably due to partial aggregation of the peptides in SDS micelles, can be explained by light scattering. Indeed, the stronger attenuation of the 192-nm signal compared with the 208- and 222-nm signals is generally considered an indication of light scattering. Finally, the FSHR-derived peptide predominantly adopted a random coil structure with very little tendency to the helical structure. Indeed, the CD curves obtained for this latter peptide did not exhibit any particular structure. Together these data show that the basolateral sorting signals of the TSHR, LHR, and FSHR have distinct secondary structures.
|
|
We also studied the polarized secretion of a truncated receptor, TSHR314–764, corresponding to a variant form deleted from the transmembrane and intracellular domains. This truncated receptor contains all the six N-glycosylation sites of the TSHR (22) and the four potential O-glycosylation sites determined using the NetOGlyc 2.0 Prediction Server (Fig. 6A). It is lacking a small part of the extracellular domain, because it has been previously shown that longer receptors are not secreted and are trapped intracellularly (23).
|
314–764 present in the medium was purified and analyzed by gel electrophoresis. As shown in Fig. 6, B and C, 55 ± 7.5% of the receptor was secreted in the apical medium, whereas 45 ± 7.5% were secreted in the basolateral compartment (means of four independent experiments). Thus, the soluble TSHR314–764 exhibits a nonpolarized secretion. We can thus conclude that the N-glysosylation sites of the receptor and the potential O-glysosylated sites found in the extracellular domain of the receptor do not contain apical sorting information, and that apical sorting information of the TSHR, revealed by deletion of the dominant basolateral sorting signal, may be located elsewhere in the receptor, in either the transmembrane or intracellular domain. |
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
In the expanded family of G protein-coupled receptors, many members have been shown to exhibit a polarized expression (3, 4, 5, 27), but the molecular mechanisms involved are poorly understood. For the 2A-adrenergic receptor, it was proposed that different parts of the transmembrane domain were implicated in the basolateral targeting of the receptor (6, 28). Basolateral sorting information has been identified in the second intracellular loop of the vasopressin receptor (13) and in the third intracellular loop of the m3-muscarinic acetylcholine receptor (12), but deletion of this last sequence did not alter the basolateral expression of the receptor. The localization of these two basolateral sorting signals varies and is also different from that of the FSHR signal.
In this study we have shown that the basolateral sorting signals of the TSHR and LHR are located between residues Leu732 and His746 and Arg672 and Asp689, respectively. Like the FSHR signal, they share the same unusual localization at the distal part of their intracellular domain, at marked distance from the membrane. However, this distance may be reduced because molecular modeling and mutagenesis of the LHR suggested that helix 7 is parallel to the inner surface of the membrane, which consequently could also be the case for the cytoplasmic tail (29). It may also be the case for the related TSHR and FSHR. In addition, palmitoylation of cysteines in the intracellular domain of the TSHR and LHR creates a fourth intracellular loop, which also contributes to reduce this distance (30, 31). Similar to the basolateral sorting signal of the FSHR, we have shown that TSHR and LHR signals are unrelated to endocytosis signals. They are independent of tyrosine and dileucine motifs. These signals share no homology with other known basolateral signals and display primary sequence differences from each other.
In the absence of the conservation of the sequences of the sorting signals, we wondered whether the secondary structure of the signals could nonetheless have been preserved. Indeed, although tyrosine-based internalization signals display little similarities in terms of amino acid sequence, structural models predict that they share a common structure involving a tight ß-turn (32). Nuclear magnetic resonance (NMR) studies of polarization signals related to endocytosis signals have indicated that the Tyr-based as well as other non-Tyr-based basolateral sorting motifs contain an apparent ß-turn structure (33, 34, 35), suggesting that it may also be a universal feature of basolateral signals (36). However, in the case of gonadotropin receptors and TSHR, CD studies revealed that each signal has a distinct conformation different from the ß-turn. To date, the structures of only two other basolateral sorting signals, independent of endocytosis signals, have been studied by two-dimensional NMR spectroscopy. It shows that the secondary structure of the basolateral sorting signal of the polymeric Ig receptor contains two structural domains, a ß-turn and a nascent -helix (33). This situation is reminiscent of what we observed for the signal of the LHR, which forms both helical and ß-sheet structures. More recently, another study on the rat Na+/taurocholate transporter showed that the two basolateral sorting signals present in its cytoplasmic domain adopt mainly a random coil structure (37), as does the signal of the FSHR. In contrast, the helical structure of the TSHR basolateral signal is unrelated to any known signals.
In the closely related family of FSHR, LHR, and TSHR, which are thought to derive from the same ancestral gene (7, 8), we show that although the unusual distal localization of the signals has been conserved during evolution, the primary and secondary structures of the signals diverge. This study, in which no common features could be detected even in the basolateral sorting signals of closely related proteins, combined with the NMR studies on the polymeric Ig receptor and Na+/taurocholate transporter signals shows the diversity of secondary structures of basolateral signals unrelated to internalization signals. This observation contrasts to what has been shown in the case of basolateral signals related to endocytosis signals. Two-dimensional NMR study and restrained molecular modeling will make it possible to confirm this diversity of structures.
The similarity between basolateral and endocytosis signals suggest that basolateral sorting signals would interact with specific regions of adaptor proteins, which mediate the incorporation of cargo protein into transport vesicles. To date, only two clathrin adaptor subunits, µ1B and µ4, have been implicated in the basolateral sorting of some proteins in epithelial cells (38, 39). However, other adaptors may be implicated, as some basolateral proteins are correctly targeted in epithelial cells deficient in µ1B or µ4 (38, 39).
Because of their different secondary structures, the three basolateral sorting signals might interact with different adaptors. In the case of the polymeric Ig receptor it was suggested that the two structural domains of its basolateral sorting signal could allow basolateral targeting by two different pathways and/or by a pathway involving different adaptors (40). Another possibility could be that the three basolateral sorting signals interact with the same adaptor, but with a different fit. In fact, it has been shown that the non-Tyr-based internalization signal of P-selectin (41) bind to µ2 adaptin in the same fashion previously described for other Tyr-based motifs, but over a larger surface area than typical motifs. In the same way, it has been suggested recently that different classes of basolateral sorting signals might bind to µ1B in different registers or by interacting with different residues (42). Further studies will be necessary to identify the cellular protein(s) and the mechanism(s) involved in the basolateral targeting of gonadotropin receptors and TSHR.
Deletion of the basolateral signals of TSHR, LHR, and FSHR yielded receptors that are targeted to the apical cell surface revealing the existence of apical sorting information. Frequently apical signals correspond to N- or O-linked oligosaccharides in the extracellular domain of proteins (14).
Truncated variants of the TSHR, corresponding to its ectodomain, have been detected (8), and it has been hypothesized that the corresponding proteins might be responsible for a soluble form of the TSHR secreted in human blood. However, it would require the truncated receptor to reach to the basolateral surface of the cells either by a polarized basolateral or a nonpolarized secretion. We have shown that the mutant TSHR314–764 corresponding to the ectodomain of the receptor with all of the N-glycosylated sites and the four putative O-glycosylated sites is secreted in a nonpolarized fashion in MDCK cells. Therefore, if the truncated variants of the TSHR are physiologically expressed they can reach the circulation. It also means that for the TSHR, extracellular N- or putative O-linked glycans do not carry apical sorting information. Thus, these information seems to be located in either the transmembrane or the intracellular domain of the receptor, as described for other proteins (43, 44). Moreover, apical sorting information have been identified in the cytoplasmic tail of two G protein-coupled receptors, the rhodopsin and the vasopressin receptor (13, 45). Thus, during the course of evolution, gonadotropin receptors and TSHR might have conserved in their structure apical sorting information, which are masked by the dominant basolateral sorting signals.
Thus, the determination of the basolateral sorting signals of the three related G protein-coupled receptors provides information on the evolution of the G protein-coupled receptors basolateral sorting signals, which remain largely unknown. Comparative NMR and two hybrid studies will allow a better understanding of the mechanisms involved.
Alteration of TSHR localization has been shown in thyroid cancers. This could be due to a loss of cellular polarization, but also to a molecular abnormality within the basolateral sorting signal (46) or to the loss of a cofactor necessary for the correct targeting of the receptor. Furthermore, natural mutations of gonadotropin receptors and TSHR have been detected in several diseases (8). In some cases, such natural mutations might alter receptor cell trafficking only in polarized cells.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
In the case of the TSHR708–764, a 501-bp fragment (position 1877–2378 of the TSHR cDNA sequence, +1 being the first base of the initial codon) containing a stop codon at position 708 was constructed. After digestion with XcmI and BstEII, the purified fragment was cloned into the pSG5-hTSHR digested with the same enzymes.
In the case of the TSHR731–764 and TSHR747–764 mutants, two 441-bp fragments (position 2044–2460 of the cDNA sequence) containing a stop codon at position 731 or 747 were constructed. After digestion with BstEII and HpaI, the purified fragments were cloned into the pSG5-TSHR vector digested with the same enzymes.
In the case of the TSHR314–764, a 214-bp fragment containing a stop codon at position 314 and a Bsu36I site at its 3' end was generated. This PCR product was digested with AflII and Bsu36I and cloned into the pSG5-TSHR digested with the same enzymes.
In the case of pLHR deletion mutants, LHR651–696, LHR666–696 and LHR690–696, three 635-bp fragments (position 1738–2373 of the cDNA sequence) containing a stop codon at positions 651, 666, and 690 were constructed. After digestion with EspI and BamHI, they were cloned into the pSG5-pLHR that was previously digested by the same enzymes.
Site-Directed Mutagenesis of the TSHR Basolateral Sorting Signal
Two point mutations were generated in the basolateral sorting signal of the TSHR using PCR-mediated oligonucleotide mutagenesis. Tyrosine 739 and leucine 741 were changed in alanine using the same strategy as for the TSHR731–764 and TSHR747–764 mutants.
Construction of Expression Vectors Encoding Chimeric Receptors between the Truncated Neurotropin Receptor NTRt and the Basolateral Sorting Signal of the TSHR or LHR
We used the pCB6 vector containing the NTR lacking its intracellular domain (six residues downstream of the transmembrane domain), which has been previously described (49). The basolateral sorting signals of the TSHR (and supplementary amino acids necessary to conserve the open reading frame) or LHR were cloned into the XbaI site yielding the NTRt/TSHR729–746 or NTRt/LHR672–689. Chimeric receptors containing a mutated basolateral sorting signal were also constructed yielding NTRt/TSHR729-Y739A-746 and NTRt/TSHR729-L741A-I742A-746 with Tyr739Ala and Leu741Ala-Ile742Ala substitutions, respectively.
Cell Culture and Expression of TSHR, LHR Mutants, and NTRt Chimera
MDCK (type II) and FRT (gift from Dr. B. Rousset, INSERM, Unité 369, Lyon, France) cells were maintained as previously described (10, 50). The different cell lines were obtained as previously described (10). Antibodies T5–317, LHR38, or ME20.4 (gift from Dr. A. Le Bivic) were used to screen the clones expressing the TSH and LH receptors or the neurotropin chimera (10, 11). We also verified that the polarized phenotype of transfected MDCK and FRT cells was not altered (10).
Indirect Immunofluorescence and Confocal Microscopy
MDCK cells were grown on 12-mm filters (0.4-µm polyester Transwell, Costar Corp., Cambridge, MA) and processed as previously described (11). Briefly they were fixed with 3% paraformaldehyde. After saturation with PBS containing 1% BSA (fraction V, Sigma-Aldrich Corp., St. Louis, MO), the apical and basolateral surfaces were incubated with the same medium containing monoclonal antireceptor antibody: T5–317 (5 µg/ml), LHR38 (5 µg/ml), or ME20.4 (dilution 1:200) directed against the extracellular domain of the TSHR, LHR, or NTR, respectively, followed by incubation with Alexa Fluor-488 goat anti-mouse IgG (Molecular Probe, Eugene, OR). After washing, the samples were mounted in fluorescent mounting medium (DAKO, Carpenteria, CA) and observed with a microscope (Axiovert 135M, Zeiss, Deerfield, IL) in conjunction with a confocal laser scanning unit (Zeiss LSM 410).
cAMP Assays
cAMP was measured as previously described (10) after 45-min incubation of transfected COS-7 cells with 100 mIU/ml bTSH (Sigma-Aldrich Corp.) or 30 IU/ml human chorionic gonadotropin (hCG; Organon, West Orange, NJ).
Surface Immunoprecipitation of Wild-Type and Mutated TSHR and LHR and of Chimeric Receptors NTRt/TSHR729–746 and NTRt/LHR672–689 Expressed in MDCK Cells
Polarized monolayers of MDCK cells were used. Receptors expressed on the cell surface were immunoprecipitated as described previously (10, 11). After a metabolic labeling of the cells, surface immunoprecipitation was performed using the polyclonal antibody TSHR19–389 (10) for the TSHR and the monoclonal antibodies LHR38 and ME20.4, for LHR and NTR, respectively, recognizing the extracellular domain of the corresponding receptor. Extraction and purification of receptor-antibody complexes were performed as described previously (10, 11). All experiments were performed at least twice with triplicate sample.
Study of the Polarized Secretion of the TSHR314–764 Receptor Mutant
MDCK cells permanently expressing the TSHR314–764 receptor mutant were grown on 24-mm polycarbonate filters, pulsed for 1 h with [35S]methionine and chased for 5 h with medium containing unlabeled amino acids. The soluble TSHR314–764 receptor secreted either in the apical or in the basolateral culture medium was then immunopurified using the monoclonal antibody T5–317.
Electrophoresis and Autoradiography
SDS-PAGE was performed as previously described (10). The gels were fixed and processed for fluorography. Densitometric scanning was used for quantification.
Receptor-Mediated TSH and LH Internalization
MDCK cells expressing the wild-type or mutated TSHR or LHR were processed as previously described (51). Briefly, they were incubated for 2 h at 4 C (for TSHR) or 10 min at 37 C (for LHR) with [125I]bTSH (0.1 µCi/ml; ERIA Diagnostic Pasteur, Sanofi, France) or [125I]hCG (325 µCi/ml; PerkinElmer, Wellesley, MA) (10, 51). Unbound ligand was removed, and the cells were incubated at 37 C for varying time periods (0–30 min). The ligand remaining on the cell surface was stripped by acidic washes. Finally, the cells were recovered by addition of trypsin, and the total radioactivity of the suspension was counted. This value corresponds to the internalized fraction. Nonspecific binding was determined in the presence of an excess of unlabeled bTSH (Sigma-Aldrich Corp.) or hCG (Organon) and with nontransfected MDCK cells. Internalized ligand was expressed as the percentage of the total radioactivity (corrected for nonspecific binding) initially bound to the cells.
Peptide Synthesis: Human TSHR, FSHR, and LHR Sequences
TSHR725–753 (THDMRQGLHNMEDVYELIENS-HLTPKKQG), FSHR671–695 (HCSSAPRVTSGSTYILVPLSHLAQN), and LHR671–694 (NKPSRSTLKLTTLQCQYSTVMDKT) were synthesized (0.1 mmol scale) by the solid phase methodology using a 433A PE Applied Biosystems peptide synthesizer (Foster City, CA) and the standard F-moc procedure as previously described (52). Synthetic peptides were purified by reverse phase HPLC on a 2.2 x 25-cm Vydac C18 column (Alltech, Templemars, France) using a linear gradient (10–50% over 45 min) of acetonitrile/TFA (99.9:0.1, vol/vol) at a flow rate of 10 ml/min. Analytical reverse phase HPLC was performed on a Vydac C18 column (0.45 x 25-cm) using a linear gradient (10–60% over 25 min) of acetonitrile/TFA at a flow rate of 1 ml/min. The purified peptides were characterized by matrix-assisted laser desorption ionization time-of-flight mass spectrometry on a Tof-Spec E spectrometer (Micromass, Manchester, UK).
CD Spectroscopy
The CD spectra of the peptides were recorded at room temperature on a CD6 spectropolarimeter (Jobin Yvon, Longjumeau, France). The instrument was routinely calibrated with an aqueous solution of (+)-10-camphorsulfonic acid. The CD spectra were obtained in the far UV, using the wavelength ranges 186–260 nm at 1-nm resolution in 0.02-cm cells. Ten scans were recorded at a speed of 1 point every 2 s. The peptide concentrations used were 112.5 µM (FSHR), 93.53 µM (TSHR), and 100.0 µM (LHR) as determined by the absorbance of tyrosine at 276 nm, using an extinction coefficient of 1450 cm-1 x M-1 (53). Mean residue ellipticity values ([
]) were expressed in degrees x square centimeters x decamoles-1 and were calculated from the equation [] = x M/10 x l x C, where is the observed ellipticity in degrees, M is the mean residue molecular weight of the peptide, l if the optical path length in centimeters, and C if the peptide concentration in grams x milliliter-1. CD spectra were recorded with peptides in SDS solution [25 mM TSHR (pH 6.6), LHR (pH 7.5), and FSHR (pH 7.7)], methanol, and TFE/H2O (90:10). Estimation of the conformations was performed using the CD reference curves reported by Yang et al. (18) and Chang et al. (19). Calculations were performed using a custom-made program applying the conventional additivity rule between 190 and 240 nm at every nanometer. All conformation combinations (i.e. helix, ß-sheet, ß-turn, and random coil) were computed from 0–100% in steps of 1%. The resulting theoretical curves were subtracted from the experimental ones, and the least square method was used to determine the best fit.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Abbreviations: CD, Circular dichroism; FSHR, FSH receptor; hCG, human chorionic gonadotropin; LHR, LH receptor; MDCK, Madin-Darby canine kidney; MeOH, methanol; NMR, nuclear magnetic resonance; NTR, neurotropin receptor; NTRt, truncated neurotropin receptor; SDS, sodium dodecyl sulfate; TFE, trifluoroethanol; TSHR, TSH receptor.
Received for publication April 10, 2003. Accepted for publication December 17, 2003.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
2-adrenergic receptor subtypes achieve basolateral localization in Madin-Darby canine kidney II cells via different targeting mechanisms. J Biol Chem 271:5017–50242A-adrenergic receptor on the basolateral membrane of Madin-Darby canine kidney cells. J Biol Chem 269:16425–16432
sorting motifs. Traffic 2:105–110[CrossRef][Medline]
This article has been cited by other articles:
 |
|
 |
|
  S. Tsui, V. Naik, N. Hoa, C. J. Hwang, N. F. Afifiyan, A. Sinha Hikim, A. G. Gianoukakis, R. S. Douglas, and T. J. Smith Evidence for an Association between Thyroid-Stimulating Hormone and Insulin-Like Growth Factor 1 Receptors: A Tale of Two Antigens Implicated in Graves' Disease J. Immunol., September 15, 2008; 181(6): 4397 - 4405. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A.-D. Qi, S. C. Wolff, and R. A. Nicholas The Apical Targeting Signal of the P2Y2 Receptor Is Located in Its First Extracellular Loop J. Biol. Chem., August 12, 2005; 280(32): 29169 - 29175. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. D. Holliday, C.-W. Lam, I. R. Tough, and H. M. Cox Role of the C Terminus in Neuropeptide Y Y Receptor Desensitization and Internalization Mol. Pharmacol., March 1, 2005; 67(3): 655 - 664. [Abstract] [Full Text] [PDF]
|
|
) or the truncated TSHR
) were incubated with [125I]bTSH, washed, and placed at 37 C for various periods of times (see Materials and Methods). After acid washes to eliminate the radiolabeled hormone remaining at the cell surface, the cells were recovered by incubation with trypsin. Trypsin-resistant radioactivity was considered to represent the internalized hormone and was expressed as a percentage of the iodinated hormone initially bound to the cells. D, COS-7 cells transfected with the wild-type (WT) or truncated TSHR were incubated for 45 min with bTSH, and the accumulation of cAMP was quantified. NT, Control cells.
) and the 4 potential O-linked sugars (
) was constructed, and its polarized secretion was studied. B, After metabolic labeling of MDCK cells permanently expressing TSHR