The Venus's-flytrap and Cysteine-rich Domains of the Human Ca2+ Receptor Are Not Linked by Disulfide Bonds*

The extracellular N-terminal domain of the human Ca2+ receptor (hCaR) consists of a Venus's-flytrap (VFT) domain and a cysteine-rich (Cys-rich) domain. We have shown earlier that the Cys-rich domain is critical for signal transmission from the VFT domain to the seven-transmembrane domain. The VFT domain contains 10 cysteines: two of them (Cys129 and Cys131) were identified as involved in intermolecular disulfide bonds necessary for homodimerization, and six others (Cys60-Cys101, Cys358-Cys395, and Cys437-Cys449) are predicted to form three intramolecular disulfide bonds. The Cys-rich domain contains nine cysteines, the involvement of which in disulfide bond formation has not been defined. In this work, we asked whether the remaining cysteines in the hCaR VFT, namely Cys236 and Cys482, form disulfide bond(s) with cysteines in the Cys-rich domain. We constructed mutant hCaRs with a unique tobacco etch virus (TEV) protease recognition site inserted between the VFT domain and the Cys-rich domain. These mutant hCaRs remain fully functional compared with the wild type hCaR. After TEV protease digestion of the mutant hCaR proteins, dimers of the VFT were identified on Western blot under nonreducing conditions. We concluded that there is no disulfide bond between the VFT and the Cys-rich domains in the hCaR.

The Ca 2ϩ receptor (CaR) 1 plays a central role in the regulation of [Ca 2ϩ ] o homeostasis (for reviews, see Refs. 1 and 2). Activation of CaR by elevated levels of [Ca 2ϩ ] o stimulates phospholipase C via the G q subfamily of G-proteins, resulting in the increase of both phosphoinositide (PI) hydrolysis and the concentration of cytosolic calcium, [Ca 2ϩ ] i . The CaR mediates the inhibitory actions of [Ca 2ϩ ] o on parathyroid hormone secretion by the parathyroid gland and on Ca 2ϩ reabsorption by the kidney.
The CaR is a member of family 3 of the G-protein-coupled receptor (GPCR) superfamily, which also includes metabotropic glutamate receptors (mGluRs) (3), ␥-amino butyric acid type B receptors (GABA B Rs) (4), some putative pheromone receptors (5), and some putative taste receptors (6). Their distinctively large extracellular domains (ECDs) consist of a "Venus's-flytrap" (VFT) domain and a cysteine-rich (Cys-rich) domain with the exception of GABA B Rs, which lack a Cys-rich domain. The three-dimensional structure of the VFT domain of mGluR1 has been determined recently by x-ray crystallography (7) showing a bilobed VFT-like structure.
The hCaR VFT contains 10 cysteines, while mGluR1 contains 9. mGluR1 forms homodimers involving an intermolecular disulfide bond through cysteine Cys 140 (8), whereas hCaR forms homodimers involving two intermolecular disulfide bonds through both Cys 129 and Cys 131 (9). The crystal structure of the mGluR1 VFT domain shows that it forms four intramolecular disulfide bonds. Based on amino acid sequence alignment, three homologous intramolecular disulfide bonds are predicted to form within the hCaR VFT, i.e. Cys 60 -Cys 101 , Cys 358 -Cys 395 , and Cys 437 -Cys 449 (7) (Fig. 1). The hCaR VFT has no cysteines corresponding to the disulfide-bonded pair Cys 289 -Cys 291 in mGluR1 VFT. Of the remaining two cysteines in the hCaR VFT, Cys 236 , which is conserved between the hCaR and mGluR1, is a free cysteine in mGluR1 VFT, whereas Cys 482 of hCaR has no counterpart in mGluR1. We reported previously that Cys 236 was critical to the function of the hCaR, but Cys 482 was not (10).
We recently reported that the Cys-rich domain plays a critical role in signal transmission from the VFT domain to the 7 TM domain (11). The hCaR Cys-rich region contains nine highly conserved cysteines in a closely spaced (about 60 amino acids long) sequence (Fig. 1A). We reported previously that each of the nine cysteines in the Cys-rich domain in hCaR is critical for the receptor's function (10). It is speculated that multiple intramolecular disulfide bonds are formed within this region, which may give rise to a tightly packed domain. However, the disulfides within the Cys-rich domain have not been characterized, and the mechanism by which an agonist signal is transmitted from the VFT domain through the Cys-rich domain to the 7 TM domain remains unknown. One possibility is that the hCaR VFT and Cys-rich domains are linked by disulfide bond(s) between Cys 236 and/or Cys 482 in the VFT and one or more cysteines in the Cys-rich domain. If this were true, conformational changes in the VFT domain after agonist binding might directly cause the conformational changes of the Cys-rich domain through the disulfide bond(s) linkage. To test this hypothesis, we constructed mutant hCaRs with a tobacco etch virus (TEV) protease recognition site (for a review of TEV protease, see Ref. 12) inserted between Glu 536 and Val 537 and studied the products of TEV protease digestion of the mutant hCaRs.

MATERIALS AND METHODS
Construction of Mutant hCaRs-Site-directed mutagenesis was performed by using the Quickchange site-directed mutagenesis kit (Stratagene Inc., La Jolla, CA), according to the manufacturer's instructions. The mutagenic oligonucleotide primer pair for introducing six new residues (NLYFQG) between Glu 536 and Val 537 in the hCaR was * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Secretion of the hCaR VFT Terminating following Residue Glu 536 -hCaR sequence from amino acids Met 1 to Glu 536 was amplified by polymerase chain reaction with the addition of a stop codon following Glu 536 and then cloned into pCEP4 vector (Invitrogen Corp., Carlsbad, CA). The construct was transformed into HEK-293 cells and stable cell lines secreting the hCaR VFT from Tyr 20 to Glu 536 , termed E536, were established by screening colonies resistant to hygromycin treatment.

Transient Transfection of CaRs in HEK-293 Cells, PI Hydrolysis Assay, and Preparation of Detergent-solubilized Whole Cell Extracts-
The CaRs were transfected in HEK-293 cells by using a DNA-Lipo-fectAMINE mixture (Life Technologies, Inc.) as described previously (11). PI hydrolysis assay and preparation of detergent-solubilized whole cell extracts were carried out as described previously (11) with one modification in that iodoacetamide was not included in lysis buffer.
TEV Protease Digestions-100 g of total protein of the whole cell lysate was diluted with 1ϫ TEV buffer (Life Technologies, Inc.), and buffer exchange was carried out by repeated dilution with 1ϫ TEV buffer and concentration using Microcon YM-100 (Millipore Corp., Bedford, MA) spin columns. For proteolysis, 100 units of recombinant TEV protease (Life Technologies, Inc.) were added to 500 l of reaction mixture containing the proteins in 1ϫ TEV buffer, and the digestion was performed at room temperature overnight. All TEV protease digestions were carried out without adding dithiothreitol. The digestion products were concentrated using Centricon YM-50 spin columns before loading onto a gel for SDS-PAGE.
Immunoblotting Analysis-Protein samples were resolved by SDS-PAGE under either nonreducing or reducing conditions (with the addition of 300 mM ␤-mercaptoethanol in sample buffer). The proteins on the gel were electrotransferred onto nitrocellulose membrane. For the ADD blot, the membrane was incubated with mouse monoclonal anti-hCaR antibody ADD (raised against a synthetic peptide corresponding to residues 214 -235 of hCaR protein) and then with a secondary goat anti-mouse antibody conjugated to horseradish peroxidase (Kirkegaard and Perry Laboratories, Gaithersburg, MD). For the His blot, a nitrocellulose membrane was incubated with rabbit polyclonal His-probe (H-15) (Santa Cruz Biotechnology, Inc., Santa Cruz, CA) and then with anti-rabbit IgG conjugated to horseradish peroxidase (Santa Cruz Biotechnology, Inc.). The hCaR protein was detected with an ECL system (Amersham Pharmacia Biotech). Blots shown in this paper were representatives from at least three independent experiments.

Construction and Functional Assay of a Mutant hCaR with a Unique TEV Protease Recognition Site Inserted between Glu 536
and Val 537 -Site-directed mutagenesis was applied to insert nucleotide sequence for six new residues (NLYFQG) between Glu 536 and Val 537 in the wild type (WT) hCaR, resulting in a unique seven-amino acid TEV protease recognition site (EN-LYFQG) (Fig. 1A). The mutant hCaR was termed hCaR(TEV). Recombinant TEV protease purified from Escherichia coli cleaves between residues Gln and Gly. We first tested whether hCaR(TEV) is capable of signal transduction using the intact cell [Ca 2ϩ ] o -stimulated PI hydrolysis assay. Nontransfected or vector only transfected HEK-293 cells (11) showed no PI hydrolysis response at [Ca 2ϩ ] o concentrations as high as 50 mM.   (Fig. 1B). Previous studies have shown that the monomeric ϳ150-kDa band represents hCaR forms expressed at the cell surface and modified with N-linked complex carbohydrates. The ϳ130-kDa band represents high mannose-modified forms, trapped intracellularly and sensitive to endoglycosidase H digestion (13)(14)(15). Under nonreducing conditions, both forms appear as poorly resolved ϳ260and ϳ300-kDa bands representing intermolecular disulfide-linked dimers (Fig. 1B). In Fig. 1B, the ϳ300-kDa form of hCaR(TEV) appears to be expressed at a somewhat higher level than that of WT, but this is not a consistent finding. Thus, the hCaR(TEV) was expressed at equivalent levels to that of WT hCaR and dimerized normally, indicating that the six-amino acid insertion between the VFT and the Cys-rich domains had no detectable effect on the receptor's folding, cell-surface expression, or function.
TEV Protease Digestion of hCaR(TEV)-Whole cell lysate from cells transiently transfected with WT hCaR or hCaR(TEV) were subjected to TEV protease digestion and analyzed by SDS-PAGE and immunoblot. Fig. 2A shows that no endogenous TEV cleavage site is present in the WT hCaR and immunoreactive bands remain unchanged from those of the WT hCaR without TEV protease digestion (compare with Fig.  1B). Following TEV protease digestion, immunoreactive bands were less distinct under nonreducing conditions, which may be because of the aggregation of the receptor after sample concentration necessitated by TEV protease digestion protocol. Fig.  2B shows that TEV protease cleaved the hCaR(TEV) protein, although the digestion was not complete. Two new lower molecular weight bands were identified under reducing conditions following TEV protease digestion. Based on the location of the epitope of the ADD monoclonal antibody (Fig. 1A), both bands represent monomers of the hCaR VFT. The upper band is the VFT domain cut from the cell surface-expressed, fully glycosylated receptor, and the lower band is the VFT domain cut from the intracellularly trapped, high mannose-modified receptor. Culture medium containing a secreted form of the VFT domain, E536 (containing VFT sequence from Tyr 20 to Glu 536 ), was run as a size marker to indicate the location of the VFT monomer cleaved from hCaR(TEV). E536 migrated on SDS-PAGE similarly to the VFT domain released from the fully processed receptor after TEV protease digestion.
Under nonreducing conditions, bands for the dimerized VFT domain, migrating similarly to the dimer of the secreted E536, appeared in the TEV protease digestion products of hCaR(TEV) (Fig. 2B). This result indicates that the VFT domain in the hCaR(TEV) is not disulfide-linked to the Cys-rich or 7 TM domains of the receptor; otherwise, the immunoreactive bands should have remained the same as those of the WT hCaR under nonreducing conditions.
Construction and TEV Protease Digestion of hCaR(TEV-HIS)-By site-directed mutagenesis, we added a 6 ϫ His tag to the intracellular C terminus of the hCaR(TEV) and termed the new construct hCaR(TEV-HIS). The PI hydrolysis assay showed its maximal response to [Ca 2ϩ ] o and EC 50 were identical to those of WT hCaR and it was expressed at the same level as WT hCaR, indicating that the 6 ϫ His tag at the C terminus had no effect on the receptor's folding, expression, or function (data not shown).Whole cell lysate from cells transiently transfected with hCaR(TEV-HIS) was digested with TEV protease and loaded onto 6% SDS-polyacylamide gel containing 8 M urea. An ADD immunoblot under reducing condition showed bands for monomeric VFT released from hCaR(TEV-HIS) after TEV protease digestion, whereas under nonreducing conditions bands for dimeric VFT appeared (Fig. 2C). After the blots were stripped and reprobed with anti-His antibody directed against the 6 ϫ His tag added to the C-tail of the hCaR(TEV-HIS), bands corresponding to the remaining part of the receptor (Cys-rich ϩ 7 TM ϩ C-tail domains) were visualized together with the residual uncut whole receptor (Fig. 2C). The bands for Cys-rich ϩ 7 TM ϩ C-tail domains stained with the anti-His probe were identical under reducing and nonreducing conditions. These results confirm that the hCaR VFT domain is not disulfide-linked to the rest of the receptor protein. hCaR (11,16,17). Our current structural information for the hCaR VFT comes largely from site-directed mutagenesis studies and from homology modeling based on the three-dimensional structure of the recently crystallized mGluR1 VFT (7). Our site-directed mutagenesis data (9) showed that both Cys 129 and Cys 131 are involved in intermolecular disulfide bonds. A homology model based on the mGluR1 VFT crystal structure predicts that Cys 60 -Cys 101 , Cys 358 -Cys 395 , and Cys 437 -Cys 449 are involved in intramolecular disulfide bonds within the hCaR VFT. Cys 236 and Cys 482 are the remaining cysteines in the hCaR VFT that have not been accounted for. It is unclear whether they are free cysteines in the hCaR. The Cys-rich domain in the hCaR ECD contains nine highly conserved cysteines in an ϳ60-amino acid-long sequence. Multiple intramolecular disulfide bonds are speculated to form within this region, but none of the disulfides has been defined yet. Given the odd number of cysteines in this domain, at least one of them should be free or disulfide-linked to a cysteine in other domains.
We recently reported that the Cys-rich domain in the hCaR is critical for signal transmission from the VFT to the 7 TM domains (11). But a key question remains unanswered: how are the VFT and Cys-rich domains structurally and functionally coordinated during agonist binding and activation of the receptor? The crystal structure determined for the mGluR1 VFT domain does not address this question, because the Cys-rich domain was not included in the mGluR1 construct expressed and crystallized (7). One possible mechanism for communication between the VFT and Cys-rich domains could involve disulfide linkage. Conformational changes of the VFT after agonist binding would then cause conformational changes in the Cys-rich domain through disulfide(s) linkage between the two domains.
It is noteworthy that the hCaR has two cysteines (Cys 677 and Cys 765 ) in the extracellular loops 1 and 2 of the 7 TM domain. However, they are less likely to pair with any cysteines in the ECD, because the two conserved cysteines in extracellular loops 1 and 2 are known to be linked by a disulfide in most GPCRs including bovine rhodopsin, the thyrotropin-releasing hormone receptor, the thromboxane receptor, the GnRH receptor, and many others (18).
To determine whether the VFT is linked to the Cys-rich domain by disulfide bond(s), we inserted a unique TEV protease recognition site between the two domains. To date, it remains unclear where the exact boundary is between the two domains. A construct of the mGluR1 VFT made by Tsuji and co-workers (7,8) that ends at Ser 522 , two amino acids ahead of the first cysteine in the Cys-rich region (corresponding to Ser 540 in hCaR), was secreted, purified, and crystallized. An alternative splicing product of hCaR, with 10 additional amino acids between Glu 536 and Val 537 (19), was shown to be fully functional in an in vitro assay. We hypothesized that Glu 536 is at or very close to the boundary of the VFT and Cys-rich domains, and the receptor may be tolerant of insertion of additional amino acids at this site. By applying site-directed mutagenesis, we made a secreted version of the hCaR VFT with a stop codon after Glu 536 . It was well expressed as a dimer under nonreducing conditions and secreted into the culture medium. Moreover, hCaR(TEV) with a TEV protease recognition sequence inserted at this site was fully functional in the PI assay, as compared with the wild type hCaR.
Our results show that, although it is difficult to achieve complete digestion in lysates of transfected HEK-293 cells, recombinant TEV protease recognizes the site inserted in the hCaR (TEV) cDNA and cleaves the receptor protein, as shown in ADD blots of hCaR(TEV) run on SDS-PAGE under reducing conditions. However, there could be two alternative results for the behavior of the products of TEV digestion on SDS-PAGE under nonreducing conditions. If the VFT is disulfide-linked to the Cys-rich domain, the TEV protease digestion product of hCaR(TEV) should remain as a holoprotein under nonreducing conditions, whereas if there is no such linkage, a dimer of the VFT domain will be released from the rest of the receptor on SDS-PAGE under nonreducing conditions. Our results show that there is no disulfide bond between the VFT and the Cysrich domains. A HIS blot of the TEV protease digestion product of hCaR(TEV-HIS) shows that the His antibody recognizes the protein with a C-terminal His tag but not the VFT released from the holoprotein, confirming that the hCaR VFT is not disulfide-linked to the Cys-rich domain. The absence of disulfide linkage between the VFT and other domains of the receptor might be common among family 3 members of GPCR, as GABA B receptors lack a Cys-rich domain altogether.
In summary, our results exclude the existence of a disulfide bond between the hCaR VFT and Cys-rich domains. We cannot, however, exclude noncovalent interactions between these two domains, and indeed such interactions could be important in the mechanism of activation of the receptor. We attempted to determine whether the hCaR VFT and Cys-rich domains are associated by noncovalent interactions by analyzing the products of TEV protease digestion under native gel conditions. Unfortunately, perhaps because of its high degree of glycosylation, the hCaR is very poorly resolved under native gel conditions 2 , making it difficult to draw definitive conclusions using this method. Further study of the structure of both the VFT and Cys-rich domains will help to reveal the mechanism by which the signal of Ca 2ϩ -induced conformational changes in the VFT is transmitted through the Cys-rich domain to the 7 TM domain, resulting in CaR activation.