Recombinant expression of a secreted form of the atrial natriuretic peptide clearance receptor.

A general structure for the atrial natriuretic peptide clearance receptor (ANP C-receptor) has been proposed based on hydropathicity analysis of the deduced amino acid sequence of this membrane protein (Fuller, F., Porter, J.G., Arfsten, A., Miller, J., Schilling, J., Scarborough, R.M., Lewicki, J.A., and Schenk, D.B. (1988) J. Biol. Chem. 263, 9395-9401). The ANP C-receptor is believed to possess a large amino-terminal extracellular domain (436 amino acids), a single hydrophobic transmembrane anchor (23 amino acids), and a short cytoplasmic tail (37 amino acids). As a means of testing the structure and proposed cellular orientation of this protein, we have employed the technique of in vitro mutagenesis to prepare a receptor mutant (anc-) lacking the transmembrane and cytoplasmic domains. Expression of this mutant in mammalian cells using a vaccinia virus vector results in secretion of a truncated soluble form of the ANP C-receptor which binds native ANP and synthetic ANP analogs with a specificity similar to that of the native ANP C-receptor. In contrast to the native ANP C-receptor that exists predominantly as a homodimer on the cell surface, the secreted receptor exists as a monomeric species. The results are consistent with the proposed structure of this receptor with the amino-terminal domain containing the ANP-binding site oriented extracellular to the plasma membrane. In addition, these data demonstrate that the receptor does not require association with the plasma membrane or its native dimeric configuration in order to bind ANP ligands with high affinity and specificity.

ized by its rigorous binding specificity for native ANPs. This receptor has been shown to possess intrinsic ANP-dependent guanylate cyclase activity that may mediate many of the biological effects of ANP (12)(13)(14)(15)(16)(17)(18)(19). The ANP B-receptor has been purified by several laboratories (15)(16)(17)(18), and the amino acid sequence of the protein has recently been deduced from analysis of cloned DNA sequences that encode the receptor (19).
A second receptor, termed the ANP C-receptor, is distinct from the B-receptor in terms of its structure, immunoreactivity, physiological role, and ligand-binding specificity (16,(20)(21)(22)(23). The ANP C-receptor exhibits a high affinity for native ANP and is also able to bind various truncated and internal ring-contracted analogs of ANP with high affinity (21-23). The ANP C-receptor represents the major ANP-binding site in cultured vascular cells and in the kidney (5,22). Although this receptor is not involved in stimulation of particulate guanylate cyclase and does not appear to mediate directly any of the known biological effects of ANP, it has been shown to mediate the metabolic clearance and degradation of this hormone (21,22).
The purification, cloning, and recombinant expression of the ANP C-receptor has been reported recently, and this has allowed a more detailed understanding of the structural and functional properties of this protein (20,24,25). The ANP Creceptor consists of a single -60,000-dalton subunit that appears to exist predominantly as a homodimer on the cell surface. From hydropathicity analysis of the deduced amino acid sequence, the receptor subunit is shown to have two extended hydrophobic regions. One, at the amino terminus, has the characteristics of a signal sequence and may be involved in translocation of the protein to the cell surface. The second region is located near the carboxyl terminus and is proposed to represent a transmembrane domain. A possible receptor structure thus consists of a large amino-terminal extracellular domain that contains the ANP-binding site, a single transmembrane region, and a short carboxyl-terminal cytoplasmic tail. We have used the technique of in vitro mutagenesis to delete the transmembrane and cytoplasmic domains and have expressed this truncated receptor in mammalian cells. The properties of the resulting soluble and secreted ANP C-receptor mutant are described in this report.

Preparation of Recombinant Plasmid
DNA-The ANP C-receptor cDNA insert ANPRc3/4 was subcloned into the expression vectors pSCll and pGEMl using standard recombinant DNA techniques yielding plasmids pSCANPRc3/4 and pGEMANPRc3/4 as described (20,25). The pGEMl vector contains the SP6 promoters which allowed the synthesis of receptor-specific RNA using SP6 polymerase (Promega Biotec). In vitro translation of this mRNA was accom-

14179
Atrial Natriuretic Peptide Clearance Receptor Structure plished with a cell-free system derived from reticulocytes (Du Pont-New England Nuclear). The pSCll construct allowed the generation of specific vaccinia recombinants that were used subsequently to infect confluent monolayers of L-M cells as described below.
Oligonucleotide-directed Mutagenesis-In vitro mutagenesis was performed according to the method of Zoller and Smith (26). The ANP C-receptor cDNA fragment ANPRc3/4 (20, 25) was isolated and subcloned into the m13 phage vector mp8. Cloning orientation as determined by sequencing, and the negative strand DNA prepared from this phage was used for synthesis of a double-stranded template using a mutagenic oligonucleotide described below and the Klenow fragment of DNA polymerase I (Bethesda Research Laboratories). Plasmid DNA from this reaction was transfected into Escherichia coli strain JM101. Phage DNAs with the desired mutation were identified by hybridization to 32P-labeled oligonucleotide primer under stringent wash conditions (3.0 M Tetramethylammonium chloride, 52 "C). Template DNA was prepared from several positively hybridizing phage, and the sequence of the recombinant mutant was confirmed by enzymatic sequencing (27). Replicative form DNA was also prepared from bacteria transfected with this phage, and the mutant cDNA insert (9247) was isolated and subcloned into the expression vectors pSCll and pGEMl using standard methods and resulting in the plasmids pSC9247 and pGEM9247.
Two oligonucleotides were designed for mutagenesis experiments. A 33-base oligonucleotide containing a single base change (G-A) was used to introduce a stop codon immediately 5' to the carboxylterminal transmembrane domain, whereas a 17-base oligonucleotide spanning the same region was used for identifying m13 phage containing mutated receptor sequences. Oligonucleotides were synthesized on an Applied Biosystems model 380A automated DNA synthesizer and purified by gel electrophoresis.
Propagation and Isolation of Recombinant Vaccinia Virus-Vaccinia virus recombinants were isolated and grown as described previously (20). Wild-type vaccinia virus (Wyeth strain) was propagated in CB-1 cells (green monkey kidney) as described (28). CV-1 cells were grown in Dulbecco's modified Eagle's medium (GIBCO) containing 10% fetal bovine serum. Human TK-143 cells were grown in Eagle's minimum essential medium containing 10% fetal bovine serum.
CV-1 cells (IO6 cells/60-mm dish) were infected with wild-type vaccinia virus and co-transfected with 5 pg of calcium phosphateprecipitated recombinant plasmid DNA containing wild-type receptor cDNA sequences (pSCANPRc3/4) or mutant receptor cDNA sequences (pSC9247) (29). Cells were incubated at 37 "C for 2 days and harvested by scraping. After freezing and thawing three times, dilutions of this virus stock were used to infect confluent monolayers of hTK-143 cells. After 2 h, cells were overlayed with agarose containing 5-bromodeoxyuridine and incubated for 2 days. Cells were then overlayed a second time with agarose containing 5-bromo-4-chloro-3indoly-0-D-galactopyranoside, and blue plaques were visualized after 4-12 h. Plaques were isolated, freeze-thawed three times, sonicated, and repurified three times in succession. Purified recombinant virus containing either native or mutated (anc-) receptor cDNA sequences were amplified in hTK-143 cells, viral DNA, and RNA isolated and recombinants confirmed by hybridization. Vaccinia virus recombined with the plasmid vector alone were also isolated and served as controls throughout these experiments.

L-M cells (mouse fibroblast) were grown in Dulbecco's modified
Eagle's medium and 10% calf serum. Cells were infected with vaccinia virus by first washing confluent monolayers in 6-well plates with serum-free medium and then incubating for 2 h with 0.5 ml of recombinant or control vaccinia virus diluted in phosphate-buffered saline, 0.05% bovine serum albumin, and phenol red. The inoculum was then removed and replaced with 1 ml of Dulbecco's modified Eagle's medium-21 without serum, and the cells were incubated for an additional 4-6 h (30). Radiolabeling experiments were performed by incubating cells overnight in 0.5 ml of medium containing 250 pCi of [35S]methionine as described previously (20). Both labeled supernatants and cell lysates were assayed by immunoprecipitation.
'251-ANP-binding Assays-Binding assays were performed in 6-or 24-well dishes on confluent cell monolayers either before infection or 6 h after infection with either control virus or recombinant vaccinia virus as described. Infected cell monolayers were washed and incubated with 'T-ANP in the presence of various concentrations (0.3-100.0 nM) of ANP or ANP analogs for 1 h at 37 "C. Cells were washed with culture medium, solubilized with NaOH, and counted as described (5). Culture medium from infected cells was also saved and assayed for specific binding of '251-ANP. Typically, 0.5 ml of infected cell medium was incubated with '251-ANP in the presence of various concentrations of unlabeled ANP for 30 min at 20 "C. Free and receptor-bound T -A N P were separated by the addition of 1 ml of a 10 mg/ml suspension of activated charcoal in 100 mM Tris (pH 7.5), 100 mM NaCl, 10 mM CaC12, and 10 mM MgC1, followed by centrifugation at 1000 rpm for 10 min. Receptor-bound lz5I-ANP was unabsorbed and was monitored by counting the resulting supernatant in a Beckman 5500 y-counter.
Peptide Synthesis and Conjugation to Carrier-A peptide (P4) homologous to a specific region of the ANP C-receptor (amino acids 118-136) was synthesized according to the amino acid sequence of this receptor predicted from the nucleotide sequence of the corresponding cDNA (25). Selection of a peptide region for synthesis was based on hydrophilicity indices determined by the method of Hopp and Woods (32). Peptides (receptor homologous peptides as well as ANP analogs) were synthesized exclusively by the solid phase method using protocols described previously (20). Purified peptides were characterized by amino acid analysis and homogeneity assessed by analytical reversed phase liquid chromatography (>98% pure). The receptor-specific peptide P4 was coupled to keyhold limpet hemocyanin using a standard procedure that consisted of treatment of 6 mg of peptide and 6 mg of keyhold limpet hemocyanin with l-ethyl-3-(3'-dimethylaminopropy1)carbodiimide HCl.
Generation of Receptor-specific Antisera-In order to characterize a region of the C-receptor by inoculating rabbits with a peptide (P4) recombinant receptor proteins, we prepared specific antisera against homologous to a distinct region of the deduced amino acid sequence (amino acids 118-136) of the cloned protein. New Zealand white rabbits were immunized by intradermal injection of 1 mg of the conjugated peptide emulsified in complete Freund's adjuvant. Each animal was subsequently boosted three times a t 2-week intervals with -600 pg of immunogen. Serum samples were collected from these animals at 2-week intervals and titered according to their ability to immunoprecipitate recombinant radiolabeled ANP C-receptor. Immunoprecipitation was performed according to the method of Kessler (33) with modifications as described previously (20). Endoglycosidase Experiments-Samples of immunoprecipitated and radiolabeled ANP C-receptor were first solubilized by boiling for 5 min in 35 pl of 0.5% SDS and 1% @-mercaptoethanol. Samples were then diluted into 20 pl of buffer containing 250 mM sodium phosphate (pH 8.6) and 2% Nonidet P-40 followed by the addition of 1-2 units of endoglycosidase F Du Pont-New England Nuclear. For endoglycosidase H reactions, immunoprecipitated and solubilized samples were diluted into 10 pl of 100 mM citrate (pH 5.5) followed by the addition of 150 ng of enzyme (Du Pont-New England Nuclear). Incubations were for 12 h at 37 "C. Reaction mixtures were then diluted into SDS sample buffer, boiled, and electrophoresed on a 12.5% polyacrylamide gel as described. Results were visualized via autoradiography.

RESULTS AND DISCUSSION
I n Vitro Translation of Receptor-specific mRNAs-A putative schematic representation of the ANP C-receptor is shown in Fig. 1. Based on the hydropathicity profile of this protein and analogies with related receptor proteins, we have hypothesized that the ANP C-receptor is comprised of a large extracellular domain (436 amino acids), a single-transmembrane region (23 amino acids), and a short carboxyl-terminal cfloplasmic tail (37 amino acids) (25). As a means of examining this structure further, we have used the technique of in vitro mutagenesis to introduce an in-frame stop codon into the cDNA just upstream from the putative hydrophobic transmembrane domain (amino acids 437-459) (Fig. 1). Subcloning of native and mutant cDNAs into the SP6 transcription vector pGEMl allowed the synthesis of specific messenger RNAs corresponding to the truncated mutant as well as full length

FIG. 2. In vitro translation of ANP C-receptor-specific
mRNA. Mutant and native receptor-specific mRNAs were synthesized using the viral SP6 polymerase and translated in vitro using a cell-free reticulocyte lysate (see "Materials and Methods"). Radiolabeled protein products were separated by SDS-polyacrylamide gel electrophoresis and the result visualized by autoradiography. The migration of molecular weight standards is indicated.
recombinant ANP C-receptor. In vitro translation of the full length receptor mRNA resulted in a -56,000-dalton primary translation product as predicted by the amino acid sequence. In contrast, translation of the mRNA derived from the mutant cDNA clone resulted in a truncated receptor protein of -50,000 daltons consistent with removal of 58 amino acids at the carboxyl terminus (Fig. 2).

Expression of Native and Mutant Receptor in L . " Fibro-
blosts-Infection of mouse fibroblasts (L-M cells) with recombinant vaccinia virus containing the native ANP C-receptor cDNA resulted in the expression of a membrane-bound receptor as measured by the specific saturable binding of lZ5I-ANP to confluent cell monolayers (20). As shown in Fig. 3A, unlabeled ANP (102-126) competed for binding to these cells with a Ki of 2.2 nM. No specific ANP binding could be detected in the medium of these cells. In addition, 1251-ANP did not bind to I," cells infected with control vaccinia virus. In contrast to cells infected with recombinant virus encoding native receptor, cells infected with vaccinia virus containing mutant receptor sequences (anc-) exhibited no specific binding to infected monolayers. However, specific binding of lZ5I-ANP was observed in culture medium from these infected cells, demonstrating that this mutant receptor is a secreted protein (Fig. 3B). Binding of lZ5I-ANP to this secreted receptor was saturable (not shown), and competitive binding experiments yielded a Ki for ANP (102-126) of 2.3 nM, a value nearly identical to that seen for the native membrane-associated receptor (5). Data presented in Fig. 3C show that the mutant-secreted receptor possesses the ability to bind actively a ring-contracted ANP analog NH2) as well as truncated linear ANP analogs (not shown) which are specific for the ANP C-receptor (22, 34). This secreted ANP receptor thus retains the affinity and specificity for ANP and C-receptor-specific ANP analogs attributed previously to this receptor in its native environment on the surface of cultured smooth muscle cells.
Immunoprecipitation of the Secreted ANP C-receptor-Immunoprecipitation experiments were performed using the P4 antisera described under "Materials and Methods" and [35S] methionine-labeled L-M cells infected with recombinant vaccinia virus bearing either the intact or truncated (anc-) receptor coding sequences. Immunoprecipitation of whole cell lysates prepared from cells infected with the native receptor construct yielded a 65,000-dalton protein upon electrophoresis. Very little immunoreactive protein was observed in the culture medium (Fig. 4). Conversely, L-M cells infected with anc-display two immunoreactive proteins in the 55,000-60,000-dalton range in the culture medium whereas only a trace amount of a 55,000-dalton protein is evident in the cell lysate itself. The appearance of receptor proteins in the culture medium of cells infected with anc-is consistent with secreted ANP-binding activity.
Experiments were performed to resolve the observed molecular heterogeneity of the secreted ANP C-receptor proteins. Potential N-linked glycosylation of the C-receptor has been suggested by the presence of canonical glycosylation signals in the receptor sequence (25). This contention is supported by the observation that purified ANP C-receptor exhibits an apparent molecular mass of 65,000 upon SDS electrophoresis, which is larger than the value predicted from the amino acid sequence (55,700) and the molecular mass observed upon SDS-gel electrophoresis of the in vitro translation product (56,000 daltons) (24). As shown in Fig. 5, it is evident that the 3sS-labeled receptor has an altered electrophoretic mobility upon treatment with endoglycosidase F or H, indicating that the protein has both high mannose and complex sugars attached to it. The mobility of the endoglycosidase F-treated receptor (-56,000 daltons) is similar to that of the in vitro translation product, suggesting that the difference between the observed molecular mass of the native protein and that predicted from the amino acid sequence is due to N-linked glycosylation. The secreted receptor also appears to be glycosylated, since both immunoreactive electrophoretic bands observed in conditioned medium are shifted to lower molecular mass upon treatment with endoglycosidase H. Both immunoprecipitable bands collapse to a single, broad -50,000dalton protein band upon treatment with endoglycosidase F. Thus, the observed heterogeneity of the secreted receptor protein upon immunoprecipitation presumably represents al- ternate glycosylation products of a single secreted protein.

Secreted ANP C-receptor Exists in a Monomeric State-
The ANP C-receptor migrates as a 120,000-dalton homodimer upon SDS-polyacrylamide gel electrophoresis performed under nonreducing conditions (24,35). However, Iz5I-ANP crosslinking experiments have suggested that the C-receptor may exist as both monomer and dimer on the cell surface, perhaps in equilibrium (20, 36, 37). Results shown in Fig. 6 support this possibility since "S-labeled recombinant receptor is seen

Atrial Natriuretic Peptide Clearance Receptor Structure 14183
to migrate as both a homodimer and smaller monomeric species when gel electrophoresis is performed under nonreducing conditions. The larger of the two small bands in the ANP C-receptor lane of the nonreduced gel may represent unprocessed receptor protein or possibly an unrelated receptor associated protein. Alternatively, these observations may be an artifact of the gel electrophoresis conditions, and conclusions cannot be definitely drawn. However, the data presented here suggest that the mutant secreted ANP C-receptor does not dimerize, as can be seen by the identical 55,000-60,000dalton receptor protein observed under reducing and nonreducing conditions (Fig. 6). The monomeric secreted receptor species can effectively bind lz5I-ANP ligand, indicating that dimerization may not be essential for its ligand-binding activity. However, it is noted that noncovalent dimerization of this receptor may occur. Conclusions-We have demonstrated previously that recombinant expression of the ANP C-receptor protein is sufficient to account for the specific ANP-binding activity observed in bovine smooth muscle cells. We have now characterized the expression of a mutant secreted ANP C-receptor protein with binding affinity for different ANP analogs identical to that observed for the native C-receptor. These data are consistent with the proposed structure of the ANP Creceptor consisting of a large amino-terminal extracellular ligand-binding domain adjacent to a single transmembrane anchor and a short hydrophilic cytoplasmic tail. The extracellular domain of this receptor is alone sufficient for specific high affinity ligand binding, with no contribution necessary from other membrane components or from the membrane or cytoplasmic domains. In addition, it is likely that overexpression of native ANP C-receptor sequences can also result in secretion of a small amount of a truncated receptor protein in addition to a large quantity of membrane-associated receptor. The detection of truncated receptor species increased with extended (>24 h) vaccinia infections (not shown), and its appearance may be the result of proteolysis of membraneassociated C-receptor. Whether release of receptor from the membrane occurs in vivo or is simply an artifact of cell death resulting from extended viral infection has not been determined. However, there are reports of ANP bound to a larger component in serum (38-40), suggesting the existence of naturally circulating ANP-binding proteins that could be derived from the ANP C-receptor.
Glycosylation of the C-receptor has been suggested previously (15,16,25,35) but is here clearly demonstrated. The electrophoretic mobility of the recombinant ANP C-receptor is increased substantially by treatment with endoglycosidase F but only slightly with endoglycosidase H. These differences suggest that the majority of attached carbohydrate is of the modified complex type. It is not clear why the truncated mutant receptor is differentially glycosylated, however transmembrane and cytoplasmic protein domains have been shown previously to be important in signaling transport of integral membrane proteins through the endoplasmic reticulum to the cell surface (41,42). Interference with this transport path may affect proper post-translational modification involving attachment and maturation of carbohydrate. It is not known if carbohydrate side chains on the extracellular domain are necessary or influence ligand-binding properties of the receptor. However, the observation that the secreted ANP Creceptor is aberrantly glycosylated but still shows high affinity for ANP-and C-receptor-specific ANP analogs suggest that specific oligosaccharide structures may play only a minor role in ligand-binding affinity and specificity. This point must be addressed in more detail.
The native C-receptor may exist in monomer-dimer equilibrium on the cell surface. However, the data presented here demonstrate that the secreted receptor does not form covalent dimers and yet binds ANP in a fashion similar to native receptor. Based on this observation, it seems likely that monomeric and dimeric forms of the native ANP C-receptor may function equally well in binding ligand on the cell surface and do not represent separate receptor forms as has been suggested previously (36). The fact that the truncated monomeric receptor is readily secreted from cells implies that covalent dimerization of the native protein is not necessary for its transport to the cell surface and may occur only after the receptor has migrated to the extracellular membrane. It is not known if dimerization of this protein is important for other possible receptor functions such as internalization.
Analogous truncations of other cell surface proteins and receptors containing a single transmembrane domain have also led to the secretion of soluble proteins with immunoreactivity or binding activity similar to that of the native membrane protein (44-46). The overall deduced structure of the ANP clearance receptor is also similar to that of other receptors that are involved primarily in ligand sequestration and internalization rather than activation of second messenger systems leading to biological effects. Examples of these are the insulin-like growth factor II/M6P receptor and the low density lipoprotein receptor (47,48). Both of these receptors contain a single transmembrane domain with a relatively large extracellular ligand-binding domain and are responsible for the internalization of ligand. Other important points that need to be addressed concerning the biological role of this ANP receptor include receptor phosphorylation and cell surface aggregation. We feel that this vaccinia virus expression system will be an efficient tool for examining other ANP Creceptor mutants which will be useful for probing the structure-function relationships of the ANP C-receptor.