Processing of Pro-atrial Natriuretic Peptide by Corin in Cardiac Myocytes*

Corin is a type II transmembrane serine protease abundantly expressed in the heart. In a previous study using transfected 293 cells, we showed that corin converted pro-atrial natriuretic peptide (pro-ANP) to atrial natriuretic peptide (ANP), suggesting that corin is likely the pro-ANP convertase. Because other serine proteases such as thrombin and kallikrein had previously also been shown to cleave pro-ANP in vitro, it remained to demonstrate that corin is indeed the endogenous pro-ANP convertase in cardiomyocytes. In this study, we examined pro-ANP processing in a murine cardiac muscle cell line, HL-5. Northern analysis showed that corin mRNA was present in HL-5 cells. In HL-5 cells transfected with a plasmid expressing pro-ANP, recombinant pro-ANP was converted to mature ANP as determined by Western analysis, indicating the presence of the endogenous pro-ANP convertase in these cells. The processed recombinant ANP was shown to be active in an enzyme-linked immunosorbent assay-based cGMP assay in baby hamster kidney cells. The processing of recombinant pro-ANP in HL-5 cells was highly sequence-specific, because mutation R98A, but not mutations R101A and R102A, in pro-ANP prevented the conversion of pro-ANP to ANP. Expression of recombinant wild-type corin enhanced the processing of pro-ANP in HL-5 cells. In contrast, overexpression of active site mutant corin S985A or transfection of oligonucleotide small interfering RNA duplexes directed against the mouse corin gene completely inhibited the processing of recombinant pro-ANP in HL-5 cells. These results indicate that corin is the physiological pro-ANP convertase in cardiac myocytes.

Atrial natriuretic peptide (ANP) 1 is a cardiac hormone stored in the dense granules of cardiac myocytes. In response to high blood pressure, ANP is secreted into the circulation. In target organs such as kidney and peripheral vessels, ANP binds to its receptor and stimulates the intrinsic guanylyl cyclase activity of the receptor, leading to production of intracellular cGMP.
The biological effects of ANP are to promote salt excretion, reduce blood volume, and relax vessel tension, thereby reducing blood pressure (1,2). The biological importance of the ANP-mediated pathway in maintaining normal blood pressure has been demonstrated in a number of studies. In knockout mice, for example, deficiency in either ANP or its receptor led to spontaneous hypertension (3,4). High plasma concentrations of ANP and brain-type natriuretic peptide (BNP) are found in patients with congestive heart failure. The levels of these natriuretic peptides are often correlated with the extent of ventricular dysfunction and development of cardiac arrhythmias (5,6). ANP and BNP have been used as therapeutic agents in patients with decompensated congestive heart failure and acute myocardial infarction to improve cardiac function and clinical status (7,8). Administration of ANP has also been used as therapy in patients with renal failure (9).
In cardiac myocytes, ANP is synthesized as a 126-amino acid prepropeptide (10,11). After the signal peptide is removed, pro-ANP is stored in the dense granules of the cell. Upon secretion from the dense granules, pro-ANP is activated on the surface of cardiac myocytes by proteolytic cleavage at residue arginine 98, generating an N-terminal propeptide and a mature 26-amino acid C-terminal peptide that is biologically active (12,13). Several studies showed that a high-molecularweight trypsin-like enzyme associated with the membrane of cardiac myocytes was responsible for the activation cleavage of pro-ANP (14 -16). Despite tremendous efforts, the identity of the pro-ANP convertase remained unknown for many years.
Recently, we cloned a unique serine protease, corin, from the human heart (17). Sequence analysis indicates that corin is a mosaic protein composed of a distinctive assortment of domains. At its N terminus, corin has a cytoplasmic domain and an integral transmembrane domain. In the extracellular region of corin, there are two frizzled-like cysteine-rich motifs, eight low density lipoprotein receptor repeats, a macrophage scavenger receptor-like domain, and a trypsin-like protease domain at the C terminus (17,18). The overall topology of corin is similar to those of other type II transmembrane serine proteases of the trypsin superfamily (19), such as hepsin (20,21), and enterokinase (22). In a functional study using transfected human embryonic kidney 293 cells, we showed that recombinant human corin converted pro-ANP to ANP, suggesting that corin is likely the long-sought pro-ANP convertase (23).
To further demonstrate that corin is indeed the endogenous enzyme responsible for the processing of pro-ANP in cardiac myocytes, we studied pro-ANP processing in a murine cardiac myocytic cell line, HL-5. Here we show that expression of recombinant human wild-type corin enhanced the processing of pro-ANP in HL-5 cells. In contrast, overexpression of active site mutant corin S985A or transfection of oligonucleotide siRNA duplexes directed against the mouse corin gene completely blocked the processing of pro-ANP. These results strongly sup-port corin being the physiological pro-ANP convertase in the heart.

EXPERIMENTAL PROCEDURES
Materials-Penicillin, streptomycin, L-glutamine, fetal bovine serum (FBS), and cell culture medium were purchased from Invitrogen (Rockville, MD). The murine cardiac myocytic cell line HL-5 was kindly provided by Dr. William C. Claycomb (Louisiana State University Medical Center, New Orleans, LA). Human embryonic kidney 293 cells and baby hamster kidney (BHK) cells were obtained from the American Type Culture Collection (ATCC) and maintained at the Core Facility at Berlex Biosciences. Anti-V5 antibody was purchased from Invitrogen (Carlsbad, CA). All other chemical reagents were obtained from Sigma Chemical Co. (St. Louis, MO).
Northern Analysis-Poly(A) ϩ RNA samples from mouse heart and lung were purchased from CLONTECH. mRNA from HL-5 cells was isolated using the Micro-Fast Track 2.0 mRNA isolation kit (Invitrogen). RNA samples (2.5 g each) were fractionated on denaturing agarose gels and transferred onto nitrocellulose membranes. Northern hybridization was performed with a 32 P-labeled mouse corin cDNA probe at 42°C overnight in a solution containing 40% formamide, 5% Denhardt's solution, 6ϫ SSC, 100 g/ml salmon sperm DNA, and 0.1% SDS. As a control for mRNA sample loading, Northern blots were reprobed with a mouse ␤-actin cDNA probe (CLONTECH).
Site-directed Mutagenesis-Plasmid constructs expressing human wild-type corin (pcDNACorin), active site mutant corin S985A (pcDNACorinS985A), human wild-type pro-ANP (pcDNAproANP), and mutant pro-ANPs R101A (pcDNAproANPR101A) and R102A (pcDNAproANPR102A) were generated by site-directed mutagenesis and described previously (23). Plasmids expressing active site mutant prothrombin S205A (pPTS205A) and active site mutant hepsin S353A (pHepsinS353A) were generated by a PCR-based mutagenesis method. Plasmid pcDNAproANPR98A expressing mutant pro-ANP R98A, in which residue Arg-98 was replaced by Ala, was constructed by sitedirected mutagenesis using the QuikChange kit (Stratagene, La Jolla, CA) with sense primer 5Ј-ATCTGCCCTCCTAAAAAGCAA-3Ј and antisense primer 5Ј-GATCTCCGCAGGCTCGGCAGGGGCAGTGAGCAG-3Ј. To facilitate detection of recombinant proteins, a tag sequence was included in the constructs that expresses a viral V5 and a His tag at the C terminus of recombinant pro-ANP. All plasmid constructs were confirmed by DNA sequencing.
Transfection and Western Analysis-Transient transfection was performed in HL-5 or 293 cells using Lipofectin (Invitrogen) according to the manufacturer's instructions. Conditioned medium was collected 48 or 72 h after transfection. To analyze pro-ANP processing, recombinant pro-ANP and its derivatives in the conditioned medium were immunoprecipitated by an anti-V5 antibody (Invitrogen). Proteins were sepa-rated by SDS-PAGE and analyzed by Western blotting using a horseradish peroxidase-conjugated anti-V5 antibody (Invitrogen).
cGMP Assay-To examine the activity of recombinant ANP, a cGMP assay was performed using an enzyme immunoassay (EIA) kit (Biotrak, Amersham Biosciences, Inc.). In this assay, BHK cells were grown in 96-well plates in MEM medium supplemented with 10% FBS and 1% of L-glutamine. Confluent cells were washed once with serum-free medium. The conditioned medium (180 l) containing recombinant pro-ANP and its derivatives from transfected HL-5 or 293 cells was added to each well and incubated at 37°C for 10 min. The cells were lysed by addition of a lysis buffer (20 l/well) containing 2% dodecyl trimethylammonium and 50 mM sodium acetate, pH 5.8. The intracellular cGMP concentration in ANP-stimulated BHK cells was determined with the Biotrak EIA kit. Each experimental condition was assayed in quadruplicate.
Effects of Protease Inhibitors-Recombinant pro-ANP was expressed in 293 cells transfected with the pro-ANP-expressing plasmid. The conditioned medium containing recombinant pro-ANP was incubated with HL-5 cells at 37°C for 4 h in the presence of one of the following: 2.5 mg/ml aprotinin, 20 mg/ml benzamidine, 25 mg/ml leupeptin, or 12.5 mg/ml soybean trypsin inhibitor. The conditioned medium was collected, and the processing of pro-ANP was analyzed by SDS-PAGE and Western blotting using an anti-V5 antibody.
Effects of RNA Interference-Oligonucleotide siRNAs were designed based on sequences specific for mouse corin cDNA (5Ј-AACCACCGC-CCUCCGUGCA-3Ј and 5Ј-AACACAAGCACCUGCAUGAAC-3Ј) (24). Antisense and sense siRNA oligonucleotides with dTdT 3Ј-overhang were synthesized and annealed by Dharmacon Research, Inc. (Lafayette, CO). Transfection of HL-5 and 293 cells with the siRNA duplexes was performed according to the manufacturer's instruction. Briefly, HL-5 cells or 293 cells expressing corin were grown in 24-well cell culture plates. Increasing concentrations of siRNA duplexes were added to a solution containing Oligofectamine (3 l) and Opti-MEM (100 l) (Invitrogen) and incubated at room temperature for 25 min. The mixture was then added to HL-5 or 293 cells and incubated at 37°C for 24 h. The cells were washed twice with serum-free medium and then incubated with the conditioned medium containing recombinant pro-ANP at 37°C for 4 h. Pro-ANP and its derivatives in the conditioned medium were analyzed by immunoprecipitation and Western blotting.

Expression of Corin mRNA in HL-5 Cells-
The murine HL-5 cell line was derived from the AT-1 atrial cardiomyocyte tumor lineage developed in transgenic mice overexpressing simian virus 40 large T antigen under the control of the ANP promoter (29,30). The AT-1 cells were maintained by serial subcutaneous grafts in C57BL/6J mice and used to establish cardiac muscle cell lines that are stable under culture conditions. Previous studies showed that a similar cell line, HL-1, derived from the same AT-1 tumor lineage, exhibited electrophysiological and pharmacological features that were characteristic of adult cardiac myocytes (31). To determine if HL-5 cells also maintain a cardiac-specific gene expression profile, the expression of a set of selected cardiac-specific genes was examined by RT-PCR. As shown in Fig. 1, transcripts of the ANP, ␣-MHC, connexin43, and ␣-cardiac actin genes were detected in RNA samples from HL-5 cells and the adult mouse heart. By Northern hybridization (Fig. 2), a corin transcript of ϳ5 kb was detected in mRNA samples from HL-5 cells and the adult heart. In contrast, ANP, ␣-MHC, connexin43, ␣-cardiac actin, and corin mRNAs were not detected in samples from the mouse lung by the RT-PCR or Northern analyses. As a control, ␣-skeletal actin mRNA was detected in samples from the lung but not HL-5 cells and the heart (Figs. 1 and 2). These results are consistent with the previous report that the AT-1 tumor-derived cells maintain characteristics of adult atrial cardiomyocytes (31) and demonstrate that corin mRNA is present in HL-5 cells.
Processing of Pro-ANP in HL-5 Cells-To study the processing of pro-ANP in HL-5 cells, transfection studies were performed using plasmids expressing human pro-ANP and corin. As reported in our previous study (23) and shown here as a control (Fig. 3A), processing of recombinant pro-ANP to ANP in 293 cells was not detected unless a corin-expressing plasmid was co-transfected with the pro-ANP-expressing plasmid. In contrast, in HL-5 cells co-transfection of the corin-expressing plasmid was not required for the conversion of recombinant pro-ANP to ANP. Processing of pro-ANP was detected in HL-5 cells transfected with only the pro-ANP-expressing plasmid (Fig. 3B), suggesting that recombinant pro-ANP was processed by endogenous corin present in HL-5 but not 293 cells. Expression of recombinant corin further enhanced the processing of pro-ANP in HL-5 cells, as shown in Fig. 3B, when the HL-5 cells were co-transfected with the corin-expressing plasmid. In a separate experiment, recombinant pro-ANP was incubated with either HL-5 cells or transfected 293 cells expressing recombinant corin. Western analysis showed similar processing of recombinant pro-ANP by both cell types (Fig. 3C). These results are consistent with the previous findings that corin is a transmembrane protease present on the surface of cardiac myocytes (18,23).
The Activity of Recombinant ANP-The biological function of ANP is mediated through its receptor that is present on the surface of targeted cells. The binding of ANP to its receptor stimulates the intrinsic guanylyl cyclase activity of the receptor, leading to generation of intracellular cGMP. To determine if the corin-processed recombinant ANP is biologically active, a cell-based cGMP assay was performed. As shown in Fig. 4, cGMP-stimulating activity was detected in the conditioned medium from 293 cells transfected with both corin and pro-ANPexpressing plasmids. In contrast, little cGMP-stimulating activity was detected in the conditioned medium from 293 cells transfected with either corin or pro-ANP-expressing plasmids alone (Fig. 4). In a parallel study, low levels of the cGMPstimulating activity were detected in the conditioned medium from parental HL-5 cells (Fig. 4). The activity was most likely due to the presence of native ANP produced in HL-5 cells being processed by the native corin. The cGMP-stimulating activity was significantly increased in the conditioned medium from HL-5 cells transfected with corin and pro-ANP-expressing plasmids, either independently or simultaneously (Fig. 4). These results are consistent with the Western analysis of the processing of recombinant pro-ANP in 293 and HL-5 cells (Fig. 3), and demonstrate that corin-processed recombinant ANP is biologically active.
Sequence Specificity of Pro-ANP Cleavage in HL-5 Cells-In the heart, pro-ANP is processed by proteolytic cleavage at residue Arg-98, generating a 28-amino acid mature C-terminal peptide that is the major circulating form of ANP. In other tissues, including brain, kidney, and testis, proteolytic cleavage of pro-ANP can occur at other residues such as Arg-101 and Arg-102 (32,33). Earlier we had shown that corin specifically cleaved pro-ANP at residue Arg-98 in transfected 293 cells (23). To determine if pro-ANP is also processed at Arg-98 in HL-5 cells, transfection experiments were performed in HL-5 cells using plasmids expressing human wild-type and mutant pro-ANPs. Recombinant pro-ANP and its derivatives in the conditioned medium were analyzed by Western blotting. As shown in Fig. 5, recombinant wild-type pro-ANP and mutant pro-ANPs R101A and R102A were processed in HL-5 cells. In contrast, no cleavage was detected when mutant pro-ANP R98A was expressed in HL-5 cells, indicating that mutation at Arg-98 prevented specific processing of pro-ANP by endogenous corin in the cardiac myocytes.
Effects of Protease Inhibitors-Previous studies showed that the processing of pro-ANP by a partially purified protease present in membrane fractions from cardiac myocytes was inhibited by high concentrations (100 mg/ml) of benzamidine, leupeptin, and aprotinin but not soybean trypsin inhibitor (15). Similar effects of the protease inhibitors were observed on recombinant corin-mediated processing of pro-ANP in 293 cells (23). To examine effects of the protease inhibitors on the processing of pro-ANP in HL-5 cells, recombinant pro-ANP was incubated with HL-5 cells in the presence of these protease inhibitors. Western analysis showed that benzamidine, leupeptin, and aprotinin but not soybean trypsin inhibitor inhibited the processing of pro-ANP by HL-5 cells (Fig. 6). The results further support the hypothesis that corin is the endogenous pro-ANP convertase in the cardiac myocytes.
Effects of Overexpression of an Active Site Mutant Corin-If corin is indeed the endogenous pro-ANP convertase, it may be possible to inhibit corin-mediated pro-ANP processing in HL-5 cells by overexpression of active site mutant corin S985A, in which the active site serine residue has been replaced by Ala. To test this hypothesis, an experiment was performed in which

FIG. 2. Northern analysis of corin mRNA expression.
Northern blots were prepared using mRNA samples derived from HL-5 cells or mouse heart and lung tissues, as described under "Experimental Procedures." Northern hybridization was performed with a mouse corin cDNA probe. Corin mRNA was detected in samples from HL-5 cells and the heart but not the lung (upper panel). As a control, the blot was re-probed with a mouse actin probe. Actin mRNA was detected in all three samples (lower panel).
plasmids expressing pro-ANP and mutant corin S985A were co-transfected in HL-5 cells. As shown in Fig. 7, the processing of recombinant pro-ANP was inhibited when HL-5 cells were co-transfected with increasing concentrations of the plasmid expressing mutant corin S985A. In controls, co-transfection of plasmids expressing active site mutant human hepsin S353A or active site mutant human prothrombin S205A did not inhibit the processing of recombinant pro-ANP in HL-5 cells (data not shown). These results indicate that active site mutant corin S985A competed with endogenous corin, thereby inhibiting the processing of recombinant pro-ANP in HL-5 cells.
Effects of siRNA-To further demonstrate the requirement of corin for the pro-ANP processing in cardiac myocytes, we used the RNA interference (RNAi) technique to block endogenous corin expression in HL-5 cells. This powerful oligonucleotide siRNA-mediated gene-silencing technique has been used successfully to prevent gene expression in cultured mammalian cells (34). In HL-5 cells, co-transfection of the plasmid expressing pro-ANP with increasing concentrations of oligonucleotide siRNA duplexes directed specifically against the mouse corin gene inhibited the pro-ANP processing in a dose-dependent manner (Fig. 8). In contrast, the siRNA duplexes had no inhibitory effects on the pro-ANP processing in 293 cells expressing recombinant human corin, demonstrating the sequence specificity of the siRNA duplexes (Fig. 8). Although we were unable to verify the down-regulation of mouse corin protein expression in the transfected HL-5 cells due to lack of antibodies against mouse corin, these results are consistent with the notion that inhibition of endogenous corin expression prevented the pro-ANP processing in the cardiac myocyte.  in panels A and B). To show that recombinant corin was expressed in transfected cells, Western analysis was performed using an anti-V5 antibody. Recombinant corin was detected in cell lysate from 293 and HL-5 cells transfected with the corinexpressing plasmid (lower sections in panels A and B). C, conditioned medium containing recombinant human pro-ANP was incubated with either HL-5 cells or 293 cells expressing recombinant corin. Processing of pro-ANP was analyzed by Western blotting using an anti-V5 antibody. At high resolution, two bands of pro-ANP were detected on the Western blots possibly caused by differences in glycosylation in transfected cells. peptide hormones, the activation cleavage is mediated by a family of subtilisin-like serine proteases, known as precursor convertases (PCs). These convertases are located in the secretory pathway inside of the cell and cleave proproteins at specific sequences commonly composed of single or paired basic amino acids (35,36). Processing of pro-ANP, however, appears to be mediated by a different mechanism. Following removal of its signal peptide, pro-ANP is stored in the dense granules of cardiac myocytes. The activation cleavage occurs on the cell surface when the pro-hormone is secreted from the cells. Earlier biochemical studies have indicated that a high-molecularweight trypsin-like protease present on the cell surface is responsible for the processing of pro-ANP in cardiac myocytes (14 -16).
Our discovery of the cardiac transmembrane serine protease, corin, and subsequent characterization of its function indicate that corin is probably the pro-ANP convertase. Corin mRNA was detected in tissues such as heart, kidney, testis, uterus, and bone where pro-ANP and other natriuretic peptides are known to be expressed (17). Biochemical studies showed that the corin protein had a molecular mass of ϳ150 kDa and was associated with the cell membrane (23). In transfected 293 cells, recombinant corin cleaved pro-ANP to produce a small peptide that was indistinguishable from the mature ANP by SDS-PAGE and Western analysis. Mutagenesis analysis showed that the corin-mediated cleavage of pro-ANP was highly sequence-specific. It remained, however, to show that corin is indeed the endogenous pro-ANP convertase in cardiac myocytes, because other serine proteases such as thrombin and  kallikrein have also been reported in the past to cleave pro-ANP in vitro (37,38).
In this study, we characterized the corin-mediated pro-ANP processing in a murine cardiac myocyte cell line, HL-5, that was derived from the murine AT-1 atrial cardiomyocyte tumor lineage. Previous studies have shown that AT-1-derived cells maintain a cardiac-specific phenotype (31). AT-1-derived cells, for example, exhibit a gene expression pattern that is very similar to that of adult atrial cardiac myocytes. AT-1-derived cells possess pro-ANP-containing dense granules and are capable of processing the prohormone when it is secreted from the cells. In RT-PCR and Northern analyses (Figs. 1 and 2), we showed that both corin and pro-ANP genes were expressed in HL-5 cells. In a cell-based cGMP assay, ANP activity was detected in the conditioned medium derived from HL-5 cells (Fig. 4), demonstrating endogenous processing of pro-ANP and constitutive secretion of active ANP from these cells. These results indicate that HL-5 cells are an excellent model for studying the role of corin in pro-ANP processing in cardiac myocytes.
In transfection studies, we showed that recombinant pro-ANP was processed by an endogenous enzyme present in HL-5 cells. Characteristically, the processing of pro-ANP in HL-5 cells was indistinguishable from the corin-mediated processing of pro-ANP in 293 cells. In both cell types, for example, the proteolytic activity was present on the cell surface (Fig. 3C), the cleavage in pro-ANP was sequence-specific (Fig. 5), and serine protease inhibitors had similar effects (Fig. 6). We also showed that transfection of a plasmid expressing corin enhanced pro-ANP processing in HL-5 cells and that the corin-processed ANP was biologically active as measured in a cell-based cGMP assay (Fig. 4). Moreover, overexpression of the active site mutant of corin S985A in HL-5 cells completely prevented the pro-ANP processing in HL-5 cells (Fig. 7). The dominant effect of mutant corin S985A in HL-5 cells appeared to be specific, because expression of two other active site mutant trypsin-like serine proteases, hepsin S353A and prothrombin S205A, did not inhibit the pro-ANP processing in HL-5 cells. In RNAi-mediated gene silencing experiments, oligonucleotide siRNA duplexes directed against the mouse corin gene inhibited the pro-ANP processing in HL-5 cells (Fig. 8). The specificity of the oligonucleotide siRNA duplexes for mouse corin and not for other sequences was demonstrated by the lack of effect on pro-ANP conversion in 293 cells expressing human corin (Fig. 8). Together, these results strongly indicate that corin is the physiological pro-ANP convertase in cardiac myocytes.
At the present time, the mechanism by which mutant corin S985A exhibited its dominant negative effect in HL-5 cells is not known, but several mechanistic possibilities exist. It is unlikely that the inhibitory effect of mutant corin S985A was caused by dimerization with the endogenous corin, because there was no evidence in Western analysis that corin formed dimers (23). Similarly, competition for occupancy of a limited number of sites in the plasma membrane is possible but unlikely. A more plausible mechanism is that the mutant corin may compete with the endogenous corin for substrate binding through either its protease domain or its distinct motifs in the propeptide domain such as frizzled-like cysteine-rich motifs, low density lipoprotein receptor repeats and macrophage scavenger receptor-like domain. The frizzled-, low density lipoprotein receptor-, and macrophage scavenger receptor-like motifs are known to be important in protein-protein interactions (39 -41). In addition to competition for substrate binding, the mutant corin may also compete with the endogenous corin for an activator. The corin amino acid sequence indicates that corin, like most other trypsin-like serine proteases, is synthesized as a zymogen and that a proteolytic cleavage at residue Arg-801 by a trypsin-like enzyme is required for its activation (17). At this time, the identity of such a protease activator is completely unknown. Further study of corin will help to better understand corin-mediated pro-ANP processing and may also lead to discovery of other proteins that regulate corin activity.