Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Cyclic guanosine monophosphate (cGMP) is produced from guanosine triphosphate (GTP) by the action of guanylyl cyclases (GCs). GCs exist in two main forms, a soluble form and a particulate, membrane-bound form (1, 2). Each form is activated by distinct agonists: soluble GC is activated by nitric oxide (NO) and NO donors, whereas particulate GC is a plasma membrane receptor for the natriuretic peptides and related hormones (1). The discovery of the role of NO as a relaxant of airway smooth muscle (ASM) and the inhibitory neurotransmitter of inhibitory nonadrenergic noncholinergic (iNANC) nerves in human airways (3) has highlighted the importance of the NO-GC-cGMP pathway as a physiologic regulator of ASM tone. Exogenously administered atrial natriuretic peptide (ANP) elicits a bronchodilator effect at physiologic plasma concentrations (4) and plasma ANP levels are significantly increased in acute asthma (5), suggesting that endogenous ANP may have a physiologic protective role as a bronchodilator (6). In addition to potentially important physiologic and pathophysiologic roles, nitrates and ANP may have a therapeutic role. Pharmacologic concentrations of nitrates given by inhalation (9) or ANP given by inhalation or intravenously (4, 7) cause bronchodilatation, suggesting that pharmacologic elevation of cGMP in ASM may be a target for the treatment of asthma.

Desensitization of adenylyl cyclases in human ASM cells (HASMC) is a well recognized phenomenon after treatment with ₂-adrenoceptor agonists that can reduce the effectiveness of these drugs in asthma. Studies of fatal asthma and in animal models of asthma have shown downregulation of adenylyl cyclase, and this can contribute to disease pathogenesis (10). Although adenylyl cyclases have been the subject of several studies in ASM, the desensitization of GCs has not been studied in ASM from any species. It is important to study the regulation of GC function in ASM because alteration of this function could be important in determining the effectiveness of GC activators as asthma therapy and may also be involved in asthma pathogenesis. It is relevant that attenuated cGMP accumulation has been described in the vasculature after continued exposure to NO donors or ANP (11, 12), suggesting that desensitization of both soluble and particulate GC can occur in other biologic systems. Desensitization can be either homologous when caused by agonist occupancy of the receptor or heterologous when initiated by agents that act through different receptors.

We have previously characterized the GCs present in cultured HASMC. These studies showed that both soluble and particulate forms of GC were abundant in HASMC and that S-nitroso-N-acetyl pencillamine (SNAP) was the most potent activator of soluble GC and ANP the most potent activator of particulate GC (13). In this investigation we used these agents to study whether desensitization of soluble and particulate GCs could occur under suitable experimental conditions. In addition, we used a number of pharmacologic tools to probe the mechanisms involved.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Materials

Fetal calf serum (FCS) was purchased from JRH Biosciences (Sera-Lab Ltd., Sussex, UK), human ANP_1-28 was purchased from NovaBiochem (Nottingham, Nottinghamshire, UK), and Complete was obtained from Boehringer Mannheim (Mannheim, Germany); otherwise, all chemicals were purchased from Sigma-Aldrich Company Ltd. (Poole, Dorset, UK). All concentrations of the reagents shown refer to the final concentration in the cell bathing solution. Plastic wares were purchased from Costar (Cambridge, MA).

Cell Culture

Primary cultures of HASMC were prepared from explants of ASM according to methods previously described (13). We have shown, using morphologic and immunohistochemical staining, that this method produces a relatively pure (> 95%) population of ASM cells. Frozen aliquots of cells were thawed before use and plated at a density of 2 × 10⁴ cells/well in 12-well culture plates containing Dulbecco's modified Eagle's medium (DMEM) with 10% FCS, penicillin G (50 U/ml), streptomycin (50 µg/ml), amphotericin B (2.5 µg/ml), and L-glutamine (4 × 10³ M). All experiments were performed in confluent HASMC that had been growth-arrested for 24 h by serum deprivation. Cells were used between passages 2 and 5. HASMC originating from the same source were used for each experimental design.

cGMP Assay

Cells were washed with sterile phosphate-buffered saline (PBS) at the end of the pretreatment period, then incubated with 10³ M SNAP, 10³ M sodium nitroprusside (SNP), or 10⁶ M ANP for 2 h in the presence of 3-isobutyl-1-methylxanthine (IBMX). IBMX, 10³ M, was used in all experiments (to augment cGMP accumulation and increase the sensitivity of the method) except in the experiments that examined phosphodiesterase (PDE) activity, in which the concentrations were varied as described in RESULTS. A 2-h incubation was the optimum time for measurement of cGMP in our previous study (13). cGMP was measured as described previously (13). Briefly, cells were removed with trypsin (0.25%). cGMP was extracted by adding 1 ml of ice-cold 0.1 M hydrochloric acid to the cell suspension. The resulting suspension was freeze-dried (SB9; Lab Plant Ltd., Huddersfield, Yorkshire, UK) before the measurement of cGMP content using a commercially available enzyme-linked immunosorbent assay (ELISA) kit, RPN 226 (Amersham Ltd., Little Chalfont, Buckinghamshire, UK). The samples were first acetylated with a mixture of acetic anhydride and triethylamine to increase the sensitivity of the assay to 2 fmol/50 µl. All samples were assayed in duplicate. The cells from six wells, not used for cGMP assay, were used to give a representative estimate of protein content in each experiment by the method of Bradford using bovine serum albumin (BSA) as a standard (14).

GC Assays

Soluble and particulate GC activities were measured using previously described methods (15, 16). Briefly, GC activity was measured by using GTP as a substrate. Cells, grown in 100-mm culture dishes, were pretreated for 16 h with vehicle or agonist (10³ M SNAP or 10⁶ M ANP). The cells were washed with PBS; scraped off the dishes in the presence of ice-cold buffer: 5 × 10² M tris(hydroxymethyl)aminomethan hydrochloride (Tris-HCl), pH 7.4, protease inhibitors cocktail (Complete); and, in the case of particulate GC assay, 10³ M NaVO₄ and 10³ M sodium pyrophosphate were used as phosphatase inhibitors. The cells were sonicated and centrifuged at 50,000 × g for 30 to 45 min at 4°C, and the supernatant fraction was used for assaying soluble GC activity; whereas the membrane pellet was washed in the same buffer, then solubilized by incubation for 30 min on ice with a buffer containing 5 × 10² M Tris-HCl (pH 7.4), 5 × 10² M NaCl, 10% glycerol, 1% triton X-100, and Complete before being assayed for particulate GC activity. GC activity assay was performed in a reaction buffer containing 5 × 10² M Tris-HCl (pH 7.4), 10³ M IBMX, and a GTP regenerating system (7.5 × 10³ M phosphocreatine/20 U/ml¹ creatine phosphokinase). The assay was initiated by adding the cell extract to the above solution and 4 × 10³ M magnesium chloride and 10³ M GTP ± 10³ M SNAP (soluble GC assay) or 10⁶ M ANP (particulate GC assay) for 10 min at 37°C. The incubation was terminated with 5 × 10² M sodium acetate, pH 4. Samples were freeze-dried and cGMP was measured by ELISA. The GC activity was expressed as picomoles of cGMP per milligram of protein per min.

Determination of Thiols

Thiol concentrations were determined by the Ellmann method (17). Briefly, cells pretreated with DMEM, 10³ M SNAP, or 10³ M SNP for 16 h were trypsinized, pelleted, and extracted with 250 µl 6% trichloroacetic acid (TCA). TCA-insoluble material was sedimented by microcentrifugation and solubilized by sonication in 0.5% sodium dodecyl sulfate containing 2 × 10² M ethylenediaminetetraacetic acid (EDTA). Separate aliquots of supernatants (250 µl) were combined with Tris-EDTA buffer and 5,5 dithio(2-nitrobenzoic acid) and optical density (OD) was determined at 410 nm using a micro plate reader (MR 5000; Dynatec, West Sussex, UK). Reduced glutathione was used as a standard.

Cytotoxicity Assays

MTT assay. Cells were treated with DMEM, 10³ M SNAP, 10³ M SNP, or 10⁶ M ANP for 24 h. (3-[4,5- dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide; Thi-azolyl blue) (MTT) assay was performed as previously described. Briefly, 20 µl of 5 mg/ml MTT were added to each well for 1 h at 37°C. After the medium was removed, 200 µl of dimethyl sulfoxide was added to solubilize the blue formazan product and the plates were shaken for 5 min. The OD at 630/570 nm was compared with control using a plate reader (MR 5000; Dynatec). Viability was set as 100% in control cells.

Cyclic adenosine monophosphate (cAMP) accumulation. cAMP accumulation produced by 10⁵ M forskolin for 10 min was measured in cells pretreated with DMEM, 10³ M SNAP, 10³ M SNP, or 10⁶ M ANP for 24 h. The intracellular levels of cAMP were measured in cell extracts by a protein binding assay (18).

Protein assay. Cells were treated with DMEM, 10³ M SNAP, 10³ M SNP, or 10⁶ M ANP for 24 h before protein content was measured by the method of Bradford using BSA as a standard (14).

Statistical Analysis

Results are shown as means ± standard error of the mean (SEM) of the indicated number of individual observations. Data are expressed as pmol cGMP/mg protein (or as % controls to allow comparison between experimental protocols). The significance of drug effect was assessed by one-way analysis of variance followed by Student's t test (19) using the SPSS software program (SPSS, Inc., Chicago, IL). P < 0.05 was regarded as significant.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Homologous Desensitization of Soluble GC

Effects of SNAP pretreatment on SNAP-stimulated cGMP accumulation. A total of 16 h pretreatment with 10³ M SNAP caused a rightward shift of the concentration response curve of SNAP-stimulated cGMP accumulation (Figure 1a). cGMP values in control and SNAP-pretreated cells were 6.6 ± 0.3 and 3.7 ± 0.1 (56 ± 1.8%) at 10⁵ M, 18.6 ± 1.6 and 12.0 ± 1.6 (64 ± 8.3%) at 10⁴ M SNAP, and 24.9 ± 0.6 and 18.6 ± 1.3 (75 ± 5.1%) at 10³ M SNAP (n = 4; P < 0.01, 0.05, and 0.01, respectively).

View larger version (16K): [in this window] [in a new window]   Figure 1.   Concentration response curve of cGMP accumulation stimulated by SNAP (a), or ANP (b) in cells pretreated with 103 M SNAP or 106 M ANP, respectively, for 16 h (filled triangles) versus control (filled circles). cGMP levels were measured after 2 h incubation with 105 to 103 M SNAP or 108 to 106 M ANP in the presence of 103 M IBMX compared to basal (IBMX alone). Data represent means ± SEM, n = 4. Where error bars are not shown they lie within the data point. Significantly different from control: *P < 0.05; **P < 0.01.

Time course of SNAP-dependent desensitization. No sig-nificant desensitization was observed after 2 h pretreatment with 10³ M SNAP; 51% reduction was seen in the cGMP response to SNAP after 8 h and 60% after 16 h pretreatment. The 10³ M SNAP-stimulated cGMP value in control cells was 7.6 ± 0.9 pmol/mg protein; and in pretreated cells, 8.8 ± 0.2 after 2 h pretreatment (n = 4, P = 0.3), 3.7 ± 0.2 pmol/mg protein after 8 h pretreatment (n = 4, P < 0.05), and 3 ± 0.1 pmol/mg protein after 16 h pretreatment (n = 4, P < 0.05) (Figure 2a).

View larger version (42K): [in this window] [in a new window]   Figure 2.   Time course of desensitization of GC. cGMP levels were measured after 2 h incubation with 103 M SNAP (a) or 106 M ANP (b) (in presence of 103 M IBMX) in cells pretreated with the same agonist for 2, 8, and 16 h. Data represent means ± SEM, n = 4. Significantly different from control: *P < 0.05.

Concentration responses of SNAP-dependent desensitization. Experiments were performed by preincubating cells with different concentrations of SNAP for 16 h. SNAP- dependent desensitization was concentration-dependent (Figure 3a), being significant only at the highest concentration of SNAP used (10³ M). cGMP values in control and pretreated cells were 17.4 ± 0.8, 17.3 ± 0.8 (100 ± 4.4%) after 10⁷ M SNAP pretreatment, 16.7 ± 1.6 (96 ± 9.3%) after 10⁶ M SNAP pretreatment, 16.4 ± 0.5 (94 ± 2.9%) after 10⁵ M SNAP pretreatment, 16.1 ± 0.1 (92 ± 0.8%) after 10⁴ M SNAP pretreatment, and 4.7 ± 0.9 (27 ± 5%) after 10³ M SNAP pretreatment (n = 4, P = 0.9, 0.7, 0.3, 0.2, and < 0.001, respectively).

View larger version (48K): [in this window] [in a new window]   Figure 3.   Concentration response of desensitization of GC. cGMP levels were measured after 2 h incubation with 103 M SNAP (a) or 106 M ANP (b) (in presence of 103 M IBMX) in cells pretreated with 107 to 103 M SNAP or 1010 to 106 M ANP, respectively, for 16 h. Data represent means ± SEM, n = 5. Significantly different from control: *P < 0.05; ***P < 0.001.

Cross-desensitization between SNP and SNAP. A total of 16 h pretreatment with either 10³ M SNAP or 10³ M SNP decreased cGMP-stimulated accumulation in response to SNP or SNAP, respectively. SNAP-stimulated cGMP accumulations in control and SNP-pretreated cells were 23 ± 0.9 and 12 ± 0.6 pmol/mg protein, respectively (n = 4, P < 0.001; Figure 4a). SNP-stimulated cGMP accumulations in control and SNAP-pretreated cells were 20 ± 0.9 and 9 ± 2.2 pmol/mg protein, respectively (n = 4, P < 0.01; Figure 4b). SNP pretreatment also decreased SNP-stimulated cGMP accumulation from 19.7 ± 1.7 to 7.9 ± 0.5 pmol/mg protein (n = 4, P < 0.001).

View larger version (23K): [in this window] [in a new window]   Figure 4.   Effect of 16 h pretreatment with 103 M SNP (a) or 103 M SNAP (b) on SNAP- and SNP-stimulated cGMP accumulation. cGMP levels were measured after 2 h incubation with 103 M SNP and 103 M SNAP in the presence of 103 M IBMX compared to basal (IBMX alone). Data represent means ± SEM, n = 4. Significantly different from control: *P < 0.05; **P < 0.01; ***P < 0.001.

Effects of SNAP and SNP pretreatment on thiol levels. A total of 16 h pretreatment with 10³ M SNAP or 10³ M SNP did not affect the protein-bound thiol content of the cells. Thiol levels were 13 ± 0.8 µM in controls, 14 ± 1.0 µM in SNAP-treated cells, and 14 ± 0.3 µM in SNP-treated cells (n = 4, P = 0.7 and 0.6 versus controls, respectively). Parallel-processed cultures pretreated at the same time showed a decreased cGMP response as compared with controls (data not shown).

Effects of hemoglobin on SNAP-induced desensitization. To determine whether NO donation was involved in the mechanism of SNAP-dependent desensitization, 10⁴ M hemoglobin (Hb), a NO scavenger, was added during the pretreatment period with 10³ M SNAP. Hb prevented the development of tolerance to subsequent stimulation with 10³ M SNAP. SNAP-stimulated cGMP values were 54 ± 1.7% and 106 ± 6.7% in the absence and presence of Hb, respectively (n = 4, P < 0.001 and 0.6, respectively).

Role of PDEs. Similar degrees of SNAP-dependent desensitization were obtained regardless of the level of PDE inhibition during after-pretreatment stimulation with SNAP. SNAP-stimulated cGMP values in pretreated cells were 74 ± 1.3% of control when no IBMX was used; and 83 ± 4.4%, 70 ± 2.7%, and 75 ± 5.1% of control when 10⁵, 10⁴, and 10³ M IBMX was used, respectively (n = 4, P = 0.09, 0.2, and 0.9, respectively).

Effects of SNAP-pretreatment on soluble GC activity. Soluble GC activity was measured in extracts of cells pretreated with 10³ M SNAP for 16 h (Figure 5a). SNAP pretreatment reduced soluble GC activity. Soluble GC activity in control and SNAP-pretreated cells were 334 ± 20.3 and 254 ± 9.2 fmol/mg protein/min (n = 4, P < 0.05).

View larger version (27K): [in this window] [in a new window]   Figure 5.   Effect of 16 h pretreatment with 103 M SNAP or 106 M ANP on soluble (a) and particulate (b) GC activity, respectively. Soluble and particulate GC were purified as described in MATERIALS AND METHODS and activity was assayed in the absence or presence of 103 M SNAP or 106 M ANP, respectively, for 10 min. GC activity is expressed as picomoles per milligram of protein per min. Data represent means ± SEM. Significantly different from control: *P < 0.05.

Homologous Desensitization of Particulate GC

Effects of ANP pretreatment on ANP-stimulated cGMP accumulation. A total of 16 h pretreatment with 10⁶ M ANP caused a rightward shift of the concentration response curve of ANP-stimulated cGMP accumulation (Figure 1b). cGMP values in control and ANP-pretreated cells were 2.4 ± 0.3 and 1.5 ± 0.2 (64 ± 6.6%) at 10⁸ M ANP, 3.0 ± 0.3 and 1.8 ± 0.2 (60 ± 6.1%) at 10⁷ M ANP, and 5.1 ± 0.4 and 2.9 ± 0.3 (56 ± 5.3%) at 10⁶ M ANP (n = 4, P = 0.07, < 0.05, and 0.05, respectively).

Time course of ANP-dependent desensitization. No significant desensitization was observed after 2 and 8 h pretreatment with 10⁶ M ANP. A 62% reduction was seen in the cGMP response to ANP after 16 h pretreatment. The cGMP response to 10⁶ M ANP in control cells was 1.7 ± 0.6 pmol/mg protein; and in pretreated cells, 1.6 ± 0.2 after 2 h pretreatment (n = 4, P = 0.9), 1.5 ± 0.3 pmol/mg protein at 8 h pretreatment (n = 4, P = 0.8), and 0.6 ± 0.2 pmol/mg protein after 16 h pretreatment (n = 4, P < 0.05) (Figure 2b).

Concentration responses of ANP-dependent desensitization. ANP-dependent desensitization was concentration-dependent (Figure 3b), being significant only at the highest concentration of ANP used (10⁶ M). cGMP values in control and pretreated cells were 46.5 ± 3.8 (113 ± 9.9%) after 10¹⁰ M ANP pretreatment, 50.2 ± 5.5 (108 ± 11.8%) after 10⁹ M ANP pretreatment, 41.2 ± 2.1 (89 ± 4.6%) after 10⁸ M ANP pretreatment, 41.9 ± 2.4 (90 ± 5%) after 10⁷ M ANP pretreatment, and 34.4 ± 2.7 (74 ± 5.8%) after 10⁶ M ANP pretreatment (n = 4, P = 0.3, 0.6, 0.3, 0.3, and < 0.05, respectively).

Role of PDEs. As with the response to SNAP, the percentage of inhibition of ANP-stimulated cGMP formation was similar regardless of whether IBMX was added during cGMP assay. In a separate series of experiments to test this, ANP-stimulated cGMP values in pretreated cells were 89 ± 2.3% of control when no IBMX was used; and 93 ± 3.5%, 86 ± 4.7%, and 91 ± 3.1% of control when 10⁵, 10⁴, and 10³ M IBMX was used, respectively (n = 4, P = 0.4, 0.9, and 0.5, respectively).

Effects of ANP-pretreatment on particulate GC activity. To determine whether ANP-dependent desensitization was due to downregulation of particulate GC activity, we measured particulate GC activity in membranes from ANP-pretreated cells. ANP pretreatment decreased particulate GC activity (Figure 5b). Particulate GC activity in control and pretreated cells was 364 ± 15.1 and 232 ± 7.9 fmol/mg protein/min (n = 4, P < 0.001).

Recovery of GC Responsiveness

SNAP- and ANP-dependent desensitization was partially or completely reversible after the agonist was washed out. When cells were treated with 10³ M SNAP or 10⁶ M ANP for 24 h, responses to subsequent stimulation with the same agonist were 55 ± 5.3% and 80 ± 2.8% of control, respectively (n = 4 and 16, P < 0.01 for both compared with control). However, when cells were maintained in DMEM for a further 24 h, cGMP responses increased to 82 ± 2.1% and 118 ± 9.0% of control, respectively (n = 4, P < 0.01 for both compared with cells not allowed to recover) (Figure 6).

View larger version (37K): [in this window] [in a new window]   Figure 6.   Recovery from SNAP- (a) and ANP- (b) dependent desensitization. cGMP levels were measured after 2 h incubation with 103 M SNAP or 106 M ANP in cells pretreated with 103 M SNAP or 106 M ANP for 24 h and then left to recover in DMEM only or DMEM + 106 M cycloheximide (CHX), 105 M dexamethasone (DEX), or 106 M actinomycin D (ACT) for another 24 h. Data represent means ± SEM, n = 4. Significantly different from desensitized cells: ##P < 0.01. Significantly different from recovered cells: *P < 0.05; ***P < 0.001.

The addition of 10⁶ M cycloheximide or 10⁵ M dexamethasone (protein synthesis inhibitors), or 10⁶ M actinomycin D (transcriptional inhibitor) after washing out the agonists (during the recovery period) inhibited recovery of cGMP response to SNAP and ANP (Figure 6). The 10³ M SNAP-stimulated cGMP accumulations were 49 ± 4.3%, 53 ± 2.6%, and 68 ± 4.0% after treatment with cycloheximide, dexamethasone, and actinomycin D, respectively (n = 4, P < 0.001, 0.001, and 0.05, respectively, compared with recovered cells) (Figure 6a). The 10⁶ M ANP- stimulated cGMP accumulations were 90 ± 5.6%, 85 ± 2.5%, and 85 ± 3.8% after treatment with cycloheximide, dexamethasone, and actinomycin D, respectively (n = 16, P < 0.05 for all compared with recovered cells) (Figure 6b). None of the three protein synthesis inhibitors had an effect on basal or stimulated cGMP levels in control cells over 24 h (data not shown).

Heterologous Desensitization

Effects of ANP pretreatment on SNAP-stimulated cGMP elevation. A total of 16 h pretreatment with 10⁶ M ANP did not decrease SNAP-stimulated cGMP accumulation (Figure 7a). cGMP responses to SNAP in control and ANP-pretreated cells were 16.6 ± 3.1 and 17.0 ± 3.6 (103 ± 21.5%) at 10⁵ M SNAP, 32.0 ± 1.8 and 36.0 ± 1.7 (112 ± 5.5%) at 10⁴ M SNAP, and 41.4 ± 1.8 and 45.5 ± 0.6 (110 ± 1.4%) at 10³ M SNAP (n = 4, P = 0.9, 0.2, and 0.1, respectively).

View larger version (16K): [in this window] [in a new window]   Figure 7.   Concentration response curve of cGMP accumulation stimulated by SNAP (a) and ANP (b) in cells pretreated with 106 M ANP, or 103 M SNAP, respectively, for 16 h (filled triangles) versus control (filled circles). cGMP levels were measured after 2 h incubation with 105 to 103 M SNAP or 108 to 106 M ANP in the presence of 103 M IBMX compared to basal (IBMX alone). Data represent means ± SEM, n = 4. Where error bars are not shown they lie within the data point. Significantly different from control: **P < 0.01.

Effects of SNAP pretreatment on ANP-stimulated cGMP elevation. A total of 16 h pretreatment with 10³ M SNAP did not decrease ANP-stimulated cGMP accumulation (Figure 7b). cGMP responses to ANP in control and SNAP-pretreated cells were 3.4 ± 0.4 and 4.0 ± 0.5 (118 ± 14.3%) at 10⁷ M ANP, and 6.3 ± 0.3 and 6.2 ± 0.8 (98 ± 12.2%) at 10⁶ M ANP (n = 4, P = 0.4 and 0.9, respectively). Pretreated cells generated higher cGMP levels compared with control cells at 10⁸ M ANP; cGMP values in control and SNAP-pretreated cells were 1.0 ± 0.2 and 2.0 ± 0.2 pmol/mg protein (202 ± 16.7%) (n = 4, P < 0.01).

Effects of zaprinast pretreatment on SNAP- and ANP-stimulated cGMP elevation. To further determine whether increased cGMP would cause heterologous desensitization, we studied the effect of pretreatment with zaprinast, a cGMP-specific PDE inhibitor, on SNAP- and ANP-stimulated cGMP accumulation. A 24-h pretreatment of cell with 10⁴ M zaprinast increased basal levels of cGMP from 9 ± 0.9 to 12 ± 0.4 pmol/mg protein (n = 4, P < 0.05). There was no evidence of impaired cGMP responses to SNAP or ANP after zaprinast pretreatment. In fact, zaprinast-pretreated cells generated more cGMP on subsequent stimulation with SNAP or ANP. cGMP values in control and zaprinast-pretreated cells were 11 ± 0.4 and 15 ± 0.7 pmol/mg protein in response to 10³ M SNAP (n = 4, P < 0.01), and 17 ± 1.0 and 26 ± 0.9 pmol/mg protein in response to 10⁶ M ANP (n = 4, P < 0.001).

Assessment of Cell Viability

There was no evidence of cytotoxicity after 24 h incubation with either SNAP, SNP, or ANP. MTT assay values were 101 ± 5.2%, 103 ± 2.9%, and 100 ± 1.4% of control in cells pretreated with 10³ M SNAP, 10³ M SNP, and 10⁶ M ANP, respectively (n = 6, P = 0.9, 0.4, and 0.7, respectively). Forskolin-stimulated cAMP accumulations were 123 ± 0.1%, 108 ± 4.0%, and 99 ± 4.3% of control in cells pretreated with 10³ M SNAP, 10³ M SNP, and 10⁶ M ANP, respectively (n = 4, P = 0.1, 0.1, and 0.8, respectively). Protein content values were 109 ± 3.5%, 116 ± 3.0%, and 103 ± 9.5% of control in cells pretreated with 10³ M SNAP, 10³ M SNP, and 10⁶ M ANP, respectively (n = 2, P = 0.3, 0.2, and 0.8, respectively).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

We have previously shown that cultured HASMC contain both soluble and particulate forms of GCs (13). In the present study we have demonstrated homologous, but not heterologous, desensitization of both soluble and particulate GC in HASMC after prolonged treatment with SNAP and ANP, respectively.

Pretreatment of HASMC with SNAP, an activator of soluble GC, led to the development of tolerance to further application of SNAP in a time- and concentration-dependent manner. Our cytotoxicity experiments showed that this was not due to nonspecific effect. We performed several experiments to study the mechanism of this homologous desensitization. Several mechanisms for tolerance to the effects of nitrates have been suggested in other biologic systems. These include: (1) impaired biotransformation of nitrates to NO due to depletion of intracellular thiols (20, 21); (2) increased cGMP breakdown due to increased PDE activity (22); and (3) desensitization of soluble GC due to either molecular alteration or decreased abundance (16, 23). The effect of SNAP pretreatment was blocked by coincubation with hemoglobin, a NO scavenger, suggesting that desensitization was NO-mediated. The fact that we saw cross-desensitization between SNAP, a thiol-independent NO donor, and SNP, a thiol-dependent NO donor (24), suggests that thiol depletion was not involved in the desensitization mechanism. Consistent with this, we found no reduction of thiol concentrations after either SNAP or SNP treatment. To determine whether altered PDE activity was involved, we performed experiments in the absence and presence of IBMX during the cGMP assay. We found a similar degree of desensitization in the absence and presence of IBMX, which excludes the possibility of altered PDE activity contributing to the desensitization we saw. Our experiments in which we measured soluble GC activity suggest that SNAP-dependent desensitization was due to an effect on GC activity. These experiments also exclude the possibility that depletion of GTP substrate for soluble GC was involved in the desensitization because these experiments were performed in an excess of GTP. Because the reversal of tolerance in HASMC was inhibited by coincubation with protein synthesis inhibitors cycloheximide and dexamethasone and the transcriptional inhibitor actinomycin D, the recovery must have required de novo synthesis of a protein. The most likely explanation is that reduced synthesis of soluble GC subunits was responsible. The time dependence of the desensitization produced by SNAP is consistent with this, as it took several hours to occur. Our results with SNAP in HASMC are consistent with previous reports with NO donors in a number of other cultured cell systems, including rat lung fibroblasts (15), rat medullary interstitial cells (16), kidney epithelial cells (25), vascular smooth-muscle cells (11), and isolated vascular smooth-muscle preparations (23, 26). These studies also showed that SNAP- dependent desensitization was due to reduction in soluble GC activity, and studies by Schroder and colleagues (15) and Ujiie and associates (16) suggested that it was due to an absolute reduction in soluble GC abundance.

We performed similar experiments looking at particulate GC. Pretreatment of HASMC with ANP attenuated cGMP accumulation upon re-exposure to ANP, showing homologous desensitization of particulate GC. These findings are similar to studies showing decreased ANP-stimulated cGMP elevation after pretreatment with ANP in cultured rat medullary interstitial cells (16) and rat vascular smooth muscle (12, 27). As with soluble GC, there are several possible explanations for tolerance to activators of particulate GC. These include reduced substrate availability, increased PDE activity, and reduction in particulate GC activity. Because a reduction in particulate GC activity was seen after ANP treatment when assayed in an excess of GTP, substrate depletion cannot explain the effect of ANP in our study. Our experiments showing a similar degree of desensitization in the presence and absence of IBMX suggest that, as with soluble GC, altered PDE activity was not a contributor. When we measured particulate GC activity after ANP stimulation we found that it was reduced, suggesting that ANP causes desensitization by reducing particulate GC activity. The fact that recovery from ANP-dependent desensitization was blocked by coincubation with protein synthesis and transcriptional inhibitors suggests that recovery required de novo protein synthesis. The time dependence of the desensitization produced by ANP was consistent with this. Collectively, these findings suggest that ANP-dependent desensitization is due to a reduction in the level of particulate GC protein.

Having shown that homologous desensitization of soluble or particulate GC could occur after stimulation with either SNAP or ANP, respectively, we performed additional experiments to determine whether heterologous desensitization could occur. To test for this, we studied cross- desensitization between SNAP and ANP. We found no evidence of cross-desensitization in either direction. This finding was supported by lack of significant recovery of GC activity at 2 h (data not shown). To test the hypothesis further we used zaprinast, a selective inhibitor of cGMP-specific PDE, to elevate cGMP. Consistent with the lack of cross-desensitization between SNAP and ANP, zaprinast had no effect on SNAP- and ANP-mediated cGMP accumulation despite its causing a significant elevation in basal cGMP levels. Our experiments therefore showed no evidence of heterologous desensitization of soluble or particulate GCs in HASMC. Several studies have examined heterologous desensitization of soluble GC in other biologic systems, with contrasting results. The lack of heterologous desensitization in our experiments is similar to studies in cultured rat lung fibroblasts (15) and rat and human vascular smooth muscle (23). In contrast, one study showed heterologous desensitization of SNP-stimulated cGMP accumulation after ANP pretreatment in cultured rat medullary interstitial cells (16). These differences most likely reflect intrinsic, cell-specific differences in the regulatory mechanisms of GCs.

What is the significance of our findings for airway function in asthma? Studies of adenylyl cyclase activators (₂-adrenoreceptor agonists) have shown that tolerance can develop when high concentrations of these agents are used in cell culture systems such as HASMC (28). This tolerance is mimicked in vivo when high doses of ₂-adrenoreceptor agonists are used in the treatment of patients with asthma. Our studies suggest that tolerance may occur to activators of GCs in HASMC. Tolerance to SNAP and ANP in our study occurred only when high concentrations were used, which is similar to the studies with ₂-adrenoreceptor agonists. Although it is not known whether tolerance will develop to the airway effects of GC activators when they are given to asthmatic patients in vivo, this may potentially limit their effectiveness when given in high doses.

In addition to having relevance to their pharmacologic effects, we have considered whether tolerance to activators of GC may be important during asthmatic inflammation. This would seem possible with soluble GC because NO has an important physiologic protective role as an iNANC relaxant neurotransmitter. Inflammatory mediators such as interleukin-1 present in asthma have been shown to induce NO synthase, increasing NO production by airway epithelium and inflammatory cells (29). NO produced by these cells would have the capacity to desensitize soluble GC in ASM and thereby impair NO-mediated relaxation by iNANC (3). The pathophysiologic relevance of the desensitization of particulate GC is less clear, as the physiologic role of ANP in the airways is unknown. ANP is, however, produced locally in the lung (30); plasma ANP levels are increased in acute asthma (5); and exogenously administered ANP is protective against methacholine- induced bronchoconstriction (8). This suggests that ANP may have a role in protecting against bronchoconstriction. Our results suggest that this protective effect could potentially become downregulated under circumstances where high ANP concentrations occur.

In conclusion, we have shown that exposure of HASMC to either SNAP or ANP causes homologous desensitization of soluble and particulate GC, respectively, but we were unable to find any evidence of heterologous desensitization. Homologous desensitization to both soluble and particulate GC appears to be due to downregulation of GC activity rather than to an increase in PDE activity. These findings are relevant for asthma pathophysiology and the pharmacologic treatment of asthma.