INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

THE PATHOGENESIS OF intimal hyperplasia, which has been implicated as the major cause of failure of prosthetic bypass grafts and angioplasty procedures, includes proliferation of medial smooth muscle cells (SMC) and their migration to the intima in response to vascular injury (35). It has been hypothesized that, after the endothelium is disrupted, SMC are directly exposed to blood flow and that their function might be modulated by changes in the local hemodynamic environment (17, 46). In vivo studies (16) with rat carotid artery balloon catheter injury have demonstrated that areas of low shear stress had significantly more intimal hyperplasia than areas of high shear stress. Work from our laboratory (29) has supported the above studies by demonstrating that the proliferation rate of cultured human aortic SMC (HASMC) decreased with increasing shear stress and that this reduction was not due to shear stress-induced cell injury or detachment. In addition, recent modeling studies (44) have indicated that, in intact arteries, underlying SMC are exposed to significant shear stresses due to interstitial flow driven by transmural pressure gradients. Although the transmural flow velocity is low, due to the existence of very thin boundary layers around the SMC associated with the fibrous matrix, the calculated wall shear stress on the SMC surface is on the order of 0.1-10 dyn/cm² (44), which is in the range known to affect endothelial cells in vitro.

Nitric oxide (NO) is a relatively short-lived molecule that is involved in many physiological functions, including vasorelaxation, reduction of platelet aggregation, and inhibition of adhesion of leukocytes to the vascular wall (7). In addition, the cytostatic and cytotoxic actions of NO on various cell types suggest that increased NO production may play a role in the pathogenesis of numerous acute inflammatory and autoimmune diseases (38). Most of the aforementioned responses are associated with the activation of soluble guanylate cyclase, which leads to accumulation of guanosine 3',5'-cyclic monophosphate (cGMP) in target cells (7, 36). To date, three major isoforms of NO-producing enzymes have been identified in various cell types. One subtype is the inducible nitric oxide synthase (NOS II), which is regulated at the transcriptional level, requires several hours to be expressed, produces high levels of NO for extended periods, and is Ca²⁺ independent. NOS II can be induced in macrophages, hepatocytes, vascular SMC, endothelial cells, and mesangial cells on stimulation with cytokines and/or bacterial lipopolysaccharides (7, 38). The other two isoforms are constitutively expressed, release low levels of NO immediately on stimulation, and are Ca²⁺ dependent. These isoforms are termed NOS I, found in neurons, epithelial cells, pancreatic islets, skeletal muscle, kidney macula densa cells, human keratinocytes, and retina tissue, and NOS III, found in vascular endothelium, epithelial cells, cardiac myocytes, platelets, monocytes, hippocampal neurons, and macrophages (30, 37, 38).

Mechanical deformation of the endothelium by defined flow-induced shear stress or cyclic stretching increases both NO production and NOS III mRNA, protein, and enzymatic activity in vitro (2, 27, 32). Studies in large arteries (42) indicate that changes in blood flow induce vasodilation that is mediated by endothelial release of NO due to NOS III and that is diminished immediately after endothelial denudation. Endothelium-independent relaxation of rat aortic rings was observed from the perfusate of interleukin-1 (IL-1)-treated cultured SMC due to NO produced from the activation of NOS II (34). In contrast, Bevan and co-workers (4, 12) demonstrated that in certain resistance arteries (such as the middle cerebral artery and small ear arteries), flow-induced relaxation was largely preserved even after endothelial denudation and was dependent on guanylate cyclase but not on NOS. It appears that in these arteries, after removal of the endothelium, fluid shear stress influences smooth muscle tone by acting directly on luminal SMC. In another report, endothelium-denuded cerebral arteries were relaxed with the use of vasoactive intestinal peptide (VIP), and this response was inhibited by NOS and guanylate cyclase inhibitors (11). Apparently, the production of VIP activates a constitutive form of NOS in SMC exposed to flow by an unknown mechanism. Furthermore, it has been shown (25) that a constitutive Ca²⁺-calmodulin NOS is present in gastric and intestinal SMC that appears to be different from NOS I and NOS III and is activated by VIP. In addition, positive immunostaining for NOS I has been observed in pulmonary artery and vein endothelial cells, SMC from newborn rats (22), and SMC of human cerebral blood vessels (40).

Using antibodies for NOS I, Topors et al. (41) recently detected a specific 100-kDa band in both unstimulated and stimulated (with endothelin-1) rat mesenteric artery SMC with enzymatic activity. In addition, a labile vascular SMC muscle-derived relaxing factor with pharmacological and chemical properties similar to NO was generated by perfusion of endothelium-denuded arterial rings after 24 h of incubation in 37°C (45). Human mesentery SMC release biologically active NO in culture that relaxes isolated rat mesenteric resistance arteries in a microperfusion system (3). For these studies, the possibility cannot be excluded that NO production is due to the induction of NOS II, possibly by contaminating lipopolysaccharides in the medium, rather than to a constitutively expressed NOS in SMC. To date, most investigators have detected NO production in vascular SMC only after stimulation with cytokines due to activation of NOS II (38).

In this study, we investigated whether vascular SMC are capable of producing NO on exposure to fluid flow and determined the NOS isoform(s) involved in the process. In pathophysiological conditions, flow-induced NO production by SMC exposed to blood may regulate the vascular tone and wound healing and may modulate the interactions of blood cells with the vessel wall. Alternatively, in the normal vascular wall, interstitial flow-induced NO production in SMC may contribute to vasodilation and to the maintenance of the quiescent, contractile phenotype of SMC. Using parallel-plate flow chambers with recirculating flow loops, we exposed cultured HASMC to different levels of shear stress, and we report herein that HASMC produced NO on exposure to flow by activation of a constitutively expressed NOS I isoform.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Materials. N^G-amino-L-arginine (L-NAA; flavianate salt), dexamethasone, cycloheximide, calmidazolium (CMZ), and isobutylmethylxanthine (IBMX) were purchased from Sigma Chemical (St. Louis, MO). 1,2-Bis(2-aminophenoxy)ethane-N,N,N',N'-tetraacetic acid acetoxymethyl ester (BAPTA-AM) "cell permeant" was purchased from Molecular Probes (Eugene, OR). Monoclonal antibodies for the three isoforms of NOS enzyme were purchased from Transduction Labs (Lexington, KY). As an immunogen for NOS I antibody, a protein fragment corresponding to amino acids 1,095-1,289 of human NOS I was used. For NOS II antibody, a protein fragment corresponding to amino acids 961-1,114 of mouse NOS II was used. Similarly, a protein fragment corresponding to amino acids 1,030-1,209 of human NOS III was used for the NOS III antibody. A rat pituitary tumor cell line and human umbilical vein endothelial cells, supplied together with the NOS antibodies from Transduction Labs, were used as positive controls for NOS I and NOS III, respectively. A human glioblastoma cell line (A-172), purchased from American Type Culture Collection (ATCC; Rockville, MD), stimulated for 24 h with 50 ng/ml tumor necrosis factor-, 10 ng/ml IL-1, 100 U/ml interferon-, and 10 µg/ml lipopolysaccharides (Salmonella minnesota) was used as a positive control for NOS II. All cytokines were purchased from R&D Systems (Minneapolis, MN). The radio immunoassay kit for [¹²⁵I]-labeled cGMP was purchased from Amersham.

Cell culture and shear stress apparatus. An HASMC line initiated from the abdominal aorta of a 9-year-old kidney transplant donor was used in all experiments performed in this study (29). The cells were characterized as SMC by the hill-and-valley pattern displayed at confluency and by positive immunostaining with a monoclonal antibody to -actin (Sigma Chemical). The culture medium was Dulbecco's modified Eagle's medium (DMEM; GIBCO BRL, Gaithersburg, MD) supplemented with 20% fetal bovine serum (Hyclone Labs, Logan, UT), 2 mM L-glutamine, 200 U/ml penicillin, and 100 µg/ml streptomycin (GIBCO BRL). Sterile techniques were rigorously used because NOS II can be induced by bacterial lipopolysaccharides. For the nitrite experiments, phenol red-free DMEM was used to prevent interference with the fluorometric assay, whereas for the cGMP experiments, serum-free DMEM was used. HASMC (passages 2-10) were plated at a subconfluent density of 2.5 × 10⁴ cells/cm² on glass slides (75 × 38 mm; Fisherbrand) coated with 1 µg/cm² human plasma fibronectin (Collaborative Biomedical Products).

Nitrite assay. The enzymatic production of NO from L-arginine by cultured HASMC was assayed by measuring nitrite, an oxidation product of NO. Recent studies have indicated that nitrite levels reflect >75% of the total NO produced by vascular SMC in culture (36). Conditioned media were collected at different time points from both stationary control and cultures exposed to flow, and nitrite levels were measured with a quantitative fluorometric assay (23). This assay is based on the reaction of nitrite with an acid form of 2,3-diaminonaphthalene (DAN; nonfluorescent; Aldrich) to form the highly fluorescent product 1-H-naphthotriazole. Twenty microliters of freshly prepared DAN (0.05 mg/ml in 0.62 N HCl) was added to 200 µl of sample diluted 1:5 with sterile water (Baxter), and the mixture was incubated in the dark at room temperature for 10 min. The intensity of the fluorescent product was maximized by the addition of 10 µl of 2.8 N NaOH, and the signal was measured using a fluorescent 96-well plate reader (Cytofluor 2350, Millipore, Bedford, MA) with excitation at 365 nm and emission at 460 nm. Nitrite concentrations were determined relative to a standard curve. Sodium nitrite standards (Sigma Chemical) were made fresh in the sample buffer (including serum components) and were kept on ice. This assay gave a linear fluorescence response from 5 µM to 50 nM and was 20 times more sensitive than the Greiss colorimetric assay for nitrite (23). Total nitrite production values were normalized to the number of cells on each slide.

Nitrite measurements in presence of various agonists. To test whether the measured nitrite levels in the conditioned media reflected increased NO production, we used a specific inhibitor (L-NAA) for that pathway. HASMC cultures were incubated with 100 µM of L-NAA for 60 min before and during exposure to 25 dyn/cm² for 24 h. To study the role of Ca²⁺ and calmodulin in flow-stimulated nitrite production, we performed experiments using BAPTA-AM and CMZ. Cultures were preincubated for 90 min with 10 µM BAPTA-AM and then exposed to 25 dyn/cm² for 30 min in the continuous presence of BAPTA-AM. Similarly, HASMC were preincubated for 45 min with 25 µM CMZ and then exposed to 25 dyn/cm² for 1 h in the continuous presence of CMZ. To examine whether the NOS II isoform was responsible for the flow-induced nitrite production, we incubated cultures with 1 µM dexamethasone or 2 µM cycloheximide 24 h before and during 6 h of exposure to 25 dyn/cm².

Growth studies. For these studies, all flow experiments were performed for 24 h. Cells were then removed from the slide with 0.05% trypsin-EDTA (GIBCO BRL) and counted using a Coulter Counter. A flow chamber template was used on stationary control cultures to normalize culture surface area (29).

Lactate dehydrogenase assay. To control for possible damage and removal of cells by flow, the release of the cytoplasmic enzyme lactate dehydrogenase (LDH) into the culture medium was monitored by an LDH-L reagent kit (Ciba-Corning). Media samples were collected at the end of each experiment from both flow and control cultures and stored at 20°C. To ensure complete cell lysis, media were sonicated for 40 s with a Sonic Dismembrator (Fisher Scientific). The sonicated media samples (150 µl) were added to 1 ml of LDH-L reagent, and the ultraviolet absorbance of the solution was measured at 340 nm by spectrophotometer (System 2600, Gilford Instruments). LDH activity (U/l) was calculated. Control experiments to investigate the effects of the flow system on LDH stability in cell-free media showed no difference in LDH activity between stationary (15.6 ± 0.54 U/l) and flow cultures (15 ± 0.2 U/l) after 24 h (29).

Western blotting. Cells from both control and cultures exposed to flow were washed twice with phosphate-buffered saline (PBS; GIBCO BRL), and total cell lysates were harvested in 150 µl of lysis buffer containing protease inhibitors [0.5% sodium dodecyl sulfate (SDS), 50 mM tris(hydroxymethyl)aminomethane (Tris) · HCl, pH 7.4, 1 mg/ml leupeptin, 1 mg/ml pepstatin, 0.1 M phenylmethylsulfonyl fluoride, and 1 mg/ml aprotinin] (Sigma Chemical). Cell lysates were immediately boiled for 5 min to reduce possible enzyme proteolytic degradation. The viscosity of the samples was reduced by brief sonication and by several passages through a 26-gauge needle. Samples were centrifuged for 5 min at 14,000 g, 4°C, to remove insoluble material, and a clear supernatant was collected. Protein concentration was measured in a small aliquot with the Micro bicinchoninic (BCA) assay reagent kit (Pierce). Samples were further diluted, at a ratio of 3:1, in a 4× sample buffer (0.2 mM Tris · HCl, pH 6.8, 4% SDS, 40% glycerol, 0.4% bromphenol blue, and 10% -mercaptoethanol), boiled for 5 min, then stored at 70°C.

Measurement of cGMP. To determine the relationship between cGMP concentration and production of NO by HASMC, cultures were exposed to 25 dyn/cm² for 1 and 5 min and cGMP concentrations were calculated at each time point in the presence of 1 mM IBMX, a phosphodiesterase inhibitor, added to the media 10 min before cell harvest. The conditioned media (500 µl) were then collected and freeze-dried. Stationary control and monolayers exposed to flow were rapidly washed (2× 5 ml) with ice-cold PBS, and cGMP was extracted twice with 750 µl acidified (10 mM HCl) ethanol. Cells were harvested and centrifuged at 14,000 revolutions/min for 15 min at 4°C. The protein content of the pellet was measured using the Micro BCA assay reagent kit. The supernatant was transferred to a fresh tube and evaporated under a 60°C vacuum oven until completely dried. Dried cell and media extracts were stored at 80°C until processed for radioimmunoassay determination of cGMP levels. cGMP concentration was normalized to protein content.

Statistical analysis. Data are expressed as means ± SE. When data from more than two groups were compared, one-way analysis of variance (ANOVA) was used followed by Fishers protected least significance differences post hoc test. To determine trends for nitrite between control and flow over time, a univariate repeated-measures ANOVA was used, and when overall significance was indicated, contrasts of means and regression-coefficient comparisons were performed to test significance in within-subject groups. Differences were considered statistically significant when P < 0.05. All calculations were performed with SuperANOVA 1.11 for the Macintosh.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

HASMC produce nitrite on exposure to flow. As shown in Fig. 1, A and B, fluid flow stimulated nitrite production in HASMC. Nitrite levels were increased after 5 min of flow exposure (shear stress: 1.1, 5, and 25 dyn/cm²), the earliest time point that nitrite was sampled, and reached significance compared with controls after 15 min of flow. During the first hour of exposure to flow, nitrite production increased 6-, 9-, and 12-fold at 1.1, 5, and 25 dyn/cm², respectively, compared with matched stationary controls (Fig. 1A). Nitrite accumulated in the conditioned media with increasing duration of shear stress exposure, and after 24 h, 13-, 13-, and 15-fold increases in nitrite levels were observed at 1.1, 5, and 25 dyn/cm², respectively (Fig. 1B). However, no significant differences in nitrite levels were observed among the different shear stress levels over the course of the experiment, indicating that nitrite release was flow and not shear stress dependent in this range of shear stress. Sixty percent of total nitrite was produced during the first 60 min of flow exposure. Nitrite release from stationary control cultures decreased slightly with time, but this trend was not statistically significant.

View larger version (20K): [in this window] [in a new window]   Fig. 1.   Time course of cumulative nitrite production in human aortic smooth muscle cells (HASMC) in presence or absence of fluid flow (shear stress: 1.1, 5, 25 dyn/cm2). A: nitrite production during 1st hour of exposure to flow (n = 9-15 stationary controls, n = 3 cultures exposed to 1.1 dyn/cm2, n = 6-15 cultures exposed to 5 dyn/cm2, and n = 7-13 cultures exposed to 25 dyn/cm2). B: nitrite production from 0 to 24 h of exposure to flow (n = 4-16 stationary controls, n = 3 cultures exposed to 1.1 dyn/cm2, n = 4-15 cultures exposed to 5 dyn/cm2, and n = 6-17 cultures exposed to 25 dyn/cm2). Data are means ± SE. Univariate repeated-measures analysis of variance (ANOVA) was used to compare control and experimental groups over time (P < 0.001). * Significantly different from respective control values (P < 0.05);  significantly different from nitrite levels at 25 dyn/cm2 for 5 min (P < 0.05);  significantly different from nitrite levels at 5 dyn/cm2 for 5 min (P < 0.05); § significantly different from nitrite levels at 25 dyn/cm2 for 1 or 2 h (P < 0.05).

1

1

1

1

1

1

1

1

View larger version (22K): [in this window] [in a new window]   Fig. 2.   Effects of stopping and then restarting flow on nitrite production. Cells were exposed to 25 dyn/cm2 for 30 min, and then flow was stopped for 30 min and restarted for an additional 30 min. Media were sampled for nitrite determination at the end of each 30-min period. Data are means ± SE; 30 min: n = 10 control and 10 experimental cultures; 60 min: n = 4 control and n = 6 experimental cultures; and 90 min: n = 4 control and n = 7 experimental cultures. One-way ANOVA was used to test difference between treated and untreated groups (P < 0.001). * Significantly different from respective control values (P < 0.05);  significantly different from nitrite levels in cultures exposed to 25 dyn/cm2 for 30 min (P < 0.05).

Role of Ca²⁺ and calmodulin in flow-mediated NO production. To study the role of intracellular Ca²⁺ in flow-induced production of nitrite, HASMC cultures were preincubated for 90 min with 10 µM BAPTA-AM, a Ca²⁺ chelator, and then exposed to 25 dyn/cm² for 30 min in the continuous presence of BAPTA-AM. The Ca²⁺ chelator blocked the nitrite production associated with 30 min of exposure to flow (Fig. 3A).

View larger version (18K): [in this window] [in a new window]   Fig. 3.   Effect of Ca2+ and calmodulin antagonists on flow-mediated nitrite production by HASMC. A: HASMC cultures were preincubated with 10 µM 1,2-bis(2-aminophenoxy)ethane-N,N,N',N'-tetraacetic acid acetoxymethyl ester (BAPTA)-AM for 90 min and then exposed to 25 dyn/cm2 for 30 min in continuous presence of BAPTA-AM. Media samples for nitrite determination were taken only at the end of each 30-min experiment. Data are means ± SE; n for each experiment is in parentheses below each bar. One-way ANOVA was used to test difference between treated and untreated groups (P < 0.001). * Significantly different from respective control values (P < 0.05);  significantly different from cultures exposed to 25 dyn/cm2 in presence of BAPTA-AM (P < 0.05). B: HASMC cultures were preincubated with 25 µM CMZ for 45 min and then exposed to 25 dyn/cm2 for 1 h in presence of calmidazolium (CMZ). Data are means ± SE; n = 3 control and 3 flow cultures in presence of CMZ, n = 9-15 stationary controls, and n = 7-13 cultures exposed to 25 dyn/cm2. Univariate repeated-measures ANOVA was used to compare control and flow groups over time (P < 0.001). * Significantly different from respective control values (P < 0.05);  significantly different from nitrite levels at 25 dyn/cm2 for 5 min of shear stress (P < 0.05).

L-NAA inhibited nitrite production but had no effect on HASMC proliferation. Experiments were performed with an NOS inhibitor, L-NAA, to determine whether the increase in nitrite levels under flow conditions reflected increased NO release. Incubation with 100 µM L-NAA for 60 min before and during exposure to 25 dyn/cm² for 24 h completely abolished the flow-induced nitrite accumulation (Fig. 4A).

View larger version (21K): [in this window] [in a new window]   Fig. 4.   A: nitrite production in HASMC exposed to 25 dyn/cm2 for 24 h in presence of 100 µM NG-amino-L-arginine (L-NAA; n = 8). B: growth of HASMC exposed to 25 dyn/cm2 for 24 h in presence of L-NAA. Data are means ± SE; n for each experiment is in parentheses below each bar. One-way ANOVA was used to test difference between treated and untreated groups (P < 0.001). * Significantly different from respective control values (P < 0.05);  significantly different from 25 dyn/cm2 in presence of L-NAA (P < 0.05).

Fluid flow increases cGMP levels in HASMC. Because the determination of nitrite levels in the incubation medium does not provide information on whether NO production occurs immediately, both cellular and media levels of cGMP were measured after 1 and 5 min of exposure to 25 dyn/cm² as a more direct index of NO production (Fig. 5, A and B). Exposure to 25 dyn/cm² for 1 and 5 min resulted in significant 1.54- and 1.76-fold increases, respectively, in the intracellular levels of cGMP compared with those in stationary control cells (Fig. 5A). Similar results were obtained in the media levels of cGMP in which significant 1.87- and 2.72-fold increases were observed after 1 and 5 min of exposure to flow, respectively (Fig. 5B). No significant differences were observed in the intracellular levels of cGMP between 1 and 5 min of flow exposure, whereas the 5-min flow exposure produced significantly higher cGMP media levels compared with those produced by the 1-min flow exposure.

View larger version (19K): [in this window] [in a new window]   Fig. 5.   Effect of fluid flow on concentration of guanosine 3',5'-cyclic monophosphate (cGMP) in HASMC cultures. A: cellular concentration of cGMP after 1 and 5 min of flow exposure at 25 dyn/cm2. B: accumulation of cGMP in conditioned media of HASMC cultures exposed to 25 dyn/cm2 for 1 and 5 min. Data are means ± SE; n = 3. One-way ANOVA was used to test difference between treated and untreated groups (P < 0.001). * Significantly different from respective control values (P < 0.05);  significantly different from media levels of cGMP at 1 min of flow exposure (P < 0.05).

NOS II and NOS III are not involved in flow-induced production of nitrite in HASMC. Cultures of HASMC were incubated with 1 µM dexamethasone or 2 µM cycloheximide 24 h before and during 6 h of exposure to 25 dyn/cm². Dexamethasone and cycloheximide had no effect on nitrite levels in medium from either control or HASMC cultures exposed to flow (Fig. 6).

View larger version (31K): [in this window] [in a new window]   Fig. 6.   Nitrite production in HASMC exposed to 25 dyn/cm2 for 6 h in presence (+) and absence () of 2 µM cycloheximide (A) or 1 µM dexamethasone (B). Data are means ± SE; n for each experiment is in parentheses below each bar. One-way ANOVA was used to test difference between treated and untreated groups (P < 0.001). * Significantly different from respective control values (P < 0.05).

View larger version (32K): [in this window] [in a new window]   Fig. 7.   A: Western blot of nitric oxide synthase (NOS) protein in total cell lysates of HASMC. Top: blot was incubated with a monoclonal NOS II antibody. Lane 1, positive control for NOS II (human glioblastoma cell line, incubated with cytokines); lanes 2-4, representative samples of a 6-h experiment; lanes 5-7, representative samples of a 24-h experiment. Middle: blot was incubated with a monoclonal NOS III antibody. Lane 1, positive control for NOS III (human endothelial cells); lanes 2-4, representative samples of a 6-h experiment; lanes 5-7, representative samples of a 24-h experiment. Bottom: blot was immunoblotted with a monoclonal NOS I antibody. Lane 1, positive control for NOS I (rat pituitary tumor cell line); lanes 2-4, representative samples of a 6-h experiment; lanes 5-7, representative samples of a 24-h experiment; lane 8, human endothelial cell (EC) lysate used as a negative control. B: densitometric analysis for NOS I protein levels for the 155-kDa band in control and cell lysates exposed to flow both at 6 and 24 h. Data are means ± SE; n = 4.

Cultured HASMC constitutively express NOS I isoform. Incubation of HASMC immunoblots with monoclonal NOS I primary antibodies gave 155- and 100-kDa molecular mass bands in all HASMC samples; a rat pituitary tumor cell line was used as a positive control, and a human endothelial cell lysate was used as a negative control for the NOS I antibody (Fig. 7A). Densitometric analysis of Western blots revealed no gross differences in the intensity of NOS I bands between control and cell lysates exposed to flow for 6 h (Fig. 7B). However, at the end of 24 h, a decrease in the intensity of the NOS I 155-kDA band was observed in lysates from both control cells and cells exposed to flow compared with intensity at the 6-h time point. Exposure to 25 dyn/cm² for 24 h resulted in an ~40% reduction in NOS I protein levels compared with the respective control cultures. Treatment with dexamethasone did not affect the intensity of the NOS I bands in both stationary control and shear stressed cultures (data not shown).

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

The important observations of this study are that fluid flow rapidly stimulates NO production in cultured HASMC in vitro that is Ca²⁺-calmodulin dependent and that the NOS enzyme responsible for the flow-induced generation of NO is a constitutively expressed neuronal isoform.

HASMC produce nitrite on exposure to flow. The observation that cumulative nitrite production did not vary significantly with increasing levels of shear stress both at early and late times of exposure (Fig. 1, A and B) suggests that nitrite production is flow and not shear stress dependent. At the onset of flow, enhanced convective and diffusive transport rates of an unidentified agonist (for example, L-arginine) at the cell surface might have initiated intracellular signaling responses that resulted in NO production. Experiments using endothelial cells have shown that fluid flow altered the boundary layer concentration of ATP and that this resulted in ATP-mediated increases in intracellular Ca²⁺ (26). However, the possibility cannot be excluded that maximum nitrite production may have been achieved with 1.1 dyn/cm² and that a dependence on shear stress may be observed at even lower shear stresses. Supporting evidence for this possibility comes from the study of Olesen et al. (28), who demonstrated that shear stress-activated K⁺ channels in endothelial cells developed a membrane current with a half-maximal activation at 0.7 dyn/cm², and from the study of Bhagyalakshmi et al. (5), who showed that flow stimulated an increase in inositol 1,4,5-trisphosphate levels, with a half-maximal response below 1.4 dyn/cm². In endothelial cells, it was shown (20) that NO increased biphasically, with an initial burst that was independent of shear stress (1.8-25 dyn/cm²) followed by a sustained phase that depended on shear stress. The signaling mechanisms that lead to different shear stress- or flow-induced NO release responses between endothelial and SMC are as yet unidentified. Differences in flow-induced responses between endothelial and SMC have also been reported by Stamatas et al. (39), who showed that fluid flow caused alkalinization in HASMC, whereas an acidification was observed for bovine aortic endothelial cells.

Flow-induced nitrite production is Ca²⁺-calmodulin dependent. The initial burst of NO production on exposure to fluid flow was Ca²⁺ dependent because an intracellular free Ca²⁺ chelator inhibited the response (Fig. 3A). However, because cultures exposed to flow were incubated with BAPTA-AM for only 30 min, we cannot conclude that the second sustained phase of NO production was Ca²⁺ dependent. To further clarify the role of Ca²⁺, a calmodulin inhibitor, CMZ, was also used. A temporal study was performed for up to 1 h after initiation of flow to address the effects of this inhibitor in the burstlike, and partially in the sustained, nitrite release. Experiments were not performed for longer than 1 h because of the toxicity of CMZ in HASMC cultures at longer incubation periods. Consistent with BAPTA-AM, the calmodulin antagonist inhibited the burstlike flow-induced nitrite production (during the initial 30 min) but had little effect at later times (1 h) (Fig. 3B). The diminished inhibitory effect of CMZ at 1 h provides some evidence that the sustained NO production at later time points of flow exposure is Ca²⁺-calmodulin independent. In shear stressed endothelial cells, the acute NO production was dependent on Ca²⁺-calmodulin, but the sustained second phase of NO production was not affected by the continuous presence of intracellular Ca²⁺-calmodulin antagonists (20). Although the burstlike flow-induced NO production appears to depend on Ca²⁺-calmodulin signaling pathways, it is controversial whether exposure to flow results in elevated cytoplasmic Ca²⁺ concentrations (20, 26). Preliminary experiments using fura 2 fluorescent ratio imaging showed that fluid flow elicited no overall cytosolic Ca²⁺ changes in HASMC (39). A possible interpretation that could reconcile these observations is that flow induces localized changes in Ca²⁺ not detectable by two-dimensional fura 2 measurements or, alternatively, that flow upregulates calcium-sensitive enzymes without directly affecting total cytosolic Ca²⁺ concentration (26).

L-NAA inhibited nitrite production but had no effect on HASMC proliferation. The observation that L-NAA blocked flow-induced increases in nitrite indicates that 1) these nitrite measurements reflect increased NO production, 2) NO is enzymatically produced by HASMC after exposure to flow, and 3) the nitrite measured is not an artifact of the assay system due to the presence of cell debris in conditioned media. L-NAA is a structural analog of L-arginine, the mechanism of action of which appears to be competitive inhibition of the conversion of endogenous L-arginine to NO. This inhibitor has equipotent affinity for both the constitutive and inducible enzyme isoforms and is 100- to 300-fold more potent than N^G-methyl-L-arginine in antagonizing relaxation of vascular rings (9, 13).

Fluid flow increased cGMP levels in cultured HASMC. Exposure to 25 dyn/cm² for 1 and 5 min resulted in a significant increase in both intracellular and media levels of cGMP (Fig. 5, A and B). The elevation of the intracellular cGMP levels is consistent with the known role of NO in activating the soluble form of guanylate cyclase by interaction with the heme moiety (Fig. 5A) (36). In addition, media concentration of cGMP was measured because cGMP accumulates in the conditioned media with time such that even when it returns to basal levels intracellularly, it will remain high in the medium (Fig. 5B). The increases in cellular and media levels of cGMP together with the nitrite accumulation in the conditioned media provide even stronger supporting evidence that fluid flow stimulates NO production in HASMC.

NOS II and NOS III are not involved in flow-induced production of nitrite. The rapid increase in nitrite production at the onset of flow, the Ca²⁺-calmodulin dependence of early flow-induced nitrite release, and the diminished production rates at longer periods of exposure to flow suggest that the NOS isoform responsible for flow-induced NO production is a constitutively expressed enzyme. Glucocorticoids such as dexamethasone inhibit the expression of NOS II but have no effect on constitutive NOS isoforms (31). Dexamethasone blocks the induction of NOS II in numerous cells by downregulating cytokine-induced activity of the transcription factor nuclear factor-B rather than reducing NOS II enzymatic activity (15, 18). It is noteworthy that treatment with 1 µM dexamethasone was ineffective in suppressing nitrite production in HASMC exposed to fluid flow (Fig. 6) but that the cytokine-induced NO in rat vascular SMC was blocked dose dependently by dexamethasone with a 50% inhibitory concentration of 6 nM (14). Although these experiments suggest that flow-induced NO production was not due to NOS II induction, there is recent evidence that dexamethasone fails to suppress NOS II under diverse experimental conditions and in many tissues (21). Our finding that cycloheximide also failed to inhibit nitrite production from cultures exposed to flow, together with the data for dexamethasone and the time course of nitrite production, strongly suggests that the flow-induced NO production in HASMC is not due to induction of NOS II.

Cultured HASMC constitutively express NOS I isoform. In our studies, NOS I monoclonal antibodies gave 155- and 100-kDa protein bands in all HASMC samples but showed no immunoreactivity with human endothelial cell lysates (Fig. 7A). The 155-kDa band had the same molecular mass as the rat pituitary tumor cell line used as a positive control. Potential explanations for the two different molecular mass forms of NOS I in HASMC are that the 155- and 100-kDa proteins may be alternative spliced products, that the 155-kDa form may result from posttranslational modification of the 100-kDa form, or that the 100-kDa may be a proteolytic degradation product of the 155-kDa protein.

NOS I protein is not upregulated by fluid flow. Our observations that the density of the NOS I bands did not differ between stationary control and cell lysates that have been exposed to flow for 6 h suggest that cultured HASMC express a constitutive NOS I protein whose enzymatic activity, rather than the protein amount, is upregulated by flow. The reduction in the intensity of NOS I bands in both control and cultures exposed to flow at 24 h suggests that 1) NOS I expression in HASMC may be regulated by the state of confluency because similar trends have been obtained with the NOS III isoform in endothelial cells (1), or 2) at 24 h, the 155-kDa NOS I protein band is more susceptible to proteolytic degradation. In studies on endothelial cells, NOS III protein levels were increased for up to 12 h of flow-induced shear stress (32, 43) in contrast to our studies in HASMC, in which we observed no NOS I protein increases with flow (Fig. 7B). Taken together, these results suggest that fluid flow may affect NOS activity and expression differently in endothelial cells compared with SMC.