Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The induction of nitric oxide synthase (iNOS) in smooth muscle tissue is believed to account for many of the untoward manifestations of irreversible shock induced in humans and other mammals by septicemia or endotoxin (Rees et al., 1990). Long recognized as an agent responsible for the bactericidal and tumoricidal actions of macrophages (Green et al., 1981a,b; Hibbs et al., 1987; Stuehr et al., 1989), nitric oxide from macrophages is now known to be synthesized from its precursor, L-arginine (LR), by an inducible enzyme (iNOS or NOS2) that has been cloned from a number of tissues including rat smooth muscle cells (Nakayama et al., 1992; Nunokawa et al., 1993). It has been determined that the iNOS induced in macrophages and other cell types is distinct in amino acid sequence and immunological cross-reactivity from the constitutively produced calcium-regulated enzymes present either in vascular endothelial cells (eNOS or NOS3) or in neural tissue (nNOS/bNOS or NOS1; Jaffrey and Snyder, 1995; Förstermann et al., 1995). In the past, most of the studies that have examined the induction of NOS in smooth muscle have used either cultured cell systems (Busse and Mulsch, 1990; Nakayama et al., 1992; Marczin et al., 1993) or tissues harvested from animals that had been pretreated with endotoxin in vivo (Knowles et al., 1990; Rees et al., 1990). More recently, several studies have been conducted using aorta tissue exposed to lipopolysaccharide (LPS) in vitro (Sirsjö et al., 1994; Moritoki et al., 1995; Duarte et al., 1997). In our own work examining the acute actions of growth factors on gastric and vascular smooth muscle preparations studied using classical organ bath procedures in vitro (Hollenberg, 1994a,b), we became interested in the possibility that iNOS might be induced in such preparations over the time course of a routine day's experiment. In exploring this question, we recently observed that iNOS was indeed induced in rat gastric circular muscle (CM) strips (but not in the longitudinal muscle) in vitro. Surprisingly, we found that in the CM tissue, iNOS was located not in the smooth muscle elements, but rather in a set of macrophage-related cells that anatomically were in proximity to the CM cells (Zheng et al., 1997). Given our own results with the CM tissue and in view of the above cited information pointing to the induction of iNOS in the aorta of endotoxin-treated animals and in LPS-treated arterial smooth muscle cultures and aorta rings, we wished to reexamine in further detail the induction of iNOS in rat aorta rings (RA) maintained in an in vitro organ bath; and for simultaneous comparison, under identical conditions in vitro, the induction of iNOS both in CM and gastric longitudinal muscle (LM) preparations. Furthermore, we wished to study, in the distinct RA and CM preparations, possible roles for tyrosine kinase/tyrosine phosphatase-signaling pathways and for NF-B involvement in the induction of iNOS. Because tyrosine kinase pathways and NF-B-mediated transcriptional events have been implicated in iNOS induction, it was our working hypothesis that inhibitors of tyrosine kinase and NF-B activation should block iNOS induction in both the RA and CM preparations. To study these issues, we examined the induction in vitro of iNOS in the RA and CM preparations both functionally (using LR-mediated relaxation as an index of iNOS induction) and biochemically [using the appearance of iNOS mRNA, as measured by a reverse transcriptase-polymerase chain reaction (RT-PCR) method]. Furthermore, with an immunohistochemical approach, we assessed the localization of iNOS induction. The induction of iNOS was monitored both in the absence and presence of inhibitors of transcription and translation, as well as in the presence of inhibitors of tyrosine kinase, tyrosine phosphatase, and NF-B activation.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Tissue Sources and Bioassay Procedures. The aorta ring tissue and the LM and CM preparations were prepared essentially as described previously (Muramatsu et al., 1988; Hollenberg et al., 1989; Laniyonu and Hollenberg, 1995). The male Sprague-Dawley rats weighing approximately 250 g used for our work were cared for according to the recommendations of the Canadian Council on Animal Care. After sacrifice by cervical dislocation, the animals were exsanguinated from the common carotid arteries, and the stomach and aorta tissues were isolated for further dissection. In brief, RA (2 mm × 3 mm) either intact or rubbed free of endothelium were mounted at 37°C in a gassed (95% O₂/5% CO₂) Krebs-Henseleit buffer (4 ml) of the following composition (mM): NaCl (118), KCl (4.7), CaCl₂ (2.5), MgCl₂ (1.2), NaHCO₃ (25), KH₂PO₄ (1.2), and glucose (10). At hourly intervals, rings were precontracted with 1 µM phenylephrine, followed by a tissue wash. When appropriate, the presence of an intact endothelium in the RA preparation was ascertained by monitoring a relaxation response to 1 µM acetylcholine. The LM and CM strips (about 3 mm × 10 mm) were prepared from a region of the fundus that was free from overlying secretory mucosa. The stomach was opened along the lesser curvature and the smooth muscle element was carefully dissected free from overlying tissue. The CM and LM strips were obtained by cutting either along or at right angles to the visible CM bundles, respectively. This procedure allows for the measurement of contractile responses of either the LM or CM elements derived from the same tissue preparation (Muramatsu et al., 1988). As for the RA tissues (above), the LM and CM preparations were equilibrated at 37°C in Krebs-Henseleit buffer under a resting tension of about 1 g. Changes in tissue tension were monitored isometrically using Grass or Statham force transducers. Routinely, tissue integrity was tested by monitoring a robust contraction in response to either 50 mM KCl or 1 µM phenylephrine, after which tissues were washed and allowed to return to baseline tension. Tissues were maintained in the organ bath with periodic washing (about 45-min intervals) for time periods of up to 8 h.

Monitoring of iNOS Induction Using the Relaxation Response to LR. After mounting the preparations (vascular RA or LM and CM) in the organ bath, tissues were precontracted with either 1 µM carbachol (LM preparation) or 1 µM phenylephrine (RA and CM preparations). At the plateau of the contractile response, LR (1 mM) was added to the organ bath; a relaxant response upon adding LR served as a pharmacologic index of iNOS induction, as used by us previously (Zheng et al., 1997) and by others (Moritoki et al., 1995). Tissues were then washed three times to remove agonist and excess LR from the organ bath. The LR-induced relaxation response has been found to provide an accurate reflection of the induction of functional iNOS (Moritoki et al., 1995; Zheng et al., 1997). The time course of the induction of iNOS was determined pharmacologically by monitoring LR-induced relaxation every hour during the incubation of tissue in the organ bath; tissues were washed at 45-min intervals. When present, actinomycin D (ACTD) or cycloheximide (CHX; both obtained from Sigma, St. Louis, MO) were present from the beginning of an experiment. The effects of tyrphostin 47 (AG213; Calbiochem, La Jolla, CA), vanadate (VAN; Sigma), tosyl-phenylalanyl-chloromethyl ketone (TPCK, Sigma), and pyrrolidinedithiocarbamate (PDTC; Sigma) were also assessed by adding these reagents at the start of a prolonged incubation period (5 h). After 5 h, the tissue was washed free from these reagents immediately before the assessment of iNOS induction by pharmacological (i.e., LR-induced relaxation), biochemical (RT/PCR evaluation of mRNA presence), or immunohistochemical procedures. The enzyme inhibitors aminoguanidine (AG; Sigma) and LY83583 (Sigma) were added to the organ bath before the evaluation of tissue responsiveness (see Fig. 1).

View larger version (22K): [in this window] [in a new window]   Fig. 1.   Induction of iNOS in vascular RA and CM tissue; appearance of LR induced relaxation and effects of inhibitors of iNOS and guanylyl cyclase. RA preparations (A and B) and CM preparations (C and D) were studied both immediately after equilibration in the organ bath (left-hand set of tracings) or after a 5-h incubation period (middle and right-hand set of tracings). Separate experiments were done to assess the actions of the iNOS inhibitor, AG (1 mM, ; A and C) and the guanylyl cyclase inhibitor, LY83583 (, LY, 20 µM; B and D), which were added to the organ bath after the 5-h induction period and 20 min before contracting the tissues with phenylephrine (, PE, 1 µM) so as to assess tissue relaxation in response to the addition of either 1 mM LR () or 10 nM SNP (*). The absence of an intact endothelium in the RA preparations (A and B) was documented by an absence of a relaxant response to 1 µM acetylcholine (, Ach) in tissues that otherwise relaxed in response to 10 nM SNP (*). The tracings are representative of six to eight separate experiments done with tissues obtained from two or more different animals.

Preparation of Tissue RNA and RT-PCR Analysis. Total tissue RNA from the CM and RA preparations was extracted with the TRI-reagent protocol (Molecular Research Center, Cincinnati, OH) both before and after prolonged incubation of tissue samples in the organ bath. The RNA was reverse-transcribed with a first strand cDNA synthesis kit using pd(N)6 primer (Pharmacia LKB Biotechnology, Uppsala, Sweden) according to manufacturer's recommendations at 37°C for 1.5 h; 3 µl of this solution was used to amplify the iNOS cDNA fragment. The sequences of the forward (5'CCAGGGGCAAG CCATGTC3') and reverse (5'CTCCAGGCCATCTTGGTGGC3') primers were based on the published rat aorta smooth muscle iNOS cDNA sequence (Nunokawa et al., 1993). Sequencing of cDNA subcloned into the pBluescript SK^- phagemid was done using the dideoxynucleotide sequencing method (Sanger et al., 1977) and a T7 DNA polymerase sequencing kit (Pharmacia). The RT-PCR signals obtained from the gastric and aorta tissue were normalized to the PCR signal generated concurrently by an actin primer pair (Watson et al., 1992) that spans an intron: forward primer (5'CGTGGGCCGCCCTAGGCACCA3'; reverse primer (5'TTGGCCT TAGGGTTCAGGGGG3'). The detection of a 243-bp PCR product using this primer pair can confirm the absence of intron sequences in the RT product obtained from tissue RNA.

Immunohistochemistry. Both before and after a 5-h incubation in the organ bath, tissues were fixed at 4°C by overnight immersion in 4% paraformaldehyde in 0.1 M PBS (pH 7.4). Fixed tissue was then washed with PBS (3 × 10 min) and cryoprotected by immersion in PBS containing 20% (w/v) sucrose. Tissues were sectioned (12 µm) in a cryostat and then processed for indirect immunofluorescence. Sections were washed with PBS containing 0.1% Triton X-100 for 30 min at room temperature and incubated with the primary antibodies for 24 to 48 h at 4°C in a moist chamber. The primary anti-iNOS antibody used (1:500 dilution) was purchased from Transduction Laboratories (Lexington, KY). The mouse anti-rat macrophage antibody (clone ED2, 1:1500 dilution) was purchased from Serotec (Oxford, UK; Dijkstra et al., 1985). After exposure to the primary antibody, sections were rinsed (3 × 10 min) in PBS and then incubated for 1 h at room temperature with secondary antibodies (donkey anti-rabbit IgG conjugated to cyanine 3 at a dilution of 1:100, Jackson Immuno Research Laboratories, West Grove, PA and/or goat anti-mouse IgG conjugated to fluorescein isothiocyanate (FITC), 1:50, Incstar, Stillwater, MN). Finally, sections were washed with PBS containing 0.1% triton X-100 (3 × 10 min) and mounted in bicarbonate-buffered glycerol (pH 8.6). Sections were examined using a Ziess Axioplan fluorescence microscope and photographs were taken with Kodak TMax 400 ASA film. Nonspecific immunoreactivity was assessed using both "noninduced" tissues as controls and nonspecific primary antisera in both control and "induced" tissues.

Monitoring NF-B Activation Using an Immunoprecipitation/Western Blot Approach. RA tissue was isolated and mounted in the oxygenated organ both exactly as for the bioassay procedure (above) either in the absence or presence of 20 µM TPCK. After 3 h at 37°C, tissues were harvested (four to five rings pooled for each condition), quick-frozen on solid CO₂, and stored at 70°C for further analysis. Nonincubated tissue samples served as a control. The frozen tissue (about 800 mg/pooled tissue sample) was rapidly thawed, diced, and homogenized (Polytron, Brinkman Instruments, Rexdale, Ontario, Canada) in 2 ml of an immunoprecipitation buffer comprising 50 mM Tris HCl, pH 7.4 supplemented with 1% v/v detergent Nonidet P40, 0.25% (w/v) sodium deoxycholate, 1 mM EDTA, 1 mM phenylmethyl sulfonyl fluoride, 1 mM sodium VAN, 1 mM NaF, 1 µg/ml each of aprotinin and leupeptin, and 150 mM NaCl. Tissue extracts were clarified by centrifugation (13,000 RPM for 20 min at 4°C in a 2-ml microfuge tube), and protein concentrations were measured using the BioRad reagent (BioRad, Richmond, CA).

Immunoprecipitation and Immunoblot Detection of NF-B. Equal protein aliquots of tissue extract (100 µg) were added to 1 ml of a saline (300 mM NaCl)- fortified 50 mM sodium phosphate buffer, pH 8.0 (PBS) and were incubated for 1 h at 4°C with 0.5 µg/ml of a monoclonal antibody targeted to the nuclear localization sequence of the p65 subunit of NF-B (catalog no. 1697838; Boehringer Mannheim, Laval, Quebec). This concentration of antibody does not displace NF-B from the NF-B inhibitory subunit, I-kB and recognizes only the active NF-B subunits free from IkB (Kaltschmidt et al., 1995). The immune complex containing activated NF-B was then harvested by the addition of protein G-sepharose beads (100 µl of a 50% v/v suspension in isotonic phosphate buffer, pH 7.4; Sigma) and incubated at 4°C overnight. Bead-bound protein was washed five times by centrifugation (microfuge) with 100 µl of the above described PBS, pH 8.0. Protein adsorbed to the washed beads was eluted at 100°C into 30 µl of SDS/mercaptoethanol containing electrophoresis sample buffer (Laemmli, 1970) and 25 µl of the eluted sample was subjected to electrophoresis in SDS containing 11% polyacrylamide gels (1 mm × 6 cm × 8 cm) for 1.5 h at 100V at room temperature. After electrophoresis, protein was transferred (30V overnight at 4°C) to nitrocellulose membranes (0.45 µm; BioRad) before immunoblot detection. Membranes were blocked with 10% (w/v) BSA for 1 h at room temperature, followed by exposure for 1 h at room temperature to a 1:500 dilution (final concentration 2 µg/ml) of the murine monoclonal antibody described above, targeted to the nuclear localization sequence of the NF-B p65 subunit. After washing in 10 mM Tris/saline pH 7.4, the blot was exposed to the second antibody (1/50,000 dilution of supplied reagent of goat anti-mouse IgG, coupled to horseradish peroxidase; Amersham, Oakville, ON) and the location of NF-B was visualized by chemiluminescence detection (Amersham ECL reagent, Amersham).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Pharmacologic Indices of iNOS Induction and Detection of iNOS mRNA. To evaluate iNOS induction in the intact tissue strips, we anticipated that under the conditions of the organ bath, tissue stores of LR might not be sufficient to cause a spontaneous relaxation of the tissue when iNOS was induced. In view of this supposition, it was expected that in contrast with freshly isolated tissue, the addition of 1 mM LR to a precontracted tissue in which iNOS had been induced would cause relaxation due to the metabolism of LR to NO. In keeping with this supposition, the addition of LR to freshly isolated vascular and gastric tissues caused no relaxation effect (left, tracings A to D, Fig. 1) in tissues that otherwise relaxed in response to 10 nM sodium nitroprusside (SNP; *, Fig. 1). In contrast, after incubation of the tissues in the organ bath for a period of 5 h, the addition of 1 mM LR to precontracted tissues caused a prompt relaxation of either the RA (middle tracings A and B, Fig. 1) or CM (middle tracings C and D) preparations, as has been observed previously by others for RA preparations (Moritoki et al., 1995). The induction of iNOS we observed was spontaneous, reproducible over a series of experiments conducted over a 12- to 16-month time period, and did not require the addition of LPS to the organ bath. To rule out any influence of eNOS, we used endothelium-free RA preparations (lack of relaxant response to 1 µM acetylcholine; left-hand portion of tracings A and B, Fig. 1). Comparable results were obtained with endothelium-intact preparations (not shown). As with our previous observations (Zheng et al., 1997), no relaxation was observed upon adding LR to a LM preparation that had been maintained for 5 h in the organ bath (not shown). Thus, the LM tissue was not studied further. In both the RA and CM preparations, the iNOS inhibitor, AG (, Fig. 1, A and C), blocked the LR-induced relaxation that had appeared after the 5-h incubation period (right-hand portions of tracings A and C, Fig. 1) but both tissues were still responsive to SNP (right-hand portions of tracings A and C, Fig. 1). In contrast, in induced RA and CM tissues, the guanylyl cyclase inhibitor, LY83583, (, tracings B and D, Fig.1) not only blocked LR-induced relaxation but also inhibited relaxation caused by SNP (right-hand portion of tracings B and D, Fig. 1). In both the RA and CM tissues, the ability of LR to cause a relaxant response became apparent after 2- to 3-h incubation in the organ bath and was maximal at about 4 to 6 h (Fig. 2). The addition of LPS (5 µg/ml) to the organ bath did not alter the time course or extent of iNOS induction in either RA or CM preparations (not shown). In keeping with our previous observations with CM tissue (Zheng et al., 1997), the appearance of LR-induced relaxation of the RA tissue coincided with the appearance of a prominent RT-PCR signal generated by the iNOS-targeted primer pairs (Fig. 3). Sequencing of the PCR product, which was maximal at 5 h, verified that it did indeed represent iNOS (not shown).

View larger version (13K): [in this window] [in a new window]   Fig. 2.   Time course of appearance of LR-induced relaxation in aorta (top) and CM tissue (top). At hourly time intervals after the initial equilibration of fresh RA and CM tissue in the organ bath, each preparation was washed and tested for a relaxant response to the addition of 1 mM LR, as illustrated in Fig. 1. The relaxation response (% MAX) was expressed as a percentage of relaxation relative to the maximal tension developed in response to 1 µM phenylephrine [% MAX = 100 × (tension in the presence of phenylephrine  tension in the presence of LR)/(tension in the presence of 1 µM phenylephrine)]. Each data point shows the average relaxant response (±S.E.M., bars) of four independently incubated tissues. The figure is representative of two independently conducted experiments using tissues from separate animals.

View larger version (33K): [in this window] [in a new window]   Fig. 3.   Appearance of iNOS mRNA in aorta tissue maintained for 5 h in the organ bath. Either fresh endothelium-free RA tissue (control) or tissue that had been maintained for 5 h in the organ bath with hourly washing (induced) was processed for RNA isolation and RT-PCR analysis using primers for iNOS (top) and actin (bottom) as outlined in Materials and Methods. The arrows denote the expected positions in the separating gel for the iNOS (top) and actin (bottom) PCR products. The positions of the oligonucleotide size markers (base pairs, BP) are shown on the right.

Immunohistochemical Localization of iNOS. In fresh tissue, iNOS was not visualized either in the RA or CM preparations (Fig. 4, A and C). Only background staining, which was equivalent to that observed using nonimmune serum, was detected (not shown). Nonetheless, in the RA tissue, after 5 h in the organ bath (a time corresponding to a maximal relaxation response caused by LR), a prominent iNOS immunoreactivity was detected in the smooth muscle layer (Fig. 4, B). In contrast, in the CM preparation, the iNOS immunoreactivity was localized primarily in a subset of cells found in the submucosal layer (Fig. 4, D, small arrows) rather than in the smooth muscle elements. Double staining of the induced CM preparations with the iNOS and macrophage-specific antibodies revealed a coincidence of immunoreactivity for the cells localized in the submucosal layer of the CM preparation (data not shown; Zheng et al., 1997).

View larger version (90K): [in this window] [in a new window]   Fig. 4.   Immunohistochemical localization of iNOS in aorta (A and B) and CM (C and D) tissue preparations either before (A and C) or after 5-h incubation in the organ bath (B and D). Vascular RA (A and B) and CM (C and D) preparations that were either fresh (A and C) or incubated for 5 h in the organ bath (B and D) were processed for immunohistochemistry as outlined in Materials and Methods. Only tissue that had been incubated with iNOS antibody showed the intense fluorescence in the smooth muscle elements of the RA tissue (B) or in sparsely distributed cells (arrows) in the submucosal (sm) region adherent to the circular muscle elements (cm). The low nonspecific fluorescence signal seen in the smooth muscle of fresh RA tissue was also seen with nonimmune sera. No immunoreactivity above background fluorescence was localized over the circular muscle cells (cm). Scale bar: 100 µm.

Effects of Inhibitors on Induction of Functional iNOS. We evaluated the effects of a variety of inhibitors on the appearance of functional iNOS in the RA and CM preparations, as assessed by the ability of LR to cause a relaxation in a precontracted tissue that had previously been maintained in the organ bath for 5 h (Fig. 5). For experiments with each inhibitor, sets of control and inhibitor-treated tissues were run in parallel; Fig. 5 shows a representative control tissue along with representative tracings for experiments done with each inhibitor. In the RA tissue, the transcriptional and translational inhibitors, ACTD and CHX, both blocked the appearance of functional iNOS (Fig. 5, B and C). In the CM preparation, the transcriptional and translational inhibitors also abrogated the appearance of functional iNOS (Fig. 5, H and I). Unexpectedly, in the RA preparation, inhibitors of both tyrosine kinase (AG213) and tyrosine phosphatase (VAN) appeared equally effective in blocking the appearance of functional iNOS (Fig. 5, D and E). In contrast, in the CM preparation, neither AG213 nor VAN affected the appearance of functional iNOS (Fig. 5, J and K). Both TPCK (Fig. 5, F) and PDTC (not shown), known for their ability to block NF-B activation via presumably different mechanisms, blocked the appearance of functional iNOS in the RA preparation. Again, in contrast with the RA tissue, neither of the two inhibitors of NF-B activation blocked the appearance of functional iNOS in the CM preparation, as assessed by LR-induced relaxation (Fig. 5, L; data not shown). In the RA preparation, the appearance of functional iNOS correlated with an increase in the amount of active NF-B, as detected by an immunoprecipitation/Western blot procedure using an antibody targeted to active NF-B (Kaltschmidt et al., 1995; Fig 6, middle lane). The increase in the abundance of activated NF-B was blocked in the presence of TPCK, in keeping with the ability of TPCK to block the appearance of functional iNOS.

View larger version (21K): [in this window] [in a new window]   Fig. 5.   Effects of inhibitors on the induction of LR- mediated relaxation in the vascular RA (left, A to F) and CM (right, G to L). The vascular RA and CM preparations were either untreated (representative tracings for untreated RA and CM preparations are shown in A and G) or were incubated for 5 h in the organ bath with the following inhibitors (abbreviations and concentrations in brackets): B and H, ACTD (1 µM); C and I, CHX (10 µM); D and J, AG213 (80 µM); E and K, VAN (0.1 mM); F and L, TPCK (20 µM). Both before and after a 5-h incubation at 37°C with or without inhibitors, the tissues were washed (W, arrows), precontracted with 1 µM phenylephrine (, PE), and tested for a relaxant response to the addition of 1 mM LR (). The response to LR was evaluated both before (left, A to L) and after (right, A to L) the 5-h incubation period (5 h, ). The scale for time and tension is shown at the bottom, to the right of tracing L.

View larger version (90K): [in this window] [in a new window]   Fig. 6.   Immunoprecipitation/Western blot detection of activated NF-B: Inhibitory effect of TPCK. Endothelium-free RA were either quick-frozen immediately (fresh) or were incubated in the organ both for 3 h at 37°C without (induced) or with (TPCK) the addition of 20 µM TPCK. Quick-frozen tissue samples were processed for immunoprecipitation and Western blot analysis as outlined in Materials and Methods. The immunobead harvests from equal amounts of protein (100 µg) extracted from the three pooled tissue samples were analyzed by electrophoresis and Western blot detection using a monoclonal antibody targeted to the nuclear localization sequence of the NF-B p65 subunit. The positions of the molecular size markers (kD) are shown on the left. The band detected is at the position of NF-B.

Effects of Inhibitors on the Appearance of iNOS mRNA. In each instance wherein an inhibitor blocked the appearance of functional iNOS (LR-induced relaxation), we wished to determine if the appearance of iNOS mRNA, as assessed by RT-PCR, was similarly affected. All PCR signals for iNOS were compared relative to the actin PCR signal obtained from the identical RT product. In all cases, the 243-bp actin PCR product indicated that the RNA used for the procedure had been free from genomic DNA. In both the RA and CM tissue, as anticipated for an induced protein, the transcriptional inhibitor, ACTD blocked the appearance of iNOS mRNA (Fig. 7, top). Similarly, in the RA tissue, as expected, the translational inhibitor CHX had no effect on the appearance of iNOS mRNA (Fig. 7, upper left). In contrast with the results using RA tissue, in the CM tissue the presence of CHX, like ACTD, markedly attenuated (by about 90%, based on densitometry of the PCR bands, relative to actin) the appearance of iNOS mRNA (Fig. 7, upper right). We next turned our attention to the RA tissue that had been treated with either of the tyrosine kinase inhibitors, genistein (GS) or AG213. Although both of these tyrosine kinase inhibitors blocked the induction of functional iNOS (Fig. 5, tracing D and data not shown), neither GS nor AG213/TP47 blocked the appearance of iNOS mRNA (Fig. 8, A). In contrast, VAN, which blocked the induction of functional iNOS in the RA preparation (Fig. 5, E) also blocked the appearance of iNOS mRNA in the tissue (Fig. 9). In keeping with the ability of the inhibitor of NF-B activation, TPCK, to block the induction of functional iNOS in the RA preparation and to attenuate the increase of active NF-B visualized by Western blot analysis (Fig. 6), TPCK also markedly attenuated the appearance of RA iNOS mRNA (Fig. 10). Comparable effects were observed with PDTC (not shown). In contrast with the results with the RA tissue, neither the tyrosine kinase/tyrosine phosphatase inhibitors nor the inhibitors of NF-B activation affected the appearance of iNOS mRNA in the CM tissue, in keeping with the lack of effect of these agents on the appearance of functional iNOS (Fig. 5, J to L, and data not shown). The effects of the various inhibitors on the appearance of functional iNOS and on the appearance of iNOS mRNA are summarized in Table 1.

View larger version (35K): [in this window] [in a new window]   Fig. 7.   Effect of ACTD and CHX on the induction of iNOS mRNA. The vascular RA and CM (RGCM) preparations were incubated for 5 h in the organ bath either in the absence (induced) or presence of either 1 µM ACTD (+ ACTD) or 10 µM CHX (+ CHX), as outlined in Fig. 5. At 5 h, all tissues were tested for the absence of a relaxant response to 1 mM LR, in keeping with data shown in Fig. 5, washed, and then processed for the isolation of RNA and RT-PCR analysis using primers for both iNOS (top) and actin (bottom). The arrows on the left denote the expected positions of the PCR products for iNOS (top) and actin (bottom); the positions of the standard oligonucleotide size markers (base pairs: BP) are shown on the right. The data are representative of two independently conducted experiments.

View larger version (22K): [in this window] [in a new window]   Fig. 8.   Effects of inhibitors of tyrosine kinase and of NF-B activation on the induction of iNOS mRNA in the aorta RA (a, upper panels) and CM (b, lower panels) preparations. The RA (a, top) and CM (b, bottom) preparations were incubated in the organ bath for 5 h as outlined in Fig. 5 either in the absence (induced) or presence of 80 µM AG213 (+ TP), 100 µM GS (+ GS), or 20 µM TPCK (+ TPCK). At 5 h, the tissues were tested for the absence or presence of a relaxant response to 1 mM LR, as shown in Fig. 5, and tissues were harvested for the preparation of RNA and for RT-PCR analysis using primer pairs for iNOS and actin. Arrows on the left denote the expected positions of the PCR products for iNOS and actin; the positions of the oligonucleotide size markers (base pairs, BP) are shown on the right. The data are representative of an experiment in which tissue samples were harvested in triplicate and pooled for RT-PCR analysis.

View larger version (32K): [in this window] [in a new window]   Fig. 9.   Effects of VAN on the induction of iNOS mRNA in aorta tissue. Endothelium-free RA tissue was incubated for 5 h in the organ bath either without (induced) or with 100 µM sodium orthovanadate (+ VAN). Tissue was then processed for RT-PCR analysis as outlined in Materials and Methods and in the legends to Figs. 7 and 8. The positions for the iNOS (top) and actin (bottom) PCR products are shown by the arrows on the left; positions of the oligonucleotide size markers (base pairs, BP) are on the right.

View larger version (36K): [in this window] [in a new window]   Fig. 10.   Effect of the inhibitor of NF-B activation, TPCK, on the induction of iNOS mRNA in the RA preparation. RA tissue was incubated for 5 h in the organ bath either without (induced) or with 20 µM +TPCK as shown in Fig. 5, F and evaluated for the absence of an LR-induced relaxation. After washing, tissue was harvested for RNA preparation and RT-PCR analysis as outlined in Materials and Methods and in the legends to Figs. 7 and 8. The very low iNOS PCR signal detected in TPCK-treated tissue was obtained from preparations that did not relax in response to 1 mM LR. The arrows on the left show the positions of the iNOS (top) and actin (bottom PCR) products. The positions of the oligonucleotide size markers (base pairs, BP) are shown on the right.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Our results both confirm and in many ways extend our preliminary study of the induction of iNOS in CM preparations (Zheng et al., 1997), particularly in terms of the differences in the signal transduction pathways governing iNOS induction in gastric versus aorta smooth muscle. Furthermore, in contrast with previous studies with intact aorta (Moritoki et al., 1995; Duarte et al., 1997), we were able to assess independently the differential effects of the inhibitors we used both on the induction of iNOS mRNA and on the induction of functional iNOS. The main finding of our new work was that in smooth muscle tissue taken from different anatomic locations, there was a marked difference in the cellular sites of iNOS induction. In the aorta preparation, iNOS induction was localized predominantly within the smooth muscle elements whereas in the CM preparation, iNOS was not detected in the smooth muscle, but was prominently expressed in a set of submucosal macrophage-related cells. In each tissue preparation, nonetheless, the induction of functional iNOS, as monitored by the ability of LR to cause an AG and LY83583-inhibitable relaxation, was closely correlated with the appearance of iNOS mRNA, and in both the gastric and vascular smooth muscle preparations, iNOS induction could be seen to have an effect on tissue contractility. This inhibitory effect would be expected to have a direct impact on the regulation of vascular and gastric smooth muscle function in vivo

It was surprising to us that the same experimental conditions and presumably the same stimuli in the identical organ bath that led to a robust induction of iNOS in the vascular smooth elements did not result in a comparable induction in the smooth muscle cells of the CM preparation as detected immunohistochemically. Differences in the induction of iNOS mRNA by LPS and interferon- have previously been observed in smooth muscle cells depending on whether the cells were in intact strips or propagated in tissue culture (Sirsjö et al., 1994). Furthermore, a variety of different stimuli have been observed to modulate iNOS induction, even in the same cultured cell type (Gilbert and Herschman, 1993a,b; Förstermann et al., 1995). However, differences in vivo between tissues with the same phenotype (i.e., smooth muscle) have yet to be reported using an immunohistochemical approach. Although it was not possible to determine the precise stimulus in the organ bath that led to iNOS induction (e.g., the trauma of tissue dissection, or very possibly, the presence of endotoxin in the organ bath), our data point to interesting differences in the iNOS induction pathway between vascular and gastric tissue. Simply adding LPS to the organ bath failed to alter the time course or extent of iNOS induction in the RA tissue. Furthermore, our work failed to identify an agent that, upon addition to the organ bath, led to an induction of iNOS in the gastric smooth muscle elements. Two possibilities that might account for the differences in iNOS induction between the two tissues may relate to differences in the coinduction of the LR transporter and differences in the profile of cytokines produced by the two tissues as a result of the dissection procedure. Because the coinduction of the LR transporter may be a prerequisite for observing a functional iNOS response (Simmons et al., 1996a,b), it would be of value in future studies to assess the induction of this transporter in the RA and CM preparations. Furthermore, it would be of interest to treat both the RA and CM preparations with a "cytokine cocktail" to determine if iNOS induction can be accelerated (RA) or triggered (CM) in the smooth muscle cells. To our knowledge, the induction of iNOS in gastric smooth muscle elements has yet to be reported. Thus, further work is warranted to examine the inducibility (or lack thereof) of iNOS in gastric smooth muscle and to identify the precise stimulus responsible for the spontaneous induction of iNOS in the RA tissue (usually presumed to be contamination with endotoxin).

It was of interest that the various inhibitors we used resulted in differential effects on the induction of iNOS in the RA tissue compared with the CM preparation (summarized in Fig. 5 and Table 1). These differences most likely reflect the distinct cell types within the two tissues where iNOS was maximally induced (i.e., the smooth muscle cells in the RA tissue versus the macrophage-related cells in the CM tissue). One interesting difference that we observed was that CHX inhibited the induction of iNOS mRNA in the CM but not in the RA tissue (Fig. 7). These data suggest that the new synthesis of protein paracrine factors may play different roles in the induction of iNOS in the RA and CM tissue. Although rapidly synthesized paracrine or intracellular factors (e.g., the transcription factor AP1) that can up-regulate iNOS transcription very likely play an important role in the modulation of iNOS in the macrophage-related cells present in the CM tissue, other mechanisms may govern iNOS induction in vascular smooth muscle. Results that reflect our observations with the CM tissue have been reported for iNOS induction in murine and rat macrophages (Chesrown et al., 1994) and in insulin-producing HIT cells (Eizirik et al., 1993). The lack of effect of the tyrosine kinase/phosphatase inhibitors and of the inhibitors of NF-B activation on iNOS induction in the CM tissue as opposed to the RA tissue (Figs. 5 and 8 and Table 1) highlight further the marked differences between the signaling pathways that regulate iNOS induction in the two tissue types.

One of our working hypotheses was that the induction of iNOS in the RA and CM tissues might involve the activation of a tyrosine kinase signaling pathway. Thus, we anticipated that tyrosine kinase inhibitors might block, and that a tyrosine phosphatase inhibitor might potentiate, the induction of iNOS. This hypothesis was in keeping with our previous studies of the role of tyrosine kinase/phosphatase pathways in regulating contractility (Laniyonu et al., 1994a; Hollenberg, 1994b). The hypothesis would also be in keeping with the ability of tyrosine kinase inhibitors to abrogate LPS-induced lethal toxicity (Novogrodsky et al., 1994), to attenuate LPS-induced induction of iNOS in murine macrophages (Dong et al., 1993), and to abrogate circulatory failure (Ruetten and Thiemermann, 1997). It was, therefore, not expected that both AG213 and VAN would block the induction of functional iNOS in the RA tissue (Fig. 5) and that neither reagent would affect iNOS induction in the CM tissue. The effects of GS agreed with the observations of Moritoki et al. (1995) who found that tyrosine kinase inhibitors blocked functional iNOS induction in RA, but who did not monitor iNOS mRNA. Our data also agreed with a study that appeared after the completion of our work (Duarte et al., 1997), showing that both GS and VAN blocked the appearance of functional iNOS in endotoxin-treated RA. Nonetheless, our work considerably extends the previous observations by demonstrating an effect of VAN at the level of mRNA induction and an effect of AG213 at an as yet unidentified post-transcriptional level. Furthermore, our work localizes the iNOS in the RA tissue specifically in the smooth muscle elements. The results with the CM tissue would suggest that in contrast with the RA tissue, tyrosine kinase and tyrosine phosphatase signal pathways may both be irrelevant for the induction of iNOS in certain tissues (e.g., the macrophage-related cells in the CM preparation). Conversely, in the RA preparation, tyrosine kinase and tyrosine phosphatase signal pathways would appear to act in concert for the induction of iNOS. The data obtained with VAN in the RA tissue (Figs. 5 and 9 and Table 1) indicate that an increase in cellular tyrosine phosphorylation (see Fig. 6 of Laniyonu et al., 1994b) leads to a suppression of iNOS transcription; yet, the AG213-mediated inhibition of ambient tyrosine kinase activity in the setting of iNOS induction did not affect the transcription of iNOS mRNA (Fig. 8, A; Table 1). Nonetheless, both tyrosine kinase inhibitors that we used blocked the induction of functional iNOS. Thus, as alluded to above, our data point to a role for multiple tyrosine kinase steps involved at both the transcriptional (blocked by VAN) and the translational/post-translational (blocked by GS/AG213) levels for iNOS induction.

A final issue raised by our results relates to the potential role for NF-B activation. The differential inhibition of iNOS induction in the RA versus the CM tissue by TPCK and PDTC would suggest that in selected tissues, such as the macrophage-related cells in the gastric tissue, transcription factors other than NF-B may be involved in iNOS induction. This suggestion would be in contrast with the data linking PDTC-sensitive NF-B activation to iNOS induction (Xie et al., 1994; Nunokawa et al., 1996). The mechanisms that underlie the differential induction of iNOS in different smooth muscle preparations via apparently distinct signaling pathways remain an intriguing topic for further study.