INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

ACTIVATION OF NITRIC OXIDE (NO) synthase (NOS) and consequent increased production of NO are believed to elicit antiproliferative effects attributed to NO itself (2, 8, 15, 17, 20, 23, 25, 28, 35, 43, 44). Additional studies suggested that increased NOS activity or added NO causes cytostasis (21, 22, 26, 30-33, 37-40). These studies are important but inconclusive in that cGMP did not account for the cytostatic effect of NO in many cell systems. Moreover, studies addressing the role of NOS activation on cell proliferation were based on the assumption that NO is the only potentially active species generated from arginine by NOS, and the effect of the NOS intermediate N^G-hydroxy-L-arginine (NOHA) was not addressed. The latter possibility takes on special significance in light of our recent finding that cytokine-activated cultured cells generate and accumulate relatively large quantities of NOHA in addition to NO or its oxidized metabolites (5). Furthermore, the accumulation of NOHA by other cell types and in plasma has been reported (19).

NOS isoforms catalyze the NADPH-dependent oxidation of arginine to NO plus citrulline

(1)

where NOHA is the principal intermediate in the reaction (21, 27, 38, 40). Although in vitro experiments with purified NOS indicated that NOHA might be a fleeting or transient intermediate in the catalytic conversion of arginine to NO plus citrulline (21, 38, 40), cell culture experiments indicated clearly that NOHA can accumulate in the cell culture medium up to 20-30% of the NO and citrulline generated (5, 18). These observations indicated that NOHA may not only serve as an intermediate in the production of NO but may also play a role as a distinct biological mediator in its own right. One such role for NOHA conceivably might be to function as an endogenous inhibitor of arginase activity, a view that is consistent with the observations that NOHA inhibits arginase activity (3, 5, 9). Arginase, which catalyzes the conversion of arginine to ornithine plus urea, is important not only in the urea cycle but also in biochemical pathways essential to cell proliferation and wound healing. Ornithine is a precursor to polyamines and proline and in mammalian cells is derived from arginine via the catalytic action of arginase

(2)

Accordingly, inhibition of arginase activity by NOHA could result in inhibition of cell proliferation and this action could account for at least part of the mechanism by which NOS activation leads to cytostasis. The influence of NO on polyamine production from arginine is unknown. The objective of the present study was to elucidate the mechanisms by which NOHA and NO inhibit the proliferation of cultured Caco-2 cells, a human colon carcinoma cell line.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Chemicals and solutions. Sources of lipopolysaccharide (LPS), cytokines, cell culture media and supplements, reagents for chemiluminescence detection, and reagents for urea determination and arginase assay and protein determination were described previously (5, 41). NOHA was obtained from Cayman Chemical, and S-ethylisothiourea (EITU) was from TCI America. 1H-[1,2,4]oxadiazolo[4,3-a]quinoxalin-1-one (ODQ) was purchased from Alexis Biochemicals. (Z)-1-[N-(2-aminoethyl)-N-(2-aminoethyl)-amino]-diazen-1-ium-1,2-diolate (DETA/NO) and sodium (Z)-1-(N,N-diethylamino)-diazen-1-ium-1,2-diolate (DEA/NO) were generously provided by Dr. David A. Wink. All other chemicals and reagents were purchased from Sigma Chemical.

Caco-2 cell culture. Caco-2 cells (human colon adenocarcinoma cell line; American Type Culture Collection) were plated in high glucose DMEM-HEPES supplemented with 10% fetal bovine serum (FBS), 2 mM glutamine, 1 mM sodium pyruvate, 100 U/ml penicillin, 100 µg/ml streptomycin, and 0.25 mg/ml Fungizone and grown until confluent, following which time cells were subcultured by trypsinization. Cell cultures were performed at 37°C in a humidified atmosphere of 5% CO₂-95% air. Caco-2 cells were plated in 12-well plates at a density of ~5,000 cells/cm² in DMEM-HEPES containing 10% FBS and 0.4 mM arginine and incubated for 3 h. The serum- and arginine-containing medium was then removed and replaced with serum-and arginine-free DMEM-HEPES supplemented with SITE+3 serum substitute growth medium (Sigma Chemical). SITE+3 is a mixture of selenium (sodium selenite), insulin, transferrin, ethanolamine, linoleic acid, oleic acid, and albumin and is commonly used to promote cell proliferation in the absence of added serum or growth factors. Cells were washed initially and again at 24 h before cell proliferation experiments were started. Incubation of cells in serum- and arginine-free medium resulted in arrest of cell proliferation, because cell proliferation is arginine dependent. Preliminary experiments revealed that Caco-2 cell proliferation was dependent on the concentration of arginine within the range of 0.1-10 µM. Peak rates of cell proliferation were observed at arginine concentrations of ~10 µM. Raising the arginine concentration above 10 µM to between 30 and 300 µM did not significantly alter cell proliferation rates compared with rates observed at 10 µM arginine. Cell growth was initiated by addition of 30 µM arginine to cell culture media, and cells were grown for 8 days. Test agents were added at the time of addition of the 30 µM arginine (referred to as the start of cell growth). Cell culture media were replaced routinely with fresh media and test agents every other day. This procedure was employed to replenish cell culture medium constituents and test agents as well as to remove any dead cells and cell debris. Greater than 95% cell viability was maintained under all conditions for up to at least 8 days of cell growth as assessed by trypan blue exclusion and release of lactate dehydrogenase.

Measurements of cell proliferation. Cell proliferation was assessed primarily by determining rates of DNA synthesis by monitoring the incorporation of thymidine into DNA. As a check on the reliability and reproducibility of this procedure to assess cell proliferation, two additional procedures were used concurrently; these were cell count and protein determination. Cell counting was performed after trypsinization of cells and resuspension of pelleted cells in PBS and transfer to a hemocytometer. Cell count represents the total number of cells remaining at the end of the 8-day growth period. Aliquots of cell suspensions used for cell counting were centrifuged, and cells were dissolved in 100 ml of 0.1 N NaOH by incubation overnight at 37°C for protein determination. Protein concentrations were determined by the Bradford Coomassie brilliant blue method (Bio-Rad), using bovine serum albumin as the standard. Cell culture media were replaced every other day to replenish the medium constituents and to remove any dead cells and cell debris. Cell viability remained at >95% throughout the 8-day growth period. Therefore, protein determinations are believed to accurately reflect the number of viable cells present. In determining rates of DNA synthesis, a modification of the [methyl-³H]thymidine uptake procedure described previously (29) was employed. Thymidine incorporation into DNA was determined during the final 24 h of an 8-day growth period in medium containing 30 µM arginine. Briefly, 0.1 µCi of [methyl-³H]thymidine (6.7 Ci/mmol; NEN) was added to each well containing cells in a 1-ml volume. Cells were incubated at 37°C for 24 h and rinsed with PBS, and 1 ml of ice-cold 10% trichloroacetic acid was added to each well. Cells were harvested, sedimented, rinsed twice with the trichloroacetic acid solution, and solubilized in 1 ml of 0.12 N NaOH containing 0.1% SDS. Samples were added to EcoLite (ICN) scintillation cocktail and counted in a liquid scintillation spectrometer.

Arginase assay. Caco-2 cells (3 × 10⁶ cells/sample) were harvested, washed with PBS, and centrifuged, and the sediments were lysed as described previously for endothelial cells (5). Supernatant fractions were assayed for arginase activity under physiological conditions of pH (7.4) and arginine concentration (0.3 mM) by monitoring the conversion of L-[guanido-¹⁴C]arginine to [¹⁴C]urea in a 10-min incubation by methods that we have described previously (5).

Urea determination. Cell culture media were collected and analyzed spectrophotometrically for urea exactly as described previously (5).

Ornithine decarboxylase assay. Ornithine decarboxylase (ODC) activity was determined by monitoring the formation of [¹⁴C]CO₂ from L-[1-¹⁴C]ornithine (14, 34). Caco-2 cells were harvested and homogenized in 0.5 M Tris · HCl, pH 7.5, at 4°C to yield preparations containing ~1 mg protein/ml. Briefly, reactions were conducted for 60 min at 37°C in 0.5 M Tris · HCl, pH 7.5, containing 0.1 mg protein (Caco-2 cell homogenate), 0.4 mM ornithine (labeled with 0.3 µCi L1-¹⁴C]ornithine; 55 Ci/mmol, NEN), 40 µM pyridoxal 5'-phosphate and test agents in a final volume of 250 ml. Reactions were conducted in sealed tubes designed to trap CO₂ on filter paper saturated with 10% KOH. Reactions were initiated by addition of Caco-2 cell extract and terminated by addition of 300 ml of 6 N HCl. Tubes were kept at 37°C for 90 min, filters were removed and placed in 5 ml of EcoLite scintillation cocktail at 25°C for 60 min, and samples were counted in a liquid scintillation spectrometer.

Coculture procedures. Rat aortic endothelial cells (RAEC) were plated on Falcon cell culture inserts on polyethylene terephthalate track-etched membranes with 3-mm pore size (Becton Dickinson). Cells were seeded at a density of 4.5 × 10⁴ cells/cm² in DMEM-HEPES containing 20% FBS, 1% endothelial cell growth supplement, 2 mM L-glutamine, 100 U/ml penicillin, 100 µg/ml streptomycin, 1 mM sodium pyruvate, and 10 U/ml heparin, and incubated at 37°C in a humidified atmosphere of 5% CO₂-95% air. Cells were allowed to attach, and the medium was changed the next day and then every other day. At confluence (3-4 days), some of the cells were activated by addition of LPS (100 µg/ml), interferon- (IFN-; 100 U/ml), interleukin-1 (IL-1; 400 U/ml), and tmuor necrosis factor- (TNF-; 1,000 U/ml) and subsequent incubation for 6 h, followed by removal of LPS and cytokines by extensive washing before coculture with Caco-2 cells as follows. The medium was replaced after it was washed twice with arginine- and serum-free medium (Specialty Media), containing 2 mM L-glutamine, 100 U/ml penicillin, 100 µg/ml streptomycin, 1 mM sodium pyruvate, and SITE+3 serum substitute. The nonactivated RAEC cultured on inserts were also washed. Caco-2 cells were seeded at a density of 10⁴ cells/cm² in six-well companion plates for the inserts in DMEM-HEPES containing 10% FBS, 2 mM L-glutamine, 100 U/ml penicillin, 100 µg/ml streptomycin, and 1 mM sodium pyruvate. After 3 h, the cells were washed twice with PBS and the medium was replaced with arginine- and serum-free supplemented medium as described above. After 48 h, the activated and nonactivated RAEC plated on the inserts were added to the wells containing the Caco-2 cells. An equal number of wells contained only Caco-2 cells. To each well and to each insert, 1 ml of serum-free supplemented medium containing 100 µM L-arginine was added. After 24 h of coculture, 0.1 µCi of [methyl-³H]thymidine was added to each well and to each insert and incubated for another 24 h, after which time DNA synthesis in Caco-2 cells was assessed.

Other determinations. Chemiluminescence detection of nitrite and nitrate and protein determinations were as described previously (5).

Statistics. The data illustrated in Figs. 1-9 were analyzed statistically using one-way ANOVA and the Bonferroni t-test for unpaired values.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

NOHA, which has a half-life of ~24 h in cell culture media (5), produced a concentration-dependent inhibition of Caco-2 cell proliferation (Fig. 1). NO, in the form of the NONOate type of NO donor, DETA/NO, which has a half-life of ~24 h (24), also inhibited Caco-2 cell proliferation in a concentration-dependent manner and was equal in potency to NOHA. The S-nitrosothiol NO donor S-nitroso-N-acetylpenicillamine (SNAP) elicited effects that were similar to those for DETA/NO (data not shown). Three different methods for analysis of cell proliferation yielded similar data in experiments in which the ornithine decarboxylase inhibitor -difluoromethylornithine (DFMO) was tested as a positive control (Fig. 2). The above experiments indicate that both NOHA and NO inhibit tumor cell growth.

View larger version (34K): [in this window] [in a new window]   Fig. 1.   Inhibitory effects of NG-hydroxy-L-arginine (NOHA) and (Z)-1-[N-(2-aminoethyl)-N-(2-aminoethyl)-amino]-diazen-1-ium-1,2-diolate (DETA/NO) on Caco-2 tumor cell proliferation. Cell proliferation was assessed by thymidine incorporation into DNA during final 24 h of an 8-day growth period in medium containing 30 µM arginine as described in text. Data were calculated as dpm per 105 cells per well and expressed as percentage of control (CTL), where CTL represents cells grown in absence of added NOHA or DETA/NO. Test agents were added at start of cell growth (addition of 30 µM arginine), and media and test agents were replaced every other day as described in text. * Statistical significance from CTL (P < 0.05). Data represent means ± SE of duplicate determinations from 6 separate experiments.

View larger version (36K): [in this window] [in a new window]   Fig. 2.   Use of 3 different procedures to verify inhibitory effects of NOHA and DETA/NO on Caco-2 tumor cell proliferation. Cell proliferation was assessed by cell count, cell protein determination, and DNA synthesis. In control experiments, number of Caco-2 cells increased from a range of 7,000-15,000 cells per well (0.5 ml) to 210,000-360,000 cells per well during an 8-day growth period in medium containing 30 µM arginine. Data for protein concentration and thymidine incorporation were normalized to 105 cells per well and expressed as percentage of control, where control represents cells grown in absence of added NOHA, DETA/NO, or -difluoromethylornithine (DFMO). Test agents were added at start of cell growth (addition of 30 µM arginine), and media and test agents were replaced every other day. Data represent means ± SE of duplicate determinations from 3 separate experiments.

The next step was to elucidate the mechanism of cytostatic action of NOHA, which was presumed to be inhibition of arginase activity. Accordingly, three observations were deemed necessary: 1) Caco-2 cells should contain arginase activity that is inhibited by NOHA in the concentration range that inhibited cell proliferation; 2) Caco-2 cells should generate ornithine and urea, which should be inhibited by NOHA; 3) the cytostatic effect of NOHA should be overcome by addition of those enzymatic reaction products that are distal to the site of inhibition by NOHA (Eq. 2), such as ornithine, putrescine, spermidine, and spermine. Proliferating Caco-2 cells were found to contain arginase activity present constitutively, and the arginase activity present in Caco-2 cell lysates was inhibited by 0.3-10 µM NOHA (Fig. 3), using physiological conditions of pH and arginine concentration (5). NOHA inhibited arginase activity by competitive mechanisms (not shown), and the approximate inhibitory constant for NOHA was 0.5 µM. Arginase activity was not inhibited by NO in the form of DEA/NO, a NONOate with a short half-life of 2 min (24), or SNAP, each at 100 µM. Growing Caco-2 cells continuously produced ornithine and urea, both of which accumulated in the cell culture medium (data for urea are shown in Fig. 4). Addition of NOHA at concentrations that inhibited cell growth (10 µM) also inhibited the intracellular production of urea (Fig. 4), whereas NO was inactive (not shown), thereby suggesting that inhibition of arginase activity may account for the action of NOHA as an inhibitor of tumor cell proliferation. NOHA was equipotent in diminishing urea production at arginine concentrations of 30 and 100 µM. When the arginine concentration was raised to 300 µM, the NOHA concentration had to be raised to 30 µM to elicit the same 50% inhibition of urea production. In view of the knowledge that NOHA is a competitive inhibitor of arginase with respect to arginine, these observations suggest that 10 µM NOHA is capable of functioning as an inhibitor of arginase at physiological levels of arginine on the order of 100 µM. Effects of NOHA similar to those illustrated in Fig. 4 were obtained on Caco-2 cell proliferation (data not shown). For example, 10 µM NOHA inhibited cell proliferation by 62 ± 8% in the presence of 30 µM arginine, by 66 ± 6% in the presence of 100 µM arginine, and by 50 ± 5% in the presence of 300 µM, and raising the NOHA concentration to 30 µM in the presence of 300 µM arginine caused 68 ± 8% inhibition of cell proliferation arginine (mean ± SE of duplicate determinations from 3 experiments). These observations suggest that 10 µM NOHA can markedly inhibit Caco-2 cell proliferation at arginine concentrations of up to at least 300 µM.

View larger version (33K): [in this window] [in a new window]   Fig. 3.   NOHA, but not nitric oxide (NO), inhibits arginase activity derived from Caco-2 tumor cells. Arginase activity was determined by monitoring conversion of L-[guanido-14C]arginine to [14C]urea as described in text. Enzyme reactions were initiated by addition of substrate, and test agents were preincubated with enzyme-containing supernatants for 1 min at 37°C before addition of substrate. * Statistical significance from CTL (P < 0.05). Data represent means ± SE of duplicate determinations from 4 separate experiments. SNAP, S-nitroso-N-acetylpenicillamine.

View larger version (27K): [in this window] [in a new window]   Fig. 4.   NOHA interferes with intracellular arginase activity in Caco-2 tumor cells. NOHA was added at start of cell growth (addition of 30 µM arginine), and media and NOHA were replaced every other day. At day 8, which represents a 48-h time interval during which medium was not disturbed, cell culture media were collected and analyzed for urea. NOHA (10-30 µM) was tested at 3 different concentrations of arginine ranging from 30 (routinely used in experiments) to 300 (upper limit of physiological range) µM. NOHA inhibited urea production by 40-60% at all arginine concentrations. * Statistical significance from corresponding control (P < 0.05). Data represent means ± SE of duplicate determinations from 5 separate experiments.

The addition of ornithine, putrescine, spermidine, or spermine to Caco-2 cell cultures grown in the presence of NOHA prevented the cytostatic action of NOHA (Fig. 5), as expected according to the reaction sequences depicted in Eq. 2, that is, the addition of reaction sequence products that are distal to the arginase reaction would be expected to prevent the cytostatic effect of NOHA, an arginase inhibitor. Moreover, this reaction sequence would predict that the cytostatic action of DFMO, a selective inhibitor of ODC, should be prevented by putrescine, spermidine, and spermine but not ornithine. Indeed, this effect was observed (Fig. 5). Moreover, the data obtained with DETA/NO were qualitatively similar to the data obtained with DFMO in that the cytostatic effects of DETA/NO and DFMO were prevented by addition of putrescine, spermidine, or spermine but not ornithine. Addition of ornithine or polyamines alone to proliferating Caco-2 cells elicited no effects on cell proliferation (Fig. 6), that is, at concentrations used to prevent the cytostatic effects of added NOHA, DETA/NO, or DFMO, neither ornithine nor the polyamines stimulated cell proliferation. These observations indicate that an arginine concentration of 30 µM under the experimental conditions employed is sufficient to maintain maximal cell proliferation, and adding back either more arginine (discussed above) or products distal to arginine in the arginine-polyamine pathway does not further increase the rate of cell proliferation. Adding back 10-30 µM ornithine, putrescine, spermidine, or spermine to Caco-2 cell cultures in the absence of arginine, under which conditions cell growth is arrested, caused an initiation of cell proliferation (data not shown). All of these observations suggest that the inhibitory effects of added ornithine or polyamines on the cytostatic action of NOHA, DETA/NO and DFMO are not attributed merely to a nonspecific increase in cell proliferation.

View larger version (65K): [in this window] [in a new window]   Fig. 5.   Prevention of cytostatic actions of NOHA and NO on Caco-2 tumor cell proliferation by products of arginine-polyamine pathway. Cell proliferation was assessed by thymidine incorporation into DNA during final 24 h of an 8-day growth period in medium containing 30 µM arginine, as described in text. Data were calculated as dpm per 105 cells per well and expressed as percentage of control, where control represents cells grown in absence of any added test agents. Test agents, including ornithine, putrescine, spermidine, and spermine, were added at start of cell growth (addition of 30 µM arginine), and media and test agents were replaced every other day. Concentrations of test agents were 100 µM ornithine, 100 µM putrescine, 300 µM spermidine, or 300 µM spermine. * Statistical significance from control growth (P < 0.05). Data represent means ± SE of duplicate determinations from 6 separate experiments.

View larger version (34K): [in this window] [in a new window]   Fig. 6.   Lack of effect of added products of arginine-polyamine pathway on Caco-2 tumor cell proliferation in presence of excess arginine. Cell proliferation was assessed by thymidine incorporation into DNA during final 24 h of an 8-day growth period in medium containing 30 µM arginine, as described in text. Data were calculated as dpm per 105 cells per well and expressed as percentage of control, where control represents cells grown in absence of any added test agents. Test agents were added at start of cell growth (addition of 30 µM arginine), and media and test agents were replaced every other day. Data represent means ± SE of duplicate determinations from 3 separate experiments.

The findings that the polyamines but not ornithine prevented the cytostatic effect of both DETA/NO and DFMO in a closely similar fashion in this special experiment suggest a common mechanism of cytostatic action for DETA/NO and DFMO. Therefore, DETA/NO was compared with DFMO as a potential inhibitor of ODC activity. With a standard ODC assay, the NO donor agents DEA/NO and SNAP were tested and found to inhibit ODC activity in Caco-2 cell homogenates with a potency that was at least 10-fold greater than DFMO (Fig. 7). NOHA did not inhibit ODC activity. Other experiments revealed that DEA/NO and SNAP also inhibited purified recombinant mammalian ODC (provided by Anthony E. Pegg, Milton S. Hershey Medical Center, Hershey, PA) at concentrations of 3-30 µM (30-70% inhibition; data not shown).

View larger version (31K): [in this window] [in a new window]   Fig. 7.   NO, but not NOHA, inhibits ornithine decarboxylase activity from Caco-2 tumor cells. Ornithine decarboxylase activity was determined by monitoring formation of [14C]CO2 from L-[1-14C]ornithine in a 10-min incubation at 37°C as described in text. Enzyme reactions were initiated by addition of substrate, and test agents were preincubated with enzyme-containing Caco-2 cell homogenates for 1 min at 37°C before addition of substrate. * Statistical significance from CTL (P < 0.05). Data represent means ± SE of duplicate determinations from 3 separate experiments.

The above observations suggest that NOHA and NO inhibit Caco-2 cell proliferation by distinct mechanisms involving inhibition of arginase activity and ODC activity, respectively. The possibility was considered, however, that the cytostatic effect of NOHA might be attributed to conversion of NOHA to NO by NOS or cytochrome P-450. Four observations indicate that the cytostatic action of NOHA on Caco-2 cells is unlikely to be mediated by NO. First, Caco-2 cells did not contain detectable NOS activity, as assessed by a sensitive technique involving the conversion of radiolabeled arginine to citrulline (4, 6). Second, incubation of cells with NOHA for up to 48 h did not result in nitrite or nitrate accumulation, as assessed by chemiluminescence detection (6). Third, 1 mM EITU, a potent NOS inhibitor (36), which abolished NOS activity in cytokine-activated murine macrophages, failed to influence the cytostatic action of NOHA. Fourth, the cytostatic action of NOHA but not DETA/NO was prevented by addition of excess ornithine (Fig. 5).

To determine, albeit indirectly, whether the cytostatic action of NO may be attributed to cGMP, ODQ, an inhibitor of soluble guanylyl cyclase (16), was tested. ODQ, at concentrations of 10-100 µM, failed to alter the cytostatic effects of DETA/NO or NOHA (Fig. 8), providing indirect evidence that the cytostatic action of NO may be cGMP independent. When tested alone on cell proliferation, ODQ was found to be without significant effects (Fig. 8). Assuming that ODQ entered the Caco-2 cells, these observations could be interpreted to mean that cGMP is not involved in Caco-2 cell proliferation or in the cytostatic action of NO.

View larger version (47K): [in this window] [in a new window]   Fig. 8.   1H-[1,2,4]oxadiazolo[4,3-a]quinoxalin-1-one (ODQ) does not inhibit cytostatic actions of NOHA and NO in Caco-2 tumor cells. Cell proliferation was assessed by thymidine incorporation into DNA during final 24 h of an 8-day growth period in medium containing 30 µM arginine, as described in text. Data were calculated as dpm per 105 cells per well and expressed as percentage of CTL, where CTL represents cells grown in absence of added NOHA or DETA/NO. Test agents, including ODQ, were added at start of cell growth (addition of 30 µM arginine), and media and test agents were replaced every other day. DETA/NO and NOHA each significantly inhibited cell proliferation, and these effects were not altered significantly by any concentration of ODQ. Similarly, ODQ alone did not influence cell proliferation. Data represent means ± SE of duplicate determinations from 3 separate experiments.

To determine whether NOS in one cell population can attenuate the proliferation of target cells, LPS/cytokine-activated RAEC (5) were cocultured with Caco-2 cells (Fig. 9). RAEC were previously shown to undergo inducible NOS (iNOS) induction in the presence of LPS plus certain cytokines (IFN-, TNF-, IL-1), resulting in the production and accumulation in the cell medium of NOHA and the oxidation products of NO (5). In the present experiment, RAEC were activated by addition of a cocktail of LPS, IFN-, TNF-, and IL-1. In describing the results, it should be noted first that coculture of Caco-2 cells with nonactivated RAEC resulted in a sevenfold increase in the proliferation rate of the Caco-2 cells (Fig. 9). This stimulation of cell proliferation may be caused by the elaboration of soluble trophic factors (transforming growth factor-, vascular endothelium growth factor, platelet-derived growth factor) from the endothelial cells (11-13). In assessing the influence of activated RAEC on Caco-2 cell proliferation, a comparison was made between activated RAEC and Caco-2 cells in coculture with nonactivated RAEC and Caco-2 cells in coculture. The control represents Caco-2 cells incubated alone and is illustrated only to show that coculture with nonactivated RAEC caused an increase in Caco-2 cell proliferation. Accordingly, Fig. 9 illustrates that activated RAEC markedly inhibited (70-75%) Caco-2 cell proliferation, which was accompanied by NO₂ and NO₃ accumulation in the cell culture medium. Activated RAEC generated NOS products as indicated by the accumulation of nitrite and nitrate in the cell culture medium. The finding that the NOS inhibitor EITU attenuated the cytostatic effect of activated RAEC on Caco-2 cell proliferation supports the hypothesis that NOS products generated by activated RAEC are at least partially responsible for the observed target cell cytostasis.

View larger version (38K): [in this window] [in a new window]   Fig. 9.   Coculture of lipopolysaccharide (LPS)/cytokine-activated rat aortic endothelial cells (RAEC) with Caco-2 tumor cells results in tumor cell cytostasis. Coculture procedures are described in text. Cell proliferation was assessed by thymidine incorporation into DNA. Caco-2 cells were grown in absence or presence of nonactivated (Nonact) or activated (Act) RAEC for 24 h in presence of 100 µM arginine. RAEC were activated by incubation with LPS (100 µg/ml), interferon- (100 U/ml), interleukin-1 (400 U/ml), and tumor necrosis factor- (1,000 U/ml) for 6 h followed by removal of LPS and cytokines by extensive washing before coculture with Caco-2 cells. At 24 h after initiation of coculture, thymidine incorporation into DNA was allowed to proceed for an additional 24 h, after which DNA synthesis in Caco-2 cells was determined. S-ethylisothiourea (EITU; 1 mM) was added to some cocultures as indicated. Coculture media were analyzed for NO2-NO3 after 48 h, and values in µM are indicated above each bar. Statistical significance: * (P < 0.05) from values for cocultures of Caco-2 cells-Nonact RAEC; ** (P < 0.05) from values for cocultures of Caco-2 cells-Act RAEC. Data represent means ± SE of duplicate determinations from 3 separate experiments.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

This study reveals two novel mechanisms by which NOS may inhibit cell proliferation. One mechanism is that NOHA, the major intermediate in the NOS-catalyzed conversion of arginine to NO plus citrulline, is a potent arginase inhibitor (9) that inhibits Caco-2 cell arginase activity both in cell-free extracts and in intact cells in culture. The inhibition of intracellular accumulation of urea by added NOHA correlates well with the antiproliferative action of NOHA, which can be prevented by addition of polyamine pathway products that are distal to arginine, such as ornithine, putrescine, spermidine, and spermine. These observations are consistent with our hypothesis that NOHA interferes with cell proliferation by inhibiting the production of arginine-derived polyamines such as putrescine, spermidine, and spermine. The second mechanism is that NO is a potent inhibitor of mammalian ODC, including the ODC present in Caco-2 cells, thereby suggesting that the antiproliferative action of NO may be cGMP independent and attributed to inhibition of polyamine formation. This view is strengthened by the observed prevention of the cytostatic effect of NO by addition of putrescine, spermidine, or spermine but not ornithine to cell cultures. cGMP may not be involved in expressing the antiproliferative effect of NO in human Caco-2 tumor cells because relatively high concentrations of the potent inhibitor of cytosolic guanylyl cyclase, ODQ (16), failed to influence the cytostatic action of NO. The observations made in this study with ODQ, however, are indirect in that cGMP levels in cells were not determined. Therefore, this study does not rule out the possible involvement of cGMP in expressing the action of NO. It is possible that NO may inhibit cell proliferation by both cGMP-independent and cGMP-dependent mechanisms. These findings are consistent with the view that although cGMP may be primarily involved in the antiproliferative actions of NO in some cell types, cGMP-independent effects may occur in certain cell types, and both effects are possible in other cell types (2, 8, 15, 17, 20, 23, 25, 28, 35, 43, 44).

DETA/NO and SNAP elicited approximately equipotent cytostatic effects in Caco-2 cells, as did NOHA. The similarity in potency between DETA/NO and NOHA may be more apparent than real, however, because DETA/NO releases its NO slowly with a half-life of ~24 h. Therefore, NO may be a more potent cytostatic agent than NOHA. The direct cytostatic action of NO gas itself was not tested because we were not equipped to deliver NO gas to cultured cells continuously for 8 days. The mechanism of arginase inhibition by NOHA is likely one of competition with arginine substrate for the catalytic site in arginase (9). The mechanism of inhibition of ODC by NO is not known. However, previous studies have shown that ODC contains two cysteines at its active site, one of which is critical for enzymatic activity (7). Therefore, NO may inhibit ODC activity by a cGMP-independent modification of a critical sulfhydryl in ODC. In light of the chemistry that has been proposed to take place between sulfhydryls and nitrogen oxide species (10, 42), it is conceivable that NO-mediated sulfhydryl group modification, to form a disulfide, nitrosothiol or related chemical species, is responsible for enzyme inhibition. In support of this view, sodium nitroprusside, which yields both NO and NO⁺ (nitrosonium), has recently been shown to inhibit ODC activity (2).

Both arginase and ODC may be rate limiting with respect to polyamine formation, at least in Caco-2 cells. Some mammalian cells can actively transport ornithine from the blood, and the assumption has been that ornithine is the critical polyamine precursor and ODC is the initial and rate-limiting enzyme in the polyamine pathway. Surprisingly little attention, however, has been paid to arginase, the enzyme that precedes ODC and produces ornithine from arginine. Caco-2 tumor cells contain high arginase activity, and enzyme inhibition is accompanied by cytostasis that can be prevented by adding back excess ornithine. This finding implicates a critical role for arginase as a primary step in the pathway to polyamine formation and cell proliferation. Additional experiments must be conducted, however, to assess more precisely the role of arginase in the regulation of cell proliferation.

Most of the cell culture experiments were performed at an arginine concentration of 30 µM, which is within the physiological range (1). We have noted that Caco-2 cell proliferation was dependent on arginine within a concentration range of 0.1-10 µM, above which concentration the rate of cell proliferation remained relatively constant. Thus the experiments on Caco-2 cell proliferation were conducted at arginine concentrations (30 µM or higher) that were apparently saturating with respect to biochemical pathways governing cell growth. Because NOHA competes with arginine substrate for binding sites on arginase (9), it is possible that NOHA would be a less effective inhibitor of arginase activity, and presumably cell proliferation, as the arginine concentration is raised. This means that the potential modulatory action of NOHA on cell proliferation might be dependent on the local arginine concentration. In the present study, we examined the influence of 30, 100, and 300 µM arginine on the inhibitory effects of 10 µM NOHA on intracellular urea production and cell proliferation and found that NOHA was very effective at all arginine concentrations within the physiological range (30-300 µM). NOHA was equieffective at 30 and 100 µM arginine and somewhat less effective at 300 µM. The data suggest that NOHA (10 µM) is capable of inhibiting cell growth within a fairly wide range of arginine concentrations from 30 to 300 µM.

Coculture experiments with RAEC as the source of iNOS-generated NOHA and NO together with Caco-2 tumor cells as the target cells suggested that NOHA and NO generated by one cell type can inhibit the proliferation of nearby target cells. RAEC were used because these cells can readily undergo induction of NOS by addition of LPS plus one or more cytokines (5). Coculture resulted in slowed target cell proliferation that appeared to be mediated by NOS products because both cytostasis and NO production were attenuated in the presence of the NOS inhibitor EITU. These observations are consistent with the hypothesis that NOHA and NO generated by one cell type are capable of interfering with target cell proliferation. NO would gain access to target cells by simple diffusion, whereas NOHA would utilize the y⁺ basic amino acid transporter to permeate cell membranes (18). Although these experiments indicate that pathophysiological high-output production of NOS products may cause cytostasis, additional experiments are needed to ascertain whether physiologically functional NOS (endothelial NOS) can generate sufficient quantities of NOHA and NO to suppress target tumor cell proliferation.

The coculture data obtained in this study should be interpreted cautiously in deriving mechanisms of cytostatic action because coculture of nonactivated RAEC with Caco-2 cells resulted in marked stimulation of target cell growth. Presumably, this is caused by the elaboration of soluble trophic factors (transforming growth factor-, vascular endothelium growth factor, platelet-derived growth factor) from the endothelial cells (11-13). Activated RAEC inhibited target cell proliferation when a comparison was made between activated RAEC plus Caco-2 cells in coculture and nonactivated RAEC plus Caco-2 cells in coculture. No cytostatic effect is apparent if the comparison is made with Caco-2 cells incubated alone. Therefore, one possible interpretation of these experimental data is that activated RAEC elaborate less growth factor(s) than do nonactivated RAEC. This effect, if applicable, may be mediated in part by NOS products because EITU partially prevented the cytostatic effect of activated RAEC. Coculture systems using various cell types are required to obtain more definitive information on the mechanisms by which LPS/cytokine-activated cells inhibit target cell proliferation.

This study contributes to our understanding of the potential role of NOS and the arginine-NO pathway in the modulation of cell proliferation in two ways. First, we report that NOHA is a potent inhibitor of cell proliferation that may function by inhibiting arginase activity. Second, the mechanism by which NO inhibits cell proliferation may be inhibition of ODC activity by cGMP-independent mechanisms. These observations and mechanisms apply to human Caco-2 tumor cells, and additional experiments are required to determine whether these concepts apply also to other tumor cells or nontumor cells. The two distinct mechanisms of antiproliferative action of NOS products are focused on sequential steps in the critical pathway of arginine conversion to polyamines required for cell proliferation. This complementarity of antiproliferative mechanisms supports the growing belief that NOS can function to slow cell growth.

Perspectives