Changes of L-arginine Metabolism in Rat‘s Colon Mucosa under the Conditions of COX/LOX Inhibition and Acute Stress Action

L-arginine is a semi-essential amino acid and a precursor of many biologically active compounds. Polyamines and NO produced from L-arginine take part in the regulation of biochemical processes in colon mucosa. Emotional stress, nonsteroidal anti-inflammatory drugs (NSAIDs) and their combined action can change the activity of L-arginine metabolizing enzymes. The aim of this study was to investigate the single action of NSAIDs with different mechanisms of action and their combination with acute stress on L-arginine metabolism in colon mucosa of rats. Animals were divided into 8 groups: control group (1), administration of nonselective, COX-2 selective and dual COX-2/5-LOX inhibitors (groups 2-4), acute stress group (5), administration of same NSAIDs as in groups 2-4 under the conditions of acute stress (groups 6-8). The activity of iNOS, cNOS, arginase, concentration of L-arginine, nitrite and nitrate was measured in colon mucosa. Nonselective COX inhibition by naproxen caused the increase in iNOS and decrease in cNOS activity in colon mucosa. Both COX-2 (celecoxib) and dual COX-2/5-LOX (2A5DHT) inhibitors enhanced cNOS and arginase acting in combination with acute stress. The concentration of L-arginine remained unchanged in most of the groups, but combination of dual COX-2/5-LOX inhibitor and acute stress raised this parameter.

L-arginine and its metabolites are important regulators of biochemical processes in colon mucosa (CM) and whole gastro-intestinal (GI) tract. The turnover of L-arginine in CM takes place in properly colonocytes 1 , it regulates the proliferation of intestinal stem cells 2 . Exertive metabolism of L-arginine was also identified in macrophages present in lamina propria 3 and in CM biofilm, where it is metabolized by bacteria 1 .
Intracellular pool of L-arginine is formed by few factors -protein degradation 4 , transport from intracellular space by amino acid and L-argininecontaining peptides transporters 5 and endogenous synthesis via intestinal-renal axis 6 . Cationic amino acids transporters (CAT) -CAT-1 and CAT-2 are responsible for L-arginine and other cationic amino acids transport 7 . Endogenous synthesis of L-arginine by intestinal-renal axis covers few steps: L-citrulline is synthesized from glutamine, glutamate and proline in the mitochondria of enterocytes, released from the small intestine and taken up by kidneys for L-arginine production via argininosuccinate synthase and argininosuccinate lyase 8 .
There are four main pathways of L-arginine metabolism -oxidative NO-synthase/ NO pathway and non-oxidative arginase/L-ornithine, arginine decarboxylase/agmatine and arginine:glycine amidinotransferase/creatine pathways 9 (Fig.  1). More than 1% of metabolized L-arginine is utilized for polyamine synthesis and near 2% for constitutive NO production in mammalian cells 6 . All the rest L-arginine can be included into protein biosynthesis 6 . L-ornithine can be used by ornithine transaminase to provide L-proline 10 , which takes part in collagen and other proteins biosynthesis. Arginine:glycine amidinotransferase takes part in creatine synthesis mainly in kidneys, but is also present in intestines 11 (Fig. 1).
The utilization of L-arginine by NOS/NO pathway in CM is highly augmented in state of stress 15 and inflammation 16 due to infiltration of CM with iNOS-expressing macrophages 17 . Polyamines like putrescine, spermidine and spermine protect CM from endotoxins and activate protein synthesis 18 , induce the secretion of immunoglobulin A, maintain barrier integrity 19 , DNA stabilization, activate ion channel transport, scavenge free radicals and induce cell proliferation 20 . However, polyamines synthesis can be disturbed under the conditions of elevated consumption of L-arginine by iNOS 1 .
Nonsteroidal anti-inflammatory drugs (NSAIDs) are still one of the most wideused medications 21 . They act by inhibition of cyclooxygenase (COX) pathway of arachidonic acid (AA) metabolism. It results in downregulation of prostaglandins (PGs), mainly prostaglandin E 2 (PG E 2 ) synthesis 22 . Basal level of PG E 2 synthesis is carried out by constitutive isoform of COX (COX-1) in epitheliocytes, located in lower half of crypts 23 . PG E 2 binds with PGE 2 receptors and activates the production of mucus by CM via cAMP dependent mechanism 24 . Thromboxane A 2 (TX A 2 ) and prostacyclin (PG I 2 ) synthesis is also catalyzed by COX-1, they ate blood flow and thrombocytes aggregation 25 . Expression of inducible COX-2 is activated in the apical epitheliocytes of villi and inflammatory cells of lamina propria in case of inflammation 26 . The large intestine shows relatively low levels of COX-1 expression compared with that of other parts in GI tract 27 . Sharply increased synthesis of PG E 2 by COX-2 can determine further inflammation reactions after binding with PGE 2 receptors located on mast cells, T-helpers 28 and other immune cells. Thus, inhibition of excessive PG E 2 synthesis is the base of anti-inflammatory action of NSAIDs.
However, it is well-known that long-term and/or high-dose administration of NSAIDs can cause ulcers of GI mucosa 29 . Ulcerogenic action of NSAIDs is connected with mechanism of their action, mainly with the COX selectivity 30 . Declined activity of COX-1 leads to lower the production of mucus 31 , so CM becomes more sensitive to endotoxins, flagellins and bacterial lipopolysaccharides (LPS) action. In this regard, COX-2 selective NSAIDs are actively synthesized and studied nowadays. However, COX inhibition can cause the shift of AA metabolism on lipoxygenase (LOX) pathway 32 , which flows out in raised production of proinflammatory leukotriene B 4 (LT B 4 ) by 5-LOX isoform 33 . Dual COX-2/5-LOX inhibitors (darbufelone, licofelone, 2,3-diarylxanthones, chebulagic acid etc.) are considered to be more effective NSAIDs, as they block both pathways of AA metabolism 34 , while LOX inhibitors can cause the switchover of AA metabolism on COX pathway.
Emotional stress is another, but no less important factor of modern life. Like NSAIDs, stress can cause ulcers in GI mucosa 35 . The mechanism of its action is associated with adrenaline 36 , which can cause the vasoconstriction, hypoxia and activation of nitroso-oxidative processes 37 . On the other hand, slow influence of stress is mediated by activation of hypothalamicpituitary-adrenal axis via hyperproduction of corticotropin releasing hormone, corticotropin and cortisol 38 . Cortisol activates the expression of lipocortin-1, which inhibits phospholipase A 2 -a key enzyme of AA synthesis 39 . So, both COX and LOX pathways become inactivated like under the action of some NSAIDs.
Taking into account cytoprotective and proinflammatory pathways of L-arginine and AA metabolism, the purpose of this study was to find the cross-links between these pathways. The main parameters of L-arginine metabolism pathways in CM in the normal state, under the conditions of acute stress and action of different COX/LOX inhibitors were defined.

Materials and Methods animals
All animal procedures were carried out in accordance with international guidelines for the use and care of laboratory animals. The permission was granted by Bioethics Committee of Danylo Halytsky Lviv National Medical University (Protocol ¹3, 16.03.2015). Total number of 80 12-week-old male outbred albino rats (Rattus norvegicus domestica), 200-240 g body weight were obtained from Danylo Halytsky Lviv National Medical University vivarium. All rats were housed in an environmentally controlled room under a 12 h light/dark cycle. There was ad libitum access to water and animals were kept on the standard diet in groups of five in cages. All the experiments were performed at the light cycle. experimental groups Animals were randomly divided into 8 groups (Fig. 2), 10 rats in every group. The first group was control, animals were administered NaCl solution and taken out of the experiment after 1 hour. Rats of groups 2-4 were receiving COX/LOX inhibitors by the next scheme: 2 nd was administered by nonselective COX inhibitor (naproxen), 3 rd -COX-2 selective inhibitor (celecoxib), 4 th -dual COX-2/5-LOX inhibitor (2A5DHT) and were taken out of the experiment after 1 hour. The remaining 40 rats were fasting for 24 hours. Water-immersion stress (WIS) was carried out on animals of group 5, after that, they were taken out the experiment. Animals of groups 6-8 were administered the same COX/ LOX inhibitors in the same dose and way as rats of groups 2-4. Animals of groups 6-8 were exposed to the WIS action after 30 min, then they were taken out the experiment. Euthanasia was done by urethane injection.

Wis model
The WIS model was executed as described by K. Takagi 40 , with 5-hours modification 41 . Rats were fasting for 24 hours before WIS modeling and had free access to the water. Animals were fixed in plastic boxes in such a way that their body remained stationary. The boxes with rats were immersed into a water bath (t=23 °C), so that their heads were upon the water surface. Animals were kept in this position during 5 hours, water temperature was controlled every 30 min. Euthanasia was performed immediately after the experiment was finished.

tested compounds, doses and ways of administration
Sterile sodium chloride solution in a concentration 0.9% (Arterium, Lviv, Ukraine) was administered intragastrically in a volume of 1 mL to a control group rats. Three COX/LOX inhibitors were used: nonselective COX-inhibitor naproxen sodium salt (Sigma-Aldrich, St. Louis, MO, USA), selective COX-2 inhibitor celecoxib (American Custom Chemicals Ltd., San Diego, CA, USA), dual COX-2/5-LOX inhibitor 2A5DHT (2-amino-5-(3,5-ditertbutyl-4-hydroxybenzylidene)-thiazol-4-one) -a structural analogue of darbufelone. Compound 2A5DHT (Fig. 3) was synthesized in the laboratory of Department of organic, bioorganic and pharmaceutical chemistry (Danylo Halytsky Lviv National Medical University, Lviv, Ukraine) by method described previously 15 . All the tested compounds were administered intragastrically in a single dose of 10 mg/kg, in a volume of 1 mL. Naproxen sodium was dissolved in distilled water, while celecoxib and 2A5DHT were dissolved in water with addition of polysorbate-80 (10% of the total solution volume) for emulsification. Urethane (Sigma-Aldrich, St. Louis, MO, USA) was used for euthanasia in a single dose 4 g/kg, injected intraperitoneally, dissolved in a distilled water.

arginase assay
Total arginase (both arginase-1 and arginase-2) activity was checked by an approach, based on the activity of urea production 43 . The reaction with á-isonitrosopropiophenone, diluted in ethanol was used for urea detection.
The concentration of urea was revealed using a calibration curve. Activity of arginase was expressed in ìmol of urea/min·mg of protein.

l-arginine assay
Sakaguchi reaction was used for L-arginine concentration determination 44 . The concentration of L-arginine was found using a calibration curve, it was expressed in ìmol/g of tissue.

nitrite-, nitrate-anion assay
The reaction with Griess reagent for nitrite-anion determination and zinc powder to reduce nitrate-anion into nitrite-anion 45 was used. All the reagents were dissolved in water for injections to avoid the false results, as distilled water may contain nitrite-/nitrate-anion. Absorbance (ë=550) was estimated and nitriteanion concentration was calculated using a calibration curve. The values in tubes with zinc corresponded to the total concentration of nitrite-/ nitrate-anions concentration, which was expressed in ìmol/g of tissue.

Protein assay
A modified assay for protein determination was used, which is based on a biuret reaction 46 . The concentration of protein was distinguished using a calibration curve. Concentration of protein was expressed in ìmol/L.

statistical analysis
All experimental values are expressed as the mean ± standard deviation. Descriptive statistics and normality test (Shapiro-Wilk test) were complied, using BioStat Pro Version v6 (BioStat, AnalystSoft Inc., Walnut, CA, USA). The distribution was normal in all the tested groups. Unequal-variance t-test was performed to designate reliability using Excel (Microsoft Office, Microsoft, Redmond, WA, USA). Values p<0.05, p<0.01, p<0.001 were considered statistically significant.

results
Administration of COX/LOX inhibitors did not change significantly the concentration of L-arginine and nitrite-anion in CM, it affected mainly on NOS and arginase activity.
Thus, nonselective COX inhibitor naproxen administration augmented the iNOS activity by 19% and decreased cNOS activity by 12% (p<0.01) (Fig. 4) in CM compared with that of a control group. Tendency of arginase activity to be lower and nitrate-anion to be higher (p>0.05) was noted.
Celecoxib, like naproxen, did not change the concentration of L-arginine and nitrite-anion. However, it caused iNOS activation by 22% (p<0.01) on the one hand, and tendency of cNOS and arginase activity to be lower (p>0.05) in CM compared to that of a control group (Fig. 4). Like naproxen, celecoxib caused the tendency of nitrateanion concentration to be elevated (p>0.05).
There were no significant changes of studied parameters in CM under the action of dual COX/LOX inhibitor A25DHT. However, we noted the tendency of arginase activity and nitrite-anion to be lower (p>0.05) compared with that of a control group.
Action of WIS caused the diminishment of L-arginine concentration by 27%, constitutive enzymes' activity -cNOS by 22% and arginase by 24% (p<0.001) in CM compared with that  (Fig. 5). At the same time iNOS activity was 2.3 times higher (p<0.001), concentration of nitrite-and nitrate-anion increased by 20% and 27% (p<0.05) compared with that of a control group (Fig. 5).
Administration of COX inhibitors in case of WIS tended to raise L-arginine concentration in CM. Though, the general enzymatic activity did not change significantly compared with that of a WIS group, some interesting changes of enzymes' activity ratio were noted.
Action of naproxen under the conditions of WIS caused the tendency of L-arginine concentration to be elevated (p>0.05), activity of iNOS was lowered by 9% (p<0.05) in CM compared with that of WIS group (Fig. 5). Activity of cNOS, arginase, concentration of nitrite-and nitrate-anion did not change significantly in this case.
The linked action of celecoxib caused the tendency of L-arginine concentration to be augmented (p>0.05), arginase activity was higher by 12% and iNOS activity -lower by 14% (p<0.05) in CM compared with that of WIS group. Nitriteand nitrate-anion concentration, cNOS activity remained unchanged (Fig. 5).
Dual COX/LOX inhibitor -2A5DHT -was the only compound, which increased L-arginine concentration by 17% (p<0.05); iNOS activity was lowered by 15% (p<0.01), while arginase activity was higher by 12% (p<0.05) in CM compared with that of a WIS group (Fig. 5). The tendency of cNOS to be higher (p>0.05) and no changes of nitrite-and nitrate-anion concentration were confirmed.

discussion
It has to be noted that CM consists of different cell types, where expression of main L-arginine and AA-metabolizing enzymes is not equable. Moreover, CM is covered by a biofilm, which is inhabited by lots of bacteria capable of metabolizing L-arginine 47 . We consider the inability to study the L-arginine metabolism separately in every cell type and bacterium species as the main limitation of our study.
The first step of this investigation was to find out the influence of a single action of COX/ LOX inhibitors on L-arginine metabolism in CM. In general, three possible levels of studied compounds action can be regarded: their action as COX/LOX inhibitors, their topical action as compounds were injected intragastrically and other mechanisms not related to COX/LOX inhibition. All the tested compounds are inhibitors of PG E 2 synthesis, so AA was metabolized by other noninhibited enzymes. None of studied compounds As a result of a nonselective mechanism of naproxen action, it causes a decrease in all the PGs concentration. This results in an elevation of LT B 4 concentration according to excess of AA and ability of LOX, mainly 5-LOX, to metabolize it 48 . PG E 2 is responsible for mucus and bicarbonates secretion in GI mucosa 31 . PG I 2 and TX A 2 are local regulators of the bloodstream 49 . LT B 4 is considered to be a potent proinflammatory agent 50 . The transition of balance between proinflammatory and cytoprotective mediators can cause the activation of inflammation by two mechanisms. The first one is related with macrophages activation by LT B 4 via LTB 4 -receptors 51 and later raise in their number in lamina propria 52 . Thus, enhanced activity of iNOS in this group may emerge from macrophages activation 53 . The second mechanism may be associated with lowered mucus production by cause of PG E 2 deficiency. Mucus protects CM against bacterial LPS, flagellins and toxins supplied by rat's own cells 54 . Moreover, due to the half maximal inhibitory concentration (IC-50) of naproxen, it shows higher affinity to COX-1 (IC-50 35.48 ìmol/l) compared with that of COX-2 (IC-50 64.62 ìmol) 55 . According to our previous investigation, naproxen did not cause histological changes in CM, but the augmentation of free radical oxidation and antioxidant enzymes activity was noted 56 . This fact supports the idea about proinflammatory action of naproxen. The reduction of cNOS and arginase activity may occur by reason of the competition with iNOS for a substrate or as a result of direct action of naproxen on their activity. More expressed changes were obtained in a gastric mucosa -activation of iNOS in gastric mucosa under the action of naproxen led to increased nitrite-anion concentration 57 .
The action of celecoxib -a selective COX-2 inhibitor -was accompanied by an elevation of iNOS activity and tended to decline cNOS activity, which may be caused by competition for L-arginine. As in a case with naproxen, the activity of 5-LOX was not inhibited, the AA was accumulating, and this underlies the raised iNOS activity. However, no changes of arginase activity were noted. It can be explained by not inhibited COX-1 activity, which contributed the normal production of other PGs. Thus, action of constitutive eicosanoids on arginase activity needs further investigation. Selective COX-2 inhibition demonstrated that the competition between NOS and arginase enzymes is not the only one reason of diversity in inducible and constitutive enzymes' activity. The investigation of celecoxib action on the gastric mucosa showed more expressed changes -enhanced iNOS activity and nitriteanion concentration, diminished or tended to be diminished cNOS and arginase activity 57 .
Dual COX/LOX inhibitor 2A5DHT caused the tendency of arginase activity to be fallen off, while the activity of NOS isoforms did not change. According to its mechanism of action, 2A5DHT is inhibiting the production of PG E 2 and LT B 4 . In this case AA metabolism may be crossed to the local vasoactive PGs production. Both PG I 2 and TX A 2 can be produced, so it remains unclear which one could decrease arginase activity and is this effect related to eicosanoids. We do not connect the lowering of arginase activity with competition for the substrate as the concentration of L-arginine and the activity of NOS isoforms did not change.
By cause of pharmañokinetic properties of tested compounds both naproxen 58 and celecoxib 59 are absorbed in 2-4 hours in upper GI tract when administered per os. Pharmacokinetic parameters of 2A5DHT are not well-studied. Giving this, we do not support the idea of topical action of naproxen and celecoxib on CM, whereas 2A5DHT theoretically could have this property. Our previous investigation showed that neither celecoxib, nor 2A5DHT cause histological changes or prooxidant action, unlike naproxen 56 . We suppose that the proinflammatory impact of COX/LOX inhibitors can be observed in case of both iNOS activation and cNOS/arginase inhibition, while one of these actions separately does not cause the inflammation.
Action of WIS caused typical changes of studied parameters 57,60 . The decline of L-arginine concentration in blood plasma was also noted 57 . We did not measure the concentration of L-arginine in blood plasma as there is a constant flux between the intracellular and extracellular space for L-arginine 61 . L-arginine concentration falling off in CM may be caused by a sharp activation of iNOS. Activation of this enzyme is mediated by CM infiltration with macrophages and activation of nonspecific antibacterial action. This process could be triggered by the action of adrenaline, which causes vasoconstriction 62 . Activation of brain-gut axis cause the inhibition of phospholipase A 2 by cortisol 63 and augments visceral sensitivity by corticotropin releasing factor 64 . Glucocorticoids affect the sympathetic nervous system, mediating sensitivity of adrenoreceptors to catecholamines 65 , what may increase their effect. According to the literature data, expression of COX-2 and increase of PG E 2 concentration are raised during the acute stress in rats 66 . Decrease in L-arginine concentration also can be caused by the defect in CAT-2 function in CM in state of inflammation 61 . The deficiency of L-arginine is contributing to impairment in epithelial restitution, protein translation, phagocytosis, and innate immunity, as well as dysregulation of apoptosis 61 . Defective L-arginine uptake can lead to the diminished ability of arginase/ornithine decarboxylase pathway to generate polyamines 61 . Polyamines have many biological functions in CM, including inhibition of Th1 cytokines production, monocyte activation, stimulation of epithelial restitution by enhancing cell migration and proliferation, regulation of apoptosis 61 . Literature data are showing controversary affection of glucocorticoids on arginase activity -it can be downregulating 67 and upregulating 68 . Although it was reported about the co-induction between both arginase-1, arginase-2 and iNOS in macrophages 68 , we noted the opposite action. It can be caused by determination of enzymatic activity in CM homogenates, which consist of different cell types, and macrophages were not the most significant among them. Going back to NOS activity, it can synthesize not only NO, but also N-hydroxyarginine and superoxideanion 69 . The last one activates arginase, whereas N-hydroxyarginine -inhibits 69 . Sharp elevation of NO production by iNOS in CM caused the elevated concentration of nitrite-/nitrate-anion respectively, what was also noted in gastric 57 and colon 70 mucosa. Increased concentration of NO can lead to enhanced formation of peroxinitrite anion, nitration of protein tyrosine, activate nitrosation processes. The last one is regarded to be one of the factors leading to augmented expression of some proinflammatory genes like tumor necrosis factor-á and p38 mitogen-activated protein kinases 69 .
Action of COX/LOX inhibitors in case of WIS action was different compared to their single action. Naproxen did not show pro-inflammatory impact, moreover the iNOS activity tended to decrease, although it was not enough to shift the L-arginine metabolism on constitutive pathways. We suppose that the reduction of iNOS activity in this case is associated with inhibition of both COX isoforms. In case of a single action, naproxen is supposed to inhibit COX-1 mainly, as the activity of COX-2 is minor. On the other hand, inhibition of COX-1 by naproxen could block the biosynthesis of TXs more than PG I 2 , thus increasing the blood flow in the ischemic CM. Related effect of combined action of naproxen and WIS was noted in gastric 57 and small intestine mucosa 60 .
The coupled action of celecoxib and WIS cannot be considered proinflammatory. However, as described above, the single action of celecoxib enhanced the activity of iNOS. This may point to the cytoprotective role of COX-1 in course of inflammation. On the other hand, WIS action elevates COX-2 activity, which is the main target for celecoxib action. Single administration of celecoxib could partially inhibit COX-1 also, thus causing biochemical changes that led to iNOS activation. Acting on the gastric mucosa under the conditions of WIS, celecoxib declined iNOS activity and did not change cNOS and arginase activity 57 or diminished arginase activity 70 .
Dual COX/LOX inhibitor 2A5DHT showed the most active regulation of L-argininemetabolizing enzymes activity in CM in state of WIS action. Uptake of L-arginine and its metabolism by arginase to L-ornithine and conversion to L-proline by ornithine aminotransferase is essential for colonic epithelial lesion repair 71 . Moreover, this compound was the only one which increased the concentration of L-arginine in CM. As it was mentioned, WIS may cause ischemia of CM due to augmented adrenaline production. Synthesis of PG I 2 , TXs and some other PGs was not inhibited by 2A5DHT. By reason of the reversed vasoactive influence of PG I 2 and TX A 2 and cytoprotective impact of 2A5DHT administration we suppose that AA was metabolized by a prostacyclin synthase mainly.
Interestingly, that the action of all the tested COX/LOX inhibitors did not change the total activity of L-arginine-metabolizing enzymes. Only the administration of dual COX/LOX inhibitor raised the concentration of L-arginine, which can be used in anabolic pathways. Moreover, according to experimental data, supplementation of rats with L-arginine attenuated the degree of tissue damage in the intestinal ischemia and promoted healing of intestinal mucosa 61 . Colonic epithelial restitution is dependent on L-arginine transport into the cells by CAT-2, which is interrupted in case of inflammation 71 . According to the literature data, global protein translation is low in the absence of L-arginine 6 . Most studies have demonstrated the beneficial action of L-arginine on GI function improving gastric ulcer healing, accelerating intestinal mucosal regeneration, enhancing bacterial clearance, and reducing histological bowel necrosis 6 . Like NOS isoforms, arginase-1 and CAT-2 are also regarded as inducible, while arginase-2 and CAT-1 -constitutive isoforms 72 . It was reported, that iNOS and argininosuccinate synthetase activities can be co-induced in LPS and interferon-ã stimulated macrophages 72 . This underlies the re-synthesis of L-arginine from L-citrulline. We did not observe this hypothetical elevation of L-arginine concentration under the single action of naproxen and WIS, which caused the highest augmentation of iNOS activity. That is why we speculate that L-arginine increasing effect of 2A5DHT may be mediated by PG I 2 . It serves for vasodilation, so the concentration of extracellular L-arginine becomes higher. As CAT-2 is an inducible L-arginine transporter localized on macrophages, it is possible that during WIS the transport of L-arginine was increased in all the groups, but catabolized immediately by activated iNOS. The argument for this assumption is lower value of Michaelis constant (Km) of iNOS (13 ìM) 73 compared with that of arginase-1 (3.3 mM) and arginase-2 (1.9 mM) 74 .
Our basic assumption is related to the shift of AA metabolism to different directions under the action of COX/LOX inhibitors, which defines proinflammatory or cytoprotective action. An alternative version is associated with accumulation of AA under the action of COX/LOX inhibitors. AA activates calcium entry into the cell, which can stimulate cNOS activity without iNOS inhibition 69 . However, literature data suggest that iNOS and arginase activities are regulated reciprocally in macrophages by cytokines, and this may guarantee the efficient production of NO 75 .

conclusions
The single action of nonselective COX inhibitor naproxen can be considered proinflammatory to colon mucosa as it enhances iNOS activity and decreases cNOS/arginase activity. Action of selective COX-2 (celecoxib) and dual COX-2/5-LOX (2A5DHT) inhibitors in combination with acute stress decreased iNOS activity and upregulated cNOS and arginase activity as well. 2A5DHT raised the concentration of L-arginine, showing the most pronounced antiinflammatory effect. The transition of arachidonic acid metabolism from inducible to constitutive pathways of metabolism can cause a complementary transition of L-arginine metabolism.