Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Arginase catalyzes the divalent cation-dependent hydrolysis of L-arginine to form L-ornithine and urea. Enzymatic activity is found at highest levels in mammalian liver (Herzfeld and Raper, 1976), where this reaction constitutes the final step of the urea cycle. Urea production provides a mechanism for excreting nitrogen in the form of a highly soluble, nontoxic compound, thus avoiding the potentially dangerous consequences of high ammonia levels. In addition to its presence in the liver, arginase activity has also been found in a number of extrahepatic tissues that lack a complete urea cycle. Such tissues include the lactating mammary gland (Yip and Knox, 1972; Jenkinson and Grigor, 1994), kidney (Kaysen and Strecker, 1973; Gotoh et al., 1996; Vockley et al., 1996; Morris et al., 1997), prostate (Vockley et al., 1996), and activated macrophages (Gotoh et al., 1996). In most of these tissues, the arginase present represents a second isozyme (arginase II) that is distinct in immunologic properties, amino acid sequence, and subcellular location from the more abundant live isozyme (arginase I). The comparative properties of the two arginase isozymes have been discussed in a number of recent reviews (Iyer et al., 1998; Perozich et al., 1998; Jenkinson et al., 1996). The biological functions of arginase II have been the subject of considerable interest; current thinking is that this isozyme provides a supply of L-ornithine for proline and polyamine biosynthesis, and along with arginase I, it serves to regulate the levels of L-arginine available for nitric oxide (NO) production.

Because both NO synthase (NOS) and arginase compete for the same substrate, the possibility of reciprocal regulation of both arginine metabolic pathways has recently been explored (Modolell et al., 1995; Wang et al., 1995). Furthermore, N-hydroxy-L-arginine (L-HO-Arg), an intermediate in the NOS reaction (Pufahl et al., 1992; Yamaguchi et al., 1992; Klatt et al., 1993; Furchgott, 1995; Pufahl et al., 1995), is an endogenous arginase inhibitor (Chenais et al., 1993; Daghigh et al., 1994; Boucher et al., 1994; Buga et al., 1996). The phenomenon of reciprocal regulation between arginase and NOS has only been examined recently (Langle et al., 1995, 1997; Chakder and Rattan, 1997). In the internal anal sphincter (IAS), it was shown that the exogenous administration of arginase causes an attenuation of the NOS-mediated nonadrenergic and noncholinergic (NANC) relaxation (Chakder and Rattan, 1997).

An excess of arginase has recently been associated with a number of pathological conditions that include gastric cancer (Wu et al., 1992, 1994; Leu and Wang, 1992; Straus et al., 1992; Ikemoto et al., 1993), certain forms of liver injury (Ikemoto et al., 1993), and pulmonary hypertension after the orthotopic liver transplantation (Langle et al., 1995, 1997). Furthermore, high levels of arginase may cause impairment in NANC relaxation of the IAS (Chakder and Rattan, 1997). Previous studies in our laboratory have shown that arginase pretreatment causes a significant suppression of the NANC nerve-mediated relaxation of the IAS (Chakder and Rattan, 1997) that is mediated primarily via the L-arginine-NOS pathway (Rattan and Chakder, 1992; Rattan et al., 1992). The impairment in NANC relaxation by excess arginase may be related to L-arginine depletion (Wang et al., 1995). Furthermore, the suppressed relaxation could be restored by the arginase inhibitor L-HO-Arg. It is possible, therefore, that patients with certain conditions associated with an increase in arginase activity may stand to benefit from treatment with arginase inhibitors. However, an arginase inhibitor such as L-OH-Arg may not be selective because it also serves as a NOS substrate (Pufahl et al., 1992, 1995; Chenais et al., 1993; Klatt et al., 1993; Boucher et al., 1994; Furchgott, 1995; Griffith and Stuehr, 1995). Because of this, the exact role of arginase in pathophysiology and the potential therapeutic actions of arginase inhibitors remains undetermined.

The purpose of the present investigation therefore was to test a newly designed and selective arginase inhibitor, 2(S)-amino-6-boronohexanoic acid (ABH) (Baggio et al., 1997) for its effectiveness in the physiologically relevant system. We also examined the effectiveness of ABH on the arginase activity of the IAS, rectum, brain, and liver.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Functional Studies

Preparation of Smooth-Muscle Strips. Studies were performed on circular smooth muscle strips of the IAS obtained from adult opossums (Didelphis virginiana) of either sex after pentobarbital anesthesia (40 mg/kg; i.p.) and subsequent exsanguination. The entire anal canal was isolated carefully using sharp dissection and transferred to a dissecting tray containing oxygenated (95% O₂ plus 5% CO₂) Krebs' solution. The composition of the Krebs' solution was 118.07 mM NaCl, 4.69 mM KCl, 2.52 mM CaCl₂, 1.16 mM MgSO₄, 1.01 mM NaH₂PO₄, 25 mM NaHCO₃, and 11.10 mM glucose. The anal canal was cleaned of the extraneous connective tissue and the blood vessels, opened by an incision along the longitudinal axis, and pinned flat with the mucosal side facing up. The mucosa and submucosa was removed by a sharp dissection. Circular smooth-muscle strips were obtained from the whole circumference of the anal canal and divided into two equal strips (1 × 8 mm). Both ends of the muscle strips were secured with silk sutures (4-0; Ethicon Inc., Sommerville, NJ) and used for the measurement of isometric tension.

Measurement of Isometric Tension. The IAS smooth-muscle strips prepared as described above were mounted onto the thermostatically controlled 2-ml muscle baths (37°C) containing oxygenated (95% O₂ and 5% CO₂) Krebs' solution. One end of the muscle strip was fixed to the bottom of the muscle bath with a tissue holder to the stand, and the other end was attached to an isometric force transducer (model FT03; Grass Instruments Co., Quincy, MA) for the measurement of isometric tension. The smooth-muscle tension was recorded on a Dynograph recorder (model R411; Beckman Instruments, Schiller Park, IL). After an equilibration period of 1 h with intermittent washings, the optimal length and the baseline of the resting tension of each smooth-muscle strip was determined as described (Moummi and Rattan, 1988). Only those smooth-muscle strips that developed spontaneous and steady tension and relaxed in response to electrical field stimulation (EFS) were used.

NANC Nerve Stimulation with EFS. EFS was delivered via a pair of platinum wires from a Grass stimulator (model S88; Grass Instruments Co., Quincy, MA) connected in series to a Med-Lab Stimu-Splitter II (Med-Lab Instruments, Loveland, CO). The Stimu-Splitter was used to amplify and measure the stimulus intensity using the optimal stimulus parameters for the neural stimulation (12 V, 0.5-ms pulse duration, 200-400 mA, 4-s train) at varying frequencies from 0.5 to 20 Hz. These parameters of EFS are known to cause relaxation of the IAS smooth muscle via the selective activation of NANC myenteric neurons. Neurally mediated relaxation of the IAS smooth-muscle strips was quantified in response to different frequencies of EFS. All of the experiments were performed in the presence of atropine (1 × 10⁶ M) and guanethidine (3 × 10⁶ M).

Drug Responses. To determine the influence of arginase on the NANC nerve-mediated relaxation of the IAS, we first determined the effects of different doses of arginase on the relaxation. Arginase at 30 U/ml was found to be the most effective in causing the attenuation of relaxation. The effectiveness of the arginase inhibitors (L-OH-Arg and ABH) on the arginase-induced attenuation of the EFS-induced IAS relaxation was then tested. The optimal doses of arginase and L-OH-Arg in the IAS have been reported (Chakder and Rattan, 1997). To determine the selectivity of the arginase inhibitors in the IAS, we tested their effects on the suppressed IAS relaxation by the NOS inhibitor N-nitro-L-arginine (L-NNA; 3 × 10⁵ M). This was then compared with the reversal by the NOS substrate L-arginine in different concentrations. To examine the physiological relevance of the arginase inhibitors in the IAS relaxation, we examined the influence of different concentrations of arginase inhibitors on the EFS-induced IAS relaxation.

Synthesis of the Arginase Inhibitor ABH. The details of the synthesis of ABH, a boronic acid-based arginine isostere, have been described elsewhere (Baggio et al., 1997). ABH is the most potent inhibitor of Mn²⁺₂-arginase reported to date, with an estimated binding constant of 0.1 µM (Baggio et al., 1997). In the hydrated form, the boronic acid moiety of ABH is believed to mimic the putative tetrahedral intermediate formed during arginine hydrolysis and thus may serve as a transition-state analog.

Enzymatic Studies

Tissue Preparation. Tissue samples of the opossum (IAS) muscle, the adjoining rectal tissue, the liver, and the brain were homogenized with an Ultraturrax tissue homogenizer (Tekmar, Cincinnati, OH) in 10 mM Tris-HCl, 150 mM KCl, and 25 mM MnCl₂, pH 7.4. The resulting homogenates were then dialyzed overnight against 10 mM Tris-HCl, 150 mM KCl, and 25 mM MnCl₂, pH 7.4. The dialyzed homogenates were centrifuged to remove insoluble material and concentrated with Amicon Centricon 30 microconcentrators (Amicon Corp., Danvers, MA) to give stock protein concentrations of 2.4 mg/ml for the IAS smooth muscle, 3.6 mg/ml for the rectal smooth muscle, 3 mg/ml for the brain, and 17. 5 mg/ml for the liver. Protein concentrations were estimated with the Pierce Coomassie protein reagent (Pierce Chemical Co., Rockford, IL) using BSA as a standard.

Arginase Assay. The arginase activity of tissue homogenates was evaluated using the radioactive L-[guanidino-¹⁴C]arginine assay of Rüegg and Russell (1980). Assays were performed in 100 mM 2-(N-cyclohexylamino)-ethanesulfonic acid-NaOH, 0.1 mM MnCl₂ at pH 9.0. The reactions were initiated by the addition of 5 µl of tissue homogenate to 45 µl of reaction mixture that contained the 2-(N-cyclohexylamino)-ethanesulfonic acid buffer, the appropriate concentration of arginine (0.5-5 mM), and 5.0 × 10⁴ cpm of L-[guanidino- ¹⁴C]arginine. The IAS, rectal, and brain homogenates were incubated with the assay mixture at room temperature for 1 h, whereas the liver sample was incubated for 5 min. The reactions were stopped by the addition of 200 µl of a solution containing 0.25 M acetic acid, 7 M urea, and 10 mM arginine at pH 4.5. Arginase has essentially no activity at the low pH of the stop solution. [¹⁴C]Urea was separated from unreacted L-[guanidino-¹⁴C]arginine by treatment with 200 µl of 1:1, v/v, slurry of Dowex 50W-X8 in water and quantitated by adding 200 µl of the supernatant from the Dowex treatment to 3 ml of Liquiscint (National Diagnostics, Manville, NJ) for liquid scintillation counting in a Beckman LS 5000CE counter (Beckman Instruments, Berkeley, CA). The data were analyzed using double-reciprocal plots of the initial velocity measurements; S.E.s were determined by regression analysis.

Arginase Inhibition Studies. Assays in the presence of the inhibitors L-HO-Arg (1-100 µM) and ABH (0.05-100 µM) were performed as described above. The amino acid D-ornithine served as a control in these experiments, because this amino acid, unlike L-ornithine, is not an inhibitor of arginase.

Drugs and Chemicals. Bovine liver arginase, L-HO-Arg, L-NNA, D-NNA, L-arginine hydrochloride, D-arginine, D-ornithine, and atropine sulfate were purchased from Sigma Chemical Co. (St. Louis, MO). Guanethidine monosulfate was purchased from Ciba Pharmaceuticals (Summit, NJ). EDTA (tetrasodium salt) was purchased from Fisher Scientific Co. (Fair Lawn, NJ). ABH was synthesized as described (Baggio et al., 1997). L-[guanidino-¹⁴C]Arginine (specific activity 2.5 GBq mmol¹) was purchased from NEN/DuPont (Boston, MA). All chemicals used were of the highest purity available. Solutions of all the chemicals were prepared in Krebs' solution fresh on the day of the experiment.

Data Analysis. The responses to EFS and other relaxants were expressed as the percentage of maximal relaxation caused by 5 mM EDTA. The results are expressed as means ± S.E. Statistical significance between different groups was determined using Student's t test, and a p value <0.05 was considered significant.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Influence of Exogenous Administration of Arginase before and after Arginase Inhibitors L-HO-Arg and ABH on NANC Nerve-Mediated Relaxation of IAS. First, we determined the effects of different concentrations of arginase on the IAS relaxation by the NANC nerve stimulation. Arginase at 30 U/ml was found to be optimal in causing the attenuation of the NANC nerve-mediated relaxation of the IAS. In experiments examining the influence of L-HO-Arg, the IAS relaxations in response to 0.5, 1, and 2 Hz of EFS in control experiments were 31.0 ± 0.7, 55.9 ± 3.0, and 67.7 ± 3.9%, respectively. After the arginase pretreatment, the IAS relaxation in response to EFS was significantly suppressed, and these values in response to 0.5, 1, and 2 Hz EFS were 12.1 ± 2.5, 28.2 ± 7.3, and 36.6 ± 7.4%, respectively (Fig. 1A; p < .05; n = 5). The pretreatment of the tissues with L-HO-Arg (1 × 10⁴ M) before the addition of arginase antagonized the effect of arginase in attenuating the EFS-induced IAS relaxation. The effect of ABH in antagonizing the arginase-suppressed IAS relaxations was similar to that of L-HO-Arg (Fig. 1B; p < .05; n = 5).

View larger version (19K): [in this window] [in a new window]   Fig. 1.   Effect of exogenous arginase before and after application of the arginase inhibitors L-HO-Arg (A) and ABH (B), on the IAS smooth-muscle relaxation by different frequencies of EFS. Note the significant suppression of the IAS relaxation by treatment with arginase alone and its reversal by L-HO-Arg and ABH. * on the right side of the frequency-response curve in the presence of arginase indicates that the entire curve was statistically different from control (p < .05; n = 5). It is noteworthy that the EFS responses following arginase + L-HO-Arg (A) or arginase + ABH (B) were reversed to near control levels.

Influence of Arginase Inhibitors on NANC Nerve-Mediated IAS Relaxation in the Presence of NOS Inhibitor L-NNA. It is well known that the NOS inhibitor L-NNA causes a marked suppression of IAS relaxation by NANC nerve stimulation. To test the selectivity of the arginase inhibitors in the IAS, we tested their effects on the IAS relaxation suppressed by the NOS inhibitor L-NNA (3 × 10⁵ M). These results were then compared with the reversal of the NANC relaxation of the IAS by the NOS substrate L-arginine at different concentrations. Interestingly, the L-NNA-suppressed IAS relaxation was completely reversed by L-HO-Arg (Fig. 2A). In control experiments, the fall in the basal IAS tension in response to 0.5, 1, 2, and 5 Hz EFS was 26.7 ± 3.6, 46.4 ± 4.05, 64.1 ± 3.0, and 74.8 ± 3.9%, respectively. L-NNA caused a significant attenuation of the IAS relaxation to 0 ± 0, 0.6 ± 0.6, 1.8 ± 1.1, and 5.4 ± 2.1%, respectively (p < .05; n = 5). The NANC nerve-mediated IAS relaxation in the presence of L-NNA plus L-HO-Arg (3 × 10⁴ M) was indistinguishable from that of control values (p > .05; n = 5). In this regard, L-HO-Arg was nearly as potent as L-arginine in causing the reversal (Fig. 2B). In contrast to L-HO-Arg, the newly synthesized arginase inhibitor ABH (3 × 10⁴ M) failed to cause any reversal of L-NNA-suppressed IAS relaxation (Fig. 2C; p > .05; n = 4).

View larger version (23K): [in this window] [in a new window]   Fig. 2.   Percent maximal fall in IAS tension by EFS before and after treatment of the tissue with the NOS inhibitor L-NNA or L-NNA plus different concentrations of L-HO-Arg (A), L-NNA or L-arginine plus L-NNA (B), and L-NNA plus ABH (C). L-NNA caused a marked attenuation of the EFS-induced IAS relaxation (*p < .05; n = 5-7). The L-NNA-attenuated IAS relaxation was reversed by L-HO-Arg and L-Arg completely by 3 × 104 M, and in a concentration-dependent manner (A and B). Note that unlike L-HO-Arg and L-Arg, ABH had no effect on the L-NNA-suppressed IAS relaxation (p > .05; n = 4).

Influence of Arginase Inhibitors on IAS Relaxation Caused by NANC Nerve Stimulation. To determine the physiological significance of arginase in the gastrointestinal smooth muscle, we examined the effects of arginase inhibitors on the NANC nerve-mediated IAS relaxation. Both L-HO-Arg (Fig. 3A) and ABH (Fig. 3B) caused significant and concentration-dependent augmentation of the NANC nerve-mediated IAS relaxation by EFS. This was particularly evident at the lower frequencies of EFS. In control experiments for these series of studies, the fall in the IAS tension with 0.5 and 1 Hz EFS before and after L-HO-Arg (3 × 10⁴ M) was 35.5 ± 7.0 and 54.7 ± 7.0% (0.5 Hz) and 57.4 ± 4.9 and 68.8 ± 4.9% (1 Hz), respectively (p < .05; n = 4). Similar data were obtained in experiments with ABH: 29.3 ± 5.7 and 49.6 ± 3.9% (0.5 Hz) and 56.5 ± 7.0 and 73.6 ± 3.5% (1 Hz), respectively, of the fall in the basal IAS tension before and after the selective arginase inhibitor (1 × 10⁴ M) (p < .05; n = 4).

View larger version (22K): [in this window] [in a new window]   Fig. 3.   Influence of L-HO-Arg (A) and ABH (B) on the NANC nerve-mediated IAS relaxation produced by EFS. Note that both L-HO-Arg (1 × 104 and 3 × 104 M) and ABH (3 × 105 and 1 × 104 M) caused a significant augmentation of the EFS-induced relaxation of the IAS in a concentration-dependent manner, and this was evident at the lower frequencies of EFS (*p < .05; n = 4).

Basal Levels of Arginase Activity in Different Tissues. A comparison of the basal arginase activity of different tissues is given in Fig. 4. Of the tissues examined, the liver was found to contain the highest levels of arginase activity (7400 nmol/min/mg protein), consistent with the role of this tissue in nitrogen metabolism and urea synthesis. Of the nonhepatic tissues tested, the IAS was found to contain the highest levels of arginase activity (7.8 nmol/min/mg protein), whereas the rectum and brain had lower levels (1.7 nmol/min/mg protein). The K_m values for each of these enzymes were similar, ranging from 1.0 to 1.9 mM. These K_m values are similar to those for the native and recombinant rat liver enzymes (Cavalli et al., 1994).

View larger version (32K): [in this window] [in a new window]   Fig. 4.   Basal arginase activity of the tissue homogenates of the liver, IAS, rectum, and brain. The basal arginase activity for the nonhepatic tissues was found to be 103 that in the liver. Of nonhepatic tissues, the basal arginase activity in the IAS smooth muscle was found to be nearly 4-fold higher than in the adjoining region of the rectum and the brain.

Influence of Arginase Inhibitors L-HO-Arg ABH on Basal Arginase Activity in Different Tissues. Of the different tissues investigated, L-HO-Arg was found to be the most potent in inhibiting arginase activity in the liver (Fig. 5A). The IC₅₀ values for the inhibition of arginase activity in liver, IAS, rectum, and brain homogenates by L-HO-Arg were 2.4, 25, 42, and 40 µM, respectively. Thus, liver arginase activity is approximately 10- to 20-fold more sensitive to inhibition by L-HO-Arg than the arginase activities in the other tissues. The ability of ABH to inhibit arginase activity in the tissues was in striking contrast to the inhibition observed with L-HO-Arg. ABH was found to be the most potent inhibitor of arginase activity in brain and rectum, followed by IAS and liver; the corresponding IC₅₀ values were 0.05, 0.05, 0.10, and 0.44 µM for the arginase activities in brain, rectum, IAS, and liver, respectively (Fig. 5B). Inhibition constants for ABH were estimated by titrating the inhibitor into assay mixtures containing an arginine concentration fixed at the K_m value and assuming competitive inhibition. These experiments yielded estimated K_i values of 0.018, 0.026, 0.05, and 0.19 µM for the arginase activities in brain, rectum, IAS, and liver, respectively. The estimated K_i for ABH inhibition of the liver enzyme is in good agreement with the K_d of 0.11 µM determined by titration calorimetry (R.B., S. L. Harper, F.A.E., D. W. Speicher, D.E.A., and D.W.C., unpublished results).

View larger version (17K): [in this window] [in a new window]   Fig. 5.   Effect of L-OH-Arg (A) and ABH (B) on the arginase activity found in hepatic versus the nonhepatic tissues. L-OH-Arg was found to be 10 times more potent in inhibiting the liver arginase activity than the arginase activities from the nonhepatic tissues. In contrast to the effects seen with L-OH-Arg, ABH was found to be more potent in the nonhepatic than in the hepatic arginase activities.

Influence of NOS Inhibitor L-NNA and L-NNA plus Arginase Inhibitors on Arginase Activity in Different Tissues. It is well known that L-HO-Arg is not only an intermediate in the biosynthesis of NO but is also a substrate for NOS. Thus, it is possible that the tissue variations in the inhibition of arginase activities by L-HO-Arg could be due to depletion of added L-HO-Arg by conversion to NO. To test this possibility, the effects of L-HO-Arg on the arginase activities in the different tissue homogenates were determined in the presence of the NOS inhibitor L-NNA. L-NNA at 80 µM had no effect on the activities of the various arginases in the presence of L-HO-Arg (Fig. 6), indicating that depletion of added L-HO-Arg by the action of NOS was not a concern in these experiments. Furthermore, the results indicate that the tissue-specific variations in arginase inhibition by L-HO-Arg are likely to result from inherent differences in the arginases expressed in these tissues.

View larger version (30K): [in this window] [in a new window]   Fig. 6.   Influence of the NOS inhibitor L-NNA, L-HO-Arg, and the combination of L-NNA plus L-HO-Arg on the arginase activity of the liver (A), IAS (B), rectum (C), and brain (D). The data show that L-NNA had no significant effect on either the basal arginase activity or on the L-HO-Arg-attenuated arginase activity in any of the tissues examined. The data suggest that the differential inhibitory effects of L-HO-Arg in these tissues do not result from variable interactions of this NOS substrate with the endogenous NOS.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The present studies have shown that the newly synthesized arginase inhibitor ABH is a potent, tissue-selective inhibitor of arginase. The data show that ABH was 5, 250, 840, and 800 times more potent than L-HO-Arg in inhibiting the arginase activity in liver, IAS, rectum, and brain homogenates, respectively. Of the tissues examined, ABH was more potent in inhibiting brain, rectum, and IAS arginase activity than the liver, with estimated K_i values of 0.018, 0.026, 0.05, and 0.19 µM, respectively, assuming competitive inhibition. Although complete inhibition patterns were not determined, previous studies have shown that ABH can displace the competitive inhibitor L-OH-Arg from the rat liver enzyme (Baggio et al., 1997). In contrast to the inhibition results with ABH, L-OH-Arg was more potent in inhibiting arginase activity in liver homogenates than in the other tissues (Fig. 5A).

Two isozymes of arginase have been described in mammals, the hepatic (type I) arginase and nonhepatic type (type II) arginase (Hecker et al., 1995; Gotoh et al., 1996; Buga et al., 1996; Boucher et al., 1994; Jenkinson et al., 1996; Daghigh et al., 1994). Type I arginase is found predominantly in mammalian liver and red blood cells, whereas the type II enzyme is thought to be expressed in macrophages and kidney and endothelial cells. Although the expression patterns for the arginase in opossum rectum, brain, IAS, and liver are not known, the differential effects of ABH and L-OH-Arg on the arginase activities in these tissues are consistent with type II enzyme expression in the nonhepatic tissues. Verification must wait the purification and characterization of the type II enzyme.

The higher potency of L-OH-Arg against hepatic arginase than the nonhepatic tissue extracts might be related to the ability of L-OH-Arg to serve as a substrate for NOS. Thus, in those tissues expressing high levels of NOS, added L-OH-Arg would be rapidly converted to NO and citrulline, lowering the effective concentration of L-OH-Arg and lowering arginase inhibition. To assess this possibility, inhibition studies with L-OH-Arg were repeated in the presence of L-NNA, a known inhibitor of NOS. Control experiments showed that L-NNA had no effect on the arginase activities of the various tissues. The combination of L-NNA and L-OH-Arg was no more effective than L-OH-Arg alone in inhibiting the arginase activities, indicating that the differential inhibition of the arginase activities in liver, brain, rectum, and IAS is not due to NOS depletion of L-OH-Arg and is therefore likely to reflect differences in affinity of the arginases for the inhibitor.

The functional data in the IAS also support the concept that ABH is more selective in inhibiting arginase than is L-HO-Arg. This was evident from the IAS studies in the presence of the NOS inhibitor L-NNA. As discussed in the introduction, L-HO-Arg may also serve as NOS substrate. The experiments were performed to examine the influence of L-HO-Arg on the NANC relaxation of the IAS that was attenuated by the NOS inhibitor. The effects of L-HO-Arg and ABH in reversing the attenuation of the NANC relaxation were compared with L-arginine, an authentic substrate for NOS. Interestingly, L-HO-Arg caused the reversal of L-NNA-attenuated NANC relaxation of the IAS with a potency comparable with L-arginine. ABH, on the other hand, had no effect on the IAS relaxation suppressed by the NOS inhibitor. This suggests that L-HO-Arg may in part be a substrate for NOS and that it may be less selective than ABH in inhibiting arginase.

Before this study, there was limited information on the physiological relevance of arginase in the NANC relaxation in the gastrointestinal smooth muscle. The present data showed that the arginase inhibitors L-HO-Arg and ABH caused an augmentation of the NANC relaxation of the IAS. Because the NOS pathway is the predominant pathway responsible for the NANC relaxation of the IAS (Rattan et al., 1992; Rattan and Chakder, 1992), the augmentation of the IAS relaxation is speculated to be caused by the up-regulation of the NOS pathway by an increase in the tissue levels of L-arginine. We previously showed that exogenously administered L-arginine in the normal tissues has no significant effect on the NANC relaxation of the IAS unless the tissues are made L-arginine-deficient (Chakder and Rattan, 1997). In the basal state, in the normal tissues, exogenous L-arginine has no significant effect on either the basal IAS tone or the NANC relaxation (Chakder and Rattan, 1997). Conversely, in the L-arginine-deficient tissues, exogenous L-arginine causes a significant fall in the basal IAS tone and reversal of the impaired NANC relaxation in the IAS smooth muscle (Rattan and Chakder, 1997; Chakder and Rattan, 1997). It is possible that the effect of exogenous L-arginine in the L-arginine-deficient tissues is caused by an increase in the uptake of the amino acid, leading to an augmentation of the NANC relaxation via the up-regulation of the NOS pathway. Such mechanisms may not be operative in the normal tissues because of the normal state of equilibrium of L-arginine levels at the cellular levels. The augmentation of the NANC relaxation in the presence of arginase inhibitors may be due to the increase in the intracellular L-arginine levels, as proposed in other nongastrointestinal tissues (Boucher et al., 1994; Hecker et al., 1995; Buga et al., 1996; Jenkinson et al., 1996).

Surprisingly, in contrast to the potencies in the arginase inhibitory activity in the IAS, where ABH was found to be 250 more potent than L-HO-Arg, in the smooth-muscle NANC relaxation experiments, the potencies of these two agents were approximately similar in augmenting the NANC relaxation. There is a plausible explanation for these findings. In the functional data, the smooth-muscle relaxation is the final outcome of the multiple pathways that involve not only arginase but NOS. The net IAS smooth-muscle relaxation in response to NANC nerve stimulation in the presence of inhibitors is the result of their interaction with different pathways available to them. L-HO-Arg acts at both the levels, as an arginase inhibitor and as a substrate for the NOS. ABH, on the other hand, is selective for arginase inhibition only. We (Rattan and Chakder, 1992; Rattan et al., 1992; Chakder and Rattan, 1993) and others (Tottrup et al., 1992; O'Kelly et al., 1993) have shown that the NOS pathway is the predominant pathway for the NANC nerve-mediated smooth-muscle relaxation. Therefore, the augmentation of the NANC relaxation in the IAS by L-HO-Arg may be due to the summation of its effects in causing an increase in the intracellular concentrations of L-arginine, as an NOS substrate plus the arginase inhibition. It is reasonable therefore that in the final analysis, the potencies of the ABH and L-HO-Arg in causing an augmentation of the IAS smooth-muscle relaxation did not turn out to be significantly different. Furthermore, because of the lack of the present knowledge for the relative role of the two pathways in the NANC relaxation, and the actions of the inhibitors on L-arginine uptake mechanisms, it is difficult to predict the influence of these agents on the NANC relaxation. In the enzymatic assays, however, the comparison is straightforward because one is only examining their influence on the arginase activity.

The increase in the levels of arginase has been associated with a number of pathological conditions, including gastric cancer (Wu et al., 1992. 1994; Leu and Wang, 1992; Straus et al., 1992; Ikemoto et al., 1993). Additionally, elevated arginase levels after human orthotopic liver transplantation have been shown to cause pulmonary hypertension and reduced hepatic blood flow (Langle et al., 1995, 1997). Higher blood levels of arginase have been found in patients with various tumors (Straus et al., 1992; Wu et al., 1992, 1994; Leu and Wang, 1992; Parajuli and Singh, 1996) or in certain forms of hepatic injury (Ikemoto et al., 1993). Arginase inhibitors, therefore, may have a significant role in the pathophysiology and potential therapy of a number of disease conditions.

In addition to its therapeutic potentials, the newly described arginase inhibitor ABH may play a novel role in the identification of isozyme-specific arginase pathways and their therapeutic potentials for the specific control of the hemodynamic effects associated with the unregulated arginase activity.

We conclude that arginase plays a significant role in the down-regulation of NOS-mediated NANC relaxation in the IAS. Whether the NOS down-regulation is involved in the pathophysiologic conditions associated with the increase in arginase levels remains to be determined. The availability of the new arginase inhibitor ABH may facilitate the distinction between different isozymes of arginase, ABH being more selective for arginase II and L-HO-Arg for arginase I. The advent of new arginase inhibitor may have important pathophysiological and specifically targeted therapeutic implications in a variety of disease conditions, including certain gastrointestinal motility disorders.