Optimizing Nitric Oxide Production by Time Dependent L-Arginine Administration in Isolated Human Corpus Cavernosum

Optimizing Nitric Oxide Production by Time Dependent L-Arginine Administration in Isolated Human Corpus Cavernosum Serap Gur, Philip J. Kadowitz, Landon Trost and Wayne J. G. Hellstrom*,† From the Departments of Urology and Pharmacology, Tulane Health Sciences Center, New Orleans, Louisiana

Purpose: We investigated the relaxant effects of repetitive administration of L-arginine, the substrate for nitric oxide, at hourly intervals and elucidated its mechanism of action in human corpus cavernosum. Materials and Methods: Samples of human corpus cavernosum were suspended in an organ chamber for measurements of isometric tension. After precontraction with phenylephrine (10 ␮M), concentration-response curves were performed for L-arginine at 2-hour intervals (1 to 10 hours). Underlying mechanisms of relaxation were evaluated by inhibitory and stimulatory agents. Results: After a brief incubation period of 1 to 4 hours L-arginine (0.1 to 1,000 ␮M) but not D-arginine induced minor changes in HCC. In contrast, when incubation time was increased to 6 to 10 hours L-arginine evoked detectable human corpus cavernosum relaxation proportional to concentration and time. Relaxation was significantly attenuated by the nitric oxide synthase inhibitor L-NAME, the blocker of soluble guanylyl cyclase ODQ and the blocker of small conductance Ca2⫹ activated K⫹ channels apamin, and partially by the inducible nitric oxide synthase inhibitor aminoguanidine and the cyclic guanosine 5=-monophosphate dependent protein kinase G inhibitor Rp-8-pCPT-cGMPS. Relaxation was potentiated in the presence of the membrane permeable cyclic guanosine 5=-monophosphate analogue 8-bromo-cyclic guanosine 5=-monophosphate, the Rho-kinase inhibitor Y-27632 and the phosphodiesterase-5 inhibitor sildenafil. Conclusions: These observations demonstrate that L-arginine induces slow and prolonged relaxation of human corpus cavernosum. This may occur by restoring the endogenous amino acid pool for nitric oxide synthesis and by nitric oxide-soluble guanylyl cyclase-protein kinase G signaling involving the activation of KCa channels or by inhibiting the up-regulated RhoA/Rho-kinase pathway. The use of sildenafil combined with L-arginine further facilitates erections and it may benefit men with more severe erectile dysfunction. Key Words: penis, impotence, sildenafil, arginine, muscle relaxation

he amino acid L-arginine contains 4 nitrogen atoms per molecule and it is the precursor of NO synthesis by constitutive and inducible isoforms of NOS in humans and animals.1 The direct effect of L-arginine has been studied in isolated vascular tissue. Previous observations show that L-arginine specifically and stereoselectively relaxes aortic rings, probably by restoring the endogenous pool of amino acids, which is likely depleted by prolonged incubation.2 L-arginine also can relax intrapulmonary arteries by an endothelium independent mechanism.3 In the penis L-arginine induced relaxation most likely occurs via NOS induced conversion of L-arginine to NO in the penile bulb.4 The NO/cGMP signaling cascade has an essential role in the relaxation of penile corporeal smooth muscle, leading to erection. NO insufficiency can occur as a result of decreased synthesis and/or impaired bioavailablity due to oxidative stress.

T

Submitted for publication December 4, 2006. Study received Tulane University Health Sciences Center institutional review board approval. * Correspondence: Section of Andrology, Department of Urology, Tulane University Health Sciences Center, 1430 Tulane Ave., SL42, New Orleans, Louisiana 70112 (telephone: 504-988-7308; FAX: 504-988-5059; e-mail: [email protected]). † Financial interest and/or other relationship with Auxilium, American Medical Systems, Bayer/GlaxoSmithKline/ScheringPlough, Indevus, Johnson & Johnson/OrthoUrology, ICOS, Mentor, Pfizer, sanofi-aventis, Vivus and King-Palatin.

0022-5347/07/1784-1543/0 THE JOURNAL OF UROLOGY® Copyright © 2007 by AMERICAN UROLOGICAL ASSOCIATION

Decreased NO synthesis can result from decreased eNOS expression, or a lack of sufficient substrate or cofactors. Hence, a hypothesized method to augment NO release from the endothelium is by supplementing with the substrate Larginine. Animal studies revealed the beneficial effect of Larginine supplementation on erectile function.5–7 In isolated rabbit CC Rajfer et al noted that the inhibitory effect of NOS inhibitors on electrical field stimulation induced relaxation is reversed by the addition of excess L-arginine.8 Similar findings were also noted in the CC in humans.6 – 8 The objective of this study was 3-fold. 1) We investigated the direct relaxant effect and the time course of L-arginine substrate on the formation of NO (1 to 10-hour period) in isolated HCC. 2) We determined the mechanisms underlying L-arginine induced relaxation using a number of pharmacological methods. 3) We evaluated the relationship between L-arginine and the cGMP/PKG pathway.

MATERIALS AND METHODS Preparation of HCC Strips Tissue samples were obtained from 46 to 70-year-old patients with ED undergoing penile prosthesis implantation surgery according to Tulane institutional review board guidelines. The tissue was immediately placed in cold Krebs solution composed of 118 mM NaCl, 4.7 mM KCl, 1.2 mM KH2PO4,

1543

Vol. 178, 1543-1548, October 2007 Printed in U.S.A. DOI:10.1016/j.juro.2007.05.121

1544

L-ARGININE MEDIATED RELAXATION OF HUMAN CORPUS CAVERNOSUM

1.2 mM MgSO4, 25 mM NaHCO3, 2.5 mM CaCl2 and 10 mM glucose, and transported to our laboratory. Organ Bath Pharmacology Strips of HCC tissue (5 ⫻ 2 mm) were manually tied at each end with cotton thread and mounted using 1 gm resting tension in 20 ml organ baths of Krebs solution. The bath was maintained at 37C with 95% O2-5% CO2. An FT3 force displacement transducer coupled to a Model 7E polygraph (Grass Instruments, Quincy, Massachusetts) was used to measure tension in HCC tissues. Samples were initially equilibrated for 1 hour and Krebs solution was replaced every 15 minutes throughout the experiment. Establishment of L-Arginine Mediated Response HCC strips were contracted with Phe (10 ␮M), followed by the administration of L-arginine (10⫺8 to 10⫺3 M) or Darginine (10⫺8 to 10⫺3 M) at 1, 2, 4, 6, 8 and 10 hours in a cumulative manner. Concentration response curves were constructed before initiation and following each treatment administration. Characterization of L-Arginine Response In some experiments the concentration-response curves to L-arginine were also obtained in the presence of L-NAME (100 ␮M), ODQ (10 ␮M), AMG (100 ␮M), the PKG inhibitor Rp-8-CPT-cGMPS (10 ␮M), TEA (100 ␮M), apamin (1 ␮M) or indomethacin (10 ␮M). In the second series of experiments 8-Br-cGMP (10 ␮M), the Rho-kinase inhibitor Y-27632 (50 ␮M) and sildenafil (10 ␮M) were added to the organ baths as pharmacological probes at 8 hours. These agents were added when Phe induced contraction attained a steady level. They were allowed to incubate for 30 minutes. Establishing Tissue Viability at the Beginning and End of the Experiments Cavernous tissues used in this study showed similar contractile responses to Phe and relaxant responses to acetylcholine at the beginning and end of the experiments (variability less than 10%). Responses with greater than 10% variability were excluded. Statistical Analysis All values are shown as the mean ⫾ SEM of at least 12 strips per tissue. Maximal effects are expressed as the percent inhibition of L-arginine induced relaxation in contracted tissue. EC50 values were calculated by the linear regression

FIG. 1. Representative relaxation traces of L-arginine obtained from HCC at 2, 4, 6 and 8 hours.

FIG. 2. Effects of L-arginine in 10 ␮M Phe precontracted human HCC strips at hourly intervals. Each point represents mean ⫾ SEM of at least 12 experiments in different strips. Asterisks indicate p ⬍0.01 vs 1 hour.

analysis of L-arginine curves and pEC50 values were [⫺log (EC50)]. Prism® statistical analysis software was used for curve fitting and statistical analysis. Comparisons were performed using 1-way ANOVA, followed by the Bonferronimultiple comparison test with differences considered significant at p ⬍0.05. Drugs Used All were obtained from Sigma® except sildenafil citrate (Pfizer, New York, New York). ODQ and sildenafil were stored as a stock solution (10 ␮M) in 100% dimethyl sulfoxide at ⫺20C and further diluted in deionized water just before use. Indomethacin was dissolved as a 10 ␮M stock solution in 5% NaHCO3 and absolute ethanol. RESULTS Relaxing Activity of L-Arginine and D-Arginine in HCC Figure 1 shows representative traces of relaxation by Larginine at 2, 4, 6 and 8-hour intervals. L-arginine did not statistically produce relaxation at 1, 2 or 4 hours (maximal relaxation 3.06% ⫾ 2.23%, 4.52% ⫾ 1.94% and 15.74% ⫾ 2.42%, respectively, p ⬎0.05, fig. 2). However, subsequent administration of L-arginine progressively decreased Phe evoked contractions by 24% ⫾ 6.09% at 6 hours, by 55.7% ⫾ 9.13% at 8 hours and by 72.3% ⫾ 19.38% at 10 hours in HCC strips (p ⬍0.001, fig. 2). D-arginine (10⫺8 to 10⫺3 M) did not show time dependent responses and it only induced mild relaxation effects (maximal response 8.16% ⫾ 1.10% at 10 hours). The pEC50 values for L-arginine were not altered at hourly intervals (fig. 3). Effect of L-NAME, ODQ and AMG on L-Arginine Induced Relaxation Pretreatment of HCC strips with L-NAME and ODQ significantly attenuated maximal relaxation responses to L-arginine but L-arginine induced relaxation was only partially inhibited by AMG at high concentrations (p ⬍0.05, fig. 4, A). These relaxations were inhibited by L-NAME, ODQ and AMG by 87%, 79% and 63%, respectively. Indomethacin

L-ARGININE MEDIATED RELAXATION OF HUMAN CORPUS CAVERNOSUM

1545

FIG. 3. L-arginine induced relaxation pEC50 values at 1 to 10-hour intervals.

pretreatment did not alter the relaxation response to Larginine (data not shown). Involvement of Kⴙ Channels in L-Arginine Induced Relaxation Incubation of tissues with the nonselective K⫹ channel inhibitor TEA did not significantly decrease L-arginine induced relaxation. Incubation with apamin, a blocker of small

FIG. 5. Concentration-response curves to 10⫺8 to 10⫺3 M L-arginine (L-arg) in 10⫺5 M Phe precontracted strips from HCC after pretreatment with 10 ␮M 8-Br-cGMP and 10 ␮M Rp-8-pCPT-cGMPS (A), and in presence of 10 ␮M sildenafil and 50 ␮M Y-27632 (B). Each point represents mean ⫾ SEM of at least 8 experiments in different strips. Single asterisk indicates p ⬍0.05 vs 1 hour. Double asterisks indicate p ⬍0.01 vs 1 hour.

conductance KCa, significantly inhibited relaxation by 91% at 8 hours (fig. 4, B). Effect of 8-Bromo-cGMP and Rp-8-pCPT-cGMPS on L-Arginine Induced Relaxation L-arginine evoked relaxation was potentiated by the cGMP analogue 8-Br-GMP (fig. 5, A). Further application of the PKG inhibitor Rp-8-pCPT-cGMPS inhibited L-arginine induced relaxation by 49% (p ⬍0.05, fig. 5, A), suggesting the involvement of PKG in the cGMP potentiating effect. Effects of the PDE-5 Inhibitor Sildenafil and the Rho-Kinase Inhibitor Y-27632 Incubation of tissue with sildenafil and Y-27632 increased L-arginine induced relaxation at 8 hours by 38% and 42%, respectively (fig. 5, B). FIG. 4. Concentration-response curves to L-arginine (L-arg) in 10 ␮M Phe precontracted strips from human CC after pretreatment with 100 ␮M L-NAME, 10 ␮M ODQ and 100 ␮M AMG (A), and in presence of K⫹ channel blockers, 100 ␮M TEA and 1 ␮M apamin (B). Each point represents mean ⫾ SEM of at least 8 experiments in different strips. Single asterisk indicates p ⬍0.05 vs 1 hour. Double asterisks indicate p ⬍0.01 vs 1 hour.

Phe Induced Contractions and Acetylcholine Induced Relaxations Phe (10⫺8 to 10⫺3 M) induced contractile responses and acetylcholine (10⫺8 to 10⫺3 M) induced relaxation responses

1546

L-ARGININE MEDIATED RELAXATION OF HUMAN CORPUS CAVERNOSUM ical function of the L-arginine-NO pathway. The application of L-arginine may have caused small increases in NO bioavailability, an effect that increased with the interval of incubation. This can best be explained by the selective and progressive enhancement of the biological pathway activated by L-arginine, leading to progressive HCC relaxation. It is reasonable to examine the relaxation response to L-arginine in various vascular tissues, such as rat aorta, rat cerebral vasculature and human vessels.2,9,10 Detectable relaxation involving iNOS was also observed following 2-hour incubation with L-arginine.10 It is noteworthy that Furchgott reported that rat aortic ring smooth muscle cells are more prone to NOS induction during long experiments (6 to 10 hours).11 Our data suggest that L-arginine selectively activates membrane associated receptors or it may interact with a saturable transport system. These receptors likely enhance the activity of cytosolic NOS, which induces the production of NO in HCC and causes relaxation (fig. 7). Similarly Schini and Vanhoutte previously suggested the presence of L-arginine receptors in the rat aorta.2 Hence, a deficiency of L-arginine in HCC may be due to alterations in L-arginine receptors or another intracellular mechanism. In this study in isolated HCC we focused on the potential intracellular mechanisms of L-arginine induced relaxation. An 8-hour incubation time was noted to be optimal for the study of mechanisms underlying relaxation to L-arginine in strips contracted by Phe. L-arginine induced relaxation was not due to an alteration in the contractile response of strips because Phe evoked similar contractions in strips incubated for several hours. L-arginine mediated corporeal relaxation was significantly inhibited by the NOS inhibitor L-NAME and the sGC inhibitor ODQ, and by pre-exposure to the PKG inhibitor Rp-8-pCPT-cGMPS. This suggests that L-arginine induced relaxation is mediated by NO-cGMP and PKG pathways in HCC. The iNOS inhibitor AMG also attenuated the effect of L-arginine in HCC, suggesting a role of iNOS in HCC relaxation.

FIG. 6. Concentration-response curves to Phe (A) phenylephrine and (B) acetylcholine (10⫺8 to 10⫺3 M) in 10⫺5 M Phe precontracted strips from HCC at 1, 4, 8 and 10 hours. Each point represents mean ⫾ SEM of at least 8 experiments in different strips. g, gm.

were not altered in 4 consecutive concentration-response curves in HCC at 1, 4, 8 and 10-hour intervals (fig. 6). DISCUSSION To our knowledge this is the first report providing information about the time dependent role of L-arginine in the regulation of HCC. In all experiments exposure to L-arginine during the first 4 hours after HCC isolation showed no significant relaxation response. However, after repeat applications of L-arginine at 6, 8 and 10-hour intervals the relaxation phenomenon developed and increased progressively with time to a steady level at 10 hours in a specific and stereoselective fashion. The relaxation evoked by L-arginine was concentration and time dependent. As such, it is likely that a decrease in the concentration of L-arginine in HCC could impair biolog-

FIG. 7. Proposed mechanism of regulation of smooth muscle tone of HCC by L-arginine. L-arginine mediates increases in intracellular NO via eNOS and iNOS activation, leading to increased cGMP, which may be involved in PKG activation, inhibition of up-regulation of RhoA/Rho-kinase system or modification of KCa channels, resulting in hyperpolarization of HCC smooth muscle. Each process leads to smooth muscle relaxation via L-arginine directly or through its receptor mechanisms.

L-ARGININE MEDIATED RELAXATION OF HUMAN CORPUS CAVERNOSUM These data are consistent with findings noted in rat studies using isolated penile bulb and aortic rings.4,10 In the current study we also examined the effects of 8-bromo-cGMP, a cell permeant analogue of cGMP and cGMP dependent protein kinase activator, with L-arginine to more fully understand sGC activation. The addition of 8bromo-cGMP to L-arginine showed a synergistic relaxant effect that was further enhanced by a PDE inhibitor. This incriminates the involvement of cGMP and suggests that the L-arginine response may involve phosphorylation of proteins via a cGMP dependent protein kinase. Prostaglandins are unlikely to be involved in the L-arginine induced relaxant effect because indomethacin did not influence the relaxant effects of L-arginine. Interestingly while the blockade of small conductance KCa by apamin strongly inhibited L-arginine evoked relaxation when administered alone, the nonspecific K⫹ channel antagonist TEA had no effect, arguing against the role of an inward rectifier K⫹ channel. This finding can be explained mechanistically since NO formed from L-arginine administration acts on KCa channels. Furthermore, the maxi-K channel subtype is similar to KCa and it is clearly the predominant K⫹ channel subtype found in corporeal smooth muscle.12 This K⫹ channel is responsible for approximately 90% of the observed whole-cell outward K⫹ current observed in cultured human cavernous smooth muscle cells.12 Since it is postulated that the main pathway for acetylcholine, PGE1 or L-arginine entry is the maxi-K channel,13 alterations in the function and regulation of this channel may have an important role in the pathophysiology of ED. Likewise, activation of this channel in corporeal smooth muscle may represent an important and attractive mechanism to control function, suggesting a possibility for future ED pharmacotherapy. The RhoA/Rho-kinase signaling pathway antagonizes the effects of endothelial derived NO and likely has a role in ED pathogenesis. In HCC smooth muscle the expression and activity of Rho-kinase have been demonstrated.14 Our results suggest that L-arginine may inhibit the up-regulation of Rho-kinase signaling by affecting post-translational modification of RhoA via the NO/cGMP signaling pathway in HCC. Based on the literature it was suggested that the NO/cyclic GMP/cyclic GMP dependent protein kinase pathway controls penile erection by acting in 2 distinct ways, that is by lowering intracellular Ca2⫹ levels, leading to cavernous and vascular muscle relaxation, and by inhibiting Rho-kinase mediated contraction.15 Evidence for inhibition by NO of Rho-kinase induced cavernous contraction during penile erection was obtained in the in vivo rat model.16 These results showed that increased NO and decreased Rhokinase activity synergize in their action to increase intracavernous pressure in the erectile response. On the other hand, a recent study by Ming et al provides evidence for the role of arginase in the stimulation of the RhoA/ROCK pathway in human endothelial cells.17 They suggested that increased arginase activity is associated with higher RhoA expression, which contributes to endothelial dysfunction in atherosclerosis. It is known that arginase competes with eNOS for the substrate L-arginine and decreases NO production. Previous data documented that diabetic HCC with ED had higher levels of arginase II protein, gene expression and enzyme activity compared to those of normal HCC.18 The decreased ability of diabetic tissue to convert L-arginine

1547

to L-citrulline via NOS was reversed by the selective inhibition of arginase by 2(S)-amino-6-boronohexanoic acid. Bivalacqua et al suggested that inhibition of RhoA/Rhokinase improves eNOS protein abundance and activity, thus, restoring erectile function in diabetic rats.19 The relationship among the Rho-kinase pathway, L-arginine depletion and arginase activity must be investigated further. Targeting these interactions in HCC represents a novel therapeutic strategy for ED. Detumescence mainly occurs by the degradation of cGMP by PDE-5. Therefore, the addition of sildenafil to L-arginine results in a synergistic relaxant response in isolated HCC via increased cGMP levels. CONCLUSIONS The results of the current study indicate that L-arginine shows a relaxing effect on corporeal smooth muscle. To our knowledge these are the first reported data demonstrating the importance of L-arginine in HCC. L-arginine relaxes HCC through various mechanisms, including the stimulation of NO/sGC/cGMP/ PKG, the opening of Ca2⫹ activated small conductance K⫹ channels and inhibition of the Rho-kinase system. This effect can be potentiated by PDE inhibition. ED management in the next decade may include substances targeting the L-arginineNO-guanylate cyclase-cGMP-PDE-5 pathway, resulting in increased cGMP. L-arginine supplementation may also have a role in ED therapy by optimizing endogenous NO production. Future investigations in this area must focus on activation of the NO-cGMP pathway.

Abbreviations and Acronyms AMG ⫽ aminoguanidine CC ⫽ corpus cavernosum cGMP ⫽ cyclic guanosine monophosphate EC50 ⫽ affinity constant ED ⫽ erectile dysfunction eNOS ⫽ endothelial NOS HCC ⫽ human CC iNOS ⫽ inducible NOS KCa ⫽ Ca2⫹ activated K⫹ channels L-NAME ⫽ N-nitro-L-arginine methyl ester NO ⫽ nitric oxide NOS ⫽ NO synthase ODQ ⫽ 1H-[1,2,4]oxadiazolo[4,3,-a] quinoxalin-1-one PDE ⫽ phosphodiesterase pEC50 ⫽ apparent affinity constant Phe ⫽ phenylephrine PKG ⫽ protein kinase G Rp-8-pCPT-cGMPS ⫽ Rp-8-[(4-chlorophenyl) thio]guanosine 3’,5’-cyclic monophosphothioate sGC ⫽ soluble guanylyl cyclase TEA ⫽ tetraethylammonium

REFERENCES 1.

Appleton J: Arginine: clinical potential of a semi-essential amino. Altern Med Rev 2002; 7: 512.

1548 2.

3.

4.

5.

6.

7.

8.

9.

10.

L-ARGININE MEDIATED RELAXATION OF HUMAN CORPUS CAVERNOSUM

Schini VB and Vanhoutte PM: L-arginine evokes both endothelium-dependent and -independent relaxations in L-arginine-depleted aortas of the rat. Circ Res 68: 209 –1991. Peng W, Hucks D, Priest RM, Kan YM and Ward JP: Ligustrazine-induced endothelium-dependent relaxation in pulmonary arteries via an NO-mediated and exogenous L-arginine-dependent mechanism. Br J Pharmacol 1996; 119: 1063. Ismailoglu BU, Sahin-Erdemli I and Ilhan M: L-arginineinduced relaxation of the rat isolated penile bulb. Eur J Pharmacol 2002; 435: 113. Ignarro LJ, Bush PA, Buga GM, Wood KS, Fukuto JM and Rajfer J: Nitric oxide and cyclic GMP formation upon electrical field stimulation cause relaxation of corpus cavernosum smooth muscle. Biochem Biophys Res Commun 1990; 170: 843. Kim N, Azadzoi KM, Goldstein I and Saenz de Tejada I: A nitric oxide-like factor mediates nonadrenergic-noncholinergic neurogenic relaxation of penile corpus cavernosum smooth muscle. J Clin Invest 1991; 88: 112. Holmquist F, Hedlund H and Andersson KE: Characterization of inhibitory neurotransmission in the isolated corpus cavernosum from rabbit and man. J Physiol 1992; 449: 295. Rajfer J, Aronson WJ, Bush PA, Dorey FJ and Ignarro LJ: Nitric oxide as a mediator of relaxation of the corpus cavernosum in response to nonadrenergic, noncholinergic neurotransmission. N Engl J Med 1992; 326: 90. Bode-Boger S, Boger RH, Galland A, Tsikas D and Frolich JC: L-arginine-induced vasodilation in healthy humans: pharmacokinetic-pharmacodynamic relationship. Br J Clin Pharmacol 1998; 46: 489. Briones B, Alonso M, Marin J and Salaices M: Role of iNOS in vasodilator responses induced by -arginine in the middle cerebral artery from normotensive and hypertensive rats. Br J Pharmacol 1999; 126: 111.

11.

12. 13.

14.

15. 16.

17.

18.

19.

Furchgott RF: Bioassays with isolated vascular tissue for endothelium-derived relaxing factor, nitric oxide and nitric oxide donors. In: Methods in Nitric Oxide Research. Edited by M Feelisch and JS Stamler. New York: Wiley 1996; p 567. Christ GJ: Frontiers in gene therapy for erectile dysfunction. Int J Impot Res 2003; 15: S33. Christ GJ, S. Richards S and Winkler A: Integrative erectile biology: the role of signal transduction and cell-to-cell communication in coordinating corporal smooth muscle tone and penile erection. Int J Impot Res 1997; 9: 69. Rees RW, Ziessen T, Ralph DJ, Kell P, Moncada S and Cellek S: Human and rabbit cavernosal smooth muscle cells express Rho-kinase. Int J Impot Res 2002; 14: 1. Mills TM, Chitaley K and Lewis RW: Vasoconstrictors in erectile physiology. Int J Impot Res, suppl., 2001; 5: S29. Mills TM, Chitaley K, Lewis RW and Webb RC: Nitric oxide inhibits RhoA/Rho-kinase signaling to cause penile erection. Eur J Pharmacol 2002; 439: 173. Ming XF, Barandier C, Viswambharan H, Kwak BR, Mach F, Mazzolai L et al: Thrombin stimulates human endothelial arginase enzymatic activity via RhoA/ROCK pathway: implications for atherosclerotic endothelial dysfunction. Circulation 2004; 110: 3708. Bivalacqua TJ, Hellstrom WJ, Kadowitz PJ and Champion HC: Increased expression of arginase II in human diabetic corpus cavernosum: in diabetic-associated erectile dysfunction. Biochem Biophys Res Commun 2001; 283: 923. Bivalacqua TJ, Champion HC, Usta MF, Cellek S, Chitaley K, Webb RC et al: RhoA/Rho-kinase suppresses endothelial nitric oxide synthase in the penis: a mechanism for diabetes-associated erectile dysfunction. Proc Natl Acad SciU S A 2004; 101: 9121.