Page 1 Articles of 38 in PresS. Am J Physiol Heart Circ Physiol (October 27, 2006). doi: /ajpheart

Page 1 Articles of 38 in PresS. Am J Physiol Heart Circ Physiol (October 27, 26). doi:1.1152/ajpheart.121.25 Overexpression of Arginase in the Aged Mouse Penis Impairs Erectile Function and Decreases Endothelial NOS Activity: Influence of In Vivo Gene Therapy of Anti-Arginase Trinity J. Bivalacqua 1, Arthur L. Burnett 1, Wayne J.G. Hellstrom 2, and Hunter C. Champion 3 1 Brady Urological Institute, Johns Hopkins Hospital, Baltimore, MD 21287 2 Department of Urology, Tulane Health Sciences Center, New Orleans, LA, 7115 3 Division of Cardiology, Department of Medicine, Johns Hopkins Hospital, Baltimore, MD 21287 Short title: Arginase and Aging in the Penis Address correspondence to: Hunter C. Champion, M.D., Ph.D. Assistant Professor of Medicine Division of Cardiology Department of Medicine Johns Hopkins Hospital 72 Rutland Avenue, Ross 85 Baltimore, MD 2125 Phone: 41-52-18 Fax: 41-52-2557 E-mail: hcc@jhmi.edu Copyright 26 by the American Physiological Society. 1

Page 2 of 38 Abstract As both increased nitric oxide synthase (NOS) abundance and diminished nitric oxide (NO) signaling have been reported in the aging penis, the role of NO in the adaptations of aging remain controversial. Here we tested the hypothesis that arginase, enzyme that competes with NOS for the substrate L-arginine, contributes to erectile dysfunction (ED) with advanced age in the B6/129 strain of mouse. Arginase protein abundance, mrna expression, and enzyme activity were elevated in aged compared with young penile endothelial cells. In addition, endothelial NOS (NOS3) protein abundance was greater in aged vs young penile endothelial cells while NOS activity and cgmp levels were reduced. Calcium-dependent conversion of L- arginine to L-citrulline and cgmp formation increased significantly in the aged mouse penes in the presence of the arginase inhibitor, 2(S)-amino-6-boronohexanoic acid (ABH). However, there was no effect on L-arginine to L-citrulline conversion or cgmp accumulation in the endothelium from young mouse penes. To assess the functional role of arginase in the inhibition of NOS pathway responsiveness in the penis, we evaluated the effects of ABH and an adeno-associated virus encoding an antisense sequence to arginase I (AAVAnti-Arginase) on erectile function in vivo. ABH and AAVAnti-Arginase enhanced endothelium-dependent erectile responses in the aged mice without altering endothelium-independent responses. Paralleling our in vitro observations, ABH or AAVAnti-Arginase did not affect vascular responses in the young mice. Inhibition of the arginase pathway improves endothelial function in the aging penile circulation, suggesting that the arginase pathway may be exploited to improve ED-associated with aging. Key Words: arginase, nitric oxide, gene therapy, enos, aging 2

Page 3 of 38 Introduction Penile erection is a neurovascular event, which depends on neural integrity, a functional vascular system, and healthy cavernosal tissue. The release of neurotransmitters from cavernous nerve terminals and corporal smooth muscle endothelium in response to sexual stimulation results in cavernosal smooth muscle relaxation and ultimately penile erection. The nitric oxide (NO)/cGMP system is the principal mediator of corporal smooth muscle relaxation and is synthesized by the neuronal and endothelial isoforms of nitric oxide synthase (NOS1 and NOS3, respectively; 13, 29, 43). Erectile dysfunction (ED) is defined as the inability to achieve and/or maintain penile erection sufficient to permit satisfactory sexual intercourse. Pathological alteration in the anatomy of the penile vasculature or impairment of any combination of neurovascular processes can result in erectile dysfunction (ED; 1, 8). Although a number of pathophysiologic processes are recognized to lead to abnormal function and responsiveness of the penile vascular bed, the natural aging process is also associated with ED with elevated vasoconstrictor tone and decreased endothelium- and neurogenic-mediated relaxation of the corpus cavernosum (1, 8, 23, 25, 33, 37). Aging-associated ED involves aberrancies at multiple levels of the NO/cGMP cascade in the penis including, reduced NANC nerve fibers in the penis, decreased constitutive NOS activity, impaired endothelial-dependent smooth muscle relaxation, and diminished NObioavailability (15, 16, 9, 24). It has been shown that dietary L-arginine supplementation as well as acute infusion of L- arginine results in improved NO release and enhanced endothelium-dependent vasodilation in the penis (26, 38). The basis for which L-arginine supplementation can improve endothelial function and NO release remains controversial. It has been reported in animal models of penile vascular aging that constitutive NOS expression is reduced (8). Alternatively, it has been reported that NOS3 expression is upregulated with advanced age in the penis and in the 3

Page 4 of 38 peripheral vasculature, but that constitutive NOS activity is reduced such that for any given concentration of L-arginine, vascular production of NO is reduced (4, 5, 27). Arginase is a metalloenzyme that converts L-arginine to urea and L-ornithine in a number of cell types. In endothelial cells, L-arginine is used as a substrate by both NOS3 and arginase. Because both NOS and arginase use L-arginine as a common substrate, arginase may downregulate NO biosynthesis by competing with NOS for L-arginine. Thus, NO production may be linked to the regulation of arginase activity (39). Arginase exists in two isoforms, the hepatic type, arginase I and the extrahepatic type, arginase II. Although arginase I is traditionally thought to exist exclusively in the liver, it has recently been reported that significant expression of arginase exists in vascular endothelium and smooth muscle cells (5, 34, 52). Recently, both arginase isoforms have been shown to exist in human corpus cavernosum and inhibition of this enzyme resulted in facilitation of corporal smooth muscle relaxation (7, 22, 32). Additionally, arginase activity was shown to be increased in human diabetic corpus cavernosum (7). Since arginase activity is upregulated in diabetes, a condition known to be associated with reduced endothelial-derived NO, this may be interpreted to suggest that the diminished erectile response due to decreased NO production found in diabetic men may be due to a combination of increased expression of arginase and decreased amounts of NOS nerve fibers. This is a similar vascular phenomenon that occurs in the aged penile vascular bed. Therefore, we hypothesized that impaired endothelial-derived NO bioactivity or signaling by increased expression of arginase contributes to the reduced endothelium-dependent responses in the penis of aging mice via the scavenging of intracellular L-arginine. To test this hypothesis, we established that the aging B6/129 mouse ages in a manner consistent with aging rats and humans with a reduction in erectile responses to cavernous nerve stimulation and endothelium-dependent erectile responses. We then measured the arginase and NOS3 4

Page 5 of 38 concentrations and activity in corporal endothelial cells isolated from penes of B6/129 mice. Moreover, we established that the pharmacologic and genetic inhibition of arginase in the mouse penis abrogates the reduction of neurogenic- and endothelium-dependent erectile responses in the aged mouse. These results demonstrate that arginase signaling is substantially elevated in penile vascular endothelium of the aged mouse which results in significant reductions in NOS3 and ED. Materials and Methods Animal Model B6/129 hybrid mice were used in the present study. For initial studies on characterization of the aging-associated reduced erectile responses in the mouse strain, mice ranging from 4 months to 26 months were used (n=35). For grouped comparisons between young and aged mice, we studied aged animals (n=87), defined as 22 to 26 months old, the age at which there is approximately 45% mortality of the colony, and young adults (n=42) 4 to 8 months old and having reached sexual maturity. Isolation of Murine Penile Endothelial Cells Magnetic cell sorting of labeled endothelial cells was performed as previously described using a minimacs separation unit (Miltenyi Biotec, Bisley, Surrey, UK, catalog number 421-1) and confirmation of the vascular endothelial cell population was performed by flow cytometry (18). Briefly, penes were excised and rinsed with PBS to remove blood and diced in a Petri dish, using sterile crossed scalpels. The resulting pieces were washed twice in PBS, subjected to low speed centrifugation (21 g, 1 min) and incubated in a solution of collagenase (.5 mg/ml) 5

Page 6 of 38 for 1 h at 37 C in a humid incubator. Following incubation, the slurry was passed through a cell strainer to remove undigested blocks, washed twice in PBS supplemented with 2.5% FCS, and incubated for a further 1 min in trypsin EDTA solution to obtain single cell suspension. The suspension was further incubated for 3 min at 4 C with murine immunoglobulins to block Fc receptors followed by washing twice in cold PBS supplemented with 2.5% FCS. The suspension was then incubated for 3 min at 4 C with rat anti-mouse CD31 and subsequently washed twice in cold PBS supplemented with.5% FCS. The cells were counted and then incubated with PBS.5% FCS (2 µl/l, 2.5 17 cells), rat anti-mouse Ig (25 µl/l, 2.5 17cells)- and streptavidin-conjugated microbeads (25 µl/l, 2.5 1 7 cells) for 15 min at 4 C (total volume 25 µl). Columns were then loaded onto the separation unit (one column for every 1 2.5 1 7 cells) and washed with 5 µl PBS.5% FCS as per manufacturer s instructions followed by the loading of each column with 25 µl cell suspension. The magnetically labeled cells are retained in the column(s) while non-labeled cells pass through. The columns were then unloaded from the magnet and the magnetically retained cells were eluted with PBS.5% FCS. Labeled cells were incubated with MACS magnetic goat anti-mouse IgG (H1L) (Miltenyi Biotec) MicroBeads and streptavidin (Miltenyi Biotec, catalog number 481-1). MicroBeads were then separated using a high gradient magnetic separation column 1 (MS columns, catalog number 422-1, Miltenyi Biotec). Adeno-Associated Viral Vectors A plasmid encoding an antisense sequence for arginase I linked to a pcmv5 promoter was designed as described previously (5, 5). Adeno-associated viruses encoding -galactosidase (AAVgal) and antisense for arginase I (AAVAnti-Arginase), both driven by a cytomegalo virus (CMV) promoter, were prepared according to standard procedures. The Arginase I oligonucleotide antisense sequence used in the present study was: 5'- 6

Page 7 of 38 ATGTGGCGCATTCACAGTCAC-3' (5). Briefly, fragments of Anti-Arginase and - galactosidase, containing the open reading frame sequence were cleaved and subcloned at the corresponding sites in the plasmid. The insert was cut at EcoRI sites and cloned into corresponding sites in raav backbone containing the CMV promoter and the bovine growth hormone polyadenylation signal flanked by the AAV inverted terminal repeat sequences. Packaging, propagation, and purification of AAV viral particles were carried out by standard procedures (47). Localization and Measurement of Penile Endothelial Arginase Quantitative PCR Real-time quantitative polymerase chain reaction (PCR) was used to determine relative expression of arginase I mrna in whole mouse penes and penile endothelial cells isolated from young and aged mouse penes. Total RNA was isolated using RNeasy kit (Qiagen), DNase treated (Ambion), and reverse transcribed using Superscript II (Life Technologies). PCR reactions were performed in a GeneAmp 79 sequence detection system (Applied Biosystems) using SYBR green PCR master mix as previously described (5). To amplify specific gene products, the following intron-spanning primers were used: arginase I are as follows: sense: CAAGCTGGGAATTGGCAAAG; antisense: GGTCCAGTCCATCAAC-ATCAAA, NOS3 sense: 5'-AAGACAAGGCAGCGGTGGAA-3', antisense: 5'-GCAGGGGACAGGAAATAGTT-3'. Mouse GAPDH was coamplified as an internal control, using primer sequences: GAPDH sense: 5'- CATCACCATCTTCCAGGAGCG-3', antisense: 5'-GAGGGGCCATCCACAGTCTTC-3'. Standard curves were performed using GAPDH to ensure appropriate RNA loading per sample. Western Blot Analysis 7

Page 8 of 38 Penile vascular endothelial cells were pelleted by centrifugation, and then resuspended in a hypotonic buffer containing 5mM Tris-HCl (ph 7.4),.1 mm EGTA, 2mM -mercaptoethanol, and protease inhibitors (1 µg/ml leupeptin, 1 µg/ml pepstatin, and 1 mm phenymethysulfonyl fluoride). After centrifugation twice at 15, g and 4 C for 2 min, protein concentration was determined by using a Bio-Rad DC protein assay (Bio-Rad, Hercules, CA), and the samples were stored at -7 C until use. For western blot analysis, the supernatant was mixed with an equal volume of 2% SDS/1% -mercaptoethanol and fractionated using 8% SDS/PAGE (7 µg/lane). Proteins were then transferred to a nitrocellulose membrane (Hybond-ECL, Amersham Life Sciences, Ghent, Belgium) by semi-dry electroblotting for 1 hr. The membranes were blocked 1 hr at room temperature with blotto-tween (5% nonfat dry milk,.1% Tween-2) and incubated with a primary mouse monoclonal arginase I IgG antibody (1:5) and primary rabbit monoclonal NOS3 IgG (1:25; Transduction Laboratories, Lexington, KY). Bound antibody was detected with labeled anti-mouse IgG secondary antibody (1:2,; Santa Cruz Biotechnology, Santa Cruz, CA) and visualized using enhanced chemiluminescence. Measurement of Penile Endothelial Arginase and NOS Enzyme Activity Endothelial cell lysate and whole mice penes were combined 1:4 (wt:vol) with ice-cold buffer composed of 2 mm HEPES (ph 7.4) and.25 M sucrose on ice in the presence of protease inhibitors (PMSF, leupeptin, aprotinin). Arginase activity was assayed as previously described (5, 7). The selective arginase inhibitor, 2(S)-amino-6-boronohexanoic acid (ABH; Kd=.11 µm) in a dose of 5 µm was added to the endothelial cell and penile supernatant to determine what effect pharmacological inhibition of arginase has on arginase activity in our samples in vitro. Enzyme activity was expressed as urea production in pmol. min -1. mg protein -1. 8

Page 9 of 38 Endothelial NOS activity was measured in vascular endothelial cells and whole mice penes from young and aged mice by L-[ 14 C]arginine to L-[ 14 C]citrulline conversion assay in the presence or absence of ABH (5 µm), as previously described (7, 18). Enzyme activity was expressed as L-citrulline production in pmol. min -1. mg protein -1. Penile Tissue cgmp Levels Quantitative assays for cgmp from endothelial cell lysates and whole mouse penes were performed using a commercial enzyme immunoassay kit (Amersham Pharmacia Biotech, Piscataway, NJ) as previously described (9, 19). Penile cgmp levels were measured in penes from young, aged, and aged mice one month after transfection with AAVCMVgal and AAVCMVAnti-Arginase into the corpus cavernosum. Endothelial cell lysate cgmp levels are expressed as percent change of cgmp from baseline while whole mouse penes cgmp concentrations are expressed as femtomole per mg of protein. In Vivo Gene Transfer to the Mouse Corpus Cavernosum Mice were anesthetized with isoflurane and intubated in the supine position on a thermoregulated surgical table as described above. Using sterile technique, the penis was exposed and the corpus cavernosum was approached by blunt dissection. 1 µl of vehicle (.5% glycerol in PBS), AAVgal (1x1 12 parts/ml), or AAVAnti-arginase (1x1 12 parts/ml) was injected into the corpus cavernosum with a 3-gauge needle attached to a microliter syringe, as previously described (6, 17, 19). Immediately before instillation, blood drainage via the dorsal veins was halted by circumferential compression at the base of the penis with an elastic band. Compression was released approximately five minutes after injection of 1 Pl of the vehicle/virus. The needle and the endotracheal tube were removed, and the animals were 9

Page 1 of 38 allowed to recover. Mice were studied thirty days after injection of virus vehicle, AAVgal, or AAVAnti-arginase. The mice did not show any overt signs of systemic (fever, dyspnea, tachycardia, unkempt appearance) or local (purulent discharge, erythema, edema) infection when observed any day after transfection. L-Arginine Supplementation to the Aged Mouse Aged B6/129 hybrid mice were fed a high L-arginine diet (75 mg/kg/d) for 2 weeks by mixing L- arginine into semi-soft rodent chow (Bacon flavored transgenic dough, Bioserv, NJ; 4-6 g/day) that also provided full all daily required nutrition. Age-matched control mice were fed a normal chow diet (Bioserv) without supplemental L-arginine. At the end of the 2 week normal and high L-arginine diets, the mice were subject to electrical field stimulation of the cavernous nerve (2 and 4 volts) and direct intracavernous injection of the endothelium-dependent vasodilator, acetylcholine (ACh; 13 and 44 nmol). Intracavernosal and systemic arterial pressures were recorded as previously described (45). In Vivo Erectile Responses in the Mouse In vivo erectile function to cavernous nerve stimulation (CNS) and intracavernous injection of vasoactive agents was studied in the young and aged anesthetized mouse. Induction of anesthesia was achieved by placing the animal in a jar containing gauze soaked with isoflurane. The mice were then intubated and placed on a thermoregulated surgical table. The animals were ventilated with 95% O 2 /5% CO 2 and 2% isoflurane using a custom-designed, constant-flow mouse ventilator with tidal volume set to 6.7 µl/g at 14 breaths/min. A carotid artery was cannulated for measurement of mean systemic arterial pressure (MAP), which was measured continuously with a Viggo-Spectramed transducer (Viggo Spectramed, Oxnard, CA) attached to 1

Page 11 of 38 a data acquisition system (Biopac Systems, Santa Barbara, CA). Heart rate was determined from the systolic pressure pulses with a tachometer (Biopac). The shaft of the penis was freed of skin and fascia, and by removing part of the overlying ischiocavernous muscle exposure of the right crus was performed. A 27-gauge needle filled with 25 U/ml of heparin and connected to PE-5 tubing was inserted into the right crura and connected to a pressure transducer to permit continuous measurement of intracavernosal pressure (ICP). The bladder and prostate were exposed through a midline abdominal incision. The right major pelvic ganglion and cavernous nerve were identified posterolateral to the prostate on one side, and an electrical stimulator with a stainless steel bipolar hook was placed around the cavernosal nerve. MAP and ICP were measured with a pressure transducer connected to a data acquisition system (Biopac) for continuous measurement of MAP and ICP pressures. The cavernous nerve was stimulated with a square pulse stimulator (Grass Instruments, Quincy, MA). Each mouse underwent CNS at a frequency of 15 Hz and pulse width of 3 seconds. The application of 1-8 volts was used in the current protocol to achieve a significant and consistent erectile response. The duration of stimulation was approximately 1 minute with rest periods of 2 to 3 mins between subsequent stimulations. This procedure has been previously described (19, 45). Johns Hopkins University Animal Care and Use Committees have approved all procedures used in the present study. In experiments in which ACh (Sigma), an endothelium-dependent vasodilator, and calcitonin gene related peptide (CGRP; Phoenix Pharmaceuticals, Inc.), an endotheliumindependent vasodilator, was utilized, intracavernosal administration via a 3-gauge needle inserted into the left corpus cavernosum was performed, as described previously. In all experiments injections of agonists were made when intracavernosal pressure was at baseline value. The erectile effects of a single injection of ACh (13 and 44 nmol) and CGRP (.1 and.3 nmol) on intracavernosal pressure were monitored until intracavernosal pressure returned to pre-injection level. The next injection was made after a 1-15-min period from the end of the 11

Page 12 of 38 receding response to ensure a stable baseline. In experiments in which ABH was injected, all erectile responses were evaluated twenty minutes after injection of ABH. There was only a transient increase in intracavernosal pressure after injection of ABH (approximately 8-1 minutes) which returned to baseline before cavernous nerve stimulation or intracavernous injection of pharmacological agents. These methods have been previously described (19). Drugs 2(S)-amino-6-boronohexanoic acid (ABH) was synthesized as previously described (5). Preliminary experiments were performed to determine the final concentrations and doses used in the present study based on selective inhibition of arginase in the mouse. These concentrations of ABH inhibited recombinant arginase I, but failed to inhibit recombinant NOS3 activity in an assay in vitro. Acetylcholine bromide (ACh; Sigma Chemical Company) and CGRP (Phoenix Pharmaceuticals, Inc.) were dissolved in.9% NaCl. The solvents for the agonists used in this study did not alter vascular intracavernosal pressures or responses to the agonists. The stock solutions were stored in a freezer in 1 ml opaque tubes and working solutions were prepared daily and kept on crushed ice during the course of the experiment. Statistical Analysis Data are presented as mean ± SEM. Comparisons between baseline variables in ABH and vehicle-treated penes were performed using paired or unpaired t tests, as appropriate. Comparisons between groups were made using ANOVA analysis with repeated measures and Neumann-Kuels post hoc test for multiple group comparisons. Statistical calculations were performed using Systat software. 12

Page 13 of 38 Results Influence of Age on Erectile Responses in the Mouse The influence of age on erectile responses to CNS and to intracavernosal injection of the endothelium-dependent vasodilator acetylcholine (ACh) and the endothelium-independent vasodilator calcitonin gene-related peptide (CGRP) were studied in B6/129 mice and these data are summarized in Figure 1. There was an age-related decrease in erectile responses to CNS and intracavernosal injections of ACh with a strong positive correlation (r 2 =.8376 for CNS and r 2 =.6832 for ACh). In contrast, there was no significant correlation between age and responses to intracavernosal injection of the endothelium-independent vasodilator, CGRP (Fig 1). Although the maximal response to ICP was not different in young and aged mice in response to CGRP, the area under the curve (AUC) was lower in the aged group (P<.5; data not shown). The influence of L-arginine supplementation on erectile responses to CNS and ACh injection was studied in this aging model in order to validate this animal model. Dietary supplementation of L-arginine via a high L-arginine chow for 2 weeks resulted in increased erectile responses to CNS and intracavernosal injection of ACh when compared to aged mice fed regular chow (Fig 2). L-arginine supplementation did not alter erectile responses to CNS or ACh injection in young mice (data not shown). Arginase Expression and Activity: Biochemical Assessment Mouse penile endothelial cells were isolated as described (18). The purity of the penile endothelial cell population was confirmed by flow cytometry (data not shown) as well as by western blot analysis confirming the presence of caveolin 1, PECAM, and the absence of - smooth muscle actin (Fig 3A). Cell lysate from mouse pulmonary vascular endothelial cells from primary culture served as positive control for endothelial cells. In addition, the expression of arginase I and NOS3 was evaluated in this endothelial cell lysate. Arginase I mrna 13

Page 14 of 38 expression and protein abundance was approximately 6% higher in penile vascular endothelial cells from aged mice when compared to young adult mice and these data are summarized in Figure 3B. Moreover, the mrna and protein concentrations of NOS3 were significantly higher in penile endothelial cells from aged mice when compared to young mice (Fig 3B). The higher concentration of arginase protein was associated with higher enzyme activity in aged mouse penile endothelial cells when compared to endothelial cells isolated from young mouse penes (Fig 4A). In some cases, additional immunostaining was seen below the 38 kd location of arginase I. Pretreatment of the membrane with a blocking peptide for the monoclonal arginase I antibody resulted in a loss of all staining suggesting that the additional immunostaining is likely secondary to degradation products of arginase I (data not shown). In vitro administration of the arginase inhibitor, ABH (5 µm), reduced arginase activity in both young and aged mouse endothelial cells as evidenced by a reduction in urea formation from L- arginine (Fig 4A). In contrast to the arginase I expression and activity, the increase in NOS3 protein and mrna was associated with a decreased conversion of L-arginine to L-citrulline, a direct measure of endothelial NOS activity, in endothelial cells isolated from young and aged penes (Fig 4B). In vitro administration of the arginase inhibitor, ABH (5 µm), increased NOS3 activity in aged mouse endothelial cells as evidenced by an increase in L-citrulline formation from L-arginine (Fig 4B). In contrast to the effect of ABH on NOS3 activity in the aged penile endothelial cells, ABH did not alter L-arginine to L-citrulline conversion in penile endothelial cells from young mice (Fig 4B). The effect of in vitro administration of ABH (1-3 µm) on cgmp accumulation in mouse penile homogenate is shown in Figure 4C. Baseline cgmp concentrations in penes from aged mice were significantly lower (952 ± 68 fmol/mg protein) when compared to young mice (28 ± 174 ; P<.5). In both young and aged mouse penes, ABH produced significant concentrationdependent increases in cgmp concentrations (Fig 4C). The magnitude of the increase in cgmp 14

Page 15 of 38 in response to ABH was significantly higher in the penes from aged mice when compared to young mice (P<.5; Fig 4C). In Vivo Physiological Erectile Response to ABH in the Mouse The effect of the arginase inhibitor ABH (18.2 nmol/kg ic) on baseline erectile function was evaluated in young and aged mice and these data are summarized in Figure 5. Intracavernosal injection of ABH resulted in a small, but significant, increase in intracavernosal pressure in aged mice (3.9 ±.2 mmhg) when compared to vehicle in the aged mice (.2 ±.1 mmhg; P<.5). Intracavernosal pressure was not altered in response to intracavernosal injection of ABH in the young mice (data not shown). In aged mice, intracavernosal pressure returned to baseline pressures over the course of 8-1 min and erectile responses were evaluated twenty minutes after injection of ABH. The transient increase in intracavernosal pressure after administration of ABH did not alter the ratio of ICP/MAP (P>.5). In aged mice treated with ABH, erectile responses to CNS were significantly increased at all voltages studied (Fig 5A left). In contrast to the increase in erectile responses to CNS observed in the presence of ABH in the aged mice, responses to CNS were not altered in young mice when compared before and after injection of ABH (18.2 nmol/kg ic; Fig 5A right). In addition to intracavernosal pressures, MAP were measured in young (12 ± 4.7 mmhg) and aged mice (112 ± 4.6 mmhg; P=.47), and there was a statistical increase in baseline MAP in aged mice. While there was a statistically significant difference in MAP in young and aged mice, the ICP/MAP ratio was not statistically different at baseline (P=.68). Moreover, intracavernosal injection of ABH resulted in a nonsignificant reduction in MAP in young (11.9 ± 5.1; P>.5) or aged mice (18 ± 4.9, P>.5). Erectile responses to the endothelium-dependent vasodilator, ACh, and the endothelium-independent vasodilator, CGRP, were evaluated in aged and young mice before and after injection of the arginase inhibitor ABH (18.2 nmol/kg ic; Fig 5B and 5C). Erectile 15

Page 16 of 38 responses to ACh and CGRP were determined approximately twenty minutes after intracavernous injection of ABH. Responses to ACh were enhanced in aged mice (Fig 5B left) treated with ABH whereas there was no difference in erectile responses to ACh in young mice (Fig 5B right). In contrast to the enhanced responses after treatment with ABH, erectile responses to the endothelium-independent vasodilator, CGRP, were unaltered by ABH in the young or aged mice (Fig 5C). In Vivo Gene Transfer of Anti-Arginase on Erectile Function in the Mouse To further evaluate the role of arginase I on age-associated erectile function, the influence of gene transfer of an antisense sequence for arginase I was studied, and these data are summarized in Figure 6. One month after injection of an adeno-associated viral vector (AAV) encoding an antisense sequence for arginase I (AAVAnti-Arginase), mrna and protein expression of arginase I in penile homogenates was significantly lower when compared to vehicle-treated aged mice (Fig 6A). Arginase I protein and mrna expression was significantly higher in aged mice penes treated with vehicle or AAVgal (Fig. 6A). Treatment of aged mice with AAVgal did not significantly alter arginase I protein expression (Fig. 6A; right) or mrna expression and arginase activity when compared to aged mice treated with vehicle (data not shown). This increase in arginase I was associated with a significant decrease in constitutive NOS activity, the conversion of L-arginine to L-citrulline in the presence of calcium, and cgmp levels in aged mice penes treated with vehicle (Fig. 6B and 6C) or AAVgal. Constitutive NOS activity was significantly increased in the penes of aged mice after AAVAnti-arginase transfection when compared to vehicle-treated (Fig 6B) and AAVgal-transfected aged mice (data not shown). cgmp concentrations in penes of aged mice transfected with AAVAnti- 16

Page 17 of 38 arginase were significantly higher when compared to vehicle-treated (Fig 6C) or AAVgaltransfected aged mice (data not shown). Erectile responses to CNS and vasoactive agents were compared in young and aged mice one month after transfection with AAVgal and AAVAnti-arginase. One month after transfection with AAVAnti-arginase, erectile responses to CNS were significantly improved when compared to mice treated with AAVgal (a gene product that has no measurable effect on erectile function; Fig 7A). In results that paralleled the acute studies with ABH, responses to ACh were improved in AAVAnti-arginase treated mice whereas responses to CGRP (data not shown) were not different from AAVgal treated mice (Fig 7B and 7C). Discussion The results of the present study demonstrate for the first time a mouse model of age-related ED and that arginase is upregulated in endothelial cells isolated from the penis of aged mice where it directly influences and decreases endothelial NOS activity. Moreover, these data are the first to identify a physiologic role for arginase in modulating erectile function in vivo by reducing endothelial NOS activity, thus decreasing penile cgmp levels. Additionally, this study demonstrates that the pharmacologic inhibitor of arginase, ABH and adeno-associated viral gene transfer of Anti-Arginase to the aged mouse penis decreases arginase I protein and mrna and restores endothelial and erectile function in vivo as a direct result of an increase in penile constitutive NOS activity and elevated penile cgmp levels. Furthermore, these data provide pharmacologic and gene transfer evidence for a role of the arginase system in mediating at least a portion of age-related endothelial and erectile dysfunction in the penile vascular bed. These findings implicate that arginase may influence erectile function through attenuation of endothelial-derived NO in the penis. 17

Page 18 of 38 The decline in erectile function that occurs with aging can be attributed to decreased NOS expression/activity, phosphorylation, and bio-availability of NO in the penile vasculature (8, 9, 4, 41, 46, 49). Overexpression of enos by adenoviral gene transfer to the penis of aged and diabetic rats restores endothelial and erectile function, suggesting that decreased production or bioavailability of NO from the endothelium plays a significant role in mediating penile vascular dysfunction (6, 1-12, 17). In the present study, we have established for a mouse model of age-associated ED. The B6/129 strain of mouse has significant impairments in endothelial and erectile function as determined by CNS and endothelium-dependent vasodilator responses to ACh with advancing age. These in vivo physiological parameters parallel a significant reduction in penile endothelial NOS activity and cgmp concentrations. These defining characteristics support the use of this aging mouse model for the study of ageassociated penile vascular dysfunction. In the present study we found a significant increase in enos protein expression in isolated penile endothelial cells from aged mice, which is consistent with previous reports in aged rat penes (4). Bakircioglu and colleagues have shown that corporal enos protein expression is significantly upregulated in the aging rat penis due to a reduction in caveolin-1 protein in the penis (4). Caveolin-1 tightly regulates enos in the vascular endothelium. Therefore, it is possible that the upregulation of enos protein expression in mouse penile endothelial cells is a result of reduced caveolin-1 expression. Although, enos protein expression is upregulated in the aged mouse endothelium, there is significant decline in enos activity, suggesting that the increased protein is not biologically active. Data to support this hypothesis is draw from the effect of adeno-associated viral gene transfer of Anti-Arginase improving enos activity after inhibition of arginase and thus restoring endothelial vascular responses in the aging mouse penis. This biochemical process occurs by increasing the amounts of enos enzyme available for synthesis of NO thus allowing more enos to be available for L-arginine. Additionally, adenoviral gene transfer of enos to the aging rat penis 18

Page 19 of 38 improves erectile function via increases in corporal enos activity and improved cgmp levels in the penis by increasing endothelial-derived NO synthesis (12). Endothelial dysfunction has been shown to be a major mechanism that contributes to ED (31, 42). Activation of NOS3 plays an important role in the relaxation of corporal smooth muscle to endogenous agonists such as Ach and shear stress (14, 28, 4, 41). Because impaired endothelium-dependent corporal smooth muscle relaxation in vitro and erectile function in vivo is restored by supplementation with L-arginine, this suggests that a reduction of L-arginine availability for NOS3 may be involved (26, 38, 44). Therefore, we postulated that arginase, if expressed in endothelial cells of the corpus cavernosum, might play a significant role in the regulation of NO-mediated vasodilation in vivo by reducing L-arginine availability to NOS3 in the aging penile vascular bed. Arginase is responsible for the hydrolysis of L-arginine to ornithine and urea, thus playing a critical role in urea synthesis from ammonia in the urea cycle. However, L-arginine is a common substrate for both arginase and NOS, and thus arginase plays a significant role in regulating NOS and NO synthesis by modulating L-arginine bioavailability. Recently, both arginase isoforms have been identified in rabbit and human corpus cavernosum and the arginase inhibitor ABH enhanced the non-adrenergic, non-cholinergic (NANC) corporal smooth muscle relaxation suggesting that arginase inhibition may enhance the substrate pool of L-arginine that is available for NOS (7, 22, 32). While inhibition of arginase activity has been shown to enhance NO production by NOS2, in macrophages, there is a relative paucity of information about the role of arginase in regulating the function of NOS3 in the aging penile vascular bed (2, 44, 48). Recently, a role for arginase in the aging rabbit penis in vitro has been elucidated (44). In this study, endothelium-dependent corporal smooth muscle relaxation in vitro was improved after pre-treatment with a selective arginase inhibitor (44). However, in vivo erectile responses were not determined. Data from the present study show arginase I mrna and protein are constitutively expressed in the mouse penile endothelium with an age-associated increase in arginase I in the 19

Page 2 of 38 B6/129 mouse penis. Arginase activity in penile endothelial homogenates was present and could be suppressed with the arginase inhibitor ABH. These data suggest that arginase expression is not limited to the liver and macrophages and imply that extrahepatic expression of arginase exists in the penile endothelium. Moreover, enzyme activity studies in the penile endothelial cells suggest that endothelial arginase is active as evidenced by the ability to convert L-arginine to urea and L-ornithine. While there is constitutive expression of arginase in young mouse penile endothelial cells, and the arginase inhibitor ABH in vitro can inhibit this arginase activity, arginase inhibition had no significant effect on erectile responses to CNS or intracavernous ACh in the young mouse. Moreover, the inhibition of arginase by ABH did not alter the calcium-dependent conversion of L-arginine to L-citrulline in the young penile endothelial homogenates. The reason for the lack of effect of ABH on the calcium-dependent conversion of L-arginine to L-citrulline in young mouse endothelial cells is uncertain. It is possible that there is a specific concentration of arginase that is required to have an effect on the ability of NOS to convert L-arginine to L-citrulline in the presence of calcium and that the concentration of arginase is not sufficient in the young mouse endothelial cells. However, in aged mouse endothelial cells the overexpression of arginase may reach this threshold to influence NOS3 and constitutive NOS activity. These findings are consistent to what was observed in isolated corporal strips from young and aged rabbits (44). There may be a constitutive role for arginase in the young mouse penis other than influencing corporal smooth muscle reactivity, such as the control of corporal smooth muscle proliferation or other signaling pathways yet undetermined (3). The concentration of ABH required to selectively inhibit arginase activity was lower than in our previous report in the diabetic human corpus cavernosum (7). The reasons for the differences are uncertain, but may reflect the differences in species or in the magnitude of the increased in arginase expression/activity in the aging process when compared to diabetes. The doses chosen for the present study were based on preliminary experiments to determine a concentration range of ABH that would effectively inhibit 2

Page 21 of 38 arginase but not NOS3 in experiments in which NOS3 protein was purified by immunoprecipitation. In the present study arginase I was the molecular target investigated in young and aged mice penile vasculature in vitro and in vivo. All biochemical and physiological analyses where determined after pharmacological inhibition with ABH, a potent non-selective arginase inhibitor with higher affinity for arginase II (3). It is possible that some of the physiological effects observed after ABH therapy may be in part due to inhibition of both isoforms, however, when arginase I was selectively inhibited using the adeno-associated virus encoding for Anti-Arginase I, similar physiological effects in vivo and significant reductions in arginase activity and improvements in enos activity were observed suggesting that selectively inhibiting one isoform of arginase (arginase I) can improve neurogenic- and endothelium-dependent erectile responses via improvements in endothelial-derived NO/cGMP signaling. The role of arginase II in the pathophysiology of age-associated ED is uncertain and warrants further investigation. To further establish a link between arginase, NOS and NO/cGMP interactions, we measured the effect of arginase inhibition on cyclic nucleotide (cgmp) production in the mouse penis. Inhibition of arginase resulted in increases in cgmp concentrations in penile homogenates from aged mice, supporting the hypothesis that inhibiting arginase results in increased NOS activity, enhanced NO production and elevation of cgmp concentrations. The observation that ABH increased cgmp concentrations in the young mouse penile homogenates suggest that the assay for studying cgmp may be more sensitive than the L-arginine to L- citrulline assay in measuring small changes in arginase activity/concentrations. Alternatively, it is possible that ABH, by an unknown mechanism, increases cgmp in a direct manner. Regardless of the mechanism by which ABH increases cgmp in young mouse penes, there was no measurable change in physiologic erectile responses. 21

Page 22 of 38 It has previously been shown that chronic administration of L-Arginine enhances erectile function in models of age-associated ED (38, 44). Additionally, oral administration of L-arginine in high doses caused significant subjective improvement in sexual function in men with organic ED (21). The mechanism by which L-arginine enhances NO signaling despite apparently adequate intracellular levels of this substrate remains controversial but has been widely documented in the setting of endothelial reactivity. It has been suggested that although total intracellular L-arginine content is relatively high, the pool available to NOS may be reduced or compartmentalized within the cell and sequestered away from NOS in that it would limit NOS s function (2, 34). Moreover, intracellular arginine can be methylated thus rendering it useless to NOS. When taken together, these data suggest that intracellular L-arginine may be tightly controlled by a number of factors. These findings may suggest that by reducing the available supply of free L-arginine in the endothelial cells of the penis, NOS activity can be reduced. By inhibiting arginase, we hypothesize that the concentration of L-arginine that is available for use by NOS3 is increased, resulting in increased NO release, and subsequent increases in penile cgmp levels. The specific contribution of the NO/cGMP signaling pathway to erectile function is further supported by our findings that the arginase inhibitor ABH and adeno-associated viral gene transfer of Anti-Arginase to the penis, like L-arginine, improves erectile function to CNS and the endothelium-dependent vasodilator ACh. In the current study, we did not investigate other potential alterations in L-arginine availability that could lead to decrease NO production. Among the possibilities are changes in L- arginine transport and/or an increase in levels of other endogenous NOS inhibitors with aging, which can be overcome with increased L-arginine (35, 36). Yet, regardless of any other ageassociated alterations in precursor availability/metabolism or downstream NO inactivation, arginase inhibition, like L-arginine administration, appears to overcome these secondary mechanisms. Thus, the response to arginase inhibition may have important therapeutic 22

Page 23 of 38 implications with regard to erectile function in aging or other disease states (diabetic- and vasculogenic-ed). It has been suggested that the mechanisms associated with the impairment of endothelium-dependent erectile responses with advancing age may be multifactorial. Thus, the inability of arginase inhibition either pharmacologically or genetically to fully restores normal erectile function could be related to additional mechanisms besides increased arginase activity. Potential mechanisms that may be intimately related to impaired endothelial and erectile function postulated are: 1) increased levels of reactive oxygen species, in particular superoxide anion, in the aging penis could directly inactivate NO, 2) downregulation of NOS1 in the cavernous nerves innervating the penis, and 3) diminished levels of NOS3 co-factors and downregulation of enos activity (8, 9, 15, 46, 51). Although these mechanisms could contribute to the impairment of endothelial-derived NO-mediated dilation in the penis of the aged mouse, the predominant mechanism in the present model appears to be related to the upregulation of arginase. Previous reports in the literature vary on the effect of age on penile constitutive NOS activity (8, 24, 46). In the present study, endothelial cells isolated from aging mouse penes have a significant decline in enos activity. Others have shown that increasing age reduces constitutive NOS activity in the penis (46). However, this is not uniformly constant depending on the age the experimental animals (8, 24, 46). Results of the present study demonstrate that isolated endothelial cells from aging mice have reduced enos activity, while this differences from previous reports, the disparity may be a result of species difference and/or cell type used (pure endothelial cells) vs. whole penes (24). In conclusion, the results of the present study provide the first evidence, at both the molecular and functional level, for a biological role of arginase in regulating erectile function in the aging penile vascular bed. With advancing age, expression and function of arginase increases in the penile vasculature and contributes to endothelial and erectile dysfunction. 23

Page 24 of 38 Penile endothelial cells isolated from the aged mouse penis overexpressed arginase and as a result decreased NOS3 activity and impaired vascular function. Inhibition of arginase via adeno-associated viral gene transfer of Anti-Arginase in the aged mouse penis enhanced penile endothelial NOS activity and cgmp levels thus restoring endothelial-derived NO vasodilation and erectile function. Taken together, our data demonstrate that arginase is an endogenous competitor of NOS for their common substrate L-arginine in the penis and consequently, an aged-related up-regulation of arginase may compromise NO-mediated vasodilation in the penis via a reduction in substrate availability. Therefore, arginase may represent a novel molecular therapeutic target for the treatment of age-associated ED (vasculogenic ED). Acknowledgments This work was supported in part by the Bernard A. and Rebecca S. Bernard Foundation (HCC), a Young Investigator Award from the International Society of Impotence Research and Pfizer, Inc. (TJB, HCC), National Kidney Foundation of Maryland (TJB), The Shin Chun-Wang Award from the American Physiological Society (HCC), and NIH grants HL494 (HCC) and NIDDK64679 (ALB). References 1. Andersson KE. Erectile Physiological and Pathophysiological Pathways Involved in Erectile Dysfunction. J Urol 17: S6-S14, 23. 2. Arnal JF, Munzel T, Venema RC, James NL, Bai CL, Mitch WE and Harrison DG. Interactions between L-arginine and L-glutamine change endothelial NO production. An effect independent of NO synthase substrate availability. J Clin Invest 95: 2565-2572, 1995. 3. Baggio R, Emig FA, Christianson DW, Ash DE, Chakder S and Rattan S. Biochemical and functional profile of a newly developed potent and isozyme-selective arginase inhibitor. J Pharmacol Exp Ther 29: 149-1416, 1999. 4. Bakircioglu ME, Sievert KD, Nunes L, Lau A, Lin CS and Lue TF. Decreased trabecular smooth muscle and caveolin-1 expression in the penile tissue of aged rats. J Urol 166: 734-738, 21. 24

Page 25 of 38 5. Berkowitz DE, White R, Li D, Minhas KM, Cernetich A, Kim S, Burke S, Shoukas AA, Nyhan D, Champion HC and Hare JM. Arginase reciprocally regulates nitric oxide synthase activity and contributes to endothelial dysfunction in aging blood vessels. Circulation 18: 2-26, 23. 6. Bivalacqua TJ, Champion HC, Mehta YS, Abdel-Mageed AB, Sikka SC, Ignarro LJ, Kadowitz PJ and Hellstrom WJ. Adenoviral gene transfer of endothelial nitric oxide synthase (enos) to the penis improves age-related erectile dysfunction in the rat. Int J Impot Res 12 Suppl 3: S8-17., 2. 7. 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 283: 923-927, 21. 8. Bivalacqua TJ, Usta MF, Champion HC, Kadowitz PJ and Hellstrom WJ. Endothelial dysfunction in erectile dysfunction: role of the endothelium in erectile physiology and disease. J Androl 24 Suppl: S17-37, 23a. 9. Bivalacqua TJ, Armstrong JS, Biggerstaff J, Abdel-Mageed AB, Kadowitz PJ, Hellstrom WJ and Champion HC. Gene transfer of extracellular SOD to the penis reduces O 2-. and improves erectile function in aged rats. Am J Physiol Heart Circ Physiol 284: H148-H1421, 23b. 1. Bivalacqua TJ, Usta MF, Champion HC, Adams D, Namara DB, Abdel-Mageed AB, Kadowitz PJ and Hellstrom WJ. Gene transfer of endothelial nitric oxide synthase partially restores nitric oxide synthesis and erectile function in streptozotocin diabetic rats. J Urol 169: 1911-1917, 23c. 11. Bivalacqua TJ, Deng W, Champion HC, Hellstrom WJ and Kadowitz PJ. Gene therapy techniques for the delivery of endothelial nitric oxide synthase to the corpora cavernosa for erectile dysfunction. Methods Mol Biol 279: 173-185, 24. 12. Bivalacqua TJ, Musicki B, Usta MF, Champion HC, Kadowitz PJ, Burnett AL and Hellstrom WJ. Endothelial nitric oxide synthase gene therapy for erectile dysfunction. Curr Pharm Des 11: 459-467, 25. 13. Burnett AL, Lowenstein CJ, Bredt DS, Chang TS and Snyder SH. Nitric oxide: a physiologic mediator of penile erection. Science 257: 41-43, 1992. 14. Burnett AL, Chang AG, Crone JK, Huang PL and Sezen SE. Noncholinergic penile erection in mice lacking the gene for endothelial nitric oxide synthase. J Androl 23: 92-97, 22. 15. Carrier S, Nagaraju P, Morgan DM, Baba K, Nunes L and Lue TF. Age decreases nitric oxide synthase-containing nerve fibers in the rat penis. J Urol 157: 188-192, 1997. 16. Cartledge JJ, Eardley I and Morrison JF. Nitric oxide-mediated corpus cavernosal smooth muscle relaxation is impaired in ageing and diabetes. BJU Int 87: 394-41., 21. 25

Page 26 of 38 17. Champion HC, Bivalacqua TJ, Hyman AL, Ignarro LJ, Hellstrom WJ and Kadowitz PJ. Gene transfer of endothelial nitric oxide synthase to the penis augments erectile responses in the aged rat. Proc Natl Acad Sci U S A 96: 11648-11652., 1999. 18. Champion HC, Georgakopoulos D, Takimoto E, Isoda T, Wang Y and Kass DA. Modulation of in vivo cardiac function by myocyte-specific nitric oxide synthase-3. Circ Res 94: 657-663, 24. 19. Champion HC, Bivalacqua TJ, Takimoto E, Kass DA and Burnett AL. Phosphodiesterase-5A dysregulation in penile erectile tissue is a mechanism of priapism. Proc Natl Acad Sci U S A 12: 1661-1666, 25. 2. Chang CI, Liao JC and Kuo L. Macrophage arginase promotes tumor cell growth and suppresses nitric oxide-mediated tumor cytotoxicity. Cancer Res 61: 11-116, 21. 21. Chen J, Wollman Y, Chernichovsky T, Iaina A, Sofer M and Matzkin H. Effect of oral administration of high-dose nitric oxide donor L-arginine in men with organic erectile dysfunction: results of a double-blind, randomized, placebo-controlled study. BJU Int 83: 269-273, 1999. 22. Cox JD, Kim NN, Traish AM and Christianson DW. Arginase-boronic acid complex highlights a physiological role in erectile function. Nat Struct Biol 6: 143-147, 1999. 23. Feldman HA, Goldstein I, Hatzichristou DG, Krane RJ and McKinlay JB. Impotence and its medical and psychosocial correlates: results of the Massachusetts Male Aging Study. J Urol 151: 54-61., 1994. 24. Garban H, Vernet D, Freedman A, Rajfer J and Gonzalez-Cadavid N. Effect of aging on nitric oxide-mediated penile erection in rats. Am J Physiol 268: H467-475, 1995. 25. Gonzalez-Cadavid NF and Rajfer J. Molecular pathophysiology and gene therapy of aging-related erectile dysfunction. Exp Gerontol 39: 175-1712, 24. 26. Gur S, Ozturk B and Karahan ST. Impaired endothelium-dependent and neurogenic relaxation of corpus cavernosum from diabetic rats: improvement with L-arginine. Urol Res 28: 14-19, 2. 27. Haas CA, Seftel AD, Razmjouei K, Ganz MB, Hampel N and Ferguson K. Erectile dysfunction in aging: upregulation of endothelial nitric oxide synthase. Urology 51: 516-522, 1998. 28. Hurt KJ, Musicki B, Palese MA, Crone JK, Becker RE, Moriarity JL, Snyder SH and Burnett AL. Akt-dependent phosphorylation of endothelial nitric-oxide synthase mediates penile erection. Proc Natl Acad Sci U S A 99: 461-466, 22. 29. 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 17: 843-85, 199. 26

Page 27 of 38 3. Ignarro LJ, Buga GM, Wei LH, Bauer PM, Wu G and del Soldato P. Role of the arginine-nitric oxide pathway in the regulation of vascular smooth muscle cell proliferation. Proc Natl Acad Sci U S A 98: 422-428, 21. 31. Jackson G, Rosen RC, Kloner RA and Kostis JB. The second Princeton consensus on sexual dysfunction and cardiac risk: new guidelines for sexual medicine. J Sex Med 3: 28-36; discussion 36, 26. 32. Kim NN, Cox JD, Baggio RF, Emig FA, Mistry SK, Harper SL, Speicher DW, Morris SM, Jr., Ash DE, Traish A and Christianson DW. Probing erectile function: S-(2- boronoethyl)-l-cysteine binds to arginase as a transition state analogue and enhances smooth muscle relaxation in human penile corpus cavernosum. Biochemistry 4: 2678-2688, 21. 33. Laumann EO, Paik A and Rosen RC. Sexual dysfunction in the United States: prevalence and predictors. JAMA 281: 537-544., 1999. 34. Li H, Meininger CJ, Hawker JR, Jr., Haynes TE, Kepka-Lenhart D, Mistry SK, Morris SM, Jr. and Wu G. Regulatory role of arginase I and II in nitric oxide, polyamine, and proline syntheses in endothelial cells. Am J Physiol Endocrinol Metab 28: E75-82, 21. 35. Lu R, Hu CP, Wu XP, Liao EY and Li YJ. Effect of age on bone mineral density and the serum concentration of endogenous nitric oxide synthase inhibitors in rats. Comp Med 52: 224-228, 22. 36. Masuda H, Tsujii T, Okuno T, Kihara K, Goto M and Azuma H. Accumulated endogenous NOS inhibitors, decreased NOS activity, and impaired cavernosal relaxation with ischemia. Am J Physiol Regul Integr Comp Physiol 282: R173-1738, 22. 37. Melman A and Christ GJ. Integrative erectile biology. The effects of age and disease on gap junctions and ion channels and their potential value to the treatment of erectile dysfunction. Urol Clin North Am 28: 217-231, vii, 21. 38. Moody JA, Vernet D, Laidlaw S, Rajfer J and Gonzalez-Cadavid NF. Effects of longterm oral administration of L-arginine on the rat erectile response. J Urol 158: 942-947, 1997. 39. Mori M and Gotoh T. Regulation of nitric oxide production by arginine metabolic enzymes. Biochem Biophys Res Commun 275: 715-719, 2. 4. Musicki B, Champion HC, Becker RE, Liu T, Kramer MF and Burnett AL. Erection capability is potentiated by long-term sildenafil treatment: role of blood flow-induced endothelial nitric-oxide synthase phosphorylation. Mol Pharmacol 68: 226-232, 25a. 41. Musicki B, Kramer MF, Becker RE and Burnett AL. Age-related changes in phosphorylation of endothelial nitric oxide synthase in the rat penis. J Sex Med 2: 347-357, 25b. 27

Page 28 of 38 42. Pegge NC, Twomey AM, Vaughton K, Gravenor MB, Ramsey MW and Price DE. The role of endothelial dysfunction in the pathophysiology of erectile dysfunction in diabetes and in determining response to treatment. Diabet Med 23: 873-878, 26. 43. 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 326: 9-94, 1992. 44. Sakai Y, Masuda H, Kihara K, Kurosaki E, Yamauchi Y and Azuma H. Involvement of increased arginase activity in impaired cavernous relaxation with aging in the rabbit. J Urol 172: 369-373, 24. 45. Sezen SF and Burnett AL. Intracavernosal pressure monitoring in mice: responses to electrical stimulation of the cavernous nerve and to intracavernosal drug administration. J Androl 21: 311-315, 2. 46. Shi JP, Zhao YM and Song YT. Effect of aging on expression of nitric oxide synthase I and activity of nitric oxide synthase in rat penis. Asian J Androl 5: 117-12, 23. 47. Snyder RO, Xiao X and Samulski RJ. Vectors for gene therapy: production of recombinant adeno-associated viral vectors. Current Protocols in Human Genetics N.C. Dracopoli, J.L. Haines, B.R. Korf, et al. (Eds.): Wiley, New York, 1996. 48. Tenu JP, Lepoivre M, Moali C, Brollo M, Mansuy D and Boucher JL. Effects of the new arginase inhibitor N(omega)-hydroxy-nor-L-arginine on NO synthase activity in murine macrophages. Nitric Oxide 3: 427-438, 1999. 49. Utkan T, Yildirim S, Sarioglu Y, Yildiz F and Yildirim K. Aging impairs nitric oxidemediated relaxant responses of rabbit corporal smooth muscle. Nitric Oxide 6: 342-346, 22. 5. White AR, Ryoo S, Li D, Champion HC, Steppan J, Wang D, Nyhan D, Shoukas AA, Hare JM and Berkowitz DE. Knockdown of arginase I restores NO signaling in the vasculature of old rats. Hypertension 47: 245-251, 26. 51. Wolin M. Interactions of oxidants with vascular signalling systems. Arterioscler Thromb Vasc Biol 2: 143-1442, 2. 52. Zhang C, Hein TW, Wang W, Chang CI and Kuo L. Constitutive expression of arginase in microvascular endothelial cells counteracts nitric oxide-mediated vasodilatory function. Faseb J 15: 1264-1266, 21. 28

Page 29 of 38 Figure Legends Figure 1. (Left) Effect of (A) cavernous nerve stimulation [CNS; 8 V], (B) intracavernous injection of acetylcholine [ACh; 44 nmol], and (C) calcitonin gene related peptide [CGRP;.1 nmol] on increases in intracavernosal pressure obtained in young and aged mice ranging from 4-26 months of age. Aged mice demonstrated a significant age-dependent decrease in erectile responses to CNS and intracavernous ACh with regression analyses of r 2 =.8376 and r 2 =.6832, respectively. (Right) Bar graphs showing the increase in intracavernosal pressure in response to (D) cavernous nerve stimulation [1-8 V], intracavernous injection of (E) ACh [13 and 44 nmol], and (F) CGRP [.1 and.3 nmol]. n indicates number of experiments; (P <.5) response significantly different compared to young mice. Figure 2. Effect of dietary supplementation of L-arginine (75 mg/kg/d) for 2 weeks on increases in intracavernosal pressure in response to cavernous nerve stimulation (CNS; 2 and 4 V) and intracavernous injection of acetylcholine (ACh; 13 and 44 nmol) in aged mice. n = 9; (P <.5) response significantly different compared to mice given free access to normal chow. Figure 3. (A) Western blot analysis of the endothelium-selective protein caveolin 1 (CAV1) and endothelium-specific protein PECAM (CD-31) in isolated mouse pulmonary vascular endothelial cells and isolated penile endothelial cells. Western blot for -smooth muscle actin (-sm-actin) confirmed no expression of this smooth muscle selective protein in isolated penile endothelial cells. Positive control (+C) is mouse vascular smooth muscle cell extract (B) Western blot analysis demonstrating expression of (left) arginase I and (right) NOS3 protein in isolated penile endothelial cells from young (lanes 1,2) and aged (lanes 3,4) mice penes. Bar graph demonstrating the densitometry analysis of arginase I and NOS3 protein normalized as the gel units of arginase I/GAPDH and NOS3/GAPDH in young and aged mice penile endothelial cells (n = 7). Additional bar graph demonstrating quantitative PCR for arginase I and NOS3 mrna 29

Page 3 of 38 (pg/sample) in young and aged mice penile endothelial cells (n = 7). RNA samples were normalized for GAPDH RNA expression to confirm RNA loading. (P <.5) significantly different than young mice. Figure 4. Effect of aging on arginase and endothelial NOS activity as well as penile cgmp levels. (A) Arginase enzyme activity as determined by the conversion of L-arginine to urea and L-ornithine in young and aged (old) endothelial cell lysates in the absence and presence of the arginase inhibitor ABH (5 µm; P<.5 compared to young; P<.5 compared to young; P<.5 compared to aged). (B) Endothelial NOS activity as determined by the conversion of L- arginine to L-citrulline in the absence and presence of the arginase inhibitor ABH (5 µm) in endothelial cell lysates (P<.5 compared to young; P<.5 compared to aged) (C) Concentration-dependent influence of the arginase inhibitor, ABH (1-3 µm), on cgmp concentrations in young and aged (old) mice penes (P<.5 compared to baseline; P<.5 compared to young). n indicates number of experiments. Figure 5. Bar graphs demonstrating the increase in intracavernosal pressure before and after intracavernous administration of ABH (dose) in response to (A) cavernous nerve stimulation [2 and 4 V] and direct intracavernous injection of (B) ACh [13 and 44 nmol] and (C) CGRP [.1 and.3 nmol] in aged (left) and young (right) mice. n indicates number of experiments; (P <.5) response significantly different compared to erectile response obtained before intracavernous injection of ABH. Figure 6. (A; left) Western blot analysis demonstrating expression of arginase I and GAPDH protein in penes of young (lanes 1,2), aged (lanes 3,4), and aged mouse penes 1 month after gene transfer of AAVAnti-Arginase (lanes 5,6). (A; right) In addition, the effect of transfection of 3

Page 31 of 38 the reporter gene (AAVgal; lanes 4-6) on the expression of arginase I was compared to vehicle (lanes 1-3) and found to be similar. Moreover, arginase activity, NOS activity, and cgmp levels were not altered by AAVgal when compared to vehicle (data not shown). Bar graph (left) demonstrating the densitometry analysis of arginase I protein normalized as the gel units of arginase I/GAPDH in young and aged mice penes (n = 7). Additional bar graph (right) demonstrating quantitative PCR for arginase I mrna (pg/sample) in young and aged mice penes (n = 6). (P <.5) significantly different than young; (P<.5) significantly different than aged. (B) Bar graph demonstrating constitutive NOS activity in penes from young, aged, and aged mice transfected with AAVAnti-Arginase (n = 9). (P <.5) indicates NOS activity significantly different when compared to young mice;(p <.5) indicates NOS activity significantly different when compared to aged mice. (C) Bar graph demonstrating penile cgmp levels in young, aged, and aged mice transfected with AAVAnti-Arginase (n = 9). (P <.5) indicates cgmp levels significantly different when compared to young mice;(p <.5) indicates cgmp levels significantly different when compared to aged mice. Figure 7. Bar graphs demonstrating the increase in intracavernosal pressure in response to (A) cavernous nerve stimulation [2 and 4 V] and (B) intracavernous injection of ACh [13 and 44 nmol] in aged (left) and young (right) mice transfected with AAVgal or AAVAnti-Arginase. In vivo erection experiments were conducted 1 month after transfection with adeno-associated viruses. n indicates number of experiments; (P <.5) response significantly different compared to mice transfected with AAVgal. 31

Page 32 of 38 Intracavernosal Pressure (mmhg) Intracavernosal Pressure (mmhg) Intracavernosal Pressure (mmhg) a b c 75 5 25 CNS (8 V) r 2 =.8376 P<.1 1 2 3 75 5 25 ACh 44 (1 nmol µg) r 2 =.6832 P<.1 1 2 3 75 5 25 CGRP (.1 nmol) r 2 =.317 P=.45 1 2 3 Age (months) d e f 75 5 25 75 5 25 75 5 25 Young (n=9) Aged (n=8-9) 1 2 4 8 Voltage 13 3 1 44 ACh (µg) (nmol)..1.3 CGRP (nmol) Figure 1

Page 33 of 38 Figure 2 13 44 ACh (nmol)

Page 34 of 38 Figure 3 a Pulmonary EC Penile EC CAV1 PECAM +C Penile EC -sm-actin b Young Aged Young Aged Arg I NOS3 GAPDH GAPDH Arginase I / GAPDH 2 1 Arginase I mrna (pg/sample) Arginase I mrna (pg/sample) 1 75 5 25 NOS3 / GAPDH 2 1 NOS3 mrna NOS3 mrna (pg/sample) (pg/sample) 1 75 5 25 Young Aged Young Aged Young Aged Young Aged Arginase I NOS3

Page 35 of 38 a Arginase Activity (pmol urea min -1 mg protein -1) pmol urea min -1 mg protein -1 75 5 25 n=9 Young Young + ABH Old Old + ABH Figure 4 b [C 14 ]-Citrulline Formation (pmol.mg L-arginine to L-Citrulline Formation (pmol hr-1 mg protein-1) protein -1.hr -1 ) 3 2 1 n=9 Arginase Activity Young Young + ABH Old Old + ABH Ca Calcium-Dependent 2+ - Conversion c Change in cgmp (%) 3 2 1 Young (n=6) Old (n=7) 1 1 1 ABH (µm)

Page 36 of 38 Figure 5 a Intracavernosal Pressure (mmhg) 5 4 3 2 1 Aged Control (n=6) ABH (n=7) Intracavernosal Pressure (mmhg) 5 4 3 2 1 Young 2 4 2 4 Voltage Voltage b Intracavernosal Pressure (mmhg) 5 4 3 2 1 Intracavernosal Pressure (mmhg) 5 4 3 2 1 13 3 1 44 13 3 1 44 ACh (nmol) (µg) ACh (nmol) (µg) c Intracavernosal Pressure (mmhg) 5 4 3 2 1 Intracavernosal Pressure (mmhg) 5 4 3 2 1..1.3..1.3 CGRP (nmol) CGRP (nmol)

Page 37 of 38 a Vehicle AAVgal AAVAnti-Arginase Arg I Vehicle AAVgal Arg I Figure 6 GAPDH Arginase I / GAPDH 2 1 Arginase I (pg/sample) 1 75 5 25 b c Young Aged Aged + AAVAnti-Arginase L-arginine to L-Citrulline Formation (pmol hr-1 mg protein-1) cgmp (fmol/mg protein) 3 2 1 Calcium-Dependent Conversion Young Young Aged Aged + AAVAnti-Arginase Aged Aged + AAVAnti-Arginase Young Aged Aged + AAVAnti-Arginase