Abstract
The goal of the present study was to investigate the antifibrotic role of inducible nitric oxide synthase (iNOS) in Peyronie’s disease (PD) by determining whether a plasmid expressing iNOS (piNOS) injected into a PD-like plaque can induce regression of the plaque. A PD-like plaque was induced with fibrin in the penile tunica albuginea of mice and then injected with a luciferase-expressing plasmid (pLuc), either alone or with piNOS, following luciferase expression in vivo by bioluminescence imaging. Rats were treated with either piNOS, an empty control plasmid (pC), or saline. Other groups were treated with pC or piNOS, in the absence of fibrin. Tissue sections were stained for collagen, transforming growth factor (TGF) β1, and plasminogen-activator inhibitor (PAI-1) as profibrotic factors; copper-zinc superoxide dismutase (CuZn SOD) as scavenger of reactive oxygen species (ROS); and nitrotyrosine to detect nitric oxide reaction with ROS. Quantitative image analysis was applied. Both iNOS and xanthine oxido-reductase (XOR; oxidative stress) were estimated by Western blot analysis. Luciferase reporter expression was restricted to the penis, peaked at 3 days after injection, but continued for at least 3 wk. In rats receiving piNOS, iNOS expression also peaked at 3 days, but expression decreased at the end of treatment, when a considerable reduction of plaque size occurred. Protein nitrotyrosine, XOR, and CuZn SOD increased, and TGFβ1 and PAI-1 decreased. The piNOS gene transfer regressed the PD plaque and expression of profibrotic factors, supporting the view that endogenous iNOS induction in PD is defense mechanism by the tissue against fibrosis.
Introduction
Peyronie’s disease (PD) is a localized fibrosis of the tunica albuginea (TA) of the penis affecting approximately 2%–4% of the male population [1]. This fibrosis is assumed to be triggered by trauma to the erect penis [2, 3], followed by vascular damage, fibrin extravasation, recruitment of inflammatory cells, and release of profibrotic factors, such as transforming growth factor (TGF) β1 and other cytokines, as well as reactive oxygen species (ROS), leading to cellular infiltration, proliferation, and differentiation. These processes cause chronic inflammation, excessive collagen deposition and disorganization, and elastin fragmentation. This can be recognized clinically as a palpable plaque and/or penile curvature during tumescence [2–4]. Although spontaneous regression of the plaque and the resulting curvature may occur in approximately 15% of these cases [1], and although some reports suggest that the injection of certain agents (e.g., verapamil) may improve the clinical condition, the majority of PD cases seem to be refractory to pharmacological therapy.
The recent description of a new rat animal model for PD, whereby the injection of fibrin into the TA produces a lesion that is indistinguishable histologically from the human condition, allows us to test pharmacological regimens in an attempt to cure or diminish the fibrosis. One approach was based on the recently recognized antifibrotic effect of nitric oxide (NO), which reduces collagen deposition and inhibits fibroblast differentiation into myofibroblasts (cells whose persistence in abnormal wound healing may lead to fibrosis) [5]. Nitric oxide was also a potent quencher of ROS, one of the presumed agents in PD that induce fibrosis. Furthermore, when inducible NO synthase (iNOS) activity is inhibited in a rat model of PD, tissue fibrosis is exacerbated and leads to an increase in oxidative stress, whereas administration of the NOS substrate l-arginine and certain phosphodiesterase inhibitors exerts an antifibrotic effect [6– 8].
In the present study, we have investigated in the rat whether direct administration of a plasmid construct of iNOS cDNA into the penile TA produces NO and whether NO would then act as an antifibrotic agent to cause regression of the PD-like plaque induced by fibrin.
Materials and Methods
Plasmid Preparation
Plasmid pcDNA3-RPiNOS (piNOS), which expresses the rat penile iNOS cDNA from the CMV promoter, is the construct cloned from a rat corpora cavernosa lambda cDNA library as previously described [9]. The second plasmid construct in the same vector, but expressing luciferase as a reporter protein, is the pSVD5′-luc (pLuc) [10]. The “plasmid control” for both constructs is the pcDNA3 vectors without any cDNA insert was termed pC. All plasmids were amplified and purified for in vivo injection using a Qiagen EndoFree Plasmid Maxi Kit (Qiagen, Inc., Valencia, CA) and were checked for purity by restriction-enzyme digestion. Concentrations were estimated by spectrophotometric determination at 260 nm and further validated by agarose gel/ethidium bromide staining against known amounts of standard plasmid DNA. Before injection, plasmids were diluted in sterile 0.9% saline at 100 μg/μl.
Animal Treatments and Tissue Processing
Male Balb mice (age, 3 mo) and male Sprague Dawley rats (age, 8– 11 mo) were purchased from Harlan Sprague Dawley (San Diego, CA) and maintained under controlled temperature and lighting according to National Institutes of Health regulations and following Institutional and Animal Care and Use Committee-approved protocols. Animals were anesthetized (isofluorane in air, delivered via a nose cone) and then injected (into the TA) with a fibrin mix preparation as previously described [4] to induce a PD-like plaque. Mice received 30 μl and rats 60 μl of the fibrin mix (Fig. 1A). At 21 days after fibrin injection, when the plaque is well-developed [4], the animals were injected into the same area of the TA with either plasmid constructs (pLuc, piNOS, or pC) or saline, followed by a second, similar injection on Day 32 (11 days after the first injection of plasmid). Electroporation was also used at the time of plasmid injection to facilitate plasmid uptake [11]. Plasmid injections 21 days after the fibrin injection in mice (n = 5 per group) were as follows: pLuc alone (25 μg), or pLuc (25 μg) and piNOS (25 μg). In the reinjections at Day 32, pLuc was increased to 75 μg, but piNOS remained at 25 μg. Injections in rats (n = 13 per group) at 21 and 32 days after fibrin injection were as follows: piNOS (100 μg), pC (100 μg), or no plasmid and saline alone (100 μl). In addition, other rats not receiving fibrin were injected in the TA with either piNOS (100 μg) or pC (100 μg) and then killed at Day 3 after plasmid injection.
Time frame of fibrin and plasmid injections to the tunica albuginea in the bioluminescence and iNOS gene transfer protocols. Animals were injected with fibrin to generate the PD-like plaque, followed by a plasmid constructs or saline injection to study their localization and expression in vivo (A; bioluminescence in mice) or to evaluate the development of the PD plaque (B; gene transfer in rats)
Male Balb-c mice (n = 5 per group; groups a and b) were evaluated for in vivo gene expression by bioluminescence imaging system analysis (Xenogen, Alameda, CA). Rat penile tissues (groups c–e) were used for histological and immunohistochemistry analysis (n = 5 per group) or Western blot analysis (n = 4 per group) and were killed at 45 days after fibrin injection (Fig. 1B). Groups f and g (n = 4 per group) were killed 3 days after plasmid injection to determine, via Western blot analysis, both to what extent iNOS protein was induced from the plasmid injection at a very early time, when no PD plaque has been induced, and the effect of iNOS on oxidative stress (xanthine oxido-reductase [XOR]).
Animals dedicated for histochemistry/immunohistochemistry were perfused through the left ventricle with saline followed by 4% formalin. Penile shafts were excised, and transverse slices (thickness, 3–4 mm) were cut around the site of injection. Tissues were postfixed overnight in 4% formalin, stored at 4°C in PBS, and used for preparing paraffin-embedded serial sections (thickness, 5 μm). The remaining animals were not perfused, and equivalent sections of the penile shaft were obtained, snap-frozen in liquid nitrogen, and then stored at −80°C.
Western Blot Analysis of iNOS and XOR Expression
Western blot analysis was performed essentially as previously described for iNOS [9] on a postmitochondrial supernatant obtained from homogenates of nonfixed tissue slices around the site of injection. Equal amounts of protein (40 μg) were run on 7.5% polyacrylamide gels and submitted to Western blot immunodetection. Membranes were reacted with monoclonal anti-iNOS immunoglobulin (Ig) G (1:500; Transduction Laboratories, Lexington, KY) and a secondary goat anti-mouse IgG linked to horse radish peroxidase (Transduction Laboratories), followed by a luminol reaction. The positive control was cytosol from rat penile TA cells induced with interferon-γ, interleukin-1β, tumor necrosis factor α and lipopolysaccharide for 72 h. A semiquantitative estimation of the 130-kDa iNOS band was performed by densitometry. The XOR was estimated with a polyclonal IgG antibody against XOR loading 30 μg of total protein (1: 5000; Abcam, Cambridge, MA) utilizing the same procedure as that for estimation of the xanthine dehydrogenase (145-kDa) and xanthine oxidase (65- to 70-kDa) bands [12].
Bioluminescence Image Analysis
Luciferase expression was monitored in live mice using the cooled IVIS animal imaging system (Xenogen, Alameda, CA) linked to a PC running the Living Image software (Xenogen) along with IGOR (Wavemetrics, Seattle, WA) under Microsoft 2000. This system provides high signal-to-noise images of luciferase signals [13]. Before imaging, 150 mg/ml of luciferin (K+ salt; Promega, Madison, WI) in PBS was injected i.p. at a dose of 126 μg/g body weight. Luciferin as a substrate, along with oxygen and ATP, causes emission of light within the mouse proportional to the level of luciferase expression. The emitted light scatters and diffuses until reaching the surface, and photons are captured by the charged-coupled device (CCD) camera as a signal measured and recorded in the computer. In addition, the software creates a color image related to the photon density.
Mice were anesthetized with ketamine and xylazine (100 and 10 mg/kg i.p., respectively). The initial surface image of each mouse was acquired by using 10-cm field overview, a 0.2-sec exposure time, a binning (resolution) factor of 2, a 16 f/stop (aperture), and an open filter. Animals were positioned to place the penile area toward the CCD camera, obtaining bioluminescence images 20 min after luciferin injection, with a binning factor of 4, a 1 f/stop, and an open filter [13].
Relative intensities of transmitted light were represented as a color image ranging from violet (least intense) to red (most intense). Signal intensities from the region of interest (ROI) were defined manually, and data were expressed as photon flux (photons sec−1 square-cm−1 steradian−1, where steradian refers to the photons emitted from a unit solid angle of sphere. Background photon flux was defined from a ROI of the same size drawn over the thorax of each animal, and these data were subtracted from signal intensities measured at the penile site [13, 14].
Histochemistry and Immunohistochemistry Determinations
The formalin-fixed penile tissue sections around the site of injection were used for Masson trichrome (Sigma Diagnostic, St. Louis, MO) staining for collagen estimation in the TA. This staining identifies collagen as blue and smooth muscle as red [4]. Other serial sections were evaluated by immunohistochemistry [4, 7] with polyclonal antibodies for TGFβ1 (1: 100; Promega, Madison, WI), plasminogen-activator inhibitor (PAI-1; 1: 200; Abcam Ltd., Cambridge, U.K.), nitrotyrosine (1:100; Upstate, Lake Placid, NY), and copper-zinc superoxide dismutase (CuZn SOD; 1:500; Oxygen, Portland, OR). Only tissue sections evaluated for TGFβ1 were treated with proteinase K (20 μg/ml), followed by quenching in 2% H202-methanol, blocked with goat serum (Vector Laboratories, Burlingame, CA), and incubated overnight at 4°C with the primary antibody at the dilution mentioned above. This was followed by reaction with biotinylated anti-rabbit IgG (Vector Laboratories) or anti-sheep biotinylated antibody (1:200) for CuZn SOD for 30 min, followed by the avidin-biotin-peroxidase complex (1:100; Calbiochem) and 3,3′-diaminobenzidine. Sections were counterstained with hematoxylin. Antibodies were validated by Western blot analysis, and negative controls in the immunohistochemical detections omitted the primary antibody or had the first antibody replaced by IgG isotype at the same concentration.
Quantitative image analysis was performed by computerized densitometry using the ImagePro 4.01 program (Media Cybernetics, Silver Spring, MD) coupled to an Olympus BHS microscope equipped with a Spot RT digital camera [4, 6, 7]. For Masson staining, the plaque was analyzed by the tunical thickness, collagen deposition, and disorganization. For immunohistochemistry, cells were considered to be positive when their color intensity was threefold greater than background, and their number was counted in a computerized grid and expressed against the total number of cells as determined by counterstaining or as positive cells per field. Four fields were measured per section, with at least three anatomically matched sections per animal being analyzed in all determinations.
Statistical Analysis
Values are expressed as the mean ± SEM. The normality distribution of the data was established using the Wilk-Shapiro test. Multiple comparisons among the different groups were analyzed by a single-factor ANOVA, followed by post-hoc comparisons with the Student-Neuman-Keuls test, according to the GraphPad Prism (version 3.0) software. Differences among groups were considered to be significant at P < 0.05.
Results
Localization and Expression of Plasmid cDNA Constructs
To determine whether injection of a plasmid construct expressing iNOS cDNA (piNOS) into the TA remains predominantly in the penis after injection and is active in expressing protein for a sufficiently long period of time, adult mice (n = 5 per group) were injected with fibrin into the TA, and at 21 days, the animals were injected in the same area of the TA with 25 μg of a plasmid construct expressing luciferase as a reporter gene (pLuc) or pLuc together with the piNOS construct (Fig. 1A). Mice were used in this experiment to facilitate detection of emitted light, and the 21-day period for starting the plasmid injection was selected based on our previous study in rats [4] and on our preliminary data in mice (not shown) demonstrating that the plaque in both mouse and rat becomes recognizable in the TA 21 days after fibrin injection.
The mice were evaluated at 0, 21, 25, 29, 32, 34, 35, 39, 42, and 45 days after fibrin injection, and the plasmids were injected at 21 days (25 μg of pLuc with or without 25 μg of iNOS) and 32 days (75 μg of pLuc with or without 25 μg of piNOS) post-fibrin injection. The combination of both plasmids, pLuc and piNOS, was intended to determine whether the effects exerted by the antifibrotic action of iNOS cDNA on the plaque itself would affect the localization and persistence of luciferase expression by the reporter gene construct. The reporter gene dose was tripled at Day 32 to facilitate light-emission detection. As a comparison, the treatment timeline for iNOS gene transfer in rats is given in Figure 1B.
The top of Figure 2 shows visualization of luciferase expression in the anesthetized mouse. Clearly, from Day 21 to Day 45 of the study, the plasmid remains localized in the lower abdominal region in a very small area consistent with the location of the penis, with negligible dissemination to other areas (Fig. 2, A–F). However, at Day 32, the intensity decreased (not shown) as compared to 29 days (B), and for this reason, it was determined that a second injection should be given with triple the initial pLuc dose to allow maintenance of the intensity of the luciferase expression.
Time course of plasmid luciferase (pLuc) and plasmid iNOS (piNOS) expression in living mice. Top) Photon flux and light emission in the images elicited 20 min after luciferin injection. Left) Pseudocolor scale: blue (minimum) to red (maximum). Right) Representative images of mice (n = 5): control (A; Day 0), injected with pLuc (75 μg) and piNOS (25 μg) at 29 days (B), mice reinjected at 32 days with pLuc with or without piNOS (C–F); 75 μg ± 25 μg). Scanning was performed as indicated. Bottom) Means and SEMs for photon flux on each day after plasmid injection. Differences among groups were considered to be significant at P < 0.05
To determine an approximate time-course of expression in the animals, a computerized image analysis was performed on the captured images (Fig. 2, bottom). This indicates a decay of luminescence after Day 29 (Fig. 2B) following the first series of plasmid injections; after the second plasmid injection (Day 32), a peak of luciferase expression was seen 2 days (Day 34) and 3 days (Day 35) later (Fig. 2, C and D), followed by a progressive decline until the last time studied (Day 45) (Fig. 2, E and F).
iNOS and XOR Expression after Plasmid iNOS cDNA Injection into Normal TA of Rat
To show that our piNOS cDNA is able to express iNOS in the normal TA [9], 100 μg of either the piNOS construct or the plasmid control (empty) were injected into the TA of rats that had not been treated with fibrin (n = 4). Animals were then killed at Day 3 of plasmid injection, coinciding with the peak of luciferase expression observed in mice with the reporter plasmid (Fig. 2). Portions of the penile shaft around the site of injection were homogenized and subjected to Western blot analysis, which showed a considerable increase in iNOS protein levels in the animals that received the piNOS as compared to the ones injected with plasmid control (Fig. 3, top). The densitometric analysis (Fig. 3, bottom) confirmed the visual observations: a significant, fourfold stimulation of iNOS protein levels in the rats injected with the piNOS construct as compared to the animals receiving the plasmid control. The control tissue was whole rat penile tissue, including the crura.
The iNOS and xanthine oxido-reductase expression at an early period 3 days after piNOS injection into the normal penile tunica albuginea of the rat as determined by Western blot analysis. Rats received 100 μg of plasmid control (pC) or piNOS (n = 4 per group). Tissue region around the site of injection (40 μg protein/lane) were subjected to Western blot analysis, with primary antibodies as indicated, followed by a secondary antibody linked to peroxidase and a luminol reaction. Top) Selected immunoreactive bands in each individual tissue homogenate for iNOS (130 kDa) and, on a separate gel, the two forms of xanthine oxido-reductase: xanthine dehydrogenase (XDH; 145 kDa) and xanthine oxidase (XO; 70 kDa). To facilitate visualization and densitometry, the film exposures are different: 2 min for iNOS, 60 min for XDH, and 10 min for XO. Control (C) was penile homogenate, including crura. Bottom) Densitometry analysis expressed as the mean ± SEM. *P < 0.05
To investigate whether plaque development is associated with ROS production, we measured in the same extracts the expression of XOR, one of the enzymes more responsible for oxidative stress [12]. This enzyme has two distinct forms: a nonglycosylated homodimer of two 145-kDa subunits, which is constitutively expressed in vivo (xanthine dehydrogenase); and a posttranscriptionally modified form originated by spontaneous proteolytic cleavage consisting of four discrete fragments, the main one at approximately 70–80 kDa (xanthine oxidase) [12]. Nonsignificant reductions were observed in xanthine oxidase and xanthine dehydrogenase in the TA of animals receiving piNOS as compared with the ones treated with the control plasmid (pC) (Fig. 3, middle and bottom).
Effects of Gene Transfer with iNOS cDNA Construct on PD-Like Plaque Development and Oxidative Stress
Based on our previous results, we decided to use rats (n = 4 per group), a validated animal model [4], for our next protocol, which was to inject fibrin into the TA and then, 21 days later, when the plaque had developed [4], inject the animals with 100 μg of piNOS, pC, or saline alone (no plasmid). Animals were reinjected at Day 32 (11 days after the first plasmid injection) as previously described (Fig. 1B).
The rats were killed 45 days after the fibrin injection, which was 24 and 13 days after the first and second plasmid/saline injections, respectively, and tissues were collected for analysis. The effects of these plasmid treatments on the development of the PD-like plaque in the rats was evaluated by the estimation of collagen deposition/disorganization, tunical thickness, and the expression of profibrotic factors, such as TGFβ1 and PAI-1, within the TA. In addition, nitrotyrosine and the antioxidant-enzyme CuZn SOD staining were estimated. The top of Figure 4 shows representative tissue sections from each group, where the disorganization of collagen fibers in the PD-like plaque of the rats injected with either saline or pC is denoted by the pale blue stain and thickening of the TA. This indicates that the plaque is still present at 45 days, which is in agreement with our previous data [4]. In contrast, the TA in the piNOS-treated animals appeared to be normal. Image analysis (Fig. 4, bottom) confirmed the visual assessment, with an 80% reduction in the PD-like plaque area, expressed as image units, in the animals treated with piNOS as compared to the ones treated with saline or pC.
Evaluation of collagen deposition and disorganization and plaque area in the fibrin-induced, PD-like plaque (rat) at the completion of treatment with piNOS. The PD-like plaque was injected at 21 and 32 days with saline, pC, or piNOS (100 μg); animals were killed at 45 days after fibrin injection (n = 5 per group). Top) Masson trichrome staining. Red arrowheads denote site of injection, solid arrows increased collagen deposition/disorganization and a thicker TA (PD-like plaque), and broken arrows “normal”-appearing TA (normal collagen deposition and TA thickness). Magnification, ×40. Bottom) Quantitative image analysis was performed on four sections from each one of the individual tissue specimens. Plaque area is defined as the area encompassed by an increase of collagen deposition/disorganization and tunical thickness. Values are the mean ± SEM and are expressed as image units (IU). Differences among groups were considered to be significant at P < 0.05
To correlate this plaque regression with iNOS expression, Western blot analysis was performed in another group of rats (n = 4) that were similarly treated and killed 13 days after the second plasmid injections (45 days after fibrin injection). Comparison between the groups injected with either piNOS or pC indicated an increase of iNOS expression in rats receiving piNOS (Fig. 5, top). The densitometric analysis (Fig. 5, bottom) confirmed the visual observations, again indicating a fourfold greater iNOS expression in the TA injected with piNOS (where the plaque had regressed) as compared with the PD-like plaque unaffected by pC. The collective bioluminescence (plasmid expression) and Western blot results (iNOS) indicate that the CMV-driven plasmid iNOS construct injected and electroporated into the TA localizes in the penis and remains active in synthesizing iNOS for at least 3.5 wk after injection, although expression peaks much earlier (3–7 days after plasmid injection) and gradually diminishes over time.
The iNOS and xanthine oxido-reductase expression at a late period 45 days after piNOS injection into the penile tunica albuginea of the rat where a PD-like plaque had been induced with fibrin as determined by Western blot analysis. The piNOS and plasmid control pC were injected as described in Figure 4. To facilitate visualization and densitometry, the film exposures for iNOS, xanthine dehydrogenase (XDH), and xanthine oxidase (XO) are different (10, 60, and 10 min, respectively). All other details are as described in Figure 3
Despite plaque regression, the tissue appears at this late stage to have experienced additional oxidative stress by the prolonged piNOS expression, because in contrast with observations at 3 days, both xanthine dehydrogenase and xanthine oxidase were significantly elevated as compared to the unaffected PD-like plaque treated with plasmid control (Fig. 5, middle and bottom).
Effects of Gene Transfer with iNOS cDNA Construct on Expression of Profibrotic Factors and anAntioxidant Enzyme
Tissue sections of the penis were immunostained for nitrotyrosine, an indicator of a cumulative production of NO and its reaction with ROS to produce peroxynitrite. The assumption was made that even if iNOS expression declined in the last stages of treatment, nitrotyrosine would be an indirect marker of cumulative NO synthesis and, hence, of iNOS induction. This, indeed, seems to be the case, as shown in the top of Figure 6, where representative tissue sections indicate an intense nitrotyrosinylation of penile proteins after piNOS treatment as compared with the controls. Quantitative image analysis (Fig. 6, bottom, left bars) demonstrated a significant, twofold increase of nitrotyrosine accumulation in terms of positive cells per field as compared to the control groups, thus supporting the assumption that iNOS was considerably expressed from the piNOS construct during plaque regression (even if at late stages its expression might have declined).
Immunohistochemical evaluation of fibrosis and oxidative stress-related markers in the fibrin-induced, PD-like plaque (rat) at the completion of treatment with piNOS. Tissue sections adjacent to the ones shown in Figure 4 were immunostained with specific antibodies as indicated, followed by a secondary antibody linked to peroxidase and 3,3′-diaminobenzidine staining (n = 5 per group). Top) Representative micrographs. Arrows point to cells with intense staining. Magnification, ×200. Bottom) Quantitative image analysis of positive cells in an average microscopic field representing four fields per section and four sections per specimen from each animal. Values are the means ± SEM and are expressed as positive cells per field (PCPF). *P < 0.05 versus control
Because TGFβ1 and PAI-1 have been identified as the main profibrotic factors in PD, we studied their expression to confirm whether plaque regression, as indicated above (Fig. 4), was actually accompanied by the reduction of these fibrotic factors. Both TGFβ1 and PAI-1 immunostaining were visually reduced at 45 days after fibrin injection in rats receiving piNOS (Fig. 4, top). This was confirmed by image analysis (Fig. 4, middle and bottom), where TGFβ1 was significantly decreased by 42% and PAI-1 was significantly decreased by 68% when compared with the animals injected with pC. A similar situation occurred when comparisons were made against the saline-injected plaques. Finally, the expression of antioxidant CuZn SOD as a compensatory reaction against ROS was evaluated (Fig. 4, top). Likely as a consequence of the observed increase in XOR in the TA of the animals receiving piNOS as compared to pC, the antagonistic-enzyme CuZn SOD was slightly elevated by this iNOS treatment. These results suggest that our iNOS plasmid construct induced considerable iNOS expression and NO synthesis in the penis, down-regulated the well-known profibrotic factors TGFβ1 and PAI-1, but did not lead to a clear predominance of antioxidant over oxidant enzymes.
To confirm that plaque size was correlated with the expression of both TGFβ1 and PAI-1, we compared our data in a plot that correlates the size of the fibrin-induced plaque (established by Masson trichrome) with the number of positive cells expressing PAI-1 and TGFβ1 among the three groups of animals at Day 24 of treatment (45 days after fibrin injection) (Fig. 7). As expected, we observed a significant, positive correlation between plaque sizes and the levels of both PAI-1 and TGFβ1 in the penile shaft area around the site of injection.
Relationship between PD-like plaque area in the rat and positive cells expressing TGFβ1 and PAI-1. The average number of positive cells was correlated with the average plaque area after saline or plasmid injection (n = 5 per group). Values for the PD-like plaque, as estimated by Masson staining as arbitrary imaging units (IU), were obtained as in Figure 4, and values for TGFβ1 and PAI-1 positive cells per field, as estimated by immunocytochemistry, were obtained as in Figure 6. R, Regression coefficient
Discussion
The present study adds to the accumulating evidence that gene therapy for disorders afflicting the penis may be a viable therapeutic approach. We have previously reported on the use of this modality to treat erectile dysfunction in the aged rat by up-regulating the expression of NOS and, hence, the levels of NO, which is the chemical mediator of penile erection, in the corporal tissue [9, 15]. Our aim in the present study was to evaluate the role of the iNOS gene transfer as a modulator of tissue fibrosis during the remodeling process in the PD-like plaque in the TA.
To our knowledge, this is the first direct demonstration that NO produced by iNOS acts as an antifibrotic agent in the penis. The present study also highlights the role of gene therapy in combating disorders of the penis, where the anatomical location of the organ makes it ideal for the delivery of gene constructs into the tissue proper itself. Our success in ameliorating the tunical fibrosis, induced by the injection of fibrin, with iNOS gene transfer suggests that the disorder of PD, which is extremely recalcitrant to pharmacological modulation, may be an adequate target for utilizing a gene therapy approach to correct this condition. We have shown that it is feasible to monitor the uptake and expression of the administered cDNA construct in the live animal by in vivo bioluminescence. In addition, our results support the concept that endogenous and exogenous iNOS constitute a tissue mechanism of defense or a viable therapeutic approach against collagen deposition and that this is associated with the reduction in the levels of TGFβ1 and PAI-1, although its effects on oxidative stress are less clear, in our PD animal model.
Our choice of gene transfer to raise the levels of iNOS protein in the rat TA was dictated by our previous experience with gene therapy to the penis as an effective modality for treating erectile dysfunction in the aged rat [9]. In those studies, we observed that a plasmid construct of iNOS similar to the one applied in the current study was satisfactorily expressed in the corpora cavernosa for at least 2 wk and, equally importantly, did not cause any detectable side effects [9]. We later documented the efficacy and safety of NOS-related gene therapy administered to the penis by injection in combination with electroporation, utilizing an adenoviral construct of penile neuronal NOS [15] and plasmid constructs of antisense sequences for a neuronal NOS-binding peptide, PIN (protein inhibitor of NOS) [16]. In this current work, we utilized, as an initial approach, the same iNOS cDNA construct that we employed previously [9], which is a plasmid-based expression system instead of a viral-based vector to avoid possible immune reactions associated with “nongutless” viral vectors that could possibly confound plaque regression.
The in vivo bioluminescence system employed in this protocol showed conclusively that expression of the luciferase reporter construct was restricted to the penis, decayed with time, and needed the boost of a second injection at Day 32 to keep levels detectable. The in vivo bioluminescence procedure has the advantage of being conducted in the live animal during the therapeutic time frame. It concurred with iNOS expression as detected by Western blot analysis that peaked at 3–4 days. Although iNOS levels were still elevated at 13 days after second plasmid injection (45 days after fibrin injection), it had decayed over this time period. Although our bioluminescence studies indicate the persistence of expression of the reporter gene even at 45 days, we cannot state with certainty how long this plasmid could have continued to be active in the penis. Previous determinations by reverse transcription-polymerase chain reaction of the persistence of similar amounts (100 μg) of the same vector, but expressing a different gene, have shown that it lasts as long as 6 months; however, in that study, no information was provided concerning protein expression [17].
The fact that iNOS expression was effective in regressing a well-formed, PD-like plaque, as indicated by a decrease of TA thickness and collagen disorganization as compared to the control groups, confirms our previous demonstration on the antifibrotic role of endogenous iNOS based on the long-term inhibition of iNOS activity by L-N-(1-iminoethy)-lysine [6, 7]. We hypothesize that increasing the NO to ROS ratio by overexpressing NO from iNOS gene transfer inhibits the profibrotic mechanisms and stimulates proper remodeling pathways in PD. Endogenous iNOS induction occurring in a PD plaque in the absence of treatment may not be sufficient, in some cases, to arrest plaque development, because the outcome will depend on the balance between NO production and the release of pro-fibrotic factors.
The clear stimulation of nitrotyrosine immunostaining after iNOS gene transfer indicates not only a higher production of NO but also an intensified reaction of NO with ROS to produce peroxynitrite, as detected by the cumulative nitrotyrosinylation of proteins in the tissue. This correlates with the oxidative stress that we observed by the xanthine oxidase and dehydrogenase expression at 13 days after plasmid iNOS injection (see below).
The levels of other key proteins involved in the fibrotic pathway, specifically TGFβ1 and PAI-1, were down-regulated by iNOS, and this suggests a possible mechanism for iNOS action, particularly considering that NO is known to inhibit the expression of both proteins [6, 7, 18, 19]. Regarding TGFβ1, this is one of the best characterized profibrotic factors [20, 21], and it is highly expressed subsequent to fibrin injection into the TA to elicit the PD-like plaque [4]. As to PAI-1, it inhibits fibrinolysis and may be a critical factor in the persistence of fibrin that triggers chronic inflammation and fibrosis within the TA [4] as well as inhibiting metalloproteinases and, therefore, collagen breakdown.
The moderate increase in cytosolic CuZn SOD on treatment with piNOS is probably related to SOD antifibrotic effects through the down-regulation of TGFβ1 expression [22] as well as through the quenching of free oxygen radicals [23, 24]. The administration of SOD by gene therapy induced an increase in cGMP levels in the corpora cavernosa of the penis as treatment of erectile dysfunction [25]. We also found that increasing cGMP/cAMP by long-term administration of phosphodiesterase inhibitors is effective in reversing the fibrosis of PD [8]. The fact that SOD was slightly increased by recombinant iNOS in conjunction with the increase in XOR levels suggests a struggle between antioxidant and oxidant pathways in which the net balance between both antagonistic processes cannot be determined from the respective protein levels.
The direct measurement of ROS levels by malonaldialdehide, reduced/oxidized glutathione, and other procedures [14, 24, 26] may clarify whether, regardless of XOR content, recombinant iNOS expression does or does not induce the desired decrease in ROS levels, because it apparently occurs with endogenous iNOS induction, as inferred from the increase in ROS observed in aging penile arteries by blockade of iNOS activity [7]. These measurements could not be conducted in the present study because of specific requirements of tissue preparation that were not compatible with the other selected assays. Therefore, although our previous studies suggested that endogenous iNOS is effective in reducing ROS, this question remains open for iNOS gene transfer. However, what can be stated is that NO induced by iNOS may induce a remodeling mechanism based on the reduction of TGFβ1 and PAI-1, two important profibrotic factors in PD.
The antifibrotic role of iNOS is supported by other, indirect evidence where NO is postulated as a protective factor, such as in an obstructive kidney model of fibrosis [27]. However, partial bladder outlet obstruction induced the expression of iNOS mRNA in the mouse bladder [28], and genetic and pharmacologic decreases of iNOS led to significantly less bladder fibrosis in mice and rats. This controversy demonstrates the complexity of the NO pathway in fibrosis of urological tissues and suggests that iNOS induction may not constitute a universal antifibrotic mechanism and, in certain cases, may actually be involved in inducing chronic and acute inflammation. This would reflect the yin-yang effects of NO [29], which would depend on the specific tissue of expression, the extent of oxidative stress, and the type of cellular environment. Nevertheless, the responsiveness of the PD plaque to iNOS gene transfer, the easy accessibility of the penis, and the relative persistence of plasmid iNOS expression could place this therapeutic approach as an attractive strategy for treatment of the PD fibrotic plaque.
In conclusion, these results showed that piNOS, when injected into the TA of the PD rat model, induced an early high expression of iNOS that declined but remained active during several weeks, as indicated by iNOS levels and nitrotyrosine accumulation as a marker of NO synthesis and its reaction with ROS. Therefore, this supports the assumption that iNOS/NO induces a decrease in collagen deposition and TA thickness in the rat penis and is associated with a decrease in TGFβ1 and PAI-1 expression, but further work is needed to demonstrate whether net ROS levels in the tissue are reduced after iNOS gene transfer. Our results further confirm the view that iNOS induction is an endogenous mechanism of defense against fibrosis in the PD plaque, and we propose that NO is a modulator of the remodeling phase of fibrosis in the TA of the rats. Nitric oxide may have a therapeutic role in human PD if the proper balance between its beneficial effects on ROS quenching (SOD) and its potentially deleterious oxidant effects is achieved, perhaps in combination with antioxidant therapy. However, both current animal models for PD may not necessarily reflect the complex development of the plaque within the human TA, and further studies are needed.
Acknowledgments
We would like to thank Dr. Robert Reiter and Dr. Isla Garraway for their kind support during performance of the bioluminescence protocol.
References
Author notes
Supported by NIH grants R01DK-53069 and G12RR-03026, the Edythe and Eli Broad Foundation, and the Peter Morton Foundation.
