Nitric oxide synthase gene transfer for erectile dysfunction in a rat model


M.B. Chancellor, MD, Suite 700 Kaufmann Building, 3471 Fifth Avenue, Pittsburgh, PA 15213, USA.



To determine whether over-expression of nitric oxide synthase (NOS) in the corpus cavernosum of the penis improves erectile function, as NO is an important transmitter for genitourinary tract function, mediating smooth muscle relaxation and being essential for penile erection.


The inducible form of the enzyme NOS (iNOS) was introduced into the corpus cavernosum of adult Sprague-Dawley rats (250–300 g) by injecting a solution of plasmid, adenovirus or adenovirus-transduced myoblast cells (adeno-myoblasts). Plasmid, adenovirus and adeno-myoblasts encoding the expression of the β-galactosidase reporter gene were also injected into rats.


Throughout the corpora cavernosum there was expression of β-galactosidase after injecting each of the three solutions. Maximum staining was greatest for adeno-myoblast, then adenovirus and then plasmid. The mean (sd) basal intracavernosal pressure (ICP) of iNOS-treated animals (adenovirus and adeno-myoblast) increased to 55 (23) cmH2O, compared with naive animals with a basal ICP of 5 (6) cmH2O (P = 0.001). Stimulating the cavernosal nerve (15 Hz, 1.5 ms, 10–40 V, 1 min) resulted in a doubling of the ICP (adenovirus and adeno-myoblast) from the basal level of the iNOS-treated animals. Direct in situ measurement of NO showed the release of 1–1.3 µmol/L in the adeno-myoblast penis.


Myoblast-mediated gene therapy was more successful for delivering iNOS into the corpus cavernosum than direct adenovirus injection or plasmid transfection. Surprisingly, implanting muscle cells into the penis is not only feasible but also beneficial. Gene therapy for NOS may open new avenues of treatment for erectile dysfunction. Control of NOS expression would be necessary to prevent priapism.


Penile erection depends on the integration of vascular, endocrine and neurological mechanisms. Neurally mediated vascular dilatation and relaxation of penile smooth muscle leads to increased blood flow and erection of the cavernosal tissue [1,2]. An essential mediator of penile erection is nitric oxide (NO), the hydrolysis product of l-arginine by NO synthase (NOS). NO operates locally as a postganglionic neurotransmitter of nonadrenergic, noncholinergic-mediated penile erection. Additionally, co-transmitters such as acetylcholine and bradykinin may also induce this neurotransmitter [2]. After release into the corpus cavernosum, NO diffuses to the neighbouring vasculature and trabecular smooth musculature of the penis, where it stimulates the activity of the enzyme guanylyl cyclase. The formation of cGMP results in decreased intracellular Ca2+ levels and thereby induces smooth muscle cell relaxation with subsequent penile vasodilatation [3–5].

Any disease or dysfunction affecting the brain, spinal cord, cavernosum and pudendal nerves or receptors in the terminal arterioles and cavernosal smooth muscle can result in penile erectile dysfunction or impotence. Vascular disease and neuropathies resulting from diabetes mellitus are frequently underlying causes of impotence in younger men. Chronic smoking diminishes NO-mediated vascular relaxation and induces vasospasm of penile arteries; long-term smoking also causes structural damage to the corporal tissue [3,6]. Commonly prescribed pharmacological agents such as H2-receptor antagonists, β-blockers and corticosteroids, also produce impotence. Although impotence is not part of the normal ageing process, erectile dysfunction commonly increases with advancing age [7].

In rat, the functional isoform of NOS in penile tissue has been reported as either an endothelial constitutive-type (ecNOS), a Ca2+ calmodulin-dependent, EGTA-sensitive form [3] or an inducible form (iNOS) with a distinctive expression pattern and Ca2+-dependence [8]. Neuronal NOS appears to have a minimal role, as transgenic animals lacking this NOS were fertile with normal erectile function [8]. In mature animals, NOS was expressed in the nerve terminals of the corpus cavernosum, while in ageing animals the dorsal nerve carried fewer NOS-containing nerve fibres and with a changed distribution [8,9]. Furthermore, both the frequency and duration of penile erection is diminished after castration, although hormone replacement restores the process [2].

The objective of the present study was to determine whether over-expression of iNOS in the corpus cavernosum of the penis would correct erectile dysfunction. As NO, the product of NOS, is essential for erection, the introduction of the enzyme may result in improved erectile function. Three methods of delivering the iNOS gene were used, i.e. plasmid transfer, adenovirus-mediated transfer and somatic cell injection. We chose iNOS because this enzyme produces NO in a sustained manner and thus allows the establishment of the ‘proof of principle’ for iNOS and ex vivo gene therapy for treating erectile dysfunction.


Three different methods of gene transfer were assessed; because of concerns with transfection efficiency using direct plasmid injection, and safety issues with the adenovirus vector, we also included the ex vivo method of gene transfer by first infecting myoblasts with the iNOS vector in culture and then injecting the myoblast expressing iNOS into the penis.

Plasmids with β-galactosidase with and with no mouse iNOS were purified by SDS lysis and gel filtration. Yields were 2–10 mg and depended on the copy number (courtesy of Dr Bruce Pitt, University of Pittsburgh). Briefly, competent Escherichia coli DH5 cells were transformed with 50 ng plasmid DNA and plated. A single colony was used to start an inoculum. Cells were harvested from 1 L of broth culture and lysed in 0.2 mol/L NaOH/2% SDS. After neutralization with acetic acid, proteins were removed with isopropanol precipitation. Nucleic acids were resuspended in 2.5 mol/L ammonium acetate in Tris-EDTA, pH 8, and applied to a gel filtration column. Plasmid DNA migrated in the void volume.

Adenovirus or plasmid injection. Injections with 20–100 µL adenovirus with a viral titre of 106−109 were delivered into the sinusoid of the corpus cavernosum (adenovirus iNOS vector kindly provided by Dr Tim Billiar, University of Pittsburgh) [10]. Different adult male Sprague Dawley rats were injected with β-galactosidase with or with no iNOS in an adenovirus vector [10]. Vehicles for injection included PBS, polybrene (8 µg/mL in PBS, Sigma Chem Co., St Louis, MO) or 20% sucrose. Injections of 10, 50 or 100 µg purified plasmid/100 µL of 20% sucrose were delivered into each corpora; the tissue was then harvested after 2, 4 or 7 days.

Myoblast preparation. Myoblasts from the mouse MDX cell line were plated at a density of 5 × 104 per well in a six-well plate, and after 24 h the cells were washed with Hank's balanced salt solution. Sufficient iNOS adenovirus stock was allocated into each well to achieve a multiplicity of infection (MOI) of 50. Plates were incubated for 2 h at 37 °C to allow infection; proliferation medium was then added to each well. Plates were placed at 34 °C for 48 h. Cells were immunostained for expression or harvested in Hank's balanced salt solution for injection, which were delivered at 1.33 × 105 to 1 × 106 cell per 10 µL.

Injection of the corpus cavernosum. All experiments were conducted on mature male Sprague Dawley rats (250–400 g) in accordance with the requirements and recommendations in the Guide for the Care and the Use of Laboratory Animals (USA Public Health Service, NIH Publication no. 85–23, 1985). Animals were certified virus-free and individually housed in an approved viral gene therapy facility. Before injection animals were anaesthetized with ketamine/rompun (45 mg/kg and 3.0 mg/kg, respectively). A midline incision was made to expose the penis, and blood flow to the penis occluded by securely tying a length of cotton string at the base of the organ. Injections were delivered into the corpus cavernosum with a 28 G disposable syringe. The ligature was removed after 30 min and the incision closed. Animals were routinely given ampicillin (2 units/200 µL, intramuscular).

Intracavernosal pressure (ICP) measurement. At 2, 4 or 7 days after injection animals were anaesthetized with nembutal (8 mg/kg, intraperitoneal and a 25 G needle inserted into the proximal corpus cavernosum with connections to a pressure transducer. The basal pressure was measured, followed by measurements of the maximum pressure after stimulating the efferent neural excitatory pathway at the level of the pelvic ganglion, at 15 Hz for 1.5 ms and 10–40 V for 1 min.

Tissue was harvested at 2, 4 or 7 days after injection and snap-frozen using 2-methylbutane pre-cooled in liquid nitrogen. Analysis of the sections included haematoxylin and X-gal staining. The β-galactosidase reporter gene was used to determine the location and function of the gene transfer area. The area around each injection site was stained, examined microscopically and photographed. β-galactosidase is a convenient ‘reporter gene’ that can be stained blue on histochemical staining. Only a virus that is biologically active can produce the β-galactosidase protein and this activity can be easily detected using the LacZ stain in the X-gal substrate. β-galactosidase is a widely used reporter gene to document not only where the plasmid, virus or myoblast resides, but also to confirm that the gene is transferred.

The cryostat sections of the injected tissue were stained for LacZ expression as follows. They were first fixed with 1.5% glutaraldehyde (Sigma) for 1 min and rinsed twice in PBS, then finally incubated in X-gal substrate (0.4 mg/mL 5-bromo-chloro-3-indolyl-β-d-galactoside, Boehringer-Mannheim, Indianapolis IN, USA), 1 mmol/L MgCl2, 5 mmol/L K4Fe(CN)6 and 5 mmol/L K3Fe(CN)6 in PBS overnight (37 °C).

The expression for iNOS from myoblasts was explored with direct iNOS staining of the injected myoblasts. Myoblasts were fixed with 4% paraformaldehyde for 20 min and permeabilized with cold methanol at − 20 °C for 10 min. Wells were blocked and reagents diluted with 1% BSA in PBS. Diluted anti-iNOS antibody (1 : 250, rabbit polyclonal, Santa Cruz Labs, CA, USA) was incubated at room temperature for 60 min. Biotinylated antimouse antibody was incubated for 30 min at room temperature, followed by incubation with streptavidin-Cy3 for 30 min.

Direct NO release was measured using a porphyrinic microsensor previously used to measure NO release in bladder strips and dorsal root ganglion cells [11]. Dissociated myoblasts or dissected corpora cavernosum were attached to glass coverslips and placed in a chamber mounted on the stage of an inverted microscope, to measure endogenous NO release (Fig. 1). All procedures were performed in Ringer's solution that contained (in mmol/L): 140 NaCl, 5.2 KCl, 1.2 MgSO4, 1 CaCl2, 10 NaH2PO4/NaHPO4 and 10 glucose, pH 7.4. The tip of a porphyrinic microsensor, which specifically detects NO, was placed directly on the cell or in the penis. Drugs were added to the perfusate or applied locally using a nano-ejector (1–3 µL). As a control, all agents were applied with no cells present to determine the baseline and with cells present to ensure that perfusate flow did not cause cellular disruption or release of intracellular substances that might induce synthesis of NO. The NOS cofactor tetrahydrobiopterin (H4B, Sigma) was added selectively to myoblast cell cultures and penis during physiological measurement, and during NO microsensor measurement (100 µmol/L).

Figure 1.

Illustration of the nitric oxide porphyrinic microsensor that can measure NO release from cells and tissue. The microsensor (diameter 5 µm, detection limit 1 nmol/L; response time 1 ms) is placed directly on the cell wall of individual cultured myoblasts.

ICPs were compared using Student's t-test and expressed as the mean (sem); P < 0.05 was considered to indicate statistical significance.


Control studies were conducted to determine whether the operative procedure or injection into the corpora affected normal erectile function. After anaesthesia, mature male rats (250–300 g) were sham-operated and received a saline injection of 100 µL, which filled the entire corpora. Furthermore, occlusion of blood flow by securing a length of cotton cord at the base of the penis reduced the blood flow for 30 min with no apparent trauma to the organ. Four days after the operation both the basal and maximal ICP after direct neural stimulation were measured.

In normal and sham-operated animals there was no change in the maximum ICP or the time to attain it between naive (untreated, seven) and the sham-operated animals (five). Staining for both β-galactosidase and iNOS were negative. Similar results of an unchanged ICP and rate of attainment were obtained with the reporter plasmid and adenovirus with the reporter gene (five).

Both the β-galactosidase and iNOS (mouse) plasmids were purified from transfected E. coli. Several levels of purified plasmid with the bacterial β-galactosidase (10, 50 and 100 µg) were injected into the corpus cavernosum and tissue harvested at 2, 4 or 7 days afterward. Of six animals injected with the plasmid, only one (100 µg at 2 days) was positive for expression of β-galactosidase. Experiments with the iNOS plasmid also showed a similar positive staining pattern at 4 days (three), which was greater than that at 2 days (three); the minimum staining was at 7 days (three).

Adenovirus vector levels as low as 106 plaque-forming units per organ resulted in expression of the β-galactosidase gene. The level of viral particles, the concentration, volume and excipient components or vehicles of the injected solution were varied. The level of infection appeared to be unaffected by the presence of excipients, e.g. 20% sucrose or polybrene. Injection volume affected the area of expression, with greater expression when the volume of the corpora was ‘filled’. Based on the intensity of X-gal staining, the level of expression was time-dependent, with maximum expression at 4 days (three). The level of expression was lower at 2 days (three) with the minimum expression at 7 days after injection (three) (Figs 2 and 3).

Figure 2.

Gross photography of rat penis after staining for β-galactosidase. The intensity of blue correlates with lac-Z staining. Left, penis myoblast plus adenovirus; middle, control saline injection; right, adenovirus injection. Tissue was harvested 4 days after transfection; the volume of injection was 100 µL in all animals.

Figure 3.

Cross-section of corpus cavernosum of a rat after injection with adenovirus plus β-galactosidase. Adenovirus (109 pfu) was injected into the corpus cavernosum in 100 µL of PBS; tissue was harvested 4 days later, fixed, stained for β-galactosidase with X-gal (blue within tissue stained pink with haematoxylin) and sectioned. One complete corpora is shown.

Transformed myoblasts were infected with adenovirus/β-galactosidase at a MOI of 50 (five). Tissue was harvested 4 days after injection and B-galactosidase staining of the tissue was strongly positive. Immunostaining of tissue after injection with iNOS-transduced myoblasts showed positive staining for antigens from both the adenovirus and the human iNOS. Staining was positive for up to 10 days while in culture (Fig. 4). The porphyrinic microelectrode showed that transduced cells had greater production of NO from calcium-independent NOS (i.e. iNOS) than did naïve cells, after adding 100 µmol/L H4B to the culture media. The NO release in the rat penis treated by iNOS transduced myoblasts was 20 µmol/L more than in the naive penis (three each; Fig. 5). The concentrations of NO recorded from control myoblasts, myoblasts transfected with iNOS, control + H4B and transfected myoblast + H4B were 11, 62, 14 and 205 µmol/L, respectively. Continuous NO production (from iNOS) did not require the presence of an agonist.

Figure 4.

Immunohistochemistry staining for iNOS in virally transduced myoblasts. Some but not all of the myoblasts and myotubules were positively stained for the polyclonal iNOS antibody with light background staining (MOI = 50).

Figure 5.

Application of noradrenaline (0.5 µmol/L) evoked transient NO release from untreated myoblasts; right, arrow indicates continuous NO release in myoblasts transfected with iNOS. The electrode was raised off the cell surface (with an immediate decrease in NO measurement) to show constant release by the transfected myoblasts.

Myoblasts expressing iNOS were injected into the corpora of rats as described for those expressing β-galactosidase; the behaviour of the rats was unaffected by the injections, and 2 days later the ICP was measured. The basal and maximum ICPs were significantly higher in the rats treated with adeno-myoblast + iNOS than in the control and myoblast-alone (P = 0.001; Table 1). Intact animals were not treated with exogenous H4B.

Table 1.  Basal and maximum ICP on cavernosal nerve stimulation in control rats, rats receiving the reporter gene, and rats injected with adeno-myoblast iNOS
GroupMean (sem) ICP, cmH2O
  • *

    P  < 0.05 for adeno-myoblast iNOS vs control and reporter gene rats.

Control (8)  5 (6)  85 (28)
Myoblast + reporter gene (6)  8 (3)  78 (31)
adeno-myoblast + iNOS (4)55 (23)*158 (36)*


These experiments were devised to explore a potential long-lasting therapy for erectile dysfunction, by addressing the physiology underlying erection, i.e. enhancing NO production through gene therapy. The preliminary results are promising, with an enhanced ICP in both the basal state and after stimulation.

iNOS gene transfer to the penis in this rat model was successful, with significant increases in physiological response. Among the three techniques of iNOS gene therapy tested, myoblast cell-mediated ex vivo gene therapy was superior to adenovirus or plasmid injection. We recognize that iNOS gene transfer could cause priapism and will probably not be clinically appropriate unless the expression can be titrated and/or regulated with an ‘on-off’ gene switch.

Adult Sprague-Dawley male rats were used as the in vivo model for these studies as we have experience with this species, and a continuing interest in investigating the role of NOS in visceral neurones [12–16]. Effects of ageing on androgen levels, the reproductive system and on behaviour have been well characterized in this small animal model [17–19]. Pharmacological responses of cavernosal tissue to papaverine and prostaglandin-E1 resemble those in humans [20].

As the corpus cavernosum is highly perfused, as are the lungs, liver and kidney, materials that are injected into the corpora rapidly enter the venous circulation. A significant percentage of the adenovirus could be carried into the portal circulation of the liver, where a viraemia could result. As an alternative to in vivo injection of virus, we explored the possibility of cellular-based ex vivo gene transfer that could minimize host expression to the virus vector. The grafting of myocytes after genetic manipulation in culture could be an alternative to direct viral or plasmid injection [21–23]. The NO microsensor and physiological pressure experiments showed that the MDX myoblast cell lacked the H4B cofactor and that adding it is necessary for the iNOS to produce NO in the myoblast culture. However, the rat penis showed in both the ICP and NO-sensor studies that the whole penis has enough endogenous H4B to allow full expression of iNOS. This has important clinical implications that will allow unrestricted NO production in cases of NOS gene therapy [24,25].

There are two recent reports of NOS gene therapy for erectile dysfunction. Garban et al.[17] reported the cloning of rat and human iNOS for penile injection, and Champion et al.[26] showed the feasibility of gene transfer of ecNOS for augmenting erectile responses in the aged rat. One unique feature of the present study is the transfer of a different NOS isoform, iNOS or type II NOS, to penile cells such as endothelial or smooth muscle cells. Distinct from the ecNOS, iNOS synthesises much more NO in a sustained manner and functions independently of intracellular calcium fluxes. Therefore, even if iNOS is expressed in relatively few cells, a wide area of cells could be exposed to NO independent of external agonists needed by the ecNOS to stimulate calcium fluxes.

Because of the technical limitations associated with penile gene-transfer strategies, a gene effective when expressed in low abundance could have distinct advantages. In these respects, iNOS could have important advantages over ecNOS as a therapeutic tool for treating impotence.

An additional and surprising interesting finding of this study was that implanting muscle cells (myoblasts) into the penis is not only feasible but also effective. This provides proof of the concept that muscle-based cellular therapy to augment the muscle within the penis is feasible. The functional consequence of this study showed that gene transfer of iNOS, regardless of the transfection method, can dramatically improve erection. Up-regulation of NOS can improve penile tumescence and may be a treatment for erectile dysfunction from vascular and neurogenic causes.

In conclusion, iNOS gene and muscle transfer to the rat penis was successful, with an increased physiological response. Among the three techniques of iNOS gene therapy tested, myoblast cell-mediated gene therapy was superior to plasmid or adenovirus injection. Gene therapy of NOS is a potentially promising cure for the underlying pathophysiology of erectile dysfunction, if the NO expression can be controlled.


We thank Dr Bruce Pitt from the Department of Pharmacology kindly provided iNOS plasmid and Dr Timothy Billiar and Dr Richard Shapiro of the Department of Surgery at the University of Pittsburgh School of Medicine kindly provided adenovirus iNOS. Supported by: NIH DK55045; DK55387; HD39768


nitric oxide


inducible/endothelial constitutive NO synthase


multiplicity of infection


intracavernosal pressure