Oxidative Stress and Neurodegeneration in the Ischemic Overactive Bladder Kazem M. Azadzoi,* Subbarao V. Yalla and Mike B. Siroky† From the Departments of Urology (KMA, SVY, MBS) and Pathology (KMA), Veterans Affairs Boston Healthcare System, Departments of Urology (KMA, MBS) and Pathology (KMA), Boston University School of Medicine and Department of Urology, Harvard University School of Medicine (SVY), Boston, Massachusetts
Purpose: The central and peripheral nervous systems are highly sensitive to ischemia and oxidative stress. We searched for markers of oxidative injury and examined neural density in the rabbit ischemic overactive bladder. Materials and Methods: Blood flow and oxygenation were recorded during cystometrogram in overactive and control rabbit bladders at weeks 8 and 16 after the induction of ischemia. Oxidative products and neural density were assessed by enzyme immunoassay and immunohistochemical staining, respectively. Reverse transcriptase-polymerase chain reaction was done to determine the gene expression of nerve growth factor and its receptor p75. The effect of acute oxidative stress was examined in tissue culture medium containing H2O2. Results: Overactivity produced repeating cycles of ischemia/reperfusion and hypoxia/reoxygenation in the ischemic bladder, leading to oxidative and nitrosative products. Neural density in the 8-week ischemic bladder was similar to that in controls, while neurodegeneration was evident after 16 weeks of ischemia. Nerve growth factor gene levels initially increased at week 8 but significantly decreased at week 16 after the induction of ischemia. Gene levels of p75 decreased after 8 weeks and remained lower than in controls after 16 weeks of ischemia. Acute oxidative stress decreased nerve growth factor protein release in culture medium. The antioxidant enzyme catalase had no significant effect on control tissues but it partially protected nerve growth factor from H2O2 injury. Conclusions: Ischemia may have a role in bladder neuropathy. Overactivity under ischemic conditions produces noxious oxidative products in the bladder. Neurodegeneration in bladder ischemia may involve a lack of nutrients, hypoxia and overactivity induced free radicals. Nerve growth factor and its receptors may regulate neural reactions to oxidative injury. Key Words: bladder; ischemia; oxidative stress; nerve degeneration; urinary bladder, overactive
oxidative and nitrosative stress as the leading cause of neuropathy in ischemic organs.5 Neuromuscular dysfunction, including muscular instability, efferent nerve excitation and unmyelinated afferent fiber activation, have been documented in the ischemic heart, jejunum and abdominal wall. In ischemia oxygen free radicals combine to form nitrosative species such as O ⫽ NOO⫺, a product of the O2⫺ and NO radical reaction.5 Oxidative and nitrosative species, and their products are known for their highly neurotoxic properties. Their involvement in neural functional deficit, neurodegeneration, structural damage and dysfunction has been documented in several organs, including the heart, penis and kidney.6,7 Antioxidants such as vitamin E and vitamin C have been used widely in clinical practice to protect the body from free radical injuries. We determined the impact of ischemia and oxidative stress on bladder innervation. We searched for markers of oxidative injury and examined neural density in the rabbit ischemic overactive bladder.
tudies of the ultrastructure of intrinsic human bladder innervation have shown axonal degeneration in nonneuropathic bladder dysfunction, and combined axonal degeneration, restricted regeneration and activated Schwann cells in neuropathic bladder dysfunction.1 The precise mechanism of neurodegeneration and the nature of neuromuscular interactions in the overactive bladder remain unknown. In previous studies we found that arterial obstructive disease, and subsequent ischemia and hypoxia led to overactive bladder.2– 4 The mechanism appeared to involve bladder innervation, muscarinic receptors, epithelium, and disproportionately increased prostaglandin and leukotriene production.2– 4 Ischemic bladder tissue was instable at resting tension and hyperreactive to electrical field neural stimulation.2– 4 Neural excitation and degeneration in ischemia are complex and multifactorial phenomena. Recent studies imply
Submitted for publication November 16, 2006. Supported by a Department of Veterans Affairs Merit Review Grant. * Correspondence: Surgery/Urology (151), 150 South Huntington Ave., Boston, Massachusetts 02130 (telephone: 857-364-5602; FAX: 857-364-4540; e-mail: [email protected]
). † Financial interest and/or other relationship with Astellas and Pfizer.
0022-5347/07/1782-0710/0 THE JOURNAL OF UROLOGY® Copyright © 2007 by AMERICAN UROLOGICAL ASSOCIATION
MATERIALS AND METHODS Overactive Bladder Model New Zealand White male rabbits weighing 3 to 3.5 kg were assigned into treatment and age matched control groups. In the treatment group atherosclerosis induced bladder isch-
Vol. 178, 710-715, August 2007 Printed in U.S.A. DOI:10.1016/j.juro.2007.03.096
OXIDATIVE STRESS AND NEURODEGENERATION IN ISCHEMIC OVERACTIVE BLADDER emia was created by partial de-endothelialization of the iliac arteries, as previously described.2– 4 To expedite arterial atherosclerosis the animals received a 0.5% cholesterol diet for 4 weeks, followed by a regular rabbit diet. Seven treated and 7 control animals were studied at week 8, and the remaining 7 treated and 7 control animals were studied at week 16 after the induction of bladder ischemia. Certain parameters were recorded. Bladder Blood Flow and pO2 In anesthetized animals the bladder was exposed and intravesical pressure was monitored with a 20 gauge catheter passed through the wall and connected to a transducer. The bladder was filled with 25 ml normal saline through a 6Fr Foley catheter. The catheter balloon was inflated to avoid leakage. The frequency of spontaneous bladder contractions was recorded. Bladder blood flow and pO2 were measured with a laser Doppler flowmeter and a polarographic oxygen sensing electrode, respectively, as previously described.2– 4 Five blood flow and pO2 measurements were obtained at random bladder sites in each animal. Subsequently the bladder was drained and the animals were sacrificed. Tissues from the bladder body and dome were processed for assays. Oxidative Products EIA of the oxidatively modified product isoprostane 8-epi PGF2␣ was performed, as previously reported.8 Briefly, ischemic and control bladder tissues were equilibrated for 2 hours in culture medium at 37C. The medium was exchanged with fresh medium every hour. After the last hour of incubation the levels of isoprostane 8-epi PGF2␣ in supernatant were assayed in triplicate with commercially available EIA kits (Cayman Chemical, Ann Arbor, Michigan). Microtiter assay plates were scanned with a SpectraMax® Plus 384 computer controlled microplate reader. The quantity of isoprostane 8-epi PGF2␣ was standardized as pg/100 mg wet weight of tissue per hour. Nitrosative Products Frozen bladder body and dome tissues were processed for cryostat sectioning and immunostaining using mouse antinitrotyrosine monoclonal antibody (Abcam®), diluted 1:100. Sections were incubated with primary antibodies for 1 hour and with biotinylated secondary antibodies for 20 minutes, and then processed by the avidin-biotin peroxidase complex with 3-amino-9-ethylcarbazole as the chromogen. Sections were counterstained with Mayer’s hematoxylin and observed under a light microscope. Negative controls were prepared in the same manner without primary antibody. The staining intensity of ischemic and control tissues was interpreted by 2 investigators in blinded fashion. Data were interpreted based on a scoring system that measured intensity as 0 —none, 1— borderline, 2—weak, 3—moderate, 4 —marked and 5—strong. Neural Density and NGF Expression Bladder body and dome tissues were processed for immunostaining, as described. Sections were incubated with 2 g/ml mouse anti-S-100 primary antibody for Schwann cells and 2 g/ml mouse anti-neurofilament 70 plus 200 primary antibody for myelinated nerve fibers (Research Diagnostics, Flanders, New Jersey) or with monoclonal anti-NGF anti-
body (Santa Cruz Biotechnology, Santa Cruz, California). The number of nerve fibers was counted at 100⫻ magnification in 10 high power fields per slide. NGF expression in 5 tissue preparations each in the 8 and 16-week ischemia, and control groups was interpreted by 2 investigators in blinded fashion based on the staining scoring system described. NGF and p75 Gene Expression Semiquantitative RT-PCR was used to examine the gene expression of NGF and its receptor p75. Total RNA (2 g) was isolated from 5 bladder tissue preparations each in the control, and 8 and 16-week ischemia groups using a Totally RNA kit (Ambion®). First strand cDNA was synthesized using SuperScript™ II for reverse transcription. RT was initiated at 25C. Reaction temperature was gradually increased to 55C at a rate of 1C per 20 seconds. The RT reaction was then diluted and stored at ⫺20C. PCR was used to determine the gene expression of NGF, p75 and GAPDH. Computer stored images of ethidium bromide stained agarose gels were analyzed by densitometry. NGF and p75 gene levels were calculated as the ratio of the corresponding intensity of the PCR product to that of GAPDH. NGF Reaction to Free Radicals To examine the free radical impact on NGF bladder tissues from control animals were equilibrated in culture media, as described. The medium was exchanged every 2 hours with fresh culture medium containing the free radical H2O2 (300 M), the antioxidant enzyme catalase (1,200 U/ml), H2O2 plus catalase or placebo (culture medium alone). After 12 hours the medium supernatant was used for EIA. The NGF protein level was assayed in triplicate with commercially available EIA kits (Promega™). Microtiter assay plates were scanned with a SpectraMax Plus 384 microplate reader. The quantity of NGF release was standardized as pg per 100 mg tissue per hour. Statistical Analysis Data are expressed as the mean ⫾ SEM. The mean values of measured parameters were compared across the 3 groups with ANOVA and post hoc pairwise comparisons of group means were then performed. Statistical significance was considered at p ⱕ0.05. RESULTS Oxidative Stress in the Overactive Bladder In control bladders spontaneous contraction caused mild decreases in blood flow and pO2, which returned to baseline levels after contraction (see table). Arterial ballooning produced atherosclerotic occlusive disease, causing significant decreases in bladder blood flow and pO2 (see table). In bladders with preexisting ischemia and hypoxia contraction decreased blood flow and pO2 to extremely low levels. This was followed by a phase of reperfusion/reoxygenation, characterized by sharp rebounds of blood flow and pO2 after contraction (see table). The frequency of bladder contractions, and concomitant blood flow and pO2 changes in the ischemic group were approximately 4-fold greater than in controls (see table).
OXIDATIVE STRESS AND NEURODEGENERATION IN ISCHEMIC OVERACTIVE BLADDER Cyclic changes in blood flow and pO2 during and after spontaneous bladder contraction Mean ⫾ SE Control
At rest: Blood flow (ml/min/100 gm) pO2 (mm Hg) At peak contraction: Blood flow (ml/min/100 gm) pO2 (mm Hg) Contraction end: Blood flow (ml/min/100 gm) pO2 (mm Hg) No. contractions/10 mins
Mean ⫾ SE Ischemia 8 wks
7.2 ⫾ 0.8 46 ⫾ 5
3.1 ⫾ 0.2 27 ⫾ 6
2.8 ⫾ 0.6 24 ⫾ 7
4.8 ⫾ 0.7 38 ⫾ 6
0.7 ⫾ 0.4 9 ⫾4
1.0 ⫾ 0.5 11 ⫾ 3
7.8 ⫾ 0.5 50 ⫾ 4 3 ⫾ 1.5
4.2 ⫾ 0.2 34 ⫾ 5 11.0 ⫾ 1.7
3.6 ⫾ 0.4 29 ⫾ 6 9.0 ⫾ 2.0
In 7 preparations per group ischemia/hypoxia was sustained and severe at rest and at peak contraction, respectively, and contraction frequency, ischemia/reperfusion and hypoxia/reoxygenation were significantly greater for ischemia vs control.
Marker of Oxidative Stress Isoprostane 8-epi PGF2␣ levels significantly increased in 8 and 16-week ischemic bladders (fig. 1). Isoprostane 8-epi PGF2␣ levels at 16 weeks of ischemia were slightly greater than at 8 weeks but the difference between the 2 ischemic groups was not statistically significant. Marker of Nitrosative Stress Immunohistochemistry showed no or mild positive nitrotyrosine staining (score 0 –1) in control bladder tissues (fig. 1). Diffuse intense staining (score 4 –5) for nitrotyrosine was evident in 8 and 16-week ischemic bladders (fig. 1). Nitrotyrosine staining intensity in 8-week ischemic tissues was comparable to that in 16-week ischemic tissues. Neurodegeneration Neural density in the 8-week ischemic bladder was slightly less than in controls but the difference was not significant. Marked neurodegeneration was evident in the 16-week ischemic bladder. The number of nerve fibers per high power field in the 16-week ischemic bladder was significantly less than control levels (fig. 2). Progressive thickening was evi-
dent in nerve fibers that survived 8 and 16 weeks of ischemia. Detailed structural and ultrastructural studies of ischemic nerves remain the focus of our future studies. Changes in NGF and p75 Gene Levels RT-PCR showed a significant increase in NGF gene levels and a significant decrease in its receptor p75 at the initial stage of ischemia at week 8 (fig. 3). However, after 16 weeks of ischemia NGF and p75 levels significantly decreased to lower than control levels (fig. 3). Changes in NGF Expression Immunohistochemical analysis showed sporadic areas of positive NGF staining in control and ischemic bladder tissues (fig. 3). Immunoreactive areas in bladder tissue progressively decreased in 8 and 16-week ischemic bladders compared with those in controls (score 2–3 and 0 –1, respectively, vs 4 –5). NGF Reaction to Acute Oxidative Stress In tissue culture acute exposure to H2O2 decreased bladder NGF protein levels (fig. 4). This effect of H2O2 was partially blocked by the antioxidant enzyme catalase. NGF release from H2O2 plus catalase treated tissues was greater than the release from those treated with H2O2 but lower than control levels (fig. 4). Catalase alone had no significant effect on NGF protein levels. DISCUSSION Our previous studies showed that the bladder becomes overactive when nutrients and oxygen are lacking due to arterial insufficiency.2– 4 Ischemia altered bladder muscarinic receptor reactivity, increased neurogenic contractions and led to smooth muscle instability at resting tension.2– 4 Long-term ischemia up-regulated constrictor eicosanoid production and ultimately led to epithelial structural damage, subepithelial fibrosis and decreased bladder compliance.2– 4 Other investigators reported that obstruction-associated functional changes closely correlated with decreased bladder blood
FIG. 1. Levels of oxidatively modified product isoprostane 8-epi PGF2␣ and nitrosative product nitrotyrosine significantly increased 8 and 16 weeks after induction of bladder ischemia vs controls in 7 preparations each. There was no significant difference between 2 ischemic groups. Top, data are presented as mean ⫾ SE. Asterisk indicates significant difference in ischemic groups. Bottom, reduced from ⫻40.
OXIDATIVE STRESS AND NEURODEGENERATION IN ISCHEMIC OVERACTIVE BLADDER
FIG. 2. Top, immunostaining of nerve fibers with anti-S-100 and anti-neurofilament 70 plus 200 primary antibodies is shown at ⫻100 magnification. Number of myelinated nerve fibers progressively decreased during bladder ischemia course. Nerve fibers (arrows) that survived 8 and 16 of weeks ischemia appeared markedly thickened vs those in controls. Bottom, there was significant decrease in number of nerve fibers after 16 weeks of ischemia. Data are presented as mean ⫾ SE. Asterisk indicates significant difference in 16-week ischemia group.
flow.8 Others found that calcium dependent proteolysis of calpain substrate ␣-spectrin preceded reperfusion injury in the partially obstructed bladder.9 Our current study suggests that overactivity under the ischemic condition may also involve noxious elements in the bladder wall that accumulate by oxidative and nitrosative stress. The mechanism of oxidative stress in bladder ischemia appears to involve recurrent hemodynamic disturbances, and rapidly fluctuating perfusion and oxygenation of the frequently contracting wall. While reperfusion and reoxygenation generate free radicals, antioxidant deficiency and malfunctioning antioxidant enzymes in ischemia and hypoxia could let the deleterious process continue unchecked. The marker of oxidative stress isoprostane 8-epi PGF2␣ was detected 8 and 16 weeks after the induction of bladder
ischemia. Oxidatively modified products, including isoprostane 8-epi PGF2␣ and eicosanoids, are known to interfere with bladder smooth muscle contraction,3,10 suggesting that free radicals generated by overactivity may result in further bladder instability. The formation of free radicals in normal smooth muscle cells and neurons is tightly regulated by homeostatic mechanisms. Biological antioxidants react with oxidant free radicals to help detoxify them. In pathological conditions such as ischemia the production of free radicals exceeds the antioxidant capacity of a cell. Excessive free radicals induce destructive changes in cellular and subcellular components, including lipids, protein and DNA, ultimately leading to cellular degeneration.11 Positive areas of nitrotyrosine detected immunohistochemically imply reactive nitrogen species and nitrosative
FIG. 3. Top, changes in ratio of specific gene product and GAPDH 8 and 16 weeks after induction of bladder ischemia vs controls in 5 preparations each. Ratios from 5 PCR reactions were averaged. NGF expression increased after 8 and decreased after 16 weeks of ischemia. Gene levels of p75 decreased after 8 weeks and remained lower than in controls after 16 weeks. Data are presented as mean ⫾ SE. Asterisk indicates significant changes in ischemic groups. Bottom, immunohistochemical staining reveals progressive decrease in NGF expression after 8 and 16 weeks of ischemia vs controls in 5 preparations each. Reduced from ⫻100.
OXIDATIVE STRESS AND NEURODEGENERATION IN ISCHEMIC OVERACTIVE BLADDER
FIG. 4. NGF protein release from control tissues decreased after acute exposure to H2O2. Catalase had no significant effect but appeared to partially protect NGF from H2O2 injury. Data are presented as mean ⫾ SE. Asterisk indicates significant difference vs control. Plus sign indicates significant difference vs control and H2O2.
stress in the ischemic bladder. Nitrosative stress occurs when oxygen free radicals accumulate in an NO containing environment.12 It was reported that O2⫺ initiates the process by interacting with NO to form the potent oxidant O ⫽ NOO⫺.12 Degradation of O ⫽ NOO⫺ induces the formation of cytotoxic radicals such as nitrite and hydroxyl radicals, leading to nitrosylation reactions. O ⫽ NOO⫺ initiates chain reaction byproducts that are damaging to cellular and subcellular components. O ⫽ NOO⫺ induces rapid oxidative injury to smooth muscle cells, nerves and vessels since it is approximately 1,000-fold more potent as an oxidizing compound than H2O2.13 Neurodegeneration has been documented in human neurogenic bladder dysfunction.1 In our model bladder overactivity emerged before neurodegeneration at week 8 after ischemia, suggesting mechanisms other than nerve fiber death at the initial stages of disease. Loss of nerve fibers in the ischemic bladder seemed to develop as an end stage phenomenon after 16 weeks. This may imply an initial intrinsic neural defensive mechanism against ischemic and oxidative injury, which seems to fail after prolonged ischemia, possibly due excessive accumulation of neurotoxic products. Ischemia may affect nerve fibers directly by nutrient deficiency and hypoxia, and indirectly by generating free radicals. The role of nerve fiber ischemia in neuropathy and its impact on muscular reflexes have been documented in humans14 and in rats with experimental diabetic neuropathy.15 These studies suggest that nerve fibers and microvessels closely communicate to assemble neural and vascular protective mechanisms.14,15 Vascular proliferation and increased vascular density in response to Wallerian degeneration was shown to induce partial nerve regeneration.16 The mechanism of nerve regeneration after spinal cord injury in rats was shown to involve vascular growth factors and angiogenesis.16,17 Nerve graft experiments have shown that angiogenesis precedes axonal regeneration.16,17 New neural outgrowth and Schwann cells were found to be more abundant in areas with dense blood vessels.16,17 Following spinal cord injury peripheral nerve angiogenesis was found to precede nerve fiber regeneration.17 The role of oxidative injury in neuropathy and neurodegeneration has been documented in human and experimental models.18 Studies in mice have shown that the loss of antioxidant genes can lead to neurodegeneration. Markers
of oxidative stress have been found in autopsy examination of brain tissue from patients with neurodegenerative disorders. Hallmarks of oxidative stress, such as lipid peroxidation, nucleic acid and protein oxidation, and changes in some antioxidants have been noted in dopaminergic neurons of samples from patients with Parkinson’s disease. The involvement of oxidative stress in the pathogenesis of age related neurodegenerative disorders has also been reported.18 Bladder neural reactions to ischemia and oxidative stress may involve multiple neurovascular, cellular and subcellular mechanisms. Our RT-PCR data suggest NGF sensitivity to ischemia and oxidative stress, and its possible role in neurodegeneration. NGF gene levels initially increased in response to ischemia and oxidative stress at week 8 but they significantly decreased after long-term ischemia at week 16. The initial up-regulation of NGF may be an intrinsic defensive mechanism to protect the bladder nerves from ischemia and oxidative injury. However, this physiological maneuver was ineffective since NGF receptor levels decreased immediately after the induction of ischemia. Impaired receptors may explain the lack of new neural outgrowth despite increased NGF gene levels at 8 weeks of ischemia. Immunohistochemical analysis showed similar NGF staining intensity in 8-week ischemic samples compared with controls, suggesting that increased transcript levels may not be capable of stimulating NGF protein synthesis under the ischemic condition. This implies that, while the NGF gene level increases in response to ischemia, protein synthesis fails due to a lack of nutrients and oxygen. After prolonged ischemia at 16 weeks decreased transcript levels correlated with decreased NGF staining and neurodegeneration. NGF has been shown to act positively on axonal regeneration and outgrowth, and partially reverse neuropathy, including sensory deficit, myelin degeneration and transmitter functional changes.19,20 This action of NGF during nerve repair has been noted to involve the induction of angiogenic factors in Schwann cells and neurons.19 An interaction between vascular endothelial growth factor and NGF to protect nerve fibers against diabetic neuropathy has been documented.20 Our studies in tissue culture confirmed NGF sensitivity to free radicals, showing a significant decrease in its protein levels after exposure to H2O2. The antioxidant enzyme catalase appeared to have no significant effect on control tissues, while providing partial protection to NGF from H2O2 injury (fig. 4). NGF protein release from catalase plus H2O2 treated tissues was significantly greater than from tissues treated with H2O2, suggesting a prophylactic role of antioxidants against bladder oxidative injury (fig. 4).
CONCLUSIONS Our data suggest that long-term arterial insufficiency leads to bladder neuropathy via ischemia, hypoxia and oxidative stress. Bladder overactivity under the ischemic condition generates neurotoxic oxidative and nitrosative products. Neurodegeneration may be an end stage phenomenon in ischemia induced bladder overactivity. NGF and its receptors may regulate neural reactions to bladder ischemia and oxidative stress.
OXIDATIVE STRESS AND NEURODEGENERATION IN ISCHEMIC OVERACTIVE BLADDER
Abbreviations and Acronyms EIA ⫽ enzyme immunoassay GAPDH ⫽ glyceraldehyde-3-phosphatasedehydrogenase NGF ⫽ nerve growth factor NO ⫽ nitric oxide O ⫽ NOO⫺ ⫽ peroxynitrite PCR ⫽ polymerase chain reaction PGF2␣ ⫽ prostaglandin F2␣ pO2 ⫽ oxygen tension RT ⫽ reverse transcriptase REFERENCES 1.
Haferkamp A, Dorsam J, Resnick NM, Yalla SV and Elbadawi A: Structural basis of neurogenic bladder dysfunction. III. Intrinsic detrusor innervation J Urol 2003; 169: 555. Azadzoi KM, Tarcan T, Kozlowski R, Krane RJ and Siroky MB: Overactivity and structural changes in the chronically ischemic bladder. J Urol 1999; 162: 1768. Azadzoi KM, Shinde V, Tarcan T, Kozlowski R and Siroky MB: Increased leukotrienes and prostaglandins production and overactivity in the chronically ischemic bladder. J Urol 2003; 169: 1885. Azadzoi KM, Heim VK, Tarcan T and Siroky MB: Alteration of urothelial-mediated tone in bladder ischemia: Role of eicosanoids. Neurourol Urodyn 2004; 23: 258. Sayer LM, Smith MA and Perry G: Chemistry and biochemistry of oxidative injury in neurodegenerative disease. Curr Med Chem 2001; 8: 721. Azadzoi KM, Schulman R, Aviram M and Siroky MB: Oxidative stress in arteriogenic erectile dysfunction: prophylactic role of antioxidants. J Urol 2005; 174: 386. Aragno M, Cutrin JC, Mastrocola R, Perrelli MG, Restivo F, Poli G et al: Oxidative stress and kidney dysfunction due to ischemia/reperfusion in rat: attenuation by dehydroepiandrosterone. Kidney Int 2003; 64: 836. Tarcan T, Krane RJ, Siroky MB and Azadzoi KM: Isoprostane 8-epi PGF2 alpha, a product of oxidative stress, is synthe-
sized in the bladder and causes detrusor smooth muscle contraction. Neurourol Urodyn 2000; 19: 43. Schroder A, Chichester P, Kogan BA, Longhurst PA, Das AK and Levin RM: Effect of chronic bladder outlet obstruction on blood flow of the urinary bladder. J Urol 2001; 165: 640. Zhao Y, Levin SS, Wein AJ and Levin RM: Correlation of ischemia/reperfusion or partial outlet obstruction-induced spectrin proteolysis by calpain with contractile dysfunction in rabbit bladder. Urology 1997; 49: 293. Voss P and Siems W: Clinical oxidation parameters of aging. Free Radical Res 2006; 40: 1339. Darley-Usmar V and White R: Disruption of vascular signaling by the reaction of nitric oxide with superoxide: implication for cardiovascular disease. Exp Physiol 1997; 82: 305. Xie Z, Wei M, Morgan TE, Fabrizio P, Han D, Finch CE and Longo VD: Peroxynitrite mediates neurotoxicity of amyloid beta-peptide-42 and lipopolysaccharide-activated microglia. J Neurosci 2002; 22: 3484. Theriault M, Dort J and Sutherland G: Local human sural nerve blood flow in diabetic and other polyneuropathies. Brain 1997; 120: 1131. Tuck RR, Schmelzer JD and Low PA: Endoneurial blood flow and oxygen tension in the sciatic nerves of rats with experimental diabetic neuropathy. Brain 1984; 107: 938. Hobson MI, Brown R and Green CJ: Interrelationship between angiogenesis and nerve regeneration. Br J Plast Surg 1997; 50: 125. Bartholdi D, Rubin BP and Schwab ME: VEGF mRNA induction correlates with changes in the vasculature upon spinal cord damage in the rat. Eur J Neurosci 1997; 9: 2549. Kim JI, Choi SI, Kim NH, Jin JK, Choi EK and Carp RI: Oxidative stress and neurodegeneration in prion disease. Ann N Y Acad Sci 2001; 928: 182. Apfel SC, Kessler JA, Adornato BT, Litchy WJ, Sanders C and Rask CA: Recombinant human nerve growth factor in the treatment of diabetic polyneuropathy. NGF Study Group. Neurology 1998; 51: 695. Unger JW, Klitzsch T, Pera S and Reiter R: NGF and diabetic neuropathy in the rat. Exp Neurology 1998; 152: 1.