The production of nerve growth factor by human bladder smooth muscle cells in vivo and in vitro


  • R. Tanner,

    1. Department of Surgery, School of Surgical and Reproductive Sciences, The Medical School, University of Newcastle, Newcastle upon Tyne, UK
    Search for more papers by this author
  • P. Chambers,

    1. Department of Surgery, School of Surgical and Reproductive Sciences, The Medical School, University of Newcastle, Newcastle upon Tyne, UK
    Search for more papers by this author
  • M.H. Khadra,

    1. Department of Surgery, School of Surgical and Reproductive Sciences, The Medical School, University of Newcastle, Newcastle upon Tyne, UK
    Search for more papers by this author
  • J.I. Gillespie

    1. Department of Surgery, School of Surgical and Reproductive Sciences, The Medical School, University of Newcastle, Newcastle upon Tyne, UK
    Search for more papers by this author

Professor J.I. Gillespie, Department of Surgery, School of Surgical and Reproductive Sciences, The Medical School, University of Newcastle, Framlington Place, Newcastle upon Tyne, NE2 4HH, UK. E-mail:


Objective To measure the concentrations of nerve growth factor (NGF) in tissue biopsies taken from subjects with a normal bladder and from patients diagnosed to have idiopathic detrusor instability (associated with a reduction in the density of motor nerves), and to use an in vitro model to study the mechanisms of NGF expression.

Materials and methods Biopsy specimens were obtained during endoscopic and open surgery from patients undergoing routine bladder surgery. The patients were divided into two categories based upon urodynamic characterization. The NGF content in samples from 11 normal bladders and seven idiopathic unstable bladders were measured using an enzyme-linked immunosorbent assay. The mechanisms influencing net NGF production were explored using detrusor cells in vitro.

Results The mean (sem) NGF content was significantly higher in unstable tissues, at 0.96 (0.05) pg/µg protein, than in the normal bladder, at 0.53 (0.05) pg/µg protein. In the cell model, acetylcholine (10 µmol/L), noradrenaline (1 and 10 µmol/L) and ATP (1 µmol/L) caused a significant increase in net NGF production; acetylcholine at 1 µmol/L had no effect. Direct stimulation of protein kinase C (PKC) by phorbol ester (33 ng/mL) or elevation of cAMP using forskolin (10 µmol/L) increased NGF, suggesting that at least two intracellular pathways (PKC- and PKA-dependent) are involved. The expression of c-Fos was increased by phorbol 12-myristate 13-acetate added before NGF, suggesting that c-Fos may be involved in regulating NGF production.

Conclusion These data suggest a role for NGF in the physiology and pathophysiology of the human bladder, and indicate some of the possible mechanisms which might regulate NGF production.


There are major structural changes in human detrusor muscle and its innervation as a consequence of outlet obstruction, and in both idiopathic and neuropathic instability [1–6]. Instability associated with outlet obstruction is accompanied by smooth muscle hypertrophy, alterations in the expression of proteins associated with contraction and changes in metabolic enzymes [7,8]. In addition, there is a reduction in the density of motor nerve terminals and decreases in the levels of neuropeptides [1,9–11]. On removing the obstruction, the bladder appears to return towards normal, with a partial reduction in hypertrophy and re-innervation [1]. However, little is known about the events which trigger or influence these structural alterations.

Less information is available on the structural changes in muscle and its innervation in idiopathic and neuropathic instability. Morphological changes are seen in bladder smooth muscle and in the motor and sensory innervation. The density of putative sensory nerves appears to be increased in the subepithelial layer of bladders with idiopathic instability [12]. It has been reported that there is a decrease in motor nerve density in idiopathic bladders, with the nerve loss being highly punctate [6]. Discrete areas of reduced innervation can be found adjacent to apparently normal areas of detrusor muscle [6]. The reasons for this localized loss of nerve fibres are unknown.

Nerve growth factor (NGF) plays an important role in maintaining the autonomic innervation of many organs [13]. This has led to the suggestion that alterations in the autonomic innervation of the pathological bladder may be accompanied by altered levels of NGF. Increased levels of NGF have been reported in the bladder tissues of both obstructed animal and human bladders where there is accompanying denervation [9]. Experimental manipulation of animal models suggests a direct link between NGF and innervation. In neonatal animals, treatment with antiserum raised to NGF produced a severe loss of nerves in the bladder [14]. Similarly, injections with NGF enhanced the innervation of rat bladder [14,15].

The mechanisms regulating the production of NGF have been studied using animal bladder smooth muscle cells in vitro. In rat bladder smooth muscle cells, platelet-derived growth factor (PDGF) and TGF-β were among the most potent stimulators of NGF production, suggesting a possible autocrine or paracrine regulatory pathway [16,17]. In rat bladder PDGF also induced its effect via protein kinase C (PKC), as PKC down-regulation by prolonged treatment with phorbol 12-myristate 13-acetate (PMA) abolished the stimulatory effects of PDGF [16]. Furthermore, it has been shown that stretch, operating through a mechanism that involves PKC and intracellular Ca2+, affects the synthesis of NGF [16–18].

As human detrusor from patients with idiopathic instability shows a punctate denervation there may be accompanying changes in the level of NGF in this pathological state. If NGF production can be altered in the human detrusor then it is important to identify the mechanism responsible for regulating NGF expression. In the present study we measured the concentrations of NGF in tissue biopsies taken from subjects with a normal bladder and from patients diagnosed to have idiopathic detrusor instability, and used an in vitro model to study the mechanisms of NGF expression.

Materials and methods

Patients were divided into two categories, a stable control group and a group with detrusor instability. The control group contained patients undergoing routine urological operations. Biopsy specimens were obtained mainly from the posterior and posterolateral wall of the bladder during endoscopic surgery, using cold-cup resection, and from the bladder dome after open surgery. No biopsies were taken from the bladder neck or the trigone area. With patients undergoing a total cystectomy for cancer, biopsies were taken well away from the tumour area, from those who had not undergone radiotherapy. Biopsies were taken from patients with idiopathic bladder instability where the diagnosis was confirmed by previous urodynamic investigation. The detrusor biopsies were taken with informed consent and with ethical approval from the Newcastle Area Health Authority.

The bladder samples were mechanically homogenized at 1:10 in homogenizing buffer (25 mmol/L Tris base, containing 0.25 mol/L sucrose and 1 mmol/L EDTA, pH 7.6). The homogenate was then centrifuged at 1000 g for 15 min, giving a crude cell lysate preparation. The protein concentration of each sample was estimated and fixed at 1 mg/mL.

The methodology and characterization of the detrusor smooth muscle cells in vitro were described previously [19]. Briefly, samples were placed in RPMI 1640 culture media (Sigma Chemical Co, Poole, UK) at 4 °C for transit to the laboratory. Tissue samples were washed in RPMI 1640 to remove excess blood, and urothelium and fatty tissues removed. Samples were finely dissected and incubated with slight agitation overnight at 37 °C in RPMI 1640 media containing 0.1% (w/v) collagenase type XI, 0.1% (w/v) trypsin inhibitor type II-S and 0.05% (w/v) hyaluronidase type III and IIS (Sigma). The cells were removed from suspension by centrifuging at 400 g for 5 min, and the pellet re-suspended in RPMI 1640 media, supplemented with 10% (v/v) heat-inactivated fetal calf serum (FCS, Sigma), penicillin (100 µg/mL), streptomycin (100 µg/mL) and additional l-glutamine (0.02 mmol/L; Gibco Life Sciences, Paisley, Scotland). The cell suspension was added to T75 cell-culture flasks or wells and the resultant cell cultures maintained at 37 °C in a 5% CO2 atmosphere. Flasks of confluent cells were passaged and the cells used between passages three and six. The purity of the culture was established using an indirect immunofluorescent antibody staining technique to human smooth muscle α-actin [19].

For pharmacological experiments, smooth muscle cells were grown to confluence in six-well culture plates and were serum-restricted 24 h before the experiment. The cells were incubated at 37 °C over a variety of periods with acetylcholine (1 and 10 µmol/L); noradrenaline (1 and 10 µmol/L); ATP (1 µmol/L); substance P; neuropeptide Y, VIP, EGF; and PDGF; all used at 50 ng/mL). Supernatant samples were removed for the analysis of NGF levels. Excess supernatant was removed and the wells washed in PBS. The underlying cells were removed from each well with a 1% trypsin solution (Gibco) centrifuged at 400 g for 5 min. The cell pellet was re-suspended and triturated in homogenizing buffer, from which samples were taken for Western blot and protein concentration analysis.

Western blot analysis

Detrusor muscle lysate preparations and cultured cell homogenates in fixed protein concentrations were loaded onto polyacrylamide gels and separated for ≈ 1 h at 40 mV, using Tris-glycine electrophoresis buffer (25 mmol/L Tris base, 250 mmol/L glycine, electrophoresis grade) and 0.1% (v/v) SDS. Proteins from the gels were transferred to nitrocellulose using an electroblotting apparatus with 48 mmol/L Tris base, 39 mmol/L glycine, 0.037% (v/v) SDS (electrophoresis grade) and 20% (v/v) methanol (pH 8.3) transfer buffer. After transfer, nitrocellulose papers were incubated in Tris-buffered saline (TBS; 25 mmol/L Tris base, NaCl, KCl and 0.05% Tween 20, pH 7.4, Sigma), containing 5% skimmed milk at room temperature for 1 h. The nitrocellulose was then incubated with antibodies specific for c-Fos and NGF (Santa Cruz Biotechnologies, Inc. Santa Cruz, CA) at 1 µg/mL (made up in the 5% milk/TBS solution) for 2 h at room temperature. As a control, primary antibodies were also incubated for 2 h with the appropriate blocking peptides (Santa Cruz, and Calbiochem Corp., San Diego, CA) at 10 times the antibody concentration, to identify the specific NGF and c-Fos bands. Blots were washed with TBS and incubated with a goat antirabbit immunoglobulin G coupled to horseradish peroxidase (Dako, Copenhagen, Denmark), diluted to 1:2000 for 2 h. After washing, antibody complexes were detected by ECL™ (Amersham Laboratories, Bucks, UK) using X-ray film and scanned on an imaging densitometer. Data were obtained when there was a linear relationship between the amount of protein loaded and the intensity of the ECL signal from the immunoblots.


The levels of total NGF released into the supernatant during the agonist-stimulation experiment were assessed using the NGF Emax™ ImmunoAssay system (Promega Corp., Madison, WI). Supernatant samples (100 µL) were added to kit-treated Maxisorb 96-well ELISA plates (Nunc, Denmark) in which NGF concentrations of 7.8–500 pg/mL were detectable using a plate reader.

The results are expressed as the mean (sem), with the significance of differences tested using Student’s t-test and P < 0.05 considered to indicate significance.


The NGF concentrations in detrusor tissue devoid of urothelium from 11 normal and seven unstable bladder samples were 0.53 (0.05) and 0.96 (0.05) pg/µg protein, respectively. There was a significant increase in NGF expression in the unstable bladder (P = 0.006). Therefore, the reported changes in detrusor innervation in idiopathic instability appear to be accompanied by significant changes in NGF levels.

To assess the cellular mechanisms that might be involved in these increased levels of NGF expression, experiments were carried out on isolated human detrusor cells in vitro. NGF release into the media of confluent smooth muscle cells was determined using the ELISA (Table 1). As the sympathetic and parasympathetic neurotransmitters in the human bladder are noradrenaline and acetylcholine, respectively, the effect of these agents on NGF output was assessed. Unstable detrusor may have an increased responsiveness to ATP [19] and therefore the effect of ATP was also examined. Acetylcholine (10 µmol/L), noradrenaline (1 and 10 µmol/L) and ATP (1 µmol/L) all significantly increased the rate of NGF release into the bathing media (P < 0.001). Acetylcholine at 1 µmol/L had no significant effect (P = 0.231; Table 1). Thus the endogenous neurotransmitters in the bladder appear to control the release of NGF from the detrusor smooth muscle.

Table 1.  The effect of the neurotransmitters acetylcholine, noradrenaline and ATP on the rate of NGF release in isolated detrusor smooth muscle cells in vitro. All values, except that using 1 µmol/L acetylcholine (P = 0.231), were significantly greater than the control (P < 0.05).
Mean ( s e m ) NGF release
(pg/µg protein/h)
Control0.22 (0.05)
Acetylcholine (1)
0.21 (0.05)
0.39 (0.02)
Noradrenaline (1)
0.58 (0.06)
0.49 (0.02)
ATP (1)0.53 (0.12)
Histamine (1)0.33 (0.05)

The effect of several other putative neurotransmitters on NGF release was also assessed. Substance P, neuropeptide Y, VIP, EGF and PDGF (all used at 50 ng/mL) failed to increase NGF release (data not shown). However, histamine at 1 µmol/L increased net NGF production significantly (P < 0.05; Table 1).

In rat detrusor smooth muscle, the activation of PKC and elevation of cAMP can alter the expression of NGF. The possibility that these pathways were also involved in the activation of NGF production in isolated human detrusor muscle was determined by direct activation of PKC with the phorbol ester PMA, and activation of adenylyl cyclase by forskolin. Exposure of cells to forskolin (10 µmol/L) for 1 h (P = 0.003) and 3 h (P = 0.001) caused a significant increase in the rate of release of NGF. When cells were stimulated with PMA there was a significant increase in the rate of release of NGF at all sample times up to 6 h (Fig. 1). PMA appeared to be a more effective stimulus than forskolin, causing a greater release of NGF. Prolonged stimulation with PMA resulted in a reduction in the rate of NGF release. Thus, activation of PKC and adenylyl cyclase, and a subsequent rise in cAMP, can stimulate NGF production in vitro in these human cells.

Figure 1.

The effects of forskolin (green) and PMA (red) on the rate of release of NGF into the culture media of human detrusor smooth muscle cells in vitro. The stimulating concentration of forskolin was 10 µmol/L and for PMA 33 ng/mL. The ordinate shows the time of incubation and the abscissa the rate of NGF release. Data are the mean (sem). The values at all times were significantly greater than the control levels (P < 0.05).

In animal cells, stimulation of NGF production may involve the activation of several oncogene pathways, including c-Fos[20], and thus the possible involvement of c-Fos activation in human bladder smooth muscle was also examined. Cells were stimulated with PMA (33 ng/mL), as this agent produced a large increase in NGF release (Fig. 2). On stimulation with PMA the expression of c-Fos was significantly increased within 1 h of application and the expression was maximal at 2 h. On longer exposure c-Fos expression declined to levels similar to that in unstimulated controls. In contrast, cellular levels of NGF increased at 2 h and were maximal after 3 h of incubation. Again, on longer exposure the amounts of NGF produced declined. This decline might be accounted for by the down-regulation of PKC by prolonged exposure to phorbol ester [16]. This decrease is therefore a consequence of the experimental approach and not a true physiological transient stimulation in NGF production. Although c-Fos increases with PMA stimulation and precedes NGF production, there is no direct evidence at present to state that an elevation in c-Fos is directly linked to NGF production.

Figure 2.

(A) Western blot analysis of intracellular c-Fos and NGF expression in PMA-treated cells (33 ng/mL). Primary c-Fos and NGF specific antibodies were used at 1 µg/mL for 2 h, with incubation with the secondary horseradish peroxidase conjugated antibody at a dilution of 1 : 2000. (B) Cellular levels of c-Fos (green) and NGF (red) in cells maintained in vitro and treated with PMA (33 ng/mL). The ordinate shows the time of incubation, and the abscissa the relative expression of c-Fos and NGF as determined by densitometric analysis of western blots. To compare different blots the level of expression at 1 h for c-Fos and at 3 h for NGF were arbitrarily set at 100%. Data are the mean (sem). Values for 1 and 2 h for c-Fos and 2, 3 and 4 h for NGF were significantly greater than the control levels (P < 0.05).


The main finding of the present study was the increased level of NGF in detrusor smooth muscle samples taken from patients with idiopathic unstable bladder. There are reportedly complex discrete structural changes in idiopathic bladders [6]. The dramatic loss of nerve profiles suggests that idiopathic instability is associated with a defect in the interactions between the motor nerves supplying the detrusor and the underlying muscle. From this information the initial defect in unstable bladder may be damage to nerves and a subsequent reduction in density. The resulting loss of either a neurotrophic stimulus or neurotransmitter-induced activity in the muscles could cause an increase in NGF production. It is therefore possible that there is a trophic loop, whereby signals pass between the nerves and muscles, to preserve both the structural integrity of the muscle and its functional differentiation. Interruption of this loop either at the level of nerve or muscle could be the trigger for the initiation of a cascade of compensatory changes to rectify any imbalance. Increased NGF production may form one part of this system. In addition, other changes might be activated, including muscle hypertrophy, changes in excitability, contractility and metabolism. Such alterations may only be temporary and normal functioning of the bladder would be reinstated. However, if these conditions persist, a pathological instability might result and may become irreversibly established. The interactions between nerve and muscle would therefore be vitally important in maintaining nerve and muscle function in the human detrusor.

In the present experiments using human cells in vitro PMA increased net NGF production, suggesting that PKC activation also plays a role in the control of NGF production. The data also showed that forskolin, an activator of adenylate cyclase, was also a potent activator of NGF secretion. Forskolin (1 µmol/L) increased cAMP in human bladder cells (Claici, Chambers and Gillespie, unpublished observations). Thus, in the human bladder there appears to be a second pathway involving cAMP, which contributes to the increased secretion of NGF.

In rat bladder smooth muscle cells in vitro, PDGF and TGF-β were the most potent stimulators of NGF production [16,17]. Carbachol weakly stimulated NGF secretion in rat cells while in contrast, isoproterenol decreased NGF secretion [17]. If isoproterenol were to act via β-adrenergic stimulation of adenylyl cyclase, then an elevation in cAMP would be inhibitory in the rat. Elevated levels of cAMP in human cells were clearly a potent activator of NGF synthesis in the present study. Thus there are clear differences in the mechanisms regulating NGF production in rat and human smooth muscle cells in vitro. In the present study, PDGF had no effect on NGF secretion in one series of experiments. The most effective stimulators of NGF production in human bladder smooth muscle cells were ATP (1 µmol/L) and noradrenaline (1 µmol/L). Acetylcholine appeared to be a weak agonist at high concentration (10 µmol/L). A specific effect on the secretion of NGF may therefore be questionable.

The observation that neurotransmitters can modulate NGF secretion suggests that the human bladder may have a mechanism to maintain and regulate the extent of bladder innervation. This neurotrophic mechanism may act in combination with stretch- and hormonally induced increases in NGF production. Thus, the ability to involve several mechanisms may convey a degree of flexibility in the processes that maintain the functional integrity of the bladder under a variety of physiological and pathological stresses.


We are grateful to Dr D. Claici and Ms P. Hatthachote for their help and advice throughout the project.


R. Tanner, BSc, Research Associate.

P. Chambers, BSc, PhD, Research Technician.

M.H. Khadra, PhD, FRACS, Senior Lecturer.

J.I. Gillespie, BSc, PhD, Professor of Human Physiology.