Professor Prem Puri, Children's Research Centre, Our Lady's Hospital for Sick Children, Crumlin, Dublin 12, Ireland. e-mail: firstname.lastname@example.org
To establish a model for investigating the pathophysiology of pelvi-ureteric junction (PUJ) obstruction, using benzalkonium chloride (BCI) treatment of the upper urinary tract of rabbits, and thus further elucidate the pathophysiology of PUJ obstruction, the most common urinary tract obstruction in children.
MATERIALS AND METHODS
Although various histological abnormalities have been described, PUJ obstruction may be functional. Defective innervation in PUJ has been suggested to be a major factor in the failure to transmit peristaltic waves across the PUJ. Previously established animal models of hydronephrosis deal mostly with surgical obstruction of the PUJ, which does not correlate with human congenital hydronephrosis. BCl has been used to ablate selectively neurones of the gastrointestinal myenteric plexus, which generated spastic segments with impaired peristalsis. Thus 12 rabbits were treated with BCl at the PUJ; the right upper urinary tract was dissected extraperitoneally and treated with a local application of 0.1% or 0.5% BCl (six each) for 15 min. The controls were four sham-operated animals treated with saline. The animals were assessed by intravenous urography (IVU) at 4 and 8 weeks after treatment, after which the animals were killed, the upper urinary tracts removed and whole-mounts prepared. Acetylcholinesterase (AChE) histochemistry, and neurofilament and tyrosine hydroxylase (TH) single-enzyme immunohistochemistry were used to detect the intrinsic innervation.
None of the animals had hydronephrosis on the IVU or at death. AChE histochemistry, TH and neurofilament immunohistochemistry showed no or very few nerve fibres within the BCl-treated PUJs in both (0.1% and 0.5%) groups. After saline treatment there was normal development of the neuronal plexus within the submucosal, muscular and adventitial layers of the upper urinary tract.
These results suggest that treatment with BCl is useful for ablating the intrinsic innervation in the upper urinary tract. Defective intrinsic innervation of the upper urinary tract did not lead to clinically or radiologically evident hydronephrosis. Further physiological studies using this model are needed to further elucidate the neuronal and myogenic influence on the development of PUJ obstruction.
Congenital hydronephrosis caused by PUJ obstruction remains an enigmatic problem. Postnatally confirmed hydronephrosis caused by PUJ obstruction necessitates surgery in 15–30% of cases for impaired or deteriorating renal function, or clinically presents as renal colic, recurrent UTI and kidney stones . Current understanding of the pathophysiology of obstructive uropathy is incomplete [2,3].
The main problem in PUJ obstruction is to clearly differentiate simple dilatation of the renal pelvis from significant renal obstruction which, if untreated, may cause progressive deterioration of renal function. Several animal models have been used to investigate the changes in renal function after surgically creating PUJ or ureteric obstruction. Surgical ligation or embedding the ureter into the psoas muscle has been used to obstruct the ureter ; by their nature such models have an acute onset and therefore do not reflect the congenital pathology.
Although various histological abnormalities have been described in congenital hydronephrosis from PUJ obstruction, many consider that the obstruction is functional. Defective innervation in PUJ obstruction has been suggested to be a major factor in the failure to transmit peristaltic waves across the PUJ [5,6]. Recent histochemical and immunohistochemical studies using whole-mount preparations show a dense intramural innervation of the upper urinary tract in various species, including humans [7,8]. Benzalkonium chloride (BCl, a cationic surfactant which generally perturbs membranes) has been used to ablate selectively neurones of the gastrointestinal myenteric plexus, which generated spastic segments with impaired peristalsis . Ablation of the rat, guinea pig or rabbit myenteric plexus with BCl has provided a model of aganglionosis. Besides ablating the intrinsic enteric nervous plexus, there is no degenerative or scarring process in the muscular layers. Thus the aim of the present study was to evaluate whether BCl treatment of the upper urinary tract causes ablation of the intramural innervation, leading to hydronephrosis.
MATERIALS AND METHODS
The animal studies were conducted according to the European Community Directive 86/609/EC, Ref. B100/3120 (Department of Health and Children, Dublin, Ireland). Altogether 16 rabbits (White New Zealand, 3–3.5 kg, female) were used in the study. Narcosis was induced with intramuscular ketamine (25 mg/kg) and xylazine (5 mg/kg), with a continuous supply of oxygen provided during the procedure. After a right flank incision the right ureter was dissected extraperitoneally (Fig. 1); the mobilization of ureter was reduced to a minimum to avoid unnecessary bleeding and damage to surrounding tissues. A ureteric segment (1 cm long) immediately distal to the PUJ was treated circumferentially with 0.1% or 0.5% BCl (six rabbits each) or saline (four) for 15 min. The BCl solution or saline was applied using wet gauze to wrap the proximal ureter. The surrounding tissues were preserved from contact with BCl solution using parts of rubber gloves around the gauze. The wound cavity was extensively irrigated with saline and the wound closed in layers.
Each animal was then assessed by IVU under general anaesthesia 4 and 8 weeks after surgery. The contrast medium was given via an ear vein at 5 mL/kg and a conventional X-ray system used, with a plain abdominal X-ray taken 5 and 15 min after administering the contrast medium.
The animals were then killed using pentobarbitone sodium (150 mg/kg) after 10 weeks, the kidneys and upper urinary tracts removed quickly, and rinsed in PBS. The resected upper urinary tracts were fixed in Zamboni's solution overnight at 4 °C, followed by rinsing three times in PBS. Whole mounts were prepared using a dissection microscope and fine forceps. The whole upper urinary tracts were separated into adventitia, the muscular layer and submucosal-mucosal layer. At the level of the PUJ the separation of an inner and outer muscle layer was possible. After additional rinsing in PBS the layers were stained using a free-floating technique.
For haematoxylin and eosin (H&E) staining, parts of the dissected upper urinary tracts were rinsed in PBS followed by fixation in 4% paraformaldehyde and processing until paraffin-wax embedding. Serial paraffin sections were processed and examined by conventional H&E staining.
For standard acetylcholinesterase (AChE) histochemistry the specimens were incubated at 37 °C for 2 h, following the method of Karnovsky and Roots . Subsequently the specimens were mounted into Glycergel mounting medium (Dako, Glostrup, Denmark) and investigated by conventional bright-field microscopy.
Single-enzyme immunohistochemistry was carried out to detect neurofilament and tyrosine hydroxylase (TH) using anti-neurofilament (mouse, Dako; 2F11, 1 : 50 in TBS-BSA 5%-Triton 0.1%) and anti-TH (sheep, Chemicon Int., Temecula, CA, 1 : 100 in TBS-BSA 5%-Triton 0.1%), with the streptavidin-alkaline-phosphatase Universal Kit (Immunotech, Inc., Marseilles, France) with the chromogen Fast Red. The specificity of the immunohistochemical staining was supported by negative controls omitting the primary antibody. The specimens were mounted into Glycergel mounting medium and evaluated using normal bright-field microscopy.
The postoperative course was uneventful in all animals; none died and the animals remained healthy with free access to water and food during the follow-up period. The IVU 4 and 8 weeks after treatment showed normal upper urinary tracts with no signs of hydronephrosis in all animals (Fig. 2). Examination after death showed no evidence of hydronephrosis in the treated or control animals.
On H&E staining the smooth muscle architecture was no different in treated or control animals, although there was slightly more interstitial connective tissue in the treated animals.
On AChE histochemistry, and neurofilament and TH immunohistochemistry, the control animals had dense neuronal plexuses within the submucosal, muscular and adventitial layers of the upper urinary tract. These neuronal plexuses contained small AChE-positive ganglia within the distal ureter. These small ganglia were also evident within the adventitial and muscular layers. The ganglia contained 2–6 ganglion cells of 20–30 µm in diameter. Thick AChE-positive, TH- and neurofilament-immunoreactive nerves were present in the adventitial layer, running longitudinally along the ureter. These nerve trunks were interconnected with short and small nerves.
Regular AChE-positive (Fig. 3a) and TH (Fig. 3b) and NF-IR (Fig. 3c) neuronal networks containing 5–10 µm thick nerves were present in the muscular layer. An additional but thinner regular network was evident within the submucosal layer, with numerous varicosities into the mucosa. The nerve diameter in the submucosal network was 4–7 µm. In general the innervation was slightly greater in the distal and proximal ureter within all layers.
In treated animals, the AChE histochemistry (Fig. 3d), and TH (Fig. 3e) and neurofilament (Fig. 3f) immunohistochemistry, showed none or markedly fewer intramural nerve plexuses within the segments of the upper urinary tract in both (0.1% and 0.5%) groups. There was a clear demarcation between the treated and untreated segments.
Treatment with BCl has been effective in creating aganglionic parts in the bowel of various animal models; the model was first described by Sato et al. and Sakata et al.. Treatment with 0.1% BCl for 30 min led to aganglionosis in the rat bowel . The agent first injures the cell membrane, causing depolarization or its impairment in active sodium transport, and may result in damage to the cell. BCl solution selectively ablates neural elements in the bowel wall through its depolarizing effect, which is more marked on the cell membranes that have a higher negative charge. The level of negative charge is known to be higher in neural tissues (− 70 to − 90 mV) than in smooth muscle (− 30 to 70 mV) .
Other studies using electron microscopy showed that in the acute phase after BCl treatment smooth muscle cells were partially damaged; later these cells recovered fully and the neural tissues were irreversibly damaged in the chronic stage. The muscle layer of the bowel wall was thickened after BCl treatment [13,14]. In the present study, the muscle layers were undamaged and only slight fibrosis was evident in the treated animals.
The rabbit was selected for the present experiments because it is a cost-effective model and the extraperitoneal approach to the kidney is easy. Furthermore, the rabbit is a robust animal and tolerates repeated anaesthesia for the X-ray studies.
Treatment with BCl in the upper urinary tract resulted in ablation of the intramural nerves. The neuronal plexuses in all layers were either markedly reduced or absent; this was clear in whole-mount preparation. Previous studies investigating the innervation of urinary tract have used conventional paraffin wax or frozen sections. The three-dimensional structure of intramural innervation is difficult to appreciate in thin transverse sections [7,8]. The whole-mount preparation is a suitable method for visualizing the three-dimensional morphology of the meshwork of intramural nerve fibres, being especially useful for morphometric investigations. In the present study the whole-mount preparations clearly showed the demarcation between the treated proximal ureter and the normal innervated parts of the upper urinary tract. Morphometric measurements were unnecessary in this preliminary study because the ablation of the intramural nerves was almost complete.
The present study further showed that despite the ablation of the intramural innervation of the upper urinary tract there was no hydronephrosis on IVU. As no electrophysiological or other functional investigations were included it is not possible to comment on the functional effect of BCl treatment. The study was limited to 10 weeks follow-up; perhaps a longer follow-up is required to develop functional changes after ablation of intramural nervous system. Additional studies using the BCl treatment model, with electrophysiological and functional investigations such as radioisotope studies, are needed to elucidate the influence of nerve ablation on pelvi-ureteric peristalsis.