Hypothesis / aims of study
Spinal cord injury (SCI) induces the loss of voluntary voiding, due to damages in neuronal circuits controlling micturition. After an initial period of little or no bladder reflex activity, post-SCI neuroplasticity results in the development of a new micturition reflex, located at the lumbosacral spinal cord, responsible for neurogenic detrusor overactivity (NDO) [1].
Much of our knowledge about the mechanisms of SCI-induced urinary impairment [2] has been obtained using rodent SCI models. In fact, the majority of data has been generated in studies using lesions caused by complete spinal cord transection (SCT). Nonetheless, SCT are rare in clinical practice, while spinal contusions are more commonly seen. Therefore, experimental studies using spinal contusions offer more translational value and are becoming the model of choice for SCI research. However, it is presently unknown if mechanisms underlying urinary dysfunction after SCT are also operating after spinal contusion and if observations made with that model can be extended to experimental spinal contusions.
Study design, materials and methods
Female Wistar rats were divided into 4 groups (n=4-5 group) and submitted to different SCI protocols under deep anaesthesia: SHAM manipulation of the spinal cord, mild spinal cord contusion (mSCC), severe spinal cord contusion (sSCC), and complete spinal cord transection (SCT) at T8/T9 level. Contusion was induced with an adapted weight drop device, which includes a braking system for post-drop weight withdrawal. The weight was dropped from 5 and 10 cm height to respectively induce mSCC and sSCC. Animals were left to recover for 4 weeks before euthanasia, period in which the volume of residual urine was recorded following daily abdominal compression.
Before euthanasia, all animals underwent 1-hour cystometry under urethane anaesthesia. The bladder, urethra and L5-S1 spinal segments were collected and processed to analyze sprouting (GAP43), sensory (CGRP), parasympathetic (VAChT) and sympathetic (TH) innervation by immunohistochemistry. The lesion site was also collected for lesion histology assessment by Formol-Thionine staining.
Results
The volume of residual urine increased over the first week after SCI in all animals, decreasing afterwards at different rates according to lesion severity. Severe contusion animals presented a longer spinal shock period and a tendency for higher residual volumes and urinary infections, followed by SCT and mSCC animals. The urine volume of mSCC animals was tendentially lower than sSCC and SCT animals during the experimental period, indicating the presence of milder urinary dysfunction in less severe contusion models. Urodynamic assessing showed that SCT animals presented higher frequency, and higher basal and peak pressures than controls (p<0.05 versus SHAM). No differences were found between contusion groups and SCT.
Formol-thionine staining of the lesion site showed that in mSCC and sSCC animals, the area of damaged tissue was significantly larger than in the SCT group (p<0.01 and p<0.001, respectively) and directly proportional to the contusion severity. In the SCT group, the area of lesioned tissue was restricted to the severed ends of the cord.
GAP43 was increased in the L5-S1 superficial laminae of the dorsal horn of mSCC animals when compared to controls (p<0.0001) and SCT group (p<0.001). In the bladder, GAP43 was abolished in the mucosa of all injured groups (p<0.001 versus SHAM) and significantly decreased in the detrusor of SCT animals (p<0.05 versus SHAM). In the urethra, GAP43 was decreased in the mucosa and IUS (internal urethral sphincter) irrespective of the SCI model (p<0.01 versus SHAM) but not in the EUS (external urethral sphincter).
At the spinal cord level, CGRP followed the same pattern as GAP43, being increased in mSCC animals when compared to controls (p<0.05) and SCT animals (p<0.05), both in the dorsal horn and intermediolateral nucleus (IML). CGRP expression was abolished in the bladder mucosa of all injured groups (p<0.0001 versus SHAM) and reduced in the detrusor (p<0.01 versus SHAM). In the urethra, CGRP was dramatically reduced in all SCI groups in the mucosa (p<0.001 versus SHAM) and IUS (p<0.05 versus SHAM). In the EUS, a significant reduction was seen in the mSCC (p<0.05) and SCT group (p<0.01).
The expression of autonomic markers was generally decreased post-SCI. At the spinal cord level, the expression of VACHT was reduced in the IML of sSCC (p<0.05 versus SHAM) and SCT animals (p<0.001 versus SHAM). In the ventral horn, a decrease was observed in SCT animals when compared to mSCC (p<0.05 versus mSCC). VACHT expression in mSCC animals was identical to SHAM. In the LUT, VACHT was abolished in the bladder mucosa of all injured groups (p<0.01 versus SHAM) and decreased in the detrusor of mSCC (p<0.001 versus SHAM), sSCC and SCT (p<0.001 versus SHAM). VAChT expression was also reduced in all SCI groups (p<0.0001 versus SHAM), IUS (p<0.0001 versus SHAM) and EUS (p<0.01 versus SHAM). Sympathetic fibres (TH-positive) were reduced in all injured groups at the lam X of the spinal cord: mSCC and sSCC (p<0.001 versus SHAM) and SCT (p<0.0001 versus SHAM). TH-positive fibres were not detected in the bladder. In the urethra, this sympathetic marker was only decreased in the IUS of mSCC animals (p<0.05 versus SHAM). In the EUS, denervation was detected in contusion groups (p<0.05 versus SHAM) but not in SCT.
Interpretation of results
Severe contusion rats presented a tendency for higher residual volumes, when compared to mSCC, as well as a higher incidence of urinary infections, suggesting that the severity of urinary dysfunction is directly proportional to contusion severity. Additionally, the spinal shock period is more prolonged in sSCC than after SCT. This might be explained by the extended area of lesioned tissue seen in sSCC, affecting more rostral and caudal segments beyond the injury site. This broader area of injury, possibly delays tissue healing and may contribute prolonging of spinal shock. Urodynamic tests failed to detect any model-specific differences in the amplitude or frequency of bladder contractions, which could be explained by the use of urethane as an anaesthetic. Nevertheless, the detection of higher basal and peak pressure in SCT animals may indicate a higher prevalent presence of detrusor-sphincter dyssynergia in these animals but not in spinal contusion rats.
Analysis of sprouting (GAP43) and sensory (CGRP) markers at the lumbosacral spinal cord level demonstrated that afferent sprouting is dependent on the injury model. Both markers were significantly higher in mSCC than in the more severe models, suggesting that less severe models have a higher ability to rearrange micturition circuits and NDO is more rapidly installed.
VACHT was generally decreased in SCI groups when compared to controls, in all the extensions of the LUT and lumbosacral spinal cord, which is consistent with the lesioning of supraspinal projections after SCI. Particularly at the spinal cord level, VACHT was decreased in the IML and ventral horn, where parasympathetic motorneurons are located. Mild contusion animals presented expression levels of VAChT similar to spinal intact animals (much higher than in SCT rats), confirming that mSCC animals maintain a certain level of supraspinal input that is beneficial to urinary recovery after SCI.
Concluding message
Our results suggest that, like SCT, spinal contusions induce urinary dysfunction. Less severe injuries result in milder dysfunction of the LUT, accompanied by signs of neuronal remodeling. In fact, the expression of autonomic markers was different between groups, correlating with the type and severity of injury. This may explain different responses to therapeutic interventions, which are similarly used among patients irrespective of the fine characteristics of spinal lesion.