Stress has a pivotal role in the initiation, maintenance, and episodic aggravation of BPS/IC symptoms. Previous studies demonstrated that an excess of norepinephrine plays a fundamental role in these processes through the activation of alpha-1A adrenoceptors. In order to investigate the pathophysiological mechanism in play several models of stress have been used, among which the maternal deprivation model (MDM) and the water-avoidance stress test (WAS) are the most commonly used. Both represent a stressful event applied during a limited period, at different stages of rodent development, during the neonatal period and at the adult life, respectively. While MDM tend to cause effects that are long-lasting, the duration of WAS effects on bladder function and sensory activity are still unclear. The aim of the present study was to further refine the above-mentioned stress models. We investigated the hypothesis that repeated stressful events aggravate bladder function and induce sensory changes in a way similar to that observed in BPS/IC patients.

First set of experiments: Adult (6M) female Wistar rats were used. Animals were divided in four groups of six animals: control 12d, control 15d, WAS 12d and WAS 15d. WAS induction was performed for 1h, for 10 days, by placing animals in a pedestal inside a cage full of water – sham-was induction. Control animals were placed in a pedestal inside a waterless cage. The experimental timeline was as follows:
Control – Sham-Was was performed from 1d to 10d. At 12d or 15d, animals were anaesthetised, urine collected and cystometry performed.
Stress naïve WAS – At baseline, grimace analysis, Hargreaves and Von Frey testing up-down method were performed. Then, WAS was induced from 1d to 10d. At 11/12d, the behaviour tests were repeated. At 12d or 15d animals were anaesthetised, urine collected and cystometry performed. Afterwards, the bladder was harvested and fixed by immersion.
Urinary noradrenaline levels were measured by HPLC; Urinary bladder were section and stained with toluidine blue to analyse mastocytosis.
Second set of experiments: After birth, litters of Wistar rats were balanced (P1) and animals were separated from mother and littermates, for 1h, from P2 to P15 (experienced-stress animals). After weaning (P21), groups of 6 females were formed. The control and WAS 15d group were balanced but animals were not separated. The experimental timeline at 6M was as follows:
Control –Sham-Was was performed from 1d to 10d. At 15d, animals were anaesthetised and cystometry performed.
Stress naïve WAS – WAS was induced from 1d to 10d. At 15d animals were anaesthetised and cystometry performed.
Adult experienced-stress 1hMDM + WAS 15d – separation P2-P15; Hargreaves testing and Von Frey testing up-down method were performed. Then, WAS was induced from 1d to 10d. At 11d, Hargreaves testing were repeated. At 12d, Von Frey testing up-down method was performed. At 15d animals were anaesthetised and cystometry performed.

At 11/12d, stress-naive WAS animals presented increased bladder reflex activity (Mann Whitney test, p = 0.0079), changes in grimace (Unpaired t test, p = 0.0010), thermal hyperalgesia and mechanical allodynia (Wilcoxon matched-pairs signed rank test, both, p = 0.0313), and bladder mastocytosis. 
However, at 15d, there was no longer bladder reflex hyperactivity (Mann Whitney test, p = 0.0996; Figure 1). Also, at this time point, animals no longer presented signs mastocytosis. 
Analysis of pooled samples revealed that controls had 0.2 nmol noradrenaline/ml urine, and animals submitted to WAS had 4.5 and 2.1 nmol noradrenaline/ml urine, at 12d and 15d respectively. Furthermore, at 15d, there was a very strong correlation between the levels of urinary noradrenaline and the frequency of bladder reflex activity (Pearson r = 0.9468, p = 0.0146).
 
When submitted to WAS, adult experienced-stress animals presented bladder reflex activity higher than control and WAS (Dunn's multiple comparisons test, p = 0,0169, preliminary results; Figure 2). 
Also, at baseline, adult experienced-stress animals also present both thermal hyperalgesia and mechanical allodynia (Wilcoxon matched-pairs signed rank test, both p = 0.0313), when compared with control animals. However, these values did not change with WAS induction ( Wilcoxon matched-pairs signed rank test, p = 0.7500 and p = 0.5000, respectively).

Animals submitted to WAS present signs of pain and bladder hyperactivity and inflammation. However, the bladder changes seem to fade away as not all WAS 15d groups animals presented bladder hyperactivity. This shows that the changes induced by WAS are transient and, therefore, care should be taken when choosing this model for protocols with longer timelines.
Curiously, the levels of urinary noradrenaline seem to be a good biomarker to predict the outcome of bladder activity in the WAS model. This may be useful if other tests, rather than cystometry, are to be performed.
When submitted to WAS, experienced-stress animals presented high bladder hyperactivity at 15d. This seems to indicate that the imprint of stress in childhood aggravated the bladder symptoms induced by stress in adulthood, which resemble the flares described by chronic pelvic pain patients. Hence, the 1hMDM+ WAS paradigm seems more appropriated for longer studies of bladder function than WAS paradigms.

The choice an animal model and timepoint for analysis should be a matter of carefully refinement. Repeated stressful events seems to be a good paradigm to mimic BPS/IC stress phenotype in what concerns bladder symptoms and pain.