Non-bladder centric rodent models of Painful Bladder Syndrome (PBS) make use of chronic stress to study the pathological mechanism involved and innovative medications necessary to treat the disease.  Refinement of such non-bladder centric models is however necessary to realize which PBS mechanisms and symptoms each model reproduces more accurately. Recent clinical studies have shown that PBS patients present changes in the brain areas receiving visceral sensory input, such as the insula. In the present work we hypothesized that different stress models induce distinct changes in the morphology and activity of the insula. We also evaluated how insula changes relates to bladder activity and to mechanical and thermal pain thresholds.

Two stress models were used: Maternal Deprivation Model (MDM) induced early in life and Water Avoidance Stress test (WAS) induce in adult life. For MDM induction, female Wistar rats were separated from mother and littermates from P2-P15. Experimental procedures were carried at P190-195. For WAS induction, adult female Wistar rats were put on a pedestal in a box full of water, for 1h/day, from P180-P189. Experiments were carried out at P190-195. That is, the age of the animals at the moment of the experiments describe below were carried on was similar for both stress models.
The same experiments were performed in the two models. Mechanical and thermal pain thresholds were investigated in L6-S1 dermatomes the sensory innervation of which project to the same spinal cord segments of the bladder sensory innervation. Thermal pain threshold was evaluated by the Hargreaves technique. Mechanical pain threshold was estimated by the up-down Von Frey method. Bladder activity was assessed by cystometry under urethane anesthesia. After cystometry animals were sacrificed and the insulas were removed. From each rat one insula was immersed-fixed in paraformaldehyde 4%, sectioned and immunoreacted against microglia markers Iba1 and Cd68. The analysis of positive cells on the anterior granular and dysgranular regions of the insula was performed using LASAF and ImageJ software. Morphometric analysis and 3D reconstructions used IMARIS software. The contralateral insula was homogenized, and the total RNA extracted. After RNA integrity tested, the total RNA was converted to single-stranded cDNA, and a multiplex real-time PCR (RT-PCR) reactions were subsequently carried out using different TaqMan probes to assess changes in the mRNA expression of c-fos, Pdyn, Sst, Olig2, iba1, CD68, GFAP, TNFa, CCR2, using GAPDH as housekeeping gene.

After WAS (n=27), the anterior granular and dysgranular regions of the insula had more microglia cells (p=0.0138; power 100%) than the control animals (n=27) and their soma was bigger in WAS (n=30) (p=0.0021; power 100%) than in controls rats (n=28). The territorial area occupied by these cells was similar between the two groups (n=29, p>0.9999; power 100%). However, the filaments of the microglia cells of WAS animals had a different arrangement when compared with the ones of the control animals. The filaments had similar length in both groups (p=0.0620; power 100%), but the WAS group presented less branching points (p= 0.0064; power 100%). Overall, the microglial filaments occupied the same volume in the two groups (p=0.8215; power 100%). The microglia cells of WAS (n=30) and control (n=25) rats had a similar expression of CD68 protein (p=0.1584; power 100%).
In the anterior granular and dysgranular regions of insula of MDM animals, microglia cells were fewer (n=27, p=0.0153; power 100%) than in controls (n=27). The filaments of the microglial cells of MDM animals (n=25) had a similar length, (p=0.4783), similar number of branching points (p=0.6861) and occupied a similar volume (p>0.9999) to the ones of the control animals (n=29). The soma size of microglia cells of the insula of MDM animals (n=29) was similar (p=0.2761) to the observed in the control animals (n=28). Microglia cells occupied the same territorial area in MDM (n=25, p=0.4496; power 100%) and in control animals (n=29).The microglia cells of the animals of MDM group (n=27) expressed more CD68 protein (p=0.0074; power 100%) that the ones from the control group (n=25).
RT-PCR analysis revealed that the insula of WAS animals (n=9) presented higher levels of expression of CCR2 (p=0.0274; power 100%) and lower levels of Olig2 mRNA (p=0.0005; power 88%) and Pdyn (p= 0.0213; power 89%) when compared to the control group (n=9). There was no difference in the mRNA expression of c-fos (p=0.7460), TNFa (p=0.2953), Sst (p=0.4936), iba1 (p= 0.2913) and CD68 (p= 0.9157) in the insula of WAS and control animals. The GFAP mRNA analysis was inconclusive (p=0.035; power 30%).
The insula of animals of the MDM group (n=9) had lower expression of c-fos mRNA (p=0.0036; power 100%). There was no difference in the mRNA expression of CCR2 (p=0.6253), Olig 2 (p=0.1000), Pdyn (p>0.9999), CCR2 (p=0.6253), TNFa (p= 0.0857), Sst (p>0.9999), iba1 (p>0.9999), CD68 (p>0.9999) and GFAP (p>0.9999) between MDM and control rats. 
WAS animals had a lower mechanical pain threshold (n=6, p=0.0066; power 92%) and an increased reflex voiding activity (n=5, p=0.0427; power 86%) than controls but the same thermal pain threshold. MDM animals had lower mechanical (n=6, p=0.0450; power 92%) and thermal pain threshold (n=6, p=0.0344; power 83%) and increased reflex voiding activity (n=5, p= 0.0080; power 86%) than controls.

WAS induces the recruitment of microglia cells to the anterior granular and dysgranular regions of the insula, increasing the number of these cells in these regions. This increase together with the observation that microglia presented an increased soma size are suggestive of neuroinflammatory events. The increase in CCR2 mRNA is in accordance with the neuroinflammatory microglia phenotype, as CCR2 is known to activate the microglia to a neurotoxic phenotype. The neurotoxic microglia may damage oligodendrocytes, which could explain the observed decrease in Olig2 mRNA expression. The observed neuroinflammation of the insula concurs to the mechanical hyperalgesia and bladder hyperactivity. However, these changes may be momentary as no changes in neuronal activity (in the expression of Fos mRNA) were observed. Also, the lower levels of Pdyn mRNA (known to be increase in stress) 5 days after the ending of WAS induction also points for transient insular changes induced by the stress in adult animals.
In contrast MDM seems to induce more permanent changes in the insula. MDM induced a decrease in the number of microglia cells in the anterior granular and dysgranular regions of the insula. In addition, the increased expression of CD64 reflect a higher phagocytic activity by microglia, which has been many times associated with synaptic stripping and shaping of neuronal circuits. In fact, the observed decrease in c-Fos mRNA suggests a decrease in neuronal activity in this brain region. These changes concurs to the mechanical and thermal hyperalgesia and to bladder hyperactivity.

The two models induce similar changes in bladder function and in the nociceptive pathways. However, the central mechanism behind such changes seems clearly different. Sub-acute stress stimulus such as WAS induced transient neuroinflammatory changes in the insula while the imprinting changes induced by MDM seems to alter the neuronal modulation environment of the insula. These findings warrant investigation of the insular regions of PBS/IC patients with functional imaging techniques.

Continence 7S1 (2023) 100759
DOI: 10.1016/j.cont.2023.100759