A hallmark symptom of inflammatory bowel disease is abnormal colon motility, largely attributed to changes in enteric nervous system (ENS) function, but studies of dysmotility are based on models of ENS-driven peristalsis in distal colon regions. Recent evidence has identified several differences between proximal and distal colon, indicating a need to define the network mechanisms responsible for motility in the proximal colon, where spontaneous colon motor complexes (CMCs) are regularly initiated. Ongoing work in our lab suggests that this rhythmicity is due to cyclical interactions among ENS motor neurons, interstitial cells of Cajal (ICC), and intrinsic primary afferent neurons (IPANs) that respond to ICC-generated ripples, and we developed a computational model that reproduces proximal colon motility in normal conditions. Therefore, the aim of this study was to measure changes in ENS and ICC activity associated with inflammation-induced dysmotility and test if our computational model predicts motility behavior based on the cellular dysfunction measured experimentally. Colon preparations from E2a-GCaMP mice treated with vehicle or 3% dextran sodium sulfate (DSS) for 5 days were removed and imaged to visualize ENS and ICC Ca²⁺ activity, and CMCs were measured by tracking displacement in the field of view. We ran simulations of our computational model to test the effect of altering ENS and ICC activity parameters on CMC behavior. Experimental results showed that DSS-inflammation produced dysmotility in the proximal colon, such that CMCs were less frequent, longer in duration, and less regular compared to control mice (p<0.05, n=8-9). ICC slow wave frequency was increased in DSS-treated mice (p<0.001, n=5-6). ENS activity was also increased (p<0.05, n=6), specifically during an “adaptation” phase that immediately follows a CMC and normally characterized by low levels of neural activity. In our computational model, increasing ICC frequency decreased CMC frequency but had no effect on other CMC variables, thus only partially reproducing dysmotility in our experimental model. Decreasing the adaptation variable in our model increased CMC duration and the variability of intervals between CMCs but did not alter CMC frequency, again only partially reproducing experimental findings. Remarkably, when changes to ICC frequency and ENS adaptation were both incorporated, all aspects of dysmotility were reproduced. Our findings suggest that ICC slow waves and the ENS are both critical for regular initiation of CMCs in the proximal colon, and that dysmotility during inflammation may be due to dysrhythmic coordination of ICC and ENS activity. Ongoing experiments are using Cre-specific mouse models for calcium imaging and optogenetics to elucidate ICC interactions with specific ENS subtypes, and results will be incorporated into a new and expanded model.

<b><u>Figure1. DSS disrupts proximal colon motor circuits.</u></b> (A) Data show frequency, variability, and duration of CMCs in the proximal colon of control (VEH) mice and mice treated with 3% DSS (n = 8-9).<br /> (B) Representative traces and quantification of ICC slow wave frequency from E2a-GCaMP mice treated with vehicle (VEH) or DSS (n = 5-6). (C) Representative traces of Ca<sup>2+</sup> responses (ΔF/F<sub>0</sub>) from enteric neurons in VEH and DSS E2a-GCaMP mice. Summary data shows the percent of responding cells after a CMC (n = 6). Data was pooled from male and female mice (individual values show males as triangle and females as circles) and was analyzed via unpaired t-test (<i>p* < 0.05, p** < 0.01, p*** < 0.001</i>).

Figure1. DSS disrupts proximal colon motor circuits. (A) Data show frequency, variability, and duration of CMCs in the proximal colon of control (VEH) mice and mice treated with 3% DSS (n = 8-9).
(B) Representative traces and quantification of ICC slow wave frequency from E2a-GCaMP mice treated with vehicle (VEH) or DSS (n = 5-6). (C) Representative traces of Ca²⁺ responses (ΔF/F₀) from enteric neurons in VEH and DSS E2a-GCaMP mice. Summary data shows the percent of responding cells after a CMC (n = 6). Data was pooled from male and female mice (individual values show males as triangle and females as circles) and was analyzed via unpaired t-test (p* < 0.05, p** < 0.01, p*** < 0.001).

<b><u>Figure2. Computational model reproduces motility changes observed experimentally</u></b>. Representative motor activity traces showing CMC intervals from vehicle (A) and DSS (B) treated mice. (C-D) CMC simulations generated in our computational model. Panel C show simulations generated under normal (control) conditions and panel D shows simulations generated with altered ICC frequency and adaptation.

Figure2. Computational model reproduces motility changes observed experimentally. Representative motor activity traces showing CMC intervals from vehicle (A) and DSS (B) treated mice. (C-D) CMC simulations generated in our computational model. Panel C show simulations generated under normal (control) conditions and panel D shows simulations generated with altered ICC frequency and adaptation.

Wilmarie Morales-Soto

Appearances