All the animal protocols were approved by the Institutional Animal Care and Use Committee of the University of Texas Medical Branch in Galveston, Texas. Experiments were performed in accordance to the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health, United States.

Male and female Sprague-Dawley rats of 8–9 weeks old (200–280 g, Harlan Sprague Dawley, Indianapolis, IN) were used in the study. Rat model of CD-like colitis was induced by intracolonic instillation of TNBS (Sigma Chemical, St. Louis, MO) as described previously (Shi et al., 2011b; Geesala et al., 2022). Prior to TNBS instillation, rats were given bowel cleanser PEG 3350 (GoLYTELY, Braintree, MA) for 24 h and fasted for 16 h. This was to cleanse the colon to minimize the impact of fecal pellets on TNBS absorption in different rats. TNBS (65 mg/kg in 250 µL of 40% ethanol) was instilled in 1 min through an intracolonic catheter to the colon 7 cm from the anus. Control rats were treated with intracolonic instillation of 250 µL saline. Control and TNBS rats were housed separately and fed ad libitum with regular pellet foods (Picolab Rodent Diet 5,053, LabDiet, St. Louis, MO) unless specified otherwise. The average daily intake of the pellet food for a control rat is nearly 20 g with ∼70 kcal (4.0 g of protein, 2.0 g of fat, 10.4 g of carbohydrate and 1.0 g of fiber). Rats were euthanized 7 days after saline or TNBS treatments. Body weight was documented over 7-day course. Inflammation was scored 1 to 3 in each rat according to criteria described previously (Shi et al., 2011b).

In the experiments involving in vivo inhibition of COX-2 activity, control and TNBS-instilled rats were treated with COX-2 inhibitor NS-398 (Cayman Chemical, Ann Arbor, MI) daily at 5 mg/kg intraperitoneally (ip) in 250 µL of 20% DMSO starting on the day inflammation was induced (Shi et al., 2011a; Shi et al., 2011b; Lin et al., 2012).

To minimize fecal accumulation and release lumen distention in inflammation, some rats were fed exclusively with Ensure liquid diet (Abbott Nutrition, Lake Forest, IL) ad libitum (Lin et al., 2021) after intracolonic instillation of saline or TNBS. The average daily intake of the liquid diet for a control rat is about 75 mL with 75 kcal (3.0 g of protein, 2.0 g of fat, 11 g of carbohydrate and 0.3 g of fiber). Rats were housed separately in wire-bottomed cages to prevent them from eating feces or bedding materials (Lin et al., 2021). All rats had continuous free access to water.

In our pilot study testing the effect of different liquid diets, we tried not only Ensure liquid diet but also a rodent liquid diet, #D11112201L made by Research Diets (New Brunswick, NJ). Although the rodent liquid diet is not identical as Ensure liquid diet in terms of nutrient contents, it had very similar effect as Ensure liquid diet in eliminating mechanical distention in the colon. We decided to use Ensure liquid diet in the study, as it has been used in the EEN treatment for CD patients and in eliminating mechanical distention in rats (Lin et al., 2021). Given that liquid diet may affect microbiota and microbial metabolites (Narula et al., 2018; Ashton et al., 2019), and nutrient contents, caloric density, food osmolarity, and fibers may affect motility, we included both control and TNBS rats in each diet group (regular pellet food or liquid diet).

Immediately after rats were euthanized, the colon was isolated, and collected in fresh and carbogenated Krebs buffer (in mmol/l: 118 NaCl, 4.7 KCl, 2.5 CaCl2, 1 NaH2PO4, 1.2 Mgcl2, 11 days-glucose, and 25 NaHCO3). After cleansing, the colon was opened along the mesenteric border, and pinned flat in a Petri dish with Sylgard base (Shi and Sarna, 2004; Shi et al., 2011b; Geesala et al., 2022). Different colonic sites (site I, site P, and site D) of TNBS treated rats were identified as shown in the figure (Figure 1), and colonic tissue of 1–2 cm-long was taken from each site for molecular and functional studies. The colonic site of inflammation where TNBS was instilled to (typically 5–7 cm from the anus) was identified as site I. The colonic portion 2–3 cm proximal to the oral margin of inflammation was identified as site P. The colonic portion 1–2 cm distal to the aboral margin of inflammation was considered site D. For saline treated control rats, the counterpart colon tissues were taken accordingly. The full-thickness tissue, mucosa/submucosa, and muscularis externa were prepared from each site as described previously (Shi and Sarna, 2004; Shi et al., 2011b; Geesala et al., 2022). The muscularis externa tissues were processed immediately for muscle bath experiments. Other tissue samples were frozen in liquid nitrogen and stored at −80°C for molecular studies. Some full thickness samples were saved in 10% formaldehyde for immunohistochemistry study.

Immunohistochemistry staining of COX-2 was performed as described previously (Shi et al., 2011a), on formalin-fixed, paraffin-embedded colon segments of sham control and rats with stenotic CD-like colitis. Sections of 4-μm thickness were blocked with 5% normal goat serum in PBS for 20 min at room temperature, and incubated with the rabbit anti-COX-2 antibody (1:200, Cayman Chemical) and a biotin-conjugated anti-rabbit secondary antibody (Vector Laboratories, Burlingame, CA). After being incubated with avidin-biotin complex (Vector kit, Vector Laboratories), the sections were stained in diaminobenzidine tetrahydrochloride with 0.03% hydrogen peroxide. As a negative control, sections of the same specimens were processed by the same method but omitting the anti-COX-2 primary antibody.

Total RNA was extracted from tissues using the Qiagen RNeasy kit (Qiagen, Valencia, CA). One microgram of total RNA was reverse transcribed by using the SuperScript III First-Strand Synthesis System (Invitrogen, Carlsbad, CA) (Shi et al., 2007; Shi and Sarna, 2008; Shi et al., 2011b). Real-time quantitative RT-PCR was performed using the Bio-Rad CFX96 Real-Time PCR system (Hercules, CA), as described previously (Lin et al., 2021; Geesala et al., 2022). The TaqMan probe for detection of rat COX-2 (Rn00568225-m1) was purchased from Invitrogen. The fold-change relative to control was calculated with the comparative C_T (ΔΔCT) method with endogenous reference 18S rRNA (Part no. 4352930E, Applied Biosystems) as the normalizer.

Rat colonic muscularis externa tissue was homogenized in cold PBS (in mmol/l 137 NaCl, 2.7 KCl, 10 Na2HPO4, KH2PO4, pH 7.4) supplemented with protease inhibitors for protein extraction. PGE₂ was measured with the PGE₂ enzyme immunoassay kit from Cayman Chemical (Ann Arbor, MI) by following the manufacturer’s protocols (Shi et al., 2011a; Lin et al., 2012; Lin et al., 2021).

Freshly obtained colon segments were opened along the mesenteric border, cleansed, and pinned flat in a Petri dish with Sylgard base in carbogenated Krebs solution. The muscularis externa layer was collected for muscle bath experiments. The smooth muscle strips (3 mm × 10 mm) were mounted along the circular muscle orientation in individual muscle baths (Radnoti Glass, Monrovia, CA) filled with 10 mL of carbogenated Krebs solution at 37°C. The contractile activity was recorded as previously described (Shi and Sarna, 2008; Lin et al., 2012; Hegde et al., 2022) with Grass isometric force transducers and amplifiers connected to Biopac data-acquisition system (Biopac Systems, Goleta, CA). The muscle strips were equilibrated in the muscle bath under 1 g tension for 60 min at 37°C before they were tested for contractile response to cholinergic activation. In the last 10 min of the equilibration period, we assessed the spontaneous contractile activity of the muscle strips by taking the average of two 2-min measurements of area under contractions (AUC) with one measurement at the time with highest contractile activity and another with lowest contractile activity of the strips. The smooth muscle contractility in response to cholinergic activation was tested by obtaining concentration-response curves to acetylcholine (ACh 10^–6 to 10^–2 M) in the muscle bath. ACh doses were tested one by one incrementally with the lowest dose first and the highest dose last with an interval of 2–3 min between each dose of ACh. The ACh-evoked contractile response was quantified as the increase of AUC during 2 min after addition of each ACh dose to the bath over the baseline AUC during 2 min before the addition of the lowest dose of ACh. The tissue dry weight of each muscle strip was used to normalize the contractile activities.

All data points are expressed as means ± SEM. Statistical analysis was performed by analysis of variance with non-repeated measures (by Student-Newman-Keuls test) for comparisons of multiple groups and Student’s t-test for comparisons of two groups. A p-value of ≤0.05 was considered statistically significant.

When rats were fed with regular pellet food, intracolonic instillation of TNBS led to localized transmural inflammation with stenosis in the instillation site (site I) and lumen distention in the site proximal to inflammation (site P). Though mechanically distended, site P did not show visible inflammation. Interestingly, the colonic site distal to inflammation (Site D) showed neither distention nor inflammation (Figures 1A,B). However, when rats were fed exclusively with liquid diet, mechanical distention in site P was eliminated in the TNBS induced colitis rats (Figures 1A,B, right panel). Images are representative of 5 rats in each group.

With the CD-like colitis model, we first determined whether colonic smooth muscle contractility was altered at the three distinct sites (I-site, P-site, D-site) in the colitis rats compared to sham controls when they were fed with regular solid food (Figures 2A–C). We found that colonic circular muscle contractile response to cholinergic agent acetylcholine (ACh 10⁻⁶ ∼ 10⁻² M)) was dramatically suppressed in the I-site compared to the corresponding site of control rats (p < 0.05 vs. sham control. N = 5 each group) (Figure 2B). This result is consistent with previous reports (Shi et al., 2011b). More importantly, the muscle contractility was significantly suppressed also in the P-site, which is mechanically distended but not with obvious inflammation (Figure 2A). However, the muscle contractility did not show any significant suppression in the D-site, which presents neither inflammation nor distention (Figure 2C). The site-specific changes of smooth muscle contractility suggest that mechanical stress, rather than inflammation itself, may be an independent pathogenic factor contributing to muscle dysfunction in Crohn’s-like colitis.

Liquid diet treatment was found to release fecal retention in constipation (Lin et al., 2021). In the present study, we found that liquid diet treatment also eliminated lumen distention in the CD-like colitis model in which inflammatory stenosis would lead to massive fecal retention and lumen distention if rats were fed with solid food (Figure 1).

We then determined the site-specific changes of muscle contractility in colitis when rats were fed exclusively with liquid diet and mechanical distention was eliminated in the colitis rats. Our results showed that muscle contractility was significantly improved in the I-site and largely normal in the P-site in inflammation when the rats were fed with liquid diet (Figures 2A,B). Interestingly, at the non-distended D-site of the colon, there was no significant change in SMC contractility in inflammation in rats fed either with regular solid diet or liquid diet (Figure 2C).

COX-2, via COX-derived prostaglandins, has profound effects on smooth muscle function in the gut (Li et al., 2012). Our previous studies demonstrated that expression of COX-2 is highly sensitive to mechanical stress in the gut, as circumferential mechanical stretch (i.e., lumen distention) leads to mechano-sensitive expression of COX-2 exclusively in gut smooth muscle cells in a model of mechanical bowel obstruction (Shi et al., 2011a; Shi, 2017). To determine if COX-2 expression is also induced in stenosis-associated distention in the colitis model, we then determined the site-specific expression of COX-2 mRNA in sites P, I, and D in the Crohn’s-like colitis model. When rats were fed with regular solid food, COX-2 mRNA expression increased 13.1 (±2.5)-fold in the I-site of colitis rats, compared to sham controls (p < 0.05 vs. sham control. N = 4 or 5) (Fig. 3Aa). Furthermore, COX-2 mRNA expression was increased significantly also in the mechanically distended P-site [16.6 (±4.3)-fold] (Figure 3A). However, when rats were given a 7-day regimen of exclusive liquid diet to eliminate mechanical stress in the model, the level of COX-2 mRNA expression was dramatically attenuated in the inflamed I-site and the distended P-site (N = 5) (Fig. 3Ab). Importantly, expression of COX-2 mRNA was not significantly changed in the non-distended D-site in the CD-like colitis rats even if the rats were fed with solid food (Figure 3A). This data indicates that expression of COX-2 in the P-site is largely dependent on mechanical stress. Lastly, we wanted to determine where COX-2 expression occurred within the colonic tissue layers in the P-site, and found that increased COX-2 expression occurred mainly in the muscularis externae (ME) layer of the distended colon segment, particularly in the colonic smooth muscle cells (N = 4) (Figures 3B,C).

PGE₂ is among the most potent prostaglandins derived from COX-2 and found to suppress colonic smooth muscle contractility in rats (Lin et al., 2012). We next determined PGE₂ levels in different sites of the colonic tissues in sham and colitis rats treated with solid food and liquid diet. When rats were fed with solid food, PGE₂ levels were increased significantly not only in the I-site (20,653 ± 4,233 pg/μg of protein, p < 0.01. N = 5 all groups) but also in the P-site (13,708 ± 3,450 pg/μg, p = 0.01. N = 5) of the colitis colon, compared to sham colon (5,776 ± 1,813 pg/μg) (Figure 4). However, the D-site (6,555 ± 1958 pg/μg) did not exhibit any significant increase of PGE₂. Interestingly, liquid diet treatment blocked the PGE₂ increase in the P-site and significantly suppressed PGE₂ level in the I-site (10,306 ± 2,046 pg/μg, p < 0.01 vs. the I-site level in solid food-treated rats) (Figure 4).

Lastly, we determined if COX-2 plays a role in smooth muscle dysfunction in the CD-like colitis model. COX-2 inhibitor NS-398 (5 mg/kg, i. p. daily) or vehicle control was administered in sham and colitis rats daily. Rats were fed with regular pellet food and euthanized 7 days later. The body weight was decreased significantly over the 7-day course in TNBS-induced colitis rats when treated with vehicle (Figure 5A P = 0.02 vs. sham. N = 4 or 5 rats in each group). However, when rats were administered with NS-398, TNBS colitis rats did not demonstrate any significant body weight decrease compared to sham control rats (Figure 5A). NS-398 administration did not significantly affect body weight change in sham control rats (p > 0.05 between sham/vehicle and sham/NS-398). We also assessed inflammation in the rats. NS-398 showed a trend to reduce inflammation in TNBS colitis rats (p = 0.064 vs. TNBS colitis with vehicle administration. N = 4 or 5) (Figure 5B). However, even with NS-398 administration, the colon inflammation in the TNBS colitis rats was still obvious (p < 0.01 vs. sham control.) (Figure 5B).

The baseline contractile activity of colonic muscle strips was significantly decreased in TNBS colitis rats in both P-site and I-site in the DMSO vehicle treatment group (p < 0.01 vs. sham control. N = 5 each group) (Fig. 6Aa and Ab). However, with NS-398 administration, the baseline contractile activity in the P-site of TNBS colitis rats was not different from sham control (p > 0.05 vs. sham control. N = 5) (Fig. Aa). The baseline activity of the I-site was still significantly reduced compared to sham control (p < 0.01) even with NS-398 treatment. However, NS-398 treatment improved contractile activity in the I-site compared to the same site with vehicle treatment (p = 0.03) (Fig. 6Ab). Representative traces of colonic baseline contractile activity were shown for each group (Figure 6B). NS-398 administration did not significantly affect colonic baseline activity in sham control rats (p > 0.05 between sham/vehicle and sham/NS-398).

We also studied muscle contractile response to cholinergic agonist ACh and found that in the TNBS-treated colitis rats given a DMSO vehicle control, there was a significant reduction of muscle contractility in response to all ACh concentrations (10^–6 to 10^–2 M) in both the I-site and the P-site (p < 0.05, N = 4 or 5) (Figure 7 Aa and Ab). Interestingly, when TNBS-treated rats were given an i. p administration of NS-398, there was no significant difference (p < 0.05, N = 4 or 5 rats) in muscle contractility at the distended P-site for all ACh concentrations, indicating that COX-2 may play a critical role in smooth muscle dysfunction in the pre-inflammation site (Fig. 7Aa). NS-398 treatment also partially improved muscle function in the inflammation site (I-site), as smooth muscle contractility at the I-site was not significantly different between sham and TNBS colitis rats for all ACh concentrations, except that of ACh 10⁻² M (p < 0.05, N = 4 or 5 rats) (Fig. 7Ab). Compared to vehicle treatment, NS-398 administration decreased the contractile response of control colons to ACh 10⁻⁴ and 10⁻³ (M) (Fig. 7Aa and Ab). Representative traces of smooth muscle contractile response to ACh doses were shown for each group (Figure 7B). We confirmed that NS-398 treatment effectively blocked PGE₂ production at P-site (6,328 ± 1,452 pg/μg vs. 3,120 ± 1,150 pg/μg in sham with NS-398, p = 0.10) and significantly decreased PGE₂ level at I-site (9,260 ± 1,610 pg/μg) compared to the I-site level in rats with no NS-398 treatment (20,653 ± 4,233 pg/μg, p < 0.05. N = 5).