Endoplasmic Reticular Stress and Pathogenesis of Experimental Colitis: Mechanism of Action of 5-Amino Salicylic Acid

Inflammatory bowel diseases (IBDs), such as Crohn’s disease and ulcerative colitis, are chronic inflammatory conditions of the gastrointestinal tract which arise from an exaggerated immune response to the normal gut microflora in genetically predisposed individuals [1‒3]. These conditions are associated with ATP deficits, increased oxidative stress, and altered protein synthesis, contributing to structural damage in the rough endoplasmic reticulum (rER), Golgi apparatus and mitochondria [4‒7]. Damage to rER impairs protein folding, leading to the onset of ER stress. To mitigate this and restore the rER homeostasis, the unfolded protein response (UPR) signaling pathway is activated. The UPR signaling is mediated by three ER membrane-bound receptors: PERK (PKR-like ER kinase), IRE1 (inositol-requiring enzyme 1), and ATF6 (activating transcription factor 6) [8‒11]. Under normal conditions, these receptors are bound and inactivated by GRP78 (glucose-regulated protein-78) chaperone [12, 13]. During ER stress, GRP78 binds to unfolded proteins, freeing these receptors to initiate UPR signaling. Once activated, PERK phosphorylates the eukaryotic initiation factor-2α (eIF2α) and produces its phosphorylated form (eIF2αP). While, eIF2αP globally inhibits protein synthesis, it selectively allows the expression of ER stress-related genes such as GRP78, ATF4, and ATF5 [14, 15]. ATF6, upon release from GRP78, translocates to the Golgi apparatus, where it is cleaved into an active N-terminal fragment (ATF6N) which together with ATF4, regulates apoptosis through the transcription factor CHOP. IRE1 once activated splices XBP1 mRNA generating a potent transcription factor that enhances cell’s ability to alleviate ER stress [16]. ER stress contributes to IBD pathogenesis by disrupting the intestinal barrier, thereby exposing the underlying lamina propria to microbial antigens [13, 14, 17, 18].

To manage IBD, patients are typically treated with 5-aminosalicylic acid (5-ASA) which induces and maintains remission [19‒21]. While 5-ASA exerts its effects by inhibiting NF-κB, TNFα, and modulating cyclooxygenase-2 and inducible nitric oxide synthase [22], its impact on UPR signaling in IBD remains poorly understood. Since GRP78 is a key regulator of the UPR signaling, we quantified its levels along with downstream mediators such as eIF2α, eIF2αP, and ATF4 in experimental colitis induced in male Sprague-Dawley rats via intrarectal administration of trinitrobenzene sulfonic acid (TNBS) dissolved in 50% ethanol. In the treatment group, an aqueous suspension of 5-ASA (100 mg/kg body weight, oral) was administered daily. On day 7 post-TNBS, animals were sacrificed, and colonic tissue samples collected and analyzed.

Trinitrobenzene sulfonic acid (Fluka, Switzerland), O-dianisidine, and hexadecyltrimethylammonium bromide (Sigma-Aldrich, Germany) were of research grade. Primary antibodies selective for IRE1, PERK, GRP78, eIF2αP, ATF4, and β-actin, and the secondary antibody anti-mouse immunoglobulin coupled with horse-radish peroxidase (IgG-HRP), anti-rabbit IgG-HRP were purchased from Abcam, UK, Anti-eIF2α antibodies were obtained from Santacruz, USA, and enzyme linked immune sorbent assay (ELISA kit, MBS3808179) for rat CHOP from LifeSpanBioScience Inc, USA.

Male Sprague-Dawley rats (200–250 g) were used in this study and maintained according to internationally accepted principles for laboratory animal care, issued by the Animal Care Facility, College of Medicine, Kuwait University. The animals were provided with a free access to feed and water.

TNBS (30 mg) dissolved in 250 μL of 50% ethanol was injected into the colon through anus and followed for 7 days [23]. The animals were studied into 3 groups, C (non-colitis control), T (colitis induced by TNBS) and TA (colitis + 5-ASA). A freshly prepared aqueous suspension of 5-ASA in phosphate buffered saline (PBS, pH 7.4) was orally administered daily, starting 2 h before inducing colitis and continuing until day 7 post-TNBS. On day 7, the animals were sacrificed by cervical dislocation, colons collected and stored at −70°C for analysis. The percentage body weight change on day 7 was calculated relative to the body weight on day 0. Colon length and weight were also measured and reported as weight per unit length (gm/cm) of colon [23].

Myeloperoxidase (MPO) activity was measured using a colorimetric method [24] and expressed as enzyme units per mg colonic tissue. Enzyme unit is defined as the amount of enzyme that converts 1 nmol of H₂O₂ into water.

Routine histochemical procedures were used to measure histologic scores in formaldehyde-fixed thin (5 µm) colonic tissue sections stained with hematoxylin and eosin and periodic acid Schiff [5, 25]. The sections were viewed under a light microscope and photographed using a microscope [OLYMPUS-BX51TF, Japan] attached with DP71 digital camera. Epithelial interruptions were quantified directly under the microscope at ×400 magnification, with six stained sections from each rat in all groups (n = 5/group). Inflammatory cells and goblet cells were counted from photomicrographs taken at ×100 magnification, while colon wall thickness was quantified from photomicrographs taken at ×20 magnification.

For inflammatory cell quantification, six micrographs per section were analyzed, counting cells in a 100 × 100 µm² area and expressing the results as number/mm². Goblet cells were counted in randomly selected photographs of six intestinal glands per section. The severity of colitis was categorized as follows.

It was characterized by a lower number of inflammatory cells in the lamina propria, an intact mucous membrane with few small discontinuities, normal sized intestinal glands with increased number of goblet cells in intestinal glands, and moderate increase in colon wall thickness.

This was characterized by inflamed mucosa with enlarged intestinal glands and prominent goblet cells, large discontinuities in mucous membrane connecting the lamina propria to the lumen, densely packed inflammatory cells extending into submucosa, and enormously increased thickness of the colon wall.

The expression of IRE1, GRP78, ATF4, eIF2α, and eIF2αP proteins was measured in crude lysates using ECL Western blot analysis. Membrane-bound PERK protein was estimated using crude microsomes prepared as described earlier [23]. Total protein concentrations in lysates and crude microsomes were determined using a protein dye-binding assay kit (BioRad). Samples (2 mg/mL protein) were electrophoresed on an 8% polyacrylamide gel and electroblotted onto PVDF membranes for probing with primary antibodies: anti-PERK (dilution 1:100), anti-IRE1 (dilution 1:200), anti-GRP78 (dilution 1:500), anti-ATF4 (dilution 1:300), anti-eIF2α (dilution 1:500), and anti-eIF2αP antibodies (dilution 1:500). After washing, the membranes were incubated with the secondary antibodies-HRP conjugates (dilution 1:2,000), and protein bands developed using an ECL kit (Amersham, UK). Band densities were recorded with a Genetic Analyzer (Alpha-Imager HP).

CHOP protein concentrations were measured in colonic tissue lysate (40 µg per well) using an ELISA kit (LifeSpanBioScience Inc., USA) according to the manufacturer’s instructions. The signal was developed using a chromogen supplied with the kit, and absorbance was measured at 450 nm using a spectrophotometer. The level of CHOP was expressed as ng/mL.

To validate the ECL Western blot data for eIF2αP expression, thin tissue sections (5 µm) were blocked, and the antigens were retrieved [25]. Sections were then incubated overnight with anti-eIF2αP primary antibody (1:100 dilution) in the PBS-Triton-BSA solution at 4°C. After washing with PBS, sections were incubated with the secondary antibody conjugated with a green-fluorescent dye (1:200 dilution, Alexa flour 488) [25]. Fluorescence levels were measured in 5 regions of interest per slide using a fluorescence microscope (Zeiss, LSM 980, Germany).

Whole tissue lysate samples (n = 5) from test conditions were pooled and solubilized with a 4x solubilization buffer [24], and incubated with 10 µL of protein-A Sepharose^® beads (Pharmacia Biotech AB, Sweden) for 20 min. After centrifugation, supernatants were incubated overnight with 10 µL of anti-GRP78 antibody, followed by a 1-h incubation with 25 µL of protein-A beads at 4°C with gentle shaking. The resulting pellets were suspended in 1x sample buffer, heated in a boiling water bath for 5 min, and then cooled on the ice [24]. The samples were separated into an 8% polyacrylamide gel, transferred to a PVDF membrane, and probed with anti-PERK antibodies by ECL Western blot analysis. Band densities were recorded to calculate the expression level of GRP-bound PERK.

Total RNA from colonic segments was extracted using a TRIzol kit (Invitrogen^™, USA) following the manufacturer’s instructions. RNA quality was analyzed using a nano-drop method, and agarose gel electrophoresis (data not shown). The level of eIF2α mRNA expression was quantified using a SYBR-green real-time RT-PCR using QuantiTect^® reverse transcription kit (QIAGEN) and specific primers with β-actin as an internal control. eIF2α selective upstream (1,501–1,515 bp) and downstream (1,723–1,740 bp) primers were selected using the published sequence (NM_001109339.1). The β-actin selective upstream (841–858 bp) and downstream (970–987 bp) primers were selected using a published sequence (NM_031144.3). C_T values were used to calculate the eIF2α:β-actin mRNA ratio.

The data have been presented as mean ± SE of duplicate determinations (n = 10). A non-parametric, unpaired, two-tailed t test was performed. The data were compared with the respective controls, and a value of p < 0.05 was considered to be statistically significant.

The animals with colitis showed a 13% body weight loss (p < 0.05) on day 7, whereas non-colitis controls gained 13% body weight (p < 0.05) compared to their day 0 body weights (Fig. 1a). 5-ASA treatment significantly (p < 0.05) inhibited the body weight loss (Fig. 1a). The weight of the inflamed colon was significantly (p < 0.05) increased compared to the non-inflamed colon (Fig. 1b), while 5-ASA treatment significantly (p < 0.05) reduced the colon weight (Fig. 1b).

MPO activity was also significantly higher in the inflamed colon compared to the non-inflamed controls (Fig. 1c). 5-ASA treatment significantly (p < 0.05) reduced MPO activity in the animals with colitis (Fig. 1c).

TNBS increased the mucosal and muscularis externa thickness, disrupted the epithelial cell lining and reduced the number of goblet cells (Fig. 2, 3). Inflammatory cell infiltration was more pronounced in the submucosa of inflamed colon compared to the non-colitis controls (Fig. 2, 3). 5-ASA treatment of colitis rats restored normal tissue architecture and increased mucin staining in goblet cells (Fig. 2, 3). TNBS induced severe colitis in 5 out of 6 animals and mild colitis in 1 out of 6 animals. Following 5-ASA treatment, 4 out of 6 animals showed mild colitis, while 2 out of 6 had severe colitis.

The expression of PERK protein (110 kDa), calculated relative to β-actin (PERK:β-actin), was significantly (p < 0.05) higher in inflamed colon compared to non-inflamed controls and was reduced following the 5-ASA treatment (Fig. 4). The expression of ATF4 (76 kDa) protein relative to β-actin (ATF4:β-actin ratio) was also significantly (p < 0.05) higher in inflamed colon compared to the non-inflamed controls (Fig. 4) and was significantly (p < 0.05) decreased by the 5-ASA treatment (Fig. 4).

Similarly, GRP78 (78 kDa) protein expression relative to β-actin (GRP78:β-actin) was significantly higher (p < 0.05) in inflamed colon compared to the non-colitis controls and was decreased by the 5-ASA treatment (Fig. 4). However, the expression of CHOP (Fig. 5a) and IRE1 (Fig. 5b) did not change in the test conditions; therefore, the focus of further experiments was on the PERK signaling pathway.

The levels of GRP78-bound-PERK were significantly decreased in TNBS-inflamed colon compared to the non-inflamed controls and were significantly reversed by treatment with 5-ASA (Fig. 6).

The eIF2α and eIF2αP proteins (36 kDa) were quantified relative to β-actin (eIF2αP:β-actin) and eIF2α (eIF2αP:eIF2α). The expression of eIF2αP protein was significantly (p < 0.05) higher in inflamed colon compared to the non-inflamed controls and was significantly reduced by treatment with 5-ASA (Fig. 7). The levels of eIF2α (eIF2α: β-actin) remained unchanged under the test conditions (Fig. 7).

Immunofluorescence microscopy validated the ECL Western blot data for eIF2αP (Fig. 8a, b). The expression of eIF2αP protein was increased significantly (p < 0.05) in the epithelial cell cytosol in inflamed colon compared to the non-inflamed controls (Fig. 8a, b); however, it was significantly (p < 0.05) reduced by 5-ASA treatment (Fig. 8a, b).

Total RNA yields (mg/gm tissue) were similar across groups: 3.8 ± 0.6 in the non-colitis controls, 3.7 ± 0.9 in colitis group of animals, and 3.6 ± 0.7 in the 5-ASA treated colitis group, indicating no change in RNA yield (p > 0.05). The quality of RNA, assessed by 18s and 28s rRNA band intensities, was consistent across test conditions (not shown). The levels of eIF2α:actin mRNA were comparable: 1.7 ± 0.16 in the inflamed colon, 1.7 ± 0.11 in non-colitic controls, and 1.8 ± 0.07 in the 5-ASA treated animals.

IBDs such as Crohn’s disease and ulcerative colitis are associated with chronic inflammation, induction of free radicals and oxidative stress which damages subcellular structures including rER, Golgi and mitochondria [5]. The accumulation of unfolded proteins in IBD overwhelms the rER, leading to “ER stress” which triggers the UPR-signaling pathway [13]. This study focused on the role of the PERK signaling, a subset branch of the UPR-pathway in an experimental colitis model. A dose (100 mg/kg body weight) of 5-aminosalicylic acid (5-ASA), which was selected based on previous studies [26] was found to be effective in mitigating colitis in this model. The confirmation of inflammation was important to ensure that the changes in ER stress protein expression were inflammation-induced. Our results confirmed the development of colitis and its inhibition by 5-ASA in the present model of colitis.

ER stress marker proteins GRP78, eIF2α, eIF2αP, and ATF4 play a key role in UPR-signaling pathway. Normally, GRP78 binds and keeps PERK inactive, but the unfolded proteins sequester GRP78, leading to PERK activation. As IRE1 and CHOP did not change, the focus of this study, therefore, was on the role of PERK signaling and its modulation by 5-ASA.

The observed increase in PERK expression, suggests accumulation of unfolded proteins in the inflamed colon. In other systems, PERK inhibition reduces ER stress-induced inflammation [10, 27], further supporting a role for PERK in the pathogenesis of colitis. Our findings showing a decreased level of GRP78-bound PERK also suggest an increased accumulation of unfolded proteins and subsequent PERK activation. PERK kinase preferentially phosphorylates the eIF2α, converting it to eIF2αP [28]; therefore, increased level of eIF2αP is consistent with enhanced PERK signaling in the present model.

Next, we investigated mechanism underlying the induction of eIF2αP by examining eIF2α mRNA levels using a real-time RT-PCR method. Normally, eIF2αP acts as a global inhibitor of translation to reduce ER stress. Our ECL Western blot findings validate the immunofluorescence microscopy data for eIF2αP. Despite increased expression of eIF2αP, the level eIF2α mRNA remained unchanged under the test conditions, suggesting a post-transcriptional regulation of the unphosphorylated eIF2α.

In light of the increased level of eIF2αP, it is important to discuss the mechanism driving the increased expression of GRP78 and PERK proteins. Cells have evolved an escape mechanism to selectively support the translation of ER stress-related genes through activation of the transcription factor ATF4 [9]. Therefore, increased ATF4 expression observed in this study likely underlies the upregulation of GRP78 and PERK proteins. These changes are inhibited by the 5-ASA treatment, indicating that its effects are mediated through the suppression of the PERK-signaling pathway in the present model of colitis.

We have demonstrated that colonic tissues used in this study were inflamed, 5-ASA exhibited anti-inflammatory properties, and TNBS induced the ER stress in this model of colitis. Our findings suggest that the ameliorative effects of 5-ASA are mediated through inhibition of the PERK-signaling pathway.

Financial support was provided by Kuwait University through a grant (YM 03/19) and the College of Graduate Studies. We acknowledge the Health Sciences Research Core Facility for using instrumentation and thank Ms. Amna Al-Shamali for technical support.

This is an animal study conducted on male Sprague-Dawley rats approved by the Health Sciences Research Ethical Committee, Kuwait University (20/VDR/EC). This study does not involve human data. All co-authors have agreed to publish this work; no other consent from any other party is required.

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this article.

The funding was provided by Kuwait University Research Sector through a grant (YM 03/19) and by the College of Graduate Studies, Kuwait University.

Zahraa A. Baydoun conducted experiments, data collection, and analysis, Muddanna Rao assisted in histochemistry, immunofluorescence experiments, and data interpretation. Islam Khan conceived the idea and supervised the overall project, interpretation, and manuscript writing in collaboration with Zahraa A. Baydoun and Muddanna Rao.

Data sets were used to calculate mean and SE and used to make Figures have been presented in this manuscript.