Hypoxia and heat stress affect epithelial integrity in a Caco-2/HT-29 co-culture

Abstract

Hypoxia and hyperthermia, which can be induced by high environmental temperature or strenuous exercise, are two common stressors that affect intestinal epithelial integrity and lead to multiple clinical symptoms. In this study, we developed an in-vitro intestinal monolayer model using two human colonic epithelial cell lines, Caco-2 and HT-29, co-cultured in Transwell inserts, and investigated the effects of heat treatment and/or hypoxia on the epithelial barrier function. The monolayer with a ratio of 9:1 (Caco-2:HT-29) showed high trans-epithelial electrical resistance (TEER), low Lucifer Yellow permeability and high mucin production. Hyperthermia and/or hypoxia exposure (2 h) triggered heat shock and oxidative stress responses. HSP-70 and HSF-1 protein levels were up-regulated by hyperthermia, which were further enhanced when hyperthermia was combined with hypoxia. Increased HIF-1α protein expression and Nrf2 nuclear translocation was only caused by hypoxia. Hyperthermia and/or hypoxia exposure disrupted the established monolayer by increasing paracellular permeability, decreasing ZO-1, claudin-3 and occludin protein/mRNA expression, while enhancing E-cadherin protein expression. Tight junction protein distribution in the monolayer was also modulated by the hyperthermia and/or hypoxia exposure. In addition, transcription levels of mucin genes, MUC-2 and MUC-5AC, were increased after 2 h of hyperthermia and/or hypoxia exposure. In conclusion, this Caco-2/HT-29 cell model is valid and effective for studying detrimental effects of hyperthermia and/or hypoxia on intestinal barrier function and related heat shock and oxidative stress pathways and can be used to investigate possible interventions to reverse hyperthermia and/or hypoxia-induced intestinal epithelial injury.

Figure 1

The monolayer integrity and mucus expression in the co-culture model at different ratios of Caco-2 and HT-29 cells. (A) TEER values of the cells grown on Transwell membranes. The cell cultures were subjected to TEER evaluation every other day since day 9. (B) Areas under the curves of TEER values of the co-culture monolayer calculated for statistical analysis. (C) Lucifer Yellow permeability assay of different combinations of Caco-2 and HT-29 cells at day 17. (D) Alcian Blue/nuclear fast red staining on Transwell membranes acquired by Nikon Plan 10 × /0.25 objective lens. Pink areas indicated nuclei and blue/purple areas indicated deposition of mucus. Mucin-related mRNA MUC-2 (E) and MUC-5AC (F) levels assessed by qRT-PCR. The target genes were normalized with reference gene β-actin. All values were presented as means ± SEM (N = 3, n = 3). Statistical differences were analyzed by Two-way analysis of variance (ANOVA), with Bonferroni post-hoc test. **p < 0.01, ***p < 0.001, ****p < 0.0001; significant different from Caco-2:HT-29 = 1:0 group.

Figure 2

Effects of hypoxia ± heat treatment on cell viability and monolayer integrity in the co-culture of Caco-2 and HT-29 cells. (A) LDH release of Caco-2 and HT-29 cells co-culture after 2 h of hypoxia ± heat treatment. In the positive control group, the cells were lysed with lysis buffer. (B) TEER values of the co-culture monolayer before and after 2 h of hypoxia ± heat treatment. (C) Lucifer Yellow permeability assay of the co-culture monolayer after 2 h of hypoxia ± heat treatment. All values were presented as means ± SEM (N = 3, n = 3). Statistical differences were analyzed by two-way ANOVA followed by the Bonferroni’s multiple comparison test. ****p < 0.0001; significant different from control (A, B and D). Means without a common letter differ at p < 0.05 (C). posit. ctrl positive control; hyp. hypoxia.

Figure 3

Effects of hypoxia and/or heat treatment on resilience pathways-related proteins in the co-culture model. Relative protein expression of HSP-70 (A), HSF-1 (C), HIF-1α (D) and cytoplasmic (E) and nuclear Nrf2 (F) assessed by Western Blot. All target proteins were normalized to reference protein β-actin (total/cytoplasmic) or histone H3 (nuclear). (B) HSP-70 mRNA levels assessed by qRT-PCR. The target genes were normalized with housekeeping gene β-actin. All values were presented as means ± SD (N = 3, n = 3). Statistical differences were analyzed by two-way ANOVA followed by the Bonferroni’s multiple comparison test. Means without a common letter differ at p < 0.05.

Figure 4

Effects of hypoxia and/or heat treatment on TJ/AJ proteins in the co-culture model. Relative protein expression of ZO-1 (A), occludin (B), claudin-3 (C) and E-cadherin (D) assessed by Western Blot. All target proteins were normalized to reference protein β-actin. ZO-1 (E), occludin (F), claudin-3 (G) and E-cadherin (H) mRNA levels were assessed by qRT-PCR. The target genes were normalized to housekeeping gene β-actin. All values were presented as means ± SD (N = 3, n = 3). Statistical differences were analyzed by two-way ANOVA followed by the Bonferroni’s multiple comparison test. Means without a common letter differ at p < 0.05. OCLD occludin; CLDN3 claudin-3; E-cad E-cadherin.

Figure 5

Effects of hypoxia and/or heat treatment on TJ localization in the co-culture model. Red or green color: anti-ZO-1, anti-OCLD or anti-CLDN3 antibody conjugated with Alexa-Fluor fluorescent secondary antibodies. The images were acquired by Leica TCS SP8 microscope with HCX IRAPO L 25 × /0.95 objective lens at 1.6 × digital magnification, pinhole: 1.5 AU. OCLD occludin; CLDN3 claudin-3.

Figure 6

Effects of hypoxia and/or heat treatment on mucin gene production in the co-culture model. MUC-2 (A) and MUC-5AC (B) mRNA levels were assessed by qRT-PCR and normalized to the housekeeping gene β-actin. All values were presented as means ± SD (N = 3, n = 3). Statistical differences were analyzed by two-way ANOVA followed by the Bonferroni’s multiple comparison test. Means without a common letter differ at p < 0.05.