Stereochemistry of Dihydroxy Bile Acids Dictates Their Role in Barrier Dysfunction in Human Colon T84 Cells

Abstract

Bile acid‐associated diarrhea and epithelial barrier dysfunction are common occurrences in patients with inflammatory bowel diseases. There are very few studies that compare the effects of the dihydroxy bile acids (BAs), chenodeoxycholic acid (CDCA) and deoxycholic acid (DCA), on barrier function, although they are known for their prosecretory properties. We have reported that CDCA induced apoptosis, released reactive oxygen species (ROS), nitrogen species (NO2/NO3) and the pro‐inflammatory cytokine IL‐8, to disrupt tight junctions, while its monohydroxy derivative, lithocholic acid (LCA) did not (FASEB J 2019 33:711.1). In this study, we examined the effects of dihydroxy BAs, DCA, CDCA, and ursodeoxycholic acid (UDCA, 7‐OH epimer of CDCA), on barrier function in human colon carcinoma cells (T84). We hypothesize that the effect of dihydroxy BAs on barrier function will be varied owing to the subtle differences in the position (3, 7, or 12) and stereochemistry (α or β) of the hydroxyl group.Confluent T84 cells (Trans Epithelial Resistance, TER >1000 Ω.cm2) were treated overnight (O/N) with 500μM of CDCA, DCA and UDCA and the effects on barrier function were studied as follows: i) apoptosis (Annexin‐V, flow cytometry), ii) mitochondrial and nuclear ROS production (CellRox, flow cytometry), iii) [NO2/NO3] (Griess Assay, Colorimetry), iv) IL‐8 release (ELISA) and v) paracellular permeability (TER and FITC‐10kD dextran fluxes (F10D)).Apical exposure of CDCA and DCA induced apoptosis while UDCA did not alter cell viability (% Annexin V+ cells, 18 hr: DMSO, 10±3; CDCA: 25±2*; DCA: 26±2*; UDCA: 13±4; n=4, *p<0.05, compared to control). CDCA and DCA increased ROS (% CellRox+ cells: DMSO: 12±0.5; CDCA: 32±0.1; DCA: 36±2, p<0.05) and [NO2/NO3] (μmol/mg protein: control: 68±6; CDCA: 170±18; DCA: 249±28). UDCA alone had no effect, but it attenuated CDCA’s action (% CellRox+ cells: UDCA: 14±0.5; CDCA+UDCA: 12±0.3; [NO2/NO3] UDCA: 65±12; CDCA+UDCA: 48±30; n=4, p<0.05). Similarly, CDCA and DCA increased IL‐8 release (ng/ml: Control: 78±4; CDCA: 578±86*; DCA: 789±98*), while UDCA decreased basal and CDCA induced IL‐8 release (UDCA: 20±5; CDCA+UDCA: 26±4; n=6, p<0.05). Time‐dependent studies (0.5–18 hr) show that CDCA and DCA decreased TER by 70% and 62%, respectively and increased F10D fluxes. In contrast, UDCA alone had no effect on paracellular permeability but restored CDCA‐ and DCA‐induced barrier dysfunction (18 hr, apparent permeability (x 10−9) cm/sec: Control: 6±1; CDCA: 96±10; DCA: 120±20; UDCA: 13±2; CDCA+UDCA: 41±12; DCA+UDCA: 18±2; n≥3, *p<0.05). Comparing the effects of hydrophobic CDCA and DCA on barrier function with that of the hydrophilic 7‐β isomer, UDCA, highlighted the importance of structure and stereochemistry of ‐OH groups in BA action. Therefore, UDCA, like LCA, could be used to ameliorate the deleterious effects of other dihydroxy BAs and serve as a target therapeutic drug for inflammatory and diarrheal diseases.Support or Funding InformationAPS‐STRIDE National Heart, Lung and Blood Institute (Grant #1 R25 HL 115473‐01) to SK and UD, APS‐UGSRF to MH, NSSRP Benedictine Funds to FD, IS, SK, MH, UD, MP, JS, DMR, NSF‐MRI: DB‐1427937 to JS