Background: Proper formation and maturation of the intestinal brush border is critical for adequate digestion, nutrition, and host defense. Mature brush borders feature specific localization of several proteins involved in actin bundling (e.g., Vil1, Eps8, Misp), plasma membrane cross-linking (e.g., Ezr), the intermicrovillar adhesion complex (e.g., Ush1c), and the terminal web (e.g., Pls1, Myh14). The Caco-2-BBE cell line has been a gold standard reagent for brush border studies; however, it was isolated from colorectal adenocarcinoma tissue. Here, we aimed to develop an in vitro brush border model using primary epithelial cells from the small intestine.
Methods: Spheroid epithelial cell lines were established from mouse ileal tissue and cultured as embedded in Matrigel with L-WRN conditioned medium to maintain proliferating cells or with base medium lacking growth factors to promote differentiation. Matrigel-coated Transwell membranes were seeded with 2x10^5 spheroid-derived cells and cultured as submerged in L-WRN medium or for up to 21 days at the air-liquid interface (ALI), i.e., medium in bottom chamber only. Protein expression was assessed by immunofluorescence and super-resolution spinning disk confocal microscopy. Gene expression was determined by qPCR. Epithelial cell function was tested by alkaline phosphatase activity assay.
Results: Vil1, Ezr, and Eps8 were properly localized (e.g., akin to villus enterocytes in vivo) in spheroids and submerged monolayers, as expected based on published reports. However, we were surprised that intermicrovillar adhesion complex and terminal web proteins did not properly localize in these culture conditions, even with medium reported to promote enterocyte differentiation. We next assessed monolayers cultured at the ALI. ALI monolayers exhibited barrier formation, with high TEER and low FITC-dextran permeability values. We found that >70% of the epithelial cells properly expressed Ezr and Vil1 by ALI Day 0, Myh14 and Ush1c by ALI Day 7, and Eps8, Pls1, and Misp by ALI Day 14. Concomitant with emergence of proper brush border protein localization, we observed microvillus lengthening (0.59 µm submerged vs. 1.10 µm ALI Day 7 and 1.26 µm ALI Day 21) and increased mRNA expression of villus enterocyte markers (Nlrp6, Ada, Alpi, and Ace2). We also observed increased alkaline phosphatase activity in ALI monolayers (average fold vs. submerged of 9.1, 7.2, and 5.0 for ALI Day 7, 14, and 21, respectively), indicating increased functional maturity of enterocytes.
Conclusions: Spheroid-derived intestinal epithelial cell monolayers cultured at the ALI represent a new in vitro model to enable functional studies of the brush border. Our data supports the use of early ALI time points (<7 days) to model immature brush borders and later ALI time points (≥14 days) to model mature brush borders.