A microengineered collagen scaffold for generating a polarized crypt-villus architecture of human small intestinal epithelium

Abstract

The human small intestinal epithelium possesses a distinct crypt-villus architecture and tissue polarity in which proliferative cells reside inside crypts while differentiated cells are localized to the villi. Indirect evidence has shown that the processes of differentiation and migration are driven in part by biochemical gradients of factors that specify the polarity of these cellular compartments; however, direct evidence for gradient-driven patterning of this in vivo architecture has been hampered by limitations of the in vitro systems available. Enteroid cultures are a powerful in vitro system; nevertheless, these spheroidal structures fail to replicate the architecture and lineage compartmentalization found in vivo, and are not easily subjected to gradients of growth factors. In the current work, we report the development of a micropatterned collagen scaffold with suitable extracellular matrix and stiffness to generate an in vitro self-renewing human small intestinal epithelium that replicates key features of the in vivo small intestine: a crypt-villus architecture with appropriate cell-lineage compartmentalization and an open and accessible luminal surface. Chemical gradients applied to the crypt-villus axis promoted the creation of a stem/progenitor-cell zone and supported cell migration along the crypt-villus axis. This new approach combining microengineered scaffolds, biophysical cues and chemical gradients to control the intestinal epithelium ex vivo can serve as a physiologically relevant mimic of the human small intestinal epithelium, and is broadly applicable to model other tissues that rely on gradients for physiological function.

Keywords: Crypt; Intestine; Microfabrication; Scaffold; Stem cell; Villus.

Fig. 1

The enteroid culture system converts the highly structured in vivo crypt-villus architecture into non-polarized enteroids with an inaccessible lumen. (A) Schematic showing the isolation crypts from human small intestinal mucosa and their culture as enteroids. Stem and progenitor cells are green while differentiated cells are red. (B) Immunofluorescence staining of freshly obtained human small intestinal tissues showing that the epithelium is composed of an array of crypts/villi with differentiated cells (KRT20⁺) located along the villi while the stem and progenitor cells are confined within the crypts (Olfm4⁺). DNA (nuclei) was stained with Hoechst 33342. (C) Enteroid culture of human small intestine crypts by embedding in Matrigel. Shown are the structures on day 0, 1, and 2 of culture at passage number 0 and 1. (D) EdU and immunofluorescence staining of human enteroids after 4 days in culture. Scale bar = 100 μm.

Fig. 2

Monolayer culture of human small intestinal cells on a cross-linked collagen hydrogel. (A) Schematic of the culture of cells on the cross-linked collagen followed by forced differentiation of the cells. (B) Cells were cultured in expansion medium (EM, 4 days) and then either stem medium (SM) or differentiation medium (DM) for 4 days. Additional reagents added to the differentiation media 5 mM butyrate (DM-B) or 10 μM DAPT (DM-D). Colors as follows: EdU, green; ALP, red; Muc2, yellow; DNA, blue. (C) The percentage of the monolayer surface area displaying fluorescence from the EdU, Muc2, and ALP stains under the various culture conditions. (D) Monolayer side view of Oflm4 (SM), integrin-β4 and actin (DM), ezrin (DM), Muc2 (DM), β-catenin and actin (DM), and top view of ZO-1 (DM). DNA (nuclei) was stained with Hoechst 33342.

Fig. 3

Micromolding a collagen scaffold on a porous membrane. (A) Schematic of the process used to generate the PDMS stamp to micromold the collagen. (B) The dimensions of microstructures in stamp #1 which replicate the features formed in the collagen. The units for the numbers are in μm. (C) A top view of stamp #1 is shown using brightfield microscopy. (D) A side view of stamp #1 is shown using electron microscopy. (E) Schematic of micromolded collagen in the modified insert. (F) Top views of the shaped collagen array acquired by brightfield microscopy. (G) A close-up view.

Fig. 4

Culture of human small intestinal cells on a collagen scaffold recreates in vitro crypt-villus architecture. (A) Schematic of cell propagation on the microstructures. (B) Cells cultured on a scaffold were tracked by brightfield microscopy over time as the cells spread across the scaffold surface. (C) A brightfield image with the focal plane at the level of the crypt showing that cells migrated into the microwell. (D) An image with the focal plane at the top of the villus showing cells migrated to the top of the micropillar. (E) Fluorescence side view image of the crypt-villus architecture displaying one villus and two crypt structures from an EdU-pulsed array.

Fig. 5

Creating a polarized tissue by applying gradients to the in vitro crypt-villus arrays. (A) Schematic of the gradient of growth factors (W: Wnt-3A; R: R-spondin 3; N: noggin). Brightfield (B) and fluorescence (C) images of a polarized crypt-villus unit under the 3-growth factor gradient. EdU⁺ cells (green) are confined to the crypt and adjacent regions, while no ALP⁺ cells (red) are observed. DNA or nuclei are shown in blue. (D) A wide field view of the polarized crypt-villus epithelium array (3 mm in diameter) under the 3-growth factor gradient. (E) Schematic showing the application of an additional gradient comprised of DAPT (gamma secretase inhibitor). Brightfield (F) and fluorescence (G) images of a polarized crypt-villus unit under the 3-growth factor gradient and opposing DAPT gradient. Mature enterocytes (red, ALP⁺) and proliferative cells (green, EdU⁺) are also marked. (H) Immunofluorescence staining (Olfm4/KRT20) of in vitro human small intestinal tissues showing tissue polarity under the combine growth factor and DAPT gradient. Scale bar = 100 μm for B,C,F,G,H.

Fig. 6

Cell migration up the crypt-villus axis of the in vitro tissue over time. Tissues were fixed at 0 (A) and 96 (B) h post-EdU pulse (24-h EdU pulse duration), and the incorporated EdU was labeled with an EdU-reactive Alex Fluor dye (green). DNA or nuclei were stained with Hoechst 33342 (blue) (C) Box plots of the relative proliferation length which was defined as the length of EdU⁺ tissue divided by the cypt + villus length (659 μm) at 0 and 96 h post-EdU pulse. 26 (0 h) and 20 (96 h) crypt-villus units were quantified. p<0.0001 based on t-test. Scale bar = 100 μm