Generation of biologically responsive colon-like intestinal tissue patches from human induced pluripotent stem cells using a rapid co-differentiation platform

Abstract

The intestinal mucosa is a complex functional layer which is formed from a diverse range of cell types that include epithelial cells (within crypts and villi) and an array of mesenchymal cells. Many intestinal diseases involve loss of the surface mucosa which can be difficult to restore, and which delays healing and return to normal function. We reason that development of a transplantable intestinal mucosal tissue graft may be a potential therapeutic strategy to aid healing. To be clinically useful, such a tissue graft would need to be capable of rapid production, avoid the risk of host rejection and be demonstrably safe. To create a potential intestinal graft, we developed a novel early-stage human induced pluripotent stem cell (hiPSC) co-differentiation platform capable of generating multiple intestinal cell lineages (epithelial, mesenchymal and endothelial) in 8 days. This protocol is simple to implement, serum-free and greatly reduces the use of animal products. We confirmed the identity of cells by demonstrating that these cells had RNA and protein expression profiles typical of intestinal cell lineages. In particular, we used bulk and single-cell RNA sequencing to characterise global cellular transcriptional profiles robustly and showed that the cells have intestinal identity with early polarisation towards colonic differentiation. The results were replicated across multiple hiPSC lines and in an independent centre. We further cultured the derived cells on collagen hydrogels to form colon-like intestinal patches (CL-IPs). When transplanted into mouse subcutis, CL-IPs formed into colon-like tissue structures, including crypts, stromal and muscle layers. They also developed human-origin vasculature which underwent anastomosis with the murine vasculature to transport murine blood into the graft. Teratoma assays and molecular analyses showed no evidence of residual pluripotency. While at an early stage, this platform shows great potential for further development as a potential source for novel intestinal mucosal regeneration therapy. In addition, the platform is physiologically relevant and thus shows promise as the basis for a new generation of in vitro models of intestinal pathobiology.

1. Introduction

Intestinal diseases are common and constitute a major health burden [1]. Many such diseases involve loss of the surface mucosa. As the mucosa is the main functional layer of the intestine – involved in absorption of nutrients and water – dysfunction results in many of the symptoms of bowel disease, such as diarrhoea, dehydration, nutrient deficiencies and haemorrhage. Loss of mucosal integrity is a prominent feature in inflammatory bowel disease (IBD) and thus restoring healthy mucosa may aid in breaking the cycle of chronic inflammation and microbial colonisation, to aid healing [2–4]. The mucosa is thus a focus for regeneration therapies [3, 5].

The intestinal mucosa is characterised by epithelial crypts supported by a range of mesenchymal cells including fibroblasts and endothelial cells and a deep band of smooth muscle, called muscularis mucosae [6–8]. These cell populations derive from endoderm and splanchnic mesoderm during embryogenesis and arise in tandem from the developing gut tube. Other cell populations, such as haemopoietic and neural cells migrate into the intestinal mucosa during embryonic life from other germ layers [8]. Interactions between epithelial and neighbouring stromal cells are particularly important in maintaining homeostasis and in disease pathogenesis.