Mechanisms for Intestinal Regeneration

Abstract

Purpose of review:

The purpose of this review is to briefly summarize the notable structures and pathways in intestinal epithelial growth before presenting the current main areas of active research in intestinal regeneration. As a rapidly advancing field, a number of breakthroughs have recently been made related to the culture of intestinal stem cells and to the engineering of intestinal tissue.

Recent findings:

Intestinal stem cells can be derived from fibroblasts and can be cultured in hydrogels under xenogeneic-free conditions. Intestinal organoids can be cultured with neural crest cells to form small intestinal tissues with neuromuscular networks. Endoluminal devices can be placed inside the native intestine to exert mechanical force to induce novel tissue growth.

Summary:

A number of recent advances in the field of intestinal regeneration are encouraging and suggest that novel therapies for a wide range of intestinal disorders may be developed in the near future. There are still a number of obstacles before such stem cell therapies can be safely used in humans.

Introduction

The intestinal epithelium is normally replaced every 4–5 days and thus has one of the highest turnover rates in the human body, underlining the impressive regenerative ability of this organ [1–4]. A major goal in studying intestinal regeneration is to develop novel therapies for intestinal diseases. The benefits could be seen in a wide range of conditions that affect the epithelium, the nerve cells, or the entire intestine. Patients with disorders such as inflammatory bowel disease, celiac disease, Hirschsprung’s disease, cystic fibrosis, and short bowel syndrome may all benefit [5]. There are a number of different aspects of intestinal regeneration that will be discussed in this review, but they share some overarching goals. First, identify stem cells so that all other differentiated intestinal cells can be generated from them. Second, expand these cells in vitro to genetically edit them as necessary. Finally, integrate the different cellular components together and engraft them back as a functioning unit [5]. With these goals in mind, we will review the current progress made in the field with a focus on recent advancements. We will start by outlining the normal cellular structures and pathways involved in the intestinal epithelial growth.

Text of Review

Cellular structure of the intestinal epithelium

Before discussing how intestinal growth and regeneration can be promoted and maximized, we will first review the native structure of the intestinal epithelium and the pathways involved in its homeostasis. The epithelium of the small intestine is folded into crypts of Lieberkühn and villi (Figure 1). Each villus is a finger-like projection into the intestinal lumen, and multiple crypts surround the bases of these projections. The crypts contain one type of intestinal stem cells (ISCs) called crypt base columnar cells (CBCs), which are distinguished by their expression of the genes for leucine-rich repeat containing G-protein coupled receptor 5 (LGR5) and olfactomedin 4 (OLFM4) among others [1–4]. These stem cells give rise to differentiated daughter cells that migrate out of the crypts to replace the epithelial cells on the villi, which are shed into the lumen via apoptosis as they approach the tip of the villi, as well as Paneth cells that move to the bottom of the crypts.

Figure 1.

Cellular structure of an intestinal crypt and villus.

Diagram of an intestinal villus and crypt. Differentiated cells such as enterocytes, goblet cells, enteroendocrine cells, and tuft cells appear in the villus. The crypt contains intestinal stem cells (ISCs). +4 cells are reserve stem cells that are able to repopulate the crypt base columnar cells (also known as LGR5+ stem cells) following injury. Paneth cells are also found in the crypt and support the growth of the ISCs. As ISCs create new progenitor cells, these cells will continue to differentiate and migrate up to the villus through the transit-amplifying zone. Mesenchymal cells support growth of the crypt base.

Other notable cells of the intestine include a number of differentiated cells and the slow cycling stem cells. The differentiated cells in the villi include absorptive enterocytes and secretory cells such as goblet cells, enteroendocrine cells, and tuft cells [3–5]. The differentiated cells in the crypts include Paneth cells and enteroendocrine cells [3–5]. Slow cycling stem cells are the “+4” cells, so named for their position as the fourth cell from the base of the crypt. These cells are identified by various marker genes Bmi1, HopX, mTert, and Lrig1 [6–9]. They are typically quiescent, but are important in their ability to replace CBCs following injury and thus represent a separate population of stem cells [10, 11]. Recent findings suggest that there is significant plasticity among intestinal epithelial cells so that the interconversion of cell types occurs when needed.

Signaling pathways in intestinal homeostasis and growth

There are several important signaling pathways that have been well studied in the intestine. Understanding these pathways is important to be able to maintain and differentiate intestinal cells in vitro. The Wnt pathway is vital for ISCs to induce crypt and epithelial proliferation and regeneration following injury [12, 13]. The Wnt ligands are expressed by Paneth cells and mesenchymal cells such as fibroblasts, neurons, and endothelial cells [14, 15]. When excessively stimulated, this pathway is also associated with an increased risk for cancer [16, 17]. The bone morphogenetic protein (BMP) pathway is necessary for differentiation of epithelial cells. Inhibitors of this pathway are secreted by mesenchymal cells to limit differentiation of cells in the crypts. In the villi, BMP agonists are expressed and promote cellular differentiation [18, 19]. The Notch pathway is another pathway involved in cellular differentiation. Activators of this pathway are expressed by Paneth cells and maintain ISC pluripotency. When the pathway is inhibited, cells will differentiate into secretory cells such as goblet cells [20, 21]. Ligands for the epidermal growth factor (EGF) pathway are also expressed by Paneth cells, and this pathway leads to crypt cell and ISC proliferation [22]. The pathway is negatively regulated by Lrig1 found on CBCs in order to limit the growth of the stem cells and to maintain homeostasis [23].

Culture and transplantation of intestinal stem cells

The ability to grow and maintain intestinal cells in vitro is a critical first step in being able to control and promote intestinal regeneration. Intestinal epithelial cells were first grown in vitro with the use of human cancer cell lines. These are cultured in two-dimensional (2D) models where the cells form a monolayer on top of a dish, membrane, or scaffold. This method also allows for co-culture of epithelial cells with goblet cells or immune cells normally found in the intestine to more closely mimic and study the natural environment of these cells [24, 25]. As an alternative to the 2D models, a different technique was developed with the use of a three-dimensional (3D) support matrix. Most commonly, this support matrix is made up of Matrigel, which is derived from the sarcoma cells of Engelbreth-Holm-Swarm mice. Matrigel mimics the extracellular matrix and contains laminin, collagen IV, enactin, and growth factors [26]. These 3D models can be created by inserting single LGR5+ ISCs from the small intestine into Matrigel and inducing these cells to proliferate and differentiate into enteroids with the use of growth factors. Enteroids are small cellular clusters with a central lumen and crypt-like budding domains. Enteroids are composed of epithelial cells normally found in the intestine, including Paneth cells, enterocytes, goblet cells, and enteroendocrine cells [27]. Growth factors used in the culture of ISCs include EGF, Noggin (a BMP pathway antagonist), and R-spondin 1 (a Wnt pathway agonist). This method of culture is robust, and enteroids can be passaged for over a year [27]. It also allows for co-culture of epithelial cells with myofibroblasts or lymphocytes [28, 29]. When epithelial cells are combined with mesenchymal cells, such mixed cell populations are called organoids. There is confusion in the current literature with respect to the nomenclature. Organoids are often used to describe populations of intestinal epithelial cells when enteroids would be the more appropriate terminology.

Both the 2D and the 3D models have some drawbacks. The use of cancer cell lines in some of the 2D models means that these models are genetically and phenotypically different than normal intestinal tissue. On the other hand the use of Matrigel is also problematic as its composition is variable, and its origin from a mouse sarcoma means it is unlikely to be safe for clinical use because of immunogenic and pathogenic concerns [26, 30, 31]. Therefore any cell lines grown on Matrigel might not be used therapeutically in patients. Recent advances have been made to overcome these challenges. We have recently shown that human ISCs can proliferate in a 2D model when grown on a thin layer of type I collagen or human laminin [31]. Furthermore, after expanding the ISCs, they can be differentiated into enteroids when resuspended in Matrigel [31].

In order to replace Matrigel, Gjorevski et al were able to develop a 3D matrix consisting of cross-linked polyethylene glycol (PEG) hydrogel enriched with a short peptide (Arg-Gly-Asp), which is the peptide motif in fibronectin recognized by cell adhesion molecules [32, 33]. By changing the composition of the matrix, they were able to first expand mouse ISCs in a stiff gel, and then allow for enteroid formation in a softer gel [32]. This 3D matrix also supported growth of human small intestinal ISCs and human colorectal cancer cells. Another recent advancement was to show that enteroids can be organized into 3D tubes when grown on collagen gels [34]. By seeding enteroids onto floating type I collagen, the epithelial cells were able to self-assemble and fuse into a macroscopic hollow tube-like structure more reminiscent of the native shape of the intestine [34]. To add other components of the intestine, another group was able to produce organoids containing cells from all three germ layers using human induced pluripotent stem cells (hiPSCs) [35]. By growing hiPSCs on a patterning glass substrate coated in PEG, they were able to produce organoids under xenogeneic-free conditions. Some of the resulting organoids were also found to be contractile and thus contained functioning muscle and nerve networks in addition to epithelial cells [35].

Finally, instead of using isolated ISCs, a recent study used mouse and human fibroblasts to create enteroids without going through a pluripotent state first [36]. Through transduction, the four transcription factors Hnf4α, Foxa3, Gata6, and Cdx2 were introduced into the fibroblasts, and the cells were thereby converted into induced fetal intestine-derived progenitor cells (iFIPCs). These cells were then cultured and differentiated in Matrigel to form enteroids [36]. Using human fibroblasts, they were able to perform the same conversion to enteroids [36]. Moreover, these enteroids could be transplanted into immunodeficient mice with a chemical induced colitis. The mice with transplanted enteroids had a greater weight gain than controls that did not receive enteroids following the mucosal injuries. Three months following the epithelial transplant, these cells could still be identified in the recipient colon and appeared structurally and morphologically indistinguishable from the surrounding native cells [36]. Taken together, these rapid advances in the field are encouraging steps towards being able to successfully expand intestinal epithelial cells in vitro, and to eventually use them therapeutically in vivo.

Tissue-engineered intestine

While the intestinal epithelium has been extensively studied as a model for tissue regeneration, the other layers of the intestine, including the lamina propria, the submucosa, the muscularis, and the serosa, have been examined in less detail. Tissue-engineered small intestine (TESI) is a model that has been developed to generate not only the epithelium but also the other components of the intestine. In this model, organoids, which contain both epithelial and mesenchymal cells, are further grown in vivo by seeding them onto omentum or under the kidney capsule with a scaffold [37, 38]. The benefit of this model is that it allows for the growth of epithelial cells together with mesenchymal cells such as fibroblasts and smooth muscle cells. The resulting TESI forms as a cyst-like structure, but this tissue can be incorporated into native intestine with a side-to-side anastomosis [38]. When tissue-engineered colon in rats is anastomosed just proximal to an end-ileostomy, it reduces weight loss, prolongs stool transit time, decreases stool moisture content, and promotes bile acid re-uptake [39]. In addition, histologic analysis shows normal architecture of the different layers [40].

While TESI has the advantage of growing epithelial cells together with mesenchymal cells, until recently this model has not been able to incorporate nerve cells to form an enteric nervous system (ENS). Now two methods have been developed to integrate nerve cells. The first method generates neural crest cells from human pluripotent stem cells (PSCs) [41]. Human intestinal organoids are grown from human embryonic and induced PSCs that are differentiated into organoids and then inserted under a mouse kidney capsule [41]. Finally, the neural crest cells and human intestinal organoids are combined and cultured in vitro in Matrigel for 4 weeks before being placed back in vivo under the mouse kidney capsule for further growth [41]. After the in vitro culture, the neural crest cells are seen to migrate into the mesenchyme of the organoids and differentiate into neurons and glial cells. Measuring calcium efflux shows evidence of rhythmic neuronal activity. Similarly, ENS-dependent rhythmic contractions can be induced with an electrical impulse, suggestive of neuromuscular interaction [41]. However, this method of ENS growth has some limitations, as the ENS here appears to be immature with smaller nerve bundles and a less developed neuro-epithelial network.

A somewhat different method of ENS integration was described by Schlieve et al. Neural crest cells were similarly developed from human PSCs. Human embryonic stem cells were differentiated into human intestinal organoids and grown in culture for 4–5 weeks. The neural crest cells and organoids were then combined on a scaffold and inserted into the omentum of immunodeficient mice [42]. After 3 months of growth, the organoids formed cystic structures and were harvested. Histology revealed submucosal and myenteric ganglia and immunostaining showed the presence of a variety of excitatory, inhibitory, and sensory neurons [42]. ENS-dependent contractions were also seen. However, a more mature ENS with a close association of neuronal cells with enteroendocrine cells was seen here after in vivo culture [42]. This was not seen by Workman et al following the in vitro culture of neural crest cells with organoids [41]. It seems that early in vivo implantation is important for optimal integration of the different cell lines, as well as implanting whole neurospheres in vivo rather than dissociated cells [42]. Together, these studies show exciting advances in the field of intestinal regeneration. As different cytologic elements of the intestine are brought together and assembled, the resulting tissue engineered intestine becomes more sophisticated and more likely to achieve a therapeutic role.

Other modalities of intestinal regeneration and future direction

Besides the various stem cell therapies already discussed that are aimed at growing de novo tissue, there are also aims to use mechanical force to regenerate intestine. Distraction enterogenesis is the use of mechanical force to stretch intestine and thereby induce novel tissue growth and lengthening. A number of devices have been studied over the last decade to achieve intestinal lengthening; these include an extraluminal expanding polymer, an endoluminal telescoping hydraulic device, and an endoluminal helical spring [43–47]. There is some variability in the lengthening achieved in mouse, rat, and pig models, although it generally ranges from a 1.5- to 3-fold increase from the starting length [43–47]. Importantly, this lengthening does not simply represent the thinning and stretching of intestine under tension. Histology of lengthened segments typically shows increased thickness of the muscularis propria and increased crypt depth when compared to controls, signifying that the mechanical forces actually induce a proliferative response at the cellular level [44–47]. Functional studies also show that the lengthened intestine has normal peristalsis as well as normal water absorption and increased glucose absorption [48, 49]. The main limitation of these various devices has been the impracticality of multiple operations to place them in or around isolated segments of intestine and then having to reoperate to replace the lengthened isolated segments. However, we recently showed that intestinal lengthening is also possible in an in-continuity segment of intestine if the endoluminal spring is adequately anchored inside the intestine [50]. Although this model is still being refined, it is an important improvement as it obviates the need for impractical anatomic configurations of the intestine during implantation. It also brings distraction enterogenesis one step closer to becoming a viable therapy for patients with short bowel syndrome.

Conclusion

Overall, while the field of intestinal regeneration has come a long way in recent years, there is still significant progress to be made. The numerous protocols for stem cell isolation, culture, and implantation will continue to be refined along with other modalities promoting intestinal regeneration. While this field has already greatly increased our understanding of the intestine, in the near future it may also provide valuable novel therapies in the form of gene-edited epithelium, transplantable nerve cells, and even intestine grown de novo. Patients suffering from a wide range of intestinal maladies are likely to benefit from the development of these new treatments.

Key Points.

* The intestine is organized into crypts and villi containing stem cells and differentiated cells, respectively, with a number of signaling pathways regulating their homeostasis.

* Intestinal stem cells can be isolated or induced, and can be used in vitro to grow enteroids that consist of all the same intestinal epithelial cells found in vivo.

* Intestinal organoids containing mesenchymal cells can be combined with neural crest cells to create tissue-engineered small intestine and colon with functioning neuromuscular networks.

* A variety of devices can be used to apply a mechanical force on native intestine to induce novel tissue growth and lengthening.