Early on within developing embryos, cells become committed into three different distinct lineages called primary germ layers, which go on to generate the different groups of organs and tissue of the foetus. Endoderm is one of these germ layers, the other two being ectoderm and mesoderm. It is sometimes referred to as the definitive endoderm, as there is a separate lineage of endodermal cells which are extra-embryonic, i.e. contributing to the placenta. The definitive endoderm is the set of cells that go on to form the epithelial cells of the internal organs comprising the digestive tract, liver, lungs and associated structures.
The story of endoderm is one of epic migration, undergoing progressive internalisation and specification through temporal and spatial cues. This article gives an overview of the emergence, migration, and specification of the definitive endoderm, including the beginning of the build up to organogenesis. Mostly focussing on development in the mouse, the most well characterised mammalian model, mention will be made of other model organisms and humans, as well as how it all relates to the advancements in stem cells.
As with all great stories, we start with chapter one. For endoderm, as well as all the other embryonic germ lineages, that first chapter is gastrulation. In the mouse, prior to gastrulation, the early embryo (known as the blastocyst) has implanted within the uterine wall, and the compacted cells within have not yet specified to any embryonic lineage. At embryonic day (E) 6 – 6.5, areas of this post-implantation stage embryo now known as the gastrula become organisers, sending signals to the surrounding cells, initiating gastrulation and beginning specification (Figure 1).
The primitive streak is the key organiser for the endoderm, located towards the posterior side. It is a group of cells that express the gene Nodal, a signalling molecule belonging to the Transforming Growth Factor-beta (TGFβ) super-family (a group of related signalling molecules influencing cell function and development). The release of NODAL in this region leads to the emergence of a group of nearby cells from within the epiblast (embryonic cells of the gastrula), which migrate outwards and populate the posterior end of the epiblast. At this point this population of cells is considered a bipotential precursor called mesendoderm, capable of generating definitive endoderm cells and some of the mesoderm.
And now begins a great migration. Between E6.5 and E7.5 cells of the visceral endoderm, extra-embryonic cells surrounding the epiblast, are triggered to migrate towards the anterior by their exposure to NODAL antagonists such as LEFTY, released from a region known as the anterior visceral endoderm (AVE). Concomitantly, NODAL and other signalling molecules such as WNT3 drive the differentiation and expansion of the mesendoderm, pushing it to an endodermal identity and driving its migration around the distal end of the embryo (Figure 1).
The expression of certain genes, often transcription factors or cell surface markers, can be used as markers to identify and characterise endodermal cells and their derivatives. Early markers like Gsc and MixL1 indicate mesendoderm, and up until E7.5, FoxA2 and Sox17 expression and a lack of Sox7 demarcate the endoderm from its neighbouring embryonic tissues. However, at this point, the endoderm undergoes its most significant genetic and positional overhaul since its initial emergence.
By E7.5, the endoderm has emerged, migrated, and is an established identifiable lineage within the embryo (Figure 1). Now, complex genetic programmes and morphogenesis begin to specify the sheet of endodermal cells as the internal tissues of the digestive tract and associated organs.
Regions become more patterned in their expression of markers from E7.5 onwards. Anterior regions can start to be discriminated by markers like Hhex and more discreet FoxA2 and Sox17 expression, whereas more posterior regions begin expressing Cdx1, 2 and Hnf1a. This patterning establishes the regions of the foregut, midgut, and hindgut, relative to their anterior-posterior position. At E8 foregut and hindgut pockets form, known as invaginations, which initiate the movement of endodermal cells into the primitive gut tube. This is the structure from which most endodermal organs bud off from.
Key signalling molecules are present as gradients during the process of patterning of the gut tube. Fibroblast growth factor (FGF)4 levels increase towards the hindgut. In the foregut region from E8.5 onwards, combined and varying levels of other molecules such as Bone Morphogenetic Protein (BMP), FGF2, FGF10 and Retinoic acid (RA), induce the formation of the cellular regions that form into different buds, corresponding to the lungs, liver and pancreas. RA and BMPs may also have a role in patterning the hindgut. At this point in time however the differentiation and patterning of endodermal cells is very dynamic, with signalling molecules having changing or different effects in various regions over short time frames.
Much of the signalling molecules which drive specification come from the surrounding tissue, often the mesenchyme (mesoderm derived tissue) such as the notochord or septum transversum. The interactions are reciprocal however, as failure in gut tube specification leads to poor specification of mesodermal tissue which surrounds the gut tube. For example, lateral plate mesoderm, which goes on to form the heart and circulatory system, requires various asymmetric signals from NODAL and its antagonists to specify properly. The complex formation of the head structures and central nervous system also relies on proper specification of the foregut and surrounding mesenchyme.
Complex and individual signalling, genetic programmes and positional rearrangements drive the organogenesis of precursor cells along the different regions of the gut tube into their final structures (see table, Figure 1). The story of endoderm development is brought to an end with the birth of viable offspring, complete with functioning digestive tract, internal organs, and metabolic regulatory systems.
Along with the mouse, understanding of vertebrate development also comes from a few other key model organisms, particularly zebrafish, Xenopus (clawed frog), and the chicken. Despite slightly different architecture, cell movement and nomenclature, the overall progression of endoderm development is fairly homologous and conserved across vertebrate species.
The exact shape and organisation of the early blastocysts is slightly different across the species, however they undergo gastrulation by similar means. In Xenopus, the Spemann organiser fulfils a similar function to the primitive streak in mice, and in chicks that role is similarly taken up by Hensen’s node. The endoderm’s association with the mesoderm via a common point of origin and common precursor cells is very similar across the species, as is the role of the homologues of Nodal, Wnt and their associated signalling molecules.
When it comes to patterning of the gut tube, Xenopus requires Wnt expression in the hindgut and a repression of Wnt via antagonists such as Dkk for foregut specification. BMP and FGF signalling similarly regulate liver and pancreas induction in zebrafish, chicken, and Xenopus. With the onset of organogenesis, there is further variation as the structure and function of certain organs diverges, as well as some homology in specific programmes in analogous structures.
Humans and mice have a relatively greater amount of conservation in their development, given their closeness phylogenetically (both being mammals).The structure of the developing embryos is different however, the human embryo being a more flattened disc structure. Hensen’s node is the organiser equivalent to the region of the primitive streak, and the endoderm begins its emergence, migration and displacement of the hypoblast (extra-embryonic tissue) around day 14-15 of gestation. The patterning and organogenesis of the endoderm is analogous in terms of which antero-posterior and dorso-ventral regions of the primitive gut tube become which structures. Given the obvious ethical difficulties of studying human embryos, not as much is known at the genetic and molecular levels. However new approaches, including the use of computer-based gene comparisons or stem cell differentiation (see below), are offering new insight into our own endodermal development.
Stem cell research has begun to efficiently recapitulate the differentiation of human embryonic stem cells (hESCs) and induced pluripotent stem cells (iPSCs) to endoderm and several of its derivatives. There has been particular focus on several organ targets.
Differentiation of stem cells towards the pancreatic beta-cells has been investigated extensively by many groups worldwide. They are the insulin-producing cells that regulate blood glucose which fail to function normally in patients with diabetes. Implantation of replacement beta-cells derived from differentiated stem cells could lead to a cure. They could provide normal internal detection and regulation of blood glucose within patients, independent of dietary management and insulin injection. Importantly, they offer a potentially unlimited and ready source of these cells, compared with organ donation.
Targeting differentiation of stem cells along the hepatic (liver) lineage is another target of great interest to the bio-pharmaceutical industry, as the liver is normally the site of complex metabolism and the breakdown of drugs and toxins. Having a ready supply of these cells would enhance drug discovery, toxicology and efficacy testing, by giving a much more accurate insight of drugs’ effects in human cells than animal models do.
To mimic what goes in vivo during development, stem cells grown in vitro are subjected to a time-course of growth factors. Cells are initiated to differentiate towards endoderm with high doses of the NODAL-related molecule Activin A, supplemented with various things, e.g. WNT. They are then exposed to sequential growth mediums containing mixtures of FGFs, Retinoic acid, and other inhibitors or factors. Stages of differentiation are characterised using the in vivo markers, and cells are generally driven to the point of pancreatic or liver precursor.
Understanding the story of endoderm development has led to profound advancements in stem cell research. There is still work to be done into characterisation, improvement in efficiency and safety with these protocols. However, targeted differentiation of stem cells is beginning to show its great potential for cell replacement therapy, toxicology, and further insight into human development.
Norris DP, Brennan J, Bikoff EK, Robertson EJ (2002). The FoxH1-dependent autoregulatory enhancer controls the level of Nodal signals in the mouse embyo. Development, 129, 3455-3468
Perea-Gomez A, Lawson KA, Rhinn M, Zakin L, Brûlet P, Mazan S, Ang S (2001). Otx2 is required for visceral endoderm movement and for the restriction of posterior signals in the epiblast of the mouse embryo. Development, 128, 753-765
Tam PPL, Khoo P, Wong N, Tsang TE, Behringer RR (2004). Regionalization of cell fates and cell movement in the endoderm of the mouse gastrula and the impact of loss of Lhx1(Lim1) function. Developmental Biology, 274, 1, 171-187
Zorn AM, Wells JM (2009). Vertebrate endoderm development and organ formation. Ann Rev Cell Dev Bio, 25, 221-251
Excellent and thorough article covering most current research and understanding of endoderm
Brief overview of midgut organogenesis, with links to other useful videos
Fastbleep © 2019.