Bugs, Guts, and Glia: How Microbiota Influence Enteric Gliogenesis and Migration

Abstract

Enteric glia are neural crest derivatives. Kabouridis et al. (2015) now show in adult animals that new glia arise within the ganglia of enteric plexuses and then migrate centripetally to colonize the mucosa. Remarkably, enteric microbiota regulate this critical migration.

The enteric nervous system (ENS) is a large independent division of the peripheral nervous system that regulates, in addition to gastrointestinal motility, secretion, immunity, and maintenance of the intestinal epithelium. The ENS is also an important node in what is now realized to be a bidirectional “gut-brain” signaling axis. The ENS comprises two ganglionated plexuses, the submucosal and myenteric, which interconnect. Glial cells outnumber neurons in the ENS by several-fold and are found within the enteric plexuses (intraganglionic glia), in association with nerves coursing within the circular muscle (intramuscular glia), and in the lamina propria underlying the epithelium (mucosal enteric glial cells, mEGCs) (Gulbransen and Sharkey, 2012). mEGCs are particularly interesting because their cellular microenvironment is unlike that elsewhere in the nervous system. The lamina propria contains the effector cells of the mucosal immune system, including lymphocytes, dendritic cells, and macrophages, as well as fibroblasts, smooth muscle, blood vessels, and lacteals (lymphatic capillaries); moreover, the lamina propria is only one cell removed from the rich microbiome within the intestinal lumen. The extensively branched processes of mEGCs have been shown to contact immune effector cells, enteroendocrine cells, and blood vessels, meaning that mEGCs are well positioned to integrate intercellular signaling (Bohórquez et al., 2014; Gulbransen and Sharkey, 2012). In the current issue, Kabouridis et al. (2015), report exciting new findings on the development and maintenance of mEGCs, showing that their postnatal development is linked to weaning, that new mEGCs are generated in an “outside-in” manner from enteric ganglia, and that the microbiome is essential for normal mEGC development.

Enteric neurons and glia arise during ontogeny from neural crest-derived precursors that colonize the bowel (Lake and Heuckeroth, 2013). These precursors, which are common to neurons and glia, express the transcription factors Sox10 and Phox2B. Mature enteric glia maintain Sox10 expression and downregulate Phox2B, while enteric neurons downregulate Sox10 and maintain Phox2B (Lake and Heuckeroth, 2013). In mice, many enteric neurons have already been born prior to birth, first in the myenteric plexus and later, following a secondary migration, in the submucosal plexus (Jiang et al., 2003). Although the generation and diversification of enteric neurons has been studied extensively, the dynamics of enteric glial development have not been investigated as thoroughly and are less well understood. To assess mEGC development, Kabouridis and colleagues quantified ileal mEGCs during postnatal development, using mice in which cells derived from Sox10-expressing precursors were lineage-traced with mosaic analysis of double markers (MADM). Kabouridis et al. (2015) found that mEGCs were largely absent from the ileum at birth, began to appear during the early postnatal period, but did not reach a full adult complement until postnatal day 27. Suspecting a link to the dietary change (weaning) that coincides with glial development, they compared numbers of ileal mEGCs before and after weaning and found that a failure to wean delayed entry of mEGCs to the lamina propria. Weaning alters the composition of macronutrients in the lumen, triggers changes in the microbiome, and is associated with major changes in the intestinal epithelium (Henning, 1981), any or all of which might play a role in mEGC development. Another intriguing possibility is that maternal separation, which accompanies weaning in mice, could also be a factor. In an analogous situation, social experience in the juvenile period is known to affect myelination in the CNS (Makinodan et al., 2012). It is therefore conceivable that social experience could similarly affect enteric glial development. This possibility will have to be explored in the future.

The intestine is a mechanically active organ with an epithelium that turns over every few days (Creamer et al., 1961). Consistent with this dynamic microenvironment, potential gliogenesis in the adult ENS is robust and upregulated by injury or inflammatory conditions (Joseph et al., 2011; Laranjeira et al., 2011). To measure turnover of mEGCs and identify the source of new mEGCs in adult gut, Kabouridis et al. (2015) employed an inducible Cre recombinase driven by the Sox10 promoter to trigger expression of the multicolor Confetti reporter in vivo (Sox10::CreER^T2; R26R-Confetti). In R26R-Confetti mice, individual cell lineages can be marked by the stochastic combinatorial expression of one of four fluorescent proteins upon Cre-mediated recombination. Four days after inducing expression of the Confetti reporter in Sox10-expressing cells, Kabouridis et al. (2015) find cells expressing each of the four colors within the enteric plexuses. In contrast, mEGCs within the lamina propria were largely unlabeled, and, where labeled mECCs appeared in the lamina propria, they were monochromatic, single, and isolated. In contrast, by 15 days post-induction, the lamina propria within 72% of villi contained labeled mEGCs, and the majority of villus cores were polychromatic. These observations suggest that Sox10-expressing cells give rise to new mEGCs within the adult myenteric plexus; mEGCs migrate from there to the mucosa. Consistent with these suggestions, lineage tracing of cells that express a second glial marker, glial fibrillary acidic protein (GFAP), also demonstrated that new mEGCs appeared in the lamina propria of the adult intestine 8–12 weeks after induction of a fluorescent reporter. In total, these data imply that new mEGCs continuously arise from Sox10- and GFAP-expressing cells of the enteric ganglia and migrate from there to repopulate the intestinal lamina propria in adult mice. Gliogenesis in the adult nervous system is not unique to the ENS; for example, recent work has shown that new oligodendrocyte generation is important for learning motor tasks (McKenzie et al., 2014). It will be exciting to explore the role of the enteric gliogenesis in gastrointestinal physiology and to understand how new mEGCs are functionally integrated into the complex and dynamic microenvironment of the bowel.

Enteric glia are unique in the nervous system for their close proximity to the microbial world. Because mEGC development is coincident with the colonization and maturation of the intestinal microbiome, Kabouridis et al. (2015) wondered whether it could be affected by the microbiome. To answer this question, they compared adult mEGCs from animals raised either in conventional or germ-free housing. Remarkably, they found that mEGCs were profoundly reduced in number in the germ-free mice; only a small number were present in the lamina propria at the level of crypts and none were in villus cores. Restoring normal gut microbiota rescued this deficiency. Antibiotic treatment of adult mice led to similar reductions in total mEGCs in the lamina propria of villi and in the number of polychromatic mEGCs in Sox10:: CreER^T2; R26RConfetti mice. These observations suggest that intestinal microbiota are required, not just for normal initial postnatal mEGC development but also for sustaining generation of new mEGCs in adult gut.

Every really good manuscript raises at least as many questions as it answers. Experiments, carried out over 20 years ago, in which explants of gut were back-transplanted into neural crest migration pathways of younger chick embryos, showed that enteric crest-derived cells recovered the ability of their ancestors to migrate and exited the explants to give rise to products of differentiation never formed in the gut, such as Schwann cells in extra-enteric peripheral nerves and catecholaminergic neurons in sympathetic ganglia (Rothman et al., 1990). Enteric crest-derived cells may have completed their migration to the bowel, but they remain plastic and uncommitted. They are not determined to differentiate only as enteric neurons or glia, but, given the opportunity, they can alter their fate in response to the microenvironment within which they find themselves. The microbiome is an important contributor to the enteric microenvironment; moreover, the work of Kabouridis et al. (2015) newly demonstrates how microbiota alter the development and the migration of enteric crest-derived cells, in this case glia. Whether the effect of the microbiome is quantitative, direct, or indirect is still unclear. The colon is far richer in microbiota than other parts of the gut; therefore, regional differences in mEGC behavior might be significant. Germ-free mice also have defects in mucosal immunity (Kabat et al., 2014), an altered intestinal epithelium (Thompson and Trexler, 1971), as well as deficits in enteric neuronal number and activity in the early postnatal period (Collins et al., 2014). Future work will be needed to determine whether microflora directly influence mEGC generation and migration, or instead if these processes are altered by the effects of microbiome on other cellular components of the intestinal environment. It would also be interesting to learn whether it is mature glial cells or retained crest-derived precursors that are the cells that respond to the presence or absence of the microbiome. Sox10, the driver used in the Kabouridis et al. (2015) study, as well as GFAP, are expressed both in uncommitted precursors and in mature glia. It would be fascinating to know whether the cells that give rise to the new mEGCs that migrate from enteric ganglia to the lamina propria express Phox2B. This distinguishing marker is downregulated in glia but expressed in their precursors. A final question is to ascertain what happens to old mEGCs to make way for the new. Given the growing evidence for the role of enteric glia in digestive and neurological disorders, it will be important to explore all of the questions that this remarkable new investigation raises.