Chemosensing in the Colon

28.1

Introduction

For centuries, it has been known that ingestion of foods rich in indigestible carbohydrates, sugars, bitter substances, and fats elicits specific cellular responses, such as releasing hormones that alter gastrointestinal (GI) function. It has been only in the past few years that the mechanisms underlying these host responses to ingested, endogenous, and microbially produced substances have been unraveled. In this chapter, we will describe the known biologic sensors for these substances termed “chemosensors,” providing data regarding their possible biologic functions, in an effort to gain further understanding of how the host responds to these biologic signals.

We have focused on the colon since the colon appears to be the main locus of the gut microbiota and luminal chemosensors. The human colon, comparable with the rodent cecum or ruminant forestomach, is the principal site in which commensal bacteria ferment luminal compounds, absorbing calories, and essential nutrients downstream of their usual small intestinal absorption sites. The colonic epithelium is also densely populated with enteroendocrine cells (EECs) that transduce luminal bioactive compounds into neurohormonal signals. Interestingly, the rectal mucosa has the densest EEC population with a predominance of glucagon-like peptide (GLP)-1, GLP-2, peptide-tyrosine-tyrosine (PYY), and 5-hyroxytryptamine (5-HT) containing cells that contribute to GI physiology and metabolic control through neurohormonal pathways. In this chapter, we will review the significance of colonic region-specific luminal bioactive compounds, their receptors, and how receptor activation affects GI function.

28.2

Gut Microbiota

The mammalian intestine is home to trillions of microbes, including more than 1000 species of bacteria, archaea, protozoa, fungi, and viruses. The gut microflora is an ecosystem that is recognized as alternative organ interacting with the host digestive and immune systems through signaling pathways. From a genetic viewpoint, the intestinal lumen includes microbial genes that are estimated to exceed the number of host genes by 100- to 400-fold, indicating that a large variety of known and unknown metabolites are produced. Microbial composition and diversity is determined by species richness and other measures, with “symbiosis” used to denote the microbial composition associated with health, and “dysbiosis” used when the diversity is altered. Dysbiosis, present in a variety of diseases, can be modified with supplementation of specific beneficial bacteria termed probiotics or by nonabsorbable antibiotics or fecal microbiota transplantation from healthy subjects. Alteration of gut microflora in this fashion significantly improves the course of inflammatory bowel disease (IBD), infectious diarrhea due to Clostridium difficile , altered insulin sensitivity, and possibly irritable bowel syndrome (IBS). In addition to GI diseases, dysbiosis has been recently implicated in the pathogenesis of obesity and central nervous disorders, such as autism and Parkinson's disease.

Human fecal production is ~ 128 g/day, increased by high dietary fiber intake. The chemical composition and pH of the fecal output are influenced by diet, with the major organic component (25%–54% of dry solid) of feces derived from bacterial biomass. Gas chromatography-mass spectrometry analysis detected more than 700 volatile organic compounds (VOCs) from human feces, which are derived from the diet, shed epithelial cells, and microbial metabolism of undigested foodstuffs. The composition of luminal VOCs is altered in diseases such as IBS, IBD, colorectal cancer, and autism, implicating that the pathogenesis of these diseases is associated with dysbiosis. Matsumoto et al. developed a novel technique for quantifying colonic luminal metabolites (metabolome) using capillary electrophoresis with time-of-flight mass spectrometry (CE-TOFMS) to compare germ-free (GF) and conventionalized GF mice, which were intragastrically inoculated with fecal suspension obtained from specific pathogen free (SPF) mice. Their method identified 179 metabolites including 77 metabolites that predominate in conventionalized mice compared with GF mice. The prominent anions are short-chain fatty acids (SCFAs) such as acetic, propionic, lactic, and butyric acids at total ~ 100 mM, and cations are polyamines (PAs) such as putrescine, spermine, and spermidine in sub-mM concentrations. The luminal metabolites include signaling and bioactive molecules such as γ-amino butyric acid (GABA), prostaglandin (PG) E ₂ , β-alanine, cadaverine, putrescine, 5-HT, tryptamine, tyramine, ornithine, and xanthine. These bacterial metabolites could influence host physiology, although some could be released from the colonic mucosa in response to microbial activity. Clarifying colonic chemosensing mechanisms for detecting and responding to luminal bioactive molecules emphasizes the importance of dysbiosis in the pathogenesis of disease.

28.3

Sensor Cells in the Colonic Mucosa

A variety of chemosensors such as G protein-coupled receptors (GPCRs) and transient receptor potential channels (TRPs) are expressed in the colonic epithelial cells, including enteroendocrine, tuft, goblet, and columnar cells ( Fig. 28.1 ). Although GPCRs are primarily cell membrane receptors usually present on the apical membrane, many are detected in the cytosol by immunohistochemistry, either an artifact or due to receptor internalization after stimulation. Overall, most colonic chemosensing GPCRs reported are “promiscuous” sensors, activated by ligands at μM-mM concentrations, differing from hormone receptors tuned at nM concentrations of very selective ligands. The expression level of some chemosensing GPCRs and their downstream transmitters is altered within a few weeks by luminal environmental changes such as altered dietary components and microbial activity. This dynamism suggests a potential value of GPCRs and sensory cells as diagnostic markers and as therapeutic targets. In addition to the epithelial cells, subepithelial tissues that include immune cells and nerve fibers possess the same sensor molecules, contributing to the communication between host and microbiome ( Fig. 28.1 ). Since those cells have no direct contact with lumen, detecting bioactive molecules that are transported across the gut epithelium, the mechanism, by which these molecules are transported across the gut, is of importance.

All types of epithelial cells differentiate from common stem cells, which are located at the bottom of the crypts, and most differentiated cells migrate to the surface epithelium. The turnover time of epithelial cells in the colon differs between cell types. For example, the turnover time of the two major types, columnar and goblet cells, is 4.6 days, while that of tuft cells is 8.2 days, and EECs is 23.3 days in mice. Columnar epithelial cells (enterocytes) and mucus-secreting goblet cells almost fill the colonic epithelium. Colonic crypts consist primarily of goblet cells, whereas the surface epithelium is composed mostly of columnar cells, indicating the possibility that goblet cells transform columnar cells, the latter which are responsible for most of the transcellular transport of ions with accompanying fluid secretion and absorption that occurs in crypts and in the surface epithelium.

Membranous epithelial or microfold (M) cells mostly occurring in solitary and aggregated lymphoid follicles throughout the large intestine are specialized in transporting luminal macromolecules and bacteria. Luminal antigens are taken up by M cells through phagocytosis and are caught by dendritic cells and lymphocytes wandering in subepithelial area; this sampling is an important function of mucosal immune system. M cell morphology and lectin binding property in the cecal patch differ from those in Peyer's patches, indicating that the surface glycoconjugate pattern of M cells differs between the small and large intestine; their target and function might vary according to intestinal regions.

The EECs are “intestinal taste cells,” a type of “paraneuron” proposed by Fujita, due to their morphological and biological features that are similar to neurons including secretory granules and transmitter release in response to distinct stimuli. The density of EECs differs among intestinal regions (1%–5%), the highest concentrations are present in the human duodenum and rectum, suggesting that even in the rectum, the “tail end” of the intestine, chemosensing is important. The morphology and gene expression profile of colonic EEC differ from those in the small intestine: colonic EEC usually possess a long basolateral rootlet, and the transmitters are less varied than small intestinal EEC. One of the major populations of colonic EEC termed L cells due to their large secretory granules, which store the hormones PYY, GLP-1, and GLP-2 ( Fig. 28.2 ). These hormones enhance energy absorption and storage via distinct pathways: PYY slows GI transit, GLP-1 potentiates glucose-induced insulin secretion, and GLP-2 increases mucosal growth. Since total colonic resection reduces postprandial plasma concentrations of GLP-2, colonic EEC contributes to systemic energy metabolism as an endocrine organ. Colonic L cells express a variety of chemosensors, including gustducin, taste receptor subunits T1R2 and T1R3, sodium-dependent glucose transporter (SGLT)-1, bitter receptor T2R38, bile acid receptor GPBA, and free fatty acid (FFA) receptors FFA1, FFA2, FFA3, FFA4, and GPR119. PYY-containing L cells express pre- and postsynaptic markers and directly link with nerve fibers, implicated in the specific efferent and afferent communication between EEC and the enteric nervous system. Another major class of EEC is 5-HT-producing enterochromaffin (EC) cells. Martin et al. demonstrated that EC cells isolated from murine colon express a variety of sensors for sugars (SGLT1, SGLT3, GLUT1, GLUT2, GLUT5, T1R3), amino acids (GPR92/93, also known as LPA5), and fatty acids (FFA1-4, GPR84, GPR119). The antisecretory hormone somatostatin is produced by distinct small EEC population, which also express FFA3. Exogenous factors regulate the individual expression of EEC; for example, EC cell numbers are decreased by the depletion of dietary fiber, whereas they are increased by the luminal microbiome or by dextran sodium sulfate-treated mucosal inflammation. As a specific example, GLP-1-producing L cells are upregulated by the presence of luminal fermentable dietary fibers. These reports suggest that there are distinct regulatory mechanisms in the differentiation of each EEC type.

Tuft cells (also called as brush, caveolated, multivesicular, and fibrillovesicular cell) are rarely present epithelial cells, marked by doublecortin-like protein kinase 1 (DCLK1) and cytokelatin-18. Although the colonic tuft cell is not well characterized, these cell types are candidate intestinal sensory cells based on the co-expression of taste transduction molecules such as gustducin and TRPM5. Immunohistochemical studies demonstrated that gastroduodenal tuft cells express neuronal nitric oxide synthase (NOS), choline acetyltransferase, cyclooxygenase (COX)-1 and -2, opioids, and uroguanylin, but not chromogranin A, suggestive of a unique transmitter releasing mechanism. Nerves that stain for the pan neuronal marker PGP9.5 ⁺ are connect to some tuft cells, suggesting a direct signal transduction pathway from tuft cells to neurons similar to that typical of other EEC. Since the depletion of epithelial DCLK1 decreases epithelial cell proliferation and survival rate after mucosal injury, DCLK1 ⁺ tuft cells may maintain epithelial homeostasis and restitution.

Antibacterial proteins, namely α-defensins, are exclusively produced in Paneth cells in the small intestine, and present in the colonic lumen in mice. Defensin-producing Paneth cells are not present in the normal colon; however, the mucosa of colonic adenomas synthesizes α-defensins, decreasing the population of commensal bacteria in patients. This change of α-defensins production sites to adenomatous epithelium may have impact on the composition of the microbiome.

28.4

Short-Chain Fatty Acids

Although the health benefits of dietary fiber intake and fermented foods have been well recognized for decades, the mechanism by which these food components benefit the host has only recently been elucidated, mostly due to the recent de-orphanization of SCFA receptors. In particular, soluble fiber improves bowel habits and improves glycemic control and energy metabolism through gut hormone release, including GLP-1, GLP-2, and PYY. These effects are not present in GF animals that lack luminal microbiota, indicating that microbial metabolites derived from dietary fiber rather than the fiber itself is beneficial.

SCFAs, such as acetate, propionate, and butyrate are among the most important microbial metabolites that interact with host cells. Up to 100 mM SCFAs are produced in the colonic lumen of non-ruminants by bacterial fermentation of dietary components that resist foregut digestion and absorption. Infant feces already contain significant amounts of acetate and lactate, but rarely propionate and butyrate, produced by bacterial fermentation from oligosaccharides contained in breast milk. Human amniotic fluid contains acetate and lactate at concentration 6.9 and 9.7 mM, respectively. These reports suggest that the human intestine is exposed to high concentrations of SCFAs throughout life and even prenatally, and SCFA sensing may be important for the development of the GI tract. Since luminal SCFAs are absorbed by colonic epithelial cells into the submucosa and the systemic circulation, a variety of SCFA signaling pathways are likely involved in acute and long-term physiological responses to luminal bacterial activity.

Luminal SCFAs acutely increase motility and transmural ion transport in rat colon in vivo and in vitro. From studies of isolated colonic segments, the contractile and secretory effects of SCFAs are primarily mediated by neural and epithelia-derived cholinergic pathways. Although isolated colonic EC cells are not directly activated by SCFAs, SCFAs induce peristalsis is accompanied with 5-HT release into the lumen and abolished by de-afferentation in vivo . SCFAs likely influence many physiological functions via local and neuronal pathways. On the other hand, repeated luminal infusion of SCFAs inhibits colonic motility and fluid secretion through PYY release. Furthermore, SCFAs reduce electrical stimulation-evoked neurally mediated contractions of isolated rat colon, which are independent of the presence of mucosa, indicating that SCFA can inversely regulate enteric neural activity. Thus, luminal SCFA sensing pathway acutely accelerates and secondarily reduces colonic motility and secretion, facilitating fermentation of the luminal content and the absorption of luminal SCFA as an additional energy source.

FFA2 and FFA3 were de-orphanized as SCFA selective receptors. Other FFA family members, FFA1 and FFA4 are long-chain fatty acid receptors not activated by SCFA. SCFAs are transported by the monocarboxylate transporter (MCT) and sodium-dependent MCT (SMCT) families, particularly basolateral MCT1 and apical SMCT1 in the large intestinal mucosa.

FFA2 (also known as GPR43) is abundantly expressed in neutrophils and in adipocytes in the GI tract, and is also present in PYY- and GLP-1-producing colonic L cells in human and rodents. Increased luminal SCFA concentrations induced by dietary fiber supplementation upregulates FFA2 expression and GLP-1-containing L cell density. In vivo and organoid studies using FFA2 deficient mice demonstrated that FFA2 activation is selectively involved in PYY, but not GLP-1 production and release into the circulation. Although colonic L cells express FFA2 and FFA3, both selective agonists have no effect on GLP-1 positive cell density in organoids, suggesting that GLP-1 upregulation requires signals different from these generated by epithelial cells. To date, ion secretion or colonic contraction induced by SCFAs has not been accomplished by the use of FFA2 selective agonists.

Experiments with FFA3- (known as GPR41) deficient and reporter protein-expressing mice indicate that GLP-1 and PYY release is mediated by FFA3 ; consistent with the FFA3 localization to L cells. FFA3 is distributed in enteric neurons, autonomic ganglia, and extrinsic afferent neurons, which contribute to GI physiological functions. The activation of FFA3, expressed on cholinergic nerves located in the mucosal and submucosal plexus of rat proximal colon, suppresses nicotinic acetylcholine receptor (nAChR)-mediated anion secretion, possibly through the inhibition of neural ACh release via the G _i/o pathway. Myenteric neurons abundantly express FFA3, and intestinal transit in FFA3 deficient mice is accelerated likely due to the lack of neuronal and PYY-mediated inhibition, suggesting that FFA3 significantly contributes to colonic motility regulation.

Microbial-derived SCFAs may influence the development of myenteric neurons, which predominantly influence intestinal motility control. A diet enriched in the fermentable carbohydrate resistant starch increases luminal SCFA production and cholinergic, but not nitrergic, neuron populations in rat distal colon. In contrast, GF mice have less nitrergic neurons but similar cholinergic neuron populations compared with conventional controls. IBS patients with motility dysfunction have variable colonic SCFA content compared with healthy controls, independent of the predominant symptoms of constipation or diarrhea. These reports suggest that dietary therapy or pre- and probiotics may improve intestinal motility disorders through alteration of the myenteric neural population via increased luminal SCFA production.

28.5