Gut bacteria products from tryptophan increase hormone-producing cell development in human colon models, with possible benefits for abnormal GLP-1 release in obesity

Intestinal epithelial enteroendocrine cells (EECs) are specialized transepithelial signal transduction conduits that respond to luminal nutrients by secreting peptide hormones that regulate the digestion, absorption, and metabolism of nutrients [1,2,3,4]. In particular, two incretin hormones, glucagon-like peptide 1 (GLP1) and glucose-dependent insulinotropic peptide (GIP), are secreted by specific EEC subtypes in response to luminal nutrients. GLP1, produced and secreted by L-type EECs predominantly localized in the colon, is transported via portal blood and neurons to various target organs [2,5,6]. For example, GLP1 stimulates insulin secretion by pancreatic β-cells (insulinotropic effect), whereas in the brain, it generates satiety signals, thereby regulating appetite and food intake [4,6,7,8,9]. Thus, altered production and release of GLP1 by L-cells could play a major role in metabolic disorders such as obesity, type-2 diabetes (T2D), and hypertension [10]. Experimental and clinical evidence suggests that impaired postprandial insulin release and dysregulated appetite control in obesity could partly be due to defective GLP1 secretion and action [11,12,13]. Indeed, the metabolic benefits of gastric bypass have been partly attributed to improved GLP1 production by EECs [14,15]. Further, GLP1 receptor agonists have recently been approved as effective obesity therapeutics for weight loss [16,17], indicating that defective EEC function could be a key factor in obesity.

Altered GLP1 in obesity could partly occur due to the impaired differentiation of EECs from intestinal stem cells (ISCs), causing reduced EEC numbers. Specific transcription factors and signaling pathways direct ISC differentiation towards the EEC lineage [18,19]. Differentiation of crypt-base-located ISCs towards the secretory lineage (enteroendocrine, goblet, and Paneth cells) is governed by the transcription factor ATOH1 (MATH1 in rodents). Next, the sequential expression of two other basic helix–loop–helix transcription factors, Neurogenin 3 (Ngn3) and NeuroD1, is required for terminal differentiation to EECs. Recent studies have shown significant downregulation of these transcription factors and a parallel decrease in CHGA-positive EECs in the intestinal mucosa of obese and T2D patients [20,21].

EECs constitute only ~1% of the total intestinal epithelial cells and are dispersed throughout the epithelial layer between enterocytes. The extremely low number of EECs has been a great hurdle in performing mechanistic studies pertaining to EEC regulation. To overcome this, new technologies have emerged over the past decade, enabling an in-depth analysis of EECs [22]. Among the newer advances, intestinal organoids have recently been used to investigate EEC function and represent a promising model for EEC research [23,24]. Organoid cultures have been shown to contain different types of EECs and to retain their regional identity about the profile of gut hormones produced [22,25]. Other studies have shown that EECs generated within intestinal organoids are responsive to a range of physiological stimuli and are secretion-competent [26].

Diet and gut microbial products play a critical role in directing ISC differentiation towards absorptive versus secretory lineage epithelial cells [27]. In this regard, recent studies have shown that microbial metabolites of tryptophan (TRP) maintain ISC homeostasis via activation of aryl hydrocarbon receptor (AhR) [28]. High dietary TRP acts as a stimulator that promotes the growth of TRP-metabolizing bacteria and the generation of AhR ligands, which drive beneficial functions, such as promoting intestinal barrier functions and maintaining ISC homeostasis [29]. A dysregulated AhR pathway has been shown to cause aberrant ISC proliferation by affecting Wnt/β-catenin signaling [30]. Gut microbiota that express tryptophanase, such as Lactobacilli, metabolize TRP into indoles. Indoles and indole derivatives, such as indole-3-aldehyde, indole-3-acetic acid, and indole-3-propionic acid, are important signaling molecules [31] and potent AhR activators that mediate AhR’s effects on ISCs. Whether the microbial TRP metabolites directly affect EEC differentiation from ISC, however, is not known. In the current study, we first determined the effects of obesity on the expression of the transcription factors (MATH1, NGN3, and NeuroD1) that govern EEC differentiation from ISC, which subsequently affect the generation of chromogranin A (CHGA)-positive EECs. We used high-fat-diet (HFD)-induced obese Sprague Dawley rats to study EEC dysregulation in obesity. Next, we sought to determine the effects of gut bacterial metabolites of TRP on EEC differentiation and to delineate the underlying mechanisms utilizing a model of human colonic organoids. Here, we report for the first time a direct stimulation of EEC differentiation in human colonic organoids by indole and the TRP metabolites present in the culture supernatant of Lactobacillus acidophilus (LA) via an AhR-dependent mechanism.

The systemic impact of EEC-secreted GLP1 on food intake, appetite, and glucose homeostasis and its role in the pathogenesis and treatment of metabolic disorders, including obesity, have been the subject of extensive investigation. However, the exact mechanisms of how GLP1 secretion is affected in obesity are not fully understood and appear to be multifactorial. The levels of GLP1 that are secreted exclusively by L-type EECs of the epithelium could be affected when (i) the L-cell number in the epithelium is decreased, (ii) GLP1 synthesis in L-cells is impaired at the transcriptional, post-transcriptional, or posttranslational level, or (iii) mechanisms that orchestrate GLP1 secretion in response to nutrient sensing by L-cells is affected. One or more of these factors could be responsible for the reduced GLP1 secretion in obesity. Until recently, most of the studies about GLP1 dysregulation in obesity, irrespective of the causative factor (physiological, dietary, or microbial), were focused on the mechanisms of altered production and release of GLP1 by mature EECs [11,36,37,38,39,40,41]. The fact that obese conditions, presumably because of altered dietary patterns and microbial dysbiosis, could also adversely impact EEC differentiation from ISCs, causing a substantial reduction of GLP1-secreting L-cells, has only recently been appreciated [21,42,43,44]. In the current study, we sought to investigate whether altered GLP1 in obesity could, at least partly, be caused by impaired EEC differentiation from ISCs, causing a reduced EEC number.

In a diet-induced rat model of obesity (HFD-induced obesity in Sprague Dawley rats), our results showed decreased mRNA and protein levels of the EEC marker CHGA, suggesting decreased EEC number. Decreased EECs in obesity were also directly demonstrated by sorting the CHGA-positive cells in the mucosa by flow cytometry and by immunofluorescence staining of CHGA-positive cells in mucosal sections of CD versus HFD rats. To specifically demonstrate that GLP1-producing L-type EECs decrease in the obese intestine, we used immunofluorescence studies in mucosal sections of CD and HFD rats with the GLP1 antibody and found that GLP1-positive L-cells also decreased in obesity.

The differentiation of ISCs towards absorptive (enterocytes) versus secretory (enteroendocrine, goblet, and Paneth cells) lineage cells is governed by various transcription factors and signaling pathways. Active Notch signaling stimulates the expression of Hairy/enhancer of split 1 (HES1), a potent repressor for the basic helix–loop–helix transcription factors ATOH1 (also known as MATH1) and neurogenin-3 (NGN3). Whereas the former is important for producing all secretory cells, the latter, along with another critical transcription factor NeuroD1, is the key regulator for EEC cell formation [45]. Mice deficient for NGN3 completely lack all EEC subtypes in the small and large intestines. Conversely, transgenic overexpression of NGN3 greatly increases the generation of all EEC lineages [45]. Our results in HFD-induced obese rats showed a significant decrease in the mucosal expression of MATH1, NGN3, and NeuroD1 compared to the respective control animals. Concomitantly, mRNA and protein levels of the EEC marker CHGA were decreased in HFD rats, suggesting an inhibitory effect of an HFD on EEC differentiation. These results support that altered GLP1 secretion and function in obesity could be partly due to decreased EEC differentiation that results in a decreased GLP1-producing L-cell number.

Currently, the most effective long-term treatment of obesity is bariatric surgery [46], which, however, is not economically viable at the population level because of its high cost [47,48]. A key mechanism of surgery-induced weight loss and diabetes remission is the increased delivery of nutrients to L-type EECs, resulting in highly elevated postprandial levels of GLP-1 that lead to improved glycemia and weight loss [14,49]. To combat dysregulated GLP1 secretion in obesity and type-2 diabetes (T2D), GLP1 mimetics have subsequently been developed as GLP1 receptor agonists [16,50]. Despite this, directly targeting L-cells with agents (nutritional and/or gut microbial) to stimulate the endogenous production and secretion of GLP1 is still highly relevant as an alternative or combined treatment of obesity for various reasons [51]. The nutritional or microbial approach has minimal or no potential adverse side effects, including GI discomfort, nausea, vomiting, and diarrhea, sometimes observed when using synthetic GLP1 receptor agonists [52,53]. The vagal/neural dependency of part of the GLP1 action (for example, in appetite regulation) might not be accessible for GLP1R agonists [54,55]. Finally, nutritional/microbial agents might lead to the increased secretion of colocalizing hormones, including PYY, GIP, and oxyntomodulin, providing an additive effect on appetite suppression and blood glucose regulation [56].

Diet, gut microbiota, and microbial metabolites, important factors regulating obesity pathogenesis, have recently been shown to play a critical role in directing ISC differentiation towards absorptive versus secretory lineage epithelial cells [32,57,58], thereby playing key roles in the dynamic balance of intestinal epithelial cell composition. For example, an HFD has been shown to cause a transcriptional shift in cell lineage trajectory from secretory to absorptive [59]. There is a substantial decrease in all secretory lineages and an increase in enterocytes over 7 days of an HFD in the proximal intestine [59]. However, the studies investigating the effects of microbiota and metabolites on ISC differentiation are relatively limited. Short-chain fatty acids, gut microbial metabolites of dietary fiber, have been shown to regulate ISC proliferation and stimulate goblet cell differentiation [27]. A recent study in pre-clinical and clinical settings showed that metabolic syndrome is associated with a reduced abundance of TRP-metabolizing bacteria in the gut microbiota [33]. Therefore, the second part of the current study was aimed at determining the effects of indole, a key TRP metabolite, on EEC differentiation, showing increased EEC differentiation (which logically increases GLP1-producing L-cell number also) in human colonic organoids. To provide the effects of the full profile of TRP metabolites present in the culture supernatant of LA, a predominant TRP-metabolizing probiotic bacteria, we used supernatants obtained from an overnight culture of LA in (1) MRS media alone (designated as LA-CS) and (2) MRS media containing 2 mM L-tryptophan (designated as LA-CS-TRP). Both LA-CS and LA-CS-TRP stimulated EEC differentiation as compared to MRS alone. However, LA-CS-TRP stimulation was significantly higher compared to LA-CS stimulation. LA-CS stimulation was presumably due to other metabolites of LA, such as SCFA, lactate, and maybe even some tryptophan metabolites. When LA grew with 2 mM TRP, the concentration of the TRP metabolites in the CS increased; therefore, the additional increase in EEC differentiation by LA-TRP appears to be due to increased TRP metabolites generated by LA from the added substrate (TRP). We named it broadly as TRP metabolites of LA. Our novel findings showing the stimulation of EEC differentiation by gut microbial TRP metabolites highlight the promising potential of targeting L-cells to develop obesity therapeutics. Intestinal organoids have emerged as a valuable tool to study the differentiation, function, and regulation of EECs. Organoid cultures have been shown to contain different types of EECs and to retain their regional identity regarding the profile of gut hormones produced [22,25]. Other studies have shown that EECs generated within intestinal organoids are responsive to a range of physiological stimuli and are secretion-competent [26].

Mechanistically, the TRP metabolite-mediated stimulation of EEC differentiation in human intestinal organoids observed in our studies showed the role of aryl hydrocarbon receptor (AhR). This ubiquitous receptor, initially identified as a conserved sensor of small xenobiotic molecules [60], was later characterized as a transcription factor activated by several endogenous and exogenous ligands, including indole-related molecules produced by the microbiota. Indeed, AhR is now considered a pivotal mediator of gut microbiota–host epithelium crosstalk [35]. Recently, increasing attention has been focused on AhR due to its role in the regulation of inflammation, immunity, intestinal barrier function, and intestinal microecology [61]. Importantly, a recent study has demonstrated that metabolic syndrome is associated with the reduced capacity of microbiota to produce TRP metabolites known to activate AhR. Indeed, supplementation with indole, an AhR agonist, or Lactobacillus reuteri, a strain with a high AhR ligand-production capacity, led to the improvement of metabolic impairments, such as glucose dysmetabolism, via improving intestinal barrier function and GLP1 production [33]. Since in our studies, indole or LA-derived TRP metabolites did not alter the expression of the EEC differentiation transcription factors (Hes 1, NGN3, and NeuroD1) (results not shown), the AhR activation by these metabolites presumably affected Wnt/Notch signaling to stimulate EEC differentiation. Previous studies have reported the AhR-mediated modulation of ISC differentiation and stem cell fate via its effects on Wnt/Notch signaling [28,62].

A major limitation of our current study is that the effects of TRP metabolites on EEC differentiation were determined in colonic organoids generated from biopsies of normal human subjects. The molecular players and signaling pathways involved in the regulation of EEC differentiation could be different in the normal and obese conditions. Studying the differential mechanisms of EEC regulation and the effects of TRP metabolites utilizing intestinal organoids derived from healthy subjects versus obese patients will be paramount to defining novel pathways to target for the development of effective obesity therapeutics.