Society: AGA

The wiring of the enteric nervous system (ENS) is a dynamic process throughout development and adulthood. In the embryo, enteric neural crest derived cells expand and differentiate into the glia and neurons that form the ENS. In mice, these early neurons are electrically active after 2 days and regulate smooth muscle contraction within 9 days, indicating rapid differentiation into neurons with functional properties. Notably, the birth of new enteric neurons occurs into adulthood and in response to injury or inflammation, supporting the potential of leveraging postnatal neurogenesis for the treatment of enteric neuropathies. Although the ENS is formed rapidly in the embryo, the environmental conditions present in the aged intestine are different in size, structure, cell composition, signalling milieu and extracellular matrix components. Despite advancing knowledge in the mechanisms of endogenous postnatal neurogenesis and the utilization of cultured enteric neural stem cells for regenerative medicine applications, the capabilities of neurons in the postnatal environment to form neuronal circuits and reinnervate the gut have not been resolved. Here, we employ the first pan neuronal-specific Cre driver mouse model (BAF53b-Cre) to understand the potential of post-mitotic enteric neurons to reinnervate the postnatal intestine and elicit functional motor activity. BAF53b expression specifically colocalized with neuronal cell bodies (HuC/D) and fibres (TUBB3), but not enteric glia (Plp1). Cell cultures derived from single cell suspensions of the intestinal muscularis indicated that BAF53b-tdT neurons reform ganglia and interconnected networks within one-week in vitro. Support from mesenchymal cells or glia was required for enteric neurons to project nerve fibres in vitro. BAF53b-Diphtheria Toxin (DT) Receptor mice were generated to perform targeted ablation of colorectal enteric neurons by local delivery of DT. RNA-Seq analysis and wholemount imaging indicated successful and specific neuronal ablation and mice exhibited delayed fecal pellet production. Transplantation of myenteric neurons to the aganglionic environment reinnervated the colon within two weeks, restored contractile activity, and reversed delayed pellet expulsion. This was not dependent on the intestinal region of neuron isolation (colon v. small bowel). Optogenetic experiments using enteric neurons generated from BAF53b- channelrhodopsin mice transplanted to the muscularis indicate that enteric neurons can innervate the muscle with functional activity within 7 days in the postnatal environment. Together, these results demonstrate that enteric neurons are proficient in forming new functional connections in the postnatal environment in a timeframe similar to that of the embryo which highlights the remarkable capabilities of neural network formation and reinnervation in the adult intestine.

Background: Current studies provide compelling evidence that the enteric nervous system (ENS) modulates neuroimmunological processes in the small intestine. Previous work revealed the essential role of enteric glial cells therein. However, the closely associated enteric neurons, which have an essential role in gut homeostasis, including regulation of barrier function and gastrointestinal motility, have not yet been studied in this context. In addition, interleukin-6 (IL-6), a crucial cytokine in the inflamed gut known for its ambivalent effects, was no focus in the context of enteric neuronal function in disease. Therefore, we investigated the role of IL-6 signaling in enteric neurons and its impact on postoperative neuroinflammation and subsequent motility disturbances, e.g. postoperative ileus (POI).
Methods: Neuroinflammation and POI were induced by surgical intestinal manipulation of different transgenic mice. To investigate the transcriptome of enteric neurons during gut inflammation, we generated a neuron-specific RiboTag mouse and performed RNA-Seq analysis. For a closer look at neuronal IL-6 signaling, we used conditional, neuronal IL-6Ra-KO mice (IL-6Ra^neuroKO). To assess differences between the genotypes, we examined gastrointestinal (GI)-motility, immune cell infiltration, and bulk RNA-Seq data in the POI animal model. Moreover, we established primary enteric neuron cultures to evaluate the molecular mechanism of IL-6 signaling in vitro and define the neuronal reaction after IL-6 treatment by qPCR, WB and immunofluorescence microscopy.
Results: IL-6 and its targets are strongly induced during POI on both RNA and protein levels. Enteric neurons show nuclear p-Stat3 activation, and RNA-Seq analysis of neuronal mRNA showed a robust, early induction of IL-6 pathway genes. IL-6Ra^neuroKO mice lacking neuronal IL-6 signaling showed a substantial delay in the recovery of GI motility, elevated neuroinflammation, and increased immune cell infiltration. In addition, the bulk-RNA-Seq analysis revealed significant changes in gene clusters associated with migration, inflammation, and metabolic processes. IL-6 treatment activates Stat3 in primary enteric neurons and induces the expression of IL-6 target genes, inflammatory mediators, and metabolic factors, thereby connecting neuroinflammation with metabolic changes.
Conclusion: IL-6Ra^neuroKO mice showed substantial immune-mediated intestinal damage and delayed recovery in the POI animal model. Both neuroinflammatory induction of p-Stat3 and enrichment of neuronal IL-6 target genes implicate the IL-6 pathway as a central regulator in restoring homeostatic enteric neuronal function in the gut.

Background: Iron plays a vital role in neurogenesis. However, disturbed iron homeostasis can lead to ferroptotic neurodegeneration. We have previously demonstrated that a high-fat diet (HFD) can lead to enteric neuronal degeneration and subsequent delayed gastrointestinal motility. In this study, we investigated the role of iron and ferroptosis in HFD and saturated fatty acid-induced enteric neurodegeneration. We further assessed the clinical relevance of iron overload in isolated networks of human myenteric ganglia and determined the presence of iron overload in myenteric ganglia in obesity.

Methods: Immortal mouse fetal enteric neuronal (IM-FEN) cell line and mouse primary enteric neurons were treated with vehicle, palmitic acid (PA, 0.5 mM), Ferrostatin-1 (Fer-1, 10 μM), or PA together with either Fer-1 or the deferoxamine (DFO). After 24 h, the cells were analyzed by Western blotting and RT-qPCR for ferroptosis markers including ferritin heavy chain 1 (FTH-1) and transferrin receptor 1 (TfR1), and 4-hydroxynonenal (4-HNE). Mitochondrial function was measured using the Seahorse XF Cell Mito Stress Test Kit. In vivo, CF-1 mice were fed a regular diet (RD) or HFD for 4 and 12 weeks. Intestinal motility was measured by dye-transit test. Mice intestines were collected for immunostaining and Western blot analysis. Using freshly isolated networks of human myenteric ganglion cells in culture we assessed changes in iron transporter expression and ferroptosis after exposure to vehicle, PA, and IL1β.

Results: Enteric neurons treated with PA had increased levels of labile iron and FTH-1 and showed evidence of ferroptosis including increased TfR1 expression and 4-HNE levels (P<0.001). PA-induced ferroptosis was prevented by the iron chelator DFO as evidenced by the reduced 4-HNE and labile iron levels. PA-induced loss of enteric neurons was also prevented by the ferroptosis inhibitor Fer-1, (P<0.05). Erastin, an inducer of ferroptosis, led to reduced mitochondrial function (P<0.01). Male and female mice fed HFD for 12 weeks gained more weight and had slower intestinal motility (P<0.001) and reduced number of nitrergic neurons than RD-fed mice. Duodenal myenteric neurons from HFD-fed male and female mice also had higher (P<0.001) FTH-1 levels relative to RD-fed mice indicative of iron overload. Isolated human ganglion networks demonstrated evidence of increased iron transporters, DMT-1, ZIP-14, and TfR1 mRNA (P<0.05) in the presence of PA (or IL1β).

Conclusions: Our findings demonstrate HFD/PA exposure leads to mitochondrial dysfunction and enteric neuronal loss. In vitro PA treatment caused systemic iron excess, lipid peroxidation, and ferroptotic cell death in murine and human enteric neurons. Chronic HFD exposure in vivo leads to iron overload and ferroptosis and can lead to the development of enteric neuropathy and gastrointestinal dysfunction.

Purpose: Intestinal inflammation damages the enteric nervous system (ENS) and a reduction of cholinergic enteric neurons is seen in patients with inflammatory bowel disease (IBD). We hypothesize that cholinergic enteric neurons possess anti-inflammatory properties and that they therefore represent a novel therapeutic target for the treatment of intestinal inflammation.

Methods: ChAT::Cre mice were crossed with R26-iDTR mice to obtain ChAT::Cre;R26-iDTR (ChAT-iDTR) mice in which all cells expressing choline acetyltransferase (ChAT) express human diphtheria toxin (DT) receptor. We also generated ChAT::Cre;R26-ChR2 (ChAT-ChR2) mice in which ChAT+ cells express light-sensitive ion channel, channelrhodopsin-2 (ChR2). Ablation of ChAT+ enteric neurons was achieved by focal injection of DT in the wall of the mid-colon in adult ChAT-iDTR mice. For optogenetics, ChAT+ neurons were activated by blue light stimulation (BLS) over 7 days using an intra-rectal probe for 30 min daily (60 cycles of 10 s of 1 ms pulses at 10 Hz, followed by a 20 s break). Controls included light-treated mice not expressing ChR2 and R26-iDTR mice receiving DT injection. All groups (n=5-7 per group) received 2.5% DSS in their drinking water for 7 days. Histology, immunohistochemistry, qPCR, organ bath studies, and electromyography (EMG) were used to determine severity of colitis, morphological changes in the ENS, and smooth muscle activity. Results of both ablated and treatment groups were compared to controls with multiple t-test analysis for significance.

Results: Immunohistochemical examination confirmed successful ablation of ChAT+ enteric neurons, which was associated with increased severity of DSS-induced colitis as indicated by significant reductions in body weight and colon length, increases in histological inflammation score, and CD45⁺ infiltration. Contractile activity of the colonic smooth muscle was markedly reduced in ablated mice as compared to control. In contrast, optogenetic activation of the cholinergic system inhibited shortening of colonic length, reduction of body weight, and splenomegaly. It also prevented morphological damage to the colon and decreased the histological score and extent of CD45⁺ infiltration. Light treatment also restored smooth muscle contractility, as indicated by an increase in EFS-induced rebound contraction compared to the control group. In vivo recordings demonstrated that BLS increased EMG amplitude and luminal pressure in the treatment group as compared to control. BLS also prevented cholinergic neuronal loss in the myenteric plexus and reduced levels of proinflammatory cytokines (IL1b, IL17a, IL6) by qPCR.

Conclusion: This study suggests that local cholinergic signaling plays an important role in attenuating colitis and improving colonic motility and should be explored further as a potential treatment for IBD.

Background: Colon motility results from activity in the enteric nervous system (ENS), interstitial cells of Cajal (ICC), and smooth muscle. While these cellular components and the motor patterns they produce have been individually studied in detail in distal colon, how they interact to produce rhythmic motility initiated in proximal regions is not understood.
Aim: Develop a calcium imaging approach to determine the relationships among activity in the ENS and ICC, ripple contractions, and colon motor complexes (CMCs) in proximal colon.
Methods: We used E2a-Cre;GCaMP6s mice with GCaMP in the ENS and ICC and Calca-Cre;GCaMPs mice with GCaMP in CGRPa neurons that represent a specialized subset of intrinsic primary afferent neurons (CGRPa-IPANs) found only in proximal colon. Calcium signals were recorded from myenteric neurons and ICC in full-thickness, full-length colons that were cut open and pinned flat. Motility was monitored by tracking displacement of the imaging field.
Results: Spontaneous CMCs were generated in unstretched colons without external sensory input in proximal but not distal regions, whereas stretch increased CMC frequency and propagation to distal regions (n=3, P<0.05), indicating intrinsic pacemaker capabilities in proximal colon. ICC slow waves and ICC-generated ripples exhibited cyclical patterns of organization and disorganization that correlated with CMC timing; higher coupling of ICC (i.e., organized waves) and larger ripples were observed before CMCs compared to after (n=4, P<0.05). Myenteric neurons exhibited 2 types of activity patterns that repeated with each CMC cycle. Type 1 neurons were active between CMCs when ICC were organized and ripples became large, suggesting mechanical sensitivity to ICC-generated ripple contractions. Type 2 neurons were active just before CMCs, and likely motor neurons. Experiments in Calca-GCaMP mice revealed that CGRPa-IPANs displayed Type 1 activity, which was suppressed by nifedipine (n=3, P<0.05), confirming that their activity was dependent on contractions and/or ICC. Finally, manipulating myenteric neuron activity with TTX or electrical stimulation altered ICC frequency and slow wave organization.
Conclusions: Based on our observations (summarized in Fig 1), we propose a novel hypothesis (Fig 2) that cyclical ICC-IPAN-ENS interactions in proximal colon act as the pacemaker to set the rhythm of spontaneous CMCs. 1) As ICC slow waves organize, ripples get larger and activate CGRPa-IPANs mechanically-sensitive to ICC-generated contractions; 2) CGRPa-IPANs activate ENS motor neurons that produce a CMC; 3) Robust ENS activity during the CMC transiently disorganizes ICC slow waves and resets the cycle, where the next CMC depends on reorganization of ICC. We have modeled this computationally and will next test our hypothesis using combined optogenetic and calcium imaging approaches.

Brian Edwards

Appearances