Crohn’s disease and ulcerative colitis are two debilitating inflammatory bowel diseases (IBD) characterized by abdominal pain, weight loss, and frequent bowel movements. The hallmarks of Crohn’s disease are the involvement of the small bowel and the transmural inflammation in contrast to ulcerative colitis, which is limited to the colon and presents as a more superficial intestinal inflammation with bloody stools. IBD often affects patients relatively early in life, commonly with onset between 15 and 30 years of age (Dignass et al., 2012; Gomollón et al., 2017). Many of the therapeutic options involve systemic immunosuppressive drugs with potential long-term adverse effects such as infections and malignancies (Dignass et al., 2012; Torres et al., 2019). A significant fraction of patients do not respond adequately to available treatment and approximately half of the people affected by Crohn’s disease still require abdominal surgery (De Simone et al., 2021). Accordingly, there is an unmet need to improve effective treatment of Crohn’s disease while minimizing risks of serious side effects. This is particularly important in young patients who will require years of immunosuppressive maintenance treatment to keep the disease in remission. Interestingly, there is increasing evidence that neural reflexes regulate gut inflammation in health and disease, suggesting that targeting neural circuits is a potential therapeutic option in IBD (Goverse et al., 2016; Stakenborg and Boeckxstaens, 2021).

The inflammatory reflex is an example of a neural circuit that involves signaling in the vagus nerve that regulates both organ-specific and systemic immune activity (Borovikova et al., 2000; Rosas-Ballina et al., 2008; Eberhardson et al., 2020). In the motor arc of this circuit, the vagus nerve, the splenic nerve, choline acetyltransferase (ChAT⁺)-expressing T cells, and alpha7 nicotinic acetylcholine receptors (α7 nAChR) on innate immune cells are essential for inhibition of systemic release of TNF-α in inflammation (Wang et al., 2003; Rosas-Ballina et al., 2008, 2011; Olofsson et al., 2012; Tarnawski et al., 2018; Caravaca et al., 2019). The discovery and functional mapping of this reflex has supported progress to clinical trials of pharmacological or electrical activation of the inflammatory reflex for treatment of diseases characterized by excessive inflammation (Bonaz et al., 2016; Koopman et al., 2016; Consolim-Colombo et al., 2017; Olofsson and Bouton, 2019; Genovese et al., 2020; Yap et al., 2020).

In parallel to the spleen-dependent neural regulation of inflammation, a direct route to the gut has been proposed, in which efferent vagus nerve fibers functionally connect with the myenteric plexus in the intestinal wall (Berthoud et al., 1990, 1991), referred to as the intestinal cholinergic anti-inflammatory pathway (Goverse et al., 2016). This pathway also inhibits release of pro-inflammatory cytokines from macrophages and other cells by cholinergic signals (Rosas-Ballina et al., 2011; Tarnawski et al., 2018; Stakenborg and Boeckxstaens, 2021).

There are many different animal models of IBD that recapitulate certain aspects of Crohn’s disease and ulcerative colitis, but none fully capture the complexity of either clinical disease (Kiesler et al., 2001; Mizoguchi, 2012). An anti-inflammatory effect of vagus nerve stimulation (VNS) was observed in a number of studies of rodent colitis, such as DSS-, oxazolone-, and TNBS-colitis (Ghia et al., 2006; Bai et al., 2007; Snoek et al., 2010; Meregnani et al., 2011; Meroni et al., 2021) and others previously reviewed (Levine et al., 2018a). However, experimental data on VNS effect on lesions in the small bowel are largely lacking. This is important since nerve-mediated treatment mechanisms and effects may differ between anatomic regions and disease models (e.g., ulcerative colitis, Crohn’s disease, postoperative ileus) (Payne et al., 2019). Whether the reduction of inflammation observed after VNS in intestinal inflammation is mediated by direct innervation of the gut or if the signals are relayed through the spleen is not completely understood (Ji et al., 2014; Matteoli et al., 2014). Furthermore, the duration of VNS effects on reduction of intestinal inflammation is not known. A better understanding of the kinetics of VNS treatment on inflammatory lesions is important to inform both design of nerve stimulators and clinical trials in IBD (D’Haens et al., 2018; Tarnawski et al., 2018). Accordingly, we studied mechanism and kinetics of VNS in indomethacin-induced acute intestinal inflammation as a small bowel disease model.

Here, we found that electrical stimulation of the vagus nerve reduced indomethacin-induced acute small bowel inflammation by a spleen-independent mechanism. Reduction of small bowel lesioning remained for 48 h after a single episode of VNS.

The findings here that electrical VNS reduced small bowel inflammation align with previous studies that show that signals in the vagus nerve regulate levels of pro-inflammatory cytokines and inflammation in a wide range of organs and scenarios (Steinberg et al., 2016). Reduction of the elevated levels of TNF that promote intestinal inflammation in Crohn’s disease has been a therapeutic success (Hanauer et al., 2006). However, after 1 year of anti-TNF treatment, more than half of the patients have stopped the treatment due to inadequate response, loss of response, or intolerance (Roda et al., 2016). Therefore, the additional evidence of vagus nerve regulation of intestinal lesion development provided here may be useful to support further development therapeutic modalities that target the neural regulation of inflammation.

The mechanism for vagus nerve regulation of intestinal inflammation is not fully understood. The vagus nerve directly innervates the gastrointestinal tract and is a key conduit for the bi-directional communication between the enteric nervous system and the central nervous system (Prechtl and Powley, 1990). The main neurotransmitter of the vagus nerve is acetylcholine, known not only to regulate visceral functions such as gastrointestinal motility, but also the release of pro-inflammatory cytokines from innate immune cells (Cailotto et al., 2014). Both afferent and efferent signals in the vagus nerve participate in the regulation of pro-inflammatory cytokines in inflammation (Olofsson et al., 2015). Data from murine models suggest that the reflex control of gut inflammation is mediated by the efferent vagus nerve and sympathetic nerve signals that culminate in regulating the activity of macrophages that reside in the viscera where vagus innervation is abundant (Luckey et al., 2003; Ji et al., 2014; Matteoli et al., 2014; Stakenborg and Boeckxstaens, 2021; Zhang et al., 2021).

Available data indicate that signals in the vagus nerve regulate gut inflammation by different routes, spleen-independent and spleen-dependent. One spleen-independent conduit involves the vagus nerve, efferent sympathetic nerves and myenteric efferent fibers that regulate macrophages in the intestinal wall (Luckey et al., 2003; Matteoli et al., 2014). Evidence of spleen-independent routes for vagus-nerve-mediated regulation of inflammation is also found in other visceral organs such as the pancreas: VNS reduced pancreatitis severity by a spleen-independent mechanism (Zhang et al., 2021).

A spleen-dependent route of vagus nerve regulation of inflammation is the inflammatory reflex, which involves efferent vagus neurons, the celiac ganglion, the splenic nerve, and the spleen (Ji et al., 2014; Kressel et al., 2020). These signals within the inflammatory reflex are propagated to immune cells in the spleen capable of biosynthesizing acetylcholine (Olofsson et al., 2012). Both spleen-dependent and spleen-independent neural regulation of inflammation may reduce macrophage accumulation and release of pro-inflammatory cytokines in the gut through activation of α7 nAChR, α4β2 nAChR or other cholinergic receptors (Wang et al., 2003; de Jonge et al., 2005; Rosas-Ballina et al., 2008, 2011; van der Zanden et al., 2009; Costantini et al., 2012). Of note, ChAT⁺ T cells, capable of acetylcholine biosynthesis, are found in the innervated Peyer’s patches of the small intestine of rodents (Vulchanova et al., 2007; Dhawan et al., 2016; Willemze et al., 2019) and it is possible that these cells partake in local regulation of immune cell activity as has been observed in spleen (Rosas-Ballina et al., 2011). Considering this, the finding here that the spleen is not essential for VNS-mediated reduction of disease intensity in indomethacin-induced small bowel inflammation is important. The data imply that for treatment of inflammatory lesions in IBD with involvement of the small intestine, targeting the cervical or gut-directed subdiaphragmatic vagus may be superior to targeting the splenic nerve or spleen.

The observation here that VNS reduced small intestinal inflammation for up to 48 h aligns with our previous finding that VNS reduced systemic TNF release in endotoxemia for up to 48 h (Huston et al., 2007; Tarnawski et al., 2018) and offers support for refining therapeutic stimulation protocols. The first published clinical trial using VNS for treatment of Crohn’s disease in biologic-naïve subjects was designed using the stimulation “duty cycle” (time on vs. time off) originally designed to treat epilepsy, stimulating up to 262 times per day (Bonaz et al., 2016). The findings here regarding the persistence of protection in the small bowel enabled a once per day VNS treatment in biologic-experienced subjects with Crohn’s disease (D’Haens et al., 2018). The growing mechanistic insights on regional differences in the neural regulation of inflammation between and within organs such as the colon, the small bowel, and other visceral organs, will be valuable for design of future IBD trials using bioelectronic medicine.

An interesting molecule in persistent inflammation is the alarmin HMGB1, which is involved in or associated with the pathogenesis of many inflammatory diseases and regulated by cholinergic signals (Huston et al., 2007; Andersson and Tracey, 2011). HMGB1 is a “late mediator” of inflammation, and excessive levels of HMGB1 may be involved in sustaining inflammation, with associated inflammatory tissue damage, organ dysfunction, and cognitive impairment (Andersson et al., 2000; Valdes-Ferrer et al., 2016; Bruck et al., 2020). Of note, anti-HMGB1 therapy was effective in significantly reducing disease severity even when administered days after onset of severe inflammation (Huston et al., 2007), and anti-HMGB1 antibodies and HMGB1-targeted sorption beads have a protective effect in experimental intestinal inflammation (Maeda et al., 2007; Yang et al., 2009; Ju et al., 2014). Based on that knowledge, it will be interesting to consider whether the link between VNS and reduced HMGB1 levels may help explain how VNS may be effective in reducing not only inflammation onset, but also ongoing inflammation (Huston et al., 2007; Bonaz et al., 2016). Accordingly, our observation here that VNS reduced serum levels of HMGB1 along with molecular markers of inflammation in the gut and small bowel lesion area is noteworthy. The potential role of HMGB1 in IBD is being further explored (Andersson and Tracey, 2011) and a pathogenetic role has been proposed for HMGB1-dependent TLR4 activation in indomethacin-induced small intestinal damage (Nadatani et al., 2012). The available data supports further consideration and study of the vagus nerve regulation of HMGB1, its potential as a biomarker and therapeutic target in IBD (Palone et al., 2016).

The current study, using a rat indomethacin-induced enteropathy model, provides evidence that electrical activation of the cervical vagus nerve reduces inflammatory lesions in the small bowel, a pathology that is commonly a critical part of the pathogenesis in Crohn’s disease (Liu et al., 2020). The rodent models of colitis that have previously been used to study vagus nerve and α7 nAChR-mediated regulation of disease severity are by many considered more relevant to ulcerative colitis than to Crohn’s disease (Snoek et al., 2010; Galitovskiy et al., 2011; Ghia et al., 2011; Ji et al., 2014). In humans, previous surgical vagotomy was associated with later development of Crohn’s disease, but not ulcerative colitis (Liu et al., 2020), suggesting that vagus nerve regulation of inflammation may be more important in small bowel inflammation than in colitis, although this remains to be comprehensively studied.

As described above, the first published study of VNS for the treatment of Crohn’s disease used high duty cycle nerve stimulation (Bonaz et al., 2016; Sinniger et al., 2020). While no major safety concerns were reported, higher frequency and intensity of electrical nerve stimulation consume more energy, thus requiring greater battery capacity for implanted stimulators. Furthering our understanding of the VNS parameter requirements for clinical benefit can therefore allow for optimization of device design and stimulation protocols. Longer periods between stimulations may also reduce patient discomfort both by limiting the stimulation-associated laryngeal muscle contraction and temporary voice changes as well as enabling smaller devices that are easier to implant and maintain (Levine et al., 2018b,c). Therefore, the observation that a short VNS period of 60 s was followed by a prolonged anti-inflammatory effect provides important information for the development of future clinical therapeutics.

Electrical stimulation of the vagus nerve, but not of the splenic nerve, reduced indomethacin-induced small intestinal enteropathy for up to 2 days. The spleen was not required for the effect of VNS on small bowel lesion area. These observations are an important step toward better understanding of the mechanisms of vagus nerve regulation of inflammation in diseases with engagement of the small bowel.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The animal study was reviewed and approved by the Feinstein Institutes for Medical Research Institutional Animal Care and Use Committee.

AC, AD, and YL planned and performed experiments. AC, AD, YL, ME, and PO analyzed and interpreted data. AC, YL, PO, and ME wrote the manuscript. All authors edited the manuscript.

YL is an employee of SetPoint Medical, Inc. AC and AD were employees of SetPoint Medical, Inc. at the time of data collection. Aside from co-author affiliation, this funder was not involved in the study design, data analyses, data interpretation, and the writing of the report. ME has received honoraria for lectures and consultancy from AbbVie, Merck (MSD), Takeda, Ferring, Orion Pharma, Otsuka, Tillotts, Novartis, Pfizer, and Janssen, received research funding from AbbVie and MSD, and has been a former shareholder of Emune AB. These funders were not involved in the study design, data analyses, data interpretation, and the writing of the report. PO has received honoraria for lectures from Ferring and Janssen and is a shareholder of Emune AB. AC and PO were supported by MedTechLabs, Stockholm. These funders were not involved in the study design, data analyses, data interpretation, and the writing of the report.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.