Activation of the Inflammatory Reflex in Rheumatoid Arthritis and Inflammatory Bowel Disease; Preclinical Evidence

Abstract

The “inflammatory reflex” is an important neural-immune mechanism that can be activated by electrical vagus nerve stimulation (VNS) to drive pleiotropic antiinflammatory effects in the periphery. It has therefore been posited that chronic inflammatory diseases, such as rheumatoid arthritis (RA) and inflammatory bowel disease (IBD), may be treated by targeted neuromodulation of the vagus nerve. Here we present wide preclinical evidence of ameliorating inflammatory disease through activation of the inflammatory reflex, providing the rationale and framework in which to study VNS as a putative treatment in clinical IBD and RA.

Keywords

Bioelectronic medicine, Cholinergic antiinflammatory pathway, Inflammatory bowel disease, Inflammatory reflex, Rheumatoid arthritis, Vagus nerve stimulation

Outline
Introduction 1493
Structure and Function of the Cholinergic Antiinflammatory Pathway 1494
- Preclinical Evidence for Cholinergic Antiinflammatory Pathway Effect in Models of Rheumatoid Arthritis 1496
- Preclinical Evidence for Cholinergic Antiinflammatory Pathway Effect in Models of Inflammatory Bowel Disease 1497
Summary 1499
References 1499

Introduction

Rheumatoid arthritis (RA) and inflammatory bowel disease (IBD) are chronic autoimmune diseases of complex etiology, affecting approximately 1%–2% of the population. Both diseases cause debilitating symptoms and progressive destruction in the joints and intestinal tract, respectively, severely affecting quality of life. Due to the exacerbating effect of ongoing systemic inflammation, RA patients also carry elevated risk of cardiovascular mortality ( ) while IBD patients have increased incidence of colorectal cancer ( ). Biologic therapies (e.g., tumor necrosis factor (TNF) inhibitors) have greatly improved the care of RA and IBD patients in recent years, yet these drugs are expensive, contain inherent risks of serious off-target effects, and are not effective in all patients. Thus, there remains a significant need for alternative therapeutic approaches.

A recent confluence of immunology, physiology, and neuroscience has led to the discovery of coordinated, autonomic nervous system (ANS) control over inflammation ( ). One way that the central nervous system regulates innate and adaptive immunity is through the “inflammatory reflex,” which senses inflammation in the periphery and responds via its efferent arm, termed the cholinergic antiinflammatory pathway (CAP) ( ). The major neural conduit for both afferent sensing and efferent activity is the vagus nerve, the tenth cranial nerve that emerges from the brainstem to innervate most organs of the viscera ( Fig. 126.1 ). On sensing infection or tissue injury, neural signals are sent through the afferent vagus nerve, processed centrally, and relayed out through efferent vagus neurons, leading to resolution of inflammation through release of humoral cortisol or locally acting neurotransmitters ( ). Intriguingly, tonic vagus nerve activity is decreased in RA, IBD, and other inflammatory diseases ( ).

The CAP has been activated electrically or pharmacologically to reduce pathological inflammation in numerous animal models of disease ( ) and electrically accessing this system via vagus nerve stimulation (VNS) has been proposed as a novel, nonpharmacologic treatment for clinical inflammatory diseases such as RA and IBD ( ).

In this chapter, we will review the biological mechanisms of the inflammatory reflex. We will then present evidence in animal models of RA and IBD of the importance of the CAP in progression of disease and evidence of the use of electrical stimulation of the vagus nerve to ameliorate disease in those models. Finally, we will discuss data from cultured human cell experiments and observational clinical studies that support exploring the use of VNS in patients with RA and IBD.

Structure and Function of the Cholinergic Antiinflammatory Pathway

The ability of the vagus nerve to restrict systemic inflammation was first demonstrating by Tracey and colleagues 15 years ago when they injected a lethal dose of endotoxin into rats with bilateral or sham cervical vagotomies ( ). Vagotomized rats produced more TNF, a key proinflammatory cytokine, and had earlier shock onset than rats with intact vagus nerves, which indicated that a physiologic restriction of cytokine production had been disabled. Electrical stimulation of the distal stumps of the transected vagus in the vagotomized rats resulted in decreased TNF production when compared to the sham rats and rescued the animals from shock. This demonstrated that activating the efferent vagus nerve electrically, not only restored normative restriction of inflammation, but importantly, achieved a new “immunological set point,” such that TNF production was further decreased with respect to sham when similarly challenged immunologically ( ).

Selective lesioning experiments in rodents demonstrate that the anatomical path of the CAP in endotoxemia is hardwired through the common celiac branch of the subdiaphragmatic vagus nerve, across the celiac plexus, through the splenic nerve and into the intact spleen ( ) ( Fig. 126.2 ). Stimulating the nerves along this path recapitulate the cytokine restriction of cervical VNS ( ). The neuronal signaling is then transmitted to effector cells within the spleen via neurotransmitter exchange to and between immune cells; nerve terminals of the catecholaminergic splenic nerve release norepinephrine that act on β-adrenoceptors ( ) on the cell surface of choline acetyltransferase positive (ChAT ⁺) CD4 ⁺CD44 ^highCD62 ^lowT cells that then produce and release acetylcholine (ACh) for tens of minutes ( ). The ACh binds the α7 nicotinic ACh receptor (nAChR) expressed on the resident macrophages. The α7 nAChR is essential for the antiinflammatory effects of the CAP in endotoxemia, as demonstrated by targeted disruption of the receptor ( ) and adoptive transfer experiments ( ).

Intracellular cascades, subsequent to α7 nAChR agonism, result in reduction in proinflammatory cytokine production through activation of the Janus-activated kinase (JAK)-2/signal transducer and activator of transcription (STAT)-3/suppressor of cytokine signaling (SOCS)-3 counterregulatory pathway ( ), inhibition of the proinflammatory transcription factor Nuclear Factor κ-light-chain-enhancer of activated B cells (NFκB) pathway ( ), and prevention of mitochondrial DNA translocation and inflammasome activation ( ). VNS and cholinergic agonists also reduce splenic marginal zone B cell trafficking to the red pulp and perifollicular areas and reduce secretion of antibodies ( ).

ACh released by ChAT ⁺lymphocytes can also act on circulating immune cells that traverse the spleen, leading to phenotypic changes in monocytes ( ) and neutrophils ( ), thereby decreasing inflammatory infiltrate to distant sites of inflammation, such as the joints in RA. Outside of the spleen, VNS reduces granulocytic infiltration to the muscularis mucosa in a model of inflammatory postoperative ileus ( ), to the pancreas in pancreatitis ( ), and to the lung following burn-induced acute lung injury ( ). Further, several studies noted that cholinergic agonism modulated and normalized the number and function of regulatory T cells, a specialized antiinflammatory T cell subset, whose characteristic function as a suppressor of other immunocytes is abnormal in many inflammatory diseases, including IBD and RA ( ). In addition, α7 nAChR agonists reduced IL-6 and IL-8 production in immunologically active synoviocytes that express α7 nAChR in man ( ), while fibroblast-like and mononuclear-like cells within the synovium express ChAT, the rate limiting enzyme critical for production and local release of ACh ( ).

In contrast to the path through the spleen, vagal efferent fibers directly innervate the gut, synapsing directly with the myenteric plexus ( Fig. 126.1 ). Densely organized nerve fibers lie in close apposition to macrophages, dendritic cells, glial cells, mast cells, and enteroendocrine cells in the gut wall. The ACh-producing T cells that are a crucial part of the CAP in the spleen have been identified in lymph nodes and Peyer’s patches ( ), yet may not be required to provide ACh to macrophages in the muscularis of the small intestine for VNS to affect an antiinflammatory response. VNS reduces intestinal inflammation in a model of postoperative ileus in recombinase-activating gene (RAG)-1 knockout animals lacking T cells and in wild type mice with surgical denervation of the spleen. Inflammation in this model was not reduced by VNS in animals with targeted disruption of the α7 nAChR, but this deficit could be reconstituted by bone marrow transplant from wild-type animals ( ). Furthermore, in a postoperative ileus model of intestinal inflammation and delayed transit, subdiaphragmatic selective vagus neurectomy, but not splenic denervation, increases proinflammatory cytokines in gut muscularis ( ). In contrast, the spleen was required for resolution of mucosal inflammation by centrally acting cholinergic agonists in a study in dextran sulfate sodium (DSS) colitis ( ). Perhaps the different splenic dependencies between these two models were due to the gastrocolonic gradient of vagus nerve innervation of the gastrointestinal tract ( ) or the presence of the innervated and lymphoid-rich Peyer’s patches in the small bowel.

Vagus nerve terminals have been shown to interface with cholinergic myenteric fibers, yet the question of precisely which, if any, nonneuronal intermediate cells complete the pathway between the vagus and macrophages in the in gut is still being resolved ( ). In addition, it is still unclear whether the CAP terminates exclusively on the α7 nAChR in the gut, as it has been demonstrated that murine α5 nAChR knockouts were significantly more susceptible to experimental colitis than wild type controls and that nicotine significantly ameliorated the colitis in these animals ( ). Further details of the anatomic, cellular, and molecular pathways involved in the inflammatory reflex and the CAP are reviewed in Chapter 125 and elsewhere ( ).

Preclinical Evidence for Cholinergic Antiinflammatory Pathway Effect in Models of Rheumatoid Arthritis

The collagen-induced arthritis (CIA) model is a rodent model of inflammatory arthritis, often used to explore molecular mechanisms of disease as well as to screen RA candidate drugs ( ). In this model, autoimmunity against the joint is induced by repeated injections of collagen, often in the presence of an immune-activating compound such as Freund’s adjuvant. Animals predictably develop joint inflammation that can be assessed by caliper measurement of joint swelling. The resulting pathology in the joint during CIA is analogous to that seen in RA patients, with infiltration of the synovium of immune effector cells, formation of an inflammatory synovial outgrowth (pannus), damage to articular cartilage, and periarticular erosions of the bone. Within the last several years, the CIA model has been utilized in several ways to demonstrate the importance of various aspects of the CAP in CIA disease progression, thereby providing rationale to study the therapeutic potential of electrical CAP activation with VNS in RA.

To explore the physiologic role of the α7 nAChR in the progression of disease, Tak and colleagues induced CIA in mice with genetic disruption of the α7 nAChR ( ). The α7 nAChR knockout group had a greater cumulative incidence of disease onset, worsened clinical disease severity and radiographic evidence of bone destruction, increased histological joint inflammation, increased systemic monocyte chemotactic peptide (MCP)-1 and TNF levels, and increased in vitro release of Th1 cytokines from cultured splenocytes, as compared to the wild-type group ( ). Conversely, when mice with CIA were treated with either nicotine or agonists specific to the α7 nAChR, disease progression was delayed and suppressed, resulting in reduced joint inflammation and bone destruction ( ) Additionally, cholinergic agonists reduced serum TNF and IL-6, joint production of TNF and HMGB1, and the translocation of HMGB1 from the nucleus to cytoplasm ( ). In addition, reported that nicotine induced a suppression of proinflammatory Th17 response and a shift toward greater Th2 versus Th1 balance. Nicotine reduced Th17 markers IL-17A in serum and splenic RORγτ expression. Furthermore, there was a significant increase in systemic levels and splenic expression of Th2 markers IL-4 and GATA3, respectively, yet no detectable changes in Th1 markers IFN-γ and T-bet.

In addition to probing the importance of the effector cell components of the CAP in CIA, there are several studies focused on the neural aspects of the reflex, and its activation through electrical VNS. A number of groups reported that, similar to mice with disrupted α7 nAChR, unilateral cervical vagotomy tended to increase the rate of CIA onset and worsened disease severity when compared to intact animals ( ), an effect perhaps self-limited by the intact contralateral nerve. Conversely, activating the CAP by directly stimulating the vagus nerve improves CIA. In an alternate approach to electrical stimulation, induced CIA in rats that have had their vagus nerve surgically suspended against the sternocleidomastoid muscle, producing a chronic mechanical stimulation that resulted in measurable bioelectric activity in the vagus nerve. When compared with the sham-operated group, animals with surgical suspension had statistically significant improvements in paw swelling, clinical arthritis score, bone erosions, periarticular inflammation, and systemic TNF levels.

Our group then expanded on these observations by demonstrating that low amounts of daily electrical VNS ameliorated disease in CIA. In that study, a percutaneous electrode with a bipolar cuff, analogous to those used in humans treated with implantable VNS devices, was implanted around the left vagus nerve of rats. Treatment (0.2 ms pulse width, 10 Hz, 3 mA) was initiated when the disease had become semiestablished, and was delivered for 60 s each day. VNS reduced clinical manifestations as compared to implanted but unstimulated animals as assessed by ankle and knee caliper measurements of joint swelling, and reduced the histological severity of inflammation, pannus formation, cartilage damage, and bone resorption ( Fig. 126.3 ). Moreover, this was accompanied by a reduction in circulating proinflammatory cytokines ( ). In addition, important mediators of osteoclastogenesis and osteoclast activity were significantly shifted away from a proresorptive state ( ).

These observations in animal models of RA are supported by experiments in human-derived tissue, cell culture, as well as clinical observations. Treatment of synovial tissue-derived fibroblast-like synoviocytes with nicotine or specific α7 nAChR agonists significantly reduces production of the proinflammatory cytokines, IL-6 and IL-8 ( ), possibly through pathways mediated by JAK2/STAT3 ( ). Furthermore, there is a clinical association between the use of nicotine-rich snuff (smokeless tobacco) by RA patients and lower disease activity ( ).

Reduction in RA-related inflammation due to CAP activation is also consistent with observations that the proinflammatory mediator HMGB1, in circulation of RA patients, is inversely related to tonic vagus nerve activity ( ). The authors noted that the negative association of proinflammatory mediators and vagus nerve activity was a “chicken and the egg” dilemma; did the low vagal tone cause the inflammation, or did the inflammation suppress the vagal tone? Tak and colleagues recently shed light on this dilemma, demonstrating in a prospective study of patients with RA and those at risk of developing RA (subjects with detectable autoantibodies but without RA symptomatology) that autonomic dysfunction preceded disease. They reported a negative association between vagus nerve activity and RA status, and decreased vagus nerve activity in at-risk patients relative to healthy controls. Importantly, they discovered that the at-risk patients who eventually developed RA had lower vagus nerve activity than those at-risk patients that did not develop RA, coupled with lower expression levels of α7 nAChR on circulating monocytes ( ). Together, these studies demonstrate the importance of the CAP in affecting the course and severity of CIA ( Fig. 126.4 ), and provide the rationale to study the therapeutic potential of VNS in RA patients. Successful clinical studies in RA will be reviewed in Chapter 127 .

Preclinical Evidence for Cholinergic Antiinflammatory Pathway Effect in Models of Inflammatory Bowel Disease

Similar to the CIA model in RA drug development, rodent models of intestinal mucosal inflammation have been widely used to explore mechanisms of action and to test preclinical efficacy of candidate drugs for studies in inflammatory bowel disease. The complex pathology of IBD is not well encompassed by any single rodent model to date, so several types have been developed to emulate aspects of the clinical disease ( ). The models fall generally into four categories: (1) direct mucosal irritant-mediated inflammation (e.g., DSS); (2) delayed type hypersensitivity-mediated inflammation caused by exposure to mucosal haptenating agents (e.g., oxalazone, dinitrobenzene or trinitrobenzene sulfonic acid (DNBS/TNBS)); (3) spontaneous colon and small intestinal inflammation occurring in IL-10 knockout mice; and (4) mucosal inflammation induced by T cell transfer into RAG knockout mice lacking native T cells ( ). The therapeutic effect of an intervention in all of these models is assessed by changes between treated and controlled group in a subset of the following: colon weight and length, animal weight loss, stool characteristics, and frequency; morphometric or semiquantitative scoring of ulceration and inflammation in whole-intestine preparations; endoscopic scoring of disease; gut histology; local and systemic levels of inflammatory mediators; and intestinal tissue levels of immune cell products (e.g., leukocyte myeloperoxidase (MPO)).

Over the last two decades, numerous studies have interrogated aspects of the CAP in models of IBD, either by activation of the α7 nAChR with selective agonists, nonselective agonists, and acetylcholinesterase inhibitors, or by vagotomy and deactivation of nAChRs with antagonists. The majority of the studies ( Table 126.1 ) report that disease was worsened by vagotomy and nAChR antagonism, and ameliorated by activation of nAChRs in general or the α7 nAChR specifically, and by inhibition of central and peripheral acetylcholinesterase ( ). However, there are exceptions to this consensus: First, noted that while nicotine improved colitis in the IL-10 knockout model, it worsened jejunal inflammation. The authors considered this duality to be consistent with the clinical observation that smoking tends to decrease disease severity in ulcerative colitis yet exacerbates symptoms in Crohn’s disease ( ). Further, , reported that two different selective α7 nAChR agonists were either ineffective or exacerbated disease activity in the TNBS colitis model, with the exception of the highest tested dose which was protective. The authors had separately demonstrated that α4β2 nAChR, rather than α7 nAChR, transduced antiinflammatory signaling from the vagus nerve to reduce NF-κB activation and proinflammatory cytokine production while increasing bacterial uptake by gut macrophages ( ). Despite these partly conflicting observations, the preponderance of evidence from vagotomy and pharmacologic intervention studies in various categories of IBD models predicts that CAP activation may have beneficial effects on intestinal mucosal inflammation, and therefore may be clinically useful in treating IBD.

Table 126.1

Only gold members can continue reading. Log In or Register to continue