Vagus Nerve Stimulation Reduces Indomethacin-Induced Small Bowel Inflammation

Abstract

Crohn's disease is a chronic, idiopathic condition characterized by intestinal inflammation and debilitating gastrointestinal symptomatology. Previous studies of inflammatory bowel disease (IBD), primarily in colitis, have shown reduced inflammation after electrical or pharmacological activation of the vagus nerve, but the scope and kinetics of this effect are incompletely understood. To investigate this, we studied the effect of electrical vagus nerve stimulation (VNS) in a rat model of indomethacin-induced small intestinal inflammation. 1 min of VNS significantly reduced small bowel total inflammatory lesion area [(mean ± SEM) sham: 124 ± 14 mm², VNS: 62 ± 14 mm², p = 0.002], intestinal peroxidation and chlorination rates, and intestinal and systemic pro-inflammatory cytokine levels as compared with sham-treated animals after 24 h following indomethacin administration. It was not known whether this observed reduction of inflammation after VNS in intestinal inflammation was mediated by direct innervation of the gut or if the signals are relayed through the spleen. To investigate this, we studied the VNS effect on the small bowel lesions of splenectomized rats and splenic nerve stimulation (SNS) in intact rats. We observed that VNS reduced small bowel inflammation also in splenectomized rats but SNS alone failed to significantly reduce small bowel lesion area. Interestingly, VNS significantly reduced small bowel lesion area for 48 h when indomethacin administration was delayed. Thus, 1 min of electrical activation of the vagus nerve reduced indomethacin-induced intestinal lesion area by a spleen-independent mechanism. The surprisingly long-lasting and spleen-independent effect of VNS on the intestinal response to indomethacin challenge has important implications on our understanding of neural control of intestinal inflammation and its potential translation to improved therapies for IBD.

Keywords: Crohn’s disease; bioelectronic medicine; cholinergic anti-inflammatory pathway; indomethacin; inflammatory bowel disease; inflammatory reflex; small bowel; vagus nerve stimulation.

FIGURE 1

VNS reduced intestinal ulcerations in indomethacin-induced enteropathy. Rats were subjected to VNS- or sham-treatment followed by subcutaneous injection of 10 mg/kg indomethacin in 5% sodium bicarbonate. (A) Representative photograph of a jejunum section in sham and VNS-treated rats. (B) Quantification of total lesion area. (C) Photomicrograph of jejunum with visible mucosal lesion after staining with hematoxylin and eosin. (D,E) Myeloperoxidase activity in the small intestine as measured by (D) peroxidation and (E) chlorination assays. Data is presented as mean ± SEM. **p < 0.01 vs. sham (B), n = 17–20 per group; *p < 0.05 vs. sham (D,E), n = 6 per group.

FIGURE 2

VNS reduced serum TNF in rats with indomethacin-induced enteropathy. Blood was collected and (A) serum TNF was measured by ELISA. n = 5–6 per group. (B) Serum HMGB1 was measured semiquantitatively by western blot and normalized to sham. n = 6–8 per group. Data is presented as mean ± SEM. **p < 0.01, ***p < 0.001 vs. sham.

FIGURE 3

Intestinal levels of select cytokines in VNS- or sham-treated animals with indomethacin-induced enteropathy. (A) Semi-quantitative measurements by western blot of intestinal IL-23 were normalized to the mean sham level. n = 9–11 per group. (B) Levels of select cytokines were measured by quantitative multiplexed electrochemiluminescence assay and normalized to the mean sham level. n = 5–6 per group. Data is presented as mean ± SEM. *p < 0.05 vs. sham.

FIGURE 4

VNS reduced intestinal ulcerations in indomethacin-induced enteropathy independent of spleen. (A,B) Rats were subjected to splenectomy (SPX), followed by VNS or sham treatment and indomethacin administration. The small intestine lesion area was quantified at 24 h. Mean and SEM values are plotted (A, n = 7–9 per group; B 9 per group). (C,D) Rats were subjected to splenic nerve stimulation (SNS) or sham SNS surgery followed by indomethacin injection. The small intestine lesion area was quantified at 24 h. Serum HGMB1 was measured semiquantitatively by western blot and normalized to mean sham values (C, n = 13–14 per group; D, n = 12 per group). Data is presented as mean ± SEM. * p < 0.05 vs. sham.

FIGURE 5

Sustained protection against indomethacin-induced intestinal ulcerations after VNS. Rats were subjected to VNS or sham surgery followed by a rest period of 0.5, 24, 48, or 72 h and subsequent subcutaneous injection of 10 mg/kg indomethacin in 5% sodium bicarbonate. Animals were injected systemically with Evans blue prior to euthanasia at 24 h after indomethacin injection. The small intestine was formalin-fixed, photographed and digitized. Total lesion area was quantified by a blinded scorer using Scion Image or ImageJ. Data is presented as mean ± SEM. *p < 0.05 vs. sham; n = 5–20 per group. Dotted line specifies the mean of the sham group.