Background: The gut- microbiota-brain axis has received increasing attention recently due to evidence that colonic microbes can affect brain function and behavior. Recent studies have demonstrated that vagal afferent neurons may be an important conduit between the gut microbiota and the brain, as it is capable of detecting mediators released from the gut microbiota. However, it is unknown whether i) the gut microbiota from healthy human stool donors affects nodose ganglion (NG) neurophysiology or ii) gut microbial dysbiosis during inflammatory bowel disease (IBD) impacts the function of vagal afferent neurons.
Hypothesis: Remodeling of gut microbiota during IBD increases secretion of mediators that change the excitability of vagal afferent neurons.
Methods: To examine the effect of IBD patients’ fecal supernatant (FS; 1:20 dilution) on the excitability of mouse vagal afferent neurons, NG neurons from C57/Bl6 mice were collected, dissociated, and incubated overnight with FS from 5 active Crohn’s disease (CD) patients, 7 active ulcerative colitis (UC) patients, and 5 healthy volunteers (HV). Current and voltage-clamp recordings were used to assess changes in neuronal excitability and ion channel function.
Results: CD and UC FS significantly increased the excitability of NG neurons through a reduction in the rheobase of 40% (CD:60 cells vs. control: 54 cells), Plt;0.0001, Mann-Whitney test) and 23% (n= 50 cells vs. control: 46 cells, P= 0.0038, Mann-Whitney test) compared to their individual vehicle control neurons, respectively. This decrease in rheobase was accompanied by a two-fold increase in the number of action potentials elicited at twice rheobase (P lt;0.01, Mann-Whitney test). However, neither resting membrane potential, nor input resistance was altered in NG neurons treated with IBD FS compared with vehicle control neurons. HV FS had no effect on NG excitability. CD and UC FS significantly reduced voltage-gated K+ currents (P= 0.0075 and P lt;0.0001, two-way ANOVA followed by Sidaks multiple comparison test, respectively), but had no effect on voltage-gated Na+ currents. The excitatory effect of CD and UC FS of NG neurons was blocked by the cysteine protease inhibitor (E64) (30 nM), but not the serine protease inhibitor (FUT175) (10 μM). In all CD patients, pre-incubation of E64 blocked the increase in excitability by CD patient FS (CD rheobase:40.6 ± 3.6 pA vs.CD+ E64 rheobase: 73.9 ± 4.5 pA) (Plt;0.0001, one-way ANOVA followed by Tukeys multiple comparison). However, E64 was only able to block the excitatory effect of FS from 5 out of 7 UC patients FS (UC rheobase:43.8 ± 4.7 pA vs. UC+ E64 rheobase: 83.5 ± 4.3 pA) (P=0.0092, one-way ANOVA followed by Tukeys multiple comparison). The protease-activated receptor 2 (PAR2) antagonist GB83 (10 μM) also blocked the effect of the CD and UC patient supernatant on NG neurons (P=0.0081 and P=0.0116, Kruskal-Wallis test, respectively).
Conclusion: FS from active IBD patients contain mediators that can excite NG neurons. Cysteine proteases directly mediates the effect of IBD FS on NG neurons by activation of PAR-2. Signaling pathways downstream of PAR-2 activation lead to inhibition of voltage-gated K+ currents.
