Specific vagus nerve stimulation parameters alter serum cytokine levels in the absence of inflammation

Abstract

Background: Electrical stimulation of peripheral nerves is a widely used technique to treat a variety of conditions including chronic pain, motor impairment, headaches, and epilepsy. Nerve stimulation to achieve efficacious symptomatic relief depends on the proper selection of electrical stimulation parameters to recruit the appropriate fibers within a nerve. Recently, electrical stimulation of the vagus nerve has shown promise for controlling inflammation and clinical trials have demonstrated efficacy for the treatment of inflammatory disorders. This application of vagus nerve stimulation activates the inflammatory reflex, reducing levels of inflammatory cytokines during inflammation.

Methods: Here, we wanted to test whether altering the parameters of electrical vagus nerve stimulation would change circulating cytokine levels of normal healthy animals in the absence of increased inflammation. To examine this, we systematically tested a set of electrical stimulation parameters and measured serum cytokine levels in healthy mice.

Results: Surprisingly, we found that specific combinations of pulse width, pulse amplitude, and frequency produced significant increases of the pro-inflammatory cytokine tumor necrosis factor (TNF), while other parameters selectively lowered serum TNF levels, as compared to sham-stimulated mice. In addition, serum levels of the anti-inflammatory cytokine interleukin-10 (IL-10) were significantly increased by select parameters of electrical stimulation but remained unchanged with others.

Conclusions: These results indicate that electrical stimulation parameter selection is critically important for the modulation of cytokines via the cervical vagus nerve and that specific cytokines can be increased by electrical stimulation in the absence of inflammation. As the next generation of bioelectronic therapies and devices are developed to capitalize on the neural regulation of inflammation, the selection of nerve stimulation parameters will be a critically important variable for achieving cytokine-specific changes.

Keywords: Inflammatory reflex; Interleukin-10; Neuromodulation; Tumor necrosis factor.

Fig. 1

Experimental design and stimulation pulse waveforms. a Experimental timeline. Electrical stimulation pulse trains were applied to the exposed left cervical vagus nerve for 4 min under anesthesia. Following stimulation, the animals were recovered for 2 h, euthanized, and whole blood was collected through cardiac puncture. b Schematic of the charge-balanced stimulation waveforms used during stimulation, with short pulse width (top) and long pulse width (bottom). The actual waveform shapes used in this study are shown to the right

Fig. 2

Specific stimulation amplitude and frequency combinations at 50 μs pulse widths reduce serum TNF levels. a A significant decrease in TNF, compared to the sham group, was observed with 30 Hz stimulation and a pulse amplitude of 200 μA. b A significant decrease in TNF was observed at 100 Hz stimulation with a pulse amplitude of 750 μA. Data is represented as individual mouse data points with mean ± SEM. n = 7–29 per group, *, P < 0.05

Fig. 3

Serum TNF is significantly increased by vagus nerve stimulation at 250 μs for a specific parameter combination. a Stimulation resulted in a significant increase in TNF at 30 Hz and 750 μA pulse amplitude, compared to the sham group. b No significant changes in serum TNF were observed with the 250 μs pulse width at 100 Hz. Data is represented as individual mouse data points with mean ± SEM. n = 7–24 per group, *** P < 0.001

Fig. 4

Serum IL-10 is increased by select parameters of electrical stimulation with 50 μs pulse width. a Stimulation for 50 μs at 30 Hz produced significant increases, compared to the sham group, in IL-10 for both 50 μA and 750 μA pulse amplitudes. b No changes in serum IL-10 were observed for 100 Hz stimulation across the four pulse amplitudes. Data is represented as individual mouse data points with mean ± SEM. n = 7–33 per group, *, P < 0.05

Fig. 5

Nerve stimulation with 250 μs pulse width increased serum IL-10 at several different parameters. a Stimulation with 250 μs pulses at 30 Hz produced a marked increase in IL-10 at the 750 μA pulse amplitude. b Stimulation with 250 μs pulses at 100 Hz produced significant increases at both 50 μA and 750 μA pulse amplitudes. Data is represented as individual mouse data points with mean ± SEM. n = 7–27 per group, *, P < 0.05; *** P < 0.001

Fig. 6

The effect of different vagus nerve stimulation parameters on heart rate. a Stimulation at 30 Hz with 50 μs pulse width resulted in bradycardia (≥10% reduction in heart rate) at only the 750 μA pulse amplitude. b At 100 Hz, bradycardia was not observed with the 50 μs pulse width. c Stimulation with the longer 250 μs pulse width resulted in bradycardia at 200 and 750 μA pulse amplitudes. d The largest decrease in heart rate was observed with the 250 μs pulse width at 100 Hz and 750 μA amplitude. Data is represented as individual mouse data points with mean ± SEM. n = 5–10 per group