Fig. 7

Implementation and validation of a model of VNS effects on HR. A Schema of computational model. VNS pattern served as input from the cervical vagus nerve (CVN) to post-ganglionic cells in the intrinsic cardiac nervous system (ICNS). ICNS included mechanisms for frequency-dependent VN-to-ICNS synaptic failure and ICNS intrinsic firing. ACh release from ICNS cells to sinoatrial node (SAN) cells was modeled using a three-compartment model; post-synaptic mechanisms for ACh-dependent modulation of ion channels and internal calcium storage were incorporated in our 10-by-10 network of SAN cells. The network of SAN cells consisted of spontaneously firing cells. The mean firing rate of the SAN cells on the periphery of the 10-by-10 network was interpreted as HR. B Model responses where SAN cell firing (i) was used as a proxy for HR (ii). To compare with in vivo data, HR during stimulation (t = 15 s to 40 s) was normalized to pre-stimulation baseline (t = 0 to 10 s) (iii). The first 5 s of stimulation were excluded from analysis due to the transient response in HR. C Stimulation amplitude was simulated by adjusting the number of SAN cells receiving VNS-evoked ACh release from the cells of the ICSN (Synapse Density) during constant frequency stimulation (5, 20, and 50 Hz) for mean (black line) and SE (grey shaded area). Higher stimulation amplitudes were implemented as higher synaptic densities. D Model outcomes (dashed lines) compared to in vivo data (solid lines). Data are presented as mean ± standard error, n = 9-10 per parameter set for in vivo data, n = 10 runs per parameter set for model data