Remarkably, the opposing effects of NA upon spontaneous and evoked inhibition were both due to noradrenergic elimination of cartwheel cell spontaneous spiking. Under control conditions, cartwheel synapses were tonically depressed by background spiking activity. By shutting off spontaneous spiking, NA relieved cartwheel synapses from depression and thereby enhanced glycine Volasertib release in response to parallel fiber stimulation. This mechanism for neuromodulation, in which synaptic output is indirectly controlled through modulation of spontaneous activity, may have distinct advantages over direct regulation of presynaptic release probability in spontaneously firing cells. We examined whether NA affects integration

of excitatory and inhibitory signals conveyed through the molecular layer circuitry of the DCN. Whole-cell voltage-clamp recordings were acquired from fusiform cells in acute slices of mouse brainstem and synaptic currents were recorded in response to activation of parallel fibers by an extracellular stimulating electrode positioned in the

DCN molecular layer (Figure 1A). Single stimuli typically elicited weak excitatory currents check details and small or undetectable inhibitory currents (see first stimulus; Figure 1B, top). Because parallel fibers exhibit strong short-term facilitation (Roberts and Trussell, 2010 and Tzounopoulos et al., 2004), brief stimulus trains (three stimuli at 20 Hz) were applied to recruit robust parallel fiber activity. When fusiform cells were clamped at −60 mV, intermediate to the reversal potentials below for Cl− conductances (−84 mV) and excitatory conductances (∼0 mV), each stimulus elicited a sequence of inward current followed

closely by outward current (Figure 1B), characteristic of direct activation of excitatory fibers followed by feed-forward recruitment of inhibitory inputs (Mittmann et al., 2005 and Pouille and Scanziani, 2001). Both inward and outward components of the responses were larger for the second and third parallel fiber stimuli due to facilitation of excitatory inputs onto both fusiform and cartwheel cells (Roberts and Trussell, 2010). Consistent with activation of disynaptic inhibition, inward and outward components of the evoked responses were largely abolished by application of NBQX (Figure 1C). When identical stimulus trains were applied in the presence of 10 μM NA, we observed a significant enhancement of the outward components of evoked currents in response to the second and third stimuli (Figures 1B, middle, and 1D; measured as total outward charge, see Experimental Procedures; stim 2 control: 701 ± 246 pA∗ms, NA: 1809 ± 561 pA∗ms, p < 0.05, n = 6; stim 3 control: 596 ± 203 pA∗ms, NA: 1680 ± 286 pA∗ms; p < 0.01, n = 6). This effect could be clearly visualized by subtracting average control responses from average currents recorded in NA (Figure 1B, bottom).