#1 Threshold for intracellular current injections In its in vivo environment the neuron generates the classical Hodgkin-Huxley action potential at its axon hillock in response to depolarization that results from neurotransmitter activated calcium channels. This action potential is generated when the cell is quickly depolarized to a voltage (threshold) where transient positive feedback occurs via fast activating sodium channels. This depolarization disturbance propagates down the axon. In our in vitro model of a neuron the same depolarization can be generated via a current pulse that transiently depolarizes the cell. This does not require the use of a multiple compartment model and can easily be modeled as a point neuron. In the simulation we ran the current pulse applied was varied in amplitude between 0.1 and 1.0nA. In all cases the membrane voltage was depolarized in response to the current injection, although we estimated the minimum current amplitude to generate an action potential as 0.9nA. The typical ‘all or non’ description of the action potential was not observed. It should be noted that this estimate is however based on a rather arbitrary estimate of the threshold required to generate an action potential. #2 Effect of Synchronization index on the CN inputs In this experiment the synchronization index of the dendritic inputs was varied between 0.3 and 0.9 while the effects on interaural phase difference (IPD) tuning of the NL cell were noted. The default number (30) and strength of dendritic CN inputs were used while the IPD of the tone stimulus was varied between 0° and 180°. It should be noted that this particular experiment could be run using a point neuron, but the multi-compartment model is a more precise since it incorporates spatial gradient in membrane voltages as depolarization travels inward from the dendrite. One would expect that the more synchronized the inputs to dendrites the more synchronized the output from the NL cell would be. Furthermore, one could expect with strict (0.9) synchrony of the inputs the NL cell’s sensitivity to IPD would be augmented. These two hypotheses can be explained with the following reasoning. CN Synchronization Index =0.9 When the vector strength of each CN input is 0.9, many synapses on both the right and left dendrite are expected to activate nearly simultaneously, with the timing of this near simultaneous event governed by the phase of the ipsalateral stimulus. This near simultaneous firing of many synapse produces a strong depolarization of the respective dendrite which travels toward the soma. When the IPD is near zero there will be two strong depolarization waves arriving nearly simultaneously from each dendrite, and the NL cell is very likely to fire immediately following the arrival of these waves. Thus, we expect a very high spike rate with a IPD near zero, and we expect the synchrony of these NL firings to reflect the strict synchrony of the arriving depolarization waves. When the IPD is near 180° both the left and right dendrites still present strong (yet subthreshold) waves of depolarization but these waves will arrive strictly out of phase. Since the synchrony of both the left and right dendrite is so strong, these two waves of depolarization will never meet each other in time, and the NL cell is unlikely to fire. This will yield sharp IDP tuning. These predictions are consistent with the plots of figure 1 which shows spike rate and NL synchrony as a function of IPD. When the IPD is near 0°, the spike rate is very high with perfect vector strength of 1. When the IPD is near 180°, no NL spikes were generated (vector strength zero). CN Synchronization Index =0.3 When the vector strength of each CN input is 0.3, the soma of the NL cell is likely to receive weaker waves of depolarization that are spread out over the stimulus cycle of the left and rightdendrites respectively. With the IPD is near 180°, we expect a low, yet non-zero, NL firing rate where firing can occur only when significant left dendrite depolarization happens to be late and right dendritic depolarization happens to be early (or vise versa). When the IPD is near 0°, we expect only a moderate increase in firing rate since the synchronization index of both the left and right stimulus is only 0.3. These expectations are consistent with the date shown in figure 1. What was rather surprising in figure 1 is the insensitivity of the NL cell synchronization index (~0.8) to IPD when the input synchronization was 0.3. However, we might hypothesize that for both the 0° and 180° IPD the NL spikes generated would be most likely to occur at a particular point in time with respect to the left and right stimulus cycles, and thus the synchronization index of the NL output could be greater than the synchronization of the CN inputs. This makes sense if we consider the left and right inputs as probability density functions distributed over the course of the stimulus cycle (period histograms) as is schematized in figure 1C for the 180° and 0° IPD cases. Multiplying the broad left and right probability distributions (VS=0.3) together can yield a probability distribution of NL firing (over the course of the stimulus cycle) which has a lower variance and thus a higher synchronization index. This explains how the vector synchronization index of the NL output could exceed that of the CN inputs. FIGURE 1A -200 -150 -100 -50 0 50 100 150 2000100200300Interaural Phase Difference (deg)Discharge Rate (sp/sec)Effect of stimVS on IPD Tuning-200 -150 -100 -50 0 50 100 150 20000.20.40.60.81Interaural Phase Difference (deg)Synchronization Index0.30.50.70.9FIGURE 1B -200 -150 -100 -50 0 50 100 150 20000.20.40.60.81Interaural Phase Difference (deg)Normalized Rate (sp/sec)Effect of stimVS on IPD Tuning0.30.50.70.9-200 -150 -100 -50 0 50 100 150 20000.20.40.60.81Interaural Phase Difference (deg)Synchronization Index FIGURE 1C#7 Effect of Inhibitory Inputs In this simulation, the effect of applying a constant inhibitory input to both dendrites was examined. The increasing inhibitory input would be expected to decrease the unit depolarization provided by one synaptic event at a CN input. Physiologically this corresponds to an increase in the conductance of the dendrite to potassium ions which tends to hyperpolarize the cell and diminish the effect of excitatory depolarizing inputs. This type of simulation can easily be run using a