Units with phasic (on or on/sustained) response patterns and low (<3 spikes/s) spontaneous rates were most prevalent in both NH and HI mice. The top panel shows the percentage distribution of response types, with the number of each type given above the bar. The bottom panel shows the distribution of spontaneous rates, among the units from NH and HI mice.

There is no apparent difference between the NH and HI gap functions. Gap recovery functions for NH and HI mice, showing the strength of the response to NB2 relative to the response to NB1 as gap duration increases. The two top panels show the normalized NB2 response of phasic units from NH and HI C57 mice. Percent recovery at 6 ms gap widths and 48 ms gap widths are displayed in box plots in the bottom panels. Note that the 6 ms gap was not presented to every unit (NH: n = 43; HI: n = 47).

Sustained units recorded from NH mice have longer latencies than sustained units recorded from HI mice. Mean first spike latency is plotted against BF for phasic units (top panel) and sustained units (bottom panel). Open circles indicate data recorded from NH mice; filled circles indicate data recorded from HI mice. A trend toward shorter latency with higher BF is seen except for 13 phasic units whose latencies were over 25 ms and were omitted from the regression analysis (six HI and seven NH units, all with BFs between 10–23 kHz, denoted by arrows). First spike latency data were available for 30 of the 38 sustained and primary-like units from NH mice, and 33 of the 57 sustained and primary-like units from HI mice.

The majority of phasic units from both NH and HI mice responded to stimuli monotonically with mean rollover coefficients near 1.0. However, sustained units recorded from both NH and HI mice responded nonmonotonically to tone bursts, with mean RCs below 0.8, but monotonically to noise bursts. Distribution of rollover coefficients for phasic and sustained units from NH (top) and HI (bottom) mice for BF tones is shown on the left and for noise bursts on the right. Each box plot extends from the first to the third quartiles, with the horizontal line representing the median of the distribution. The whiskers indicate the 10th and 90th percentiles. The number of units included is shown in each box.

No systematic age-related changes in MGT are apparent in the sample. In the top panel, the MGTs are displayed in box plots according to the age of the mouse from which each unit was recorded. Note that logarithmic coordinates are used to display MGTs; all outliers are included in the graph. Symbols are defined the same way as in Figure 5, with the data points indicating outliers. The median and 25th percentile MGT for the 21- to 30-week-old mice were both 3 ms. The bottom panel shows the percentage of units recorded from NH and HI mice that had MGTs from 1–2, 3–4, 5–6, and >7 ms. The incidence of MGTs from 1–2 and >7 ms was greater for NH mice, but the incidence of MGTs from 3–4 and 5–6 ms was greater for HI mice, so that no systematic age-related trend was found.

The paucity of high BF units in HI mice is not because recordings were limited to dorsal recording locations within the IC. The top panel shows the unit BFs plotted against tone threshold for the units recorded from NH (open circles) and HI (filled circles) mice. It is clear that the HI sample is missing units with BFs above 18 kHz, and that more units from NH mice have thresholds under 5 dB SPL. In addition, there was more than a 2:1 difference, 78 vs 28%, in the number of units with BFs < 12 kHz in HI mice as compared to NH mice. Data were missing from six HI units. The bottom panel shows the distribution of depths at which units were recorded from NH and HI mice. This distribution does not differ systematically between the NH and HI samples.

MGTs varied with response type similarly in NH and HI IC units, and MGT did not vary with noise threshold. Mean MGT of units with the different response patterns to noise (OnS = on/sustained, PRI = primary-like, SUS = sustained, INB = inhibitory, BU = build-up) recorded from NH and HI mice are displayed in the top panel. The shortest MGTs are measured in OnS, ON, PRI, SUS, and BU units, and this did not depend on whether the unit was recorded from an NH or HI mouse. The bottom panel shows mean MGT of units with different noise thresholds; no systematic differences are apparent. The noise threshold categories of 0–10 and 81–90 dB were excluded from the figure and also from the ANOVA because of the small number of units in them: There was one unit with a noise threshold of 0 dB and an MGT of 12 ms, one with a noise threshold of 83 dB and an MGT of 2 ms, and one with a noise threshold of 85 dB and an MGT of 24 ms. These three units were all recorded from HI mice, and their data are consistent with the conclusion that MGT was unrelated to noise threshold.

Temporal responses patterns elicited by the gap stimuli were comparable for units recorded from NH and HI mice. The left column shows the responses of unit #438 (BF = 26.068 kHz, tone threshold = 27 dB) from an NH C57 mouse to noise bursts with (from top to bottom) 0, 6, 12, and 48 ms gaps. The timing of the gap stimuli are indicated below each PSTH. Note that the NB2 response grows with the width of the gap. The right column show responses of unit #319 (BF = 13.920 kHz, tone threshold = 30 dB) from an HI C57 mouse to the same stimuli. Although unit #319 had a higher noise threshold (45 dB) than unit #438 (noise threshold 35 dB), the gap series was 25–30 dB above threshold for each unit, and the responses of the two units are almost indistinguishable. It is interesting to note that although the unit from the HI mouse had a lower BF, the units were recorded at similar depths (1,876 μm for #438 and 1,804 μm for # 319) and had tone thresholds within 3 dB of each other. Both units responded to noise bursts with an on/sustained pattern (top row; responses to 150 ms noise bursts with 0 ms gaps), and both have low spontaneous rates (#438(NH), 0.18 spikes/s; #319(HI), 1.60 spikes/s). Their first spike latencies are similar (#438(NH): 1.00 ms, SD = 0.36; #319(HI): 3.87 ms, SD = 3.31), as are their rollover coefficients (#438(NH): tone = 0.80, noise = 1.13; #319(HI): tone = 0.79, noise = 1.16).

An example of the composite post-stimulus time histogram (PSTH) analysis is shown for a sustained unit (left) and phasic unit (right). For clarity, the PSTHs for just three of the nine gap widths are shown. The composite histogram is shown at the bottom of each unit. For the sustained unit, it was constructed by adding the spike counts at identical bin locations from all of the individual histograms in a gap series. The bin containing the end of the response to NB1 was used as the start of the gap window (GW), which had a variable duration equal to the gap duration being tested, and as denoted by the gray box. The response in the GW of each histogram was compared to the same time period in the control PSTH (lines denote comparison times), provided that the GW ended before the control response was over. Longer GWs were truncated to the end of the control response to exclude non-driven activity in the control PST from entering into the comparison. The phasic unit’s composite histogram was generated by shifting each gap PSTH “back in time” (to the left) by the gap duration and then summing all the spikes over the nine shifted gap PSTHs for gaps of 1 to 96 ms. The onset and duration of the response to NB2 were visually determined from the composite histogram (gray box) and used to define the NB2 window (NW). The onset of the NW in each gap PSTH was adjusted by adding the gap width to the onset time established from the composite PSTH. For each gap width, spikes were counted in this window and this count was compared to the spike count within the same window in the control (0 ms) PSTH as shown by the dashed lines. A significant difference (p < 0.05) between the NW of the control and gap PSTHs periods indicated that the unit responded to the stimulus gap. The onset of this control window for each of the three PSTHs in the example is indicated by the dotted grey lines. Note that because the NW is adjusted by the gap width in each PSTH, the corresponding NW in the control PSTH is different for each gap width. The smallest gap producing a significant difference (p < 0.05) from the control was considered to be the gap threshold, providing that the next largest gap width also produced a significant difference.