Synaptic inhibition in the LLDp is largely GABAergic. A, The LLDp (arrow) is located laterally within a 300-μm-thick coronal brainstem slice with ∼4 mm between the two LLDps. B, The LLDp is readily identifiable as a heavily myelinated nucleus medial and slightly ventral to the semilunar nucleus (SLu) in fresh tissue slices. C, Schematic of the experimental setup highlighting ipsilateral recording site (blue, left), and medial electrical stimulation of fibers projecting from the contralateral LLDp. D, Bath application of gabazine (10 μm), a GABAA receptor antagonist, abolished the eIPSC, whereas the eIPSC amplitude was not affected by strychnine (1 μm), a glycine receptor antagonist. E, Averaged traces of the eIPSC during control (black), strychnine (1 μm, blue), gabazine (10 μm, red), and wash (gray). F, In the majority of cells, eIPSCs were abolished by gabazine alone (n = 10), although occasionally an additional weak strychnine component was observed (n = 3). G–I, Population data of eIPSC amplitude, 20–80% rise time, and decay time constant (tau) for GABAergic eIPSCs (n = 10). For this and subsequent figures, mean ± SEM values are shown. d, dorsal; m, medial; ctrl, Control; stim, stimulation.
LLDp neurons have functional GABAA and glycine receptors. A, Schematic of the experimental setup highlighting direct puff application of receptor agonists to the recorded cell. B, C, In all neurons recorded, puff application of muscimol (GABAA receptor agonist, 10 μm, red; B) and glycine (glycine receptor agonist, 500 μm, blue; C) evoked IPSCs, which were blocked by their respective antagonists, gabazine (10 μm) and strychnine (1 μm). D, Amplitude of IPSCs did not significantly differ among puff application of muscimol (n = 7), glycine (n = 11), or a mixture of muscimol and glycine (n = 5), suggesting that GABA and glycine may interfere with each other. E, F, sIPSCs (top, black) were pharmacologically isolated into GABAergic sIPSCs (E, red) or glycinergic sIPSCs (F, blue) with bath application of strychnine (1 μm) or gabazine (10 μm), respectively. Averaged sIPSCs (bottom, thick lines) show distinct decay kinetics between GABAergic and glycinergic sIPSCs. G–I, Population data of sIPSC amplitude, 20–80% rise time, and decay tau. Population data for decay tau (I) shows significant difference between GABAergic (n = 3) and glycinergic sIPSCs (n = 3; p = 0.008, ANOVA with post hoc Fisher’s exact test). For this and subsequent figures: *p < 0.05, **p < 0.01, and ***p < 0.001. ctrl, Control; gly, glycine.
IPSCs can be evoked electrically or chemically from activation of the contralateral LLDp. A, Schematic of the experimental setup for electrical activation of contralateral LLDp, highlighting ipsilateral recording site (blue), and the following two stimulating locations: medial (left) and in the contralateral LLDp (right). B, C, Averaged eIPSCs from medial LLDp (B) and contralateral LLDp (C) stimulation in the same neuron. Note the larger amplitude and shorter latency in the medial eIPSC. Bath application of gabazine (10 μm, red) completely abolished both medial and contralateral eIPSCs. D–G, The contralateral eIPSC population data for amplitude, latency, 20–80% rise time, and decay tau (n = 10). H, Schematic of experimental setup for chemical activation of contralateral LLDp, highlighting ipsilateral recording site (blue), and direct puff application of glutamate in the contralateral LLDp (green, right). I, For a single neuron, puff application of glutamate (150 μm, 5–10 psi, 10 s) on the contralateral LLDp produced PSCs that were abolished by bath application of gabazine (10 μm) and strychnine (1 μm). Enlarged views of PSCs are shown to the right during puff glutamate (top, green), puff glutamate with inhibition blockers (middle, purple), and wash (bottom, gray). J, Overlay of individual PSCs (thin lines) and their average (thick line) during control (left, black) and during glutamate puff application in the following conditions: ACSF (top), inhibition blockers (middle), and wash (bottom). K, Overlay of averaged PSCs. L, Cumulative probability of PSC amplitude shows that PSCs during puff glutamate (green) have larger amplitudes compared with control (black) and puff glutamate with inhibition blockers (purple) conditions in a sample cell. M, Distribution of PSC amplitude shows a bimodal distribution of PSCs during glutamate application (green), with a population of events >100 pA, which is not seen in the control or during the inhibition blockers condition. N, Cumulative probability of the IEI between PSCs shows a decrease in IEI during puff glutamate (green) and wash (gray) compared with control (black), and puff glutamate compared with inhibition blockers (purple). O–Q, Sample traces from three individual neurons with varying degrees of responsiveness to contralateral glutamate puff (green bar). ctrl, Control.
Intrinsic regulation of the Erev for Cl− channels in LLDp neurons. A, eIPSCs from a sample neuron were inward initially (black) and shifted polarity (blue) during whole-cell recording. B, The eIPSC amplitudes are plotted over time showing that the shift occurred at about 8 min after whole-cell recording began. C, Population data of eIPSC amplitude over time (n = 16). eIPSCs were largely observed as inward currents initially, but in many cells the current became outward over time during whole-cell recordings. The shift in polarity generally occurred within 20 min. D, After the eIPSC became outward, bath application of furosemide (500 μm), a KCC2 antagonist, returned the eIPSC to an inward current. Inset, eIPSC traces correspond to the following conditions: control (a, 1 min), after the polarity shift (b, 10 min), and during furosemide application (c, 28 min). E, The Erev during control (left), after the polarity shift (middle, +10 min), and during furosemide application (right) was determined by stepping the holding potential from −113 up to −33 mV (increment of 20 mV) during whole-cell recordings. The dashed lines approximately indicate the Erev. F, Average Erev was 11.8 mV more hyperpolarized than control after the polarity shift (n = 5, blue) and returned to near control levels during furosemide application (n = 5, orange). For the four of five cells that were affected by furosemide, the Erev after the polarity shift was significantly different from control and furosemide application (p = 0.036, RM-ANOVA with post hoc Bonferroni test). G, Schematic of experimental setup for gramicidin-perforated patch recording, highlighting the gramicidin-containing internal solution (with high Cl− concentration, 145 mm) and native Cl− concentration within the LLDp neuron. H, I, With gramicidin-perforated patch recording, sIPSCs were recorded under different membrane holding potentials (−125 up to −45 mV, increment of 20 mV) to determine Erev. Individual and averaged sIPSCs are shown to the right. J, The calculated Erev for sIPSCs (n = 5) was 21.6 mV more negative than the average RMP (p = 0.022, paired t test). K, The calculated Cl− concentration was relatively low (n = 5). ctrl, Control.
Contralateral synaptic inhibition reduces firing in LLDp neurons. A, Schematic of the experimental setup, highlighting the electrical stimulation of the contralateral LLDp (right) during whole-cell current-clamp recordings. B, Schematic of the current-clamp protocol used to evaluate the effects of contralateral inhibition on spiking activity. A depolarizing current step (200 ms, 100 pA above threshold) was injected into the cell body to evoke APs. This was followed by or overlapped with a contralateral electrical stimulation (100 Hz, 200 ms, 20 pulses) for the baseline condition and the experimental condition, respectively. E+I, With overlapping excitation and inhibition. The intensity of the electrical stimulation was increased in a stepwise fashion to elicit 0 to maximal eIPSP. C, Example of depolarizing eIPSPs (left, blue) and hyperpolarizing eIPSPs (right, red) in response to the contralateral stimulation (top, single-pulse; bottom, 100 Hz stimulation). D, The contralateral stimulation decreased the number of APs (top, right) compared with baseline (top, left). Bath application of the inhibition blockers gabazine (10 μm) and strychnine (1 μm; bottom, right) eliminated the effect. Stimulus artifacts from contralateral stimulation are truncated for clarity and shown in black. E, The normalized spike probability was sensitive to both hyperpolarizing and depolarizing eIPSPs during increasing contralateral stimulation levels. Normalized spike probability was significantly reduced during mid, high, and max normalized eIPSP amplitude levels, compared with control levels (n = 11, p < 0.001, two-way ANOVA with post hoc Bonferroni test). Normalized spike probability was significantly lower for hyperpolarizing inhibition (n = 3) than depolarizing inhibition (n = 8) at high and max normalized eIPSP amplitudes (†p = 0.002). F, Normalized spike probability plotted against eIPSP amplitude of both polarities. The horizontal dashed line indicates a 50% reduction of APs, and the vertical dashed line at 0 mV separates the hyperpolarizing (negative) and depolarizing (positive) eIPSP amplitudes. Contra Stim, Contralateral stimulation.
LLDp neurons have diverse cell morphology and are GAD65/67 positive. A, Biocytin-filled LLDp neurons have diverse morphology. Most cells have large somas but exhibit different dendritic branching patterns. B–E, Serial sections (50 µm in thickness) of a Nissl stain through the LLD (from left to right: caudal to rostral). LLDp neurons appear more densely distributed at caudal levels, with the number of cells becoming sparser in rostral sections. A magnified view of an oligodendrocyte (arrow) is shown in the inset (C). F–I, Serial sections of a GAD65/67 stain through the LLD (left to right: caudal to rostral). Corresponding magnified view is shown below, highlighting the punctate staining of GAD-positive terminals (f–i). GAD65/67 staining highlights the segregation of the anterior and posterior LLDs, with denser staining on the posterior portion associated with ILD-coding neurons in chicken. At rostral levels, dorsal cells exhibit weaker GAD staining than ventral cells (I). d, dorsal; m, medial.
Avian ILD circuit and theoretical models for ILD coding in LLDp. A. LLDp neurons receive excitatory input from the contralateral cochlear NA and inhibitory input from the contralateral LLDp, and may also receive inhibitory input from the SON, which is driven by excitatory inputs from the ipsilateral NA. LLDp neurons predominantly send inhibitory projections contralaterally to the mesencephalicus lateralis, pars dorsalis (MLd), the auditory midbrain (Wild et al., 2010). B–D, Models of the origin of reciprocal inhibition for ILD coding in the LLDp. The reciprocal inhibition could arise from a single cell type that both encodes ILD and inhibits the contralateral LLDp (B); two cell types, one of which encodes the ILD (black circle) and another cell type that is specialized for fast synaptic transmission and provides the reciprocal inhibition (red triangles; C); or local interneurons that convert excitation into inhibition (D). In all three hypothetical circuits, a dorsal–ventral gradient of inhibition, but not excitation, could provide a topographic readout of ILD to create a space map of sound location in the MLd. Hypothetical schematics of the timing of synaptic inputs (middle column) for EPSPs (blue) and IPSPs (red) show the timing and amplitude for each model when ipsilateral sounds are louder (a) and when contralateral sounds are louder (b). Hypothetical ILD curves (right column) based on the respective model circuit and relative timing of excitation and inhibition. The model in panel C (two cell types) may offer a full-range dynamic coding of ILD (for details, see Discussion). VIII n., 8th nerve; contra, contralateral; ipsi, ipsilateral.