Characterization of Anxiety-Like Behaviors and Neural Circuitry following Chronic Moderate Noise Exposure in Mice

Abstract

Background: Commonly encountered nontraumatic, moderate noise is increasingly implicated in anxiety; however, the neural substrates underlying this process remain unclear.

Objectives: We investigated the neural circuit mechanism through which chronic exposure to moderate-level noise causes anxiety-like behaviors.

Methods: Mice were exposed to chronic, moderate white noise [85 decibel (dB) sound pressure level (SPL)], 4 h/d for 4 wk to induce anxiety-like behaviors, which were assessed by open field, elevated plus maze, light-dark box, and social interaction tests. Viral tracing, immunofluorescence confocal imaging, and brain slice patch-clamp recordings were used to characterize projections from auditory brain regions to the lateral amygdala. Neuronal activities were characterized by in vivo multielectrode and fiber photometry recordings in awake mice. Optogenetics and chemogenetics were used to manipulate specific neural circuitry.

Results: Mice chronically (4 wk) exposed to moderate noise (85 dB SPL, 4 h/d) demonstrated greater neuronal activity in the lateral amygdala (LA), and the LA played a critical role in noise-induced anxiety-like behavior in these model mice. Viral tracing showed that the LA received monosynaptic projections from the medial geniculate body (MG) and auditory cortex (ACx). Optogenetic excitation of the $\text{MG}\to \text{LA}$ or $\text{ACx}\to \text{LA}$ circuits acutely evoked anxiety-like behaviors, whereas their chemogenetic inactivation abolished noise-induced anxiety-like behavior. Moreover, mice chronically exposed to moderate noise were more susceptible to acute stress, with more neuronal firing in the LA, even after noise withdrawal.

Discussion: Mice exposed to 4 wk of moderate noise (85 dB SPL, 4 h/d) demonstrated behavioral and physiological differences compared to controls. The neural circuit mechanisms involved greater excitation from glutamatergic neurons of the MG and ACx to LA neurons under chronic, moderate noise exposure, which ultimately promoted anxiety-like behaviors. Our findings support the hypothesis that nontraumatic noise pollution is a potentially serious but unrecognized public health concern. https://doi.org/10.1289/EHP12532.

Figure 1.

Experimental overview, auditory brainstem response and behavioral outcomes in mice exposed to white noise for 4 wk. The mice were exposed to 85 dB SPL white noise for 4 wk (4 h/d), and auditory brainstem responses were used to determine the hearing sensitivity, and OFT and EPM tests were used to examine anxiety-like behaviors. (A) Timeline for noise exposure and behavioral tests (top) and schematic for noise exposure in a sound-proof chamber (bottom). (B) Representative waveforms for ABR evoked by click stimuli at different sound pressure levels in 10 dB steps. (C) Summarized data for click-evoked ABR thresholds ($U=17$, $p=0.9740$). (D) The representative heatmaps for moving trajectory in OFT (top) and EPM (bottom). (E and F) Summarized data for time spent in the center of the OFT (E, 1W, $t_{(14)}=0.5916$, $p=0.5635$; 2W, $U=20$, $p=0.3132$; 3W, $U=5$, $p=0.0047$; 4W, $t_{(14)}=12.36$, $p<0.0001$), and in the open arms of the EPM (F, 1W, $t_{(11)}=0.3527$, $p=0.731$; 2W, $U=10$, $p=0.1375$; 3W, $t_{(11)}=3.185$, $p=0.0087$; 4W, $t_{(11)}=6.040$, $p<0.0001$) at different time points of noise exposure. (G) Representative path tracks in OFT (top) and EPM (bottom). (H and I) Summarized data for the total travel distance in the OFT (H, 1W, $t_{(14)}=0.3454$, $p=0.7350$; 2W, $t_{(14)}=0.2046$, $p=0.8408$; 3W, $t_{(14)}=0.3555$, $p=0.7275$; 4W, $t_{(14)}=0.1598$, $p=0.8753$), and the EPM (I, 1W, $t_{(11)}=0.3425$, $p=0.7384$; 2W, $t_{(11)}=0.4941$, $p=0.6310$; 3W, $t_{(11)}=0.7119$, $p=0.4913$; 4W, $t_{(11)}=0.3190$, $p=0.7557$) at different time points during the four treatment period. (J and K) Representative heatmaps of track paths in light–dark box test (J) and summarized data for time spent in the dark side (K, $t_{(17)}=14.38$, $p<0.0001$). (L and M) Representative heatmaps of track paths in the social interaction tests (L) and summarized data for interaction time with stranger mice (M, $t_{(17)}=5.775$, $p<0.0001$). The data are expressed as the $\text{mean}\pm \text{SEM}$. **$p<0.01$; ***$p<0.001$. Two-way ANOVA analysis for (H) and (I). Unpaired Student’s t-test for 1W and 4 W in (E), (F), (H), (I), (K) and (M). Mann-Whitney U-test for (C), 2W and 3W in (E), and (F). For (C) $n=6$ control mice, 6 noise mice. For (E) $n=6$ control mice, 10 noise mice. For (F) $n=6$ control mice, 7 noise mice. For (H) $n=6$ control mice, 10 noise mice. For (I) $n=6$ control mice, 7 noise mice. For (K) $n=9$ control mice, 10 noise mice. For (M) $n=9$ control mice, 10 noise mice. The numerical data underlying this figure are shown in Excel Tables S1–S2. Note: ABR, auditory brainstem response; ANOVA, analysis of variance; dB, decibel; EPM, elevated plus maze; NS, not significant; OFT, open-field tests; SEM, standard error of the mean; SPL, sound pressure level.

Figure 1.

Experimental overview, auditory brainstem response and behavioral outcomes in mice exposed to white noise for 4 wk. The mice were exposed to 85 dB SPL white noise for 4 wk (4 h/d), and auditory brainstem responses were used to determine the hearing sensitivity, and OFT and EPM tests were used to examine anxiety-like behaviors. (A) Timeline for noise exposure and behavioral tests (top) and schematic for noise exposure in a sound-proof chamber (bottom). (B) Representative waveforms for ABR evoked by click stimuli at different sound pressure levels in 10 dB steps. (C) Summarized data for click-evoked ABR thresholds ($U=17$, $p=0.9740$). (D) The representative heatmaps for moving trajectory in OFT (top) and EPM (bottom). (E and F) Summarized data for time spent in the center of the OFT (E, 1W, $t_{(14)}=0.5916$, $p=0.5635$; 2W, $U=20$, $p=0.3132$; 3W, $U=5$, $p=0.0047$; 4W, $t_{(14)}=12.36$, $p<0.0001$), and in the open arms of the EPM (F, 1W, $t_{(11)}=0.3527$, $p=0.731$; 2W, $U=10$, $p=0.1375$; 3W, $t_{(11)}=3.185$, $p=0.0087$; 4W, $t_{(11)}=6.040$, $p<0.0001$) at different time points of noise exposure. (G) Representative path tracks in OFT (top) and EPM (bottom). (H and I) Summarized data for the total travel distance in the OFT (H, 1W, $t_{(14)}=0.3454$, $p=0.7350$; 2W, $t_{(14)}=0.2046$, $p=0.8408$; 3W, $t_{(14)}=0.3555$, $p=0.7275$; 4W, $t_{(14)}=0.1598$, $p=0.8753$), and the EPM (I, 1W, $t_{(11)}=0.3425$, $p=0.7384$; 2W, $t_{(11)}=0.4941$, $p=0.6310$; 3W, $t_{(11)}=0.7119$, $p=0.4913$; 4W, $t_{(11)}=0.3190$, $p=0.7557$) at different time points during the four treatment period. (J and K) Representative heatmaps of track paths in light–dark box test (J) and summarized data for time spent in the dark side (K, $t_{(17)}=14.38$, $p<0.0001$). (L and M) Representative heatmaps of track paths in the social interaction tests (L) and summarized data for interaction time with stranger mice (M, $t_{(17)}=5.775$, $p<0.0001$). The data are expressed as the $\text{mean}\pm \text{SEM}$. **$p<0.01$; ***$p<0.001$. Two-way ANOVA analysis for (H) and (I). Unpaired Student’s t-test for 1W and 4 W in (E), (F), (H), (I), (K) and (M). Mann-Whitney U-test for (C), 2W and 3W in (E), and (F). For (C) $n=6$ control mice, 6 noise mice. For (E) $n=6$ control mice, 10 noise mice. For (F) $n=6$ control mice, 7 noise mice. For (H) $n=6$ control mice, 10 noise mice. For (I) $n=6$ control mice, 7 noise mice. For (K) $n=9$ control mice, 10 noise mice. For (M) $n=9$ control mice, 10 noise mice. The numerical data underlying this figure are shown in Excel Tables S1–S2. Note: ABR, auditory brainstem response; ANOVA, analysis of variance; dB, decibel; EPM, elevated plus maze; NS, not significant; OFT, open-field tests; SEM, standard error of the mean; SPL, sound pressure level.

Figure 2.

Calcium signals, spontaneous firing, synaptic transmission in the LA of noise-exposed mice, and behavioral effects of LA chemical inactivation. In vivo fiber photometry, silicon probe extracellular recordings, and in vitro brain slice patch-clamp recordings were used to investigate the LA neuronal activity and synaptic transmission. Observations of elevated excitability in the LA led us to investigate whether blocking LA activity could abolish noise-evoked anxiety-like behaviors. (A) Schematic for fiber photometry recording in the LA. (B) Schematic (top) and typical image (bottom) of AAV-GCaMP6m viral injection and optical fiber implantation in the LA. Scale bars: $200\mathrm{\mu }\mathrm{m}$. (C) Heatmap of ${\text{LA}}^{\text{Glu}}$ GCaMP6m calcium signals evoked by white noise (100-ms duration, 85 dB SPL). (D) Averaged traces for $\mathrm{\Delta }\mathrm{F}/\mathrm{F}$ of ${\text{LA}}^{\text{Glu}}$ GCaMP6m signals. (E) Schematic for extracellular recordings using a silicon probe in head-fixed mice. (F) Representative voltage traces (top) and summarized data (bottom) of spontaneous firings recorded in the LA ($U=165$, $p=0.0004$). (G) Sample current traces of mEPSCs recorded in LA neurons of control and noise-treated mice. (H and I) Summarized data for the frequency (H, $t_{(26)}=3.565$, $p=0.0014$) and amplitude (I, $t_{(26)}=1.084$, $p=0.2882$) of the mEPSCs. (J) Sample current traces of mIPSCs recorded in LA neurons of control and noise-treated mice. (K and L) Summarized data for the frequency (K, $t_{(16)}=2.756$, $p=0.0141$) and amplitude (L, $t_{(16)}=2.517$, $p=0.0229$) of the mIPSCs. (M) Representative image of a coronal brain slice showing cannula tracks above the LA (top) and the timeline for noise exposure, chemical inactivation, and behavioral tests. Scale bars: $500\mathrm{\mu }\mathrm{m}$. (N and O) Typical voltage traces recorded in the LA following local application of muscimol or saline (N) and summarized data for spontaneous firing rates in naive mice (O, saline vs. muscimol, $U=4$, $p<0.0001$; Baseline vs. saline, $t_{(19)}=0.1078$, $p=0.9153$). (P–R) Summarized data for time spent in center (P, $t_{(16)}=2.701$, $p=0.0158$) and total travel distance (Q, $t_{(16)}=0.0002$, $p=0.9999$) in the OFT, and time in open arms of EPM (R, $t_{(18)}=3.168$, $p=0.0053$) in noise-exposed mice treated with muscimol or saline. The data are expressed as the $\text{mean}\pm \text{SEM}$. *$p<0.05$; **$P<0.01$; ***$P<0.001$. Unpaired t-test for (H), (I), (K), (L); Baseline vs. saline in (O), (P), (Q), and (R). Mann-Whitney U-test for (F) and saline vs. muscimol in (O). For (F) $n=23$ control cells, 32 noise cells. For (H) $n=14$ control cell, 14 noise cells. For (I) $n=14$ control cells, 14 noise cells. For (K) $n=9$ control cells, 9 noise cells; For (L) $n=9$ control cells, 9 noise cells. For (O) $n=10$ baseline cells, 11 saline cells, 11 muscimol cells. For (P) $n=9$ saline mice, 9 muscimol mice. For (Q) $n=9$ saline mice, 9 muscimol mice. For (R) $n=10$ saline mice, 10 muscimol mice. The numerical data underlying this figure are shown in Excel Tables S1–S3. Note: AAV, adeno-associated virus; ANOVA, analysis of variance; CaMKII, calcium-calmodulin (CaM)-dependent protein kinase II; dB SPL, decibel sound pressure level; EPM, elevated plus maze; GCaMP6m, green fluorescent protein-based genetically encoded calcium indicator; LA, lateral amygdala; ${\text{LA}}^{\text{Glu}}$, lateral amygdala glutamatergic neurons; mEPSC, miniature excitatory postsynaptic currents; mIPSCs, miniature inhibitory postsynaptic currents; OFT, open-field tests; NS, not significant; SEM, standard error of the mean.

Figure 2.

Calcium signals, spontaneous firing, synaptic transmission in the LA of noise-exposed mice, and behavioral effects of LA chemical inactivation. In vivo fiber photometry, silicon probe extracellular recordings, and in vitro brain slice patch-clamp recordings were used to investigate the LA neuronal activity and synaptic transmission. Observations of elevated excitability in the LA led us to investigate whether blocking LA activity could abolish noise-evoked anxiety-like behaviors. (A) Schematic for fiber photometry recording in the LA. (B) Schematic (top) and typical image (bottom) of AAV-GCaMP6m viral injection and optical fiber implantation in the LA. Scale bars: $200\mathrm{\mu }\mathrm{m}$. (C) Heatmap of ${\text{LA}}^{\text{Glu}}$ GCaMP6m calcium signals evoked by white noise (100-ms duration, 85 dB SPL). (D) Averaged traces for $\mathrm{\Delta }\mathrm{F}/\mathrm{F}$ of ${\text{LA}}^{\text{Glu}}$ GCaMP6m signals. (E) Schematic for extracellular recordings using a silicon probe in head-fixed mice. (F) Representative voltage traces (top) and summarized data (bottom) of spontaneous firings recorded in the LA ($U=165$, $p=0.0004$). (G) Sample current traces of mEPSCs recorded in LA neurons of control and noise-treated mice. (H and I) Summarized data for the frequency (H, $t_{(26)}=3.565$, $p=0.0014$) and amplitude (I, $t_{(26)}=1.084$, $p=0.2882$) of the mEPSCs. (J) Sample current traces of mIPSCs recorded in LA neurons of control and noise-treated mice. (K and L) Summarized data for the frequency (K, $t_{(16)}=2.756$, $p=0.0141$) and amplitude (L, $t_{(16)}=2.517$, $p=0.0229$) of the mIPSCs. (M) Representative image of a coronal brain slice showing cannula tracks above the LA (top) and the timeline for noise exposure, chemical inactivation, and behavioral tests. Scale bars: $500\mathrm{\mu }\mathrm{m}$. (N and O) Typical voltage traces recorded in the LA following local application of muscimol or saline (N) and summarized data for spontaneous firing rates in naive mice (O, saline vs. muscimol, $U=4$, $p<0.0001$; Baseline vs. saline, $t_{(19)}=0.1078$, $p=0.9153$). (P–R) Summarized data for time spent in center (P, $t_{(16)}=2.701$, $p=0.0158$) and total travel distance (Q, $t_{(16)}=0.0002$, $p=0.9999$) in the OFT, and time in open arms of EPM (R, $t_{(18)}=3.168$, $p=0.0053$) in noise-exposed mice treated with muscimol or saline. The data are expressed as the $\text{mean}\pm \text{SEM}$. *$p<0.05$; **$P<0.01$; ***$P<0.001$. Unpaired t-test for (H), (I), (K), (L); Baseline vs. saline in (O), (P), (Q), and (R). Mann-Whitney U-test for (F) and saline vs. muscimol in (O). For (F) $n=23$ control cells, 32 noise cells. For (H) $n=14$ control cell, 14 noise cells. For (I) $n=14$ control cells, 14 noise cells. For (K) $n=9$ control cells, 9 noise cells; For (L) $n=9$ control cells, 9 noise cells. For (O) $n=10$ baseline cells, 11 saline cells, 11 muscimol cells. For (P) $n=9$ saline mice, 9 muscimol mice. For (Q) $n=9$ saline mice, 9 muscimol mice. For (R) $n=10$ saline mice, 10 muscimol mice. The numerical data underlying this figure are shown in Excel Tables S1–S3. Note: AAV, adeno-associated virus; ANOVA, analysis of variance; CaMKII, calcium-calmodulin (CaM)-dependent protein kinase II; dB SPL, decibel sound pressure level; EPM, elevated plus maze; GCaMP6m, green fluorescent protein-based genetically encoded calcium indicator; LA, lateral amygdala; ${\text{LA}}^{\text{Glu}}$, lateral amygdala glutamatergic neurons; mEPSC, miniature excitatory postsynaptic currents; mIPSCs, miniature inhibitory postsynaptic currents; OFT, open-field tests; NS, not significant; SEM, standard error of the mean.

Figure 3.

Electrophysiological recording and behavioral tests in mice with optogenetic or chemogenetic activation of the LA. To investigate the role of the LA in noise-evoked anxiety, neuronal activity in the LA was optogenetically or chemogenetically activated by direct injection of the LA with AAV-CaMKII-ChR2 or AAV-CaMKII-hM4Di viruses in C57 BL/6 mice. Brain slice recordings were used to validate the expression of each transfected virus, and anxiety-like behaviors were examined by OFT and EPM tests following optogenetic or chemogenetic manipulation of the LA. (A) Schematic for viral injection and optogenetics (top) and a typical image showing the optical fiber track above the mCherry-expressing LA (bottom). Scale bars: $200\mathrm{\mu }\mathrm{m}$. (B) Sample traces of action potentials evoked by $473\text{\hspace{0.17em}nm}$ light (blue bars) recorded in ChR2-positive LA neurons in acute brain slices. (C and D) Summarized data for time spent in the center (C, $\text{virus}\times \text{light}$ interaction, $F_{(\mathrm{1,21})}=36.12$, $p<0.0001$; main effect of light, $F_{(\mathrm{1,21})}=27.10$, $p<0.0001$) of the OFT and in the open arms of the EPM (D, $\text{virus}\times \text{light}$ interaction, $F_{(\mathrm{1,18})}=37.34$, $p<0.0001$; main effect of light, $F_{(\mathrm{1,18})}=27.60$, $p<0.0001$) before (pre) and during (light) light stimulation. (E) Schematic for bilateral viral injection for chemogenetic experiments. (F) Change in membrane potential in response to CNO perfusion ($\text{time}\times \text{virus}$ interaction, $F_{(\mathrm{10,66})}=3.913$, $p=0.0003$; main effect of virus, $F_{(\mathrm{1,66})}=116.1$, $p<0.0001$). (G and H) Summarized data for time spent in the center (G, $\text{virus}\times \text{drug}$ interaction, $F_{(\mathrm{1,20})}=8.741$, $p=0.0078$; main effect of drug, $F_{(\mathrm{1,20})}=4.657$, $p=0.0433$) of the OFT and in the open arms of the EPM (H, $\text{virus}\times \text{drug}$ interaction, $F_{(\mathrm{1,20})}=13.26$, $p=0.0016$; main effect of drug, $F_{(\mathrm{1,20})}=13.11$, $P=0.0017$) in mice treated with CNO or saline. The data are expressed as the $\text{mean}\pm \text{SEM}$. **$p<0.01$; ***$p<0.001$. Two-way ANOVA with Bonferroni post hoc analysis for (C), (D), (F), (G), and (H). For (C) $n=11$ mCherry mice, 12 ChR2 mice. For (D) $n=10$ mCherry mice, 10 ChR2 mice. For (F) $n=4$ mCherry cells, 4 hM3Dq cells. For (G) $n=6$ mCherry mice, 6 hM3Dq mice. For (H) $n=6$ mCherry mice, 6 hM3Dq mice. The numerical data underlying this figure are shown in Excel Tables S1 and S3. Note: AAV, adeno-associated virus; ANOVA, analysis of variance; ChR2, channelrhodopsin-2; CNO, clozapine-N-oxide; EPM, elevated plus maze; hM3Dq, human M3 muscarinic receptor; LA, lateral amygdala; OFT, open-field tests; NS, not significant; Pre, before light stimulation; SEM, standard error of the mean.

Figure 4.

Viral tracing and patch-clamp recording for morphological and functional characterization of auditory inputs to the LA. Retrograde viral tracing at the LA was used to identify auditory brain regions with LA-projecting neurons, and anterograde viral tracing at these auditory nuclei was performed to morphologically confirm the connections; functional synapses were examined using electrophysiological recordings and optogenetic stimulation in brain slices. (A) Schematic of viral injection for retrograde tracing. (B) Representative image of a successful viral injection in the LA. Scale bars: $500\mathrm{\mu }\mathrm{m}$. (C) Retrogradely traced neurons in the auditory thalamus. Scale bars: $500\mathrm{\mu }\mathrm{m}$. (D) Cell type identification of traced EGFP-positive MG neurons; most traced neurons were immunoreactive to antibody against glutamate (top) but none to antibody against GABA (bottom). Scale bars: $20\mathrm{\mu }\mathrm{m}$. (E) The percentage of glutamate and GABA neurons among the traced EGFP-positive MG neurons. (F) The retrogradely traced neurons in the temporal regions. Scale bars: $500\mathrm{\mu }\mathrm{m}$. (G) Cell type identification of traced EGFP-positive ACx neurons; most traced neurons were immunoreactive to the antibody for glutamate (top) but few to the antibody for GABA (bottom). Scale bars: $20\mathrm{\mu }\mathrm{m}$. (H) The percentage of glutamate-positive and GABA-positive neurons among the traced EGFP-positive ACx neurons. (I) Tracing strategy using a Cre-loxP system to dissect the $\text{ACx}\to \text{LA}$ circuit and an Flp-FRT system to dissect the $\text{MG}\to \text{LA}$ circuit. (J) Representative images of anterogradely traced LA neurons (green, EYFP) postsynaptic to MG projection fibers and those (red, mCherry) postsynaptic to ACx fibers. Scale bars: $100\mathrm{\mu }\mathrm{m}$. (K) Typical images showing predominant co-localization of traced fluorescent neurons with glutamate immunofluorescence signals but few co-localizations with GABA immunofluorescence signals in the LA. Scale bars: $20\mathrm{\mu }\mathrm{m}$. (L) The percentage of EYFP-positive cells or mCherry-positive cells containing glutamate or GABA in the LA. (M) Schematic for whole-cell recordings combined with optogenetics. (N) Sample traces of action potentials (top, current-clamp mode) and postsynaptic currents (bottom, voltage-clamp mode) evoked by a series of $473\text{\hspace{0.17em}nm}$ light (blue bars) stimuli of $\text{MG}\to \text{LA}$ fibers. (O) Schematic for whole-cell recordings combined with optogenetics. (P) Sample traces of action potentials (top, current-clamp mode) and postsynaptic currents (bottom, voltage-clamp mode) evoked by a series of $473\text{\hspace{0.17em}nm}$ light (blue bars) stimuli of $\text{ACx}\to \text{LA}$ fibers. (Q) Quantification of light-evoked EPSCs. (R) Sample current traces of light-evoked EPSCs recorded at a holding potential of $-70\text{\hspace{0.17em}mV}$ and of light-evoked IPSCs recorded at a holding potential of $0\text{\hspace{0.17em}mV}$ by illuminating $\text{MG}\to \text{LA}$ fibers (left), and quantification of the latency for light-evoked EPSCs and IPSCs (right, $t_{(18)}=19.03$, $p<0.0001$). The blue bar denotes $473\text{\hspace{0.17em}nm}$ light, and the arrowhead denotes the beginning of light-evoked postsynaptic currents. (S) Sample current traces of light-evoked EPSCs recorded at a holding potential of $-70\text{\hspace{0.17em}mV}$ and of light-evoked IPSCs recorded at a holding potential of $0\text{\hspace{0.17em}mV}$ (left) generated by illuminating $\text{ACx}\to \text{LA}$ projection fibers, and quantification of the latency of light-evoked EPSCs and IPSCs (right, $t_{(18)}=9.504$, $p<0.0001$). The blue bar denotes $473\text{\hspace{0.17em}nm}$ light, and the arrowhead denotes the beginning of light-evoked postsynaptic currents. (T) Schematic diagram showing delivery of auditory monosynaptic excitation and di-synaptic inhibition to the LA. The data are expressed as the $\text{mean}\pm \text{SEM}$. ***$p<0.001$. Unpaired Student’s t-test for (R) and (S). For (E) $n=5$ slices from 3 mice (glutamate), 6 slices from 3 mice (GABA). For (H) $n=10$ slices from 3 mice (glutamate), 9 slices from 3 mice (GABA). For (L) $n=6$ slices from 3 mice (${\text{EGFP}}^{+}$ and ${\text{glutamate}}^{+}$), 6 slices from 3 mice (${\text{EYFP}}^{+}$ and ${\text{GABA}}^{+}$), 6 slices from 3 mice (${\text{mCherry}}^{+}$ and ${\text{glutamate}}^{+}$), 6 slices from 3 mice (${\text{mCherry}}^{+}$ and ${\text{GABA}}^{+}$). For (Q) $n=14$ cells ($\text{MG}\to \text{LA}$), 14 cells ($\text{ACx}\to \text{LA}$). For (R) $n=10$ cells (EPSC), 10 cells (IPSC). For (S) $n=10$ cells (EPSC), 10 cells (IPSC). The numerical data underlying this figure are shown in Excel Tables S3–S4. Note: AAV, adeno-associated virus; ACSF, artificial cerebrospinal fluid; ACx, auditory cortex; AMPARs, alpha-Amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptors; Cre-loxP, cyclization recombinase- locus of X(cross)-over in P1; DAPI, 4′,6-diamidino-2-phenylindole; DNQX, 6,7-dinitroquinoxaline-2, 3-dione; EGFP, enhanced green fluorescent protein; EPM, elevated plus maze; EPSC, excitatory postsynaptic currents; EYFP, enhanced yellow fluorescent protein; Flp-FRT, flippase recombination enzyme-flp recognition target; GABA, gamma-aminobutyric acid; ${\text{GABA}}_{\mathrm{A}}\text{Rs}$, gamma-aminobutyric acid type A receptors; Glu, glutamate; IPSCs, inhibitory postsynaptic currents; LA, lateral amygdala; MG, medial geniculate body; SEM, standard error of the mean; VH, holding potential.

Figure 4.

Viral tracing and patch-clamp recording for morphological and functional characterization of auditory inputs to the LA. Retrograde viral tracing at the LA was used to identify auditory brain regions with LA-projecting neurons, and anterograde viral tracing at these auditory nuclei was performed to morphologically confirm the connections; functional synapses were examined using electrophysiological recordings and optogenetic stimulation in brain slices. (A) Schematic of viral injection for retrograde tracing. (B) Representative image of a successful viral injection in the LA. Scale bars: $500\mathrm{\mu }\mathrm{m}$. (C) Retrogradely traced neurons in the auditory thalamus. Scale bars: $500\mathrm{\mu }\mathrm{m}$. (D) Cell type identification of traced EGFP-positive MG neurons; most traced neurons were immunoreactive to antibody against glutamate (top) but none to antibody against GABA (bottom). Scale bars: $20\mathrm{\mu }\mathrm{m}$. (E) The percentage of glutamate and GABA neurons among the traced EGFP-positive MG neurons. (F) The retrogradely traced neurons in the temporal regions. Scale bars: $500\mathrm{\mu }\mathrm{m}$. (G) Cell type identification of traced EGFP-positive ACx neurons; most traced neurons were immunoreactive to the antibody for glutamate (top) but few to the antibody for GABA (bottom). Scale bars: $20\mathrm{\mu }\mathrm{m}$. (H) The percentage of glutamate-positive and GABA-positive neurons among the traced EGFP-positive ACx neurons. (I) Tracing strategy using a Cre-loxP system to dissect the $\text{ACx}\to \text{LA}$ circuit and an Flp-FRT system to dissect the $\text{MG}\to \text{LA}$ circuit. (J) Representative images of anterogradely traced LA neurons (green, EYFP) postsynaptic to MG projection fibers and those (red, mCherry) postsynaptic to ACx fibers. Scale bars: $100\mathrm{\mu }\mathrm{m}$. (K) Typical images showing predominant co-localization of traced fluorescent neurons with glutamate immunofluorescence signals but few co-localizations with GABA immunofluorescence signals in the LA. Scale bars: $20\mathrm{\mu }\mathrm{m}$. (L) The percentage of EYFP-positive cells or mCherry-positive cells containing glutamate or GABA in the LA. (M) Schematic for whole-cell recordings combined with optogenetics. (N) Sample traces of action potentials (top, current-clamp mode) and postsynaptic currents (bottom, voltage-clamp mode) evoked by a series of $473\text{\hspace{0.17em}nm}$ light (blue bars) stimuli of $\text{MG}\to \text{LA}$ fibers. (O) Schematic for whole-cell recordings combined with optogenetics. (P) Sample traces of action potentials (top, current-clamp mode) and postsynaptic currents (bottom, voltage-clamp mode) evoked by a series of $473\text{\hspace{0.17em}nm}$ light (blue bars) stimuli of $\text{ACx}\to \text{LA}$ fibers. (Q) Quantification of light-evoked EPSCs. (R) Sample current traces of light-evoked EPSCs recorded at a holding potential of $-70\text{\hspace{0.17em}mV}$ and of light-evoked IPSCs recorded at a holding potential of $0\text{\hspace{0.17em}mV}$ by illuminating $\text{MG}\to \text{LA}$ fibers (left), and quantification of the latency for light-evoked EPSCs and IPSCs (right, $t_{(18)}=19.03$, $p<0.0001$). The blue bar denotes $473\text{\hspace{0.17em}nm}$ light, and the arrowhead denotes the beginning of light-evoked postsynaptic currents. (S) Sample current traces of light-evoked EPSCs recorded at a holding potential of $-70\text{\hspace{0.17em}mV}$ and of light-evoked IPSCs recorded at a holding potential of $0\text{\hspace{0.17em}mV}$ (left) generated by illuminating $\text{ACx}\to \text{LA}$ projection fibers, and quantification of the latency of light-evoked EPSCs and IPSCs (right, $t_{(18)}=9.504$, $p<0.0001$). The blue bar denotes $473\text{\hspace{0.17em}nm}$ light, and the arrowhead denotes the beginning of light-evoked postsynaptic currents. (T) Schematic diagram showing delivery of auditory monosynaptic excitation and di-synaptic inhibition to the LA. The data are expressed as the $\text{mean}\pm \text{SEM}$. ***$p<0.001$. Unpaired Student’s t-test for (R) and (S). For (E) $n=5$ slices from 3 mice (glutamate), 6 slices from 3 mice (GABA). For (H) $n=10$ slices from 3 mice (glutamate), 9 slices from 3 mice (GABA). For (L) $n=6$ slices from 3 mice (${\text{EGFP}}^{+}$ and ${\text{glutamate}}^{+}$), 6 slices from 3 mice (${\text{EYFP}}^{+}$ and ${\text{GABA}}^{+}$), 6 slices from 3 mice (${\text{mCherry}}^{+}$ and ${\text{glutamate}}^{+}$), 6 slices from 3 mice (${\text{mCherry}}^{+}$ and ${\text{GABA}}^{+}$). For (Q) $n=14$ cells ($\text{MG}\to \text{LA}$), 14 cells ($\text{ACx}\to \text{LA}$). For (R) $n=10$ cells (EPSC), 10 cells (IPSC). For (S) $n=10$ cells (EPSC), 10 cells (IPSC). The numerical data underlying this figure are shown in Excel Tables S3–S4. Note: AAV, adeno-associated virus; ACSF, artificial cerebrospinal fluid; ACx, auditory cortex; AMPARs, alpha-Amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptors; Cre-loxP, cyclization recombinase- locus of X(cross)-over in P1; DAPI, 4′,6-diamidino-2-phenylindole; DNQX, 6,7-dinitroquinoxaline-2, 3-dione; EGFP, enhanced green fluorescent protein; EPM, elevated plus maze; EPSC, excitatory postsynaptic currents; EYFP, enhanced yellow fluorescent protein; Flp-FRT, flippase recombination enzyme-flp recognition target; GABA, gamma-aminobutyric acid; ${\text{GABA}}_{\mathrm{A}}\text{Rs}$, gamma-aminobutyric acid type A receptors; Glu, glutamate; IPSCs, inhibitory postsynaptic currents; LA, lateral amygdala; MG, medial geniculate body; SEM, standard error of the mean; VH, holding potential.

Figure 5.

Behavioral effects of optogenetic activation or chemogenetic inactivation of auditory inputs to the LA. Optogenetic activation was used to selectively excite auditory inputs to the LA to determine whether activating these inputs could mimic noise-evoked anxiety-like behaviors. Chemogenetic blocking was used to selectively silence auditory inputs to the LA during noise exposure to determine whether they were necessary for noise-evoked anxiety-like behavior. (A) Schematic for viral injection and optogenetics. (B) Typical images of the viral injection in the MG (left) and optical fiber track above the LA with ChR2-expressing MG fibers (right). Scale bars: $200\mathrm{\mu }\mathrm{m}$. (C and D) Summarized data for time spent in the center of OFT (C, $\text{virus}\times \text{light}$ interaction, $F_{(\mathrm{1,20})}=52.22$, $p<0.0001$; main effect of light, $F_{(\mathrm{1,20})}=32.65$, $p<0.0001$) and time in the open arms of the EPM (D, $\text{virus}\times \text{light}$ interaction, $F_{(\mathrm{1,20})}=25.67$, $p<0.0001$; main effect of light, $F_{(\mathrm{1,20})}=23.41$, $p<0.0001$) before (pre) and during (light) light stimulation of $\text{MG}\to \text{LA}$ fibers. (E) Schematic for viral injection and optogenetics. (F) Typical images of viral injection in the ACx (left) and optical fiber track above the LA with ChR2-experssing ACx fibers (right). Scale bars: $200\mathrm{\mu }\mathrm{m}$. (G and H) Summarized data for time spent in the center of OFT (G, $\text{virus}\times \text{light}$ interaction, $F_{(\mathrm{1,22})}=64.09$, $p<0.0001$; main effect of light, $F_{(\mathrm{1,22})}=36.02$, $p<0.0001$) and time in the open arms of the EPM (H, $\text{virus}\times \text{light}$ interaction, $F_{(\mathrm{1,18})}=27.47$, $p<0.0001$; main effect of light, $F_{(\mathrm{1,18})}=32.77$, $p<0.0001$) before (pre) and during (light) light stimulation of $\text{ACx}\to \text{LA}$ fibers. (I) Timeline for chemogenetic experiments. (J–L) Summarized data for time spent in the center (J, $t_{(14)}=8.097$, $p<0.0001$) and total distance traveled (K, $t_{(14)}=0.7899$, $p=0.4427$) in the OFT and time in the open arms of EPM (L, $t_{(14)}=4.386$, $p=0.0006$). The data are expressed as the $\text{mean}\pm \text{SEM}$. ***$p<0.001$. Two-way ANOVA with Bonferroni post hoc analysis for (C), (D), (G), and (H). Unpaired t-test for (J), (K), and (L). For (C) $n=10$ mCherry mice, 12 ChR2 mice. For (D) $n=10$ mCherry mice, 12 ChR2 mice. For (G) $n=12$ mCherry mice, 12 ChR2 mice. For (H) $n=10$ mCherry mice, 10 ChR2 mice. For (J) $n=8$ mCherry mice, 8 hM4Di mice. For (K) $n=8$ mCherry mice, 8 hM4Di mice. For (L) $n=8$ mCherry mice, 8 hM4Di mice. The numerical data underlying this figure are shown in Excel Tables S1. Note: AAV, adeno-associated virus; ACx, auditory cortex; ANOVA, analysis of variance; CaM, calcium-calmodulin; CaMKII, CaM-dependent protein kinase II; ChR2, channelrhodopsin-2; CNO, clozapine-N-oxide; EPM, elevated plus maze; hM4, human M4 muscarinic; hM4Di, human M4 muscarinic receptor; LA, lateral amygdala; MG, medial geniculate body; NS, not significant; OFT, open-field tests; Pre, before light stimulation; SEM, standard error of the mean.

Figure 6.

Experimental overview, behavior, and electrophysiological effects of acute noise exposure in mice after 4 wk of noise withdrawal. All chronic noise-exposed mice or control were subjected to 4-wk noise withdrawal, and then their behaviors were tested before and after 2-h noise exposure to examine their susceptibility to acute noise exposure. Anxiety-like behaviors and elevated spontaneous neuronal firing in the LA were observed. (A) Timeline of the experimental protocol. (B and C) Summarized data for the time in the center of OFT (B, 8W, $U=17$, $p=0.1806$; 2-h 85 dB SPL, $U=0$, $p=0.0002$) and time in the open arms of EPM (C, $\text{time}\times \text{noise}$ interaction, $F_{(\mathrm{1,14})}=10.47$, $p=0.0060$; main effect of noise, $F_{(\mathrm{1,14})}=10.99$, $p=0.0051$) at the indicated time points. (D and E) Representative voltage traces of spontaneous firings recorded in the LA (D) and summarized data (E, $\text{time}\times \text{noise}$ interaction, $F_{(\mathrm{1,64})}=4.401$, $p=0.0399$; main effect of noise, $F_{(\mathrm{1,64})}=6.415$, $p=0.0138$). (F and G) Summarized data for time spent in the center of the OFT (F, $t_{(10)}=4.015$, $p=0.0025$), and time in the open arms of the EPM (G, $t_{(10)}=2.890$, $p=0.0161$) in recovered mice reexposed to 2-h 75 dB SPL noise. The data are expressed as the $\text{mean}\pm \text{SEM}$. *$p<0.05$; **$p<0.01$; ***$p<0.001$. Mann-Whitney U-test for (B). Two-way ANOVA with Bonferroni post hoc analysis for (C) and (E). Unpaired Student’s t-test for (F) and (G). For (B) $n=6$ control mice, 10 noise mice. For (C) $n=6$ control mice, 10 noise mice. For (E) $n=16$ cells (8W control), 17 cells (8W noise), 15 cells (2h 85 dB control), 20 cells (2h 85 dB noise). For (F) $n=5$ control mice, 7 noise mice. For (G) $n=5$ control mice, 7 noise mice. The numerical data underlying this figure are shown in Excel Tables S1–2. Note: 8W, eighth week; ANOVA, analysis of variance; dB SPL, decibel sound pressure level; EPM, elevated plus maze; LA, lateral amygdala; OFT, open-field tests; SEM, standard error of the mean.