Sensory overamplification in layer 5 auditory corticofugal projection neurons following cochlear nerve synaptic damage

Abstract

Layer 5 (L5) cortical projection neurons innervate far-ranging brain areas to coordinate integrative sensory processing and adaptive behaviors. Here, we characterize a plasticity in L5 auditory cortex (ACtx) neurons that innervate the inferior colliculus (IC), thalamus, lateral amygdala and striatum. We track daily changes in sound processing using chronic widefield calcium imaging of L5 axon terminals on the dorsal cap of the IC in awake, adult mice. Sound level growth functions at the level of the auditory nerve and corticocollicular axon terminals are both strongly depressed hours after noise-induced damage of cochlear afferent synapses. Corticocollicular response gain rebounded above baseline levels by the following day and remained elevated for several weeks despite a persistent reduction in auditory nerve input. Sustained potentiation of excitatory ACtx projection neurons that innervate multiple limbic and subcortical auditory centers may underlie hyperexcitability and aberrant functional coupling of distributed brain networks in tinnitus and hyperacusis.

Fig. 1

Auditory corticofugal neurons that innervate the inferior colliculus have other widespread targets throughout the forebrain. a A canine adenovirus vector with efficient retrograde transport (CAV2) was injected into the IC to express cre-recombinase in neurons that project into the injection zone. A cre-dependent AAV was then injected into the ipsilateral ACtx to express a fluorescent marker throughout the entire axon field of CCol neurons. Photomicrographs show the expected labeling of layer 5 ACtx neurons and their IC axon terminals, with additional strong axon labeling in the dorsal nucleus of the medial geniculate body, lateral amygdala and posterior regions of the dorsal striatum. wm white matter. b Schematic of virus strategy used for in vivo Ca²⁺ imaging in corticofugal axons. EC and CN are external cortex and central nucleus of the IC, respectively. MGB subdivisions d, v and m are dorsal, ventral and medial, respectively. LA and BLA are lateral and basolateral amygdala, respectively. S is striatum. c Strong labeling of L5 pyramidal neuron cell bodies, apical dendrites and CCol axon terminals are observed approximately 5 weeks after injection of the GCaMP6s virus in ACtx. All scale bars = 0.1 mm

Fig. 2

Sound-evoked corticocollicular axon response increases monotonically with sound level and remains stable over 1 week of imaging. a A chronic preparation for epifluorescence imaging of GCaMP6s in CCol axons via a cranial window (scale bars = 0.5 mm) in awake, head-fixed mice. Red rectangle denotes region of interest for CCol imaging. L is lateral and C is caudal. Mouse schematics in Figs. 2a, 3a, and 4a are adapted with permission from Aronoff et al., 2010. All rights reserved. These images are not covered under the CC BY license for this article. Scientific data and content are original. b Time course of mean fractional change in the CCol response amplitude evoked by a 50 ms white noise burst from a single imaging session. Gray box denotes stimulus timing and duration. c The monotonic growth of CCol peak response amplitude falls off steeply when the region of interest is shifted away from the IC. Data represent mean ± SEM. d Top: CCol response growth functions from a single mouse across seven daily imaging sessions. Data represent mean ± SEM. Bottom: Scatterplots depict the mean CCol response amplitude (x-axis) at each sound level measured from the first two imaging sessions (defined as baseline) against the CCol response amplitude (y-axis) measured on the day specified. The slope (m) of the linear fit provides an estimate of daily changes in response gain, where m = 1 indicates a matched response growth relative to baseline, m < 1 indicates a divisive flattening of the growth function and m > 1 indicates a multiplicative enhancement relative to baseline. Shading represents the 95% confidence interval of the fit. e As per d, averaged across all control mice (n = 5)

Fig. 3

Moderate intensity noise exposure induces a temporary shift in cochlear and brainstem response thresholds but a permanent loss of auditory nerve afferent fibers. a Schematic of noise exposure and auditory brainstem response (ABR) measurement protocols. Example ABR waveforms evoked with a 32 kHz tone bursts before, 24 h after and 2 weeks after noise exposure. b, c Elevations in ABR and distortion product otoacoustic emission (DPOAE) thresholds (b and c, respectively) are observed 1 day following noise exposure (orange) but have returned to baseline 2 weeks following noise exposure (red). d ABR wave 1 (w1) growth functions. NS indicates no significant difference with pre-exposure. Asterisk indicates significant main effect for ABR amplitude between pre-exposure and post-exposure. Data represent mean ± SEM, n = 10 mice in pre-exposure and 24 h post conditions, n = 8 mice for 2 weeks post. e, f Schematic (e) and actual (f) visualizations of cochlear nerve afferent synapses on inner hair cells. Red (open) and green (closed) arrowheads depict orphaned presynaptic ribbons and postsynaptic GluA2 receptor patches, respectively. Combined red (open) and green (closed) arrowhead identifies primary afferent cochlear synapses as appositions of the CtBP2 and GluA2 pre- and postsynaptic markers, respectively. Dashed white line depicts the boundary of a single inner hair cell. g Quantification of cochlear afferent synapses in control mice, 24 h and 2 weeks following noise exposure. Synaptic counts are expressed as percent survival by comparison to normative standards from age- and strain-matched mice^,. Asterisk indicates significant difference with an unpaired t-test after correcting for multiple comparisons. Synaptic counts were made from 20.77 – 21.73 individual inner hair cells at a fixed position in the cochlear frequency map between the 20 and 30 kHz region in each ear, 3 ears per group and 2 cochlear sections per ear

Fig. 4

Opposing changes in auditory nerve and corticocollicular response growth functions following cochlear synaptopathy. a Auditory nerve growth functions were measured under anesthesia every other day according to the change in ABR wave 1 (blue circle) amplitude to white noise bursts of varying level (n = 6). CCol response growth functions were measured daily in a separate cohort of awake mice (n = 10) also using white noise bursts, per previous figures. b ABR wave 1 and CCol responses were both measured for 2 days (d) prior to moderate noise exposure and for 7 days following noise exposure. In a subset of noise-exposed mice (n = 5), CCol imaging was extended for an additional week after noise exposure. c, d As per Fig. 2e, ABR wave 1 (c) and CCol response (d) growth functions (top rows) and scatterplots of linear fits for baseline vs post-exposure growth functions (bottom rows) are provided for all mice. Data represent mean ± SEM. Linear fits of the five highest sound levels are illustrated by the solid black line with corresponding slope (m) and 95% confidence interval (blue and red shading)

Fig. 5

ABR threshold recovery belies ongoing dynamics in auditory nerve and corticocollicular response gain. a Moderate noise exposure induces a temporary shift in the ABR wave 1 threshold to white noise bursts that resolved after 2 days. b Daily changes in response gain for CCol measurements in unexposed control (gray, n = 5) and noise-exposed (red, n = 10) mice are contrasted with daily changes in the response gain of ABR wave 1 in noise-exposed mice (blue, n = 6). In all cases, gain is calculated as the slope of the fit line applied to sound level growth functions measured during baseline and subsequent days. c Daily changes in CCol response gain over an extended 2-week imaging period in a subset of noise-exposed mice (n = 5). For a–c, data represent mean ± SEM. d Gain estimates from individual imaging sessions in unexposed control mice (gray) are contrasted with gain estimates measured during the first imaging session following noise exposure (hours after), or during 3-day epochs occurring on D2–D4, D7–D9 or D12–D14. Thick horizontal bars represent sample means. Individual circles represent all individual data points. Asterisks and NS denote statistically significant differences and lack thereof, respectively, for pairwise comparisons indicated by thin horizontal lines after correcting for multiple comparisons