Sensory striatum is permanently impaired by transient developmental deprivation

Summary

Corticostriatal circuits play a fundamental role in regulating many behaviors, and their dysfunction is associated with many neurological disorders. In contrast, sensory disorders, like hearing loss (HL), are commonly linked with processing deficits at, or below, the level of auditory cortex (ACx). However, HL can be accompanied by non-sensory deficits such as learning delays, suggesting the involvement of regions downstream of ACx. Here, we show that transient developmental HL differentially affected ACx and its downstream target, sensory striatum. Following HL, both juvenile ACx layer 5 and striatal neurons displayed an excitatory- inhibitory imbalance and lower firing rates. After hearing was restored, adult ACx neurons recovered balanced excitatory-inhibitory synaptic gain and control the firing rates, but striatal neuron synapses and firing properties did not recover. Thus, a brief period of abnormal cortical activity may induce striatal impairments that persist into adulthood, and contribute to neurological disorders that are striatal in origin.

eToc Blurb

In many neurocognitive disorders, striatal deficits emerge slowly and correlate with the onset of neurological symptoms. Mowery et al. demonstrate that early abnormal sensory experience leads to a persistent change in striatal function, despite the recovery of normal cortical output.

Introduction

Sensory cortex development can be profoundly affected by genetic and congenital disorders, and this is often accompanied by perceptual deficits. However, the secondary manifestations of cognitive impairments generally emerge slowly during childhood. One explanation for the gradual appearance of cognitive symptoms is that cellular dysfunction propagates from sensory cortex to downstream regions that mediate cognitive skills. For example, changes to cortical physiology exist pre-symptomatically in a model of Parkinson’s, but behavioral symptoms emerge only after striatal neurons succumb to increased cortical activity (Cepeda et al., 2007; Estrada-Sanchez and Rebec 2013; Fusco et al., 1999). Here, we asked whether a brief period of developmental auditory deprivation propagates from primary sensory cortex to dorsal striatum, yielding permanently altered cellular properties that could contribute to reported auditory learning delays.

Developmental HL imposes a risk for deficiencies in cognitive skills (Adesman et al., 1990; Burkholder-Juhasz et al., 2007; Watson et al., 2007; Pisoni et al., 2011; Beer et al., 2014; Kronenberger et al., 2014; von Trapp et al., 2016). Even a transient period of developmental HL diminishes perceptual skills long after the deprivation has ended (Hall et al., 1995; Caras and Sanes, 2015). Since many neurological disorders are associated with corticostriatal dysfunction (Shepherd, 2013), HL-induced changes to this circuit could plausibly account for some non- sensory deficits associated with deprivation. Therefore, HL-induced changes to the corticostriatal circuit during developmental HL could plausibly account for some long-term cognitive deficits associated with auditory deprivation.

To address these issues, we induced a temporary period of sound attenuation during juvenile development, similar to that experienced by children with middle ear infections (Whitton and Polley, 2011). Our findings suggest that ACx recovered when hearing was restored, but auditory-recipient striatum remained impaired into adulthood. These disparate outcomes suggest that abnormal activity during development can cause permanent striatal dysfunction, possibly explaining cognitive deficits that accompany congenital neurological disorders.

Results

Auditory cortex projection to dorsal striatum

To generate a corticostriatal slice preparation we first injected an anatomical tracer, fluororuby (FR), into L5/6 core ACx (n=4, Figure 1A), and visualized the projections to dorsal striatum. This resulted in dense anterograde labeling of L6 corticothalamic boutons and retrograde labeling of medial geniculate nucleus (MG) cell bodies (Figure 1A, inset). The MG labeling validated the ACx injection site. Dense anterograde labeling was also observed in dorsolateral striatum (Figure 1B), with sparse labeling in rostral dorsomedial striatum (Figure 1C). To demarcate the corticorecipient region of dorsal striatum, the area containing ACx-labeled terminals was calculated for each striatal section. Figure 1D shows that the average pattern of labeling was most intense proximal to ACx, and dropped off abruptly ≈1.0 mm from the rostral pole of the hippocampus (n=4). To confirm that this pathway was present in the auditory corticostriatal slice preparation, AAV-CamKII-mCherry was injected into ACx layer 5 (L5), and allowed to transport for 3 weeks (n=4, Figure 1E). Brains were subsequently sectioned in the same perihorizontal plane used to generate the corticostriatal brain slice preparation (Figure 1F), and the ACx L5 injection site was confirmed. ACx efferent fiber labeling was abundant in the MG (Figure 1G) and the caudal striatum (Figure 1H), with terminals apposed to striatal cell bodies (Figure 1H, inset).

Figure 1.

Anatomical and functional mapping of the auditory corticostriatal projection. A) Fluorescent image showing ACx injection site and retrograde labeling in the medial geniculate nucleus (MG), scale 25 µm. B) Fluorescent image of coronal slice showing dense anterograde labeling from ACx to dorsolateral striatum, scale 1 mm; inset 25 µm. C) Fluorescent image of a coronal slice containing sparse ACx labeling within dorsomedial striatum, scale 1 mm; inset 25 µm. D) Scatter-plot showing the percent of total area within striatum labeled by ACx injections (y-axis) as a function of distance from the rostral pole of the hippocampus (x-axis). E) Diagram showing the ACx site of AAV injection and the plane of the slice preparation containing the injection site. F) Schematic of the corticostriatal slice preparation with ACx L5/6 injection site and labeled pathways. G) Fluorescent image showing ACx L5/6 injection site and labeling in MG, but no labeling in the lateral geniculate nucleus. H) Fluorescent image showing anterograde labeled fibers of passage and terminals from L5 pyramidal neurons synapsing onto striatal MSNs (inset). I) Diagram showing the slice preparation and experimental design for evoking PSPs in striatal MSNs. J) Micrograph showing the region of striatum used for functional mapping. Numbers represent the site of a patched neuron and are accompanied by traces of PSPs following stimulation of L5. K) Micrograph showing the approximate mediolateral-rostrocaudal location of recorded striatal neurons used for functional mapping, and the peak amplitude of the PSP for each cell (colored bar, left). L) A scatter-diagram showing the peak PSP amplitude as a function of distance from the rostral pole of the hippocampus.

To validate the anatomical tracing experiments, the topography of ACx-evoked postsynaptic potentials (PSP) were recorded from putative MSNs within dorsal striatum (Figure 1I). The striatal regions that displayed the densest fiber labeling (Figure 1D), also exhibited the largest L5-evoked PSP. Figure 1J shows an exemplar brain slice in which large PSPs (red) were recorded caudally at sites 1–3. L5-evoked PSPs were small (cyan) or absent at rostral sites 4–6. Figure 1K and L summarize the results of these recordings (n=10 slices), and illustrate the largest PSPs were observed in the same striatal region that received the densest ACx projection (Figure 1D).

Differential response to HL along the corticostriatal circuit: neuron excitability

To test whether neuron excitability was sensitive to transient developmental HL (Figure 2A, top), we assessed membrane and discharge properties in adult ACx L5 pyramidal neurons and their target MSNs in ACx-recipient striatum (Figure 2A, bottom). Striatal cell phenotype was distinguishable under IR-DIC by differences in cellular properties of MSNs and fast-spiking interneurons (Kawaguchi 1993; Supplemental Results, Figure S1A-D). Cortical cell phenotype was characterized by cell type-specific discharge properties (Hattox and Nelson 2007; Supplemental Materials, Figure S1E-H, Table S2). In adult striatal MSNs, resting potential was significantly hyperpolarized following transient HL (Figure 2B; CTL: -60.7 ± 0.9 mV, HL: -67.8 ± 0.5 mV, p<0.001), and there was a commensurate increase in action potential amplitude (CTL: 83.5 ± 3.2 mV, HL: 94.5 ± 2.1 mV, p<0.01). Furthermore, we observed a significantly lower membrane resistance (Figure 2C; CTL: 119 ± 13 MΩ, HL: 67 ± 8 MΩ, p<0.01) and a faster time constant (CTL: 6.4 ± 0.8 ms, HL: 3.3 ± 0.4 ms, p<0.01). However, neither action potential threshold (CTL: -32.9 ± 2.4 mV, HL: -37.9 ± 3.6 mV, p>0.1), nor half-width (CTL: 1.09 ± 0.06 ms, HL: 0.92 ± 0.03 ms, p>0.05), were significantly affected by transient HL. Together, these HL- induced changes to striatal MSN intrinsic properties resulted in a reduced current-evoked discharge rate (Figure 2E; Linear regression analysis w repeated measures; F(1,43)=10.81, p<0.01). Striatal membrane properties exhibited the same effects during the juvenile period (Table S1). Thus, the reduced excitability and lower firing rates (Linear regression analysis w repeated measures; F(1,38)=19.4, p<0.001) following HL were present in striatum at an early age and persisted into adulthood.

Figure 2.

The effect of transient developmental HL on cortical and striatal intrinsic properties. A) Diagram illustrating the experimental design (top) and examples of a recorded L5 pyramidal neuron (bottom left) and a recorded medium spiny neuron (bottom right) filled with fluorescent biotin. B) Bar graph shows the average resting membrane potential in control (Ctl) and HL animals for L5 pyramidal neurons (L5, orange) and striatal MSNs (Str, green). C) Bar graph shows the average membrane resistance in response to a -30 pA current pulse in control and HL animals for L5 pyramidal neurons and striatal MSNs. D) Representative traces are shown for a Ctl and a HL L5 neuron in response to depolarizing and hyperpolarizing current pulses (left). The plot compares the average input-output functions for ACx L5 neurons in Ctl and HL animals. The Ctl:HL ratio of maximum firing rate at an injected current level of 600 pA was 0.94 (dashed box). E) Representative traces are shown for a Ctl and a HL striatal MSNs in response to depolarizing and hyperpolarizing current pulses (left). The plot compares the average input-output functions for striatial MSNs in Ctl and HL animals. The Ctl:HL ratio of maximum firing rate at an injected current level of 600 pA was 0.51 (dashed box). Data are represented as mean ± SEM. ** p<0.01.

Adult ACx L5 pyramidal cells responded quite differently to transient auditory deprivation. Following HL, resting potential was depolarized (Figure 2B; CTL: -66.0 ± 0.6 mV, HL: -61.5 ± 1.1 mV, p<0.01), action potential amplitude was smaller (CTL: 94.6 ± 2.3 mV, HL: 86.3 ± 3.1 mV, p<0.05), action potential half-width was longer (CTL: 0.87 ± 0.03 ms, HL: 1.31 ± 0.11 ms, p<0.01), and action potential threshold was lower (CTL: -28.7 ± 1.8 mV, HL: -38.5 ± 4.1 mV, p<0.05). There was no significant difference in membrane resistance (Figure 2C; CTL: 115 ± 10 mΩ, HL: 163 ± 29 mΩ, p>0.05) or time constant (CTL: 10.8 ± 1.2 ms, HL: 10.1 ± 0.9 ms, p>0.1). Therefore despite significant changes to many intrinsic properties, adult ACx L5 neurons maintained control-like current evoked discharge rates (Figure 2D; Linear regression analysis w repeated measures; F(1,22)=0.35, p>0.1). Unlike striatal MSNs, many L5 cortex properties displayed a differential response to HL during the juvenile period (Table S1). Most notably, firing rate was significantly reduced at high current injection levels (linear regression analysis with repeated measures; F(1,28)=4.3, p<0.05).

Differential response to HL along the corticostriatal circuit: inhibitory synaptic gain

To investigate the effects of HL on inhibitory transmission in adults, spontaneous (s) and minimum evoked (me) inhibitory postsynaptic currents (IPSC) were recorded from ACx L5 pyramidal neurons and their target MSNs under voltage-clamp. Following a period of recovery from transient HL, adult striatal MSNs displayed reduced inhibitory gain. As shown in Figure 3, adult striatal MSNs displayed a reduction of sIPSC amplitude (CTL: -22.3 ± 1.5 pA, HL: -15.2 ± 0.8 pA, p<0.001), me-IPSC amplitude (CTL: -11.5 ± 0.5 pA, HL: -9.0 ± 0.5 pA, p<0.01), and evoked IPSCs obtained at twice the threshold stimulus current (CTL: -304 ± 28 pA, HL: -160 ± 20 pA, p<0.001). There was also an associated increase in sIPSC time constant (CTL: 21.5 ± 1.4 ms, HL: 28.0 ± 2.0 ms, p<0.05). Furthermore, adult MSNs displayed a reduced sIPSC frequency (CTL 11.8 ± 1.2 Hz, HL: 4.5 ± 0.8 Hz, p<0.001) and increased paired pulse ratios following transient HL (CTL: 1.0 ± 0.02, HL: 1.24 ± 0.08, p<0.05). Juvenile MSNs did not exhibit a significant change in inhibitory gain (Table S1), indicating that this effect of transient HL emerged later in development.

Figure 3.

The effect of transient developmental HL on cortical and striatal inhibitory gain. A) Representative traces showing L5 (top, orange) and striatal (bottom, green) sIPSCs from Ctl and HL animals. The bar graph shows average sIPSC amplitude in Ctl and HL animals for ACx L5 and striatal MSNs. B) Representative traces showing L5 (top, orange) and striatal (bottom, green) me-IPSCs. The bar graph shows average me-IPSC amplitude in Ctl and HL animals for ACx L5 and striatal MSNs. C) Representative traces showing L5 (top, orange) and striatal (bottom, green) evoked IPSCs at 2x the threshold stimulus current. The bar graph shows average IPSC amplitude in Ctl and HL animals for ACx L5 and striatal MSNs. We computed the Ctl:HL ratio of evoked IPSC amplitude. The value was 2.8 for L5 neurons and 0.52 for striatal MSNs (dashed boxes). Data are represented as mean ± SEM. **p<0.01. ***p<0.001.

Adult ACx L5 neurons displayed increased inhibitory gain following transient HL. As shown in Figure 3, we observed an increase in sIPSC amplitude (Ctrl -14.1 ± 0.5 pA, HL: -34.4 ± 3.8 pA, p<0.001), me-IPSC amplitude (CTL: -6.3 ± 0.4 pA, HL: -10.9 ± 0.5 pA, p<0.001), and evoked IPSCs obtained at twice the threshold stimulus current (CTL: -131 ± 13 pA, HL: -371 ± 54 pA, p<0.001). There was also an associated decrease in sIPSC time constant (CTL: 15.0 ± 0.9 ms, HL: 9.8 ± 0.7 ms, p<0.001). Furthermore, sIPSC frequency was higher (CTL: 7.5 ± 0.9 Hz, HL: 15.3 ± 0.7 Hz, p<0.001) and paired pulse ratios were smaller (CTL: 1.10 ± 0.02, HL: 0.77 ± 0.04, p<0.001), suggesting an increased GABA release probability. Although increased inhibitory strength was observed in juvenile L5 neurons, the effect was much smaller (Table S1). Together, Figure 3 suggests that transient HL leads to a dramatically different effect on inhibitory gain in adult L5 cortex and striatal MSNs. Similar results were found with a shorter duration of HL and timed onset of HL suggests the existence of a critical period (Table S3).

Differential response to HL along the corticostriatal circuit: increased excitatory synaptic gain

To assess excitatory gain in adults following a period of recovery from developmental transient HL, channelrhodopsin was expressed in ACx pyramidal neurons (see Methods). This permitted the selective activation of L5 excitatory synapses in both striatum and neighboring ACx L5 cells (Figure 4A). Figure 4B shows a patched pyramidal cell receiving synaptic terminals from transfected L5 cells (top) and a patched MSN also receiving synaptic terminals from transfected L5 projections (bottom). Following transient HL, the amplitude of excitatory postsynaptic potentials (EPSP) was significantly larger, both in MSNs (Figure 4D; Linear regression analysis w repeated measures: F(1,29)=4.84, p<0.05) and ACx L5 neurons (Figure 4C; Linear regression analysis w repeated measures: F(1,23)=33.2, p<0.001). Therefore, monosynaptic excitatory connections in both cortex L5 pyramidal neurons and striatal MSNs were stronger following transient HL. Once again, whereas the excitatory gain in ACx emerged slowly after hearing was restored, the striatal MSN excitatory gain became large during the juvenile period and remained unchanged through adulthood (Table S1).

Figure 4.

The effect of developmental HL on cortical and striatal excitatory properties. A) Diagram showing the experimental design for recording light-evoked EPSPs. B) Examples of a patched striatal MSN (top) and a L5 pyramidal cell (bottom) receiving synaptic terminals from ACx L5 pyramidal cells transfected with channelrhodopsin and mCherry fluorescent protein. Insets show light-evoked responses for these cells. C) The graph compares average light-evoked responses in ACx L5 neurons for Ctl and HL animals. The Ctl:HL ratio of maximum evoked EPSP amplitude was 2.9 (dashed box). D) The graph compares average light-evoked responses in striatal MSNs for Ctl and HL animals. The Ctl:HL ratio of maximum evoked EPSP amplitude was 3.8 (dashed box). E) Example traces showing light-evoked AMPA receptor-mediated EPSP (top), recorded at the neurons resting potential (black) and at a holding potential of -80 mV (blue). F) The bar graph shows the average maximum light-evoked AMPA receptor-mediated EPSP amplitude in L5 pyramidal neurons and striatal MSNs for Ctl and HL animals. Data are represented as mean ± SEM. ** p<0.01.

To determine whether increased excitatory synaptic gain was mediated by AMPA receptors, we acquired light-evoked EPSPs at a holding potential of -80 mV. We found that the maximum evoked AMPAergic EPSP amplitude (Figure 4E; CTL: 2.3 ± 0.3 mV, HL: 11.8 ± 1.6 mV, p<0.001) and duration (CTL: 20.4 ± 2.9 ms, HL 34.5 ± 3.1 ms, p<0.01) were each significantly greater in adult striatal MSNs following transient HL. Similarly, peak AMPAergic EPSP amplitude was significantly larger (Figure 4E; CTL: 9.1 ± 1.8 mV, HL 16.2 ± 1.2 mV, p<0.001) and duration was significantly longer (Figure 4F; CTL: 39.5 ± 3.3 ms, HL: 68.5 ± 7.9 ms, p<0.01) in ACx L5 neurons. Thus, Figure 4 shows that the increased excitatory synaptic gain in both L5 neurons and MSNs that is mediated, in part, by AMPA receptors.

Discussion

The perceptual impairments that attend developmental HL are generally attributed to processing deficits within the primary auditory neuraxis. However, clinical studies suggest that sensory deprivation can lead to diminished cognitive performance (e.g., auBuchon et al., 2015; Rhodes et al., 2016). Here, we report that a transient HL caused very different outcomes in ACx projection and MSN recipient neurons. The onset of HL caused juvenile ACx neurons to display an imbalance of excitatory and inhibitory gain and a reduced firing rate. However, when hearing was restored, ACx neurons readjusted their cellular properties such that excitatory and inhibitory strength became balanced and firing rate was control-like by adulthood. In contrast, juvenile MSNs displayed a reduced inhibition, enhanced excitation and a persistent reduction in firing rate, that persisted unabated through adulthood (Figure 2–4). Thus, striatal neurons may be especially vulnerable to abnormal sensory input during development.

The opposing changes that we observed between cortex and striatum are consistent with a report showing that early cortical activity regulates corticostriatal circuit maturation (Peixoto et al., 2016). Whereas they report that cortical hyperactivity leads to decreases in striatal excitatory gain and increased firing rates, we found that lower juvenile L5 firing rates were associated with increased striatal excitatory gain and lower firing rates (see Table S1). Together these results demonstrate a developmental mechanism by which early cortical activity inversely regulates the maturation of striatal synaptic and cellular properties. However, cortical activity does not continue to exert the same level of influence in adulthood. If this were the case, then striatal synapses and firing properties would have normalized, as we observed for L5 neurons. Instead, the results indicate that striatal cellular properties are permanently established during a brief developmental critical period. That is, MSNs compensate in response to the onset of developmental HL, but fail to compensate when hearing is restored after the critical period closes (Supplemental Material, Table S3).

Our findings raise the possibility that the etiology of striatal dysfunction in sensory and other neurological disorders may depend, in part, on abnormal cortical activity during an early critical period of development. Thus, early dysfunctional connectivity between the cortex and the striatum could lead to the gradual emergence of striatal related symptoms for ADHD (Braz et al., 2015), autism (Peixoto et al., 2016), schizophrenia (Morris et al., 2015), Parkinson's Disease (Stephens et al., 2005), Huntington's Disease (Cepeda et al., 2007), and Tourette's Syndrome (Israelashvili and Bar-Gad, 2015). Likewise, a compromised corticostriatal pathway may contribute to some non-sensory deficits that accompany developmental HL (Reichman and Healey 1983; Teele et al. 1990; Schönweiler et al. 1998; Pisoni et al. 2011; Beer et al. 2014; Kronenberger et al. 2014; Caras and Sanes, 2015; von Trapp et al., 2016). Our results provide one mechanism by which early changes to striatal cellular physiology during a critical period could lead to long-term cognitive impairments in a broad range of neurological and sensory disorders.

Experimental Procedures

Experimental animals

We recorded from 215 cortical and 330 striatal neurons in 136 (71 male, 65 female) gerbils (Meriones unguiculatus) born from breeding pairs (Charles River), and aged postnatal days (P) 35–126. All procedures were approved by the Institutional Animal Care and Use Committee at New York University.

Neuroanatomical tracer injections

Injections of the anatomical tracer fluororuby (FR) into core ACx (n=8 gerbils, P55–65) were carried out as previously reported (Mowery et al., 2016).

Transient HL with bilateral earplugs

Earplugging was carried out as previously reported in detail (Mowery et al., 2015; 2016). This manipulation produces a threshold shift of 15–50 dB, depending on frequency, as measured with auditory brainstem responses (Caras and Sanes, 2015), and ~25 dB at 4 kHz as measured behaviorally (Mowery et al., 2015).

Brain Slice Preparation

We generated perihorizontal brain slices (400 µm) that retain corticostriatal connectivity between the ACx and ACx-recipient striatum. Slices were transferred to a chamber containing oxygenated artificial cerebrospinal fluid (ACSF) at 32°C (30–45 min). Slices remained at room temperature for at least 1 hr before recordings were obtained at 32°C in oxygenated ACSF. The ACSF contained (mM): 125 NaCl, 4 KCl, 1.2 KH₂PO₄, 1.3 MgSO₄, 26 NaHCO₃, 15 glucose, 2.4 CaCl₂, and 0.4 L-ascorbic acid, pH=7.4 (when bubbled with 95% O₂/5% CO₂).

Brain slice recordings

Whole cell current and voltage clamp recordings were carried out and analyzed as reported previously (Mowery et al., 2015; 2016). Excitatory properties were obtained in one of two ways. For juvenile striatal recordings, depolarizing postsynaptic potentials (PSPs) were measured by stimulating L5 pyramidal neurons with incremental current steps. The peak PSP amplitude was obtained in response to maximum stimulus level, and duration was calculated as the time from the PSP onset to the return to baseline. For adult recordings, excitatory postsynaptic potentials (EPSPs) were generated via optogenetic activation (1 ms, 470 nm, 1–10 mW) of ACx L5 pyramidal neuron terminals located in ACx- recipient striatum or ACx L5 pyramidal neurons adjacent to the injection site. EPSP amplitudes and durations were obtained with a holding potential of -80 mV using light pulses ranging from 1 to 10 mW. The peak amplitude and duration of each EPSP were measured at 10 mW light intensity. To confirm that AMPAergic potentials were measured, AP5 was added to the bath in some experiments to show that no NMDAergic component was present, and DNQX was added at the end to show that the EPSP was completely eliminated.

Optogenetic experiments

Gerbils (n=16, P55–65) were deeply anesthetized (isoflurane 2%), placed in a stereotaxic frame, and the left temporal bone was exposed. A craniotomy was made in the temporal bone at the level of core ACx and a durotomy was made, as described previously (Mowery et al., 2016). Glass electrodes were loaded with adenovirus containing a CaMKII promotor that infects pyramidal neurons (AAV1.CaMKIIa.hChR2(H134R)- mCherry.WPRE.hGH; Penn Vector core, Addgene26975) and a Nanoject (Drummond) was then used deliver 50 nL of virus at 800 µm from the pial surface. The craniotomy was covered with sterile bone wax, the surgical site closed with (Vetbond), and the animal was allowed to recover for 3 weeks prior to in vitro recording. For optical stimulation of channelrhodopsin (ChR2), whole cell recordings were obtained, and brief pulses (1 ms) of blue light (470 nm; 1–10 mW; Plexbright) were used to activate the terminals of infected ACx L5 pyramidal neurons.

Statistical Analyses

All data are presented as means ± SEM. Statistical tests were performed using statistical software (JMP; SAS Institute). For F–I curves, a 2-way mixed-model ANOVA (Linear regression analysis w repeated measures) was used to verify a main effect of HL on evoked firing rate during incremental current injection steps. PSPs and EPSPs were compared using a 2-way mixed-model ANOVA (Linear regression analysis w repeated measures) to verify a main effect of HL on evoked PSP amplitude or duration during incremental current injection steps. ANOVA and t-test were used for all data that was distributed normally.

Highlights.

• Early hearing loss causes permanent imbalance to striatal cellular properties

• Corticostriatal projection neurons display balanced synapses and normal firing

• Striatal impairments may explain the slow emergence of neurological symptoms

Abbreviations

ACx

auditory cortex

HL

hearing loss

L5

layer 5

MSN

medium spiny neuron

P

postnatal day