Abstract
Accumulating evidence suggests a role for inhibitory neurotransmitter dysfunction in the pathology of tinnitus. Opposing hypotheses proposed either a pathologic decrease or increase of GABAergic inhibition in medial geniculate body (MGB). In thalamus, GABA mediates fast synaptic inhibition via synaptic GABAA receptors (GABAARs) and persistent tonic inhibition via high-affinity extrasynaptic GABAARs. Given that extrasynaptic GABAARs control the firing mode of thalamocortical neurons, we examined tonic GABAAR currents in MGB neurons in vitro, using the following three groups of adult rats: unexposed control (Ctrl); sound exposed with behavioral evidence of tinnitus (Tin); and sound exposed with no behavioral evidence of tinnitus (Non-T). Tonic GABAAR currents were evoked using the selective agonist gaboxadol. Months after a tinnitus-inducing sound exposure, gaboxadol-evoked tonic GABAAR currents showed significant tinnitus-related increases contralateral to the sound exposure. In situ hybridization studies found increased mRNA levels for GABAAR δ-subunits contralateral to the sound exposure. Tin rats showed significant increases in the number of spikes per burst evoked using suprathreshold-injected current steps. In summary, we found little evidence of tinnitus-related decreases in GABAergic neurotransmission. Tinnitus and chronic pain may reflect thalamocortical dysrhythmia, which results from abnormal theta-range resonant interactions between thalamus and cortex, due to neuronal hyperpolarization and the initiation of low-threshold calcium spike bursts (Walton and Llinás, 2010). In agreement with this hypothesis, we found tinnitus-related increases in tonic extrasynaptic GABAAR currents, in action potentials/evoked bursts, and in GABAAR δ-subunit gene expression. These tinnitus-related changes in GABAergic function may be markers for tinnitus pathology in the MGB.
Introduction
Inhibitory neurotransmission is essential for accurate processing of acoustic information. It modulates input/output functions, alters frequency tuning, and sharpens temporal response accuracy while gating neurotransmission to primary auditory cortex (AI; Wang et al., 2008; Edeline, 2011; Richardson et al., 2012). Downregulation of inhibitory function has been suggested to underpin tinnitus pathology in dorsal cochlear nucleus (DCN; Wang et al., 2009; Pilati et al., 2012a; Pilati et al., 2012b), inferior colliculus (IC; Bauer et al., 2008; Dong et al., 2009), and AI (Noreña and Eggermont, 2003; Yang et al., 2007, 2011; Roberts et al., 2012) where tinnitus is accompanied by increased spontaneous activity, bursting, enhanced sound-evoked responses, and reduction in markers of inhibitory neurotransmission (Eggermont and Roberts, 2004; Middleton et al., 2011). This downregulation of inhibitory function is thought to reflect altered homeostatic plasticity compensating for the loss of peripheral excitatory drive (Noreña, 2011; Richardson et al., 2012).
Few studies have examined tinnitus pathophysiology in auditory thalamus/medial geniculate body (MGB). MGB gates the percept of sound as it projects to AI (Banks and Smith, 2011) and limbic structures, which are thought to be involved in individuals most disturbed by their tinnitus (Rauschecker et al., 2010; Leaver et al., 2011). While MGB was historically considered a simple relay of auditory signals between IC and AI, recent studies suggest additional signal processing in MGB (Bartlett and Smith, 1999; Winer et al., 2005; Edeline, 2011; Bartlett, 2013). MGB has intricate reciprocal connectivity with the GABAergic thalamic reticular nucleus (TRN) that regulates attention and sleep (Jones, 1975; Rouiller et al., 1985). The unique connectivity and function of MGB strongly implicates it as a key link in the tinnitus network (Rouiller et al., 1985; Rauschecker et al., 2010; Leaver et al., 2011).
Data from sodium salicylate models of tinnitus suggest altered excitation and inhibition in rat MGB (Su et al., 2012). Proper functioning of MGB is dependent on normal GABAergic inhibition, which in sensory thalamus may account for ∼90% of total inhibitory conductance (Belelli et al., 2005; Cope et al., 2005, 2009). In rodents, MGB receives GABAergic projections from IC (Winer et al., 1996; Peruzzi et al., 1997; Ito and Oliver, 2012) and TRN (Montero and Scott, 1983; Rouiller et al., 1985), while GABAergic interneurons comprise <1% of MGB neurons (Winer and Larue, 1996). GABAergic inputs to MGB act via postsynaptic GABAA receptors (GABAARs) that are heterogeneous (Sur et al., 1999; Richardson et al., 2012). MGB synapses express low-affinity, fast-desensitizing GABAARs (α1, β2, γ2 subunits) that mediate fast synaptic transmission and enable high temporal precision (Mozrzymas, 2004, 2010; Farrant and Nusser, 2005; Barberis et al., 2011). In addition, MGB expresses high levels of GABAARs containing α4- and δ-subunits that mediate non-desensitizing tonic inhibition (Richardson et al., 2011). The presence of the δ-subunit in GABAARs results in extrasynaptic location and high agonist affinity, which favors activation by low ambient levels of GABA in extracellular space, which may arise from synaptic spillover and/or release from glia (Farrant and Nusser, 2005; Brickley and Mody, 2012). These non-desensitizing GABAARs can generate long-lasting tonic currents that hyperpolarize and shift thalamocortical neurons from tonic to burst-firing mode. This shift could alter the salience of phantom acoustic signals to AI.
GABAergic inhibition is critically involved in determining higher-frequency thalamocortical oscillations as well as slow-wave sleep (Llinás and Steriade, 2006). Llinás et al. (2005) postulated that excessive inhibition in thalamus could be the origin of pathologically increased hyperpolarization of thalamic neurons, resulting in deinactivation of T-type Ca2+ channels. Upon depolarization, these channels generate low-threshold burst spikes in the wakeful state. A proposed thalamocortical dysrhythmia model suggests that elevated inhibition and the resultant focal burst activity associated with abnormal low-frequency oscillations in thalamus produce ectopic disinhibition of fast oscillatory activity in AI that could correlate with tinnitus (Llinás and Steriade, 2006).
Using in vitro patch-clamp recordings, we sought to clarify the role of GABAA inhibition in MGB as it relates to tinnitus pathophysiology in a sound-exposure behavioral model of tinnitus. In contrast to findings from other auditory structures, MGB circuits showed an increase in GABAAR function in animals with behavioral evidence of tinnitus.
Materials and Methods
Young-adult (3–4 months) Long–Evans male rats (Harlan Laboratories) were housed on a 12 h reversed light/dark cycle with ad libitum access to food and water. Forty-five Long–Evans rats were used for electrophysiological recordings in the MGB slice preparation and 13 Fischer Brown Norway (FBN) rats from earlier tinnitus experiments (Wang et al., 2009) were used for MGB in situ hybridization studies. Both strains have similar life expectancy, with young adult animals used in both studies.
All experiments were conducted in strict accordance with a protocol approved by the Animal Care and Use Committee of Southern Illinois University.
Noise exposure and induction of tinnitus.
Tinnitus was induced between 3 and 4 months of age, using methods similar to those described by Bauer et al. (1999) and Bauer and Brozoski (2001). Briefly, acoustic exposure was a 16 kHz octave band at 116 dB SPL into one ear of a ketamine/xylazine anesthetized rat for 1 h. This exposure typically results in an ipsilateral 30–40 dB temporary threshold shift [measured as pre-to-post noise exposure shift in auditory brainstem response (ABR) thresholds] across frequencies, with increased threshold elevations at frequencies near the 16 kHz exposure (Wang et al., 2009). The exposure also produces a small, variable, permanent ipsilateral threshold shift as measured by ABR (Wang et al., 2009). Acoustic nerve fiber losses were not measured in this set of sound-exposed rats (Bauer et al., 2007; Kujawa and Liberman, 2009). Prepulse inhibition (PPI) was assessed in parallel with gap detection performance. ABR and prepulse findings suggest that gap detection was not altered by unilateral hearing loss at the levels used in these studies (Turner et al., 2006).
Behavioral testing.
The gap detection reflex method was adapted from Turner et al. (2006) and Turner (2007), and was based upon the ability of the acoustic startle reflex to be reduced when preceded by a silent gap in a constant acoustic background. As detailed in Wang et al. (2009), behavioral testing used Kinder Scientific startle reflex hardware and software, which was customized for this application. Background sounds were presented through one speaker (XT25TG30-04, Vifa) with startle stimuli presented through a second speaker (CTS KSN-1005, PowerLine). A piezoelectric transducer provided the measure of startle force applied to the floor. An animal holder was suspended above the floor, allowing the animal to freely turn around while minimizing excessive movement. The sound-gap envelope consisted of 60 dB peak (SPL) broadband noise (BBN) or filtered noise (one-third octave bandpass, 48 dB/octave roll off, Model 3988, Krohn-Hite) centered at 10, 12.5, 16, 20, and 25 kHz, calibrated at 60 dB peak SPL. PPI was measured in parallel as a hearing level control and to test for hyperacusis (Turner and Parrish, 2008). Animals with threshold elevation and/or showing hyperacusis were eliminated from the study. Two months after noise exposure, animals were behaviorally tested twice on different occasions, and the results of two sessions were averaged to determine tinnitus. Gap/PPI test sessions began with a 1 min acclimation period followed by two trials consisting of the startle-eliciting noise burst (115 dB SPL, 20 ms duration), which served to habituate the startle response to a more stable baseline. Data from these two initial trials were not used in the gap detection analysis. The remainder of the session consisted of startle-only trials mixed with gap or PPI trials (Turner et al., 2006; Kalappa et al., 2014).
The intertrial interval varied throughout test sessions, with targeted frequencies and BBN stimuli for both gap and PPI presented in ascending order from 10 kHz to BBN (380 trials per session). Gap and prepulse stimuli occurred 100 ms before the startle stimulus and were 50 ms long with a 1.0 ms rise/fall time (Turner et al., 2006). Acoustic stimuli were calibrated using a cloth model rat and a Brüel & Kjær Pulse System with a 0.5 inch free-field microphone (model 4191, Brüel & Kjær). Ambient sound levels in test chambers (with background test noise turned off) were <20 dB SPL in the 4–40 kHz range.
Tinnitus determination.
Data from gap detection and PPI testing were expressed as the ratio of mean startle response for a gap or PPI trial relative to a proximal startle-only trial (for details, see Kalappa et al., 2014). A response of 1.0 meant that the startle reflex was the same whether or not it was preceded by a gap or prepulse stimulus. The lower the ratio is from 1.0, the greater the startle inhibition. An individual subject's average data from the two behavioral test sessions were inspected for freedom from artifacts, such as a high level of ambulatory movement. Tinnitus index scores were calculated by subtracting the PPI ratio from the gap ratio. Higher scores are indicative of tinnitus-like deficits in processing silent gap cues. This index score provides a metric that describes the response of an animal to silent gap cues relative to control sounds and helps control for several potential confounds in the gap ratio as a standalone metric for tinnitus. The index score helps to control for hearing loss as an inability to hear or process the 60 dB test stimulus, which would result both in poor gap ratios and poor PPI ratios, hence a small score. The index also helps to control for startle reflex amplitude differences. If an animal has such a small startle reflex that it cannot be inhibited by a silent gap cue, the same impact would be found in the PPI testing where small startle reflexes would also be expected, again resulting in a small index score. This metric controls for temporal processing deficits, which should be present in both the gap and PPI conditions, given their identical timing features. Finally, this index corrects for nonspecific sensory gating/filtering failures, which should impact both gap and PPI response ratios. Mean tinnitus index scores for control animals were pooled to obtain the average control group startle ratio, with 95% confidence intervals determined for each frequency and BBN.
The average tinnitus index score for each noise-exposed subject was compared with the average control tinnitus index score for that frequency. Noise-exposed subject responses that fell within the 95% confidence interval for the control group were deemed no different from those of control subjects and were without tinnitus for that given frequency or BBN. Noise-exposed subjects with tinnitus index scores reduced beyond the 95% confidence interval of the companion control group were considered significantly different. These subjects were considered to have tinnitus at that given frequency or BBN.
In situ hybridization.
Adult (4 months old) male FBN rats were used for in situ hybridization experiments. The animals were repurposed from a prior series of similarly conducted tinnitus experiments with the identical sound exposure conditions. Previously, we published patch-clamp recordings of gaboxadol (GBX)-evoked tonic GABAAR currents from MGB neurons in adult FBN rats (Richardson et al., 2011, 2013). The recorded parameters of gaboxadol-evoked tonic GABAAR currents in FBN rats were comparable to those presented here for Long–Evans rats. Animals were divided into the following two groups: sound exposed and control/sham exposed. These animals were not assigned a tinnitus score. In situ hybridization studies were conducted as described by Ling et al. (2005). Oligonucleotide probes for α4 (Wisden et al., 1991; 5′-CAA GTC GCC AGG CAC AGG ACG TGC AGG AGG GCG AGG CTG ACC CCG-3′) and δ (provided by Merck Inc.; 5′-GAC CTT GGC TTT CCG TTT CTT CCT GTA GTC AGC ATT-3′) were synthesized and purified by Sigma-Genosys. The probes were then 3′-end labeled with 35S-dATP (NEN-PerkinElmer Life Science) by 3′-terminal deoxynucleotidyl transferase using a commercially available kit (Fisher Scientific), and purified and desalted through a micro BioSpin-30 column (Bio-Rad). These labeled probes were used to detect GABAAR subunits α4 and δ in MGB. Additional methods, visualization, and quantification were performed as in the study by Richardson et al. (2013).
All slides were blinded to ensure unbiased measurement. The accumulation of silver grains over cell bodies (∼10 times above slide background) was interpreted as positive hybridization labeling for the α4 and δ GABAAR subunits to their specific mRNAs. Only grains within the perimeter of the cell were counted. Data were collected from cells only if they were distinguishable as individual cells possessing a nucleolus. The numbers of grains over somata of MGB neurons (silver grains/100 μm2) were counted, and values were corrected by subtracting background labeling obtained from five random off-tissue areas on each slide.
Preparation of thalamic brain slices for electrophysiology.
Two months following sound exposure and within 2 weeks of behavioral testing, Long–Evans rats (5–7 months old) were anesthetized with isoflurane and decapitated. Brains were rapidly removed, and horizontal brain slices (300 μm) were cut using a VT 1000P vibratome (Leica Microsystems) in ice-cold (<4°C) sucrose-rich artificial CSF (aCSF) containing the following (in mm): sucrose 250, KCl 3, NaH2PO4 1.23, MgCl2 5, CaCl2 0.5, NaHCO3 26, and glucose 10, pH 7.4, saturated with carbogen (95% O2/5% CO2). Slices were retained in a holding chamber filled with standard aCSF containing the following (in mm): NaCl 124, KCl 3, KH2PO4 1.25, MgCl2 1.3, CaCl2 2, NaHCO3 26, and glucose 10, pH 7.4, saturated with carbogen and warmed to 32°C for 20 min before being allowed to equilibrate to room temperature (∼22°C) for at least 1 h before recording. Slices were transferred to a submersion recording chamber mounted on the stage of an upright Olympus BX-51WI microscope, maintained at room temperature, and continuously perfused with standard aCSF saturated with carbogen, at a flow rate of 2 ml/min using a peristaltic P720 pump (Instech Laboratories).
Electrophysiology.
Using infrared differential interference contrast microscopy, visually guided, somatic, whole-cell patch-clamp recordings were obtained from MGB neurons using fire-polished borosilicate glass electrodes (4–5 MΩ). Two intracellular solutions were used in this study. In voltage-clamp experiments, patch electrodes were filled with the following (in mm): 120 CsMeSO3, 2 CsCl, 5 NaCl, 1.5 MgCl2, 10 HEPES, 10 EGTA, 5 TEA-Cl, 2.5 Mg-ATP, 0.3 Na-GTP, and 5 QX-314, pH 7.3, adjusted with CsOH, with osmolarity of 290 ± 10 mOsm. In current-clamp experiments, a more “physiological” solution was used, as follows: K-gluconate140, NaCl 1, MgCl2 2, Mg-ATP 2, Na-GTP 0.3, and HEPES 10, pH 7.3–7.4, adjusted with KOH, with osmolarity of 290 ± 10 mOsm. Holding potentials were not corrected for the liquid junction potential. Voltage-clamp and current-clamp recordings were obtained using a MultiClamp-700B amplifier and Digidata-1440A analog-to-digital converter (Molecular Devices). The signal was low-pass filtered at 5 kHz and digitized at 10 kHz. Data were acquired using pClamp 10 software (Molecular Devices). For all recordings, series resistance was <20 MΩ and was monitored continuously.
To isolate GABAergic currents, 20 μm 6,7-dinitroquinoxaline-2,3-dione (DNQX) and 50 μm d(−)-2-amino-5-phosphonopentanoic acid (d-AP5) were added to the standard aCSF. Extrasynaptic tonic GABAAR currents were evoked by bath application of the selective agonist, 0.1–10 μm gaboxadol [4,5,6,7-tetrahydroisoxazolo[5,4-c]pyridine-3(2H)-one], in the presence of glutamatergic receptor blockers. GABAA tonic currents and spontaneous IPSCs (sIPSCs) were recorded at −10 and +20 mV, respectively. GABAA currents were blocked with 40 μm gabazine.
Drugs.
Drugs were dissolved in distilled water to a concentration of 1–10 mm, and the stock solutions were divided into small aliquots and frozen. The following stock drug solutions were used to prepare fresh final solutions every day: QX-314 from Tocris Bioscience; DNQX, d-AP-5, and gabazine from Abcam; with all other chemicals from Sigma-Aldrich.
Analysis.
Postrecording analyses of extrasynaptic GABAAR currents and action potential firing evoked using current steps were analyzed using pClamp 10. Dose–response curve fitting was performed in Origin 8 (OriginLab) using an iterative process to minimize the χ2 test. Spontaneous IPSCs were analyzed using MiniAnalysis version 6.0.3 software (Synaptosoft). All spontaneous IPSPs were visually verified on the basis of event kinetics. Between 500 and 1500 individual events were analyzed for each cell (4 min of gap-free acquisition).
All statistical values were evaluated with Origin 8 (OriginLab) or SAS/STAT (SAS Institute). Values are presented either as the mean ± SEM or mean ± SD (box plot). ANOVA with Tukey's post hoc analysis was used when multiple comparisons were made. Pairwise group differences were evaluated using a Student's t test. Statistical significance was established at p ≤ 0.05.
Results
For all voltage-clamp studies examining tonic and synaptic GABAA currents, analyses were performed on the following three groups: (1) 10 control (Ctrl) rats; (2) 14 sound-exposed rats with behavioral evidence of tinnitus (Tin); and (3) 4 sound-exposed without tinnitus (Non-T). Current-clamp experiments examined membrane properties and burst properties, used eight animals with behavioral evidence of Tin, which compared with nine Ctrls (Fig. 1). Current-clamp studies compared only two groups because of the low availability of exposed non-hearing loss animals reliably identified as Non-T. Figure 1 shows the pooled behavioral data for both experimental groups with significant differences for the broadband noise gap background condition for all groups.
Mean tinnitus scores for BBN in animals. Mean tinnitus scores for Long–Evans rats used in current-clamp experiments [red, Ctrl (n = 8); blue, Tin (n = 9)] and voltage-clamp experiments [red, Ctrl (n = 10); blue, Tin (n = 14); green, Non-T (n = 4)]. Left panels correspond to a first cohort of animals used for measuring MGB neurons passive membrane and firing properties (current clamp). Right panel represents a second set of animals used for measuring GABAAR currents (voltage clamp). Tinnitus index scores were calculated by subtracting PPI ratios from gap ratios (see Materials and Methods). Higher tinnitus scores suggest a deficit processing silent gap cues relative to other sounds. Error bars indicate ±SEM.
The MGB can be divided into the following three distinctive subdivisions: lemniscal ventral MGB (vMGB), non-lemniscal dorsal MGB (dMGB), and medial MGB (mMGB; Fig. 2). In the present study, patch-clamp recordings were localized to vMGB or dMGB, contralateral to the sound exposure.
Thionin-stained coronal section through MGB. dMGB, vMGB, and mMGB divisions are demarcated by dotted black lines. The gray dashed line represents the horizontal sectioning plane for making acute brain slices containing MGB. Inset represents a recorded MGB neuron that was labeled with neurobiotin and traced using Neurolucida. The red star mark defines the location of the neuron in MGB.
Membrane properties of MGB neurons
Immediately following entry into whole-cell mode, the basic membrane parameters, including resting membrane potential (Vrest), input resistance (Rm), whole-cell membrane capacitance (Cm), and membrane time constant (τ) were measured. Measurements of Vrest were obtained by briefly switching to current-clamp mode. Rm and Cm were measured in voltage-clamp mode at 0 holding current with 5 mV hyperpolarizing pulses that were 50 ms in duration. Although Vrest, Rm, Cm, and τ were measured in all neurons, the mean data shown in Table 1 were calculated using the K-gluconate internal solution (see Materials and Methods), as this solution most faithfully replicates physiological conditions. Pairwise differences were evaluated using a two-tailed Student's t test. No significant tinnitus-related changes were observed in the resting membrane potential of MGB neurons (dMGB: Ctrl, −70.7 ± 1.7 mV, n = 18; and Tin, −68.8 ± 1.6 mV, n = 15, p = 0.44; vMGB: Ctrl, −72.5 ± 2.0 mV, n = 17; and Tin, −72.5 ± 2.31 mV, n = 15, p = 0.99). Membrane input resistance had a slight tendency to be higher in Tin animals, but differences were nonsignificant (dMGB: Ctrl, 377.5 ± 37.4 MΩ, n = 18; and Tin, 422.9 ± 39.1 MΩ, n = 15, p = 0.41; vMGB: Ctrl, 407.9 ± 39.6 MΩ, n = 17; and Tin, 420.1 ± 44.7 MΩ, n = 15, p = 0.84). Similarly, the Cm and τ of MGB neurons were not affected by traumatic sound exposure and tinnitus pathology. Basic passive membrane properties of MGB neurons also did not differ significantly between dorsal and ventral MGB subdivisions (Table 1).
Membrane properties and discharge mode of MGB neurons
Tinnitus MGB neurons generate increased numbers of spikes per burst
Changes in the action potential firing properties of MGB neurons related to tinnitus were investigated in the experimental set with K-gluconate internal solution. In agreement with previous reports, the majority of neurons in both Ctrl and Tin groups responded with burst-firing mode upon somatic positive current injection. Representative traces of burst-firing neurons from Ctrl and Tin animals are shown in Figure 3. In accord with previous reports, bursting neurons had a more negative resting membrane potential compared with regular spiking neurons. Although there were no significant group differences in the number of burst-firing neurons (Ctrl, 62.9%; Tin, 66.7%; Table 1), Tin rats tended to have more bursting neurons in the vMGB division (Ctrl, 58.8%; Tin, 73.3%; Table 1). Most importantly, Tin rats showed a significant increase in the number of spikes per burst (4.2 ± 0.4, n = 20) compared with Ctrl (2.9 ± 0.4, n = 24) when evoked using suprathreshold current steps (Fig. 3). Tin versus Ctrl differences in spikes per burst were more evident in the vMGB (vMGB: Ctrl, 2.4 ± 0.5, n = 11; and Tin 4.6 ± 0.7, n = 11; dMGB: Ctrl, 3.4 ± 0.5, n = 13; and Tin 3.7 ± 0.4, n = 9). This tinnitus-related increase in the number of spikes per burst was not directly related to higher input resistance or a difference in resting membrane potential (Table 1).
MGB neurons from the Tin group demonstrated increased spiking during bursting. The injection of positive current, 500 ms, activates a low-threshold Ca2+ spike that generates burst firing in a majority of MGB neurons. Representative recordings from MGB neurons showing burst firing from a control animal (red; Vrest, −76 mV; 80 pA current step) and from a rat with behavioral evidence of tinnitus (blue; Vrest, −76 mV; 50 pA current step). The number of spikes per burst was significantly increased in tinnitus animals. Burst per spike latency was not different among groups. Representative traces with different burst/spike latency were selected for clarity of demonstration only. Insert showing the mean number of spikes per burst in the Ctrl and Tin groups of rats. Error bars indicate ±SEM. *p < 0.05, t test.
We further examined whether changes in the low-threshold Ca2+ conductance might underpin the tinnitus-related increase in the number of spikes per burst. Bursts generated after the offset of hyperpolarizing current injection showed no tinnitus-related differences in amplitude or area of low-threshold Ca2+ spikes (LTSs; Fig. 4A,B). This suggests that tinnitus-related altered excitability may involve additional mechanisms and that further investigation will be required.
LTSs underlying MGB neuronal bursting showed no tinnitus-related differences in amplitude or area. A, Representative recording from a Ctrl animal. MGB neurons were held at −50 mV to inactivate low-threshold Ca2+ channels. Subsequent hyperpolarizing voltage steps, 500 ms, remove channel inactivation and deactivate the channels. LTS current was generated after the offset of the hyperpolarizing voltage step. Inset, Magnified portion of the current traces after the hyperpolarizing voltage steps. B, LTS current amplitude and area [red, Ctrl (n = 39 cells); blue, Tin (n = 30)] plotted against varying hyperpolarizing steps. Data are presented as the mean ± SEM.
Gaboxadol-evoked tonic GABAAR currents are elevated in MGB neurons from Tin rats
We evaluated the tinnitus-associated behavior of GABAAR tonic inhibition in MGB neurons via application of the GABA analog GBX, which at low (micromolar) concentrations selectively activates GABAARs containing α4δ-subunits. Similar to previous reports (Richardson et al., 2011, 2013), bath application of 5 μm GBX activated extrasynaptic GABAARs on MGB neurons and resulted in a negative ∼8 mV hyperpolarizing membrane potential shift that changed MGB neuronal discharge properties from tonic firing to burst mode in response to identical depolarizing current steps (Fig. 5).
Gaboxadol tonically activates extrasynaptic GABAARs, hyperpolarizes resting membrane potential, and converts MGB neurons to burst mode. Top trace shows a representative recording from a tonically firing MGB neuron. Bath application of 5 μm GBX hyperpolarizes (Δ8 mV) the membrane potential, deactivating low-threshold Ca2+ channels, and converted this MGB neuron to burst-firing mode.
Extrasynaptic GABAAR currents were evoked by bath application of increasing concentrations of GBX (0.1–10 μm) in the presence of glutamate receptor blockers (Fig. 4A). Voltage-clamp recordings were performed at −10 mV to increase the Cl− driving force. Each incremental dose was applied for 5 min, with peak amplitude of GABAAR currents measured over a 30 s period. To verify the GABAergic nature of the evoked current, 40 μm gabazine, a nonselective GABAAR antagonist, was applied at the end of each GBX dose series. The peak amplitude of the GABAAR tonic current was measured with reference to current recorded following gabazine application. Most MGB neurons showed current saturation at the maximum (10 μm) GBX concentration used in these experiments. Higher GBX concentrations were not used to avoid concomitant activation of synaptic GABAARs.
We recorded from 32 MGB neurons from 10 Ctrl rats, 43 neurons from 14 Tin rats, and 11 neurons from 4 Non-T rats. All recordings were contralateral to the sound exposure and were parsed by subdivision into recordings from dMGB and vMGB (dMGB/vMGB, respectively: Ctrl, 16/16 neurons; Tin, 20/23 neurons; Non-T, 5/6 neurons).
Repeated-measures ANOVA revealed significant differences among three experimental groups in the dose–response amplitude of GABAAR tonic currents, measured with bath application of 0.1–10 μm GBX (F = 3.15; p = 0.048; Fig. 6). There was significant interaction between GBX concentrations and groups (p = 0.037). Additional t test analysis demonstrated significant increases in the amplitude of GABAAR tonic currents for the Tin group compared with the Ctrl group at 1–10 μm GBX concentrations (1 μm GBX, p = 0.034; 2 μm GBX, p = 0.022; 5 μm GBX, p = 0.012; 10 μm GBX, p = 0.019; Fig. 6). No differences were found between the Non-T group and either the Ctrl or Tin groups. Figure 6D shows the significant 25.5% increase in mean current amplitude in the Tin group compared with the Ctrl group at 10 μm GBX (Ctrl, 217.11 ± 14.86 pA; Tin, 272.54 ± 16.66 pA; Non-T, 259.08 ± 20.02 pA). Likewise, a 19.1% greater current density occurred in the Tin group (p = 0.02) relative to Ctrl (Ctrl, 1.83 ± 0.11 pA/pF; Tin, 2.18 ± 0.08 pA/pF; Non-T, 2.09 ± 0.1 pA/pF; Fig. 6E). No significant differences were identified between the Non-T group and either the Ctrl or Tin groups at a 10 μm GBX concentration. GABAAR tonic currents were comparable in dMGB and vMGB across groups [e.g., the 25% tinnitus-related increase in maximal current was detected in dMGB and vMGB at 10 μm GBX, but only reached significance in dMGB (p = 0.046) when comparing current density in the smaller parsed groups].
Extrasynaptic GABAAR tonic inhibitory currents in MGB neurons increased in tinnitus rats. A, Sample recording from an MGB neuron showing increasing tonic GABAA inhibitory current in response to bath application of increasing GBX doses (0.1–10 μm). The magnitude of this current is revealed by GABAAR blockade with gabazine (40 μm). B, Dose–response semi-log plot of tonic GABAA current mean amplitudes in control (n = 32 cells), sound-exposed nontinnitus (n = 11 cells), and sound-exposed tinnitus (n = 43 cells) groups of rats. Data were fitted with the logistic equation: y = Imin + (Imin − Imax)/(1 + (x/xh)p), where Imin and Imax are the minimal and maximal measured responses, respectively, xh is the half-maximal effective concentration, and p is the absolute value of the slope of the curve. Tin rats showed significantly larger responses to increasing GBX doses. Data are presented as the mean ± SEM. C, D, For the highest dose tested, the maximal evoked current was significantly (*p < 0.05) different between control and tinnitus rats. Box charts for the highest dose show mean (square), median (middle line), 95% range (box limits), and SD (error bars). All recordings were performed contralaterally to noise exposure.
When the dose–response curves were normalized to the maximal current recorded with 10 μm GBX, a simple logistic function was used to fit dose–response data from which the EC50 and activation slope for GABAAR tonic current were calculated for each individual neuron. Importantly, no decline in GABAAR function was observed for any of the functional parameters examined. No significant group differences were seen for mean EC50 (Ctrl, 2.60 ± 0.36 μm; Tin, 1.95 ± 0.21 μm; Non-T, 2.29 ± 0.49 μm) or activation slope (Ctrl, 1.6 ± 0.06; Tin, 1.61 ± 0.05; Non-T, 1.42 ± 0.06; Fig. 6C). Since no group differences were found in current kinetics, it was concluded that tinnitus pathology does not substantially alter the subunit makeup of extrasynaptic GABAARs.
This tinnitus-associated augmentation of GBX-evoked currents suggests an increase in the number of extrasynaptic GABAARs. The significant increase in tonic currents in Tin animals differs dramatically from the near 50% loss of similarly evoked tonic currents recorded previously from aged MGB neurons (Richardson et al., 2013).
δ-Subunit expression is upregulated contralateral to sound exposure in Tin group
To assess the possibility of GABAAR subunit changes underpinning the tinnitus-related increases in MGB neuronal GABAAR tonic currents, in situ hybridization was used to examine the expression of the GABAAR α4- and δ-subunits, markers of tonic–current extrasynaptic GABAAR constructs. Previous radioligand binding and immunohistochemical studies using [3H]gaboxadol and antibodies for α4- and δ-subunits suggested exceptionally high levels of the α4δ-subunit containing extrasynaptic GABAARs in MGB (Richardson et al., 2012; Richardson et al., 2013). Similarly, the present in situ hybridization study found high expression levels of α4δ GABAAR subunits throughout MGB (Fig. 7).
Tinnitus-related increase in MGB neuronal gene expression levels for α4 and δ GABAAR subunits. A, Silver grains representing α4 GABAAR subunit mRNA are seen over neurons in the vMGB. Note the higher numbers of grains over neurons from sound-exposed animals. Scale bar, 10 μm. B, Histogram shows α4 and δ MGB neuron GABAAR subunit mRNA changes (grain counts) from control (n = 7) and sound-exposed (n = 6) FBN rats. GABAAR δ-subunit gene expression showed significant tinnitus-related increases both contralateral (Contra) and ipsilateral (Ipsi) to the sound exposure (p < 0.001). MGB changes in δ gene expression were observed both in dorsal and ventral MGB subdivisions. GABAARs formed by δ-subunits are associated with extrasynaptic tonic inhibitory currents and extrasynaptic locations of GABAARs. Interestingly, the α4 mRNA level in MGB was significantly increased only ipsilateral to the sound exposure (p < 0.001).
Again, contrary to what was predicted, t test hypothesis testing revealed a significant increase in mRNA expression of the GABAAR δ-subunit contralateral to sound exposure both in dMGB and vMGB (dMGB: Ctrl, 2.9 ± 0.34 silver grains/100 μm2; Tin, 5.6 ± 0.27 grains/100 μm2; p < 0.001; vMGB: Ctrl, 3.0 ± 0.29 grains/100 μm2; Tin, 5.5 ± 0.22 grains/100 μm2; p < 0.001). Replacing the γ-subunit with the δ-subunit results in GABAARs with extrasynaptic receptor location and tonic inhibitory currents. Figure 7 summarizes changes in α and δ GABAAR subunit mRNA levels associated with behavioral evidence of tinnitus. These gene expression data support the patch-clamp findings described above, suggesting that tinnitus pathology may result in an increase in extrasynaptic GABAAR density. Interestingly, α-containing GABAARs showed significant tinnitus-related mRNA increases in MGB ipsilateral to the sound exposure (dMGB: Ctrl, 6.0 ± 0.2 grains/100 μm2; Tin, 9.0 ± 0.25 grains/100 μm2; p < 0.001; vMGB: Ctrl, 5.5 ± 0.21 grains/100 μm2; Tin, 9.7 ± 0.25 grains/100 μm2; p < 0.001) with only modest, nonsignificant increases contralateral to the sound exposure (Fig. 7).
Ambient GABA and resting membrane potential of MGB neurons
Given the presence of high-affinity extrasynaptic GABAARs on MGB neurons, it is feasible that these receptors would be tonically active in the presence of low concentrations of ambient GABA in the extracellular space. We evaluated the possible role of endogenous tonic GABAAR currents on resting membrane potential and firing properties of MGB neurons, with bath application of gabazine. MGB neurons from Ctrl and Tin groups were recorded in current-clamp mode with no correcting currents applied to clamped neurons, and blockers of glutamatergic synaptic transmission omitted from the recording solution. Ten minutes after obtaining whole-cell configuration, baseline resting membrane potential was recorded for 5–10 min followed by recording in aCSF solution with gabazine (40 μm) for another 15 min. The membrane potential change (ΔV) calculated by subtracting Vgabazine (10 min following gabazine application) from Vrest (measured immediately before gabazine application). Vrest and Vgabazine were mean values calculated over a period of 1 min. Bath application of gabazine to MGB slices blocked synaptically mediated IPSPs, and in some recordings resulted in a small depolarizing shift (3–4 mV) in Vrest. However, the depolarizing shift of gabazine on baseline Vrest was not significant in both the Ctrl and Tin groups.
Synaptic sIPSCs in auditory thalamus are not affected by sound exposure
We further examined whether GABAergic changes at extrasynaptic receptors were associated with changes in synaptic GABA transmission. For this purpose, we compared sIPSCs in MGB neurons in Ctrl and Tin animal groups. sIPSC activity was recorded for 4–5 min from MGB neurons contralateral to sound exposure. Recordings were from the same set of neurons in which extrasynaptic GABAAR currents were tested before GBX dose–response experiments.
Typical recordings of sIPSCs from MGB neurons are demonstrated in Figure 6A. Recordings were made at +20 mV to maximize outward inhibitory Cl− current in aCSF containing glutamate receptor antagonists (20 μm CNQX, 50 μm d-AP5). sIPSC ANOVA analysis revealed no significant group difference in phasic GABAA synaptic event frequency. Mean sIPSC frequencies were as follows: Ctrl, 4.64 ± 0.42 Hz (n = 32); Tin, 5.28 ± 0.41 Hz (n = 43); Non-T, 5.25 ± 0.67 Hz (n = 11). Furthermore, neither sIPSC amplitude (peak amplitude: Ctrl, 22.32 ± 1.54 pA, n = 32; Tin, 21.03 ± 0.59 pA, n = 43; Non-T: 18.64 ± 1.43 pA, n = 11) nor sIPSC duration (half-width: Ctrl, 12.70 ± 0.51 ms, n = 32; Tin, 12.29 ± 0.50 ms, n = 43; Non-T, 12.69 ± 0.71 ms, n = 11) demonstrated significant tinnitus-related changes. Cumulative probability plots and corresponding Kolmogorov–Smirnov statistical analysis further confirmed no significant differences in sIPSC properties, among any of the three groups (Fig. 8B). Adding 40 μm gabazine to the aCSF medium abolished sIPSCs in all recordings, confirming GABAAR mediation of the sIPSCs.
GABAergic synaptic inhibition is not altered in MGB neurons of tinnitus rats. A, Representative traces of sIPSC recordings from MGB neurons from the Ctrl, Non-T, and Tin groups. Recordings were performed at +20 mV in the presence of 20 μm DNQX and 50 μm AP-5. Bath application of gabazine (GBZ) blocked all inhibitory GABAAR-mediated synaptic events. B, Averaged cumulative probability plots of sIPSC interevent intervals, amplitude, and half-width of MGB neurons in the Ctrl and Tin groups. Statistical analysis revealed no significant difference in these three GABAergic synaptic event parameters between the Ctrl and Tin groups (Ctrl, n = 29 cells; Tin, n = 29 cells). All recordings were contralateral to the sound exposure. Data are reported as the mean ± SEM.
Discussion
In the present study, we examined two hypotheses regarding GABAergic inhibitory function in the MGB that could underpin tinnitus produced by a peripheral acoustic insult. In MGB, either upregulation or downregulation of inhibitory function could underlie tinnitus-related thalamocortical hyperactivity and burst alterations in projections to AI, TRN, and amygdala (Kalappa et al., 2014).
Downregulation of inhibitory neurotransmission within the central auditory system is characteristic of partial peripheral deafferentation and is one feature associated with tinnitus-related pathology at multiple levels of the auditory neuraxis in different animal models (Caspary et al., 2008; Noreña, 2011; Gold and Bajo, 2014). Reduced inhibition is associated with abnormal neuronal hyperactivity, increased spontaneous firing rates, and, in some cases, bursting and synchronous oscillations. In DCN, tinnitus is associated with decreased inhibition, resulting in increased spontaneous and driven activity of projection neurons to IC (Brozoski et al., 2002; Wang et al., 2009; Roberts et al., 2010; Dehmel et al., 2012). Wang et al. (2009) found decreased neurochemical markers for glycinergic inhibition in DCN that was supported by patch-clamp studies showing decreased glycinergic function following sound exposure (Pilati et al., 2012b). In IC, markers for GABAergic neurotransmission were downregulated in models of tinnitus (Dong et al., 2010a,b; Gold and Bajo, 2014). Deafferentation and tinnitus models show increased spontaneous activity, peak spike rates within bursts, and rhythmicity of spiking in IC neurons (Bauer et al., 2008). Tinnitus-related downregulation of inhibition in the AI is thought to involve cortical receptive field reorganization, and cortical hyperactivity and bursting (Mühlnickel et al., 1998; Lanting et al., 2009; Yang et al., 2011; Brozoski et al., 2012).
Until recently, the role of GABAergic inhibition has not been thoroughly examined in MGB. MGB neurons function to filter and enhance incoming representation and perception of auditory features (Edeline, 2011; Bartlett, 2013). MGB engages high-level processing of auditory information through thalamocortical/corticothalamic interactions (Sherman, 2012). Winer et al. (2005) noted that descending cortical axons projecting to thalamus are more numerous than ascending projections (Jones, 1975; Ojima and Rouiller, 2010; Sherman, 2012). MGB has intricate connections with the limbic system that are likely involved in the emotional aspect of tinnitus (Sah et al., 2003; LeDoux, 2007; Kraus and Canlon, 2012). A potential role of MGB in tinnitus pathology has been outlined by Leaver et al. (2011) and Rauschecker et al. (2010), and the role of GABA in the pathology of chronic pain and tinnitus at the level of thalamus is integral to the thalamocortical dysrhythmia hypothesis (Llinás et al., 1999, 2005; Llinás and Steriade, 2006).
GABAergic inhibition plays a critical role in shaping MGB neuronal response properties (Cotillon-Williams et al., 2008; Cai et al., 2014; Mellott et al., 2014). Apart from synaptic GABAARs that mediate fast signal transmission, MGB neurons show robust slow tonic inhibitory currents, mediated by α4β2δ-containing GABAARs (Richardson et al., 2011, 2013; Yagüe et al., 2013). These high-affinity extrasynaptic GABAARs appear to detect fluctuations in ambient GABA levels and may mediate ∼90% of total inhibitory current in sensory thalamus (Belelli et al., 2005; Cope et al., 2005). Both hypotheses regarding GABAergic dysfunction in tinnitus-related pathology in MGB suggest altered GABAergic function related to deafferentation, but differ in the direction of tinnitus-related changes. DCN, IC, and AI circuits in tinnitus models show downregulation of inhibitory neurotransmission. Neurons in these structures show increased spontaneous activity, abnormal sound-evoked activity, spontaneous neuronal hyperactivity, and decreased neurochemical markers of GABA and/or glycine.
Here we examined whether the loss or gain of GABAergic inhibition, underpinning thalamocortical dysrhythmia (Llinás et al., 1999, 2005; Llinás and Steriade, 2006), is the operative condition in MGB tinnitus pathology. The thalamocortical dysrhythmia hypothesis suggests that increased GABAergic inhibition leads to deinactivation of T-type Ca+2 channels for a subset of thalamocortical neurons, resulting in increased/abnormal bursting and increased MGB output (Kalappa et al., 2014). The present study found no evidence to support a downregulation of GABAergic inhibitory function in MGB. We show for the first time that 2 months following a tinnitus-inducing traumatic sound exposure, thalamocortical MGB neurons contralateral to sound exposure showed significantly increased duration of evoked bursting and increased gaboxadol-evoked tonic currents. The increase in maximum gaboxadol-evoked tonic current was evident in both the vMGB and dMGB, and suggests increased expression/density of functional extrasynaptic GABAARs. This GBX-evoked increase in tonic current does not appear to be a simple consequence of the sound exposure since tonic GABAAR currents recorded in Non-T animals were not significantly different from Ctrl group. Considering that tonic extrasynaptic GABAergic inhibition may account for ∼80–90% of total inhibition in thalamus (Belelli et al., 2005; Cope et al., 2005), this moderate increase in GABAA tonic current in MGB neurons could significantly alter the response properties of thalamocortical neurons exposed to ambient GABA or GABA from synaptic spillover, altering auditory gating (Kalappa et al., 2014).
In support of our patch-clamp results, a significant tinnitus-related increase in GABAAR δ-subunit mRNA level was found contralateral to the sound exposure. δ-Subunits are associated with extrasynaptic GABAARs and tonic inhibition. The tinnitus-related upregulation of GABAAR tonic inhibition markers are in stark contrast to a recent study (Richardson et al., 2013) that showed a near 50% age-related loss of functional and neurochemical markers of extrasynaptic GABAARs in MGB.
It is also plausible that α5-subunit-containing GABAARs could contribute to tinnitus-related pathology by increasing tonic inhibition observed in MGB. We feel that a tinnitus-related upregulation is unlikely to account for the observed changes in the present study since α5-subunit expression in thalamus is low compared with that in δ-containing GABAARs (Sur et al., 1999; Pirker et al., 2000). It is especially low when expression levels in hippocampus and cortical layer 5 are compared (Pirker et al., 2000; Atack et al., 2005; Walker and Semyanov, 2008). In contrast, MGB has some of the highest expression levels of α4δ-receptors in the brain (Richardson et al., 2011; Wafford, 2014). Gaboxadol is a selective agonist for α4δ-containing extrasynaptic GABAARs at the concentrations used in the present study (Belelli et al., 2005). The extrasynaptic action of gaboxadol is supported by its lack of significant effect on spontaneous or miniature IPSCs in MGB neurons. Future studies will examine the role of α5-containing constructs in the pathology of tinnitus.
The kinetics of gaboxadol-evoked extrasynaptic GABAAR currents were not different among experimental groups, suggesting that GABAAR extrasynaptic subunit composition may not be altered in animals with behavioral evidence of tinnitus.
Analysis of synaptic currents, sIPSCs, revealed no significant differences among groups. This suggests that synaptic MGB GABAA neurotransmission was not downregulated in tinnitus-related pathology. These data differ from findings with sodium salicylate-induced tinnitus that demonstrated reduced evoked synaptic transmission in MGB (Su et al., 2012).
In the present study, passive membrane properties of MGB neurons, including resting membrane potential, were not affected by tinnitus-related pathology. While extrasynaptic GABAARs can be activated at low ambient concentrations of GABA, we were unable to consistently demonstrate an endogenous tonic current with GABAAR blockade. The lack of significant gabazine effect on MGB neuron Vrest could be a function of aCSF flow rate in our slice preparation, resulting in low exogenous concentrations of ambient GABA. All recordings were made at room temperature, which could negatively impact ambient GABA levels (Iversen et al., 1968). A similar lack of endogenous GABAAR tonic current was extensively examined for hyperglossal motor neurons (Numata et al., 2012). The present study did not test the possibility that changes in GABA uptake (GABA transporter, GAT1 or GAT3/4) could underpin tinnitus-related pathology in light of studies in models of absence epilepsy, where changes in GAT1 function have been described (Pirttimaki et al., 2013; Walker and Pavlov, 2014). Future studies will examine changes in GAT function.
The majority of MGB neurons responded to somatic depolarization with burst firing in Ctrl and Tin groups. However, neurons from Tin animals fired significantly more spikes per burst relative to control MGB neurons, suggesting a tinnitus-related increase in intrinsic membrane excitability, perhaps attributed to altered K+, Na+, or Ca2+ conductance. The tinnitus-related increase in the number of spikes per burst was more common in vMGB, which also showed a trend toward an increase in the number of bursting neurons. The present in vitro findings are consistent with the recent in vivo single-unit findings from awake animals showing tinnitus-related increases in burst properties of MGB neurons that directly correlated with an increase in tinnitus severity score (Kalappa et al., 2014). Animals used in the study by Kalappa et al. (2014) were similarly sound exposed and behaviorally tested.
Thalamic neurons can switch from tonic firing to bursting (Llinás and Jahnsen, 1982; Carbone and Lux, 1984; Jahnsen and Llinás, 1984). Bursting is elicited when thalamic neurons are hyperpolarized and is supported by the activation of low-voltage T-type calcium channels that, upon activation, generate LTSs. However, there was no evidence found to support the LTS difference between the Ctrl and Tin groups. An increase in GABAergic inhibition might indirectly affect Ca2+ balance and, in turn, modulate potassium channels. Brief activation of GABAergic inhibitory synapses in hippocampus results in elevated neuronal excitability by modulating intracellular calcium level and reducing the sensitivity of calcium-activated potassium channels (Nelson et al., 2003). A tinnitus-related increase in tonic GABAA inhibition in MGB might similarly, but indirectly, increase MGB neuron intrinsic excitability. Future studies will examine the role of tinnitus-related changes of ion channel conductances underpinning increased burst parameters of MGB neurons.
While a large body of literature suggests that inhibition is reduced in tinnitus pathology, our results in MGB support the possibility that excessive tonic inhibition in MGB may be involved in thalamocortical dysrhythmia (Llinás et al., 2005; De Ridder et al., 2015). Recent data supporting the thalamocortical dysrhythmia hypothesis come from a human magnetoencephalography study (Adjamian et al., 2012) that found increased oscillation in tinnitus patients in the delta band. Thalamocortical bursting mode of operation in local circuitry during vigilance leads to increase in a low-frequency oscillation. This creates the phenomenon of an edge effect when low-frequency firing may lead to cortical disinhibition (Llinás et al., 2005).
Enhanced tonic inhibition in the MGB neurons is one of many factors possibly initiating thalamocortical dysrhythmia. The shift to abnormal periodic bursting activity is hypothesized to be related to partial deafferentation of specific sensory thalamic nuclei and possibly reduced excitation, which could result in thalamic neurons becoming hyperpolarized (Llinás et al., 2005). In turn, reduced ascending thalamic input could lead to a decrease in the excitatory drive/input to the TRN and reduced firing rates at the cortical level. Apart from the observed tinnitus-related increase in tonic inhibitory tone, other changes in thalamic neuronal membrane currents could mediate abnormal hyperpolarization associated with bursting activity. For example, resting membrane potential of thalamic neurons depends on the activity of two-pore-domain potassium (KCNK) channels, which are major targets for endogenous modulators (Talley et al., 2001; Musset et al., 2006). An increase in KCNK resting potassium current due to modulation and concomitant membrane hyperpolarization could effectively shift thalamic neurons into burst mode.
Another candidate for abnormal increased thalamocortical neurons bursting is hyperpolarization-activated cation (HCN) channels. HCN channels act as pacemaker channels where HCN current produces a depolarization of the membrane potential followed by LTS. Mice lacking HCN2 channels suffer from absence epilepsy, probably due to the increased responsiveness of thalamocortical neurons (Ludwig et al., 2003). Lack of HCN2 channels leads to hyperpolarization of thalamocortical neurons, promoting the recovery of T-type calcium channels from inactivation. This results in increased burst firing and oscillatory activity. Correspondingly, abnormal reduction in HCN current in thalamocortical neurons could be a mechanism generating thalamocortical dysrhythmia. Finally, an alteration in T-type calcium channel kinetics similarly modulates the bursting activity of thalamic neurons. It would be important to further explore other functional and molecular mechanisms at the MGB level involved in the thalamocortical dysrhythmia model and in tinnitus pathology.
In conclusion, the present study demonstrated that contrary to findings in other central auditory structures that have a tinnitus-related downregulation of functional and neurochemical markers of inhibitory neurotransmission, MGB circuits show no reduction of function. Rather, there was evidence of increased tonic GABAAR currents and an increased number of spikes per burst in animals with behavioral evidence of tinnitus. These results agree with, and partially support, the thalamocortical dysthymia model of tinnitus.
Footnotes
This research was supported by Office of Naval Research Grant N00014-12-1-0214 and National Institute on Deafness and Other Communication Disorders Grant DC-000151. We thank Drs. Tom Brozoski, Brandon Cox, and Steve Sandstrom for their careful reading and comments on the manuscript.
J.G.T. is the cofounder of OtoScience Labs, which conducts hearing and tinnitus research, and licenses the gap-based testing technology from Southern Illinois University School of Medicine. The authors declare no other competing financial interests.
- Correspondence should be addressed to Donald M. Caspary, Southern Illinois University, School of Medicine, Department of Pharmacology, 801 North Rutledge Street, P.O. Box 19629, Springfield, IL 62702. dcaspary{at}siumed.edu
