Exp Neurol. 2017 Jun;292:1-10.

Saturation of longterm potentiation in the dorsal cochlear nucleus and its pharmacological reversal in an experimental model of tinnitus.

Tagoe T, Deeping D, Hamann M.

Department of Neurosciences, Psychology and Behaviour, University of Leicester, UK.



Animal models have demonstrated that tinnitus is a pathology of dysfunctional excitability in the central auditory system, in particular in the dorsal cochlear nucleus (DCN) of the brainstem. We used a murine model and studied whether acoustic over-exposure leading to hearing loss and tinnitus, affects longterm potentiation (LTP) at DCN multisensory synapses. Whole cell and field potential recordings were used to study the effects on release probability and synaptic plasticity, respectively in brainstem slices. Shifts in hearing threshold were quantified by auditory brainstem recordings, and gap-induced prepulse inhibition of the acoustic startle reflex was used as an index for tinnitus. An increased release probability that saturated LTP and thereby induced metaplasticity at DCN multisensory synapses, was observed 4-5days following acoustic over-exposure. Perfusion of an NMDA receptor antagonist or decreasing extracellular calcium concentration, decreased the release probability and restored LTP following acoustic over-exposure. In vivo administration of magnesium-threonate following acoustic over-exposure restored LTP at DCN multisensory synapses, and reduced gap detection deficits observed four months following acoustic over-exposure. These observations suggest that consequences of noise-induced metaplasticity could underlie the gap detection deficits that follow acoustic over-exposure, and that early therapeutic intervention could target metaplasticity and alleviate tinnitus.

KEYWORDS: Auditory; Central auditory system; Longterm potentiation; Release probability; Synapse; Synaptic plasticity

Full text 



Tinnitus, the phantom perception of sound in the absence of a corresponding acoustic stimulus affects 10 to 15% of the adult population worldwide. Despite this prevalence, there are currently no effective drug therapies due to limited understanding of the mechanisms that underlie tinnitus onset following triggers such as exposure to loud sound. Many studies support the prevailing idea that tinnitus arises in response to enhanced multisensory synaptic transmission to the dorsal cochlear nucleus (DCN), the first relay in the auditory brainstem integrating acoustic and multisensory inputs. Such findings include a correlation between tinnitus and aberrant neural activity in the DCN (Kaltenbach 2007), altered excitability in the DCN initiating a complex sequence of events in the auditory pathway (Brozoski et al., 2002) and an increased synchrony in the network activity (Shore et al., 2016). We investigate the role of plasticity in the DCN at the early stages following acoustic overexposure and address the question whether targeting identified plastic changes can help alleviating symptoms of tinnitus in a rodent model.

We use a combination of in vitro field potential recordings (Fig 1A) and whole cell recordings (Fig. 1E) in slices containing the DCN to measure an increased release probability following acoustic overexposure (Fig. 1B inset,F-H). Long-term potentiation normally induced by a high frequency stimulation, is absent following acoustic overexposure (Fig 1C). We next test whether the absence of LTP after acoustic over-exposure is due to an increased release probability (Fig.1D, 1F). Elevating extracellular calcium increases release probability and abolishes LTP in the control conditions (Fig. 2A), whereas decreasing extracellular calcium decreases release probability in the acoustic overexposure condition and restores LTP (Fig. 1D). In each of these instances, an increase (Fig. 1D inset) or a decrease (Fig. 2A) in paired pulse facilitation provide evidence of a decrease or an increase release probability respectively.

NMDA receptors are known to be essential triggers for LTP at many excitatory synapses. In vitro perfusion of the NMDA receptor antagonist – DAP5 after HFS abolishes the LTP (Fig. 2D). Conversely, perfusion of NMDA in the control condition (Fig. 2B) or of TBOA, a glutamate transporter blocker (Fig. 2C) abolish LTP, providing strong evidence that increasing release probability of glutamate activating NMDA receptors is essential to mediating the deficits induced by acoustic overexposure (Fig. 2E,F).

Magnesium is an endogenous blocker of NMDA receptors and previous studies have shown that magnesium-threonate effectively alters synaptic plasticity and learning (Slutsky et al. 2010). To translate our findings into an in vivo model, we used magnesium threonate, a compound which has previously been shown to effectively elevate the interstitial magnesium concentration (Slutsky et al. 2010). Rats were put on a high magnesium diet using magnesium-threonate, and field potentials were recorded four weeks after acoustic overexposure. In this condition, high magnesium diet restores the induction of LTP (Fig 3A). An increased paired pulse facilitation is shown following the diet (Fig. 3A) indicating a decrease in the basal release probability essential to LTP induction (Fig 3B).

To test the effect of magnesium-threonate diet on tinnitus, we used the gap induced pre-pulse inhibition of the acoustic startle reflex (Turner et al. 2006, Fig. 4A) which provides links to frequency discrimination deficits observed in tinnitus (Fournier and Hebert 2013). We tested whether a magnesium-threonate diet allows recovery from gap detection deficits induced by acoustic over-exposure. Animals were put on the diet immediately following acoustic over-exposure. Eighteen weeks after acoustic over-exposure, the high magnesium diet significantly reduced the gap detection deficits which is indicative of a decreased perception of tinnitus (Fig 4B,C) in this rodent model.

Altogether, we have demonstrated a pathological deficit in plasticity in the DCN that occurs in the early days after acoustic overexposure. We also showed that a diet containing Mg2+ threonate prevents the tinnitus onset when administered at the early stages following acoustic over-exposure.



Fig1. Acoustic overexposure increases release probability and abolishes the induction of LTP. (A) Schematic representation of the dorsal cochlear nucleus showing the excitatory (green triangles) and inhibitory connections (red triangles) between the major cell types: granule cells (GC), cartwheel cells (CwC) and fusiform cells (FC). Stimulation of the multisensory inputs elicits field potentials which are captured by the recording electrode placed in the fusiform cell layer. (Inset) Sample field potential recordings are shown in in normal extracellular medium (black) and in the presence of 10 µM NBQX (red) which blocks AMPA receptors and the postsynaptic field potential (PSFP, N2) without affecting the presynaptic component of the field potential (N1). (B) High frequency stimulations (red arrow) induce LTP as the normalized PSFP amplitudes increase from 1.00 ± 0.16 to 1.39 ± 0.09 (n = 7, N = 5; Z = 3.46, P < 0.001, Wilcoxon). Perfusion of NBQX (10 µM) abolishes the PSFP (and the LTP). (Inset left) Sample traces showing the presence of paired pulse facilitation before high frequency stimulation, and its absence after high frequency stimulations. N2 and N2’ represent the first and the second PSFP respectively elicited at a 60 ms pulse interval. The dashed line indicates the amplitude of the first PSFP (N2). (Inset right). Summary histograms representing the decreased paired pulse ratios during LTP, from 1.32 ± 0.08 before high frequency stimulations to 1.11 ± 0.13 during LTP (n=7, N=5; Z = -2.7, P = 0.004, Wilcoxon). (C) High frequency stimulations (arrow) induce LTP of the PSFPs in unexposed conditions, increasing PSFP (N2) amplitudes by 39 ± 09 % (black circles, n = 16, N = 15; Z =3.46, Wilcoxon P < 0.001). High frequency stimulations fail to elicit LTP after acoustic over-exposure (normalised PSFP amplitude of 0.96 ± 0.06; n = 20, N = 13, Z = -0.63, Wilcoxon P = 0.55, blue squares, *** P < 0.001, U = 47, Mann Whitney test). (D) LTP can be induced following acoustic overexposure by lowering Ca2+ concentration to 1 mM (normalized PSFP amplitudes increasing from 1.01 ± 0.01 to 1.27 ± 0.09 after high frequency stimulations in presence of 1 mM Ca2+, n = 11, N=5; Z = 2.93, P = 0.02, Wilcoxon). PSFP amplitudes are significantly higher than values recorded 30 mins after high frequency stimulation in 2 mM Ca2+ (n = 20, N = 13, U = 47, * P = 0.01 Mann Whitney). (Inset) Paired pulse ratios increase from 0.97 ± 0.07 to 2.45 ± 0.29 (n = 4, N = 2, Wilcoxon, P < 0.01) when extracellular Ca2+ is decreased from 2 mM to 1 mM (E) Schematic representation of the dorsal cochlear nucleus (same as A) showing the recording of EPSCs from fusiform cells. (Inset) Sample EPSC recorded in normal extracellular medium (black) or in the presence of NBQX (red). (F, left) EPSC variance-mean plots fitted with a parabola function (unexposed: black circle; acoustic over-exposure: blue square). Acoustic over-exposure increases the release probability (t(16)=-2.6  P = 0.019, unpaired t test) while leaving the quantal size (q; unexposed: 2.6 ± 0.4; exposed: 2.1 ± 0.4; t(16) = 0.8; P = 0.43, unpaired t test) and the number of release sites (n; unexposed: 365 ± 78; exposed: 257 ± 89; t(16) = 0.9; P = 0.38, unpaired t test) unaffected. Unexposed: n = 9, N=9; Exposed: n = 9, N=7. Schematic representation illustrating the modulation of LTP at DCN multisensory synapses in the unexposed (G) and exposed (H) conditions.



Fig 2: Elevating release probability under control conditions abolishes LTP induction. (A left) In the unexposed condition, LTP is induced in 2mM Ca2+ and absent in 3 mM Ca2+ (Mann Whitney test, P <0.05). (A right) Paired pulse facilitation is absent in 3 mM Ca2+ (3 mM Ca2+: n = 5, N = 4; 2 mM Ca2+: n = 10, N = 4; U = 10; ** P = 0.002, Mann Whitney). (B left) Absence of LTP in presence of 500 nM NMDA (n = 5, N = 3). (B right) NMDA decreases paired pulse ratios before high frequency stimulation (n = 5, N=3; x r2 (2) = 8.4, P = 0.01, RM ANOVA on Ranks) and after high frequency stimulations. (C left) Absence of LTP in presence of 10 μM TBOA (n = 5, N=3; x r2 (2) = 2.8, NS, RM ANOVA on Ranks). (C right) Similar to ‘B’, paired pulse ratios are decreased when tested in the presence of TBOA before high frequency stimulation (n = 5, N=3; x r2 (2) = 7.6, P = 0.01, RM ANOVA on Ranks), and after high frequency stimulation. (D left) Induction of LTP is abolished by D-AP5 (25 µM), returning PSFP amplitudes to baseline levels as PSFP normalised amplitudes decrease from 1.58 ± 0.17 to 1.02 ± 0.18 (n = 8, N=6, Z = -2.02, P = 0.04, Wilcoxon). (D right) Histograms showing the increased paired pulse ratios in the presence of 25 µM AP5 (from 1.17 ± 0.07 to 1.67 ± 0.19, n = 8, N = 6; Z = 2.24, P = 0.023). (E) Schematic representation showing that D-AP5 maintains a low release probability which can be increased by high frequency stimulations. (F) Schematic representation illustrating that 3 mM Ca2+, NMDA and TBOA elevate the release probability and prevent LTP induction by high frequency stimulations.


Fig 3: A diet rich in magnesium threonate restores LTP following acoustic over-exposure. (A left) LTP induction by high frequency stimulations is restored when Mg2+ is administered immediately after acoustic over-exposure and for a period of 4 weeks (Exposed (blue squares), n = 6, N = 4; Z = 0.9, NS, Wilcoxon; Exposed + Mg2 (green diamonds), n= 12, N=5; Z =2.46, P = 0.008, Wilcoxon). (*) PSFP amplitudes measured 30 minutes after high frequency stimulations are different in the two conditions (Mann Whitney, U = 10, P = 0.024).  (A right) In vivo administration of Mg2+-threonate after acoustic overexposure restores in vitro paired pulse facilitation under basal stimulating conditions (n = 6, N=4; t (5) = 3.5, P = 0.02, paired t test). Paired pulse ratios in the “Exposed + Mg2+” group are higher than those recorded in the “Exposed” group (Exposed: n = 8, N = 6; Exposed + Mg2+: n = 6, N = 4; Mann Whitney, U = 8, * P < 0.05). (B) Schematic representation illustrating how a high basal release probability in the exposed condition prevents LTP induction. A high magnesium diet (below) maintains a low basal release probability thereby allowing the induction of LTP.



Fig 4: A diet rich in magnesium threonate abolishes the effects of acoustic over-exposure on tinnitus. (A top panel) In control (unexposed) condition, a startle sound (broadband, 110dB SL) embedded in a 75 dB SPL background tone triggers a startle response. A gap embedded in the background tone (lower trace) triggers a smaller startle response leading to startle response ratios (Gap/No Gap) of less than 1. (A bottom panel) Tinnitus fills in the gap (schematised in grey), decreasing its perception and leading to similar startle responses in presence and in absence of gaps (Gap/No Gap) equals 1 (the affected frequency  is 10 KHz in the example shown). (B) Eighteen weeks after acoustic over-exposure, gaps are detected when embedded in broadband noise (BBN) or in 8 kHz background sound but remain undetected when embedded in a 16 kHz background sound (N=9). Administration of Mg2+ following acoustic over-exposure prevents gap discrimination deficits at 16 kHz (N=9, * P < 0.05 One Way ANOVA). (C) Gap detection ratios for 8 kHz, 16 kHz and BBN calculated for week 0 (left symbols) and week 18 (right symbols) show that Mg2+ administration prevents gap discrimination deficits otherwise present at 16 KHz at week 18 after acoustic over-exposure (* P < 0.05, linear mixed model, pairwise comparison; unexposed: N=7; Exposed: N=9; Exposed+Mg2+: N=9).





Brozoski, T.J., Bauer, C.A., Caspary, D.M., 2002. Elevated fusiform cell activity in the dorsal cochlear nucleus of chinchillas with psychophysical evidence of tinnitus. J Neurosci 22, 2383-2390.

Fournier P, Hébert S. 2013. Gap detection deficits in humans with tinnitus as assessed with the acoustic startle paradigm: does tinnitus fill in the gap? Hear Res. 295, 16-23.

Kaltenbach, J.A., 2007. The dorsal cochlear nucleus as a contributor to tinnitus: mechanisms underlying the induction of hyperactivity. Prog Brain Res 166, 89-106.

Shore S.E., Roberts L.E., Langguth B., 2016. Maladaptive plasticity in tinnitus–triggers, mechanisms and treatment. Nat Rev Neurol. 12, 150-160.

Slutsky, I., Abumaria, N., Wu, L.J., Huang, C., Zhang, L., Li, B., Zhao, X., Govindarajan, A., Zhao, M.G., Zhuo, M., Tonegawa, S., Liu, G., 2010. Enhancement of learning and memory by elevating brain magnesium. Neuron 65, 165-177.

Turner, J.G., Brozoski, T.J., Bauer, C.A., Parrish, J.L., Myers, K., Hughes, L.F., Caspary, D.M., 2006. Gap detection deficits in rats with tinnitus: a potential novel screening tool. Behavioral Neuroscience 120, 188-195.