Tinnitus, die Wahrnehmung eines Phantomgeräuschs, geht in den meisten Fällen mit Hörverlust einher. Es ist jedoch unbekannt, wie Hörverlust zu Tinnitus führen könnte. In Tierversuchen wurde gezeigt, dass Verhaltensanzeichen für Tinnitus nach Hörverlust mit erhöhten spontanen Feuerraten von Neuronen im zentralen auditorischen System korreliert sind. Zunächst untersuchen wir ob sich bei lärmbedingtem Hörverlust die Audiogramme von Patienten mit und ohne Tinnitus unterscheiden. Im Vergleich zu Patienten ohne Tinnitus haben Tinnituspatienten im Mittel weniger Hörverlust, einen steileren Abfall des Audiogramms, und die Audiogrammkante befindet sich bei höheren Frequenzen. Mit einem theoretischen Modell zeigen wir, wie tinnitusartige Hyperaktivität durch eine Stabilisierung der mittleren Feuerrate von Neuronen im zentralen Hörsystem mittels homöostatischer Plastizität entstehen kann: verringerte Aktivität von Hörnervfasern nach Hörverlust wird kompensiert durch eine Erhöhung der neuronalen Verstärkung. Dies stabilisiert die mittlere Rate, kann jedoch zu einer Erhöhung der spontanen Feuerraten führen, die dann von Art und Stärke der cochlearen Schädigung abhängen. Wir testen das Modell, indem wir es auf die Audiogramme von Patienten mit tonalem Tinnitus und Lärmschwerhörigkeit anwenden. Für jedes Audiogramm sagen wir mit dem Modell Veränderungen in der Spontanaktivität von auditorischen Neuronen vorher. Das resultierende Hyperaktivitätsmuster hat typischerweise eine deutliche Spitze, die mit einem steilen Abfall des Audiogramms einhergeht. Wenn solch eine Spitze als Grundlage für einen tonalen Tinnitus interpretiert wird, dann sagt das Modell Tinnitusfrequenzen nahe den empfundenen Tinnitustonhöhen vorher. Unser Modell stellt also eine plausible Hypothese, wie Hörverlust zu Tinnitus führen könnte, dar. Basierend auf dem Modell zeigen wir außerdem wie Hyperaktivität und somit eventuell auch Tinnitus, durch zusätzliche akustische Stimulation reduziert werden könnte. / Tinnitus is a phantom auditory sensation that is associated with hearing loss, but how hearing loss can lead to tinnitus has remained unclear. In animals, hearing loss through cochlear damage can lead to behavioral signs of tinnitus and can increase the spontaneous firing rates of central auditory neurons. To study the relation between hearing loss and tinnitus, we first analyze audiometric differences between patients with hearing loss and tinnitus and patients with hearing loss but without tinnitus. We find that tinnitus patients have on average less hearing loss, a steeper slope of the audiogram, and the audiogram edge is located at higher frequencies compared to patients without tinnitus. We then derive a computational model that demonstrates how tinnitus-related hyperactivity could arise as a consequence of a stabilization of the mean firing rates of central auditory neurons through homeostatic plasticity: decreased auditory nerve activity after hearing loss is counteracted through an increase of the neuronal response gain. This restores the mean rate, but can also lead to increased spontaneous firing rates, which depend on the type and degree of cochlear damage. Finally, we test the ability of our model to predict tinnitus pitch by applying it to audiograms from patients with noise-induced hearing loss and tone-like tinnitus. Given an audiogram, the model is used to predict changes in the spontaneous firing rates of central auditory neurons. The resulting hyperactivity pattern typically exhibits a distinct peak that is associated with a steep drop in the audiogram. If such a peak is interpreted as the basis for a tone-like tinnitus sensation, the model predicts a tinnitus frequency that is close to the patient''s tinnitus pitch. Thus, our model presents a plausible hypothesis of how hearing loss could lead to tinnitus. Based on this model, we also show how hyperactivity, and possibly also tinnitus, could be alleviated through additional acoustic stimulation.

Neuronal activity in the brain is tightly regulated. During operation in real time, for instance, feedback and feedforward loops limit excessive excitation. In addition, cell autonomous processes ensure that neurons’ average activity is restored to a setpoint in response to chronic perturbations. These processes are summarized as homeostatic plasticity (Turrigiano in Cold Spring Harb Perspect Biol 4:a005736–a005736, 2012). In the basal ganglia, information is mainly transmitted through disinhibition, which already constraints the possible range of neuronal activity. When this tightly adjusted system is challenged by the chronic decline in dopaminergic neurotransmission in Parkinson’s disease (PD), homeostatic plasticity aims to compensate for this perturbation. We here summarize recent experimental work from animals demonstrating that striatal projection neurons adapt excitability and morphology in response to chronic dopamine depletion and substitution. We relate these cellular processes to clinical observations in patients with PD that cannot be explained by the classical model of basal ganglia function. These include the long duration response to dopaminergic medication that takes weeks to develop and days to wear off. Moreover, dyskinesias are considered signs of excessive dopaminergic neurotransmission in Parkinson’s disease, but they are typically more severe on the body side that is more strongly affected by dopamine depletion. We hypothesize that these clinical observations can be explained by homeostatic plasticity in the basal ganglia, suggesting that plastic changes in response to chronic dopamine depletion and substitution need to be incorporated into models of basal ganglia function. In addition, better understanding the molecular mechanism of homeostatic plasticity might offer new treatment options to avoid motor complications in patients with PD.