Richard Kempter

Cross-frequency phase-phase coupling between theta and gamma oscillations in the hippocampus

Neuronal oscillations allow for temporal segmentation of neuronal spikes. Interdependent oscillators can integrate multiple layers of information. We examined phase–phase coupling of theta and gamma oscillators in the CA1 region of rat hippocampus during maze exploration and rapid eye movement sleep. Hippocampal theta waves were asymmetric, and estimation of the spatial position of the animal was improved by identifying the waveform-based phase of spiking, compared to traditional methods used for phase estimation. Using the waveform-based theta phase, three distinct gamma bands were identified: slow gammaS (gammaS; 30–50 Hz), midfrequency gammaM (gammaM; 50–90 Hz), and fast gammaF (gammaF; 90–150 Hz or epsilon band). The amplitude of each sub-band was modulated by the theta phase. In addition, we found reliable phase–phase coupling between theta and both gammaS and gammaM but not gammaF oscillators. We suggest that cross-frequency phase coupling can support multiple time-scale control of neuronal spikes within and across structures.

Development of tinnitus-related neuronal hyperactivity through homeostatic plasticity after hearing loss: a computational model

Tinnitus, the perception of a sound in the absence of acoustic stimulation, is often associated with hearing loss. Animal studies indicate that hearing loss through cochlear damage can lead to behavioral signs of tinnitus that are correlated with pathologically increased spontaneous firing rates, or hyperactivity, of neurons in the auditory pathway. Mechanisms that lead to the development of this hyperactivity, however, have remained unclear. We address this question by using a computational model of auditory nerve fibers and downstream auditory neurons. The key idea is that mean firing rates of these neurons are stabilized through a homeostatic plasticity mechanism. This homeostatic compensation can give rise to hyperactivity in the model neurons if the healthy ratio between mean and spontaneous firing rate of the auditory nerve is decreased, for example through a loss of outer hair cells or damage to hair cell stereocilia. Homeostasis can also amplify non‐auditory inputs, which then contribute to hyperactivity. Our computational model predicts how appropriate additional acoustic stimulation can reverse the development of such hyperactivity, which could provide a new basis for treatment strategies.

Intrinsic Stabilization of Output Rates by Spike-Based Hebbian Learning

We study analytically a model of long-term synaptic plasticity where synaptic changes are triggered by presynaptic spikes, postsynaptic spikes, and the time differences between presynaptic and postsynaptic spikes. The changes due to correlated input and output spikes are quantified by means of a learning window. We show that plasticity can lead to an intrinsic stabilization of the mean firing rate of the postsynaptic neuron. Subtractive normalization of the synaptic weights (summed over all presynaptic inputs converging on a postsynaptic neuron) follows if, in addition, the mean input rates and the mean input correlations are identical at all synapses. If the integral over the learning window is positive, firing-rate stabilization requires a non-Hebbian component, whereas such a component is not needed if the integral of the learning window is negative. A negative integral corresponds to anti-Hebbian learning in a model with slowly varying firing rates. For spike-based learning, a strict distinction between Hebbian and anti-Hebbian rules is questionable since learning is driven by correlations on the timescale of the learning window. The correlations between presynaptic and postsynaptic firing are evaluated for a piecewise-linear Poisson model and for a noisy spiking neuron model with refractoriness. While a negative integral over the learning window leads to intrinsic rate stabilization, the positive part of the learning window picks up spatial and temporal correlations in the input.

Course of hearing loss and occurrence of tinnitus

Chronic tinnitus is often accompanied by a hearing impairment, but it is still unknown whether hearing loss can actually cause tinnitus. The association between the pitch of the tinnitus sensation and the audiogram edge in patients with high-frequency hearing loss suggests a functional relation, but a large fraction of patients with hearing loss does not present symptoms of tinnitus. We therefore, investigated how the occurrence of tinnitus is related to the shape of the audiogram. We analyzed a sample where all patients had noise-induced hearing loss, containing 30 patients without tinnitus, 24 patients with tone-like tinnitus, and 17 patients with noise-like tinnitus. All patients had moderate to severe high-frequency hearing loss, and only minor to moderate hearing loss at low frequencies. We found that tinnitus patients had less overall hearing loss than patients without tinnitus. Moreover, the maximum steepness of the audiogram was higher in patients with tinnitus (-52.9+/-1.9 dB/octave) compared to patients without tinnitus (-43.1+/-2.4 dB/octave). Differences in overall hearing loss and maximum steepness between tone-like and noise-like tinnitus were not significant. For tone-like tinnitus, there was a clear association between the tinnitus pitch and the edge of the audiogram, with tinnitus pitch being on average 1.48+/-0.12 octaves above the audiogram edge frequency, and 0.81+/-0.1 octaves above the frequency with the steepest slope. Our results suggest that the occurrence of tinnitus is promoted by a steep audiogram slope. A steep slope leads to abrupt discontinuities in the activity along the tonotopic axis of the auditory system, which could be misinterpreted as sound.

Extracting oscillations: neuronal coincidence detection with noisy periodic spike input

How does a neuron vary its mean output firing rate if the input changes from random to oscillatory coherent but noisy activity? What are the critical parameters of the neuronal dynamics and input statistics? To answer these questions, we investigate the coincidence-detection properties of an integrate-and-fire neuron. We derive an expression indicating how coincidence detection depends on neuronal parameters. Specifically, we show how coincidence detection depends on the shape of the postsynaptic response function, the number of synapses, and the input statistics, and we demonstrate that there is an optimal threshold. Our considerations can be used to predict from neuronal parameters whether and to what extent a neuron can act as a coincidence detector and thus can convert a temporal code into a rate code.

Single-Trial Phase Precession in the Hippocampus

During the crossing of the place field of a pyramidal cell in the rat hippocampus, the firing phase of the cell decreases with respect to the local theta rhythm. This phase precession is usually studied on the basis of data in which many place field traversals are pooled together. Here we study properties of phase precession in single trials. We found that single-trial and pooled-trial phase precession were different with respect to phase-position correlation, phase-time correlation, and phase range. Whereas pooled-trial phase precession may span 360°, the most frequent single-trial phase range was only ∼180°. In pooled trials, the correlation between phase and position (r = −0.58) was stronger than the correlation between phase and time (r = −0.27), whereas in single trials these correlations (r = −0.61 for both) were not significantly different. Next, we demonstrated that phase precession exhibited a large trial-to-trial variability. Overall, only a small fraction of the trial-to-trial variability in measures of phase precession (e.g., slope or offset) could be explained by other single-trial properties (such as running speed or firing rate), whereas the larger part of the variability remains to be explained. Finally, we found that surrogate single trials, created by randomly drawing spikes from the pooled data, are not equivalent to experimental single trials: pooling over trials therefore changes basic measures of phase precession. These findings indicate that single trials may be better suited for encoding temporally structured events than is suggested by the pooled data.

Acoustic stimulation treatments against tinnitus could be most effective when tinnitus pitch is within the stimulated frequency range.

Acoustic stimulation with hearing aids or noise devices is frequently used in tinnitus therapy. However, such behind-the-ear devices are limited in their high-frequency output with an upper cut-off frequency of approximately 5-6 kHz. Theoretical modeling suggests that acoustic stimulation treatments with these devices might be most effective when the tinnitus pitch is within the stimulated frequency range. To test this hypothesis, we conducted a pilot study with 15 subjects with chronic tinnitus. Eleven subjects received hearing aids and four subjects noise devices. Perceived tinnitus loudness was measured using a visual analog scale, and tinnitus-related distress was assessed using the Tinnitus Questionnaire. After six months of device usage, reductions of perceived tinnitus loudness were seen only in subjects with a tinnitus pitch of less than 6 kHz. When subjects were grouped by tinnitus pitch, the group of patients with a tinnitus pitch of less than 6 kHz (n = 10 subjects) showed a significant reduction in perceived tinnitus loudness (from 73.4 +/- 6.1 before to 56.4 +/- 7.4 after treatment, p = 0.012), whereas in subjects with a tinnitus pitch of 6 kHz or more (n = 5 subjects) tinnitus loudness was slightly increased after six months of treatment (65.0 +/- 4.7 before and 70.6 +/- 5.9 after treatment), but the increase was not significant (p = 0.063). Likewise, tinnitus-related distress was significantly decreased in the low-pitch group (from 31.6 +/- 4.3 to 20.9 +/- 4.8, p = 0.0059), but not in the group with high-pitched tinnitus (30.2 +/- 3.3 before and 30.0 +/- 5.1 after treatment, p = 1). Overall, reductions in tinnitus-related distress in our study were less pronounced than those reported for more comprehensive treatments. However, the differences we observed between the low- and the high-pitch group show that tinnitus pitch might influence the outcome of acoustic stimulation treatments when devices with a limited frequency range are used.

State-dependencies of learning across brain scales

Learning is a complex brain function operating on different time scales, from milliseconds to years, which induces enduring changes in brain dynamics. The brain also undergoes continuous “spontaneous” shifts in states, which, amongst others, are characterized by rhythmic activity of various frequencies. Besides the most obvious distinct modes of waking and sleep, wake-associated brain states comprise modulations of vigilance and attention. Recent findings show that certain brain states, particularly during sleep, are essential for learning and memory consolidation. Oscillatory activity plays a crucial role on several spatial scales, for example in plasticity at a synaptic level or in communication across brain areas. However, the underlying mechanisms and computational rules linking brain states and rhythms to learning, though relevant for our understanding of brain function and therapeutic approaches in brain disease, have not yet been elucidated. Here we review known mechanisms of how brain states mediate and modulate learning by their characteristic rhythmic signatures. To understand the critical interplay between brain states, brain rhythms, and learning processes, a wide range of experimental and theoretical work in animal models and human subjects from the single synapse to the large-scale cortical level needs to be integrated. By discussing results from experiments and theoretical approaches, we illuminate new avenues for utilizing neuronal learning mechanisms in developing tools and therapies, e.g., for stroke patients and to devise memory enhancement strategies for the elderly.

Formation of temporal-feature maps by axonal propagation of synaptic learning

Computational maps are of central importance to a neuronal representation of the outside world. In a map, neighboring neurons respond to similar sensory features. A well studied example is the computational map of interaural time differences (ITDs), which is essential to sound localization in a variety of species and allows resolution of ITDs of the order of 10 μs. Nevertheless, it is unclear how such an orderly representation of temporal features arises. We address this problem by modeling the ontogenetic development of an ITD map in the laminar nucleus of the barn owl. We show how the owls ITD map can emerge from a combined action of homosynaptic spike-based Hebbian learning and its propagation along the presynaptic axon. In spike-based Hebbian learning, synaptic strengths are modified according to the timing of pre- and postsynaptic action potentials. In unspecific axonal learning, a synapses modification gives rise to a factor that propagates along the presynaptic axon and affects the properties of synapses at neighboring neurons. Our results indicate that both Hebbian learning and its presynaptic propagation are necessary for map formation in the laminar nucleus, but the latter can be orders of magnitude weaker than the former. We argue that the algorithm is important for the formation of computational maps, when, in particular, time plays a key role.

Quantifying circular-linear associations: hippocampal phase precession.

When a rat crosses the place field of a hippocampal pyramidal cell, this cell typically fires a series of spikes. Spike phases, measured with respect to theta oscillations of the local field potential, on average decrease as a function of the spatial distance traveled. This relation between phase and position of spikes might be a neural basis for encoding and is called phase precession. The degree of association between the circular phase variable and the linear spatial variable is commonly quantified through, however, a linear-linear correlation coefficient where the circular variable is converted to a linear variable by restricting the phase to an arbitrarily chosen range, which may bias the estimated correlation. Here we introduce a new measure to quantify circular-linear associations. This measure leads to a robust estimate of the slope and phase offset of the regression line, and it provides a correlation coefficient for circular-linear data that is a natural analog of Pearsons product-moment correlation coefficient for linear-linear data. Using surrogate data, we show that the new method outperforms the standard linear-linear approach with respect to estimates of the regression line and the correlation, and that the new method is less dependent on noise and sample size. We confirm these findings in a large data set of experimental recordings from hippocampal place cells and theta oscillations, and we discuss remaining problems that are relevant for the analysis and interpretation of phase precession. In summary, we provide a new method for the quantification of circular-linear associations.