Neuronal identification of acoustic signal periodicity

An auditory feature detection circuit for sound pattern recognition

From human language to birdsong and the chirps of insects, acoustic communication is based on amplitude and frequency modulation of sound signals. Whereas frequency processing starts at the level of the hearing organs, temporal features of the sound amplitude such as rhythms or pulse rates require processing by central auditory neurons. Besides several theoretical concepts, brain circuits that detect temporal features of a sound signal are poorly understood. We focused on acoustically communicating field crickets and show how five neurons in the brain of females form an auditory feature detector circuit for the pulse pattern of the male calling song. The processing is based on a coincidence detector mechanism that selectively responds when a direct neural response and an intrinsically delayed response to the sound pulses coincide. This circuit provides the basis for auditory mate recognition in field crickets and reveals a principal mechanism of sensory processing underlying the perception of temporal patterns.

Frequency selectivity emerging from spike-timing-dependent plasticity

Periodic neuronal activity has been observed in various areas of the brain, from lower sensory to higher cortical levels. Specific frequency components contained in this periodic activity can be identified by a neuronal circuit that behaves as a bandpass filter with given preferred frequency, or best modulation frequency (BMF). For BMFs typically ranging from 10 to 200 Hz, a plausible and minimal configuration consists of a single neuron with adjusted excitatory and inhibitory synaptic connections. The emergence, however, of such a neuronal circuitry is still unclear. In this letter, we demonstrate how spike-timing-dependent plasticity (STDP) can give rise to frequency-dependent learning, thus leading to an input selectivity that enables frequency identification. We use an in-depth mathematical analysis of the learning dynamics in a population of plastic inhibitory connections. These provide inhomogeneous postsynaptic responses that depend on their dendritic location. We find that synaptic delays play a crucial role in organizing the weight specialization induced by STDP. Under suitable conditions on the synaptic delays and postsynaptic potentials (PSPs), the BMF of a neuron after learning can match the training frequency. In particular, proximal (distal) synapses with shorter (longer) dendritic delay and somatically measured PSP time constants respond better to higher (lower) frequencies. As a result, the neuron will respond maximally to any stimulating frequency (in a given range) with which it has been trained in an unsupervised manner. The model predicts that synapses responding to a given BMF form clusters on dendritic branches.

Walking in Fourier's space: algorithms for the computation of periodicities in song patterns by the cricket Gryllus bimaculatus

Is discrimination of the envelope of an acoustic signal based on spectral or temporal computations? To investigate this question for the cricket Gryllus bimaculatus, pattern envelopes were constructed by the addition of several sine waves and modified by systematic phase changes. The phonotactic response of female crickets towards such sinusoidal but also rectangular pulse patterns was quantified on a locomotion compensator. Envelope patterns that exhibited a modulation frequency of 25 Hz as the dominant frequency were attractive and although changes of phase modified the temporal pattern, the values of attractiveness remained unaffected. Removal of the 25-Hz component reduced the phonotactic scores. Patterns in which other frequency components exhibited a larger amplitude than the 25-Hz component were less attractive. However, the combination of an unattractive pulse period with the attractive modulation frequency of 25 Hz in a pattern revealed that such stimuli were unattractive despite the presence of the 25-Hz component. A comparison of the attractiveness of all patterns revealed that female crickets evaluated the duration of pulse period over a wide range of duty cycles. The combined evidence showed that pattern envelopes were processed in the time- and not in the spectral domain.

Neuronal identification of signal periodicity by balanced inhibition

Many animals, including men, use periodicity information, e.g., amplitude modulations of acoustic stimuli, as a vital cue to auditory object formation. The underlying neuronal mechanisms, however, still remain a matter of debate. Here, we mathematically analyze a model for periodicity identification that relies on the interplay of excitation and delayed inhibition. Our analytical results show how the maximal response of such a system varies systematically with the time constants of excitation and inhibition. The model reliably identifies signal periodicity in the range from about ten to several hundred Hertz. Importantly, the model relies on biologically plausible parameters only. It works best for excitatory and inhibitory neuronal couplings of equal strength, the so-called 'balanced inhibition'. We show how balanced inhibition can serve to identify low-frequency signal periodicity and how variation of a single parameter, the inhibitory time constant, can tune the system to different frequencies.

Intracellular Responses of Neurons in the Mouse Inferior Colliculus to Sinusoidal Amplitude-Modulated Tones

Changes in the temporal envelope are important defining features of natural acoustic signals. Many cells in the inferior colliculus (IC) respond preferentially to certain modulation frequencies, but how they accomplish this is not yet clear. We therefore made whole cell patch-clamp recordings in the IC of anesthetized mice while presenting sinusoidal amplitude-modulated (SAM) tones. The relation between the number of evoked spikes and modulation frequency was used to construct rate modulation transfer functions (rMTFs). We observed different types of rate tuning, including band-pass (16%), band-reject (13%), high-pass (6%), and low-pass (6%) tuning. In the high-pass rMTF neurons and some of the low-pass rMTF neurons, the tuning characteristics appeared to be already present in the inputs. In both band-pass and band-reject rMTF neurons, the nonlinear relation between membrane potential and spike probability ensured preferential spiking during only a small part of the modulation period. Band-pass rMTF neurons had rapidly rising excitatory postsynaptic potentials, allowing good phase-locking to brief tones and intermediate modulation frequencies. At low modulation frequencies, adaptation of their spike threshold contributed to the onset response. In contrast, band-reject rMTF neurons responded with small excitatory or inhibitory postsynaptic potentials to brief tones. In these cells, a power law could describe the supralinear relation between average membrane potential and spike rate. Differences in timing of synaptic input and presence or absence of spike adaptation therefore define band-pass and band-reject rate tuning to SAM tones in the mouse IC.

The role of frequency, phase and time for processing of amplitude modulated signals by grasshoppers

Acoustic signals consist of pressure changes over time and can thus be analyzed in the frequency- or in the time-domain. With behavioural experiments we investigated which frequency components (FC) are necessary for the recognition of the periodic envelope of the conspecific song by females of the grasshopper Chorthippus biguttulus. Further, we determined up to which frequency component phase information is required which would indicate processing in the time domain. Responses of females revealed that signals composed of FC between 10 and 50 Hz are sufficient for recognition of the song envelope. A systematic reduction in the number of FC showed that no single frequency component was required; signals without the fundamental frequency were still highly attractive and only three FC may be sufficient for song recognition. Phase changes for frequencies up to 40 Hz strongly changed the attractiveness of song signals but only little at 50 Hz. Females were also tested with rectangular signals in which pause duration was varied. Evidently, and despite the high attractiveness of song signals with a "missing fundamental", females evaluated the attractiveness of signals in the time-domain, since the selectivity for pause duration predicted the responses to signals composed from FC well.

Modeling the cochlear nucleus: a site for monaural echo suppression?

Echo suppression plays an important role in identifying and localizing auditory objects. One can distinguish between binaural and monaural echo suppression, although the former is the one commonly referred to. Based on biological findings we introduce and analyze a mathematical model for a neural implementation of monaural echo suppression in the cochlear nucleus. The model's behavior has been verified by analytical calculations as well as by numerical simulations for several types of input signal. It shows that in the perception of a pair of clicks the leading click suppresses the lagging one and that suppression is maximal for an interclick interval of 2-3 ms. Similarly, ongoing stimuli will be affected by the suppression mechanism primarily a couple of milliseconds after onset, resulting in a reduced perception of a sound shortly after its start. Both effects match experimental data.