Neural integration of electrocommunication signals encoded in parallel sensory processing streams.

• Applicant: Dr. Jan Grewe
• Subject Area: Cognitive, Systems and Behavioural Neurobiology
• Term: from 2019 to 2023
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 426809286

Multisensory integration joins information provided by different senses into a unified percept. In situations of conflicting information it may lead to perceptional shifts such as the ventriloquist illusion. Combining congruent information from multiple senses, however, often improves behavioral performances by reduced reaction times or improved stimulus detection. Often, this integration is Bayes-optimal, that is, information channels are weighted according to their respective reliability. On the neuronal level, increased responsiveness, reduced response latencies, and improved phase locking have been described as effects of multisensory integration. How information from multiple channels is integrated has been investigated in a multitude of studies. Mathematical as well as phenomenological models have been developed to understand the guiding principles. However, our understanding of the physiolgical processes underlying these computational principles is still weak. In particular, electrophysiological data allowing to investigate the subthreshold mechanisms are missing. We will address this gap between theory and physiology in the context of the encoding of electrocommunication signals in the weakly electric fish Eigenmannia virescens. In Eigenmannia, communication signals, called chirps, are modulations of the fish's own electric organ discharge which stimulates three parallel electrosensory pathways. Initially, information is processed separately, but is eventually joined in the torus semicircularis (TS) of the midbrain, the teleost homolog to the mammalian inferior colliculus. Our previous analyses on the level of the electroreceptors suggest that chirp detection should be improved by combining information from the parallel pathways. By means of in vivo intracellular recordings of neurons in the parallel channels in the hindbrain and of integrating and non-integrating neurons in the TS we will gather the required data of pre- and postsynaptic activity to analyze the mechanism of integration in the subthreshold regime. With the proposed study we aim at open questions regarding the encoding of electrocommunication signals in integrating and non-integrating neurons in the hind- and midbrain. Based on the acquired data we will analyze the subthreshold mechansims involved in the integration processes in multichannel neurons in the TS. These will be interpreted in the context of the known theories of multisensory integration.

• DFG Programme: Research Grants