Neural processing of communication signals in the auditory brainstem
Final Report Abstract
Human speech and many animal communication sounds are temporally segmented into trains of syllables. It is sufficient for the intelligibility of speech signals to preserve just the low frequency fluctuations in the signal envelope, a fact majorily exploited in modern cochlear implant speech processors. Highly specialized neurons in the brainstem process crucial auditory cues. Yet, it is still unknown whether there exists a specialized structure in the brainstem to extract the rich temporal structure underlying speech signals. Recently, Prof. Anna Magnusson (my host) and other researchers have suggested that neurons in the superior paraolivary nucleus (SPON) extract the slow temporal envelope as they follow amplitude modulations and typically show on- and offset spiking, i.e. respond to strong changes in stimulus energy, both in vivo and in vitro. My project was focused on exploring whether SPON neurons are tuned to slowly fluctuating stimuli, discovering the specializations that allow SPON to follow fluctuations in stimulus energy and investigate the underlying neural mechanisms that shape their response properties. SPON neurons in mice can be divided into three subpopulations based on their responses to depolarizing current steps in vitro: onset, burst and adapting neurons. Subthreshold currents involved in the characteristic rebound spiking response are the hyperpolarizationactivated inward current Ih and the inward Ca2+ -current IT. As a new finding, many neurons featured the transient outward current Ia. Using both current- and voltage clamp protocols I quantified the three hyperpolarization-activated currents and found a strong correlation not only with the step-response type, but also with the neurons’ frequency tuning. Pharmacological blocking experiments suggested that Ih together with the low voltage activated K+ -current IKLT act as a high pass filter and lead to tuning to frequencies around 30 Hz in onset neurons. Burst neurons possess strong IT currents, which shifts their tuning to lower frequencies (ca. 10 Hz), whereas adapting neurons that possess Ia were tuned to the lowest frequencies (<5 Hz). They sometimes displayed a second resonance point resembling a notch filter. In collaboration with Dr. Bertrand Fontaine (University of Leuven, Belgium), we designed a single compartment neural model for each cell type illustrating the interaction of the subthreshold currents and their effect on frequency tuning. Using stimuli of different periodicity and shape shed new light on the functional properties under more realistic stimulation. Inspite of being high frequency tuned, onset neurons turned out to prefer periodic stimuli of long duty cycles, whereas burst neurons coped better with short duty cycles, both requiring rather fast changes in input current. Adapting neurons on the other side showed graded activity and responded to very slow changes in input current. I thus propose that the three subpopulations found in SPON process different aspects of communication sounds in parallel: onset neurons extract the rough structure and segmentation of the auditory input, burst neurons are sensitive to smaller transients in loudness that may characterize certain syllables and adapting neurons encode even slow and small modulations, allowing for identification of each sound element. My findings are a first step in understanding the specializations of SPON neurons and their contributions to auditory processing. The proposed interpretation will motivate several other lines of research as it needs confirmation from in-vivo and behavioral experiments. In my current position as research adjunct at the Karolinska University, I will pursue or supervise some of these experiments and hope to initiate a clinical collaboration on brainstem activity during speech processing.
Publications
- (2013) Development of on-off spiking in Superior Paraolivary Nucleus neurons of the mouse. Journal of Neurophysiology 109: 2691-704
Felix, RA., Vonderschen, K., Berrebi, AS., Magnusson, AK.
- (2014) Detection of interaural time differences and remodeling their representation. Trends in Neuroscience 37: 289-300
Vonderschen, K., Wagner, H.
(See online at https://doi.org/10.1016/j.tins.2014.03.002)
