The computation and analysis of information that arrives from the outside world is an incredibly complex and fascinating task that our brains perform every day. How is, for example, a complicated, multi-frequency and constantly changing sound wave arriving at our ears converted and processed such that we can extract relevant pieces of information from it such as “The truck is approaching from the right, and it is coming fast”??

The auditory system is a great model system to study sensory processing because a significant part of the computations in this system are performed by a series of distinct and anatomically separate brain areas (the auditory brain stem nuclei), and neurons in each of these nuclei perform a set of restricted integration steps – which are virtually identical among all the neurons in a given nucleus. Thus, the very complex question “how does the auditory system process information” can be broken down into many smaller and simpler questions such as “how does brain nucleus x process information”?

Specifically, our laboratory is interested in the brain nuclei that form the sound localization pathway. Sound localization is obviously a useful skill for a predator hunting prey, or a prey trying to avoid a predator. However, the same skill is also used by modern humans on a daily basis to isolate a sound source of interest from other sound sources that are active at the same time. The sound localization circuit separates these competing sound sources from each other and allows us to focus on the one we are interested in, while ignoring the others. Because of this skill we can, for example, have a conversation with somebody in a crowded bar and clearly understand what the person is saying - although many other people are talking at the same time and the music is playing as well.

Most of our laboratory’s current research revolves around a nucleus called the “medial nucleus of the trapezoid body” (MNTB). This nucleus provides fast and extremely well timed inhibition to other nuclei in the sound localization pathway, and has some extreme adaptations for speed and temporal fidelity. One of the most striking adaptations is the synapse that conveys excitatory inputs to MNTB neurons – the calyx of Held.

The calyx of Held is giant synapse that carries a very large and well-timed incoming excitatory signal, which is converted into a well-timed outgoing inhibitory signal by the MNTB principal neuron. This inhibitory signal projects to other nuclei of the sound localization pathway, and participates in the computational process of sound localization, as well as in other tasks.

One of the projects we are pursuing is the question how this giant synapse performs when challenged with long streams of afferent information – as it typically occurs in real life situations. What are the rules of information processing at this synaptic station during biologically relevant activity?

Interestingly, MNTB neurons not only receive the excitatory inputs from the calyx of Held, but substantial inhibitory inputs as well (image 3). In a second set of projects we are pursuing the question, what the functional role of these inhibitory inputs might be. Where do they originate, what type of information do they carry, and how do they interact with the excitatory currents?

As mentioned above, the output of MNTB neurons integrates with excitatory projections from other sources in the actual process of sound localization. The localization of low frequency sounds and localization of high-frequency sounds is performed via different mechanisms and by two separate brain nuclei. Although the MNTB output controls both of these nuclei, we are mostly interested in the nucleus that performs low frequency localization, called medial superior olive (MSO). For many years, it was thought that the sound localization performed by MSO neurons is well understood, and that neural inhibition only plays a minor role in this localization process. However, more recent work suggests that fast inhibition, especially fast inhibition from the MNTB, may critically control the localization process at the MSO. We are testing this idea with a combination of optogenetics, in-vivo physiology and behavioral testing.

The sound localization nuclei participate not only in the pure process of sound localization, but also in the establishment of spatial channels. Being able to establish ‘channels of space’ from the incoming sound information, and being able to focus on a certain channel of interest, allows us to hear in an environment where many sound sources are active at the same time, or where background noise is present – for example the crowded bar on a Friday night. As people get older, they function progressively less well in that crowded bar, i.e. they have more and more trouble following a conversation when background noises are present. This condition, which is one form of age-related hearing loss (presbycusis), is very common and affects about half of the population by retirement age.

We would like to understand how fast inhibition in the auditory brain stem changes with age, and how these changes in fast inhibition are related to presbycusis. We study this question both in mice and humans with a combination of electrophysiology, hearing tests, and auditory brain stem response measurements. The clinical aspects of this project are performed in collaboration with Dr. Herman Jenkins, MD, and Dr. Kristin Uhler, PhD from the UCD Department of Otolaryngology.

Methods used in the lab:

Our laboratory uses a combination on in-vitro electrophysiology, in-vivo electrophysiology, anatomical and immunohistochemical methods, and optogenetics. In-vitro electrophysiology (patch clamp, image 4) is a great tool to study information processing in neurons on a cellular and subcellular level. We use patch clamp recordings to study synapses, ion channels, and the interaction of excitation and inhibition on a subcellular level. By contrast, in-vivo electrophysiology (extracellular recordings) allows us to study on a systems level, how sound information is processed (image 5, left device). Immunohistochemistry and neural tracing methods allow us to study the connections between different brain areas, and thus understand the ‘wiring’ of the auditory system. Viral manipulations and optogenetics allow us to manipulate neural circuits with light. We express novel, light sensitive, ion channels in auditory neurons, and then either turn off these cells with light (when an inhibitory protein, e.g. halorhodopsin, was expressed), or turn on the cells with light (when an excitatory protein, e.g. channelrhodopsin, was expressed). Light is delivered to deep brain nuclei via glass fibers that are connected to lasers or LEDs (image 5, right device). In order to understand, how light that is delivered by these fibers will spread in brain tissue, we also study the light scattering properties of brain tissue.