Balancing resource and energy usage for optimal performance in a neural system

The plasticity of the brain is one of the great scientific challenges and is of enormous interest in the general community because of the implications it has for the brain being able to repair itself, or be lent "a helping hand" by appropriate neural therapies and prostheses (see for example the popular book, "The Brain that Changes Itself" by Norman Doidge, Penguin 2007). Our work will provide a focussed, but hopefully significant new insight into the processes by which the brain can adjust itself to changing circumstances.
We will use computer simulations of mathematical models built from experimental data to explore the operation of an early stage of the mammalian auditory system. We will study how this brain region dynamically configures itself to meet the operational demands of incoming 'information' about sounds in the environment, encoded by the activity of neurons in the cochlear nucleus. The brain is a complex and dynamic information processing system that is built from a large, but finite set of noisy components (cells and associated extracellular and intracellular signalling systems) and must operate in an energy efficient way. We will test the hypothesis that specific plasticity mechanisms adjust neurons in this brain region differently depending on whether they are processing high or low frequency sounds. Further, we postulate that plasticity is also trying to minimise the energy used by the neurons, and that this might be in conflict with the optimum processing of incoming auditory information.
The increased understanding of the brain's intrinsic plasticity resulting from this project will ultimately have implications for the development of neural therapies. Treatments for neural dysfunction inevitably invoke intrinsic neural plasticity mechanisms that might enhance or even hinder the treatment. Of specific interest here is the development of cochlear implants to treat impaired hearing that cannot be compensated for by conventional hearing aids. These implants generate electrical signals in response to sounds and stimulate either the auditory nerve (most commonly) or the cochlear nucleus. Remarkable results have already been achieved with implants whose signals have only a fraction of the resolution and dynamic range of an intact cochlear (Wilson & Dorman (2008) Cochlear implants: Current designs and future possibilities, Journal of Rehabilitation Research & Development 45:695-730). This is entirely due to the brain's ability to adapt. Despite this success, improvements in cochlear implants will come through an improved understanding of the intrinsic plasticity mechanisms that are being invoked by the implant's stimulation. To quote from Wilson & Dorman (2008): "Cochlear implants work as a system, in which all parts are important, including the microphone, the processing strategy, the transcutaneous link, the receiver/stimulator, the implanted electrodes, the functional anatomy of the implanted cochlea, and the user's brain. Among these, the brain has received the least attention in implant designs to date." Our work will provide data on the mechanisms and theories of the implications of intrinsic plasticity in the brainstem auditory system.
A further aspect of this project that needs increased public awareness is our use of a "systems" approach to studying a neural system. This has two aspects: (1) taking a holistic view of neural function that includes aspects such as activity-dependent regulation, noise and energy consumption, and (2) a tightly integrated programme of experiments and computational modelling. People are familiar with the use of computers in weather forecasting and climate change predictions, but there is less awareness of their use in computational biology and neuroscience. Appropriate dissemination of our work can give a snapshot of how computers and experiments together can provide insight into the detailed workings of the nervous system.

We will use a systems biology approach, consisting of a tightly integrated programme of experiments and computational modelling, to study activity-dependent regulation in the medial nucleus of the trapezoid body (MNTB) in the mammalian auditory brainstem, which plays a key role in sound source localisation (SSL). We will examine how different intrinsic plasticity mechanisms, evoked by incoming neural activity, obtain satisfactory functional performance in this nucleus from a limited set of noisy resources (neurons, ion channels, synapses etc) while minimising energy usage.
Experimental recordings will be made in tissue slices from mouse. A combination of electrophysiology, pharmacology, immunohistochemistry and genetic manipulation will provide data on the resource distribution in the MNTB neurons and associated calyx of Held synapse, and the regulation of these resources by activity. The experimental data will be used to fit the parameters of a computational model, which will be in the form of a Hodgkin-Huxley-style compartmental model of an MNTB neuron and its synapse. The model will include heterogeneous distributions of identified ion channels types, stochastic neurotransmitter release, and multiple mechanisms of short-term synaptic plasticity. Advanced statistical optimisation techniques will be used to fit model parameters. The model will be analysed to determine information transmission through this system and associated energy usage as estimated by ATP consumption. We postulate that the amount of information transmitted as a fraction of energy used will be different between the high sound frequency and low sound frequency poles of the MNTB.

This project will deliver new insights into how the brain manages limited resources to maximise information transmission and achieve specific physiological functions. This topic is of increasing academic importance, because it adds a new dimension to constraining neuronal models of brain function and highlights how metabolic limitations (or signalling resources) are crucial for the overall brain 'economy'. This work will have impact for both auditory specialists and broader neuronal models of cortical function, by defining intrinsic plasticity pathways in one sensory stream that will be of broad application across neurobiology. By understanding the limitations of a young healthy brain and obtaining pointers to changes occurring in mature and older brains, we are setting the stage for a better understanding of metabolic limitations in an aged brain.
The Auditory group at Leicester has achieved impact in the auditory field because our biophysical approaches in native neurons within the auditory brainstem provide one of the few examples where auditory processing can be addressed at the molecular level in the brain. The spinoff company, Autifony Therapeutics, emerging from GSK and run by Charles Large, is developing therapeutic compounds acting on potassium channels for the treatment of tinnitus. This work achieves heightened importance when integrated with the modelling strategies of our colleagues in computational neurobiology (as we have planned in this application) because these models help generalise specific insights into the broader context of brain function. Our work is relevant to cochlea biophysicists and to those exploring cognitive and behavioural studies of hearing in the UK (e.g. Nottingham MRC Unit - Institute for Hearing Research, audiologists in Nottingham and Leicester ENT Departments, groups in London and Oxford) and internationally. Several charities also specifically target this area: AgeUK, Deafness Research UK, RNID. The demonstration that synaptic activity can regulate the target neuron excitability is important for future development of cochlea implants, because our data suggests that the implanted area, which may be hyper- or hypo-excitable, can be manipulated. Consideration of resource limits within the brain will become more important as machine-brain interfaces are increasingly explored.
It is recognised, by the BBSRC and by biological research groups around the world, that taking a "systems approach" is essential in helping us to understand how biological systems work. This approach is 'holistic' in the sense of studying all components of contained systems and also exploring how these systems fit within the complex surrounding ecosystem. Such a "systems" understanding requires mathematical models and computer simulations to be developed from experimental data. This is exactly the approach we are taking here and we will thoroughly train two young researchers (one from a predominantly biological background and the other from a quantitative or physical science background) in this combined experimental/modelling approach. A strong appreciation of the synergy between experiment and modelling is still rare, and our trained RAs will be able to push the "systems" agenda in their future research careers.
General public awareness of this sort of "systems approach" is also limited but should have great intuitive appeal and increase the understanding of how science is done and the importance of computational modelling in biological research. People are familiar with the use of computers in weather forecasting and climate change predictions, but there is less awareness of their use in computational biology and neuroscience. Appropriate dissemination of our work through press releases, public lectures can give a snapshot of how computers and experiments together can provide insight into the detailed workings of the nervous system.