Associate Professor | Neurology & Neurosurgery, Medicine (Dept. & Faculty)
Researcher | Research Institute of the McGill University Health Centre
Associate member | Dept of Physiology
Research interests in brief
- Mechanisms and phenomenology of synaptic plasticity learning rules
- Information storage and memory in the brain
- The organization of connectivity in cortical circuits in health and disease
- Advanced optical approaches in neuroscience research
Synaptic plasticity of cortical circuits
How does the visual cortex learn to interpret visual information from the outside world? How is information stored in the brain? After all, the brain has no central processing unit to control actions and no memory storage banks to keep track of information. There is only a vast network of interconnected neurons.
Today, there is good evidence supporting the notion that learning and memory occur at the synaptic connections that exist between neurons of the brain. Events in the outside world cause particular patterns of activity in subpopulations of neurons in the cortex. In turn, these activity patterns bring about changes in connective strengths among these neurons. Such changes, which are known as synaptic plasticity, are a means of storing information in neuronal circuits, and the particulars of the activity patterns that determine synaptic plasticity are known as cellular learning rules. The storage of information via changes in connective strength thus result in re-wiring of cortical sub-networks, in turn leading to particular connectivity motifs that shape certain computational features and that in effect store specific components of information. Consequently, there exist close links between spatio-temporal activity patterns in neuronal assemblies, connectivity motifs, and the cortical computations that bring about detection and recall.
To understand how the brain works, it is therefore essential to know the properties and the mechanistic underpinnings of cellular learning rules, as well as their functional impact in the intact brain. In addition, we need to know the connectivity patterns that ensue from particular learning rules and how these are shaped by stimuli in the outside world. Only then can the impact on brain functioning of drugs such as marijuana, or pathology such as epilepsy, be fully appreciated. Our starting point is Spike-Timing-Dependent Plasticity (STDP) in neocortical microcircuits. In the STDP learning paradigm, whether synaptic strengthening and weakening is brought about depends critically on the relative millisecond timing of spiking activity in connected pairs of cells. Although the precise outcome depends on the brain region that is investigated, typically pre before postsynaptic spiking activity repeated within a couple of tens of milliseconds results in synaptic strengthening, whereas the opposite temporal order evokes weakening of synaptic connections. Because spikes are in some sense the "atoms" of neuronal activity, the STDP paradigm provides exquisite experimental control, which is a first critical step toward dissecting the complexities of cortical learning rules. However, our research is not limited to the mechanisms and impact of STDP; indeed, several forms of plasticity depend critically on local dendritic spikes in the absence of the more global action potentials that underlie STDP. To understand synaptic plasticity in visual cortical circuits, my team employs several state-of-the-art approaches: quadruple whole-cell recordings, two-photon laser scanning microscopy, neurotransmitter uncaging, optogenetics, and computer simulations.
Illustration of the quadruple whole-cell recording approach for finding monosynaptically connected pairs of neurons in acute neocortical slices. Left: 2-photon laser-scanning microscopy of cells dye filled via recording pipette reveals three pyramidal cells with characteristic apical dendrites (A, B, and C), and one interneuron (D). Right: A train of action potentials in the inhibitory interneuron D elicits spike-locked IPSPs in pyramidal cells A and B, but not in C.