Past events

A fundamental problem in brain research is how distributed brain systems work together to give rise to behavior. I seek to advance our understanding of principles underlying the dynamic interaction between multiple neural systems, how the different systems co-operate and/or compete to give rise to goal-directed behavior, and the dynamics of the system when one or more of its components fail. Magnetic resonance imaging (MRI) methods allow us to simultaneously measure the function of multiple brain systems. In humans we can characterize the functional organization and specialization, and compare the system between health and disease. In animal models we can further dissect the circuits underlying these dynamics. In my work I aim to identify functional networks that span multiple cortical and subcortical regions, characterize their responses in the presence and absence of overt behavior, and modulate the observed dynamics. To advance these goals, I am developing new tools that will allow us to study large-scale neural systems across species. In this talk, I will review recent studies that use functional neuroimaging in humans and animal models. I will describe how spontaneous fluctuations of the blood oxygenation level-dependent (BOLD) signal measured with MRI in awake resting humans, reveal functional subdivisions in the medial temporal lobe memory system and parietal and prefrontal cortical components linked to it. I will describe results from non-human primates demonstrating that this functional organization persists across the species, highlighting cortical components that have undergone considerable areal expansion in humans relative to non-human primates, how this method can be used to identify homologue regions, and more generally, what can be learned from a comparative perspective. In the second part of my talk I will describe recent efforts to selectively modulate system dynamics. A lentivirus was used to target excitatory neurons in the rat cortex with light-activated cation channel channelrhodopsin-2. Using photostimulation to activate these neurons we were able to drive the BOLD response locally and in regions anatomically connected to the infected site in a variety of stimulation paradigms. I will discuss implications for understanding the BOLD signal and prospects for this approach in studying the microcircuit as well as large-scale brain dynamics. Finally, I will discuss the challenges and promises of whole-brain imaging in small animals, and how this work can provide avenues to bridge between a basic understanding of human behavior, large-scale neural dynamics, and psychiatric disorders where such dynamics are disrupted.

Diminished levels of docosahexaenoic acid (DHA, 22:6n-3), the major polyunsaturated fatty acid (FA) synthesized from alpha linolenic acid (ALA, 18:3n-3), have been implicated in changes in neurotransmitter production, ion channels disruption and impairments of a variety of cognitive, behavioural and motor functions in the developing and the adult mammal. We studied neuronal migration in the cortex and hippocampus of newborn and postnatal rats after ALA-deficiency, beginning on the 2nd day after conception and continuing for three weeks after birth. A marked decrease in the migration of bromodeoxyuridine(+)/NeuN(+)/Neurofilament(+) and glia fibrilary acidic protein(-) neuronal cells to the dense cortical plate was accompanied by a corresponding abundance of non-migrating cells in several regions such as cortical layers IV-VI, corpus callosum and the sub-ventricular zone of ALA-deficient newborn. Similarly, a delayed migration of cells to CA1 and dentate gyrus areas was noticed while most cells were retained in the subicular area adjacent to the hippocampus. The delay in migration was transient most likely due to a temporary reelin disorganization. In addition to these changes a drastic reduction in tyrosine hydroxylase (TH) and vesicular monoamine transporter-2 (VMAT-2) levels, both of which are prerequisites for appropriate synthesis and transport of DA were noticed by RNA subtractive hybridization and proteomic techniques. Concomitantly, a large increase in DA receptors DAR1 and DAR2 were noticed. The transient impairment induced by ALA deprivation may compromise the organization of neuronal assemblies and result in aberrant neuronal connectivity (lateral connections) to enhance the risk of neurodevelopmental disorders including cerebral palsy.

The neural representation of value is a matter of great debate. In particular, it is not clear whether multiple valuation systems exist, each representing value under different conditions, or whether a single system that uses a “common currency” for the representation of value under many different conditions can be identified. I will present two studies in which we combined experimental methods from behavioral economics with functional MRI to study the representation of value in the human brain. The first study compared choices under two terms of uncertainty: risk, when probabilities of different outcomes are known, and ambiguity, when such probabilities are not known. Our results show that although subjects exhibit markedly different choice behaviors under these two conditions, a single system, consisting of the striatum and the medial prefrontal cortex (MPFC) encodes choice values in both cases. In the second study we used MPFC activation elicited by passive viewing of goods in the scanner to predict subsequent choices between these goods made outside of the scanner. Our predictions were significantly above chance, suggesting that the same valuation system is engaged whether or not choice is required. Based on these results together with previous studies we suggest that the striatum and the MPFC are the final common pathway for valuation – other areas may be differentially involved in encoding value under different conditions, but all of these areas should transfer their output to the final system to guide choice behavior.

Although prominent theories of vision have emphasized dissociations between two visual streams specialized for perception and action, in some situations, the two streams must interact. One such situation is the performance of actions upon remembered objects. Neuropsychological evidence from two patients with occipitotemporal lesions suggests that while immediate actions can be performed using only the dorsal vision-for-action stream, delayed actions require integrity of the ventral vision-for-perception stream. My lab has investigated the interactions between the two streams during delayed grasping using functional magnetic resonance imaging and transcranial magnetic stimulation. Our results suggest that delayed actions re-recruit information about object properties such as shape, size and orientation from the ventral stream and early visual areas at the time the delayed action is performed

It is frequently proposed that conscious perceptual decisions are produced by recurrent interactions among multiple brain areas. Sensory stimuli, which are close to psychophysical threshold or perceptually bistable, induce fluctuating percepts in the face of constant sensory input. Thus, these stimuli provide ideal tools for probing the intrinsic neural mechanisms underlying perceptual decisions, in the absence of extrinsic stimulus changes. I will present human neuroimaging (MEG and fMRI) studies, in which we used this approach for probing the large-scale neural mechanisms underlying decisions about the presence or absence of simple visual features. Our results suggest that neural population activity in parietal, prefrontal, and premotor areas reflects the decision process, and that population activity in extrastriate ventral visual cortex reflects perception. Further, cooperative and competitive long- range interactions, across multiple levels of the cortical processing hierarchy, both seem to underlie simple perceptual decisions.

The off-line processing of acquired sensory information in the mammalian cortex is an example for the unique way biology creates to compute and store information which guides behavior. The relatively short temporal phase in the process (i.e. hours following acquisition) is defined biochemically by its sensitivity to protein synthesis inhibitors. Until recently this negative definition of molecular consolidation did not reveal the details of the endogenous processes taking place, minutes to hours, in the neurons and circuit underlying a given memory. We use taste learning paradigms in order to study this process of molecular consolidation in the gustatory cortex. Recent results, from our laboratory, obtained from genetic, pharmacological, biochemical, electrophysiological and behavioral studies demonstrate that translational control, at the initiation and elongation phases of translation, plays a key role in the process of molecular consolidation. Moreover, this spatially and temporally regulated translation control modifies both general and synaptic protein expression that is crucial for memory stabilization. We propose a model to explain the interplay between regulation of initiation and elongation phases of translation and demonstrate that in certain situations cognitive enhancement can be achieved.