Cortical Microcircuits and Dendritic Spines
Summary: Rafael Yuste's personal goal is to reverse engineer the cortex: to understand the structural and functional design principles of neocortical circuits, by taking apart and manipulating their elementary components. To do this, his group focuses on two basic questions: What is the structure of the cortical microcircuit? What is the function of the dendritic spines that mediate the majority of these connections? Yuste's laboratory uses nonlinear optical methods, many of which they developed themselves, to pursue in mouse neocortex the answers to these two questions, which they view as deeply interrelated. In addition, they apply these same optical tools to understand how epileptic seizures propagate within and across cortical circuits, aiming to prevent seizure spread by activating, or inactivating, specific cell types. This work could contribute to an understanding of cortical function and to the design of targeted therapeutics for epilepsy.
Cortical Microcircuits: The "Impenetrable Jungles" of a Universal Computer
Understanding how the neocortex functions is arguably the main challenge of modern science and medicine. The cortex constitutes the largest part of the brain and is the primary site of mental functions and mental disease. No unified theory of how the cortex works exists yet. Nevertheless, the neocortex develops in a stereotyped fashion and is similar, at least superficially, in different cortical areas and in different mammals, and has apparently not changed much since its appearance in evolution. Because of this, it is conceivable that it has a modular structure, duplicated multiple times by evolution, and that each of its modules implements the same basic design and algorithm. The widely different types of computations performed by the neocortex suggest that this algorithm, as the common denominator, must be simple. Thus, evolution may have found simple circuit design that, like a computational "double helix", could make the neocortex a universal computer, analogous to a Turing machine.
Despite a century of anatomical research, the cortical microcircuitry, defined as the intra- and interlaminar projections in a local region of cortex (a column or vertical strip with a cross section of <1 mm2), is still basically unknown. Its circuits' daunting complexity—which Cajal referred to as the "impenetrable jungles where many investigators have lost themselves"—is a result of a heavily interconnected and intermixed network of different cell types. Although it is known that cortical circuits include many subtypes of excitatory pyramidal neurons and of GABAergic inhibitory interneurons, it is still unclear how many cortical cell types actually do exist. Moreover, many cell types, such as the chandelier cells, are so distinct that they seem likely to have a specific function in the circuit. Therefore, the central hypothesis in cortical research is that there is a modular, or canonical microcircuit, built out of many cell types, connected in particular ways, with each presumably implementing a defined circuit function.
Our goal is to use mouse neocortical preparations to search for this canonical microcircuit. To do this, we are reconstructing the inter- and intralaminar connections and testing whether the microcircuits in different parts of the cortex follow the same modular design. Indeed,, we consider the cortical circuit at its most basic nature, as a system of interconnected elements and attempt to infer the basic principles implemented in this circuit from the rules of connectivity and the spontaneous or evoked dynamics generated by isolated cortical preparations. This reductionistic approach is reverse engineering: we take apart the black box, examine the wires to identify the logic that has been implemented in the circuit, and make educated guesses about its function. Although we do not ignore behavior, it is possible that the purpose of the cortex is not to generate behavior directly but to generate mental states, internal representations of the world that are mentally manipulated by the animal and can result in behavior, or not. If the cortex is indeed internally driven and has an intrinsic function—one related to, but not strictly determined by, the outside world—it could be particularly useful to study its microcircuit in isolation. Because of this, we carry out most of our work in brain slices, although we are also interested in how these circuits function in vivo, under sleep or in the awake state.
We use imaging techniques to tackle these questions. A major limitation in past work has been the difficulty of revealing functional connections in large numbers. Over the past 20 years, we have developed optical methods using calcium indicators that enable the imaging of the activity of populations of hundreds or even thousands of neurons simultaneously. This technique relies on the faithful generation of somatic calcium transients by action potentials and is quickly becoming widely used in neuroscience. More recently, we have developed two-photon methods to photoactivate or inactivate neurons, with single-cell resolutions, and also methods to perform these optical manipulations of the circuit in arbitrary spatiotemporal patterns. These experiments are analogous to "playing the piano" with the circuit, and may yield important insights in understanding its transfer function.
Dendritic Spines: The Key to the Logic of the Cortical Circuit?
Our second research project is focused on understanding the function of dendritic spines. These spines, which cover dendrites of most mammalian neurons, are particularly prominent in the cortex, where they exist in large numbers (>20,000 per pyramidal cell. In the neocortex, spines receive almost all excitatory axons, which suspiciously avoid contacting dendritic shafts. Spines therefore mediate essentially all the excitation, so their role is likely to be crucial. We suspect that we will not understand what cortical circuits do until we work out the function of spines. Moreover, cortical circuits are but one example of a larger motif of circuits with spiny neurons (e.g., hippocampus, cerebellum, basal ganglia, cochlear nucleus). Unraveling the function of spines could have wide repercussions in neuroscience.
What is the function of the spine? There are three major hypotheses. Traditionally, spines have been viewed as merely structural devices, enhancing the capacity of dendrites to accommodate more, or more varied, synaptic inputs. This structural function has received support from imaging data that demonstrate that spines move during synaptogenesis and interact with passing axons. A second potential function is biochemical compartmentalization, particularly that of calcium. With this compartmentalization, spines provide the biochemical isolation necessary to implement input-specific synaptic plasticity. This view, which is probably the most popular, identifies dendritic spines as LTP (long-term potentiation) "machines." But there is a third hypothesis: that spines have an electrical function. According to this hypothesis, besides calcium, spines could also compartmentalize voltage, due to a high spine neck resistance. This could alter excitatory potentials, isolating them from each other. The purpose of this electrical isolation could be to prevent the saturation of input integration, yielding a linear arithmetic for input summation.
Based on our recent data, we hypothesize that spines combine these three functions to achieve a single overarching goal: implementing a distributed, integrating circuit. Our argument is that, when viewed from the perspective of the circuit in which they operate, spines enhance connectivity, enabling the linear integration of multiple inputs while endowing them with individual plasticity. Thus, the three aspects of spine function intertwine to build biological neural networks. We think that a better understanding of the function of the spines could "unlock" the cortical circuits, by enabling deep insights into their structure and functional logic.
We use nonlinear optical methods to characterize the structure and function of spines in living preparations and to activate or inactivate spines optically, and we are working on developing novel methods for voltage imaging of spines. Our aim is to better understand their electrical function and how this impacts dendritic integration and synaptic plasticity.
Epilepsy: Mapping and Optically Manipulating Seizures
Our final research project is focused on understanding how epileptic seizures propagate through cortical circuits. The cortex is involved in many seizure disorders, and cortical regions that are apparently healthy are often recruited in seizures that have a distant focal region elsewhere and that secondarily become generalized. Given that cortical circuits are very heterogeneous, it is important to understand exactly how seizures propagate through the cortex: which cell types are recruited, or which cell types prevent the recruitment of a given area? To explore these questions, we apply the same tools that we use in the rest of our work to optically map how different cortical cell types participate in epileptic seizures in vitro and in vivo, with the goal of preventing the spread of seizures by optically activating or inactivating particular key cell types. This information could provide invaluable insights into mechanisms of seizure propagation and lead to better therapeutic approaches for seizure disorders, particularly for those that do not respond to standard pharmacological strategies.
This work was supported in part by a grant from National Eye Institute.
As of May 30, 2012