|
Scientific Understanding of Consciousness |
Neural Network Local Circuit Connections
Nature 471, 170–172 (10 March 2011) Neuroscience: Towards functional connectomics H. Sebastian Seung Department of Brain and Cognitive Sciences, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139, USA. [paraphrase] To understand the brain, the thousands of synaptic connections made by each of billions of neurons should be mapped and related to neuronal function. Neurons are classified into cell types, which are traditionally defined by location and shape. Rules of synaptic connection that depend on cell type, although important, are not sufficient for understanding neuronal function, because cells of the same type can have diverse functions. In this issue, Bock et al. (see below) and Briggman et al. (see below) report an exciting and pioneering approach to finding rules of connection between neurons that depend on their functional properties as well as their cell type. Serial electron microscopy has been updated in recent years, raising hopes of determining the connectomes of mammalian brains. Briggman et al. and Bock et al. applied improved versions of this method to the mouse retina and primary visual cortex, respectively. They also achieved a new feat — imaging the activity of neurons in a functioning network using two-photon microscopy and calcium-sensitive indicators before mapping their connectivity. Combining serial electron microscopy with two-photon microscopy is a way of directly studying the relationship of a neuron's function to its connections — an example of what might be called 'functional connectomics'. The mammalian retina contains about 50–60 neuronal cell types. Most of what we know about the retina's connectivity is formulated as rules governing these cell types. For example, starburst amacrine cells (SACs) make inhibitory synapses onto direction-selective ganglion cells (DSGCs). Both cell types are involved in computing the direction of movement. Each DSGC is preferentially activated by visual stimuli moving in a particular direction, and different DSGCs have different preferred directions. SACs exhibit functional diversity even within a single cell: each of the dendrites of a SAC functions independently, and each responds selectively to motion in a different preferred direction. Although a DSGC can receive synapses from SAC dendrites with any preferred direction, the number of synapses tends to be much larger for certain directions. Therefore, methods that merely establish whether a SAC dendrite is connected to a DSGC cannot reveal the specificity of connectivity. It is essential to quantify the strength of interaction by counting the number of synapses involved, as Briggman and co-workers have done. In the neocortex, most neurons are excitatory and almost all of these are of the pyramidal type. There are also many types of inhibitory neurons and many rules of neocortical connectivity based on cell type. But pyramidal neurons, even in the same cortical location and layer, can differ in their functional properties. For example, a pyramidal neuron in the primary visual cortex is preferentially activated by visual stimuli of one orientation, and the preferred orientations of pyramidal neurons are diverse. Neural-network theorists have long speculated that inhibitory neurons receive indiscriminate connections from pyramidal neurons and send back inhibition to prevent runaway excitation or to sharpen response selectivity. If this idea is correct, inhibitory neurons have a supporting role in visual computations: they are not primarily responsible for generating selectivity to visual features, but rather they help pyramidal neurons to achieve it. These papers have introduced a general approach to relating the structure of neural networks to their function: search for rules of connectivity that depend on functional properties of neurons. Finding such rules will be more arduous than finding connections between brain regions, or rules of connection between neuronal cell types. But it is crucial for testing the claim that “Nothing defines the function of a neuron more than its connections with other neurons". [end of paraphrase]
Nature 471, 177–182 (10 March 2011) Network anatomy and in vivo physiology of visual cortical neurons Department of Neurobiology, Harvard Medical School, Boston, Massachusetts 02115, USA Davi D. Bock, Wei-Chung Allen Lee, Aaron M. Kerlin, Mark L. Andermann, Sergey Yurgenson, Hyon Suk Kim & R. Clay Reid The Center for Brain Science, Harvard University, Cambridge, Massachusetts 02138, USA Davi D. Bock, Wei-Chung Allen Lee, Edward R. Soucy, Hyon Suk Kim & R. Clay Reid National Resource for Biomedical Supercomputing, Pittsburgh Supercomputing Center, Carnegie Mellon University, Pittsburgh, Pennsylvania 15213, USA Greg Hood & Arthur W. Wetzel [paraphrase] In the cerebral cortex, local circuits consist of tens of thousands of neurons, each of which makes thousands of synaptic connections. Perhaps the biggest impediment to understanding these networks is that we have no wiring diagrams of their interconnections. Even if we had a partial or complete wiring diagram, however, understanding the network would also require information about each neuron's function. Here we show that the relationship between structure and function can be studied in the cortex with a combination of in vivo physiology and network anatomy. We used two-photon calcium imaging to characterize a functional property—the preferred stimulus orientation—of a group of neurons in the mouse primary visual cortex. Large-scale electron microscopy of serial thin sections was then used to trace a portion of these neurons’ local network. Consistent with a prediction from recent physiological experiments, inhibitory interneurons received convergent anatomical input from nearby excitatory neurons with a broad range of preferred orientations, although weak biases could not be rejected. Recent studies suggest that there could be higher-order patterns of connections in cortical networks, such as mutually interconnected triplets of cells, and subnetworks of neurons that are highly connected within a group, but not between groups. Electron microscopy (EM) is an ideal tool for characterizing the highly interconnected structure of cortical networks. Electrophysiology combined with light microscopy and serial-section EM has allowed the inspection of structure–function relationships of single cells within neural circuits, such as the hippocampus, retina, thalamus and cortex. Serial-section EM can be used to follow the three-dimensional contours of neuronal membranes, so that any given axon or dendrite can be traced over hundreds of micrometres. Serial-section EM has been used to examine small volumes of cortical tissue, typically numbering in the thousands of cubic micrometres, in which portions of multiple dendrites and axons were examined for their synaptic relationships. Here we exploited recent improvements in computer speed and storage capacity to perform serial-section EM of a volume that encompasses millions of cubic micrometres, sufficient to contain large portions of the dendritic and axonal arbors of more than 1,000 cells. With this data set, we could attempt a sampling—targeted to a subset of functionally imaged cells—of the dense interconnections found in a cortical network. In particular, we tested a somewhat controversial prediction from recent physiological work that inhibitory interneurons in the mouse primary visual cortex receive dense, convergent input from nearby excitatory (pyramidal) neurons with widely varying preferred stimulus orientations. Here we explored how both the geometry and the function of cortical neurons influence the patterns of connections between them. In the case of excitatory input to local inhibitory interneurons, geometry appeared to dominate over function. This finding may provide an anatomical substrate for a prediction from recent physiological studies of mouse visual cortex, in which inhibitory neurons were found to be less selective than excitatory neurons. Inhibitory interneurons that pool excitatory input could be used to set the gain of orientation-selective pyramidal cells, they might also be involved in modulation of brain state or in attention-dependent normalization of cortical activity. Until it is possible to fully reconstruct large EM volumes, analysis of network connectivity will be limited to a partial sampling of the underlying anatomy. Here we concentrated on reconstructing the axons of functionally characterized pyramidal cells and their postsynaptic targets. Within our sample, we found that 51% of synapses were onto inhibitory targets, despite the preponderance of excitatory neurons in the cortex, and despite reports that 10–20% of the synapses made by pyramidal cells are onto inhibitory targets in cat and macaque. Whether the higher percentage we observed is due to a species difference, or to the fact that we sampled synapses from proximal portions of the pyramidal cell axonal arbors, it resulted in our ability to sample a large number of convergences onto inhibitory targets. We anticipate that the size of serial EM volumes will increase substantially in the near future, owing to increases in imaging throughput and series length made possible by automated techniques. It is fortunate that increases in the dimensions of an EM-imaged volume, and the number of physiologically characterized cells within it, produce combinatorial increases in the number of network motifs that can be analysed in a single experiment. In particular, if a population of neurons is sparsely sampled, the number of interconnections found between them increases as the square of the sampling density. With moderate gains in the number of functionally imaged cells, or in the volumes encompassed by EM reconstructions, insight into the functional logic of cortical networks should therefore increase at an accelerating pace. [end of paraphrase]
Nature 471, 183–188 (10 March 2011) Wiring specificity in the direction-selectivity circuit of the retina Max Planck Institute for Medical Research, Department of Biomedical Optics, Heidelberg 69120, Germany Kevin L. Briggman, Moritz Helmstaedter & Winfried Denk [paraphrase] The proper connectivity between neurons is essential for the implementation of the algorithms used in neural computations, such as the detection of directed motion by the retina. The analysis of neuronal connectivity is possible with electron microscopy, but technological limitations have impeded the acquisition of high-resolution data on a large enough scale. Here we show, using serial block-face electron microscopy and two-photon calcium imaging, that the dendrites of mouse starburst amacrine cells make highly specific synapses with direction-selective ganglion cells depending on the ganglion cell’s preferred direction. Our findings indicate that a structural (wiring) asymmetry contributes to the computation of direction selectivity. The nature of this asymmetry supports some models of direction selectivity and rules out others. It also puts constraints on the developmental mechanisms behind the formation of synaptic connections. Our study demonstrates how otherwise intractable neurobiological questions can be addressed by combining functional imaging with the analysis of neuronal connectivity using large-scale electron microscopy. The computation of motion direction by direction-selective retinal ganglion cells (DSGCs), discovered almost 50 years ago has defied comprehensive explanation, partly because the wiring diagram of the neuronal circuit underlying this computation is still not known in sufficient detail. DSGCs respond strongly to motion oriented along a preferred direction, but not to null-direction (where null direction is 180° from the preferred direction) motion. The magnitude of the inhibitory synaptic input to DSGCs is spatially asymmetric. The main source of this inhibition is starburst amacrine cells (SACs), retinal interneurons that are necessary in the direction-selectivity circuit and release both GABA (γ-aminobutyric acid) and acetylcholine. Asymmetric connectivity between SACs and DSGCs forms the basis of most models for how direction selectivity is computed. [end of paraphrase]
Return to — Neural Network |