There has never been a wholesale way of identifying neurons that are monosynaptically connected either to some other cell group or, especially, to a single cell. The best available tools, transsynaptic tracers, are unable to distinguish weak direct connections from strong indirect ones. Furthermore, no tracer has proven potent enough to label any connected neurons whatsoever when starting from a single cell. Here we present a transsynaptic tracer that crosses only one synaptic step, unambiguously identifying cells directly presynaptic to the starting population. Based on rabies virus, it is genetically targetable, allows high-level expression of any gene of interest in the synaptically coupled neurons, and robustly labels connections made to single cells. This technology should enable a far more detailed understanding of neural connectivity than has previously been possible.

The role of different amygdala nuclei (neuroanatomical subdivisions) in processing Pavlovian conditioned fear has been studied extensively, but the function of the heterogeneous neuronal subtypes within these nuclei remains poorly understood. We used molecular genetic approaches to map the functional connectivity of a subpopulation of GABAergic neurons, located in the lateral subdivision of the central amygdala (CEl), which express protein kinase C-delta (PKCδ). Channelrhodopsin-2 assisted circuit mapping in amygdala slices and cell-specific viral tracing indicate that PKCδ+ neurons inhibit output neurons in the medial CE (CEm), and also make reciprocal inhibitory synapses with PKCδ− neurons in CEl. Electrical silencing of PKCδ+ neurons in vivo suggests that they correspond to physiologically identified units that are inhibited by the conditioned stimulus (CS), called CEloff units (Ciocchi et al, this issue). This correspondence, together with behavioral data, defines an inhibitory microcircuit in CEl that gates CEm output to control the level of conditioned freezing.

SUMMARY To establish the mouse as a genetically-tractable model for high-order visual processing, we characterized fine-scale retinotopic organization of visual cortex, and determined functional specialization of layer 2/3 neuronal populations in seven retinotopically-identified areas. Each area contains a distinct visuotopic representation and encodes a unique combination of spatiotemporal features. Areas LM, AL, RL, and AM prefer up to three times faster temporal frequencies and significantly lower spatial frequencies than V1, while V1 and PM prefer high spatial and low temporal frequencies. LI prefers both high spatial and temporal frequencies. All extrastriate areas except LI increase orientation selectivity compared to V1, and three areas are significantly more direction selective (AL, RL, AM). Specific combinations of spatiotemporal representations further distinguish areas. These results reveal that mouse higher visual areas are functionally distinct, and separate groups of areas may be specialized for motion-related versus pattern-related computations perhaps forming pathways analogous to dorsal and ventral streams in other species.

We have constructed a deletion-mutant rabies virus encoding EGFP and find it to be an excellent tool for studying detailed morphology and physiology of neurons projecting to injection sites within the mammalian brain. The virus cannot spread beyond initially infected cells yet, unlike other viral vectors, replicates its core within them. The cells therefore fluoresce intensely, revealing fine dendritic and axonal structure with no background from partially or faintly labeled cells.A fundamental question regarding brain organization is how the various structures are connected to each other. A standard means of addressing it has been to inject into a region of interest some 'retrograde tracer': a substance taken up by the axon terminals of neurons that project to the injection site and which allows either visualization of their anatomy or, in the case of fluorescent tracers, their targeting for physiological study 1,2 . The rise of molecular biology has allowed the use of retrogradely infectious viruses expressing genetically encoded fluorophores 3,4 . In no published cases, however, has the resulting fluorescence been bright enough to consistently provide detailed anatomical information without immunohistochemical amplification and therefore to permit high-resolution identification of live labeled neurons for subsequent physiological study. Here we present a newly created virus that, because of several unique characteristics, does achieve these goals.Rabies virus, which has been used as a trans-synaptic tracer 5 , infects neurons through axon terminals and spreads between synaptically coupled neurons in an exclusively retrograde direction. In earlier work, K.K.C. and colleagues produced a virus that had the envelope glycoprotein gene deleted from its genome but that was grown in complementing cells so that the glycoprotein itself was incorporated into the viral particles'membranes despite the lack of its coding sequence in the viral genome 6 . Such a virus can infect contacted cells normally and, because the glycoprotein plays no role in transcription and replication, can still express its remaining genes and proliferate the viral core within initially infected cells. However, with no means of synthesizing glycoprotein in these cells, the newly created progeny are unable to infect other cells 6,7 , transforming the virus into a first-order retrograde tracer instead of a transsynaptic one. To reveal detailed morphology, we substituted the gene for enhanced green COMPETING INTERESTS STATEMENTThe authors declare competing financial interests (see the Nature Methods website for details).Reprints and permissions information is available online at http://npg.nature.com/reprintsandpermissions/ Note: Supplementary information is available on the Nature Methods website. (Fig. 1a). The result should be a virus that can infect initially contacted cells, replicate its core to high copy number and express high levels of EGFP, but be unable to spread beyond these initially infected cells. We termed this virus, which ...

Understanding the principles of information processing in neural circuits requires systematic characterization of the participating cell types and their connections, and the ability to measure and perturb their activity. Genetic approaches promise to bring experimental access to complex neural systems, including genetic stalwarts such as the fly and mouse, but also to nongenetic systems such as primates. Together with anatomical and physiological methods, cell-type-specific expression of protein markers and sensors and transducers will be critical to construct circuit diagrams and to measure the activity of genetically defined neurons. Inactivation and activation of genetically defined cell types will establish causal relationships between activity in specific groups of neurons, circuit function, and animal behavior. Genetic analysis thus promises to reveal the logic of the neural circuits in complex brains that guide behaviors. Here we review progress in the genetic analysis of neural circuits and discuss directions for future research and development.

Photochemical release (uncaging) of bioactive messengers with three-dimensional spatial resolution in light-scattering media would be greatly facilitated if the photolysis could be powered by pairs of IR photons rather than the customary single UV photons. The quadratic dependence on light intensity would confine the photolysis to the focus point of the laser, and the longer wavelengths would be much less affected by scattering. However, previous caged messengers have had very small cross sections for two-photon excitation in the IR region. We now show that brominated 7-hydroxycoumarin-4-ylmethyl esters and carbamates efficiently release carboxylates and amines on photolysis, with one-and two-photon cross sections up to one or two orders of magnitude better than previously available. These advantages are demonstrated on neurons in brain slices from rat cortex and hippocampus excited by glutamate uncaged from N-(6-bromo-7-hydroxycoumarin-4-ylmethoxycarbonyl)-L-glutamate (Bhc-glu). Conventional UV photolysis of Bhc-glu requires less than one-fifth the intensities needed by one of the best previous caged glutamates, ␥-(␣-carboxy-2-nitrobenzyl)-L-glutamate (CNB-glu). Two-photon photolysis with rasterscanned femtosecond IR pulses gives the first threedimensionally resolved maps of the glutamate sensitivity of neurons in intact slices. Bhc-glu and analogs should allow more efficient and three-dimensionally localized uncaging and photocleavage, not only in cell biology and neurobiology but also in many technological applications.

Preface Incoming sensory information is sent to the brain along modality-specific channels corresponding to the five senses. Each of these channels further parses the incoming signals into parallel streams to provide a compact, efficient input to the brain. Ultimately, these parallel input signals must be elaborated upon and integrated within the cortex to provide a unified and coherent percept. Recent studies in the primate visual cortex have greatly contributed to our understanding of how this goal is accomplished. Multiple strategies including retinal tiling, hierarchical and parallel processing and modularity, defined spatially and by cell type-specific connectivity, are all used by the visual system to recover the rich detail of our visual surroundings.

The specificity of cortical neuron connections creates columns of functionally similar neurons spanning from the pia to the white matter. Here we investigate whether there is an additional, finer level of specificity that creates subnetworks of excitatory neurons within functional columns. We tested for fine-scale specificity of connections to cortical layer 2/3 pyramidal neurons in rat visual cortex by using cross-correlation analyses of synaptic currents evoked by photostimulation. Recording simultaneously from adjacent layer 2/3 pyramidal cells, we find that when they are connected to each other (20% of all recorded pairs) they share common input from layer 4 and within layer 2/3. When adjacent layer 2/3 neurons are not connected to each other, they share very little (if any) common excitatory input from layers 4 and 2/3. In contrast, all layer 2/3 neurons share common excitatory input from layer 5 and inhibitory input from layers 2/3 and 4, regardless of whether they are connected to each other. Thus, excitatory connections from layer 4 to layer 2/3 and within layer 2/3 form fine-scale assemblies of selectively interconnected neurons; inhibitory connections and excitatory connections from layer 5 link neurons across these fine-scale subnetworks. Relatively independent subnetworks of excitatory neurons are therefore embedded within the larger-scale functional architecture; this allows neighbouring neurons to convey information more independently than suggested by previous descriptions of cortical circuitry.