Search

To date, various aspects of connectivity have been inferred from electron microscopy (EM) of synaptic contacts, light microscopy of axonal and dendritic arbors, and correlations in activity. However, until now it has not been possible to relate the complex structural wiring between neurons to the function of individual cells. Using a combination of functional imaging and three-dimensional serial EM reconstruction at unprecedented scale, two papers now describe the connectivity of single cells in the mouse visual system. This study examines how the selectivity of directionally selective retinal ganglion cells may arise from their asymmetry in the wiring with amacrine cells.

In recent decades, advances in technology have enabled the structure of the nervous system to be dissected in greater detail than ever before. In this Opinion article, Denk and colleagues outline why structural information is so important for our understanding of the function of neural circuits and describe new tools and approaches that are improving the structural information that we can acquire.

Advances in cellular imaging have been crucial for improving our understanding of many aspects of neuroscience. Kerr and Denk describe how sophisticated optical imaging techniques allow us to image activity in single neurons or neuron populations in living animals.

A method for staining and embedding the entire mouse brain for electron microscopy is reported. The method results in uniform myelin staining and will allow reconstructions of myelinated long-range axons using serial block-face electron microscopy.

This Technical Report describes an automated algorithm to trace densely labeled neurons and reconstruct their structure, thus providing a new tool in functional connectome analysis.

Flood-filling networks are a deep-learning-based pipeline for reconstruction of neurons from electron microscopy datasets. The approach results in exceptionally low error rates, thereby reducing the need for extensive human proofreading.

Improved electron microscopy methods are used to map a mammalian retinal circuit of close to 1,000 neurons; the work reveals a new type of retinal bipolar neuron and suggests functional mechanisms for known visual computations.

Repeated imaging of the same individual neuron for over a year in mice allows the authors of a new study in this issue to show that presynaptic axon terminals become progressively more stable as the animals age, changing little after 6 to 12 months.

Measurement of in vivo neuronal activity with single neuron and single action potential resolution is important for studying neuronal function. Delivery of a FRET-based fluorescent Ca²⁺ indicator protein using adeno-associated virus results in high expression levels allowing in vivo detection of single action potentials at low firing rates. Griesbeck et al., also in this issue, describe the use of a similar sensor for recording neuronal activity in vivo.

Motion detection by the retina is thought to rely largely on the biophysics of starburst amacrine cell dendrites; here machine learning is used with gamified crowdsourcing to draw the wiring diagram involving amacrine and bipolar cells to identify a plausible circuit mechanism for direction selectivity; the model suggests similarities between mammalian and insect vision.

An improved method for preparing mouse brains for electron microscopy allows reliable tracing of neurites and identification of synapses and appears suitable for whole-brain connectomic reconstruction.

This Resource presents a serial block-face EM dataset of the whole larval zebrafish brain, including automated segmentation of neurons, detection of synapses and reconstruction of circuitry for visual motion processing.

Two-photon calcium imaging has previously only been useful for imaging ongoing neuronal activity in the superficial cortical layers in vivo. Here the authors describe technology that enables imaging of sensory-evoked neuronal activity in layer 5 of adult mouse somatosensory cortex.