Translation elongation factor eEF1A has a well-defined role in protein synthesis. In this study, we demonstrate a new role for eEF1A: it participates in the entire process of the heat shock response (HSR) in mammalian cells from transcription through translation. Upon stress, isoform 1 of eEF1A rapidly activates transcription of HSP70 by recruiting the master regulator HSF1 to its promoter. eEF1A1 then associates with elongating RNA polymerase II and the 3'UTR of HSP70 mRNA, stabilizing it and facilitating its transport from the nucleus to active ribosomes. eEF1A1-depleted cells exhibit severely impaired HSR and compromised thermotolerance. In contrast, tissue-specific isoform 2 of eEF1A does not support HSR. By adjusting transcriptional yield to translational needs, eEF1A1 renders HSR rapid, robust, and highly selective; thus, representing an attractive therapeutic target for numerous conditions associated with disrupted protein homeostasis, ranging from neurodegeneration to cancer.

Pixel and superpixel classifiers have become essential tools for EM segmentation algorithms. Training these classifiers remains a major bottleneck primarily due to the requirement of completely annotating the dataset which is tedious, error-prone and costly. In this paper, we propose an interactive learning scheme for the superpixel classifier for EM segmentation. Our algorithm is 'active semi-supervised' because it requests the labels of a small number of examples from user and applies label propagation technique to generate these queries. Using only a small set (< 20%) of all datapoints, the proposed algorithm consistently generates a classifier almost as accurate as that estimated from a complete groundtruth. We provide segmentation results on multiple datasets to show the strength of these classifiers.

The origin of chordates has been debated for more than a century, with one key issue being the emergence of the notochord. In vertebrates, the notochord develops by convergence and extension of the chordamesoderm, a population of midline cells of unique molecular identity. We identify a population of mesodermal cells in a developing invertebrate, the marine annelid Platynereis dumerilii, that converges and extends toward the midline and expresses a notochord-specific combination of genes. These cells differentiate into a longitudinal muscle, the axochord, that is positioned between central nervous system and axial blood vessel and secretes a strong collagenous extracellular matrix. Ancestral state reconstruction suggests that contractile mesodermal midline cells existed in bilaterian ancestors. We propose that these cells, via vacuolization and stiffening, gave rise to the chordate notochord.

Most neurons function in the context of pathways that process and propagate information through a series of stages, e.g., from the sensory periphery to cerebral cortex. Because activity at each stage of a neural pathway depends on connectivity at the preceding one, we hypothesized that during development, axonal output of a neuron may regulate synaptic development of its dendrites (i.e., retrograde plasticity). Within pathways, neurons often receive input from multiple partners and provide output to targets shared with other neurons (i.e., convergence). Converging axons can intermingle or occupy separate territories on target dendrites. Activity-dependent competition has been shown to bias target innervation by overlapping axons in several systems. By contrast, whether territorial axons or dendrites compete for targets and inputs, respectively, has not been tested. Here, we generate transgenic mice in which glutamate release from specific sets of retinal bipolar cells (BCs) is suppressed. We find that dendrites of silenced BCs recruit fewer inputs when their neighbors are active and that dendrites of active BCs recruit more inputs when their neighbors are silenced than either active or silenced BCs with equal neighbors. By contrast, axons of silenced BCs form fewer synapses with their targets, irrespective of the activity of their neighbors. These findings reveal that retrograde plasticity guides BC dendritic development in vivo and demonstrate that dendrites, but not territorial axons, in a convergent neural pathway engage in activity-dependent competition. We propose that at a population level, retrograde plasticity serves to maximize functional representation of inputs.

Reconstructing neuronal circuits at the level of synapses is a central problem in neuroscience and becoming a focus of the emerging field of connectomics. To date, electron microscopy (EM) is the most proven technique for identifying and quantifying synaptic connections. As advances in EM make acquiring larger datasets possible, subsequent manual synapse identification ({\em i.e.}, proofreading) for deciphering a connectome becomes a major time bottleneck. Here we introduce a large-scale, high-throughput, and semi-automated methodology to efficiently identify synapses. We successfully applied our methodology to the Drosophila medulla optic lobe, annotating many more synapses than previous connectome efforts. Our approaches are extensible and will make the often complicated process of synapse identification accessible to a wider-community of potential proofreaders.

Mapping the connectivity of neurons in the brain (i.e., connectomics) is a challenging problem due to both the number of connections in even the smallest organisms and the nanometer resolution required to resolve them. Because of this, previous connectomes contain only hundreds of neurons, such as in the C.elegans connectome. Recent technological advances will unlock the mysteries of increasingly large connectomes (or partial connectomes). However, the value of these maps is limited by our ability to reason with this data and understand any underlying motifs. To aid connectome analysis, we introduce algorithms to cluster similarly-shaped neurons, where 3D neuronal shapes are represented as skeletons. In particular, we propose a novel location-sensitive clustering algorithm. We show clustering results on neurons reconstructed from the Drosophila medulla that show high-accuracy.

In this work, we propose a learning framework for identifying synapses using a deep and wide multi-scale recursive (DAWMR) network, previously considered in image segmentation applications. We apply this approach on electron microscopy data from invertebrate fly brain tissue. By learning features directly from the data, we are able to achieve considerable improvements over existing techniques that rely on a small set of hand-designed features. We show that this system can reduce the amount of manual annotation required, in both acquisition of training data as well as verification of inferred detections.

Pixel and superpixel classifiers have become essential tools for EM segmentation algorithms. Training these classifiers remains a major bottleneck primarily due to the requirement of completely annotating the dataset which is tedious, error-prone and costly. In this paper, we propose an interactive learning scheme for the superpixel classifier for EM segmentation. Our algorithm is "active semi-supervised" because it requests the labels of a small number of examples from user and applies label propagation technique to generate these queries. Using only a small set (<20%) of all datapoints, the proposed algorithm consistently generates a classifier almost as accurate as that estimated from a complete groundtruth. We provide segmentation results on multiple datasets to show the strength of these classifiers.

The basal ganglia plays a significant role in transforming activity in the cerebral cortex into directed behavior, involving motor learning, habit formation and the selection of actions based on desirable outcomes, and the organization of the basal ganglia is intimately linked to that of the cerebral cortex. In this chapter, we focus primarily on the neocortical part of the basal ganglia. A general canonical organizational plan of the neocortical-related basal ganglia is described. An understanding of the canonical organization of the neostriatal part of the basal ganglia, provides a framework for determining the general organizational principles of the parts of the basal ganglia connected with allocortical areas and the amygdala, and this is discussed. While it has been proposed that the basal ganglia provide interactions between disparate functional circuits, another approach might be that there are parallel functional circuits, in which distinct functions are for the most part maintained, or segregated, one from the other. This chapter, however, is biased toward the view that there is maintenance of functional parallel circuits in the organization of the basal ganglia, but that the circuit contains neuroanatomical features that provide for considerable interaction between adjacent circuits.