BRAIN Publication Roundup – January 2018

Behavioral modeling and optogenetics elucidate mechanisms of subjective, history-dependent decision bias… Advanced transgenic approach improves light-based study of neuronal circuit dynamics… Current state of computational methods in single-cell functional genomics… Sleep promotes communication between association cortices and hippocampus…

Experimentally manipulating neural activity in behaving mice implicates the posterior parietal cortex in history-based decision bias

Making decisions based on past outcomes is a key adaptive behavior, but how do neural circuits track choice-outcome history and form subjective bias to inform decision-making? At the University of California, San Diego, Dr. Takaki Komiyama and colleagues explored the neural circuit mechanisms that underlie subjective, history-dependent decision bias. They used two-photon calcium imaging to record neural activity in the posterior parietal cortex (PPC), while mice performed an action selection task. The researchers sought to identify idiosyncratic relationships between choice-outcome history and subsequent action selection bias. They showed that action decisions were swayed by past outcomes, and that this behavior was modulated by a subpopulation of PPC neurons. Furthermore, to examine the temporal specificity of these effects, they used precise optogenetic application on PPC (i.e., light directed towards PPC to inactivate the region). Inactivation of PPC before but not during the trial diminished the extent to which an animal depended on past outcomes to make a current choice. This study is the first to demonstrate an essential role for PPC neural circuitry in using an animal’s past history of choices to influence subjective biases on subsequent actions.

A key focus in neuroscience is to decipher neuronal circuit dynamics, for the sake of understanding how the brain drives perception, emotion, cognition, and behavior. Recent progress in genetically-encoded voltage and calcium indicators (GEVIs and GECIs, respectively) has benefited this effort by allowing in vivo monitoring of large identified neuronal populations. Indeed, advanced transgenic approaches now achieve high levels of indicator expression. However, targeting non-sparse cell populations leads to dense expression patterns, preventing researchers from assigning optical signal read-out to individual neurons. At Imperial College London, Dr. Thomas Knöpfel and colleagues have developed a genetic technique for sparse but strong cellular indicator expression, which allows for the resolution and segregation of individual cells and their processes within densely populated neuronal networks. The expression is termed “strong” because the expression level in individual cells is high, and it is termed “sparse” because only a fraction of the neuronal population of interest expresses the fluorescent indicator. The concept resembles Golgi staining, an important histological technique for imaging a low percentage of neurons—in their entirety—within a dense network. Utilizing GCaMP6f (GECI) or the voltage-sensitive fluorescent protein (VSFP) Butterfly 1.2 (GEVI), the researchers achieved strong intensity but sparsely-distributed indicator expression in cortical layer II/III pyramidal neurons within the mouse brain. Morphologies of individual neurons and subcellular structures were successfully resolved. Furthermore, using these fluorescent proteins and a modular transgenic approach, the team produced dual GEVI/GECI neuronal labelling, which enables monitoring of concurrent voltage and calcium activity in either the same neuron or in neighboring neurons. By enabling successful visualization of large neuronal populations, this methodological approach has the translational potential to inform animal models of neurological disease by providing rapid evaluation of cellular indicators of dysfunction.

Although cells of the same ‘type’ can exhibit significant heterogeneity, most genomic profiling studies have analyzed cell populations rather than single cells. However, technological advances have enabled genome-wide profiling of molecular information in single cells. While the field is still in its infancy, single-cell genomics are beginning to make it feasible to create a comprehensive atlas of human cells. At Harvard University, Dr. Aviv Regev and colleagues reviewed the current state of computational methods in single-cell functional genomics, focusing on single-cell RNA sequencing (scRNA-seq). The authors report that biological and technical aspects merge to determine the measured genomic profiles of cells. Sources of variation that affect single-cell genomics data are (1) technical (unwanted) factors that reflect variance due to the experimental process, and random biological factors that are either (2) intrinsic or (3) extrinsic to the molecular mechanisms of gene expression. Experimental, statistical, and computational strategies are used to reduce technical variation, so that biological variation can then be more confidently studied. The authors define a cell’s identity, which is reflected in its molecular profile, as the instantaneous intersection of all factors affecting it. These factors include multiple time-dependent processes that take place simultaneously, the cell’s response to local environmental signals, and the precise tissue location in which the cell resides. The more permanent aspects of a cell’s identity indicate its type (i.e., cell types are organized into taxonomies), but this approach also provides snapshots of the dynamic temporal transitions that cells undergo as well. Combining single-cell genomics with computational models that relate cells to one another in space and time could eventually yield an integrated understanding of how these cells function in health and disease. Importantly, this approach holds potential in contributing to a complete census of cell types in the complex mammalian brain.

Hippocampal sharp-wave ripples, which are high-frequency oscillations seen on an electroencephalograph (EEG) during sleep, have long been thought to be involved in the conversion of short-term memories into long-term memories. Little is known about how ripples transfer hippocampal information to the neocortex, which is required for the consolidation of memories that can be consciously, intentionally recalled (i.e., declarative memories). At New York University, Dr. György Buzsáki and colleagues have developed a microelectrode system (NeuroGrid) for simultaneous electrophysiological monitoring of multiple sites in the rat neocortex. NeuroGrid is a conformable array of tiny linked electrodes that can be laid across a brain area, for large-scale and spatially-continuous monitoring of a large population of neurons. With this technology, the researchers made a novel observation in rats during non-rapid eye movement (NREM) sleep, the longest stage of sleep. Ripples in the association neocortex—a brain area responsible for processing complex sensory information—and the hippocampus occurred concurrently, suggesting communication between the two regions. Then, the team monitored NREM sleep brain activity in rats trained in a cheeseboard maze versus untrained rats that explored the maze randomly. The maze training actually increased the hippocampal-cortical ripple coupling, suggesting that such communication is important for memory consolidation. This is a fundamental discovery, with potential importance to the understanding of memory disorders.