Fast Learners Have High-Dimensional Representations of Neural Activity

Post by Shireen Parimoo

What's the science?

People naturally group new pieces of information into categories to differentiate between them, eliciting distinct patterns of brain activity. One measure of quantifying how information is coded in the brain is by estimating the geometric representation or dimensionality of neural responses. For example, stimuli can be coded in a psychological dimension and represented based on their size and color. Patterns of brain activity can be represented in a high dimension, meaning that the information is coded along many dimensions and that neural responses to each stimulus are likely more distinct from one another. Alternatively, low-dimensionality of neural representations would suggest that there are fewer dimensions along which new information can be coded in the brain, resulting in less distinction between stimuli. Is there a relationship between the dimensionality of neural responses and how fast people can learn new information? This week in Nature Neuroscience, Tang and colleagues used functional magnetic resonance imaging (fMRI) to examine how the dimensionality of neural representations is related to the speed with which people learn new information.

How did they do it?

Nineteen young adults completed various tasks for four days while undergoing fMRI scanning. On each day, they first completed a value-learning task, in which they learned the monetary value of twelve stimuli that ranged from $1 and $12. They were then tested on the stimulus-value associations in a value-judgment task and their response accuracy was used as a measure of learning effectiveness. In the value-judgment task, participants were shown the stimuli and had to indicate if their monetary value was in the top 50% ($7-$12) or the bottom 50% ($1-$6) of the range. In the fMRI data, dimensionality was estimated from label assortativity and separability values of the neural responses. Label assortativity provides a measure of how easily neural responses can be distinguished from each other. In general, higher assortativity is associated with greater dimensionality, but it is also possible for high dimensional representations to have low assortativity. Assortativity was calculated using a linear support vector machine, which is a machine learning technique used to classify information into categories. On the other hand, separability values provide a more direct measure of dimensionality; this was calculated by assigning binary labels to neural activity for each stimulus and using a support vector machine to classify the neural responses according to these labels. High separability values indicate that neural responses are organized in a larger number of dimensions, whereas low values indicate lower dimensionality of representations in neural responses.

The authors examined the relationship between dimensionality and how effectively participants learned the stimulus-value associations. They also conducted a virtual lesioning analysis, in which neural activity in specific brain regions was systematically excluded from the analysis to see which regions contributed to the association between separability and learning effectiveness. Finally, as high dimensionality of neural responses is associated with greater reliance on neural resources, the authors investigated whether resources are allocated efficiently through a balance between high- and low-dimensionality of representations during learning. Participants’ response accuracy was correlated with the dimensionality of representations to specific stimuli (stimulus dimension) during learning and to neural responses independent of the stimulus (embedding dimension), which represents the overall dimensionality of participants’ neural representations.

What did they find?

Response accuracy on the value-judgment task improved over the course of four days, and fast learners – participants who performed above average on the first day – had high accuracy on subsequent days. Assortativity of brain activity was higher among participants who learned fast. That is, stimulus-specific neural responses of fast learners could be distinguished from each other more easily than that of slower learners. Fast learners also had higher dimensionality of neural representations, which is another indication that their representations of the different stimuli were more distinct from one another. In fact, high performance on the first day of the experiment was correlated with high dimensionality on the fourth day, suggesting that effective learning facilitates more precise coding of information in the brain. Interestingly, there was a positive correlation between response accuracy and dimensionality in the stimulus dimension, but a negative correlation in the embedding dimension. This means that fast learners allocate neural resources efficiently by representing all stimuli in an overall low dimensional space, but within that low-dimensional embedding space, specific stimulus-value associations are represented by higher dimensional, and thus more distinct patterns of activity.

What's the impact?

This is the first study to demonstrate how the efficiency of neural responses is optimized during learning, depending on whether it is associated with learning stimulus associations or independently of specific stimuli. The approach of quantifying the dimensionality of representations in neural responses and the finding that dimensionality is correlated with behavioral performance both have important implications for our understanding of how information is represented in the brain and how that affects behavior.

Tang et al. Effective learning is accompanied by high-dimensional and efficient representations of neural activity. Nature Neuroscience (2019). Access the original scientific publication here.

BIN1 Interacts with Tau Protein and Rescues Memory deficits in a Mouse Model of Tauopathy

Post by Amanda McFarlan

What's the science?

Alzheimer’s disease (AD) is a neurodegenerative disorder known to cause deficits in short-term memory, long-term memory and spatial memory. Neurofibrillary tangles, that arise due to the aggregation of hyperphosphorylated Tau proteins, are one of the main biomarkers of AD. Recent studies have shown that the bridging integrator 1 gene (BIN1) is associated with late-onset forms of AD and interacts directly with the Tau protein. This week in the Acta Neuropathologica, Sartori and colleagues investigated the role of overexpressed BIN1 in a mouse model of Tauopathy as well as the underlying molecular mechanisms regulating BIN1-Tau interactions.

How did they do it?

In the first set of experiments, the authors assessed the role of BIN1 expression levels on cognitive function using male and female mice from three different genetic strains: Tau mice (overexpressed the human MAPT gene to produce a Tauopathy model), Tau/BIN1 mice (overexpressed both human MAPT and human BIN1 genes) and control mice. They performed the novel object recognition and Morris water maze at 3, 6, 9, 12, and 15 months to assess the effect of BIN1 overexpression on short-term, non-spatial memory and long-term spatial memory, respectively. In the second set of experiments, the authors investigated the underlying mechanisms that modulate the interaction of BIN1 and Tau — they performed immunolabelling to quantify the level of Tau phosphorylation in the hippocampus. Next, they used proximity ligation assay and primary neuronal cultures to assess the effect of BIN1 overexpression on the amount and localization of BIN1-Tau complexes. It is known that phosphorylation of Tau prevents its interaction with BIN1. Therefore, the authors developed a semi-automated high-content screening approach to identify specific compounds in signaling pathways that may be involved in Tau phosphorylation. Finally, in the third set of experiments, the authors quantified the levels of total and phosphorylated BIN1 in human brain samples from 28 individuals (10 controls, 18 diagnosed with Alzheimer’s disease) with varying degrees of Tau pathology.

What did they find?

The authors found that short-term memory deficits were induced in male and female Tau mice starting at 9 months, while Tau/BIN1 mice showed short-term memory deficits as early as 3 months. Conversely, they determined that male Tau mice displayed long-term and spatial memory deficits at 12 months, while male Tau/BIN1 mice did not display any long-term or spatial memory deficits at any age. Together, these results suggest that overexpression of BIN1 worsens Tau pathology phenotypes for short-term memory deficits but rescues long-term and spatial memory deficits. Next, they revealed that Tau/BIN1 mice had significantly lower levels of Tau phosphorylation in the hippocampus compared to Tau mice (as determined by fewer cells with intracellular inclusions) and that Tau/BIN1 mice had a strong increase in the proximity ligation assay signal (amount of BIN1-Tau complexes) compared to Tau mice and controls.

Together, these results suggest that overexpression of BIN1 increases the number of BIN1-Tau complexes in the hippocampus which decreases the amount of phosphorylated Tau that can form toxic intracellular inclusions (i.e. protective against neurofibrillary tangles). Next, the authors determined that the signaling pathways regulated by Cyclosporin A (an inhibitor of the serine/threonine protein phosphatase Calcineurin) were important for mediating the interaction of BIN1 and Tau. They showed that dephosphorylation of BIN1 by Calcineurin on a cyclin-dependent kinase phosphorylation site at T348 promoted the open conformation of BIN1. Phosphorylation at this site increased the likelihood of BIN1 and Tau interactions. These findings suggest that Cyclosporin A mediates the interaction of BIN1 and Tau via the dephosphorylation of T348 by Calcineurin. Finally, the authors determined that although global levels of BIN1 were unchanged in AD conditions, a higher proportion of overall BIN1 levels were phosphorylated in individuals with AD compared to controls.

What's the impact?

This is the first study to show that the complex regulation of the interaction between BIN1 and Tau is involved in AD pathology. Mouse models revealed that overexpression of BIN1 had neuroprotective effects for Tau phenotypes including long-term and spatial memory deficits, and that this may be regulated by the interaction between BIN1 and Tau. Altogether, these findings provide important insight into the underlying mechanisms leading to AD pathology.

Sartori et al. BIN1 recovers tauopathy-induced long-term memory deficits in mice and interacts with Tau through Thr348 phosphorylation. Acta Neuropathologica (2019). Access the original scientific publication here.

Cell Type-Specific Molecular Changes in Autism Spectrum Disorder

Post by Stephanie Williams

What's the science?

Identifying molecular differences that distinguish the brains of individuals diagnosed with autism spectrum disorders (ASD) from non-autistic brains is important for understanding how the brain develops and functions differently in autism. This week in Science, Velmeshev, Kriegstein and colleagues analyzed the transcriptomes of single cells to identify cell-type-specific molecular changes in ASD.

How did they do it?                                            

The authors analyzed the transcriptomes of neural and glial cells from post mortem brain tissue of children and young adults (aged 4 to 22) diagnosed with autism (N=15) and healthy controls (N=16). The authors used single nucleus RNA sequencing (snRNA-seq) to analyze the transcriptomes of single cells in tissue samples from prefrontal cortex and anterior cingulate cortex, two areas known to be affected by ASD. The snRNA-seq technique allowed the authors to analyze the molecular profile of individual cells. Some patients in the ASD cohort had comorbid sporadic epilepsy, which allowed the authors to create an additional age matched group of controls to compare with this group for further analysis. They performed this analysis to tease apart the differences between epilepsy-related molecular changes and ASD-specific molecular changes. The authors then used data from structured interviews to test whether their cell-type specific findings were related to symptom severity.

What did they find?

The authors identified changes in 17 cell types, and found dysregulated development and signaling of upper layer cortical neurons along with activated astrocytes in the ASD group. The genes that the authors found to be most differentially expressed were in layer 2/3 excitatory neurons and vasoactive intestinal polypeptide-expressing interneurons - specifically, genes responsible for synaptic and neurodevelopment. In non-neuronal cells, the top genes differentially expressed were up-regulated in protoplasmic astrocytes and microglia. The authors found that ASD samples contained more protoplasmic astrocytes. Changes in layer 2/3 neurons and microglia were correlated with symptom severity. This correlation suggests that the molecular changes the authors find in the upper layer cortical neurons are responsible for the behavioral symptoms observed in ASD. Analysis of differences between patients who had comorbid ASD and epilepsy with healthy controls revealed changes in L5/6 corticofugal projection neurons and parvalbumin neurons, confirming that the molecular changes observed in the ASD sample were related to ASD pathogenesis and not seizure activity.

What's the impact?

The authors provide a detailed account of specific cell types that contribute to neural pathways affected in the brains of individuals with ASD. Broadly, the authors replicate and extend previous observations about circuit level dysfunction in ASD. Previous work had shown that there was convergence of ASD on specific cell types during development, and the authors extended this finding by showing that there are also convergent transcriptional changes in adult ASD patients. The convergence of the observed molecular changes in the ASD group onto specific cell types in adults has far-reaching implications as it confirms that there may be a common set of targets for therapeutic treatments.

Velmeshev, et al. Single-Cell genomics identifies cell type-specific molecular changes in autism. Science (2019). Access the original scientific publication here.

The authors’ data can be viewed interactively here