Seminars

:Abstract

Animal fitness largely depends on its ability to perform certain behaviors and physiological processes at particular times of the daily cycle. Such temporal regulation is mainly the outcome of an internal timing mechanism known as the circadian clock. In fish, and other non-mammalian vertebrates, the pineal gland contains an intrinsic circadian oscillator that drives the circadian rhythms of its hormonal signal (melatonin), and has been considered a key element in the circadian system. Employing a dominant-negative strategy we generated a transgenic zebrafish line in which the molecular clock is selectively blocked in the melatonin-producing cells of the pineal gland. As a result, clock-controlled rhythms of melatonin production in the adult pineal gland were disrupted and the rhythmic expression pattern of the majority of clock-controlled genes in the adult pineal gland was abolished. Importantly, the amplitude of behavioral rhythms was substantially reduced, but not completely eliminated. Thus, the pineal clock plays a key role in modulating circadian rhythms of behavior, but it is not the only regulatory component. To determine the role of other tissues in the circadian clock system and the hierarchal relationship among them, we are currently employing the same dominant-negative transgenic approach to block the clock at other tissues and cell types

:Abstract

Tatyana will describe results showing that neural responses in the hippocampus have a low-dimensional hyperbolic geometry and that their hyperbolic size is optimized for the number of available neurons. It was also possible to analyze how neural representations change with experience. In particular, neural representations continued to be described by a low-dimensional hyperbolic geometry as the animal explored the environment but the radius increased logarithmically with time. This time dependence matches the maximal rate of information acquisition by a maximum entropy discrete Poisson process, further implying that neural representations continue to perform optimally as they change with experience

:Abstract

Autism Spectrum Disorder (ASD) is a complex neurodevelopmental condition with high heritability. Genetic factors are associated with the autistic phenotype in 30–40% of children diagnosed with ASD. Despite substantial advances in genetic research, the precise cause of ASD remains unclear, suggesting that environmental factors may also play a crucial role. Stressors such as ischemia, maternal immune activation (MIA), and toxins have been linked to the development of ASD, either through direct effects on the brain or by inducing mitochondrial dysfunction and endoplasmic reticulum (ER) stress. These processes, in turn, are associated with perturbations in neuronal differentiation, potentially disrupting synaptic protein function and impairing neural circuit formation.
Extracellular vesicles (EVs) are membrane-surrounded nanovesicles routinely excreted by cells, containing proteins, lipids, nucleic acids, and metabolites, and play a role in cell-to-cell communication. EVs mainly originate from the Golgi system, ER, or, to a lesser extent, direct exocytosis from cellular or mitochondrial membranes. The content of EVs extracted from plasma reflects physiological processes occurring in body tissues, including the brain. Thus, disrupted processes, in the brain, characteristic of ASD may be evident in their protein profiles.

:Abstract

:Abstract

:Abstract

:Abstract

:Abstract

:Abstract

:Abstract

:Abstract

:Abstract

Placebo effects demonstrate the powerful influence of beliefs on health outcomes, yet their underlying mechanisms remain only partially understood. In this talk, I will explore how placebo effects shape pain perception, how treatment beliefs are shaped, and how personalization may enhance placebo responses. By integrating insights from neuroscience, cognitive science, and computational modeling, I aim to deepen our understanding of placebo mechanisms and explore novel ways to harness them for improved treatment outcomes.

:Abstract

In recent years, inter-brain synchrony has been increasingly recognized as a fundamental mechanism underlying various forms of social connectedness during interactions. In a recent study using functional near-infrared spectroscopy (fNIRS) we demonstrated that inter-brain synchrony between clients and psychotherapists dynamically changes in the inferior frontal gyrus (IFG) over the course of psychotherapy. However, it remains unclear whether inter-brain synchrony can be actively trained and whether such training translates into increased social connectedness and empathy.

To address this question, we developed a dyadic neurofeedback platform using fNIRS to provide real-time feedback targeting the IFG. Ninety-two participants were randomly assigned into dyads and allocated to either an experimental group or a control group. The training consisted of three sessions over one week.

Results revealed that ratings of social connectedness increased more steeply in the experimental group compared to the control group. Crucially, a significant group-by-session interaction was observed in the IFG, such that the experimental group exhibited a progressive increase in IFG synchrony relative to baseline across sessions, whereas the control group showed no significant change in inter-brain synchrony. This evidence advances our understanding of inter-brain plasticity and highlights its potential applications in fostering socio-emotional well-being.

:Abstract

A group of neurodegenerative diseases, such as Parkinson's disease (PD) and dementia with Lewy bodies (DLB) are often referred to as alpha-synucleinopathies, due to the appearance of alpha-synuclein (aSyn) deposits and inclusion bodies in dead neurons as a molecular phenotype. For many years, these aSyn species or their precursors have been considered as neurotoxic, with a potential link to the early cellular processes that support initiation of the diseases. Regardless of whether these phenotypes are the cause of cell death or side effects related to cell death, one thing is clear: alpha-Synuclein carries multiple physiological functions in neurons, with one famously known to be involved in multiple steps of synaptic vesicle (SV) recycling. aSyn is an intrinsically disordered protein (IDP) that, much like other IDPs, can fold to different structures upon binding to different targets, and by that attain different functions. Specifically, aSyn exhibits a structural bifurcation between its binding-structure-function state when attaching to the outer leaflet of high-curvature membranes (e.g., that of SVs), and its self-associative form that can nucleate into oligomers and amyloid-like fibrils. Under the assumption that the latter state is not contributing to physiology, but rather can influence pathophysiological pathways, it would be great if the "healthy" physiological state of aSyn can be enriched, and one way of conceiving this would be the use of specific molecules that bind aSyn and stabilize it in this state. To do so using a rational design approach, one first needs to ensure there is a target with a specific structure, to which a specific molecule binds preferentially and with high affinity. However, how can there be such a well defined target of the protein in its unbound form is disordered?
In my talk, I will show results from a hybrid experimental & computational approach, which show that the simplest forms of aSyn (e.g., free-form monomer & dimer) have several, yet not many, distinct conformations, associated with its specific functions even before it binds its relevant target that will stabilize that structure. By that, we suggest that aSyn can indeed serve as a target for small molecule design that will function to stabilize physiological forms and destabilize potentially-pathophysiological forms.

​

:Abstract

Hippocampal place cells form a cognitive map of space. My lab studies the role of the local hippocampal microcircuit in place cell formation and stabilization. We use a combination of genetically encoded voltage indicators, holographic illumination, and high-speed microscopes that enable in vivo recordings of supra- and subthreshold voltage dynamics in behaving mice. In this talk, I will describe the technology and present two stories in which we study the interplay of excitation and inhibition in CA1 during behavior. First, by combining voltage imaging with orthogonal optogenetic modulation, we recorded the intracellular dynamics and F-I curves of 4 major hippocampal cell types. We revealed behavioral state-dependent modulation of local network firing patterns, subthreshold oscillatory activity, and neuronal gain. Second, we recorded the intracellular activity of CA1 pyramidal cells and Somatostatin-positive (SST) interneurons during spatial navigation in familiar and novel virtual environments. Repeated imaging in familiar spaces revealed stable and dynamic tuning properties in the intracellular activity of both cell types. Transition to a novel space induced remapping as well as an increase in the global firing of the SST cells. Overall, these studies deepen our understanding of hippocampal microcircuit dynamics and plasticity during spatial navigation and highlight the new prospects opened by recently developed voltage imaging technology.

:Abstract

Recent large-scale recordings have revealed that decision-making dynamics are more complex than captured by classical circuit models. Task-related variables like choice and stimulus are distributed across cortical regions, and traditional two-population models fall short of explaining the observed activity. A growing approach to incorporate the complexity of the observed neural dynamics into network models is to train networks to reproduce recorded neural activity. These models can then be analyzed to uncover circuit mechanisms. I will present two studies using this data-driven modeling framework. First, I will describe a spiking neural network in which only a subset of neurons was trained to match motor cortex activity during a decision task. Task-related activity nonetheless emerged throughout the network, driven by strong, task-independent synaptic couplings. This suggests that local plasticity, together with existing task-independent connectivity, enables distributed task-coding throughout the cortex. Second, I will describe a new class of network models we recently developed that captures both latent coding dynamics and full population activity. Applied to decision-making data, these models reveal that the coding dimensions, although accounting for most of the neural variance and are correlated with the animal’s choice, have little causal effect on behavior, consistent with recent perturbation experiments. Surprisingly, residual dimensions—orthogonal to the coding subspace and accounting for only a small fraction of the variance—exert strong causal influence on decisions. These findings challenge the prevailing view that decision-making is governed solely by low-dimensional coding dynamics. Taken together, our work shows how task-variable coding can spread throughout the brain via non-coding synapses and how low-variance residual non-coding dimensions can causally drive decisions.

:Abstract

Much of the thinking about coding of information in the brain was motivated in the past several decades by reports on neurons with highly stereotyped tuning to the encoded variable. Indeed, neurons are often observed to exhibit smooth, unimodal tuning to the stimulus, centered around preferred stimulus values that vary across the neural population. Recent experiments, however, have uncovered neural response functions that are much less stereotyped and regular than observed previously. Some of the most striking examples have been observed in brain areas associated with the representation of an animal's position in space, such as the hippocampus and its associated brain areas. The classical view has been that hippocampal place cells are active only in a compact region of space and exhibit a stereotyped tuning to position. In contrast to this expectation, however, place cells in large environments typically fire in multiple locations, and the multiple firing fields of individual cells, as well as those of the whole population, vary in size and shape. We recently discovered that a remarkably simple mathematical model, in which firing fields are generated by thresholding a realization of a random Gaussian process, accounts for the statistical properties of place fields in precise quantitative detail. The model captures the statistics of field sizes and positions, and generates new quantitative predictions on the statistics of field shapes and topologies. These predictions are quantitatively verified in multiple recent data sets from bats and rodents, in one, two, and three dimensions, in both small and large environments. Together, these results imply that common mechanisms underlie the diverse statistics observed in the different experiments. The model further suggests that synaptic projections to CA1 are predominantly random.

:Abstract

In the first part of my talk, I will present our recent findings showing that rodents use their whiskers for active sensing in a way that is not purely tactile. We found that whisking against objects produces audible sounds within the mice’s hearing range. Even in the absence of tactile input, auditory cortex neurons responded to these sounds and encode the identity of the touched objects, and mice could identify objects based solely on them. These results suggest that the vibrissa system serves both tactile and auditory functions during exploration, guiding behavioral decisions.
In the second part, I will introduce a novel vibrotactile detection task with reversed contingencies, designed to dissociate perception from action. A cortex-wide optogenetic screen revealed that the premotor cortex is critical for encoding stimulus sensitivity rather than motor execution. However, decision-related signals were not independent of action, indicating that perceptual decisions are formed in an action-coupled, yet context-dependent, manner

:Abstract

Myelin has only recently been appreciated to be a plastic structure, which constantly remodels in response to experience. Therefore, the cellular and molecular mechanisms governing myelin plasticity and how it contributes to aging-related cognitive decline remain mostly unknown. To capture these dynamic changes at the protein level we used BioOrthogonal non-canonical amino acid tagging (BONCAT) which enables dynamic proteome tagging by expression of a mutant tRNA synthetase which incorporates azide-bearing non-canonical amino acids (ncAA) into newly synthesized proteins. Using click-based chemistry, tagged proteins could be identified by mass spectrometry or located spatially by fluorescent imaging. Here we generated oligodendrocyte specific BONCAT mice (OL-BONCAT) and found that BONCAT tagging results in abundant labeling of OPC and mature oligodendrocyte proteins and that these proteins can be found in myelin. Next, we tagged the proteome of OL-BONCAT mice undergoing motor learning on the complex wheel paradigm. Nascent proteins enriched in the learning group were associated with synaptic scaffolding and extracellular vesicle transport and could potentially mediate structural and functional plasticity at the myelin interface. Gaining deep understanding of these mechanisms could provide novel targets for brain rejuvenation by improving myelin health and integrity.

:Abstract

The map of synaptic connectivity between neurons shapes the computations that neural circuits carry - making the identification of the design principles of neural “connectomes” crucial for understanding brain development, learning, information processing, and behavior. 
We present a class of probabilistic generative models for the connectomes of different brain areas in zebrafish, worm, and mouse. Our models accurately replicate a wide range of circuit properties - synapse existence and strength, neuronal in-degree and out-degree, and sub-network motif frequencies -  using surprisingly small sets of biological and physical architectural features. We then show that simulated synthetic circuits generated by our models recapitulate the neural activity and computation performed by the real ones. We extend these generative models to study the development of connectomes over time, and show they accurately replicate the “developmental trajectory” of the connectome of C. elegans, revealing a simpler set of functional cell types than commonly assumed, and identifying distinct developmental epochs. Our findings suggest that connectomes across species follow surprisingly simple design principles and offer a general computational framework for analyzing connectomes, linking their structure to function.

:Abstract

The study of neural mechanisms in AI models provides a powerful, observable lens to generate testable predictions about corresponding mechanisms in the human brain. In this talk, I will present our recent findings on a specific, relatively simple, linguistic task: word repetition. While classic cognitive models explain this task using a well-established dual-route system (lexical and sublexical pathways), the underlying specific neural mechanisms remain unknown. We addressed this by training a fully observable Deep Neural Network (our Neural Word Repetition (NWR) model) on the word repetition task, subjecting it to both behavioral and neural studies. Behaviorally, the NWR model spontaneously reproduced key human psycholinguistic effects, such as the Length Effect and Primacy and Recency Effects. Furthermore, in simulated brain damage experiments (ablation studies), our 'patient' NWR models mirrored typical error profiles observed in human patients following events like a stroke. Finally, we investigated an open question in neuroscience: how are phoneme sequences neurally encoded? This led to novel organizational principles, which generate new, testable predictions about corresponding neural encoding in the human auditory cortex.

:Abstract

RNA editing is a post-transcriptional process that allows for diversification of proteomes beyond the genomic blueprint, a phenomenon called "recoding". However, it is infrequently used among animals for this purpose. I will review the state-of-the-art understanding of recoding by editing and its functional impact, and discuss at length recent results showing that recoding is particularly common in behaviorally sophisticated coleoid cephalopods (e.g. squid and octopus). Combinatorial editing may lead to thousands of protein isoforms from a single gene. Interestingly, recoding is temperature-dependent, allowing for tuning neurophysiological function in response to temperature change. Finally, the trade-off between genome evolution and transcriptome plasticity will be suggested as a partial explanation for the rarity of recoding in most animal species.