Meet the Rising Stars Scholars Below!

Talk title: Joints: form, function, and the foundation of vertebrate motion

Abstract withheld

Talk title: Mechanisms of myelin membrane expansion in development and neuroplasticity

Key words (5): myelin, exocytosis, neuron-glia interaction, oligodendrocyte

Research Abstract
Myelin sheaths allow fast signaling between neurons across long distances. To make myelin in the central nervous system, oligodendrocytes extend cellular branches that each insulate axons in spiraling layers of membrane to propagate action potentials and provide metabolic support. Loss of myelin leads to impaired cognition and physical disability in diseases, such as multiple sclerosis. Remarkably, myelin in the adult brain is capable of remodeling and regenerating in response to neuronal activity. This plasticity, also known as activity-dependent myelination, can include increasing the number of oligodendrocytes, adding new myelin sheaths, and even adjusting pre-existing sheaths, all of which require spatiotemporal coordination of membrane trafficking in oligodendrocytes.

How does neuronal activity regulate membrane trafficking in oligodendrocytes, and can these mechanisms be harnessed to regenerate myelin and influence learning? My early postdoctoral work determined that the exocytosis machinery VAMP2 and VAMP3 are required for oligodendrocyte membrane expansion during developmental myelination (PMID: 36151203). I used transgenic mouse models (iBot) to conditionally deplete two major drivers of exocytosis, VAMP2 and VAMP3, from oligodendrocytes. Using live-cell imaging of primary oligodendrocytes and electron microscopy of tissue, I found that VAMP2/3 delivers vesicles to the innermost layer of the myelin sheath to drive wrapping and to sheath edges to drive sheath elongation. Through mass spectrometry, I determined that VAMP2/3 incorporates critical myelin-axon adhesion proteins at the axon interface. Altogether, my findings position VAMP2 and VAMP3 to potentiate myelin addition in response to neuronal activity.

To investigate how neuronal activity affects myelin exocytosis, I established a tractable co-culture system to measure exocytosis in oligodendrocytes while manipulating the activity of neurons. Excitingly, my results suggest that activity from glutamatergic neurons doubles the rate of VAMP3 exocytosis in oligodendrocytes. To test how oligodendrocyte exocytosis sculpts myelin in mice, I used optogenetic stimulation of neurons to induce activity-dependent myelin remodeling and measured sheath changes through sparse fluorescent labeling of oligodendrocytes. My data show that oligodendrocyte VAMP2 and VAMP3 are necessary to lengthen myelin sheaths in response to neuronal activity, revealing a mechanism for neuron-glia communication that allows spatiotemporal control for new myelin addition. My future directions will focus on identifying the neuron-derived factor(s) that stimulate oligodendrocyte exocytosis and on investigating how oligodendrocyte exocytosis modulates learning and memory using mouse behavioral assays. Understanding the role and regulation of oligodendrocyte exocytosis will uncover important mechanisms of myelin membrane expansion as an emerging facet of neuroplasticity.

Talk title: Mechanism of olfactory caregiver recognition by social tadpoles

Key words: associative learning, in vivo imaging,

Research Abstract: Infants’ ability to distinguish caregivers from strangers is essential for survival, with olfactory cues driving these behaviors across many species. In neonatal mammals, including humans, caregiver odors trigger proximity-seeking and reduce stress. In developmental disorders like autism, olfactory perception can be disrupted. Proper olfactory cue integration helps newborns secure care and avoid danger, but its neural basis remains largely unclear. This gap in understanding is due to the lack of robust behaviors in animal models. For example, interpreting maternal recognition in mouse pups is complicated by factors like nest separation anxiety. In fact, leaders of scientific research on infant attachment learning have identified a critical need for new animal models to increase our understanding of these neural circuits.

To bridge this gap, I have established in vivo brain imaging in Ranitomeya imitator tadpoles to study the neural mechanisms of parental recognition. Amphibian brains are ancestral to mammals, but nearly all of the core brain regions governing motor patterns and complex social behaviors are conserved. Ranitomeya imitator frogs form monogamous pairs, with both parents involved in tadpole care. Tadpole are visited periodically by both parents and fed an unfertilized egg by their mother which is their primary source of nutrition (much like lactation in mammals). They display a unique behavior to their caregivers that is required to solicit food, typically referred to as begging. However, it is vital for them to identify caregiver from predator before making themselves known with their conspicuous begging display. These tadpoles rely on olfactory cues to recognize and beg to caregivers as blind tadpoles can distinguish between caregiver and a predatory species, whereas ablating the olfactory epithelium causes them to no longer preferentially beg to their caregiver.

The combination of being semi-transparent and having a robust and easily quantifiable behavior towards parents makes R. imitator an ideal system for dissenting the olfactory recognition of caregivers. My work shows that exposure to different olfactory cues during development affects who R. imitator tadpoles differentially beg to suggesting olfactory associative learning. I have found that tadpoles exposed to olfactory cues from a different species, no longer preferentially beg to a female of their own species, suggesting that they may learn the identity of their caregiver with exposure during development. Tadpoles raised with exposure to sibling olfactory cues beg more to frogs of their own species as opposed to a conspecific species (Ranitomeya variabilis), who place their predatory tadpoles into R. imitator pools. However, tadpoles raised in isolation do not differentially beg to either species. To begin to understand the neural circuitry underlying this behavior, I perform in vivo multiphoton imaging of neuronal activity in R. imitator tadpole brains by bulk loading of calcium sensitive dyes into neurons. I  also use neural tracing to map the olfactory circuit projections that may enable this behavior.

This work addresses an important gap in our knowledge of rules of life governing social behavior of infants that are critical for survival and well-being, by spanning levels of analysis from organismal behavior to individual neuron.

Talk title: Barcoding of episodic memories in the hippocampus of a food-caching bird

Key words: Episodic Memory, Neuroethology, Hippocampus,

Research Abstract : Episodic memory is an extraordinary component of cognition which allows a constellation of features, originally co-occuring in a single experience, to later be recalled and flexibly guide behavior. I study episodic memory using chickadees – birds which cache food at scattered locations and use hippocampal-dependent memory to find their caches later in time. My work identifies a neural code which appears transiently during memory formation, uniquely encodes the identity of individual memorable events, and is reactivated during memory recall. Specifically, I developed methods for high-density hippocampal recordings in freely moving chickadees, along with algorithms to automatically track and parse natural caching behavior. Each caching event was represented by spike bursts in a sparse, unique set of neurons. Surprisingly, this ‘barcode’-like activity was uncorrelated with both the place code for the location of a cache, and barcode activity for caches at other locations. Barcodes during a cache were later reactivated when that item was retrieved. Thus, barcodes may be signatures of episodic memories reactivated during memory recall, which could become associated with the contents of an experience during memory formation. These patterns assign a unique identifier to each event and may be a mechanism for rapid formation and storage of many non-interfering memories. This functional hypothesis was examined using a RNN model of episodic memory, which stored memories as attractor states by associating barcodes with activity representing place and food items. The model establishes a simple mechanism to generate barcodes using recurrent dynamics, and demonstrates computational advantages of barcode-mediated memory storage.

Episodic memories are often recalled during decision making to bias choices. My current work exploits the discovery of hippocampal ‘barcodes’ to investigate how memory signals influence navigation decisions. Using a task where chickadees make discrete choices between paths to retrieve a cached item, I am examining hippocampal representations which emerge during deliberation, and their relationship to eventual choice. I am also examining targets downstream of hippocampus which may interface between memory and motor systems, using electrical recording and causal perturbation.

Talk title: Big tau: a neglected tau isoform with pathology-resisting properties
Key words: neurodegenerative disease, Alzheimer’s disease, tauopathies, tau, protein aggregation

Research Abstract: I am interested in finding effective ways to prevent or reverse abnormal protein aggregation in Alzheimer’s disease (AD) and other neurodegenerative diseases. This research interest prompted me to ask: how do certain brain regions remain protected from developing protein aggregates in disease? Every neurodegenerative disease is characterized by pathological changes in its disease protein only in selective brain regions. For example, in AD, the microtubule-binding protein tau becomes aggregated and hyperphosphorylated in the cortex and hippocampus. However, the cerebellum and brainstem remain largely spared from tau pathology despite the universal expression of tau, suggesting that tau may either become cleared more efficiently or resist aggregation in these brain regions.

In my recent work (Chung et al., bioRxiv, 2024; Chung et al., Alzheimer & Dementia, 2022), I found that the understudied splice isoform of tau called “big tau” (Fig. 1A) is highly expressed in the cerebellum and brainstem compared to the forebrain. I found that the big tau protein has unique biochemical properties that can resist key pathological changes occurring to regular tau isoforms in AD, likely due to its unique protein structure. Using both cellular and animal models, I demonstrated that big tau is (1) significantly less prone to aggregation and hyperphosphorylation, (2) more highly ubiquitinated and rapidly degraded (less susceptible to abnormal accumulation in the system), and (3) more strongly binding to microtubules (its canonical function of microtubule-binding less likely to be disrupted), compared to regular tau isoforms (Fig. 1B, C). Moreover, I found that AD patients have a higher level of big tau in their cerebellum (minimal tau pathology), but not in the cortex (severe tau pathology) compared to healthy controls, which suggests an interesting link between the higher level of pathology-resistant big tau and the lack of tau pathology in the brain. Taken together, these findings suggest that alternative splicing leading to the elevated expression level of big tau in the cerebellum and brainstem may contribute to the sparing of these brain regions from tau pathology.

Until now, big tau has not been extensively studied for its biology or disease relevance in the central nervous system, representing a highly uncharted research area. Based on my recent work, I speculate that pathology-resisting properties of big tau can be potentially utilized to prevent tau pathology. Thus, the initial research goals of my lab are to understand the physiological functions and splicing regulators of big tau and to leverage its pathology-resisting properties in designing tau-targeting therapies. Future investigation into big tau pathophysiology will showcase the importance of studying disease protein isoforms and provide new insights into regional brain vulnerability that is common across neurodegenerative diseases beyond AD. Ultimately, I want to lead a rigorous research program that can identify molecular players underlying protein aggregation in neurodegenerative diseases and facilitate the development of effective therapeutic options targeting these protein aggregates.

Talk title: Solenogastres (Mollusca, Aplacophora): biodiversity breakthroughs and insights into coral associations.

Key words: Biodiversity, taxonomy, ecological interactions,

Research Abstract: Solenogastres are an intriguing group of mollusks that represent a significant yet underexplored component of marine biodiversity. They comprise approximately 300 described species, with estimates suggesting the true number could be ten times higher. Solenogastres are distinguished by a worm-like body and absence of a shell, but having the body covered by calcareous sclerites. They are found in all oceans and depths, with many species showing intricate relationships with corals. Furthermore, they are considered key to understanding molluscan evolution. Despite their ecological and evolutionary importance, detailed knowledge about their biodiversity and ecological roles remains limited due to sampling biases and taxonomic challenges. This talk will review the current state of knowledge on Solenogastres, addressing gaps in our understanding of their diversity, distribution, and ecological roles. The aim of this presentation is to highlight the significance of Solenogastres, emphasizing the diverse research opportunities, as well as the range of techniques (from cutting-edge molecular approaches to morphological studies using SEM, histology, micro-CT, etc.) related with their study.

Talk title: Imaging corticospinal correlates of aberrant pain processing and modulation in fibromyalgia using simultaneous spinal cord-brain functional magnetic resonance imaging

Key words: corticospinal neuroimaging, spinal cord-brain fMRI, fibromyalgia, pain modulation, motor-induced analgesia

Research Abstract: Introduction Chronic widespread pain represents the cardinal symptom of fibromyalgia (FM). While physical therapy was found to relieve pain, its modulatory mechanisms in the central nervous system remain understudied. Brain and more recently spinal cord (SC) functional magnetic resonance imaging (fMRI) have advanced knowledge of neural correlates of pain processing. Regions showing increased pain-related activity in FM include thalamus, insula, somatosensory cortices, and SC dorsal horns. Moreover, dysfunctions in cerebral pain-modulatory systems were reported, but their link to the SC is missing. We currently lack understanding of the intricate interplay between spinal and supraspinal networks during pain perception as brain and SC fMRI are conventionally performed separately. Here, we characterize corticospinal mechanisms of aberrant nociceptive processing in FM, by means of combined brain-SC fMRI, and explore descending pain-modulatory effects related to physical activity.

Methods This preliminary study includes 6 female FM patients (age range: 22-62) and 4 female healthy volunteers (HV, age range: 33-62 years). The imaging protocol consisted of the combined brain-SC sequence (EPI pulse sequence; 3T GE SIGNA 750 scanner; 16-channel neurovascular array coil; TR=2.5s, TE=30ms, GRAPPA=2), 30 brain slices (3.4×3.4×5.0mm³) and 13 cervical SC slices (1.25×1.25×5.0mm³) centered at C6 spinal level, axial and sagittal field maps for shimming, and anatomical scans (brain: T1w (1.0×1.0x1.0mm³); SC: T2w (0.7×0.5×0.5mm³)). The experimental design included four conditions (8x each, 13s) in randomized order: stimulation ON/OFF paired with motor task ON/OFF. Thermal stimuli were applied at the right volar forearm (C6 dermatome) at individual temperatures producing a 6 on the Numeric Rating Scale (0-10), using an ATS thermode (Medoc, 3x3cm² surface). The motor task consisted of repetitive isometric right-hand gripping with visual feedback (60% of maximal voluntary contraction) using a Dynamometer (BIOPAC). Blocks were followed by 5s pain rating and 6s rest periods. SC and brain images were processed (slice-time correction, motion correction, spatial normalization, smoothing, physiological noise correction) using FSL and Spinal Cord Toolbox. Activation maps (fixed effects) were cluster-corrected for the brain and uncorrected for the SC, with a threshold of p=0.05.

Results While average pain ratings decreased by 0.6 points for stimulations during motor task vs. stimulations at rest for HV, they only dropped by 0.2 points for FM. Upon stimulation during motor task, greater activity was found within regions implicated in sensory pain aspects (S1, ACC, insula, SC dorsal horns (at stimulation level C6)) in FM compared to HV (red). Enhanced activity was demonstrated in regions implicated in pain regulation (PFC, PAG, pgACC, PCC, motor cortices) in HV compared to FM (green).

Conclusion Our findings expand previous work detecting elevated activity in pain-related brain regions upon noxious thermal stimulation by showing the same pattern in the SC. Crucially, executing a motor task during stimulation reduced pain ratings and engaged brain regions implicated in descending pain modulation in HV. FM, in contrast, exhibited greater activity in brain regions processing sensory pain and SC dorsal horns. Using combined brain-SC fMRI, we identified neural correlates of aberrant pain modulation in FM which can advance development of objective biomarkers of chronic pain.

Talk title: Left, right and everything in-between: evolutionary and biomedical studies in veiled chameleons (Chamaeleo calyptratus)

Key words: Chameleons, left-right patterning, limb patterning, sex determination, genome

Research Abstract: My education started in a small village classroom in Belarus, with five first graders and two second graders, all taught by the same teacher, so it feels surreal to now be applying to faculty positions in the US. My research vision is to investigate the evolution of the early developmental processes and morphological adaptation in amniotes (reptiles and mammals) with a long-term goal of informing the study of human disease. The specific directions my research will take are 1) early embryonic development in amniotes; 2) evolutionary novelties and adaptations; and 3) genome evolution.

Current research on amniote development predominantly focuses on mice and chickens. Unfortunately, the early developmental processes and embryonic morphology of both are highly divergent from human and are not representative of most amniotes. Reptiles showcase an incredible range of morphological and ecological adaptations that span every niche on the planet. The basis for this diversity is established during the earliest embryonic stages, yet we know very little about early development in reptiles, since at the time of egg laying embryos are typically well into organogenesis. As such, veiled chameleons are perfect for the study of early development and evolution in non-avian reptiles, since they are inexpensive to keep, breed well in captivity, and lay large clutches of eggs year-round, with embryos at pre-gastrulation stages at oviposition.

As a postdoc, I discovered that unlike most deuterostomes, which use motile cilia to establish left-right (L-R) asymmetry, chameleons, and reptiles in general, have motile cilia-independent mode of L-R patterning. Instead, large-scale morphological changes precede and likely trigger asymmetric gene expression of the Nodal L-R patterning cascade. To improve genetic tractability of veiled chameleons, I generated a chromosome-level assembly of their genome, which uncovered two Nodal genes, retained from a duplication event in jawed vertebrates. Additional genome analysis revealed a retention of two members of the EGF-CFC family of proteins, which are obligate Nodal binding partners, and are presumed to have roles in both gastrulation and L-R patterning. I further established that veiled chameleons utilize XY based sex determination, independent of environmental temperature. I am currently exploring the exact mechanisms driving sex determination in veiled chameleons.

My future research program will take advantage of the strengths of veiled chameleons as a research organism, which align well with my planned research directions. Thus, 1) We will use earliest stage embryos to investigate the evolution of left-right patterning, which is fundamentally different in chameleons, from most other model organisms (R00 funding); 2) we will determine the mechanisms driving limb patterning in chameleons – an example of evolutionary novelty and morphological adaptation that parallels human split-hand syndrome – a developmental anomaly; 3) we will use the newly sequenced genome to identify the mechanisms governing sex determination in veiled chameleons, adding to our understanding of sex chromosome evolution in other reptiles and amniotes. I am excited to address fundamental questions about amniote development and evolution and I am committed to making veiled chameleon an accessible research organism for both undergraduate and graduate students.

Talk title:The molecular basis of visual system development in chelicerates: from eye loss to major transitions in eye types

Key words: Retinal Determination Gene Network; daddy longlegs; horseshoe crabs; RNAi; CRISPR-Cas9

Research Abstract: Animal eyes are excellent systems to investigate the interplay between genotype and phenotype during major evolutionary transitions: they have appeared and been lost repeatedly across the tree of life; their design converges to common themes of compound- or camera-type optics; and their development shares a core of conserved transcription factors, such as Pax6. Arthropod eyes are of particular interest because these animals have two visual systems (median and lateral), may have both compound and camera-type eyes, and their eyes are diverse in form and function. My research explores one of the megadiverse radiations of Arthropoda, the chelicerates (horseshoe crabs, spiders, daddy longlegs and others), given their phylogenetic position as sister group to the rest of Arthropoda, and that little is known about the genetic patterning of their eyes. In my current research, I aim to understand both the molecular basis behind the diversity of eye types in chelicerates, as well as behind the reduction of eyes in some arachnids (ex: daddy longlegs; cave spiders)

In this talk, I’ll first show that daddy-longlegs (Opiliones), a group once thought to possess only one pair of eyes, in fact additionally retain a pair of vestigial median eyes and a pair of vestigial lateral eyes. This inference is based on neuroanatomical gene expression surveys of eye-patterning transcription factors, opsins, and other structural proteins in the daddy-longlegs Phalangium opilio, as well as gene silencing experiments. The existence of lateral eyes and additional median eyes daddy-longlegs, as well as the extensive characterization of gene expression and function across ontogeny, have important implications for our understanding of the ancestral types of eyes in Chelicerata and the evolution of eye regression in arachnids.

Next, I’ll share my ongoing postdoctoral work about advancing the Atlantic horseshoe crab Limulus polyphemus as a model to study eye patterning in chelicerates. Horseshoe crabs are the only chelicerates with compound eyes, and as such are in a poised to clarify the ancestral mode of eye patterning that gave rise to advanced camera-type eyes in land chelicerates (ex: jumping spiders). Horseshoe crabs are also important models for the neurobiology of vision, but the molecular mechanisms that control their eye development remain largely unknown. I first built a spatial gene expression atlas of conserved eye genes in embryonic eyes and showed that Pax6 is expressed in lateral eyes, but not median eyes. Next, I produced the first genome-edited horseshoe crabs, using CRISPR-Cas9 against a pigmentation gene, resulting in clear eyes. This work sets the stage for understanding the evolution of developmental controls in arthropod visual systems and establishes modern genetic tools for the next frontier of biological discoveries in horseshoe crabs.

My long-term goal is to decipher what types of changes in genetic networks underlie major transitions in the sensory systems. For that end, I aim to reconstruct the gene regulatory network patterning the median and lateral eye visual systems in chelicerates under a broad comparative framework. My future lab will incorporate modern multiomics techniques with experimental genetic manipulations in non-model organisms.

Talk title: Temperature effects on community assembly using nectar microbes of Sticky Monkeyflower

Key words: nectar microbes, priority effects, community assembly, temperature, environmental variation

Research Abstract: Community assembly is an integral process to the development and maintenance of ecological communities. Priority effects, whereby early arriving species outcompete late-arriving ones, are key in shaping community composition. At the same time, the strength of priority effects can be altered by environmental variation through the effects of temperature on each individual species’ growth rate. The major impacts that ongoing climate change has on ecological communities makes it urgent to understand how warming influences community dynamics. We seek to address how temperature and environmental variation affect community assembly by using microbial communities of sticky monkeyflower (Diplacus aurantiacus). In this system, the interactions between yeast and bacteria are governed by the order of their arrival to newly opened flowers. For our experiments, we used the yeast Metschnikowia reukaufii and the bacteria Acinetobacter nectaris, two nectar microbes associated with D. aurantiacus. We introduced priority effects in nectar microbe communities by manipulating the initial population densities of yeast and/or bacteria in artificial flowers kept at three different constant temperatures (15, 25, and 35 °C). We also inoculated sticky monkeyflowers in the field with the same density treatments during the 2023 flowering season (April-June) while measuring the ambient temperature that the flowers experienced.

When looking at priority effects in artificial flowers incubated under controlled environments, we find that the strength of priority effects of bacteria on yeast remains unchanged when temperature increases. On the other hand, the strength of priority effects of yeast on bacteria is weakened at the high temperature treatment (35 °C). Under field conditions, where ambient temperatures fluctuate between 15 and 40 °C during the sticky monkeyflower flowering season, both yeast and bacteria are found in nectar. Preliminary analyses of temperature data find a significant relationship between mean weekly temperature and the concentration (CFU/µl) of yeast found in floral nectar, but not with the concentration of bacteria. Our lab and field results highlight the role that warming has on priority effects and the importance of daily fluctuations in the coexistence of nectar microbes. Future work will consist of comparing community composition by identifying and quantifying the species present in experimental flowers through molecular approaches. This work will help unveil the relationship between priority effects, community composition, and temperature to understand the consequences of climate change in floral nectar and other systems.

A microscopy-based arrayed CRISPR screen reveals the essential host genome for growth and development of intracellular parasite Cryptosporidium

 Key words: host-pathogen interactions, parasitology, cell biology, high-throughput screening, microscopy

 Research Abstract: Cryptosporidium (“Crypto”) is an obligate intracellular parasite of the gut. It is a leading cause of diarrheal-related deaths and disease in children and the immunocompromised, however no vaccine or suitable drug treatments are currently available. It is part of the same Apicomplexan family of parasites which includes the deadly malaria-causing Plasmodium and brain cyst-forming Toxoplasma. However, unlike these other family members that must traverse multiple different hosts to complete their complex life cycles, Crypto does so in the intestinal epithelium of a single host. It is therefore a powerful model to study parasite-host cell interactions in a singular, simpler system. Crypto homes in on epithelial cells of the small intestine, where it creates an intracellular vacuole on the apical side of its host cell, and a host actin-derived “pedestal” directly underneath this vacuole. Here, it undergoes several rounds of asexual replication, after which it must develop into its ‘male’ and ‘female’ sexual forms to complete its life cycle. However, as research on the cell biology of Crypto infections is still a nascent field, the host conditions that enable Crypto to invade, survive, and progress to its sexual stages are largely unknown.

We therefore devised a novel microscopy-based CRISPR-Cas9 screen to examine the effect of the loss of every protein-coding human gene during a Crypto infection. As the screen results are microscopy-based, multiple phenotypes of the infection process could be simultaneously assessed during the knockout of each gene. These include parasite growth (parasite number and size), parasite sexual development (percentage of female parasites), host cell viability, and host F-actin recruited to parasite vacuoles. With this vast and robust dataset, several insights into the basic cell biology of infection of this parasite have been revealed for the first time.

While we discovered several host genes essential for parasite growth and development, we also found many host genes whose loss enhanced infection. Crucially, we discovered a critical junction in the host cholesterol biosynthesis pathway controlling Cryptosporidium growth and sexual development. Surprisingly, infection can either be prohibited or promoted by manipulating a single step in this pathway, by the reduction or accumulation of a host metabolite called squalene. In defining a mechanism for this discovery, we found a critical dependence of this intracellular parasite on host glutathione. Pharmacological inhibition of host squalene synthesis to treat infection is showing great translational potential in mouse models of Crypto infection. Several other host pathways of interest remain to be followed up on from this screen for the study of Crypto-epithelial cell interactions. For example, unexpected hits in host cytoskeletal networks lay the groundwork to dissect both this parasite’s invasion process as well as host proteins required to create its actin pedestal – the mechanisms for which are currently unknown and will be the basis for future work. This microscopy-based CRISPR screen represents a new direction in the study of host-pathogen interactions, where multiple infection parameters can be simultaneously assessed to reveal new insights into host and parasite cell biologies.

Talk title: The FXR1 network acts as signaling scaffold for actomyosin remodeling

Key words: RNA-binding protein; FXR1; structural role of mRNA; non-canonical roles of mRNA; biomolecular condensates

Research Abstract : It is currently not known whether mRNAs fulfill structural roles in the cytoplasm. Here, we report the FXR1 network, an mRNA-protein (mRNP) network present throughout the cytoplasm, formed by FXR1-mediated packaging of exceptionally long mRNAs. These mRNAs serve as underlying condensate scaffold and concentrate FXR1 molecules. The FXR1 network contains multiple protein binding sites and functions as a signaling scaffold for interacting proteins. We show that it is necessary for RhoA signaling-induced actomyosin reorganization to provide spatial proximity between kinases and their substrates. Point mutations in FXR1, found in its homolog FMR1, where they cause Fragile X syndrome, disrupt the network. FXR1 network disruption prevents actomyosin remodeling—an essential and ubiquitous process for the regulation of cell shape, migration, and synaptic function. Our findings uncover a structural role for cytoplasmic mRNA and show how the FXR1 RNA-binding protein as part of the FXR1 network acts as organizer of signaling reactions. 

Talk title: Genome-wide in vivo dynamics of cohesin-mediated loop extrusion and its role in transcription activation

Key words: 3D genome, cohesin, mitosis, epigenetics, transcription

Research Abstract: The organization of the genome in three-dimensional space is highly dynamic, yet how these dynamics are regulated and the role they play in genome function is poorly understood. At the sub-chromosomal level, 3D genome organization is comprised primarily of topologically associating domains (TADs) and chromatin loops, the formation of which is regulated by the interplay of the cohesin complex and CTCF. Most cohesin is thought to reside on chromatin for only ~20 minutes, with certain chromatin loops similarly found to persist for 10-30 minutes. However, the longevity of chromatin loops has not been explored genome-wide, limiting our ability to understand the factors that regulate loop stability and to determine how this stability influences genomic function.

Cohesin facilitates the formation of 3D genome structures through the process of loop extrusion. The cohesin accessory protein NIPBL is necessary for loop extrusion in vitro. Based on this, we engineered a cell line for acute depletion of NIPBL, with the goal of specifically disrupting chromatin loop formation to understand looping dynamics genome-wide. Consistent with the previously observed transient chromatin loops, we found that NIPBL depletion resulted in widespread disruption to looping. However, we also identified a subset of large chromatin loops that were disproportionately stable in the absence of NIPBL. Surprisingly, we found that NIPBL was required for the establishment of these persistent chromatin loops, based on their failure to form when NIPBL was depleted during mitotic exit. Thus, our findings that these persistent chromatin loops require NIPBL for their formation, but not maintenance, suggest that they may be long-lived, contrasting with the widespread view that all chromatin loops have similar longevity. Furthermore, the anchors of these persistent chromatin loops tended to retain cohesin when NIPBL was depleted and were associated with repressive chromatin, indicating that chromatin state may influence cohesin turnover and chromatin loop stability.

Furthermore, we determined the contribution of the 3D genome to the widespread transcription activation that occurs during mitotic exit. From this, we found that NIPBL is necessary for the activation of lineage-defining genes, and likely facilitates activation by enabling a high-level of connectivity with neighborhood enhancers and by reducing insulation of transcription start sites. Indeed, our findings may suggest that NIPBL and cohesin support higher-order genomic structures, such as multi-way contacts between a promoter and multiple nearby enhancers. Given recent evidence that cohesin depletion has a limited effect on enhancer-promoter contacts, such higher-order genomic structures may be the primary mechanism by which cohesin supports and provides specification for transcription regulation.

Our findings provide unique insights into cohesin dynamics genome-wide, and demonstrates the role of NIPBL in regulating cohesin, genome organization, and transcription. In the future, I plan to characterize the interplay of chromatin looping dynamics and chromatin state in the regulation of transcription. Given the association of repressive chromatin with persistent chromatin loops and the link between genome organization and developmental disorders, this direction of research has exciting potential to further our understanding of the requirement for appropriate cohesin dynamics in both healthy and disease states.

Talk title: Curing Diabetes: Hematopoietic stem cell transplantation to promote replacement islet tolerance

Key words: mixed chimerism, transplant tolerance, islet transplant, diabetes, hematopoietic cell transplant

Research Abstract: Type 1 diabetes mellitus (T1D) is an autoimmune characterized by the destruction of insulin-producing cells in the pancreatic islet. Islet transplantation offers an alternative to insulin therapy for type 1 diabetes (T1D), but broader application of this approach has been hindered by multiple challenges, including recurrent autoimmunity, the need for allogeneic donor islet tolerance, and the toxicity of systemic immunosuppression. Mixed hematopoietic cell chimerism after hematopoietic cell transplant (HCT) can induce tolerance to allogeneic tissues, including islets, and correct autoimmunity. However, the toxicity of conditioning protocols for HCT has limited the use of mixed chimerism strategies for T1D and tissue transplantation. Previously, we developed a non-myeloablative conditioning regime using anti-CD117 and other monoclonal antibodies combined with 200cGy total body irradiation (TBI), to clear the hematopoietic stem cell niche, transiently deplete T cells, and promote durable mixed hematopoietic chimerism and allogeneic tolerance. To further reduce conditioning TBI and advance combined allogeneic HCT + islet transplantation as an option for diabetes, we systematically evaluated the effects of JAK1/2 inhibitor baricitinib, Bcl2 inhibitor venetoclax, CD47 blockade, and very low-dose TBI on mixed chimerism in diabetic, immunocompetent B6 mice. Following transplantation of BALB/c bone marrow cells and islets, we assessed hematopoietic chimerism and allogeneic islet graft tolerance in B6 mice across fully mismatched major histocompatibility complex (MHC) barriers. Inclusion of baricitinib, venetoclax, and CD47 with CD117-based conditioning and 10cGy TBI enabled durable mixed hematopoietic chimerism and allogeneic tolerance in diabetic immunocompetent mice and reversed diabetes for >150 days. Thus, we have developed a conditioning regime that supports allogeneic mixed hematopoietic chimerism and islet allotolerance across fully MHC barriers in immunocompetent mice after nearly complete elimination of conditioning radiation, a significant achievement. Adopting these methods clinically could reduce the morbidity associated with allogeneic HCT and islet transplantation, advancing this strategy for diabetes reversal and solid organ and tissue transplantation.

Talk title: Shedding light on the dark-yolk phenotype: Identifying novel regulators of ApoB-containing lipoproteins metabolism using forward genetics in zebrafish

Key words: ApoB-lipoprotein, SLC3A2, screen, zebrafish, dark-yolk

Research Abstract: Apo-B containing lipoproteins (B-lps) are critical for transporting energy-rich lipids through circulation from the liver and intestine to peripheral tissues. B-lps play a key role in the etiology and disease progression of metabolic syndrome which accounts for ~30% of deaths worldwide. However, factors regulating B-lp biogenesis (size, number, turnover, etc.) are incompletely understood. These knowledge gaps have been difficult to investigate due to the limitations of cell culture models for studying multi-organ phenomena and the relative inaccessibility of mammalian whole-animal models to visualize lipoprotein dynamics at the cellular level. The zebrafish presents a unique solution to these challenges. In embryonic and larval stages, maternally deposited yolk lipid is packaged into B-lps in the yolk syncytial layer and then secreted for distribution to peripheral tissues. Larvae deficient in B-lp biogenesis inefficiently mobilize lipid resulting in abnormal lipid storage that reduces light transmission, giving the yolk sac a dark appearance that is easily identified by low-magnification light microscopy. We leveraged this unique and recently appreciated phenotype to identify additional genes involved in lipoprotein processing using a forward genetic screening approach of over 1000 mutagenized families (~0.5M embryos screened). Of the 28 dark yolk mutants identified, 1 could not be recovered, 10 are at existing dark yolk loci (mttp, dgat2, apobb.1, mia2), while the other 17 mutants represent at least 13 unique and novel loci. In order to identify novel loci, we developed WheresWalker, a mapping tool that utilizes whole-genome sequencing data to define a minimal genomic region of interest and provide polymorphic markers for traditional recombinant mapping. Using this tool, we have identified the causative gene for 4 of the unknown mutants. The mutant, zion, maps to slc3a2a, the heavy chain component of slc7 amino acid transporters. Interestingly, several SLC7 genes are associated with abnormal serum lipid profiles in human GWAS datasets. By targeting each slc7 family member in zebrafish using a rapid F0 CRISPR approach, we found that only slc7a7 targeting phenocopies the zion phenotype. Using B-lp and lipid droplet reporter zebrafish, we show that zion mutants produce fewer B-lps and accumulate lipid droplets, consistent with a defect in B-lp biogenesis. Investigations into the mechanism by which slc3a2a/slc7a7 modulates B-lp is ongoing, but are shedding light on a new player in the biogenesis of B-lps. Together, this collection of dark yolk mutants is revealing unappreciated modulators of B-lp synthesis and cell biology and is providing additional candidates for the study of metabolic disease pathology.

Talk Title: QKI regulated alternative splicing hints at a novel mechanism for enhanced exon inclusion

Key words:  QKI, Cardiomyocytes, eCLIP, Alternative Splicing, and U6 snRNP.

Research Abstract: RNA-binding proteins (RBPs) play significant roles in all aspects of RNA processing and consequently gene expression. We and others have recently observed that Quaking (QKI), a known regulator of cell fate, differentially regulates alternative splicing of genes with significant roles in terminal cardiomyocyte differentiation, including contractility and sarcomere formation. The mechanisms through which QKI regulates splicing, however, remain unclear. Utilizing an enhanced CLIP-seq (eCLIP) method and knockout (KO) RNAseq data of differentiating cardiomyocytes at days 0, 4 and 8, we show that QKI predominantly binds in intronic regions to regulate exon inclusion in a position-dependent manner. We find that QKI binding near the 5’ splice site (5’ss) region enhances recognition of exons with suboptimal 5’ ss and particularly poor complementarity to U6 snRNA, a component of the major spliceosome. Intriguingly, we also observed that QKI showed uniquely strong binding to U6 snRNA, when compared to other spliceosomal snRNAs and all 150 RBPs profiled in ENCODE. This led us to hypothesize that the mechanistic basis of exon recognition mediated by QKI may directly involve the U6 small nuclear ribonucleoprotein (snRNP) and its role in the spliceosome. Using minigenes, we mutated the 5’ss to increase or decrease binding potential to U6 snRNA and observed that improved complementarity to U6 snRNA led to significantly enhanced inclusion of alternative exons in a QKI-independent manner. Additionally, QKI co-immunoprecipitated with the tri-snRNP proteins, TXNL4A, RBM42 and SNRNP27 implying a functional role for QKI association  with the tri-snRNP. Gradient co-sedimentation analyses showed that the QKI-U6 snRNA interaction occurs early during U6 snRNP maturation, prior to formation of the U4/U6-di-snRNP or U4/U6.U5 tri-snRNP. Interestingly, mutating the QKI RNA-binding domain abolishes intronic binding but not binding to U6 snRNA, suggesting that QKI utilizes a different binding modality to interact with U6 snRNA. Taken together, our findings allude to a novel mechanism in which QKI binding simultaneously to introns and U6 snRNA serves to stabilize the 5’ss-U6 snRNP interaction facilitating splicing progression, and ultimately splice site selection.

Title: Cognitive Lego, Building Compositional Tasks with Shared Neural Subspaces

Humans, and other animals, can combine simple behaviors to create more complex behaviors. For example, once we learn to discriminate whether a piece of fruit is ripe, we can reuse this ability as a component of a variety of foraging, cooking, and eating tasks. Theoretical modeling has shown artificial neural networks trained to perform multiple tasks will re-use representations and computational components across tasks. By composing tasks from these sub-components, an agent can flexibly switch between tasks and rapidly learn new tasks. This ability to compositionally combine behaviors is thought to be central to generalized intelligence in humans and a necessary component for artificial neural networks to achieve human-level intelligence. Yet, the neurocomputational mechanisms underlying this ability in the brain remain unclear.

In this talk, I will demonstrate how the brain compositionally combines shared subspaces of neural activity to represent task-relevant information across multiple tasks. We trained monkeys to alternate between three compositionally related tasks. Each task required the animal to discriminate one of two features of a stimulus (shape or color) and then respond with an associated eye movement along one of two response axes (Axis 1 or Axis 2). Through large-scale, simultaneous neural recordings in prefrontal, parietal, basal ganglia, and temporal cortices, we found that task-relevant information—about both stimulus features and motor actions—was encoded within shared neural activity subspaces.

As monkeys flexibly transitioned between these three tasks, the degree of shared subspace engagement was linked to the monkeys’ internal belief in the current task state. Importantly, changes in the geometry of representation allowed this flexibility: compression of representation selectively enhanced/suppressed task relevant/irrelevant feature representations.

In sum, these findings elucidate the brain’s ability to seamlessly compose tasks by combining sub-task-specific elements and offer insights into flexible cognitive processes that could inform both neural and artificial network models.

Talk title: “Precision Mapping of E/I Ratio Distributions: Strategies for Signal Propagation Across Cortical Layers”

Key words: connectivity mapping, E/I ratio, network E/I distribution, 2p holographic optogenetics, patch-clamp

Research Abstract: In the cortex, excitatory and inhibitory neurons continuously coordinate to decode sensory inputs. Understanding the roles of various converging and diverging motifs in a circuit, influenced by their proportions on the synaptic currents level, remains experimentally challenging. Previously, techniques such as one-photon optogenetics or electrical stimulation, yielded an average view of the network response, consistently portraying a balanced E/I ratio and a mean-field description of the network. However, these techniques lack specificity, activating large, indiscriminate neuronal groups. The mean overlooks data nuances; distribution extremes can reveal computationally powerful E/I imbalances. Consequently, the detailed distribution of E/I ratios to a single neuron remains unexplored, limiting current circuit models.

To address this, I developed a precise and high-throughput synaptic mapping method combining single-cell resolution two-photon holographic optogenetics and whole-cell patch clamp. I employed a new generation of opsins (ChroME2f/s) for reliable and time-accurate spike generation in presynaptic candidates, enhancing map precision. Additionally, I validated computational tools for unprecedented speed and scale in connectivity mapping. Within minutes, the local connectivity of a few hundred presynaptic candidates to a single postsynaptic cell is inferred using multi-cell targeted photostimulation, compressed sensing, and neural network denoising. Knowing the single-cell connectome of neurons in the Primary Visual Cortex (V1) allows for tracking signal flow across and within cortical layers, among both connected and unconnected excitatory neurons. I recorded excitatory and di-synaptic inhibitory inputs from thousands of differentially organized neuronal ensembles, obtaining the full network E/I ratio distribution. I discovered that this distribution has a long tail, with ensembles that provide uniquely strong excitatory or inhibitory inputs. Previous methods could not detect or test the potentially high computational impact of these ensembles. Furthermore, I found that a single connected neuron can contribute to forming dramatically different E/I ratios, which may be useful for encoding diverse types of information. Finally, I observed that the E/I distribution can shift abruptly toward imbalanced scenarios depending on the distance of the stimulated excitatory ensemble. These connectivity rules describe E/I ratio-based encoding strategies and vary between cortical layers 2/3 and layer 4 of V1. The variability in neuronal E/I ratio distribution may enhance the computational capabilities of the cortex, allowing for greater flexibility and parallelization of cortical processes. In future work, I will investigate the functional role of highly excitatory and inhibitory neuronal ensembles in encoding visual stimuli in the mouse V1.