ERP

From B-cell decisions to antibody repertoires

PROJECT SUMMARY/ABSTRACT Vaccine responses are highly variable across the population and not without risk for debilitating side-effects. Antibody-mediated immunity is generated by a Darwinian process to generate B-cells that contain B-cell receptors (BCR) that have high affinity for the pathogen-derived antigen, while also eliminating B-cells that happen to react to self-antigens. This process depends on cell fate decisions such as (i) death vs survival, (ii) entry into a proliferative program, (iii) differentiation into antibody-secreting plasma cells. According to clonal selection theory, B-cell fate decisions are made based on the genetically encoded affinity of the the BCR to the antigen (Signal 1) and the cognate T-cells’ TCR to the antigen peptide (Signal 2). However, single-cell resolution studies have revealed that fate decisions of genetically identical B-cells are remarkably heterogeneous. Our studies of the previous funding period revealed that B-cell epigenetic heterogeneity is in fact dynamically controlled: it is generated during the selection process but remains largely stable during the proliferative burst. This leads to our newly proposed Aim 1 to examine how the dynamic control of epigenetic state variability affects antibody responses. An innovative multi-scale model of Darwinian evolution directs and interprets experimental studies by life cell video microscopy in vitro and in immunization studies in vivo. Our previous studies also found that B-cells are capable of sensing the time gap between signal 1 and 2, suggesting a temporal proofreading mechanism for negative selection. This leads to newly proposed Aim 2 which seeks to identify the regulatory circuits that control the stringency of negative selection, as well as contextual germinal center (GC) cytokines that could be manipulable in vivo. These in silico and in vitro studies are followed by in vivo immunization to extend their physiological relevance. Finally, in Aim 3, we will ask what determines the time-gap of signal1 and signal 2, which occur in the immune- induced structure of the GC. We will develop a new model that simulates B-cell fate decisions as a function of their interactions with antigen-presenting stromal cells and T-cells that may be cognate or non-cognate. Model simulations will be used to interpret spatial transcriptomic data to test different adjuvants and predictions will be tested in in vivo immunization studies. With mouse models of inflammation and aging we will examine how adjuvants alter vaccine efficacy and risk.

BKCa Channel Contributions to Cerebellar Regulated TSC-Associated Neuropsychiatric Disorders

Project Summary TSC is associated with neurodevelopmental disability including cognitive disability and autism spectrum disorders (ASD) that make up part of TSC associated neuropsychiatric disorders (TAND). The mechanisms for TAND remain poorly understood but studies have increasingly implicated cerebellar dysfunction in the pathogenesis of cognitive and behavioral deficits in both TSC and other neurodevelopmental disorders. A shared feature is cerebellar Purkinje cell (PC) dysfunction. Changes in intrinsic properties of PCs results in both motor and cognitive/ behavioral changes in disease models and in individuals afflicted by these disorders. Mechanistic underpinnings of these altered properties remain unknown, but a significant emerging body of data implicate ion channel dysfunction as the primary etiology of these deficits. The current proposal seeks to delineate the ion channel contribution to PC dysfunction and to TAND-relevant behaviors. In doing so, these studies will produce significant both short- and long-term impact. Short-term: These proposed studies will provide a mechanistic understanding of the contribution of ion channels to the neuronal dysfunction in the cerebellum that has been demonstrated to be causally linked to abnormal TAND-relevant behaviors. In addition, we will target specific ion channels both genetically and pharmacologically to evaluate the benefits of ion channel restoration on both electrophysiological abnormalities but also the TAND-relevant behaviors observed in the model. Long-term: These studies, thus, provide a framework for subsequent clinically-relevant therapeutic development for TAND. First, these studies will uncover the ability for TAND-relevant behaviors to be improved upon targeting ion channel alterations in TSC. These studies will also define molecular targets on which therapeutic development can be targeted, thereby potentially providing a molecular-informed pipeline for therapeutic development. In addition, these studies will utilize clinically-available, FDA-approved pharmacological agents to target ion channel function and investigate the potential therapeutic benefits for these agents for TAND-relevant behaviors. Thus, these studies will address a core gap in knowledge to achieve a better mechanistic understanding of TAND and to develop therapeutic opportunities to address TAND. These studies will not only reveal previously understudied and novel mechanistic underpinnings for these behaviors but will provide pre-clinical insights into the therapeutic utility of clinically-utilized agents for the treatment of TAND-related behaviors, thus potentially providing both immediate and long-term opportunities for the treatment of TAND. Moreover, although these studies focus on TSC, these mechanisms may prove generalizable beyond TSC and provide a shared basis and therapeutic opportunity for other neuropsychiatric/developmental conditions.

TACTIC: Tuberculosis Active Case Tracking via Interpersonal Connections

PROJECT SUMMARY/ABSTRACT Tuberculosis (TB) remains the leading infectious cause of death worldwide. Interruption of transmission is the most effective strategy to reduce incident infections, yet current approaches often fail to reach individuals for timely testing and treatment. This study addresses that gap by leveraging social networks to identify individuals at highest risk of transmitting TB, specifically, people who use drugs (PWUD). We will evaluate respondent-driven sampling (RDS), a peer7 based community recruitment strategy, to identify TB cases among PWUD and the household contacts (HHCs) of those with TB disease (RDS-TB) in Kampala, Uganda. Conducting this work in a high-prevalence setting such as Kampala where our team has established expertise allows us to overcome recruitment challenges common in settings in the United States while generating findings that are directly translatable. This is particularly relevant given that higher TB prevalence and larger outbreaks in the United States have been associated with the use of methamphetamine, heroin, and crack/cocaine, drugs that we will study. In Aim 1, we will compare the effectiveness and reach of RDS-TB with a traditional clinic-based index case HHC approach for TB case finding. We will screen 2,000 PWUD and their HHCs, estimate the number needed to screen to identify one case of TB disease, and compare the demographic and network characteristics of RDS-TB recruits with clinic-based HHCs. Whole genome sequencing will be used to characterize transmission dynamics. In Aim 2, we will compare the yield of individual and combined TB diagnostic strategies for community-based active case finding. Participants will undergo chest radiography with computer-aided detection, tongue swab testing for TB nucleic acid amplification tests (NAAT), and sputum testing for NAAT and mycobacterial culture. We will identify the minimal combination of tests needed to meet World Health Organization target product profile thresholds for screening. In Aim 3, we will define the conditions under which RDS-based screening can effectively interrupt TB transmission. We will develop an agent-based model informed by social network data from individuals with and without TB, incorporating drug use patterns and demographic characteristics. This project will generate a practical, scalable roadmap for social network–based TB active case finding in high28 risk communities. The approach will be readily adaptable to settings in the United States and will inform strategies to interrupt transmission and advance progress toward TB elimination, in alignment with the NIH Strategic Plan for TB Research.

Targeting disulfidptosis in cancer: mechanisms and preclinical translation

Project Summary Studying regulated cell death is critical for our understanding of cellular homeostasis and tumor suppression. We recently discovered disulfidptosis as a new form of regulated cell death induced by disulfide stress under NADPH-depleting conditions in SLC7A11-high cancer cells. However, in contrast to our deep understanding of other cell death modalities such as apoptosis and ferroptosis, the molecular and metabolic underpinnings of disulfidptosis, along with its therapeutic implications, remain largely unexplored. The objectives of this application are to elucidate the mechanisms underlying disulfidptosis and to therapeutically target this form of cell death in SLC7A11-high cancers. The proposed studies will make extensive use of human cancer cell lines and integrated human cellbased molecular analyses, including metabolomics, proteomics, CRISPR screening, and biochemical studies, to define the metabolic and signaling mechanisms governing disulfidptosis. In addition, select in vivo studies are incorporated in the therapeutic validation components of the project, where tumor growth response, systemic drug exposure and tolerability, tumor microenvironmental influences, and host immune/stromal interactions must be evaluated in an organismal context to ensure translational rigor. Alternative in vitro systems such as organoids may provide useful complementary information on tumor-intrinsic responses, but they cannot fully recapitulate the systemic metabolic stress, pharmacologic exposure, and organism-level therapeutic efficacy required for these studies. It is expected that our proposed studies will reveal novel mechanisms underlying disulfidptosis and identify effective therapies to induce this form of cell death in SLC7A11-high cancers. Our proposal is highly innovative because it focuses on a previously unexplored cell death pathway in cancer therapy. Our proposed studies will have significant impact on both our understanding of the fundamental mechanisms of disulfidptosis and our ability to target this cell death pathway in cancer treatment.

The role of endogenous chimeric mRNA encoded GasderminD fusion proteins in immunity

Project Summary: Programmed inflammatory cell death, or pyroptosis, is a crucial innate defense mechanism that protects hosts against infection and orchestrates subsequent immune responses. Central to this process is Gasdermin D (GSDMD), a protein that forms plasma membrane pores upon activation, enabling the release of pro- inflammatory cytokines such as IL-1β and driving cell lysis. Although GSDMD-mediated pyroptosis has been conventionally understood to be controlled mainly at the post-translational level, through proteolytic cleavage by inflammatory caspases, we have discovered compelling evidence that alternative RNA processing may introduce additional, previously unappreciated complexity in GSDMD regulation. Our laboratories have developed and optimized a highly innovative long-read direct RNA sequencing pipeline, which bypasses conventional cDNA synthesis to avoid artifacts and enables unbiased discovery of native chimeric mRNA (chRNA) in mammalian cells. Using this approach, we have uncovered a remarkably diverse repertoire of chRNA species, including over a thousand unique fusions in murine macrophages and more than two thousand in human inflamed tissues. Among the chRNA found in mice, we identified a chRNA joining the effector domain of GSDMD with a novel C-terminal region encoded by Tmem106a, giving rise to the GSDMD:TMEM106A fusion protein. Functional studies demonstrate that GSDMD:TMEM106A is not only produced in response to inflammatory signals in macrophages but is critical for GSDMD-dependent cytokine release and optimal pyroptosis. Genetic loss of GSDMD:TMEM106A in mice results in reduced cytokine secretion and increased susceptibility to bacterial infection, while in vivo delivery of Gsdmd:Tmem106a mRNA is sufficient for protective immunity. Intriguingly, we have also identified a putative human counterpart, GSDMD:S100A6, which is highly inducible in colon biopsies from patients with inflammatory bowel disease. In this application, we propose a comprehensive exploration of this newly defined class of naturally occurring GSDMD fusion proteins. The specific aims are: (1) to elucidate the subcellular localization, protein-protein interactions, and pore-forming function of GSDMD:TMEM106A during canonical and non-canonical inflammasome activation; (2) to determine the transcriptomic, proteomic, and physiological consequences of GSDMD chRNA expression in vivo during infection, sepsis, and inflammatory disease, and to validate and functionally characterize GSDMD:S100A6 in relevant immune and barrier cell populations. Collectively, this work will establish chimeric splicing as a fundamental source of immunoregulatory protein diversity, redefining the landscape of cell death control in the immune system. By revealing new layers of gasdermin regulation and function, our studies have the potential to identify novel therapeutic strategies for infectious, auto-inflammatory, and immune-mediated diseases.

Targeting the Molecular Crosstalk Between EZHIP and PRC2 in PFA Ependymoma

Project Summary: PFA ependymoma is a rare and aggressive pediatric brain tumor with a poorly understood molecular mechanism. Unlike many cancers, PFA ependymoma exhibits very few genetic alterations. Instead, it is thought to be driven primarily by epigenetic dysregulation. A key player in this disease is the EZH1/2 inhibitory protein EZHIP, which is normally expressed only in germ cells. EZHIP is aberrantly expressed in PFA ependymoma, where it disrupts the function of Polycomb Repressive Complex 2 (PRC2), a master epigenetic regulator of developmental gene repression through deposition of the trimethylated histone H3 lysine 27 (H3K27me3) repressive histone mark. EZHIP-mediated dysregulation of PRC2 involves both enzymatic inhibition and physical stalling of PRC2 on CpG island (CGI) chromatin, leading to a global loss of H3K27me3 levels, an epigenetic hallmark of PFA ependymoma. PRC2 itself is a highly dynamic and intricate complex that assembles into two functional variants, PRC2.1 and PRC2.2. These two variants share a core composed of the catalytic subunits EZH1/2, along with EED, SUZ12, and RBBP4/7, and differ by incorporating distinct accessory subunits. PRC2.1 includes PHF1/MTF2/PHF19, EPOP, and PALI1/2, while PRC2.2 features AEBP2 and JARID2. Our preliminary data reveal intriguing molecular crosstalk between EZHIP and multiple PRC2 components, suggesting potential competitive or cooperative interplay. The ability of EZHIP to inhibit PRC2 partly stems from its mimicry of the oncohistone H3K27M, which harbors a lysine-to-methionine mutation that causes diffuse midline glioma, another devastating brain tumor in children, where PRC2 activity is also globally suppressed. However, the precise, EZHIP-specific mechanisms behind PRC2 dysregulation in PFA ependymoma remain largely unexplored. Our work aims to uncover these elusive mechanisms using a powerful combination of structural biology, biochemistry, and genomics approaches. Ultimately, we aim to identify therapeutic strategies that disrupt the pathogenic EZHIP–PRC2 crosstalk and restore the normal H3K27me3 epigenetic landscape. Specifically, in Aim 1, we will determine the structural and biochemical mechanisms underlying the enzymatic inhibition of the PRC2 core complex by EZHIP. In Aim 2, we will elucidate the molecular basis of EZHIP-mediated stalling of PRC2 on CGI chromatin, involving PRC2 functional variants. In Aim 3, we will explore an exciting mechanism-based therapeutic strategy to overcome PRC2 enzymatic inhibition and chromatin stalling induced by EZHIP.

Th17 plasticity in rheumatoid arthritis

ABSTRACT The objective of this grant application is to explore the plasticity of Th17 in arthritis. Interleukin-17A (IL-17A) producing Th17 are present in the blood and synovium of patients with rheumatoid arthritis (RA). However, targeting of IL17A has been insufficient to control joint inflammation of RA patients. One potential scenario is that in the context of worsening RA joint inflammation, Th17 undergo conversion into pathogenic IL17A- negative cell populations, collectively called exTh17. The conversion of Th17 into exTh17 has been documented in the context of neuroinflammation, colitis, and infection. However, the occurrence of Th17 plasticity in autoimmune arthritis and its potential role in perpetuating synovial inflammation has remained mostly unexplored. We generated a novel fate-mapping mouse model of autoimmune arthritis, which allows to follow the conversion of Th17 into exTh17, and collected preliminary data suggesting that Th17 undergo significant loss of IL17A expression and conversion into exTh17 in the context of synovial inflammation. We also identified exTh17 signatures which might help exTh17 perpetuate joint inflammation despite their loss of IL17A expression. Here our objective is to further elucidate intrinsic (Aim 1) and extrinsic (Aim 2) mechanism of Th17-exTh17 conversion and exTh17-mediated joint inflammation, and explore the potential role of exTh17 in RA interstitial lung disease (ILD, Aim 3) a feared and often untreatable complication of established RA. Our long-term goal is to leverage the knowledge of local immune cell phenotypes and how they change at various stages of disease to enable stage-specific and personalized therapies of RA which minimize non- specific immunosuppression.

Investigating the nonlinear complex dynamics of the tuft cell-microbiome cross-talk: the impact of feedback loops on immune regulation, microbial modulation and response to tissue insults

Project Abstract Tuft cells (TCs) are specialized chemosensory epithelial cells that are emerging as critical regulators of intestinal homeostasis. Named over 70 years ago based on their distinct morphology, a defined function for TCs was only elucidated in the last decade. TCs in the small intestine sense succinate from helminths to initiate type 2 immune responses that mediate parasite expulsion. Recently, we discovered a novel physiologic function for TCs in the colon, where their role had been considered minimal. Succinate, a key microbial metabolite, is produced by colonic microbiota as both a precursor to other metabolites and a cross-feeding fuel source for pathogens. TCs respond to succinate by secreting interleukin-25 (IL-25), which activates type 2 cytokine- producing lymphocytes (T2Ls), amplifying TC expansion and reinforcing barrier function. We recently demonstrated that this SPB–TC–IL-25–T2L feedback loop is essential for protection against pathogen-induced colitis. Our preliminary data further suggest that TCs actively promote colonization by succinate-producing bacteria (SPBs), establishing positive feedback on TC-supporting microbes, while other epithelial cells such as goblet cells (GCs) and Paneth cells (PCs) may exert complementary or counterbalancing influences. Supported by new modeling insights, we hypothesize that these epithelial–immune–microbiome interactions form coordinated feedback loops that collectively optimize intestinal resilience. These loops may create a dynamic, multi-stable system that flexibly transitions between homeostatic and hyperplastic states, buffering against microbial fluctuations and pathogenic insults while preventing uncontrolled type 2 inflammation. Using a combination of mathematical modeling and experimental validation, we will develop a multi- layered systems framework to explore how epithelial–immune–microbial feedbacks shape resilience or breakdown in clinically relevant models of colonic infection and inflammation. Our three Aims will (1) develop, calibrate, and validate a mathematical model that integrates TCs, GCs, PCs, SPBs, and SCBs; (2) define the immunological circuits governing epithelial–microbiome equilibrium; and (3) determine how epithelial feedbacks regulate microbial community structure and resilience. In line with NIH’s new initiative to prioritize human-based research, our proposal combines computational modeling, human colonic organoids, and complementary mouse models. Organoid experiments will provide human-relevant data for model calibration, while in vivo studies validate systemic predictions, ensuring both rigor and translational relevance while minimizing reliance on animal models. This work will generate interoperable models that integrate epithelial, microbial, and immune networks, providing predictive insight into intestinal outcomes under homeostatic, infectious, and inflammatory conditions and informing therapeutic strategies for microbiome-targeted interventions.

Validating Causality of Disputed Mitochondrial Variants in Inborn Errors of Metabolism

PROJECT SUMMARY Primary mitochondrial disease (PMD) encompasses multi-systemic disorders caused by impaired mitochondrial function. PMDs arise from pathogenic variants in either nuclear genes encoding mitochondrial proteins, or in the mitochondrial DNA (mtDNA) genome. Clinical diagnosis is challenging due to phenotypic heterogeneity, underscoring the importance of genetic diagnosis. ACMG/AMP guidelines provide a well-established framework for interpreting nuclear DNA variants while diagnosing genetic diseases. Their application to mtDNA variants, however, remains challenging due to unique features of mtDNA: maternal inheritance, heteroplasmy, threshold effects, and effect of transfer or ribosomal RNA rather than coding variants. To address these challenges, the ClinGen Mitochondrial Disease Nuclear and Mitochondrial Variant Curation Expert Panel, co-chaired by the Multi-PIs of this study, developed widely adopted ACMG/AMP revised guidelines for mtDNA variant interpretation. Over the past five years, this global expert panel has curated more than 280 mtDNA variant. Because of the lack of functional data of individual mtDNA variants in the literature, 23 previously reported pathogenic (P) variants were classified as Variants of Uncertain Significance (VUS), hindering definitive PMD diagnoses and therapeutic development. This R01 project aims to resolve the pathogenicity of these 23 mtDNA VUS through functional validation, leveraging advanced mtDNA base editing and single-cell genomics in in vitro and in vivo models. In Aim 1, we will create human 143B cell line models for 20 VUS using cutting-edge mtDNA editing techniques, optimized for efficiency and minimal off-target effects. Single-cell genomics (mtscATAC-seq and scRNA-seq) will assess heteroplasmy and genomic changes, while functional assays will evaluate mitochondrial ATP production, oxidative phosphorylation, membrane potential, and redox stress. Aim 2 will develop zebrafish models for 17 conserved VUS, characterizing phenotypic and mitochondrial outcomes to corroborate in vitro findings and PMD patient phenotypes. This study will clarify longstanding uncertainties regarding the pathogenicity of these mtDNA VUSs which were nonetheless reported to be pathogenic with often strong genetic evidence but limited functional data. The study will also establish valuable cell and zebrafish models and provide mechanistic insights of PMDs. The resulting resources will be shared with the scientific community to accelerate research and therapeutic advancements for novel precision medicine approaches for PMDs.

Examining the foundations of reading comprehension: a longitudinal study of brain and behavior starting in infancy

SUMMARY Reading comprehension (RC) is one of the most complex skills that we utilize daily and is crucial for functioning in modern society, but despite its significance for academic achievement, employment prospects, and mental health, many children and adults do not exhibit proficient RC abilities. New theoretical models aiming to explain variability in RC suggest a dynamic interplay and co-development among ‘precursor’ foundational and cognitive- linguistic skills, interacting with environmental and socio-ecological factors across the developmental timeline of learning to read. Behavioral and neuroimaging studies in school-age children have demonstrated critical mechanistic support for these multifactorial RC models by identifying the developmental trajectories of precursor skills and further showing that brain areas, tracts, and networks typically underlying language and cognitive skills are also involved in RC. Nevertheless, the precursor skills that support RC start developing in infancy and the brain correlates underlying these precursors begin to develop in utero, which suggests that typical and atypical RC developmental trajectories could diverge long before school age. As such, examining RC development using a multifactorial, longitudinal approach that includes brain and behavior starting in infancy is critical for developing theoretical frameworks that can inform early preventative and intervention strategies. Here, we propose a comprehensive longitudinal study of RC development in which we examine direct and indirect effects on RC from brain, behavioral, familial risk, and environmental data from infancy to adolescence. To achieve this goal, we will combine two existing longitudinal cohorts, one ranging from infancy to late childhood (n = 174) and the other from preschool to early adolescence (n = 137). By applying state-of-the-art pediatric neuroimaging analyses, multiple indicator growth model structural equation models, and an innovative behavior- brain co-development measurement index to this unique, combined dataset, we will be able to identify brain and behavioral measures in infancy that directly and indirectly support subsequent RC development (Aim1). We will further characterize how longitudinal trajectories of behavioral measures as well as brain structure, function, and white matter organization contribute to RC development and how familial risk and environmental factors shape these trajectories (Aim 2). Finally, we will examine how the co-development of brain and behavior, as measured with an innovative co-development index, relates to subsequent RC (Aim 3). If successful, we will contribute the first multifactorial longitudinal model of RC development comprising direct and indirect effects from brain, behavior, brain-behavior co-development, familial risk, and environmental measures beginning in infancy. Understanding RC development using a multifactorial longitudinal lens will be crucial for building theoretical models and developing experimental designs focused on early preventative and intervention approaches long before the start of formal schooling.

TAR RNA binding to INI1/SMARCB1 and its role in HIV-1 transcription and latency reactivation

Abstract The goal of this application is to study the role of interplay between the components of chromatin remodeling SWI/SNF (BAF complex) and HIV-1 transcription machinery, focusing on the interaction of a BAF component, INI1 (Integrase Interactor 1) with TAR RNA. HIV-1 reservoirs are a mixture of latent cells harboring proviruses silenced at transcriptional level. Cure strategies need a deeper understanding of HIV-1 transcriptional regulation. HIV-1 transcription, initiated by RNA Pol II, pauses producing short TAR transcripts. pTEFb recruitment to TAR by Tat overcomes this transcriptional pause, facilitating elongation. Beyond Tat, the action of chromatin remodeling complexes (CRCs) is required to facilitate elongation. The BAF complexes CBAF and PBAF play distinct roles. While CBAF represses proviral transcription by maintaining nucleosomes in an unfavorable state, PBAF remodels nucleosomes to facilitate elongation. INI1 is a component of both CBAF and PBAF, and its role in transcription is not fully understood. INI1 was identified as a binding partner for HIV-1 integrase (IN) and exerts multifacted roles in virus assembly, production and morphogenesis. INI1 has multiple functional domains. IN binding Rpt1 domain structurally mimics TAR RNA & is necessary for late events. We have made a novel observation that another domain of INI1, the N-terminal Winged Helix DNA binding domain (WHD) specifically binds to TAR RNA and that this interaction is necessary for mediating HIV-1 transcriptional elongation. These exciting results suggest that different functional domains of INI1(Rpt1 and WHD) involved in “TAR RNA mimicry” or “TAR RNA binding” regulate distinct stages of replication. We hypothesize that INI1 WHD domain-TAR interaction is necessary for recruitment of PBAF to HIV-1 LTR for transcriptional elongation and latency reactivation. Disrupting this interaction results in transcriptional repression. We will investigate the role of this novel INI1:TAR RNA interaction in HIV-1 transcription and latency reactivation. This is a multi-PI application involving Drs. Kalpana (HIV-1 virologist), Heng (NMR biophysicist) and Zou (computational biologist/protein-RNA structure). In Aim 1, we will characterize INI1-WHD:TAR interaction in vitro and in vivo via molecular/genetic analyses (Kalpana/Heng). We will employ alanine scanning mutagenesis based on WHD NMR structure to test WHD:TAR interaction. We will use biophysical & biochemical approaches to probe TAR structural elements required for this interaction. In Aim 2, we will employ computational modeling and NMR to determine the structure of INI1- WHD:TAR RNA complex (Zou/Heng). In Aim 3, we will determine the role of INI1:TAR interactions in HIV-1 transcription, latency reactivation and mechanism of action (Kalpana). We will analyze the effect of TAR- Interaction-Defective (TID) INI1 mutants on transcription of LTR-reporters and full-length HIV in INI1-/- cells. Latent cells in which TID-INI1 mutants are knocked in (KI) will be used to assess effect on reactivation via RNA-FISH and qRT-PCR assays. Our studies will establish INI1:TAR interaction as a drug target. Inhibiting this interaction could block latency reactivation promoting deep latency and advancing cure strategies.

Engineering inducible morphotype switching control in Mycobacterium abscessus for investigating infection outcomes and discovering pathophysiological-targeted treatments

PROJECT SUMMARY Antibiotic-resistant nontuberculous mycobacteria (NTM) infections are rising at a rate of 8% each year and account for ~$1.7 billion in annual U.S. healthcare costs. Mycobacterium abscessus (Mabs), the most common rapidly growing NTM infection, is notoriously nicknamed the “antibiotic nightmare” for its extensive intrinsic and inducible broad-range multidrug resistance to antibiotic countermeasures. As part of its natural infection cycle, Mabs undergoes a morphotypical conversion from smooth to rough, characterized by irreversible genetic changes resulting in the loss of cell envelope glycopeptidolipids (GPLs). This morphotypic conversion is intimately associated with disease progression, ultimately leading to debilitating, refractory Mabs pulmonary disease. Specific stimuli triggering Mabs morphotypical conversion are unknown, thus preventing directed investigations into morphotype-specific immunological responses and the discovery of morphotype-specific therapeutic targets. This project leverages cutting-edge molecular genetic tools, including CRISPR (clustered regularly interspersed short palindromic repeats) interference (CRISPRi) and inducible knockdown control of CRISPRi via the anhydrotetracycline-inducible TetR-regulated promoter-operator system, to create six unique, reversible Mabs smooth to conditional rough morphotype strains. These molecular morphoswitchable strains allow precise investigator-mediated on-off control of Mabs surface GPLs, enabling investigations into Mabs morphological plasticity, unique pathophysiology traits associated with each morphotype, and the complex interplay between Mabs and morphotype-specific immunological responses. In Aim 1, we implement CRISPRi inducible knockdown tunable control of Mabs morphotype switching by targeting six, independent genetic targets directly involved in GPL biosynthesis (mps1, mps2) or transport (mmpS4, mmpL4a, mmpL4b, gap) and validate in vitro morphoswitching. In Aim 2, we establish and confirm Mabs morphoswitching and intracellular growth in infected THP-1 macrophages. Subsequently, we evaluate differential and distinct innate cellular immune responses elicited by Mabs smooth and Mabs conditional rough morphotypes during intracellular infection in human primary monocyte-derived macrophages. Collectively, these studies create a suite of characterized and reversible Mabs smooth and conditional rough morphoswitchable strains with controlled, regulated, and on- demand expression of Mabs surface GPLs. By enabling precisely timed and controlled induction of the Mabs conditional rough morphotype during intracellular growth, we can molecularly dissect and investigate fundamental Mabs host-pathogen interactions and immunological responses that so substantially influence negative clinical outcomes.

A dynamic regulatory mechanism controlling bacterial persister formation and resuscitation within biofilms

PROJECT SUMMARY Persisters present a major challenge in clinical infection treatment and recurrent infection management. A continued effort towards a better understanding of the molecular mechanisms of persister formation and resuscitation is needed to provide novel treatment strategies for the control of chronic infections and problems related to persisters. Unlike resistant bacteria, persisters are genetically identical to their susceptible counterparts, and this phenotypic state is inherently transient and shifts in response to environmental conditions. Therefore, it is essential to use an approach tailored to the transient and rare nature of this phenomenon. Pseudomonas aeruginosa (Pa) is an important human pathogen frequently implicated in both acute and chronic infections. Persisters have been identified in both Pa planktonic and biofilm modes of growth, with higher frequencies of persister formation being observed in biofilm, especially in the interior of the mature biofilm structure. In this study, we obtained the first high-resolution single-cell transcriptomes of persister and resuscitated cells isolated directly from the interior of mature biofilms. The results led to the identification of a previously uncharacterized transcriptional regulator that controls persister formation and resuscitation. This regulator, named PriR here, is conserved in Pseudomonas species and has homologs in two critical bacterial pathogens, Acinetobacter baumannii and Enterobacter cloacae. We showed that PriR has a dynamic spatiotemporal gene expression profile, and its expression directly correlates with and causes persister resuscitation. In this application, we propose two specific aims to investigate this novel regulation mechanism of persister formation and resuscitation. Aim 1 will identify the physiological effects of this novel regulatory system on antibiotic tolerance in vitro and in hosts using the Drosophila melanogaster biofilm infection model. Aim 2 will determine its molecular regulatory mechanism via ChIP-seq and RNA-seq, and analyze the putative PriR- controlled genes on persister formation and resuscitation in additional clinically-relevant Pa strains. The insights gained from this proposal will provide crucial new information about the dynamic regulatory mechanism of persister formation and resuscitation. The PriR-controlled resuscitation mechanism could be a promising target for persister eradication approaches by re-sensitizing persister cells to conventional antimicrobials or preventing persister formation. Understanding this novel regulatory system that controls bacterial persister formation and resuscitation could provide new drug targets and/or treatment strategies for persistent infections.

Chromatin-Based Mechanisms Linking Transcriptional Dysregulation to Genome Instability in Neurodevelopmental Disorders.

PROJECT SUMMARY/ABSTRACT Neurons depend on a finely tuned interplay between chromatin regulation and genome maintenance, yet they are acutely vulnerable to DNA damage generated during activity-dependent transcription of long, synaptic genes. Disruption of this balance is increasingly recognized as a driver of neurodevelopmental disorders (NDDs) such as autism spectrum disorder (ASD), intellectual disability, and epilepsy. High-confidence genetic studies converge on regulators of histone H3 lysine 4 (H3K4) methylation, such as the writers ASHIL and Klv1T2C and the eraser KDNISB, as recurrently mutated loci in NTIDs. The overarching goal of this study is to investigate how dysregulated H3K4 methylation compromises genome integrity in human neurons, thereby contributing to the pathogenesis of NDDs. The central, hypothesis is that coordinated II3K4 methylation safeguards neuronal genomes by maintaining an open chromatin architecture that permits the efficient detection and repair of transcription-coupled DNA lesions. The rationale/Or this study is to define the epigenetic control of DNA repair, which will illuminate a shared pathogenic hub across multiple ~I)D-linked genes. During the mentoredK99 phase, I will define how ASHIL, KMT2C, and KDM5B regulate chromatin structure and DNA repair at baseline and during transcriptional stress. Aim-1: I will use isogenic iPSC-derived cortical neurons with patient-relevant mutations or CRrSPRi knockdowns of these regulators, applying an integrated multi-omic pipeline: CUT&Tag and Micro-C to map H3K4 methylation and 3D chromatin topology. Aim-2: I will use Paired-Damage-seq, and CUT&RUN to chart oxidative lesions, repair synthesis, and recruitment of key repair factors; and RNA-seq to relate damage hotspots to altered gene expression. Aims l and 2 will be performed under the guidance of Dr. Lizarraga and Dr. Morrow, experts in the field of neurodevelopmental biology. My advisory team brings unique and complementary skills, enhancing my knowledge in 3D chromatin structure, transcription-coupled repair, gene editing, and multi-omics analysis. I will utilize these skills in the R00 phase (Aim 3), expanding the framework to include additional H3K4 regulators (e.g., LSD1, KMT2A) and broader neural lineages, thereby developing a comprehensive model. This study is innovative in its integration of single-cell D.NA damage mapping with chromatin topology and transcriptional profiling, enabling a direct and mechanistic connection between disrupted H3K4 methylation and genome instability. By uncovering how H3.K4 methylation prevents transcription-coupled genome instability in the developing brain, this research will address a critical gap in our understanding of NDD mechanisms. This award will enable me to launch an independent research program dedicated to determining mechanisms of chromatin-based processes that maintain genome stability in the developing human brain.

Stability in disrupted maternal representations over the perinatal period: Contributors and consequences

Abstract High-quality mother-infant relationships promote social, emotional, and cognitive development while protecting against poor child behavioral, health, and psychological adaptation that create risk for long- term negative outcomes. As mothers transition to parenthood, their own experiences of being cared for influence their emerging views of parenting and representations of their developing child. Evidence suggests that ‘disrupted’ maternal representations of the child, i.e., representations characterized by mixed communication, role merging, extreme withdrawal, and other unusual psychological processes, are tied to both poor child socioemotional adjustment and both insecure and disorganized attachment. However, it is unclear whether disrupted representations that emerge during pregnancy remain stable across the first several years of the child’s life. In addition, to date, research has not examined how change/stability in these representations may affect maternal caregiving and subsequent child adaptation. Using data from a longitudinal, multi-method study, this proposed project will examine the stability of maternal representations of the child for 99 women living in high risk contexts using the Working Model of the Child Interview during the third trimester and again when the child is two years of age. Mothers’ demographic characteristics (i.e. SES and relationship status), interpersonal violence experiences (i.e. child maltreatment or intimate violence exposure), psychological health (i.e. depressive, anxious, and PTSD symptoms), and parenting stress (i.e. perceptions of the child as difficult and parent-child interactions as dysfunctional) are measured as well to examine influences on representation stability. Finally, the observed quality of maternal caregiving and child adaptation are measured and examined in relation to stability in maternal representations of the child. Findings from this study have the potential to identify which mother-child dyads are at greatest risk for poor adaptation across the perinatal period and to delineate the contributors and consequences of maternal representational stability. These findings will serve as an important step towards informing the development or modification of existing prevention/intervention approaches that are targeted specifically towards mother-child dyads who are most at need.

Targeted Prodrug Cytokines for Metastatic Breast Cancer Immunotherapy

Project Summary. Our approach directly addresses key limitations in targeting and treating metastatic breast cancer, where we propose the selective activation of modular immune-modulating cytokines within the hypoxic and ROS-active TME for delivery across the BBB, providing the necessary pre-clinical data for future clinical translation. The in vitro and in vivo investigations of this novel immunotherapeutic in immunocompetent models will allow our team to study the interplay between tumor-driven immune activation, cytokine signaling, and anti-tumor immunity in both primary and metastatic sites, and establish a robust groundwork for subsequent clinical validation within the OSUCCC. This proposal addresses two key challenges in developing a novel immunotherapy strategy for breast cancer by answering two hypotheses: (1) can a modular immunotherapy platform with tumor-selective activation of prodrug recombinant cytokines overcome these limitations in drug delivery, and (2) can the development of nanobody-cytokine fusions that can selectively target primary breast cancer tumors and cross the BBB to reach metastatic tumor sites? The first hypothesis focuses on achieving tumor environment-specific activation of prodrug-based recombinant cytokines. Protein cytokines are highly potent, and while others have tried to block their activity using a fused genetic linker to ‘mask’ functionality, no one has yet attempted to use a non-canonical-based chemical strategy to achieve this inhibition. Immune-modulating cytokines will be recombinantly expressed with integrated ncAAs that block cytokine activity until the function is regenerated in the breast cancer TME. Once the cytokine activity is controlled, our second hypothesis will be to achieve selective delivery of the cytokine via fusion to nanobodies. While success has been found in targeting primary tumors in drug and protein delivery, a key challenge remains in reaching secondary metastatic tumors in hard-to-reach sites (i.e., brain). Engineered nanobodies, with affinity for breast cancer tumors and the ability to bind to BBB transcytosis receptors, will enable selective delivery to metastatic breast-to-brain tumors, resulting in tumor- specific activation, immune responses, and improved therapeutic outcomes. This system can significantly improve therapeutic outcomes for patients with mBC by integrating selective activation and delivery mechanisms to reduce off-target effects and enhance tumor-specific immune responses in both primary and secondary metastatic tumor sites. Optimizing drug delivery systems to tune immune responses could offer more effective and less invasive treatment options when compared to traditional and engineered cell-based approaches. Our momentum towards precision medicine and targeted therapies holds significant promise for improving outcomes for mBC patients, and has the potential to serve as a pan-cancer treatment for aggressive metastatic cancers from the following aims: (1) generating a modular platform for tumor-specific activation of prodrug cytokines, (2) evaluating cytokine delivery and anti-cancer immune phenotypes in mBC.

Primary cilia protein IFT88 governs smooth muscle phenotype and vascular remodeling

Project Summary/Abstract Cardiovascular disease remains the leading cause of death in the United States, accounting for nearly 1 million deaths in 2022. Vascular diseases such as atherosclerosis, aneurysm, and coronary artery disease are regulated largely by smooth muscle cells (SMCs) residing in the blood vessel wall. The central dogma of vascular SMC biology is that differentiated cells can de-differentiate and give rise to a spectrum of alternative phenotypes promoting invasion, proliferation, fibrosis, and inflammation, but the mechanisms regulating SMC phenotypic transitions are poorly understood. Intraflagellar transport 88 (IFT88) is an essential protein for the formation of primary cilia, centriole-associated plasma membrane organelles that project into the extracellular milieu and regulate cell cycle reentry and responses to stimuli like growth factors and mechanical strain. Non- ciliary functions of IFT88 also include progression of the cell cycle checkpoint and polarized motility, both of which are functionally critical for SMC-mediated vascular remodeling. Little is known about the functional role of the primary cilia in SMCs and the role of the essential cilia protein IFT88 in regulating SMC phenotype. To address this gap in knowledge, my postdoctoral studies focus on the role of IFT88 in the context of intimal hyperplasia (K99). During the independent phase (R00), I will apply these findings to arteriovenous fistula (AVF) maturation, a surgical intervention often required for dialysis individuals with polycystic kidney disease (PKD), an IFT88 loss-of-function disease. I will test my central hypothesis that cilia are key regulators of SMC phenotype in three Specific Aims: 1) determine the role of IFT88-dependent SMC primary cilia in mechanotransduction of extracellular matrix (ECM) stiffness (K99), 2) determine the role of IFT88 in pathological intimal hyperplasia (K99), and 3) test whether SMC IFT88 expression is required for adaptive remodeling of grafted veins following AVF placement (R00). Overall, we propose that IFT88+ ciliated SMC represent an unidentified subclass of the SMC phenotype spectrum that is primarily responsible for vascular remodeling and is an attractive potential target for treatment of vascular diseases. Building on strong existing collaborations, we have formed a research and mentoring team with expertise in SMC pathophysiology, primary cilia biology, mechanobiology, AVF surgery, and PKD to complete the proposed aims. The additional training in cell-ECM interactions (Aim 1, K99), in vivo murine ligation injury and in vivo cilia imaging (Aim 2, K99), and AVF surgery and PKD pathology (Aim 3, R00) will be indispensable for preparing the PI, Dr. O’Brien, for his career as an independent investigator. Completion of the proposed aims will also contribute directly to an understanding of the function of IFT88-dependent primary cilia in SMCs and may likely identify novel therapeutic targets for treatment of vascular diseases.

Decoding stress vulnerability

Although stress can be considered as an ongoing process that helps an organism to cope with present and future challenges, when it is too intense or uncontrollable, it can lead to adverse consequences for physical and mental health. Social stress specifically, is a highly prevalent traumatic experience, present in multiple contexts, such as war, bullying and interpersonal violence, and it has been linked with increased risk for major depression and anxiety disorders. Nevertheless, not all individuals exposed to strong stressful events develop psychopathology, with the mechanisms of resilience and vulnerability being still under investigation. During this talk, I will identify key gaps in our knowledge about stress vulnerability and I will present our recent data from our contextual fear learning protocol based on social defeat stress in mice.

Spike train structure of cortical transcriptomic populations in vivo

The cortex comprises many neuronal types, which can be distinguished by their transcriptomes: the sets of genes they express. Little is known about the in vivo activity of these cell types, particularly as regards the structure of their spike trains, which might provide clues to cortical circuit function. To address this question, we used Neuropixels electrodes to record layer 5 excitatory populations in mouse V1, then transcriptomically identified the recorded cell types. To do so, we performed a subsequent recording of the same cells using 2-photon (2p) calcium imaging, identifying neurons between the two recording modalities by fingerprinting their responses to a “zebra noise” stimulus and estimating the path of the electrode through the 2p stack with a probabilistic method. We then cut brain slices and performed in situ transcriptomics to localize ~300 genes using coppaFISH3d, a new open source method, and aligned the transcriptomic data to the 2p stack. Analysis of the data is ongoing, and suggests substantial differences in spike time coordination between ET and IT neurons, as well as between transcriptomic subtypes of both these excitatory types.

Understanding reward-guided learning using large-scale datasets

Understanding the neural mechanisms of reward-guided learning is a long-standing goal of computational neuroscience. Recent methodological innovations enable us to collect ever larger neural and behavioral datasets. This presents opportunities to achieve greater understanding of learning in the brain at scale, as well as methodological challenges. In the first part of the talk, I will discuss our recent insights into the mechanisms by which zebra finch songbirds learn to sing. Dopamine has been long thought to guide reward-based trial-and-error learning by encoding reward prediction errors. However, it is unknown whether the learning of natural behaviours, such as developmental vocal learning, occurs through dopamine-based reinforcement. Longitudinal recordings of dopamine and bird songs reveal that dopamine activity is indeed consistent with encoding a reward prediction error during naturalistic learning. In the second part of the talk, I will talk about recent work we are doing at DeepMind to develop tools for automatically discovering interpretable models of behavior directly from animal choice data. Our method, dubbed CogFunSearch, uses LLMs within an evolutionary search process in order to "discover" novel models in the form of Python programs that excel at accurately predicting animal behavior during reward-guided learning. The discovered programs reveal novel patterns of learning and choice behavior that update our understanding of how the brain solves reinforcement learning problems.

Understanding reward-guided learning using large-scale datasets

Understanding the neural mechanisms of reward-guided learning is a long-standing goal of computational neuroscience. Recent methodological innovations enable us to collect ever larger neural and behavioral datasets. This presents opportunities to achieve greater understanding of learning in the brain at scale, as well as methodological challenges. In the first part of the talk, I will discuss our recent insights into the mechanisms by which zebra finch songbirds learn to sing. Dopamine has been long thought to guide reward-based trial-and-error learning by encoding reward prediction errors. However, it is unknown whether the learning of natural behaviours, such as developmental vocal learning, occurs through dopamine-based reinforcement. Longitudinal recordings of dopamine and bird songs reveal that dopamine activity is indeed consistent with encoding a reward prediction error during naturalistic learning. In the second part of the talk, I will talk about recent work we are doing at DeepMind to develop tools for automatically discovering interpretable models of behavior directly from animal choice data. Our method, dubbed CogFunSearch, uses LLMs within an evolutionary search process in order to "discover" novel models in the form of Python programs that excel at accurately predicting animal behavior during reward-guided learning. The discovered programs reveal novel patterns of learning and choice behavior that update our understanding of how the brain solves reinforcement learning problems.

Relating circuit dynamics to computation: robustness and dimension-specific computation in cortical dynamics

Neural dynamics represent the hard-to-interpret substrate of circuit computations. Advances in large-scale recordings have highlighted the sheer spatiotemporal complexity of circuit dynamics within and across circuits, portraying in detail the difficulty of interpreting such dynamics and relating it to computation. Indeed, even in extremely simplified experimental conditions, one observes high-dimensional temporal dynamics in the relevant circuits. This complexity can be potentially addressed by the notion that not all changes in population activity have equal meaning, i.e., a small change in the evolution of activity along a particular dimension may have a bigger effect on a given computation than a large change in another. We term such conditions dimension-specific computation. Considering motor preparatory activity in a delayed response task we utilized neural recordings performed simultaneously with optogenetic perturbations to probe circuit dynamics. First, we revealed a remarkable robustness in the detailed evolution of certain dimensions of the population activity, beyond what was thought to be the case experimentally and theoretically. Second, the robust dimension in activity space carries nearly all of the decodable behavioral information whereas other non-robust dimensions contained nearly no decodable information, as if the circuit was setup to make informative dimensions stiff, i.e., resistive to perturbations, leaving uninformative dimensions sloppy, i.e., sensitive to perturbations. Third, we show that this robustness can be achieved by a modular organization of circuitry, whereby modules whose dynamics normally evolve independently can correct each other’s dynamics when an individual module is perturbed, a common design feature in robust systems engineering. Finally, we will recent work extending this framework to understanding the neural dynamics underlying preparation of speech.

Neurosurgery & Consciousness: Bridging Science and Philosophy in the Age of AI

Overview of neurosurgery specialty interplay between neurology, psychiatry and neurosurgery. Discussion on benefits and disadvantages of classifications. Presentation of sub-specialties: trauma, oncology, functional, pediatric, vascular and spine. How does an ordinary day of a neurosurgeon look like; outpatient clinic, emergencies, pre/intra/post operative patient care. An ordinary operation. Myth-busting and practical insights of every day practice. An ordinary operation. Hint for research on clinical problems to be solved. The coming ethical frontiers of neuroprosthetics. In part two we will explore the explanatory gap and its significance. We will review the more than 200 theories of the hard problem of consciousness, from the prevailing to the unconventional. Finally, we are going to reflect on the AI advancements and the claims of LLMs becoming conscious

Enhancing Real-World Event Memory

Memory is essential for shaping how we interpret the world, plan for the future, and understand ourselves, yet effective cognitive interventions for real-world episodic memory loss remain scarce. This talk introduces HippoCamera, a smartphone-based intervention inspired by how the brain supports memory, designed to enhance real-world episodic recollection by replaying high-fidelity autobiographical cues. It will showcase how our approach improves memory, mood, and hippocampal activity while uncovering links between memory distinctiveness, well-being, and the perception of time.

Genetic and epigenetic underpinnings of neurodegenerative disorders

Pluripotent cells, including embryonic stem (ES) and induced pluripotent stem (iPS) cells, are used to investigate the genetic and epigenetic underpinnings of human diseases such as Parkinson’s, Alzheimer’s, autism, and cancer. Mechanisms of somatic cell reprogramming to an embryonic pluripotent state are explored, utilizing patient-specific pluripotent cells to model and analyze neurodegenerative diseases.

LLMs and Human Language Processing

This webinar convened researchers at the intersection of Artificial Intelligence and Neuroscience to investigate how large language models (LLMs) can serve as valuable “model organisms” for understanding human language processing. Presenters showcased evidence that brain recordings (fMRI, MEG, ECoG) acquired while participants read or listened to unconstrained speech can be predicted by representations extracted from state-of-the-art text- and speech-based LLMs. In particular, text-based LLMs tend to align better with higher-level language regions, capturing more semantic aspects, while speech-based LLMs excel at explaining early auditory cortical responses. However, purely low-level features can drive part of these alignments, complicating interpretations. New methods, including perturbation analyses, highlight which linguistic variables matter for each cortical area and time scale. Further, “brain tuning” of LLMs—fine-tuning on measured neural signals—can improve semantic representations and downstream language tasks. Despite open questions about interpretability and exact neural mechanisms, these results demonstrate that LLMs provide a promising framework for probing the computations underlying human language comprehension and production at multiple spatiotemporal scales.

Unmotivated bias

In this talk, I will explore how social affective biases arise even in the absence of motivational factors as an emergent outcome of the basic structure of social learning. In several studies, we found that initial negative interactions with some members of a group can cause subsequent avoidance of the entire group, and that this avoidance perpetuates stereotypes. Additional cognitive modeling discovered that approach and avoidance behavior based on biased beliefs not only influences the evaluative (positive or negative) impressions of group members, but also shapes the depth of the cognitive representations available to learn about individuals. In other words, people have richer cognitive representations of members of groups that are not avoided, akin to individualized vs group level categories. I will end presenting a series of multi-agent reinforcement learning simulations that demonstrate the emergence of these social-structural feedback loops in the development and maintenance of affective biases.

Decomposing motivation into value and salience

Humans and other animals approach reward and avoid punishment and pay attention to cues predicting these events. Such motivated behavior thus appears to be guided by value, which directs behavior towards or away from positively or negatively valenced outcomes. Moreover, it is facilitated by (top-down) salience, which enhances attention to behaviorally relevant learned cues predicting the occurrence of valenced outcomes. Using human neuroimaging, we recently separated value (ventral striatum, posterior ventromedial prefrontal cortex) from salience (anterior ventromedial cortex, occipital cortex) in the domain of liquid reward and punishment. Moreover, we investigated potential drivers of learned salience: the probability and uncertainty with which valenced and non-valenced outcomes occur. We find that the brain dissociates valenced from non-valenced probability and uncertainty, which indicates that reinforcement matters for the brain, in addition to information provided by probability and uncertainty alone, regardless of valence. Finally, we assessed learning signals (unsigned prediction errors) that may underpin the acquisition of salience. Particularly the insula appears to be central for this function, encoding a subjective salience prediction error, similarly at the time of positively and negatively valenced outcomes. However, it appears to employ domain-specific time constants, leading to stronger salience signals in the aversive than the appetitive domain at the time of cues. These findings explain why previous research associated the insula with both valence-independent salience processing and with preferential encoding of the aversive domain. More generally, the distinction of value and salience appears to provide a useful framework for capturing the neural basis of motivated behavior.

Probing neural population dynamics with recurrent neural networks

Large-scale recordings of neural activity are providing new opportunities to study network-level dynamics with unprecedented detail. However, the sheer volume of data and its dynamical complexity are major barriers to uncovering and interpreting these dynamics. I will present latent factor analysis via dynamical systems, a sequential autoencoding approach that enables inference of dynamics from neuronal population spiking activity on single trials and millisecond timescales. I will also discuss recent adaptations of the method to uncover dynamics from neural activity recorded via 2P Calcium imaging. Finally, time permitting, I will mention recent efforts to improve the interpretability of deep-learning based dynamical systems models.

Characterizing the causal role of large-scale network interactions in supporting complex cognition

Neuroimaging has greatly extended our capacity to study the workings of the human brain. Despite the wealth of knowledge this tool has generated however, there are still critical gaps in our understanding. While tremendous progress has been made in mapping areas of the brain that are specialized for particular stimuli, or cognitive processes, we still know very little about how large-scale interactions between different cortical networks facilitate the integration of information and the execution of complex tasks. Yet even the simplest behavioral tasks are complex, requiring integration over multiple cognitive domains. Our knowledge falls short not only in understanding how this integration takes place, but also in what drives the profound variation in behavior that can be observed on almost every task, even within the typically developing (TD) population. The search for the neural underpinnings of individual differences is important not only philosophically, but also in the service of precision medicine. We approach these questions using a three-pronged approach. First, we create a battery of behavioral tasks from which we can calculate objective measures for different aspects of the behaviors of interest, with sufficient variance across the TD population. Second, using these individual differences in behavior, we identify the neural variance which explains the behavioral variance at the network level. Finally, using covert neurofeedback, we perturb the networks hypothesized to correspond to each of these components, thus directly testing their casual contribution. I will discuss our overall approach, as well as a few of the new directions we are currently pursuing.

Executive functions in the brain of deaf individuals – sensory and language effects

Executive functions are cognitive processes that allow us to plan, monitor and execute our goals. Using fMRI, we investigated how early deafness influences crossmodal plasticity and the organisation of executive functions in the adult human brain. Results from a range of visual executive function tasks (working memory, task switching, planning, inhibition) show that deaf individuals specifically recruit superior temporal “auditory” regions during task switching. Neural activity in auditory regions predicts behavioural performance during task switching in deaf individuals, highlighting the functional relevance of the observed cortical reorganisation. Furthermore, language grammatical skills were correlated with the level of activation and functional connectivity of fronto-parietal networks. Together, these findings show the interplay between sensory and language experience in the organisation of executive processing in the brain.

The quest for brain identification

In the 17th century, physician Marcello Malpighi observed the existence of distinctive patterns of ridges and sweat glands on fingertips. This was a major breakthrough, and originated a long and continuing quest for ways to uniquely identify individuals based on fingerprints, a technique massively used until today. It is only in the past few years that technologies and methodologies have achieved high-quality measures of an individual’s brain to the extent that personality traits and behavior can be characterized. The concept of “fingerprints of the brain” is very novel and has been boosted thanks to a seminal publication by Finn et al. in 2015. They were among the firsts to show that an individual’s functional brain connectivity profile is both unique and reliable, similarly to a fingerprint, and that it is possible to identify an individual among a large group of subjects solely on the basis of her or his connectivity profile. Yet, the discovery of brain fingerprints opened up a plethora of new questions. In particular, what exactly is the information encoded in brain connectivity patterns that ultimately leads to correctly differentiating someone’s connectome from anybody else’s? In other words, what makes our brains unique? In this talk I am going to partially address these open questions while keeping a personal viewpoint on the subject. I will outline the main findings, discuss potential issues, and propose future directions in the quest for identifiability of human brain networks.

Dyslexia, Rhythm, Language and the Developing Brain

Recent insights from auditory neuroscience provide a new perspective on how the brain encodes speech. Using these recent insights, I will provide an overview of key factors underpinning individual differences in children’s development of language and phonology, providing a context for exploring atypical reading development (dyslexia). Children with dyslexia are relatively insensitive to acoustic cues related to speech rhythm patterns. This lack of rhythmic sensitivity is related to the atypical neural encoding of rhythm patterns in speech by the brain. I will describe our recent data from infants as well as children, demonstrating developmental continuity in the key neural variables.

Using Adversarial Collaboration to Harness Collective Intelligence

There are many mysteries in the universe. One of the most significant, often considered the final frontier in science, is understanding how our subjective experience, or consciousness, emerges from the collective action of neurons in biological systems. While substantial progress has been made over the past decades, a unified and widely accepted explanation of the neural mechanisms underpinning consciousness remains elusive. The field is rife with theories that frequently provide contradictory explanations of the phenomenon. To accelerate progress, we have adopted a new model of science: adversarial collaboration in team science. Our goal is to test theories of consciousness in an adversarial setting. Adversarial collaboration offers a unique way to bolster creativity and rigor in scientific research by merging the expertise of teams with diverse viewpoints. Ideally, we aim to harness collective intelligence, embracing various perspectives, to expedite the uncovering of scientific truths. In this talk, I will highlight the effectiveness (and challenges) of this approach using selected case studies, showcasing its potential to counter biases, challenge traditional viewpoints, and foster innovative thought. Through the joint design of experiments, teams incorporate a competitive aspect, ensuring comprehensive exploration of problems. This method underscores the importance of structured conflict and diversity in propelling scientific advancement and innovation.

Event-related frequency adjustment (ERFA): A methodology for investigating neural entrainment

Neural entrainment has become a phenomenon of exceptional interest to neuroscience, given its involvement in rhythm perception, production, and overt synchronized behavior. Yet, traditional methods fail to quantify neural entrainment due to a misalignment with its fundamental definition (e.g., see Novembre and Iannetti, 2018; Rajandran and Schupp, 2019). The definition of entrainment assumes that endogenous oscillatory brain activity undergoes dynamic frequency adjustments to synchronize with environmental rhythms (Lakatos et al., 2019). Following this definition, we recently developed a method sensitive to this process. Our aim was to isolate from the electroencephalographic (EEG) signal an oscillatory component that is attuned to the frequency of a rhythmic stimulation, hypothesizing that the oscillation would adaptively speed up and slow down to achieve stable synchronization over time. To induce and measure these adaptive changes in a controlled fashion, we developed the event-related frequency adjustment (ERFA) paradigm (Rosso et al., 2023). A total of twenty healthy participants took part in our study. They were instructed to tap their finger synchronously with an isochronous auditory metronome, which was unpredictably perturbed by phase-shifts and tempo-changes in both positive and negative directions across different experimental conditions. EEG was recorded during the task, and ERFA responses were quantified as changes in instantaneous frequency of the entrained component. Our results indicate that ERFAs track the stimulus dynamics in accordance with the perturbation type and direction, preferentially for a sensorimotor component. The clear and consistent patterns confirm that our method is sensitive to the process of frequency adjustment that defines neural entrainment. In this Virtual Journal Club, the discussion of our findings will be complemented by methodological insights beneficial to researchers in the fields of rhythm perception and production, as well as timing in general. We discuss the dos and don’ts of using instantaneous frequency to quantify oscillatory dynamics, the advantages of adopting a multivariate approach to source separation, the robustness against the confounder of responses evoked by periodic stimulation, and provide an overview of domains and concrete examples where the methodological framework can be applied.

Identifying mechanisms of cognitive computations from spikes

Higher cortical areas carry a wide range of sensory, cognitive, and motor signals supporting complex goal-directed behavior. These signals mix in heterogeneous responses of single neurons, making it difficult to untangle underlying mechanisms. I will present two approaches for revealing interpretable circuit mechanisms from heterogeneous neural responses during cognitive tasks. First, I will show a flexible nonparametric framework for simultaneously inferring population dynamics on single trials and tuning functions of individual neurons to the latent population state. When applied to recordings from the premotor cortex during decision-making, our approach revealed that populations of neurons encoded the same dynamic variable predicting choices, and heterogeneous firing rates resulted from the diverse tuning of single neurons to this decision variable. The inferred dynamics indicated an attractor mechanism for decision computation. Second, I will show an approach for inferring an interpretable network model of a cognitive task—the latent circuit—from neural response data. We developed a theory to causally validate latent circuit mechanisms via patterned perturbations of activity and connectivity in the high-dimensional network. This work opens new possibilities for deriving testable mechanistic hypotheses from complex neural response data.

From controlled environments to complex realities: Exploring the interplay between perceived minds and attention

In our daily lives, we perceive things as possessing a mind (e.g., people) or lacking one (e.g., shoes). Intriguingly, how much mind we attribute to people can vary, with real people perceived to have more mind than depictions of individuals, such as photographs. Drawing from a range of research methodologies, including naturalistic observation, mobile eye tracking, and surreptitious behavior monitoring, I discuss how various shades of mind influence human attention and behaviour. The findings suggest the novel concept that overt attention (where one looks) in real-life is fundamentally supported by covert attention (attending to someone out of the corner of one's eye).

Social and non-social learning: Common, or specialised, mechanisms? (BACN Early Career Prize Lecture 2022)

The last decade has seen a burgeoning interest in studying the neural and computational mechanisms that underpin social learning (learning from others). Many findings support the view that learning from other people is underpinned by the same, ‘domain-general’, mechanisms underpinning learning from non-social stimuli. Despite this, the idea that humans possess social-specific learning mechanisms - adaptive specializations moulded by natural selection to cope with the pressures of group living - persists. In this talk I explore the persistence of this idea. First, I present dissociations between social and non-social learning - patterns of data which are difficult to explain under the domain-general thesis and which therefore support the idea that we have evolved special mechanisms for social learning. Subsequently, I argue that most studies that have dissociated social and non-social learning have employed paradigms in which social information comprises a secondary, additional, source of information that can be used to supplement learning from non-social stimuli. Thus, in most extant paradigms, social and non-social learning differ both in terms of social nature (social or non-social) and status (primary or secondary). I conclude that status is an important driver of apparent differences between social and non-social learning. When we account for differences in status, we see that social and non-social learning share common (dopamine-mediated) mechanisms.

Sleep deprivation and the human brain: from brain physiology to cognition”

Sleep strongly affects synaptic strength, making it critical for cognition, especially learning and memory formation. Whether and how sleep deprivation modulates human brain physiology and cognition is poorly understood. Here we examined how overnight sleep deprivation vs overnight sufficient sleep affects (a) cortical excitability, measured by transcranial magnetic stimulation, (b) inducibility of long-term potentiation (LTP)- and long-term depression (LTD)-like plasticity via transcranial direct current stimulation (tDCS), and (c) learning, memory, and attention. We found that sleep deprivation increases cortical excitability due to enhanced glutamate-related cortical facilitation and decreases and/or reverses GABAergic cortical inhibition. Furthermore, tDCS-induced LTP-like plasticity (anodal) abolishes while the inhibitory LTD-like plasticity (cathodal) converts to excitatory LTP-like plasticity under sleep deprivation. This is associated with increased EEG theta oscillations due to sleep pressure. Motor learning, behavioral counterparts of plasticity, and working memory and attention, which rely on cortical excitability, are also impaired during sleep deprivation. Our study indicates that upscaled brain excitability and altered plasticity, due to sleep deprivation, are associated with impaired cognitive performance. Besides showing how brain physiology and cognition undergo changes (from neurophysiology to higher-order cognition) under sleep pressure, the findings have implications for variability and optimal application of noninvasive brain stimulation.

Brain network communication: concepts, models and applications

Understanding communication and information processing in nervous systems is a central goal of neuroscience. Over the past two decades, advances in connectomics and network neuroscience have opened new avenues for investigating polysynaptic communication in complex brain networks. Recent work has brought into question the mainstay assumption that connectome signalling occurs exclusively via shortest paths, resulting in a sprawling constellation of alternative network communication models. This Review surveys the latest developments in models of brain network communication. We begin by drawing a conceptual link between the mathematics of graph theory and biological aspects of neural signalling such as transmission delays and metabolic cost. We organize key network communication models and measures into a taxonomy, aimed at helping researchers navigate the growing number of concepts and methods in the literature. The taxonomy highlights the pros, cons and interpretations of different conceptualizations of connectome signalling. We showcase the utility of network communication models as a flexible, interpretable and tractable framework to study brain function by reviewing prominent applications in basic, cognitive and clinical neurosciences. Finally, we provide recommendations to guide the future development, application and validation of network communication models.

Epilepsy genetics 2023: From research to advanced clinical genetic test interpretation

The presentation will provide an overview of the expanding role of genetic factors in epilepsy. It will delve into the fundamentals of this field and elucidate how digital tools and resources can aid in the re-evaluation of genetic test results. In the initial segment of the presentation, Dr. Lal will examine the advancements made over the past two decades regarding the genetic architecture of various epilepsy types. Additionally, he will present research studies in which he has actively participated, offering concrete examples. Subsequently, during the second part of the talk, Dr. Lal will share the ongoing research projects that focus on epilepsy genetics, bioinformatics, and health record data science.

Mechanisms Underlying the Persistence of Cancer-Related Fatigue

Cancer-related fatigue is a prominent and debilitating side effect of cancer and its treatment. It can develop prior to diagnosis, generally peaks during cancer treatment, and can persist long after treatment completion. Its mechanisms are multifactorial, and its expression is highly variable. Unfortunately, treatment options are limited. Our research uses syngeneic murine models of cancer and cisplatin-based chemotherapy to better understand these mechanisms. Our data indicate that both peripherally and centrally processes may contribute to the developmental of fatigue. These processes include metabolic alterations, mitochondrial dysfunction, pre-cachexia, and inflammation. However, our data has revealed that behavioral fatigue can persist even after the toxicity associated with cancer and its treatment recover. For example, running during cancer treatment attenuates kidney toxicity while also delaying recovery from fatigue-like behavior. Additionally, administration of anesthetics known to disrupt memory consolidation at the time treatment can promote recovery, and treatment-related cues can re-instate fatigue after recovery. Cancer-related fatigue can also promote habitual behavioral patterns, as observed using a devaluation task. We interpret this data to suggest that limit metabolic resources during cancer promote the utilization of habit-based behavioral strategies that serve to maintain fatigue behavior into survivorship. This line of work is exciting as it points us toward novel interventional targets for the treatment of persistent cancer-related fatigue.

Richly structured reward predictions in dopaminergic learning circuits

Theories from reinforcement learning have been highly influential for interpreting neural activity in the biological circuits critical for animal and human learning. Central among these is the identification of phasic activity in dopamine neurons as a reward prediction error signal that drives learning in basal ganglia and prefrontal circuits. However, recent findings suggest that dopaminergic prediction error signals have access to complex, structured reward predictions and are sensitive to more properties of outcomes than learning theories with simple scalar value predictions might suggest. Here, I will present recent work in which we probed the identity-specific structure of reward prediction errors in an odor-guided choice task and found evidence for multiple predictive “threads” that segregate reward predictions, and reward prediction errors, according to the specific sensory features of anticipated outcomes. Our results point to an expanded class of neural reinforcement learning algorithms in which biological agents learn rich associative structure from their environment and leverage it to build reward predictions that include information about the specific, and perhaps idiosyncratic, features of available outcomes, using these to guide behavior in even quite simple reward learning tasks.

The centrality of population-level factors to network computation is demonstrated by a versatile approach for training spiking networks

Neural activity is often described in terms of population-level factors extracted from the responses of many neurons. Factors provide a lower-dimensional description with the aim of shedding light on network computations. Yet, mechanistically, computations are performed not by continuously valued factors but by interactions among neurons that spike discretely and variably. Models provide a means of bridging these levels of description. We developed a general method for training model networks of spiking neurons by leveraging factors extracted from either data or firing-rate-based networks. In addition to providing a useful model-building framework, this formalism illustrates how reliable and continuously valued factors can arise from seemingly stochastic spiking. Our framework establishes procedures for embedding this property in network models with different levels of realism. The relationship between spikes and factors in such networks provides a foundation for interpreting (and subtly redefining) commonly used quantities such as firing rates.

Precise spatio-temporal spike patterns in cortex and model

The cell assembly hypothesis postulates that groups of coordinated neurons form the basis of information processing. Here, we test this hypothesis by analyzing massively parallel spiking activity recorded in monkey motor cortex during a reach-to-grasp experiment for the presence of significant ms-precise spatio-temporal spike patterns (STPs). For this purpose, the parallel spike trains were analyzed for STPs by the SPADE method (Stella et al, 2019, Biosystems), which detects, counts and evaluates spike patterns for their significance by the use of surrogates (Stella et al, 2022 eNeuro). As a result we find STPs in 19/20 data sets (each of 15min) from two monkeys, but only a small fraction of the recorded neurons are involved in STPs. To consider the different behavioral states during the task, we analyzed the data in a quasi time-resolved analysis by dividing the data into behaviorally relevant time epochs. The STPs that occur in the various epochs are specific to behavioral context - in terms of neurons involved and temporal lags between the spikes of the STP. Furthermore we find, that the STPs often share individual neurons across epochs. Since we interprete the occurrence of a particular STP as the signature of a particular active cell assembly, our interpretation is that the neurons multiplex their cell assembly membership. In a related study, we model these findings by networks with embedded synfire chains (Kleinjohann et al, 2022, bioRxiv 2022.08.02.502431).

Dynamic endocrine modulation of the nervous system

Sex hormones are powerful neuromodulators of learning and memory. In rodents and nonhuman primates estrogen and progesterone influence the central nervous system across a range of spatiotemporal scales. Yet, their influence on the structural and functional architecture of the human brain is largely unknown. Here, I highlight findings from a series of dense-sampling neuroimaging studies from my laboratory designed to probe the dynamic interplay between the nervous and endocrine systems. Individuals underwent brain imaging and venipuncture every 12-24 hours for 30 consecutive days. These procedures were carried out under freely cycling conditions and again under a pharmacological regimen that chronically suppresses sex hormone production. First, resting state fMRI evidence suggests that transient increases in estrogen drive robust increases in functional connectivity across the brain. Time-lagged methods from dynamical systems analysis further reveals that these transient changes in estrogen enhance within-network integration (i.e. global efficiency) in several large-scale brain networks, particularly Default Mode and Dorsal Attention Networks. Next, using high-resolution hippocampal subfield imaging, we found that intrinsic hormone fluctuations and exogenous hormone manipulations can rapidly and dynamically shape medial temporal lobe morphology. Together, these findings suggest that neuroendocrine factors influence the brain over short and protracted timescales.

Asymmetric signaling across the hierarchy of cytoarchitecture within the human connectome

Cortical variations in cytoarchitecture form a sensory-fugal axis that shapes regional profiles of extrinsic connectivity and is thought to guide signal propagation and integration across the cortical hierarchy. While neuroimaging work has shown that this axis constrains local properties of the human connectome, it remains unclear whether it also shapes the asymmetric signaling that arises from higher-order topology. Here, we used network control theory to examine the amount of energy required to propagate dynamics across the sensory-fugal axis. Our results revealed an asymmetry in this energy, indicating that bottom-up transitions were easier to complete compared to top-down. Supporting analyses demonstrated that asymmetries were underpinned by a connectome topology that is wired to support efficient bottom-up signaling. Lastly, we found that asymmetries correlated with differences in communicability and intrinsic neuronal time scales and lessened throughout youth. Our results show that cortical variation in cytoarchitecture may guide the formation of macroscopic connectome topology.

Are place cells just memory cells? Probably yes

Neurons in the rodent hippocampus appear to encode the position of the animal in physical space during movement. Individual ``place cells'' fire in restricted sub-regions of an environment, a feature often taken as evidence that the hippocampus encodes a map of space that subserves navigation. But these same neurons exhibit complex responses to many other variables that defy explanation by position alone, and the hippocampus is known to be more broadly critical for memory formation. Here we elaborate and test a theory of hippocampal coding which produces place cells as a general consequence of efficient memory coding. We constructed neural networks that actively exploit the correlations between memories in order to learn compressed representations of experience. Place cells readily emerged in the trained model, due to the correlations in sensory input between experiences at nearby locations. Notably, these properties were highly sensitive to the compressibility of the sensory environment, with place field size and population coding level in dynamic opposition to optimally encode the correlations between experiences. The effects of learning were also strongly biphasic: nearby locations are represented more similarly following training, while locations with intermediate similarity become increasingly decorrelated, both distance-dependent effects that scaled with the compressibility of the input features. Using virtual reality and 2-photon functional calcium imaging in head-fixed mice, we recorded the simultaneous activity of thousands of hippocampal neurons during virtual exploration to test these predictions. Varying the compressibility of sensory information in the environment produced systematic changes in place cell properties that reflected the changing input statistics, consistent with the theory. We similarly identified representational plasticity during learning, which produced a distance-dependent exchange between compression and pattern separation. These results motivate a more domain-general interpretation of hippocampal computation, one that is naturally compatible with earlier theories on the circuit's importance for episodic memory formation. Work done in collaboration with James Priestley, Lorenzo Posani, Marcus Benna, Attila Losonczy.

Autopoiesis and Enaction in the Game of Life

Enaction plays a central role in the broader fabric of so-called 4E (embodied, embedded, extended, enactive) cognition. Although the origin of the enactive approach is widely dated to the 1991 publication of the book "The Embodied Mind" by Varela, Thompson and Rosch, many of the central ideas trace to much earlier work. Over 40 years ago, the Chilean biologists Humberto Maturana and Francisco Varela put forward the notion of autopoiesis as a way to understand living systems and the phenomena that they generate, including cognition. Varela and others subsequently extended this framework to an enactive approach that places biological autonomy at the foundation of situated and embodied behavior and cognition. I will describe an attempt to place Maturana and Varela's original ideas on a firmer foundation by studying them within the context of a toy model universe, John Conway's Game of Life (GoL) cellular automata. This work has both pedagogical and theoretical goals. Simple concrete models provide an excellent vehicle for introducing some of the core concepts of autopoiesis and enaction and explaining how these concepts fit together into a broader whole. In addition, a careful analysis of such toy models can hone our intuitions about these concepts, probe their strengths and weaknesses, and move the entire enterprise in the direction of a more mathematically rigorous theory. In particular, I will identify the primitive processes that can occur in GoL, show how these can be linked together into mutually-supporting networks that underlie persistent bounded entities, map the responses of such entities to environmental perturbations, and investigate the paths of mutual perturbation that these entities and their environments can undergo.

How Children Design by Analogy: The Role of Spatial Thinking

Analogical reasoning is a common reasoning tool for learning and problem-solving. Existing research has extensively studied children’s reasoning when comparing, or choosing from ready-made analogies. Relatively less is known about how children come up with analogies in authentic learning environments. Design education provides a suitable context to investigate how children generate analogies for creative learning purposes. Meanwhile, the frequent use of visual analogies in design provides an additional opportunity to understand the role of spatial reasoning in design-by-analogy. Spatial reasoning is one of the most studied human cognitive factors and is critical to the learning of science, technology, engineering, arts, and mathematics (STEAM). There is growing interest in exploring the interplay between analogical reasoning and spatial reasoning. In this talk, I will share qualitative findings from a case study, where a class of 11-to-12-year-olds in the Netherlands participated in a biomimicry design project. These findings illustrate (1) practical ways to support children’s analogical reasoning in the ideation process and (2) the potential role of spatial reasoning as seen in children mapping form-function relationships in nature analogically and adaptively to those in human designs.

Integration of 3D human stem cell models derived from post-mortem tissue and statistical genomics to guide schizophrenia therapeutic development

Schizophrenia is a neuropsychiatric disorder characterized by positive symptoms (such as hallucinations and delusions), negative symptoms (such as avolition and withdrawal) and cognitive dysfunction1. Schizophrenia is highly heritable, and genetic studies are playing a pivotal role in identifying potential biomarkers and causal disease mechanisms with the hope of informing new treatments. Genome-wide association studies (GWAS) identified nearly 270 loci with a high statistical association with schizophrenia risk; however each locus confers only a small increase in risk therefore it is difficult to translate these findings into understanding disease biology that can lead to treatments. Induced pluripotent stem cell (iPSC) models are a tractable system to translate genetic findings and interrogate mechanisms of pathogenesis. Mounting research with patient-derived iPSCs has proposed several neurodevelopmental pathways altered in SCZ, such as neural progenitor cell (NPC) proliferation, imbalanced differentiation of excitatory and inhibitory cortical neurons. However, it is unclear what exactly these iPS models recapitulate, how potential perturbations of early brain development translates into illness in adults and how iPS models that represent fetal stages can be utilized to further drug development efforts to treat adult illness. I will present the largest transcriptome analysis of post-mortem caudate nucleus in schizophrenia where we discovered that decreased presynaptic DRD2 autoregulation is the causal dopamine risk factor for schizophrenia (Benjamin et al, Nature Neuroscience 2022 https://doi.org/10.1038/s41593-022-01182-7). We developed stem cell models from a subset of the postmortem cohort to better understand the molecular underpinnings of human psychiatric disorders (Sawada et al, Stem Cell Research 2020). We established a method for the differentiation of iPS cells into ventral forebrain organoids and performed single cell RNAseq and cellular phenotyping. To our knowledge, this is the first study to evaluate iPSC models of SZ from the same individuals with postmortem tissue. Our study establishes that striatal neurons in the patients with SCZ carry abnormalities that originated during early brain development. Differentiation of inhibitory neurons is accelerated whereas excitatory neuronal development is delayed, implicating an excitation and inhibition (E-I) imbalance during early brain development in SCZ. We found a significant overlap of genes upregulated in the inhibitory neurons in SCZ organoids with upregulated genes in postmortem caudate tissues from patients with SCZ compared with control individuals, including the donors of our iPS cell cohort. Altogether, we demonstrate that ventral forebrain organoids derived from postmortem tissue of individuals with schizophrenia recapitulate perturbed striatal gene expression dynamics of the donors’ brains (Sawada et al, biorxiv 2022 https://doi.org/10.1101/2022.05.26.493589).

Interplay between circuits that mediate spontaneous retinal waves and early light responses during retinal development

Cell-type specific alterations underpinning convergent ASD phenotypes in PACS1 neurodevelopmental disorder

Understanding Machine Learning via Exactly Solvable Statistical Physics Models

The affinity between statistical physics and machine learning has a long history. I will describe the main lines of this long-lasting friendship in the context of current theoretical challenges and open questions about deep learning. Theoretical physics often proceeds in terms of solvable synthetic models, I will describe the related line of work on solvable models of simple feed-forward neural networks. I will highlight a path forward to capture the subtle interplay between the structure of the data, the architecture of the network, and the optimization algorithms commonly used for learning.

Children-Agent Interaction For Assessment and Rehabilitation: From Linguistic Skills To Mental Well-being

Socially Assistive Robots (SARs) have shown great potential to help children in therapeutic and healthcare contexts. SARs have been used for companionship, learning enhancement, social and communication skills rehabilitation for children with special needs (e.g., autism), and mood improvement. Robots can be used as novel tools to assess and rehabilitate children’s communication skills and mental well-being by providing affordable and accessible therapeutic and mental health services. In this talk, I will present the various studies I have conducted during my PhD and at the Cambridge Affective Intelligence and Robotics Lab to explore how robots can help assess and rehabilitate children’s communication skills and mental well-being. More specifically, I will provide both quantitative and qualitative results and findings from (i) an exploratory study with children with autism and global developmental disorders to investigate the use of intelligent personal assistants in therapy; (ii) an empirical study involving children with and without language disorders interacting with a physical robot, a virtual agent, and a human counterpart to assess their linguistic skills; (iii) an 8-week longitudinal study involving children with autism and language disorders who interacted either with a physical or a virtual robot to rehabilitate their linguistic skills; and (iv) an empirical study to aid the assessment of mental well-being in children. These findings can inform and help the child-robot interaction community design and develop new adaptive robots to help assess and rehabilitate linguistic skills and mental well-being in children.

Visual Perception in Cerebral Visual Impairment (CVI)

The impact of emerging technologies and methods on the interpretation of genetic variation in autism and fetal genomics

Network inference via process motifs for lagged correlation in linear stochastic processes

A major challenge for causal inference from time-series data is the trade-off between computational feasibility and accuracy. Motivated by process motifs for lagged covariance in an autoregressive model with slow mean-reversion, we propose to infer networks of causal relations via pairwise edge measure (PEMs) that one can easily compute from lagged correlation matrices. Motivated by contributions of process motifs to covariance and lagged variance, we formulate two PEMs that correct for confounding factors and for reverse causation. To demonstrate the performance of our PEMs, we consider network interference from simulations of linear stochastic processes, and we show that our proposed PEMs can infer networks accurately and efficiently. Specifically, for slightly autocorrelated time-series data, our approach achieves accuracies higher than or similar to Granger causality, transfer entropy, and convergent crossmapping -- but with much shorter computation time than possible with any of these methods. Our fast and accurate PEMs are easy-to-implement methods for network inference with a clear theoretical underpinning. They provide promising alternatives to current paradigms for the inference of linear models from time-series data, including Granger causality, vector-autoregression, and sparse inverse covariance estimation.

Universal function approximation in balanced spiking networks through convex-concave boundary composition

The spike-threshold nonlinearity is a fundamental, yet enigmatic, component of biological computation — despite its role in many theories, it has evaded definitive characterisation. Indeed, much classic work has attempted to limit the focus on spiking by smoothing over the spike threshold or by approximating spiking dynamics with firing-rate dynamics. Here, we take a novel perspective that captures the full potential of spike-based computation. Based on previous studies of the geometry of efficient spike-coding networks, we consider a population of neurons with low-rank connectivity, allowing us to cast each neuron’s threshold as a boundary in a space of population modes, or latent variables. Each neuron divides this latent space into subthreshold and suprathreshold areas. We then demonstrate how a network of inhibitory (I) neurons forms a convex, attracting boundary in the latent coding space, and a network of excitatory (E) neurons forms a concave, repellant boundary. Finally, we show how the combination of the two yields stable dynamics at the crossing of the E and I boundaries, and can be mapped onto a constrained optimization problem. The resultant EI networks are balanced, inhibition-stabilized, and exhibit asynchronous irregular activity, thereby closely resembling cortical networks of the brain. Moreover, we demonstrate how such networks can be tuned to either suppress or amplify noise, and how the composition of inhibitory convex and excitatory concave boundaries can result in universal function approximation. Our work puts forth a new theory of biologically-plausible computation in balanced spiking networks, and could serve as a novel framework for scalable and interpretable computation with spikes.

Beyond Biologically Plausible Spiking Networks for Neuromorphic Computing

Biologically plausible spiking neural networks (SNNs) are an emerging architecture for deep learning tasks due to their energy efficiency when implemented on neuromorphic hardware. However, many of the biological features are at best irrelevant and at worst counterproductive when evaluated in the context of task performance and suitability for neuromorphic hardware. In this talk, I will present an alternative paradigm to design deep learning architectures with good task performance in real-world benchmarks while maintaining all the advantages of SNNs. We do this by focusing on two main features – event-based computation and activity sparsity. Starting from the performant gated recurrent unit (GRU) deep learning architecture, we modify it to make it event-based and activity-sparse. The resulting event-based GRU (EGRU) is extremely efficient for both training and inference. At the same time, it achieves performance close to conventional deep learning architectures in challenging tasks such as language modelling, gesture recognition and sequential MNIST.

Nonlinear computations in spiking neural networks through multiplicative synapses

The brain efficiently performs nonlinear computations through its intricate networks of spiking neurons, but how this is done remains elusive. While recurrent spiking networks implementing linear computations can be directly derived and easily understood (e.g., in the spike coding network (SCN) framework), the connectivity required for nonlinear computations can be harder to interpret, as they require additional non-linearities (e.g., dendritic or synaptic) weighted through supervised training. Here we extend the SCN framework to directly implement any polynomial dynamical system. This results in networks requiring multiplicative synapses, which we term the multiplicative spike coding network (mSCN). We demonstrate how the required connectivity for several nonlinear dynamical systems can be directly derived and implemented in mSCNs, without training. We also show how to precisely carry out higher-order polynomials with coupled networks that use only pair-wise multiplicative synapses, and provide expected numbers of connections for each synapse type. Overall, our work provides an alternative method for implementing nonlinear computations in spiking neural networks, while keeping all the attractive features of standard SCNs such as robustness, irregular and sparse firing, and interpretable connectivity. Finally, we discuss the biological plausibility of mSCNs, and how the high accuracy and robustness of the approach may be of interest for neuromorphic computing.

No Free Lunch from Deep Learning in Neuroscience: A Case Study through Models of the Entorhinal-Hippocampal Circuit

Research in Neuroscience, as in many scientific disciplines, is undergoing a renaissance based on deep learning. Unique to Neuroscience, deep learning models can be used not only as a tool but interpreted as models of the brain. The central claims of recent deep learning-based models of brain circuits are that they shed light on fundamental functions being optimized or make novel predictions about neural phenomena. We show, through the case-study of grid cells in the entorhinal-hippocampal circuit, that one may get neither. We rigorously examine the claims of deep learning models of grid cells using large-scale hyperparameter sweeps and theory-driven experimentation, and demonstrate that the results of such models are more strongly driven by particular, non-fundamental, and post-hoc implementation choices than fundamental truths about neural circuits or the loss function(s) they might optimize. We discuss why these models cannot be expected to produce accurate models of the brain without the addition of substantial amounts of inductive bias, an informal No Free Lunch result for Neuroscience.

From Machine Learning to Autonomous Intelligence

How could machines learn as efficiently as humans and animals? How could machines learn to reason and plan? How could machines learn representations of percepts and action plans at multiple levels of abstraction, enabling them to reason, predict, and plan at multiple time horizons? I will propose a possible path towards autonomous intelligent agents, based on a new modular cognitive architecture and a somewhat new self supervised training paradigm. The centerpiece of the proposed architecture is a configurable predictive world model that allows the agent to plan. Behavior and learning are driven by a set of differentiable intrinsic cost functions. The world model uses a new type of energy-based model architecture called H-JEPA (Hierarchical Joint Embedding Predictive Architecture). H-JEPA learns hierarchical abstract representations of the world that are simultaneously maximally informative and maximally predictable.

From Machine Learning to Autonomous Intelligence

How could machines learn as efficiently as humans and animals? How could machines learn to reason and plan? How could machines learn representations of percepts and action plans at multiple levels of abstraction, enabling them to reason, predict, and plan at multiple time horizons? I will propose a possible path towards autonomous intelligent agents, based on a new modular cognitive architecture and a somewhat new self-supervised training paradigm. The centerpiece of the proposed architecture is a configurable predictive world model that allows the agent to plan. Behavior and learning are driven by a set of differentiable intrinsic cost functions. The world model uses a new type of energy-based model architecture called H-JEPA (Hierarchical Joint Embedding Predictive Architecture). H-JEPA learns hierarchical abstract representations of the world that are simultaneously maximally informative and maximally predictable. The corresponding working paper is available here:https://openreview.net/forum?id=BZ5a1r-kVsf

General purpose event-based architectures for deep learning

Biologically plausible spiking neural networks (SNNs) are an emerging architecture for deep learning tasks due to their energy efficiency when implemented on neuromorphic hardware. However, many of the biological features are at best irrelevant and at worst counterproductive when evaluated in the context of task performance and suitability for neuromorphic hardware. In this talk, I will present an alternative paradigm to design deep learning architectures with good task performance in real-world benchmarks while maintaining all the advantages of SNNs. We do this by focusing on two main features -- event-based computation and activity sparsity. Starting from the performant gated recurrent unit (GRU) deep learning architecture, we modify it to make it event-based and activity-sparse. The resulting event-based GRU (EGRU) is extremely efficient for both training and inference. At the same time, it achieves performance close to conventional deep learning architectures in challenging tasks such as language modelling, gesture recognition and sequential MNIST

The Secret Bayesian Life of Ring Attractor Networks

Efficient navigation requires animals to track their position, velocity and heading direction (HD). Some animals’ behavior suggests that they also track uncertainties about these navigational variables, and make strategic use of these uncertainties, in line with a Bayesian computation. Ring-attractor networks have been proposed to estimate and track these navigational variables, for instance in the HD system of the fruit fly Drosophila. However, such networks are not designed to incorporate a notion of uncertainty, and therefore seem unsuited to implement dynamic Bayesian inference. Here, we close this gap by showing that specifically tuned ring-attractor networks can track both a HD estimate and its associated uncertainty, thereby approximating a circular Kalman filter. We identified the network motifs required to integrate angular velocity observations, e.g., through self-initiated turns, and absolute HD observations, e.g., visual landmark inputs, according to their respective reliabilities, and show that these network motifs are present in the connectome of the Drosophila HD system. Specifically, our network encodes uncertainty in the amplitude of a localized bump of neural activity, thereby generalizing standard ring attractor models. In contrast to such standard attractors, however, proper Bayesian inference requires the network dynamics to operate in a regime away from the attractor state. More generally, we show that near-Bayesian integration is inherent in generic ring attractor networks, and that their amplitude dynamics can account for close-to-optimal reliability weighting of external evidence for a wide range of network parameters. This only holds, however, if their connection strengths allow the network to sufficiently deviate from the attractor state. Overall, our work offers a novel interpretation of ring attractor networks as implementing dynamic Bayesian integrators. We further provide a principled theoretical foundation for the suggestion that the Drosophila HD system may implement Bayesian HD tracking via ring attractor dynamics.

At the nexus of genes, aging and environment: Understanding transcriptomic and epigenomic regulation in Parkinson's disease

Parkinson’s Disease (PD), the most common neurodegenerative movement disorder, is based on a complex interplay between genetic predispositions, aging processes, and environmental influences. In order to better understand the gene-environment axis in PD, we pursue a multi-omics approach to comprehensively interrogate genome-wide changes in histone modifications, DNA methylation, and hydroxymethylation, accompanied by transcriptomic profiling in cell and animal models of PD as well as large patient cohorts. Furthermore, we assess the plasticity of epigenomic modifications under influence of environmental factors using longitudinal cohorts of sporadic PD cases as well as mouse models exposed to specific environmental factors. Here, we present gene expression changes in PD mouse models in context of aging as well as environmental enrichment and high-fat diet.

FY15 ERP Funding Opportunities Now Available!

Consumer Led Flexibility for the Clean Energy Superpower Mission

Neuronal bursting from an interplay of fast voltage and slow concentration dynamics mediated by the Na+/K+-ATPase

Bernstein Conference 2024

Emergence of time persistence in an interpretable data-driven neural network model

COSYNE 2022

The interplay between prediction and integration processes in human perception

COSYNE 2022

The interplay between prediction and integration processes in human perception

COSYNE 2022

Interpretable behavioral features have conserved neural representations across mice

COSYNE 2022

Interpretable behavioral features have conserved neural representations across mice

COSYNE 2022

An interpretable dynamic population-rate equation for adapting non-linear spiking neural populations

COSYNE 2022

An interpretable dynamic population-rate equation for adapting non-linear spiking neural populations

COSYNE 2022

An interpretable spline-LNP model to characterize feedforward and feedback processing in mouse dLGN

COSYNE 2022

An interpretable spline-LNP model to characterize feedforward and feedback processing in mouse dLGN

COSYNE 2022

Modeling Hippocampal Spatial Learning Through a Valence-based Interplay of Dopamine and Serotonin

COSYNE 2022

Modeling Hippocampal Spatial Learning Through a Valence-based Interplay of Dopamine and Serotonin

COSYNE 2022

Supervised learning and interpretation of plasticity rules in spiking neural networks

COSYNE 2022

Supervised learning and interpretation of plasticity rules in spiking neural networks

COSYNE 2022

Using 1D-convolutional neural networks to detect and interpret sharp-wave ripples

COSYNE 2022

Using 1D-convolutional neural networks to detect and interpret sharp-wave ripples

COSYNE 2022

“Attentional fingerprints” in conceptual space: Reliable, individuating patterns of visual attention revealed using natural language modeling

COSYNE 2023

Semi-supervised quantification and interpretation of undirected human behavior

COSYNE 2023

Sparse Component Analysis: An interpretable dimensionality reduction tool that identifies building blocks of neural computation

COSYNE 2023

New tools for recording and interpreting brain-wide activity in C. elegans

COSYNE 2023

Automated discovery of interpretable cognitive programs underlying reward-guided behavior

COSYNE 2025

Connectome Interpreter: a toolkit for efficient connectome exploration and hypothesis generation

COSYNE 2025

Data-driven evaluation of interpretive framework for model-based planning

COSYNE 2025

Ethological foraging fingerprints reveal heterogeneous effects of serotonergic neuromodulation

COSYNE 2025

A flexible and interpretable statistical model of distributed neural computation

COSYNE 2025

Interpretable “component-encoding” models for multi-experiment integration

COSYNE 2025

Mixtures of decoders for interpreting dynamic neural-behavioral mappings

COSYNE 2025

NetFormer: An interpretable model for recovering dynamical connectivity in neural populations

COSYNE 2025

Astrocyte-neuron interplay is critical for Alzheimer's disease pathogenesis and is rescued by TRPA1 channel blockade

Auditory tetanization effects in human event-related potential (ERP): long-term potentiation (LTP) and lateral inhibition

Behavioral effects of the early life stress, dopaminergic imbalance, and their interplay in male and female prepubertal mice

Characterizing neural selectivity in sensory systems using a data-driven interpretable feature finding method

ERP components of multisensory spatial representation in the visual cortex

Compression-enabled interpretability of voxel-wise encoding models

A cross-network oscillatory motif underpinning cocaine-paired memory retrieval

Defining specific fingerprints of neurodegenerative diseases in the retina

Developing 2-D and 3-D models to disentangle the interplay between mitochondrial vulnerability and inflammation in Parkinson’s disease

Developmental and adult memory capacity control via interplay between non-conventional GluN3A-NMDA receptors and mTOR signaling

Discovering low-dimensional interpretable dynamics from high-dimensional neuronal activity

Model Selection in Sensory Data Interpretation

Bernstein Conference 2024