Icahn School Of Medicine At Mount Sinai

NIH Research Projects · FY 2025 · 2025-06

PROJECT SUMMARY Multiple Myeloma (MM) is an incurable plasma cell malignancy and is the second most common hematologic malignancy. SETD8 (also known as SET8, PR-Set7, or KMT5A) is the lysine methyltransferase that is responsible for monomethylation of histone H4 lysine 20 (H4K20me), an important histone methylation mark. SETD8 also monomethylates non-histone substrates such as the tumor suppressor p53 and replication factor PCNA, causing suppression of p53-dependent transcriptional activation in cancer cells and promoting cancer cell proliferation. Consistent with this, SETD8 is overexpressed in numerous cancers. In particular, we recently reported that SETD8 is overexpressed in relapsed primary MM and high SETD8 expression is associated with poor prognosis. Importantly, SETD8 knockdown (KD) effectively suppressed the proliferation of SETD8-high MM cells, but not SETD8-low MM cells and normal cells. Moreover, primary malignant plasma cells are particularly addicted to the methyltransferase activity of SETD8. We therefore hypothesized that pharmacological inhibition of SETD8 could be a novel and effective therapeutic strategy for the treatment of MM. In our preliminary studies, we discovered UNC0379 (the first SETD8 selective inhibitor), MS453 (the first SETD8 selective covalent inhibitor), and MS2928 (the most potent and selective SETD8 inhibitor to date). Inhibition of SETD8 by UNC0379 reduced H4K20me, induced cell-cycle defects and apoptosis, and suppressed the growth of MM cell lines and primary MM cells without significant toxicity in non-myeloma cells. Furthermore, MS2928 reduced H4K20me more effectively than UNC0379 and effectively inhibited the growth of SETD8-high MM cells, but not SETD8-low MM cells and normal cells, thereby phenocopying the effect of SETD8 KD. Importantly, MS2928 was bioavailable in mouse pharmacokinetic studies and significantly inhibited tumor growth in vivo in two MM cell line xenograft mouse models without apparent toxicity. Encouraged by these promising preliminary results, we propose to optimize our SETD8 inhibitors into in vivo chemical probes and evaluate SETD8 selective inhibitors in MM cellular and mouse models to further test and validate our therapeutic hypothesis. The optimized SETD8 inhibitors to be generated in this project will also be invaluable chemical tools for assessing the therapeutic potential of SETD8 inhibition in other SETD8-overexpressing cancers, and can be further optimized into drug candidates in the future and ultimately translated in the clinic for cancer patients.

NIH Research Projects · FY 2025 · 2025-06

Aromatase Inhibitors (AIs) letrozole, anastrozole, and exemestane are used to treat hormone receptor positive breast cancer in post-menopausal women. Despite showing a survival benefit, many patients discontinue treatment because of painful and incapacitating side effects. Most of these patients cannot tolerate any of the three AIs. There is a significant unmet clinical need to reduce AI toxicity to vastly improve patient quality of life and decrease the incidence of patients discontinuing therapy. However, other drugs that block the effects of estrogen (e.g., tamoxifen) do not induce the same muscle and joint pain and up to 40% of patients sometimes tolerate one of the three AIs better than the others. Unfortunately, there is currently no way to identify which patients should avoid a specific AI. Many patients try each of the three in a painful, frustrating, and emotionally draining trial-and-error process. There is a significant clinical need to identify genetic susceptibility to off- target toxicities of letrozole, anastrozole, and exemestane to match patients with a “best fit” AI from treatment start. This is particularly important in the current clinical landscape when studies are showing significant disease- free survival benefits of continuing AI therapy for up to 10 years. To address these unmet clinical needs, we developed a Drosophila (fruit fly) model to study Aromatase Inhibitor toxicity. Flies can be easily tested in motor function assays to assess muscular issues such as pain or other impairments. Flies can also be put into a variety of “pain” assays where they are tested for avoidance of noxious stimuli; greater avoidance typically reflects increased pain. Feasibility studies using a wild-type control strains show motor impairment and/or greater noxious temperature avoidance upon AI treatment. Excitingly, 15 divergent wild-type strains tested had different sensitivity AIs as seen with human patients (from no motor impairment to motor impairment in reaction to specific AIs). The rich set of Drosophila assays and tools useful in addressing pain and muscle issues and the growing body of work developing, evaluating, and optimizing therapeutics taken together with our feasibility studies make Drosophila ideally suited to dissect genetic susceptibility to AI side effects that cause patients to discontinue therapy. Our goals are to use unbiased genome-wide screens (Deficiency and Drosophila Genomic Reference Panel screens), systematic candidate screens, and FDA-approved compound screens to identify candidate therapeutics and genetic variants AI toxicity. Identifying genetic predisposition to adverse side effects for each AI will allow clinicians to match patients with a specific AI most likely to be tolerated from the outset, thus eliminating AI switching. We will identify candidate therapeutics as co-therapies to reduce AI side effects and also genetic modifiers that suppress these adverse side effects, thus identifying additional candidate therapeutic targets. This R21 focuses on Drosophila; we have assembled a multi-disciplinary team including breast cancer clinicians and researchers and genomics experts to take our findings into the mammalian models and to validate our candidates in patient datasets in the future.

NIH Research Projects · FY 2025 · 2025-06

PROJECT SUMMARY/ABSTRACT While Centers for Disease Control and Prevention clinical practice guidelines recommend delivering HIV prevention services to all adolescents seeking sexually transmitted infection (STI) testing and treatment, only 55% of youth seeking STI care receive HIV testing and 8% are counseled on HIV pre-exposure prophylaxis (PrEP). These gaps in service delivery have persisted over the last decade, despite both STI rates being at an all-time historic high in adolescents and HIV disproportionally affecting adolescents and young adults. Innovative implementation strategies—discrete tools and practices to close evidence to implementation gaps— are needed to enhance HIV testing and PrEP delivery in adolescents. One such strategy is point of care testing (POCT) for gonorrhea and chlamydia (GC/CT), which delivers test results within 30 minutes, rather than the standard 2-3 days for laboratory-based testing. The first GC/CT POCT device was approved by the Food and Drug Administration in 2021, but, as yet there are no clinical or implementation outcome data for GC/CT POCT in routine pediatric settings. This new testing modality can increase opportunities for delivery of HIV testing and PrEP services by heightening HIV risk perception for patients and clinicians and allowing adolescents same- day access to HIV prevention services when they are already in clinic. The proposed research will test our overarching hypothesis—that GC/CT POCT will improve HIV testing, PrEP counseling, and PrEP delivery and be a feasible, acceptable, and appropriate implementation strategy in adolescents seeking STI testing in pediatric settings—through a pragmatic pilot feasibility trial. Leveraging funding for the POCT equipment and supplies provided through a Penn Centers for AIDS Research equipment grant, in Aim 1, we will use a phased implementation design to assess population and patient-level associations between test modality (POCT versus laboratory-based) through analysis of electronic health record data from patients seeking STI testing (n~12000) at three diverse adolescent health clinics. We will use causal inference methods, including interrupted time series analysis and inverse probability regression adjustment, to strengthen the rigor of our statistical models. In Aim 2, we will conduct an embedded mixed methods study using surveys and semi-structured interviews grounded in the Health Beliefs Model and Consolidated Framework for Implementation Research to assess determinants and outcomes of feasibility, acceptability, and appropriateness of POCT as an HIV prevention implementation strategy in clinicians and clinic staff (n~40) and adolescent patients (n~175). We will integrate these data into an Implementation Research Logic Model that will guide design of a Type II effectiveness-implementation trial. This R21 will provide critical funding to support this clinical and implementation outcomes research, which will yield foundational data to inform a large-scale trial testing the clinical effectiveness of GC/CT POCT, paired with various implementation strategies identified in this R21 research, as a strategy to improve HIV testing and PrEP delivery outcomes in pediatric settings.

NIH Research Projects · FY 2026 · 2025-06

Abstract The brain and heart are vital co-dependent organs and proper inter-organ communication is necessary for health. Brain-heart connectivity occurs through neuronal innervation and vascular supply of immune cells and molecules. This heart-brain axis is calibrated by many factors and one of the most important is sleep. Sleep influences the immune system and heart function through top-down efferent signaling. What is unknown, however, is if the heart or cardiovascular injury influences sleep and the brain’s sleep circuits through bottom-up proprioceptive mechanisms. As described herein, our preliminary data suggest that after ischemic myocardial infarction (MI) monocytes are rapidly recruited to the brain’s superior thalamus by reactive microglia. In the brain, monocytes generate TNF that signals through thalamic glutamatergic neurons to increase sleep. Our data purport that augmented sleep after MI limits cardiac sympathetic input to supports heart healing. Our goal is to document the neuroimmune pathways that regulate sleep after MI and the consequential brain-heart outputs that mediate cardiac recovery. Our specific hypothesis is that after MI microglia recruit TNF-producing monocytes to the thalamus to augment sleep which restricts sympathetic outflow from the brain to the heart and promotes cardiac healing. We will achieve our goal through three independent aims utilizing murine models and advanced tools in neuroscience, immunology, and cardiology. In Aim 1 we will investigate how thalamic neuroinflammation after MI promotes sleep. Using cellular tracking, functional, and reprogramming assays we will document changes in neuroinflammation, monocyte influx, and the blood brain barrier after MI. Telemetric recording and scoring of electroencephalogram (EEG) signals coupled with spectral analysis will quantify increases in sleep pressure, drive, and alterations in micro-architecture after MI. Supported by our preliminary data, microglia- specific deletion of Ccl2 or pharmacological blockade of TNF in the thalamus will link these pathways to sleep regulation post-MI. In Aim 2 we will assess how neuroinflammation after MI limits sympathetic output from the brain to the heart. We will stereotactically deliver anti-TNF compounds to the thalamus or delete Ccl2 from microglia and assess the impact on cardiac sympathetic responses post-MI by quantifying heart sympathetic nerve activity and abundance, catecholamine synthesis, and heart rate variability and arrhythmia by telemetric echocardiogram analysis. Finally, in Aim 3, we will modify sleep after MI and test the outcome on heart healing, inflammation, and sympathetic signaling. Mice will be subjected to sleep fragmentation after MI and cardiac flow cytometry, immunofluorescent imaging, and transcript sequencing will evaluate heart inflammation, fibrosis, and remodeling, while MRI will quantify heart function. Heart sympathetic responses will be evaluated; and we will test the specific hypothesis that sleep modulates cardiac inflammation by altering macrophage Adrβ2 catecholamine signaling. Supported by our multidisciplinary team, our discoveries will identify novel fundamental biology linking the brain, heart, and sleep, and advocate for rigorous sleep management in post-MI clinical care.

NIH Research Projects · FY 2026 · 2025-06

PROJECT SUMMARY Alzheimer's disease (AD) is a devastating neurodegenerative disorder affecting millions worldwide, characterized by memory loss and cognitive decline. The Apolipoprotein E (ApoE) gene is closely linked to AD risk, with APOE4 associated increased susceptibility while APOE2 exhibits protective effects, although the mechanisms of this protection remain poorly understood. Nucleic acid-based therapeutics offer promising avenues for treating AD by targeting pathological genes like APOE and promoting therapeutic protein expression. However, their efficacy is hindered by challenges in crossing the blood-brain barrier (BBB). To address this, we aim to develop novel lipid nanoparticles (LNPs) designed to interact with specific BBB-related receptors for efficient delivery of nucleic acids to the brain. Our objectives include synthesizing novel BBB-crossing lipids (BLs), formulating a library of LNPs, and elucidating their trafficking pathways across the BBB. Additionally, we will engineer siRNA and mRNA sequences to modulate APOE4 and APOE2 expression, optimizing their design and assessing their potency. Lastly, we will evaluate the therapeutic effects, pharmacokinetics, and safety of LNP- RNA formulations in mouse AD models, including examination of brain delivery using PET/MRI imaging. This comprehensive approach aims to advance our understanding of AD pathogenesis and develop innovative therapies targeting APOE for the treatment of this debilitating disease.

NIH Research Projects · FY 2026 · 2025-06

Late onset Alzheimer’s disease (LOAD), the most common form of dementia, and Major Depressive Disorder (MDD) frequently co-exist, but the co-morbidity is not explained by common genetic variants. We have therefore sought to identify shared pathophysiological mechanisms, and VGF is implicated in both AD and MDD. Several AMP-AD groups exhaustively profiled gene expression in brain regions from multiple cohorts of AD and control subjects, and then performed systems biology analyses which collectively identified VGF (acronymic) and its network as a key driver (regulator) implicated in LOAD. Additional evidence supporting the relevance of VGF to AD includes: 1) identification of reduced VGF levels in the brains and CSF of patients with neurodegenerative disease including AD; 2) progressive decrease of VGF CSF levels with AD progression, suggesting its use as a biomarker; 3) protective effect against the LOAD risk of an ApoE4 haplotype by high levels of VGF; and 4) our data showing that hippocampal and cortical VGF levels are reduced in subjects with MDD, in mouse models of depression-like behavior, and in the 5xFAD mouse AD model. Our preclinical studies show that 1) VGF overexpression in 5xFAD hippocampus reduces cortical and hippocampal amyloid deposition, microgliosis, astrogliosis, and cognitive impairment, and rescues neurogenesis deficits, and 2) chronic icv infusion of the VGF- derived neuropeptides TLQP-21 or TLQP-62 (named by the N-terminal 4 amino acids and length) have similar effects. As not all VGF activities are easily explained by its known receptors, we therefore identified additional receptors and active members of the VGF network. Screening of a proprietary expression library identified TLQP- 21 binding to the neuropilins, NRP1 and NRP2, and TLQP-62 binding to the fibroblast growth factor receptor 2 (FGFR2). In Aim 1, we propose to better characterize these novel binding interactions biochemically and computationally, and to determine their functional role(s) and downstream signaling, in vivo, in the APP/PS1 AD mouse model, including in depression-like behavior. In vitro, we will assess their effects on microglial migration, activation, and amyloid uptake. In Aim 2, to better understand VGF actions that may be transduced via its causal network, which we constructed from AD and control subjects with and without MDD, we will investigate the role in AD pathogenesis of the RAB3A interactor called rabphilin 3A (RPH3A), a key node in the VGF protein network and regulator of neurotransmitter release and synaptic plasticity. VGF and RPH3A also share an association with cognitive resilience in aged populations. We will assess the outcomes of RPH3A overexpression or ablation on neuropathology, synaptic function, transcriptomics, and AD- and MDD-related behavioral phenotypes in the APP/PS1 and the humanized hAbetaKI (LOAD) mouse models, to determine similarities and differences between this nodal gene and VGF. In Aim 3, we will explore novel subnetworks and key drivers, comparing expression in AD and AD+MDD brain, and cross-cell-type interaction networks that regulate neuron-microglial VGF crosstalk. Our studies will elucidate new mechanisms and targets underlying comorbid AD and MDD.

NIH Research Projects · FY 2025 · 2025-06

Project Summary Physical activity (PA) has been shown to improve the overall health of human beings. The molecular mechanisms responsible for the diverse benefits of PA are not well understood. The Molecular Transducers of Physical Activity Consortium (MoTrPAC) is being formed to advance knowledge in this area. We propose to conduct comprehensive analyses of the MoTrPAC human PA intervention samples, contribute these data to public databases, help identify candidate molecular transducers of PA and elucidate new PA response mechanisms, and help develop predictive models of the individual response to PA. The Genomic, Epigenomic, and Transcriptomic assay site at Icahn School of Medicine at Mount Sinai and New York Genome Center (ISMMS GET) provide the infrastructure, assay and analysis expertise and experience to support this large-scale, comprehensive analysis of molecular changes associated with endurance and resistance exercise. ISMMS GET aims are to 1. Work with the MoTrPAC Steering Committee and Biospecimens Committee to finalize analysis plans and protocols for randomized human clinical trial study; 2. Generate high throughput GET data such as high-depth RNA-seq and Whole Genome Sequencing (WGS), supplemented by additional assay types such as ATAC-seq, Iso-Seq, and CITE-seq based on clinical human samples; 3. QC data and transfer to MoTrPAC Bioinformatics Center (BIC) and perform initial analyses to help ensure data quality; 4. Collaborate with BIC and other MoTrPAC sites to analyze animal and human data from GET and other MoTrPAC analysis sites to identify candidate PA transducers and molecular mechanisms; 5. Integrate various omics data and develop predictive models of PA capacity and response to training in humans; 6. Participate in the preparation of manuscripts and presentations and make the results of the study available to the public using peer-reviewed publications in scientific journals; 7. Make data generated in the study available to the scientific community through MoTrPAC DataHub and Gene Expression Omnibus. The success of GET Sites and the MoTrPAC program will transform insight into the molecular networks that transduce PA into health, create an unparalleled comprehensive public PA data resource, and can provide the foundation for profound advances in the prevention and treatment of many major human diseases.

NIH Research Projects · FY 2025 · 2025-06

PROJECT SUMMARY Flaviviruses infect up to hundreds of millions of people worldwide annually. The spectrum of diseases caused by these viruses display a range of symptoms from mild illness to hemorrhagic fever, encephalitis, and congenital defects. To establish infection and cause pathogenesis, viruses need to suppress their vertebrate host’s innate immune response. Flaviviruses use their non-structural protein 5 (NS5) to suppress type I interferon (IFN) signaling. While this function is highly conserved, there is diversity in the host factors that are targeted. ZIKV, DENV, and YFV NS5s target signal transducer and activator of transcription 2 (STAT2) for degradation/sequestration. These flaviviruses are transmitted by Aedes mosquitoes and circulate in a sylvatic cycle between non-human primates and humans. TBEV, LIV, and LGTV NS5s target tyrosine kinase 2 (TYK2) to inhibit its catalytic activity. These viruses are tick-borne and mainly circulate in rodents, with humans acting as dead-end hosts. TBEV, LGTV, and WNV NS5s target prolidase (PEPD) to block the expression of IFNAR1 on the cell surface. WNV, transmitted by Culex mosquitoes, circulates in birds with humans acting as dead-end hosts. These findings suggest that some flaviviruses evolved NS5s that encode multiple modes of IFN antagonism. While tick-borne flaviviruses and Culex-borne flaviviruses primarily amplify in rodents and birds respectively, their host ranges are much broader. Perhaps this extended host range applied variable evolutionary pressures that resulted in diverse modes of antagonism. My proposal will expand the knowledge of NS5- mediated IFN antagonism evolution and provide a framework for categorizing these mechanisms. I will achieve these goals by testing my central hypothesis that the evolution of flaviviruses is constrained by the structural and molecular determinants that govern their IFN antagonism functions in essential reservoir vertebrate hosts. In Aim 1, I will use structure-function analyses to identify determinants that define each mode of NS5-mediated IFN antagonism. Cryo-EM and extensive mutagenesis will serve to create a molecular “fingerprint” for each mechanism that will be used to screen a range of flavivirus NS5s to identify their mode of antagonism. I will explore the overlap in antagonism and replication determinants by conducting multicycle growth curves of various NS5 mutants in vitro, since not only is NS5 evolutionarily restricted by its function in innate immune suppression but also in the replication of the RNA genome. In Aim 2, I will examine the coevolution of type I IFN signaling in vertebrate hosts and flavivirus NS5. On the viral side, chimeric NS5s that swap regions from flaviviruses with divergent antagonism mechanisms will be tested for both host factor interaction and antagonism capabilities to. On the host side, I will replace endogenous PEPD and TYK2 with orthologs of various vertebrate species and study the ability of various flavivirus NS5s to antagonize these proteins. Together, the proposed work will provide a broader understanding of flavivirus NS5-mediated antagonism both in distinct structural, molecular terms, as well as provide clues about the evolutionary history as well as potential of these viruses.

NIH Research Projects · FY 2025 · 2025-05

Abstract: Affordable and available modern computational and data resources for AI-driven biomedical research projects are in short supply. AI Mount Sinai (AIMS) addresses this gap by providing state-of-the-art GPU capability and capacity for 42 research projects at 26 institutions. Research areas include: (1) AI for medical imaging and multi-modal clinical data, (2) structural and chemical biology, and (3) genetics and genomic sciences. Resulting research products will be disseminated through publications and national data repositories, including but not limited to: dbGAP, TOPMed, AMP-AD, CommonMind, PsychENCODE, BRAIN, Autism Sequencing Consortium and GEO, with the overarching goal of catalyzing translational science. AIMS consists of six Lenovo ThinkSystem SR780a V3 servers for a total of 48 NVIDIA B200 GPUs connected via NVLink and Infiniband NDR400. There are 9 terabytes of memory available on the Graphics Processing Units (GPUs), and an additional 12 terabytes available on the server nodes. Local high-speed NVME storage and DDR5 RAM enables caching of intermediate results by 672 compute cores on the compute nodes. Other features include a job scheduler, parallel file system and archival storage system. Mount Sinai is supplying networking infrastructure, short- and long-term storage and other essential supplies. Five years of operating costs, including expert personnel, facilities (power, cooling and space), hardware and software maintenance is also being funded by Mount Sinai to meet the acute scientific need for AIMS. Expert PhD and MD computational and data scientists' collaborate with biomedical researchers through extensive training and researcher engagement activities to maximize researcher productivity. The broad, long-term objectives of AIMS focus on improving our collective understanding of human disease in the following areas: Alzheimer's, autism, schizophrenia and related behavioral disorders, cardiovascular disease, kidney disease, diabetes type 2, drug addiction, depression and cancer. Without AIMS, there would be slower progress towards the diagnosis and treatment of these conditions affecting human health.

NIH Research Projects · FY 2026 · 2025-05

PROJECT SUMMARY With the unprecedented increase in age of the global population, there is a critical need for improving our understanding of the cellular, molecular, and spatial tissue underpinnings of aging biology. New insight into the age-related mechanisms and pathways in the human lung have the potential to uncover new therapeutic opportunities to modulate the cellular microenvironment, decelerate the lung aging process, and decrease the susceptibility of the elderly to numerous respiratory diseases. We propose to combine highly-parallelized, single-cell RNA-sequencing (scRNA-seq) and spatial transcriptomics profiling with advanced computational analyses to characterize the effect of age on the human lung and peripheral blood cellular microenvironment. As part of the Human Lung Cell Atlas network, we have helped build an integrated scRNA-seq census of hundreds of healthy individuals. The integrated scRNA-seq data set will provide us with the necessary statistical power to decipher the cell composition, cell type expression, and regulatory differences between different age groups of healthy lung and blood donors. We will also use machine learning to model the effect of cell type specific immunomodulators/pathways on lung cell compostion and to identify novel therapeutic targets in order to better understand the underlying aging mechanisms and how they associate with age. Finally, we will validate our computational modeling by measuring the spatial cellular organization of normal lung tissue samples spanning different age groups as well as testing the anti-aging effect of the most promising therapeutic targets on lung function and cellular communication. Our study promises to improve our understanding of the effect of age on cell-cell interactions, gene expression programs, and multicellular communities in the human lung and peripheral blood, identify/validate new treatment strategies, and uncover new biomarkers that can improve the quality of life for elderly individuals. Finally, our data-driven, computational approach can be extended to study the relationship between age and the cellular microenvironment in other tissues or organs and help combat the effect of aging throughtout the human body.

NIH Research Projects · FY 2026 · 2025-05

Deep brain stimulation (DBS) of the subcallosal cingulate (SCC) white matter is an evolving treatment strategy for treatment resistant depression (TRD) with published studies demonstrating sustained long-term antidepressant effects in 40-60% of implanted patients. Converging evidence from positron emission tomography (PET), electroencephalography (EEG) and diffusion tractography (DTI) strongly suggests that DBS mediates its clinical benefits by direct modulation of the SCC -- a key hub in an aberrant neural circuit. Despite encouraging sustained long-term effects in this notoriously difficult to treat patient population, randomized controls trials of SCC DBS and other DBS targets for TRD are now on hold as initial results failed to meet predefined clinical endpoints. In our recently completed first UH3, we were successful in developing a generalizable LFP brain-based biomarker in a non-commercial prototype research device that defines a depression specific clinical depression state (sick vs. well) in all patients. This biomarker can further guide needed dose adjustments in individual patients during ongoing DBS treatment. In this project, we propose to develop a control policy that integrates this novel biomarker with additional measures to fully capture ongoing clinical decisions during long term DBS treatment, allowing for implementation and scaling of the use of biomarkers in SCC DBS into the next generation commercial clinical DBS sensing devices. We will leverage previously acquired UH3 data (LFP, EEG, imaging and video diary) from 3 consecutive experimental trials of SCC DBS involving 3 independent TRD cohorts similarly recruited, implanted and treated with chronic high frequency stimulation and studied using 3 models of the Medtronic DBS stim/sense DBS system (PC+S, RC+S, Percept) during long-term clinical monitoring and ongoing data collection, to design control policies that minimize subjective decision-making during the adjustment of stimulation settings and adjunctive therapy decisions. We will then test utility of the control policy over the course of 52 weeks of SCC DBS in a new cohort of 10 patients who will be implanted with the Medtronic Percept RC system to anticipate and/or improve on DBS and adjunctive treatment adjustment decisions. These methods will be assessed relative to previous cohorts receiving SCC DBS treatment-as-usual in which all clinical decision making was made based on the intuition of the treating psychiatrist. The primary outcome measure will be to determine if comparable efficacy can be achieved while minimizing subjective decision-making. If successful, the data- driven model and control strategy will enable objective, rational clinical programming of DBS stimulation for depression and provide a new model and approach for stimulation initiation and long-term monitoring and management of patients receiving this treatment.

NIH Research Projects · FY 2026 · 2025-05

Project Summary Frontal cortical circuits undergo a protracted developmental process from adolescence to early adulthood to establish cognitive functions such as attention and working memory. This developmental process is disrupted in neurodevelopmental and psychiatric disorders. It remains poorly understood how key frontal circuits and their associated brain networks are differentially recruited during cognitive behavior between adolescents and adults, and how the dysregulation of frontal circuit development contributes to cognitive impairments. Identifying the specific types of neurons in the frontal cortex that are crucial for attention development could lead to the discovery of new therapeutic targets for treating cognitive impairments. Our long-term goals are to characterize the neural circuit mechanisms of frontal cortex maturation that are essential for cognitive function, and to identify therapeutic targets for mitigating cognitive impairments. Attention requires recruitment of the anterior cingulate cortex area (ACA), a subregion of the frontal cortex. In adults, an increase in gamma oscillations between the frontal cortex and sensory cortical areas is also crucial for attention. Among various cell types within the ACA, parvalbumin- expressing interneurons (PVIs) exhibit neural activity that synchronizes with gamma oscillations, and this synchronous activity is known to play a crucial role in attention control. Notably, ACA-PVIs undergo maturation from adolescence to early adulthood, coinciding with the acquisition of improved attentional abilities, and adolescent stress is known to negatively affect the maturation of ACA-PVIs. However, assessing attentional performance in young mice is technically challenging because it typically takes months for mice to learn conventional attentional behavior tasks. As a result, it is entirely unknown how the developmental changes of ACA-PVIs are associated with in vivo network activity and attentional behavior across adolescence, and how stress during adolescence affects PVIs and associated downstream neural activity during attentional behaviors in adulthood. Our preliminary study overcame this challenge by applying an improved fast attentional testing protocol which enabled the recording and manipulation of neural activity in adolescent mice performing the attentional behavior task. Here, we will test the hypothesis that maturation of attentional performance following adolescence depends on increasing recruitment of ACA-PVIs to synchronize ACA and visual cortex (VIS) activity. We will also test the hypothesis that social stress during adolescence reduces recruitment of ACA-PVIs in adulthood, leading to a decline in attentional performance, and that stress-induced attention deficits can be alleviated through genetic manipulation of ACA-PVI maturation. To test these hypotheses, we will integrate techniques for manipulating and monitoring cell-specific activity in developing mice with the aforementioned attentional testing protocol.

NIH Research Projects · FY 2026 · 2025-05

Over the past decade, the US has been affected by a re-emerging stimulant use public health crisis and alarming increases in crack/cocaine-related overdose deaths. In contrast to other types of addiction, there are no FDA approved treatments for crack/cocaine use disorder (CUD). Developing and testing evidence-based treatment options for this population, and exploring the underlying neural substrates, are therefore urgently needed. Core symptoms of addiction are craving and enhanced reactivity to drug cues, attributed to impairments in prefrontal cortical-mediated functions that include reduced inhibitory control. Using a battery-powered portable device that can be used safely, remotely and repetitively, we have shown in a Phase-1 pilot trial that 15 sessions of real (vs. sham) transcranial direct current stimulation (tDCS) over the dorsolateral prefrontal cortex (dlPFC) reduced craving in individuals with CUD. Combining tDCS with cognitive/emotional training is optimal, as supported by human (and mechanistic) studies. Therefore, here tDCS will be administered during a dlPFC-dependent learning strategy, namely cognitive reappraisal (CR) of drug cues, shown to reduce motivated processing of and spontaneous attention bias to these salient cues in CUD, improving clinical outcomes in drug addiction. With four groups (total N=120), comprised of real- vs. sham dlPFC tDCS crossed over with CR vs. no CR, we will perform a double-blind randomized trial in the natural setting where individuals with CUD receive inpatient treatment. We will also test the feasibility of training in self-administration (for the ultimate design of remotely delivered trials). Our working hypothesis is that synergistically enhancing dlPFC function by combining tDCS with CR, a cognitive training, will decrease craving and cue-reactivity in CUD. Specifically, we hypothesize a real-tDCS vs. sham beneficial effect. In exploratory analyses, we will test whether this effect is modulated by CR (vs. no CR). Our main outcome measure, craving (and drug use measures), will be assessed at all study sessions and at a one month follow-up. Using a reliable and valid inhibitory control fMRI task before and after the tDCS trial (in all four groups) will allow ascertaining whether dlPFC recovery (post>pre tDCS) underlies the expected craving reductions (expected to be strongest during real-tDCS+CR) and identify biomarkers predictive of outcomes. Thus, in this project we will explore an entirely novel learning paradigm approach, using electrically enhanced neurocognitive training to modulate the neural networks underlying inhibitory control with the therapeutic goal of reducing craving (and hence drug use and relapse) in treatment-seeking inpatients with CUD. In addition to testing this novel intervention, the identification at baseline of predictive MRI-based measures could later be used for the design of timely prevention efforts. Establishing feasibility of self-administration and harnessing ecological validity by venturing to the clinic to enhance recovery in inpatients with CUD at the time/place it is most needed, we aim to empower patients with drug addiction with real-time lifesaving tools.

NIH Research Projects · FY 2026 · 2025-05

PROJECT SUMMARY Dysregulation in stress responsivity, encompassing mesocorticolimbic and hypothalamus-pituitary-adrenal-axis pathways, is a psychiatry-transdiagnostic fundamental phenomenon. Current anxiolytic pharmacotherapies are limited in their efficacy and could cause dependence, low tolerance, and sexual dysfunction. With the growing number of vulnerable individuals suffering from stress-related disorders in society, there is an urgent need for novel therapeutic modalities particularly for early intervention and to improve treatments of stress-related disorders. Convergent evidence from animal and human studies of cannabidiol (CBD), a nonintoxicating and well-tolerated cannabinoid, has shown to have anxiolytic effects on stress reactivity especially in responses to environmental stimuli (cue-sensitized states). CBD has multiple pharmacological targets acting as an allosteric modulator of cannabinoid receptors and a glutamate-modulating agent. Despite recent surges in the use of CBD to alleviate stress symptoms, its pathophysiological mechanisms in human stress-system pathways remain largely unknown. Understanding the mechanisms of action of CBD in stress responsivity may lead to novel therapeutic strategies for stress-related disorders. Based on its safety and the growing evidence of CBD to reduce cue-induced reactivity, we propose to evaluate the mechanisms underlying CBD’s roles in stress response in a clinical high-risk population of young adults with early life adversity (ELA), known to exhibit this phenotype. This proposal leverages our clinical and research expertise with high-risk populations with ELA and our established experience with CBD clinical trials. Specifically, our neuroimaging study demonstrated prefrontal neuroanatomical impairments associated with enhanced clinical symptomatology in individuals with ELA. In relation to CBD, we have shown in healthy individuals that oral CBD is safe and well tolerated. We have demonstrated that CBD reduced cue-induced anxiety in individuals with psychopathology and that the effects persisted even when the cannabinoid was no longer detectable in the body. Moreover, in our randomized, double-blind placebo-controlled trial we not only replicated the original findings that CBD decreased cue-induced anxiety but also showed that it concomitantly reduced physiological stress responsivity marks (cortisol and heart rate). Importantly, its protracted effects were again evident a week after the last administration. Notably, our pharmacokinetic trial replicated our previous findings showing rapid bioavailability in oral CBD dose (400mg) known to have behavioral efficacy. Our hypotheses are that during stress responsivity CBD will: 1) downregulate the neural reactivity in mesocorticolimbic regions; 2) reduce influence of limbic regions within the stress network dependency hierarchy; 3) reduce neuronal viability in mesocorticolimbic regions; and 4) decrease endocrine and behavioral hyperreactivity, in clinical high-risk individuals. This study of unmedicated young adult males and females (N=160) with ELA, will examine neurobiological mechanisms of stress as manipulated by acute CBD vs. placebo (400mg, oral; using a double-blind, randomized, placebo-control design), as well as determine the relationship between CBD-sustained (7-day) change and endocrine and behavioral acute stress responsivity.

NIH Research Projects · FY 2025 · 2025-05

Project Summary Irreversible damage to salivary glands (SGs) and their diverse cell types, is a debilitating condition caused by irradiation treatment for head and neck cancer, and the autoimmune disease Sjögrens Syndrome. Loss and remodeling of SG epithelial cells results in hyposalivation, leading to significant clinical issues including chronic dry mouth, increased dental caries, and oral infections. Currently, preventative and/or curative treatments do not exist, yet cell-based bioengineering therapies hold significant promise for SG regeneration. A significant hindrance to the progress of therapeutic optimization, however, is our incomplete understanding of the cellular cues that regulate SG progenitor dynamics, differentiation and the establishment of a functional secretory organ. The specialized branched architecture of the SG maximizes the total area of cell-cell contact between the epithelium and the components of its microenvironment. This project aims to elucidate how immune cells of the SG niche regulate coordinated epithelial cell processes. Our preliminary work using transgenic models defines intimate interactions between tissue-resident macrophages and SG epithelial cells during critical timepoints of SG formation, and preliminary cell depletion studies confirm the necessity of macrophages in guiding SG establishment. Here, we will leverage our expertise in SG development, discrete progenitor dynamics and niche- epithelial interactions to test the hypothesis that multiple aspects of SG morphogenesis are regulated by diverse functionalities of tissue-resident macrophages. Using genetic mouse models and transcriptomic approaches, we will determine how macrophages govern unique cell dynamics and the establishment of heterogenous SG cell identities. We will further test the hypothesis that macrophage-derived signaling factors govern SG progenitor cells and their directed cell expansion. Using biased cell-specific in vivo gene targeting and unbiased spatial single cell transcriptomic techniques, we will determine common immune-epithelial signaling networks, and potential species-specific differences, required for SG epithelial patterning across human and mouse development. Outcomes from this study will identify novel immune-specific mediators of multiple SG epithelial cellular and molecular processes and contribute at large to the advancement of therapeutic targets for salivary gland repair, regeneration, and restoration.

NIH Research Projects · FY 2026 · 2025-05

In the US alone, there was 3.7 million births in 2022. This number declined by 43% in women aged 15-24 since 2007 but increased in older women; 2% increase in women 35-39, 6% in women 40-44 and 12% in women 45- 49. Pregnancy at a younger age (< 25 year old) is associated with a reduction in breast cancer, while pregnancy starting after the age of 30 is associated with long-term (over 2 decades post-partum) increased risk, peaking at 6 years post-partum. Post-partum breast cancers (PPBC) are more metastatic compared to sub-type, size etc. matched breast cancers in never pregnant patients (NPBC). However, breastfeeding is generally protective against breast cancer but this is not the case for all women. Why lactation is protective in some women but not in others is unknown. The identification of mouse models where lactation is or is not protective respectively but where multiple variables (diet, frequency, length, number of pregnancies etc.) can be eliminated, may enable the identification of the mechanism. This application provides such models. At the sub-cellular level, lactation involves an exceptionally high level of secretion of proteins and synthesis of lipids. The endoplasmic reticulum (ER) is the entry point of the classical secretory pathway. The ER is in close physical contact with the mitochondria so that stress in the ER affects the mitochondria and vice versa. Our lab uses mice that share the same nuclear genome (C57BL/6) but have different mitochondrial genomes (C57 or NZB) (therefore referred to as BL/6C57 or BL/6NZB mice). Our model is relevant to humans as it recapitulates mitochondrial diversity observed in humans. We have characterized lactation in these mice and found that while the BL/6C57 females activates a potent pro-apoptotic stress response of the ER and mitochondria during lactation, these stress responses are transient and do not promote apoptosis in BL/6NZB females. Further, RNAseq data revealed an entirely different genes and pathways expression pattern during lactation. Importantly, these differences in gene expression levels translate in a pro- and anti-tumorigenic environment during lactation in the BL6/C57 and BL/6NZB females respectively. Bulk RNAseq and scRNAseq also revealed overlap of gene signatures between PPBC in humans and BL/6NZB lactating mammary gland and identified a PPBC-like population in BL/6NZB females during lactation. Since γ-Tocotrienol was reported to induced ER-mediated apoptosis, we treated BL/6NZB females late during lactation and found that it can reverse the pro-tumorigenic effect of lactation and PPBC formation. This finding raises the exciting possibility to use γ-Tocotrienol as a prevention strategy against PPBC. Therefore, our goal is to obtain data to support the design of a clinical trial for the prevention of PPBC through the following aims: Specific aim 1: Establish the most effective regimen of administration of γ-Tocotrienol for the prevention of PPBC. Specific aim 2: Investigate the protective effect of γ-Tocotrienol on late pregnancy in older mice. Specific aim 3: Deepening our understanding of the mechanism of action of γ-Tocotrienol on various cell types in the mammary gland.

NIH Research Projects · FY 2025 · 2025-04

SUMMARY ABSTRACT Accelerated biological aging poses significant risks to health and mortality, yet the underlying factors remain poorly understood. Telomere length (TL), a biomarker of biological aging, is inversely associated with chronological age and indicative of aging-related disease susceptibility. Telomeres shorten more rapidly during the first two decades of life due to accelerated physical growth, particularly in infancy and adolescence. Childhood exposures have been linked to TL in later life, emphasizing the importance of these developmental periods in influencing later-life health trajectories. While prenatal toxic metal exposures are associated with TL in infancy, there is limited understanding of the association between essential metals and TL. The gut microbiome (GM) is another important component of healthy aging, with recent studies indicating an association between GM composition in childhood and TL. GM composition is influenced by metal exposures, and the GM also influences xenobiotic toxicity. Specific metal- and metal-microbial cliques, subgroups of metals and microbes, may be associated with TL attrition, as not all metals and all gut microbes likely influence TL jointly. The concept of cliques provides a unique measure to assess associations lying between individual components and mixtures. However, few, if any, studies have investigated the complex associations between metals, microbes, and longitudinal TL attrition from birth to late childhood. Given the potential links between metals, microbes, and TL, we hypothesize that mixtures of metals and specific metal-microbial cliques are associated with TL attrition. The long-term goal of this work is to better understand the interactions between environmental exposures and GM on health and aging across the lifespan. Our overall objective is to investigate and model the complex interactions among prenatal metal exposures, GM, and their influence on TL attrition from early to late childhood. The proposed study will employ a robust methodology, utilizing mixture methods, interpretable machine learning, and causal inference, and will draw on data from the well-characterized pediatric cohort in Mexico City. Specific objectives include identifying associations between prenatal metal exposure mixtures and biological age acceleration, pinpointing subgroups of children susceptible to accelerated TL aging based on metal cliques, exploring interactions between prenatal metals and childhood gut microbiome cliques in late childhood, and their association with TL in childhood. Our study team is highly motivated and well-positioned to complete the proposed study. Despite the complexity of the project, its significance lies in paving the way for targeted interventions aimed at preventing adverse health outcomes from an early developmental stage by identifying early targets for metal and microbial interventions. This research is part of a broader continuum, aiming to address the complexities of accelerated biological aging for long-term health benefits.

NIH Research Projects · FY 2026 · 2025-04

PROJECT SUMMARY Cancer is a leading cause of death worldwide, with over 1.9 million new cases in 2022 in the United States, with 10% of that been lung cancer. Cancer risks is associated with multitude of factors including hereditary mutations, environment factors and the recently described bad luck hypothesis. The establishment of immune checkpoint inhibitors as treatment of several cancers, emphasizes the importance of the immune system in tumor control. However, we are yet to fully understand how the immune system contributes to cancer risks as well as cancer heterogeneity. Cancer Immune surveillance is the posits that lymphocytes, particularly T cells, are constantly surveying tissues for neoplastic cells, by recognizing neoantigens presented by the MHC molecules through their T cell receptor (TCR). Our group discovered using population level genetic datasets, that heterozygosity at the HLA class II (HLA-II) loci –rather than the traditionally emphasized HLA-I – is linked with a decreased risks of developing lung cancer among smokers. I plan to follow up on this work and dissect the mechanisms involved in HLA zygosity on tumor development as well as understand the emerging role of HLA II and CD4 T cells in tumor immunosurveillance. Aim 1 focuses on using a murine model that to explore MHC heterozygosity in the development of lung cancer. The murine model, paired with carcinogen induced lung cancer, will be used understand the mechanism behind MHC heterozygosity protective effect against lung cancer as well as explore what T cell type contributes to this protectiveness. Aim 2 seeks to utilize genetically modified cell lines to study exclusively MHC class I versus II heterozygosity in reducing tumor burden and further I will use genetically engineered mouse models to interrogate the role of MHC class II in tumor immunosurveillance and assess the interplay between MHC class II on professional versus nonprofessional antigen presenting cells in controlling tumor growth. However, consistently murine models do not completely reconcile to human biology and will not fully capture the diversity that exists in the immune system. Therefore, Aim 3 will use anonymize and publicly available human lungs single cell datasets to study the HLA heterozygosity on TCR repertoire diversity as well as elucidate clonotypic differences in immune subpopulations. Overall, this proposal combines in vivo models with human datasets to investigate the hypothesis that HLA genetics, specifically HLA-II, shapes cancer immunosurveillance, which can be used to identify biomarkers and inform future treatment strategies.

NIH Research Projects · FY 2026 · 2025-04

Rationale: Non–muscle invasive bladder cancer (NMIBC) accounts for 70-80% of newly diagnosed bladder cancer cases worldwide. Intravesical mycobacterium bovis Bacillus Calmette–Guérin (BCG) is the only US Food and Drug Administration (FDA)-approved first-line treatment option for high-risk NMIBC. Intravesical BCG targets the tumors by activating the mucosal anti-tumor immunity in the bladder tumor microenvironment (TME) and this involves intricate molecular interactions within and between distinct cell populations manifesting in different regions of the tissues. To this end, a comprehensive molecular atlas of NMIBC TME is needed to dissect the cellular and spatial features, and ultimately the crucial signaling cascades amongst these features that underlie the BCG resistance. Single-cell and spatially resolved sequencings hold promises to reveal the near-cellular resolution snapshots of the BCG resistant TME. Current studies on NMIBC TME, however, are often focused on the after-mass of BCG treatments, dominated by the muscle-invasive bladder cancers, are based on small numbers of samples (< 15), and mostly focus on each –omics data modality. Overall, a holistic, data-driven molecular model of NMIBC BCG resistance is lacking, which hinders the systematic search for the therapeutic axes. Method: In this study, we will generate tissues from BCG-treated NMIBC tumors from the responsive and unresponsive patients before BCG (pre-BCG) and after BCG (post-BCG) to explore these therapeutic axes. Specifically, we will elucidate the spatial and cellular landscapes of BCG-resistant NMIBC TME through i) matched single-nuclei RNA and ATAC sequencing, and near-cell resolution spatial sequencing across 36 samples, and ii) bulk RNA sequencing across 132 samples to validate single-nuclei and spatial findings and provide robust transcriptional readouts with the larger sample size. These data will be systematically utilized to construct cell type-resolved gene network models to survey de novo BCG resistance mechanisms and regulators, and repurpose pre-clinical or FDA-approved drugs to abolish multi-facetted BCG resistance. Expected Results: To our knowledge, this is the first effort to simultaneously study cellular genomic, epigenomic and spatial features of BCG-unresponsive NMIBCs. The altered cellular and spatial changes from each type of sequencing data will be integrated into the cell type-resolved multi-scale network model to elucidate their inter-wined signaling cascades as novel BCG resistance mechanisms and their regulators. Ultimately, these mechanistic insights will inform the identification of pre-clinical and FDA- approved drugs to target the BCG resistance regulators, and these will be validated in immune competent tumor explants model to observe their anti-tumor efficacies. Overall, this study will yield valuable resources of high resolution sequencing data to elucidate the NMIBC TME, and address the clinical unmet needs to tackle BCG resistances in NMIBC.

NSF Awards · FY 2025 · 2025-04

This I-Corps project is based on the lab to market translation of an open-source web application that allows doctors and researchers to share digital resources and accelerate the development of artificial intelligence (AI) models in medicine. Although AI in medicine is rapidly advancing, current approaches face challenges such as limited data sharing and restricted access to models, hindering its widespread adoption in medical research settings. This solution addresses these issues by creating a collaborative space for doctors and researchers to develop, share, and refine medical AI models. By fostering an open and cooperative environment, this project seeks to make advanced AI tools more widely accessible, helping to develop more robust, adaptable, and impactful medical AI solutions for medical research. This I-Corps project utilizes experiential learning coupled with a first-hand investigation of the industry ecosystem to assess the translation potential of a digital platform to create a collaborative space for healthcare professionals and researchers to work together towards the development and deployment of cutting-edge medical artificial intelligence (AI) tools for use in medical research. The benefits of this approach include interactive medical data viewers, an open-source medical model hub, and a code-free development space, all designed to streamline the creation and deployment of advanced medical research-related AI models. Furthermore, by facilitating accessibility to AI-powered medical research solutions, this technology has the potential to accelerate new discoveries in data-intensive fields of medicine and clinical care, such as personalized medicine and digital twin technologies. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NIH Research Projects · FY 2026 · 2025-04

PROJECT SUMMARY Despite great progress has been made in identifying hundreds of risk genes in autism spectrum disorders (ASD), we know little about modifiers that may exacerbate or ameliorate disease severity and thus can explain this highly heterogenous and extremely complex condition. Here we put forward the innovative “two-hit” hypothesis that posits that coexpression of molecular variants in individual risk genes associated with non- syndromal autism (“hit one”) and molecular variants underlying suboptimal mitochondrial function i.e., impaired neuronal bioenergetics (“hit two”) leads to impaired neurodevelopment, synaptic plasticity, and neuronal network phenotypes explaining the etiology of ASD. First, we will establish “two-hit” hiPSC lines, presenting with impaired neuronal bioenergetics caused by 30% of mtDNA heteroplasmy and deficiency in PPP2R5D or SHANK3 or GRIN2B autism relevant risk genes. Next, we will test our hypothesis in two specific aims. In specific aim 1, we will resolve the relationship between suboptimal mitochondrial function and the impact of molecular genetic defect in PPP2R5D, SHANK3 or GRIN2B genes on neuronal development and synaptic plasticity. We will image “two hit” iNeurons and 3D brain organoids to reconstruct neurons and neuronal networks as well as quantify soma size and dendritic morphology. We will also assess the number of synapses by quantifying presynaptic Synapsin 1/2 puncta on MAP2-positive dendrites. Finally, in 3D brain organoids, we will assess the production of radial glial cells and the production of cortical neuron subtypes expressing markers found in cortical layers. In specific aim 2, we will delineate that neuronal network-level phenotypes observed when disease-associated autism risk variants (PPP2R5D, SHANK3 or GRIN2B ) are expressed in neurons with suboptimal mitochondrial function. We will first examine and compare the spontaneous activity of neuronal networks, assess difference in the level and pattern of synchronous activity, neuronal network burst firing rate, mean firing rate, burst duration, inter-burst intervals, and the percentage of random spikes using multiple electrode array assays. This innovative hypothesis reinterprets the complexity of the genetics and pathophysiology of ASD in the context of neuronal bioenergetics. The proposed studies will also establish improved disease relevant in vitro neuronal models that will (A) overcome current limitations, (B) facilitate the development of novel testable etiological theories and (C) provide insights into disease modifiers, a knowledge that is currently lacking.

NIH Research Projects · FY 2026 · 2025-04

PROJECT SUMMARY Lipid nanoparticles (LNPs) are a promising delivery method for nucleic acid-based therapeutics in the brain. LNPs are composed of ionizable lipids and helper components, which are biocompatible and biodegradable, and can protect RNA from degradation and improve its stability. Additionally, LNPs can be engineered to cross the blood-brain barrier (BBB) and deliver nucleic acids to brain cells, including neuronal populations that are implicated in neurological and psychiatric disorders. Particularly, amino acid and glucose transporters that play critical roles in molecular trafficking have been explored for nucleic acid delivery in multiple animal models. Preclinical data demonstrate the importance of both LNP formulations and these cell surface receptors for effectively crossing the BBB. Despite these important findings, effective and safe delivery of nucleic acids to various cell types in the brain remains a significant challenge in the delivery field. To overcome these challenges, we propose to integrate BBB-crossing lipid nanoparticles (BLNPs) and engineered nucleic acids for brain cell delivery. In preliminary studies, we developed several classes of BLNPs, which showed great potential to cross the BBB and deliver mRNAs into the mouse brain. Moreover, we constructed mRNAs to include specific sequences that lead to their de-targeting to mouse liver. Based on these results, the goal of this proposed project is to develop BBB-crossing lipid nanoparticles (BLNPs) as functional nanomaterials capable of efficiently delivering nucleic acids in vivo, consequently generating strong functions such as gene silencing or protein expression. The following specific aims will be carried out to accomplish our goal: 1) To synthesize and characterize BBB-crossing lipid nanoparticles (BLNPs); 2) To identify lead BLNPs and elucidate their trafficking pathways in mouse models; and 3) To determine delivery efficiency and safety profiles of BLNPs in mouse and non-human primate (NHP) models. The teams at the Icahn School of Medicine at Mount Sinai (ISMMS) and Biogen Inc. will work closely on the discovery and development of the delivery platform that can potentially be applied to several therapeutic indications. We anticipate, if successful, that this project will greatly expand the knowledge for brain delivery and advance nucleic acid-based therapeutics. Multiple early-stage clinical trials will be initiated within the coming decade for the treatment of many common and severe neurological and psychiatric disorders.

NIH Research Projects · FY 2026 · 2025-04

Project Summary/Abstract Translating genetic information into clinical and public health practice requires a detailed understanding of how genetics influence disease risk. While many monogenic disease risk alleles have been identified, the true level of risk these alleles confer remains unknown for many alleles. Many people with disease risk alleles never develop disease, especially for rare diseases. This disconnect between genetic risk and actual pathogenesis limits the clinical usefulness of pure genetic information. To implement precision medicine requires new methods and techniques that integrate non-genetic factors with genetic risk to produce a complete picture of individual risk. In this proposal, we propose to develop new methods to predict disease risk among individuals who harbor rare monogenic risk alleles, refining the accuracy of genetic diagnostics and increasing the clinical usefulness of genetic information. These methods are based on existing approaches that use deep learning (DL) to impute missing phenotype data. DL methods can leverage information about the context in which diagnoses are seen to learn information about the relationships between diagnoses. This allows them to make accurate predictions about even very rare diseases by analyzing the contexts in which specific diagnoses are seen. At the same time, they can transform these raw data into biologically meaningful signatures, which can be extracted and analyzed to gain insight about disease biology. We will develop and apply these methods in electronic health record (EHR) data, metabolomics data, and proteomics data, using data from the UK Biobank, the NIH All of Us program, the Mount Sinai BioMe Biobank, and the Mount Sinai Million Health Discoveries Program. The result will be robust, validated, and clinically useful prediction methods for disease risk based on DL phenotype imputation, as well as new insights into the biology and genetic architecture of rare genetic diseases.

NIH Research Projects · FY 2026 · 2025-04

Project Summary/Abstract. The goal of this project is to provide a long overdue psychometrically valid and biologically informed update of cognitive status in relapsing remitting and progressive multiple sclerosis (MS). The existing framework used to characterize cognition in MS assumes slowing of cognitive processing speed (COG SPD) is the primary cognitive disability and slowing causes impairments in other cognitive domains (e.g., memory). This framework has never been prospectively tested, does not incorporate the past two decades of advances in disease management, including treatment discoveries that slow disease progression alongside improved diagnostic sensitivity leading to ascertainment of milder cases and has not been updated in over 30 years. The speed- centric framework depends on invalid polyfactorial psychometric COG SPD assessment tools. There is no way to know if low scores indicate impaired COG SPD or impairments in other cognitive domains. Cognitive psychology and computational modeling offer strong frameworks to investigate a psychometrically valid, theoretically derived, and empirically testable model of the time-course of cognitive processing in MS. Specifically, the diffusion model (DM) is an empirically validated mathematical method that extracts 'less contaminated' dissociable components of cognitive processing. Components include rate of information accumulation (COG SPD), criteria necessary to make a cognitive decision, and stimulus encoding/motor response execution. Our preliminary psychometrically valid computational modeling research strongly supports a revised cognitive profile in MS that is likely a result of the clinical advances in disease treatment and diagnosis. In aim 1, we translate and integrate cognitive psychology and computational modelling with confirmatory factor analysis to reappraise and prospectively test the longstanding speed-centric model of MS cognition and determine whether there are domain general vs. domain specific changes in DM cognitive processes across MS phenotypes. We test the specific hypothesis that MS cognitive impairment is primarily due to changes in stimulus encoding/motor response execution with larger differences in the progressive compared to relapsing remitting disease course. We also test the hypothesis that COG SPD is impacted in more advanced progressive disease where neurodegeneration is more prominent. In aim 2, we investigate the predictive value of DM components of cognitive processing and use structural equation modeling to evaluate relationships between DM cognitive processes and brain pathology (T2 lesion and normalized brain volumes) along with patient reported outcomes (PRO: psychological/psychiatric and cognitive symptoms) and performance in other cognitive domains (derived from extensive neuropsychological battery). This will lead to a comprehensive scientific understanding of the dynamic interplay across what cognitive domains are impaired in present-day MS, at what stages of disease and pathology, and moderating effects of PROs that will allow us to closely monitor and develop treatments that specifically target those domains and/or moderating variables.

NIH Research Projects · FY 2026 · 2025-04

SUMMARY Immunotherapies have changed the landscape of cancer treatment, but remain mostly focused on CD8 T cells. In liver cancer patients, we found that in addition to expansion of PD-1+ effector- like CD8 T cells, expansion of CD4 T cells with similarities to follicular helper cells (Tfh) was associated with response to PD-1 targeted therapy. Tfh-like were found in close association to progenitor exhausted CD8 T cells and activated dendritic cells. Overall, our data in patients suggest that CD4-helpers may promote intratumoral differentiation of progenitor exhausted CD8 T cells into effector-like CD8 T cells that control tumor growth. The role of CD4-help in priming is well established, but the role of conventional CD4 T cells in sustaining effective anti-tumor responses and contributing to PD-1 targeted therapy needs to be better understood. In this study, using an autochthonous immunogenic mouse model of liver cancer, we propose to define the key cell-intrinsic features and cell interactions of anti-tumor CD4 T cells that promote anti-tumor immunity and enhance responses to PD-1 targeted therapy. In Aim 1, we will study transcriptional programs and secreted factors in CD4-helper T cells that aid anti-tumor immune responses. In Aim2, we will address how PD-1 signaling modulates tumor-specific CD4 T cells. In Aim3, we will identify CD4-helper interacting partners, and how CD4-help affects cell interactions of tumor-specific CD8 T cells. Given that the frequency Tfh-like in patients' tumors was corelated with plasma cells, we will specifically address the contribution of B cells (and the role of antibodies) in modulating anti-tumor responses, including phenotype of tumor-specific CD4 T cells. Finally, we will address the role of dendritic cells interactions with CD4 T cells, beyond T cell priming, to assess whether CD4-licensing is required for effective response to PD-1 targeted immunotherapy. Understanding what are the key interactions and signals that promote persistent T cell responses in tissues is critical to improve therapies for conditions mediated by chronic T cell stimulation. .