At the Mount Sinai Alzheimer’s Disease Research Center, our translational research integrates human genetics, stem cell–based models, systems neuroscience, and digital assessment to identify mechanisms of risk and early disease in ADRD. We investigate cell-type–specific genetic and tau-related pathways and develop scalable neurophysiological and digital biomarkers to improve early detection, stratification, and intervention across the ADRD continuum.
Rec Fellow Research
2026-2027 Fellows
Francesca Garretti, PhD
Project Title: Investigating the impact of Alzheimer’s disease risk gene EIF2B3 on microglial cell function
Introduction: Human genetic studies provide a powerful means to identify genes and pathways underlying Alzheimer’s disease (AD). Apolipoprotein E4 (APOE4) is the strongest genetic risk factor for sporadic AD and is associated with earlier age at onset (AAO). However, AAO varies widely among APOE4 carriers, suggesting that additional genetic modifiers influence APOE4-associated risk. To identify such modifiers, we performed a genome-wide association study of APOE4 carriers at the extremes of the AAO distribution and identified a single nucleotide polymorphism in EIF2B3 encoding the S404A substitution as a candidate risk modifier. EIF2B3 encodes a catalytic subunit of eIF2B, a key regulator of protein translation and the integrated stress response (ISR). Cellular stress induces eIF2 phosphorylation, inhibiting eIF2B and suppressing global protein synthesis while promoting adaptive stress-response programs. Our lab has shown that APOE4 induces reduced translation in murine and human iPSC-derived microglia. APOE is highly upregulated in microglia in aged and diseased brains, and AD risk variants are enriched in myeloid-specific regulatory regions, supporting investigation of EIF2B3 function in microglia. Using CRISPR-edited APOE3/3 THP1 macrophages, we found that EIF2B3S404A reduces phagocytosis, endolysosomal proteolysis, lysosomal mass, and acidification. We hypothesize that these defects are exacerbated in APOE4 microglia. The overall goal of this proposal is to define how EIF2B3S404A alters microglial function in vitro and in vivo to modulate AAO and AD risk.
Aim 1. Use hiPSC-derived microglia (iMGL) to assess the impact of EIF2B3S404A on APOE4 microglia cell function in vitro.
Aim 2. Use hiPSC-derived hematopoietic progenitor cells (HPCs) and AD chimeric mice to assess the impact of EIF2B3S404A microglia cell function in vivo.
PI: Kristen Whitney, PhD
Project Title: Elucidating sporadic tauopathy mechanisms using patient-iPSC derived mini brain organoids
Introduction: Tauopathies, with Alzheimer’s disease (AD) being most common, are disorders showing abnormal accumulation of tau protein in the brain. Progressive supranuclear palsy (PSP) is an amyloid-independent primary tauopathy and a compelling context to investigate sporadic disease. While changes in proteostasis, cell stress responses, and other mechanisms are implicated, more human studies are required to clarify how tau triggers degeneration. To understand how genetic association confers disease risk and to test candidate mechanisms, there is a critical need for model systems that recapitulate the complex cellular environment of the human brain and retain the genetic complexity of human disease. Induced pluripotent stem cell (iPSC) and organoid models are a powerful tool to study disease pathogenesis, yet sporadic tauopathies have yet to be investigated with these models. To address this, we assembled a collection of sporadic PSP brains and iPSC lines and generated over 9,000 midbrain organoids, enabling us to make direct comparisons. The objective is to utilize a PSP organoid model and human autopsy brain tissue to understand the extent to which the model enables identification of tauopathy-associated candidate pathways disrupted in the earliest stages of PSP. We will characterize molecular changes by performing biochemical, histological, and transcriptomic analyses to yield novel insights relevant to molecular correlates of the degenerating human midbrain.
Aim 1. To investigate tau proteinopathy in a human midbrain organoid model of sporadic PSP.
Aim 2. To validate changes in PSP midbrain organoids against paired human autopsy brain tissue.
2025-2026 Fellows
Ola Alsalman, PhD
Project Title: Multi-rhythmic Stimulation in MCI
Introduction: Alzheimer’s disease (AD) is a progressive neurodegenerative disorder characterized by early disruption of large-scale brain networks that support memory, attention, and executive function. A growing body of evidence indicates that impaired neural oscillations, particularly in the alpha (8–12 Hz) and gamma (~40 Hz) bands, emerge early in the disease course and may contribute mechanistically to cognitive decline by degrading attentional control, sensory integration, and memory encoding. Importantly, 40 Hz sensory stimulation has demonstrated potential to engage neurophysiological and neuroimmune pathways and reduce AD related pathology in preclinical models. Building on this foundation, our preliminary human data suggest that multi-frequency visual stimulation can enhance sustained attention and working memory beyond unimodal stimulation, with accompanying improvements in reaction time, response accuracy, and broader frontal parietal engagement. This REC supported study evaluates a physiologically grounded dual frequency pairing of 10 Hz (alpha) and 40 Hz (gamma), designed to target complementary oscillatory functions. Alpha rhythms are central to top-down attentional gating and regulation of cortical excitability, while gamma rhythms support local synchrony and perceptual binding. We will test whether dual frequency visual stimulation produces stronger EEG-based neural synchrony, including alpha gamma coherence (connectivity) and phase amplitude coupling, than 40 Hz-only or sham stimulation, and whether stimulation induced increases in neural synchrony are associated with measurable gains in attention and executive function performance.
Aim 1. To determine whether dual-frequency visual stimulation more effectively enhances alpha gamma neural synchrony in older adults with MCI compared with single frequency, or sham stimulation. Using a within-subjects crossover design with repeated EEG recordings, we will test the hypothesis that dua frequency stimulation produces stronger entrainment of alpha and gamma oscillations and more robust cross frequency coupling than unimodal or sham conditions.
Aim 2. To determine whether stimulation-induced increases in alpha gamma neural synchrony translate into measurable cognitive benefits. We hypothesize that participants exhibiting greater enhancement in EEG based coherence and phase amplitude coupling will demonstrate corresponding improvements in sustained attention and cognitive control on standardized tasks.
Katherine Hackett, PhD
Project Title: Multimodal digital assessment of behavior and speech to characterize cognitive, behavioral, and biological risk in early neurodegenerative disease
Introduction: Early detection of Alzheimer’s disease and related dementias (ADRD) is essential for timely initiation of disease-modifying treatments, yet existing assessment tools have limitations. Smartphones offer a low-burden approach to measure early cognitive and functional decline by passively capturing everyday behaviors outside the clinic. With 76% of older adults owning smartphones, these devices can continuously monitor functional domains including mobility, socialization, and technology use linked to dementia risk (i.e., those with incident decline and AD pathology spend more time at home, reduce social interactions, and experience technology use difficulties). Pilot data shows that older adults with worse cognition exhibit less active and less variable smartphone-derived GPS mobility patterns. Together with findings from other passive monitoring studies, results suggest that digital metrics of routine behaviors may serve as sensitive indicators of AD risk. Naturalistic speech analysis offers a complementary digital approach. Brief speech samples can be processed into acoustic, lexical and syntactic features capturing clinically meaningful signs like word-finding pauses, reduced linguistic complexity, or diminished emotional valence that can be used to distinguish clinical and biologic risk groups.
Importantly, few studies have examined the combined utility of these low-burden digital assessments. This study will collect passive smartphone phenotypes in a well characterized ADRC sample with normal cognition (NC), mild cognitive impairment (MCI), subjective cognitive decline (SCD), and mild behavioral impairment (MBI).
Aim 1. Identify digital phenotypes of everyday behavior and speech that distinguish clinical (NC vs. MCI vs. SCD vs. MBI) and biological (pTau 217 +/-) groups.
Aim 2. Determine the added utility of digital metrics beyond a common cognitive screener.