2026 Theses Doctoral

Characterization of Human Induced Pluripotent Stem Cell-Derived Microglial Responses in Alzheimer’s Disease

Cardona, Christopher L.

Alzheimer’s disease (AD) is a progressive neurodegenerative disease that involves the accumulation of extracellular amyloid deposits and intracellular tau tangles. Genome-wide association studies (GWAS) have identified key genetic variants that significantly alter the risk of developing AD. Importantly, many of these genes are primarily expressed in microglia and converge on microglial pathways, such as amyloid processing, lipid homeostasis, and phagocytosis. However, the mechanisms by which these variants affect microglial transcriptomic and metabolic states remain poorly understood.

Here, our objectives were to: (i) characterize transcriptomic and metabolic changes associated with loss of TREM2 in human microglia, and (ii) develop a high-throughput CRISPR screening method to identify key risk genes that regulate microglial inflammation. TREM2 has been implicated in AD as a critical regulator of microglial functions and loss of function mutations in TREM2, such as TREM2 R47H, increase the risk of developing AD. Given the diverse functional roles of TREM2, we sought to unravel how TREM2 affects downstream pathways using human induced pluripotent stem cell (hiPSC)-derived microglia-like cells (iMGLs) in TREM2 knockout (KO) lines and isogenic controls. We leveraged bulk RNA sequencing and targeted lipidomics approaches to uncover transcriptional and metabolic alteration mediated by TREM2 KO. At baseline, we identified a set of TREM2-dependent genes, which correspond to the disease-associated microglial (DAM) signature described in mice.

Additionally, loss of TREM2 resulted in increased expression of cholesterol biosynthesis genes. In response to oligomeric amyloid-β 1-42 (oAβ), we found that the transcriptional response was largely TREM2 independent and corresponded to a strong inflammatory response. However, we identified pathways that were unique to either controls or TREM2 KO microglia, suggesting subtle differences in gene expression driven by TREM2. Given that the inflammatory response was TREM2-independent, we sought to identify the key drivers of inflammation. Pharmacological inhibition of toll-like receptor signaling demonstrated that inflammatory responses to oAβ were primarily driven by TLR signaling rather than TREM2. Cytokine production in response to oAβ was significantly increased to similar levels in both controls and TREM2 KO microglia. Excitingly, lipidomic profiling of microglia revealed altered lipid composition between controls and TREM2 KO microglia, further highlighting a role for TREM2 in promoting metabolic adaptation.

Taken together, we found that TREM2 primarily regulates lipid and metabolic adaptation and lipid-associated DAM gene expression that than acute inflammatory activation. Greater than a hundred genetic risk factors for AD have been identified by GWAS, but the pathogenic role for most of them remains unclear. We sought to assess at scale how AD GWAS hits influence microglial inflammatory responses. Thus, we conducted CRISPR inhibition (CRISPRi) screens of 119 AD GWAS hits in human iMGLs, with reactive oxygen species (ROS) produced in response to the viral mimic poly(I:C) as a readout for inflammation. Top hits that either decreased or increased ROS levels in response to poly(I:C) when knocked down were then interrogated in CROP-seq experiments. We identified nine microglial clusters that significantly overlapped with previously identified AD-relevant microglial clusters, including a poly(I:C) driven inflammatory cluster. Distinct AD GWAS genes reshape inflammatory (MS4A6A and EED) and metabolic microglial states (INPP5D and RABEP1) in response to pro-inflammatory stimuli.

In addition, KD of MS4A6A reduced the proportion of iMGLs in the DAM cluster under all conditions, suggesting that MS4A6A may modulate the DAM response. Further, INPP5D KD, under all conditions, led to an enrichment of a poly(I:C)-independent HLA inflammatory cluster. Despite INPP5D or RABEP1 having disparate biological roles, their perturbation led to similar changes in gene expression, which included changes related to metabolism, suggesting a shared conversion point by which AD GWAS hits affect metabolic cell state to reduce inflammatory activation. Overall, our data begins to elucidate the functional roles played by various AD GWAS hits, including how they regulate inflammatory responses observed in AD.

Collectively, these studies demonstrate that AD risk genes reshape microglial transcriptional, inflammatory, and metabolic states, with many converging on pathways involved in lipid homeostasis and cellular energetics. Our findings position microglial metabolic adaptation as a central feature of AD-relevant microglial dysfunction and provide a framework for understanding how AD genetic risk influences disease-associated microglial responses.