Modeling AD Risk and Protective Genetic Factors in Human CNS Cell Types

One large focus of the lab is modeling risk and protective factors for Alzheimer’s disease (AD) using human pluripotent stem cell (hPSC) models. hPSCs are differentiated into various central nervous system (CNS) cell types: neurons, microglia, astrocytes, and brain microvascular endothelial cells (BMECs). We assess the phenotypic effects of various manipulations of AD-related genetic factors by various methods including knocking in mutations into control hPSC backbones using CRISPR/Cas9, using CRISPRi/a to manipulate gene expression, and more conventional siRNA and lentiviral overexpression approaches.

Protective genetic factors for PSEN1 G206A: Our collaborator Dr. Joseph Lee has identified protective genetic factors for the PSEN1 G206A mutation.  PSEN1 G206A, while an autosomal-dominant mutation sufficient to cause familial AD (FAD), can have a range of onset over 20 years in an affected Caribbean Hispanic cohort. We have been studying a protective genetic variant that can delay onset of FAD by almost 10 years. Our data suggests that its manipulation has a stronger effect in microglia rather than neurons, and acts at least in part by dampening the pro-inflammatory response via reduced JNK and NFkB signaling (in preparation for resubmission).

High throughput identification of regulators of neuroinflammatory processes in AD: This collaborative formerly ADRC-funded Development Project between the Chavez, Hargus, and Sproul labs is using CRISPRi screens to assess over 100 GWAS-identified AD genetic risk factors in the context of microglial inflammation as measured by reactive oxygen species (ROS). Top hits were further analyzed by CROP-seq for effects on microglial cell state (Cardona et al, in press Journal of Neuroinflammation, 2024). We plan on expanding this pooled screening approach to other inflammatory readouts as well as test how different AD GWAS risk factors work in an epistatic manner to drive inflammatory processes.

PSEN G206A KI Neurons

FACS on different levels of microglial inflammatory response

Generating Novel Microfluidic Human Blood Brain Barrier Models of AD

The blood brain barrier (BBB) is a critical component of the CNS which tightly regulates influx of molecules and cells into the brain. Brain microvascular endothelial cells (BMECs) are the core component for this barrier, and along with astrocytes and pericytes comprise the neurovascular unit (NVU).We recently contributed to a study where it was demonstrated that the most commonly used stem cell differentiation protocols used to produce BMECs instead generates cells that have an epithelial rather than endothelial identity, albeit with good barrier properties (Lu et al., PNAS 2021). In collaboration with Dr. Dritan Agalliu, we developed a novel BMEC differentiation protocol (rBMECs) using a combination of FOXF2 and ZIC3 transcription factors and developmental patterning. We combined rBMECs with hiPSC-derived astrocytes and perictyes in microfluidic devices to make a functional human NVU using the MIMETAS platform which have been using to model aspects of AD (Cui et al. in press ATVB 2025). We are also starting to expand this NVU model by incorporating neurons and immune components including peripheral monocytes and microglia.

TEER measurement of barrier tightness

We have additional collaborative projects, two of which are highlighted below.

HTS of compounds to promote synaptic function and mediate AD phenotypes: Synaptic loss correlates well with AD progression, and is a potential target for therapeutic intervention. In collaboration with Dr. Andrew Teich we have just completed a screen of ~1000 compounds in hiPSC-derived neurons for beneficial effects on synaptic function and other AD-related pathways, using bar-coded RNA-Seq (PLATE-Seq) as a readout. In order to interpret these lower read depth profiles, we have identified master regulators (MRs) in human neurons generated by Ngn2-mediated transdifferentiation, which are associated with the expression of many other genes for which we can use MR as a proxy (manuscript in preparation).

Selective vulnerability of excitatory vs. inhibitory neurons to tauopathy: We are collaborating with Dr. Karen Duff (now at UCL) to better understand why neurofibrillary tangles selectively form in excitatory (EX) glutamatergic neurons but not in inhibitory GABAergic neurons during tauopathies such as AD or frontotemporal dementia (FTD). In order to better study this in human neurons, we developed a permanent line method to generate either EX or IN neurons from hPSCs by addition of doxycycline to drive transgenes and neuronal media (Song et al, Current Protocols 2021). Interestingly, we find that the APPLon mutation differentially affects EX and IN neurons in terms of transcriptional and lipidomic profiles (manuscript in preparation).

Other Collaborative Projects

Evoked APs in excitatory and inhibitory transdifferentiated neurons