Untangling the shared and unique single-cell transcriptomic signatures of different MAPT mutations in iPSC-organoid models and human brain

Background

Up to 20% of all familial cases of Frontotemporal dementia (FTD) are associated with mutations in the MAPT gene, which encodes for the protein tau. To date, 111 unique MAPT mutations have been identified. While many of these mutations are on one specific exon (Exon 10), pathogenic mutations have also been identified on other exons as well as intronic regions that modify Exon 10 splicing. Consequently, each mutation differs in its severity, pathology and clinical presentation. It is therefore difficult to reconcile which mutation-associated mechanisms are of most relevance to the sporadic disease, and which are unique to each modification.

Despite this diversity in MAPT mutation-associated pathology, the vast majority of research surrounding mechanisms of FTD have been focused on just one or two mutations. Given the locations of different mutations on the MAPT gene, we hypothesise that each mutation would alter MAPT expression and function in unique ways, as well as share some elements of cellular dysfunction that fundamentally lead to the development of tauopathy. 

We have a large panel of human iPSC MAPT mutation lines with CRISPR-corrected isogenic controls. We have differentiated these lines into 3D cerebral organoids in order to study the very earliest effects these mutations have in a complex human system. We employ single-cell RNA-sequencing as a powerful tool to characterise the effect of MAPT mutations on the transcriptional profiles of different cell types. Using this approach, we have recently identified that the MAPT V337M mutation accelerates synaptic maturity by upregulating glutamate receptor expression in deep layer excitatory cortical neurons (Bowles et al., 2021). Our most recent work has been focused on comparing two different categories of splicing mutation that increase Exon 10 expression; missense (S305N, S305I) and synonymous/intronic (S305S, IVS10+16)(Bowles et al., 2023). Consistent with our hypothesis, our preliminary analyses indicate distinct pathogenic mechanisms downstream of each mutation in different cortical layers and neuronal sub-types, which also differ from our V337M work.      

Aims

1. Integrate single-cell data from different MAPT mutation iPSC-organoid models in order to identify shared and divergent transcriptomic effects of each mutation

Hypothesis: Different MAPT mutation types will result in both shared and distinct transcriptomic signatures across multiple different neural cell types and brain regions

Approach: Identification, generation and integration of available single-cell RNA-seq datasets from iPSC-organoid models using publicly available data processing tools, e.g. Seurat, followed by differential gene expression analyses in different cell types and mutations and pathway enrichment to infer effects on cellular function. Additional datasets from ongoing experiments in the lab will be generated using organoid dissociation, cell hashing and library preparation for sequencing.

2. Comparison of iPSC-organoid transcriptomic signatures with human brain tissue from MAPT mutation cases and sporadic disease

Hypothesis: MAPT-mutation iPSC-organoids will share some transcriptomic changes with human brain, and the pathways common across all mutation classes will be most similar to sporadic disease.

Approach: Collation of in-lab and publicly available human post-mortem transcriptomic data (single nuclei, bulk RNA-sequencing) from MAPT mutation cases, sporadic tauopathy cases and healthy controls. Analysis of common and unique transcriptomic effects across mutations and disease, with a focus on cell-type specific effects using common RNA-seq processing tools e.g. Limma, WGCNA. These analyses will be contrasted with iPSC-organoid data in order to identify the most common mechanisms of cellular dysfunction in tauopathies, regardless of mutation or pathology.

Training Outcomes

• Expert knowledge in processing and analysis of transcriptomic data.
• Comprehensive knowledge of Unix and R programming languages
• Experience with handling large, computationally intensive datasets
• Planning and experimental design
• Organoid maintenance and single-cell library preparation.
• Critical scientific literature review
• Effective scientific communication