tRNA Epitranscriptome and Translational Dysfunction in ADRD

Boston University Medical Campus, Boston MA

Investigators

Abstract

This proposal examines tRNA modifications in Alzheimer’s Disease (AD) and Related Disorders (ADRD). tRNA is important because the amount of protein produced in a cell depends on tRNA availability and function, both of which are highly regulated. tRNA biology is complicated but demands investigation. Humans have up to 12 tRNAs (termed isoacceptors) for each amino acid, with some tRNAs able to interact with different codons depending on the tRNA modification; thus, there are ~250 different tRNAs. Each tRNA is regulated by up to 50 different kinds of modifications with up to 10 modifications per each tRNA molecule. These modifications control tRNA levels and codon specificity. Our preliminary results from the PS19 P301S MAPT ADRD mouse model reveal >25 tRNA modifications in degenerating brains that exhibit dysregulation (up or down) disproportionate to aging. For instance, we observe increased levels of tRNA modifications such as 2’-O-methyl-adenosine (Am), and decreased modifications such as queusine (AD vs. controls). Proteomic studies show corresponding changes in the enzymes producing the modifications. Studies of bacteria, yeast, and cell lines show that the tRNA epitranscriptome and tRNA pool change with stress, which modifies the speed of tRNA translation for particular codons (termed codon bias), whic modifies the pattern of protein synthesis. Indeed, our results show striking levels of codon bias that correlate with the major changes in brain proteomes of the mouse AD model and human AD. In states of dysfunction, mismatches between codon usage and tRNA modification or abundance can dramatically alter tRNA translation which can enhance pathological protein misfolding. We hypothesize that disease stresses alter the system of tRNA modifications and tRNA levels, contributing to mistranslation, protein misfolding, and proteotoxic stress, which can be ameliorated by genetic modulation of tRNAs and modification writers. The hypothesis is addressed with three aims, delineating tRNA metabolism and then testing effects of observed changes with forward genetics. Aim 1 will determine patterns of tRNA expression and modifications associated with disease states in European (EU) and African American (AA) cases in PFC. We will characterize how the levels and modifications of tRNAs change with disease stage among EU and AA brains collected from the highly characterized ROSMAP cohort. Results will be cross-correlated with A) neuropathologies, B) insoluble protein levels, including Tau, A, TDP-43, and SNCA (Aim 2), C) disease-linked genetic polymorphisms (e.g., ApoE), and D) other endophenotypes available on NIA genomic, proteomic, clinical and imaging databases. Aim 2 will characterize codon biases in genes for soluble and insoluble proteomes in ADRD disease states among EU and AA cases. Major pathologies will be examined as outcomes. Aim 3 will determine whether manipulating tRNA modifications or codon bias alters the aggregation of Tau, TDP-43, or other elements of the insoluble proteome. The aim begins by measuring tRNA half-life in 3D human neuron/astrocyte assembloids (which develop AD pathology) in basal conditions (A) or with stress (B). Then we will (C) use forward genetics in 3D iPS-N/A to correct disease-linked activity changes of tRNA-modifying enzymes implicated or identified in Aims 1 & 2. We will also test if human iPSC lines CRISPR’d with synonymous codon (i.e., silent) mutations modulate AD-linked changes in levels or solubility of proteins linked to ADRD.