Somatic genome instability in Tauopathies: a single cell DNA- and RNA-seq approach in humans, mice and flies to understand the disease aetiology

Somatically acquired genetic variation has been described in mammalian developing as well as elderly brain. It has been suggested that neurons with genetic anomalies due to genomic instability (GIN) accumulate during neurodevelopment, while these aberrant neurons are only partially retained in neurotypical adult and elderly brain following a selection process. Furthermore, the functional impact of these genomic aberrations in neurons and its contribution to disease remains largely elusive. Interestingly, in a Drosophila 4Rtauopathy model showing neurodegeneration, altered tau expression during neurodevelopment caused mitotic defects leading to an increased incidence of genomic anomalies, suggesting that neurodegeneration is already determined early in life.

We therefore aimed to gain a better understanding of the extent and consequences of GIN in healthy and diseased brain across the different stages of life. To this end, we established and optimized an automated pipeline for single-nucleus genome-plus-transcriptome sequencing (snG&T-seq) of brain tissue to allow the quantification of DNA content variation (DCV) in neurons and the simultaneous investigation of the consequences of these genomic aberrations on the cellular function.

We applied scG&T-seq on developing and post-mitotic single neurons derived from a Drosophila tauopathy model of tau-mediated neurodegeneration and with that, generated the first single-cell DNA copy number profiles for Drosophila neurons. Resulting data indicated that 4Rtau expression only causes aneuploidies when expressed during neurodevelopment, and moreover, that these aneuploidies persist into adulthood. Subsequently, snG&T-seq applied to neurotypical foetal human brain cells hinted at a peak incidence of genomic instability around the end of neurogenesis. Next, applying this approach to adult and elderly post-mortem human frontal cortex tissue unveiled that DCV is more abundant in neurons derived from 4Rtauopathy compared to neurotypical brain. Lastly, using immunofluorescence, fluorescence-activated nucleus sorting (FANS) and single-nucleus sequencing in Alzheimer’s disease (AD) brain, we reported an oxidative stress-associated accumulation of nuclear hypophosphorylated tau in a subpopulation of cycling neurons confined in S phase in AD brains, near amyloid plaques.

Taken together, we provided trustworthy quantification of DCV in healthy and diseased human brain at different timepoints. Furthermore, we revealed that GIN in neurotypical brain is likely determined through mitotic and neuronal differentiation events. In diseased brain on the other hand, we showed that GIN in 4Rtauopathies is triggered by developmental-onset overexpression of 4Rtau, while GIN emerges in AD when tau accumulates in response to oxidative stress and permits the cell cycle to progress to S-phase to evade apoptosis. In sum, our results indicate that GIN in the brain has a developmental component and contributes to neurodegeneration.