Activation of transposable elements during aging and neuronal decline in Drosophila

Abstract

We found that several transposable elements were highly active in Drosophila brain during normal aging. In addition, we found that mutations in Drosophila Argonaute 2 (Ago2) resulted in exacerbated transposon expression in the brain, progressive and age-dependent memory impairment, and shortened lifespan. These findings suggest that transposon activation may contribute to age-dependent loss of neuronal function.

Figure 1. Age dependent increases in expression of LINE-like and LTR retrotransposons in Drosophila brain

(A) Levels of transcripts of R2 and gypsy quantified by QPCR from young (2–4-day) and aged (~14-day, ~21-day and ~28-day) wild type (WT) heads. Transcript levels normalized to Actin and shown as fold changes relative to WT (means ± SEM). (B) ENV immunofluorescence is elevated in brains from older animals (13-day, 23-day, 34-day) relative to ~2–4-day old animals. Projection through the central brain are shown. ENV signal is detected throughout the cortex layer that includes most of the cell bodies as well as in neuropil areas of axons and dendrites (see also individual confocal sections in Figs. 3B and S5B).

Figure 2. “Gypsy-TRAP” reporter detects de novo integration in neurons in aged animals

A ~500bp fragment from the ovo regulatory region containing 5 Ovo binding sites is inserted between Tub promoter and GAL80 gene. A mutated “gypsy-TRAP” construct contains mutations that disrupt each of the 5 Ovo binding sites. In the absence of gypsy insertions, GAL80 expression suppresses GAL4, and UAS::mCD8::GFP is not expressed. In the presence of gypsy integration into the “gypsy-TRAP”, GAL80 expression is blocked, and UAS::mCD8::GFP is turned on (see Fig. S2). (A) Approximately 800 mushroom body Kenyon cell neurons per brain hemisphere are labeled by MB247-GAL4-driven UAS::mCD8::GFP. (B) An example brain from 2–4-day old mutated “gypsy-TRAP”; UAS::mCD8::GFP/+; MB247/+. No GFP labeled neurons seen. (C) An example brain from ~28-day old mutated “gypsy-TRAP”;UAS::mCD8::GFP/+; MB247/+. No GFP labeled neurons seen. (D) An example brain from ~2–4-day old “gypsy-TRAP”; UAS::mCD8::GFP/+; MB247/+. No GFP labeled neurons seen. (E) Example brains from ~28-day old “gypsy-TRAP”; UAS::mCD8::GFP/+; MB247/+. Several GFP-labeled MB neurons seen in each brain. See Table S1 and Fig. S2 for statistical summary and additional example images.

Figure 3. Age-dependent TE expression contributes to memory decline and age-dependent mortality

(A) Levels of transcripts of R1, R2 and gypsy were quantified from young (2–4-day) and aged (~28-day) WT and dAgo2 mutant animal heads. Within all genotypes, aged animals have significantly elevated levels of each of the transposon transcripts (R1, R2, and gypsy), compared to young animals (*, p<0.05, N=4 for both young and aged dAgo2⁴¹⁴ groups, N=7 for both young and aged WT and dAgo2^51B groups), except for the comparison between young and aged groups within dAgo2^51B (p=0.085) for gypsy, which also is elevated in young animals. For R2 and gypsy, transcript levels in dAgo2^51B young groups are as high as in WT aged groups. ~28-day old dAgo2^51B animals exhibit dramatically increased levels of R2 compared to aged WT group (*, p<0.05). For R2, the 5′ probe set was used in this experiment (see Online Methods) (B) ENV immunoreactivity is detected throughout the cortex layer that includes most of the somata as well as in neuropil (see also Figs. 1 and S5B). Central projections are shown for whole mount brains. Brains from dAgo2⁴¹⁴ mutants exhibit higher levels of ENV immunolabeling in ~14-day old and ~30-day old animals, as also is observed with other dAgo2 alleles (Fig. S5B). (C) Western blot detection with ENV monoclonal antibody shows age-dependent accumulation in heads from dAgo2 mutant animals (see also Fig. S5A). Levels for dAgo2^51B appear increased although somewhat variable. (D) LTM performance (means ± SEM) shown for 2–4-day old and ~20-day old WT and dAgo2⁴¹⁴ mutant animals. 2–4-day old dAgo2⁴¹⁴ mutants exhibit significantly reduced LTM performance relative to 2–4-day old WT animals, and show a dramatic further reduction in performance in the 20-day old dAgo2⁴¹⁴ mutant group (*, p<0.05 and N=15). (E) Lifespan is significantly shortened for dAgo2⁴¹⁴ and dAgo2^51B animals relative to WT (log-rank test). (F) Knocking down loki gene expression with lokiRNAi in neurons significant delays mortality (Gehan-Breslow-Wilcoxon test) of the elav/+; lokiRNAi/+ animals compared to heterozygous controls for transgenes (elav/+ and lokiRNAi/+), as well as the onset of age-dependent memory decline (Fig. S7).