Systematic assessment of transcriptomic and metabolic reprogramming by blue light exposure coupled with aging

Abstract

The prevalent use of light-emitting diodes (LEDs) has caused revolutionary changes in modern life, but the potential hazards to health of blue light are poorly understood. N⁶-methyladenosine (m⁶A) is the most prevalent posttranscriptional modification in eukaryotes and can modulate diverse physiological processes by regulating mRNA fate. Here, to understand the effects and molecular mechanisms of daily low-intensity blue light exposure (BLE) and ascertain whether m⁶A methylation plays a role in BLE-induced phenotypes, we constructed a series of Drosophila models under different durations of daily low-intensity BLE and obtained multiomics profiles. Our results revealed that BLE could induce transcriptomic, m⁶A epitranscriptomic, and metabolomic reprogramming in Drosophila along with aging process. Importantly, the m⁶A methylation sites enriched in the 5' untranslated regions (UTRs) of Drosophila transcripts showed strong age specificity and could be altered by BLE. We experimentally validated that aging-related gene Tor and circadian rhythm-related gene per were regulated by 5' UTR-enriched m⁶A methylation. Overall, our study provides a systematic assessment of m⁶A RNA methylome reprogramming by BLE and aging in Drosophila model.

Keywords: RNA methylome; aging; blue light; circadian rhythm.

Fig. 1.

Workflow and preliminary experimental results of this study. A) Schematic illustration of the samples, treatments, and omics techniques used in this study. One day after eclosion, Drosophila melanogaster w¹¹¹⁸ fly colonies were reared in a climatic chamber under a photoperiodic cycle of 12 h low-intensity BLE: 12 h darkness (BD) and constant darkness (DD), respectively. To investigate the mRNA profiles of the F₁ generation under the influence of parental photoperiod and aging, we collected eggs from the 8-day-old and 23-day-old flies, reared them under DD until 3 days after eclosion, and sampled the whole 3-day-old adult male flies for RNA-seq. We then sampled the whole 10-day-old and 25-day-old adult male flies for MeRIP-seq and untargeted metabolome quantification by LC-MS/MS, and we collected adult male heads for RNA-seq. Survival rates of the sampled adult male (B) and female flies (C) under BD compared with DD. ***P < 0.001, ****P < 0.0001, measured with the log-rank test. D) m⁶A levels of poly(A)-RNA quantified by LC-MS/MS. BD10 and BD25: 10-day-old and 25-day-old adult flies reared under the BD photoperiodic cycle, respectively; DD10 and DD25: 10-day-old and 25-day-old adult flies reared under DD, respectively. Mean ± SD; *P < 0.05, **P < 0.01, measured with two-way ANOVA.

Fig. 2.

Poly(A)-RNA-seq of adult male heads. A) t-SNE dimension reduction analysis based on 11,350 genes. B) Venn diagram displaying the common and unique DEGs identified in the single-factor comparisons between groups. An FDR-corrected P-value <0.05 and fold change ≥2 were used as criteria for identifying DEGs. C) Volcano plot between the BD25 and DD25 heads based on 11,057 genes. P < 0.05 and a fold change ≥2 were set as the thresholds of significance. The symbols representing some of the significant DEGs are shown. D) Enrichment and clustering analyses showing the top 20 GO terms of the DEGs identified in the single-factor comparisons between groups. E) The most significant PPI network based on the DEGs identified in the single-factor comparisons between groups.

Fig. 3.

Poly(A)-RNA-seq and untargeted metabolomic analysis of whole adult male flies. A) t-SNE dimension reduction analysis based on 13,636 genes. B) Venn diagram displaying the common and unique DEGs identified in the single-factor comparisons between groups. An FDR < 0.05 and a fold change ≥2 were used as criteria for identifying DEGs. C) KEGG enrichment analysis based on the DEGs identified in the single-factor comparisons between groups. D) PCA based on 208 quantified compounds. E) Heatmap and clustering analysis of the top 50 compounds with high levels.

Fig. 4.

MeRIP-seq of whole adult male flies. A) t-SNE dimension reduction analysis based on 12,619 genes. B) Venn diagram displaying the common and unique methylated genes among groups. C) Annotation locations in genes of the significant (differential) m⁶A peaks. D) Proportions of the m⁶A methylated and DMGs with different m⁶A peak counts. E) m⁶A peak distribution across the mRNA meta-transcript, showing differences in enrichment around annotated locations. F) m⁶A read count frequency near TSSs. G) KEGG enrichment analysis based on the m⁶A DMGs identified in the single-factor comparisons between groups. Nine-quadrant diagrams showing that the m⁶A epitranscriptome and transcriptome are weakly negatively correlated (H) between the BD25 and DD25 flies and (I) between the DD25 and DD10 flies. A fold change ≥2 (1.5), P < 0.01, and an FDR < 0.01 were used as criteria for identifying significant (differential) m⁶A peaks.

Fig. 5.

Differential m⁶A enrichment in the 5′ UTRs with RT-qPCR and Western blot validation. IGV tracks displaying the differential read coverage across (A) Tor and (B) per according to the MeRIP-seq IP and input data of whole adult male flies and (C) Tor and (D) per according to the MeRIP-seq IP and input data of the yw and MTC RNA interference (RNAi) strains (26). The data ranges of all the IGV tracks in each panel have been scaled to the same level. E) RT-qPCR results of aging-related and circadian rhythm-related mRNAs in whole adult male flies. F) Western blotting of Tor and per in whole adult male flies. G) Quantification of relative Tor and per protein expression levels based on the integrated density values (IDVs) from the Western blot images in (F). Mean ± SD; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, measured with two-way ANOVA.

Fig. 6.

Experimental validations in Drosophila strains and Tor-RNAi S2 cells. A) Confocal immunofluorescence visualization of midguts of adults from the Act5C-Gal4/+, Act5C-Gal4/Mettl3^HMS06028, and fl(2)d²/+ strains fluorescently stained with DAPI (left panels) and the m⁶A primary antibody (middle panels) and corresponding merged images (right panels). B) Western blotting of Tor, per, and Mettl3 in the Act5C-Gal4/+ and Act5C-Gal4/Mettl3^HMS06028 whole adult male flies. C) Quantification of relative Tor, per, and Mettl3 protein expression levels based on the IDVs from the Western blot images in (B). D) Western blotting of Tor, per, and fl(2)d in the wild-type control w¹¹¹⁸ (+/+) and fl(2)d²/+ whole adult male flies. E) Quantification of relative Tor, per, and fl(2)d protein expression levels based on the IDVs from the Western blot images in (D). F) RT-qPCR results of the Tor, fl(2)d, Mettl3, and per in the green fluorescent protein (GFP) control and Tor-RNAi S2 cells. G) Western blotting of Tor and per in the GFP control and Tor-RNAi S2 cells. H) Quantification of relative Tor and per protein expression levels based on the IDVs from the Western blot images in (G). Mean ± SD; *P < 0.05, **P < 0.01, measured with the unpaired, two-tailed Student's t test.

Fig. 7.

Rescue experiments and proposed working model in this study. A) Survival rates of male flies under blue light treated with methylation inhibitor DAA or mTOR inhibitor rapamycin, compared with untreated control (n = 3 each). B) Survival rates of female flies under blue light treated with DAA or rapamycin, compared with untreated control (n = 3 each). C) Western blotting of METTL3/14 in adult flies under blue light after DAA treatment, compared with untreated control (n = 3 each). D) Western blotting of METTL3/14 in adult flies under blue light after rapamycin treatment, compared with untreated control (n = 3 each). E) Quantification of relative METTL3/14 expression levels in DAA and the control, respectively. F) Quantification of relative METTL3/14 expression levels in rapamycin and the control, respectively. G) Proposed working model in this study. Our results from this study suggest that aging-induced 5′ UTR-enriched m⁶A epitranscriptomic reprogramming could regulate the fate of aging-related mRNAs, which could further exacerbate the aging processes as a feedback response. Mean ± SD; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001; (A) and (B) measured with two-way ANOVA; (E) and (F) measured with the unpaired, two-tailed Student's t test.