Crosstalk of clock gene expression and autophagy in aging

Abstract

Autophagy and the circadian clock counteract tissue degeneration and support longevity in many organisms. Accumulating evidence indicates that aging compromises both the circadian clock and autophagy but the mechanisms involved are unknown. Here we show that the expression levels of transcriptional repressor components of the circadian oscillator, most prominently the human Period homologue PER2, are strongly reduced in primary dermal fibroblasts from aged humans, while raising the expression of PER2 in the same cells partially restores diminished autophagy levels. The link between clock gene expression and autophagy is corroborated by the finding that the circadian clock drives cell-autonomous, rhythmic autophagy levels in immortalized murine fibroblasts, and that siRNA-mediated downregulation of PER2 decreases autophagy levels while leaving core clock oscillations intact. Moreover, the Period homologue lin-42 regulates autophagy and life span in the nematode Caenorhabditis elegans, suggesting an evolutionarily conserved role for Period proteins in autophagy control and aging. Taken together, this study identifies circadian clock proteins as set-point regulators of autophagy and puts forward a model, in which age-related changes of clock gene expression promote declining autophagy levels.

Keywords: C. elegans; aging; autophagy; circadian clock; lin-42; primary human skin fibroblasts.

Conflict of interest statement

The authors declare no conflicts of interest.

Figures

Figure 1. Autophagic flux is reduced in aged primary human fibroblasts

(A) Correlation of age and LC3-II protein levels in primary dermal human fibroblasts from differently aged donors. AU: artificial units. (B) Immunostaining of LC3B protein in cells from young and old donors after Bafilomycin treatment. In the bottom row cells stained only with secondary antibody (Cy3) and DAPI are shown. Total signal strength (C) and the number of LC3 positive punctae in a defined area (D) were quantified.

Figure 2. Clock gene expression is deregulated in aged primary human fibroblasts

(A) Circadian expression of Bmal1 mRNA (squares) and Per2 mRNA (circles) in primary human dermal fibroblasts from one young (age 21) and one old (age 67) donor of the cohort. (B) mRNA levels of core clock genes in cell lines from differently aged donors. Data obtained for age groups 20-30 years (white), 40-50 years (grey) and 60-70 years (black) are shown as mean ± SEM of five individual donors per age group. Data are normalised to the mean of age group 20-30 years. Asterisks indicate statistically significant differences between age groups 20-30 years and 60-70 years (unpaired t-test, two-tailed). (C) PER2 and BMAL1 protein levels in age groups 20-30 years (white) and 60-70 years (black). (D) Correlation of PER2 mRNA expression with LC3-II protein levels in the same cell lines. Solid and dashed lines indicate the linear regression curve and the 95% confidence band. Numbers next to data points show donor ages. Data are shown as mean values ± SEM, n=4.

Figure 3. Autophagy in Cryptochrome-deficient MEFs

(A) Circadian accumulation of Bmal1 mRNA (squares) and Per2 mRNA (circles) in synchronised wild type and Cry1/2−/− MEFs. (B) Bmal1, Per2, Nr1d1 and Per1 mRNA levels in Cry1/2−/− MEFs normalised to the corresponding values in wild type MEFs. Data are shown as mean values ± SEM, n=8. (C) Autophagic flux in wild type MEFs (white) and Cry1/2−/− MEFs (black) determined by quantification of LC3-II protein levels in the absence or presence of bafilomycin (Bafilo).

Figure 4. Cell-autonomous regulation of autophagy in mouse fibroblasts

(A) Western blots of LC3-II without (upper panel) or with (lower panel) the lysosome inhibitor Chloroquine in synchronised NIH 3T3 Bmal1-Luc fibroblasts. The time after Dexamethasone treatment is indicated on top. (B) Quantification of LC3-II protein levels (black) and Bmal1-Luc reporter gene activity (grey). (C) Levels of endogenous Bmal1 mRNA (squares), Per2 mRNA (closed circles) and LC3-II protein (open circles). Data are normalised to the respective mean value at 26 h. (D) Circadian luciferase activity in NIH 3T3 Bmal1-Luc fibroblasts after transfection with Bmal1-specific siRNA (black) or an equivalent dose of Non-target (Nt) siRNA (grey). (E) Expression of LC3-II protein at time points indicated in (D). (F) Quantification of LC3-II protein levels. Data are normalised to the mean value at 26 h. Asterisks designate statistically significant differences between siBmal1 and siNt (unpaired t-test, two-tailed). All data are shown as mean values ± SEM, n=4.

Figure 5. PER2 regulates autophagy

(A) Per2 mRNA and protein levels after transfection with Per2-or Bmal1-specific siRNA or an equivalent dose of non-target siRNA. C: untransfected control sample. (B) Bmal1 mRNA and protein levels. (C) LC3-II protein levels. (D) ULK1 protein levels. (E) Ulk1 mRNA levels. All data are shown as mean values ± SEM, n=4. Asterisks designate statistically significant differences of specific siRNA treatment versus control treatment with siNt (unpaired t-test, two-tailed). (F) Autophagy value (LC3-II in inhibitor treated cells – LC3-II in untreated cells) in two different cell lines from donors aged 60-70 years expressing only GFP (white) or PER2-GFP (black). Asterisks designate statistically significant differences of PER2-GFP versus GFP expression (unpaired t-test, two-tailed), n=3.

Figure 6. Lin-42 regulates lifespan and autophagy in C. elegans

Survival analyses of wild-type animals (WT) compared to lin-42(n1089) mutants (A) or to lin-42 (+) transgenic animals (B). Survival analyses of wild-type animals (WT) and lin-42 (+) transgenic animals fed bacteria transformed with empty-vector (con) or with vector expressing dsRNA against bec-1 (C) or atg-5 (D). (E) Table summarizing survival data analysis from (A-D). (F) Quantification of LGG-1/LC3::GFP positive foci in the DA2123 transgenic strain fed bacteria transformed with empty-vector (con) or with vector expressing dsRNA against lin-42 (lin-42 RNAi). Results are plotted as mean ± SEM of GFP foci (***) represent the P-value (<0.0001) calculated by performing the t-test between the 2 conditions. (G) P62/SQST-1::GFP translational reporter strain (HZ589) fed bacteria transformed with empty-vector (con) or with vector expressing dsRNA against lin-42 (lin-42 RNAi). Top panels: green fluorescence channel (GFP) images. Bottom panels: differential (Nomarski) interference contrast images (DIC). Red squares indicate selected areas used for fluorescence foci quantification. White bars = 20μm. (H) Quantification of GFP positive foci in the anterior pharyngeal bulb of HZ589 fed bacteria transformed with empty-vector (con) or with vector expressing dsRNA against lin-42 (lin-42 RNAi). Results are plotted as normalized mean ± SEM of GFP foci, relative to the control. (*) represent the P-value (<0.05) calculated by performing the t-test between the 2 conditions.