Figure 3.

Inhibition of Autophagy Flux at Lysosomal Level and Decrease in Autophagy Initiation Signaling in DRPLA

(A–C) The ratio of LC3II to LC3I was used to quantify autophagic flux in western blot analysis of full-length GFP-LC3 in the supernatant fraction of cerebellar lysates at 14 weeks of age (A). The anti-LC3 antibody recognizes a doublet between 35 and 55 kDa (Figure S4B), consistent with GFP-LC3-I (upper) and cleaved GFP-LC3-II (lower). The level of Atg5-12 conjugate was used to quantify the events of autophagy initiation. Densitometric analysis shows a decreased relative abundance of cleaved GFP-LC3-II to full-length GFP-LC3-I (B) in ATN1-FL-65Q;GFP-LC3 (65Q) mice compared to ATN1-FL-26Q;GFP-LC3 (26Q) and WT;GFP-LC3 (wt) mice. Atg5-12 conjugate (C) is also decreased in ATN1-FL-65Q;GFP-LC3 (65Q) compared to WT;GFP-LC3 (wt). Student’s t test, mean ± SEM, ^∗∗p < 0.01, ^∗p < 0.05.

(D–F) The accumulation of GFP cleavage product and autophagy receptor p62 was analyzed as a measure of autophagy flux blockage in western blot assay of the cerebellar lysates at 14 weeks of age (D). Mouse anti-GFP antibody recognizes only GFP-LC3-II (Figure S4B) and cleaved GFP after longer exposure. ^∗Shorter exposure of anti-GFP signal. Densitometric analysis of the relative abundance of cleaved GFP to GFP-LC3-II (E) as well as the abundance of p62 relative to α-tubulin (F) in WT;GFP-LC3 mice (wt), ATN1-FL-26Q;GFP-LC3 (26Q) and ATN1-FL-65Q;GFP-LC3 (65Q) mice. Student’s t test, mean ± SEM, ^∗p < 0.05.

(G) Accordingly, western blot analysis of autophagy shows a stall in autophagy flux in end-stage ATN1-FL-65Q mice compared to wild-type (WT) as evidenced by relative decrease of GFP-LC3-II as well as increase of p62 in the pellet fraction. ^∗Anti-p62 antibody revealed an additional band 20 kDa above the expected band at around ∼60 kDa in the supernatant fractions of the cerebellum in end-stage mice.

(H) qPCR analysis of Tfeb mRNA levels in the cerebellum of wild-type (wt, white), ATN1-FL-26Q (26Q, blue) and ATN1-FL-65Q mice (65Q, red) at presymptomatic (3 weeks) and early symptomatic (10 weeks) time points. Relative levels normalized to β-actin and Hprt1 are given as a fold change of wild-type. Two-way ANOVA, v1, genotype; v2, age, mean ± SEM (n = 6), ^∗∗∗p < 0.001.

(I and J) Levels of Tfeb protein in the supernatant fraction of cerebellar lysates at 14 weeks of age (I). Densitometric analysis (J) reveals a significant decrease in ATN1-FL-65Q;GFP-LC3 (65Q) compared to WT;GFP-LC3 (wt). One-way ANOVA, mean ± SEM, ^∗∗p < 0.05.

(K) qPCR analysis of Ctsb and Prkag mRNA levels in the cerebellum of wild-type (wt, white), ATN1-FL-26Q (26Q, blue) and ATN1-FL-65Q mice (65Q, red) at an early symptomatic (10 weeks) time point. Relative levels normalized to Hprt1 are given as a fold change of wild-type. One-way ANOVA, mean ± SEM (n = 6), ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.001.

(L) Whole-cell lysates of human fibroblasts from healthy control and DRPLA patients were subjected to western blot analysis for endogenous LC3-I and LC3-II as well as p62. Induction of autophagy with Rap and block with BafA1 for 6 hr resulted in increase of LC3II compared to DMSO in control fibroblasts, while there was no acute response observed in DRPLA patient samples. No changes in p62 levels were evident after 6 hr acute treatment.

(M and N) Analysis of the autophagy flux in control and DRPLA fibroblasts (DRPLA 17) transfected with the tandem RFP-GFP-LC3B reporter. Starvation in Hank’s balanced salt solution (HBSS) medium was used to induce autophagy, BafA1 treatment to inhibit lysosomal degradation. Autophagosomes are marked by yellow signal as a result of combined RFP and GFP double fluorescence. Due to quenching of the GFP signal in acidic environment autolysosomes show RFP fluorescence only. Representative images acquired with Nikon spinning disc confocal microscope display a greater amount of autophagosomes and autolysosomes in control fibroblasts compared to DRPLA after starvation in HBSS for 3 hr (M). Quantification in (N) demonstrates a significant increase in both autophagosomes and autolysosomes in control but not in DRPLA cells after starvation. Addition of BafA1 to starvation medium resulted in a greater number of GFP⁺RFP⁺ puncta as compared to starvation only in control cells. No changes were significant in DRPLA patient fibroblasts (see also Figure S6H). Significance values in the columns show differences for fed versus starved condition and starved versus +BafA1 condition for RFP⁺ or GFP⁺RFP⁺ puncta. One-way ANOVA, mean ± SEM, ^∗p < 0.05, ^∗∗∗p < 0.001. Significance values between control and DRPLA cells for overall puncta are shown above the horizontal bars, two-way ANOVA ^∗∗∗p < 0.001; v1, genotype; v2,- total GFP⁺ and GFP⁺RFP⁺ puncta. Scale bar, 20 μm.

See also Figures S3 and S4.