Glycation Potentiates Neurodegeneration in Models of Huntington's Disease

Abstract

Protein glycation is an age-dependent posttranslational modification associated with several neurodegenerative disorders, including Alzheimer's and Parkinson's diseases. By modifying amino-groups, glycation interferes with folding of proteins, increasing their aggregation potential. Here, we studied the effect of pharmacological and genetic manipulation of glycation on huntingtin (HTT), the causative protein in Huntington's disease (HD). We observed that glycation increased the aggregation of mutant HTT exon 1 fragments associated with HD (HTT72Q and HTT103Q) in yeast and mammalian cell models. We found that glycation impairs HTT clearance thereby promoting its intracellular accumulation and aggregation. Interestingly, under these conditions autophagy increased and the levels of mutant HTT released to the culture medium decreased. Furthermore, increased glycation enhanced HTT toxicity in human cells and neurodegeneration in fruit flies, impairing eclosion and decreasing life span. Overall, our study provides evidence that glycation modulates HTT exon-1 aggregation and toxicity, and suggests it may constitute a novel target for therapeutic intervention in HD.

Figure 1. Glycation increases HTT levels and inclusion formation in yeast.

(a) Fluorescence micrographs of BY4741 (Ctrl), glo1Δ and tpiΔ yeast strains transformed with HTT 25Q, 72Q or 103Q variants fused to GFP (scale bar 20μm). (b) % of yeast cells displaying HTT inclusions (at least n = 3 per condition). (c) Yeast protein extracts were immunoblotted with an anti-GFP antibody. Arrow indicates HTT aggregates. (d) Corresponding HTT levels are presented in arbitrary units (at least n = 3 per condition). Data in all panels are average ± SD, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. For (D), unpaired two-tailed t-test with equal SD.

Figure 2. MGO induces HTT aggregation and toxicity in H4 cells.

(a) Human H4 cells expressing HTT 25Q or 104Q were treated with vehicle (Ctrl) or MGO (0.5 mM) for 16 h. Cells were lysed and immunoprecipitated (IP) with GFP-trap (bottom panels). The whole cell lysates (WCL) and IP samples were probed for AGEs (left panels) or GFP (right panels). Arrow indicates HTT 25Q and HTT 104Q MW. Corresponding loading controls (β-actin) are presented (n = 3). (b) H4 cells expressing HTT 25Q or 104Q fused with GFP were treated with vehicle (Ctrl) or MGO (0.5 mM) for 16 h. After treatment, cells were probed with Hoechst and imaged in vivo. Fluorescence micrographs and (c) corresponding % of cells with HTT inclusions are presented. Scale barm 50 μm. (d) Representative filter trap assay and (e) corresponding HTT levels of cells expressing HTT 104Q treated with vehicle or MGO for 16 h and immunoblotted with an anti-GFP antibody. (f) Protein extracts were immunoblotted with an anti-GFP antibody. (g) The corresponding HTT levels are presented (at least n = 3 per condition). (h) Toxicity of vehicle (Ctrl) or MGO measured by LDH release (n = 3) and normalized to 25Q. Data in all panels are average ± SD, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. For (c,e,g,h), unpaired two-tailed t-test with equal SD.

Figure 3. Glycation impairs HTT clearance.

H4 cells expressing HTT 25Q (a,b) or 104Q (c,d) for 24 h were pre-treated with vehicle (Ctrl) or MGO (0.5 mM) for 16 h. Cells were treated with vehicle or MGO for 24 h together with CHX. Protein extracts were probed for GFP and β-actin, for normalization, and protein levels are presented (at least n = 3). HTT 25Q (e,f) and 104Q (g,h) expressing cells were pre-treated with vehicle (Ctrl) or MGO (0.5 mM) for 16 h. Cells were treated with vehicle or MGO for 2 h together with vehicle (−) or NH₄Cl (+). Protein extracts were probed for LC3 (I and II) and β-actin. LC3-II levels (lower band) were normalized to β-actin and as a metric for autophagy induction, the difference between NH₄Cl and vehicle treatments was calculated. The ratio between MGO and Ctrl is presented as autophagy induction ratio (at least n = 3). (i) HTT 25Q and GFPu 104Q and GFPu cells were treated with vehicle (−) or MGO (0.5 mM) (+) for 16 h. Protein extracts were probed for GFP and β-actin (at least n = 3). Arrow indicates GFPu. (j) Normalized GFPu levels are presented. (k) HTT 25Q or 104Q were treated with vehicle (−) or MGO (0.5 mM) (+) for 16h. Fresh media was conditioned for 6 h in the same cells (−) or (+) and probed in a dotblot system for GFP (n = 3). (l) Normalized Htt released levels are presented. Data in all panels are average ± SD, *p < 0.05, **p < 0.01. For (b,d,f,h,j), unpaired two-tailed t-test with equal SD.

Figure 4. Knockdown of Glo1 or Tpi induces neurotoxicity and decreases lifespan and survival in HTT93Q expressing flies.

(a) 100 heads of flies expressing Htt93Q and knocked down for Tpi, Glo1 or 3 M (as control) were lysed and immunoprecipitated with anti-HTT antibody (bottom panels). The whole protein lysates (WPL) (n = 3) and IP samples (n = 2) were probed for AGEs (left panels) or HTT (right panels). Arrow indicates HTT93Q MW. Corresponding loading controls (α-tubulin) are shown. (b) Adult flies expressing HTT93Q were treated with different concentrations of MGO. Quantification of mean rhabdomeres per ommatidium is presented. (c) Flies were treated during development with MGO. Quantification of rhabdomeres per ommatidium upon eclosion is presented. Number of rhabdomeres per ommatidium in HTT expressing flies with pan-neuronal knockdown of Glo1 (d) or Tpi (e) is presented at day 0 or 7 post-eclosion. 3M + Htt93Q and GFP + Htt93Q were used as titration controls for Glo1 and Tpi silencing lines, respectively. Data in all panels are mean ± SEM, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001; one-way ANOVA with Newman-Keuls post-hoc test.

Figure 5

Knockdown of Glo1 or Tpi impairs development and reduces lifespan in flies. RNAi silencing of Glo1 (a) or Tpi (b) caused a reduction in the percentage of flies emerging from the pupal case. Flies carrying a single copy of the driver (elavGAL4) are shown as a control. Data in panels (a,b) are mean ± SEM. Survival rate was evaluated in flies with pan-neuronal knockdown of Glo1 (c) or Tpi (d) in mutant HTT backgrounds (n = 100–150 flies per genotype). 3M + Htt93Q and GFP + Htt93Q were used in experiments as titration controls for GloI and Tpi silencing lines, respectively. Htt93Q + Glo1 RNAi (mean = 6); Htt93Q + 3M (mean = 12); Htt93Q + Tpi a RNAi (mean = 12); Htt93Q + Tpi b RNAi (mean = 11.5); Htt93Q + GFP (mean = 13). *p < 0.05, **p < 0.01, ****p < 0.0001; for (a,b) Ordinary one-way ANOVA with Newman-Keuls multiple comparisons test; for (c,d) Log-rank (Mantel-Cox) test.

Figure 6. Schematic representation of the effects of MGO glycation in models of HD.

Glycation increases HTT intracellular levels and inclusion formation in yeast and H4. Ultimately, it increases HTT-dependent toxicity, leading to neurodegeneration and reduced viability and lifespan in flies.