Transcriptional induction of the heat shock protein B8 mediates the clearance of misfolded proteins responsible for motor neuron diseases

Abstract

Neurodegenerative diseases (NDs) are often associated with the presence of misfolded protein inclusions. The chaperone HSPB8 is upregulated in mice, the human brain and muscle structures affected during NDs progression. HSPB8 exerts a potent pro-degradative activity on several misfolded proteins responsible for familial NDs forms. Here, we demonstrated that HSPB8 also counteracts accumulation of aberrantly localized misfolded forms of TDP-43 and its 25 KDa fragment involved in most sporadic cases of Amyotrophic Lateral Sclerosis (sALS) and of Fronto Lateral Temporal Dementia (FLTD). HSPB8 acts with BAG3 and the HSP70/HSC70-CHIP complex enhancing the autophagic removal of misfolded proteins. We performed a high-through put screening (HTS) to find small molecules capable of inducing HSPB8 in neurons for therapeutic purposes. We identified two compounds, colchicine and doxorubicin, that robustly up-regulated HSPB8 expression. Both colchicine and doxorubicin increased the expression of the master regulator of autophagy TFEB, the autophagy linker p62/SQSTM1 and the autophagosome component LC3. In line, both drugs counteracted the accumulation of TDP-43 and TDP-25 misfolded species responsible for motoneuronal death in sALS. Thus, analogs of colchicine and doxorubicin able to induce HSPB8 and with better safety and tolerability may result beneficial in NDs models.

Conflict of interest statement

The authors declare no competing financial interests.

Figures

Figure 1. HSPB8 enhances the clearance of TDP-43 and TDP-25 (A,B) Immunofluorescence microscopy analysis (63x magnification and 2.5x zoom) of mouse motoneuronal NSC34 cells transiently overexpressing GFP-TDP-43 or GFP-TDP-25 plasmids.

Cells were stained for the endogenous autophagic markers p62/SQSTM1 (A) and LC3 (B). Nuclei were stained with DAPI. (C) Immunofluorescence analysis of NSC34 cells transiently overexpressing GFP-TDP-43 or GFP-TDP-25 and pCDNA3 or hHSPB8. Where indicated, cells were treated with 10 mM 3-methyladenine (3-MA) for the last 48 hrs prior to fixation. DAPI: nuclei staining. RED: human transfected HSPB8. (D) Representative Filter Retardation Assay (FRA) and FRA quantification (left insets), and Western Blot (WB; right insets) of the PBS extracts of NSC34 cells transiently overexpressing the GFP-TDP-43 or GFP-TDP-25 and pCDNA3 or hHSPB8 plasmids. The histogram represents densitometric analysis of three independent replicates for each sample (**p < 0.01 vs GFP-TDP-43; °°p < 0.01 vs GFP-TDP-25).

Figure 2. Development of the cellular system for the screening of the HSPB8 promoter modulators (A) Validation of the GFP-hPromB8LUC plasmid.

NSC34 cells were transiently transfected with GFP-hPromB8LUC and pRL-TK plasmids. 24 hrs after transfection, cells were treated with DMSO or 10 μM MG132 for the last 16 hrs prior to luminescence analysis. The histogram represents the Luminescence Counts Per Second (LCPS) of the inducible firefly luciferase normalized by the LCPS of renilla luciferase of six independent replicates (**p B) Clones selection of NSC34-GFP-hPromB8LUC (#Ns) stably transfected cells. Different #Ns clones were analysed for their luminescence levels, 48 hrs after plating. NSC34 untransfected cells were used as negative control. The histogram represents the firefly luciferase LCPS mean normalized by the total protein level mean of six independent replicates for each clone (**p < 0.01 vs NSC34 untransfected cells). (C) Optimization of cell number and signal to background ratio for HTS assay. Raw data of relative luminescence units detected by NSC34 clones after 24 hrs seeding at the indicated cell number. SD refers to two independent experiments. (D) Validation of the #N4 selected clone. NSC34-GFP-hPromB8LUC #N4 cells were analysed for their ability to respond to MG132 treatment. 24 hrs after plating, cells were treated with DMSO or 10 μM MG132 for the last 16 hrs prior to luminescence analysis. The histogram represents the firefly inducible luciferase LCPS mean normalized by the total protein level mean of six independent replicates (**p < 0.01 vs DMSO).

Figure 3. Primary screening parameters and ranking of the 2,000 library compounds.

(A) Distribution of positive (MG132) and negative (DMSO) controls according to firefly units normalized to GFP signal. #N4 cells were treated with 5 μM of MG132 or DMSO for 24 hrs in two columns (1–2 and 23-24, respectively) of each 384-well plate ran (n = 25). (B) Parameters calculated from positive and negative controls validating robustness of the screening. (C) Distribution of GFP signals relative to each compound tested; values near to 100% indicate un-perturbing activity of compound on cellular or GFP-associated metabolism. (D) Plot of compounds ranked according to progressive Z score values (compounds are listed in Supplementary Table 2). (E) Chemical ontology (SODIAC) annotation according to the recognized bioactivity of HSPB8 activating compounds. *indicates less than 1% over-representation; NC = non classified.

Figure 4. Counterscreening of 18 hits by dose-response effects.

(A,B) NSC34 #N4 cells were treated for 24 hrs with four doses of indicated compounds. Relative cytotoxic activity was evaluated by Alamar blue assay following manufacture’s instruction. (C,D) Relative luminescence units detected in cells treated as in (A,B), normalized to cell viability (n = 2).

Figure 5. Hits validation on endogenous HSPB8 expression in human neuronal cells (A) Q-PCR analysis of HSPB8 mRNA levels performed on SH-SY5Y cells.

24 hrs after plating, cells were treated for the last 24 hrs prior to extraction with 1 μM of different compounds or with DMSO (negative control) or 10 μM MG132 (positive control). HSPB8 mRNA levels were normalized with RplP0 mRNA levels (relative levels); histogram represents the analysis of 4 independent replicates (**p B–E) Dose-dependent response analysis of colchicine (B,C) and doxorubicin (D,E) treatments performed on SH-SY5Y cells treated with 100 nM, 500 nM or 1 μM of colchicine or doxorubicin, and with DMSO or 10 μM MG132 for 24 hrs. (B,D) Q-PCR analyses of HSPB8 mRNA levels. The histograms represent the HSPB8 mRNA relative levels mean of 4 independent replicates (*p < 0.05 vs DMSO; **p < 0.01 vs DMSO). (C,E) WB and WB quantification of HSPB8 protein levels. The histograms represent densitometric analysis of three independent replicates for each sample (*p < 0.05 vs DMSO; **p < 0.01 vs DMSO).

Figure 6. Analysis of the heat shock response after colchicine and doxorubicin treatments (A–F) Q-PCR analyses of HSF1 (A,B), BAG3 (C,D) and BAG1 (E,F) mRNA levels performed on SH-SY5Y cells treated for the last 24 hrs prior to extraction with 100 nM, 500 nM or 1 μM of colchicine (A,C,E) or doxorubicin (B,D,F), and with DMSO or 10 μM MG132.

Where indicated, cells where heat shocked for 1hr at 42 °C, and collected after the shock (no recovery) or after a recovery period of 3hrs at 37 °C. Analysed mRNA levels were normalized with RplP0 mRNA levels and expressed as relative levels mean of 4 independent replicates (*p 

Figure 7. Colchicine and doxorubicin effects on the autophagic pathway (A–F) Q-PCR analyses of TFEB (A,B), p62/SQSTM1 (C,D) and LC3 (E,F) mRNA levels performed on SH-SY5Y cells treated as described in Fig. 6.

The histograms represent analysed mRNA relative levels mean of 4 independent replicates (colchicine treatment, A, C and E; doxorubicin treatment, B, D and F) (*p < 0.05 vs DMSO; **p < 0.01 vs DMSO).

Figure 8. Colchicine and doxorubicin effects on HSPB8 expression are TFEB-independent (A–F) Analysis of the TFEB silencing on HSPB8 expression performed on SH-SY5Y cells transfected with an ON-TARGETplus TFEB siRNA or scramble siRNA, and treated with 500 nM of colchicine or doxorubicin, and with DMSO for 24 hrs. (A) Q-PCR analyses of TFEB mRNA levels.

TFEB mRNA relative levels were normalized on the levels of RplP0 and expressed as mean of 4 independent replicates (**p #p < 0.05 and ##p < 0.01 vs treatment-corresponding scramble samples). (B) The histogram represents the quantification of the TFEB silencing (expressed as percentage). (C) Q-PCR analyses of HSPB8 mRNA levels (**p < 0.01 vs scramble/DMSO; °°p < 0.01 vs siRNA-TFEB/DMSO). (D) WB and WB quantification of TFEB (E) and HSPB8 (F) protein levels. The histograms represent densitometric analysis of three independent replicates for each sample (*p < 0.05 and **p < 0.01 vs scramble/DMSO; °p < 0.05 and °°p < 0.01 vs siRNA-TFEB/DMSO, ##p < 0.01 vs treatment-corresponding scramble samples).

Figure 9. Colchicine and doxorubicin effects on TDP-43 and TDP-25 aggregation (A) Representative immunofluorescence microscopy analysis (63x magnification and 2.5x zoom) performed in SH-SY5Y cells transiently transfected for 24 hrs with GFP-TDP-43 or GFP-TDP-25 plasmids and then treated for 24 hrs with 1 μM colchicine, doxorubicin or DMSO.

Left panels: GFP-TDP-43; right panels: GFP-TDP-25. DAPI: nuclei staining. RED: HSPB8. (B) Quantification of GFP-TDP-43 and GFP-TDP-25 positive cells with GFP-spots using Operetta Imaging System and Harmony software analysis (Perkin Elmer) on living SH-SY5Y cells transiently transfected and treated as described in A. Spots are defined with a range of size from 1 to 10 μm2 in SH-SY5Y treated as in A. SD refers to four independent replicates in HCS. (*p < 0.05 and **p < 0.01 vs untreated control). (C) Quantification of spots area in GFP-TDP-43 and GFP-TDP-25 positive SH-SY5Y cells after drugs treatment obtained by Harmony software analysis (Perkin Elmer).(n = 4). (*p < 0.05 and **p < 0.01 vs untreated control). (D,E) Representative FRA and FRA quantifications of the PBS extracts of SH-SY5Y cells transiently overexpressing the GFP-TDP-43 or GFP-TDP-25 plasmids treated or not with 500 nM of colchicine (left insets) or doxorubicin (right insets) the day after transfection, for the last 24 hrs prior to extraction. (D) The histograms represent densitometric analysis of three independent replicates for each sample (**p < 0.01 vs untreated GFP-TDP-43; °p < 0.05 vs untreated GFP-TDP-25). (E) The histogram represents the quantification of the GFP-TDPs insoluble species reduction percentage after treatment with colchicine or doxorubicin.