ROCK-phosphorylated vimentin modifies mutant huntingtin aggregation via sequestration of IRBIT

Abstract

Background: Huntington's Disease (HD) is a fatal hereditary neurodegenerative disease caused by the accumulation of mutant huntingtin protein (Htt) containing an expanded polyglutamine (polyQ) tract. Activation of the channel responsible for the inositol-induced Ca²⁺ release from ensoplasmic reticulum (ER), was found to contribute substantially to neurodegeneration in HD. Importantly, chemical and genetic inhibition of inositol 1,4,5-trisphosphate (IP3) receptor type 1 (IP3R1) has been shown to reduce mutant Htt aggregation.

Results: In this study, we propose a novel regulatory mechanism of IP3R1 activity by type III intermediate filament vimentin which sequesters the negative regulator of IP3R1, IRBIT, into perinuclear inclusions, and reduces its interaction with IP3R1 resulting in promotion of mutant Htt aggregation. Proteasome inhibitor MG132, which causes polyQ proteins accumulation and aggregation, enhanced the sequestration of IRBIT. Furthermore we found that IRBIT sequestration can be prevented by a rho kinase inhibitor, Y-27632.

Conclusions: Our results suggest that vimentin represents a novel and additional target for the therapy of polyQ diseases.

Figure 1

Vimentin modifies mutant Htt aggregation. A. Representative confocal images show distribution of normal (16Q) and pathogenic (60Q and 150Q) tNHtt (green) and RFP or RFP-vimentin (red) in inducible tNHtt-polyQ-EGFP Neuro2a cells. Note the cages formed by vimentin in 60Q and 150Q Neuro2a cells. Nuclei were stained with DAPI (blue). Scale bar, 5 μm. B. RFP-vimentin expression increased and vimentin knock-down reduced polyQ aggregation and levels of total mutant Htt in 150Q Neuro2a cells as compared to the control. C. The effect of RFP-vimentin on Htt levels is polyQ length-dependent. While tNHtt-60Q-EGFP and tNHtt-150Q-EGFP accumulated as the insoluble forms at the gel top, tNHtt-16Q-EGFP levels remained unchanged upon RFP-vimentin transfection.

Figure 2

Vimentin affects the mutant Htt inclusion formation in 150Q Neuro2a cells and mediates the effect of Y-27632. A. Immunoblot demonstrating inhibition of vimentin phosphorylation at Ser71 and Ser38 by ROCK inhibitor Y-27632 (20 μM) in Neuro2a cells. B. tNHtt-60Q-EGFP (green) was transfected to Neuro2a cells stably expressing RFP-vimentin (red). Treatment of these cells with 20 μM Y-27632 resulted in filament-like distribution of vimentin and disruption of vimentin cages observed around tNHtt-60Q-EGFP inclusions in the untreated cells. Nuclei were stained with DAPI (blue). Scale bar, 15 μm. C. The effect of Y-286432 on polyQ inclusion formation depends on vimentin level (high vimentin levels enhance inclusion formation). 150Q Neuro2a cells were transfected with vimentin shRNA and 48 hrs later, the cells were induced and treated with 20 μM Y-27632. After 24 hrs, cells were fixed and the inclusion formation was quantified by ArrayScan. *p = 0.00003, ** p = 0.0012, ***p = 0.00003, ****p = 0.0018. D. The effect of vimentin knock-down and Y-27632 on polyQ cytotoxicity. Cells were incubated with PI and the PI-positivity was quantified by ArrayScan. *p = 0.0022, ** p = 0.002, ***p = 0.001, ****p = 0.017. E. The effect of Y-286432 on polyQ inclusion formation is dependent on vimentin phosphorylation (vimentin phosphorylation enhances inclusion formation). 150Q Neuro2a cells were transfected with RFP, WT and phospho-mutant (A2 and E2) forms of RFP-vimentin. Cells were induced and treated with 20 μM Y-27632 for 24 hrs, fixed and analyzed by ArrayScan. *p = 0.00016, **p = 0.0003, ***p = 0.0004, ^ˆp = 0.0021, ^ˆˆp = 0.018, ^#p = 0.04, ^##p = 0.022. F. The effect of vimentin over-expression and Y-27632 on polyQ-induced cell death. Dead cells were detached and removed during samples preparation for analysis. Cells that remained attached were counted by ArrayScan. *p = 0.049, **p = 0.008, ^ˆp = 0.044, ^ˆˆp = 0.045, ^#p = 0.032, ^##p = 0.034. Bars in C-F represent relative mean values ± s.d. from three independent experiments, with levels under control conditions normalized to a value of 1.

Figure 3

Vimentin influences the IRBIT-IP3R1 interaction by sequestering IRBIT in perinuclear inclusions. A. HeLa cells were transfected with RFP or tested forms of RFP-vimentin. 24 hrs later, cells were lysed and immunopreciptitation was performed using 10A6 anti-IP3R1 antibody. For immunoblotting, KM1112 anti-IP3R1, anti-IRBIT and anti-RFP antibodies were used. B. Quantification of IRBIT-IP3R1 interaction. The densities of immunoprecipitated IRBIT were normalized to the densities of corresponding immunoprecipitated IP3R1. *p = 0.0002, **p = 0.009, ***p = 0.0001, ^#p = 0.0017, ^##p = 0.029, ^###p = 0.0009. Bars represent relative mean values ± s.d. from three independent experiments, with levels of IRBIT-IP3R1 interaction under control conditions normalized to a value of 1. C. Representative confocal images show distribution of immobile fraction of IRBIT in presence of RFP or tested RFP-vimentin forms. Neuro2a cells were transfected with RFP or RFP-vimentins, 24 hrs later permeabilized with saponin and immunostained for endogenous IRBIT (green). Fluorescence of RFP (red) and DAPI (blue) are also shown. Note that RFP was washed out from cells during saponin permeabilization. Scale bar, 5 μm.

Figure 4

Vimentin phosphorylation affects IRBIT subcellular distribution. Effect of ROCK and UPS inhibitors. Representative confocal images show subcellular distribution of IRBIT in presence of RFP or tested RFP-vimentin forms. Neuro2a cells were transfected and treated as indicated for 24 hrs. Cells were fixed, permeabilized and immunostained for enogenous IRBIT (green). A. Control untreated cells (DMSO, 1/1000). B. Cells treated with 20 μM Y-27632. C. Cells treated with 5 μM MG132. Fluorescence of RFP (red) and DAPI (blue) are also shown. Scale bar, 5 μm.

Figure 5

IRBIT modifies inclusion formation in 150Q Neuro2a cells. A. 48 hrs after transfection of control or IRBIT siRNA and RFP or tested RFP-vimentin forms, cells were induced for 24 hrs, fixed, and analyzed by ArrayScan. *p = 0.028, **p = 0.0008, ***p = 0.00005, ****p = 0.00004, ^ˆp = 0.0018, ^ˆˆp = 0.0021, ^ˆˆˆp = 0.00034, ^#p = 0.0007, ^##p = 0.0004, ^###p = 0.00002, ^#167;p = 0.016, ^#167;#167;p = 0.0001. B. Images generated and analyzed by ArrayScan representing each experimental condition in A. Green, tNHtt-150Q-EGFP; blue, Hoechst 33258. Magnification, 20x. C. Cells were co-transfected with mock or flag-IRBIT and RFP or tested RFP-vimentin forms and induced for 24 hrs before analyzed by ArrayScan. *p = 0.042, **p = 0.0002, ***p = 0.049, ^ˆp = 0.0002, ^ˆˆp = 0.0099, ^ˆˆˆp = 0.00013, ^#p = 0.00005, ^##p = 0.015, ^###p = 0.00009, ^####p = 0.0008, ^#####p = 0.0005, ^######p = 0.0092, ^#167;p = 0.0026, ^#167;#167;p = 0.0016. D. Images generated and analyzed by ArrayScan representing each experimental condition in C. Green, tNHtt-150Q-EGFP; blue, Hoechst 33258. Magnification, 20x. Bars in A and C represent relative mean values ± s.d. from four independent experiments, with levels of inclusion formation under control conditions normalized to a value of 1.

Figure 6

Schematic representation of the proposed mechanism of the vimentin effect on mutant Htt aggregation. A. In HD, mutant Htt and dopamine stimulation may activate ROCK, which in turn phosphorylates vimentin at Ser71 and Ser38 leading to IRBIT sequestration, reduced interaction of IRBIT with IP3R1, and activation of IP3R1. Mutant Htt expression and aggregation may increase aggresome formation enhancing IRBIT sequestration. B. Blocking vimentin phosphorylation by ROCK inhibitors may lead to reduced IRBIT sequestration by vimentin and consequently to decreased accumulation and aggregation of the pathogenic polyQ protein.