Cdk5 phosphorylation of huntingtin reduces its cleavage by caspases: implications for mutant huntingtin toxicity

Abstract

Huntington's disease (HD) is a neurodegenerative disorder caused by an expanded polyglutamine (polyQ) tract in the huntingtin (htt) protein. Mutant htt toxicity is exposed after htt cleavage by caspases and other proteases release NH(2)-terminal fragments containing the polyQ expansion. Here, we show htt interacts and colocalizes with cdk5 in cellular membrane fractions. Cdk5 phosphorylates htt at Ser434, and this phosphorylation reduces caspase-mediated htt cleavage at residue 513. Reduced mutant htt cleavage resulting from cdk5 phosphorylation attenuated aggregate formation and toxicity in cells expressing the NH(2)-terminal 588 amino acids (htt588) of mutant htt. Cdk5 activity is reduced in the brains of HD transgenic mice compared with controls. This result can be accounted for by the polyQ-expanded htt fragments reducing the interaction between cdk5 and its activator p35. These data predict that the ability of cdk5 phosphorylation to protect against htt cleavage, aggregation, and toxicity is compromised in cells expressing toxic fragments of htt.

Figure 1.

Htt interacts with Cdk5. (A) COS-7 cells were cotransfected with empty vector and cdk5 (lanes 1–3) or Flag-tagged htt1-588 (htt588) and cdk5 (lane 4). In all experiments, empty vector controls have been performed where appropriate and the amount of empty vector DNA was equivalent to the amount of experimental DNA (in this case, htt588). Cell lysates were immunoprecipitated with anti-Flag. Anti-Flag immunoprecipitates were subjected to immunoblotting with anti-cdk5 (top) or anti-Flag (middle). Total lysates were immunoblotted with anti-cdk5 (bottom). (B) COS-7 cells were cotransfected with empty vector and cdk5 (lane 1); Flag-tagged, wild-type, full-length htt (wtFLhtt) and cdk5 (lane 2); Flag-tagged, wild-type htt588 (httwt588) and cdk5 (lane 3); Flag-tagged, mutant, full-length htt (muFLhtt) and cdk5 (lane 4); and Flag-tagged, mutant htt588 (httmu588) and cdk5 (lane 5), respectively. Cell lysates were immunoprecipitated with anti-Flag and immunoblotted with anti-cdk5 (top) and anti-Flag (middle), respectively. Total lysates were immunoblotted with anti-cdk5 (bottom). (C) Bacterially expressed GST (lane 2) and GST-htt (5–56 aa; lane 3) were incubated with cdk5-transfected COS-7 cell lysate. The cdk5 that associated with GST-htt (5–56 aa) was detected with anti-cdk5. The input signal represents 5% cdk5 in the total cell lysate (lane 1). (D, i) Mouse brain lysate was immunoprecipitated with anti-Flag as a control (NS; lane 1) or anti-htt (181–800; lane 2). Cdk5 associated with htt as shown with anti-cdk5 antibody probing of Western blot of immunoprecipitate (top). (ii) Mouse brain lysate was immunoprecipitated with anti-Myc as a control (NS; lane 1) or anti-cdk5 (C8; lane 2). Htt associated with cdk5 as shown with anti-htt antibody probing of Western blot of immunoprecipitate (top).

Figure 2.

Htt associates with cdk5 at LM. (A) Cdk5 was cotransfected with empty vector/cdk5 (2:1) or Flag-htt1-551 (htt551)/cdk5 (2:1) in COS-7 cells. After 24 h, transfected cells were harvested. Light membranes (LM) (including endosomes and all ER vesicles) were isolated. Total cellular (lanes 1 and 2) and LM (lanes 3 and 4) lysates were resolved by SDS-PAGE and transferred to PVDF membrane, and then probed with anti-cdk5 (top), anti-Flag (middle), and anti-actin (bottom). The blots were quantified with ChemiImager. The ratios of cdk5/actin in total lysates are set as 1. The relative values of cdk5/actin in LM are shown. Three independent experiments were performed. Error bars are SD; *, P = 0.01. (B) Mouse brain lysate was immunodepleted with anti-Flag for control (lane 1) or anti-htt (lane 2). The lysate was separated as cytosolic and LM fractions. Cytosolic and LM fractions were resolved by 10% SDS-PAGE and transferred to PVDF membranes. Anti-cdk5 (top), anti-htt (middle), and anti-actin (bottom) were probed for cdk5, htt, and actin, respectively. (C) PC-12 cells were induced to differentiate with 100 ng/ml NGF for 3 d. Confocal immunofluorescence of PC-12 cells: green, anti-htt, Alexa 488; red, anti-cdk5, Alexa 594; blue, nuclei.

Figure 3.

Cdk5 phosphorylates htt in vitro and in vivo. (A) GST and GST-tagged htt1-588 (GST-htt588) (wild-type) were purified from E. coli. Both proteins were phosphorylated by 0.1 μg of p35–cdk5 complexes. Top panel shows phosphorylated GST (lane 1) and GST-htt588 (lane 2). Bottom panel shows purified GST (lane 1) and GST-htt588 (lane 2). (B) p35–cdk5 was cotransfected to COS-7 cells. We immunoprecipitated p35–cdk5 with anti-cdk5. Httwt588 or httmu588 were pulled down with anti-Flag from distinct COS-7 cells transfected with these constructs. The figure shows in vitro kinase assays (top) and anti-Flag blot (bottom) from p35–cdk5 incubated with httwt588 and γ-[³²P]ATP (lane 1) and p35–cdk5 incubated with httmu588 and γ-[³²P]ATP (lane 2). The mixtures were resolved with 10% SDS-PAGE, and then transferred to PVDF membrane and subjected to autoradiography (top). The PVDF membrane was blotted with anti-Flag (bottom). (C) PC-12 cells were starved for 24 h, and then induced to differentiate with 100 ng/ml NGF for 48 h. Cells were treated with 20 μM of the cdk5 inhibitor roscovitine (Rosco) or DMSO (control) when cells were induced to differentiate with NGF. After 48 h of treatment, PC-12 cells were lysed in buffer A containing phosphatase inhibitors. Htt was immunoprecipitated with anti-htt antibody, 2166, and then detected with the antiphosphoserine antibody 16B4 (top) and anti-htt (middle). p35–cdk5 complex was pulled down from the differentiated PC-12 cells and an in vitro kinase assay was performed in the presence of DMSO (lane 1) or roscovitine (lane 2) using histone H1 as a substrate (bottom).

Figure 4.

Cdk5 phosphorylates htt at Ser434 in vitro and in vivo. (A, i) htt1-551 (htt551), htt1-415 (htt415), and htt1-314 (htt314) were immunoprecipitated with anti-Flag from COS-7 cells, and then γ-[³²P]ATP and 0.1 μg of recombinant p35–cdk5 complexes were added to htt551 (lane 1), htt415 (lane 2), and htt314 (lane 3) to phosphorylate these htt variants. The mixtures were subjected to SDS-PAGE and transferred to PVDF membrane, subjected to autoradiography (top), and blotted with anti-Flag (bottom). (ii) htt1-588 (htt588; lane 1) and htt588 S434A mutant (htt588 S434A; lane 2) were pulled down, phosphorylated (top), and blotted with anti-Flag (bottom) as above. Note that only htt forms with S434 (lane 1) were efficiently phosphorylated. (B) Htt551/empty vector (lane 1), htt551/p35 (lane 2 and 4), and htt551 S434A/p35 (lane 3) were cotransfected into cdk5-expressing HeLa cells. After 24 h, cells were labeled with [³²P]orthophosphate for 3 h, and one of the htt551/p35 transfections (lane 4) was treated with 20 μM of the cdk5 inhibitor roscovitine (Rosco) during this process. Htt551 was pulled down with anti-Flag and subjected to SDS-PAGE and transferred to PVDF membrane for autoradiography (top) and immunoblotting with anti-Flag. (C) An alignment of htt from mini-pig, human, mouse, rat, fugu, and zebrafish. The conserved phosphorylation-determining amino acid residues are highlighted. (D) Purified GST or GST-htt588 were phosphorylated by p35–cdk5 complex in vitro as described in Materials and methods. 1/100 of the mixtures were subjected to SDS-PAGE and Western blot using phospho-htt antibody, pS434 (top), and anti-GST. (E) Htt551/empty vector (lane 1), htt551/p35 (lane 2), htt551 S434A/empty vector (lane 3), and htt551 S434A/p35 (lane 4) were transfected into HeLa cells. After 24 h, cells were harvested, lysates were subjected to SDS-PAGE, transferred to PVDF membrane, and then probed with anti-pS434 (top). The same membrane was probed with anti-Flag (bottom) after stripping. (F) Htt551-138Q (muhtt) was cotransfected with empty vector (lane 1) or p35 (lane 2). After 24 h, cell lysates were subjected to Western blot and probed with anti-pS434 (top) and anti-Flag (bottom). (G) 5 μg of empty vector, cdk5, and cdk5DN were cotransfected to PC-12 cells in 10-cm dishes with 1 μg EGFP. The transfected cells were starved for 24 h and treated with NGF for 4 h. FACS was used to sort GFP-positive (i.e., transfected) cells. Those GFP-positive cell lysates were subjected to Western blot probed with anti-pS434 (top) and anti-htt (bottom), respectively. Newborn mouse brain lysate (lane 5) was probed with anti-pS434 (top) and anti-htt (bottom).

Figure 5.

Cdk5-phosphorylating htt blocks caspase cleavage and regulates mutant htt toxicity. (A) Httmu588/vector, httmu588/p35, httmu588 S434A/vector, and httmu588 S434A/p35 were transfected to cdk5-expressing neuroblastoma SK-N-SH cells. After 48 h, cells were fixed and immunostained with anti-Flag and p35 antibodies. Htt-expressing cells were scored for the presence of aggregates and abnormal nuclei. Data are from three independent experiments. Each experiment was performed in triplicate. Error bars represent SD. ***, P < 0.0001; **, P < 0.001. (B) p35 was cotransfected with htt551 (lanes 1 and 2) or htt551 S434A (lanes 3 and 4), or empty vector was cotransfected with htt551 (lanes 5 and 6) or htt551 S434A (lanes 7 and 8) into HeLa cells. After 24 h, cells were treated with 1μM staurosporine (STS) for 0 (lanes 1, 3, 5, and 7) or 6 h (lanes 2, 4, 6, and 8). Cell lysates were resolved with 10% SDS-PAGE and transferred to PVDF membrane for Western blotting. The variation in the amount of htt expressed in this transient transfection experiment can vary from well to well depending on the transfection efficiency in the specific well (as well as survival when treated with STS). We are primarily concerned with the proportion of the protein that is cleaved. Quantitation of the intact/cleaved(513) htt ratios from of the scanned original gel is shown. The phenomena shown in this figure were reproduced in independent experiments. (C) p35 was cotransfected with muhtt551 (lanes 1 and 2), or empty vector was cotransfected with htt551 (lanes 3 and 4) into cdk5-expressing HeLa cells. After 24 h, cells were treated with 1 μM staurosporine (STS) for 0 (lanes 1 and 3) or 6 h (lanes 2 and 4). Cell lysates were resolved with 10% SDS-PAGE and transferred to PVDF membrane for Western blot. (D) Htt551 and htt551 S434A were in vitro translated with ³⁵S-labeling. The in vitro–translated htt551 (lane 3) and htt551 S434A (lane 6) were phosphorylated in vitro in the presence of recombinant p35–cdk5 complex and ATP-γ-S. The unphosphorylated (lane 2) and phosphorylated htt551 (lane 3) and unphosphorylated (lane 5) and phosphorylated htt551 S434A (lane 6) were subjected to caspase-3 cleavage (100 ng/ml). 10% SDS-PAGE was performed. Data show a representative of three independent experiments.

Figure 6.

Inhibition of cdk5 activity promotes htt cleavage, and mutant htt aggregate formation in differentiated htt stable PC-12 cells. (A) Wild-type, full-length htt PC-12 cells were differentiated with 100 ng/ml NGF for 48 h. 1 μg/ml doxycycline was added to induce htt expression for 5 d. At same time, indicated samples in duplicate were also treated with 20 μM of the cdk5 inhibitor roscovitine. Cells were harvested and subjected to 7.5% SDS-PAGE. Anti-Flag (M2) was used for Western blot. This antibody detects a 3× Flag tag at the NH₂-terminal of the htt transgene. (B) Mutant full-length htt PC-12 cells were differentiated with 100 ng/ml NGF for 48 h. 1 μg/ml doxycycline was added to induce htt expression for 5 d. At the same time as inducing transgene expression, cells were also treated with 20 μM of the cdk5 inhibitor roscovitine, where indicated. Cells were fixed and anti-Flag (M2) was used for immunocytochemistry. This antibody detects a 3× FLAG tag at the NH₂-terminal of the htt transgene. (i) The proportions of cells with aggregates are shown. Experiment was performed in triplicate. Error bars represent SD. *, P < 0.01. (ii) Mutant, full-length htt PC-12 cells were differentiated with NGF, induced with doxycycline, and treated with roscovitine as wild-type, full-length htt PC-12 cells. Cells were harvested and subjected to 10% SDS-PAGE. Anti-Flag (M2) was used for Western blot. This antibody detects a 3× Flag tag at the NH₂-terminal of the htt transgene.

Figure 7.

Mutant htt impairs cdk5 activity by interfering p35–cdk5 interaction. (A) Cdk5 was pulled down from the lysates of wild-type littermate control (Ctrl; lane 1) or HD (lane 2) mice whole brains (12 wk). Histone H1 was used as a substrate for kinase assays. Radiophotographs and Western blots were quantified. Six independent experiments were performed. Error bars represent SD. *, P < 0.05. Note that kinase activity is expressed as a function of cdk5 levels (specific activity). (B) Whole brain lysates of 13-wk wild-type littermate control (Ctrl; lane 1) and HD (lane 2) mice were subjected to SDS-PAGE, and Western blots were probed successively with anti-p35 (top), -cdk5 (middle), and -tubulin (bottom panel) antibodies. (C) Anti-cdk5 (J3, monoclonal) was used to pull down p35 from 13-wk-old wild-type littermate control (lane 1) and HD (lane 2) mouse whole brain lysates. Anti-p35 (C-19, polyclonal) was used to detect p35 (top). After stripping, the same membrane was probed with anti-cdk5 (C8, polyclonal) to measure cdk5 levels (second panel). P35 (third panel) and cdk5 (bottom) levels in total control (lane 1) and HD (lane 2) mouse brain lysates are shown. Data show a representative experiment from three independent experiments. (D) Anti-cdk5 (J3, monoclonal) was used to pull down p35 from p35/cdk5/GFP-httEx1-23Q- (1:1:2) (lane 1) and p35/cdk5/GFP-httEx1-74Q- (1:1:2) (lane 3) transfected COS-7 cell lysates. Anti-p35 (C-19, polyclonal) was used to detect p35 (top). After stripping, the same membrane was probed with anti-cdk5 (C8, polyclonal) to measure cdk5 levels (second panel). P35 (third) and cdk5 (bottom) levels in total p35/cdk5/GFP-httEx1-23Q- (lane 1) and p35/cdk5/GFP-httEx1-74Q- (lane 3) transfected COS-7 cell lysates are shown. Data show a representative experiment from four independent experiments. (E) p35–cdk5 (0.75 μg each) were cotransfected with 0 (lane 1), 0.5 (lane 2), 1 (lane 3), 2 (lane 4), and 3 μg httEx1-74Q (lane 5), respectively. Note that the total amount of DNA transfected was kept constant by adding empty vector DNA, where necessary. After 48 h, anti-cdk5 was used to immunoprecipitate p35 in each transfected cell. IP products (left) and total cell lysates (right) were detected with anti-GFP, anti-p35, and anti-cdk5, respectively. (F, left) 8-wk-old HD (lane 2) and nontransgenic (lane 1) mouse whole brain lysates were subjected Western blot and probed with anti-pSer434 (top) and anti-htt (bottom). One of the phtt/htt ratios was set as 1 and data were quantified. Error bars represent SD. *, P < 0.05. (right) 17-wk-old HD (lanes 3 and 4) and nontransgenic (lanes 1 and 2) mouse whole brain lysates (40 μg) were subjected to Western blot and probed with anti-pS434 (top) and anti-htt (bottom). One of the phtt/htt ratios was set as 1. Western blot was quantified. Error bars represent SD. *, P < 0.05.

Figure 8.

Htt588 or full-length htt stabilizes p35–cdk5 interaction, and expanded polyQ of mutant htt compromises this stablization. (A) p35–cdk5 (0.75 μg each) were cotransfected into HeLa cells with 1 μg of empty vector (lane 1), 0.5 μg httwt588 (+0.5 μg of empty vector, to ensure similar amounts of DNA transfected in all lanes) (lane 2), 1 μg httwt588 (lane 3), 0.5 μg httmu588 (+0.5 μg empty vector) (lane 4), and 1 μg httmu588 (lane 5). After 24 h, cell lysates were subjected to IP with mouse anti-cdk5 (J3). The IP products were probed with anti-p35, anti-cdk5, and anti-Flag (for htt), respectively. Data represent a representative of experiments repeated four times with similar trends. (B) p35–cdk5 (0.75 μg each) were cotransfected with empty vector (lane 1), httwt588 (lane 2), httmu588 (lane 3), wtFLhtt (lane 4), muwtFLhtt (lane 5), httEx1-23Q (lane 6), and httEx1-74Q (lane 7) into HeLa cells; the ratio of p35/cdk5/htt is 1:1:2. The cell lysates were immunoprecipitated with anti-cdk5. The IP products were then probed with anti-p35, anti-cdk5, anti-Flag (for htt588 and full-length htt), and anti-GFP (for GFP-httEx1). Similar results were observed in another two independent experiments.