Expression of protein phosphatase 2A mutants and silencing of the ...

8 downloads 159338 Views 1MB Size Report
Captured images were transferred to Adobe. Photoshop and Illustrator CS for printing (Adobe Systems. Incorporated, San Jose, CA, USA). Table 1 Comparative ...
Journal of Neurochemistry, 2007, 101, 959–971

doi:10.1111/j.1471-4159.2007.04503.x

Expression of protein phosphatase 2A mutants and silencing of the regulatory Ba subunit induce a selective loss of acetylated and detyrosinated microtubules Viyada Nunbhakdi-Craig,* Stefan Schuechner,  Jean-Marie Sontag,* Lisa Montgomery,* David C. Pallas,à Claudia Juno,  Ingrid Mudrak,  Egon Ogris  and Estelle Sontag* *Department of Pathology, University of Texas Southwestern Medical Center, Dallas, Texas, USA  Department of Medical Biochemistry, Max F. Perutz Laboratories, Medical University of Vienna, Austria àDepartment of Biochemistry and Winship Cancer Institute, Emory University School of Medicine, Atlanta, Georgia, USA

Abstract Carboxymethylation and phosphorylation of protein phosphatase 2A (PP2A) catalytic C subunit are evolutionary conserved mechanisms that critically control PP2A holoenzyme assembly and substrate specificity. Down-regulation of PP2A methylation and PP2A enzymes containing the Ba regulatory subunit occur in Alzheimer’s disease. In this study, we show that expressed wild-type and methylation- (L309D) and phosphorylation- (T304D, T304A, Y307F, and Y307E) site mutants of PP2A C subunit differentially bind to B, B¢, and B¢¢-type regulatory subunits in NIH 3T3 fibroblasts and neuro2a (N2a) neuroblastoma cells. They also display distinct binding affinity for microtubules (MTs). Relative to controls, expression of the wild-type, T304A and Y307F C subunits in N2a cells promotes the accumulation of acetylated and

detyrosinated MTs. However, expression of the Y307E, L309D, and T304D mutants, which are impaired in their ability to associate with the Ba subunit, induces their loss. Silencing of Ba subunit in N2a and NIH 3T3 cells is sufficient to induce a similar breakdown of acetylated and detyrosinated MTs. It also confers increased sensitivity to nocodazole-induced MT depolymerization. Our findings suggest that changes in intracellular PP2A subunit composition can modulate MT dynamics. They support the hypothesis that reduced amounts of neuronal Ba-containing PP2A heterotrimers contribute to MT destabilization in Alzheimer’s disease. Keywords: cytoskeleton, methylation, microtubule, protein phosphatase 2A, silencing. J. Neurochem. (2007) 101, 959–971.

The microtubule (MT) cytoskeleton plays essential roles in cell polarity, differentiation, and protein trafficking. MT destabilization and disruption of MT-based transport are associated with a large number of diseases, including neurodegenerative disorders like Alzheimer’s disease (AD) (reviewed in Mandelkow et al. 2003). MTs are highly dynamic structures (reviewed in Desai and Mitchison 1997), whose stability is regulated by the complex rearrangement of the constituent ab-tubulin heterodimers and binding of microtubule-associated proteins (MAPs). MTs are subject to many regulatory post-translational modifications, including tyrosination and detyrosination (Gundersen et al. 1987), acetylation (Black and Keyser 1987), and phosphorylation (Piras and Piras 1975), which regulate their functions (reviewed in Westermann and Weber 2003). For instance, MT acetylation controls motor-protein trafficking (Reed et al. 2006), while its detyrosination plays an important role

in cellular morphogenesis (Bulinski and Gundersen 1991). In cells, a tubulin tyrosine ligase maintains the pool of unassembled a-tubulin in the tyrosinated state (‘Tyr-tubulin’) (Webster et al. 1987; Barra et al. 1988). Conversely, a tubulin tyrosine carboxypeptidase preferentially detyrosinates a-tubulin assembled into MTs (Kumar and Flavin

Received August 17, 2006; revised manuscript received November 17, 2006; accepted November 18, 2006. Address correspondence and reprint requests to Estelle Sontag PhD, Department of Pathology, University of Texas Southwestern Medical Center, 5323 Harry Hines Blvd, Dallas, TX 75390-9073, USA. E-mail: [email protected] Abbreviations used: AD, Alzheimer’s disease; Glu-tubulin, detyrosinated tubulin; MAPs, microtubule-associated proteins; MT, microtubule; N2a, neuro-2a; OA, okadaic acid; PP2A, protein phosphatase 2A; PP2A-C, PP2A catalytic subunit; small t, small tumor antigen; Tyrtubulin, tyrosinated tubulin; WT, wild-type.

 2007 The Authors Journal Compilation  2007 International Society for Neurochemistry, J. Neurochem. (2007) 101, 959–971

959

960 V. Nunbhakdi-Craig et al.

1981). Although detyrosination does not directly stabilize MTs, detyrosinated a-tubulin, also called ‘Glu-tubulin,’ accumulates in stabilized MTs (Gundersen et al. 1987). The Ser/Thr-specific protein phosphatase 2A (PP2A) binds to and regulates MTs in an isoform-dependent manner (Sontag et al. 1995, 1996, 1999; Merrick et al. 1997; Hiraga and Tamura 2000). The mammalian PP2A holoenzyme is a multiprotein complex consisting of a catalytic (C), a structural (A), and one of a number of variable regulatory (B) subunits (Sontag 2001). To date, several cellular families of B-type subunits have been identified: B/B55/PR55, B¢/B56/PR61, B¢¢/PR59/PR72/PR130, and striatin/SG2NA (putative B¢¢¢). Each class is encoded by multiple genes, resulting in a large number of isoforms. PP2A functional specificity is tightly regulated in vivo. Regulatory B subunits directly modulate PP2A catalytic activity and substrate specificity. They contribute to direct the phosphatase to distinct intracellular multi-protein complexes and microenvironments, thereby allowing exquisite control of the enzyme–substrate relationship. In addition, various cellular or viral proteins can associate with and regulate PP2A. For instance, simian virus 40 small tumor antigen (small t) forms a complex with the (AC) core enzyme of PP2A, resulting in changes in PP2A substrate specificity (reviewed in Sontag 2001). Furthermore, the 301TRRTPDYFL309 sequence of PP2A catalytic subunit (PP2A-C) can undergo reversible post-translational modifications, including carboxymethylation on Leu309 and phosphorylation on Tyr307 and putative Thr304 residues. Phosphorylation is transient and inactivates PP2A (Brautigan 1995). Studies in yeast and fibroblasts have demonstrated that PP2A-C methylation and phosphorylation control PP2A holoenzyme assembly, which in turn affects PP2A substrate specificity and functions (Ogris et al. 1997; Bryant et al. 1999; Evans and Hemmings 2000; Tolstykh et al. 2000; Wei et al. 2001; Yu et al. 2001; Gentry et al. 2005). Of particular interest in these studies is the observation that PP2A-C methylation is required for formation of Ba-containing heterotrimers. We have reported that both neuronal PP2A methylation and Ba expression levels are down-regulated in AD-affected brain regions (Sontag et al. 2004a,b). Yet, the functional significance of these alterations for specific cellular functions remains to be clarified. In this study, we show in cultured cell models that expression of PP2A-C methylationand phosphorylation-site mutants and Ba subunit knockdown affect the stability of selective MT populations.

Material and methods Antibodies Antibodies used in this study comprised: Anti-HA ‘16B12’ (Covance Research Products, Berkeley, CA, USA); anti-small t ‘Pab108’ (Nunbhakdi-Craig et al. 2003); mouse anti-a-actin, anti-a, -b, -Tyr, and -acetylated tubulin (Sigma, St Louis, MO, USA); rabbit

anti-Glu-tubulin (Gundersen et al. 1984); goat anti-A (Santa Cruz Biotechnology, Santa Cruz, CA, USA); mouse anti-C (BD Biosciences-Pharmingen, San Jose, CA, USA); anti-B55 ‘2G9’ (Nunbhakdi-Craig et al. 2002). The pan anti-B¢ antiserum was raised against a peptide sequence in the human B¢a1 subunit that is conserved in six B¢ isoforms, but not in B subunits (Zhu et al. 1997). The rabbit anti-B¢/B56a antibody was raised against residues 1–278 of human B56a (McCright and Virshup 1995) and affinitypurified. It primarily recognized B56a and slightly cross-reacted with B56c, B56d, and B56e, but not B56b. The rabbit anti-B¢/PR59 antibody was raised against mouse PR59 lacking the N-terminal 113 residues (Voorhoeve et al. 1999) and specifically recognized the 59 kD protein. Cell culture, infection, and transfection NIH 3T3 and Neuro-2a (N2a) cells were maintained in Dulbecco’s Modified Eagle’s Medium (Invitrogen, Carlsbad, CA, USA) containing 10% fetal bovine serum (Hyclone, Logan, UT, USA). All the experiments were performed in growing cells incubated in regular medium to prevent differentiation-induced changes in MT stability and protein expression. Cells were transfected using metafectene following the manufacturer’s instructions (Biontex, Germany). When indicated, NIH 3T3 cells were infected as described previously (Fellner et al. 2003b). ‘Control’ cells were mock-transfected or -infected. Stable N2a clones were generated after transfection with the following plasmids: pcDNA3.1 (Invitrogen), pcDNA 3.1 encoding HA-tagged WT or mutated PP2A-C (Ogris et al. 1997; Goedert et al. 2000; Yu et al. 2001), and Rc/ CMV-small t (Sontag et al. 1993), followed by selection with 600 lg/mL G418 (Invitrogen). The expression levels of transfected proteins were constantly monitored by both immunoblotting and immunofluorescence. At least four distinct stable clones were used throughout our studies with similar results. B subunit binding analyzes The ability of HA-tagged PP2A-C mutants to bind to various B-type subunits was analyzed by western blotting in HA-immunoprecipitates prepared from infected NIH 3T3 or transfected N2a cells, following the protocol described previously (Fellner et al. 2003a,b). The relative amounts of B-type subunit bound to PP2A-C were determined after densitometry scanning of the immunoblots and normalization for the amounts of immunoprecipitated HA-tagged PP2A-C subunits present in each lane. Cell lysis, extraction, and MT purification Total cell extracts were prepared as described previously (Sontag et al. 1995). Detergent-soluble and -insoluble fractions were obtained after selective cell extraction with a MT-stabilizing buffer containing 0.1% Nonidet P-40 (Sontag et al. 1995). Endogenous taxol-stabilized MTs were purified by centrifugation for 30 min at 20C at 30 000 g, as described previously (Sontag et al. 1995, 1999). The MT pellet was resuspended in the same volume of buffer that had been initially used for cell homogenization. Electrophoresis and western blotting analyzes Protein concentrations were determined using the Bio-RadTM protein assay kit (Bio-Rad Laboratories, Hercules, CA, USA). Equivalent volumes (100 lL) of the MT pellet and post-MT

 2007 The Authors Journal Compilation  2007 International Society for Neurochemistry, J. Neurochem. (2007) 101, 959–971

Regulation of microtubule stability by PP2A 961

supernatant or equivalent amounts of proteins (50 lg) from total cell extracts or detergent-soluble/insoluble fractions were resolved by sodium dodecyl sulfate–polyacrylamide gel electrophoresis using 10% polyacrylamide gels (GE Healthcare-Amersham Biosciences, Pittsburgh, PA, USA). Extended electrophoretic conditions were utilized to separate ectopically expressed from endogenous PP2A-C bands. Proteins were transferred to nitrocellulose membranes, and blotted with the following antibodies: antiHA (1 : 1000), anti-C (1 : 2000), anti-B55 ‘2G9’ (1 : 10 000), anti-B56a (1 : 10 000), anti-B¢ (1 : 5000), pan anti-B¢ (1 : 5000), anti-small t (1 : 20), anti-A (1 : 1000), and all anti-tubulin antibodies (1 : 2000). When indicated, blotting with anti-actin (1 : 2000) was used as a control for protein loading. When appropriate, Ponceau red staining of the blots was used to verify that equivalent amounts of MTs had been pelleted in each condition. Blots were developed using the SuperSignal West Pico chemiluminescence detection system (Pierce Biotechnology Inc., Rockford, IL, USA). They were scanned by densitometry and quantified using both Eastman Kodak Image softwareTM (Eastman Kodak Company, Rochester, NY, USA) and a Molecular Dynamics Densitometer with ImageQuant 4.0 software (Molecular Dynamics, Sunnyvale, CA, USA). Although representative blots are shown in figures, several exposures were assessed for quantification to ensure that analyzed signals were within the linear range. Silencing experiments Ba silencing was achieved by using previously validated target sequences (Adams et al. 2005; Van Kanegan et al. 2005) to create vectors expressing specific shRNAs: shBa1 (5¢-AA TCCAGTCTCATAGCAGAGG-3¢) and shBa2 (5¢-AAGTGGCA AGCGAAAGAAAGA-3¢). Using these sequences as templates, oligonucleotides (sense and antisense) were synthesized (Integrated DNA Technologies, Coralville, IA, USA) following published guidelines (Brummelkamp et al. 2002). Sense and antisense fragments were hybridized, phosphorylated, and cloned

Table 1 Comparative binding of PP2A-C mutants to regulatory B-type subunits

a

into the BamHI–EcoRI sites of a linearized U6 promoter-driven RNAi-Ready pSIREN-DNR-DsRed-Express vector (Clontech, Mountain View, CA, USA). Vectors inserts were verified by sequencing. N2a and NIH3T3 cells were transiently transfected with the silencing vectors using metafectene. After 48 h posttransfection, cells were split and plated at the appropriate density on both 60-mm dishes and glass coverslips and grown for another 24 h in regular medium prior to being processed for western blotting or immunofluorescence studies. Immunofluorescence Cells were grown overnight in regular medium on uncoated glass coverslips to prevent differentiation. When indicated, control cells were treated for 1 h with 100 nmol/L okadaic acid (OA; LC Laboratories, Woburn, MA, USA) prior to fixation. In some experiments, prior to fixation, cells were incubated for 20 min with 5 lmol/L of nocodazole (Sigma). Intact cells were washed quickly at 37C with PEM buffer (0.1 mol/L PIPES, pH 6.9, 2 mmol/L EGTA, and 5 mmol/L MgCl2) prior to fixation. Intact cells or detergent-resistant cytoskeletons were fixed for 5 min with absolute methanol at )20C (Sontag et al. 1995, 1996). However, for silencing experiments, methanol fixation resulted in the loss of DsRed immunofluorescence, and cells were thus fixed for 20 min with 4% p-formaldehyde phosphate-buffered saline, permeabilized for 5 min in phosphate-buffered saline containing 0.1% Triton X-100, then washed and labeled exactly as described previously (Sontag et al. 1995). Cells were stained for 1 h with indicated anti-tubulin antibodies followed by incubation for 1 h with Alexa Fluor 488conjugated goat antibodies (Invitrogen-Molecular Probes, Carlsbad, CA, USA). The samples were mounted with Fluoromount (Fisher Scientific, Pittsburgh, PA, USA) and examined on a Carl Zeiss Inc. microscope (Carl Zeiss Inc., Thornwood, NY, USA) using a 63· objective and appropriate filters for detecting fluorescein isothiocyanate or Ds-Red. Captured images were transferred to Adobe Photoshop and Illustrator CS for printing (Adobe Systems Incorporated, San Jose, CA, USA).

PP2A-C

Methylation competence (%)

WT T304A

100 71–100

T304D

71–100

Y307F

41–70

Y307E