- Email: [email protected]
p97) in untransformed osteoblast-like cells
ELSEVIER
Journal of Orthopaedic Research
Journal of Orthopaedic Research 23 (2005) 618-624
www.elsevier.cod1ocatelorthres
Identification and characterization of valosin-containing protein (VCP/p97) in untransformed osteoblast-like cells Keyvan Behnam
a,
Samuel S. Murray
b3c, Elsa
J. Brochmann
b7c7*
Department .J‘Ph~wicilogiculSciences, Uniuersitv .f Culifi,rniu. Los Angeles. C A 90024. USA Department of Medicine, University of Culiforniu, Los Angeles. CA 90024. USA Geriutric Reseurch, Educution & Clinicul Center ( I I - E ) . VA Greuter Los Angeles Heultkcure System, 16111 Plunimer Srreer, Sepuhedu. C A 91343, USA
Received 1 December 2004; accepted 22 December 2004
Abstract
A 97-kDa protein called valosin-containing protein (VCP) has been implicated in osteosarcoma metastasis and Paget’s disease of bone, two conditions that complicate the course and outcome of orthopaedic surgery. High VCP gene expression is associated with high metastatic potential in osteosarcoma cells, while loss-of-function VCP mutations cause inclusion body myopathy associated with Paget’s disease of bone and frontotemporal dementia (IBMPFD). VCP protein expression and regulation have not been examined in normal osteoblasts. The purpose of these studies was to characterize VCP protein expression in control and stressed untransformed osteoblasts. Proteins from confluent MC3T3-E 1 mouse osteoblast-like cells were separated by 2D IEF/SDS-PAGE. An abundant spot with a M,.of 94 kDa and a plof 5.4 was identified as VCP by MALDI/ToF and peptide mass fingerprint analysis. High constitutive VCP protein expression in subconfluent and confluent resting and mildly physiologically stressed MC3T3-El cells was confirmed by Western blotting. When assessed by indirect immunofluorescence in fixed cells or Western blotting of subcellular fractions, VCP was more abundant in the cytoplasm than in the nucleus. Induction of mild physiological stress sufficient to stimulate the ubiquitin-proteasome pathway, which is partially dependent on VCP-mediated targeting of polyubiquitinylated substrates, did not affect steady-state VCP levels or distribution. Thus, VCP is a constitutively abundant protein in untransformed osteoblastic cells under all conditions tested. Such high levels of VCP protein expression in untransformed osteoblastic cells argue against a major causative role for it in metastasis, while the occurrence of Paget’s disease in patients with missense VCP mutations supports a major role for VCP in normal osteoblast proliferation and regulation. 0 2005 Orthopaedic Research Society. Published by Elsevier Ltd. All rights reserved. Keyirurds: Valosin-containing protein; p97; Osteoblast; Proteomics
Introduction Orthopaedic surgery is frequently complicated by comorbid conditions, and a better understanding of the
* Corresponding author. Address: Geriatric Research, Education & Clinical Center ( I I-E). VA Greater Los Angeles Healthcare System, 16111 Plummer Street. Sepulveda, CA 91343, USA. Tel.: + I 818 895 931 I ; fax: + I 818 895 9519. E-nmil uddress: Elsa .Murray@med .va.gov (E.J. Brochmann) .
biochemical basis for these conditions may improve treatment. For example, osteosarcoma is the most common malignant tumor of bone, and metastases to the lung are a frequent complication [l]. High constitutive levels of VCP (valosin-containing protein) gene expression occur in highly metastatic osteosarcoma sublines, but not in parental cell lines that do not form pulmonary metastasis [l]. The VCP gene codes for a 97-kDa member of the AAA ATPase (ATPases associated with a variety of cellular activities) superfamily of proteins that regulate mitosis, membrane fusion, apoptosis, and UPP
0736-0266/$ - see front matter 0 2005 Orthopaedic Research Society. Published by Elsevier Ltd. All rights reserved doi: 10.1016/j.orthres.2004.12.012
K . Belinarn el ul. I Journul of Ortliopaedic Reseurcli 23 (2005) 6 1 8 4 2 4
(ubiquitin-proteasome pathway)-mediated protein turnover [25,30,31]. Proteins destined for UPP degradation are usually polyubiquitinylated by the enzymes of the ubiquitin pathway and subsequently bind to and are degraded by the 26 S proteasome that catalyzes most of the intracellular regulatory proteolysis in eucaryotes [23]. VCPI p97 binds both polyubiquitin chains and the proteasome [3], thus targeting ubiquitinylated substrates to the proteasome for degradation [24]. When stimulated with tumor necrosis factor-alpha, VCP-transfected osteosarcoma cells rapidly degrade the phosphorylated inhibitor of NF-KB-alpha (p-IKB-a) and exhibit enhanced metastasis due to the anti-apoptotic effects of constant NF-KB activation [l]. Dunn, MG-63, SaOS-2, and HOS osteosarcoma cells express VCP mRNA, but protein expression has only been illustrated in Dunn cells 113. Conversely, missense mutations in the ubiquitinbinding or ATPase domains of VCP cause the autosoma1 dominant progressive disorder known as inclusion body myopathy associated with Paget's disease of bone and frontotemporal dementia (IBMPFD) [25]. Despite the key roles of VCP in many aspects of cellular physiology and its association with two major disorders of bone, little is known about its abundance and subcellular distribution in untransformed osteoblastic cells during the cell cycle or in response to physiological stimuli. We addressed the following questions related to VCP protein expression in untransformed murine MC3T3El osteoblastic cells: (a) how abundant is VCP, (b) where is it found in the cell, and (c) do its expression or subcellular localization change during the cell cycle or under conditions of physiological stress sufficient to stimulate the UPP? Materials and methods Muteriuls
Rabbit anti-VCP was from Santa Cruz Biotech (Santa Cruz, CA) and anti-polyubiquitin was from Lampire Biologicals (Pipersville, PA). Routine biochemicals were from Sigma (St. Louis, MO). Protein assay kits, ALP (alkaline phosphatase)-conjugated 2" antibody, and ALP substrate were from Pierce (Rockford, IL). Western blotting supplies to detect polyubiquitin conjugates were from Vector Laboratories (Burlingame. CA). ID SDS-PAGE electrophoresis supplies, molecular weight standards, and DNA assay kits were from BioRad (Hercules, CA).
Culture and induction of physiological stress in MC3 T3-El osteoblust-like cells
MC3T3-EI osteoblast-like cells were cultured in Dulbecco's modified Eagle's media containing 4.5 gll glucose, 1oUh newborn calf serum, glutamine, and antibiotics [12]. For induction of mild physiological stress sufficient to stimulate the UPP, cells were randomized to a stationary shelf in the incubator (controls) or placed on a reciprocating shaker set to 35 cycles per minute (shaken cells) [7] for up to 48 h.
619
At the end of the culture periods. the cells were harvested by gentle scraping into PBS and pelleted by centrifugation. Estirnution of relutive VCP levels in MC3T3-EI
cells by 2D SDS-PAGE Cellular proteins (30 pg) were separated by 2D SDS-PAGE by the method of OFarrell [I51 (Nancy Kendrick, Ph.D., Kendrick Laboratories, Madison, WI) [2]. The gel was silver-stained by the method of OConnell and Stults [14]. The spot containing VCP/p97 was selected, excised, eluted, reduced in 5% j-mercaptoethanol, alkylated, trypsin digested, and subjected to peptide mass fingerprint analysis by MALDIITOF MS (Mary Ann Gawinowicz, Ph.D., Protein Chemistry Core Facility, Howard Hughes Medical Institute, Columbia University, New York, NY) [2]. Methionines weren't oxidized. Peptides with a mass [M + Hf] > 1000 Da were analyzed. The peptide fingerprint was compared to those in the SWISS-PRO and NCBI databanks [2]. Immunoblotting of VCP
The proteins in 50-pg aliquots of cell extracts were separated by 15% PAGE and transblotted to nitrocellulose [I 21. The transblot was
hydrated in TBS (Tris-buffered saline; 50 mM Tris-HCI, pH 7.6, and 150 mM NaCI) for 30 rnin at 23 "C and incubated in BB (blocking buffer: TBS plus 0.25'1/0 calf skin gelatin and 0.05'1/0NaN3) overnight at 4 "C. The blot was incubated with I" antibody (0.2 pg rabbit antiVCP/ml in BB) for 3 h and rinsed 3x with TTBS (TBS plus O.I'%l Surfact-Amps 20) for 5 rnin at 23 "C. The blot was incubated in ALP-conjugated donkey anti-rabbit IgG [1:2,5000 (vlv) in BB] for I h at 23 "C, rinsed in TTBS, and incubated in BCIP/NBT reagent for 20 min. The M, of the immunoreactive band was determined by comparison with prestained standards. Indirect iiivnunofluorescent localization of VCP in control arid
shaken celh Slides inscribed with 10-mm circles were subbed in 0.05'%~ potassium dichromate, coated with poly-L-lysine (0.1 mglml in water), and dried. MC3T3-EI cells ( I x lo5celld0.l ml) were applied in each circle. cultured to near confluence and then incubated under control or shaken conditions for 1 h. The cells were fixed in 2'fi paraformaldehyde in PBS for 15 min, rinsed 3x with PBS for 10 min and incubated in PBS plus 10%goat serum for 1 h at 23 "C to block nonspecific binding. The cells were incubated with rabbit anti-VCP [ I 5 0 (vlv) in PBS plus 3.0% goat serum] overnight at 4 "C and rinsed 3x with PBS for 5 rnin at 23 "C. The cells were incubated in goat anti-rabbit Cy3-conjugated 2" antibody (5 pglml in PBS with 3% goat serum) for 1 h, and rinsed 3x with PBS and 2x with water for 10 rnin at 23 "C. The cells were viewed by epifluorescence to visualize VCP (red). Negative controls, consisting of cells incubated in the absence of primary, secondary, or primary and secondary antibodies. were included. Verification of stress-induced stimulation of WPP activity and its effects on ALP
Ubiquitinylation and 26 S proteasome activity were measured as previously described [ 121. Cell pellets were suspended in a buffer containing 135 mM Tris-acetate, pH 7.5, 12.5 mM KCI, 80 pM EGTA, and 6.25 mM P-mercaptoethanol, and sonicated. Protein contents were determined by bicinchoninic acid colorimetric assay. Equal amounts of protein (50 pg) were separated by SDS-PAGE on 15% acrylamide gels and transblotted to PVDF [12]. High-molecular weight polyubiquitinylated protein levels were determined by Western blotting using a commercial ABC kit and a 1:1,000 (vlv) dilution of rabbit antibody to polyubiquitin-protein conjugates [12]. Digital images were captured. The M,'s and integrated optical densities of the high-molecular (> 100 kDa) immunopositive bands were determined with Gel-Pro analyzer (v. 3.0; Media Cybernetics, Silver Spring, MD). Proteasome activity was determined by incubating cell extract with 100mM Hepes-HCI (pH 7.5) containing 50 pM Ala-Ala-Phe-7-amido-4-methylcoumarin for 30 rnin at 37 "C [12]. The reaction was terminated by
620
K. Belinam et al. l Journal of Orthopaedic Research 23 (2005) 6 1 8 4 2 4
adding 220 mM sodium acetate buffer. Undegraded protein was precipitated for 30 rnin at 4 "C. The fluorescence o f the supernatant was measured at an excitation wavelength of 370nm and an emission wavelength of 430 nm. Alkaline phosphatase activity was measured spectrophotometrically at 41 0 nm based on the release of p-nitropheno1 from p-nitrophenylphosphate in alkaline buffer [I 31. Preparation of nuclear and ryroplasmir extracts [I61
Adherent cells were washed twice with 10 ml of PBS, incubated on ice for 10 rnin in 0.5 ml of Buffer A (10 mM Hepes-HCI, 10 mM KCI. 0.1 mM EDTA, 0.1% NP-40, 1 mM DTT. and 1"h (vlv) protease inhibitor cocktail 111 [Calbiochern]), and scraped. The lysates were incubated for 10 rnin on ice. and spun at 5000g for I rnin to pellet nuclei. After removing the cytosolic supernatant, the pellet was extracted 2x with 50 p1 of Buffer C (50 mM Hepes-HCI, pH 7.5, 420 mM KCI, 0.1 rnM EDTA, 5 mM MgC12, 20% (vlv) glycerol, 1 mM DTT. and 1% protease inhibitor cocktail). After the first addition of Buffer C , the tube was vortexed, held on ice for 30min, and spun at 14,OOOg for 30 rnin at 4 "C.The supernatant was removed, and a second aliquot of Buffer C was added. The final extract containing the nuclear pellet was sonicated for 15 s at 30% power at 4 "C. DNA (the nuclear marker) was assayed fluorometrically in the presence of Hoechst dye. Lactate dehydrogenase (LDH, the cytosolic marker) was assayed spectrophotometrically at 340 nm based on the conversion of NADH to NAD [20]. Sfuristical unuljxis
The mean and SEM were calculated using InStat (GraphPad Software, San Diego, CA). Pairs of means were compared using the twotailed Welch t-test. Representative results from repetitive experiments are presented.
Fig. 1. 2D map of silver-strained proteins from MC3T3-EI cells. An aliquot of protein (30 pg) was separated by 2D IEFISDS-PAGE. The area containing VCP is boxed. The outer boundaries o f the excised area are indicated with a dark line (-). The internal molecular weight standard (tropomyosin; M, = 33 kDa; P I =5 . 2 ) is indicated with a filled triangle.
Results 2D SDS-PAGE of confluent MC3T3-El cellular proteins resulted in the resolution of a spot with a p l of about 5.4 and a M , of 94 kDa that was present at levels equal to or greater than those of most of the other 500+ proteins in the cell [13] (Fig. 1). When it was excised and subjected to MALDUTOF MS for peptide fingerprinting, a total of 24 different peptides were identified that corresponded to discrete peptides contained in the sequence of murine VCP (Table 1). This permitted rapid unambiguous identification of high levels of VCP in MC3T3-El cells. VCP can undergo phosphorylation and subcellular redistribution in response to cell cycle progression or physiologic stimuli [4,10]. Therefore, the subcellular distribution of VCP was assessed in control and physiologically stimulated (shaken) MC3T3-El cells. VCP was found in the cytoplasm and perinuclear Golgi, as well as the nuclei, of control (Fig. 2A) and shaken cells (Fig. 2B). No immunofluorescence was observed in the absence of 1"/2" antibodies (data not shown). Since VCP helps target polyubiquitinylated proteins for UPP-mediated degradation [3], we induced mild physiological stress (shaking) to stimulate the UPP and assessed its effects on steady-state VCP protein levels and cytosolidnuclear distribution. Shaking MC3T3El cells for 24 h increased the steady-state levels of high
molecular weight polyubiquitinylated proteins, stimulated 26 S proteasome activity, and reduced ALP activity (Table 2). Thus, the activity of both major arms of the UPP increased, while the levels of ALP (a UPP-substrate) fell in response to 24 h of shaking at 35 cpm. Proteasome activity and high molecular weight polyubiquitinylated protein levels returned to normal after 48 h of shaking (data not shown). When assessed by Western blotting, steady-state VCP levels were similar in cells at near confluence (time-zero) or after 24 or 48 h of further culture under control or shaken conditions (Fig. 3). Thus, VCP appears to be constitutively expressed at similar levels under control and mildly stressed conditions sufficient to stimulate the UPP in MC3T3-El cells. Finally, shaking had no effect on the subcellular distribution of VCP between the cytosol and the nucleus (Fig. 4). Discussion VCP is an 806-amino acid residue AAA ATPase with a calculated mass of 89.3 kDa and a p l of 5.14 [NCBI database]. Modern proteomic methods were used to unambiguously identify VCP in untransformed MC3T3-El osteoblast-like cells. With an observed M , of 94 kDa and a p l of 5.4, and confirmation of the
K . Behnurn et ul. I Journal q/’ Ortlivpacdic Reseurch 23 (200.5) 6 1 8 4 2 4
62 1
Table 1 Identification of peptide fragments of valosin-containing protein (VCP) in MC3T3-EI cell extracts after 2D IEFISDS-PAGE, excision of the 94-kDa spot from the 2D IEFISDS-PAGE gel, tryptic digestion, MALDmOF MS, and bioinformatic data reduction Observed peptide mass (Da)
Theoretical peptide mdSS (Da)
VCP residue #
Peptide sequence
1190.56 1267.03 1330.43 1416.70 1502.38 1594.06 1630.85 1680.87 1779.25 1811.74 1815.11 1937.97 1981.10 1992.10 2166.28 2166.28 2187.23 2778.94 2809.46 3655.19 3688.70 3817.31 3873.86 4058.45
1191.32 1266.51 1330.50 1416.63 1502.70 1593.85 1630.88 1680.99 1779.12 1811.17 1815.05 1937.20 1981.23 1992.43 2166.52 2166.52 2187.23 2779.10 2809.13 3655.08 3689.08 3817.25 3873.09 4058.79
733-741 5 13-524 454465 697-708 755-766 530-543 754766 600-6 14 4660 94109 678493 544560 487-502 232-251 566-584 568-586 192-2 10 587-614 21 4 454486 390424 390-425 767-806 616651
RDHFEEAMR GVLFYGPPGCGK WALSQSNPSALR LAIRESIESEIR YEM FAQTLQQSR A1ANECQANFISIK KYEMFAQTLQQSR VINQILTEMDGMSTK MDELQLFRGDTVLLK VRLGDVISIQPCPDVK MTNGFSGADLTEICQR GPELLTMWFGESEANVR RELQELVQYPVEHPDK AIGVKPPRGILLYGPPGTGK ARQAAPCVLFFDELDSI AK QAAPCVLFFDELDSIAKAR EDEEESLNEVGYDDVGGCR GGNIGDGGGAADRVINQILT EMDGMSTK NRPNRLIVDEAINEDNSVVS LSQPK WALSQSNPSALRETVVEVPQ VTWEDIGGLEDVK LADDVDLEQVANETHGHVGA DLAALCSEAALQAIR LADDVDLEQVANETHGHVGA DLAALCSEAALQAIRK GFGSFRFPSGNQGGAGPSQGSGGGTGGSVYTEDNDDDLY G NVFIIGATNRPDIIDPAILR PGRLDQLIYIPLPDEK
Fig. 2. Photomicrograph of MC3T3-El osteoblast-like cells under control conditions (A) or after 1 h of shaking at 35 cpm (B). Intense staining for VCP (red) is observed in the cytoplasm, cytoskeleton, and perinuclear Golgi, as well as the nuclei, of control and shaken cells. Original magnification 200x.
K. Behnaln et al. I Journal of’ Ortliopaedie Research 23 (200.5) 618424
622
Table 2 Effects of 24 h of mild physiological stress (shaking) on ubiquitin-proteasome pdthwdy and alkaline phosphatase activity in MC3T3-El cells Parameter
Control cells
Shaken cells
p value
Proteasome activity (relative fluorescence unitslmg protein) High molecular weight (> 100 kDa) polyubiquitinylated protein immunoreactivity (integrated optical density units) Alkaline phosphatase activity (nmol pNP liberatedlmg protein)
8283 f 152 (4) 2921 f 338 (3)
9706 f 531 (4) 6517 f 855 (3)
p < 0.05
108.3 f 2.5 (3)
82.8 f 1.4 (4)
p = 0.003
p = 0.017
Data are expressed as mean f S.E.M. (n).
Mr 133 40
32
19
7
Time (Hours)
0
24
24
48
48
-
+
-
+
Shaking
Fig. 3. Western blots showing a single immunopositive VCP band ( M r = 9 7 k D a ) in MC3T3-EI cell extracts (50-pg protein) at near confluence or after 24 (confluence) or 48 (post-confluence) additional hours of culture under control or shaken conditions. Lane 1: Molecular weight standards (myosin: 199-kDa; galactosidase: 133kDa; BSA: 87-kDa; carbonic anhydrase: 40-kDa. soybean trypsin inhibitor: 32-kDa; lysosyme: 19-kDa; and aprotinin: 7-kDa). Equivalence of loading was confirmed by Coomassie blue staining (results not shown).
P97 Fraction Shaking
-
C
C
+
NP
NP
+
Fig. 4. Western blot showing the steady-state constitutive distribution of VCP in the cytoplasm and nucleus of control and shaken MC3T3El cells after 24 and 48 h of treatment. Distribution of markers: 96.7% f 1.1% (n = 4) of the LDH was found in the cytosolic [C] fraction, while 92.5% f 0.7% (n = 5) of the DNA was found in the nuclear pellet [NP].
presence of 24 discrete peptide fragments identical to those in murine VCP [21], there is little possibility that a related protein, such as the 855 amino acid residue nuclear VCP-like protein or its 307 amino acid residue fragment [5], has been identified. However, these proteins may be present in spots distinct from the one that was analyzed. Definitive localization of VCP in the 2D database of osteoblastic proteins should facilitate comparative proteomic analyses in the future. Immunofluorescent staining of intact cells and classical Western blotting showed that the level of VCP is remarkably stable in the cytosol and nuclei of MC3T3-
El osteoblastic cells at confluence and following induction of mild physiological stress for 24 or 48 h in postconfluent cells. Stable constitutive expression of VCP in MC3T3-El cells differs significantly from that observed in osteosarcoma cells, where VCP declines to very low levels at confluence [l], or in yeast, where VCP levels are cell cycle dependent [lo]. VCP is immunolocalized to the cytosol, cytoskeleton, perinuclear Golgi, and nuclei of confluent MC3T3-E1 osteoblastic cells, but appears to be confined primarily to the cytoplasm of preconfluent, but not post-confluent, Dunn osteosarcoma cells [l]. VCP is abundant in the nucleus and ER membranes of epidermoid carcinoma and COS-7 kidney cells [26]. Endothelial and well-differentiated gastric carcinoma cells exhibit strong cytoplasmic VCP immunostaining that is absent in undifferentiated tumor cells [28]. VCP is present at high concentrations in the endomysial capillaries and lipofuscin granules of normal muscle cells, where cytoplasmic levels of VCP are low [25]. In contrast, VCP is localized in both large and small focal inclusions in the muscle tissue of IBMPFD patients [25]. Thus, VCP appears to be more widely distributed in MC3T3-El cells than it is in the other mammalian cell types examined to date. The high levels of stable constitutive expression and complex subcellular distribution of VCP in MC3T3-El cells are consistent with significant roles for VCP in its previously defined diverse functions in suppression of apoptosis [1,3], ERAD of aberrant proteins in the secretory pathway [18]; protein transport from the ER (endoplasmic reticulum) to the cytosol [29]; membrane fusion during Golgi, ER, and nuclear membrane assembly [11,221; receptor-mediated endocytosis and Golgi sorting [ 171; and transcriptioncoupled DNA repair [31] in these cells. VCP contains three major domains, including two 250-amino acid residue ATPase domains (D1 and D2), each of which contains the Walker A and B motifs required for ATP binding and hydrolysis, as well as an amino-terminal domain of approximately 200 amino acids that binds substrates, such as polyubiquitinylated proteins [24]. The entire molecule is required for 26 S proteasome binding and UPP degradation [19]. Lossof-function missense mutations clustered in the N-terminal polyubiquitin-binding domain or in the D1 domain of VCP cause IBMPFD [25], a disorder characterized by coarse trabeculation, cortical thickening, and spotty sclerosis in the spine, pelvis, shoulder, and skull [8].
K. Behnatn el al. I Journul of Orthopedic Research 23 (2005) 618--624
VCP RNA interference in HeLa cells is associated with the accumulation of polyubiquitinylated proteins in dispersed aggregates, cytosolic vacuolization, cell cycle progression blockage, mitotic abnormalities, and apoptosis [27]. However, these features were not reported for mineralized tissue from IBMPFD patients [8,25], suggesting that redundant biochemical pathways may reduce the severity of the effects of VCP mutations in bone. Mutations in the ubiquitin-binding domain of sequestosome I (SQSTM 1) cause autosomal dominant Paget’s disease of bone 3 (PDB3) [6,9], and it has been suggested that both PDB3 and IBMFPD are characterized by compromised ubiquitin-binding and targeting of similar cellular pathways or proteins [25]. However, since both up-regulation of VCP and impaired VCP function are associated with dysregulated over-growth of osteoblastic cells (e.g., osteosarcoma and Paget’s disease of bone), it is clear that additional research will be required to identify VCP binding partners and UPP substrates in osteoblasts and other bone cells and to determine their roles in signal transduction and cellular physiology in normal and pathological states. Acknowledgments We acknowledge the expert technical contributions and invaluable advice of Drs. Nancy Kendrick (Kendrick Laboratories), Mary Ann Gawinowicz (Columbia University), and Raffi Vartanian (Sepulveda). This work was supported, in part, by the Geriatric Research, Education, and Clinical Centers and the Research Service of the Department of Veterans Affairs (SSM, EJB). References [I] Asai T, Tomita Y, Nakatsuka S, et al. VCP (p97) regulates NF-KB signaling pathway, which is important for metastasis of osteosarcoma cell line. Jpn J Cancer Res 2002;93:296304. [2] Behnam K, Murray S, Whitelegge J, Brochmann E. Identification of the molecular chaperone alpha B-crystallin in demineralized bone powder and osteoblast-like cells. J Orthop Res 2002;20: 1190-6. [3] Dai R, Li C. VCP is a multi-ubiquitin chain targeting factor required in ubiquitin-proteasome degradation. Nature Cell Biol 200 1;3:7404 [4] Ficarro S, Chertihin 0, Westbrook VA, et al. Phosphoproteome analysis of capacitated human sperm. J Biol Chem 2003;278: 1 1579-89. [5] Germain-Lee E, Obie C, Valle D. NVL: a new member of the AAA family of ATPases localized to the nucleus. Genomics 1997;44:22-34. [6] Hocking L, Lucas G, Daroszewska A, et al. Domain-specific mutations in sequestosome 1 (SQSTMI) cause familial and sporadic Paget’s disease. Hum Mol Genet 2002;11:2735-9. [7] Jessop L, Rawlinson S, Pitsillides A, Lanyon L. Mechanical strain and fluid movement both activate extracellular regulated kinase (ERK) in osteoblast-like cells but via different signaling pathways. Bone 2002;31:18694.
623
[8] Kimonis V. Kovach M, Waggoner B, et al. Clinical and molecular studies in a unique family with autosomal dominant limbgirdle muscular dystrophy and Paget disease of bone. Genet Med 2000;2:23241. [9] Laurin N, Brown J, Morissette J, et al. Recurrent mutation of the gene encoding sequestosome 1 in Paget disease of bone. Am J Hum Genet 2002;70:1582-8. [lo] Madeo F, Schlauer J, Zischka H, et al. Tyrosine phosphorylation regulates cell cycle-dependent nuclear localization of Cdc48p. Mol Biol Cell 1998;9:13141. [Ill Meyer HH, Yang Y, Warren G. Direct binding of ubiquitin conjugates by the mammalian p97 adaptor complexes, p47 and Udfl-Npl4. EMBO J 2002;21:564-52. [12] Murray E, Bentley G, Grisanti M, Murray S. The ubiquitinproteasome system and cellular proliferation and regulation in osteoblastic cells. Exp Cell Res 1998;242:460-9. 131 Murray E, Murray S, Manolagas S. 2D gel autoradiographic analyses of the effects of 1,25-dihydroxycholecalciferolon protein synthesis in clonal rat ROS cells. Endocrinology 1990: 126: 2679-92. 141 OConnell K, Stults J. Identification of mouse liver proteins on 2D electrophoresis gels by MALDI MS of in situ enzymatic digests. Electrophoresis 1997;18:349-59. 151 OFarrell PH. High-resolution 2D electrophoresis of proteins. J Biol Chem 1975;250:4007-21. [I61 Ogasawara A, Arakawa T, Kaneda T, et al. Fluid shear stressinduced cyclooxygenase-2 expression is mediated by ClEBP p. CAMP-response element-binding protein, and AP-I in osteoblastic MC3T3-EI cells. J Biol Chem 2001;2767048-54. [I71 Pleasure I, Black M, Keen J. VCP is a ubiquitous clathrin-binding protein. Nature 1993;365:459-62. [18] Rabinovich E, Kerem A, Frohlich K, et al. AAA-ATPase p97l Cdc48p, a cytosolic chaperone required for ER-associated protein degradation. Mol Cell Biol 2002;22:62634. [I91 Song C, Wang Q. Li C. ATPase activity of p97-VCP. D2 mediates the major enzyme activity, and DI contributes to the heat-induced activity. J Biol Chem 2003;278:3648-55. [20] Storrie B, Madden EA. Isolation of subcellular organelles. In: Deutscher MP. Guide to protein purification. Method Enzymol 1990;182-220. [21] Strausberg R, Feingold E, Grouse L, et al. Generation and initial analysis of more than 15,000 full-length human and mouse cDNA sequences. Proc Natl Acad Sci USA 2002;99:16899-903. [22] Uchiyama K, Jokitalo E, Kano F, et al. VCIP135, a novel essential factor for p97lp47-mediated membrane fusion, is required for Golgi and ER assembly in uiuo. J Cell Biol 2002; 159:855-66. [23] Volker C, Lupas A. Molecular evolution of proteasomes. The proteasome-ubiquitin protein degradation pathway. New York: Springer; 2002. [24] Wang Q , Song C, Li C. Hexamerization of p97-VCP is promoted by ATP binding to the DI domain and required for ATPase and biological activities. Biochem Biophys Res Commun 2003:300: 25340. [25] Watts G, Wymer J, Kovach M, et al. Inclusion body myopathy associated with Paget disease of bone and frontotemporal dementia is caused by mutant VCP. Nat Genet 2004;36:37781. [26] Waugh M, Minogue S, Anderson J, et al. Localization of a highly active pool of type I1 PI4K in a p97NCP-rich fraction of the ER. Biochem J 2003;373:57-63. [27] Wojcik C, Yano M, DeMartino G. RNA interference of VCP reveals multiple cellular roles linked to ubiquitinlproteasomedependent proteolysis. J Cell Sci 2004;117:281-92. [28] Yamamoto S, Tomita Y, Hoshida Y, et al. Expression level of VCP is strongly associated with progression and prognosis of gastric carcinoma. J Clin Oncol 2003;21:25374.
624
K. Behnum et al. I Journal of Orthopaedic Research 23 (2005) 618424
[29] Ye Y, Meyer H, Rapoport T. The AAA ATPase Cdc48/p97 and its partners transport proteins from the ER to the cytosol. Nature 200 1;4 14:652-6. [30] Zalk R, Shoshan-Barmatz V. ATP binding sites in brain p97NCP (valosin-containing protein), a multifunctional AAA ATPdse. Biochem J 2003:374473-80.
[31] Zhang H, Wang Q, Kajino K, et al. VCP, a weak ATPase involved in multiple cellular events, interacts physically with BRCAI in the nucleus of living cells. DNA Cell Biol 2000;19: 253-63.
Journal of Orthopaedic Research
Journal of Orthopaedic Research 23 (2005) 618-624
www.elsevier.cod1ocatelorthres
Identification and characterization of valosin-containing protein (VCP/p97) in untransformed osteoblast-like cells Keyvan Behnam
a,
Samuel S. Murray
b3c, Elsa
J. Brochmann
b7c7*
Department .J‘Ph~wicilogiculSciences, Uniuersitv .f Culifi,rniu. Los Angeles. C A 90024. USA Department of Medicine, University of Culiforniu, Los Angeles. CA 90024. USA Geriutric Reseurch, Educution & Clinicul Center ( I I - E ) . VA Greuter Los Angeles Heultkcure System, 16111 Plunimer Srreer, Sepuhedu. C A 91343, USA
Received 1 December 2004; accepted 22 December 2004
Abstract
A 97-kDa protein called valosin-containing protein (VCP) has been implicated in osteosarcoma metastasis and Paget’s disease of bone, two conditions that complicate the course and outcome of orthopaedic surgery. High VCP gene expression is associated with high metastatic potential in osteosarcoma cells, while loss-of-function VCP mutations cause inclusion body myopathy associated with Paget’s disease of bone and frontotemporal dementia (IBMPFD). VCP protein expression and regulation have not been examined in normal osteoblasts. The purpose of these studies was to characterize VCP protein expression in control and stressed untransformed osteoblasts. Proteins from confluent MC3T3-E 1 mouse osteoblast-like cells were separated by 2D IEF/SDS-PAGE. An abundant spot with a M,.of 94 kDa and a plof 5.4 was identified as VCP by MALDI/ToF and peptide mass fingerprint analysis. High constitutive VCP protein expression in subconfluent and confluent resting and mildly physiologically stressed MC3T3-El cells was confirmed by Western blotting. When assessed by indirect immunofluorescence in fixed cells or Western blotting of subcellular fractions, VCP was more abundant in the cytoplasm than in the nucleus. Induction of mild physiological stress sufficient to stimulate the ubiquitin-proteasome pathway, which is partially dependent on VCP-mediated targeting of polyubiquitinylated substrates, did not affect steady-state VCP levels or distribution. Thus, VCP is a constitutively abundant protein in untransformed osteoblastic cells under all conditions tested. Such high levels of VCP protein expression in untransformed osteoblastic cells argue against a major causative role for it in metastasis, while the occurrence of Paget’s disease in patients with missense VCP mutations supports a major role for VCP in normal osteoblast proliferation and regulation. 0 2005 Orthopaedic Research Society. Published by Elsevier Ltd. All rights reserved. Keyirurds: Valosin-containing protein; p97; Osteoblast; Proteomics
Introduction Orthopaedic surgery is frequently complicated by comorbid conditions, and a better understanding of the
* Corresponding author. Address: Geriatric Research, Education & Clinical Center ( I I-E). VA Greater Los Angeles Healthcare System, 16111 Plummer Street. Sepulveda, CA 91343, USA. Tel.: + I 818 895 931 I ; fax: + I 818 895 9519. E-nmil uddress: Elsa .Murray@med .va.gov (E.J. Brochmann) .
biochemical basis for these conditions may improve treatment. For example, osteosarcoma is the most common malignant tumor of bone, and metastases to the lung are a frequent complication [l]. High constitutive levels of VCP (valosin-containing protein) gene expression occur in highly metastatic osteosarcoma sublines, but not in parental cell lines that do not form pulmonary metastasis [l]. The VCP gene codes for a 97-kDa member of the AAA ATPase (ATPases associated with a variety of cellular activities) superfamily of proteins that regulate mitosis, membrane fusion, apoptosis, and UPP
0736-0266/$ - see front matter 0 2005 Orthopaedic Research Society. Published by Elsevier Ltd. All rights reserved doi: 10.1016/j.orthres.2004.12.012
K . Belinarn el ul. I Journul of Ortliopaedic Reseurcli 23 (2005) 6 1 8 4 2 4
(ubiquitin-proteasome pathway)-mediated protein turnover [25,30,31]. Proteins destined for UPP degradation are usually polyubiquitinylated by the enzymes of the ubiquitin pathway and subsequently bind to and are degraded by the 26 S proteasome that catalyzes most of the intracellular regulatory proteolysis in eucaryotes [23]. VCPI p97 binds both polyubiquitin chains and the proteasome [3], thus targeting ubiquitinylated substrates to the proteasome for degradation [24]. When stimulated with tumor necrosis factor-alpha, VCP-transfected osteosarcoma cells rapidly degrade the phosphorylated inhibitor of NF-KB-alpha (p-IKB-a) and exhibit enhanced metastasis due to the anti-apoptotic effects of constant NF-KB activation [l]. Dunn, MG-63, SaOS-2, and HOS osteosarcoma cells express VCP mRNA, but protein expression has only been illustrated in Dunn cells 113. Conversely, missense mutations in the ubiquitinbinding or ATPase domains of VCP cause the autosoma1 dominant progressive disorder known as inclusion body myopathy associated with Paget's disease of bone and frontotemporal dementia (IBMPFD) [25]. Despite the key roles of VCP in many aspects of cellular physiology and its association with two major disorders of bone, little is known about its abundance and subcellular distribution in untransformed osteoblastic cells during the cell cycle or in response to physiological stimuli. We addressed the following questions related to VCP protein expression in untransformed murine MC3T3El osteoblastic cells: (a) how abundant is VCP, (b) where is it found in the cell, and (c) do its expression or subcellular localization change during the cell cycle or under conditions of physiological stress sufficient to stimulate the UPP? Materials and methods Muteriuls
Rabbit anti-VCP was from Santa Cruz Biotech (Santa Cruz, CA) and anti-polyubiquitin was from Lampire Biologicals (Pipersville, PA). Routine biochemicals were from Sigma (St. Louis, MO). Protein assay kits, ALP (alkaline phosphatase)-conjugated 2" antibody, and ALP substrate were from Pierce (Rockford, IL). Western blotting supplies to detect polyubiquitin conjugates were from Vector Laboratories (Burlingame. CA). ID SDS-PAGE electrophoresis supplies, molecular weight standards, and DNA assay kits were from BioRad (Hercules, CA).
Culture and induction of physiological stress in MC3 T3-El osteoblust-like cells
MC3T3-EI osteoblast-like cells were cultured in Dulbecco's modified Eagle's media containing 4.5 gll glucose, 1oUh newborn calf serum, glutamine, and antibiotics [12]. For induction of mild physiological stress sufficient to stimulate the UPP, cells were randomized to a stationary shelf in the incubator (controls) or placed on a reciprocating shaker set to 35 cycles per minute (shaken cells) [7] for up to 48 h.
619
At the end of the culture periods. the cells were harvested by gentle scraping into PBS and pelleted by centrifugation. Estirnution of relutive VCP levels in MC3T3-EI
cells by 2D SDS-PAGE Cellular proteins (30 pg) were separated by 2D SDS-PAGE by the method of OFarrell [I51 (Nancy Kendrick, Ph.D., Kendrick Laboratories, Madison, WI) [2]. The gel was silver-stained by the method of OConnell and Stults [14]. The spot containing VCP/p97 was selected, excised, eluted, reduced in 5% j-mercaptoethanol, alkylated, trypsin digested, and subjected to peptide mass fingerprint analysis by MALDIITOF MS (Mary Ann Gawinowicz, Ph.D., Protein Chemistry Core Facility, Howard Hughes Medical Institute, Columbia University, New York, NY) [2]. Methionines weren't oxidized. Peptides with a mass [M + Hf] > 1000 Da were analyzed. The peptide fingerprint was compared to those in the SWISS-PRO and NCBI databanks [2]. Immunoblotting of VCP
The proteins in 50-pg aliquots of cell extracts were separated by 15% PAGE and transblotted to nitrocellulose [I 21. The transblot was
hydrated in TBS (Tris-buffered saline; 50 mM Tris-HCI, pH 7.6, and 150 mM NaCI) for 30 rnin at 23 "C and incubated in BB (blocking buffer: TBS plus 0.25'1/0 calf skin gelatin and 0.05'1/0NaN3) overnight at 4 "C. The blot was incubated with I" antibody (0.2 pg rabbit antiVCP/ml in BB) for 3 h and rinsed 3x with TTBS (TBS plus O.I'%l Surfact-Amps 20) for 5 rnin at 23 "C. The blot was incubated in ALP-conjugated donkey anti-rabbit IgG [1:2,5000 (vlv) in BB] for I h at 23 "C, rinsed in TTBS, and incubated in BCIP/NBT reagent for 20 min. The M, of the immunoreactive band was determined by comparison with prestained standards. Indirect iiivnunofluorescent localization of VCP in control arid
shaken celh Slides inscribed with 10-mm circles were subbed in 0.05'%~ potassium dichromate, coated with poly-L-lysine (0.1 mglml in water), and dried. MC3T3-EI cells ( I x lo5celld0.l ml) were applied in each circle. cultured to near confluence and then incubated under control or shaken conditions for 1 h. The cells were fixed in 2'fi paraformaldehyde in PBS for 15 min, rinsed 3x with PBS for 10 min and incubated in PBS plus 10%goat serum for 1 h at 23 "C to block nonspecific binding. The cells were incubated with rabbit anti-VCP [ I 5 0 (vlv) in PBS plus 3.0% goat serum] overnight at 4 "C and rinsed 3x with PBS for 5 rnin at 23 "C. The cells were incubated in goat anti-rabbit Cy3-conjugated 2" antibody (5 pglml in PBS with 3% goat serum) for 1 h, and rinsed 3x with PBS and 2x with water for 10 rnin at 23 "C. The cells were viewed by epifluorescence to visualize VCP (red). Negative controls, consisting of cells incubated in the absence of primary, secondary, or primary and secondary antibodies. were included. Verification of stress-induced stimulation of WPP activity and its effects on ALP
Ubiquitinylation and 26 S proteasome activity were measured as previously described [ 121. Cell pellets were suspended in a buffer containing 135 mM Tris-acetate, pH 7.5, 12.5 mM KCI, 80 pM EGTA, and 6.25 mM P-mercaptoethanol, and sonicated. Protein contents were determined by bicinchoninic acid colorimetric assay. Equal amounts of protein (50 pg) were separated by SDS-PAGE on 15% acrylamide gels and transblotted to PVDF [12]. High-molecular weight polyubiquitinylated protein levels were determined by Western blotting using a commercial ABC kit and a 1:1,000 (vlv) dilution of rabbit antibody to polyubiquitin-protein conjugates [12]. Digital images were captured. The M,'s and integrated optical densities of the high-molecular (> 100 kDa) immunopositive bands were determined with Gel-Pro analyzer (v. 3.0; Media Cybernetics, Silver Spring, MD). Proteasome activity was determined by incubating cell extract with 100mM Hepes-HCI (pH 7.5) containing 50 pM Ala-Ala-Phe-7-amido-4-methylcoumarin for 30 rnin at 37 "C [12]. The reaction was terminated by
620
K. Belinam et al. l Journal of Orthopaedic Research 23 (2005) 6 1 8 4 2 4
adding 220 mM sodium acetate buffer. Undegraded protein was precipitated for 30 rnin at 4 "C. The fluorescence o f the supernatant was measured at an excitation wavelength of 370nm and an emission wavelength of 430 nm. Alkaline phosphatase activity was measured spectrophotometrically at 41 0 nm based on the release of p-nitropheno1 from p-nitrophenylphosphate in alkaline buffer [I 31. Preparation of nuclear and ryroplasmir extracts [I61
Adherent cells were washed twice with 10 ml of PBS, incubated on ice for 10 rnin in 0.5 ml of Buffer A (10 mM Hepes-HCI, 10 mM KCI. 0.1 mM EDTA, 0.1% NP-40, 1 mM DTT. and 1"h (vlv) protease inhibitor cocktail 111 [Calbiochern]), and scraped. The lysates were incubated for 10 rnin on ice. and spun at 5000g for I rnin to pellet nuclei. After removing the cytosolic supernatant, the pellet was extracted 2x with 50 p1 of Buffer C (50 mM Hepes-HCI, pH 7.5, 420 mM KCI, 0.1 rnM EDTA, 5 mM MgC12, 20% (vlv) glycerol, 1 mM DTT. and 1% protease inhibitor cocktail). After the first addition of Buffer C , the tube was vortexed, held on ice for 30min, and spun at 14,OOOg for 30 rnin at 4 "C.The supernatant was removed, and a second aliquot of Buffer C was added. The final extract containing the nuclear pellet was sonicated for 15 s at 30% power at 4 "C. DNA (the nuclear marker) was assayed fluorometrically in the presence of Hoechst dye. Lactate dehydrogenase (LDH, the cytosolic marker) was assayed spectrophotometrically at 340 nm based on the conversion of NADH to NAD [20]. Sfuristical unuljxis
The mean and SEM were calculated using InStat (GraphPad Software, San Diego, CA). Pairs of means were compared using the twotailed Welch t-test. Representative results from repetitive experiments are presented.
Fig. 1. 2D map of silver-strained proteins from MC3T3-EI cells. An aliquot of protein (30 pg) was separated by 2D IEFISDS-PAGE. The area containing VCP is boxed. The outer boundaries o f the excised area are indicated with a dark line (-). The internal molecular weight standard (tropomyosin; M, = 33 kDa; P I =5 . 2 ) is indicated with a filled triangle.
Results 2D SDS-PAGE of confluent MC3T3-El cellular proteins resulted in the resolution of a spot with a p l of about 5.4 and a M , of 94 kDa that was present at levels equal to or greater than those of most of the other 500+ proteins in the cell [13] (Fig. 1). When it was excised and subjected to MALDUTOF MS for peptide fingerprinting, a total of 24 different peptides were identified that corresponded to discrete peptides contained in the sequence of murine VCP (Table 1). This permitted rapid unambiguous identification of high levels of VCP in MC3T3-El cells. VCP can undergo phosphorylation and subcellular redistribution in response to cell cycle progression or physiologic stimuli [4,10]. Therefore, the subcellular distribution of VCP was assessed in control and physiologically stimulated (shaken) MC3T3-El cells. VCP was found in the cytoplasm and perinuclear Golgi, as well as the nuclei, of control (Fig. 2A) and shaken cells (Fig. 2B). No immunofluorescence was observed in the absence of 1"/2" antibodies (data not shown). Since VCP helps target polyubiquitinylated proteins for UPP-mediated degradation [3], we induced mild physiological stress (shaking) to stimulate the UPP and assessed its effects on steady-state VCP protein levels and cytosolidnuclear distribution. Shaking MC3T3El cells for 24 h increased the steady-state levels of high
molecular weight polyubiquitinylated proteins, stimulated 26 S proteasome activity, and reduced ALP activity (Table 2). Thus, the activity of both major arms of the UPP increased, while the levels of ALP (a UPP-substrate) fell in response to 24 h of shaking at 35 cpm. Proteasome activity and high molecular weight polyubiquitinylated protein levels returned to normal after 48 h of shaking (data not shown). When assessed by Western blotting, steady-state VCP levels were similar in cells at near confluence (time-zero) or after 24 or 48 h of further culture under control or shaken conditions (Fig. 3). Thus, VCP appears to be constitutively expressed at similar levels under control and mildly stressed conditions sufficient to stimulate the UPP in MC3T3-El cells. Finally, shaking had no effect on the subcellular distribution of VCP between the cytosol and the nucleus (Fig. 4). Discussion VCP is an 806-amino acid residue AAA ATPase with a calculated mass of 89.3 kDa and a p l of 5.14 [NCBI database]. Modern proteomic methods were used to unambiguously identify VCP in untransformed MC3T3-El osteoblast-like cells. With an observed M , of 94 kDa and a p l of 5.4, and confirmation of the
K . Behnurn et ul. I Journal q/’ Ortlivpacdic Reseurch 23 (200.5) 6 1 8 4 2 4
62 1
Table 1 Identification of peptide fragments of valosin-containing protein (VCP) in MC3T3-EI cell extracts after 2D IEFISDS-PAGE, excision of the 94-kDa spot from the 2D IEFISDS-PAGE gel, tryptic digestion, MALDmOF MS, and bioinformatic data reduction Observed peptide mass (Da)
Theoretical peptide mdSS (Da)
VCP residue #
Peptide sequence
1190.56 1267.03 1330.43 1416.70 1502.38 1594.06 1630.85 1680.87 1779.25 1811.74 1815.11 1937.97 1981.10 1992.10 2166.28 2166.28 2187.23 2778.94 2809.46 3655.19 3688.70 3817.31 3873.86 4058.45
1191.32 1266.51 1330.50 1416.63 1502.70 1593.85 1630.88 1680.99 1779.12 1811.17 1815.05 1937.20 1981.23 1992.43 2166.52 2166.52 2187.23 2779.10 2809.13 3655.08 3689.08 3817.25 3873.09 4058.79
733-741 5 13-524 454465 697-708 755-766 530-543 754766 600-6 14 4660 94109 678493 544560 487-502 232-251 566-584 568-586 192-2 10 587-614 21 4 454486 390424 390-425 767-806 616651
RDHFEEAMR GVLFYGPPGCGK WALSQSNPSALR LAIRESIESEIR YEM FAQTLQQSR A1ANECQANFISIK KYEMFAQTLQQSR VINQILTEMDGMSTK MDELQLFRGDTVLLK VRLGDVISIQPCPDVK MTNGFSGADLTEICQR GPELLTMWFGESEANVR RELQELVQYPVEHPDK AIGVKPPRGILLYGPPGTGK ARQAAPCVLFFDELDSI AK QAAPCVLFFDELDSIAKAR EDEEESLNEVGYDDVGGCR GGNIGDGGGAADRVINQILT EMDGMSTK NRPNRLIVDEAINEDNSVVS LSQPK WALSQSNPSALRETVVEVPQ VTWEDIGGLEDVK LADDVDLEQVANETHGHVGA DLAALCSEAALQAIR LADDVDLEQVANETHGHVGA DLAALCSEAALQAIRK GFGSFRFPSGNQGGAGPSQGSGGGTGGSVYTEDNDDDLY G NVFIIGATNRPDIIDPAILR PGRLDQLIYIPLPDEK
Fig. 2. Photomicrograph of MC3T3-El osteoblast-like cells under control conditions (A) or after 1 h of shaking at 35 cpm (B). Intense staining for VCP (red) is observed in the cytoplasm, cytoskeleton, and perinuclear Golgi, as well as the nuclei, of control and shaken cells. Original magnification 200x.
K. Behnaln et al. I Journal of’ Ortliopaedie Research 23 (200.5) 618424
622
Table 2 Effects of 24 h of mild physiological stress (shaking) on ubiquitin-proteasome pdthwdy and alkaline phosphatase activity in MC3T3-El cells Parameter
Control cells
Shaken cells
p value
Proteasome activity (relative fluorescence unitslmg protein) High molecular weight (> 100 kDa) polyubiquitinylated protein immunoreactivity (integrated optical density units) Alkaline phosphatase activity (nmol pNP liberatedlmg protein)
8283 f 152 (4) 2921 f 338 (3)
9706 f 531 (4) 6517 f 855 (3)
p < 0.05
108.3 f 2.5 (3)
82.8 f 1.4 (4)
p = 0.003
p = 0.017
Data are expressed as mean f S.E.M. (n).
Mr 133 40
32
19
7
Time (Hours)
0
24
24
48
48
-
+
-
+
Shaking
Fig. 3. Western blots showing a single immunopositive VCP band ( M r = 9 7 k D a ) in MC3T3-EI cell extracts (50-pg protein) at near confluence or after 24 (confluence) or 48 (post-confluence) additional hours of culture under control or shaken conditions. Lane 1: Molecular weight standards (myosin: 199-kDa; galactosidase: 133kDa; BSA: 87-kDa; carbonic anhydrase: 40-kDa. soybean trypsin inhibitor: 32-kDa; lysosyme: 19-kDa; and aprotinin: 7-kDa). Equivalence of loading was confirmed by Coomassie blue staining (results not shown).
P97 Fraction Shaking
-
C
C
+
NP
NP
+
Fig. 4. Western blot showing the steady-state constitutive distribution of VCP in the cytoplasm and nucleus of control and shaken MC3T3El cells after 24 and 48 h of treatment. Distribution of markers: 96.7% f 1.1% (n = 4) of the LDH was found in the cytosolic [C] fraction, while 92.5% f 0.7% (n = 5) of the DNA was found in the nuclear pellet [NP].
presence of 24 discrete peptide fragments identical to those in murine VCP [21], there is little possibility that a related protein, such as the 855 amino acid residue nuclear VCP-like protein or its 307 amino acid residue fragment [5], has been identified. However, these proteins may be present in spots distinct from the one that was analyzed. Definitive localization of VCP in the 2D database of osteoblastic proteins should facilitate comparative proteomic analyses in the future. Immunofluorescent staining of intact cells and classical Western blotting showed that the level of VCP is remarkably stable in the cytosol and nuclei of MC3T3-
El osteoblastic cells at confluence and following induction of mild physiological stress for 24 or 48 h in postconfluent cells. Stable constitutive expression of VCP in MC3T3-El cells differs significantly from that observed in osteosarcoma cells, where VCP declines to very low levels at confluence [l], or in yeast, where VCP levels are cell cycle dependent [lo]. VCP is immunolocalized to the cytosol, cytoskeleton, perinuclear Golgi, and nuclei of confluent MC3T3-E1 osteoblastic cells, but appears to be confined primarily to the cytoplasm of preconfluent, but not post-confluent, Dunn osteosarcoma cells [l]. VCP is abundant in the nucleus and ER membranes of epidermoid carcinoma and COS-7 kidney cells [26]. Endothelial and well-differentiated gastric carcinoma cells exhibit strong cytoplasmic VCP immunostaining that is absent in undifferentiated tumor cells [28]. VCP is present at high concentrations in the endomysial capillaries and lipofuscin granules of normal muscle cells, where cytoplasmic levels of VCP are low [25]. In contrast, VCP is localized in both large and small focal inclusions in the muscle tissue of IBMPFD patients [25]. Thus, VCP appears to be more widely distributed in MC3T3-El cells than it is in the other mammalian cell types examined to date. The high levels of stable constitutive expression and complex subcellular distribution of VCP in MC3T3-El cells are consistent with significant roles for VCP in its previously defined diverse functions in suppression of apoptosis [1,3], ERAD of aberrant proteins in the secretory pathway [18]; protein transport from the ER (endoplasmic reticulum) to the cytosol [29]; membrane fusion during Golgi, ER, and nuclear membrane assembly [11,221; receptor-mediated endocytosis and Golgi sorting [ 171; and transcriptioncoupled DNA repair [31] in these cells. VCP contains three major domains, including two 250-amino acid residue ATPase domains (D1 and D2), each of which contains the Walker A and B motifs required for ATP binding and hydrolysis, as well as an amino-terminal domain of approximately 200 amino acids that binds substrates, such as polyubiquitinylated proteins [24]. The entire molecule is required for 26 S proteasome binding and UPP degradation [19]. Lossof-function missense mutations clustered in the N-terminal polyubiquitin-binding domain or in the D1 domain of VCP cause IBMPFD [25], a disorder characterized by coarse trabeculation, cortical thickening, and spotty sclerosis in the spine, pelvis, shoulder, and skull [8].
K. Behnatn el al. I Journul of Orthopedic Research 23 (2005) 618--624
VCP RNA interference in HeLa cells is associated with the accumulation of polyubiquitinylated proteins in dispersed aggregates, cytosolic vacuolization, cell cycle progression blockage, mitotic abnormalities, and apoptosis [27]. However, these features were not reported for mineralized tissue from IBMPFD patients [8,25], suggesting that redundant biochemical pathways may reduce the severity of the effects of VCP mutations in bone. Mutations in the ubiquitin-binding domain of sequestosome I (SQSTM 1) cause autosomal dominant Paget’s disease of bone 3 (PDB3) [6,9], and it has been suggested that both PDB3 and IBMFPD are characterized by compromised ubiquitin-binding and targeting of similar cellular pathways or proteins [25]. However, since both up-regulation of VCP and impaired VCP function are associated with dysregulated over-growth of osteoblastic cells (e.g., osteosarcoma and Paget’s disease of bone), it is clear that additional research will be required to identify VCP binding partners and UPP substrates in osteoblasts and other bone cells and to determine their roles in signal transduction and cellular physiology in normal and pathological states. Acknowledgments We acknowledge the expert technical contributions and invaluable advice of Drs. Nancy Kendrick (Kendrick Laboratories), Mary Ann Gawinowicz (Columbia University), and Raffi Vartanian (Sepulveda). This work was supported, in part, by the Geriatric Research, Education, and Clinical Centers and the Research Service of the Department of Veterans Affairs (SSM, EJB). References [I] Asai T, Tomita Y, Nakatsuka S, et al. VCP (p97) regulates NF-KB signaling pathway, which is important for metastasis of osteosarcoma cell line. Jpn J Cancer Res 2002;93:296304. [2] Behnam K, Murray S, Whitelegge J, Brochmann E. Identification of the molecular chaperone alpha B-crystallin in demineralized bone powder and osteoblast-like cells. J Orthop Res 2002;20: 1190-6. [3] Dai R, Li C. VCP is a multi-ubiquitin chain targeting factor required in ubiquitin-proteasome degradation. Nature Cell Biol 200 1;3:7404 [4] Ficarro S, Chertihin 0, Westbrook VA, et al. Phosphoproteome analysis of capacitated human sperm. J Biol Chem 2003;278: 1 1579-89. [5] Germain-Lee E, Obie C, Valle D. NVL: a new member of the AAA family of ATPases localized to the nucleus. Genomics 1997;44:22-34. [6] Hocking L, Lucas G, Daroszewska A, et al. Domain-specific mutations in sequestosome 1 (SQSTMI) cause familial and sporadic Paget’s disease. Hum Mol Genet 2002;11:2735-9. [7] Jessop L, Rawlinson S, Pitsillides A, Lanyon L. Mechanical strain and fluid movement both activate extracellular regulated kinase (ERK) in osteoblast-like cells but via different signaling pathways. Bone 2002;31:18694.
623
[8] Kimonis V. Kovach M, Waggoner B, et al. Clinical and molecular studies in a unique family with autosomal dominant limbgirdle muscular dystrophy and Paget disease of bone. Genet Med 2000;2:23241. [9] Laurin N, Brown J, Morissette J, et al. Recurrent mutation of the gene encoding sequestosome 1 in Paget disease of bone. Am J Hum Genet 2002;70:1582-8. [lo] Madeo F, Schlauer J, Zischka H, et al. Tyrosine phosphorylation regulates cell cycle-dependent nuclear localization of Cdc48p. Mol Biol Cell 1998;9:13141. [Ill Meyer HH, Yang Y, Warren G. Direct binding of ubiquitin conjugates by the mammalian p97 adaptor complexes, p47 and Udfl-Npl4. EMBO J 2002;21:564-52. [12] Murray E, Bentley G, Grisanti M, Murray S. The ubiquitinproteasome system and cellular proliferation and regulation in osteoblastic cells. Exp Cell Res 1998;242:460-9. 131 Murray E, Murray S, Manolagas S. 2D gel autoradiographic analyses of the effects of 1,25-dihydroxycholecalciferolon protein synthesis in clonal rat ROS cells. Endocrinology 1990: 126: 2679-92. 141 OConnell K, Stults J. Identification of mouse liver proteins on 2D electrophoresis gels by MALDI MS of in situ enzymatic digests. Electrophoresis 1997;18:349-59. 151 OFarrell PH. High-resolution 2D electrophoresis of proteins. J Biol Chem 1975;250:4007-21. [I61 Ogasawara A, Arakawa T, Kaneda T, et al. Fluid shear stressinduced cyclooxygenase-2 expression is mediated by ClEBP p. CAMP-response element-binding protein, and AP-I in osteoblastic MC3T3-EI cells. J Biol Chem 2001;2767048-54. [I71 Pleasure I, Black M, Keen J. VCP is a ubiquitous clathrin-binding protein. Nature 1993;365:459-62. [18] Rabinovich E, Kerem A, Frohlich K, et al. AAA-ATPase p97l Cdc48p, a cytosolic chaperone required for ER-associated protein degradation. Mol Cell Biol 2002;22:62634. [I91 Song C, Wang Q. Li C. ATPase activity of p97-VCP. D2 mediates the major enzyme activity, and DI contributes to the heat-induced activity. J Biol Chem 2003;278:3648-55. [20] Storrie B, Madden EA. Isolation of subcellular organelles. In: Deutscher MP. Guide to protein purification. Method Enzymol 1990;182-220. [21] Strausberg R, Feingold E, Grouse L, et al. Generation and initial analysis of more than 15,000 full-length human and mouse cDNA sequences. Proc Natl Acad Sci USA 2002;99:16899-903. [22] Uchiyama K, Jokitalo E, Kano F, et al. VCIP135, a novel essential factor for p97lp47-mediated membrane fusion, is required for Golgi and ER assembly in uiuo. J Cell Biol 2002; 159:855-66. [23] Volker C, Lupas A. Molecular evolution of proteasomes. The proteasome-ubiquitin protein degradation pathway. New York: Springer; 2002. [24] Wang Q , Song C, Li C. Hexamerization of p97-VCP is promoted by ATP binding to the DI domain and required for ATPase and biological activities. Biochem Biophys Res Commun 2003:300: 25340. [25] Watts G, Wymer J, Kovach M, et al. Inclusion body myopathy associated with Paget disease of bone and frontotemporal dementia is caused by mutant VCP. Nat Genet 2004;36:37781. [26] Waugh M, Minogue S, Anderson J, et al. Localization of a highly active pool of type I1 PI4K in a p97NCP-rich fraction of the ER. Biochem J 2003;373:57-63. [27] Wojcik C, Yano M, DeMartino G. RNA interference of VCP reveals multiple cellular roles linked to ubiquitinlproteasomedependent proteolysis. J Cell Sci 2004;117:281-92. [28] Yamamoto S, Tomita Y, Hoshida Y, et al. Expression level of VCP is strongly associated with progression and prognosis of gastric carcinoma. J Clin Oncol 2003;21:25374.
624
K. Behnum et al. I Journal of Orthopaedic Research 23 (2005) 618424
[29] Ye Y, Meyer H, Rapoport T. The AAA ATPase Cdc48/p97 and its partners transport proteins from the ER to the cytosol. Nature 200 1;4 14:652-6. [30] Zalk R, Shoshan-Barmatz V. ATP binding sites in brain p97NCP (valosin-containing protein), a multifunctional AAA ATPdse. Biochem J 2003:374473-80.
[31] Zhang H, Wang Q, Kajino K, et al. VCP, a weak ATPase involved in multiple cellular events, interacts physically with BRCAI in the nucleus of living cells. DNA Cell Biol 2000;19: 253-63.