View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

Virology 386 (2009) 1–5

Contents lists available at ScienceDirect

Virology

journal homepage: www.elsevier.com/locate/yviro

Rapid Communication Proteomic analysis reveals the as cellular target for the human papillomavirus type 8

Baki Akgül a,b,⁎, Paola Zigrino c, David Frith d, Sarah Hanrahan d, Alan Storey a,1

a Institute of Cell and Molecular Science, Skin Tumour Laboratory, Cancer Research UK, 4 Newark Street, London E12AT, UK b Institute of Virology, University of Cologne, Fürst-Pückler-Strasse 56, 50935 Cologne, Germany c Department of Dermatology, Kerpener Strasse 62, University of Cologne, 50937 Cologne, Germany d Analysis Laboratory, London Research Institute, Cancer Research UK, 44 Lincoln's Inn Fields, London WC2A 3PX, UK

article info abstract

Article history: Recent studies strongly support a role of human papillomavirus type 8 (HPV8) in non-melanoma skin cancer Received 26 August 2008 development. In this study, a quantitative two-dimensional (2D) differential expression (DiGE) gel Returned to author for revision approach combined with mass spectrometry has been used to identify that are abundantly 25 September 2008 deregulated in primary human epidermal expressing HPV8 sequences. Twenty six protein spots Accepted 16 January 2009 showed significant changes in the level of expression between keratinocytes expressing E7 or the complete Available online 24 February 2009 early region (CER) of HPV8 compared to extracts from cells lacking HPV8 . No differences Keywords: between HPV8 E7 alone and HPV8 CER expressing cells were observed. The 26 protein spots that were HPV8 differentially expressed corresponded to 20 different proteins, of which 14 actin-associated proteins were Actin cytoskeleton downregulated except for calponin-2, which was the only actin-binding protein that was overexpressed. transformation Besides changes in actin modulating proteins, an upregulation of (CK) 5, 6 and 14 was also noted. This study suggests that the actin and cytoskeleton modulating proteins are targets for HPV8. © 2009 Elsevier Inc. All rights reserved.

Human papillomaviruses (HPVs) are strictly epitheliotropic, with of the human keratin-14 promoter, where animals developed specific viral types infecting stratified at defined body sites. spontaneous skin tumours without any exogeneous treatment with Epidemiological and experimental evidence has defined a role for HPV physical or chemical carcinogens (Schaper et al., 2005). Overall in the development of specific human cancers, in particular cervical however, there is still little information that exists concerning cancer (zur Hausen, 2002). Several lines of evidence contributed to the modulation of cellular protein expression pattern by HPV8 causing hypothesis that HPV plays also a role in non-melanoma skin cancer keratinocyte transformation and invasion (reviewed in (Akgül et al., development (reviewed in (Majewski and Jablonska, 2006; Pfister, 2006)). 2003). However, in contrast to the body of knowledge describing the Advances in proteomic technologies are providing tools to help roles for alpha-HPV early proteins involved in cervical carcinogenesis unravel the pathways involved in cancer development. Proteomic (Munger and Howley, 2002), comparatively little is known about the analysis, utilizing high-resolution 2-D coupled with MS, is a powerful functions of the equivalent proteins derived from beta-HPV and their means to identify differential protein expression between biological possible roles in cell transformation and the development of non- samples. A recent advance has come from the introduction of CyDye melanoma skin cancer (NMSC). Recently, it has been demonstrated fluorophores for prelabeling of protein samples. 2-D DiGE (differential that the E6 and E7 proteins of the beta-HPV type 38 immortalize gene expression) technology adds a quantitative component to human primary cells in vitro (Caldeira et al., 2003) and experiments in conventional 2-D gel analysis, allowing for comparison of protein organotypic skin culture model systems have demonstrated the ability expression changes across multiple samples simultaneously without of the high-risk beta-HPV type 8-E7 to induce human keratinocyte gel-to-gel variation, and hence with greater statistical confidence invasion (Akgül et al., 2005). Further evidence supporting a direct role (Tonge et al., 2001; Von Eggeling et al., 2001). for HPV in NMSC development was provided by transgenic mice We hypothesized that expression of HPV8 early region may expressing the complete early region (CER) of HPV8 under the control alter the expression profile of proteins that play a role in keratinocyte transformation and contribute towards tumourigenesis in HPV- associated lesions. To investigate this, in this study we analysed ⁎ Corresponding author. Institute of Virology, University of Cologne, Fürst-Pückler- changes in abundantly expressed proteins in HPV8-E7 and HPV8-CER- Str. 56, 50935 Cologne, Germany. Fax: +49 221 4793904. expressing cells compared to non-HPV8-expressing cells by 2-D DiGE E-mail address: [email protected] (B. Akgül). 1 Present address: Molecular Oncology Department, University of Oxford, John (Fig. 1A). In our experiments, we used primary human adult skin Radcliffe Hospital, Headington, Oxford OX3 9DS, UK. keratinocytes, which are the natural target cells of cutaneous HPVs

0042-6822/$ – see front matter © 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.virol.2009.01.036 2 Rapid Communication

Fig. 1. (A) 2D gel electrophoresis pattern of proteins from adult keratinocytes retrovirally transduced for HPV8-E7, HPV8-CER or empty vector. For the first dimension IEF, 24 cm pH 3–10 (linear) Immobiline DryStrip (IPG) strips (GE Healthcare) were cuploaded with 150 μg of CyDye labeled proteins (50 μg of each fluor Cy2, Cy3 and Cy5 — for sample combinations see Table 1). Isoelectric focusing was performed on an Ettan IPGphor IEF System using the Ettan IPGphor Cup Loading Manifold (both GE Healthcare). Equilibrated IPG strips were transferred onto 1 mm thick 25×20 cm 12% polyacrylamide gels without a stacking gel, cast between low-fluorescence glass plates. Gels were electrophoresed in an Ettan Dalt six gel tank (GE Healthcare) at 20 °C, at 1.5 W/gel constant power overnight. Following SDS-PAGE, fluorescently labeled proteins were visualized while still between the low-fluorescence glass plates using a Typhoon 9400 Variable Mode Imager and Typhoon Scanner Control Software, Version 3.0 (GE Healthcare). For the Cy2 images a 488 nm laser and an emission filter of 520 nm with a band pass (BP) of 40 nm was used. Cy3 images were obtained using a 532 nm laser and an emission filter of 580 nm/BP30. Cy5 images were obtained using a 633 nm laser and an emission filter of 670 nm/BP30. (B) Protein association network of the proteins differentially expressed due to HPV8 expression based on STRING Network Display. Protein–protein interaction prediction from the identified proteins. Stronger associations are represented by thicker lines. ACTA1: Actin, alpha skeletal muscle; KRT14: Keratin-14; VCL: ; KRT5: Keratin-5; IMMT: Mitochondrial inner (Mitofilin); CNN2: Calponin-2; ENSP00000264785: WD repeat domain 1 (WDR1); MSN: Moesin; GSN: ; FLNB: B; HSPA4: Heat shock 70; TBCA: -specific chaperone A TCP1; PDCD6IP: Programmed cell death 6-interacting protein; LMNA: A/C; RDX: Radixin; TPM1: -1; TPM3: Tropomyosin-3; TPM4: Tropomyosin-4. that were retrovirally infected with pLXSN-based recombinant retro- 3–10) containing a protease inhibitor cocktail (Complete Mini, Roche) viruses coding for HPV8-E7 or the HPV8-CER. Cells infected with the and endonuclease (Benzonase, Merck) for 1 h at RT. To remove empty retrovirus pLXSN were used as controls. The construction of unsolubilised cell debris the cell lysates were centrifuged at 4 °C for retrovirus pLXSN-8-E7 have been described in Akgül et al. (2005) and 5 min at 14,000×g. Prior to labeling with CyDye DIGE Fluors, all the construction of pLXSN-8-CER and RT-PCRs demonstrating that the protein samples were cleaned using the 2-D Clean-up kit (GE HPV8 early genes E1, E2, E6 and E7 are expressed by this construct has Healthcare). The resulting pellets were resolubilised in DiGE labeling been reported in Akgül et al. (2007). Cells were infected with the buffer (7 M urea, 2 M thiourea, 4% (w/v) CHAPS, 30 mM Tris–Cl, pH different recombinant retrovirus in triplicate, selected with G418, 8.5) to give a final protein concentration of between 5 and 10 mg/ml. expanded in passage 2 and pooled separately to generate 3 biological Experiments were performed according to the Ettan DIGE System User replicates for each type of infection. Cells were lysed in 2-D lysis buffer Manual. Analysis of gel images was performed using DeCyder (7 M urea, 2 M thiourea, 4% CHAPS, 65 mM DTT, 1% pharmalytes pH Differential Analysis Software Version 5.0 (GE Healthcare). The Rapid Communication 3

Biological Variation Analysis (BVA) module was used to compare were down regulated in HPV8-E7 and HPV8-CER cells when compared abundant changes across the three control and 6 test samples. to keratinocytes lacking HPV gene expression. DeCyder also performs Student's t-tests to determine statistically Tropomyosins are a family of actin-binding proteins that stabilize significant changes, as well as calculating average abundant changes microfilaments and are expressed with a high degree of tissue between the three sample populations. Approximately 2500 spots specificity (Pawlak and Helfman, 2001). Tropomyosin 1 reorganizes were detected on each gel and 1200 of these were matched on all gels. actin filaments and functions as a suppressor of malignant transfor- Only 100 protein spots, represented in all gels (n=9) with expression mation (Mahadev et al., 2002) and the down-regulation by HPV8 E7 level greater than 1.25-fold and pb0.05, were defined as being noted here may therefore contribute towards malignant progression. differentially expressed. Preparative gels for spot picking and Gelsolin is a calcium-regulated actin-binding protein, which functions subsequent protein identification by mass spectrometry were pre- to regulate actin dynamics both by capping microfilament ends and pared by running 400–500 μg of protein on 2-D gels. Due to the higher inducing actin depolymerization, as well as serving to nucleate new protein load, samples were applied to the first dimension strips using growth of actin filaments (Kwiatkowski, 1999). Knockdown of gelsolin an in-gel rehydration method. Samples were diluted in rehydration has been described to induce epithelial–mesenchymal transition, a solution (8 M Urea, 0.5% CHAPS, 65 mM DTT, 0.5% pharmalyte pH 3– process occurring during oncogenesis by which epithelial cells obtain 10). Protein spots were visualized using colloidal Coomassie blue and a fibroblast-like properties and show reduced cell adhesion (Tanaka digital image of the preparative gel was imported into BVA (Biological et al., 2006). Calponin-2 is the only actin associated protein over- Variation Analysis) workspace and matched to the analytical gels. expressed in HPV8 positive keratinocytes. Previous studies suggested Twenty-six distinct spots were able to be visualized and these were that Calponin-2 may play a role in the organization of the actin manually excised from the gels, digested with Trypsin and analysed as cytoskeleton (Danninger and Gimona, 2000). Interestingly calponin-2 described by Craven et al. (2006). In addition, electrospray mass has been shown to be over-expressed in human cutaneous SCC (Dang spectra were recorded using a 4000 Q-TRAP mass spectrometer et al., 2006). Notably also is that actin is also present in the nucleus, (Applied Biosystems) interfaced to an additional LC Packings Inte- where it is involved in activities including chromatin remodelling, grated nano-lc system. stress responses, nuclear export and transcription (Pederson and Aebi, The 26 protein spots showing significant change in the level of 2002). In the nucleus it can interact with lamin A/C, proteins which are expression between sample and control extracts, corresponded to 20 components of the lamina network with roles in DNA organization different proteins as identified by mass spectrometry (Table 1). Based and replication (Spann et al., 1997). Lamin A/C expression has been on STRING protein–protein interaction predictions, an association shown to be downregulated in human cutaneous SCC (Dang et al., network of the proteins, which have their expression modified due to 2006). It is possible that E7 proteins of other viral types may also HPV8 expression, was created (Fig. 1B). This bioinformatics analysis increase calponin-2 and alter actin expression and that this may indicated that the major group of proteins identified represent actin- account for the observed differences in SCC. associated proteins. Actin is one of the main components of the The third group of differentially expressed proteins comprises cytoskeleton in all cell types and is required for cell division, motility, cytokeratins (CK) CK5, CK6 and CK14. Keratin filaments are a class of mechanical resilience and cell integrity. One of the hallmarks of intermediate filaments that is selectively expressed in epithelia. The malignant transformation is the alteration of the cellular cytoskeleton, basal cells of all stratified squamous epithelia express CK5 and CK14 resulting in decreased adhesion and modified adhesion-dependent (Purkis et al., 1990). Under hyperproliferative conditions epidermal responses (Ben-Ze'ev, 1997). These alterations involve structural keratinocytes express CK6 (Kopan and Fuchs, 1989). The expression of components of the cytoskeleton or proteins that interact with these these cytokeratins was increased in HPV8 expressing cells when structures (Janmey and Chaponnier, 1995). Since the actin cytoskele- compared to the control cells infected with empty retrovirus, ton is controlled by its binding proteins (dos Remedios et al., 2003), suggesting that the virus induces a hyperproliferative, basaloid state the downregulation of actin and actin-binding-proteins would be which confirms our previous observation of specific being predicted to facilitate neoplastic conversion. Consistent with this overexpressed in HPV8 positive keratinocytes (Akgül et al., 2007). Not hypothesis, all actin-modulating proteins identified in this study, with associated with the cellular microfilaments are mitofilin and vinculin. the exception of Calponin 2, and discussed further in the text below Mitofilin is an integral protein of the inner mitochondrial membrane

Table 1 Identification of picked protein spots by mass spectroscopy and description of their function towards actin

Spot no. T-test Av. ratio Protein Mol. weight Accession no. Function towards actin 1/6 0.0000012 +2.71 51.6 NP_000517 Intermediate filament 2 0.00064 −1.74 Tropomyosin 1 32.8 AAH07433 Actin binding and stabilization 3 0.0016 −1.82 Tropomyosin 4 28.5 NP_999500 Actin binding and stabilization 4 0.0015 −1.86 29.0 NP_705935 Actin binding and stabilization 5 0.02 −1.29 alpha-Actin 32.0 CAI19051 Component of the actin cytoskeleton 7 0.026 −1.54 Heat shock protein 70 70.0 NP_005337 Modulation of actin formation and dynamics 8/9/10 0.000095 −1.69 Mitofilin 79.7 BAA04654 Inner mitochondrial membrane protein 11 0.0017 −1.54 Gelsolin 85.7 NP_000168 calcium-regulated actin binding protein 12/13 0.00088 −1.61 Moesin 67.8 NP_002435 Links actin to plasmamembrane 14 0.0037 −1.60 Radixin 68.6 NP_002897 Links actin to plasmamembrane 15 0.0013 −1.48 Vinculin 116.7 NP_003364 Associates with cell–cell junctions 16/17 0.00017 −1.47 Filamin A 104.2 AAH67111 Crosslinking of actin filaments 18 0.00070 −1.42 Programmed cell death 6 interacting protein (PDCD6IP) 96.0 NP_037506 depletion causes accumulation of unusual actin structures 19 0.00016 −1.44 Filamin B 71.2 AAF97046 Crosslinks actin filaments 20 0.00084 −1.67 Tryptophan-aspartate-repeat protein-1 (WDR1) 58.0 AAD05045 Actin binding, facilites actin turnover 21 0.0058 −1.55 Lamin A/C 65.0 CAI15521 Interacts with actin in the nucleus 22 0.0064 −1.36 Chaperonin TCP1 57.5 NP_006422 Chaperone for actin folding 23/24 0.000055 +4.99 62.5 P13647 Intermediate filament 25 0.00011 +1.79 Keratin 6 60.0 AAH69269 Intermediate filament 26 0.00022 +1.72 Calponin 2 33.7 NP_004359 Organization of actin cytoskeleton 4 Rapid Communication whose downregulation results in a drastic change in the organization of the inner mitochondrial membrane (John et al., 2005). But its role in pathophysiology of human diseases is still unclear. Vinculin is the only cytoskeletal protein found linked with cell–cell and cell–matrix junctions (Rudiger, 1998). Vinculin null cells are highly metastatic (Lifschitz-Mercer et al., 1997), motile (Xu et al., 1998a, 1998b) and show reduced susceptibility to apoptosis (Subauste et al., 2004). To further verify 2D-DIGE results we performed Western blot analysis on total cell extracts from control cells and keratinocytes expressing HPV8-E7 lysed in RIPA buffer. The blots were probed with antibodies to three selected altered proteins, Tropomyosin-1 (rabbit polyclonal to Tropomyosin-1, Abcam, Cambridge, UK), alpha-actin (clone C4, Chemicon, Temecula, CA, USA), Keratin-5 (clone XM26, Diagnostic BioSystems, Pleasanton, CA, USA) and GAPDH (clone mAbcam9484, Abcam, Cambridge, UK) as loading control. Fig. 3 shows the result of this experiments confirming the measured expression level changes in these extracts. The observation that most of the dysregulated proteins have been implicated in the organization of the actin filaments suggested that HPV8-E7 induced changes in cell cytoskeleton occurred. To address this Fig. 3. Western blots showing the protein expressions of alpha-, alpha-actin, we analysed actin cytoskeleton organization of HPV8-E7 expressing Tropomyosin-1, Keratin-5 and GAPDH in total cell extracts of adult keratinocytes primary keratinocytes and compared this to control keratinocytes retrovirally transduced for HPV8-E7 or empty vector pLXSN. infected with empty retrovirus. After adhesion of the cells in SFM to glass slides for 4 h, cells were fixed 10 min with 4% PFA in PBS and the actin striated pattern that crossed the cell body as well as bundles present in cytoskeleton was visualized by using Cy3-phalloidin (10 μg/ml; Sigma) the cell periphery. In contrast in HPV8-E7 expressing cells fewer fibers for 1 h at 37 °C. Nuclei were counterstained by DAPI (1 μg/ml). were formed and a diffuse staining pattern was observed in the central Interestingly, in control cells actin formed long stress fibers with a part of the cell. In addition, those fibers that were formed in the HPV8 E7 cells were more diffusely packed than those seen in control cells and were more disorganized (Fig. 2). To further confirm the observation, that actin fibers are not correctly formed in HPV8-E7 cells we also performed Western blot for the abundant actin modulating protein alpha-actinin, which bundles and cross-links actin filaments into stress fibers (Naumanen et al., 2008). The fact, that less alpha-actinin is detected in extracts of E7 cells by the anti alpha-actinin antibody (clone H2, Santa Cruz, Heidelberg, Germany) (Fig. 3) further supports the observation that actin fibers are diffusely organized in HPV8 positive cells. Overall, our data is in line with observations made with alpha-HPV type 16, where the amount of alpha-actin was decreased in cells expressing HPV16-E7 (Nishida et al., 1995; Rey et al., 2000). It is interesting to note that no difference was seen between the 2-D DIGE images of cells expressing either HPV8-E7 alone or in context of the complete early region with simultaneous expression of E1, E2 and E6 indicating E7 as the only viral gene modulating the expression of cytoskeletal proteins of primary keratinocytes. However, we cannot rule out the possibility that other differentially expressed proteins, when comparing E7 to CER expressing cells, may be at a level below the detection sensitivity of our experiments. Since the cytoskeletal network plays an important role in cell shape and cell locomotion, which in turn is thought to be involved in growth control, invasion and , changes in the intermediate filaments actin and keratins may contribute to the transformed cell induced by HPV8. The changes in cellular protein expression brought about by HPV8 await further detailed analysis and the precise manner by which a disrupted actin cytoskeleton can induce neoplastic conversion of virus infected cells in human skin tumours remains to be determined.

Acknowledgments

This study was supported by grants from the Deutsche For- schungsgemeinschaft (PF123/5-1) and Cancer Research UK.

References

Akgül, B., Garcia-Escudero, R., Ghali, L., Pfister, H.J., Fuchs, P.G., Navsaria, H., Storey, A., Fig. 2. Actin cytoskeletal organization in HPV8-E7 and control cells. The actin 2005. The E7 protein of cutaneous human papillomavirus type 8 causes invasion of cytoskeleton is visualized by using Cy3-phalloidin (red staining). Nuclei were counter- human keratinocytes into the dermis in organotypic cultures of skin. Cancer Res. 65 stained by DAPI (blue staining). (6), 2216–2223. Rapid Communication 5

Akgül, B., Cooke, J.C., Storey, A., 2006. HPV-associated skin disease. J. Pathol. 208 (2), Nishida, M., Miyamoto, S., Kato, H., Miwa, T., Imamura, T., Miwa, K., Yasumoto, S., Barrett, 165–175. J.C., Wake, N., 1995. Transcriptional repression of smooth-muscle alpha-actin gene Akgül, B., Ghali, L., Davies, D., Pfister, H., Leigh, I.M., Storey, A., 2007. HPV8 early genes associated with human papillomavirus type 16 E7 expression. Mol. Carcinog. 13 (3), modulate differentiation and cell cycle of primary human adult keratinocytes. Exp. 157–165. Dermatol. 16 (7), 590–599. Pawlak, G., Helfman, D.M., 2001. Cytoskeletal changes in cell transformation and Ben-Ze'ev, A., 1997. Cytoskeletal and adhesion proteins as tumor suppressors. Curr. tumorigenesis. Curr. Opin. Genet. Dev. 11 (1), 41–47. Opin. Cell Biol. 9 (1), 99–108. Pederson, T., Aebi, U., 2002. Actin in the nucleus: what form and what for? J. Struct. Biol. Caldeira, S., Zehbe, I., Accardi, R., Malanchi, I., Dong, W., Giarre, M., de Villiers, E.M., 140 (1–3), 3–9. Filotico, R., Boukamp, P., Tommasino, M., 2003. The E6 and E7 proteins of the Pfister, H., 2003. Chapter 8: Human papillomavirus and skin cancer. J. Natl. Cancer Inst. cutaneous human papillomavirus type 38 display transforming properties. J. Virol. Monographs (31), 52–56. 77 (3), 2195–2206. Purkis, P.E., Steel, J.B., Mackenzie, I.C., Nathrath, W.B., Leigh, I.M., Lane, E.B., 1990. Craven, R.A., Hanrahan, S., Totty, N., Harnden, P., Stanley, A.J., Maher, E.R., Harris, A.L., Antibody markers of basal cells in complex epithelia. J. Cell. Sci. 97 (Pt 1), Trimble, W.S., Selby, P.J., Banks, R.E., 2006. Proteomic identification of a role for the 39–50. von Hippel Lindau tumour suppressor in changes in the expression of mitochon- Rey, O., Lee, S., Baluda, M.A., Swee, J., Ackerson, B., Chiu, R., Park, N.H., 2000. The E7 drial proteins and septin 2 in renal cell carcinoma. Proteomics. oncoprotein of human papillomavirus type 16 interacts with F-actin in vitro and in Dang, C., Gottschling, M., Manning, K., O'Currain, E., Schneider, S., Sterry, W., Stockfleth, vivo. Virology 268 (2), 372–381. E., Nindl, I., 2006. Identification of dysregulated genes in cutaneous squamous cell Rudiger, M., 1998. Vinculin and alpha-: shared and unique functions in adherens carcinoma. Oncol. Rep. 16 (3), 513–519. junctions. BioEssays 20 (9), 733–740. Danninger, C., Gimona, M., 2000. Live dynamics of GFP-calponin: isoform-specific Schaper, I.D., Marcuzzi, G.P., Weissenborn, S.J., Kasper, H.U., Dries, V., Smyth, N., Fuchs, P., modulation of the actin cytoskeleton and autoregulation by C-terminal sequences. Pfister, H., 2005. Development of skin tumors in mice transgenic for early genes of J. Cell. Sci. 113 (Pt 21), 3725–3736. human papillomavirus type 8. Cancer Res. 65 (4), 1394–1400. dos Remedios, C.G., Chhabra, D., Kekic, M., Dedova, I.V., Tsubakihara, M., Berry, D.A., Spann, T.P., Moir, R.D., Goldman, A.E., Stick, R., Goldman, R.D., 1997. Disruption of nuclear Nosworthy, N.J., 2003. Actin binding proteins: regulation of cytoskeletal microfila- lamin organization alters the distribution of replication factors and inhibits DNA ments. Physiol. Rev. 83 (2), 433–473. synthesis. J. Cell Biol. 136 (6), 1201–1212. Janmey, P.A., Chaponnier, C., 1995. Medical aspects of the actin cytoskeleton. Curr. Opin. Subauste, M.C., Pertz, O., Adamson, E.D., Turner, C.E., Junger, S., Hahn, K.M., 2004. Cell Biol. 7 (1), 111–117. Vinculin modulation of paxillin–FAK interactions regulates ERK to control survival John, G.B., Shang, Y., Li, L., Renken, C., Mannella, C.A., Selker, J.M., Rangell, L., Bennett, and motility. J. Cell Biol. 165 (3), 371–381. M.J., Zha, J., 2005. The mitochondrial inner membrane protein mitofilin controls Tanaka, H., Shirkoohi, R., Nakagawa, K., Qiao, H., Fujita, H., Okada, F., Hamada, J., cristae morphology. Mol. Biol. Cell 16 (3), 1543–1554. Kuzumaki, S., Takimoto, M., Kuzumaki, N., 2006. siRNA gelsolin knockdown induces Kopan, R., Fuchs, E., 1989. The use of retinoic acid to probe the relation between epithelial–mesenchymal transition with a cadherin switch in human mammary hyperproliferation-associated keratins and cell proliferation in normal and epithelial cells. Int. J. Cancer 118 (7), 1680–1691. malignant epidermal cells. J. Cell Biol. 109 (1), 295–307. Tonge, R., Shaw, J., Middleton, B., Rowlinson, R., Rayner, S., Young, J., Pognan, F., Hawkins, Kwiatkowski, D.J., 1999. Functions of gelsolin: motility, signaling, apoptosis, cancer. E., Currie, I., Davison, M., 2001. Validation and development of fluorescence two- Curr. Opin. Cell Biol. 11 (1), 103–108. dimensional differential gel electrophoresis proteomics technology. Proteomics 1 Lifschitz-Mercer, B., Czernobilsky, B., Feldberg, E., Geiger, B., 1997. Expression of the (3), 377–396. adherens junction protein vinculin in human basal and squamous cell tumors: Von Eggeling, F., Gawriljuk, A., Fiedler, W., Ernst, G., Claussen, U., Klose, J., Romer, I., relationship to invasiveness and metastatic potential. Hum. Pathol. 28 (11),1230–1236. 2001. Fluorescent dual colour 2D-protein gel electrophoresis for rapid detection of Mahadev, K., Raval, G., Bharadwaj, S., Willingham, M.C., Lange, E.M., Vonderhaar, B., differences in protein pattern with standard image analysis software. Int. J. Mol. Salomon, D., Prasad, G.L., 2002. Suppression of the transformed phenotype of breast Med. 8 (4), 373–377. cancer by tropomyosin-1. Exp. Cell Res. 279 (1), 40–51. Xu, W., Baribault, H., Adamson, E.D., 1998a. Vinculin knockout results in heart and brain Majewski, S., Jablonska, S., 2006. Current views on the role of human papillomaviruses defects during embryonic development. Development 125 (2), 327–337. in cutaneous oncogenesis. Int. J. Dermatol. 45 (3), 192–196. Xu, W., Coll, J.L., Adamson, E.D., 1998b. Rescue of the mutant phenotype by reexpression Munger, K., Howley, P.M., 2002. Human papillomavirus immortalization and transfor- of full-length vinculin in null F9 cells; effects on cell locomotion by domain deleted mation functions. Virus Res. 89 (2), 213–228. vinculin. J. Cell. Sci. 111 (Pt 11), 1535–1544. Naumanen, P., Lappalainen, P., Hotulainen, P., 2008. Mechanisms of actin stress fibre zur Hausen, H., 2002. Papillomaviruses and cancer: from basic studies to clinical assembly. J. Microsc. 231 (3), 446–454. application. Nat. Rev., Cancer 2 (5), 342–350.