Cell Tissue Res (2013) 351:59–72 DOI 10.1007/s00441-012-1500-y

REGULAR ARTICLE

Epidermal growth factor-induced modulation of cytokeratin expression levels influences the morphological phenotype of head and neck squamous cell carcinoma cells Galina Makarova & Michael Bette & Ansgar Schmidt & Ralf Jacob & Chengzhong Cai & Fiona Rodepeter & Thomas Betz & Johannes Sitterberg & Udo Bakowsky & Roland Moll & Andreas Neff & Andreas Sesterhenn & Afshin Teymoortash & Matthias Ocker & Jochen A. Werner & Robert Mandic Received: 2 September 2011 / Accepted: 7 September 2012 / Published online: 1 November 2012 # Springer-Verlag 2012

Abstract The migratory ability of tumor cells requires cytoskeletal rearrangement processes. Epidermal growth factor receptor (EGFR)-signaling tightly correlates with tumor progression in head and neck squamous cell carcinomas (HNSCCs), and has previously been implicated in the regulation of cytokeratin (CK) expression. In this study, HNSCC cell lines were treated with EGF, and CK expression levels were monitored by Western blot analysis. Changes in cellular morphology were documented by fluorescence- and atomic force microscopy. Some of the cell lines demonstrated an EGF-dependent modulation of CK expression levels. Interestingly, regression

of some CK subtypes or initial up-regulation followed by downregulation at higher EGF-levels could also be observed in the tested cell lines. Overall, the influence of EGF on CK expression levels appeared variable and cell-type-dependent. Real-time cellular analysis of EGF-treated and -untreated HNSCC cell lines demonstrated a rise over time in cellular impedance. In three of the EGF-treated HNSCC cell lines, this rise was markedly higher than in untreated controls, whereas in one of the cell lines the gain of cellular impedance was paradoxically reduced after EGF treatment, which was found to correlate with changes in cellular morphology rather than with relevant changes in

Electronic supplementary material The online version of this article (doi:10.1007/s00441-012-1500-y) contains supplementary material, which is available to authorized users. This study was partly funded by the “Alfred und Ursula Kulemann Stiftung”. G. Makarova : C. Cai : F. Rodepeter : A. Sesterhenn : A. Teymoortash : J. A. Werner : R. Mandic (*) Department of Otorhinolaryngology, Head and Neck Surgery, University Hospital Giessen and Marburg, Campus Marburg, 3. BA, +3/08070, Baldingerstrasse, 35033 Marburg, Germany e-mail: [email protected]

M. Bette Department of Molecular Neuroscience, Institute of Anatomy and Cell Biology, Philipps University, Marburg, Germany A. Schmidt : R. Moll Institute of Pathology, Philipps University of Marburg, Marburg, Germany

R. Jacob Department of Cell Biology and Cell Pathology, Philipps University, Marburg, Germany T. Betz : J. Sitterberg : U. Bakowsky Department of Pharmaceutical Technology and Biopharmacy, Philipps University, Marburg, Germany A. Neff Department of Oral and Maxillofacial Surgery, University Hospital Giessen and Marburg, Campus Marburg, Medical Faculty of Philipps University, Marburg, Germany M. Ocker Institute for Surgical Research, Philipps University, Marburg, Germany

60

cellular viability or proliferation. After treating HNSCC cells with EGF, CK filaments frequently appeared diffusely distributed throughout the cytoplasm, and in some cases were found in a perinuclear localization, the latter being reminiscent to observations by other groups. In summary, the data points to a possible role of EGFR in modulating HNSCC cell morphology. Keywords Head and neck squamous cell carcinoma . Epidermal growth factor receptor . Cytokeratin . Morphology . Human

Introduction For a long time, pathologists and cell biologists have been able to observe that tumor cells, although frequently retaining many phenotypic characteristics of the original tissue, have attained morphological features that are quite different from the tissue of origin. Furthermore, these morphological changes appear to correlate with the ability of tumor cells to invade neighbouring tissues and to metastasize to distant sites (Hanahan and Weinberg 2011). The process of invasion and metastasis is dependent on the tumor cells' ability to cross tight tissue barriers (Friedl and Wolf 2003). Any changes in cellular morphology usually require cytoskeletal rearrangement processes that are driven by programs frequently found over-activated in cancer cells (Ridley et al. 2003). Actin filament assembly and disassembly typically underlies cellular movement processes in normal physiology such as organ development but also during pathological events such as cancer cell migration and invasion (Friedl and Wolf 2003). Head and neck squamous cell carcinomas (HNSCCs) are the most frequent malignancies of the upper aero-digestive tract, accounting worldwide for more than 400,000 deaths each year (Dobrossy 2005). They are characterized by a primarily lymphatic metastatic spread that typically precedes the appearance of distant metastases. The epidermal growth factor receptor (EGFR) gene is amplified in up to 100 % of all HNSCC tumors, and expression levels closely correlate with tumor growth, metastasis, and prognosis of the patient (Weichselbaum et al. 1989; Mandic et al. 2001; Chung et al. 2004, 2006). Early observations found that activation of the EGFR signaling pathway induces overexpression of the hyperproliferative cytokeratins (CKs) 6 and 16 (Jiang et al. 1993). Also, Beil et al. reported that tumor cells treated with the bioactive lipid sphingosylphosphorylcholine (SPC) attained an invasive phenotype due to cytoskeletal CK reorganization (Beil et al. 2003). The aim of the present study, therefore, was to investigate the dependence of CK expression and cellular morphology as a response to EGFR activation in HNSCC cells.

Cell Tissue Res (2013) 351:59–72

Materials and methods Cell lines and cell culture The HNSCC cell lines UM-SCC-3 and UT-SCC-26A were kindly provided by T. Carey (University of Michigan, USA) and R. Grénman (University of Turku, Finland), respectively (Lansford et al. 1999). The UMB-SCC-864 and -969 HNSCC cell lines were as described earlier (Mandic et al. 2005). The identity of the cell lines was confirmed either by genotyping (German Biological Resource Centre Human and Animal Cell Lines, Braunschweig, Germany) or by confirmation of previously published gene mutations (Zhao et al. 2011; Lansford et al. 1999; Mandic et al. 2005). The HaCaT and A431 control cell lines were kindly provided by W.W. Franke (German Cancer Research Center, Heidelberg, Germany). All cell lines were cultured in Dulbecco’s Modified Eagle’s Medium (DMEM) supplemented with 10 % fetal calf serum penicillin and streptomycin as described earlier (Zhu et al. 2007). Cells were incubated at 37 °C, 5 % CO2, 95 % humidity, and were allowed to grow to 90 % confluence prior to incubation (18 h) with EGF (0.1, 1, 10, 100, 1,000 ng/ml). Antibodies The pan-CK mouse monoclonal antibody (clone K8.13), that was used for Western blot analysis, was obtained from SigmaAldrich, Inc. (St. Louis, MO, USA, cat# C6909), and reportedly is able to detect CK1, CK5, CK6, CK7, CK8, CK10, CK11 and CK18 (Gigi et al. 1982). A pan-CK mouse monoclonal antibody (clone MNF116, DAKO, Hamburg, Germany) able to detect CK5, CK6, CK8, CK17 and perhaps CK19 was used for immunofluorescence analysis (Prieto et al. 1996). Mouse monoclonal antibodies directed against CK4 (clone 6B10; Euro Diagnostica AB, Malmö, Sweden, cat# M6B10), CK5 (clone RCK103; Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA, cat# sc-32721), CK6 (clone LHK6B, Lab Vision, Fremont, CA, USA, cat# MS-766), CK7 (clone OVTL 12/30, Monosan®, Uden, Netherlands, cat# MON3014), CK8 (clone Ks8.7, PROGEN Biotechnik GmbH, Heidelberg, Germany, cat# 61038), CK13 (clone 2D7, Monosan®, cat# MON3018), CK14 (clone LL002, Meridian Life Science Inc., Saco, ME, USA, cat# M54197M), CK19/16/14/(+ other acidic type I keratins) (clone AE1, PROGEN Biotechnik GmbH, cat#61804), CK18 (clone Ks18.04, PROGEN Biotechnik GmbH, cat# 65028) and CK19 (clone Ks19.1 (A53-B/A2), PROGEN Biotechnik GmbH, cat# 61010) were used to further characterize cytokeratin subgroups. Polyclonal antibodies directed against plakophilin-3 (PKP3; own production) were produced in rabbits. Briefly, polyclonal antiserum against PKP3 was generated by a standard immunization protocol with a bacterial and affinity-purified recombinant polypeptide (corresponding to amino acids 1-426 of human PKP3) as antigen, and was subsequently affinity-purified as described

Cell Tissue Res (2013) 351:59–72

elsewhere (Hammerl et al. 1993). Mouse monoclonal antibodies against diphosphorylated ERK-1&2 (clone MAPK-YT, cat# M8159) were from Sigma-Aldrich, Inc. Goat polyclonal antibodies directed against phosphorylated Stat3 (p-Stat3 (Tyr 705), cat# sc-7993) as well as the secondary HRP-coupled and IgG control antibodies were from Santa Cruz Biotechnology, Inc.. Mouse monoclonal antibodies against desmoplakin (clone DP 2.17, cat# 61024) were purchased from PROGEN Biotechnik GmbH. Fluorochrome-labeled secondary (Cy2-labeled donkeyanti-mouse, Cy3-labeled goat-anti-rabbit) antibodies were from Dianova (Hamburg, Germany). The anti-β-actin antibody was purchased from Sigma-Aldrich, Inc. (clone AC-74, cat# A5316). For real-time cellular analysis, neutralizing mouse-monoclonal anti-EGF antibodies (clone 10825.1, cat# MAB36) were obtained from R&D Systems Inc. (Minneapolis, MN, USA). SDS-PAGE and Western blot analysis To evaluate protein expression levels, SDS-PAGE and Westernblot analysis were performed with the indicated CK-specific antibodies. HNSCC cell lines were harvested 18 h after EGF treatment, washed twice in cold PBS and resuspended in lysis buffer (1 % Nonidet P40, 137 mM NaCl, 2 mM ethylene diamine tetracetic acid, 20 mM TRIS/HCl (pH7.5), 10 % glycerol), supplemented with protease and phosphatase inhibitors (Sigma). SDS-PAGE was performed under standard conditions as described earlier (Mandic et al. 2005) using a discontinuous 10 % acrylamide gel. Thirty-five μg of whole cell lysate protein was loaded per lane. The Precision Plus Protein™ Standard (161-0373) from Biorad (Hercules, CA, USA) was used for size comparison. After SDS-PAGE, the proteins were transferred to nitrocellulose membranes. Subsequently, the membranes were blocked with 3 % nonfat dry milk/PBS and incubated with primary antibody (pan-CK 0 1:500, CK4 0 1:200, CK5 0 1:200, CK6 0 1:200, CK7 0 1:200, CK8 0 1:1,000, CK13 0 1:200, CK14 0 1:100, CK19/16/14/(+ other acidic type I keratins) 0 1:50, CK18 0 1:200, CK19 0 1:200, Erk-1/2-phospho 0 1:500, p-Stat3 (Y 705) 0 1:200, β-actin 0 1:4,000) overnight at 4 °C. Then, membranes were washed thrice for 10 min in blocking buffer and were incubated with an HRP-coupled secondary antibody (1:2,000) for 1 h at room temperature. Membranes were washed and signals were visualized on X-ray film (Agfa, Cologne, Germany) using the enhanced chemiluminescence (ECL) method (Amersham Biosciences, Buckinghamshire, United Kingdom). One separate nitrocellulose membrane was used for each tested antibody. Subsequently, membranes were tested for β-actin expression that served as an input control. Real-time cellular analysis EGF-dependent real-time cellular impedance was evaluated with the xCELLigence system (Roche, Germany). 5×103 HNSCC cells were suspended in 150 μl DMEM media and

61

added into a specially designed 96-well microtiter plate (E-Plate, Roche, Germany). Twenty-four h after cell plating, 10 ng/ml EGF was added into the medium. To confirm the specificity of the measured EGF effect, an anti EGF-specific antibody was added to the culture media together with EGF (control). In contrast, for measurement of the EGF effect, cells only received an unspecific IgG control antibody instead. Cellular impedance was monitored every 15 s for 6 h, and then every 15 min for the next 72 h or until maximum impedance was reached. XTT assay The XTT (sodium 3’-[1-(phenylaminocarbonyl)-3,4-tetrazolium]-bis (4-methoxy-6-nitro) benzene sulfonic acid hydrate) assay [Cell Proliferation Kit II (XTT), Roche Diagnostics GmbH, Mannheim, Germany] was used to measure proliferation and viability of EGF-treated cells. In short, 25×103 cells suspended in 100 μl regular media, containing the respective concentration of EGF (0, 0.1, 1, 10, 100, 1,000 ng/ml) were added as quadruplicates to a clear 96-well tissue culture plate with flat bottom. Cells were incubated for 18 h at 37 °C, 5 % CO2. Subsequently, 50 μl of XTT labeling mixture containing XTT and the electron coupling reagent PMS (N-methyl dibenzopyrazine methyl sulfate) was added to each well and incubation was continued for 4 more hours at 37 °C, 5 % CO2. Absorbance was measured with a microplate reader (DTX880, Beckman Coulter, Inc., Fullerton, CA, USA) at 450 nm with a reference wavelength of 620 nm. Immunofluorescence analysis For immunofluorescence staining, EGF-treated (10 ng/ml, 18 h) or untreated cells were grown on glass cover slips and fixed for 5 min in pre-cooled (−20 °C) methanol and acetone for 1 min afterwards. Following rehydration of cells in phosphate-buffered saline (PBS)-buffer, primary antibodies were incubated for 1 h in suitable dilutions on cells in a humid chamber. Primary antibodies applied were a monoclonal mouse anti-cytokeratin antibody (clone MNF116, diluted 1:100, DAKO) in combination with a rabbit polyclonal antiplakophilin-3 (PKP3) antiserum (own production) at a dilution of 1:8,000. After three wash-steps with PBS for 15 min each, fluorochrome-labeled secondary antibodies were incubated for 20 min (Cy2-labeled donkey-anti-mouse, 1:400, Cy3-labeled goat-anti-rabbit, 1:1,000). Staining of DNA was performed by adding 4’,6-diamidino-2-phenylindole (DAPI, 1 mg/ml, diluted 1:10,000 in H2O) for the final 3 min of incubation. Subsequent to three wash-steps with PBS, cover slips were mounted with Fluoromount-G (Southern Biotech, Birmingham, AL, USA) and analyzed by confocal fluorescence microscopy. Confocal images of living cells were acquired on a Leica TCS SP2 AOBS microscope using a 40× oil planapochromat lens (Leica Microsystems, Germany), essentially as described

62

Fig. 1 Treatment of HNSCC cell lines with EGF influences cytokeratin expression levels. The four HNSCC cell lines UM-SCC-3, UMBSCC-864, -969 and UT-SCC-26A were treated with rising levels of EGF (0.1, 1, 10, 100 and 1,000 ng/ml). Western blot analysis of whole cell lysates was performed as described in the materials and methods section. Pan-CK as well as CK selective antibodies were used to monitor for changes in the respective CK expression levels. A431 (C1) and HaCaT (C2) cell lysates were used as positive controls (for details, please see main text). Please note the sporadic co-appearance of CK bands, particularly those of the small-sized CKs, next to the actin

Cell Tissue Res (2013) 351:59–72

input control that was applied at the same membrane after the respective CK antibody. Contrast, brightness and size of the images have been adjusted for better comparison. # The antibody AE1 is able to recognize CK19, CK16, CK14 and other acidic type I CKs. Shown are the two major bands consistent with CK19 and CK16. Note: Evaluating protein expression by Western blot analysis solely intended to evaluate changes in the protein expression levels after EGF treatment. If no CK bands are detectable at a specific EGF concentration it does not necessarily imply that cells do not express the protein

Cell Tissue Res (2013) 351:59–72

before (Astanina and Jacob 2010). Specificity of the used anti-PKP3 antibody for detection of desmosomal structures was confirmed by co-labeling with a commercially available validated antibody (clone DP 2.17) against the desmosomal component desmoplakin (Moll et al. 2006) (see supplemental figure). Atomic force microscopy (AFM) AFM was performed on a JPK NanoWizard (JPK, Berlin, Germany) under ambient conditions. Cells were grown on Petri dishes, fixed with methanol, and measured in “Intermittent contact air” to minimise the impact of the tip on the sample’s surface. Commercially available NSC 16 AlBS tips (Micromash, Estonia) with a cantilever length of 230 nm and a spring constant of 40 N/m were used. Scan size of an overview was 100×100 μm; magnifications had a size of 40×40 μm. Scan frequency was applied to 0.6 Hz with a resolution of 512 pixel/line. The JPK image processing software was used to achieve 3D images. Fig. 2 Real-time cellular analysis of HNSCC cells after treatment with EGF. After exposure to EGF (10 ng/ml), three of the HNSCC cell lines (UM-SCC-3, UMB-SCC-864, 969) responded with a rise in impedance that was significantly higher than in the untreated control cells and consistent with an induced proliferation rate after EGF-treatment (a–c). However, the HNSCC cell line UT-SCC-26A responded paradoxically to EGF treatment, exhibiting a markedly lower gain of impedance than the untreated control cell lines (d). To demonstrate that the observed effect was EGF-specific the same cell lines were co-incubated with an EGF-neutralizing antibody (e–h). In all cases, the previously observed EGF effect could be reverted to control levels, except for one cell line (UMB-SCC-969) that even exhibited lower impedance levels than in the control cells

63

Results Four HNSCC cell lines were used to evaluate the effect of EGFR-activation on CK expression levels. Cells were longterm (18 h) exposed to rising EGF levels (0.1, 1, 10, 100 and 1,000 ng/ml), and protein expression was evaluated by Western blot analysis (Fig. 1). Using a pan-CK antibody, a slight increase in CK expression levels was noted in UT-SCC-26A cells, whereas UMSCC-3 cells rather exhibited a reduction at higher EGF levels. Notably, the CK signals seen in UT-SCC-26A were typically weaker than the bands seen in the other three cell lines. CK selective antibodies were used to further define expression of cytokeratin subgroups or individual CKs. Only weak CK4 signals were detected in the UMB-SCC969 and UT-SCC-26A cell lines. Cytokeratin 13 was expressed in UMB-SCC-864, -969 and UT-SCC-26A cells, and appeared up-regulated in a dose-dependent manner in UMB-SCC-864 and UT-SCC-26A cells. A marked upregulation of CK5 after treatment with EGF was observed

64

in UMB-SCC-864 and UM-SCC-3, reaching a maximum at 10 ng EGF/ml in the latter cell line. UMB-SCC-969 cells already demonstrated high basal (0 ng EGF/ml) CK5 levels that appeared reduced after EGF treatment. High CK14 levels were found in UM-SCC-3 cells, and did not seem to be visibly changed after EGF treatment. A slight EGF-dependent induction of CK14 was noted in UMB-SCC-969 cells. Significant co-expression of cytokeratins 6 and 16 were found in UMBSCC-969 and UM-SCC-3 cells, whereas no apparent coexpression was seen in the other two cell lines. Cytokeratin 19, representing the smallest (40 kDa) of the CKs, was only expressed in UMB-SCC-864 and UT-SCC-26A cells, and exhibited a dose-dependent rise in expression after EGF treatment, which was particularly visible in UT-SCC-26A cells. Here it is of note that when using an antibody with broader specificity for acidic type I CKs (clone AE1) that particularly should recognize CKs 19, 16, and 14, a band consistent with CK19 could be observed in untreated (0 ng EGF/ ml) UM-SCC-3 cells, and virtually vanished after exposure to even low levels of EGF (0.1 ng/ml). However, this polypeptide was not reliably detectable with the CK19-specific antibody. An EGF-dependent rise of CK18 was most obvious in the UT-SCC-26A cell line, which also demonstrated co-expression of CK8 at 10 ng EGF/ml. A concomitant rise in phosphorylated Erk-1/2 and Stat-3 proteins after EGF treatment was most pronounced in the UTSCC-26A cell line, whereas UMB-SCC-969 exhibited high basal expression levels of these phosphoproteins. No relevant change in phospho-Stat-3 levels were seen in the latter cell line after EGF treatment, whereas phospho-Erk-1/2 appeared reduced at higher EGF levels. The xCELLigence real-time cell analyzer system was used to study changes in cellular impedance of the respective HNSCC cells after treatment with EGF. Next to the level of proliferation, changes in the cellular morphology of EGF-treated cells should also influence cellular impedance. Cells were either not treated or treated with 10 ng EGF/ml. For this, HNSCC cell lines were grown on special 96-well plates, and cellular impedance was measured over time (Fig. 2). Interestingly, without addition of EGF, all tested cell lines responded with a gain in cellular impedance consistent with a specific level of proliferation characteristic for each cell line (Fig. 2a–h). After EGF addition, three of the cell lines (UM-SCC-3, UMB-SCC-864, and -969) responded with a pronounced rise of cellular impedance (Fig. 2a–c). In contrast, EGF addition resulted in paradoxically reduced cellular impedance in the UT-SCC-26A cell line compared to untreated cells of the same cell line (Fig. 2d). To verify that the observed effect is specific for EGF, cells were incubated with an anti-EGF specific antibody at the time of EGF addition. As seen in Fig. 2(e–h), antibody addition could virtually completely abolish the EGF-induced effect in all cell lines. Interestingly, antibody

Cell Tissue Res (2013) 351:59–72

addition to UMB-SCC-969 cells even led to a reduction of cellular impedance below control levels (Fig. 2g). Since changes in cellular impedance during real-time cellular analysis could be affected by the degree of cellular proliferation and cellular death, as well as by morphological cell shape changes, cell viability was evaluated with the XTT-viability assay. For this, cells were incubated with rising levels of EGF (0.1, 1, 10, 100, 1,000 ng/ml), and viability was evaluated 18 h after EGF addition (Fig. 3a–d), corresponding to a time were EGF-mediated effects on cellular impedance were clearly detectable (Fig. 2, 42 h). As depicted in Fig. 3, there were no distinct changes in cell viability. A slight rise in cellular viability followed by a moderate reduction at higher EGF levels was noted in

Fig. 3 Cellular viability of HNSCC cells appears not markedly affected after EGF treatment. To monitor for changes in cellular viability that could account for the observed differences in real-time cellular analysis, an XTT-viability assay was performed on all tested HNSCC cells. All HNSCC cell lines were treated for 18 h with rising EGF levels (0.1, 1, 10, 100 and 1,000 ng/ml) and viability was assessed thereafter. None of the tested cell lines responded with a marked change in cellular viability (a–d). A slight EGF-dependent elevation in viability was noted in UT-SCC-26A cells (d)

Cell Tissue Res (2013) 351:59–72

65

Fig. 4 The keratinocytederived HaCaT cell line, known to express abundant levels of cytokeratin (CK) and plakophilin-3 (PKP3), was used as a positive control to demonstrate the typical subcellular localisation of these two proteins. Cytokeratins are shown in green (a,a’) and Plakophilin-3 in red (b,b’). Nuclei were counterstained with Dapi and appear blue (c,c’). Images were overlayed (Merge) for better comparison of the signals (d,d’). A scale bar is depicted for size comparison

UM-SCC-3 cells (Fig. 3a). Interestingly, UT-SCC-26A, the cell line with the paradoxical drop in cellular impedance after EGF treatment (Fig. 2d), not only did not show a drop in cellular viability but even exhibited a moderate rise (Fig. 3d). Fig. 5 Influence of EGF treatment on cytokeratin and plakophilin-3 expression and organization in UM-SCC-3 cells. The HNSCC cell line UM-SCC-3 was either treated (10 ng/ml EGF) or not treated (no EGF) with EGF. Cytokeratin expression appears more diffusely distributed after ligand treatment (e,e’), whereas no obvious change in the subcellular distribution was noted for plakophilin-3 (f,f’). Nuclei were counterstained with Dapi and appear blue c,c’,g,g’). Images were overlayed (Merge) for better comparison of the signals (d,d’,h,h’). A scale bar is depicted for size comparison

Immunofluorescence analysis was deployed to evaluate changes in the subcellular arrangement of CK filaments. Labelling of cell lines was performed with a pan-CK antibody, together with a PKP3-specific antibody denoting desmosomal structures (Figs. 4, 5, 6, 7 and 8). Prior to staining,

66

Cell Tissue Res (2013) 351:59–72

Fig. 6 Influence of EGF treatment on cytokeratin and plakophilin-3 expression and organization in UMB-SCC-864 cells. The HNSCC cell line UMB-SCC-864 was either treated (10 ng/ml EGF) or not treated (no EGF) with EGF. Cytokeratins appear more diffusely distributed after ligand treatment, losing their typical cable-like structures (e,e’), whereas no obvious change in the subcellular distribution was noted for plakophilin-3 (f,f’). Nuclei were counterstained with Dapi and appear blue (c,c’,g,g’). Images were overlayed (Merge) for better comparison of the signals (d,d’,h,h’). A scale bar is depicted for size comparison

HNSCC cells were either treated or not treated with 10 ng EGF/ ml. HaCaT cells were carried as a control for staining patterns of antibodies directed against CKs and desmosomes (Fig. 4a–d and a’–d’). Interestingly, after EGF treatment, CKs frequently did not exhibit typical cable-like structures but rather appeared in a diffuse cytoplasmic staining (Figs. 5, 6 and 7e and e’). On the other hand, in some cases a pronounced perinuclear staining was detectable (Fig. 8e and e’, asterisk in e’). When applying atomic force microscopy, the effect of EGF-treatment (10 ng/ml) on cellular shape was most strikingly visible in UT-SCC-26A cells. This cell line responded with a major change in its cellular morphology, developing cytoplasmic filopodia-like extensions associated with shrinkage of the cell body (Fig. 9h and h’). Similar observations, although less pronounced, could be made on UMBSCC-969 cells (Fig. 9f and f’).

Discussion Around 20 CKs were characterized in previous studies (Moll et al. 1982), whereas meanwhile a total of 54 genes

encoding human CKs could be identified in the human genome, being located on chromosomes 12 and 17 (Rogers et al. 2005; Hesse et al. 2004; Moll et al. 2008). Twentyeight of the genes encode so-called acidic type I keratins, whereas 26 belong to the neutral or basic type II keratins (Moll et al. 1982; Schweizer et al. 2006; Moll et al. 2008). Typically, a type I keratin pairs with a type II keratin in vivo and in vitro (Herrmann and Aebi 2004; Hatzfeld and Franke 1985). Several diseases were found to be associated with mutations affecting CK genes or with a disturbance of CK function (Omary et al. 2004). Furthermore, Jiang et al. reported that CK mRNA transcripts, particularly those of the hyperproliferative cytokeratins CK6 and CK16, were up-regulated in human epidermal keratinocytes after exposure to EGF or TGF-α, the two major ligands of the EGF receptor (Jiang et al. 1993). The present study aimed to investigate the consequences which EGF treatment of HNSCC cells had on CK expression levels and subcellular distribution, cellular morphology, and viability. The expression levels of CK4, CK5, CK6, CK7, CK8, CK13, CK14, CK16, CK18, and CK19 were evaluated by Western blot analysis

Cell Tissue Res (2013) 351:59–72

67

Fig. 7 Influence of EGF treatment on cytokeratin and plakophilin-3 expression and organization in UMB-SCC-969 cells. The HNSCC cell line UMB-SCC-969 was either treated (10 ng/ml EGF) or not treated (no EGF) with EGF. Similarly as observed in UMB-SCC-864 cells (Fig. 6), cytokeratins appear more diffusely distributed after ligand treatment losing their typical cable-like structures (e,e’), whereas no obvious change in the subcellular distribution was noted for plakophilin-3 (f,f’). Nuclei were counterstained with Dapi and appear blue (c,c’,g,g’). Images were overlayed (Merge) for better comparison of the signals (d,d’,h,h’). A scale bar is depicted for size comparison

CK4 and CK13 CK4 and CK13 are co-expressed in suprabasal regions of non-keratinized epithelia. Mutations in CK4 and CK13 are associated with the development of white sponge nevus in the oral and esophageal mucosa (Chao et al. 2003; Fuchs 1996). In the present study, co-expression of the two CKs could be observed in UMB-SCC-969 and UT-SCC-26A cells, and moderate EGF-dependent upregulation of CK13 was seen in UMB-SCC-864 and UT-SCC-26A cells at high EGF levels. In a study by Yanagawa et al., loss of CK13 expression in tongue squamous cell carcinomas predicted for a higher risk of local recurrence (Yanagawa et al. 2007). CK5 and CK14 CK5 and CK14 are typically found co-expressed in the basal progenitor cell layer of the epidermis and mucosa (Lee and Coulombe 2009). Mutations in either of the two keratin genes are typically found in patients with epidermolysis bullosa

simplex (EBS) (Yiasemides et al. 2008). CK14 is also implicated as a useful marker (together with other markers) for the selection of epithelial cancer stem cells such as those present in HNSCC (Prince et al. 2007; Sterz et al. 2010). In a report by van der Velden et al., CKs 14 and 16 were found to be overexpressed in benign lesions of the vocal cord (van der Velden et al. 1996). In the present study, expression of CK5 was detected in all tested cell lines, particularly in UMB-SCC969 and UM-SCC-3. EGF-dependent up-regulation of CK5 was noted particularly in UMB-SCC-864 and UM-SCC-3 cells, whereas UMB-SCC-969 cells responded rather with a reduction of CK5 levels. The most pronounced expression of CK14 was found in UM-SCC-3 cells, whose expression did not seem to respond to EGF-treatment, which is in contrast to UMB-SCC-969 cells that expressed only low CK14 levels but exhibited a slight induction of CK14. CK14 levels appeared very low in UMB-SCC-864 and UT-SCC-26A despite expression of CK5 in the same cells. In this context, it is interesting that Yiasemides et al. suspected the presence of an alternative CK5 binding partner in EBS that could compensate for loss of CK14 (Yiasemides et al. 2008).

68

Cell Tissue Res (2013) 351:59–72

Fig. 8 Influence of EGF treatment on cytokeratin and plakophilin-3 expression and organization in UT-SCC-26A cells. The HNSCC cell line UT-SCC-26A was either treated (10 ng/ml EGF) or not treated (no EGF) with EGF. Cells appeared slightly contracted after ligand treatment (compare a’ and e’), whereas no obvious change in the subcellular distribution was noted for plakophilin-3 (f,f’). Perinuclear cytokeratin staining appeared prominent after EGF treatment (asterisk in e’). Nuclei were counterstained with Dapi and appear blue (c,c’,g,g’). Images were overlayed (Merge) for better comparison of the signals (d,d’,h,h’). A scale bar is depicted for size comparison

CK6 and CK16

CK7 and CK19

Cytokeratins 6 and 16 are frequently co-expressed in squamous epithelia, and represent hyperproliferative CKs. Their expression was tested in the present study, since previous work on human epidermal keratinocytes observed induction of CK6 and CK16 transcripts after stimulation with the EGFR ligands EGF and TGF-α (Jiang et al. 1993) and since CKs 6 and -16, were found up-regulated in HNSCC cell lines and tissues (Sesterhenn et al. 2005). CK6 has three distinct, highly homologous isoforms that are encoded by separate genes (Schweizer et al. 2006). Mutations in CK6 as well as in CK16 and CK17, the latter representing an alternative binding partner of CK6, are found in skin diseases such as pachyonychia congenita (Smith et al. 1998; McLean et al. 1995; Bowden et al. 1995). In the present study, CK6 and CK16 appeared co-expressed in UMB-SCC-969 and UMSCC-3 cells. EGF-dependent CK16 induction was particularly visible in UM-SCC-3 cells. Expression of CK6 appeared to increase in UMB-SCC-969 and UM-SCC-3 cells and to decrease again at higher EGF levels. Overall, expression of these CKs appear to vary among the tested cell lines.

CK19 is the smallest cytokeratin (40 kDa), and is found in several epithelial tissues. Overexpression of this CK is frequent in thyroid carcinoma of the papillary type (Cameron and Berean 2003). Of particular interest for head and neck cancer research is that the marker CYFRA21-1, which represents a soluble CK19 fragment, is used to monitor disease progression, specifically the appearance of distant lung metastasis (Kuropkat et al. 2002). Interestingly, two of the tested HNSCC cell lines (UMB-SCC-864 and UT-SCC-26A) expressed high CK19 levels, which appeared up-regulated after EGF treatment. Previously, we reported that these two cell lines were highly resistant to a treatment with cisplatin (CDDP). In contrast, the other two tested cell lines responded more sensitive to this chemotherapeutic agent (Mandic et al. 2005). In addition, it was noted that treatment of UT-SCC-26A cells with EGF, in the same manner as described in the present study, could render this cell line CDDP-sensitive (Mandic et al. 2009). Except for a faint signal that was observed in EGFuntreated UM-SCC-3 cells there was no clearly detectable signal for CK7, which is typically found in simple epithelia.

Cell Tissue Res (2013) 351:59–72

69

Fig. 9 Atomic force microscopy demonstrates a pronounced change in cellular morphology of UT-SCC-26A cells after EGF treatment. For further evaluation of EGF-dependent changes in the cell shape of the four tested HNSCC cell lines, cells were either treated (b,b’,d,d’,f,f’,h, h’) or not treated (a,a’,c,c’,e,e’,g,g’) with EGF. Subsequently, cells

were fixed and subjected to atomic force microscopy as described in materials and methods. The most severe change in cellular morphology was observed in UT-SCC-26A cells, which appeared contracted after EGF-treatment, thereby covering a markedly reduced surface area compared with the untreated control cells (h,h’)

CK8 and CK18

Keratin and desmosomal protein patterns

Expression of CK8 and CK18 is found in simple, singlelayer epithelia (Oshima et al. 1996; Fuchs 1996). These CKs also play a major role in the development of liver diseases. In this context, it has been found that mutations in either CK8 or CK18 predispose patients to liver injury, and that changes in the ratio of both CKs toward CK8 promote the development of Mallory–Denk bodies (Ku et al. 2007). Interestingly, Yamashiro et al. observed in metastatic and non-metastatic transformed mouse keratinocytes that the metastatic line, in contrast to the nonmetastatic one, expressed CK8 and CK18, and that coexpression of both CKs in the non-metastatic cell line resulted in enhanced invasiveness (Yamashiro et al. 2010).

Downregulation of desmosomes is found in poorly differentiated tumors such as transitional cell carcinomas and HNSCCs (Garrod 1995). Previous immunohistochemical studies investigated the expression of the desmosomalassociated plakophilins in diverse tumor tissues, including oral and pharyngeal squamous cell carcinomas (Schwarz et al. 2006). In the study, it was found that, PKP1 and -3 not only exhibited a desmosomal staining pattern but also that their expression levels inversely correlated with the aggressiveness of the tumor. Interestingly, plakophilins can also be found in a nuclear, e.g., associated with RNA polymerase III, as well as cytoplasmic localisation (Schmidt and Jager 2005; Schmidt et al. 1997; Mertens et al. 2001). The observation that PKPs are not exclusively associated with

70

desmosomes is consistent with the observations in this paper, where PKP3 signals could also be detected in the cytoplasm (Fig. 8b and b’, f and f’). Beil et al. reported about a change in the CK architecture of PANC-1 epithelial cells after exposure to the bioactive lipid SPC. In their study, the authors found that perinuclear accumulation of CKs 8 and 18 resulted in a gain of visco-elastic properties of the cells, correlating with an enhanced migration rate through a size-limited pore filter, which appeared independent of F-actin or microtubules (Beil et al. 2003). In the present study, real-time cellular analysis demonstrated EGFdependent changes in the cellular impedance which were probably caused by an EGF-induced alteration in the morphology of HNSCC cells, leading to a reduced surface coverage of the electrode at the bottom of the well and reduced gain of impedance compared with untreated cells (Fig. 2d). This effect assumedly was not due to reduced proliferation or viability of the cells, since the viability of the cells was not markedly affected after EGF addition. Interestingly, after EGF treatment, some HNSCC cells (Fig. 8e and e’) demonstrated a pronounced perinuclear CK staining, which is reminiscent to the observations of Beil et al., who observed this effect after SPC treatment (Beil et al. 2003). In this context, it is interesting that only UT-SCC-26A cells responded to EGF treatment with a substantial rise in CK18 and to a lesser extent CK8 levels (Fig. 1). Liao et al. observed that post-translational modifications such as phosphorylation of CK18 on serine 52 was increased during S and G2/M phase of the cell cycle, and proposed that this cell-cycle-dependent phosphorylation is associated with cytoskeletal reorganization events (Liao et al. 1995). The same group was further able to show that phosphorylation on serine 33 in CK18 is required for its binding to 14-3-3 proteins, which could be implicated in the subcellular distribution of this keratin (Ku et al. 1998). During EGF treatment, some of the HNSCC cells, particularly UT-SCC-26A and UMB-SCC-969 (Fig. 9), developed filopodia-like extensions. Interestingly, other groups have proposed an actin-mediated transport of keratin particles from growing lamellipodia via actin stress fibers to central regions of the cell, where they are integrated in the CK network (Kolsch et al. 2009; Windoffer et al. 2011). A similar mechanism could be assumed to play a role in EGF-dependent CK reorganization. To elucidate what kind of signal pathways are responsible for this cytoskeletal reorganization in HNSCC cells will be an interesting task for future studies. Role of EGFR signalling in HNSCC disease The EGFR gene is amplified in the vast majority of HNSCC tumors (Weichselbaum et al. 1989), and its level of signalling correlates tightly with tumor progression, therapy

Cell Tissue Res (2013) 351:59–72

response, and disease outcome (Chung et al. 2004, 2006). For this reason, several therapeutic approaches were undertaken to inhibit EGFR-signalling in HNSCC and other EGFR-expressing tumors, e.g., by using monoclonal antibodies directed against the extracellular ligand-binding site of the receptor such as cetuximab (Erbitux®), or smallmolecule kinase inhibitors such as gefitinib (Iressa®) aiming to inhibit the intracellular kinase domain of the receptor. In previous studies, we were able to demonstrate EGFR expression and activation levels in several HNSCC cell lines, including the cell lines used in the present study (Mandic et al. 2001, 2005, 2006, 2009; Krohn et al. 2011). UT-SCC26A cells that most severely responded to EGF treatment were previously also found to be particularly resistant to cisplatin (CDDP), representing the major chemotherapeutic agent in HNSCC treatment (Mandic et al. 2005). Furthermore, it was demonstrated that EGF treatment could render UT-SCC-26A cells susceptible to CDDP treatment (Mandic et al. 2009). Although this is an interesting observation, it remains to be elucidated if and how the observed EGFassociated changes in morphology relate to other EGFinduced cellular features such as higher sensitivity to cisplatin. Future studies need to determine if histopathological and -morphological characteristics observed in HNSCC tumors are associated with an EGFR-activity-dependent CK reorganization, and how this would relate to tumorrelevant features such as migration, invasion and metastasis. Acknowledgments Technical assistance of Ms. Roswitha Peldszus and Ms. Grazyna Sadowski (both Department of Otorhinolaryngology, Head and Neck Surgery, University Hospital Giessen and Marburg, Campus Marburg, Marburg, Germany) and Ms. Viktoria Morokina (Institute of Pathology, University Hospital Giessen and Marburg GmbH, Campus Marburg, Marburg, Germany) was greatly appreciated.

References Astanina K, Jacob R (2010) KIF5C, a kinesin motor involved in apical trafficking of MDCK cells. Cell Mol Life Sci 67:1331–1342 Beil M, Micoulet A, von Wichert G, Paschke S, Walther P, Omary MB, Van Veldhoven PP, Gern U, Wolff-Hieber E, Eggermann J, Waltenberger J, Adler G, Spatz J, Seufferlein T (2003) Sphingosylphosphorylcholine regulates keratin network architecture and viscoelastic properties of human cancer cells. Nat Cell Biol 5:803–811 Bowden PE, Haley JL, Kansky A, Rothnagel JA, Jones DO, Turner RJ (1995) Mutation of a type II keratin gene (K6a) in pachyonychia congenita. Nat Genet 10:363–365 Cameron BR, Berean KW (2003) Cytokeratin subtypes in thyroid tumours: immunohistochemical study with emphasis on the follicular variant of papillary carcinoma. J Otolaryngol 32:319–322 Chao SC, Tsai YM, Yang MH, Lee JY (2003) A novel mutation in the keratin 4 gene causing white sponge naevus. Br J Dermatol 148:1125–1128 Chung CH, Parker JS, Karaca G, Wu J, Funkhouser WK, Moore D, Butterfoss D, Xiang D, Zanation A, Yin X, Shockley WW, Weissler MC, Dressler LG, Shores CG, Yarbrough WG, Perou

Cell Tissue Res (2013) 351:59–72 CM (2004) Molecular classification of head and neck squamous cell carcinomas using patterns of gene expression. Cancer Cell 5:489–500 Chung CH, Ely K, McGavran L, Varella-Garcia M, Parker J, Parker N, Jarrett C, Carter J, Murphy BA, Netterville J, Burkey BB, Sinard R, Cmelak A, Levy S, Yarbrough WG, Slebos RJ, Hirsch FR (2006) Increased epidermal growth factor receptor gene copy number is associated with poor prognosis in head and neck squamous cell carcinomas. J Clin Oncol 24:4170–4176 Dobrossy L (2005) Epidemiology of head and neck cancer: magnitude of the problem. Cancer Metastasis Rev 24:9–17 Friedl P, Wolf K (2003) Tumour-cell invasion and migration: diversity and escape mechanisms. Nat Rev Cancer 3:362–374 Fuchs E (1996) The cytoskeleton and disease: genetic disorders of intermediate filaments. Annu Rev Genet 30:197–231 Garrod DR (1995) Desmosomes and cancer. Cancer Surv 24:97–111 Gigi O, Geiger B, Eshhar Z, Moll R, Schmid E, Winter S, Schiller DL, Franke WW (1982) Detection of a cytokeratin determinant common to diverse epithelial cells by a broadly cross-reacting monoclonal antibody. EMBO J 1:1429–1437 Hammerl P, Hartl A, Thalhamer J (1993) Particulate nitrocellulose as a solid phase for protein immobilization in immuno-affinity chromatography. J Immunol Methods 165:59–66 Hanahan D, Weinberg RA (2011) Hallmarks of cancer: the next generation. Cell 144:646–674 Hatzfeld M, Franke WW (1985) Pair formation and promiscuity of cytokeratins: formation in vitro of heterotypic complexes and intermediate-sized filaments by homologous and heterologous recombinations of purified polypeptides. J Cell Biol 101: 1826–1841 Herrmann H, Aebi U (2004) Intermediate filaments: molecular structure, assembly mechanism, and integration into functionally distinct intracellular scaffolds. Annu Rev Biochem 73:749–789 Hesse M, Zimek A, Weber K, Magin TM (2004) Comprehensive analysis of keratin gene clusters in humans and rodents. Eur J Cell Biol 83:19–26 Jiang CK, Magnaldo T, Ohtsuki M, Freedberg IM, Bernerd F, Blumenberg M (1993) Epidermal growth factor and transforming growth factor alpha specifically induce the activation- and hyperproliferationassociated keratins 6 and 16. Proc Natl Acad Sci U S A 90:6786–6790 Kolsch A, Windoffer R, Leube RE (2009) Actin-dependent dynamics of keratin filament precursors. Cell Motil Cytoskeleton 66: 976–985 Krohn V, Wiegand S, Werner JA, Mandic R (2011) EGFR codon 497 polymorphism—implications for receptor sensitivity to inhibitors in HNSCC cell lines. Anticancer Res 31:59–65 Ku NO, Liao J, Omary MB (1998) Phosphorylation of human keratin 18 serine 33 regulates binding to 14-3-3 proteins. EMBO J 17:1892–1906 Ku NO, Strnad P, Zhong BH, Tao GZ, Omary MB (2007) Keratins let liver live: mutations predispose to liver disease and crosslinking generates Mallory–Denk bodies. Hepatology 46:1639–1649 Kuropkat C, Lippert BM, Werner JA (2002) Follow-up with serum Cyfra 21-1 in patients with squamous cell carcinomas of the head and neck. Oncology 63:280–285 Lansford C, Grenman R, Bier H, Somers KD, Kim S-Y, Whiteside TL, Clayman G, Carey TE (1999) Head and neck cancers. In: Masters J (ed) Human cell culture vol 2, cancer cell lines part 2. Kluwer Academic Press, Dordrecht, pp 185–255 Lee CH, Coulombe PA (2009) Self-organization of keratin intermediate filaments into cross-linked networks. J Cell Biol 186:409–421 Liao J, Lowthert LA, Ku NO, Fernandez R, Omary MB (1995) Dynamics of human keratin 18 phosphorylation: polarized distribution of phosphorylated keratins in simple epithelial tissues. J Cell Biol 131:1291–1301

71 Mandic R, Eikelkamp N, Peldszus R, Sadowski M, Sesterhenn AM, Dunne AA, Werner JA (2001) Variations of EGF-R surface expression in squamous cell carcinomas of the head and neck region. Anticancer Res 21:3413–3418 Mandic R, Schamberger CJ, Muller JF, Geyer M, Zhu L, Carey TE, Grenman R, Dunne AA, Werner JA (2005) Reduced cisplatin sensitivity of head and neck squamous cell carcinoma cell lines correlates with mutations affecting the COOH-terminal nuclear localization signal of p53. Clin Cancer Res 11:6845–6852 Mandic R, Rodgarkia-Dara CJ, Zhu L, Folz BJ, Bette M, Weihe E, Neubauer A, Werner JA (2006) Treatment of HNSCC cell lines with the EGFR-specific inhibitor cetuximab (Erbitux) results in paradox phosphorylation of tyrosine 1173 in the receptor. FEBS Lett 580:4793–4800 Mandic R, Rodgarkia-Dara CJ, Krohn V, Wiegand S, Grenman R, Werner JA (2009) Cisplatin resistance of the HNSCC cell line UT-SCC-26A can be overcome by stimulation of the EGFreceptor. Anticancer Res 29:1181–1187 McLean WH, Rugg EL, Lunny DP, Morley SM, Lane EB, Swensson O, Dopping-Hepenstal PJ, Griffiths WA, Eady RA, Higgins C (1995) Keratin 16 and keratin 17 mutations cause pachyonychia congenita. Nat Genet 9:273–278 Mertens C, Hofmann I, Wang Z, Teichmann M, Sepehri CS, Schnolzer M, Franke WW (2001) Nuclear particles containing RNA polymerase III complexes associated with the junctional plaque protein plakophilin 2. Proc Natl Acad Sci U S A 98:7795–7800 Moll R, Franke WW, Schiller DL, Geiger B, Krepler R (1982) The catalog of human cytokeratins: patterns of expression in normal epithelia, tumors and cultured cells. Cell 31:11–24 Moll R, Holzhausen HJ, Mennel HD, Kuhn C, Baumann R, Taege C, Franke WW (2006) The cardiac isoform of alpha-actin in regenerating and atrophic skeletal muscle, myopathies and rhabdomyomatous tumors: an immunohistochemical study using monoclonal antibodies. Virchows Arch 449:175–191 Moll R, Divo M, Langbein L (2008) The human keratins: biology and pathology. Histochem Cell Biol 129:705–733 Omary MB, Coulombe PA, McLean WH (2004) Intermediate filament proteins and their associated diseases. N Engl J Med 351: 2087–2100 Oshima RG, Baribault H, Caulin C (1996) Oncogenic regulation and function of keratins 8 and 18. Cancer Metastasis Rev 15:445–471 Prieto VG, Lugo J, McNutt NS (1996) Intermediate- and lowmolecular-weight keratin detection with the monoclonal antibody MNF116. An immunohistochemical study on 232 paraffin-embedded cutaneous lesions. J Cutan Pathol 23: 234–241 Prince ME, Sivanandan R, Kaczorowski A, Wolf GT, Kaplan MJ, Dalerba P, Weissman IL, Clarke MF, Ailles LE (2007) Identification of a subpopulation of cells with cancer stem cell properties in head and neck squamous cell carcinoma. Proc Natl Acad Sci U S A 104:973–978 Ridley AJ, Schwartz MA, Burridge K, Firtel RA, Ginsberg MH, Borisy G, Parsons JT, Horwitz AR (2003) Cell migration: integrating signals from front to back. Science 302:1704–1709 Rogers MA, Edler L, Winter H, Langbein L, Beckmann I, Schweizer J (2005) Characterization of new members of the human type II keratin gene family and a general evaluation of the keratin gene domain on chromosome 12q13.13. J Investig Dermatol 124:536–544 Schmidt A, Jager S (2005) Plakophilins—hard work in the desmosome, recreation in the nucleus? Eur J Cell Biol 84:189–204 Schmidt A, Langbein L, Rode M, Pratzel S, Zimbelmann R, Franke WW (1997) Plakophilins 1a and 1b: widespread nuclear proteins recruited in specific epithelial cells as desmosomal plaque components. Cell Tissue Res 290:481–499 Schwarz J, Ayim A, Schmidt A, Jager S, Koch S, Baumann R, Dunne AA, Moll R (2006) Differential expression of desmosomal

72 plakophilins in various types of carcinomas: correlation with cell type and differentiation. Hum Pathol 37:613–622 Schweizer J, Bowden PE, Coulombe PA, Langbein L, Lane EB, Magin TM, Maltais L, Omary MB, Parry DA, Rogers MA, Wright MW (2006) New consensus nomenclature for mammalian keratins. J Cell Biol 174:169–174 Sesterhenn AM, Mandic R, Dunne AA, Werner JA (2005) Cytokeratins 6 and 16 are frequently expressed in head and neck squamous cell carcinoma cell lines and fresh biopsies. Anticancer Res 25:2675–2680 Smith FJ, Jonkman MF, van Goor H, Coleman CM, Covello SP, Uitto J, McLean WH (1998) A mutation in human keratin K6b produces a phenocopy of the K17 disorder pachyonychia congenita type 2. Hum Mol Genet 7:1143–1148 Sterz CM, Kulle C, Dakic B, Makarova G, Bottcher MC, Bette M, Werner JA, Mandic R (2010) A basal-cell-like compartment in head and neck squamous cell carcinomas represents the invasive front of the tumor and is expressing MMP-9. Oral Oncol 46: 116–122 van der Velden LA, Schaafsma HE, Manni JJ, Link M, Ruiter DJ, Ramaekers FC, Kuijpers W (1996) Cytokeratin and vimentin expression in normal epithelium and benign lesions of the vocal cords. Acta Otolaryngol 116:325–331 Weichselbaum RR, Dunphy EJ, Beckett MA, Tybor AG, Moran WJ, Goldman ME, Vokes EE, Panje WR (1989) Epidermal growth factor receptor gene amplification and expression in head and neck cancer cell lines. Head Neck 11:437–442

Cell Tissue Res (2013) 351:59–72 Windoffer R, Beil M, Magin TM, Leube RE (2011) Cytoskeleton in motion: the dynamics of keratin intermediate filaments in epithelia. J Cell Biol 194:669–678 Yamashiro Y, Takei K, Umikawa M, Asato T, Oshiro M, Uechi Y, Ishikawa T, Taira K, Uezato H, Kariya K (2010) Ectopic coexpression of keratin 8 and 18 promotes invasion of transformed keratinocytes and is induced in patients with cutaneous squamous cell carcinoma. Biochem Biophys Res Commun 399:365–372 Yanagawa T, Yoshida H, Yamagata K, Onizawa K, Tabuchi K, Koyama Y, Iwasa S, Shimoyamada H, Harada H, Omura K (2007) Loss of cytokeratin 13 expression in squamous cell carcinoma of the tongue is a possible sign for local recurrence. J Exp Clin Cancer Res 26:215–220 Yiasemides E, Trisnowati N, Su J, Dang N, Klingberg S, Marr P, Melbourne W, Tran K, Chow CW, Orchard D, Varigos G, Murrell DF (2008) Clinical heterogeneity in recessive epidermolysis bullosa due to mutations in the keratin 14 gene, KRT14. Clin Exp Dermatol 33:689–697 Zhao M, Sano D, Pickering CR, Jasser SA, Henderson YC, Clayman GL, Sturgis EM, Ow TJ, Lotan R, Carey TE, Sacks PG, Grandis JR, Sidransky D, Heldin NE, Myers JN (2011) Assembly and initial characterization of a panel of 85 genomically validated cell lines from diverse head and neck tumor sites. Clin Cancer Res 17:7248–7264 Zhu L, Werner JA, Mandic R (2007) Implications of tropomyosinrelated kinase B (TrkB) in head and neck cancer. Anticancer Res 27:3121–3126