overexpression of γ-tubulin in non-small cell lung cancer

Summary. We and others have previously shown thatincreased expression and altered compartmentalizationof γ-tubulin may contribute to tumorigenesis and tumorprogression (J. Cell Physiol. 2009;223:519-529; CancerBiol. Ther. 2010;9:66-76). Here we have determined byimmunohistochemistry the localization and cellulardistribution of γ-tubulin in clinical tissue samples from109 non-small cell lung cancer (NSCLC) cases. Theexpression and distribution of γ-tubulin protein andtranscripts was also determined in the NSCLC tumor celllines NCI-H460 (HTB-177) and NCI-H69 (HTB-119) byimmunocytochemistry, quantitative immunoblotting andreverse transcription quantitative real-time PCR (RT-qPCR). Polyclonal and monoclonal anti-peptideantibodies recognizing epitopes in the C- or N-terminaldomains of γ-tubulins and human gene-specific primersfor γ-tubulins 1 (TUBG1) and 2 (TUBG2) were used. Innon-neoplastic cells of the airway epithelium in situ, γ-tubulin exhibited predominantly apical surface andpericentriolar localizations. In contrast, markedlyincreased, albeit heterogeneous and variously prominentγ-tubulin immunoreactivity was detected in clinicaltumor specimens and in the NCI-H460 and NCI-H69cell lines, where tumor cells exhibited overlappingmulti-punctate and diffuse patterns of localization. Co-expression of γ-tubulin and Ki-67 (MIB-1) was detected

in a population of proliferating tumor cells. Astatistically significant increase of γ-tubulin expressionwas found in Stage III compared to lesser stage tumors(p<0.001 v. Stages I/II) regardless of histologicalsubtype or grade. By quantitative immunoblotting NCI-H460 and NCI-H69 cells expressed higher levels of γ-tubulin protein compared to small airway epithelial cells(SAEC). In both tumor cell lines increase in TUBG1 andTUBG2 transcripts was detected by RT-qPCR. Ourresults reveal for the first time an increased expressionof γ-tubulin in lung cancer. Key words: Gamma-tubulin, Microtubules, Non-smallcell lung cancer, NSCLC, Adenocarcinoma, Squamouscell carcinoma

Introduction

Lung cancer is a major public health issue and aformidable therapeutic challenge. Most lung tumors areof epithelial origin and belong to the category of non-small cell lung cancer (NSCLC), which is responsiblefor 80 % of all lung cancers. According to the WorldHealth Organization (WHO), NSCLC encompasses threemain histologic types, adenocarcinoma (AC), squamouscell carcinoma (SCC), and the less common large cellcarcinoma (LCC) (Travis et al., 2004). Recent trendsindicate that tumors of the NSCLC category aregenerally treated according to similar strategies withoutthe need for further distinction among histologic types

Overexpression of γγ-tubulin in non-small cell lung cancerNicoletta F. Maounis1,2*, Eduarda Dráberová3*, Eleni Mahera4, Maria Chorti5, Valentina Caracciolo6, Tetyana Sulimenko3, Dimitra Riga4, Nikolaos Trakas7, AphroditeEmmanouilidou2, Antonio Giordano6,8, Pavel Dráber3 and Christos D. Katsetos1,91Department of Pathology and Laboratory Medicine, Drexel University College of Medicine, Philadelphia, Pennsylvania, USA,2Department of Clinical Cytology, “Sismanoglion” General Hospital, Athens, Greece, 3Institute of Molecular Genetics, Academy ofSciences of the Czech Republic, Prague, Czech Republic, 4Department of Pathology, “KAT” General Hospital, Athens, Greece,5Department of Pathology, “Sismanoglion” General Hospital, Athens, Greece, 6Sbarro Institute for Cancer Research and MolecularMedicine, Center for Biotechnology, College of Science and Technology, Temple University, Philadelphia, Pennsylvania, USA,7Department of Clinical Biochemistry, “Sismanoglion” General Hospital, Athens, Greece, 8Department of Human Pathology andOncology, University of Siena, Siena, Italy and 9Department of Pediatrics, Drexel University College of Medicine, St. Christopher’sHospital for Children, Philadelphia, Pennsylvania, USA*These authors contributed equally to this work

Histol Histopathol (2012) 27: 1183-1194

Offprint requests: Prof. Christos D. Katsetos, MD, PhD, FRCPath,Drexel University College of Medicine, St. Christopher’s Hospital forChildren, 3601 A Street, Philadelphia, Pennsylvania 19134, USA. e-mail: [email protected]

http://www.hh.um.esHistology andHistopathology

Cellular and Molecular Biology

(Hirsch et al., 2008). Others contend that a histo-pathologic distinction between SCC and AC remainsrelevant in dictating chemotherapeutic strategies inadvanced NSCLC (Selvaggi and Scagliotti, 2009). Atinitial diagnosis, more than half of NSCLCs correspondto advanced clinical stage and patients with advancedNSCLC are candidates for systemic chemotherapy (Sèveand Dumontet, 2005).

The role of tubulin in NSCLC has been the focus ofprevious translational research (Katsetos et al., 2000;Dumontet et al., 2005; Sève et al., 2005a,b, 2007). Todate, large clinical trial studies have focused solely onßΙΙΙ-tubulin and its roles in tumor prognosis and taxanechemoresistance (Dumontet et al., 2005; Sève et al.,2005a,b, 2007, 2010). The elucidation of microtubuleabnormalities in lung cancer and the development ofnew therapeutic approaches would be further enhancedby the characterization of altered expression of tubulinspecies involved in microtubule nucleation, which mayalso potentially serve as clinical biomarkers andmolecular targets for cancer therapy.

γ-Tubulin is the principal cytoskeletal constituent ofthe pericentriolar material of centrosomes, the cell'smicrotubule organizing centers (MTOCs), where it playsa central role in microtubule nucleation and in theregulation of cell cycle progression (Zhou et al., 2002).Experimental depletion of γ-tubulin leads to a depletionof microtubules and to growth arrest. In mammaliancells, two γ-tubulin genes TUBG1 and TUBG2 exist,encoding two closely related isotypes (Wise et al., 2000).TUBG1 is ubiquitously expressed, whereas TUBG2 hasbeen found mainly in the brain (Yuba-Kubo et al., 2005;Vinopal et al., 2012).

Although γ-tubulin is localized on MTOCs, a largeramount of γ-tubulin is in soluble form (Moudjou et al.,1996; Sulimenko et al., 2002). γ-Tubulin appears in twomain complexes: the large γ-tubulin ring complex(γTuRC) (Moritz et al., 1995; Zheng et al., 1995) and theγ-tubulin small complex (γ-TuSC) (Moritz et al., 1998).The existence of γTuRCs correlates with the ability ofcentrosomes to nucleate microtubules (Schnackenbergand Palazzo, 2001). Additionally, γ-tubulin is associatedwith cellular membranes (Chabin-Brion et al., 2001;Dryková et al., 2003; Bugnard et al., 2005; Sulimenko etal., 2006) where it can give rise to non-centrosomalmicrotubule nucleation (Chabin-Brion et al., 2001;Efimov et al., 2007; Macurek et al., 2008). Nuclear-specific functions of γ-tubulin were recently reported(Hořejší et al., 2012; Höög et al., 2011).

The role of γ-tubulin in microtubule nucleation andmicrotubule nucleation-independent functions in non-transformed cells, as well as the altered expression,distribution and subcellular sorting of this protein incancer cells have been reviewed (Katsetos et al., 2009).We have previously shown overexpression and ectopiccompartmentalization of γ-tubulin in malignant braintumors and have suggested that this can be one of theputative mechanisms of tumorigenesis and tumorprogression (Katsetos et al., 2006, 2007, 2009;Caracciolo et al., 2010). Moreover, recent studies have

shown increased expression and ectopic solubledistribution of γ-tubulin with decoupling/ dissociationfrom the centrosomes in aggressive breast cancer celllines (Cho et al., 2010).

To get a broader insight into the expression of γ-tubulin in lung tumors, we have analyzed the expressionand cellular distribution of this protein in a large cohortof NSCLC pathological specimens with comparison tothe expression of γ-tubulin protein and mRNA in thehuman NSCLC cell lines NCI-H460 and NCI-H69.Materials and methods

Cells

Human lung cell carcinoma cell lines NCI-H460(Catalog. No. ATCC-HTB-177) and NCI-H69 (CatalogNo. ATCC-HTB-119) were obtained from the AmericanType Culture Collection (Manassas, VA). Cells werecultured in Dulbecco’s modified Eagle’s medium(DMEM) containing 10 % fetal bovine serum, penicillin(100 units/ml), and streptomycin (0.1 mg/ml). Cells weregrown at 37°C in 5 % CO2 in air and passaged every 2or 3 days. In the case of adherent H460 cells 0.25 %trypsin/0.01% EDTA in PBS, pH 7.5 was used.Proliferating non-immortalized, non-transformed humansmall airway epithelial cells (SAEC) were obtained fromLonza (Cologne, Germany), and were maintained insupplemented Small Airway Epithelial Cell BasalMedium according to the manufacturer’s directions. Antibodies

The following anti-peptide antibodies against humanγ-tubulin were used: mouse monoclonal antibodies TU-30 and TU-32 to the sequence 434-449 (Nováková et al.,1996); mAb GTU-88 to the sequence 38-53 (Sigma,T6657); rabbit antibodies to sequences 433-451 (DQ-19;Sigma, T3195) and 38-53 (Sigma, T5192). Themonoclonal antibody supernatants TU-30 and TU-32were stored in freeze-dried state in the presence oftrehalose (Dráber et al., 1995). Monoclonal antibodyGTU-88 and polyclonal antibody (Sigma, T 5192) wereraised against peptide sequence (38-53) which isidentical both in human γ-tubulin 1 and human γ-tubulin2. Monoclonal antibodies TU-30, TU-32 and polyclonalantibody DQ-19 were raised against peptides that differonly in one amino acid when sequences of human γ-tubulin 1 and γ-tubulin 2 were compared. A rabbitantibody to actin was purchased from Sigma (A2066)and Cy3-conjugated anti-mouse antibody from JacksonImmunoresearch Laboratories (West Grove, PA).Horseradish peroxidase (HRP)-conjugated anti-mouseand anti-rabbit antibodies were obtained from Promega(Madison, WI).Reverse transcription quantitative real-time PCR (RT-qPCR analysis)

Total RNA from cultured cells was isolated by the

1184γ-tubulin in NSCLC

RNeasy Mini kit (QIAGEN, Valencia, CA). Aliquots of1 µg of total RNA in 20 µl reaction mixture wereconverted to cDNA using the ImProm-II RT kit(Promega) with oligo(dT) primers. Twenty microliters ofthe cDNA reaction mixture were diluted 5-times inDEPC-treated water to prevent inhibition of Taqpolymerase in subsequent PCR reaction. One microliterof diluted cDNA product was used in each PCR reaction.Amplifications were performed in 10 µl PCR reactionmixtures containing QuantiTect SYBR Green PCRMaster Mix (QIAGEN) and 0.5 µM of each humangene-specific primers for γ-tubulin 1 (TUBG1, NCBIRefseq ID: NM_001070), γ-tubulin 2 (TUBG2, NCBIRefseq ID: NM_016437) or ß-actin (ACTB, NCBIRefseq ID: NM_001101). Primer sequences were asfollows: γ-tubulin 1, forward 5’-CCCTCATCTGCCTTACTGGTTG -3’, γ-tubulin 1, reverse 5’-AGGTCCCTGATCTGTGCTCTGA -3’; γ-tubulin 2, forward 5’-GGAGCTCATTGATGAGTACCATG-3’, γ-tubulin 2,reverse 5’-AGGAGAAGGAGTAGTGGGGAG-3’ andß-actin forward 5’-TCCTTCCTGGGCATGGAGT-3’, ß-actin, reverse 5’-AAAGCCATGCCAATCTCATC-3’.Oligonucleotides were from East Port (Prague, CzechRepublic). RT-qPCRs were performed on Master-cyclerrealplex (Eppendorf, Wesseling-Berzdorf, Germany) asdescribed (Dráberová et al., 2008). Experiments wereperformed twice with triplicate samples (cDNA isolatedfrom three separate cultures of tested cells). Theexpression of analyzed genes was normalized to theexpression of ß-actin. Levels of ß-actin did not differsignificantly between tested cell types. PCR efficiencies(E) for probed genes were calculated from calibrationcurves by the Realplex 1.5 software. Calculation of thenormalized relative quantity (NRQ) of evaluatedtranscripts was based on the following formula:

taking into account different PCR efficiencies of thegene of interest (goi) and reference gene (ref) used fornormalization (Hellemans et al., 2007). Identity of PCRproduct was verified by sequencing. Statistical analysiswas performed with the Student’s two-tailed unpaired t-test using Microsoft Excel.Gel electrophoresis and immunoblotting

Whole-cell extracts were prepared by washing thecells in cold PBS, solubilizing them in hot SDS-samplebuffer (Laemmli, 1970) and boiling for 5 min. SDS-PAGE on 7.5 % gels, electrophoretic transfer ofseparated proteins onto nitrocellulose and details of theimmunostaining procedure have been describedelsewhere (Dráberová et al., 1986). Quantification ofproteins in SDS-sample buffer by colloidal silvermethod was described previously (Dráber, 1991).Polyclonal antibodies to γ-tubulin and DQ19 (SigmaT5192 and T3195 respectively) were diluted 1:10,000.Monoclonal antibodies to γ-tubulin TU-30 and TU-32were used as 10 times diluted supernatants, while

monoclonal antibody GTU-88 and polyclonal antibodyto actin were diluted 1:10,000 and 1:3,000 respectively.Bound primary antibodies were detected after incubationof the blots with HRP-conjugated secondary antibodiesdiluted 1:10,000. HRP signal was detected withchemiluminiscence reagents (Pierce, Rockford, IL) andquantified using LAS 3000 imaging system (Fujifilm,Tokyo, Japan) and AIDA (ver. 4) software (Raytest,Straubenhardt, Germany). Statistical analysis wasperformed with the Student’s two-tailed unpaired t-testusing Microsoft Excel.Immunofluorescence on cells in culture

Immunofluorescence microscopy on methanol-fixedcells was performed as described previously (Dráberováand Dráber, 1993). TU-30 antibody was used asundiluted supernatant. Cy3-conjugated anti-mouseantibodies was diluted 1:500. The preparations weremounted in MOWIOL 4-88 (Calbiochem, San Diego,CA) and examined with an Olympus A70 Provismicroscope. Conjugate alone did not give any detectableimmunoreactivity.Clinical tumor samples

Formaldehyde-fixed, paraffin-embedded biopsy/resection samples of NSCLC (n=109) were collectedretrospectively from “KAT” and “Sismanoglion”General Hospitals, Athens, Greece. Ninety-three patientswere male (86.32 %) the mean being 63 years (range 40to 81 years). The lung tumors were classifiedhistologically according to the WHO criteria as follows:AC (n=56), SCC (n=49), adenosquamous carcinoma(n=3) and LCC (n=1). Among ACs, 52 were classifiedas “not otherwise specified,” 2 as bronchioalveolar, 1 asclear cell, and 1 as fetal. Furthermore, the degree ofdifferentiation was evaluated, distinguishing betweenwell differentiated (n=12), moderately differentiated(n=41) or poorly differentiated tumors (n=56) (grades Ithrough III respectively). From the total number of casesexamined, clinical staging information was available in63 cases as follows: Stage I (n=29), Stage II (n=11) andStage III (n=23). Control tissues included 10 age-matched non-neoplastic lung biopsy specimens, 5showing varying degrees of interstitial fibrosis and 5with interstitial pneumonitis. Histologic preparationswere evaluated independently by four pathologists, (NM,EM, MC, CDK), who were unaware of the originalpathological diagnosis. In the case of disagreementbetween observers, the diagnosis was assigned byconsensus.Immunohistochemistry on clinical tumor samples

Immunohistochemistry was performed according tothe avidin biotin complex (ABC) peroxidase methodusing Rabbit and Mouse IgG ABC Elite® detection kits(Vector Labs, Burlingame, CA) as previously described


(Katsetos et al., 2006, 2007). Prior to the performance ofthe immunohistochemical procedure, 5 µm thickhistological sections from paraffin-embedded tissueblocks were subjected to non-enzymatic antigenunmasking in 0.01 M sodium citrate buffer (pH 6.0) for10 minutes in a microwave at medium power. Bothrabbit polyclonal antibodies to γ-tubulin (Sigma T5192,T3195) were used at 1:500 dilution, whereas anti-γ-tubulin monoclonal antibodies GTU-88 and TU-32 wereused at a 1:300 dilution, or as undiluted supernatants,respectively. Negative controls included omission ofprimary antibody and substitution with nonspecificmouse IgG1 and IgG2a, which were used asimmunoglobulin class-specific controls (correspondingto the immunoglobulin subclasses of the primaryantibodies employed in this study) (Becton Dickinson,Franklin Lakes, NJ). Preparations using non-conjugatedisotype matched control monoclonal antibodies did notshow any non-specific binding of the secondary rabbitanti-mouse IgG1 and IgG2a antibodies. For theevaluation of immunoperoxidase staining, a NikkonEclipse E400 brightfield microscope equipped with adigital Leica camera (model BFC 280) was used.Cell counting

Manual cell counting of labeled tumor cells wasperformed by three observers independently (NM, EM,CDK). Cell counting and statistical analysis wereperformed with all antibodies employed. The cases weregrouped according to histological subtype, grade andstage. To assess the fraction of immunolabeled tumorcells in each specimen, the labeling index (LI), definedas the percentage (%) of γ-tubulin-labeled cells out ofthe total number of tumor cells counted in each case andfor each antibody, was determined as previouslydescribed (Katsetos et al., 2001). Between 258 and 805tumor cells were evaluated per case, in 10 non-overlapping high-power (40x) fields, and LIs weregenerated per each specimen. The criteria for theidentification of a γ-tubulin-positive cell were thedetection of 3 or more punctate/dot-like immunoreactivesignals and/or robust diffuse staining in the cytoplasm ofindividual tumor cells as previously described (Katsetoset al., 2006). The level of interobserver agreement wasquantitated using generalized kappa and pairwise kappastatistics (Fleiss, 1981; Landis and Koch, 1997).Interobserver agreement was substantial (k=0.65). In thecase of disagreement (+/- 10 % in the recording of the LIbetween the two observers) the immunohistochemicalpreparations were re-reviewed at a multi-headedmicroscope in order to achieve consensus. For purposesof LI recording, the consensus opinion was consideredas conclusive. For statistical analysis SPSS software wasused (SPSS for Windows, version 13, IBM SPSS,Chicago, IL). All p-values were two-tailed and 5% waschosen as the level of statistical significance. Correlationof γ-tubulin with baseline characteristics was performedusing Mann-Whitney, Kruskal-Wallis and χ2 tests. A p

value of less than 0.05 was considered as statisticallysignificant. Because of the small number ofadenosquamous and LCC included in this study, thesecases were excluded from the statistical analysis. Immunofluorescence on clinical tumor samples

Deparaffinized histological sections from selectedsurgically excised NSCLC tissue specimens wereutilized for immunofluorescence microscopy. Prior tothe performance of the immunological staining, thetissue sections were subjected to non-enzymatic antigenunmasking in 0.01 M sodium citrate buffer (pH 6.0) for10 min in a microwave at medium power. The slideswere then rinsed with PBS and incubated for 1 hour inblocking solution (6% BSA in PBS). The mousemonoclonal antibody to Ki-67 (clone MIB-1) (sc-101861) (Santa Cruz Biotechnology, Inc., Santa Cruz,CA) was used at 1:100 dilution in PBS containing 1%BSA. The rabbit polyclonal antibody to γ-tubulin (DQ-19) was used at 1:500 dilution in PBS containing 1%BSA. Secondary fluorochrome conjugated antibodieswhich included goat anti-mouse AlexaFluor 488 (green)and goat anti-rabbit AlexaFluor 594 (red) (Invitrogen,Carlsbad, CA) were diluted 1:600 with PBS containing1% BSA. Double staining was performed in twosequential sessions of immunofluorescence staining. Atthe end of the staining the slides were incubated for 15minutes in a freshly made PBS solution containing1mg/ml of sodium borohydride (Sigma) to decrease theautofluorescence, and subsequently they were rinsedextensively with PBS. 4’,6-diamidino-2-phenylindole(DAPI) was used to label cell nuclei. Slides were cover-slipped with Vectashield mounting medium (VectorLabs, Burlington, CA) and visualized with an OlympusIX81 deconvolution fluorescence microscope.Results

Immunohistochemical localization in clinical tumorsamples

In non-neoplastic cells of the ciliated airwayepithelium, γ-tubulin exhibited predominantly apicalsurface (Fig. 1a-c) and pericentriolar/centrosomallocalizations (Fig. 1d –thick arrow) and to a lesser extenta soluble distribution (Fig. 1b,c). Apical surface labelingwas localized to the base of the cilia in the form of alinear appearance corresponding to the basal bodies (Fig.1a,b -curved arrow, 1c - thick arrows). In contrast, thecilia themselves were uniformly γ-tubulin-negative (Fig.1a,b –straight arrow, 1c).

Overall, a variously prominent, albeit distinctlyheterogeneous γ-tubulin distribution was detected in allNSCLC specimens (Figs. 2,3). Varying degrees of γ-tubulin labeling were detected in all histological types(SCCs and ACs) and tumor grades (I-III). In SCCs, γ-tubulin labeling was noted focally either amongindividual tumor cells (Fig. 2a - straight arrow) or in


clusters of neoplastic cells within tumor sheets (Fig. 2b).Occasional tumor cells exhibited staining in theperiphery of the cytoplasm (Fig. 2c - straight arrow).Strong cytoplasmic staining was noted in large,pleomorphic cells, surrounding areas of necrosis,including in tadpole-shaped tumor cells (Fig. 2d-f), somewith γ-tubulin-positive nuclear pseudoinclusions (Fig. 2f- thick arrow). γ-Tubulin immunoreactive cells displayedan overall tendency for the periphery of tumor sheetsand for perivascular areas (not shown). Spindle cellswere less immunoreactive as compared to largepleomorphic cells (not shown).

In ACs, there was variable γ-tubulin staining inglandular (Fig. 3a-c, 3e-short arrows) and papillary areas(Fig. 3d) with a trend for increased immunoreactivity inless differentiated areas (Fig. 3e- straight arrows, 3f-i).Strong γ-tubulin labeling was observed in small clustersof invasive poorly differentiated tumor cells (Fig. 3g -straight arrows). However, in some tumors, well-differentiated glandular areas exhibited robust anddiffuse cytoplasmic staining (Fig. 3a-c). In gland-

forming tumor cells, localization was diffuse but in sometumor cells was more pronounced in the apical surfaceportion of the cytoplasm (Fig. 3c - short arrow)recapitulating the superficial apical localization of thisprotein in normal ciliated bronchial epithelium localizedto the base of the cilia (Fig. 1a-d). In 2 well-differentiated papillary ACs, apical protrusions incolumnar cells of tumor glands, resembling cilia, were γ-tubulin negative (Fig. 3c and 3e-short arrows). In bothSCCs and ACs immunoreactivity for γ-tubulin washeterogeneous and, where present, it was characterizedby overlapping multi-punctate and diffuse stainingpatterns (Figs. 2, 3e,f,h - thick arrow, and 3i). Identicalimmunoreactivity profiles were obtained with thedifferent polyclonal and monoclonal anti-γ-tubulinantibodies employed.

The mean LIs for SCCs versus ACs relative tohistological grade are tabulated on Table 1 and themedian LIs of combined NSCLCs relative to stage areshown in box plot form on Figure 4. Grade-for-grade, atrend toward increased immunoreactivity was observed


Fig. 1. Immunolocalization ofγ-tubulin in the bronchiolar (a-c) and bronchial (d) epitheliausing monoclonal antibodyGTU-88. Note localizations inthe apical surface at the baseof the cilia (a, b -curved arrow,and c -thick arrows) withabsence of labeling in cilia (a,b-straight arrow, c, and d).Also, note juxtanuclearpericentriolar/centrosomallocalization (d -thick arrow).Avidin biotin complex (ABC)peroxidase with hematoxylincounterstaining. Scale bars:20 µm.

in grade 3 as compared to grade 1 tumors, regardless ofhistological subtype, which, however, was notstatistically significant. Conversely, a statisticallysignificant difference between expression of γ-tubulinand clinical stage -- regardless of histological subtype ortumor grade-- was found, with the highest mean LIdetected in Stage III tumors (p<0.001 vs. Stage I andStage II) (Fig. 4).

Co-expression of γ-tubulin and Ki-67 (MIB-1) wasdetected in populations of non-proliferating (Fig. 5a-c)and proliferating tumor cells (Fig. 5d-f) from allhistological types and grades of NSCLC. Thelocalization of γ-tubulin in tumor cells waspredominantly cytoplasmic (diffuse and punctatepatterns) (Fig. 5a,c) in contrast to the localization of Ki-67, which was predominantly nuclear (Fig. 5b,e).Nuclear co-localization of γ-tubulin and Ki-67 (MIB-1)was also detected in a subpopulation of proliferatingtumor cells (Fig. 5d-f).Expression and cellular localization of γ-tubulin in H460and H69 versus SAEC

Immunofluorescence microscopy revealed moreprominent γ-tubulin staining in cytoplasm and MTOCsin H460 cells as compared to SAEC (Fig. 6A, panels “b”

and “a” respectively). It is well established that anti-γ-tubulin antibodies intensively label dots inparanuclear/juxta-nuclear positions, which correspond toMTOCs (Nováková et al., 1996), as confirmed also bystaining with anti-pericentrin antibody (Katsetos et al.,2006). It was difficult to compare γ-tubulin staining inH69 with that in SAEC, as H69 cells are substantially


Fig. 2. Immunolocalization of γ-tubulin in squamous cell carcinomas (SCCs) using rabbit polyclonal antibody (Sigma, T5192) (a-c) and mousemonoclonal antibody GTU-88 (Sigma, T6657) (d-f) both of which are specific for the sequence 38-53 of γ-tubulin. c is a higher magnification of panel a. ABC peroxidase with hematoxylin counterstaining. Scale bars: a, 100 µm; b, c, 50 µm; d-f, 20 µm.

Table 1. γ- Tubulin labeling index versus histologic type and grade inNSCLC

HISTOLOGIC TYPE GRADE Number Mean LI (%) Std. Deviation

SCC 1 5 49.22 0.372 18 67.29 0.233 26 76.60 0.19

Total 49 70.38 0.38Adenocarcinoma 1 6 56.58 0.34

2 22 78.63 0.153 28 76.38 0.15

Total 56 75.14 0.18

LI: Labeling index, defined as the percentage (%) of γ-tubulin-labeledcells out of the total number of tumor cells counted in each case and foreach antibody, as previously described (Katsetos et al., 2006). SCC:Squamous cell carcinoma

smaller and grow in clusters in suspension. However,when H69 cells were attached to poly-L-lysine coatedcoverslips, prominent γ-tubulin staining in cytoplasmand MTOC was observed (not shown). Immunoblotanalysis with monoclonal antibody TU-32 revealeddifferential staining of blots prepared from total celllysates of SAEC, H460 and H69 cells (Fig. 6B). Arepresentative example of immunoblot staining anddensitometric quantification of immunostaining ispresented in adjacent panels in Figure 6B revealing thatH460 and H69 cells expressed substantially higher levelsof γ-tubulin protein as compared to SAEC. Similardifferential staining of blots was observed with the othermonoclonal antibodies (TU-30, GTU-88), as well aswith both polyclonal antibodies (Sigma, T3195, T5192)to γ-tubulin (not shown).

By RT-qPCR, γ-tubulin transcripts for TUBG1 and

TUBG2 genes were detected in SAEC, H460, and H69cells. However, transcripts for both TUBG1 and TUBG2genes were more abundant in H460 and H69 cells whencompared with SAEC (Fig. 7). Altogether the dataclearly indicate, both at the protein and mRNA levels,that γ-tubulins are overexpressed in NSCLC cell lines ascompared to SAEC.Discussion

Expression and cellular distribution of γ-tubulin in NSCLCcells

In the present study we have produced novelevidence for an increased γ-tubulin expression insurgically excised NSCLC specimens and in the NCI-H460 and NCI-H69 NSCLC cell lines. This is, to our


Fig. 3. Immunolocalization of γ-tubulin in adenocarcinomas (ACs) using monoclonal antibodies TU-32 specific for sequence 434-449 (a and b) andGTU-88 (c–i). h is a higher magnification from the same field as the one depicted in panel g. a is representative of a putative bronchioalveolarcarcinoma and the rest correspond to pulmonary ACs (not otherwise specified) with different degrees of differentiation. ABC peroxidase withhematoxylin counterstaining. Scale bars: a, b, d-f, 50 µm; c, h, i, 20 µm; g, 100 µm.

knowledge, the first study to demonstrate increasedexpression of γ-tubulin in lung cancer.

Diploid cells contain one or two juxtanuclearcentrosomes typified by pericentriolar staining for γ-tubulin (Katsetos et al., 2006). This pattern is faithfullyreproduced in cells of the human respiratory epitheliumwherein in addition to pericentriolar γ-tubulin stainingthere is evidence of localization in apical adluminalcytoplasmic region near the ciliated brush border. Thislocalization corresponds to putative basal bodies at thebase of the cilia in multiciliated epithelia of the bovinetrachea and oviduct (Muresan et al., 1993), retinalphotoreceptors (Muresan et al., 1993) and in mousecochlear epithelial cells (Mogensen et al., 1997).

Although NSCLC cells recapitulate -in part- thisapical surface and pericentriolar patterns of distribution,they also exhibit variously prominent, diffusecytoplasmic γ-tubulin staining indicating that thisprotein, which is normally associated with basal bodiesand centrosomes in ciliated epithelia, undergoes alteredcompartmentalization in tumors where it is eitherincorporated into insoluble aggregates, associated withmembraneous components, or becomes part of anincreased soluble pool in the cytoplasm of tumor cells.Cho and coworkers have shown that aggressive breastcancer cell lines with metastatic potential exhibit a


Fig. 4. Box plot of γ-tubulin labeling indices versus stage (p<0.001).Boxes correspond to the values between the 25th and 75th percentiles[Q1-Q3]; the dark line in the middle is the median value; upper andlower whiskers to the highest and lower values of distribution; values incircles are at least 1.5 greater than the 75th percentile; and value inasterisk is at least 3.0 greater than the 75th percentile.

Fig. 5. Double labeling of γ-tubulin and Ki-67 (MIB-1) in clinical tumor samples as determined by immunofluorescence microscopy. Note localization ofγ-tubulin (a and d, red) and Ki-67 (MIB-1) (b and e, green) in tumor cells. a-c depict distinct localizations of γ-tubulin and Ki-67 (MIB-1) in populations ofnon-proliferating and proliferating tumor cells respectively. Conversely, d-f illustrate a pattern of (predominantly cytoplasmic) γ-tubulin and (nuclear) Ki-67 (MIB-1) co-expression in individual tumor cells (representative field from a case of poorly differentiated adenocarcinoma, Grade 3, Stage III). Also,note co-localization of both antigens in the nuclei of a subpopulation of poorly differentiated tumor cells. Scale bar: 50 µm.

diffuse soluble, and more dispersive, subcellularlocalization of γ-tubulin as compared to non-invasivecell lines in which the localization of this protein islargely centrosomal (Cho et al., 2010). In addition to itstraditional role in microtubule nucleation at thepericentriolar material of centrosomes, γ-tubulin can befound in cytosolic non-centrosomal nucleating-sites(Chabin-Brion et al., 2001; Efimov et al., 2007; Macureket al., 2008) as well as in nuclei/nucleoli (Hořejší et al.,2012), as was also demonstrated in NSCLC cells in thepresent study.Translational studies on γ-tubulin overexpression incancer

Centrosome abnormalities in cancer arecharacterized by increased, or otherwise altered,expression of centrosomal proteins, including γ-tubulin

and pericentrin (Pihan et al., 1998, 2001; Sato et al.,1999; Kuo et al., 2000; Setoguchi et al., 2001; Katsetoset al., 2006, 2007, 2009). We and others have previouslyshown γ-tubulin overexpression in glioblastomas(Katsetos et al., 2006, 2007, 2009; Loh et al., 2010) andmedulloblastomas (Caracciolo et al., 2010). The presentstudy has revealed a statistically significant increase ofγ-tubulin expression in Stage III as compared to lesserstage tumors without statistically significant differencesin the expression of this protein according to histologicalsubtype or tumor grade. This suggests that γ-tubulin mayserve potentially as a prognostic/predictive biomarker inthe setting of NSCLC. However, the latter requiresfuture validation in a larger cohort of cases with survivaldata. A recent study has shown that γ-tubulin may be apromising marker of recurrence of SCC of the larynx(Syed et al., 2009). Interestingly, increased γ-tubulinimmunostaining in laryngeal cancer has been found to be


Fig. 6. Comparison of γ-tubulin expression innon-transformed small airway epithelial cells(SAEC) and NSCLC cell lines (H460 and H69).A. Immunofluorescence staining of SAEC (a)and H460 cells (b) with monoclonal antibodyTU-30 to γ-tubulin. Images were collected andprocessed in exactly the same manner. Fixationwith methanol. Scale bar: 20 µm. B.Immunoblot analysis of whole cell extracts fromSAEC, H460, and H69 cells. Blots wereimmunostained with monoclonal antibody TU-32 to γ-tubulin and polyclonal antibody to actin.Densitometric quantification of immunoblot isshown on the right. Data are presented as themean fold change ± SD obtained from threeindependent experiments with tr ipl icatesamples. Mean values different from SAEClevels are indicated: ***, p<0.001.

a better predictor of tumor recurrence than the T stage,especially in node-negative (N0) tumors (Syed et al.,2009).

Increased γ-tubulin expression occurs in in-situ andinfiltrating breast carcinomas (Liu et al., 2009; Niu et al.,2009). Alterations of γ-tubulin expression andsubcellular sorting may also play a role in earlytumorigenesis (Nguyen et al., 2007).Expression of the two γ-tubulin functional genes inNSCLC

Compared to the multiple genes encoding for α andß-tubulin isotypes (Banerjee et al., 2008), there are onlytwo functional genes in mammalian cells (TUBG1,TUBG2) that code very similar γ-tubulin species (Wiseet al., 2000; Yuba-Kubo et al., 2005). TUBG1 isubiquitously expressed, whereas TUBG2 is expressedmainly in the brain (Yuba-Kubo et al., 2005). Consistentwith this differential tissue distribution pattern betweenthe two functional γ-tubulin genes, in the present studywe have observed an increase both in TUBG1 andTUBG2 by RT-qPCR. Conversely, molecular profilinghas revealed overexpression of TUBG2 in prostatecancer cells (Li et al., 2005) and in glioblastoma celllines (Katsetos et al., 2010). In thyroid carcinomas(Montero-Conde et al., 2008), gliomas (Rickman et al.,2001), pediatric pilocytic astrocytomas (Potter et al.,2008), and in medulloblastomas (Caracciolo et al., 2010)an increase in TUBG1 was detected. Moreover, changesin the expression of both TUBG1 and TUBG2 geneshave been reported in breast cancer cells (Orsetti et al.,2004).Conclusion

In summary, our results show for the first time in thecontext of lung cancer that NSCLC cells from surgicallyexcised tumor samples are accompanied by highlyprominent expression and ectopic cellular distribution of

γ-tubulin, as compared to non-neoplastic cells of therespiratory epithelium. However, the significance of thisoverexpression and altered compartmentalizationremains to be determined in future functional studiesfocusing on the role of this protein in ectopicmicrotubule nucleation, and cell cycle progression.

Given the well-established overexpression of ßIII-tubulin in NSCLCs (Katsetos et al., 2000; Sève andDumontet, 2005; Sève et al., 2010), it is unclear whetherincreased co-expression of ßIII-tubulin/γ-tubulinenhances or diminishes chemotherapeutic efficacy inNSCLC subsets. This provides the rationale for theperformance of further studies focusing on the effect ofnovel tubulin binding agents on γ-tubulin-overexpressingor silenced human NSCLC cells in vitro or in xenografts.In addition, large clinical randomized studies arenecessary to further elucidate the prognostic orpredictive value of γ-tubulin in the context of differentclinical stages, histological types, tumor grades, andtreatment settings.Acknowledgements. CDK dedicates this research article to his formermentor in Pulmonary Pathology, Professor Carl Teplitz, MD, AM (adeundem - Brown), Emeritus Chairman, Department of Pathology andLaboratory Medicine, Beth Israel Medical Center, New York City, on theoccasion of his 80th birthday. Grant information: Contract grant sponsor:Grant Agency of the Czech Republic; Contract grant number:204/09/1777 (PD) and P302/10/1701 (ED). Contract grant sponsor:Ministry of Education, Youth and Sports of the Czech Republic; Contractgrant numbers: LC545 (TS): Contract grant sponsor: InstitutionalResearch Support; Contract number: AVOZ 50520514 (PD); Contractgrant sponsor: St. Christopher’s Foundation for Children (Philadelphia,USA); Contract number: 203 (CDK)

References

Banerjee A., Jensen-Smith H., Lazzell A., Prasad V., Elguezabal G.,Hallworth R. and Ludueña R.F. (2008). Localization of ßv tubulin inthe cochlea and cultured cells with a novel monoclonal antibody.


Fig. 7. Transcription of genes for γ-tubulin 1 (TUBG1) and γ-tubulin2 (TUBG2) in lung cell carcinoma cell lines H460 and H69 relativeto the level in non-transformed small airway epithelial cells(SAEC). Data are presented as the mean fold change ± SDobtained from two independent experiments with triplicatesamples. Mean values different from SAEC levels are indicated: **,p<0.005.

Cell Motil. Cytoskeleton 65, 505-514.Bugnard E., Zaal K.J. and Ralston E. (2005). Reorganization of

microtubule nucleation during muscle differentiation. Cell Motil.Cytoskeleton 60, 1-13.

Caracciolo V., D’Agostino L., Dráberová E., Sládková V., Crozier-Fitzgerald C., Agamanolis D.P., de Chadarévian J.-P., Legido A.,Giordano A., Dráber P. and Katsetos C.D. (2010). Differentialexpression and cellular distribution of γ-tubulin and ßIII-tubulin inmedulloblastomas and human medulloblastoma cell lines. J. Cell.Physiol. 223, 519-529.

Chabin-Brion K., Marceiller J., Perez F., Settegrana C., Drechou A.,Durand G. and Poüs C. (2001). The Golgi complex is a microtubule-organizing organelle. Mol. Biol. Cell 12, 2047-2060.

Cho E.H., Whipple R.A., Matrone M.A., Balzer E.M. and Martin S.S.(2010). Delocalization of γ-tubulin due to increased solubility inhuman breast cancer cell lines. Cancer Biol. Ther. 9, 66-76.

Dráber P. (1991). Quantitation of proteins in sample buffer for sodiumdodecyl sulfate-polyacrylamide gel electrophoresis using colloidalsilver. Electrophoresis 12, 453-456.

Dráber P., Dráberová E. and Nováková M. (1995). Stability ofmonoclonal IgM antibodies freeze-dried in the presence oftrehalose. J. Immunol. Methods 181, 37-43.

Dráberová E. and Dráber P. (1993). A microtubule-interacting proteininvolved in coalignment of vimentin intermediate filaments withmicrotubules. J. Cell Sci. 106, 1263-1273.

Dráberová E., Dráber P., Havlíček F. and Viklický V. (1986). A commonantigenic determinant of vimentin and desmin defined bymonoclonal antibody. Fol. biol. (Praha) 32, 295-303.

Dráberová E., Del Valle L., Gordon J., Marková V., Smejkalová B.,Bertrand L., de Chadarévian J.-P., Agamanolis D.P., Legido A.,Khalili K., Dráber P. and Katsetos C.D. (2008). Class III ß-tubulin isconstitutively coexpressed with glial fibrillary acidic protein andnestin in midgestational human fetal astrocytes: Implications forphenotypic identity. J. Neuropathol. Exp. Neurol. 67, 341-54.

Dryková D., Cenklová V., Sulimenko V., Volc J., Dráber P. and BinarováP. (2003). Plant γ-tubulin interacts with alphabeta-tubulin dimers andforms membrane-associated complexes. Plant Cell 15, 465-480.

Dumontet C., Isaac S., Souquet P.J., Bejui-Thivolet F., Pacheco Y.,Peloux N., Frankfurter A., Ludueña R. and Perol M. (2005).Expression of class III ß-tubulin in non-small cell lung cancer iscorrelated with resistance to taxane chemotherapy. Bull. Cancer 92,E25-30.

Efimov A., Kharitonov A., Efimova N., Loncarek J., Miller P.M.,Andreyeva N., Gleeson P., Galjart N., Maia A.R., McLeod I.X., YatesJ.R. III, Maiato H., Khodjakov A., Akhmanova A. and Kaverina I.(2007). Asymmetric CLASP-dependent nucleation ofnoncentrosomal microtubules at the trans-Golgi network. Dev. Cell12, 917-930.

Fleiss J.G. (1981). Statistical methods for rates and proportions. Wiley.New York. pp 225-232.

Hellemans J., Mortier G, De Paepe A, Speleman F and VandesompeleJ. (2007). qBase relative quantification framework and software formanagement and automated analysis of real-time quantitative PCRdata. Genome Biol. 8, R19.

Hirsch F.R., Speafico A., Novello S. and Wood M.D., Simms L. andPapotti M. (2008). The prognostic and predictive role of histology inadvanced non-small cell lung cancer: a literature review. J. Thorac.Oncol. 3, 1468-1481.

Höög G., Zarrizi R., von Stedingk K., Jonsson K. and Alvarado-Kristensson M. (2011). Nuclear localization of γ-tubulin affects E2F

transcriptional activity and S-phase progression. FASEB J. 25,3815-3827.

Hořejší B., Vinopal S., Sládková V., Dráberová E., Sulimenko V.,Sulimenko T., Vosecká V., Philimonenko A., Hozák P., KatsetosC.D. and Dráber P. (2012). Nuclear gamma-tubulin associates withnucleoli and interacts with tumor suppressor protein C53. J. Cell.Physiol. 227, 367-382

Katsetos C.D., Kontogeorgos G., Geddes J.F., Herman M.M., Tsimara-Papastamatiou H., Yu Y., Sakkas L.I., Tsokos M., Patchefsky A.S.,Ehya H., Cooper H.S., Provencio J., Spano A.J. and Frankfurter A.(2000). Differential distribution of the neuron-associated class III ß-tubulin in neuroendocrine lung tumors. Arch. Pathol. Lab. Med. 124,535-544.

Katsetos C.D., Del Valle L., Geddes J.F., Assimakopoulou M., LegidoA., Boyd J.C., Balin B., Parikh N.A., Maraziotis T., de ChadarevianJ.-P., Varakis J.N., Matsas R., Spano A., Frankfurter A., HermanM.M. and Khalili K. (2001). Aberrant localization of the neuronalclass III ß-tubulin in astrocytomas. Arch. Pathol. Lab. Med. 125, 613-624.

Katsetos C.D., Reddy G., Dráberová E., Šmejkalová B., Del Valle L.,Ashraf Q., Tadevosyan A., Yelin K., Maraziotis T., Mishra O.P., MörkS., Legido A., Nissanov J., Baas P.W., de Chadarévian J.-P. andDráber P. (2006). Altered cellular distribution and subcellular sortingof γ-tubulin in diffuse astrocytic gliomas and human glioblastoma celllines. J. Neuropathol. Exp. Neurol. 65, 465–77.

Katsetos C.D., Dráberová E., Šmejkalová B., Reddy G., Bertrand L., deChadarévian J.-P., Legido A., Nissanov J., Baas P.W. and Dráber P.(2007). Class III ß-tubulin and γ-tubulin are co-expressed and formcomplexes in human glioblastoma cells. Neurochem. Res. 32, 1387-1398.

Katsetos C.D., Dráberová E., Legido A. and Dráber P. (2009). Tubulintargets in the pathobiology and therapy of glioblastoma multiforme.II. γ-Tubulin. J. Cell. Physiol. 221, 514–20.

Katsetos C.D., Dráberová E., Sládková V., Hořejší B. and Dráber P.(2010). Overexpression of γ-tubulin transcripts in humanglioblastoma cell lines. J Neuropathol. Exp. Neurol. 69, 539 (Abstr.)

Kuo K.K., Sato N., Mizumoto K., Maehara N., Yonemasu H., Ker C.G.,Sheen P.C. and Tanaka M. (2000). Centrosome abnormalities inhuman carcinomas of the gallbladder and intrahepatic andextrahepatic bile ducts. Hepatology 31, 59-64.

Laemmli U.K. (1970). Cleavage of structural proteins during theassembly of the head of bacteriophage T4. Nature 227, 680-5.

Landis J.R. and Koch G.G. (1977). An application of hierarchical kappa-type statistics in the assessment of majority agreement amongmultiple observers. Biometrics 33, 363-374.

Li Y., Hussain M., Sarkar S.H., Eliason J., Li R. and Sarkar F.H. (2005).Gene expression profiling revealed novel mechanism of action ofTaxotere and Furtulon in prostate cancer cells. BMC Cancer 5, 7.

Liu T., Niu Y., Yu Y., Liu Y. and Zhang F. (2009). Increased γ-tubulinexpression and P16INK4A promoter methylation occur together inpreinvasive lesions and carcinomas of the breast. Ann. Oncol. 20,441-448.

Loh J.K., Lieu A.S., Chou C.H., Lin F.Y., Wu C.H., Howng S.L., ChioC.C. and Hong Y.R. (2010) Differential expression of centrosomalproteins at different stages of human glioma. BMC Cancer 10, 268.

Macurek L., Dráberová E., Richterová V., Sulimenko V., Sulimenko T.,Dráberová L., Marková V. and Dráber P. (2008). Regulation ofmicrotubule nucleation from membranes by complexes ofmembrane-bound γ-tubulin with Fyn kinase and phosphoinositide 3-kinase. Biochem. J. 416, 421-430.


Mogensen M.M., Mackie J.B., Doxsey S.J., Stearns T. and Tucker J.B.(1997). Centrosomal deployment of γ-tubulin and pericentrin:evidence for a microtubule-nucleating domain and a minus-enddocking domain in certain mouse epithelial cells. Cell Motil.Cytoskeleton 36, 276-90.

Montero-Conde C., Martín-Campos J.M., Lerma E., Gimenez G.,Martínez-Guitarte J.L., Combalía N., Montaner D., Matías-Guiu X.,Dopazo J., de Leiva A., Robledo M. and Mauricio D. (2008).Molecular profiling related to poor prognosis in thyroid carcinoma.Combining gene expression data and biological information.Oncogene 27, 1554-1561.

Moritz M., Braunfeld M.B., Sedan J.W., Alberts B. and Agard D.A.(1995). Microtubule nucleation by γ-tubulin-containing rings in thecentrosome. Nature 378, 638-640.

Moritz M., Zheng Y., Alberts B.M. and Oegema K. (1998). Recruitmentof the γ-tubulin ring complex to Drosophila salt-stripped centrosomescaffolds. J. Cell Biol. 142, 775-786.

Moudjou M., Bordes N., Paintrand M. and Bornens M. (1996). γ-Tubulinin mammalian cells: the centrosomal and the cytosolic forms. J. CellSci. 109, 875-887.

Muresan V., Joshi H.C. and Besharse J.C. (1993). γ-Tubulin indifferentiated cell types: localization in thc vicinity of basal bodies inretinal photoreceptors and ciliated epithelia. J. Cell Sci. 104, 1229-1237.

Nguyen C.L., Eichwald C., Nibert M.L. and Münger K. (2007). Humanpapillomavirus type 16 E7 oncoprotein associates with thecentrosomal component γ-tubulin. J. Virol. 81, 13533-13543.

Niu Y., Liu T., Tse G.M., Sun B., Niu R., Li H.M., Wang H., Yang Y., YeX., Wang Y., Yu Q. and Zhang F. (2009). Increased expression ofcentrosomal α, γ-tubulin in atypical ductal hyperplasia andcarcinoma of the breast. Cancer Sci. 100, 580-587.

Nováková M., Dráberová E., Schürmann W., Czihak G., Viklický V. andDráber P. (1996). γ-Tubulin redistribution in taxol-treated mitotic cellsprobed by monoclonal antibodies. Cell Motil. Cytoskeleton 33, 38-51.

Orsetti B., Nugoli M., Cervera N., Lasorsa L., Chuchana P., Ursule L.,Nguyen C., Redon R., du Manoir S., Rodriguez C. and Theillet C.(2004). Genomic and expression profiling of chromosome 17 inbreast cancer reveals complex patterns of alterations and novelcandidate genes. Cancer Res. 64, 6453–6460.

Pihan G.A., Purohit A., Wallace J., Knecht H., Woda B., Quesenberry P.and Doxsey S.J. (1998). Centrosome defects and genetic instabilityin malignant tumors. Cancer Res. 58, 3974-3985.

Pihan G.A., Purohit A., Wallace J., Malhotra R., Liotta L. and DoxseyS.J. (2001). Centrosome defects can account for cellular andgenetic changes that characterize prostate cancer progression.Cancer Res. 61, 2212–2229.

Potter N., Karakoula A., Phipps K.P., Harkness W., Hayward R.,Thompson D.N.P., Jacques T.S., Harding B., Thomas D.G.T..,Palmer R.W., Rees J., Darling J. and Warr T.J. (2008). Genomicdeletions correlate with underexpression of novel candidate genesat six loci in pediatric pilocytic astrocytoma. Neoplasia 10, 757-777.

Rickman D.S., Bobek M.P., Misek D.E., Kuick R., Blaivas M., KurnitD.M., Taylor J. and Hanash S.M. (2001). Distinctive molecularprofi les of high-grade and low-grade gliomas based onoligonucleotide microarray analysis. Cancer Res. 61, 6885-6891.

Sato N., Mizumoto K., Nakamura M., Nakamura K., Kusumoto M.,Niiyama H., Ogawa T. and Tanaka M. (1999). Centrosomeabnormalities in pancreatic ductal carcinoma. Clin. Cancer Res. 5,963-970.

Schnackenberg B.J. and Palazzo R.E. (2001). Reconstitution ofcentrosome nucleation in spisula. Methods Cell Biol. 67, 149-165.

Selvaggi G. and Scagliotti G.V. (2009). Histologic subtype in NSCLC:does it matter? Oncology (Williston Park) 23, 1133-1140.

Setoguchi A., Okuda M., Nishida E., Yazawa M., Ishizaka T., Hong S.H.,Hisasue M., Nishimura R., Sasaki N., Yoshikawa Y., Masuda K.,Ohno K. and Tsujimoto H. (2001). Results of hyperamplification ofcentrosomes in naturally developing tumors of dogs. Am. J. Vet.Res. 62, 1134-1141.

Sève P. and Dumontet C. (2005). Chemoresistance in non-small lungcancer. Curr. Med. Chem. Anticancer Agents 5, 73-88.

Sève P., Isaac S., Trédan O., Souquet P.J., Pachéco Y., Pérol M.,Lafanéchère L., Penet A., Peiller E.L. and Dumontet C. (2005a).Expression of class III ß-tubulin is predictive of patient outcome inpatients with non-small cell lung cancer receiving vinorelbine-basedchemotherapy. Clin. Cancer Res. 11, 5481-5486.

Sève P., Mackey J., Isaac S., Trédan O., Souquet P.J., Pérol M., Lai R.,Voloch A. and Dumontet C. (2005b). Class III ß-tubulin expression intumor cells predicts response and outcome in patients with non-small cell lung cancer receiving paclitaxel. Mol. Cancer Ther. 4,2001-2007.

Sève P., Reiman T., Lai R., Hanson J., Santos C., Johnson L., DabbaghL., Sawyer M., Dumontet C. and Mackey J.R. (2007). Class III ß-tubulin is a marker of paclitaxel resistance in carcinomas of unknownprimary site. Cancer Chemother. Pharmacol. 60, 27-34.

Sève P., Reiman T. and Dumontet C. (2010). The role of ßIII-tubulin inpredicting chemoresistance in non-small cell lung cancer. LungCancer 67, 136-143.

Sulimenko V., Sulimenko T., Poznanovic S., Nechiporuk-Zloy V., BöhmK., Macůrek L., Unger E. and Dráber P. (2002). Association of brainγ-tubulins with αß-tubulin dimers. Biochem. J. 365, 889-895.

Sulimenko V., Dráberová E., Sulimenko T., Macurek L., Richterová V.,Dráber P. and Dráber P. (2006). Regulation of microtubule formationin activated mast cells by complexes of γ-tubulin with Fyn and Sykkinases. J. Immunol. 176, 7243-7253.

Syed M.I., Syed S., Minty F., Harrower S., Singh J., Chin A., McLellanD.R., Parkinson E.K. and Clark L.J. (2009). γ-Tubulin: a promisingindicator of recurrence in squamous cell carcinoma of the larynx.Otolaryngol. Head Neck Surg. 140, 498-504.

Travis W.D., Brambilla E., Muller-Hermelink H.K. and Harris C.C.(2004). Pathology and genetics of tumours of the lung, pleura,thymus and heart. In: World Health Organization Classification ofTumours. IARC Press. Lyon. 9-124 p.

Vinopal S., Černohorská M., Sulimenko V., Sulimenko T., Vosecká V.,Flemr M., Dráberová E. and Dráber P. (2012) γ-Tubulin 2 nucleatesmicrotubules and is downregulated in mouse early embryogenesis.PLoS ONE 7, e29919.

Wise D.O., Krahe R. and Oakley B.R. (2000). The γ-tubulin gene familyin humans. Genomics 67, 164-170.

Yuba-Kubo A., Kubo A., Hata M. and Tsukita S. (2005). Gene knockoutanalysis of two γ-tubulin isoforms in mice. Dev. Biol. 282, 361-373.

Zheng Y., Wong M.L., Alberts B. and Mitchison T. (1995). Nucleation ofmicrotubule assembly by a γ-tubulin-containing ring complex. Nature378, 578–583.

Zhou J., Shu H.B. and Joshi H.C. (2002). Regulation of tubulin synthesisand cell cycle progression in mammalian cells by γ-tubulin mediatedmicrotubule nucleation. J. Cell Biochem. 84, 472-483.

Accepted March 28, 2012


overexpression of γ-tubulin in non-small cell lung cancer

Documents