TUBB3 bIII-Tubulin Acts through the PTEN/AKT Signaling ...Matrix-assisted laser desorption/ionization–time-of-ﬂight mass spectrometry Protein spots were excised from SYPRO Ruby–stained

Therapeutics, Targets, and Chemical Biology

TUBB3/bIII-Tubulin Acts through the PTEN/AKTSignaling Axis to Promote Tumorigenesis andAnoikis Resistance in Non–Small Cell Lung CancerJoshua A. McCarroll1,2, Pei Pei Gan1, Rafael B. Erlich1, Marjorie Liu1, Tanya Dwarte1,Sharon S. Sagnella1,2, Mia C. Akerfeldt1, Lu Yang1, Amelia L. Parker1, Melissa H. Chang1,Michael S. Shum1, Frances L. Byrne1, and Maria Kavallaris1,2

Abstract

bIII-tubulin (encoded by TUBB3) expression is associated withtherapeutic resistance and aggressive disease in non–small celllung cancer (NSCLC), but the basis for its pathogenic influence isnot understood. Functional and differential proteomics revealedthat bIII-tubulin regulates expression of proteins associated withmalignant growth and metastases. In particular, the adhesion-associated tumor suppressor maspin was differentially regulatedby bIII-tubulin. Functionally, bIII-tubulin suppression altered cell

morphology, reduced tumor spheroid outgrowth, and increasedsensitivity to anoikis. Mechanistically, the PTEN/AKT signalingaxis was defined as a critical pathway regulated by bIII-tubulin inNSCLC cells. bIII-Tubulin blockage in vivo reduced tumor inci-dence and growth.Overall, ourfindings revealed howbIII-tubulininfluences tumor growth in NSCLC, defining new biologic func-tions and mechanism of action of bIII-tubulin in tumorigenesis.Cancer Res; 75(2); 415–25. �2014 AACR.

IntroductionLung cancer is a lethal adult cancer accounting for the most

cancer-related deaths worldwide (1). The most common form,non–small cell lung carcinoma (NSCLC) represents >80% of allcases (2). Over half of the patients with NSCLC have developedmetastasis at the time of diagnosis and 5-year survival is approx-imately 14% (2). Advanced NSCLC is poorly responsive totherapy and the mechanisms responsible for its resistance andaggressive behavior are not well defined.

Alterations in expression of microtubule proteins in cancercells are emerging as important contributors to chemotherapyresistance (3, 4). Microtubules are cytoskeletal proteins thatcomprise a- and b-tubulin heterodimers and are involved inmany important cellular processes including maintenance ofshape, intracellular transport, and mitosis (reviewed in ref. 4).These critical roles make microtubules attractive drug targetsfor anticancer therapies. Indeed, chemotherapy agents includ-ing the microtubule stabilizing (e.g., taxanes) and destabiliz-

ing (e.g., vinca alkaloids) drugs exert their toxic effect bybinding the b-tubulin subunit of microtubules to induce apotent mitotic block (reviewed in ref. 4). b-Tubulin has sevenisotypes that display tissue-specific expression. For example,bI-tubulin is constitutively expressed in many tissues, whereasbII-tubulin and bIII-tubulin are expressed in neuronal tissues,and bIV-tubulin is restricted to hematopoietic tissues (3, 4).Clinical studies show that high bIII-tubulin expression corre-lates with chemoresistance and poor survival in differenttumor types, including breast, ovarian, gastric, and NSCLC(5, 6). Previously, we demonstrated a functional role for bIII-tubulin in regulating chemosensitivity in NSCLC using RNAinterference to silence bIII-tubulin expression in NSCLC cells(7–10). Knockdown of bIII-tubulin increased sensitivity viaan increase in apoptosis to chemotherapeutic agents both invitro and in vivo (7–10). However, despite evidence to linkbIII-tubulin expression to chemotherapy drug sensitivity, itsroles in the tumorigenic and metastatic potential of NSCLCare unknown. Recent clinical data have suggested that highbIII-tubulin expression is correlated to aggressive and meta-static tumors (11, 12). In addition, we recently showed thatsilencing bIII-tubulin expression in the absence of chemo-therapy delayed tumor growth in a mouse model of lungcancer (10). However, how bIII-tubulin exerts its effect ontumor growth and its role in metastases of tumor cells remainunknown.

In this study, we show for the first time that stable suppres-sion of bIII-tubulin in NSCLC cells alters the expression of keyproteins involved in regulating tumorigenic and metastaticpotential, alters cell morphology, increases anoikis sensitivity,and modulates PTEN/AKT signaling. Finally, we demonstratethat knockdown of bIII-tubulin in NSCLC cells decreases theincidence and growth of lung tumors in two different preclin-ical mouse models.

1Children's Cancer Institute, LowyCancer Research Centre, Randwick,University of New South Wales Australia, Sydney, New South Wales,Australia. 2ARCCentre of Excellence in Convergent Bio-Nano Scienceand Technology, Australian Centre for NanoMedicine, University ofNew South Wales Australia, New South Wales, Australia.

Note: Supplementary data for this article are available at Cancer ResearchOnline (http://cancerres.aacrjournals.org/).

J.A. McCarroll, P.P. Gan, and R.B. Erlich contributed equally to this article.

CorrespondingAuthor:Maria Kavallaris, TumorBiology andTargetingProgram,Children's Cancer Institute, Lowy Cancer Research Centre, University of NewSouthWales Australia, P.O. Box 81, Randwick, New SouthWales 2031, Australia.Phone: 612-9385-2151; Fax: 612-9662-6583; E-mail: [email protected]

doi: 10.1158/0008-5472.CAN-14-2740

�2014 American Association for Cancer Research.

CancerResearch

www.aacrjournals.org 415

on July 12, 2021. © 2015 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst November 20, 2014; DOI: 10.1158/0008-5472.CAN-14-2740

http://cancerres.aacrjournals.org/

Materials and MethodsCell culture

H460 and A549 NSCLC cells were obtained from ATTCand grown as described (8–10). Upon receipt from ATCC, cellmaster stocks were prepared and cells for experiments werepassaged for less than 6 months. H460 master stock cells werevalidated by PCR (STR profiling) in December 2012 byCell Bank Australia (Sydney, Australia). Stable bIII-tubulinshRNA expressing clones were validated by PCR (STR profiling)in October 2014 by the Molecular Genetics Facility at theGarvan Institute (Sydney, Australia). H460 (pRS/bIIISH4 andpRS/bIIISH59) and A549 (pRS/bIIISH61) cell clones stablyexpressing shRNA targeting bIII-tubulin or control (nonfunc-tional) shRNA (pRS/CtrlSH1, pRS/CtrlSH2, pRS/CtrlSH27), andtwo H460 bIII-tubulin "rescue" cell clones (pRS/bIIISH4/R6 andpRS/bIIISH4/R17) were developed and characterized asdescribed (10). All cells were routinely screened and found tobe free of mycoplasma.

Two-dimensional differential gel electrophoresisTwo-dimensional differential gel electrophoresis (2D-DIGE)

is a fluorescent-based proteomics approach that uses chargeand size-matched CyDye fluorophores that covalently attach tolysine residues for protein labeling before fractionation. A keystrength of this assay is the ability to directly compare proteinsamples using a multiplex approach, while incorporating astandardized internal control (13, 14). To identify differen-tially expressed proteins in NSCLC cells with stable suppres-sion of bIII-tubulin, we performed 2D-DIGE on H460 bIII-tubulin shRNA (pRS/bIIISH4 and pRS/bIIISH59) or controlshRNA (pRS/CtrlSH1 and pRS/CtrlSH2) cells (SupplementaryFig. S1). 2D-DIGE was performed as we previously describedwith modifications (15). Refer to Supplementary Materials andMethods for detailed methodology. For analysis of the effectsof bIII-tubulin knockdown, H460 pRS/bIIISH4 and pRS/bIIISH59 cells were grouped and compared with controls,H460 pRS/CtrlSH1 and pRS/CtrlSH2 cells. These comparisonswere performed for both nuclear and cytoplasmic fractions attwo pH ranges, pH 4.5–5.5 (narrow) or pH 4–7 (broad).Analysis of the fluorescent gel images was performed usingthe Batch Processor and Biological Variation Analysis (BVA)Module of the DeCyder software version 5.0.The BVA modulewas used to compare each group and data are expressed asaverage expression ratios and Student t tests of individualprotein spots.

Matrix-assisted laser desorption/ionization–time-of-flightmass spectrometry

Protein spots were excised from SYPRO Ruby–stained gelsand matrix-assisted laser desorption/ionization–time-of-flight(MALDI-TOF) mass spectrometry performed as described pre-viously (15). Peptide masses were searched against SwissProt/TREMBL protein databases using the PeptIdent tool onExPASY (http://tw.expasy.org/tools/peptident.html) for pro-tein identification.

Western blottingWestern blot analysis was performed using the following anti-

bodies: bIII-tubulin (clone TUJ1, Chemicon), maspin (BD Bio-sciences), profilin 2 (AbCam), perioxiredoxin 4 (AbCam), total

b-tubulin (AbCam), lamin B1 (AbCam), p-AKT (S473), pAKT(T308), total AKT, PTEN (Cell Signaling Technology), andGAPDH (Abcam) as described (10, 16).

Immunofluorescence microscopyThe actin cytoskeleton of NSCLC cells was visualized by phal-

loidin staining and imaged using a Zeiss immunofluorescencemicroscope (Zeiss).

Real-time quantitative PCR analysisThe expression of bIII-tubulin and maspin mRNA in NSCLC

cells was examined by quantitative real-time PCR (qPCR) asdescribed (10, 16). All data were normalized to the housekeepinggene b2-microglobulin.

Spheroid outgrowth assayTumor multicellular spheroids were generated in 96-well

round bottom plates as described (17). After formation, spher-oids of uniform size were transferred to 6-well tissue cultureplates and left to attach and grow for 24 hours. Images werecaptured on a Zeiss Axiovert 200M inverted microscopeand cell outgrowth quantified using the AxioVision softwarepackage.

Anoikis assayTo measure anchorage-independent cell death, bIII-tubulin

shRNA expressing (pRS/bIIISH4 and pRS/bIIISH59), control(pRS/CtrlSH1 and pRS/CtrlSH2), or bIII-tubulin rescue (pRS/bIIISH4/R17) cells were seeded onto poly-(2-hydroxyethylmethacrylate)-coated tissue culture plates. Forty-eight hoursafter seeding cells, cell death was measured as described(10, 18).

In vivo mouse modelsExperimental metastasis model. All animal experiments wereapproved by the Animal Ethics committee, UNSW (ACEC no.12/41A). Balb/c nude mice received a tail vein injection of 5 �105 cells stably expressing bIII-tubulin (pRS/bIIISH4 and pRS/bIIISH59) or control (pRS/CtrlSH2 and pRS/CtrlSH1) shRNA. Thedetection of tumors within the lung was monitored by micro-CT imaging.

Tumorigenicity model. BALB/c nude mice received subcutaneousinoculations into the flank with 1 � 106 NSCLC cells expressingbIII-tubulin (pRS/bIIISH4 and pRS/bIIISH59) or control (pRS/CtrlSH1 and pRS/CtrlSH2) shRNA was resuspended in PBS tomimic anchorage independence or in Matrigel to mimicanchorage dependence as described (10, 19). Tumors weremeasured as described (10, 19).

Micro-CT tumor analysisThe presence of lung tumors in the mice was monitored by

micro-CT scanning. The micro-CT data were acquired using anInveon system (Siemens) at 71.88 pixel size, 220 projections,250 ms integration time, 50 keV photon energy, and 450 mAcurrent. Image data were evaluated using the Inveon softwarepackage (Siemens). Tumor margins were identified by contrastthresholding, which allowed the tumor margins (soft tissuedensity) to be defined against the surrounding lung (airdensity).

McCarroll et al.

Cancer Res; 75(2) January 15, 2015 Cancer Research416




Statistical analysisData are expressed as the mean � SE and analyzed using

ANOVA or Student t test followed by the nonparametric Dunnetttest using the GraphPad Prism program. A P value of <0.05 wasconsidered statistically significant.

ResultsSuppression of bIII-tubulin alters the expression of proteinsinvolved in regulating tumorigenesis and metastases

Previously, we established NSCLC cells that have stable andpotent suppression of bIII-tubulin (10). These cells displayedincreased chemosensitivity and decreased tumor incidence (10).However, despite clinical reports correlating increased levels ofbIII-tubulin in NSCLC with aggressive disease, studies describingits role in tumorigenicity and metastases are limited. To investi-gate the role of bIII-tubulin, 2D-DIGE proteomics was performedonboth the cytoplasmic andnuclear fractions of two independentTUBB3/bIII-tubulin shRNA–expressing H460 NSCLC cell clones(pRS/bIIISH4 and pRS/bIIISH59) and two independent controlshRNA H460 NSCLC cell clones (pRS/CtrlSH1 and pRS/CtrlSH2

(Supplementary Figs. S1 and S2). Differences in protein expres-sion between the bIII-tubulin shRNA–expressing and control cellswere quantitatively determined using theDeCyder image analysissoftware, which provided statistical information on each individ-ual protein identifiedon the2D-DIGEgel. Protein spots of interestwere excised and identified by mass spectrometry. Eleven out of atotal of 963 proteins [1.1%; isoelectric point (pI) 4-7, broadrange] and 56 (21 upregulated and 35 downregulated) out of atotal of 753 proteins (7%; pI 4.5–5.5, narrow range) were signif-icantly altered in the cytoplasmic fractions of H460 pRS/bIIISH4

and SH59 cells compared with controls (pRS/CtrlSH1 and SH2).Moreover, 33 out of a total of 1,331 proteins (2.5%; pI 4–7,broad range) and 42 (9 upregulated and 33 downregulated) outof a total of 502 proteins (8.4%; pI 4.5–5.5, narrow range) weresignificantly altered in the nuclear fractions of pRS/bIIISH4 and SH59

cells compared with controls. A number of proteins identified asbeing differentially expressed in the bIII-tubulin knockdown cells,are involved in modulating tumor biology (SupplementaryTables S1 and S2). To confirm changes in protein expressionand mass spectrometry analysis, a select number of proteinswere validated by Western blotting as being differentiallyexpressed in both individual bIII-tubulin shRNA cell clones(pRS/bIIISH4 and SH59) compared with controls (pRS/CtrlSH1 and

SH2; Supplementary Fig. S4).

bIII-Tubulin levels influence the expression of the tumorsuppressor maspin in NSCLC cells

Maspin also known as Serpin B5 is a member of the serineprotease inhibitor/noninhibitor superfamily (reviewed in ref. 20)and is classified as a tumor suppressor in different cancer types(20). Maspin was identified by 2D-DIGE as being upregulated(2.29-fold, P < 0.05) in the bIII-tubulin shRNA–expressing cells(Supplementary Table S1). Western blot analysis confirmed theincrease in maspin protein levels in pRS/bIIISH4 and SH59 cellswhen compared with controls (Fig. 1A and B). Recently, studieshave described the localization of maspin in the cytoplasm or thenucleus as being an important determinant in its tumor suppres-sor activity (21, 22). To examine whether there was any differencein the cellular distribution of maspin in bIII-tubulin shRNA–expressing cells, cytoplasmic and nuclear extracts were collected

and maspin expression measured by Western blot analysis. Con-trol shRNA–expressing cells hadhighermaspin expression in theircytoplasm when compared with the nucleus (Fig. 1C). bIII-Tubu-lin shRNA–expressing cells also had increased maspin expressionin their cytoplasmic fraction compared with the nuclear fraction(Fig. 1C). However, the expression of maspin was increasedin both the cytoplasmic and nuclear fractions of the bIII-tubulinshRNA cells compared with control cells (Fig. 1C). To determinewhether this difference was reflected at the gene level, maspingene expression was measured by qPCR. Maspin mRNA levelswere increased in both bIII-tubulin shRNA expressing cell clones(pRS/bIIISH4 and SH59) compared with controls (pRS/CtrlSH1 and

SH2; Fig. 1D). The increase in maspin expression was also con-firmed in anotherH460NSCLCbIII-tubulin shRNA–independentcell clone (pRS/bIIISH60), which also displayed potent knock-down of bIII-tubulin (Supplementary Fig. S5). Notably, this cellclone was not used in the 2D-DIGE experiments, thus addingfurther support to the correlation between bIII-tubulin knock-down and increased maspin expression.

To determinewhether bIII-tubulin expression can directly influ-ence maspin expression, bIII-tubulin was restored back into thebIII-tubulin shRNA–expressing cells as described (10) andmaspinexpression measured. Restoration of bIII-tubulin expression intwo individual bIII-tubulin rescue clones (pRS/bIIISH4/R6 andpRS/bIIISH4/R17) completely abolished the increased expressionofmaspin (Fig. 2AandB). In contrast, the emptypREP vector clone(pRS/EV) that maintained suppressed bIII-tubulin expression (Fig.2A) did not alter maspin expression (Fig. 2A and B). Collectively,these results show for the first time that bIII-tubulin levels directlyinfluence the expression of a tumor suppressor protein.

bIII-Tubulin suppression induces cell morphology changes inNSCLC cells

Malignant transformation of tumor cells is associated withchanges in organization of the cell cytoskeleton, decreased celladhesion, increased cell migration and resistance to anchorage-independent apoptosis (23). Given that suppression of bIII-tubu-lin led to alterations in the expression of proteins involved inregulating tumor growth and metastases, we assessed whethersuppression of bIII-tubulin led to changes in cell morphology ofNSCLC cells. bIII-Tubulin knockdown produced marked changesin cellular morphology when compared with controls. The con-trol cells had a rounded morphology and poorly defined actinstress fibers (Fig. 3A and B). In contrast, the bIII-tubulin shRNAcells were flatter and displayed prominent actin stress fibers asevidenced by phalloidin staining (Fig. 3C and D). To establishwhether the reduced expression of bIII-tubulin was responsiblefor the change in cell shape, we also examined the cellularmorphology of two individual H460 NSCLC bIII-tubulin rescueclones. Strikingly, the actin stress fibers were markedly reduced inthese cells and their shape appeared to be very similar to thecontrols (Fig. 3E and F).

bIII-Tubulin suppression inhibits tumor spheroid outgrowth,cell migration, and sensitizes NSCLC cells to anoikis

To determine the functional significance of the altered cellmorphology in bIII-tubulin knockdown NSCLC cells, we firstmeasured their ability to grow out from tumor spheroids in 3Dculture. In this system, spheroids are formed by the self-assemblyof clusters of cell colonies under anchorage-independent condi-tions where cell–cell interactions dominate (24). Notably, after

TUBB3/bIII-Tubulin Regulates NSCLC Tumorigenicity

www.aacrjournals.org Cancer Res; 75(2) January 15, 2015 417




transfer of the spheroids to a standard tissue culture plate, H460bIII-tubulin shRNA–expressing cells failed to attach and grow outfrom the spheroids (Fig. 4A, panel II). In contrast, control shRNAcells readily attached and grew out from the spheroids (Fig. 4A,panel I). The importance of high bIII-tubulin levels in regulatingcell attachment and spheroid outgrowth was further highlightedin another independent NSCLC cell line (A549) that stablyexpresses TUBB3/bIII-tubulin shRNA. These cells displayedpotent and stable knockdown of bIII-tubulin (SupplementaryFig. S7). In agreement with H460 bIII-tubulin shRNA cells, A549bIII-tubulin knockdown cells (pRS/bIIISH61) also showed a sig-nificant delay in attachment and spheroid outgrowth when com-paredwith controls (Fig. 4B, panels I and II). The ability of bIII-tubulin knockdown cells to adhere to an extracellular matrix(ECM) protein fibronectin (a major ECM protein expressed inthe lung and tumor stroma; ref. 25) was also measured.Detachment of tumor cells from the ECM is a critical step inthe formation of metastases. We showed that suppression ofbIII-tubulin in NSCLC cells induced a modest but significant

increase in adhesion to fibronectin (Supplementary Fig. S6A).Next, we examined whether suppression of bIII-tubulin wouldinfluence cell migration. Briefly, A549 NSCLC cells stablyexpressing bIII-tubulin shRNA (pRS/bIIISH61) or control shRNA(A549 pRS/CtrlSH27) were seeded into the wells of modifiedBoyden chambers, and cell migration was assessed 24 hourslater (Supplementary Methods). Stable suppression of bIII-tubulin expression markedly reduced the migratory capacityof NSCLC cells (pRS/bIIISH61 migration index of 3.95 � 0.70)when compared with controls (A549 pRS/CtrlSH27 migrationindex of 30.93 � 4.81; Supplementary Fig. S6B). Collectively,these results provide strong evidence that bIII-tubulin plays animportant role in regulating cell–cell and cell–matrix interac-tions as well as the migration capacity of NSCLC cells

Typically, detachment of normal epithelial cells from ECMresults in programmed cell death "anoikis". Tumor cells withhigh metastatic potential are resistant to anoikis, which allowsthem to survive under nonadherent conditions and travel in theblood or lymph to form metastases. To determine whether

Figure 1.Suppression of bIII-tubulin increases the expression of maspin in NSCLC cells. A, representative 2D-DIGE densitometric volume of maspin is shown in a three-dimensional (3D) view generated by the DeCyder Image Analysis software package. Arrows, 3D densitometric peak for maspin expression in both control(H460 pRS/CtrlSH1 and SH2) and bIII-tubulin (H460 pRS/bIIISH4 and SH59) shRNA–expressing NSCLC cells. B, Western blot and densitometry graph demonstratingthat potent suppression of bIII-tubulin expression in two individual bIII-tubulin shRNA NSCLC cell clones (H460 pRS/bIIISH4 and SH59) is associated with asignificant increase inmaspin protein expressionwhen comparedwith two individual control clones (H460pRS/Ctrl SH1 and SH2); n¼ 3–5 independent experiments. Allsampleswere normalized to the housekeeping protein GAPDH (� , P <0.001; H460 pRS/bIIISH4 and SH59 vs. H460pRS/CtrlSH1 and SH2). C,Western blot analysis showingan increase inmaspinprotein expression in both the cytoplasmic andnuclear fractions ofNSCLC cells stably expressingbIII-tubulin shRNA (H460pRS/bIIISH4 and SH59)when compared with two individual control clones (H460 pRS/CtrlSH1 and SH2); n¼ 3 independent experiments. D, a graph showing a significant increase in maspinmRNA expression in two individual cell clones of NSCLC cells stably expressing bIII-tubulin shRNA (H460 pRS/bIIISH4 and SH59) when compared with controlclones (pRS/CtrlSH1 and SH2); n¼ 4 separate experiments. All sampleswere normalized to the housekeeping gene b2-microglobulin (� , P <0.001, H460 pRS/bIIISH4 and

SH59 vs. pRS/CtrlSH1 and SH2).

McCarroll et al.





suppression of bIII-tubulin modulated anoikis sensitivity, bIII-tubulin shRNA cells were plated in poly-HEMA–coated cultureplates to prevent cell adhesion. Cell death was then measured byAnnexin V staining. Suppression of bIII-tubulin did not inducecell death in adherent conditions (Fig. 4C). In contrast, when keptin nonadherent conditions, bIII-tubulin knockdown significantlyincreased anoikis sensitivity in both the pRS/bIIISH4 and pRS/bIIISH59 cell clones when compared with controls (Fig. 4C).To confirm a direct correlation between bIII-tubulin expressionand increased anoikis sensitivity, the same experiment was per-formed using the bIII-tubulin rescue (pRS/bIIISH4/R17) cells.Reexpression of bIII-tubulin abolished the increased sensitivityto anoikis (Fig. 4D).

bIII-Tubulin modulates AKT activity and PTEN expression inNSCLC cells

To investigate themolecularmechanisms thatmediate increasedsensitivity to anoikis inbIII-tubulinknockdowncells,weexaminedthe activationof thePI3Kdownstream substrate AKT kinase. AKT iscommonly dysregulated in tumor cells and is important in pro-moting cell survival and resistance to anoikis (26). To examinewhether there was any difference in AKT activity in bIII-tubulinshRNA–expressing cells, we measured AKT phosphorylation (p-AKT) at residues S473 and T308 (phosphorylation at both sites isrequired for full AKT activation) between control (pRS/CtrlSH2 and

SH27) and bIII-tubulin knockdown (pRS/bIIISH4 and SH61) H460and A549 NSCLC cells. Both bIII-tubulin shRNA–expressingNSCLC cell lines displayed a clear reduction in p-AKT expressionat bothS473andT308 residueswhen comparedwith controls (Fig.

5A). Remarkably, restoration of bIII-tubulin back into the H460bIII-tubulin shRNA cells completely restored p-AKT levels at bothphosphorylation sites (Fig. 5B). Next, the expression of theupstream tumor suppressor PTEN, which is a key regulator of AKTactivation, was measured. Loss of PTEN is observed in many typesof cancer including NSCLC, which often leads to increased p-AKT(27). Control cells had little PTEN expression (Fig. 5C). Strikingly,NSCLC cells with stable bIII-tubulin knockdown displayedmarkedly increased PTEN expression (Fig. 5C). Moreover, resto-ration of bIII-tubulin expression abolished the increase in PTENexpression, thus confirming that the low levels of bIII-tubulin wereresponsible for the differential expression of PTEN (Fig. 5C).Finally, to determine whether the decrease in p-AKT activity inthe bIII-tubulin knockdown cells, was in part mediated viaincreased PTEN expression, we transfected bIII-tubulin shRNAcells with PTEN siRNA (Fig. 6D). Cells transfected with nonfunc-tional siRNA served as controls. Notably, the decrease in p-AKTexpression levels was restored in the bIII-tubulin knockdown cellsto the level of control cells (Fig. 5D). These results suggest that highlevels of bIII-tubulin in NSCLC cells influence p-AKT activity viaPTEN.

To examine whether AKT activity is altered when NSCLC cellsare exposed to nonadherent conditions, p-AKT levels were com-pared between the bIII-tubulin shRNA (pRS/bIIISH4 and SH61) andcontrol shRNA (pRS/CtrlSH2 and SH27) H460 and A549 cells atdiffering time points when in suspension. In H460 cells, aprogressive time-dependent decrease in p-AKT expression wasobserved in both control and bIII-tubulin shRNA cells (Fig. 6A).However, p-AKT levels in bIII-tubulin shRNA (pRS/bIIISH4) cells

Figure 2.Rescue of bIII-tubulin reverses the increase in maspin expression in bIII-tubulin shRNA NSCLC–expressing cells. A, representative Western blot and densitometrygraph showing that in two individual bIII-tubulin rescue cell clones (H460 pRS/bIII R6 and pRS/bIII R17), maspin protein expression is restored to control(pRS/CtrlSH1 and SH2) levels when compared with bIII-tubulin shRNA (H460 pRS/bIIISH4) or empty vector control (H460 pRS/EV); n¼ 6 independent experiments. Allsamples were normalized to the housekeeping protein GAPDH (� , P < 0.001, H460 pRS/bIIISH4 and pRS/bIIIEV vs. control clones H460 pRS/CtrlSH1 and SH2; #,P < 0.001, H460 pRS/bIII R6 and R17 vs. pRS/bIIISH4 and pRS/bIIIEV). B, a graph showing that maspin mRNA expression levels are restored to control (H460pRS/CtrlSH1 and SH2) and empty vector control (H460 pRS/EV) in two individual bIII-tubulin rescue cell clones (H460 pRS/bIII R6 and pRS/bIII R17); n¼ 4 independentexperiments. All samples were normalized to the housekeeping gene b2-microglobulin (� , P < 0.001, H460 pRS/bIIISH4 and pRS/bIIIEV vs. control clonespRS/CtrlSH1 and SH2; #, P < 0.001, pRS/bIII R6 and R17 vs. pRS/bIIISH4 and pRS/bIII EV).






became undetectable as early as 10 minutes after exposure tononadherent conditions, p-AKT activation in control (pRS/CtrlSH2) cells was still evident after 20 minutes in suspension(Fig. 6A). In A549 cells, a marked time-dependent increase in p-AKT activation was observed in control cells (pRS/CtrlSH27),whereas little to no p-AKT expression was observed in the bIII-tubulin shRNA cells (pRS/bIIISH61; Fig. 6A). To determine wheth-er the increased levels of PTEN in thebIII-tubulin knockdown cellswas responsible for the suppression of AKT activation when thecells were in suspension, PTEN was silenced in bIII-tubulinknockdown cells (pRS/bIIISH4) using siRNA. Notably, knock-down of PTEN fully restored AKT activation at all of the timepoints assessed (5, 10, and 20 minutes) in the bIII-tubulinknockdown cells (pRS/bIIISH4) cells when compared with cellstreatedwith control siRNA (Fig. 6B). Together, these results clearlyshow that the activity of a prosurvival signaling protein is sup-pressed in NSCLC cells with low bIII-tubulin expression whenexposed to nonadherent conditions, and that PTEN appears tomediate the suppression of AKT activation.

bIII-Tubulin suppression increases anoikis sensitivity in vivoTo expand our findings into a clinically relevant context, and to

investigatewhether increased sensitivity to anoikis and reduced p-

AKT activity induced by bIII-tubulin knockdown would affect thetumorigenic and/or metastatic potential of NSCLC cells, weinjected control shRNA or bIII-tubulin shRNA cells into the tailvein of mice and monitored the formation of metastatic lungtumors by micro-CT imaging. After 50 days, 3 of 5 (60%) miceinjected with control shRNA cells (pRS/CtrlSH2) developed lungtumors, whereas only 1 of 5 (20%)mice injectedwithbIII-tubulinshRNA cells (pRS/bIIISH4) developed lung tumors (Fig. 6C). Thepresence of tumors within the lungs of mice was confirmed bymicro-CT and histology (Fig. 6D and E). This finding suggests thatNSCLC cells with low bIII-tubulin expression have reducedtumorigenic and metastatic potential when exposed to the circu-latory system.

Finally, to reenforce that bIII-tubulin knockdown cells have areduced capacity to form tumors in the absence of adhesion toECM, we assessed tumor growth in mice following subcutaneousimplantation of control shRNA or bIII-tubulin shRNA cells sus-pended in PBS tomimic an ECM-free environment, or inMatrigelto mimic attachment to ECM. After 21 days, tumors derived fromcontrol shRNA (pRS/CtrlSH2) cells that were administered in PBSweremarkedly bigger as comparedwith tumors derived from bIII-tubulin shRNA (pRS/bIIISH4) cells in PBS (Fig. 7A). In contrast,when bIII-tubulin shRNA cells were suspended in ECM weobserved no significant difference in the size of the tumorsgenerated by control (pRS/CtrlSH2) or bIII-tubulin shRNA cells(pRS/bIIISH4; Fig. 7B). Similar results were observed when com-paring another individual bIII-tubulin shRNA cell clone (pRS/bIIISH59) and its control (pRS/CtrlSH1; data not shown). Collec-tively, the data provide strong evidence in two independentmouse models of NSCLC that bIII-tubulin levels influence tumorgrowth via dependence of bIII-tubulin–depleted cells on celladhesion.

DiscussionOverexpression of TUBB3/bIII-tubulin in tumor cells is often

associated with resistance to chemotherapeutic drugs. Recently,attention has turned to its clinical correlation with an aggressivetumor phenotype. However, despite reports highlighting bIII-tubulin as a potential biomarker for tumor aggressiveness, studiesdescribing its functional role in tumorigenicity andmetastases arelimited. Herein, we report for the first time novel roles for bIII-tubulin in (i) altering the expression of proteins involved inpromoting tumor growth and metastases; (ii) anoikis sensitivity;(iii) modulating PTEN/AKT signaling; and (iv) promoting tumorincidence and growth in vivo.

Cancer cells with high tumorigenic and metastatic potentialdifferentially express a host of proteins including tumor suppres-sors, oncogenes, and regulators of the cell cytoskeleton, whichenable them to escape apoptosis and achieve rapid growth andmotility. To gain an understanding as to whether NSCLC cellswith high bIII-tubulin levels have an altered proteomic profile, weperformed 2D-DIGE proteomic analysis on isolated cytoplasmicand nuclear fractions from NSCLC cells with stable and potentbIII-tubulin knockdown.Our data demonstrated thatbIII-tubulincan alter the expression of key proteins involved in modulatingtumor growth andmetastases. We demonstrated for the first timethat the levels of bIII-tubulin in NSCLC cells are critical inmodulating the expression of tumor suppressor proteins. Forinstance, maspin amember of the serpin family of serine proteaseinhibitors was originally identified as a tumor suppressor that is

Figure 3.Suppression of bIII-tubulin alters the cell morphology of NSCLC cells.Representative microscopy images of phalloidin (F-actin) staining on bIII-tubulin shRNA–expressing NSCLC (H460 pRS/bIIISH4 and SH59) cells (C and D)demonstrating a marked alteration in cell morphology as evidenced by aflattened appearance with prominent actin stress fibers (arrows)when compared with control shRNA expressing NSCLC cells (H460 pRS/CtrlSH1 and SH2A and B). Rescue of bIII-tubulin expression (E and F; H460 pRS/bIII R6 and R17) reduced the presence of actin stress fibers. Scale bar, 20 mm;n ¼ 3 independent experiments.

McCarroll et al.





expressed in normal mammary epithelial cells, but is reduced orabsent in breast tumor cells (22). Recent studies have explored thepotential of maspin as a prognostic marker in different tumortypes. Several reports have shown high maspin expression to be afavorable predictor for different tumors, including NSCLC (20).However, for other tumor types such as pancreatic, gallbladder,colorectal, and thyroid, high maspin expression has been associ-ated with a poor prognosis (20). These differences are thought tobe attributed to its differential expression within the cell (i.e.,cytoplasmic vs. nuclear; ref. 20). Nevertheless, in NSCLC there isan increasing body of evidence to suggest that tumors with highmaspin expression have a favorable prognosis (28–30). Thesestudies have been supported using in vitro and in vivo models oflung cancer. Beltran and colleagues (31) demonstrated that res-toration of maspin expression in NSCLC cells using artificial

transcription factors combined with chromatin modifier com-pounds, reduced NSCLC metastatic behavior (31). Maspin wasalso shown to be involved in regulating the survival of lung cancercells to chemotherapy drugs (32). Our studies showed thatsuppression of bIII-tubulin led to an increase in maspin expres-sion, and that rescue of bIII-tubulin expression back into thesecells was sufficient to bring maspin gene and protein expressionback to control levels. This result confirms that the levels of bIII-tubulin are directly responsible for the differential expression ofmaspin in NSCLC cells. The regulation ofmaspin in tumor cells isthought to involve a number of processes including control bytranscription factors such as p53, ATF-2, PTEN, and Snail (33–35).Interestingly, reports have demonstrated that the carboxy-termi-nal tail of bIII-tubulin is subject to posttranslational modifica-tions, which can allow it to form protein–protein complexes with

Figure 4.Suppression of bIII-tubulin decreases tumor spheroid outgrowth and induces anoikis in NSCLC cells. A and B, representative photomicrograph images and graphsshowing a significant delay in the attachment and outgrowth of H460 and A549 NSCLC cells stably expressing bIII-tubulin shRNA (H460 pRS/bIIISH4 and A549 pRS/bIIISH61) compared with controls (H460 pRS/CtrlSH2 and A549 pRS/CtrlSH27) 24 hours after attachment; n ¼ 3 experiments; �, P < 0.05. C, graph demonstrating asignificant increase in anoikis (cell detachment–induced cell death) in NSCLC cells stably expressing bIII-tubulin shRNA (H460 pRS/bIIISH4 and SH59) whencompared with controls (H460 pRS/CtrlSH1 and SH2) 48 hours after seeding; n¼ 3 independent experiments (� , P < 0.001, adherent vs. suspension; #, P < 0.001 pRS/bIIISH4 and SH59 vs. pRS/CtrlSH1 and SH2 in suspension). D, a graph showing that reexpression of bIII-tubulin (H460 pRS/bIII R17) abolished the increased sensitivityto anoikis in bIII-tubulin shRNA H460 NSCLC cells (H460 pRS/bIIISH4); n ¼ 3 independent experiments (�, P < 0.05 H460 pRS/bIIISH4 vs. pRS/CtrlSH2;#P < 0.05 pRS/bIIISH4 vs. pRS/bIII R17).






signaling proteins such as small GTPases and/or PKC in cancercells (36, 37). Furthermore, binding to these proteins initiatedsignaling cascades, which promoted cellular survival understress conditions. Therefore, it is possible that high levels ofbIII-tubulin in NSCLC cells enhance protein–protein interac-tions, which in turn activate signaling cascades that control theexpression of maspin. Further studies aimed at understandingthe molecular link between bIII-tubulin and maspin in NSCLCcells are required.

Our findings showing that the levels of bIII-tubulin inNSCLC cells affects the expression of proteins involved inmodulating tumorigenic and metastatic potential promptedus to explore the functional significance of bIII-tubulin in theseprocesses. We showed that suppression of bIII-tubulin had aprofound effect on the cell morphology of NSCLC cells asevidenced by their flattened appearance and prominent actinstress fibers. The altered morphology of the bIII-tubulin shRNAcells was not associated with any significant change in theexpression of total actin or tubulin (results not shown). Reor-ganization of the cell cytoskeleton due to loss or gain in theexpression of proteins associated with the cytoskeleton hasbeen reported in tumor cells. For instance, Bharadwaj andcolleagues (38) demonstrated that breast cancer cells with hightumorigenic and metastatic potential had reduced amounts ofthe actin-binding protein tropomyosin 1 (TM1; ref. 38). Rein-troduction of TM1 into these cells produced a striking alter-ation in the cytoskeleton with a flattened phenotype and

prominent actin stress fibers. This correlated with a reductionin their tumorigenic potential by increased sensitivity to anoi-kis (38). Notably, our proteomic analysis identified a signifi-cant increase in TM1 expression in the bIII-tubulin shRNA cells(Supplementary Table S1). In addition, studies have alsoreported that maspin can modulate changes in the expressionof proteins associated with the cytoskeleton that correlate toincreased cell adhesion and a reduced migratory phenotype(39). We cannot exclude the possibility that maspin may beacting in concert with other proteins and it is possible that bothTM1 and maspin may in part be involved in the reorganizationof the cell cytoskeleton in NSCLC cells with suppressed bIII-tubulin.

Cancer cells must detach from an ECM matrix, survive underanchorage-independent conditions to travel in the blood orlymphatic system, and adhere and proliferate at new organ sites(40). Therefore, tumor cells with metastatic potential haveacquired altered mechanisms of cellular adhesion as well asresistance to anoikis (40, 41). We demonstrated that suppressionof bIII-tubulin prevented/delayed the ability of NSCLC cells toadhere and grow out from multicellular tumor spheroids. Inaddition, we showed that bIII-tubulin knockdown increased celladhesion tofibronectin. These results support the notion thatbIII-tubulin plays an important role in regulating cell–cell and cell–matrix interactions. We also demonstrated that suppression ofbIII-tubulin sensitized NSCLC cells to anoikis. Importantly, anoi-kis was reversed by reexpression of bIII-tubulin, indicating a direct

Figure 5.Suppression of bIII-tubulin modulates AKT activity and PTEN expression in NSCLC cells. A, representative Western blots demonstrating that bIII-tubulin shRNA–expressing cells (H460pRS/bIIISH4 and A549pRS/bIIISH61) have significantly decreased phosphorylation ofAKT at S473 and T308 residues versus control (H460pRS/CtrlSH2 and A549 pRS/CtrlSH27) cells. Total AKT levels were unchanged. B, representativeWestern blot analysis showing that reexpression of bIII-tubulin (H460 pRS/bIII R17) back into bIII-tubulin shRNA (H460 pRS/bIIISH4) cells restores p-AKT activity at both S473 and T308 residues back to control (H460 pRS/CtrlSH2). C,representative Western blot analysis demonstrating increased PTEN expression in bIII-tubulin shRNA (H460 pRS/bIIISH4) cells when compared with control(H460 pRS/CtrlSH2) cells. Restoration of bIII-tubulin expression (H460 pRS/bIII R17) returns PTEN expression back to control (H460 pRS/CtrlSH2). Total AKT levelswere unchanged. D, representativeWestern blot analysis and densitometry graph showing restoration of p-AKT S473 expression in bIII-tubulin shRNA (H460 pRS/bIIISH4) cells 72 hours after transfection with PTEN siRNA. Cells transfected with nonfunctional siRNA served as controls. Total AKT levels were unchanged.� , P < 0.05. GAPDH was used a protein loading control for all Western blots. All data are representative of three independent experiments.

McCarroll et al.





correlation between bIII-tubulin levels and anoikis sensitivity. Togain an understanding into themechanisms that link bIII-tubulinknockdown and increased anoikis, we examined the activation ofthe prosurvival signaling protein AKT in bIII-tubulin knockdowncells. There are numerous reports that highlight the importance ofthe AKT signaling pathway in tumor cells as a mechanism thatpromotes resistance to various forms of apoptosis (40). Interest-ingly, we showed that suppression of bIII-tubulin correlated witha decrease in AKT phosphorylation at both phosphorylationsites (S473 and T308). Strikingly, AKT phosphorylation wasfully restored following reexpression of bIII-tubulin. One of thekey regulators of AKT activity is the tumor suppressor PTEN.The principal catalytic function of this phosphatase is to dephos-phorylate phosphatidylinositol-triphosphate, which is a potentactivator of AKT (42). Loss of PTEN function leads to AKT

signaling hyperactivation and is a common feature in a widerange of tumors (43). In agreement with our finding that inhi-bition of bIII-tubulin expression correlates with decreased AKTphosphorylation, we showed that expression of PTEN ismarkedlyincreased in bIII-tubulin shRNA–expressing cells. Furthermore,knockdown of PTEN using siRNA in the bIII-tubulin shRNA cellsrestored p-AKT activity back to controls. Finally, we showed thatwhen kept in nonadherent conditions, which is a surrogatemarker for tumorigenicity and is an environment in which cellsare susceptible to anoikis, AKT activation in bIII-tubulin shRNAcells became undetectable much earlier compared with controlcells. The question that arises is how increased levels of bIII-tubulin influence the activity of an important prosurvival signal-ing pathway in NSCLC cells. Although more studies will beneeded to fully answer this question, it should be noted, that

Figure 6.Suppression of bIII-tubulin reduces tumor incidence in vivo. A, representative Western blots from bIII-tubulin shRNA (H460 pRS/bIIISH4 and A549 pRS/bIIISH61)NSCLC cells showing a decrease in p-AKT (S473) expression levels at different time (5, 10, 20 minutes) points when exposed to nonadherent conditions. Total AKTlevels were unchanged. GAPDH was used as a protein loading control; n ¼ 3 independent experiments. B, representative Western blot analysis showing thatknockdown of PTEN in bIII-tubulin shRNA H460 (pRS/bIIISH4) cells using siRNA completely reverses the decrease in p-AKT (S473) expression over time (5, 10,20 minutes) when exposed to nonadherent condition. Cells (pRS/bIIISH4) treated with nonfunctional siRNA served as controls; n ¼ 3 independent experiments.C, graphic representation of tumor incidence for control (H460 pRS/CtrlSH2) and bIII-tubulin shRNA (H460 pRS/bIIISH4) cells; n ¼ 5 animals per group. D,representativemicro-CT images of lung tumors (arrowheads) on the axial, coronal, and sagittal planes 50 days after tail vein injection of control (H460 pRS/CtrlSH2)or bIII-tubulin shRNA (H460 pRS/bIIISH4) NSCLC cells. E, representative hematoxylin and eosin stain showing the presence of tumors within the lung of miceinjected systemically with control shRNA (H460 pRS/CtrlSH2) or bIII-tubulin shRNA (H460 pRS/bIIISH4) NSCLC cells. T, tumor; NL, normal lung; and H, heart.






silencing PTEN in the bIII-tubulin shRNA cells restored AKTphosphorylation when the cells were exposed to adherent ornonadherent conditions. A recent study by Nam and colleagues(32) showed that overexpression of maspin in NSCLC cellsresulted in a significant reduction in AKT phosphorylation thatcorrelated with increased expression of PTEN (32). Therefore, it istempting to speculate that the expression levels of bIII-tubulin,maspin, and PTEN are important determinants in modulatingAKT activity in NSCLC cells. This in turn may not only beinfluencing tumor cell survival under stressful conditions suchas nutrient and oxygen deprivation, but also under chemother-apeutic insult. Given the complex multifunctional nature of bIII-tubulin it is highly likely there are other additional biologic effectson NSCLC cells.

Finally, the importance of bIII-tubulin in promoting tumori-genicity and metastases in NSCLC cells was highlighted in twodifferent lung cancermousemodels. Thefirstmodelmeasured theability of bIII-tubulin shRNA cells to formmetastatic lung tumorsfollowing dissemination through the vascular system. In thismodel, tumor cells need to be able to survive in an anchorage-independent environment before extravasating and seeding inorgans. In this case, expression of bIII-tubulin correlated withincreasedmetastatic potential as suppression of this protein led todecreased tumor dissemination. The second model assessedtumor growth when the cells were administered in the presenceor absence of ECM.Here, it became clearly evident that differencesin the tumorigenic potential between bIII-tubulin shRNA andcontrol shRNAcells couldonly beobserved if cells are exposed to anonadherent environment.

Collectively, this work provides a valuable biologic insight intothe multifunctional role of bIII-tubulin in regulating tumorsuppressor and metastasis pathways as well as prosurvival sig-naling activity to directly influence NSCLC tumor growth andincidence.

Disclosure of Potential Conflicts of InterestM. Kavallaris received an industry research grant from Benitec Biopharma

to develop a lung cancer therapeutic. This grant did not fund this research.M. Kavallaris has ownership interest in a patent on methods for detecting andmodulating the sensitivity of tumor cells to antimitotic agents. No potentialconflicts of interest were disclosed by the other authors.

Authors' ContributionsConception and design: J.A. Mccarroll, P.P. Gan, R.B. Erlich, S. Sagnella,M. Akerfeldt, M.S. Shum, M. KavallarisDevelopment of methodology: J.A. Mccarroll, R.B. Erlich, T. Dwarte,M. Akerfeldt, M.S. Shum, M. KavallarisAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): J.A. Mccarroll, P.P. Gan, R.B. Erlich, M. Liu,T. Dwarte, S. Sagnella, M. Akerfeldt, A.L. Parker, M.H. Chang, M.S. Shum,F.L. ByrneAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): J.A. Mccarroll, P.P. Gan, R.B. Erlich, T. Dwarte,S. Sagnella, M. Akerfeldt, L. Yang, A.L. Parker, M.H. Chang, M.S. Shum,F.L. Byrne, M. KavallarisWriting, review, and/or revision of the manuscript: J.A. Mccarroll, R.B. Erlich,M.S. Shum, M. KavallarisAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases): L. Yang, M.H. ChangStudy supervision: M. KavallarisOther (involvement in the initial phase of the study): M.H. Chang

AcknowledgmentsThe authors thankDr.MatthewMcKay, Australian ProteomeAnalysis Facility

Ltd., for the mass spectrometry analysis.

Grant SupportThis work was supported by the Children's Cancer Institute Australia,

which is affiliated with the University of New South Wales (UNSW Australia)and Sydney Children's Hospital and by grants from the National Health andMedical Research Council (NHMRC APP1008719; M. Kavallaris and J.A.Mccarroll), Cancer Council New South Wales (M. Kavallaris), Cancer Ins-titute New South Wales Career Development Fellowship (J.A. Mccarroll),and NHMRC Senior Research Fellowships (#658611 and APP1058299;M. Kavallaris). M. Kavallaris is funded by the Australian Research CouncilCentre of Excellence in Convergent Bio-Nano Science and Technology(project number CE140100036). This work was also supported by a CancerInstitute of New South Wales infrastructure award, and facilitated by accessto the Australian Proteome Analysis Facility Ltd. established under theAustralian Commonwealth Government's Major National Research FacilitiesScheme.

The costs of publication of this articlewere defrayed inpart by the payment ofpage charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Received September 16, 2014; revised November 5, 2014; acceptedNovember 5, 2014; published OnlineFirst November 20, 2014.

Figure 7.Suppression of bIII-tubulin reduces tumor growth in an anchorage-independent environment. A, graph showing tumor volumes (mm3) following subcutaneousinjection of control shRNA (pRS/Ctrl SH2) or bIII-tubulin shRNA NSCLC cells (pRS/bIIISH4) when administered in an anchorage-independent environment (PBS).B, graph demonstrating tumor volumes (mm3) following subcutaneous injection of control shRNA (pRS/Ctrl SH2) or bIII-tubulin shRNA NSCLC cells (pRS/bIIISH4)when administered in an anchorage-dependent environment (Matrigel). n ¼ 3–5 mice for bIII-tubulin knockdown and control cells in PBS; n ¼ 6–7 animalsfor bIII-tubulin knockdown and control cells in Matrigel; values presented as mean � SEM.

McCarroll et al.





References1. Jemal A, Bray F, Center MM, Ferlay J, Ward E, Forman D. Global cancer

statistics. CA Cancer J Clin 2011;61:69–90.2. AisnerDL,Marshall CB.Molecular pathology of non-small cell lung cancer:

a practical guide. Am J Clin Pathol 2012;138:332–46.3. JordanMA,Wilson L.Microtubules as a target for anticancer drugs. Nat Rev

Cancer 2004;4:253–65.4. Kavallaris M. Microtubules and resistance to tubulin-binding agents. Nat

Rev Cancer 2010;10:194–204.5. Reiman T, Lai R, Veillard AS, Paris E, Soria JC, Rosell R, et al. Cross-

validation study of class III beta-tubulin as a predictive marker for benefitfrom adjuvant chemotherapy in resected non-small-cell lung cancer: anal-ysis of four randomized trials. Ann Oncol 2012;23:86–93.

6. Vilmar AC, Santoni-Rugiu E, Sorensen JB. Class III beta-tubulin inadvanced NSCLC of adenocarcinoma subtype predicts superior outcomein a randomized trial. Clin Cancer Res 2011;17:5205–14.

7. Gan PP, McCarroll JA, Byrne FL, Garner J, Kavallaris M. Specific beta-tubulin isotypes can functionally enhance or diminish epothilone Bsensitivity in non-small cell lung cancer cells. PLoS One 2011;6:e21717.

8. Gan PP, McCarroll JA, Po'uha ST, Kamath K, Jordan MA, Kavallaris M.Microtubule dynamics, mitotic arrest, and apoptosis: drug-induced differ-ential effects of betaIII-tubulin. Mol Cancer Ther 2010;9:1339–48.

9. Gan PP, Pasquier E, KavallarisM. Class III beta-tubulinmediates sensitivityto chemotherapeutic drugs in non small cell lung cancer. Cancer Res2007;67:9356–63.

10. McCarroll JA, Gan PP, Liu M, Kavallaris M. betaIII-tubulin is a multifunc-tional protein involved in drug sensitivity and tumorigenesis in non-smallcell lung cancer. Cancer Res 2010;70:4995–5003.

11. Egevad L, Valdman A, Wiklund NP, Seve P, Dumontet C. Beta-tubulin IIIexpression in prostate cancer. Scand J Urol Nephrol 2010;44:371–7.

12. Katsetos CD, Draber P, Kavallaris M. Targeting betaIII-tubulin in glioblas-toma multiforme: from cell biology and histopathology to cancer thera-peutics. Anticancer Agents Med Chem 2011;11:719–28.

13. Alban A, David SO, Bjorkesten L, Andersson C, Sloge E, Lewis S, et al. Anovel experimental design for comparative two-dimensional gel analysis:two-dimensional difference gel electrophoresis incorporating a pooledinternal standard. Proteomics 2003;3:36–44.

14. UnluM,MorganME,Minden JS. Difference gel electrophoresis: a single gelmethod for detecting changes in protein extracts. Electrophoresis1997;18:2071–7.

15. Verrills NM, LiemNL, Liaw TY,Hood BD, Lock RB, KavallarisM. Proteomicanalysis reveals a novel role for the actin cytoskeleton in vincristineresistant childhood leukemia–an in vivo study. Proteomics 2006;6:1681–94.

16. Byrne FL, Yang L, Phillips PA, Hansford LM, Fletcher JI, Ormandy CJ, et al.RNAi-mediated stathmin suppression reduces lungmetastasis in an ortho-topic neuroblastoma mouse model. Oncogene 2014;33:882–90.

17. Sagnella SM, Duong H,MacMillan A, Boyer C,Whan R,McCarroll JA, et al.Dextran-based doxorubicin nanocarriers with improved tumor penetra-tion. Biomacromolecules 2014;15:262–75.

18. Raval GN, Bharadwaj S, Levine EA, Willingham MC, Geary RL, Kute T,et al. Loss of expression of tropomyosin-1, a novel class II tumorsuppressor that induces anoikis, in primary breast tumors. Oncogene2003;22:6194–203.

19. Boyer C, Teo J, Phillips P, Erlich RB, Sagnella S, Sharbeen G, et al. Effectivedelivery of siRNA into cancer cells and tumors using well-defined biode-gradable cationic star polymers. Mol Pharm 2013;10:2435–44.

20. Berardi R, Morgese F, Onofri A, Mazzanti P, Pistelli M, Ballatore Z, et al.Role of maspin in cancer. Clin Transl Med 2013;2:8.

21. Frey A, Soubani AO, Adam AK, Sheng S, Pass HI, Lonardo F. Nuclear,compared with combined nuclear and cytoplasmic expression of maspin,is linked in lung adenocarcinoma to reduced VEGF-A levels and in Stage I,improved survival. Histopathology 2009;54:590–7.

22. Goulet B, ChanG,Chambers AF, Lewis JD. An emerging role for the nuclearlocalization of maspin in the suppression of tumor progression andmetastasis. Biochem Cell Biol 2012;90:22–38.

23. Gupta GP, Massague J. Cancer metastasis: building a framework. Cell2006;127:679–95.

24. Elliott NT, Yuan F. A review of three-dimensional in vitro tissuemodels fordrug discovery and transport studies. J Pharm Sci 2011;100:59–74.

25. Mina LA, Sledge GW Jr. Rethinking the metastatic cascade as a therapeutictarget. Nat Rev Clin Oncol 2011;8:325–32.

26. Sheng S, Qiao M, Pardee AB. Metastasis and AKT activation. J Cell Physiol2009;218:451–4.

27. Stambolic V, Suzuki A, de la Pompa JL, Brothers GM, Mirtsos C, Sasaki T,et al. Negative regulation of PKB/Akt-dependent cell survival by the tumorsuppressor PTEN. Cell 1998;95:29–39.

28. Katakura H, Takenaka K, Nakagawa M, Sonobe M, Adachi M, Ito S, et al.Maspin gene expression is a significant prognostic factor in resected non-small cell lung cancer (NSCLC). Maspin in NSCLC. Lung Cancer2006;51:323–8.

29. Takanami I, Abiko T, Koizumi S. Expression of maspin in non-small-celllung cancer: correlation with clinical features. Clin Lung Cancer 2008;9:361–6.

30. Zheng HC, Saito H, Masuda S, Wang ZG, Takano Y. Cytoplasmic andnuclear maspin expression in lung carcinomas: an immunohistochemicalstudy using tissue microarrays. Appl Immunohistochem Mol Morphol2008;16:459–65.

31. Beltran AS, Blancafort P. Reactivation of MASPIN in non-small cell lungcarcinoma (NSCLC) cells by artificial transcription factors (ATFs). Epige-netics 2011;6:224–35.

32. Nam E, Park C. Maspin suppresses survival of lung cancer cells throughmodulation of Akt pathway. Cancer Res Treat 2010;42:42–7.

33. Kim S,Han J, Kim J, Park C.Maspin expression is transactivated by p63 andis critical for the modulation of lung cancer progression. Cancer Res2004;64:6900–5.

34. Maekawa T, Sano Y, Shinagawa T, Rahman Z, Sakuma T, Nomura S, et al.ATF-2 controls transcription of Maspin and GADD45 alpha genes inde-pendently from p53 to suppress mammary tumors. Oncogene 2008;27:1045–54.

35. Neal CL, Henderson V, Smith BN, McKeithen D, Graham T, Vo BT, et al.Snail transcription factor negatively regulates maspin tumor suppressor inhuman prostate cancer cells. BMC Cancer 2012;12:336.

36. De Donato M, Mariani M, Petrella L, Martinelli E, Zannoni GF, Vellone V,et al. Class III beta-tubulin and the cytoskeletal gateway for drug resistancein ovarian cancer. J Cell Physiol 2012;227:1034–41.

37. Mhaidat NM, Thorne RF, Zhang XD, Hersey P. Regulation of docetaxel-induced apoptosis of human melanoma cells by different isoforms ofprotein kinase C. Mol Cancer Res 2007;5:1073–81.

38. Bharadwaj S, Thanawala R, BonG, Falcioni R, PrasadGL. Resensitization ofbreast cancer cells to anoikis by tropomyosin-1: role of Rho kinase-dependent cytoskeleton and adhesion. Oncogene 2005;24:8291–303.

39. Qin L, Zhang M. Maspin regulates endothelial cell adhesion andmigration through an integrin signaling pathway. J Biol Chem 2010;285:32360–9.

40. Taddei ML, Giannoni E, Fiaschi T, Chiarugi P. Anoikis: an emerginghallmark in health and diseases. J Pathol 2012;226:380–93.

41. Zhong X, Rescorla FJ. Cell surface adhesion molecules and adhesion-initiated signaling: understanding of anoikis resistance mechanisms andtherapeutic opportunities. Cell Signal 2012;24:393–401.

42. Maehama T, Dixon JE. The tumor suppressor, PTEN/MMAC1, depho-sphorylates the lipid second messenger, phosphatidylinositol 3,4,5-tri-sphosphate. J Biol Chem 1998;273:13375–8.

43. Song MS, Salmena L, Pandolfi PP. The functions and regulation of thePTEN tumour suppressor. Nat Rev Mol Cell Biol 2012;13:283–96.






2015;75:415-425. Published OnlineFirst November 20, 2014.Cancer Res Joshua A. McCarroll, Pei Pei Gan, Rafael B. Erlich, et al. Lung Cancer

Small Cell−Promote Tumorigenesis and Anoikis Resistance in Non III-Tubulin Acts through the PTEN/AKT Signaling Axis toβ/TUBB3

Updated version

10.1158/0008-5472.CAN-14-2740doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2014/11/25/0008-5472.CAN-14-2740.DC2

Access the most recent supplemental material at:

Cited articles

http://cancerres.aacrjournals.org/content/75/2/415.full#ref-list-1

This article cites 43 articles, 8 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/75/2/415.full#related-urls

This article has been cited by 6 HighWire-hosted articles. Access the articles at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected]

To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/75/2/415To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/lookup/doi/10.1158/0008-5472.CAN-14-2740

http://cancerres.aacrjournals.org/content/suppl/2014/11/25/0008-5472.CAN-14-2740.DC2

http://cancerres.aacrjournals.org/content/75/2/415.full#ref-list-1

http://cancerres.aacrjournals.org/content/75/2/415.full#related-urls

http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]

http://cancerres.aacrjournals.org/content/75/2/415


Documents

TUBB3 bIII-Tubulin Acts through the PTEN/AKT Signaling ...Matrix-assisted laser desorption/ionization–time-of-ﬂight mass spectrometry Protein spots were excised from SYPRO Ruby–stained