Cell Cycle and Senescence

PKC« Controls Mitotic Progression by RegulatingCentrosome Migration and Mitotic SpindleAssemblySilvia Martini1,2, Tanya Soliman2, Giuliana Gobbi1, Prisco Mirandola1, Cecilia Carubbi1,Elena Masselli1, Giulia Pozzi1, Peter J. Parker2,3, and Marco Vitale1,4

Abstract

To form a proper mitotic spindle, centrosomes must be dupli-cated and driven poleward in a timely and controlled fashion.Improper timing of centrosome separation and errors in mitoticspindle assemblymay lead to chromosome instability, a hallmarkof cancer. Protein kinaseC epsilon (PKCe) has recently emerged asa regulator of several cell-cycle processes associated with theresolution of mitotic catenation during the metaphase–anaphasetransition and in regulating the abscission checkpoint. However,an engagement of PKCe in earlier (pre)mitotic events has not beenaddressed. Here, we now establish that PKCe controls prophase-to-metaphase progression by coordinating centrosomemigrationandmitotic spindle assembly in transformed cells. This control isexerted through cytoplasmic dynein function. Importantly, it is

also demonstrated that the PKCe dependency of mitotic spindleorganization is correlated with the nonfunctionality of theTOPO2A-dependent G2 checkpoint, a characteristic of manytransformed cells. Thus, PKCe appears to become specificallyengaged in a programme of controls that are required to supportcell-cycle progression in transformed cells, advocating for PKCe asa potential cancer therapeutic target.

Implications: The close relationship between PKCe dependencyfor mitotic spindle organization and the nonfunctionality of theTOPO2A-dependent G2 checkpoint, a hallmark of transformedcells, strongly suggests PKCe as a therapeutic target in cancer.Mol Cancer Res; 16(1); 3–15. �2017 AACR.

IntroductionBipolar spindle assembly is a highly coordinated process that

requires the separation of the centrosome in prophase along thenuclear envelope and the interaction of microtubules emanat-ing from the spindle poles with the kinetochores. This arrange-ment allows the duplicated chromosomes to be pulled to theopposite sides of the cell upon sister chromatid separation, thusleading to the formation of two daughter cells with an equalnumber of chromosomes. Improper timing of centrosomeseparation and errors in mitotic spindle assembly have beendemonstrated to be a potentially frequent source of aneuploidyin human cancers (1–3).

Cytoplasmic dynein is a large motor protein involved inmany cellular functions including microtubule organization,cell motility, intracellular trafficking, and organelle positioning(4). Powering the movement along microtubules toward their

minus ends, dynein plays critical roles in cell division, includ-ing early centrosome separation and bipolar spindle formation(5–7), cargo transport along microtubules (8, 9), spindlepositioning (10), and chromosome movement (11). Becauseof these multiple functions, dynein localizes to centrosomes,nuclear envelope (NE), mitotic spindle microtubules, kineto-chores (KT), and the cell cortex during G2 and mitosis (12–14).

Protein kinase C epsilon (PKCe) is a serine/threonine kinaseinvolved in tumor cell invasion and metastasis (15). PKCe isfound frequently overexpressed in a wide variety of humancancers and has been implicated in malignant transformationof cells, including invasion and metastasis (16–19). Besides itsoncogenic role, PKCe has been also implicated in proliferationand differentiation in different cell types (20–25). Furthermore,an emergent role for PKCe has been described in transformedcell models, in the resolution of mitotic catenation during themetaphase–anaphase transition and in regulating the comple-tion of cytokinesis (26–30). Although it has been reportedpreviously that the hypophosphorylated PKCe associates withCG-NAP in the centrosome area (31), a role in early (pre)mitotic events has not been established. Here we demonstratethat PKCe activity is required for proper centrosome migrationin prophase, thus regulating prophase-to-metaphase progres-sion. We identify cytoplasmic dynein as a PKCe binding partnerand show that dynein inhibition phenocopies the inhibition ofthe kinase. In this novel scenario, by influencing dynein func-tion, PKCe is positioned as a coordinator of prophase-to-metaphase progression. Importantly, we further demonstratethat PKCe has this role in mitotic spindle organization intransformed cell models, but not in "normal" diploid cells,

1Department ofMedicine and Surgery, University of Parma, Parma, Italy. 2ProteinPhosphorylation Laboratory, Francis Crick Institute, London, United Kingdom.3King's College London, New Hunts House, Guy's Campus, London, UnitedKingdom. 4CoreLab, Azienda Ospedaliero-Universitaria di Parma, Parma, Italy.

Note: Supplementary data for this article are available at Molecular CancerResearch Online (http://mcr.aacrjournals.org/).

Corresponding Author: Marco Vitale, University of Parma, Ospedale Maggiore,14 Via Gramsci, Parma, 43126, Italy. Phone: 39-0521-033032; Fax: 39-0521-033033; E-mail: marco.vitale@unipr.it.

doi: 10.1158/1541-7786.MCR-17-0244

�2017 American Association for Cancer Research.

MolecularCancerResearch

www.aacrjournals.org 3

on July 14, 2020. © 2018 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from

Published OnlineFirst October 11, 2017; DOI: 10.1158/1541-7786.MCR-17-0244

consistent with the phenotype of the mouse knockout andsuggesting this emergent PKCe/dynein pathway as a potentialtherapeutic target.

Materials and MethodsReagents

All reagentswereobtained fromSigmaAldrich unless otherwisespecified. ATM/ATR inhibitors, NaPP1, BIM-1, EHNA, RO-3306,and Ciliobrevin were obtained from Calbiochem. Blu577was obtained from Dr. Jon Roffey, Cancer Research Technology(London, United Kingdom).

Cell linesCervix adenocarcinoma HeLa cells and colon adenocarci-

noma DLD1-(GFP-PKCe-M486A) cells were routinely culturedin DMEM containing 10% FCS, antibiotics (50 U/mL penicil-lin, 50 mg/mL streptomycin) and were incubated at 37�C and10% CO2.

Erythroleukemia HEL 92.1.17 cells were routinely cultured inRPMI containing 10%FCS, antibiotics and incubated at 37�C and5% CO2.

Normal retina epithelial RPE1-hTERT were routinely cul-tured in DMEM containing 10% FCS, antibiotics, and MEMnonessential amino acids (Gibco), incubated at 37�C and 10%CO2. HeLa, RPE-1, HEL 92.1.17, and DLD-1 cells were obtainedfrom the cell services of the Francis Crick Institute, where theywere routinely screened for mycoplasma. Cells were passagedtwo or three times weekly at ratios between 1:5 and 1:10 andculture for the described experiments for 8/12 weeks betweeneach collection.

Normal skin-derived immortalized human keratinocytes,clone NCTC 2544, were obtained from the ATCC and culturedin EMEM (Euroclone) containing 10% FCS, L-glutamine (2mmol/L). Cells were passaged two or three times weekly at ratiosbetween 1:5 and 1:10. Cell cultures were routinely assayed formycoplasma contamination by Mycoalert mycoplasm detectionkit (Lonza).

Normal human fibroblasts were kindly provided by ProfessorBussolati (Department of Medicine and Surgery, University ofParma, Parma, Italy) and established as previously described (32).Cells were cultured in DMEM containing 20% FCS, antibiotics,and incubated at 37�C and 5%CO2. Cells were passaged at ratiosbetween 1:2 and 1:5, kept in culture, and used in the describedexperiments for 1 month after the collection. Cell cultures wereassayed for mycoplasm contamination by Mycoalert MycoplasmDetection Kit (Lonza).

Cell synchronization and flow cytometryHEL cells were synchronized in G1–S border by culturing in

growth medium supplemented with 2 mmol/L thymidine for16 hours. Cells were washed and released into growth mediumfor 6 hours and arrested in G2–M border adding 10 mmol/L RO-3306 for 16 hours. Cells were then released into growthmedium for 90 minutes to obtain the population enriched inmetaphase. To assess the efficiency of synchronization, aliquotsof HEL cells blocked in G1–S and G2–M phases were permea-bilized with 70% ethanol for 1 hour at 4�C, washed with PBS,and incubated with PBS containing 20 mg/mL propidiumiodide (PI) and 100 mg/mL RNAse-A for 15 minutes in darkat room temperature before analysis. Analysis of samples was

performed by FC500 flow cytometer (Beckman Coulter) andthe Expo ADC software (Beckman Coulter).

DLD1-(GFP-PKCe-M486A) and HeLa cells were synchro-nized in metaphase by a sequential block and release withthymidine and RO-3306. Cells were monitored using the LeicaDM IL phase contrast microscope (40�/0.5NA) and cells werecollected when 50%–60% were in metaphase. DLD1-(GFP-PKCe-M486A) cell line was cultured with doxycycline(100 ng/mL) overnight to induce expression of GFP-PKCe.

Formaldehyde crosslinkingSynchronized cells were scraped, pelleted, and incubated

using formaldehyde 1% at 4�C for 10 minutes. Following acentrifugation at 1,800 rpm for 3 minutes, the cross-linkingreaction was quenched with 1.25 mol/L glycine/PBS, as de-scribed by Klockenbush and Kast (33).

Immunoprecipitation and immunoblottingFor coimmunoprecipitation (co-IP) experiments, HeLa cells

were collected in ice-cold RIPA buffer (50 mmol/L Tris pH 7.4,150mmol/LNaCl, 1mmol/L EDTA, 1%Nonidet P-40, 0.1%SDS,and 0.5% sodium deoxycholate) supplemented with fresh pro-tease inhibitors and 1 mmol/L phenylmethylsulfonyl fluorideand incubated for 30 minutes at 4�C. Lysed proteins were immu-noprecipitated with Dynabeads Protein G (Invitrogen) for 1 hourat 4�C after coating with control IgM. Protein complexes wereeluted with SDS 2� sample buffer after washing with RIPA buffer.Specifically, 10 mg mouse anti-dynein IC (Sigma Aldrich) andmouse-IgG (Santa Cruz Biotechnology) were used for immuno-precipitation. Proteins were resolved in 10% polyacrylamide gelsand analyzed by immunoblotting using specific primary antibo-dies diluted as described by manufacturer's protocol, specifically,rabbit anti-PKCe (Merck Millipore) and mouse anti-dynein IC(Sigma Aldrich).

For the endogenous protein expression analysis, the followingantibodies were used: rabbit anti-PKCe (Merck Millipore),rabbit anti-PKCd (Cell Signaling Technology), rabbit anti-PKCbII(Santa Cruz Biotechnology), rabbit anti-PKCi (Abcam), rabbitanti-PKCq (Abcam), mouse anti-GAPDH (Merck Millipore).Membranes were washed and incubated for 1 hour at roomtemperature with peroxidase-conjugated anti-rabbit (ThermoFisher Scientific) or peroxidase-conjugated anti-mouse IgG(Thermo Fisher Scientific) and resolved by ECL Supersignal WestPico Chemiluminescent Substrate detection system (ThermoFisher Scientific). Protein densitometric analysis was performedby using the ImageJ software system.

Pharmacologic inhibitionDLD1-(GFP-PKCe-M486A) cell line was cultured with doxycy-

cline (100 ng/mL) 1 mmol/L NaPP1 to inhibit PKCe activity, asdescribed previously (27). In HeLa, DLD1-(GFP-PKCe-M486A),HEL, RPE-1, NCTC2544, and human fibroblasts, PKCe wasinhibited using 1 mmol/L BIM-1 or 0.5 mmol/L Blu577, asdescribed previously (29). Dynein activity was inhibited using100 mmol/L EHNA or 20 mmol/L ciliobrevin.

Co-IP using GFP-trap, proteolytic digestion, andnano-LC-MS/MS analysis

All chemicals were purchased from Sigma at the highestgrade possible unless otherwise stated. All solvents andnano-LC-MS/MS additives were purchased as LC-MS grade

Martini et al.

Mol Cancer Res; 16(1) January 2018 Molecular Cancer Research4

on July 14, 2020. © 2018 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from

Published OnlineFirst October 11, 2017; DOI: 10.1158/1541-7786.MCR-17-0244

from Thermo Fisher Scientific. DLD1-(GFP-PKCe-M486A) cellssynchronized in G1–S, G2–M and metaphase were scraped,pelleted, and incubated for cross-linking reaction as describedabove. All samples were coimmunoprecipitated using GFP-Trap(Chromotek). Proteins were separated using a SDS-PAGE gel[NuPAGE 3%–8% Tris-Acetate gel (Invitrogen)] for short gellane extraction for MS analysis and 8 bands from each lanewere excised and processed further using the previouslydescribed in gel digestion procedure adapted for a Janus LiquidHandling System (Perkin Elmer). Ten microliters of gel-extracted peptides in 1 % acetonitrile and 1 % formic acid inwater were analyzed by nano-LC-MS/MS) using an AcquityUPLC (Waters) connected to a LTQ Orbitrap XL Mass Spec-trometer (Thermo Scientific). Raw files containing MS spectrawere processed using a protein database search, carried outusing MaxQuant (version 1.5.2.8) as described previously,against a UniProt human database. Intensity-based absolutequantification (iBAQ) was utilized for label-free quantitationof the reported proteins. Data were then further analyzed usingPerseus (version 1.4.0.2).

Live-cell imagingFor video microscopy experiments, cells were cultured on

LabTek chambered coverglass slides (Nunc) in Liebovitz CO2-independent media (Gibco). All experiments were performed atlow light level inverted microscope (Nikon TE2000) imagingsystem equipped with a lamina-flow heater maintaining a con-stant temperature of 37� 0.01�C, a Plan-Fluor 40X DIC lens, andaXenon lamp forfluorescence excitation. Imageswere taken usinga high quantum efficiency charge-coupled device camera (AndorIxon) every 3 minutes.

Immunofluorescence microscopy and image analysisAdherent cells were grown on 13-mm glass coverslips and fixed

and permeabilized with PHEM buffer (60 mmol/L PIPES pH6.8,25 mmol/L HEPES pH7.4, 10 mmol/L EGTA pH8, 4 mmol/LMgSO4, 4% paraformaldeyhyde, and 0.1% Triton X-100) for 20minutes. HEL samples were fixed in methanol for 5 minutes at�20�C, centrifuged onto 1 mg/mL poly-L-lysine–coated 10-mmcoverslips, and permeabilized with 0.5% Triton X-100 for 7minutes. Cells were then blocked using 3% BSA/PBS and probedusing the following primary antibodies in 3% BSA/PBS: mouseanti-dynein IC (Sigma Aldrich), mouse anti-p150(Glued; BDTransduction Laboratory), mouse anti-BubR1 (Novus Biologi-cals), rabbit anti-phosphoCENP-A (S7; Cell Signaling Technolo-gy), human anti-centromere (ACA; Antibodies Inc.15-234-0001),mouse anti-atubulin (Sigma Aldrich), rabbit anti-atubulin(Abcam), and rabbit anti-PKCe (Abcam). The secondary antibo-dies used were obtained from Thermo Fisher Scientific and wereall used diluted in 3% BSA/PBS: goat anti-rabbit and goat anti-mouse Alexa Fluor 549, goat anti-rabbit goat anti-mouse AlexaFluor 488, and goat anti-human Alexa Fluor 647 (Thermo FisherScientific). All coverslips were mounted using ProLong Gold withDAPI (Invitrogen).

HeLa images were acquired using Carl Zeiss LSM 780 confocalmicroscope equipped with a X63 Plan-APOCHROMAT DIC oil-immersionobjective and serial 1-mmZ-sectionswere taken. Imageanalysis was carried out using Metamorph image analysis soft-ware. HEL, DLD1-(GFP-PKCe-M486A), NCTC2544, and humanfibroblasts were examined with a Nikon Eclipse 80i fluorescentmicroscope equippedwithNikon Plan color 40X/0.75 andNikon

Plan Apo VC 100X/1.4 oil immersion objective. Images wereobtained using Nikon Camera DS-JMC and images acquisitionwas performed using Nis element F2.30 (Nikon). Image analysiswas performed using ImageJ software.

In situ proximity ligation assayThe DuoLink In Situ PLA Kit (Olink Bioscience) with the

DuoLink in situ Detection Reagent Orange (Sigma Aldrich) wasused to detect PKCe/dynein interaction according to the manu-facturer's protocols. HeLa cells were grown on 8-well Culture-Slides (Falcon) overnight and treated with or without Blu557 for1 hour. Cells were subsequently fixed and permeabilized withPHEM buffer and blocked with 3% BSA/PBS. The slides weredirectly used for the assay and primary antibody mix solutioncontaining anti-PKCe (Abcam) and anti-dyneinIC (SigmaAldrich) diluted in 3% BSA/PBS was added to each sample andincubated in a humidity chamber for 1 hour at room temperature.Mouse IgG and Rabbit IgG (Santa Cruz Biotechnology) were usedas negative controls. The assay was subsequently performedfollowing the manufacturer's instructions. The slides weremounted with ProLong Gold with DAPI (Invitrogen). Imageswere acquired using Carl Zeiss LSM 780 confocal microscopeequipped with X63 Plan-APOCHROMAT DIC oil-immersionobjective and analyzed, and serial 1-mm Z-sections were takenusing ZEN image analysis software. Z-sections were summed andPKCe/dynein interaction was quantified counting the number ofsignals each field counted using ImageJ.

PKC« downregulationFor short hairpin RNA (shRNA)-based gene silencing, HEL cells

were infected and were subsequently cultured in the presence ofpuromycin (2 mg/mL), to select infected, puromycin-resistantcells. Cellswere then collected after 5 days of puromycin selection.The pLKO.1 lentiviral vector encoding shRNA against humanPKCe (NM_005400; shPKCe) were obtained from Open Biosys-tems (Thermo Fisher Scientific). As control (shRNA CT), we usedthe MISSION pLKO.1-puro Non-target shRNA Control Plasmid,not targeting any known genes (Sigma-Aldrich). The shRNA-expressing viruses were produced in 293TL cells according tostandard protocols. For siRNA transfection, HiPerfect (Qiagen)was used according to the manufacturer's recommendations;all siRNAs were used at the final concentration of 10 nmol/L.siScramble1 and siPKCe1: PKCe expression levels were down-regulated by transfection of double-stranded siRNA (dsRNA)designed to target sequences corresponding to nt 223–244,429–450, 942–963, and 1158–1179 on human PKCemRNA(NM005400). The target sequences are the following: 50-AAGAT-CAAAATCTGCGAGGCC-30, 50-AAGAT CGAGC TGGCTG TCTTT-30, 50-AACTA CAAGG TCCCT ACCTTC-30, and 50-AAAAAGCT-CATTGCTGGTGCC-30. The respective sense and antisense RNAsequences were synthesized by the Silencer siRNA ConstructionKit (Ambion). siScramble 2 was purchased from Qiagen: catalogno. 1027310 (50-AATTCTCCGAACGTGTCACGT-30). siPKCe2was purchased as SmartPool from Dharmacon: siGenome PKCesiRNA Cat. D-004653(50-GGGCAAAGAUGAAGUAUAU-30) andcatalog no. J-004653-08-0050 (50-GACGUGGACUGCACAAU-GA-30).

Statistical analysisStatistical analyses were performed by using one-way ANOVA

or t test when applicable. Prism software (GraphPad)was used for

Early Mitotic Progression is Regulated by PKCe

www.aacrjournals.org Mol Cancer Res; 16(1) January 2018 5

on July 14, 2020. © 2018 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from

Published OnlineFirst October 11, 2017; DOI: 10.1158/1541-7786.MCR-17-0244

all the calculations. The level of statistical significance is repre-sented as follows: n.s., P > 0.05; �, P < 0.05; ��, P < 0.01; ���, P <0.001; ����, P < 0.0001.

ResultsPKC« inhibition affects mitotic spindle morphology

A role for PKCe in mitosis has been established previously(26–30). However, a cell-cycle function preceding onset ofmitosis is not documented. We used immunofluorescencemicroscopy to investigate PKCe subcellular localization in theearly stages of mitosis. PKCe localizes to the centrosomes fromG2–M phase to metaphase, in a pattern identical to that ofg-tubulin (Fig. 1A). To assess the effects of PKCe inhibition,HeLa cells were treated for 1 hour with bisindolylmaleimide1 (BIM-1), a protein kinase C inhibitor, or Blu577, a structur-ally unrelated and more selective PKCe inhibitor. PKCe local-ization at the centrosome is not affected by the treatment witheither inhibitor (Fig. 1B). In contrast to control cells, in whichthe microtubules were organized into well-defined radialarrays, Blu577- and BIM-1–treated cells showed an unfocusedand disoriented mitotic spindle with misaligned chromosomes(Fig. 1C). Strikingly, PKCe inhibition by Blu577 results in asignificant higher percentage of cells (52%� 4%) which displayabnormal mitotic spindle morphology compared with thecontrol (15.66% � 2.88%; P < 0.001 vs. CT; (Fig. 1D)]. BIM-1 treatment results in a stronger effect on mitotic spindleorganization than Blu577 (72% � 4.36%; P < 0.01 vs. Blu577).This might be due to the broader effect of BIM-1, which is ableto inhibit not only PKCe but other PKC isoforms includingPKCb, which have previously been implicated in microtubuleorganization and spindle function (34). To confirm this find-ing, we treated HeLa cells with two different siRNAs to down-regulate PKCe expression for 48 hours (Fig. 1E and F). Note thatno effect of PKCe knockdown on other PKC isoforms was found(Supplementary Fig. S1A and S1B). Consistently, followingPKCe knockdown, we observed an increase in the number ofcells in metaphase with defects in mitotic spindle structure insiPKCe-transfected cells, compared with siScramble controls.Several others PKC isoforms have been reported to be centro-some-associated (34–37). However, using siRNA, chemicalbiology, and drugs, we here demonstrate the involvement ofPKCe in mitotic spindle organization.

Inhibition of PKC« results in a delayed metaphase entry dueto prolonged centrosome migration

We imaged unsynchronized HeLa cells stably expressingmCherry-H2B and GFP-Tubulin by time-lapse video microscopyand recorded the time taken from prophase, when the centro-somes are duplicated and the DNA is condensed, to metaphasealignment (Fig. 2A; Supplementary Fig. S1C). Most control cells(58%) took from 3 to 9 minutes to reach metaphase alignment.This frequency is dramatically reduced upon treatment withBlu577 or BIM-1 (2.2% � 2% Blu577, 2.5% � 1% BIM-1; Fig.2B). As shown in Fig. 2A, a representative control cell reachesmetaphase alignment in 9 minutes from prophase, whereas thecell treated with Blu577 is still in prophase/prometaphase at thistime point. Consistently, PKCe inhibition causes a delay inmetaphase entry, as illustrated in the cumulative frequency graphsin Fig. 2B and Supplementary Fig. S2A, showing an increase in thenumber of cells taking from 10 to 20 minutes (42% � 7.6%

Control vs. 57.4% � 3% Blu577 and 58.3% � 0.5 BIM-1) orbetween 21 and 39 minutes to transit from prophase to meta-phase (3%� 0.4% Control vs. 40.4%� 1.3% Blu577 and 39.3%� 1.7% BIM-1). In line with the delay to metaphase entry, wealso observed a lag in the timing of centrosomemovement towardthe opposite poles. We scored the time that centrosomes tookfrom their duplication, until they are positioned to the oppositepoles assembling the bipolar spindle. Duplicated centrosomes ofuntreatedHeLa cells did not takemore than 6minutes tomigrate,whereas upon Blu577 and BIM-1 treatment, centrosomes tookmore than 10 minutes [17� 5% (P< 0.001) and 30.5� 6.5%(P < 0.001 vs. Control) minutes, respectively)(Fig. 2c)]. In addi-tion, we observed in cells treatedwith PKCe inhibitors an aberrantmorphology of the spindle that we refer to as "bending" pheno-type (Fig. 2A and C, white arrow), which correlates with the delayin centrosome movement and consequent mitotic spindle disor-ganization observed by confocal imaging (Fig. 2A and D). Inaccordance with the critical role of PKCe in cytokinesis (26), thenumber of cells unable to complete cytokinesis is increased uponPKCe inhibition (Supplementary Fig. S1D).

Identification of cytoplasmic dynein and PKC« as bindingpartners in G2–M

We used MS/MS fingerprinting to identify potential PKCe-binding partners throughout mitosis. Among the known spin-dle-associated proteins reported, dynein heavy chain 1 was foundtobindPKCe inG2–Mandmetaphase (Fig. 3A).We confirmed thePKCe interaction with dynein by coimmunoprecipitation of theendogenous proteins inHeLa cells. PKCe physically interacts withdynein in cells synchronized in metaphase (Fig. 3B). To confirmand determine the subcellular localization of PKCe/dynein inter-action,weused an in situproximity ligation assay (PLA). This assayrevealed that PKCe interacts with dynein around the nuclearenvelope and chromatin in prophase, and in the mitotic spindleregion in prometaphase and metaphase (Fig. 3C). No PLA signalwas detected in the negative control (Fig. 3D). Our findingssuggested an involvementof PKCe inmitotic spindle organizationand mitotic progression. As dynein plays critical roles in centro-some separation and bipolar spindle assembly, we hypothesizedthat PKCe supports dynein function and regulates prophase–metaphase progression in these cells.

Dynein ATPase activity inhibition phenocopies inhibition ofPKC«

To investigate whether PKCe regulates dynein function, wetreated HeLa cells with EHNA (Erythro-9-3-(2-hydroxynonyl)adenine), to inhibit dynein ATPase activity and interfere withdynein binding with microtubules (38). One hour inhibition ofdynein activity caused aberrations in mitotic spindle morpho-logy comparable with the spindle defect seen when HeLa cellswere treated with Blu577 and BIM-1 (Fig. 4A). Cells treatedwith EHNA showed a defective mitotic spindle when comparedwith control (84% � 6%; P < 0.001; Fig. 4B). HeLa cells werethen imaged by time-lapse microscopy following treatmentwith EHNA (Fig. 4D). As expected, dynein inhibition resultedin a delay of centrosome movement, thereby delaying mitoticprogression. Indeed, prophase to metaphase alignment transi-tion is delayed in most EHNA-treated cells: 61% � 4% of cellstook from 10 to 20 minutes to transit and 33% � 5% of cellstook from 21 to 39 minutes to reach metaphase alignment(Fig. 4C; Supplementary Fig. S2B). Interestingly, the very slow

Martini et al.

Mol Cancer Res; 16(1) January 2018 Molecular Cancer Research6

on July 14, 2020. © 2018 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from

Published OnlineFirst October 11, 2017; DOI: 10.1158/1541-7786.MCR-17-0244

Figure 1.

A, Localization of PKCe at the centrosomes in G2–M, prophase and metaphase. HeLa cells were stained with anti-PKCe (green), anti-g-tubulin (red), and DAPIfor DNA (blue). Scale bars, 5 mm. B, Blu577 or BIM-1 treatment do not affect PKCe localization at the centrosome. Cells were stained with anti-PKCe(green), anti-g-tubulin (red), and DAPI (blue). Scale bars, 5 mm. C, Representative images of HeLa cells treated with Blu577 and BIM-1. Cells were stainedwith anti-PKCe (green), anti-a-tubulin (red), and DAPI (blue). Scale bars, 5 mm. D, The percentage of cells with a disrupted mitotic spindle is increasedupon treatment with PKCe inhibitors Blu577 and BIM-1. Quantification of cells in metaphase with a defective mitotic spindle. Chart shows mean ofthree experiments� SD, n > 100 per condition per experiment. ��� , P < 0.001 versus CT. BIM-1 has stronger effect on mitotic spindle organization (��P < 0.01 vs.Blu577). E, PKCe silencing using siRNA alters mitotic spindle organization in HeLa cells. Representative images and quantification (F) of HeLa cellstreated for 48 hours with two different siRNA targeting PKCe; cells were stained with anti-PKCe (green), anti-g-tubulin (red) and DAPI (blue). Scale bars, 5 mm.Chart shows mean of three experiments � SD, n > 100 per condition per experiment. �� , P < 0.01; ���, P < 0.001 vs. SiScramble.

Early Mitotic Progression is Regulated by PKCe

www.aacrjournals.org Mol Cancer Res; 16(1) January 2018 7

on July 14, 2020. © 2018 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from

Published OnlineFirst October 11, 2017; DOI: 10.1158/1541-7786.MCR-17-0244

transiting cells (21–39 minutes) were unable to properly com-plete mitosis, indicating that the multiple functions exerted bydynein throughout mitosis are essential for the proper com-pletion of cell division. Furthermore, we noted the "bending"behavior of GFP-Tubulin, which was very similar to the phe-notype seen with PKCe inhibition (Fig. 4E and F). The samebending behavior was also observed in cells treated with adifferent dynein inhibitor, Ciliobrevin (ref. 39; SupplementaryFig. S2). Collectively, these data indicate that inhibition ofdynein phenocopies PKCe inhibition and indicates that PKCeacts with dynein in the prophase to metaphase transition.

Dynein accumulates at kinetochores upon PKC« inhibitionImmunofluorescence microscopy studies demonstrated that

dynein localizes to kinetochores prior to MTs attachment,where it regulates initial interactions with spindle fibers andcoordinates the early aspects of chromosome movement in

prometaphase (40, 41). As chromosomes achieve bipolarattachment, kinetochore dynein staining becomes less promi-nent and is undetectable once the chromosomes are aligned.Loss of dynein at the kinetochores coincides with an enhancedlabelling along the spindle fibers and spindle poles. As PKCeinhibition results in a prolonged prometaphase, we investigat-ed dynein localization using immunofluorescence in HeLa cellstreated for 1 hour with Blu577 or BIM-1. Untreated HeLa cellsin metaphase displayed dynein labeling on the spindle fibersand at the spindle poles, at the plus end of MTs and at the cortexat one side of the cell (Fig. 5A). In contrast, HeLa cells treatedwith PKCe inhibitors showed a disorganized spindle morphol-ogy with chromosomes in a prometaphase-like state and anenrichment of dynein at the kinetochores and at the plus end ofunattached MTs. Indeed, double staining of dynein and thechromatin-associated protein phospho-CENP-A (Ser7) uponPKCe inhibition confirmed the dynein localization at the

Figure 2.

A, HeLa cells that stably express mCherry-H2B and GFP-Tubulin were imaged by time-lapse microscopy. Stills taken from time-lapse imaging of HeLa cellsupon treatment with Blu577 and Control; time in minutes is marked in white. The yellow arrow indicates the centrosome delay. The white arrowindicates the "bending" phenotype. As shown in the yellow field, representative control cell is in metaphase at 9 minutes, whereas cells treated with Blu577are still in prophase/prometaphase. B, Graph shows the cumulative frequency of cells that took 3–9 minutes (quick), 10–20 minutes (medium) and21–39 minutes (slow) from prophase to metaphase alignment upon PKCe inhibition. Chart shows the average of three experiments � SD, n > 30 percondition per experiment. ���, P < 0.001 versus CT. C, Graph represents the percentage of cells which centrosome took more than 10 minutes to reachthe opposite poles (black charts) and the cells presenting the "bending" phenotype (gray). Charts show the average of three experiments � SD, n > 30 percondition per experiment. ���, P < 0.001 Blu577 and BIM-1 versus CT.

Martini et al.

Mol Cancer Res; 16(1) January 2018 Molecular Cancer Research8

on July 14, 2020. © 2018 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from

Published OnlineFirst October 11, 2017; DOI: 10.1158/1541-7786.MCR-17-0244

kinetochores (Fig. 5B). In an unsynchronized populationof HeLa cells, dynein localization at the kinetochores is detect-able in the 2% � 1% of cells compared with 37% � 4% and the50% � 2% of cells treated with Blu577 and BIM-1, respectively(P < 0.001; Fig. 5C). This is in line with our previous obser-vation that inhibition of PKCe in an unsynchronized popula-tion of cells results in a delay of metaphase entry, sustaining thenotion that PKCe is required for prophase to metaphase pro-gression. In addition, dynein localization at the kinetochorescan be explained as a perturbation of dynein streaming fromthe plus ends of microtubules to the spindle poles, in line withthe hypothesis that dynein activity is regulated by PKCe. Wetherefore examined dynein localization in cells treated withEHNA. As expected, EHNA treatment resulted in a higherpercentage of cells with a disorganized mitotic spindle anddynein enrichment at the kinetochores. This phenotype issimilar to that seen upon Blu577 and BIM-1 treatment, con-cordant with our observation that dynein inhibition pheno-copies PKCe inhibition (Fig. 5C; Supplementary Fig. S3A).EHNA treatment revealed also a new localization of PKCe atthe kinetochores, as confirmed by the double staining with thecentromere antibody CREST (Supplementary Fig. S3B andS3C). We showed that dynein interacts with PKCe in prome-taphase and the inhibition of dynein and PKCe result in anaccumulation of both proteins at the kinetochores. This sug-gests that in prometaphase, the PKCe/dynein complex is

trapped at the kinetochores when either of the proteins areinhibited. We also examined the localization of p150 (Glued),the largest subunit of dynactin complex, which links dyneinto cargos and increases its processivity (42). As for dynein,but with a less profound effect, dynactin localized at thekinetochores to a significantly higher extent in cells treatedwith PKCe inhibitors indicating that not only dynein, butthe dynein/dynactin complex, becomes more stably localizedat the kinetochores when PKCe is inhibited (SupplementaryFig. S3D and S3E).

A prediction arising from the delayed movement and organi-zation of productive kinetochore–microtubule (K-MT) engage-ment, is that in unsynchronized cells, there will be an increase inthe steady-state presence of APC/C regulators such as BubR1 (43)at the kinetochore (BubR1 poleward streaming is K-MT anddynein dependent and is abolished upon dynein inhibition;refs. 44, 45). We assessed therefore whether PKCe inhibitioninfluences BubR1 retention at kinetochores following treatmentwith Blu577 or BIM-1 (Supplementary Fig. S3F). PKCe inhibitionresulted in a higher percentage of cells with a kinetochore accu-mulation of BubR1 compared with controls, consistent with thedelay in productive K-MT formation(Fig. 5D and E). This iscomparable with EHNA treatment which, as expected, caused anincrease in the number of cells retaining both BubR1 and PKCe atthe kinetochores. Note that this delayed K-MT organization isdistinct from the delayed release of BubR1 associated with a

Figure 3.

A, Bar graph comparing PRKCE (blue) and DYNC1H1 (red) at G1–S, G2–M and metaphase stages of the cell cycle using log10-transformed intensity basedabsolute quantification (iBAQ) values. An order of magnitude difference is observed for DYNC1H1 at G1 compared to G2 and metaphase in contrast toPRKCE. B, Co-IP assay was performed using anti-Dynein IC antibody in HeLa cells synchronized in prometaphase/metaphase and cross-linked using 1%formaldehyde. IgG-Mouse used as control, total lysate and Dynein IP were analyzed by Western blot analysis and probed for PKCe and for Dynein IC to assessthe efficiency of the assay. Under the extraction conditions used, there is an approximately 1:2 relative ratio of PKCe:Dynein IC in cells enriched inprometaphase/metaphase. C, Detection of PKCe and Dynein IC interaction in HeLa cells using the in situ PLA. The images show a maximum intensityprojection of the raw image; PLA signals are shown in red and DNA in gray. Scale bars, 5 mm. D, PLA negative control reaction using rabbit anti-PKCeantibody with anti-IgG mouse (left) and mouse anti-Dynein IC antibody with anti-IgG rabbit (right).

Early Mitotic Progression is Regulated by PKCe

www.aacrjournals.org Mol Cancer Res; 16(1) January 2018 9

on July 14, 2020. © 2018 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from

Published OnlineFirst October 11, 2017; DOI: 10.1158/1541-7786.MCR-17-0244

Figure 4.

A, Confocal images of HeLa cells untreated or treated for 1 hour with EHNA. a-tubulin is represented in red and DNA in blue (DAPI). Scale bar, 5 mm.B, The amount of mitotic HeLa cells treated with EHNA (red) with a disrupted spindle is significantly higher compared with control (black). Chart showsmean of three experiments � SD, n > 100 per condition per experiment. ��� , P < 0.001 versus CT. C–F, Mitotic events were live-imaged in HeLa cellsstably expressing mCherry-H2B (red) and GFP-tubulin (green). C, The graph shows the cumulative frequency of cells that took 3–9 minutes (quick),10–20 minutes (medium), and 21–39 minutes (slow) from prophase to metaphase alignment upon Blu577, BIM-1, and EHNA treatment. Chart shows theaverage of three experiments � SD, n > 30 per condition per experiment. ���, P < 0.001 versus CT. D, Stills taken from time-lapse imaging of HeLa cells with orwithout EHNA treatment; as shown in the yellow box and in the zoom, representative control cell is in metaphase at 9 minutes, whereas the celltreated with EHNA is still in prophase. The time in minutes is marked in white. The yellow arrow indicates the centrosome delay, white arrow indicatesthe "bending" phenotype. E, The graphs describes that EHNA treatment (red) causes an increase of the percentage of cells unable to complete mitosis, anincrease of the number of cells with delayed centrosome movement (>10 minutes) and with the "bending" phenotype. Charts show the average ofthree experiments � SD, n > 30 per condition per experiment. ��� , P < 0.001 versus CT. F, Representative images of the centrosome delay (upper lane) andthe "bending" phenotype (lower lane) taken from live imaging of HeLa cells upon EHNA, Blu577, and BIM-1 treatments.

Martini et al.

Mol Cancer Res; 16(1) January 2018 Molecular Cancer Research10

on July 14, 2020. © 2018 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from

Published OnlineFirst October 11, 2017; DOI: 10.1158/1541-7786.MCR-17-0244

Figure 5.

A, Confocal images of HeLa cells treated with Blu577 and BIM-1 for 1 hour, fixed, and stained with anti-Dynein IC (green), a-tubulin (red), and DAPI(blue). In a control cell in metaphase (upper panel), Dynein IC is localized at the plus end of the microtubules (inset), at the cortex (arrow) andalong the microtubules of the mitotic spindle. In HeLa cells treated with Blu577 and BIM-1, the mitotic spindle is disorganized, the chromatidsare condensed prophase/prometaphase-like and Dynein is localized at the kinetochores (inset). B, Representative images of HeLa cells treatedwith Blu577 and BIM-1 for 1 hour, fixed, and stained with anti-Dynein IC (green), pCENP-A(S7) (red) and DAPI (blue). In the control, Dynein IClocalizes with pCENP-A(S7) at the plus end of the microtubules (top), whereas it is coupled with the kinetochores when PKCe is inhibited(Blu577, middle; BIM-1, bottom). C, Quantification of mitotic HeLa cells with Dynein IC labelled at the spindle (gray charts) and at the kinetochores(black charts). Chart shows mean of three experiments � SD, n > 100 per condition per experiment. ��� , P < 0.001 versus CT. D, Quantification ofthe amount of HeLa cells in metaphase with BubR1 retained at the kinetochores. Chart shows mean of three experiments � SD, n > 50 per condition perexperiment. ��� , P < 0.001 versus CT. E, Representative images of BubR1 staining (green) and DNA (blue) in metaphase cells upon the differenttreatments. Scale bars, 5 mm.

Early Mitotic Progression is Regulated by PKCe

www.aacrjournals.org Mol Cancer Res; 16(1) January 2018 11

on July 14, 2020. © 2018 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from

Published OnlineFirst October 11, 2017; DOI: 10.1158/1541-7786.MCR-17-0244

catenation triggered metaphase/anaphase transition (the latter isassociated with the loss of Mad2 from kinetochores while thedelay observed here is associated with retention of kinetochoreMad2; data not shown).

PKC« regulates mitotic spindle assembly in transformed cellmodels

It has been demonstrated previously that PKCe is a regulatorof the metaphase catenation response, following escape fromthe G2 Topoisomerase-2–dependent checkpoint (27). Defectsin the G2 checkpoint are a characteristic of many transformedcells (28, 46). Here we showed that PKCe inhibition results indefects in mitotic spindle organization with consequent delayin mitotic progression in HeLa cells. Therefore, we sought todetermine whether the abnormal mitotic spindle morphologycaused by PKCe inhibition was restricted to transformed cellmodels. As observed in HeLa cells, PKCe localized at thecentrosomes in the additional transformed cell lines HEL andDLD-1 (PKCeM486A) (Supplementary Fig. S4A and S4B). ThePKCe dependency for mitotic spindle assembly was assessed bytreating HEL and DLD-1(PKCeM486A) for 1 hour with Blu577(Fig. 6A). Similarly to what was observed in HeLa cells, thenumber of cells in metaphase with defective mitotic spindles issignificantly higher in HEL (43.10% � 0.34%) and DLD-1 cells(52% � 5.29%) treated with Blu577, compared with controls(9.44% � 0.79% HEL and 16% � 2.65% DLD1; P < 0.001). Toconfirm this finding, we used specific shRNA and siRNA todownregulate the expression of PKCe in HEL cells and the ATPanalogue NaPP1 to inhibit the gate-keeper modified kinase inthe DLD-1 cell line (27, 47). Consistently, in both cell lines, weobserved an increase in the number of mitotic cells withabnormal spindle geometry upon PKCe inhibition, confirmingthat PKCe is necessary for mitotic spindle assembly (Fig. 6B–D;Supplementary Fig. S4F). We further assessed PKCe localizationin "normal" diploid cell lines: normal human keratinocytes(NCTC 2544), primary human fibroblasts, and nontrans-formed RPE1-hTERT cells showed PKCe enrichment at thecentrosomes as the transformed cell lines (Supplementary Fig.S4G). Therefore, we investigated PKCe inhibition response innontransformed cell lines by treating cells with Blu577 underthe same conditions as the transformed cells. Although PKCeinhibition resulted in a higher number of cells with spindledefects, compared with control cells (Supplementary Fig. S4H),this percentage is significantly lower compared with trans-formed cell lines (Fig. 6E); only 6.2% � 0.68% RPE1- hTERTcells, respond to PKCe inhibition with an abnormal spindlemorphology. These results suggest that the PKCe dependency ofmitotic spindle organization is related to the transformed statusof cells.

DiscussionWe have investigated the effects of PKCe inhibition in early

mitotic events by using different methods of inhibition inmultiple transformed cell models (HeLa human cervical carci-noma, HEL human erythroleukemia and DLD1 human colo-rectal adenocarcinoma). Consistently, cells in prometaphasefrom all cell lines displayed an abnormal mitotic spindlemorphology compared with controls, implicating the involve-ment of PKCe in bipolar spindle assembly. Interestingly, thiswas not the case in the "normal" keratinocytes NCTC2544,

primary human fibroblasts, and the nontransformed retinalpigment epithelium RPE1-hTERT cell lines. The G2–M cate-nation checkpoint is defective in several cell lines, includingHeLa cells, whereas it is functional in "normal" cells (46).Brownlow and colleagues have previously demonstrated thatcells with a leaky G2 catenation checkpoint are dependent onPKCe for catenation resolution in mitosis (28). In an interestingalignment with this finding, we suggest that the dependence onPKCe for mitotic spindle assembly is related to the transfor-mation status of the cell. These finding suggests that PKCeinhibition leads to a delay of metaphase transit in transformedcells, increasing the risk of failure in MTs/kinetochores attach-ment and potentially contributing to genomic instability.

Live-cell imaging of asynchronous HeLa cells revealed thatPKCe inhibition causes a delay of centrosome migration to theopposite sides of the nucleus. Nevertheless, increasing microtu-bule nucleation at centrosomes occurs and a mitotic spindle isassembled albeit showing an abnormalmorphology, a "bending"phenotype. Thismight reflect the dominance of chromosome armengagement and the kinesin-10 dependent chromosome organi-zation, that is, the polar ejection force (48). The centrosomemovement delay and the "bending" phenotype caused by PKCeinhibition are likely to interfere with the kinetochore capture bythe mitotic microtubules, leading to a delay to metaphase entry.

Among the several motor proteins implicated in cell division,cytoplasmic dynein is amicrotubule motor complex, which playskey roles in multiple processes, such as centrosome separationand spindle organization. How cytoplasmic dynein is able tofulfill such awide range of processes in cell division is still unclear,making it challenging to address functions attributed to specificdynein pools. However, it has been demonstrated that the inter-action with multiple adaptor proteins is essential to orchestratethe timing and localization of dynein functions (49). Here wedemonstrate that PKCe physically and functionally interacts withcytoplasmic dynein from G2–M phase to metaphase. Indeed, thephenotypes observed upon dynein inhibition are comparablewith the effects of PKCe inhibition, suggesting that PKCe anddynein are functionally related.

After nuclear envelope breakdown, dynein is recruited to thekinetochores where it regulates proper microtubule attachmentand chromosome alignment in prometaphase. Defects in K-MTattachment and the spindle assembly checkpoint (SAC) duringcell division are strongly associated with genomic instability.The SAC is active during prometaphase and once the chromo-somes are aligned and the mitotic spindle is correctly assem-bled in metaphase, dynein moves from the kinetochores alongthe microtubules streaming the proteins involved in the SAC,such as BubR1.

It has been shown that PKCe plays a role partially antagonisticwith dynein streaming in G2 checkpoint–compromised cells,where delayed metaphase progression is triggered by sister chro-matid catenation consequent to treatment with ICRF-193 (28).Here cells are delayed in progressing to anaphase by PKCe action,characterized by the retention of BubR1 at the kinetochores and anonsilenced SAC. We show evidence here that PKCe inhibitioncauses the accumulation of both dynein and BubR1 at the kine-tochores in asynchronous HeLa cells, indicating that PKCe inhi-bition compromises K-MT attachment thereby influencing thedynein-dependent streaming and delaying sister chromatidsalignment in metaphase. The interaction of PKCe and dynein (asdetermined by coimmunoprecipitation and PLA) suggests that

Martini et al.

Mol Cancer Res; 16(1) January 2018 Molecular Cancer Research12

on July 14, 2020. © 2018 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from

Published OnlineFirst October 11, 2017; DOI: 10.1158/1541-7786.MCR-17-0244

Figure 6.

A, PKCe inhibition using Blu577 results in defects in mitotic spindle organization in HeLa (black), DLD1-PKCeM486A (white) and HEL (red) cells.��� , P < 0.001 versus Control. B, Quantification of DLD1 cells in metaphase with defective mitotic spindle, with (red) or without (black) Na-PP1 toinhibit PKCe. ��� , P < 0.001 versus Control. C, Quantification of HEL cells in metaphase with defective mitotic spindle, treated with shControl (black) orshPKCe (red). �� , P < 0.01 versus shControl. D, Representative images of DLD1 (left) and HEL cells (right), treated with or without the respectivePKCe inhibitor. Cells were stained using anti-a-tubulin antibody (green) and DAPI (blue). Scale bar, 10 mm. E, Quantification of defective mitoticspindle in transformed cell lines HeLa (black), DLD1 (red), and HEL (yellow) compared with nontransformed cell lines NCTC 2544 (green), humanfibroblast (blue), and RPE1-hTERT (white). All cell lines were treated with the PKCe inhibitor Blu577 for 1 hour. ���, P < 0.001. Charts in all panels showthe mean of three experiments � SD, n > 100 per condition per experiment using HeLa DLD1-PKCeM486A, NCTC 2544, human fibroblast, and RPE1-hTERTcells, n ¼ 30 per condition per experiment using HEL cells.

Early Mitotic Progression is Regulated by PKCe

www.aacrjournals.org Mol Cancer Res; 16(1) January 2018 13

on July 14, 2020. © 2018 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from

Published OnlineFirst October 11, 2017; DOI: 10.1158/1541-7786.MCR-17-0244

PKCe may act as a general regulator of dynein function–modify-ing cargo selection perhaps both in prometaphase (as describedhere) and at anaphase entry (28).

Taken together, the presented data demonstrate that PKCemodulates prophase-to-metaphase progression in transformedcell models, by regulating centrosome migration and mitoticspindle organization via dynein interaction. These findingsand those reported previously (26–30) establish PKCe as a keyregulator of cell-cycle progressionwhere there areDNA-associatedstresses encountered. In conclusion, the close relationshipbetween PKCe dependency for mitotic spindle organization andthe nonfunctionality of G2 checkpoint(s), (a hallmark of trans-formed cells), is a strong prescription for PKCe as a therapeutictarget in cancer.

Disclosure of Potential Conflicts of InterestP.J. Parker has ownership interest (including patents) in Fastbase Solutions

and is a consultant/advisory boardmember for NovintumBioscience, CRT, andMRC. No potential conflicts of interest were disclosed by the other authors.

Authors' ContributionsConception and design: S. Martini, G. Gobbi, P.J. Parker, M. VitaleDevelopment of methodology: C. CarubbiAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): C. Carubbi

Analysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): S. Martini, T. Soliman, P. Mirandola, E. Masselli,G. PozziWriting, review, and/or revision of the manuscript: S. Martini, T. Soliman,G. Gobbi, P. Mirandola, P.J. Parker, M. VitaleStudy supervision: M. Vitale

AcknowledgmentsThe authors thank Luciana Cerasuolo, Vincenzo Palermo, Domenico

Manfredi, and Davide Dallatana, University of Parma, Italy, for technicalsupport. We also thank the Francis Crick Institute (LRI) facilities for thevaluable support, in particular the Light Microscopy and the ProteinAnalysis and Proteomics technology platforms, Francis Crick Institute(London, United Kingdom).

Grant SupportThis work was supported by Regione Emilia-Romagna Area 1 – Strategic

Program 2010-2012 (to G. Gobbi).

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 May 5, 2017; revised August 4, 2017; accepted October 4, 2017;published OnlineFirst October 11, 2017.

References1. Silkworth WT, Nardi IK, Paul R, Mogilner A, Cimini D. Timing of centro-

some separation is important for accurate chromosome segregation.Mol Biol Cell 2012;23:401–11.

2. Fukasawa K. Oncogenes and tumour suppressors take on centrosomes.Nat Rev Cancer 2007;7:911–24.

3. Nam HJ, van Deursen JM. Cyclin B2 and p53 control proper timing ofcentrosome separation. Nat Cell Biol 2014;16:538–49.

4. Roberts AJ, Kon T, Knight PJ, SutohK, Burgess SA. Functions andmechanicsof dynein motor proteins. Nat Rev Mol Cell Biol 2013;14:713–26.

5. Vaisberg EA, Koonce MP, McIntosh JR. Cytoplasmic dynein plays a role inmammalian mitotic spindle formation. J Cell Biol 1993;123:849–58.

6. Tanenbaum ME, Macu� rek L, Galjart N, Medema RH. Dynein, Lis1 andCLIP-170 counteract Eg5-dependent centrosome separation during bipolarspindle assembly. EMBO J 2008;27:3235–45.

7. Raaijmakers JA, van Heesbeen RG, Meaders JL, Geers EF, Fernandez-GarciaB, Medema RH, et al. Nuclear envelope-associated dynein drives prophasecentrosome separation and enables Eg5-independent bipolar spindleformation. EMBO J 2012;31:4179–90.

8. Young A, Dictenberg JB, Purohit A, Tuft R, Doxsey SJ. Cytoplasmic dynein-mediated assembly of pericentrin and gamma tubulin onto centrosomes.Mol Biol Cell 2000;11:2047–56.

9. Silva PM, Reis RM, Bolanos-Garcia VM, FlorindoC, Tavares �AA, BousbaaH.Dynein-dependent transport of spindle assembly checkpoint proteinsoff kinetochores toward spindle poles. FEBS Lett 2014;588:3265–73.

10. Kiyomitsu T, Cheeseman IM. Chromosome- and spindle-pole-derivedsignal generate an intrinsic code for spindle position and orientation.Nat Cell Biol 2012;14:311–7.

11. Collins ES, Balchand SK, Faraci JL, Wadsworth P, Lee WL. Cell cycle-regulated cortical dynein/dynactin promotes symmetric cell division bydifferential pole motion in anaphase. Mol Biol Cell 2012;23:3380–90.

12. Whyte J, Bader JR, Tauhata SB, Raycroft M, Hornick J, Pfister KK, et al.Phosphorylation regulates targeting of cytoplasmic dynein to kinetochoresduring mitosis. J Cell Biol 2008;183:819–34.

13. Dujardin DL, Vallee RB. Dynein at the cortex. Curr Opin Cell Biol2002;14:44–9.

14. Tanenbaum ME, Akhmanova A, Medema RH. Dynein at the nuclearenvelope. EMBO Rep 2010;11:649.

15. Gorin MA, Pan Q. Protein kinase C epsilon: an oncogene and emergingtumor biomarker. Mol Cancer 2009;8:9.

16. Basu A, Sivaprasad U. Protein kinase Cepsilon makes the life and deathdecision. Cell Signal 2007;19:1633–42.

17. Gobbi G, Masselli E, Micheloni C, Nouvenne A, Russo D, Santi P,et al. Hypoxia-induced down-modulation of PKCepsilon promo-tes trail-mediated apoptosis of tumor cells. Int J Oncol 2010;37:719–29.

18. Gutierrez-UzquizaA, Lopez-HaberC, JerniganDL, Fatatis A,KazanietzMG.PKCe is an essential mediator of prostate cancer bone metastasis.Mol Cancer Res 2015;13:1336–46.

19. Garg R, Benedetti LG, Abera MB, Wang H, Abba M, Kazanietz MG. Proteinkinase C and cancer: what we know and what we do not. Oncogene2014;33:5225–37.

20. Queirolo V, Galli D,Masselli E, Borz��RM,Martini S, Vitale F, et al. PKCe is aregulator of hypertrophic differentiation of chondrocytes in osteoarthritis.Osteoarthritis Cartilage 2016;24:1451–60.

21. Galli D, Carubbi C, Masselli E, Corradi D, Dei Cas A, Nouvenne A, et al.PKCe is a negative regulator of PVAT-derived vessel formation. ExpCell Res2015;330:277–86.

22. Gobbi G, Mirandola P, Carubbi C, Galli D, Vitale M. Protein kinase C e inhematopoiesis: conductor or selector? Semin Thromb Hemost 2013;39:59–65.

23. Di Marcantonio D, Galli D, Carubbi C, Gobbi G, Queirolo V, Martini S,et al. PKCe as a novel promoter of skeletal muscle differentiation andregeneration. Exp Cell Res 2015;339:10–9.

24. Carubbi C, Masselli E, Martini S, Galli D, Aversa F, Mirandola P, et al.Human thrombopoiesis depends on Protein kinase Cd/protein kinase Cefunctional couple. Haematologica 2016;101:812–20.

25. Masselli E, Carubbi C, Gobbi G, Mirandola P, Galli D, Martini S, et al.Protein kinase Ce inhibition restores megakaryocytic differentiation ofhematopoietic progenitors fromprimarymyelofibrosis patients. Leukemia2015;29:2192–201.

26. Saurin AT, Durgan J, Cameron AJ, Faisal A, Marber MS, Parker PJ. Theregulated assembly of a PKCepsilon complex controls the completion ofcytokinesis. Nat Cell Biol 2008;10:891–901.

27. Saurin AT, BrownlowN, Parker PJ. Protein kinase C epsilon in cell division:control of abscission. Cell Cycle 2009;8:549–55.

28. Brownlow N, Pike T, Zicha D, Collinson L, Parker PJ. Mitotic catenation ismonitored and resolved by a PKCe-regulated pathway. Nat Commun2014;5:5685.

Martini et al.

Mol Cancer Res; 16(1) January 2018 Molecular Cancer Research14

on July 14, 2020. © 2018 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from

Published OnlineFirst October 11, 2017; DOI: 10.1158/1541-7786.MCR-17-0244

29. Pike T, Brownlow N, Kjaer S, Carlton J, Parker PJ. PKCe switches AuroraB specificity to exit the abscission checkpoint. Nat Commun 2016;7:13853.

30. H€am€alist€o S, Pouwels J, de FranceschiN, SaariM, Ivarsson Y, ZimmermannP, et al. A ZO-1/a5b1-integrin complex regulates cytokinesis downstreamof PKCe in NCI-H460 cells plated on fibronectin. PLoS One 2013;8:e70696.

31. TakahashiM,MukaiH,Oishi K, IsagawaT,OnoY.Associationof immaturehypophosphorylated protein kinase Cepsilon with an anchoring proteinCG-NAP. J Biol Chem 2000;275:34592–6.

32. Marchesi C, Dall'Asta V, Rotoli BM, Bianchi MG, Maggini C, Gazzola GC,et al. Chlorpromazine, clozapine andolanzapine inhibitanionic aminoacid transport in cultured human fibroblasts. Amino Acids 2006;31:93–9.

33. Klockenbusch C, Kast J. Optimization of formaldehyde cross-linkingfor protein interaction analysis of non-tagged integrin beta1. J BiomedBiotechnol 2010;2010:927585.

34. Chen D, Purohit A, Halilovic E, Doxsey SJ, Newton AC. Centrosomalanchoring of protein kinase C betaII by pericentrin controls microtubuleorganization, spindle function, and cytokinesis. J Biol Chem 2004;279:4829–39.

35. Ma W, Koch JA, Viveiros MM. Protein kinase C delta (PKCdelta)interacts with microtubule organizing center (MTOC)-associated pro-teins and participates in meiotic spindle organization. Dev Biol 2008;320:414–25.

36. Quann EJ, Liu X, Altan-Bonnet G, Huse M. A cascade of protein kinase Cisozymes promotes cytoskeletal polarization in T cells. Nat Immunol2011;12:647–54.

37. Atwood SX, Li M, Lee A, Tang JY, Oro AE. GLI activation by atypical proteinkinase C i/l regulates the growth of basal cell carcinomas. Nature 2013;494:484–8.

38. Yamada M, Toba S, Yoshida Y, Haratani K, Mori D, Yano Y, et al. LIS1 andNDEL1 coordinate the plus-end-directed transport of cytoplasmic dynein.EMBO J 2008;27:2471–83.

39. Roossien DH, Miller KE, Gallo G. Ciliobrevins as tools for studying dyneinmotor function. Front Cell Neurosci 2015;9:252.

40. Yang Z, Tulu US, Wadsworth P, Rieder CL. Kinetochore dynein is requiredfor chromosome motion and congression independent of the spindlecheckpoint. Curr Biol 2007;17:973–80.

41. Pfarr CM, Coue M, Grissom PM, Hays TS, Porter ME, McIntosh JR.Cytoplasmic dynein is localized to kinetochores during mitosis. Nature1990;345:263–5.

42. Schroer TA. Dynactin. Annu Rev Cell Dev Biol 2004;20:759–79.43. Musacchio A, Salmon ED. The spindle-assembly checkpoint in space

and time. Nat Rev Mol Cell Biol 2007;8:379–93.44. Howell BJ, McEwen BF, Canman JC, Hoffman DB, Farrar EM, Rieder CL,

et al. Cytoplasmic dynein/dynactin drives kinetochore protein transportto the spindle poles and has a role in mitotic spindle checkpoint inacti-vation. J Cell Biol 2001;155:1159–72.

45. Varma D, Monzo P, Stehman SA, Vallee RB. Direct role of dynein motorin stable kinetochore-microtubule attachment, orientation, and align-ment. J Cell Biol 2008;182:1045–54.

46. Damelin M, Bestor TH. The decatenation checkpoint. Br J Cancer 2007;96:201–5.

47. Knight ZA, Shokat KM. Chemical genetics: where genetics and pharma-cology meet. Cell 2007;128:425–430.

48. Zhong A, Tan FQ, Yang WX. Chromokinesin: Kinesin superfamilyregulating cell division through chromosome and spindle. Gene 2016;589:43–8.

49. Raaijmakers JA, Tanenbaum ME, Medema RH. Systematic dissection ofdynein regulators in mitosis. J Cell Biol 2013;201:201–15.

www.aacrjournals.org Mol Cancer Res; 16(1) January 2018 15

Early Mitotic Progression is Regulated by PKCe

on July 14, 2020. © 2018 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from

Published OnlineFirst October 11, 2017; DOI: 10.1158/1541-7786.MCR-17-0244

2018;16:3-15. Published OnlineFirst October 11, 2017.Mol Cancer Res   Silvia Martini, Tanya Soliman, Giuliana Gobbi, et al.   Migration and Mitotic Spindle Assembly

Controls Mitotic Progression by Regulating CentrosomeεPKC

  Updated version

  10.1158/1541-7786.MCR-17-0244doi:

Access the most recent version of this article at:

  Material

Supplementary

  http://mcr.aacrjournals.org/content/suppl/2017/10/10/1541-7786.MCR-17-0244.DC1

Access the most recent supplemental material at:

   

   

  Cited articles

  http://mcr.aacrjournals.org/content/16/1/3.full#ref-list-1

This article cites 49 articles, 16 of which you can access for free at:

  Citing articles

  http://mcr.aacrjournals.org/content/16/1/3.full#related-urls

This article has been cited by 1 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

  .pubs@aacr.org

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://mcr.aacrjournals.org/content/16/1/3To request permission to re-use all or part of this article, use this link

on July 14, 2020. © 2018 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from

Published OnlineFirst October 11, 2017; DOI: 10.1158/1541-7786.MCR-17-0244