Non–small cell lung cancer (NSCLC) is notorious for its paltry responses to first-line therapeutic regimens. In contrast to acquired chemoresistance, little is known about the molecular underpinnings of the intrinsic resistance of chemo-naïve NSCLC. Here we report that intrinsic resistance to paclitaxel in NSCLC occurs at a cell-autonomous level because of the uncoupling of mitotic defects from apoptosis. To identify components that permit escape from mitotic stress–induced death, we used a genome-wide RNAi-based strategy, which combines a high-throughput toxicity screen with a live-cell imaging platform to measure mitotic fate. This strategy revealed that prolonging mitotic arrest with a small molecule inhibitor of the APC/cyclosome could sensitize otherwise paclitaxel-resistant NSCLC. We also defined novel roles for CASC1 and TRIM69 in supporting resistance to spindle poisons. CASC1, which is frequently co-amplified with KRAS in lung tumors, is essential for microtubule polymerization and satisfaction of the spindle assembly checkpoint. TRIM69, which associates with spindle poles and promotes centrosomal clustering, is essential for formation of a bipolar spindle. Notably, RNAi-mediated attenuation of CASC1 or TRIM69 was sufficient to inhibit tumor growth in vivo. On the basis of our results, we hypothesize that tumor evolution selects for a permissive mitotic checkpoint, which may promote survival despite chromosome segregation errors. Attacking this adaptation may restore the apoptotic consequences of mitotic damage to permit the therapeutic eradication of drug-resistant cancer cells. Cancer Res; 74(14); 3857–69. ©2014 AACR.

Paclitaxel is a first-line chemotherapeutic agent that inhibits the dynamic instability of microtubules, thus preventing bi-orientation of chromosomes during mitosis (1). Although the reduction of breast and ovarian tumor burden following taxane-based therapies demonstrates efficacy in these settings, responses in non–small cell lung cancer (NSCLC) are rarely curative, as only 30% of patients exhibit a partial response at best, indicating a widespread intrinsic resistance to antimitotic agents (2, 3). Thus, given the promise, yet limitation, of current antimitotic therapies, the identification of mechanisms supporting intrinsic resistance to paclitaxel in NSCLC is essential. Our goal here was to identify the cell autonomous components that permit escape from mitotic stress–induced cell death in a paclitaxel-resistant NSCLC setting.

Efficacy of paclitaxel and other antimitotic agents hinges on the coupling of mitotic defects to cell death. By inhibiting the dynamic instability of microtubules, paclitaxel disrupts chromosome alignment, thereby preventing satisfaction of the spindle assembly checkpoint (SAC). The SAC is composed of sentinel proteins, including MAD2 and BUBR1, which, in the absence of proper microtubule-kinetochore attachments, inhibit the activity of the anaphase promoting complex/cyclosome (APC/C; refs. 4 and 5). Nearly all tumor cells are sensitive to paclitaxel-induced mitotic defects and engage the SAC. However, live-cell imaging studies have revealed that the length and outcome of this mitotic arrest is variable within and among tumor cell lines. For example, apoptosis may be activated directly from mitotic arrest. Otherwise, arrested cells undergo mitotic slippage, defined as an aberrant exit in the presence of misaligned chromosomes thereby forming micronucleated cells that can either die, arrest, or reenter a subsequent division cycle (6–8). This slippage from an SAC-mediated mitotic arrest has been implicated as a survival mechanism, as delaying mitotic exit, either by inhibiting an activator of the APC/C, CDC20, or overexpressing cyclin B1, can increase mitotic dwell time and cell death during or following mitosis (6, 9). The prolonged mitotic arrest may allow for accumulation of death signals that trigger apoptosis (10, 11). However, the molecular components that promote mitotic slippage, particularly in chemoresistant tumor cell lines are unclear.

Previously, we established a pan-genomic, siRNA-based screening platform to identify gene products that modulate sensitivity of NSCLC cells to paclitaxel (12). Here, we use this platform to investigate the molecular components supporting resistance to paclitaxel in a NSCLC system that exhibits no loss of viability following exposure to doses as high as 1 μmol/L. At single cell resolution, we find that resistant cells undergo mitotic slippage and survive as micronucleated cells, indicative that resistance is because of a regulatory setting that permits the deflection of mitotic stress–induced cell death. In agreement with this notion, a global screen discovered components, which when depleted, sensitize these chemorefractory cells to paclitaxel. Suppression of a cohort of these chemosensitizers induced a protracted mitotic arrest, which we find is essential for postmitotic cell death. Functional elaboration of these chemosensitizers reveals that prolonging mitotic arrest can be accomplished either by direct inhibition of the APC/C, or through collateral spindle damage, because of the depletion of CASC1 or TRIM69, which we reveal here are novel regulators of mitotic spindle assembly in tumor cells. Thus, prolonging mitotic arrest, which can be achieved through multiple avenues, is a dominant mechanism to sensitize otherwise paclitaxel-resistant NSCLC cells.

Cells and reagents

Human bronchial epithelial cells (HBEC), NSCLC, 293, and HeLa cells were a gift from J. Minna and M. White (University of Texas Southwestern Medical Center, Dallas, TX). HeLa, H1299, HCC4017, HCC1171, H1299, H1155, and HBEC lines were validated by short tandem repeat analysis. NSCLC cell lines were maintained in RPMI medium (Gibco) with 5% fetal bovine serum (FBS). HBEC cell lines were maintained in keratinocyte medium plus supplements (Gibco). HeLa and 293 cells were maintained in DMEM with 10% FBS. Paclitaxel (Sigma or Tocris), proTAME (Boston Biochem), and Nocodazole (Sigma) were dissolved in dimethyl sulfoxide (DMSO). Cell-Titer Glo (CTG) and APO-ONE were obtained from Promega. Dharmafect transfection reagents were obtained from Thermo-Fisher. Control siRNA transfections were performed with either non-targeting siRNA pool (Dharmacon) or a pool targeting DLNB14 (12).

Paclitaxel dose curves

Cell lines were plated in 96-well plates to reach 50% confluence at 48 hours, when they were exposed to paclitaxel for 48 hours. Cell viability assessed using the Cell-Titer Glo assay.

Immunoblotting

Cells were lysed in boiling 2× Laemmli sample buffer and resolved by SDS-PAGE. Antibodies used as follows: actin, ANAPC5, CASC1, GAPDH, MCL-1, PDE3B, Myc-A14, Myc-9E10, GST (Santa Cruz); cleaved caspase-3 (Epitomics); MAD2L1 (Covance); TRIM69, β-tubulin (Sigma); pericentrin (Abcam); phospho-histone-H3 (Millipore).

Immunofluorescence

Cells were grown on glass coverslips, fixed in 3.7% formaldehyde, permeabilized with 0.5% Triton X-100, and blocked with PBS, 0.1% Tween-20, 10 mg/mL bovine serum albumin (PBTA). Slips were incubated with primary antibodies in PBTA and, following PBTA washes, incubated with Alexa Fluor–conjugated secondary antibodies (Invitrogen). For BUBR1 staining, cells were fixed in cold methanol. For microtubule preservation in Fig. 3C, cells were extracted with BRB80 (80 mmol/L PIPES pH 6.8, 1 mmol/L MgCl2, 5 mmol/L EGTA) and 0.5% Triton X-100 for 30 seconds followed by a 0.5% glutaraldehyde fix and 0.1% NaBH4 quench. Slides were imaged on an Axioimager upright microscope (Zeiss) equipped with a charge-coupled device camera or a Leica CTR5500 upright scope with a monochrome digital camera (Leica DFC340 FX). Quantitation of fluorescence was performed with ImageJ software and normalized to cytoplasmic fluorescence.

Stable cell line production

Cell lines stably expressing green fluorescent protein-histone H2B (GFP-H2B) or myc-TRIM69A were generated by retroviral transduction. Retrovirus was produced by transfection of 293GP cells with vesicular stomatitis virus G protein (VSV-G) and either pCLNCX-GFP-H2B (gift from Gray Pearson) or pLPCX-myc-TRIM69A. Virus was harvested 48 hours after transfection and transduced cells were selected with geneticin (pCLNCX-GFP-H2B) or puromycin (pLPCX-myc-TRIM69A).

High-content live-cell imaging

Cell lines stably expressing GFP-H2B were transfected in a 96-well plate and exposed to vehicle or paclitaxel 48 hours after transfection and, unless otherwise indicated, imaged immediately following the addition of paclitaxel. Imaging was performed on a BD Pathway 855 Bioimager using a 40× or a 20× high-numerical aperture objective. Images were taken every 20 minutes for 48 to 72 hours. Image sequences were generated using ImageJ and manually quantified.

siRNA screen

A 2-condition, triplicate analysis screen was performed in a 96-well plate format using a Thermo-Fisher library that targets 21,127 unique genes (12). HCC366 cells were transfected with 40 nmol/L siRNAs for 48 hours, followed by exposure to vehicle or 10 nmol/L paclitaxel for an additional 48 hours. Subsequently, CTG was added and values were read on an EnVision Plate Reader. The screening protocol was identical to that previously described with the following exceptions: HCC366 cells were seeded at 1 × 104, transfected with Dharmafect 2 complexed siRNAs in RPMI medium, and cells were treated at 48 hours after transfection with vehicle or paclitaxel and a final FBS concentration of 10% (12). High interest chemosensitizers were ranked by z-score (Supplementary Methods).

siRNA transfections

siRNA transfections were performed using Thermo-Fisher or Sigma siRNA pools. Pools were used at a final concentration of 50 nmol/L for indicated times.

Quantitative reverse transcription PCR

Total RNA was collected using the GenElute Mammalian Total RNA Miniprep Kit (Sigma). Two micrograms of total RNA was used in subsequent reverse transcription using the High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems). Quantitative reverse transcription PCR (qRT-PCR) was performed with TaqMan gene expression assays and ribosomal protein L27 (RPL27) was the endogenous control.

Colony formation assays

HCC366 cells were transfected in a 24-well format and exposed to 0 and 10 nmol/L paclitaxel 48 hours after transfection. Ninety-six hours after transfection, 2 × 103 viable cells were replated in a 6-well format. Cells were fed biweekly for up to 3 weeks before formaldehyde fixation and Geimsa staining. Colonies were counted manually.

Microtubule regrowth assay

Cells were exposed to 11 μmol/L nocodazole for 1 hour, 72 hours after transfection. Cells were then rinsed with PHEM buffer (60 mmol/L PIPES, 25 mmol/L HEPES, 10 mmol/L EGTA, 2 mmol/L MgCl2, 1 μmol/L paclitaxel) and placed in fresh media to recover for the time indicated. Cells were then permeabilized with 0.2% Triton X-100 for 1 minute then fixed and immunostained.

In vivo polymerized tubulin assay

Seventy-two hours after transfection, cells were lysed in microtubule stabilizing buffer (100 mmol/L PIPES, 2 mol/L glycerol, 0.1 M MgCl2, 2 mmol/L EGTA, 0.5% Triton X-100, 5 μmol/L paclitaxel). Subsequently, whole-cell lysates samples were collected and centrifuged for 30 minutes at 4°C and 16,000 × g. The supernatant (monomeric tubulin) was collected and pellet (polymerized tubulin) resuspended in microtubule stabilizing buffer. Fractions were resolved by SDS-PAGE.

cDNA expression and plasmids

Cells were transfected with cDNA expression vectors using Lipofectamine 2000 (Invitrogen). Plasmids used: Tomato-H2B, pCMV-myc (Clontech), pCMV-myc-TRIM69A, pCMV-myc-TRIM69B, and pCMV-myc-TRIM69A (C50S/C53S).

Tumor xenografts

pLKO.1 vectors expressing short hairpin RNA (shRNA) were used to generate lentivirus to infect HCC366 cells. Target knockdown was assessed 72 hours after infection. Cells were harvested 96 hours after infection and 2 × 106 cells were injected into the flank of female NSG (NOD.Cg-Prkdcscid Il2rgtm1Wjl/SzJ JAX) mice. The protocol used 10 week old mice weighing 20–26 grams. A total of 3 × 106 cells in 4-week old mice weighing 15–20 grams were used for paclitaxel studies. Mice were bred at University of North Carolina at Chapel Hill, Chapel Hill, NC. All mice were maintained in sterile conditions according to an approved Institutional Animal Care and Use Committee protocol and abiding by all University of North Carolina Animal Welfare guidelines. Tumor growth was monitored by caliper measurement at indicated time points. In paclitaxel experiments, mice were treated with 20 mg/kg of paclitaxel starting 10 days after injection, at a frequency of 2 days per week for 4 weeks.

Hematoxylin and eosin staining

Tumors from xenograft mouse studies were excised and formalin fixed and paraffin embedded. Tissue was processed by routine microtomy into 5 to 6 μm sections for automated staining. At least 7 fields were manually quantitated for micronucleated cells.

TUNEL staining

Terminal deoxynucleotidyl transferase–mediated dUTP nick end label (TUNEL) staining for apoptotic cells in tissue sections was performed using the DeadEnd Fluorometric TUNEL System (Promega) standard protocol. Apoptotic cells were labeled with fluorescein, and the sections were counterstained with propidium iodide. At least 6 fields were counted per sample.

Pan-genomic screening reveals components that liberate tumor cells from mitotic stress–induced death

As a discovery platform for identifying mechanisms that promote resistance to paclitaxel-induced cell death, we selected HCC366 cells, a member of a cohort of NSCLC cell lines, which undergo limited viability loss and no activation of caspase-3 following exposure to escalating doses of paclitaxel (Fig. 1A and B). Although cell lines in this cohort exhibit minimal viability loss following paclitaxel exposure, a 10 nmol/L dose is sufficient to induce multipolar spindles, misaligned chromosomes, and generate micronucleated progeny (Fig. 1C and Supplementary S1A), indicating that these cells have an intact SAC. Live-cell imaging studies demonstrated that paclitaxel induced a MAD2- and BUBR1-dependent mitotic arrest that uniformly led to mitotic slippage and the generation of micronucleated progeny (Fig. 1D). Single-cell lineage tracing of HCC366 micronucleates, beginning 72 hours after paclitaxel exposure, revealed that the majority (92%) of these cells continued to survive for 40 hours and 8% successfully underwent a subsequent division cycle (Fig. 1E and Supplementary Fig. S1B and Supplementary Video S1). Replating assays revealed that 40% to 50% of treated cells were capable of colony formation, indicating that these cells, despite mitotic damage, maintain the capacity to proliferate (Supplementary Fig. S1C). In contrast, a representative sensitive cell line, H1155, undergoes a dramatic mitotic arrest resolved by either cell death or micronucleation at 100 nmol/L (Supplementary Fig. S1D). The HCC366 mitotic slippage response was also recapitulated in vivo, as tumors grown subcutaneously in immune-deficient mice exposed to paclitaxel were composed of micronucleated cells (Fig. 1F). This finding suggests that HCC366 cells harbor a regulatory context in which mitotic stress is uncoupled from tumor cell death, leading to paclitaxel resistance and continued tumor cell proliferation.

Figure 1.

A, paclitaxel dose curves in indicated NSCLC cell lines. Each point is the mean ± standard error of the mean (SEM) for three independent experiments. B, whole-cell lysates from HCC366 and H1155 exposed to paclitaxel for 48 hours were immunoblotted as indicated. C, immunostaining of HCC366 cells treated for 24 hours. Arrowheads, micronucleated cells. Scale bars, 5 or 15 μm (right). D, single-cell lineage tracing of HCC366 cells stably expressing GFP-H2B (HCC366-GFP-H2B) for 48 hours post-paclitaxel exposure. Each circle represents a single cell. Bar is mean mitotic transit time, which is also indicated. Right, cells were transfected with indicated siRNAs for 48 hours before drug exposure. E, quantitation of single-cell lineage tracing of HCC366-GFP-H2B micronucleates after exposure to 10 nmol/L paclitaxel. Bars represent mean ± range for 100 cells over two experiments. F, H&E staining of HCC366 subcutaneous tumor xenografts treated as indicated. Arrowheads, micronucleated cells. Scale bars, 40 μm. Zoomed images are of white-boxed area.

Figure 1.

A, paclitaxel dose curves in indicated NSCLC cell lines. Each point is the mean ± standard error of the mean (SEM) for three independent experiments. B, whole-cell lysates from HCC366 and H1155 exposed to paclitaxel for 48 hours were immunoblotted as indicated. C, immunostaining of HCC366 cells treated for 24 hours. Arrowheads, micronucleated cells. Scale bars, 5 or 15 μm (right). D, single-cell lineage tracing of HCC366 cells stably expressing GFP-H2B (HCC366-GFP-H2B) for 48 hours post-paclitaxel exposure. Each circle represents a single cell. Bar is mean mitotic transit time, which is also indicated. Right, cells were transfected with indicated siRNAs for 48 hours before drug exposure. E, quantitation of single-cell lineage tracing of HCC366-GFP-H2B micronucleates after exposure to 10 nmol/L paclitaxel. Bars represent mean ± range for 100 cells over two experiments. F, H&E staining of HCC366 subcutaneous tumor xenografts treated as indicated. Arrowheads, micronucleated cells. Scale bars, 40 μm. Zoomed images are of white-boxed area.

Close modal

To comprehensively investigate the molecular basis for deflection of paclitaxel-induced death, we performed a pan-genomic, siRNA-based loss of function screen in the presence and absence of 10 nmol/L paclitaxel in HCC366 cells (Supplementary Fig. S1E and S1F). This analysis returned 49 candidate siRNA pools with a z-score of <−2.5 that correspond to annotated genes in the human GENE database (see Supplementary Methods, Table S1 and S2).

We further stratified these siRNAs using a secondary screening strategy to measure caspase-3/-7 activity and mitotic transit time for 23 candidates selected from diverse functional categories (Fig. 2A; Supplementary Fig. S2A and S2B, and Supplementary Table S1). Target mRNA knockdown was also evaluated (Supplementary Fig. S2C). This counter screen returned 9 targets whose inhibition activated apoptotic signaling, 4 of which increased mitotic transit time at least 20%. Colony formation assays revealed a 50% decrease in replating efficiency, indicating that the cell death observed was sufficient to reduce survival of the population (Fig. 2B). This finding indicated that even in slippage prone NSCLC, a number of molecular mechanisms exist to enhance SAC robustness. Thus, we focused on this cohort of ANAPC5, PDE3B, CASC1, and TRIM69 for further mechanistic elaboration. All 4 of these siRNA pools met our off-target validation criteria of 2 independent siRNAs or nonredundant pools reproducing the apoptotic phenotype (Supplementary Fig. S2D).

Figure 2.

A, graph of relative mitotic transit time (y) vs. relative caspase-3/-7 activity (APO-ONE) in HCC366 cells exposed to paclitaxel. A subset of siRNA gene targets is indicated. B, colony formation assay in HCC366s exposed to 10 nmol/L paclitaxel for 48 hours. Bars represent mean ± SEM from n = 3 independent experiments. C, HCC366-GFP-H2B cells were transfected with indicated siRNAs for 48 hours followed by exposure to 10 nmol/L paclitaxel. Live-cell imaging commenced with paclitaxel treatment and single-cell lineage traces of 50 cells from two independent experiments were measured. D, as in C, but in the presence of siMAD2. E, whole-cell lysates from HCC366 cells transfected for 48 hours followed by exposure to vehicle or 10 nmol/L paclitaxel for 48 hours were immunoblotted with indicated antibodies. F, whole-cell lysates from nontumorigenic HBEC3KT (N) and HCC366 (T) cells treated as in 2E were immunoblotted with indicated antibodies. G, left, relative cleaved caspase-3/-7 activity (APO-ONE) in HCC366s treated with vehicle, 10 nmol/L paclitaxel, or 2.5 μmol/L proTAME for 24 hours. Bars represent mean ± SEM for n = 9 independent experiments. **, P < 0.01 and ***, P < 0.0001 (unpaired 2-tailed Student t test). Right, immunoblot of WCLs of HCC366s treated as in the left.

Figure 2.

A, graph of relative mitotic transit time (y) vs. relative caspase-3/-7 activity (APO-ONE) in HCC366 cells exposed to paclitaxel. A subset of siRNA gene targets is indicated. B, colony formation assay in HCC366s exposed to 10 nmol/L paclitaxel for 48 hours. Bars represent mean ± SEM from n = 3 independent experiments. C, HCC366-GFP-H2B cells were transfected with indicated siRNAs for 48 hours followed by exposure to 10 nmol/L paclitaxel. Live-cell imaging commenced with paclitaxel treatment and single-cell lineage traces of 50 cells from two independent experiments were measured. D, as in C, but in the presence of siMAD2. E, whole-cell lysates from HCC366 cells transfected for 48 hours followed by exposure to vehicle or 10 nmol/L paclitaxel for 48 hours were immunoblotted with indicated antibodies. F, whole-cell lysates from nontumorigenic HBEC3KT (N) and HCC366 (T) cells treated as in 2E were immunoblotted with indicated antibodies. G, left, relative cleaved caspase-3/-7 activity (APO-ONE) in HCC366s treated with vehicle, 10 nmol/L paclitaxel, or 2.5 μmol/L proTAME for 24 hours. Bars represent mean ± SEM for n = 9 independent experiments. **, P < 0.01 and ***, P < 0.0001 (unpaired 2-tailed Student t test). Right, immunoblot of WCLs of HCC366s treated as in the left.

Close modal

Single-cell lineage tracing revealed that the prolonged mitotic arrest in target-depleted cells resolved as mitotic slippage to generate micronucleated progeny; 40% to 50% of which subsequently died as measured by cell blebbing and a loss of movement (Fig. 2C). Indicative of a functional conservation of these genes across multiple genetic backgrounds, their inhibition enhanced paclitaxel-induced mitotic arrest in HCC1171, HCC515, and H2887, cell lines that also exhibit a deflection of mitotic stress–induced death (Supplementary Fig. S2E). Thus, prolonging mitotic arrest induced a postmitotic apoptotic response.

Restoring paclitaxel sensitivity requires a protracted mitotic arrest

Of these validated genes, ANAPC5 is the best characterized as encoding a scaffold in the multiprotein APC/C complex, whose inhibition attenuates the proteolytic cascades required for anaphase onset (13). Although roles for CASC1, TRIM69, and PDE3B in mitosis have not been described, their identification in a cohort with ANAPC5 suggests that they may contribute to APC/C inhibition. Indeed, codepletion of CASC1, TRIM69, or PDE3B with MAD2 significantly reduced mitotic transit time in the presence of paclitaxel (Supplementary Fig. S3A). Codepletion also rescued micronucleated cell death as assessed by live-cell imaging and cleaved caspase-3 protein levels (Fig. 2D and Supplementary Fig. S3B).

The prosurvival protein, MCL-1, is degraded during a prolonged mitosis and has been implicated as a regulator of apoptosis following exposure to antimitotic therapeutics (10, 11). Depletion of CASC1 or TRIM69 leads to a reduction in MCL-1 levels in a paclitaxel-dependent manner whereas PDE3B inhibition led to a reduction in MCL-1 even in the absence of paclitaxel. Consistent with previous reports indicating that MCL-1 is degraded by residual APC/C activity during mitosis, siANAPC5 did not decrease MCL-1 levels (Fig. 2E; ref. 10). Cell death despite MCL-1 stabilization in ANAPC5-depleted cells indicates existence of MCL-1–independent apoptotic mechanisms. Taken together, these findings suggest that enhancing mitotic arrest by 20% to 50% in the presence of paclitaxel can be sufficient to engage a programmed cell death response in chemoresistant NSCLC.

Inhibition of APC/C is synthetic lethal with paclitaxel in chemorefractory NSCLC

Findings thus far indicate that, in principle, prolonging mitotic arrest by inhibiting the APC/C may be a therapeutic strategy to restore paclitaxel sensitivity in NSCLC. Importantly, suppression of ANAPC5 in nontumorigenic cells (HBEC3KT) did not enhance the impact of paclitaxel, suggesting that tumor cells may have an enhanced sensitivity to APC/C inhibition (Fig. 2F). The difference in response is unlikely because of cell-cycle kinetics as the HBEC3KT and HCC366 cells proliferate at approximately the same rate (Supplementary Fig. S3C). Together these data suggest that this combinatorial approach may exhibit a therapeutic window. Indeed, combining a sublethal dose of the APC/C inhibitor, proTAME, with paclitaxel in HCC366 cells enhanced the accumulation of mitotic cells and induced cleaved caspase-3/-7 above levels detected for either agent alone (Fig. 2G and Supplementary Fig. S3D; ref. 14).

To further examine molecular mechanisms that contribute to the duration of mitotic arrest, we focused on CASC1 and TRIM69, whose inhibition displayed minimal, if any, single-agent activity, but which collaborated with paclitaxel to both increase mitotic transit time and reduce MCL-1 levels.

CASC1 regulates microtubule function

Although limited information is available on the function of human CASC1, a polymorphism in the murine Casc1 gene at codon 60 [AGT (Ser) → AAT (Asp)] is associated with susceptibility to urethane-induced lung carcinogenesis in mouse models (15, 16). The significance of these polymorphisms has not been realized in human cancer as none of the 5 human CASC1 isoforms harbor Asn or Ser at residue 60, and other polymorphisms in CASC1 have not been linked to NSCLC susceptibility or progression (17). However, TCGA analysis revealed an amplification of CASC1 in 15%, 11%, and 7% of lung (adenoma and squamous), ovarian, and breast tumors, respectively (Supplementary Fig. S4A; 18). This amplification typically co-occurs with KRAS, which is adjacent to CASC1 on chromosome 12. Although the function of human CASC1 has not been documented, returning this gene as a top paclitaxel sensitizer in a genome-wide screen implies it may have essential roles in supporting cell division. We assessed the consequences of CASC1 depletion alone and in the presence of an otherwise innocuous dose of paclitaxel in a panel of NSCLC lines with differing paclitaxel sensitivity. In resistant cell lines, where we detected mitotic arrest phenotypes, we found a dependence on CASC1 for viability alone or in the presence of paclitaxel (Fig. 3A). In addition, siCASC1 is monogenic lethal or synthetic lethal with paclitaxel in 5 additional NSCLC cell lines, a defect we did not observe in an immortalized human bronchial epithelial (HBEC3KT) cell line (Fig. 3A). Thus, CASC1 seems to be essential in both resistant and sensitive tumor settings, indicating that tumor cells develop a dependence on this protein for mitotic progression. Examination of individual cells indicated that CASC1 is essential for mitotic spindle integrity as inhibition led to (i) exacerbation of the frequency of spindles with >2 pericentrin positive foci in HCC366 cells, (ii) generation of multipolar or miniature spindles in H1299 cells, and (iii) induction of aberrant mitotic spindles in the presence of an otherwise innocuous dose of paclitaxel in H1155 cells (Fig. 3B and Supplementary Fig. S4B). Live-cell imaging of H1155 cells indicated that depletion of CASC1 prolonged mitotic arrest, which resolved as either aberrant mitotic exit or cell death in the presence of paclitaxel (Supplementary Fig. S4C).

Figure 3.

A, whole-cell lysates of NSCLC cells immunoblotted with indicated antibodies following siRNA transfection for 48 hours and subsequent treatment as indicated for 48 hours. siPLK1 was used as a positive control for cleaved caspase-3 detection in HBEC3KT cells. B, as in A, except paclitaxel treatment was 24 hours before immunostaining. Scale bars, 5 μm. C, top, H1299 cells were transfected for 72 hours with indicated siRNAs, then fixed in glutaraldehyde and immunostained. Scale bars, 15 μm. Bottom, immunoblots of H1299 cells following an in vivo polymerized tubulin assay. D, in vivo polymerized tubulin assay in H1299 cells transfected for 72 hours. Lysates were immunoblotted with indicated antibodies. E, HCC366 cells, treated as in B, were fixed in cold methanol and immunostained as indicated. Scale bars, 5 μm. BUBR1 foci were counted by manual inspection in 60 cells per experiment in three independent experiments. Bar indicates mean. ***, P < 0.001, Mann–Whitney t test.

Figure 3.

A, whole-cell lysates of NSCLC cells immunoblotted with indicated antibodies following siRNA transfection for 48 hours and subsequent treatment as indicated for 48 hours. siPLK1 was used as a positive control for cleaved caspase-3 detection in HBEC3KT cells. B, as in A, except paclitaxel treatment was 24 hours before immunostaining. Scale bars, 5 μm. C, top, H1299 cells were transfected for 72 hours with indicated siRNAs, then fixed in glutaraldehyde and immunostained. Scale bars, 15 μm. Bottom, immunoblots of H1299 cells following an in vivo polymerized tubulin assay. D, in vivo polymerized tubulin assay in H1299 cells transfected for 72 hours. Lysates were immunoblotted with indicated antibodies. E, HCC366 cells, treated as in B, were fixed in cold methanol and immunostained as indicated. Scale bars, 5 μm. BUBR1 foci were counted by manual inspection in 60 cells per experiment in three independent experiments. Bar indicates mean. ***, P < 0.001, Mann–Whitney t test.

Close modal

Previous reports indicate that murine and chlamydomonas CASC1 (IC97) orthologues can bind microtubules (15, 19). Given that human and mouse CASC1 proteins are 66.7% identical and 92% identical in the microtubule binding region, we reasoned that the mitotic defects observed following CASC1 depletion could be because of a microtubule defect. We examined the microtubule network in H1299 cells, where siCASC1 alone had a potent impact on mitotic spindle formation and cell viability. siCASC1 diminished the polymerized tubulin network in interphase cells and attenuated microtubule regrowth after nocodazole washout (Fig. 3C and D). Recently, SAC signaling has been correlated with levels of unattached kinetochores. Thus, increasing the number of unattached kinetochores only slightly could prolong mitotic arrest (20–22). An increase of BUBR1 positive kinetochores in HCC366 cells exposed to paclitaxel and depleted of CASC1 suggest that CASC1 may be supporting mitotic slippage by promoting microtubule–kinetochore attachments and reducing SAC strength (Fig. 3E).

We further examined the impact of siCASC1 on microtubules in HBEC3KT alone, or following stable depletion of p53 in the absence (HBEC3KT-p53) and presence of oncogenic K-RAS (HBEC3KT-p53RASV12; ref. 23). Here, we found that CASC1 was essential for microtubule polymerization in all settings, but only essential for viability following depletion of p53, thus suggesting the tumor-specific dependency on CASC1 for viability can be driven by a loss of tumor suppressor activity (Supplementary Fig. S4D and S4E). Taken together, these findings suggest that combinatorial regimens that incorporate multiple antitubulin agents may enhance or restore sensitivity to NSCLC.

TRIM69 is a centrosomal and microtubule-associated protein essential for mitotic spindle formation

TRIM69 is a member of the TRIpartite Motif (TRIM) containing protein family, which contains a RING E3 ligase domain, B box, and coiled coil domains (24). Two isoforms exist, A and B, with only the TRIM69A isoform containing a RING domain (Supplementary Fig. S5A). Indicative that TRIM69A is an E3 ligase, we detected autoubiquitination of TRIM69A in an in vitro ubiquitination assay (Supplementary Fig. S5B). We found that TRIM69 depletion, similar to CASC1, activated cell death signaling alone or in the presence of paclitaxel in multiple sensitive and resistant NSCLC backgrounds, with the exception of H2887. This defect was not observed in normal HBEC3KT cells (Fig. 4A), thus implicating a dependency on TRIM69 for tumor cell survival. At the single cell level, we observed a number of mitotic defects, including (i) micronucleation (A549), (ii) paclitaxel-enhanced micronucleation (H1299), and (iii) multipolar spindles in the presence of paclitaxel (HCC366 and H1299; Fig. 4B and Supplementary Fig. S5C). Depletion of TRIM69 also increased BUBR1 occupancy at kinetochores in paclitaxel-treated HCC366 cells (Fig. 4C). Thus, we conclude that TRIM69 is essential for mitotic fidelity and may be essential for the proper attachment of microtubules to kinetochores.

Figure 4.

A and B, as in 3A and 3B; scale bars, 10 μm for A549 and H1299, 5 μm for HCC366. C, as in 3E. **, P < 0.05.

Figure 4.

A and B, as in 3A and 3B; scale bars, 10 μm for A549 and H1299, 5 μm for HCC366. C, as in 3E. **, P < 0.05.

Close modal

To further elucidate the mitotic function of TRIM69, we examined the localization of both isoforms of TRIM69 during the cell cycle. Examination of cells overexpressing myc-TRIM69A revealed localization to microtubules during interphase and centrosomes during mitosis (Fig. 5A). Overexpressed myc-TRIM69A also induced microtubule bundling, rendering microtubules resistant to nocodazole and leading to the generation of micronucleated cells (Fig. 5A and Supplementary Fig. S5D). The localization of myc-TRIM69A to microtubules and centrosomes was dependent upon an intact RING domain as mutation of canonical cysteines in the zinc finger of TRIM69A abrogated microtubule and centrosomal localization of the overexpressed construct, indicating that the localization of TRIM69A is dependent upon E3-ligase activity and/or interaction partners (Fig. 5A). Consistent with this finding, myc-TRIM69B, which lacks the RING domain, was dispersed throughout the cytoplasm during the cell cycle and did not display microtubule localization or induce alterations in microtubule stability (Fig. 5A and Supplementary Fig. S5D). These data suggest that TRIM69 may play a critical role in supporting mitotic fidelity.

Figure 5.

A, HeLa cells transfected with indicated cDNAs for 48 hours were fixed and immunostained. Scale bars, 5 μm. Right, quantitation of cells from A. Bars represent mean ± range for n = 2 independent experiments. B, immunostaining of H1299-myc-TRIM69A representing each stage in the cell cycle based on nuclear morphology. Scale bars, 5 μm. C, H1299 cells stably expressing myc-TRIM69A were transfected with indicated siRNA for 72 hours and immunostained with indicated antibodies. Plots represent relative fluorescence of myc-TRIM69A at centrosomes in >150 cells across three independent experiments. ***, P < 0.0001 by 2-tailed Mann–Whitney test. Scale bars, 5 μm. D, left, H1299 cells were immunostained with indicated antibodies following siRNA transfection for 72 hours. Scale bars, 5 μm. Middle, quantitation of cells from D. Bars represent mean ± range from two independent experiments. Right, bars represent mean mRNA expression levels of HAUS1 72 hours after transfection.

Figure 5.

A, HeLa cells transfected with indicated cDNAs for 48 hours were fixed and immunostained. Scale bars, 5 μm. Right, quantitation of cells from A. Bars represent mean ± range for n = 2 independent experiments. B, immunostaining of H1299-myc-TRIM69A representing each stage in the cell cycle based on nuclear morphology. Scale bars, 5 μm. C, H1299 cells stably expressing myc-TRIM69A were transfected with indicated siRNA for 72 hours and immunostained with indicated antibodies. Plots represent relative fluorescence of myc-TRIM69A at centrosomes in >150 cells across three independent experiments. ***, P < 0.0001 by 2-tailed Mann–Whitney test. Scale bars, 5 μm. D, left, H1299 cells were immunostained with indicated antibodies following siRNA transfection for 72 hours. Scale bars, 5 μm. Middle, quantitation of cells from D. Bars represent mean ± range from two independent experiments. Right, bars represent mean mRNA expression levels of HAUS1 72 hours after transfection.

Close modal

We next studied the dynamics of myc-TRIM69A localization during the cell cycle, using H1299 cells with stable expression of myc-TRIM69A. In this setting, we found that myc-TRIM69A is recruited to centrosomes during prometaphase where it remains through telophase (Fig. 5B and Supplementary Fig. S5E). myc-TRIM69A also associated with endogenous pericentrin in cells arrested in mitosis (Supplementary Fig. S5F). The centrosomal localization of TRIM69A was further bolstered by proteomics datasets that indicate interactions between TRIM69 and a number of mitotic proteins, including centrosomal and cytoskeletal components (Supplementary Fig. 5G; refs. 25–36). Indeed, depletion of 2 centrosomal proteins that putatively interact with TRIM69, MYPT1, and GNAI3, attenuate centrosomal localization of myc-TRIM69A during mitosis (Fig. 5C).

Tumor cells frequently amplify centrosomes, which can be induced by tumor suppressors and/or oncogenes. As multiple centrosomes can lead to multipolar mitosis, mitotic defects, and cell death, tumor cells depend upon mechanisms that promote centrosomal clustering for cell survival (37). Proteomic studies suggest that TRIM69 interacts with the 8-member Augmin complex (also known as HAUS), which is essential for centrosomal clustering (31, 37). To evaluate a functional interaction between TRIM69 and the HAUS complex, we depleted HAUS1 in H1299 cells in the presence and absence of myc-TRIM69A. We find that spindle multipolarity induced by inhibition of HAUS1 can be reversed by overexpression of myc-TRIM69A, indicating that TRIM69A can promote centrosomal clustering (Fig. 5D). Thus, inhibition of TRIM69A may enhance the vulnerability of tumor cells to paclitaxel by attenuating centrosome clustering, which may prolong mitotic arrest and allow for the reengagement of cell death signaling.

CASC1 and TRIM69 are required for in vivo tumor growth

We next determined whether CASC1 and TRIM69 were essential for growth of tumors in vivo. Specifically, to achieve stable depletion of CASC1 and TRIM69, we introduced shRNAs targeting GFP, CASC1, or TRIM69 into HCC366 cells (Fig. 6A). These cells were subcutaneously injected into immunodeficient mice, and we observed attenuation in growth of tumors expressing shTRIM69 and shCASC1 (Fig. 6B and C), indicating that the depletion of CASC1 or TRIM69 in vivo was sufficient to reduce mitotic fidelity, hematoxylin and eosin (H&E) staining on excised tumors revealed extensive micronucleation, and aberrant mitotic figures (Fig. 6D). Furthermore, TUNEL staining on tissue sections revealed enhanced cell death (Fig. 6E). Taken together, these findings indicate that CASC1 and TRIM69 suppression is sufficient to affect mitotic fidelity and viability in vivo.

Figure 6.

A, bars represent mean mRNA expression levels ± range of CASC1 and TRIM69 in HCC366 cells expressing indicated hairpins from two independent infections. B, tumor growth curves for mice harboring HCC366s expressing indicated shRNAs. shGFP n = 9, shCASC1 n = 9, and shTRIM69 n = 7. Mean ± SD. C, tumors expressing indicated short hairpins were excised 96 days following injection. Left, bars represent mean mass of tumors ± SEM; shGFP n = 7, shCASC1 n = 10, shTRIM69 n = 8. Right, representative images of excised tumors. D, tumors from B were sectioned and H&E stained. Left, quantitation of micronucleated cells in tumor sections. Six independent tumors were examined for each indicated shRNA. Right, representative images of H&E–stained tumor sections. Scale, 50 μm. White arrowheads, micronucleated cells. Blue arrowheads, aberrant mitotic cells. E, tumors from B were sectioned and TUNEL stained. Bars indicate mean TUNEL-positive cells ± SEM from 6 tumors for each shRNA condition. For all panels, **, P < 0.01 and ***, P < 0.0001 (unpaired 2-tailed Student t test).

Figure 6.

A, bars represent mean mRNA expression levels ± range of CASC1 and TRIM69 in HCC366 cells expressing indicated hairpins from two independent infections. B, tumor growth curves for mice harboring HCC366s expressing indicated shRNAs. shGFP n = 9, shCASC1 n = 9, and shTRIM69 n = 7. Mean ± SD. C, tumors expressing indicated short hairpins were excised 96 days following injection. Left, bars represent mean mass of tumors ± SEM; shGFP n = 7, shCASC1 n = 10, shTRIM69 n = 8. Right, representative images of excised tumors. D, tumors from B were sectioned and H&E stained. Left, quantitation of micronucleated cells in tumor sections. Six independent tumors were examined for each indicated shRNA. Right, representative images of H&E–stained tumor sections. Scale, 50 μm. White arrowheads, micronucleated cells. Blue arrowheads, aberrant mitotic cells. E, tumors from B were sectioned and TUNEL stained. Bars indicate mean TUNEL-positive cells ± SEM from 6 tumors for each shRNA condition. For all panels, **, P < 0.01 and ***, P < 0.0001 (unpaired 2-tailed Student t test).

Close modal

Although genomic studies have begun to isolate cohorts of patients that may be selectively sensitive to precision therapeutics, the majority of patients with NSCLC are administered standard chemotherapeutic regimens that are comprised of DNA damage agents and antimitotics. These regimens are noncurative and in 70% of cases fail to reduce tumor burden (2, 3). The underlying basis for the intrinsic resistance of NSCLC to standard chemotherapy is poorly understood. Here, we use a genome-wide RNAi approach to define the molecular components that deflect cell death in the presence of mitotic damage. This work resulted in a number of new principles about chemoresistance and mitosis in tumor cells. First, a common feature to uncouple tumor cell death from an insurmountable spindle defect is mitotic slippage. Second, inhibition of specific proteins can increase mitotic transit time >20%, which is sufficient to activate cell death. Third, CASC1 is a novel regulator of microtubule dynamics that is essential for tumor, but not normal cell survival. Fourth, TRIM69 is a novel mitotic centrosomal resident protein that supports centrosomal clustering. In contrast to a previous screen in H1155 sensitive cells, we did not return multiple components of the γTURC or kinetochore complexes, suggesting that perturbing spindle assembly in general is not sufficient to enhance sensitivity in NSCLC cells resistant to paclitaxel (12). These findings suggest a new therapeutic strategy for chemoresistant NSCLC: combining antimitotic agents, such as paclitaxel, with agents, such as small molecule inhibitors of the APC/C may enhance the potency of these first-line agents with limited impacts on normal tissues.

Many tumor cells exhibit a brief SAC-dependent arrest and mitotic escape following exposure to spindle poisons in vitro and in vivo (6, 38). The frequency of this behavior suggests that tumor evolution may generally select for variants that are capable of bypassing prolonged mitotic arrest. Oncogenic changes such as p53 loss, EGFR activation, or K-RAS mutation can induce mitotic spindle and centrosomal alterations, which could conceivably disrupt mitotic progression and lead to cell death in pre-tumorigenic lesions (39–41). Thus, those preneoplastic variants that bypass mitotic arrest could have a selective advantage, if they retain fecundity. A collateral consequence of this relationship would be an intrinsic resistance to antimitotic chemotherapeutics. Our work supports the hypothesis that re-engaging a prolonged mitotic arrest can enhance sensitivity to antimitotics (6, 9). We find that elevation of mitotic transit time by 20% to 50% is sufficient to switch a paclitaxel-resistant setting to a sensitive one. However, frank inhibition of the APC/C, through depletion of CDC20, is highly lethal in tumor cells and likely normal cells, given the requirement for CDC20 to activate the APC/C (9). Thus, single agent targeting of the APC/C could have a narrow or nonexistent therapeutic window. Our finding that ANAPC5 depletion can exhibit a synthetic lethal interaction with paclitaxel, suggests that tumor cells may have an enhanced sensitivity to APC/C inhibition that may be exploited by combining sublethal inhibition of APC/C with low-dose paclitaxel to produce a synthetic phenotype.

We also define a mitotic and microtubule role for the human CASC1 protein in tumor cells. Although this gene has been linked with murine lung tumorigenesis, any functional role in human tumorigenesis has not been demonstrated. We find that tumor cells require CASC1 for microtubule polymerization and inhibition of CASC1 potentially impacts microtubule availability for the alignment of chromosomes, thereby increasing the number of unattached kinetochores and prolonging the SAC. The heightened dependency on tubulin regulation in tumor cells could be driven by oncogenic stress that may increase centrosomal clustering defects and attachment errors, thus creating an environment in which mild disruptions to the microtubule network can wreak havoc on chromosome segregation in tumor cells. As a recent report suggest that mitosis is compromised in K-RAS backgrounds, the genetic link between CASC1 and K-RAS may allow for mitigation of transformation induced mitotic defects by increasing the expression of a microtubule regulatory protein in concert with an oncogene (40). Importantly, these findings suggest that one mechanism for prolonging the SAC in tumor cells may be to compromise microtubule function in parallel, but independent, manners by combining antitubulin agents.

TRIM69A is an E3 ligase with only limited characterized function, which we implicate as critical for mitotic fidelity. Specifically, TRIM69A seems to promote centrosomal clustering, which is an essential process to prevent multipolar spindles and decreased mitotic fidelity (37, 42, 43). Thus, interfering with centrosomal clustering may be a route to sensitize resistance cells to paclitaxel. Furthermore, the dependency of TRIM69 on MYPT1 and GNAI3 for centrosomal localization implicates TRIM69 in the regulation of phosphatase and receptor-independent G(ialpha) signaling during mitosis. MYPT1, which is localized to the spindle, kinetochores and centrosomes, is thought to regulate SAC checkpoint silencing though dephosphorylation of Aurora B and BUB1 (44–46). GNAI3 and other G(ialpha) subunits have been localized to the centrosome and depletion of GNAI3 induces cytokinesis defects, whereas overexpression increases mitotic transit time (47). Thus, TRIM69A activity may be essential for coordinating multiple molecular events that promote bipolar spindle formation and allow for rapid checkpoint silencing. Investigation into the role of TRIM69A regulation may reveal additional mechanisms governing spindle alignment and centrosome clustering.

No potential conflicts of interest were disclosed.

Conception and design: A.W. Whitehurst

Development of methodology: R. Sinnott, A.W. Whitehurst

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): R. Sinnott, L. Winters, B. Larson, D. Mytsa, P. Taus, K.M. Cappell, A.W. Whitehurst

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): R. Sinnott, D. Mytsa, A.W. Whitehurst

Writing, review, and or revision of the manuscript: R. Sinnott, P. Taus, A.W. Whitehurst

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): L. Winters, K.M. Cappell, A.W. Whitehurst

Study supervision: A.W. Whitehurst

The authors thank C. Ross for support with animal studies, S. Rogers for assistance with microtubule assays, and N. Sciaky for technical support with live-cell imaging.

This work was supported by Public Health Service grants CA154699 and CA128926 to A.W. Whitehurst from the National Cancer Institute and a research grant from Uniting against Lung Cancer. R. Sinnott and P. Taus were supported by general medicine training grant T32GM007040-37. K.M. Cappell was supported by the National Cancer Institute training grant CA071341-14. P. Taus was supported by general NCI training grant NIH-CA071341-14.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Jordan
MA
,
Kamath
K
. 
How do microtubule-targeted drugs work? An overview
.
Curr Cancer Drug Targets
2007
;
7
:
730
42
.
2.
Schiller
JH
,
Harrington
D
,
Belani
CP
,
Langer
C
,
Sandler
A
,
Krook
J
, et al
Comparison of four chemotherapy regimens for advanced non-small-cell lung cancer
.
N Engl J Med
2002
;
346
:
92
8
.
3.
Ichiki
M
,
Kawasaki
M
,
Takayama
K
,
Ninomiya
K
,
Kuba
M
,
Iwami
F
, et al
A multicenter phase II study of carboplatin and paclitaxel with a biweekly schedule in patients with advanced non-small-cell lung cancer: Kyushu thoracic oncology group trial
.
Cancer Chemother Pharmacol
2006
;
58
:
368
73
.
4.
Peters
JM
. 
The anaphase promoting complex/cyclosome: a machine designed to destroy
.
Nat Rev Mol Cell Biol
2006
;
7
:
644
56
.
5.
Yu
H
. 
Regulation of APC-Cdc20 by the spindle checkpoint
.
Curr Opin Cell Biol
2002
;
14
:
706
14
.
6.
Gascoigne
KE
,
Taylor
SS
. 
Cancer cells display profound intra- and interline variation following prolonged exposure to antimitotic drugs
.
Cancer Cell
2008
;
14
:
111
22
.
7.
Hornick
JE
,
Bader
JR
,
Tribble
EK
,
Trimble
K
,
Breunig
JS
,
Halpin
ES
, et al
Live-cell analysis of mitotic spindle formation in taxol-treated cells
.
Cell Motil Cytoskeleton
2008
;
65
:
595
613
.
8.
Rieder
CL
,
Maiato
H
. 
Stuck in division or passing through: what happens when cells cannot satisfy the spindle assembly checkpoint
.
Dev Cell
2004
;
7
:
637
51
.
9.
Huang
HC
,
Shi
J
,
Orth
JD
,
Mitchison
TJ
. 
Evidence that mitotic exit is a better cancer therapeutic target than spindle assembly
.
Cancer Cell
2009
;
16
:
347
58
.
10.
Harley
ME
,
Allan
LA
,
Sanderson
HS
,
Clarke
PR
. 
Phosphorylation of Mcl-1 by CDK1-cyclin B1 initiates its Cdc20-dependent destruction during mitotic arrest
.
EMBO J
2010
;
29
:
2407
20
.
11.
Wertz
IE
,
Kusam
S
,
Lam
C
,
Okamoto
T
,
Sandoval
W
,
Anderson
DJ
, et al
Sensitivity to antitubulin chemotherapeutics is regulated by MCL1 and FBW7
.
Nature
2011
;
471
:
110
4
.
12.
Whitehurst
AW
,
Bodemann
BO
,
Cardenas
J
,
Ferguson
D
,
Girard
L
,
Peyton
M
, et al
Synthetic lethal screen identification of chemosensitizer loci in cancer cells
.
Nature
2007
;
446
:
815
9
.
13.
Schreiber
A
,
Stengel
F
,
Zhang
Z
,
Enchev
RI
,
Kong
EH
,
Morris
EP
, et al
Structural basis for the subunit assembly of the anaphase-promoting complex
.
Nature
2011
;
470
:
227
32
.
14.
Zeng
X
,
Sigoillot
F
,
Gaur
S
,
Choi
S
,
Pfaff
KL
,
Oh
DC
, et al
Pharmacologic inhibition of the anaphase-promoting complex induces a spindle checkpoint-dependent mitotic arrest in the absence of spindle damage
.
Cancer Cell
2010
;
18
:
382
95
.
15.
Liu
P
,
Wang
Y
,
Vikis
H
,
Maciag
A
,
Wang
D
,
Lu
Y
, et al
Candidate lung tumor susceptibility genes identified through whole-genome association analyses in inbred mice
.
Nat Genet
2006
;
38
:
888
95
.
16.
Zhang
Z
,
Futamura
M
,
Vikis
HG
,
Wang
M
,
Li
J
,
Wang
Y
, et al
Positional cloning of the major quantitative trait locus underlying lung tumor susceptibility in mice
.
Proc Natl Acad Sci U S A
2003
;
100
:
12642
7
.
17.
Galbiati
F
,
Pettinicchio
A
,
Dragani
TA
,
Manenti
G
. 
Allelic effects of mouse Pas1 candidate genes in human lung cancer cell lines
.
Cancer Lett
2006
;
244
:
176
81
.
18.
Cerami
E
,
Gao
J
,
Dogrusoz
U
,
Gross
BE
,
Sumer
SO
,
Aksoy
BA
, et al
The cBio cancer genomics portal: an open platform for exploring multidimensional cancer genomics data
.
Cancer Discov
2012
;
2
:
401
4
.
19.
Wirschell
M
,
Yang
C
,
Yang
P
,
Fox
L
,
Yanagisawa
HA
,
Kamiya
R
, et al
IC97 is a novel intermediate chain of I1 dynein that interacts with tubulin and regulates interdoublet sliding
.
Mol Biol Cell
2009
;
20
:
3044
54
.
20.
Collin
P
,
Nashchekina
O
,
Walker
R
,
Pines
J
. 
The spindle assembly checkpoint works like a rheostat rather than a toggle switch
.
Nat Cell Biol
2013
;
15
:
1378
85
.
21.
Dick
AE
,
Gerlich
DW
. 
Kinetic framework of spindle assembly checkpoint signalling
.
Nat Cell Biol
2013
;
15
:
1370
7
.
22.
Heinrich
S
,
Geissen
EM
,
Kamenz
J
,
Trautmann
S
,
Widmer
C
,
Drewe
P
, et al
Determinants of robustness in spindle assembly checkpoint signalling
.
Nat Cell Biol
2013
;
15
:
1328
39
.
23.
Sato
M
,
Vaughan
MB
,
Girard
L
,
Peyton
M
,
Lee
W
,
Shames
DS
, et al
Multiple oncogenic changes (K-RAS(V12), p53 knockdown, mutant EGFRs, p16 bypass, telomerase) are not sufficient to confer a full malignant phenotype on human bronchial epithelial cells
.
Cancer Res
2006
;
66
:
2116
28
.
24.
Hatakeyama
S
. 
TRIM proteins and cancer
.
Nat Rev Cancer
2011
;
11
:
792
804
.
25.
Zhu
M
,
Settele
F
,
Kotak
S
,
Sanchez-Pulido
L
,
Ehret
L
,
Ponting
CP
, et al
MISP is a novel Plk1 substrate required for proper spindle orientation and mitotic progression
.
J Cell Biol
2013
;
200
:
773
87
.
26.
Yamashiro
S
,
Yamakita
Y
,
Totsukawa
G
,
Goto
H
,
Kaibuchi
K
,
Ito
M
, et al
Myosin phosphatase-targeting subunit 1 regulates mitosis by antagonizing polo-like kinase 1
.
Dev Cell
2008
;
14
:
787
97
.
27.
Wang
W
,
Zhang
W
,
Han
Y
,
Chen
J
,
Wang
Y
,
Zhang
Z
, et al
NELIN, a new F-actin associated protein, stimulates HeLa cell migration and adhesion
.
Biochem Biophys Res Commun
2005
;
330
:
1127
31
.
28.
Snyers
L
,
Thines-Sempoux
D
,
Prohaska
R
. 
Colocalization of stomatin (band 7.2b) and actin microfilaments in UAC epithelial cells
.
Eur J Cell Biol
1997
;
73
:
281
5
.
29.
Mulder
J
,
Ariaens
A
,
van den Boomen
D
,
Moolenaar
WH
. 
p116Rip targets myosin phosphatase to the actin cytoskeleton and is essential for RhoA/ROCK-regulated neuritogenesis
.
Mol Biol Cell
2004
;
15
:
5516
27
.
30.
Levi
M
,
Maro
B
,
Shalgi
R
. 
The conformation and activation of Fyn kinase in the oocyte determine its localisation to the spindle poles and cleavage furrow
.
Reprod Fertil Dev
2011
;
23
:
846
57
.
31.
Lawo
S
,
Bashkurov
M
,
Mullin
M
,
Ferreria
MG
,
Kittler
R
,
Habermann
B
, et al
HAUS, the 8-subunit human Augmin complex, regulates centrosome and spindle integrity
.
Curr Biol
2009
;
19
:
816
26
.
32.
Kao
SC
,
Chen
CY
,
Wang
SL
,
Yang
JJ
,
Hung
WC
,
Chen
YC
, et al
Identification of phostensin, a PP1 F-actin cytoskeleton targeting subunit
.
Biochem Biophys Res Commun
2007
;
356
:
594
8
.
33.
Hauge
H
,
Patzke
S
,
Aasheim
HC
. 
Characterization of the FAM110 gene family
.
Genomics
2007
;
90
:
14
27
.
34.
Hasegawa
H
,
Hyodo
T
,
Asano
E
,
Ito
S
,
Maeda
M
,
Kuribayashi
H
, et al
The role of PLK1-phosphorylated SVIL in myosin II activation and cytokinetic furrowing
.
J Cell Sci
2013
;
126
(
Pt 16
):
3627
37
.
35.
Grimm-Gunter
EM
,
Milbrandt
M
,
Merkl
B
,
Paulsson
M
,
Plomann
M
. 
PACSIN proteins bind tubulin and promote microtubule assembly
.
Exp Cell Res
2008
;
314
:
1991
2003
.
36.
Fischer
RS
,
Yarmola
EG
,
Weber
KL
,
Speicher
KD
,
Speicher
DW
,
Bubb
MR
, et al
Tropomodulin 3 binds to actin monomers
.
J Biol Chem
2006
;
281
:
36454
65
.
37.
Leber
B
,
Maier
B
,
Fuchs
F
,
Chi
J
,
Riffel
P
,
Anderhub
S
, et al
Proteins required for centrosome clustering in cancer cells
.
Sci Transl Med
2010
;
2
:
33ra8
.
38.
Orth
JD
,
Kohler
RH
,
Foijer
F
,
Sorger
PK
,
Weissleder
R
,
Mitchison
TJ
. 
Analysis of mitosis and antimitotic drug responses in tumors by in vivo microscopy and single-cell pharmacodynamics
.
Cancer Res
2011
;
71
:
4608
16
.
39.
Fukasawa
K
,
Choi
T
,
Kuriyama
R
,
Rulong
S
,
Vande Woude
GF
. 
Abnormal centrosome amplification in the absence of p53
.
Science
1996
;
271
:
1744
7
.
40.
Luo
J
,
Emanuele
MJ
,
Li
D
,
Creighton
CJ
,
Schlabach
MR
,
Westbrook
TF
, et al
A genome-wide RNAi screen identifies multiple synthetic lethal interactions with the Ras oncogene
.
Cell
2009
;
137
:
835
48
.
41.
Mardin
BR
,
Isokane
M
,
Cosenza
MR
,
Kramer
A
,
Ellenberg
J
,
Fry
AM
, et al
EGF-induced centrosome separation promotes mitotic progression and cell survival
.
Dev Cell
2013
;
25
:
229
40
.
42.
Quintyne
NJ
,
Reing
JE
,
Hoffelder
DR
,
Gollin
SM
,
Saunders
WS
. 
Spindle multipolarity is prevented by centrosomal clustering
.
Science
2005
;
307
:
127
9
.
43.
Ganem
NJ
,
Godinho
SA
,
Pellman
D
. 
A mechanism linking extra centrosomes to chromosomal instability
.
Nature
2009
;
460
:
278
82
.
44.
London
N
,
Ceto
S
,
Ranish
JA
,
Biggins
S
. 
Phosphoregulation of Spc105 by Mps1 and PP1 regulates Bub1 localization to kinetochores
.
Curr Biol
2012
;
22
:
900
6
.
45.
Meadows
JC
,
Shepperd
LA
,
Vanoosthuyse
V
,
Lancaster
TC
,
Sochaj
AM
,
Buttrick
GJ
, et al
Spindle checkpoint silencing requires association of PP1 to both Spc7 and kinesin-8 motors
.
Dev Cell
2011
;
20
:
739
50
.
46.
Posch
M
,
Khoudoli
GA
,
Swift
S
,
King
EM
,
Deluca
JG
,
Swedlow
JR
. 
Sds22 regulates aurora B activity and microtubule-kinetochore interactions at mitosis
.
J Cell Biol
2010
;
191
:
61
74
.
47.
Cho
H
,
Kehrl
JH
. 
Localization of Gi alpha proteins in the centrosomes and at the midbody: implication for their role in cell division
.
J Cell Biol
2007
;
178
:
245
55
.

Supplementary data