Inhibition of clathrin by pitstop 2 activates the spindle assembly checkpoint and induces cell death in dividing HeLa cancer cells
© Smith et al.; licensee BioMed Central Ltd. 2013
Received: 3 September 2012
Accepted: 3 January 2013
Published: 17 January 2013
During metaphase clathrin stabilises the mitotic spindle kinetochore(K)-fibres. Many anti-mitotic compounds target microtubule dynamics. Pitstop 2™ is the first small molecule inhibitor of clathrin terminal domain and inhibits clathrin-mediated endocytosis. We investigated its effects on a second function for clathrin in mitosis.
Pitstop 2 did not impair clathrin recruitment to the spindle but disrupted its function once stationed there. Pitstop 2 trapped HeLa cells in metaphase through loss of mitotic spindle integrity and activation of the spindle assembly checkpoint, phenocopying clathrin depletion and aurora A kinase inhibition.
Pitstop 2 is therefore a new tool for investigating clathrin spindle dynamics. Pitstop 2 reduced viability in dividing HeLa cells, without affecting dividing non-cancerous NIH3T3 cells, suggesting that clathrin is a possible novel anti-mitotic drug target.
KeywordsClathrin Pitstop Spindle assembly checkpoint Metaphase Cell death Cancer
Cell division (mitosis) results in equal segregation of duplicated chromosomes into two independent daughter cells. Premature chromosome segregation (metaphase-anaphase transition) results in aneuploidy, a hallmark of many human cancers . This adverse situation is avoided by activation of the spindle assembly checkpoint (SAC). This is a signalling pathway consisting of a number of protein complexes that monitor proper mitotic spindle assembly, delaying anaphase onset until all chromosomes are stably attached to kinetochores (KTs) . KTs are a large protein assembly around the centromere of chromosomes that mediates the attachment of chromosomes to the spindle microtubules (MTs) so they can complete segregation. Both SAC activators and inhibitors are well-known targets for several chemotherapeutic agents. Drugs that target MTs such as the taxanes and vinca alkaloids are extensively used for cancer treatment . They stop APC/C-mediated proteolysis and block anaphase onset in a SAC-dependent manner. Despite success in the clinic, drug resistance and toxicity have limited their effectiveness . Interference with KT assembly, impairment of MT motors (e.g. dynein) and interference with MT dynamics also activate the SAC . Thus, a new class of chemotherapeutics are being developed that specifically target key mitotic proteins to either activate or inhibit the SAC, such as aurora A kinase and kinesin spindle protein, respectively . These targeted inhibitors prevent proliferation of most tumour cells in vitro and reduce tumour volume in vivo by inhibiting growth and/or triggering cell death following SAC activation/ inhibition [3, 4]. Many are in cancer clinical trials, such as the aurora A protein kinase inhibitor MLN8054 . They are expected to have a more favourable therapeutic window than current chemotherapeutic agents , as they would spare non-dividing cells. The anti-cancer efficacy of these mitotic inhibitors is dependent on their ability to induce apoptosis following mitotic insult. However, they do not always result in cell death . Thus, there is scope for identification of new anti-mitotic targets and the development of new anti-cancer compounds with greater efficacy.
Clathrin is a protein complex of three identical 190 kDa clathrin heavy chains (CHCs) arranged in a trimer (called a triskelion) of three “legs” connected by their C-termini at a central vertex [7, 8]. A globular N-terminal β-propeller domain (TD) is found at the end of each clathrin leg (i.e. at the N-terminus of the protein sequence). Clathrin can interact with multiple adapter proteins like amphiphysin via its TD . Clathrin is best known for its roles in endocytosis and TGN/ endo-lysosmal sorting, however, in recent years it has been assigned another non-trafficking function in mitosis. For clathrin-mediated endocytosis (CME), clathrin cycles between the cytoplasmic triskelion and a polymerised coat on vesicles or membranes. During mitosis, clathrin localizes to the mitotic spindle [10–12] where it is involved in organizing and stabilizing spindle MTs [11–13]. It dissociates from MTs during telophase, as the Golgi reforms to participate in its reassembly . The role of clathrin at the mitotic spindle is dependent on both its TD  and ability to trimerise as well as its interaction with TACC3 (transforming acidic coiled-coil-containing protein 3) . Aurora A kinase phosphorylates and localises TACC3 to the spindle [15, 16]. Phospho-TACC3 recruits clathrin and ch-TOG to the spindle MTs  where they bridge together two or three kinetochore fibres (K-fibres) to aid chromosome congression  with TACC3 directly interacting with MTs [17, 18]. Depletion of clathrin by siRNA causes defective chromosome congression to the metaphase plate and persistent SAC activation [11, 19–21]. This is analogous to the effect of aurora A inhibitors which are also SAC activators [22–26]. Aurora A inhibitors also block clathrin recruitment to the spindle by blocking TACC3 recruitment . Thus, it is possible that SAC activation and the anti-cancer properties of aurora A inhibitors may be partly due to blocking clathrin function at the mitotic spindle.
Clathrin requires its TD to associate with the mitotic spindle , although the protein(s) mediating its recruitment remains unclear. Preventing this interaction leads to defective congression of chromosomes to the metaphase plate and persistent activation of the SAC. We have recently developed the first small molecule inhibitors of clathrin, pitstop 1 and pitstop 2, which target the TD . These two chemically unrelated small molecules inhibit the association of CHC-TD with clathrin box motif-containing endocytic proteins such as amphiphysin and AP180/CALM, with no affect on three other protein-protein interactions or on dynamin GTPase activity, demonstrating their relative specificity. Pitstop 2 is a potent inhibitor of transferrin (Tfn) uptake in cells and is reversible, with CME being fully restored after a 30 min drug washout, while pitstop 1 is not readily cell permeable. Here, we investigated the effect of pitstop 2 on mitosis to determine if its ability to block the TD function in CME is also true for its second function in the mitotic spindle. We assessed if it possesses anti-mitotic and anti-cancer properties analogous to other SAC activating compounds. We report that pitstop 2 induces mitotic phenotypes consistent with inhibition of clathrin and is anti-proliferative and cytotoxic in a dividing cancer cell, but not a dividing non-cancer cell. Our findings provide strong support to the specificity of pitstop 2 in targeting the TD and indicate that it is a new member of the SAC activating class of anti-mitotic compounds. The use of pitstop 2 has revealed that clathrin has the potential to be a new therapeutic target.
Clathrin, dynamin and aurora kinase inhibitors
Dynole 34–2 and pitstop 2 (in house synthesis) were prepared as 30 mM stock solutions in 100% DMSO and stored at −20°C. Pitstop 2 was purified to >98% purity (HPLC analysis at 254 nm) and NMR 400 MHz H spectrum in DMSO shows <1% of any other impurity; NMR 100 MHz C spectrum in DMSO shows <1% of any other impurity; TLC analysis using a solvent ratio of 9:1 DCM/MeOH shows a single spot visualised by UV. MLN8237 (Life Research Pty Ltd) was prepared as a 3 mM stock in 100% DMSO and stored at −20°C. Drugs were diluted directly into RPMI 1640 medium supplemented with 10% foetal bovine serum (FBS) and 5% penicillin/streptomycin (P/S). Pitstop 1, Pitstop 2, and Pitstop trademarks of Freie Universität Berlin, Children’s Medical Research Institute and Newcastle Innovation Ltd; Dynole 34-2, Dynole, Dyngo–4a and Dyngo are trademarks of Children’s Medical Research Institute and Newcastle Innovation Ltd are they are available from Abcam Biochemicals® (Cambridge, UK).
Cell culture, cell synchronization and drug treatment
HeLa cells were maintained in RPMI 1640 medium supplemented with 10% FBS and 5% P/S and grown at 37°C in a humidified 5% CO2 atmosphere. For mitotic synchronization, cells were synchronized at the G2/M boundary by treatment with the selective cdk1 small-molecule inhibitor, RO–3306 (9 μM) for 18 h as previously described [29, 30]. Cells were allowed to progress through mitosis following RO-3306 wash-out. Immediately following RO-3306 removal (i.e. release from the G2/M boundary), cells were treated with pitstop 2, dynole 34–2, MLN8237, drug-free medium or 0.1% DMSO vehicle.
All specific siRNA oligonucleotides were synthesised by Invitrogen. Sequences of the siRNAs targeting the indicated proteins are as follows: clathrin heavy chain: 5’-GCAATGAGCTGTTTGAAGA-3’ and epsin 5’-GGAAGACGCCGGAGTCATT -3’ [12, 31]. Luc siRNA, 5’-CGUACGCGGAAUACUUCGATT -3’, was used as a negative control.
For siRNA analysis, cells were seeded at 50-60% confluence (6–well plate-1 × 105 cells per well; 12–well plate-0.2 × 105 cells per well) and transfected with 1000 pmol of siRNA (per 10 cm dish for immunoblotting), or with 200 pmol (per well of a 6–well plate for immunofluorescence microscopy experiments) or 100 pmol (per well of a 12-well plate for immunofluorescence microscopy experiments), using Lipofectamine 2000 (Invitrogen) according to the manufacturer’s instructions.
Cell lysates were prepared as described previously . In brief, cells were collected by centrifugation, washed with PBS, then resuspended in ice-cold lysis buffer (20 mM Tris-HCL (pH 7.4), 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1% Triton X–100 and EDTA-free Complete protease inhibitor cocktail (Roche)) for 30 mins. The supernatant (cell lysate) was collected following centrifugation at 13,000 rpm for 30 min at 4°C. Cell lysates (50 μg) were fractionated by SDS-PAGE for immunoblot analysis. Antibodies targeting the following proteins were used: cleaved PARP (Cat No. 9542S, Cell Signalling) and actin (A3854, Sigma). Antibody bound to the indicated protein was detected by incubation with a horseradish peroxidase-conjugated secondary antibody (Sigma). Blotted proteins were visualised using the ECL detection system (Pierce).
Cells were fixed in ice-cold 100% methanol and immunostaining was carried out as described previously  using the following antibodies: anti-γ-tubulin (GTU88; Sigma-Aldrich), and anti-α-tubulin (clone DM1A, Sigma-Aldrich), anti–Clathrin heavy chain (BD Biosciences), anti-Centrin 2 (a gift from Eric Nigg) anti-tubulin (Pan, Cytoskeleton) anti-MAD2 (Covance), anti-phospho-BubR1S676 (a gift from Sabine Elowe ), anti-HURP (Abcam). Cells were viewed and scored with a fluorescence microscope (Olympus BX51). Fluorescent images were captured and processed using an Olympus IX80 inverted microscope using 60× or 100× oil immersion lenses and Metamorph software. Images were deconvolved using AutoQuant X2 (AutoQuant Imaging, Watervliet, NY). Spindle width and cell width was determined using the line measurement tool in Metamorph software.
MT regrowth after cold depolymerization
Asynchronously growing or metaphase synchronized cells were placed on ice for 30 min to depolymerize all MTs. Cells were then incubated at 37°C for 1, 5 or 30 min to allow the MTs to regrow. Cells were then fixed with ice cold methanol, stained for α-tubulin and γ-tubulin and fluorescent images were obtained as described above. The extent of spindle MT regrowth was assessed by determining the length of the longest microtubule using deconvolved images and the line measurement tool in Metamorph software.
Lactate dehydrogenase (LDH) cytotoxicity assay
Cytotoxicity was assayed by determination of lactate dehydrogenase (LDH) activity. HeLa cells (4000 cells per well) were seeded in 96 well plates. Asynchronously growing and G2/M-synchronized cells were treated in triplicate with pitstop 2, dynole 34–2 or MLN8237 at the indicated concentration for 20. The supernatant (50 μl) was added to 150 μl of LDH assay reagent (Sigma-Aldrich) and the reaction was allowed to developed for 20 min. Absorbance was measured at 490 nm and 690 nm (plate background absorbance). Values were normalised to drug/media background value and toxicity was calculated as a percentage of a control cell lysed with 20% Triton–X–100.
Trypan blue exclusion assay
HeLa were seeded in 6-well plates (0.5 × 105 cells per well). On day 0 (24 h after seeding), cells in duplicate were synchronized at the G2/M boundary by treatment with RO–3306. Immediately upon mitotic entry (RO–3306 release), cells were treated in the presence or absence of pitstop 2 or dynole 34–2, at a concentration of 1, 3, 10, 30, or 100 μM or with MLN8237 at 0.03, 0.3, or 1 μM. After 24 h, the total cell number and viability were measured using a Vi-CELL XR cell viability analyser as previously described . Where indicated, cells were treated a second time with the same compound for an additional 24 h and then analysed.
NIH3T3 cells were seeded in 6-well plates (0.5 × 105 cells per well). On day 0 (12–16 h after seeding), the media was replaced with low serum DMEM (0.5% FCS) to synchronize the cells at the G1/S boundary. After 24 h, the low serum media was replaced with DMEM containing 10% FCS to allow cells to synchronously enter the cell cycle. After 10 h, when cells had begun to enter mitosis, cells were treated and scored for viability as described above for HeLa cells.
Quantitative high-throughput receptor-mediated endocytosis (RME) assays were performed by an automated process using Texas Red-Transferrin (T×R-Tfn) as previously described [35, 36] in HeLa cells pre-treated with increasing concentrations of pitstop 2 and dynole 34–2 for 30 min. In brief, HeLa cells were grown in fibronectin-coated (5 μg/mL) 96–well plates. The cells were serum-starved overnight (16 h) in DMEM minus FCS then incubated with dynole 34-2, pitstop2 or vehicle for 30 min prior to addition of 4 μg/mL of Tf-×R for 8 min at 37°C. Cell surface-bound Tfn was removed by an ice-cold acid wash (0.2 M acetic acid + 0.5 M NaCl, pH 2.8) for 10 min then rinsed with ice-cold PBS for 5 min. Cells were immediately fixed with 4% PFA for 10 min at 37°C. Nuclei were stained using DAPI. Quantitative analysis of the inhibition of T×R-Tfn endocytosis in HeLa cells was performed on large numbers of cells by an automated acquisition and analysis system (Image Xpress Micro, Molecular Devices, Sunnyvale, CA). Nine images were collected from each well, averaging 40–50 cells per image. Average integrated intensity of the Tfn signal/cell was calculated using the IXM software and expressed as a percentage of DMSO-vehicle treated control cells. IC50 values were calculated using Prism 5 (GraphPad Software Inc.) and data are expressed as mean ± 95% confidence intervals (CI) for triplicates and ~1200 cells.
Pitstop 2 inhibits the mitotic spindle
We next asked whether the reduced number of mitotic spindle fibres observed in pitstop 2–treated metaphase cells was due to disruption of K-fibres. CHC stabilizes the mitotic spindle by specifically bridging K-fibres [11, 17, 18]. As would be predicted from this, we found that metaphase cells depleted of CHC also have abnormal staining of the K-fibre marker, hepatoma upregulated protein (HURP; Figure 2C and D). Pitstop disrupted HURP staining in 39.2 ± 4.6% of metaphase cells (Figure 2C and E). MLN8237 and epsin depletion also disrupted HURP staining (Figure 2C-E), while dynole 34–2 had no effect (Figure 2C and E). Therefore pitstop 2 disrupts mitotic spindle organisation by disrupting K-fibres, resulting in impaired chromosome congression to the metaphase plate.
The mitotic effects induced by pitstop 2 cannot be attributed to a block in endocytosis as both pitstop 2 and dynole 34–2 blocked uptake of transferrin in a dose-dependent manner (Additional file 4: Figure S4), as previously reported [28, 35]. However only pitstop 2 disrupted the mitotic spindle and chromosome alignment (Figure 1 and 2). This is consistent with the mitotic role of CHC being known to be independent of its endocytic function .
Pitstop 2 induces multipolar spindles and loss of centriole cohesion
Pitstop does not impair clathrin-dependent mitotic spindle regrowth
Pitstop 2 activates the spindle assembly checkpoint
Pitstop 2 induces apoptosis and inhibits cell growth in dividing cancer cells
We next sought to determine if pitstop 2 also possessed anti-proliferative properties by scoring the total viable cell number using a trypan blue exclusion assay. G2/M synchronized HeLa cells were treated for 24 h immediately following release from the G2/M block. The total number of viable HeLa cells following pitstop 2, MLN8237 or dynole 34–2 treatment was markedly reduced in a dose-dependent manner (Figure 6B). At high concentrations of pitstop 2 (30 μM) the total viable cell number was less than the number of cells scored immediately prior to administration of the compound (dashed line). This indicates that pitstop 2 not only inhibits cell proliferation but also causes cell death. This was also evident following MLN8237 and dynole 34–2 treatment (Figure 6B). Taken together this indicates that pitstop 2 blocks proliferation and causes a much greater rate of cell death in dividing cells than non-dividing cells.
Clathrin plays a non-endocytic function during mitosis that is downstream of the Aurora A kinase . Inhibitors of Aurora A are currently being assessed in human clinical trials of cancer with promising success . Here, we used the first clathrin inhibitor, pitstop 2, to illustrate that clathrin is also a valid target for the development of inhibitors that (i) are useful molecular tools to assess clathrin function and (ii) have the potential to be anti-cancer agents. Pitstop 2 induced mitotic phenotypes consistent with inhibition of clathrin, which included an increase in the mitotic index and width of the metaphase plate, loss of K-fibres and mitotic spindle integrity, chromosome mis-alignment and activation of the SAC. In an analogous manner to the Aurora A inhibitor, MLN8237, pitstop 2 inhibited cell proliferation and induced cell death exclusively in dividing cancer cells. Non-tumourigenic fibroblasts were not affected by this compound. Thus, our findings suggest that the clathrin TD inhibitor, pitstop 2, possesses anti-mitotic and anti-cancer properties consistent with other SAC activating compounds.
We confirm the mitotic role of clathrin in maintaining integrity of the mitotic spindle by stabilizing K-fibres. However, we reveal an additional mitotic role for clathrin in maintaining spindle pole integrity by participating in centriole cohesion. Depletion of CHC and pitstop 2 treatments caused an increase in the number of multipolar spindles whereby the spindle poles frequently contained only one centriole. Cyclin G-associated kinase (GAK) is a binding partner of CHC and required for CME . Like CHC, it has recently been reported to also function during metaphase in an endocytic-independent manner for correct chromosome segregation. Like CHC, GAK localizes to the mitotic spindle and GAK-depleted cells phenocopy the mitotic defects induced by depletion of CHC . One of the mitotic phenotypes observed in GAK-depleted cells is the presence of multipolar spindles that contain only one centriole at the spindle pole. It was proposed that this phenotype was due to microtubule forces, as mis-aligned chromosomes would generate pushing and pulling forces on the mitotic spindle resulting in fragmentation of the spindle poles. CHC is lost from the mitotic spindle in GAK-depleted cells  and thus a similar hypothesis could also explain the loss of centriole cohesion observed in CHC-depleted and pitstop 2–treated cells. Therefore CHC appears to play an indirect role in maintaining spindle pole integrity and centriole cohesion via its role in bridging K-fibres for correct chromosome congression and segregation.
The mitotic spindle localization of clathrin was not perturbed by pitstop 2. This was surprising given that pitstop 2 binds the TD of CHC , which is within the N-terminal region of clathrin known to be required for its mitotic spindle localization [11, 21]. The crystal structure of the CHC-TD (residues 1–364) reveals a classical seven-bladed β-propeller that contains a clathrin-box LϕXϕ [DE] (where ϕ is a bulky hydrophobic amino acid, X is any and brackets enclose alternatives) and W-box motifs for association with binding partners [47, 48], however some proteins may bind to other blades of the TD such that there may be up to 4 separate protein interaction sites in the TD . Pitstop 2 docks into the cleft comprising the clathrin-box between blades 1 and 2 of the β-propellor . Clathrin does not bind MTs directly, suggesting that clathrin is tethered to the mitotic spindle via other interactions that may involve other parts of clathrin such as the W-box. In support of this idea, phospho-TACC3 binds CHC via its linker domain and first CHC repeat and does not appear to require the TD of clathrin . B-Myb  and GAK  have been implicated in recruiting clathrin to the mitotic spindle and thus represent potential W-box binding partners during mitosis. Clathrin is thought to recruit TACC3 to the mitotic spindle via interactions that do not dependent on the TD . Consistent with this idea, we show that in contrast to depletion of CHC, pitstop 2 does not affect TACC3 recruitment to the mitotic spindle. Nor does it impair microtubule regrowth and initiation of mitotic spindle formation, which is in contrast to that caused by depletion of CHC. Thus, pitstop 2 does not induce an aberrant mitotic phenotype by blocking recruitment of clathrin to the mitotic spindle but rather appears to inhibit clathrin function once it is stationed at the spindle.
Anti-mitotic compounds that activate the SAC, such as MLN8237, are being developed as a new class of anti-cancer agents due to their ability to prolong metaphase arrest and subsequently inhibit cell proliferation and induce cell death in dividing cancer cells [3, 4]. As a result, many of these targeted inhibitors reduce tumour volume in vivo and are being assessed in cancer clinical trials . We show that pitstop 2 possesses anti-cancer properties since it phenocopies SAC activating compounds like MLN8237 or CHC knockdown. Aurora A mediates its function through many other proteins in addition to clathrin and therefore Aurora A inhibitors induce a plethora of phenotypes . This could potentially result in unwanted side effects and reduced efficacy for cancer patients. Therefore, we predict that clathrin TD inhibitors may be more targeted, raising the opportunity for them to be potentially more efficacious compounds for the treatment of cancer. Pitstop 2 was not as potent as MLN8237 at inducing aberrant mitotic phenotypes, inhibiting cell proliferation and inducing cell death, and new generation analogues are required. Our findings provide proof-of-concept that clathrin is a valid target for the development of small molecule inhibitors that can be exploited as new strategies to design anti-cancer therapeutics. Pitstop 2 therefore represents a new lead compound amenable to drug development.
We wish to thank Swetha Perera, Ngoc Chau, Ainslie Whiting and Scott L Page for their technical assistance. We also thank Erich Nigg and Sabine Elowe for the centrin 2 and phospho-BubR1S676 antibodies, respectively. This work was supported by grants from the National Health and Medical Research Council (NH&MRC) of Australia (MC & PJR), an NH&MRC Career Development Award Fellowship (MC), and by the German funding agency DFG (SFB765/B4, FOR806-HA2686/3-2 to VH).
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.