SEARCH

SEARCH BY CITATION

Keywords:

  • dichlorvos;
  • pesticide;
  • mitosis;
  • chromosome instability;
  • chromosome segregation;
  • monopolar spindle

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. AUTHOR CONTRIBUTIONS
  9. REFERENCES
  10. Supporting Information

The molecular mechanism(s) mediating long-term adverse effects of dichlorvos, a widely used insecticide, are still unclear. Our work uncovered a new cellular effect of dichlorvos in cultured human cells, i.e. its capacity to induce extremely aberrant mitotic spindles with monopolar microtubule arrays that were associated with hypercondensed chromosomes and pyknotic chromatin masses. Monopolar spindles produced by dichlorvos treatment were characterized by the delocalization of the depolymerizing kinesin Kif2a from spindle poles. Dichlorvos-induced spindle monopolarity could be reversed by promoting microtubule stabilization through chemical treatment or by inhibiting the depolymerizing function of the kinesin MCAK at kinetochores. These findings demonstrate that dichlorvos inhibits the depolymerizing activity of Kif2a at centrosomes and thereby disrupts the balance of opposing centrosomal and kinetochore forces controlling spindle bipolarity during prometaphase. Dichlorvos-induced defects in spindle bipolarity may be responsible for the previously reported induction of aneuploidy by this chemical. Collectively, these results indicate that environmental chemicals, such as dichlorvos, may promote chromosome instability by interfering with the cell division machinery. Environ. Mol. Mutagen. 54:250–260, 2013. © 2013 Wiley Periodicals, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. AUTHOR CONTRIBUTIONS
  9. REFERENCES
  10. Supporting Information

Pesticides and herbicides are extensively used to increase crop productivity both in industrial agriculture and in developing countries. However, their indiscriminate use has aroused a great concern about their adverse health effects in both target as well as nontarget species and for their environmental impact. Dichlorvos or DDVP (2, 2-dichlorovinyl dimethyl phosphate) is a common organophosphate insecticide with anti-cholinesterase activity. From the early 1960s, DDVP has been widely used to control flies, caterpillars and other insects for domestic purposes and to keep livestock, tobacco, and greenhouse-grown food crops free of pests. Currently, the use of DDVP in Western countries is limited on the basis of concerns about its possible carcinogenic effects and the availability of less toxic pesticides; however, its use in most developing countries is still extensive and unrestricted. Importantly, organophosphate pesticide self-poisoning is an important clinical problem in the developing world, and kills an estimated 200,000 people every year [Eddleston et al., 2008]. Moreover, human exposure to DDVP is also expected by the medical use of trichlorfon, a drug used in Alzheimer's disease treatment, which spontaneously metabolizes to DDVP [Ringman and Cummings, 1999].

Although substantial information is available regarding the environmental and ecological impact of DDVP [IARC, 1991], its potential harmful effects in humans are still unclear. Acute exposure to this pesticide results in neurological poisoning [Marrs, 1993], probably due to the inhibition of acetylcholinesterase activity [Qujeq et al., 2012], and poisoning by this chemical may lead to convulsions and coma [Marrs, 1993]. However, the adverse consequences of chronic exposure to DDVP and the mechanism(s) mediating long-term effects of this chemical are still not clearly defined. The International Agency for Research on Cancer (IARC) has classified DDVP as a possible (group 2B) human carcinogen on the basis of US EPA and National Toxicology Program carcinogenicity data in mice and rats [NTP, 1989; US EPA, 2002]. However, a recent evaluation of the available rodent carcinogenicity data has posed questions on data interpretation [Ishmael et al., 2006]. Some epidemiological investigations have reported association of DDVP exposure with lymphatic and prostate cancers among agricultural workers [Brown et al., 1990; Alavanja et al., 2003; Mills and Yang, 2003; Flower et al., 2004], and further studies are still at an intermediate follow-up stage [Koutros et al., 2008]. Experimental data concerning DDVP genotoxicity are still inconclusive since in vitro mutagenicity has been observed in several model systems [reviewed in Mennear, 1998], including human cells [Doherty et al., 1996; Booth et al., 2007], but its in vivo mutagenic activity is still unclear [Booth et al., 2007]. In light of the large use of this chemical in agriculture and the association of its exposure to cancer, investigations on the biological mechanism mediating its cancer-related effects are urgently needed.

We previously demonstrated that DDVP exposure produces micronuclei that exclusively contained entire chromosomes in the metabolically competent human lymphoblast cell line AHH-1 [Mattiuzzo et al., 2006]. The chemical was also effective in promoting chromosome nondisjunction in the same (karyotypically normal) cell line, as assessed by fluorescence in-situ hybridization experiments with chromosome specific centromeric probes [Mattiuzzo et al., 2006]. These findings highlighted the capacity of this pesticide to promote chromosome missegregation and aneuploidy.

Aneuploidy and its associated chromosome instability are considered driving forces behind multistep carcinogenesis, contributing to tumorigenesis and/or tumor progression [Holland and Cleveland, 2009]. Moreover, available evidence demonstrates that defects in chromosome segregation at mitosis are critical events in the generation of aneuploid cells [Cimini and Degrassi, 2005; Perez de Castro et al., 2007]. Thus, a crucial issue in carcinogenic risk prevention is the investigation on the capacity of environmental compounds to interfere with mitotic structures and control mechanisms regulating chromosome segregation, since exposure to these agents may promote aneuploidy, chromosome instability and cancer. In this study, we thoroughly analyzed the influence of DDVP on cell division and spindle assembly. We now show that DDVP delocalizes the mitotic kinesin Kif2a from centrosomes, producing monopolar spindles and extremely aberrant mitotic progression.

Abbreviations
DDVP

dichlorvos

MT

microtubule

phospho H3

Ser 10-phosphorylated histone H3

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. AUTHOR CONTRIBUTIONS
  9. REFERENCES
  10. Supporting Information

Cell Culture and Chemicals

AHH-1 lymphoblast cells [Crespi and Thilly, 1984] were obtained from ATCC and maintained in RPMI 1640 medium containing GlutaMAX (Gibco), 10% fetal bovine serum (Cambrex Bio Science), 100 IU/ml penicillin, and 100 μg/ml streptomycin. HeLa cells (from ATCC) were grown in DMEM supplemented with 10% fetal bovine serum, 1% l-glutamine and antibiotics. Cells were grown at 37°C in a humidified atmosphere containing 5% CO2. For each experiment, DDVP (Fluka, purity > 99.4%) was freshly prepared at a concentration of 10 μg/ml in dimethylsulfoxide (DMSO). 10 μM taxol or 100 mM nocodazole stock solutions were prepared in DMSO and stored at −20°C. All chemicals except DDVP were from Sigma-Aldrich.

Treatment Schedule and siRNA Transfection

For each experimental point, 1 × 106 AHH-1 cells were inoculated in 5 ml medium or 2 × 105 HeLa cells were plated in 35 mm Ø Petri dishes containing coverslips. The following day, cultures received <1%, v/v DMSO (control) or different amounts of DDVP dissolved in DMSO and were harvested 18 h later. In silencing experiments, 2 × 105 HeLa cells were plated in 35 mm Ø Petri dishes the day before RNA interference. 60 nM siRNA targeting MCAK (5′-GATCCAACGCAGTAATGGT-3′) [Cassimeris and Morabito, 2004] or control siRNA against firefly luciferase (GL2) sequence (5′-CGTACGCGGAATACT TCGA-3′) [Elbashir et al., 2001] were supplied to cells using Lipofectamine 2000 (Invitrogen) as previously described [Mattiuzzo et al., 2011]. Cells were harvested 48 hr from transfection and received DDVP for the last 18 hr of growth, where indicated. Data presented are mean ± SEM of at least three independent experiments.

Immunostaining, Image Acquisition, and Image Analysis

AHH-1 cells were rinsed in PHEM buffer (60 mM PIPES, 25 mM Hepes, 10 mM EGTA, 2 mM MgCl2), cytocentrifuged on polylysine-coated coverslips, fixed for 15 min in 3.7% formaldehyde in PHEM, permeabilized 5 min in 0.2% Triton X-100 in PHEM and postfixed in cold methanol for 3 min. HeLa cells grown on coverslips were rinsed in PHEM and fixed as described above. After blocking in 20% goat serum for 30 min, coverslips were incubated with the following antibodies: rabbit anti-Ser 10 phospho-H3 (Millipore), mouse anti-Bub1 (Chemicon), mouse anti-α-tubulin (DM1α; Sigma); rabbit anti-γ-tubulin (Sigma), rabbit anti-Kif2a (a gift of dr. Compton). For p150-glued antibody (BD Transduction Laboratory) immunostaining, cells were fixed in cold methanol. Secondary antibodies conjugated to Alexa-488 (Molecular Probes) or Rhodamine-RedX (Jackson Laboratories) were chosen as appropriate and used as recommended by the suppliers. DNA was counterstained with 4'-6-Diamidino-2-phenylindole (DAPI) and coverslips were sealed in antifade solution (Vector Laboratories). All preparations were examined under an Olympus Vanox microscope equipped with a 100× (1.35 NA) oil immersion objective. Images were acquired using a SPOT CCD camera (Diagnostic Instruments) controlled by ISO 2000 software (DeltaSistemi) and processed using Adobe Photoshop 7.0. For Bub1 and p150 glued kinetochore localization, z-stacks of optical sections were acquired at 0.3 μm intervals, and images were deconvolved and reconstructed as maximum projections using AutoDeblur 9.3 software (AutoQuant Imaging). Measurements of fluorescence intensity were obtained from maximum projections of images acquired under identical exposure settings using NIH ImageJ 1.3 software.

Tubulin Polymerization Assay and Immunoblotting

To separate cytosolic and cytoskeletal-associated proteins, HeLa cells were rinsed twice in PIPES-EGTA-MgCl2 (PEM) buffer (85 mM PIPES pH 6.94, 10 mM EGTA, 1mM MgCl2, 2 M glycerol, 1 mM phenylmethylsulfonylfluoride, 0.1 mM leupeptin, 1 μM pepstatin, 2 μg/mL aprotinin), lysed at room temperature for 10 min with PEM buffer supplemented with 0.1% v/v Triton X-100 and rinsed in PEM buffer [Appierto et al., 2009]. These Triton X-100–soluble fractions were then diluted 3:1 with 4 × SDS-PAGE sample buffer. The insoluble material that remained attached to the dish was scraped into SDS-PAGE sample buffer containing protease inhibitors. Proteins from the two fractions were resolved in SDS-PAGE with NuPAGE® Bis-Tris precasted gels, run in NuPAGE® MOPS buffer (Invitrogen) under reducing conditions and transferred onto nitrocellulose membranes (Schleicher and Schuell). Membranes were blocked and incubated with anti-β-tubulin antibody (Sigma). Goat anti-mouse HRP (Chemicon) antibody was successively applied and antigens on the membrane were revealed by enhanced chemiluminescence (ECL plus, Amersham). In immunoblotting experiments HeLa cells were lysed for 30 min on ice in RIPA buffer (50 mM Tris, 150 mM NaCl, 1 mM EGTA, 1 mM EDTA, 1 mM sodium deoxycholate, 1% NP40) supplemented with protease inhibitors. After determination of protein concentration, 40 μg of total proteins were resolved in SDS-PAGE as described above. Membranes were incubated with rabbit anti-Kif2a and goat anti-rabbit HRP (Santa Cruz) antibodies and revealed by ECL plus. Densitometric analysis of immunoblots was performed using ImageJ 1.3 software.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. AUTHOR CONTRIBUTIONS
  9. REFERENCES
  10. Supporting Information

DDVP Disrupts Mitotic Progression and Chromosome Condensation

We previously showed that DDVP acts as an aneuploidy-inducing agent in AHH-1 human lymphoblast cells [Mattiuzzo et al., 2006]. To identify the possible mechanism(s) of the aneuploidy-inducing activity of this pesticide, we analyzed mitotic progression after DDVP treatment. A strong dose-dependent increase in the number of mitoses was observed after a DDVP exposure lasting approximately one cell cycle in AHH-1 cells, as evidenced by the number of cells positive to Ser 10-phosphorylated histone H3 (phospho H3), an early marker of mitotic chromosome condensation (Figs. 1A and 1B). Concomitantly, ana-telophase figures decreased with DDVP treatment and were absent at the two higher concentrations (Fig. 1B). Metaphase figures also disappeared from the mitotic population at the same concentrations (Fig. 1A). This suggested that DDVP-treated cells did not proceed into anaphase because chromosomes were unable to fully congress to the metaphase plate and mitotic checkpoint was activated. Interestingly, phospho-H3 immunostaining revealed prominent defects in chromatin condensation of prometaphase cells following DDVP exposure. Mitotic cells were hypercondensed, i.e. possessed highly condensed chromosomes that appeared ill-defined and collapsed in a stacked central group (Fig. 1C, hypercondensed) or appeared as pyknotic mitosis, i.e. phospho-H3 positive mitoses in which the chromosomes had collapsed in one or several over-compacted chromatin masses (Fig. 1C, pyknotic). A quantitative analysis of these figures showed that the vast majority of prometaphase cells observed after 40 ng/ml DDVP were hypercondensed and that DDVP-induced condensation defects were even more dramatic at the highest concentration (50 ng/ml) where about one-third of mitoses were pyknotic.

image

Figure 1. DDVP blocks mitotic progression and produces hypercondensed chromosomes in AHH-1 cells. (A) Immunofluorescence detection of mitotic cells through reactivity to phospho-H3 histone antibody in control and 40 ng/ml DDVP-treated cells. (B) Quantification of mitosis (black squares) and ana-telophase (black triangles) frequencies following 18 h DDVP exposure. Data are mean ± SEM of three independent experiments in which 3000 interphase cells and 300 mitoses were analysed for each condition on phospho-H3 or Giemsa stained slides. *P < 0.05; **P < 0.01; ***P < 0.001, t-test. (C) Examples of DDVP induced chromosome condensation defects as detected by anti-phospho-H3 immunostaining. (D) Frequencies of chromosome condensation defects after DDVP exposure. Data are mean ± SEM of three independent experiments in which 300 mitoses were analysed for each condition on phospho-H3 or Giemsa stained slides. *P < 0.05; **P < 0.01; ***P < 0.001, t-test.

Download figure to PowerPoint

DDVP Induces Monopolar Spindles That Elicit a Checkpoint-mediated Mitotic Arrest

Absence of late mitotic stages in the mitotic population at the highest concentrations (Fig. 1B) and chromosome hypercondensation (Fig. 1C) were reminiscent of the mitotic effects induced by well known drugs that inhibit (nocodazole) or stabilize (taxol) MT assembly, thereby activating the mitotic checkpoint, the control mechanism that arrest cells in prometaphase until all kinetochores interact with spindle MTs [Musacchio and Salmon, 2007]. To investigate the influence of DDVP exposure on microtubules and mitotic spindle structure, immunofluorescence detection of α-tubulin and γ-tubulin was performed to visualize spindle MTs and centrosomes in human lymphoblasts (Fig. 2A). In cells treated with the lowest DDVP concentration mitotic spindles were bipolar but spindle structure was extremely aberrant with abnormally long MT fibres radiating toward different directions, often reaching the cell cortex beyond the opposite spindle pole (Fig. 2A, second row), resembling the MT morphology observed after inhibition of MT depolymerases [Ganem and Compton, 2004; Manning et al., 2007]. Strikingly, at 40 ng/ml DDVP several mitotic cells had monopolar spindles, in which the chromosomes were arranged around a single spindle pole, and exhibited extremely long MTs extending past the chromosomes (Fig. 2A, third row). Centrosome staining demonstrated that duplication and initial separation of centrosomes at prophase occurred in treated cells since very often two adjacent centrosomes were visible in monopolar spindles (Fig. 2A, third and fourth row). Spindle defects represented a major effect of DDVP treatment, since abnormally long MTs in bipolar spindles were present in more than 60% of mitotic cells at 20 ng/ml and monopolar spindles were efficiently induced at the two higher concentrations, and monopolar spindles represented the vast majority of mitoses at 50 ng/ml DDVP (Fig. 2B).

image

Figure 2. DDVP induces monopolar spindles in AHH-1 cells. (A) Immunostaining of centrosomes (γ-tubulin, green), mitotic spindle (α-tubulin, red) and chromosomes (blue) in control and DDVP-treated cells. The figure depicts a bipolar mitosis with abnormally long MTs that reach the cell cortex (second row), a monopolar mitosis showing exaggerated long MTs passing the chromosomes (third row) and a monopolar mitosis showing pyknotic chromatin masses (fourth row) after DDVP treatment. (B) Quantitative measurement of MT and spindle pole defects in control and DDVP-treated cells. Data are the mean ± SEM of three independent experiments in which 325–400 mitotic cells were analysed for each condition. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Download figure to PowerPoint

Mitotic checkpoint dependent prometaphase arrest is mediated by the accumulation on kinetochores not interacting with spindle fibres of several spindle checkpoint proteins including Mad1, Mad2, Bub1, BubR1, Bub3, and dynein [Howell et al., 2004]. To investigate mitotic checkpoint activation after induction of monopolar spindles by DDVP treatment, we used adherent growing HeLa cells, since immunostaining of these cells allowed us to unambiguously localize and quantify mitotic proteins on kinetochores. Preliminary results had showed that DDVP effects on chromosome condensation and induction of monopolar spindles in HeLa cells recapitulated the ones observed in suspension cultures of AHH-1 cells (Supporting Information 1). Kinetochore localization of the mitotic checkpoint protein Bub1 was then examined in HeLa mitotic cells. Bub1 immunodetection showed a checkpoint-dependent dynamics of this protein in untreated cells: the protein accumulated on early prometaphase unattached kinetochores, partially delocalized on aligned kinetochores in late prometaphase cells and was totally absent on anaphase cells (Fig. 3A). Quantitative measurements in untreated cells showed that Bub1 intensity at kinetochores was reduced to 50% in late prometaphase cells that exhibited mostly aligned chromosomes, as compared with early prometaphase cells in which most chromosomes were mono-oriented and empty kinetochores accumulated mitotic checkpoint proteins, including Bub1 (Fig. 3B). DDVP-treated bipolar prometaphases accumulated Bub1 on their kinetochores at a level that was similar to that of early prometaphase control cells and this Bub1 accumulation was even higher on kinetochores in monopolar spindles (Figs. 3A and 3B). This finding indicates that the mitotic checkpoint is activated after DDVP treatment and that kinetochores in monopolar cells strongly activate this control mechanism. We then analyzed the localization of the dynein/dynactin complex subunit p150 glued, a marker for unattached kinetochores [King et al., 2000] in DDVP-treated cells. Immunostaining experiments showed that the dynein complex decorated kinetochores as double dots prior to MT attachment and then moved on kinetochore fibres and to spindle poles after chromosome alignment, i.e., on metaphases from untreated cultures (Fig. 3C), implicating dynein function in mitotic checkpoint inactivation [King et al., 2000; Bader and Vaughan, 2010]. Following DDVP treatment, dynein localized as close double dots on most kinetochores in both bipolar and monopolar mitoses (white arrows in Fig. 3C), indicating that sister kinetochores were not subjected to inter-kinetochore tension. This observation strengthens the conclusion that DDVP-treated cells were arrested in prometaphase by mitotic checkpoint activation. To unambiguously ascertain the kinetochore attachment status, we took advantage of a calcium protocol that depolymerise unstable MTs and maintain K-fibers [Mitchison et al., 1986]. Most MTs in DDVP-treated cells were resistant to calcium treatment. However, kinetochores in monopolar cells were not under tension, since kinetocores interacted only laterally with MTs (Supporting Information 2). This finding explains the checkpoint-mediated prometaphase arrest and suggests a DDVP-induced stabilization of spindle MTs. Altogether, this set of data demonstrated the pesticide's ability to specifically produce monopolar spindles associated with mitotic checkpoint activation and prometaphase arrest.

image

Figure 3. DDVP treated HeLa cells accumulate in prometaphase by activation of the mitotic checkpoint. (A) Immunostaining of Bub1 (red), α-tubulin (green), and chromosomes (blue) on different mitotic stages in control cultures and on bipolar and monopolar prometaphases from DDVP-treated cells. Arrows point to high intensity Bub1 signals at kinetochores of polar chromosomes in a late prometaphase cell from a control culture. Bub1 signals in the central region of the spindle are weaker. Maximum projections from deconvolved z-stacks are shown. (B) Quantitative analysis of Bub1 intensity (arbitrary units) at kinetochores in control and DDVP- treated cells. Data are mean ± SEM of values on early prometaphase (N = 188/3 cells) and late prometaphase (N = 167/3 cells) kinetochores from control cultures and on prometaphase kinetochores in bipolar spindles (N = 140/2 cells), and monopolar spindles (N = 153/2 cells) from DDVP-treated cultures. ***P < 0.001 comparing mean Bub1 intensity on late prometaphase vs. early prometaphase kinetochores in control cells. P = 0.33 comparing mean Bub1 intensity on control early prometaphases versus DDVP-treated monopolar cells. (C) Immunostaining of p150 glued (green), γ-tubulin (red), and chromosomes (blue) on different mitotic stages in control cultures and on bipolar and monopolar prometaphases from DDVP-treated cultures. Arrows point to high p150 glued double signals on kinetochores from an early prometaphase cell and on sister kinetochores on polar chromosomes in a late prometaphase cell from control cultures and to high p150 glued signals on sister kinetochores from DDVP-induced bipolar and monopolar mitoses. Maximum projections from deconvolved z-stacks are shown.

Download figure to PowerPoint

To investigate the final outcome of DDVP-induced mitotic alterations, mitotic progression was analyzed by live cell imaging of HeLa cells at a concentration that allowed residual mitotic progression (Supporting Information 1). About 80% of cells entering mitosis in 20 ng/ml DDVP remained blocked in prometaphase during the observation time (Fig. 4A) and all recorded cells remained blocked in prometaphase at higher doses (not shown), indicating that a permanent mitotic arrest is the more frequent outcome of a DDVP-treated mitosis, possibly promoting apoptosis-mediated cell death. In the few cells that completed mitosis at 20 ng/ml DDVP, the prometaphase-metaphase duration was significantly lengthened and cells experienced a very long telophase/cytokinesis time (Fig. 4B). Collectively, these findings demonstrates that cells undergoing high dose DDVP treatment are permanently arrested in mitosis by the presence of monopolar spindles. At lower DDVP doses mitotic progression is extremely abnormal.

image

Figure 4. DDVP arrests mitosis and impair mitotic progression in HeLa cells. (A) Percentage of cells are arrested in prometaphase in control (N = 32) and 20 ng/ml DDVP (N = 32) treated cultures in live imaging acquisitions (recording time= 180 min). (B) Length of the different stages of mitosis in cells progressing through mitosis after 18 h exposure to 0.2 % DMSO (control, N = 25) or 20 ng/ml DDVP (DDVP, N = 7), ***P < 0.001, t-test.

Download figure to PowerPoint

DDVP Mitotic Defects Are Mediated by the Displacement of the Depolymerizing Kinesin Kif2a From Spindle Poles

To identify potential targets of DDVP action we decided to further examine the MT stabilization effect observed in mitotic cells (Supporting Information 2). Therefore, we investigated the influence of DDVP on MT stability in HeLa cells over the dose-range previously analyzed for spindle defects. Fractionation of soluble and cytoskeletal proteins allowed the separation of free and microtubule-polymerized tubulin fractions that were successively revealed by western blot analysis [Appierto et al., 2009]. This assay showed that DDVP significantly decreased the free tubulin fraction in a dose-dependent manner (Figs. 5A and 5B), with only 20% of tubulin recovered in the soluble fraction after 40 ng/ml DDVP (Fig. 5B). Taxol, a well known MT stabilizing agent, displayed a similar effect in this assay (Figs. 5A and 5B). These results demonstrate that DDVP acts by stabilizing MTs and suggests that the pesticide may stimulates tubulin polymerization by suppressing MT dynamics.

image

Figure 5. DDVP enhances tubulin polymerization in HeLa cells. (A) Western blot analysis of soluble cytosolic (S) or cytoskeletal (P) tubulin from cells exposed for 18 h to 0.4% DMSO (0), 10, 20, 30, 40 ng/ml DDVP, 10 nM taxol (TAX) or 200 μg/ml nocodazole (NOC). (B) The bar graph shows mean ± SEM of β-tubulin intensities (arbitrary units) in cytosolic (S) or cytoskeletal (P) fractions, obtained by the densitometric analysis of immunoblots from three independent experiments. *P < 0.05, **P < 0.01, t test comparing mean β-tubulin intensity in soluble fractions from DDVP treated versus control cells.

Download figure to PowerPoint

In addition to MT dynamics brought about by tubulin's intrinsic GTPase activity, MT assembly/disassembly is controlled by a balance between MT-stabilizing and destabilizing proteins that bind along MTs and by the coordinated action of two major classes of MT-dependent motor proteins, collectively grouped as kinesins and dyneins [Gatlin and Bloom, 2010]. Previous work on motor proteins has suggested that maintenance of spindle bipolarity depends on the balance of opposing forces generated by the MT depolymerizing activity of the kinesin-13 family members Kif2a at spindle poles and MCAK at kinetochores [Ganem and Compton, 2004; Manning et al., 2007]. Interestingly, depletion of the centrosomal Kif2a kinesin has been previously shown to specifically induce monopolar spindles, whose formation could be rescued by altering MT dynamics using MT drugs or depleting MCAK depolymerising activity at kinetochores [Ganem and Compton, 2004].

We readily noticed that DDVP induced mitotic spindle defects were markedly reminiscent of the ones observed after Kif2a depletion since exaggeratedly long MTs and induction of monopolar spindles with MTs extending beyond chromosomes to the cell cortex were characteristic of both DDVP treatment (Fig. 2) and Kif2a silencing [Ganem and Compton 2004; Manning et al., 2007]. Therefore, we tested whether DDVP biological activity was connected to its ability to interfere with the centrosomal Kif2a depolymerase. We compared Kif2a localization in control and DDVP treated mitoses. Combined α-tubulin and Kif2a immunostaining demonstrated that kinesin massively localized to centrosomes and to spindle MTs close to the poles in control metaphase cells. On the contrary, Kif2a was barely detectable at the center of the monopolar array of MTs on DDVP-induced monopolar (Fig. 6A) or bipolar cells (not shown). A dramatic decrease in Kif2a fluorescence intensity (mean ± SEM) at centrosomes was observed in DDVP-treated bipolar and monopolar cells as compared with control bipolar cells (Fig. 6B, P < 0.001, t test). Measurements on centrosomes in DDVP-induced bipolar and monopolar spindles were pooled together since they did not differ significantly, suggesting that delocalization of Kif2a from bipolar spindles may drives the formation of monopoles. DDVP specificity in de-localizing Kif2a from centrosomes was demonstrated by the fact that Kif2a heavily accumulated at the center of monopolar spindles induced by monastrol, a specific inhibitor of the Eg5 kinesin (Figs. 6A and 6B) and that Aurora A, a centrosomal kinase involved in mitotic spindle assembly, was not displaced from centrosomes in DDVP-treated cells (Supporting Information 3). Furthermore, Kif2a also accumulated on spindle poles in taxol-treated cells (Supporting Information 3), indicating that DDVP-induced Kif2a displacement from centrosomes was not caused by the presence of stabilized MTs. Finally, western blot analysis of whole cell proteins showed that intracellular Kif2a levels were unaltered in DDVP-treated cells (Fig. 6C), leading to the conclusion that induction of monopolar spindles by DDVP is due to the pesticide's capacity to displace the depolymerising kinesin Kif2a from spindle poles.

image

Figure 6. DDVP produces monopolar spindles by impairing the balance of centrosome and kinetochore spindle forces in HeLa cells. (A) Immunostaining of chromosomes (DNA), mitotic spindle (α-tubulin), and Kif2a kinesin (Kif2a) in control (control), 40 ng/ml DDVP (DDVP) or 100 μM monastrol (monastrol)-treated cells. (B) Quantitative analysis of Kif2a intensity (arbitrary units) on spindle poles in control cells, DDVP treated bipolar and monopolar cells and monastrol treated monopolar cells. The graph shows individual cell values and means ± SEM Kif2 intensity values from control (N = 35), DDVP- (N = 29), and monastrol- (N = 30) treated cells. ***P < 0.001 comparing mean Kif2a intensity in DDVP-treated vs. control cells. (C) Immunoblot showing Kif2a levels after 18 h exposure to 0.4% DMSO (0), 10, 20, 30, 40 ng/ml DDVP, 10 nM taxol (TAX) or 200 μg/ml nocodazole (NOC). (D) Frequencies of monopolar spindles after 18 h exposure to 40 ng/ml DDVP alone (DDVP) or DDVP in combination with 5 h exposure to 100 nM nocodazole (DDVP+NOC), or in combination with 60 nM control siRNA (DDVP+siGL2) or 60 nM specific MCAK siRNA (DDVP+siMCAK). Data are mean ± SEM of three-four independent experiments. 200–600 mitotic cells analysed for each condition. ***P < 0.001, t test comparing DDVP alone to DDVP+NOC, **P < 0.01, t test, comparing DDVP+siGL2 to DDVP+siMCAK.

Download figure to PowerPoint

In light of the coordination of depolymerising activities at MT plus and minus ends, we be hypothesized that functional inhibition of the depolymerising Kif2a activity at centrosomes of bipolar spindles, subsequent to its delocalization, could allow MT depolymerases present on kinetochores to extensively depolymerise MT plus-ends, thus pulling the centrosomes together and leading to spindle collapse and monopolarity. To test this hypothesis, we added the MT inhibitor nocodazole (NOC) for the last 5 hr of DDVP exposure at a concentration that suppressed MT dynamics [Ganem and Compton, 2004]. Monitoring of spindle polarity showed that monopolar spindles were significantly reduced and bipolar spindles increased in DDVP-treated cells entering mitosis in the presence of NOC as compared with cells exposed to DDVP alone (Fig. 6D). This suggested that dynamic MTs are required for DDVP action in spindle collapse. To further corroborate this hypothesis we used short interfering RNA (siRNA) to knock-down expression of the kinetochore MT depolymerase MCAK [Mattiuzzo et al., 2011] and found that MCAK depletion was also able to rescue spindle bipolarity in DDVP-treated cells (Fig. 6D). Collectively, these data demonstrates that DDVP suppresses MT dynamics at centrosomes by Kif2a displacement and disrupts the balance of opposing centrosomal and kinetochore forces controlling spindle bipolarity during prometaphase.

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. AUTHOR CONTRIBUTIONS
  9. REFERENCES
  10. Supporting Information

The long-term effects of the pesticide DDVP are matter of public concern due to the wide spread use of this chemical and its potential association with increased cancer risks among exposed populations [Brown et al., 1990; Alavanja et al., 2003; Mills and Yang, 2003; Flower et al., 2004; Koutros et al., 2008]. Our results uncover a new cellular effect of this pesticide in cultured human cells, i.e., its ability to interfere with mitotic spindle bipolarity. A relevant feature of DDVP effect on mitosis is the induction of extremely aberrant mitotic spindles with monopolar MT arrays, resulting in a checkpoint mediated mitotic arrest. Interestingly, monopolar spindles have been shown to block mitotic progression and promote apoptosis through a checkpoint mediated mechanism also following treatment with inhibitors of the Eg5 kinesin, indicating that interfering with the different pathways that govern spindle bipolarity may disrupt mitosis producing abnormal chromosome segregation and cell death [Marcus et al., 2005]. We also show that DDVP treated cells undergoing a prolonged mitotic arrest exhibit chromosome hypercondensation and phospho-H3 positive pyknotic chromatin masses. These aberrant chromosome morphologies have been reported to be intermediate stages in the execution of apoptosis or apoptosis-like processes that are associated with the induction of mitotic catastrophe, the form of cell death associated with mitotic failure [Ribas et al., 2006; Pratt et al., 2006; Vitale et al., 2011]. Indeed, our previous study on DDVP demonstrated its capacity to efficiently induce apoptosis in human lymphoblast cells [Mattiuzzo et al., 2006]. In this study, we demonstrate that DDVP-induced apoptosis is mediated by the pesticide capacity to induce monopolar spindle-associated mitotic arrest, which in turn promotes apoptosis either directly from mitosis, as evidenced by the presence of pycnotic chromatin, or after a compromised mitotic exit. Beside apoptosis, induction of grossly aneuploid or polyploid cells is a possible outcome of DDVP-induced mitotic disruption [Mattiuzzo et al., 2006]. Altered chromosome numbers in the form of aneuploidy or polyploidy are an ubiquitous feature of human tumor cells [Cimini, 2008] and strong evidence supports a crucial role for defective chromosome segregation and the ensuing chromosome instability in the initiation and/or progression of cancer [Cimini and Degrassi, 2005; Perez de Castro et al., 2007; Cimini, 2008; Holland and Cleveland, 2009]. In light of the role of aneuploidy in cancer, the finding that DDVP-induced spindle defects is extremely relevant to identify the mechanisms producing chromosome instability following exposure to the pesticide.

This study also demonstrates that induction of spindle monopolarity upon pesticide exposure is associated with the delocalization of the depolymerizing kinesin Kif2a from spindle poles and that spindle monopolarity can be reversed by promoting MT stabilization through chemical treatment or by inhibiting the depolymerising function of the kinesin MCAK at kinetochores. A growing body of evidence indicates that spindle bipolarity is established and maintained by interplay of multiple redundant mechanisms based on nonmotor and motor proteins [Tanenbaum and Medema, 2010]. Initial centrosome separation is driven by the MT sliding activity of the kinesin-5 protein Eg5, in association with the kinesin-12 motor protein Kif15 and MT-generated forces [Tanenbaum et al., 2009; Tanenbaum and Medema, 2010]. After nuclear envelope disassembly, Kif15 motor activity cooperates with Eg5 and with an Eg5-independent mechanism influencing MT dynamics to control spindle bipolarity. This second pathway operates on the basis of a balance of forces at MT plus- and minus-ends generated by kinesin-13 motor proteins to maintain the fusiform spindle shape [Ganem and Compton, 2004; Manning and Compton, 2007]. Indeed, the two family members Kif2a and MCAK have been implicated in the maintenance of spindle bipolarity through their coordinated action at centrosomes and kinetochores [Ganem and Compton, 2004; Manning and Compton, 2007]. Spindle bipolarity is also controlled by other spindle activities, since recent results indicate that over-stabilization of non-kinetochore MTs promotes the formation of monopolar spindles by increasing the number of overlapping anti-parallel MTs on which minus-end motors can act to pull centrosomes together [Kollu et al., 2009; Tanenbaum et al., 2009]. In this work, we demonstrate that DDVP action impinges on this finely tuned balance of forces that maintain spindle bipolarity during prometaphase. This conclusion is supported by several pieces of evidence. Beside Kif2a delocalization, the finding of DDVP-induced MT stabilization, as obtained by tubulin fractionation experiments, and the cytological observation of very long and prominent MT bundles after DDVP exposure concur in supporting this conclusion. All these DDVP effects on MT organization may be brought about by the lack of Kif2a depolymerising activity at centrosomes, secondary to the delocalization of this motor protein from centrosomes. Interestingly, recent work has demonstrated that Kif2a depolymerising activity is dependent on its centrosome localization, which is controlled by the antagonistic activity of two major mitotic kinases, i.e. Aurora A and Plk1, so that Plk1 depletion reduces spindle pole Kif2a localization without altering its cellular protein content [Jang et al., 2008]. In light of these results, it will be important to investigate the influence of DDVP on Plk1 localization and function.

In conclusion, this study uncovers and thoroughly dissects a new biological effect of the worldwide used pesticide DDVP, i.e. alteration of spindle bipolarity through delocalization of the Kif2a kinesin from centrosomes. DDVP-induced alterations in spindle organization can be the cause for the previously reported induction of aneuploidy by DDVP treatment of human cells [Mattiuzzo et al., 2006]. It is now widely accepted that aneuploid and polyploid cells are inherently chromosomally unstable producing further aneuploid and/or polyploid cells [Thompson and Compton, 2008]. Therefore, DDVP-induced chromosome instability could greatly contribute to the cancer promoting activity of this pesticide. Future work should be addressed to clarify whether disruption of mitotic spindle assembly and induction of monopolar spindles are characteristic of in vivo exposure to DDVP. In conclusion, our findings identifies a direct action of a chemical present in the environment on spindle architecture and mitotic division, opening the possibility that many other environmental agents may exert their toxic action through their interference with the mitotic machinery.

AUTHOR CONTRIBUTIONS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. AUTHOR CONTRIBUTIONS
  9. REFERENCES
  10. Supporting Information

MF, MM, and FD conceived and designed the experiments. MF, MM, GM, and PT performed the experiments, collected the data and prepared the figures. FD prepared the manuscript draft. All authors approved the final manuscript.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. AUTHOR CONTRIBUTIONS
  9. REFERENCES
  10. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGMENTS
  8. AUTHOR CONTRIBUTIONS
  9. REFERENCES
  10. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
em21769-sup-0001-suppinfo.doc361KSupplementary Figure 1.
em21769-sup-0002-suppinfo.doc1250KSupplementary Figure 2.
em21769-sup-0003-suppinfo.doc188KSupplementary Figure 3.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.