Thymoquinone disrupts the microtubule dynamics in fission yeast Schizosaccharomyces pombe

Abstract Mad2 deletion strain of Schizosaccharomyces pombe was found to be sensitive to thymoquinone, a signature molecule present in Nigella sativa in a dose-dependent manner. Mad2 protein is an indispensable part of mitotic spindle checkpoint complex and is required for the cell cycle arrest in response to the spindle defects. Although the expression of α tubulin was not affected in thymoquinone treated cells, but the expression of β-tubulin was reduced. Further, the absence of microtubule in thymoquinone treated cells suggests its involvement in tubulin polymerization. Molecular docking studies revealed that thymoquinone specifically binds to β-tubulin near the Taxotere binding site of Tub1 (Tubulin α-β dimer). These studies additionally showed that thymoquinone interacts with the residues present in chain B, which is an inherent part of Mad2 protein of mitotic checkpoint complex (MCC). We concluded that the thymoquinone disrupts the microtubule polymerization that leads to the requirement of spindle checkpoint protein for the cell survival.


Introduction
Schizosaccromyces pombe come up to an elevated level of perpetuation in cellular processes mimicking mammalian cells with numerous advantages including simple growth requirements, rapid cell division and ease of genetic maneuvring [1]. Cell cycle checkpoints are inherent safety guard of the cell warranting proper growth, and defect in these cell cycle checkpoints can lead to the tumor progression. Spindle Assembly Checkpoint (SAC), Centrosome Duplication Checkpoint (CDC) and DNA Damage Checkpoint are the major checkpoints operating during the cell cycle progression [2][3][4][5][6]. SAC is a conserved surveillance mechanism that arrests cells in mitosis in response to malfunctioning spindle [7]. SAC prevents the anaphase onset, mitosis exit, cytokinesis initiation and is triggered by either existence of unattached kinetochore or by the lack of tension on kinetochores [8,9]. The core proteins involved in SAC are Bub1, Bub3, Mad1, Mad2 and Mad3 [10,11]. These proteins are responsible for regulating Anaphase Promoting Complex (APC) activity. APC inhibition takes place after the suppression of the Cdc20 (APC activator) by Mad2, thereby averting premature cyclin B and securin degradation [12][13][14]. The S. pombe SAC protein Mad2 interacts with APC and blocks anaphase. In response to the defective microtubule, the Mad2-dependent SAC delays mitotic exit [10]. In S. pombe, Mad2 over-expression imitates activation of the SAC and results in the arrest of cell cycle at the metaphase-anaphase transition [15,16]. Tumor cells frequently exhibit Chromosomal Instability (CIN) instead of undergoing apoptosis and endure a convinced level of aneuploidy. During development and aging, polyploidy is programmed in selected tissues but sometimes due to stress and particular disease condition like cancer, random polyploidization occurs [17]. Inhibition of Mad2 or BubR1 checkpoint proteins in aneuploid cancer cells restrains pathways necessary to maintain CIN [18].
Thymoquinone (TQ) is a mitochondria-targeted antioxidant, anti-inflammatory, antineoplastic agent that regulates apoptosis by either amending PI3K/Akt and p38kinase pathways or by inducing ROS generation [19,20]. TQ affects the α/β tubulin degradation with upregulation of tumor suppressor p73 and G2/M phase apoptosis in U87 astrocytoma cancer cells and Jurkat T-lymphoblastic leukemia cells with no effects in normal fibroblast cells suggesting its role as an anti-microtubule drug [21]. In this report, cell cycle checkpoint inhibition was studied with selected monoterpenes present in many plants including Nigella sativa. We reported that the thymoquinone (TQ) disrupts the microtubule assembly that leads to growth inhibition in spindle checkpoint defective mutants. Furthermore, in silico studies revealed that TQ specifically binds to β-tubulin and mitotic checkpoint complex (MCC) through chain B, which is an inherent part of Mad2 protein.

Yeast strains and growth conditions
Haploid wild-type (SP6: hleu-32), mad2 deletion (NW1419: h -leu1-32 ura4DS/E mad2::ura4 ade6-210) and bub1 deletion (SH495: h -leu1-32 bub1::kan R ) strains of S. pombe were used. For the spotting experiment, wild type and mad2 deleted strains were grown at 30 °C up to mid-log phase, 10 7 cells were serially diluted, spotted on plates containing thymoquinone and incubated at 30 °C for 3 days before taking photographs. For cell survival assay, wild-type and mad2 deleted strains were grown in the presence of different concentrations of thymoquinone for 2 h at 30 °C, for each sample ~1000 cells were placed on YEA plates and incubated at 30 °C until colonies appear. A graph was plotted after colony count.

Preparation of cell lysate and western blot analysis
Cells were grown to mid-log phase and shifted to YEA media containing no drug or 30 µg/mL thymoquinone, incubated at 30 °C for 2 h. Cells were harvested by centrifugation and lysed using glass beads and a Fast Prep (Bio 101) vortex machine. Lysate in Phosphate Buffered Saline (PBS) was centrifuged at 10000 rpm in a microfuge for 5 min at 4 °C. Protein was estimated from the supernatant collected using Bradford assay. For western blot analysis, about 200 µg of total cell lysate was run on 10% SDS-PAGE, transferred to a nitrocellulose membrane, probed with anti-α-tubulin or anti β-tubulin antibody and horseradish peroxidase (HRP) -conjugated anti-mouse antibody. ECL western blotting detection reagents (Pierce) were used for protein detection using hyper film ECL (GE healthcare).

Immunofluorescence studies
Exponentially growing cells already treated with thymoquinone were used for immunofluorescence studies as described previously [22]. The cells were fixed with freshly prepared 3.7% formaldehyde for 30 min with shaking, washed by PEM buffer for 3 times and spheroplasting was performed using zymolyase. Primary antibody (Anti α-tubulin) was used at a dilution of 1:50, incubated overnight at room temperature. Cells were washed, treated with FITC-conjugated secondary antibody at a dilution of 1:100 and incubated for 4 h at room temperature. Analysis of the cells was done using a fluorescence microscope and processing was performed by Adobe Photoshop.

In silico docking studies
To comprehend the interactions of Thymoquinone with MCC, docking studies were performed. The 3D structures of Thymoquinone (CID_10281) were retrieved from PubChem database at NCBI web server (http://pubchem.ncbi.nlm.nih.gov). X-ray crystallographic 3D structures of Tubulin α-β dimer (PDB ID: Tub1) and Mitotic checkpoint complex of S. pombe (PDB ID: 4AEZ) were retrieved from Brookhaven Protein Data Bank (http://www.pdb.org). Docking was executed by using AutoDock 1.5.4 software [23] as the version uses the Lamarckian Genetic Algorithm and empirical free energy scoring function, typically providing reproducible docking results for ligands with approximately 10 flexible bonds. AutoDock 1.5.4 uses a semiempirical free energy force field to evaluate conformations in terms of inhibition constant (Ki) and binding energy (B.E.) during docking simulations. The docked proteins were visualized through UCSF Chimera, version 1.6. [24] Also to understand the association between H-bonding and hydrophobic interactions, Ligplot 1.4.5 (http://www.ebi.ac.uk/thrntosrv/software/LIGPLOT/) was used to engender schematic diagrams of protein-ligand interactions [25].

Statistical analysis
Experiments were executed in duplicate each time and values are mean ± SD of three independent analyzes. Results were subjected to analysis of variance (ANOVA), and differences between means at p < 0.05 level were considered significant compared to control in student's t-test.

TQ affects the spindle assembly checkpoint
Some checkpoint deficient mutant/knockout strains of S. pombe were assayed for their ability to grow on the plates containing TQ. Cells containing mad2 and bub1 deletion were found to be more sensitive than the wild-type strains at 20 and 30 µg/mL concentration ( Figure 1A). While chk1 and cds1 deletion strains do not exhibit such growth defect suggesting that abrogation of only spindle assembly checkpoint and not DNA damage checkpoint response or S phase checkpoint response affect cell survival in response to thymoquinone. Furthermore, only 18% mad2Δ cells were able to form colonies at 20 µg/mL concentration after 2 h exposure with TQ. The survival of mad2Δ cells was further reduced to 1.6% at 30 µg/mL concentration ( Figure 1B). Similarly the survival of bub1Δ cells was reduced to 13% and 0.8% at 20 and 30 µg/mL concentration of thymoquinone respectively ( Figure 1B). In contrary, the survival of wild-type cells was 83% and 45% at 20 and 30 µg/mL concentration respectively suggesting a dose-dependent inhibition of growth in mad2Δ and bub1Δ cells. As a control, the sensitivity of mad2Δ and bub1Δ strain on a plate containing thiabendazole (TBZ), a known microtubule inhibitor has also been shown ( Figure 1A, last panel).

The expression of β-tubulin was reduced after treatment with thymoquinone
Earlier studies suggest the involvement of thymoquinone in the degradation of α/β tubulin in cancer cell line [21]. Further to elucidate whether the growth defect in mad2Δ was due to the differential expression of tubulin protein, an important component of the microtubule, we performed western blot analysis after treating the wild type and mad2Δ cells with 30 µg/mL of TQ for 2 h. As presented in Figure 2A, the expression of α-tubulin was not affected in wild-type as well as in mad2Δ strain (data not shown) suggesting that TQ does not influence the expression of α-tubulin. Interestingly the expression level of β-tubulin was reduced after exposing the cells to thymoquinone for 2 h ( Figure 2B) that might be due to the degradation of β-tubulin protein as has also been reported in cancer cell line [21].

Thymoquinone disrupts spindle microtubule dynamics
Since the growth of mad2Δ cells was inhibited after treatment with TQ, we hypothesized that these phyto molecules might be affecting the formation of the microtubule network. In order to analyze this possibility, we performed immunofluorescence studies in the cells treated with 30 µg/mL of TQ for 2 h. The wild type untreated cells exhibited typical elongated microtubule structure ( Figure 2C, left panel). Interestingly, these microtubule structures were absent in the cells treated with TQ ( Figure 2C, right panel) clearly indicating that thymoquinone might be affecting the formation of microtubule structure.    Figure 3A, B, C).  Figure 4B). It means that residues present surrounding the docked ligand belongs to chain B which corresponds to Mad2 protein.

In silico docking analysis showed that thymoquinone interacts with β-tubulin and Mad2
Binding energy estimated for TQ docked with 4AEZ was −4.61 kcal/mol and inhibition constant was 414.71 μM ( Figure 4A, B). Other spindle poison TBZ does not shows any affinity with Mad2 protein but it does interacts with Slp1 and Mad3 protein within MCC complex (data not shown) suggesting that binding of spindle poison with MCC complex might also affect the cell survival in response to spindle poison.
Crystal Structure of Mitotic Checkpoint Complex (PDB: 4AEZ) containing Mad2, Mad3 and Slp1 of S. pombe is available while the Bub1 protein is not available for S. pombe, instead there is a Saccharomyces cerevisiae protein structure of Bub3-Bub1 bound to phospho-Spc105 (PDB: 4BL0). We docked our ligand, TQ with this protein and the B.E. and Ki are −6.38 kcal/mol and 21.14 µM respectively ( Figure 5A). The docked residues were: GLN 8.A, ALA 9.A, PRO 10.A, LYS 11.A, ASP 32.A, SER 34.A, THR 36.A, TYR 38.A and LYS 11.D ( Figure 5B). In this the A and D represents the chain while the number represents the position of the residue. This protein complex (PDB ID: 4BL0) consists of 3 proteins: Bub3 (chain A, D); Bub1 (chain B, E) and Spc105 (chain C, F) ( Figure 5A). Thus it is clear that TQ binds to Bub3 and not with Bub1 in protein complex ( Figure 5). Further fission yeast Bub3 is required for normal spindle dynamics, but not for SAC [27]. Bub1 is a fission yeast kinetochore scaffold protein, and is sufficient to recruit other spindle checkpoint proteins to ectopic sites on chromosomes [28]. Thus in fission yeast spindle assembly checkpoint complex, TQ binds and affects only Mad2 and does not affect other proteins necessary for spindle assembly checkpoint. These results suggest that the growth defects observed in bub1Δ strain ( Figure 1A) might be due to the defects in the microtubule polymerization.

Effect of Mad2 over-expression was not affected by thymoquinone
Mad2 over-expression leads to the metaphase arrest and hence cells are unable to complete the cell division that affects its survival [29]. In order to check the role of thymoquinone on the over-expression of mad2, we checked the survival of mad2 over-expressing cells under the thiamine repressible promoter in the presence of thymoquinone. As shown in Figure 6, cells containing mad2 gene under thiamine repressible nmt promoter were growing well in the presence of thiamine but in the absence of thiamine these cells were unable to form colonies due to the over-expression of mad2 gene. We found there was no change in the survival of mad2 over-expressing cells when thymoquinone was added to the media ( Figure 6). While cells containing vector control were growing well on all the plates ( Figure 6). These results suggest that the binding of TQ with the proteins required for spindle assembly checkpoint pathway (Mad2 and Bub3) might be responsible for growth defect. Alternatively, binding of TQ with tubulin leads to the requirement of these proteins for proper cell division and in the absence of proper spindle assembly checkpoint function cell survival is affected.

Conclusion
Yeast cells mimic eukaryotic mammalian cells in their mechanism while being unicellular and application of genetic approach ultimately helps to study the complex cellular function. Using phenotype-based screens the yeast model system can lead us to identify the exact mechanism of how a chemical inhibitor works within the cell [1]. In order to identify the potential target of thymoquinone, we studied the growth inhibition of some deletions or mutant strains of fission yeast S.pombe (particularly cell cycle checkpoint mutants) in the presence of thymoquinone. A dose-dependent inhibition of growth in mad2Δ cells by thymoquinone was observed indicating poor survival in response to TQ in the absence of functional SAC. Mad2 is a key component of the SAC that delays the start of anaphase until all kinetochores are attached to the spindle [30]. It attaches to Cdc20, a co-activator of APC or cyclosome and prevents it from advancing destruction of securin [31,32]. In yeast, microtubules are a critical component of the mitotic spindle that split the genetic material during cell division. We also observed that the expression of β-tubulin was affected in thymoquinone treated cells and the microtubule assembly was impaired. We hypothesized that due to the defective microtubule structure the mad2Δ and bub1Δ cells were unable to grow in the presence of thymoquinone since these cells required a functional spindle checkpoint for their survival.
Microtubules are formed by heterodimers of alpha and beta tubulin subunits in a 1:1 ratio. In silico docking studies indicated that thymoquinone binds to β-tubulin of αβ tubulin dimer near the Taxotere binding site that strengthen our hypothesis that a sophisticated system of microtubule formation and dissociation was affected by thymoquinone. Recently, thymoquinone has been reported as a new anti-microtubule agent in cancer cell lines. In the presence of thymoquinone a concentration and time-dependent degradation of α/β tubulin has been reported in A87 and Jurkat cell lines while no effect was observed in normal human fibroblast cells [21]. Further it has been observed that thymoquinone depolymerizes the microtubule network of A549 cells [33]. The tubulin polymerization in the presence of thymoquinone was also inhibited in a cell-free system with an IC 50 value of 27 µM [34]. Most of the anti-mitotic drugs like Paclitaxel, Vincristine, Colchicine itself binds to β-tubulin and ultimately disrupts the equilibrium [33]. The present finding indicating the reduced survival of SAC compromised cells in response to TQ could be responsible for driving the tumour cells towards apoptosis and hence impart the anticancer activity.