Targeting the Cytoskeleton with Plant-Bioactive Compounds in Cancer Therapy

In this overview we describe the main plant-derived bioactive compounds used in cancer therapy which has the cell cytoskeleton as therapeutic target. Three major classes of these compounds are described: antimitotics with microtubule-destabilizing and—stabilizing effects, plant-bioactive compounds that interact with intermediate filaments/actin, and plant-bioactive compounds that interact with intermediate filaments like keratins and vimentin. We also focus on the molecular aspects of interactions with their cellular targets: microtubules, intermediate filaments, and microfilaments. Some critical aspects of cardiac side effects of cancer chemotherapy are also discussed, focusing on cardiac cytoskeleton and protective effect of plant-derived compounds. The application of plant bioactives in the treatment of cancer has resulted in increased therapeutic efficacy through targeting the cytoskeleton, respectively, prevention of the injury of cytoskeletal components elicited by chemotherapeutics. efficacy through targeting the cytoskeleton and prevention of cytoskeletal injuries due to chemotherapy side effects. Research results testify both the evolution of knowledge coming from pharmacognosy and the great possibilities of future progress by means of a rational approach of natural product-based drug discovery or new pharmaceutical formulations.


Introduction
Chemotherapy is routinely used for cancer treatment. Since tumor cells lose many of the regulatory pathways of the normal cells, they continue to divide without control. Chemotherapeutic drugs try to solve these abnormalities, but sometimes the toxicity of allopathic treatments creates a significant problem.
The cytoskeleton constitutes the supporting framework of the cell, and it is composed of three types of cytosolic filaments: microtubules, intermediate filaments, and microfilaments. The entire cytoskeletal network is a dynamic structure which regulates the cell structure, and it is involved in many cellular functions such as movement, transport, or cell division [1]. The cytoskeleton is one of the main therapeutic targets in cancer cells [2].
Various cancer therapies use plant-derived bioactive products. There are four classes of plant-derived anticancer drugs currently used in oncotherapy: vinca alkaloids (vinblastine, vincristine), epipodophyllotoxins (etoposide and teniposide), taxanes (paclitaxel and docetaxel), and camptothecin derivatives (camptothecin and irinotecan) [3]. To date, new generations of vinca alkaloids, camptothecins, and epothilones as well as a novel class of taxanes have been developed. Some of these are in clinical use, others in clinical trials.
The major inconvenience in using antimicrotubule agents in oncotherapy is that these compounds cause significant side effects such as neutropenia and neurotoxicity and because of their limited efficacy as single agents [3].
This review describes the main natural compounds identified in the last year as potential anticancer agents, which have cell cytoskeleton as therapeutic target. We focus on the interactions of plant-derived anticancer drugs with all three types of cytosolic filaments: microtubules, intermediate filaments, and microfilaments. In addition, we summarize the most recent advances in the understanding of the molecular aspects of these interactions.
Some critical aspects of cardiac side effects of cancer chemotherapy are also discussed, focusing on cardiac cytoskeleton and protective effects of plant-derived compounds.
The dynamic ability of microtubules to polymerize and depolymerize is essential for cellular division and chromosome segregation during mitosis. Due to their crucial roles in dividing cells, microtubules have been considered a major target for cancer therapy. Microtubuleinteracting plant-derived biomolecules, namely, antimitotics, can be classified into two main groups based on their apparent mechanisms of action: microtubule-destabilizing agents act as tubulin polymerization inhibitors, and microtubule-stabilizing agents act as tubulin polymerization promoters [11].

Microtubule-destabilizing agents
Vinca alkaloids and colchicines prevent the polymerization of tubulin and promote the depolymerization of microtubules.
Vinca alkaloids are a series of biologically active agents isolated from Catharanthus roseus (Vinca rosea) with a potent antitumor activity, related to their ability to inhibit the polymerization of microtubules and preventing cell division [12]. There are approximately 130 vinca alkaloids distributed in different vegetal tissues: vincristine, vinblastine, and yohimbine in the aerial parts; catharanthine and vindoline in leaves; and almalicine and reserpine in roots [13]. They have demonstrated clinical efficacy in a broad spectrum of cancers, both as single agents and in combination. Vincristine, vinblastine, and vindesine are the first vinca alkaloids used as antitumor drugs. Vinorelbine is the first new second-generation vinca alkaloid, while vinflunine, a bis-fluorinated vinorelbine derivative, was synthesized by superacid chemistry and studied in phase I-III clinical trials [14,15].
The vinca alkaloids are dimeric compounds consisting of two multi-ringed subunits, vindoline and catharanthine, linked by a carbon-carbon bridge [16]. They act by binding specifically to β-tubulin and block its ability to polymerize with α-tubulin into microtubules, thus disrupting the mitotic spindle. This blocks mitosis and kills actively dividing cells. The results indicate that vinorelbine and vinflunine affect microtubule dynamics differently from vinblastine and proved to be weak binders [17].
Vincristine is used in the treatment of hematological and lymphatic neoplasms, whereas vinblastine in breast cancer, testicular cancer, choriocarcinoma, and vindesine in non-small cell lung cancer or breast cancer. Vinorelbine is useful for the treatment of non-small-cell lung cancer, and vinflunine has been used in the treatment of bladder, non-small-cell lung, and breast cancers [17].
Similar to Vinca alkaloids, colchicine extracted from plants of the genus Colchicum (autumn crocus) is a microtubule-destabilizing agent at high concentrations and stabilizes microtubule dynamics at low concentrations [18]. It first binds to soluble tubulin, leading to a complex that copolymerizes into the ends of the microtubules and prevents the elongation of the microtubule polymer. It is severely toxic to normal tissues at high dose, which limits its use in cancer therapies [19]. Colchicine showed different antitumoral effects which include inhibition of metastatic potential [20] and angiogenesis [21], cell blebbing through a Rho/Rho effector kinase (ROCK)/ myosin light-chain kinase (MLCK) pathway [22], decrease of ATP influx into mitochondria [23].
Novel microtubule-destabilizing plant-bioactive compounds are summarized in Table 1

Microtubule-stabilizing agents
Taxanes are the main class of microtubule-stabilizing agents, which prevent the depolymerization of microtubules and promote the polymerization of tubulin to microtubules.
One of the most important plant compounds in the fight against cancer was discovered in the bark of Taxus brevifolia-taxol, now named paclitaxel, which has become one of the most Active substance/herbal formulation

Mechanism of action Therapeutic use References
Angelica shikokiana methanol extract (AME) -AME and all isolated compounds inhibited tubulin polymerization -Angelicin and kaempferol-3-Orutinoside were the most active compounds -Phenolic compounds and furanocoumarins showed binding affinity to colchicine-binding site -Quercetin, kaempferol, luteolin, chlorogenic acid, and methyl chlorogenate exhibited the strongest activity against histone deacetylase 8 (HDAC8) and the highest affinity to trichostatin A-binding site. effective drugs against breast and ovarian cancer and has been approved for the clinical treatment of cancer patients. Since the first discovery of paclitaxel in the 1960s, a variety of other microtubule-stabilizing agents have been derived primarily from natural resources [37]. The molecular mechanism includes polymerization of tubulin to stable microtubules and also interacts directly with microtubules, stabilizing them against depolymerization and thereby blocks cells in the G2/M phase of the cell cycle [38]. The binding of taxol to β-tubulin in the polymer results in cold-stable microtubules even in the absence of exogenous GTP. Hydrogen/deuterium exchange (HDX) coupled to liquid chromatography-electrospray ionization MS demonstrated a marked reduction in deuterium incorporation in both β-and α-tubulin in the presence of taxol and contributed to increased rigidity in taxol microtubules and complementary to that due to GTP-induced polymerization [39].
Initially obtained from Taxus brevifolia bark, paclitaxel is now a semisynthetic product of 10-deacetylbaccatin III, which is extracted from the needles of the Taxus baccata. Similarly, docetaxel, a second-generation taxane, was directly obtained semisynthetically by esterification from the inactive taxane precursor 10-deacetylbaccatin III [40]. Paclitaxel and docetaxel bind to the specific binding sites of tubulin, which is different from the binding site of guanosine triphosphate, vinblastine, colchicine, and podophyllotoxin [41].
Docetaxel has a 1.9-fold higher affinity for the site than paclitaxel and induces tubulin polymerization at a 2.1-fold lower critical tubulin concentration. The effect on the cell cycle is different: paclitaxel inhibits the cell cycle traverse at the G2/M phase junction [42], while docetaxel produces its maximum cell-killing effect against cells in the S phase [43].
To decrease the toxicity and enhance delivery and distribution, new taxane formulations of micelles were investigated, including nanoparticles, emulsions, and liposomes [44]. Compounds such as Abraxane, CT-2103, and docosahexaenoic acid (DHA)-paclitaxel are examples of new taxanes with higher activity than paclitaxel in taxane-resistant cancers, as well as in tumors that have been unresponsive to paclitaxel [16].
Protopine is a benzylisoquinoline alkaloid isolated from Opium poppy, Corydalis tubers, and Fumaria officinalis. It stabilizes tubulin polymerization process but has no affinity to taxolbinding site. It induces a marked increase of tubulin polymerization in a dose-dependent manner in human hormone-refractory prostate cancer (PC-3 cells), similar to paclitaxel. It enhances microtubule assembly and formation of mitotic spindles in PC-3 cells [45].
Taccalonolides are plant steroids possessing a C2-C3 epoxide group and an enol-lactone isolated from Tacca leontopetaloides, Tacca plantaginea, Tacca chantrieri, Tacca plantaginea, Tacca integrifolia, etc. They act as microtubule stabilizers by binding to another microtubule site than taxol resulting in the formation of microtubule bundles and leading to cell cycle arrest and apoptosis. It is also reported that taccalonolides bind to β-tubulin near the lumen of microtubule, which is different from the taxol-binding site stabilizers which bind to α-tubulin protofilaments [46][47][48][49].
Recent study shows that the dietary flavonoid fisetin binds to tubulin and stabilizes microtubules with binding characteristics far superior than paclitaxel. It induces upregulation of microtubuleassociated protein (MAP)-2 and microtubule-associated protein (MAP)-4 and increases α-tubulin acetylation, an indicator of microtubule stabilization [50].

Microfilaments as chemotherapeutic targets of plant-derived bioactives
Actin filaments are composed of globular actin (G-actin) which polymerizes into filamentous (F) actin and participates in many important cellular processes including cell division and cytokinesis, cell signaling, vesicle and organelle movement, cell junction establishment, and maintenance.
Like microtubules, actin microfilaments can change rapidly their structure in response to external stimuli. Actin polymerization is stimulated by nucleating factors such as the Arp2/3 complex, which mimics a G-actin dimer in order to stimulate actin polymerization [51]. Actin binds ATP to stabilize microfilament formation and hydrolysis [52]. The growth of microfilaments is regulated by thymosin, which binds G-actin to lead the polymerizing process, whereas profilin binds G-actin and catalyzes the exchange of ADP to ATP, promoting monomeric addition to the plus end of F-actin [53].
During cytokinesis, disruption of actin polymerization can effect cellular structure. Cytokinesis inhibitors such as cytochalasin B disrupt the actin cytoskeleton, and the cell is unable to divide [54] but is still able to initiate another mitotic event, continuing to form nuclei and eventually becoming enlarged and multinucleated [55,56]. Cell lines derived from bladder, kidney, and prostate carcinomas become multinucleated when grown in cytochalasin B-supplemented medium, whereas cells from corresponding normal tissue remain monoor binucleate under comparable conditions [55]. These particular features make tumor cells ideal targets for chemotherapy, as they have reduced cytoskeletal integrity and multiple nucleation and increased mitochondrial activity [57].
Actin filaments are also of substantial importance to cancer cell migration. Cancer cell migration can convert between mesenchymal and amoeboid types. This latter can occur, e.g., when cells are exposed to protease inhibitors [58] and thereby mesenchymal cancer cell invasion is repressed by specific targeting of protease function. Inhibiting RhoA/ROCK signaling promotes the formation of multiple competing microfilament-derived lamellipodia that suppress amoeboid migration of tumor cells [59]. Tumor cells unable to move through amoeboid migration will switch to mesenchymal migration [60]. However, tumor cells exposed to protease inhibitors will move mainly through amoeboid migration. Using microfilament disrupting RhoA/ROCK inhibitors in combination with protease inhibitors would simultaneously block both types of cell migration.
Phytomedicine developed actin-targeted potential drugs, designed for cancer therapy ( Table 2). Oleuropein -Disrupt actin filaments in a dose-dependent manner Sarcoma [62] Alkaloid mixture derived from Senna spectabiliscassine and spectaline -Altered normal distribution pattern of F-actin filaments Liver cancer [63] Deoxyelephantopin (DET) -Affects the actin cytoskeleton network and downregulates calpain-mediated proteolysis of several actin-associated proteins -Inhibition of proteolysis of actin cytoskeletonassociated proteins identified by differential proteomic profiling Lung metastasis of mammary adenocarcinoma [64] Cucurbitacin E -Disruption of the F-actin cytoskeleton -Increases the filamentous or polymerized actin fraction Prostate carcinoma cells [65] Cucurbitacin E -Damaged F-actin without affecting beta-tubulin 95D lung cancer cells [66] Cucurbitacin I -Induced the co-aggregation of actin with phospho-myosin II by stimulation of the RhoA/ ROCK pathway and inhibition of LIM-kinase HeLa cells [67] Cucurbitacin B -Induced rapid and improper polymerization of the F-actin network Myeloid leukemia cells [68] Jasplakinolide (JAS) -Rearranged the actin cytoskeleton -JAS has a phalloidin-like action -Distribution of actin filaments was different from that induced by cytochalasin D

Cancer cells [69]
Ganoderma lucidum extracts -Inhibits growth and induce actin polymerization Bladder cancer cells [70] 4-Hydroxycoumarin -Disorganized the actin cytoskeleton correlated with reductions in cell adhesion to four extracellular matrix proteins and inhibition of random motility Melanoma cell line B16-F10 [71] by three linker regions [72]. The N-and C-terminus segments of intermediate filaments are non-alpha-helical regions of polypeptide sequences, associated with head to tail into protofilaments that pair up laterally into protofibrils; four of these protofibrils form an intermediate filament.
Whereas microfilaments and microtubules are actin or tubulin polymers, intermediate filaments are composed of 50 different proteins classified into six types based on similarities in amino acid sequence [72]. In regard to potential chemotherapeutic targets, the most promising intermediate filaments are keratins, nestin, and vimentin.

Anti-keratin agents
Keratin and cytokeratin are intermediate filaments found in the cytoskeleton of epithelial tissue. There are twenty different keratin polypeptides (K1-K20) identified and classified into type I (K9-K20) and type II (K1-K8) intermediate filaments [73]. Keratins of importance to cancer therapy are keratin 8 (K8) and keratin 18 (K18), the most common and characteristic members of intermediate filaments expressed in single-layer epithelial tissues [74,75]. Oncogenes, which activate Ras signaling, stimulate expression of K18 through transcription factors [76]. However, aberrant K8 and K18 expression has been noticed in particularly invasive carcinomas [77,78]. K18 was found to be a substrate of the cysteine-aspartic proteases during epithelial apoptosis [77].
Based on aberrant keratin expression found in many cancers, these intermediate filaments present a novel chemotherapeutic target that need to be investigated.
Crude acetone extract of Bupleurum scorzonerifolium (AE-BS) showed antiproliferative activity, induced cell arrest in G2/M phase, and apoptosis in A549 human lung cancer cells [79]. In a further study, Chen et al. [73] noticed K8 phosphorylation after AE-BS treatment of A549 cells. The association of ERK1/2 activation with K8 phosphorylation may be related to the apoptotic effect of AE-BS.

Anti-vimentin agents
Vimentin functions as a regulator in cancer cells undergoing epithelial-mesenchymal transition (EMT), an important change during tumor progression where cells detached from their original tissue become highly motile and invasive. Studies have shown that quercetin prevented epidermal growth factor (EGF)-induced EMT, migration, and invasion of prostate cancer cells by suppressing the expression of vimentin and N-cadherin [80]. Genistein, an isoflavone found in soybeans, fava beans, and lupine, has been shown to downregulate mesenchymal markers ZEB1, slug, and vimentin and therefore cause reversal of EMT in gemcitabine-resistant pancreatic cancer cells [81]. Similarly, this flavonoid was able to decrease protein expression of vimentin, cathepsin D, and MMP-2 and thus suppressed epithelialmesenchymal transition and migration capacity of BG-1 ovarian cancer cells [82]. Other natural compounds, like silibinin, induced the morphological reversal of mesenchymal phenotype to epithelial phenotype through downregulation of vimentin and MMP-2 and upregulation of cytokeratin-18 [83]. Moreover, silibinin meglumine, a water-soluble form of milk thistle silymarin, impedes the EMT in EGFR-mutant non-small-cell lung carcinoma cells by upregulation of the relative mRNA expression of CDH1 (E-cadherin) accompanied by downregulation of vimentin [84]. Berberine, a plant alkaloid present in various plants like Berberis, decreased the expression of the mesenchymal markers vimentin and fibronectin and restored the epithelial marker E-cadherin, thereby contributing to the reversal of EMT [85].
Piplartine, a biologically active component from Piper species (Piperaceae), also suppresses tumor progression and migration by disruption of the p120-ctn/vimentin/N-cadherin complex, which plays a critical role in tumor progression and invasion/metastasis [86].
Phenethyl isothiocyanate (PEITC), the main bioactive compound present in cruciferous vegetables, decreases breast and prostate tumor growth inhibition through vimentin suppression [87]. Cucurbitacin E induced disruption of vimentin cytoskeleton in prostate carcinoma cells, while microtubules were unaffected [65]. The natural product withaferin A (WFA) exhibits antitumor activity by binding to vimentin and covalently modifying its cysteine residue, which is present in the highly conserved helical coiled coil 2B domain [88]. Penduletin and casticin, flavonoids from the Brazilian plant Croton betulaster, induced changes in the pattern of expression of the cytoskeletal protein vimentin and thereby inhibit the growth of human glioblastoma cells [89].

Protective effect of plant-bioactive compounds on anthracycline-induced cardiac cytoskeletal toxicity
Cardiotoxicity is the most serious side effect of antitumoral anthracyclines, which include adriamycin, doxorubicin, mitoxantrone, daunorubicin, or epirubicin [90]. The main cause of toxicity is their effect on the cardiac cytoskeleton, consisting of myofibrils disarray [91], including both structural and functional changes: troponin I and troponin C phosphorylation mediated by a doxorubicin-induced protein kinase C activation [92,93] and decrease of troponin I, and changes of α-actin, creatine kinase, and myosin light-chain 2 expression [93]. In other studies, degradation of cardiac cytoskeletal proteins, including titin [94] and dystrophin [95], was observed. Recently, changes in the cardiac distribution of desmin have been detected, with areas of decreased expression in the cytoplasm and protein aggregation after mitoxantrone treatment [96,97]. The use of plant bioactives might protect against the oxidative stress caused by anthracycline drugs, including cytoskeleton injuries. Our group recently demonstrated that the flavonoid chrysin inhibits mitoxantrone-triggered cardiomyocyte apoptosis via multiple pathways, including decrease of the Bax/Bcl-2 ratio and caspase-3 expression along with preservation of the desmin disarray [96].

Conclusions
Plant-derived bioactive molecules constitute promising tools for the treatment of cancer. The application of plant bioactives in the treatment of cancer has resulted in increased therapeutic efficacy through targeting the cytoskeleton and prevention of cytoskeletal injuries due to che-