Some leopards can change their spots: potential repositioning of stem cell reprogramming compounds as anti-cancer agents
Acute myeloid leukemia
Glycogen synthase kinase-3β
Induced pluripotent stem cells
Janus activated kinase/signal transducer and activator of transcription 3
Mitogen activated protein kinase
Mouse embryonic fibroblast
Mitogen-activated protein kinase kinase
Tumor necrosis factor (TNF) α-related apoptosis inducing ligand
Small molecule compounds have been used to facilitate the generation of induced stem cells (Jung et al. 2014c). These cells are created using techniques that artificially modulate the epigenetic status of target cells (Krause et al. 2015). In addition to small molecule-based approaches, other established methods for producing induced stem cells include nuclear transfer into enucleated oocytes (Gurdon and Wilmut 2011), cell fusion with pluripotent/totipotent stem cells (Do et al. 2007), or the addition of exogenous agents, such as vectors encoding reprogramming transcription factors (Takahashi and Yamanaka 2006), microRNAs (Anokye-Danso et al. 2011), and recombinant proteins (Zhou et al. 2009). Of these approaches, small molecule-based methodologies have received significant attention from the research community (reviewed in (Yu et al. 2014)). These methodologies aimed to replace one or more of the classical “Yamanaka” reprogramming transcription factors (Oct-3/4, Sox-2, c-Myc, and Klf4 (Takahashi and Yamanaka 2006)) with small molecules. This approach has value because small molecules possess a number of advantages as tools to induce stem cell phenotypes (reviewed in (Zhang et al. 2012)). In brief, small molecules (classified as less than 800 (Dougherty et al. 2012) or 500 Da (Lipinski 2003)) are relatively cheap to produce and require relatively simple storage and quality control requirements, compared to other reagents such as recombinant proteins or synthetic RNAs. Their molecular weight limit allows oral bioavailability, which is advantageous for subsequent drug development. Moreover, an individual small molecule has the potential to produce numerous effects in the target cell, via binding to multiple protein targets (for example, retinoic acid, which targets different nuclear receptors). Prominent examples of small molecules that are employed in the production of induced stem cells include RepSox, an inhibitor of transforming growth factor-β, which can substitute for Sox-2 in mouse embryonic fibroblast (MEF) reprogramming to induced pluripotent stem cells (iPSCs), and kenpaullone, which can replace Klf4 in MEF reprogramming (Lyssiotis et al. 2011; Ichida et al. 2009). Currently, there are numerous small molecules that can facilitate the iPSC reprogramming process or completely substitute for the Yamanaka transcription factors (reviewed in (Jung et al. 2014c; Yu et al. 2014; Lin and Wu 2015)).
In the case of small molecule-based methodologies for stem cell induction, an ultimate aim is to provide a source of precursor cells that can be used to treat degenerative diseases, such as Alzheimer’s disease or heart failure. Ideally, these small molecules could even be administered as drugs that directly induce tissue repair in vivo (discussed in (Langle et al. 2014)). However, recently, it has become apparent that some of these reprogramming compounds have the potential to be repositioned for pre-clinical development as anti-cancer agents. This may appear counterintuitive, because cancer progression also involves the acquirement of stem cell characteristics (Hernandez-Vargas et al. 2009). However, the target mechanisms of certain cell reprogramming molecules are also relevant for carcinogenesis. Numerous examples have been reported, such as the histone deacetylase inhibitor, valproic acid and the glycogen synthase kinase-3β (GSK-3β) inhibitor, SB-216763 (De Souza and Chatterji 2015; Mazor et al. 2004). A discussion of all of these molecules would be beyond the scope of this commentary. Herein, we focus on two interesting examples: 6-bromoindirubin-3′-oxime (BIO) and 2-(4-morpholinoanilino)-6-cyclohexylaminopurine (reversine).
Reversine: synthetic purine that targets cell division for epigenetic reprogramming and anti-cancer activity
Further analysis of the biological mechanism of reversine revealed that the active target may not be non-muscle myosin II heavy chain and MEK1, but rather inhibition of aurora kinases A and B, which localize to the centrosome during mitosis and carry out pivotal functions during cell division (Amabile et al. 2009). Inhibition of aurora kinases results in abnormal formation of the mitotic spindle, improper alignment of segregating chromosomes, and reduced phosphorylation of the histone H3 target (D’Alise et al. 2008; Santaguida et al. 2010). These effects also suggest a potential mechanism for reversine-induced stem cell reprogramming: the epigenetic remodeling of chromatin structure.
Aurora kinase is also an anti-cancer target (D’Assoro et al. 2015). Previously characterized aurora kinase inhibitors, such as VX-680, have been used in clinical trials for cancer treatment and shown treatment efficiency (Cheung et al. 2014). The anti-cancer activity of reversine was first demonstrated in a panel of human cancer cell types, such as HeLa cervical carcinoma, PC-3 bladder adenocarcinoma, and primary acute myeloid leukemia (D’Alise et al. 2008; Hsieh et al. 2007). Moreover, reversine was shown to inhibit signaling mediated by focal adhesion kinase (Bijian et al. 2013), which is a key regulator of cancer cell invasion and migration (Avallone et al. 2015). It has been demonstrated that the reversine target, aurora kinase, also regulates focal adhesion kinase activity (Romain et al. 2014).
Box 1: Summary of current knowledge about the use of reversine in cancer research
(1) Induction of autophagy in human follicular thyroid cancer cells via inhibition of Akt/mTOR/p70S6 K-related pathways (Lu et al.)
(2) Preferentially cytotoxic for p53-deficient cancer cells (Jemaa et al.)
(3) Suppresses breast cancer tumor growth and metastasis in vivo by reducing tumor stromalization (collagen deposition, recruit activated stromal cells) (Shen et al.)
(4) Blocking human breast cancer cell proliferation via cell cycle arrest, induction of polyploidy, and apoptosis (Kuo et al.)
(5) Suppression of oral squamous cell proliferation by cell cycle arrest and induction of autophagy via inhibition of Akt/mTORC1 (Lee et al.)
(6) Inhibition of differentiated and undifferentiated thyroid cancer cell proliferation via cell cycle arrest or apoptosis (Hua et al.)
(7) Synergy with aspirin for growth inhibition and apoptosis in human cervical cancers cells (Qin et al.)
(8) Inhibition of focal adhesion disassembly and turnover to reduce breast cancer cell migration via focal adhesion kinase inhibition (Bijian et al.)
(9) Specific anti-cancer effect in various cancer cell lines via cell cycle arrest and induction of apoptosis; not observed in normal fibroblasts (Piccoli et al.)
(10) Inhibition of protein kinase monopolar spindle 1 (MPS1) to preferentially kill tetraploid tumor cells (Jemaa et al.)
(11) Induction of growth arrest and polyploidy in human cancer cell lines via increased expression of p21 (WAF1)/down-regulation of cyclin B and CDK1 (Hsieh et al. 2007)
(12) Inhibition of colony formation in human acute myeloid leukemia (D’Alise et al. 2008)
(13) Modulation of tumor cell-stroma interactions to reduce the development of chemoresistance (McMillin et al. 2010)
BIO: a mollusk-derived compound for stem cell renewal, cardiogenesis, and inhibiting tumorigenesis
The small molecule, BIO ((2′Z,3′E)-6-bromoindirubin-3′-oxime; Fig. 2), was first characterized in 2003 (Meijer et al. 2003). It is a cell permeable derivative of the natural product, 6-bromoindirubin that is produced by predatory rock snails, such as Hexaplex trunculus, which was used in ancient times to produce the highly prized “Tyrian purple” dye. BIO was shown to selectively inhibit the multifunctional enzyme, GSK-3β, leading to activation of the Wnt signaling pathway. This pathway maintains the undifferentiated state of stem cells (Willert et al. 2003), and it was shown that BIO treatment produced developmental defects in zebrafish embryos (Meijer et al. 2003). Subsequently, it was demonstrated that BIO treatment could maintain the pluripotency of stem cell cultures and prevent spontaneous cell differentiation (Sato and Brivanlou 2006; Nagai et al. 2014; Holmes et al. 2008). Notably, it was shown that periodic activation of Wnt signaling by BIO treatment enhanced somatic cell reprogramming to pluripotent stem cells using cell fusion methodology (Lluis et al. 2008). GSK-3β inhibition by BIO is also an important component of small molecule-based approaches to derive functional cardiomyocytes from embryonic stem cells and iPSCs, in which it is employed at an early stage to derive mesodermal lineage cells (Naito et al. 2006; Jung et al. 2014c). Interestingly, BIO treatment could also induce proliferation in post-mitotic adult cardiomyocytes, which involved Wnt signaling activation and down-regulation of the cell cycle inhibitor, p27 (cyclin-dependent kinase inhibitor 1B/ Kip1) (Tseng et al. 2006).
Although BIO was initially characterized as a specific inhibitor of GSK-3β, further research revealed that this compound has activity against the Janus activated kinase/signal transducer and activator of transcription 3 (JAK/STAT3) signaling protein (Liu et al. 2011). In cancer cells, the JAK/STAT3 pathway is activated and has been shown to contribute to carcinogenesis, providing a promising drug target (Ghoreschi et al. 2009). In this study, inhibition of JAK/STAT3 induced apoptosis via down-regulation of the anti-apoptosis factor, Mc-1. The in vivo effect of BIO against tumorigenesis in melanoma cells was confirmed in a mouse xenograft model. Of note, BIO was effective in this study using oral delivery (50 mg/kg daily).
Metastasis is the main cause of cancer mortality (Vatandoust et al. 2015). Thus, drugs that can block the metastatic spread of cancer cells would have a major impact on cancer patient survival. Employing the 4T1 mouse model of aggressive breast cancer, it was observed that 1 mg/kg BIO pre-treatment dramatically reduced lung metastasis 8 days after intravenous delivery of cancer cells (Braig et al. 2013). BIO treatment reduced the cell invasiveness by down-regulating expression of the pro-migratory factors, C-terminal tensin-like protein, and matrix metalloproteinase 2, which was also linked to inhibition of the JAK/STAT3 pathway. Using RNAi-mediated reduction of target gene expression, it shown that BIO inhibited the signaling molecule 3-phosphoinositide-dependent protein kinase-1 (PDK1), in addition to JAK/STAT3 and GSK-3β, to produce this anti-metastatic effect. Thus, BIO can be considered as an interesting example of “polypharmacology” (drugs that affect multiple targets or disease-related pathways) and illustrates one of the advantages of developing pleiotropic small molecules as pharmaceutical candidates.
This concept of BIO as a polypharmacology agent for anti-cancer therapy was reiterated in a recent study of drug resistance in cancer cells. Tumor necrosis factor (TNF) α-related apoptosis inducing ligand (TRAIL) is an inducer of cell death that has been shown to induce apoptosis in a variety of different cancer cell types without affecting normal cells (Fesik 2005). However, many tumors, such as breast and bladder cancer, develop resistance to TRAIL-induced death, which is linked to survival mechanisms, such as the BIO target, GSK-3β (Koschny et al. 2007). Using concentrations of BIO that are not cytotoxic for breast and bladder cancer cells, it was shown that the TRAIL pathway became reactivated and the cells were sensitized to TRAIL-induced apoptosis (Braig et al. 2014). Thus, BIO could be employed in combined therapy with TRAIL-inducing agents, such as SuperKillerTRAIL, to overcome chemoresistance in refractory tumors.
Box 2: Summary of current knowledge about the use of BIO in anti-cancer research
(1) Reduction of melanoma cell proliferation and migration, without affecting invasion, chemotoxicity, or apoptosis (Chon et al.)
(2) Induction of human melanoma cell apoptosis by functioning as a pan-JAK inhibitor selectively inhibiting STAT3 signaling (Liu et al.)
(3) Reduction of migration and promotion of the cytoskeletal rearrangement of stress fibers and focal adhesions in pediatric glioma (Cockle et al.)
(4) Suppression of ovarian cancer cell development via up-regulation of p21 expression (Yu and Zhao)
(5) Simultaneous inhibition of JAK/STAT3, PDK1, and GSK-3β to induce anti-metastatic activity in vivo (Braig et al.)
(6) Up-regulation of p21 to induce G2/M cell cycle arrest and activate caspase-dependent and caspase-independent apoptosis in invasive breast cancer cells (Nicolaou et al.)
(7) Reduction of pro-tumorigenic telomerase activity via modulation of multiple gene regulatory networks in a mathematical model; validated in vivo (Bilsland et al.)
(8) Reduction of osteolytic regions in multiple myeloma via targeting osteogenesis in bone marrow mesenchymal stem cells (Gunn et al.)
(9) Augmentation of TRAIL-induced apoptosis in various cancer cell lines (Braig et al.)
(10) Preservation of hematopoietic stem cell activity and inhibition of leukemic cell growth via down-regulation of survivin (Holmes et al. 2008)
(11) Suppression of cell growth and induction of apoptosis in human leukemia cell lines of diverse origin via modulation of the cell death regulator, Bcl-2 (Song et al. 2010)
(12) Modulation of MYCN expression to inhibit neuroblastoma cell viability via multiple pathways (Duffy et al. 2014)
The zebrafish model: swimming into view as a simple and powerful method to validate anti-cancer activity in vivo
As mentioned above, the cell reprogramming compounds, BIO and reversine, have both been shown to possess anti-cancer activity that was validated in animal models. In both cases, this was reported some years after initial characterization of the compound. However, rapid and experimentally convenient animal models have been established for anti-cancer analysis, which could allow “in-house” testing for repositioning, rapid publication, and patenting. An attractive model is the zebrafish cancer xenograft system, which has multiple advantages, such as simple housing requirements, rapid development, transparency for microscopic analysis of organ systems, and high genetic homology to humans compared with other non-mammalian models (approximately 80 % for zebrafish, compared to ≈60 % for the fruit fly, Drosophila melanogaster, and ≈36 % for the roundworm, Caenorhabditis elegans) (Mackay and Anholt 2006; Barbazuk et al. 2000; C.elegans 1998). These features have allowed the zebrafish to be used for studying pivotal aspects of carcinogenesis and metastasis (Ignatius et al. 2012; Blackburn et al. 2014). For example, zebrafish, mice, and humans develop tumors that show histological and genetic similarities. APC mutant tumors from these three species all form in the liver and intestine and show constitutive activation of Wnt signaling (Haramis et al. 2006).
Summary and future perspectives
In the commentary, we have discussed the link between stem cell reprogramming small molecules and their potential repositioning as anti-cancer compounds. Two examples were discussed: the compounds BIO and reversine, which were initially characterized as stem cell reprogramming agents and have been subsequently repositioned and patented as anti-cancer agents. We also present the zebrafish human tumor xenograft model as a rapid validation system for testing anti-cancer candidate compounds. The use of small molecules to control biological systems is increasing in scope, and this is especially relevant for the stem cell biology field. Very recently, novel small molecule cocktails have been developed that allow the generation of chemically iPSCs (ciPSCs), which are produced solely by small molecule treatment (Hou et al. 2013; Zhao et al. 2015; Ye et al. 2016). Novel small molecules and optimized small molecule-based methodologies for somatic cell reprogramming into stem cells are in continuous development and now aim to directly modulate patient cells in vivo (reviewed in (Anwar et al. 2016; Davies et al. 2015)). Many of these small molecules modulate cellular targets that are also linked to carcinogenesis, such as aurora kinases, GSK-3β, JAK/STAT3, and histone deacetylases.
It is now established that cancer cells possess many similar properties to normal stem cells (Hong et al. 2015). Therefore, it may seem counterintuitive that small molecules modulating stem cell phenotype can also possess anti-cancer activity. However, these anti-cancer effects can be explained by the pleiotropic nature of bioactive small molecules and the importance of cell context. In the case of the aurora kinase inhibitor, reversine, communication between cancer cells and stromal fibroblasts is disrupted, leading to reduced stromalization in developing tumors. BIO was initially characterized as a GSK-3β inhibitor, but the anti-cancer activity of BIO was discovered to be additionally related to the inhibition of Jak/STAT3 signaling. Thus, the cellular and disease scenario are pivotal influences on small molecule activity and their potential repositioning. Given the current high priority placed on compound repositioning in the drug development and discovery process, especially in cancer therapeutics (Wurth et al. 2016), this may encourage researchers developing small molecules in stem cell research to also consider the possible value of their compounds as anti-cancer agents. Repositioning within the same research institution may also simplify any intellectual property issues surrounding the original compound.
Currently, the drug discovery pipeline is contracting (Bailey et al. 2014; Dean et al. 2014; Spellberg et al. 2015). Strategies that facilitate the repositioning and development of novel small molecule therapeutics are attractive strategies to tackle this problem because other technologies, such as gene therapy and cell therapy, have not yet realized their full potential (Hulot et al. 2016; Lysy et al. 2016). In this commentary, we have discussed the investigation of anti-cancer activity in cell reprogramming compounds, based on the “traditional” laboratory approach of cell-based assays and animal model testing. However, there are alternative strategies to facilitate repositioning, such as in silico-based methodologies utilizing large-scale virtual screening of compound libraries or compounds approved for human use against large numbers of protein targets (high-throughput shotgun repurposing) (Wang et al. 2013). We hope that this commentary illustrates the link between some cell reprogramming compounds and potential anti-cancer activity, based on the modulation of target proteins that are important regulators in both cell reprogramming and carcinogenesis.
This work was supported by the Bio & Medical Technology Development Program and Basic Science Research Program of the NRF funded by the Korean government, MSIP (NRF-2015M3A9C6030838 and NRF-2015R1A2A2A11001597). This work was also supported by the GIST Research Institute (GRI) in 2016.
- Bijian K, Lougheed C, Su J, Xu B, Yu H, Wu JH, Riccio K, Alaoui-Jamali MA. Targeting focal adhesion turnover in invasive breast cancer cells by the purine derivative reversine. Br J Cancer. 2013;109:2810–8.Google Scholar
- Bilsland AE, Stevenson K, Liu Y, Hoare S, Cairney CJ, Roffey J, Keith WN. Mathematical model of a telomerase transcriptional regulatory network developed by cell-based screening: analysis of inhibitor effects and telomerase expression mechanisms. PLoS Comput Biol. 2014;10:e1003448.Google Scholar
- Braig S, Bischoff F, Abhari BA, Meijer L, Fulda S, Skaltsounis L, Vollmar AM. The pleiotropic profile of the indirubin derivative 6BIO overcomes TRAIL resistance in cancer. Biochem Pharmacol. 2014;91:157-67.Google Scholar
- Braig S, Kressirer CA, Liebl J, Bischoff F, Zahler S, Meijer L, et al. Indirubin derivative 6BIO suppresses metastasis. Cancer Res. 2013;73:6004–12.Google Scholar
- Chon E, Flanagan B, De Sa Rodrigues LC, Piskun C, Stein TJ. 6-Bromoindirubin-3’oxime (BIO) decreases proliferation and migration of canine melanoma cell lines. Vet J, 2015;205:305-12.Google Scholar
- Cockle JV, Picton S, Levesley J, Ilett E, Carcaboso AM, Short S, Steel LP, Melcher A, Lawler SE, Bruning-Richardson A. Cell migration in paediatric glioma; characterisation and potential therapeutic targeting. Br J Cancer. 2015;112:693-703.Google Scholar
- Dougherty TJ, Pucci MJ, Macielag M. Chemical properties of antimicrobials and their uniqueness. New York: Antibiotic Discovery and Development, Springer; 2012.Google Scholar
- Gunn WG, Krause U, Lee N, Gregory CA. Pharmaceutical inhibition of glycogen synthetase kinase-3beta reduces multiple myeloma-induced bone disease in a novel murine plasmacytoma xenograft model. Blood. 2011;117:1641-51.Google Scholar
- Gunn WG, Conley A, Deininger L, Olson SD, Prockop DJ, Gregory CA. A crosstalk between myeloma cells and marrow stromal cells stimulates production of DKK1 and interleukin-6: a potential role in the development of lytic bone disease and tumor progression in multiple myeloma. Stem Cells. 2006;24:986–91.CrossRefPubMedGoogle Scholar
- Gurdon JB, Wilmut I. Nuclear transfer to eggs and oocytes. Cold Spring Harb Perspect Biol. 2011;3.Google Scholar
- Hua SC, Chang TC, Chen HR, Lu CH, Liu YW, Chen SH, Yu HI, Chang YP, Lee YR. Reversine, a 2,6-disubstituted purine, as an anti-cancer agent in differentiated and undifferentiated thyroid cancer cells. Pharm Res. 2012;29:1990-2005.Google Scholar
- Hulot JS, Ishikawa K, Hajjar RJ. Gene therapy for the treatment of heart failure: promise postponed. Eur Heart J. 2016. doi:10.1093/eurheartj/ehw019
- Jemaa M, Galluzzi L, Kepp O, Boileve A, Lissa D, Senovilla L, Harper F, Pierron G, Berardinelli F, Antoccia A, Castedo M, Vitale I, Kroemer G. Preferential killing of p53-deficient cancer cells by reversine. Cell Cycle. 2012;11:2149-58.Google Scholar
- Jemaa M, Manic G, Lledo G, Lissa D, Reynes C, Morin N, Chibon F, Sistigu A, Castedo M, Vitale I, Kroemer G, Abrieu A. Whole-genome duplication increases tumor cell sensitivity to MPS1 inhibition. Oncotarget. 2016;7:885-901.Google Scholar
- Krause MN, Sancho-Martinez I, Izpisua Belmonte JC. Understanding the molecular mechanisms of reprogramming. Biochem Biophys Res Commun. 2015;473:693–697.Google Scholar
- Kuo CH, Lu YC, Tseng YS, Shi CS, Chen SH, Chen PT, Wu FL, Chang YP, Lee YR. Reversine induces cell cycle arrest, polyploidy, and apoptosis in human breast cancer cells. Breast Cancer. 2014;21:358-69.Google Scholar
- Lee YR, Wu WC, Ji WT, Chen JY, Cheng YP, Chiang MK, Chen HR. Reversine suppresses oral squamous cell carcinoma via cell cycle arrest and concomitantly apoptosis and autophagy. J Biomed Sci. 2012;19:9.Google Scholar
- Liu L, Nam S, Tian Y, Yang F, Wu J, Wang Y, Scuto A, Polychronopoulos P, Magiatis P, Skaltsounis L, Jove R. 6-Bromoindirubin-3’-oxime inhibits JAK/STAT3 signaling and induces apoptosis of human melanoma cells. Cancer Res. 71:3972-9.Google Scholar
- Lu CH, Liu YW, Hua SC, Yu HI, Chang YP, Lee YR. Autophagy induction of reversine on human follicular thyroid cancer cells. Biomed Pharmacother. 2012;66:642-7.Google Scholar
- Nicolaou KA, Liapis V, Evdokiou A, Constantinou C, Magiatis P, Skaltsounis AL, Koumas L, Costeas PA, Constantinou AI. Induction of discrete apoptotic pathways by bromo-substituted indirubin derivatives in invasive breast cancer cells. Biochem Biophys Res Commun. 2012;425:76-82.Google Scholar
- Parker WB, Secrist 3RD JA, Waud WR. Purine nucleoside antimetabolites in development for the treatment of cancer. Curr Opin Investig Drugs. 2004;5:592–6.Google Scholar
- Piccoli M, Palazzolo G, Conforti E, Lamorte G, Papini N, Creo P, Fania C, Scaringi R, Bergante S, Tringali C, Roncoroni L, Mazzoleni S, Doneda L, Galli R, Venerando B, Tettamanti G, Gelfi C, Anastasia L. The synthetic purine reversine selectively induces cell death of cancer cells. J Cell Biochem. 2012;113:3207-17.Google Scholar
- Polakis P. Wnt signaling in cancer. Cold Spring Harb Perspect Biol. 2012;4.Google Scholar
- Qin HX, Yang J, Cui HK, Li SP, Zhang W, Ding XL, Xia YH. Synergistic antitumor activity of reversine combined with aspirin in cervical carcinoma in vitro and in vivo. Cytotechnology. 2013;65:643-53.Google Scholar
- Shen K, Luk S, Hicks DF, Elman JS, Bohr S, Iwamoto Y, Murray R, Pena K, Wang F, Seker E, Weissleder R, Yarmush ML, Toner M, Sgroi D, Parekkadan B. Resolving cancer-stroma interfacial signalling and interventions with micropatterned tumour-stromal assays. Nat Commun. 2014;5:5662.Google Scholar
- Trede NS, Heaton W, Ridges S, Sofla H, Cusick M, Bearss D, Jones D, Fujinami RS. Discovery of biologically active oncologic and immunologic small molecule therapies using zebrafish: overview and example of modulation of T cell activation. Curr Protoc Pharmacol. 2013. Chapter 14, Unit14 24.Google Scholar
- Tulotta C, Stefanescu C, Beletkaia E, Bussmann J, Tarbashevich K, Schmidt T, et al. Inhibition of signaling between human CXCR4 and zebrafish ligands by the small molecule IT1t impairs the formation of triple-negative breast cancer early metastases in a zebrafish xenograft model. Dis Model Mech. 2016;9:141–53.CrossRefPubMedPubMedCentralGoogle Scholar
- Yu AS, Zhao L. Effects of the GSK-3beta inhibitor (2Z,3E)-6-bromoindirubin-3’-oxime upon ovarian cancer cells. Tumour Biol. 2016;37:4857–64Google Scholar