Mutation Research 771 (2015) 56–69

Contents lists available at ScienceDirect

Mutation Research/Fundamental and Molecular Mechanisms of Mutagenesis journal homepage: www.elsevier.com/locate/molmut Community address: www.elsevier.com/locate/mutres

Step-wise and punctuated genome evolution drive phenotype changes of tumor cells Aleksei Stepanenko a,∗ , Svitlana Andreieva a , Kateryna Korets a , Dmytro Mykytenko a , Nataliya Huleyuk b , Yegor Vassetzky c , Vadym Kavsan a,1 a

Department of Biosynthesis of Nucleic Acids, Institute of Molecular Biology and Genetics, National Academy of Sciences of Ukraine, Kyiv 03680, Ukraine Institute of Hereditary Pathology, National Academy of Medical Sciences of Ukraine, Lviv 79008, Ukraine c CNRS UMR8126, Université Paris-Sud 11, Institut de Cancérologie Gustave Roussy, Villejuif 94805, France b

a r t i c l e

i n f o

Article history: Received 8 July 2014 Received in revised form 14 December 2014 Accepted 18 December 2014 Available online 27 December 2014 We would like to acknowledge Prof. Vadym Kavsan who passed away on October, 2014 and dedicate the paper to his memory. Keywords: Chromosome instability Aneuploidy Cancer gene Tumor heterogeneity YKL-40

a b s t r a c t The pattern of genome evolution can be divided into two phases: the step-wise continuous phase (stepwise clonal evolution, stable dominant clonal chromosome aberrations (CCAs), and low frequency of non-CCAs, NCCAs) and punctuated phase (marked by elevated NCCAs and transitional CCAs). Depending on the phase, system stresses (the diverse CIN promoting factors) may lead to the very different phenotype responses. To address the contribution of chromosome instability (CIN) to phenotype changes of tumor cells, we characterized CCAs/NCCAs of HeLa and HEK293 cells, and their derivatives after genotoxic stresses (a stable plasmid transfection, ectopic expression of cancer-associated CHI3L1 gene or treatment with temozolomide) by conventional cytogenetics, copy number alterations (CNAs) by array comparative genome hybridization, and phenotype changes by cell viability and soft agar assays. Transfection of either the empty vector pcDNA3.1 or pcDNA3.1 CHI3L1 into 293 cells initiated the punctuated genome changes. In contrast, HeLa CHI3L1 cells demonstrated the step-wise genome changes. Increased CIN correlated with lower viability of 293 pcDNA3.1 cells but higher colony formation efficiency (CFE). Artificial CHI3L1 production in 293 CHI3L1 cells increased viability and further contributed to CFE. The opposite growth characteristics of 293 CHI3L1 and HeLa CHI3L1 cells were revealed. The effect and function of a (trans)gene can be opposite and versatile in cells with different genetic network, which is defined by genome context. Temozolomide treatment of 293 pcDNA3.1 cells intensified the stochastic punctuated genome changes and CNAs, and significantly reduced viability and CFE. In contrast, temozolomide treatment of HeLa CHI3L1 cells promoted the step-wise genome changes, CNAs, and increased viability and CFE, which did not correlate with the ectopic CHI3L1 production. Thus, consistent coevolution of karyotypes and phenotypes was observed. CIN as a driving force of genome evolution significantly influences growth characteristics of tumor cells and should be always taken into consideration during the different experimental manipulations. © 2014 Elsevier B.V. All rights reserved.

1. Introduction Chromosome instability (CIN) refers to the rate of genome changes in a given cell population (cell-to-cell variation) and implies clonal chromosome aberrations (CCAs) and non-clonal chromosome aberrations (NCCAs). Generally, CIN correlates with

Abbreviations: aCGH, array comparative genome hybridization; CIN, chromosome instability; CCAs, clonal chromosome aberrations; CNAs, copy number alterations; NCCAs, non-clonal chromosome aberrations; CFE, colony formation efficiency; TMZ, temozolomide. ∗ Corresponding author. Tel.: +380 44 526 34 98; fax: +380 44 526 07 59. E-mail address: [email protected] (A. Stepanenko). 1 Deceased. http://dx.doi.org/10.1016/j.mrfmmm.2014.12.006 0027-5107/© 2014 Elsevier B.V. All rights reserved.

tumorigenic potential of cells, tumor progression, patient survival, treatment sensitivity, and the risk of acquired therapy resistance [1–9]. The studies of the evolutionary process of cancer [4–6,10] revealed that the pattern of genome evolution can be divided into two phases: the Darwinian step-wise continuous phase (marked by step-wise clonal evolution, stable dominant CCAs, and low frequency of NCCAs) and the discontinuous/punctuated phase (marked by elevated NCCAs and transitional CCAs). Shifts between phases are induced by system stress. The high genome-level heterogeneity in the punctuated phase provides the genetic underpinnings of high heterogeneity and clonal diversity universally detected in cancers. NCCAs reflect system instability and drive cancer evolution by increasing population diversity, whereas CCAs actually represent chromosome stability. Furthermore, genome

A. Stepanenko et al. / Mutation Research 771 (2015) 56–69

heterogeneity in cell population was linked to tumorigenicity [11–13]. Tumorigenicity of cell lines correlated with level of genome heterogeneity regardless of the genetic background of cells. All lines with low tumorigenicity displayed distinctly lower frequency of structural NCCAs [11]. The changing NCCAs/CCAs pattern is a driving force, which ensures genetic heterogeneity, phenotypic plasticity, and population diversity that is essential to cancer evolution. The sequencing studies confirmed that cancer evolution occurs as multiple cycles of the stepwise and punctuated branched polyclonal evolution [14,15]. Depending on the phase, system stresses (the diverse CIN promoting factors) may lead to the very different responses [6]. During the step-wise phase, additional CIN destabilizes the established CCAs and abolishes growth advantage that stemmed from the initial CCAs. Selection and spread throughout the population of the advantageous chromosome aberrations result eventually in fixation of the newly formed dominant CCAs able to continue cancer evolution. During the punctuated phase, additional CIN results in even more increased instability. In such highly stressful conditions most of unstable cells are not viable [6]. Actually, high-level aneuploidy has a negative impact on cellular fitness and generates non-neoplastic and nonviable cells [16,17]. However, chaotic chromosome changes increase the population genome diversity, probability of emergence of the advantageous CCAs, and evolutionary potential of tumor. These phases of cancer cell evolution may shed light on the tumor-promoting and suppressing effects of CIN in tumorigenesis as well as the “paradoxical” relationship between excessive CIN and improved survival outcome in cancer [18,19]. Many cancer-associated genes showed both tumor promoting and suppressing effects (antagonistic functional duality) in the diverse cancer models [20]. The cell type-dependent effects on proliferation, migration, anti-apoptosis, and tumor growth of CHI3L1 (encoding a secretory glycoprotein chitinase-3-like protein 1, alternatively known as YKL40) were also reported [21–24]. The effect and function of gene can be opposite and versatile in cells with different genomes as gene function is dependent on the genetic network (gene content, RNA and protein expression and their interaction), which is defined by genome context (a number and structure of chromosomes and their nucleus topology) [5,6]. Temozolomide (TMZ) is the cytotoxic alkylating agent for treatment of patients with glioma, melanoma, lymphoma, pancreatic cancer and other types of cancer. TMZ induces methylation of adenine and guanine. If unrepaired, methylguanine mispaires with thymidine, which is recognized by mismatch repair system (MMR). MMR performs errors-prone futile repair cycles resulting in the formation of the secondary lesions, which block DNA replication in the next replication cycle. This leads eventually to DNA doublestrand breaks [25,26]. Therapy-mediated stress can significantly change tumor evolution by generating novel phenotypes through induction of genome changes. For example, analysis of chromosomal changes, tumor cell aggressiveness, and chemosensitivity of cell lines established from primary glioma tumors and consecutive recurrences developed under therapy showed genome and phenotype evolution [27]. Similarly, analysis of low-grade gliomas and recurrent TMZ-resistant tumors revealed TMZ-driven hypermutation and evidence of evolution to more aggressive high-grade gliomas [28]. To address these important issues and establish the correlation between changes of karyotype and phenotype, we tracked CCAs/NCCAs pattern, copy number alterations (CNAs), and growth characteristics of 293 and HeLa cell lines after genotoxic stresses (transfection of the empty vector pcDNA3.1, pcDNA3.1 CHI3L1, or treatment with TMZ). 293 and HeLa cell lines were reported to be characterized by relatively high karyotype stability during the longterm culture passaging [29,30]. Both cell lines were authenticated (Fig. S1).

57

2. Materials and methods 2.1. Cell lines Cell lines were grown in DMEM (HyClone, Thermo Scientific, Logan, UT, USA) supplemented with 10% FBS (HyClone) and 100 U/ml penicillin/100 ␮g/ml streptomycin (Sigma, USA) in an environment of 95% air/5% CO2 . 293 (HEK293), 293T, and HeLa cell lines were generously gifted by Dr. I. Gout (Department of Structural and Molecular Biology, University College London, London, UK). 293 cells stably transfected with pcDNA3.1 (293 pcDNA3.1 variant 1, 70 passages) and 293 cells stably transfected with CHI3L1 cDNA in pcDNA3.1 vector (293 CHI3L1, 70 passages) were kindly provided by Dr. A. Iershov, clones of HeLa CHI3L1 cells by Dr. O. Balynska (Department of Biosynthesis of Nucleic Acids, Institute of Molecular Biology and Genetics, IMBG, Kiev, Ukraine), and 293 pcDNA3.1 (variant 2) cells by Dr. V. Grishkova (Department of Cell Signaling, IMBG). HeLa CHI3L1 (clone 1) and (clone 2) were derived from HeLa CHI3L1 heterogeneous cell culture on 20th passage of culturing after pcDNA3.1 CHI3L1 vector transfection (Dr. O. Balynska). 0.8 mg/ml geneticin G418 sulphate (Sigma, St. Louis, MO, USA) was used to select stable cell lines. Antibiotic G418 was withdrawn from cell lines before and during in vitro tests. U87 cell line was kindly gifted by Dr. M. Sanson (INSERM, U711, Biologie des Interactions Neurones and Glie, Paris, France). U251 cell line was received from the Bank of Cell Lines from Human and Animal Tissues, R.E. Kavetsky Institute of Experimental Pathology, Oncology and Radiobiology (Kyiv, Ukraine). 2.2. Long-term treatment with temozolomide (TMZ) 293 pcDNA3.1 (variant 2) and HeLa CHI3L1 (clone 2) cells were treated with 100% DMSO-dissolved TMZ (Sigma) during 10 weeks as follows: 10, 20, 40, 60, 80, and 100 ␮M TMZ for 6 weeks with each concentration two times per week, then 120 ␮M TMZ two times per week for other 4 weeks, followed by 1 month of washout (TMZ-free medium) before in vitro tests. In culture plates DMSO did not exceed 0.4% by volume. 2.3. Conventional cytogenetics Chromosome samples were prepared by accumulation of metaphases in the presence of 0.5 ␮g/ml colcemide for 1.5–2 h, followed by treatment with 0.075 M potassium chloride (20 min) and fixation in methanol:acetic acid (3:1, v/v) two times for 30 min each. Fixed nuclei were spread on the wet glass slides (Marienfeld, Germany), air dried and then placed in 5% Giemsa stain (Merck, USA) plus trypsin (HyClone, ThermoScientific, USA) for 4 min. Images of metaphase plates were captured with Olympus BX40 microscope (Japan) and evaluated with Lucia computer analysis system for karyotyping, version 1.6.1 (Laboratory Imaging, Praha, Czech Republic). 100–200 well-spread metaphases were scored per variant to determine chromosome modal number and 20 metaphases were described for chromosome abnormalities. Karyotypes were described according to the International System for Human Cytogenetic Nomenclature (ISCN 2009). Modal chromosome number was defined as the most frequent number of chromosomes per cell found across all metaphases studied in a given cell population. CCAs were defined as aberrations found at least in two cells among 20 examined metaphases, whereas NCCAs as aberrations detected in only one cell. Dominant CCAs were ascribed to aberrations with ≥50% clonality (10 or more cells harbored). The frequency of NCCAs in cell line was calculated by using all cells displaying NCCAs divided by the total number of examined cells. Only structural NCCAs were considered.

58

A. Stepanenko et al. / Mutation Research 771 (2015) 56–69

2.4. Array comparative genome hybridization (aCGH)

DNA ladder (GeneRuler, ThermoScientific, USA) and visualized with ethidium bromide staining under UV light.

Total DNA was isolated using NucleoSpin® Blood DNA extraction kit (Macherey-Nagel, Germany) according to the manufacturer’s instructions. DNA purity was assessed using NanoDrop 1000 Spectrophotometer (Thermo Scientific, USA). 1.5 ␮g of each test and control DNA (Human Genomic DNA: Male, Promega, USA) sample was digested with AluI and RsaI enzymes (Promega, USA). They were then labeled by random priming using a CytoSure Genomic DNA Labeling Kit (Oxford Gene Technologies, UK) with the appropriate cyanine dye (Cy3 and Cy5, respectively) according to the manufacturer’s protocol. The labeled samples were purified by Clean-up Columns (Oxford Gene Technologies, UK) followed by combination of Cy3- and Cy5-laybeled samples, condensation (Concentrator Plus, Eppendorf, Germany) and denaturation with hybridization buffer. Competitive co-hybridization was performed on the CytoSure Aneuploidy Array 15k (Oxford Gene Technologies, UK). After 24-hour hybridization at 65 ◦ C, slides were washed with Wash 1 and Wash 2 buffers (Agilent, USA), dried, and scanned using Innocsan 710 scanner (Innopsys, France) at 5 nm resolution. Image analysis was carried out with CytoSure Analysis Software (Oxford Gene Technologies, UK). Threshold factor for losses was defined as 0.6, for gains – 0.3.

To analyze ectopic CHI3L1 production, cells were seeded into 6-well tissue culture plate in DMEM/10% FBS and allowed to grow to near-confluence. After denaturing in 200 ␮l loading 2× Laemmli sample buffer by heating at 99 ◦ C for 1 min, the whole cell lysates were separated by 10% SDS-polyacrylamide gel electrophoresis (30 ␮l of total cell lysate per gel lane) and transferred to Hybond nitrocellulose membrane (Amersham Biosciences, USA). The membranes were blocked in 5% non-fat milk for 1hr at room temperature, followed by incubation in primary antibody overnight. The antibodies were goat polyclonal antiCHI3L1, clone S-18 (1:2000 dilution, Santa Cruz Biotechnology, USA), and mouse monoclonal anti-␤-actin, clone AC15 (1:5000 dilution, Sigma), as well as HRP Conjugated anti-Mouse IgG and anti-Goat IgG (Invitrogen, USA). Tris-buffered saline (TBS) supplemented with 0.1% Tween-20 was used to dilute the antibodies and wash the membranes. Proteins were visualized using the ECL system (Amersham Biosciences, USA) according to the supplier’s instructions.

2.5. Cell viability assay

2.9. Statistics

Cells were seeded into 96 well plate at density 500 cells/well (or 2000 cells/well for 293pcTMZ1/2) and grown in DMEM supplemented with 10% FBS. Cells viability was measured using MTT reagent (Sigma) on 1, 2, 3, 5, 7 (and 9th where designated) days of seeding. On the day of measurement, medium was carefully replaced on fresh DMEM + 10% FBS with diluted MTT (1:10, stock 5 mg/ml), and incubated for 3.5 h. Color change was measured at 570 nm using ELx800 absorbance microplate reader (BioTek Instruments, Winooski, VT, USA). Test for each cell line was repeated at least six times in quadruplexes.

The Student’s t-test was used to analyze data sets. Results with *P ≤ 0.05, **P ≤ 0.01, and ***P ≤ 0.001 were considered significant. All experimental data are reported as mean and the error bars represent the experimental standard error (±standard deviation, SD).

2.8. Western blot analysis

3. Results 3.1. Stable transfection of either the empty vector pcDNA3.1 or pcDNA3.1 CHI3L1 induced the punctuated genome evolution and phenotype changes of 293 cells

2.6. Soft agar colony formation assay 5 × 103 (or 2 × 103 for 293pcTMZ1/2) cells were placed in 1.5 ml of 0.35% low gelling temperature agarose (Gibco, Life Technologies, Grand Island, NY, USA) with DMEM supplemented with 10% FBS. 0.35% top agarose was poured on 1.5 ml of solidified 0.5% base agarose/10% FBS/DMEM. Cells were seeded in triplicates in a 35mm diameter dish and grown at 37 ◦ C for 21 days to allow colony formation. Colonies were visualized by staining with 0.005% crystal violet, photographed, counted using TotalLab software (Newcastle upon Tyne, UK), and expressed as means of triplicates of four independent experiments. 2.7. Polymerase chain reaction (PCR) analysis Total DNA was isolated using NucleoSpin Blood DNA extraction kit (Macherey-Nagel, Germany) according to the manufacturer’s instructions. PCR analysis in a total volume of 25 ␮l with 200 ng of total DNA was carried out in TPersonal Thermocycler (Biometra GmbH, Goettingen, Germany). The following primers were used: E1 AD5: forward 5 -TTATTACCGAAGAAATGGCCGC-3 and reverse 5 -CAAACATGCCACAGGTCCTCATATA-3 (Tm = 59 ◦ C); T-large SV40: forward 5 -TTAGCAATTCTGAAGGAAAGTCCTTG-3 and reverse 5 AGCAGTGGTGGAATGCCTTTCATGAGG-3 (Tm = 59 ◦ C). The thermal profiles were as follows: predenaturation for 2 min at 95 ◦ C, denaturation for 30 s at 95 ◦ C, annealing for 30 s at 57 ◦ C, elongation for 30–45 s at 72 ◦ C, and a final elongation for 5 min at 72 ◦ C. After 40 thermal cycles, the PCR products were resolved by gel electrophoresis in 1.5% agarose TAE gels in parallel with a 1-kb plus

To evaluate correlation between patterns of genome evolution and phenotype changes of cell lines, the following karyotype parameters of 20 metaphases were analyzed: total number of CCAs, line-specific CCAs, total number of dominant CCAs, total number of NCCAs, frequency of NCCAs, and variation of NCCAs between cells. Modal numbers of chromosomes in 293 pcDNA3.1 cells (variant 1) (68–70 chromosomes/cell), 293 pcDNA3.1 cells (variant 2) (68–72 chromosomes/cell), and 293 CHI3L1 cells (67–71 chromosomes/cell) were lower than that in parental 293 cells (72–76 chromosomes/cell) (Fig. 1A). 293 pcDNA3.1 cells (variant 1) and independently derived and longer passaged 293 pcDNA3.1 cells (variant 2) increased number of CCAs and number and frequency of NCCAs compared to 293 cells (Fig. 1B and C; Table S1). Both 293 pcDNA3.1 cell line variants acquired a number of new cell line specific CCAs indicating on random karyotype evolution after the same type of stress. 293 CHI3L1 cells also demonstrated acquisition of many CCAs as compared to 293 cells. The total CCAs number was comparable to both 293 pcDNA3.1 cell line variants, whereas the frequency and number of NCCAs in 293 CHI3L1 cells was close to 293 cells and less than in both 293 pcDNA3.1 variants (Fig. 1B and C; Table S1). In 293 cell line derivatives, the average number of markers (the highly structurally abnormal chromosomes, which are indicative of genome chaos and cannot be unambiguously characterized by conventional banding cytogenetics) increased 1.6–1.7 times per cell, underscoring aggravated genome chaos. Despite the obvious increase in whole and structural CIN, array comparative genome hybridization (aCGH) analysis showed

A. Stepanenko et al. / Mutation Research 771 (2015) 56–69

59

Fig. 1. Either stable transfection of the empty vector pcDNA3.1 or pcDNA3.1 CHI3L1 initiated the punctuated genome evolution of 293 cells. (A) A distribution of chromosomes across 100 metaphases analyzed for each designated cell line. (B) The karyographs show the degree of clonality and variability of chromosomes in each cell line by comparing the copy numbers of intact and abnormal chromosomes of 20 metaphases to each other. The karyotype differences between cell lines are visualized by alignment and comparison of karyographs for each cell line. All chromosomes depicted on the x-axis are listed in the table in the same order. (C) The table lists all CCAs and summarizes karyotypic parameters of each cell line. The most obvious differences in CCAs between cell lines are marked. NCCAs specific for each cell line are listed in Table S1.

60

A. Stepanenko et al. / Mutation Research 771 (2015) 56–69

Fig. 2. 293 cells derivatives have many common CNAs but essentially different phenotypes. (A) CNAs were identified by aCGH. Chromosomal ideograms with the drawn green and red bars aligned along chromosomes show the areas of copy number gain and loss, respectively. The chromosome CNAs loci are detailed in Table S2. (B) Comparison of viability curves derived from MTT assay. Data points represent mean (±SD) of six individual experiments in quadruplexes. (C) The representative photographs of plates with stained colonies grown in soft agar for three weeks. (D) Comparison of CFE in soft agar colony assay. Bar graph shows mean (±SD) of four individual experiments in triplicates. **P ≤ 0.01.

that majority of CNAs in 293 CHI3L1, 293 pcDNA3.1 variants, and parental 293 cells were common with minor cell line-specific CNAs (Fig. 2A and Table S2). Producing an average profile of genetic changes in a sample, aCGH did not mirror complexity and heterogeneity of changes revealed by karyotyping. 293 CHI3L1 viability was higher than those of both variants of 293 pcDNA3.1 or parental 293 cells. In turn, viability of both variants of 293 pcDNA3.1 cells was much lower than that of parental 293 cells (Fig. 2B). Colony formation efficiencies (CFEs) of 293 CHI3L1 cells and both variants of 293 pcDNA3.1 cells were significantly higher than that of parental 293 cells. 293 CHI3L1 cells gave the largest total number

of colonies followed by 293 pcDNA3.1 (variant 1) and then 293 pcDNA3.1 (variant 2). 293 cells formed large in size colonies, whereas their derivatives gave colonies mainly of a smaller size (Fig. 2C and D). Thus, transfection of plasmid DNA and selection of the stable cell lines resistant to cytotoxic antibiotic G418 induced significant changes of CCAs/NCCAs pattern and phenotype of 293 cells. Increased CIN correlated with lower viability but higher CFE. Artificial ectopic CHI3L1 production increased viability (counterbalancing the negative effect of elevated CIN) and further contributed to CFE.

A. Stepanenko et al. / Mutation Research 771 (2015) 56–69

3.2. The long-term TMZ-treatment of 293 pcDNA3.1 cells intensified the punctuated genome evolution, promoted CNAs and phenotype changes To reveal karyotype-phenotype correlation after additional genotoxic stress but to avoid interference of ectopic CHI3L1 production with phenotype, we used 293 pcDNA3.1 (variant 2) cells, which moreover had more stabilized karyotype than 293 pcDNA3.1 cells (variant 1). We established two sub-lines (namely 293pcTMZ1 and 293pcTMZ2) by repetitively exposing 293 pcDNA3.1 (variant 2) cells to the incremental concentrations of DNA-methylating drug TMZ followed by the one-month wash-out period after last treatment to capture a simulated “end run” of karyotype evolution after drug application (total additional 25–30 passages). Karyotype evolution of 293pcTMZ1 cells was accompanied by loss of 15 CCAs, acquisition of 11 new CCAs and high frequency and total number of NCCAs. Similarly, karyotype changes of 293pcTMZ2 cells was characterized by loss of 22 CCAs, acquisition of 15 new CCAs, and high but reduced total number and frequency of NCCAs (Fig. 3A and B, Table S3). Most of the newly acquired CCAs were individual in the TMZ-treated cell lines pointing to stochastic karyotype evolution under the same type of stress. The average number of marker chromosomes per cell was moderately increased for 293pcTMZ1 cells but significantly reduced for 293pcTMZ2. We have no logical explanation for the latter case as both 293pcTMZ1 and 293pcTMZ2 cells were characterized by massive CCAs/NCCAs, derived in parallel from the same initial culture, and the same drug treatment protocol was applied. Analysis of aCGH data showed individual CNAs in 293pcTMZ1 and 293pcTMZ2 cells, which were distinct from CNAs in 293 pcDNA3.1 (variant 2) (Fig. 4A). Viability and CFE of both 293pcTMZ1 and 293pcTMZ2 cells were decreased (Fig. 4B–D). Taken together, sequential application of genotoxic stresses of different nature (stable vector transfection and TMZ treatment) to 293 cells shifted genome evolution from the step-wise phase to the obvious punctuated phase with negative influence on viability and CFE. 3.3. Clones of HeLa cells transfected with pcDNA3.1 CHI3L1 demonstrated step-wise genome evolution, distinct CNAs and phenotype changes The modal chromosome numbers of both clones of HeLa CHI3L1 cells (69–73 and 72–74 chromosomes/cell, respectively) were higher than that of parental HeLa cells (68–71 chromosomes/cell) (Fig. 5A). Thus, there were the opposite patterns of whole CIN in derivatives of 293 and HeLa cell lines. The shorter passaged HeLa CHI3L1 cells (clone 1, 40 passages after pcDNA3.1 CHI3L1 transfection, 20 passages after cloning) had decreased CCAs number and absence of NCCAs versus parental HeLa cells. In contrast, HeLa CHI3L1 cells (clone 1) on passage 70 (after pcDNA3.1 CHI3L1 transfection) showed acquisition of 9 new CCAs and slightly increased number of NCCAs (Fig. 5B and C). This late, stepwise accumulation of chromosome aberrations may indicate the influence of intrinsic chromosome stability errors and CHI3L1 production rather than the direct side effects of pcDNA3.1 vector integration and G418 antibiotic resistance acquisition. HeLa CHI3L1 cells (clone 2, passage 60 after pcDNA3.1 CHI3L1 transfection) had 9 new CCAs and increased number of NCCAs comparing to parental HeLa cells. Most of CCAs revealed in HeLa CHI3L1 clones differed, pointing to the independent evolution of karyotype changes (Fig. 5B and C, Table S4). Interestingly, the average number of marker chromosomes per cell in HeLa CHI3L1 clones decreased 1.7–2.0 times. This fact indicates that single cell cloning of the step-wise genome evolving cells can significantly reduce population genetic complexity and heterogeneity, which did not reverse

61

even after the long-term passaging of clones. Actually, the relative stability of karyotypes of single cell clones derived from cell lines with the moderate ongoing structural and numerical CIN was reported [31]. On the other hand, increased karyotypic heterogeneity of single cell clones was observed even within a short period of time for highly genomic unstable cell populations, where non-clonal chromosome aberrations (NCCAs) greatly outnumbered clonal chromosome aberrations (CCAs), clearly indicating that highly unstable cell populations were not clonable [10]. aCGH analysis revealed many differences in CNAs between HeLa CHI3L1 clones themselves and parental HeLa cells as the cumulative result of plasmid integration, stable cell line selection, single-cell cloning, and CHI3L1 production (Fig. 6A; Table S5). Despite constitutive ectopic CHI3L1 production in the cloned HeLa CHI3L1 cells (Fig. 6B), their viability and CFE were lower compared to parental HeLa cells (Fig. 6C–E). Thus, the opposite growth characteristics were observed between HeLa CHI3L1 clones and 293 CHI3L1 cells. 3.4. The long-term TMZ treatment of HeLa CHI3L1 cells promoted the step-wise genome evolution, CNAs and phenotype changes To reveal karyotype-phenotype correlation after additional genotoxic stress, we established two sub-lines (namely CL2TMZ1 and CL2TMZ2) from HeLa CHI3L1 (clone 2) by repetitively exposing cells to the incremental TMZ concentrations followed by the one-month wash-out period after last treatment to capture a simulated “end run” of karyotype evolution after drug application (total additional 25–30 passages). Both CL2TMZ1 and CL2TMZ2 cells had reduced number of CCAs and no obvious changes in number and frequency of NCCAs compared to HeLa CHI3L1 (clone 2) cells (Fig. 7A and B, Table S4). Karyotype changes were accompanied mainly by loss of non-dominant CCAs. The average number of markers per cell did not change significantly. Analysis of aCGH data revealed CNAs specific for CL2TMZ1 or CL2TMZ2 cells and distinct from HeLa CHI3L1 (clone 2) CNAs (Fig. 8A, Table S5). The ectopic CHI3L1 production was reduced after TMZ treatment (Fig. 8B). Viability of CL2TMZ1 cells did not change, whereas viability of CL2TMZ2 cells increased, despite lower plating efficiency (point 24 h in viability curve) (Fig. 8C). Both CL2TMZ1 and CL2TMZ2 formed more colonies in soft agar than HeLa CHI3L1 (clone 2) cells (Fig. 8D and E). Taken together, changes in CCAs and CNAs were accompanied by phenotype changes, which did not correlate with ectopic CHI3L1 production. The opposite growth characteristics were observed between CL2TMZ1/2 cells and 293pcTMZ1/2 cells in response to the long-term TMZ treatment. 4. Discussion The cell lines are a prevailing tool of choice in investigation of gene functions, oncogenic properties, and drug treatment responses/resistance. However, the experimental manipulations (e.g., transgene overexpression, gene knock out/down, chemical treatments, changes in culture conditions, etc.) act as a system stress increasing a level of the genome dynamics and heterogeneity (genome chaos), especially when the stress is high or the examined system is unstable [5,32]. In support, modulation of expression/activity of the diverse cancer-associated genes in 293 cells (review in preparation) as well as in other cell lines and tumor models [7,33] was documented to promote CIN. However, a concern of possibility of an empty vector-mediated genome destabilization was rarely reported due to use of empty vector-transfected cells as the only control without karyotype analysis of parental/wild type cells. We revealed that either transfection of the empty

62

A. Stepanenko et al. / Mutation Research 771 (2015) 56–69

Fig. 3. The long-term TMZ treatment intensified the punctuated genome evolution of 293 pcDNA3.1 (variant 2) cells. (A) The karyographs show the degree of clonality and variability of chromosomes in each cell line by comparing the copy numbers of intact and abnormal/marker chromosomes of 20 metaphases to each other. The karyotype differences between cell lines are visualized by alignment and comparison of karyographs for each cell line. All chromosomes depicted on the x-axis are listed in the table in the same order. (B) The table lists all CCAs and summarizes karyotypic parameters of each cell line. The most obvious differences in CCAs between cell lines are marked. NCCAs specific for each cell line are listed in Table S2.

vector pcDNA3.1 or pcDNA3.1 CHI3L1 into 293 cells induced significant changes of karyotype stability and heterogeneity. Bylund et al. [29] also reported that wild type 293 cells had hypotriploid karyotype, whereas the empty vector transfected 293 pVgRXR cells were hyperdiploid. Despite less chromosome number in 293 pVgRXR cells, they had higher number of aberrant chromosomes and only one derivative chromosome was common with 293 cells. Similarly, the empty vector-transfected MCF-7 cells differed from parental cells in total chromosome number, number of chromosome deletions/duplications, and pattern of translocations

[34]. Generally, DNA transfer into mammalian cells can result in the insertion mutagenesis, chromosome abnormalities, and alterations of the host DNA methylation [35–37]. A fundamental importance of karyotype examination of the empty vector transfected-, transgene- or drug-treated cells becomes clear in the light of data that karyotype changes and CNAs caused by these manipulations coincided with the growth characteristic changes. Consistent coevolution of karyotypes and phenotypes in diverse models was reported [2,3,10,38,39,16,40,41]. The reason is the chromosomes, by preserving karyotype, ensure the

A. Stepanenko et al. / Mutation Research 771 (2015) 56–69

63

Fig. 4. 293pcTMZ1/2 cells demonstrated the TMZ-promoted distinct CNAs, low viability and CFE. (A) CNAs were identified by aCGH. Chromosomal ideograms with the drawn green and red bars aligned along chromosomes show the areas of copy number gain and loss, respectively. The chromosome CNAs loci are detailed in Table S3. (B) Comparison of viability curves derived from MTT assay. Data points represent mean (±SD) of four individual experiments in quadruplexes. (C) The representative photographs of plates with stained colonies grown in soft agar for three weeks. (D) Comparison of CFE in soft agar colony assay. Bar graph shows mean (±SD) of four individual experiments in triplicates. **P ≤ 0.01, ***P ≤ 0.001.

maintenance of system inheritance (the order of genes along the chromosomes and within the genome). Numerical and/or structural chromosomal alterations disrupt system inheritance and result in extensive gene expression changes. The more dynamic karyotype changes and heterogeneity (genome chaos) are the more dramatic and dynamic changes of gene expression occur [10,32,42] as correlation between a number of chromosome(s) or genes copy number and mRNA/protein levels is generally proportional in aneuploid cells [3,7,43–46]. Changes in genome topology (altered distribution of chromatin) and expression of transcription factors may magnify global changes in gene expression and lead to further disruption of cellular function. Since the genome context (a number and structure of chromosomes and their nucleus topology) drives the genetic network (gene content, RNA and protein expression and their interaction), the transcriptome, gene isoform expression, protein abundance, subcellular localization, and protein–protein interactions are under change in cells with unstable genome. Consequently, genome alterations rewire and create new genetic networks, reprogramming signaling pathways and their function,

and generate biochemical individuality of cells [5,10,47]. As a proof of concept, a number of studies observed that stable transfection of empty vectors into different cell lines led to phenotype changes such as gene expression/cell signaling, proliferation, apoptosis, migration, growth in soft agar and as xenografts in mice [48–54]. PC3 pcDNA3.1 prostate cells proliferated slower in comparison to parental PC3 cells [48]. Loss of corticotropin-releasing factor (CRF) binding to membranes and inability of CRF agonists to stimulate cAMP production was observed in 293 pcDNA3.1 cells due a strong reduction of CRF expression in comparison to parental 293 cells [49]. The spontaneous apoptotic rate of RAW pcDNA3.1 mouse macrophage cells was higher than that of parental RAW cells [50]. Increased vimentin and metalloproteinase expression was observed in 293 pEFIRES-P cells in comparison to parental 293 cells [51]. PEO1 pEF6/V5 ovarian cells formed fewer colonies and their rate of growth in nude mice was less than that of parental PEO1 cells. Another two independently established PEO1 pEF6/V5 cell lines also grew faster or slower compared to parental PEO1 cells in nude mice [52]. MCF-7 pU1 breast cells demonstrated a

64

A. Stepanenko et al. / Mutation Research 771 (2015) 56–69

Fig. 5. The stable pcDNA3.1 CHI3L1 vector transfection promoted the step-wise genome evolution of HeLa CHI3L1 cells. (A) Distribution of chromosomes across 200 metaphases analyzed for each cell line. (B) Karyotypic differences between cell lines are visualized by alignment and comparison of karyographs for each cell line. All chromosomes depicted on the x-axis are listed in the table on the right. №P/№P depicts number of passages after pcDNA3.1 CHI3L1 transfection/single cell cloning, respectively. (C) The table lists all CCAs and summarizes karyotype parameters of each cell line. The most obvious differences in CCAs between cell lines are marked. NCCAs specific for each cell line are listed in Table S4.

A. Stepanenko et al. / Mutation Research 771 (2015) 56–69

65

Fig. 6. Clones of HeLa CHI3L1 cells have significantly different CNAs and are phenotypically less aggressive than parental HeLa cells. (A) CNAs were identified by aCGH. Chromosomal ideograms with the drawn green and red bars aligned along chromosomes show the areas of copy number gain and loss, respectively. The chromosome CNAs loci are detailed in Table S4. (B) Western blot analysis of CHI3L1 production in cell lines. Cell lysates blotted in duplicates were prepared independently. U87 cells express endogenously a high level of CHI3L1 and were used as a positive control. (C) Comparison of viability curves derived from MTT assay. Bar graph shows mean (±SD) of six individual experiments in quarduplexes. (D) The representative photographs of plates with stained colonies grown in soft agar for three weeks. (E) Comparison of CFE in soft agar colony assay. Bar graph shows mean (±SD) of four individual experiments in triplicates. **P ≤ 0.01, ***P ≤ 0.001.

lower/lagged migration speed in response to hepatocyte growth factor stimulation in comparison to parental MCF-7 cells [53]. Overexpression of syndecan family members caused significant reduction of doubling time only in comparison to HT-1080 pEGFPN1 fibrosarcoma cells but not parental HT-1080 cells [54]. Empty vector transfections gave a nudge to immortalized cells to make them obviously tumorigenic in mice [35], whereas retroviral vectors can potentially transform normal cells rising concern of their use in clinical trials [36]. Although the empty vector transfected HeLa pcDNA3.1 cells were unavailable for this study, however, HeLa CHI3L1 cells (clone 1) at early passages demonstrated the high karyotype stability after pcDNA3.1 CHI3L1 transfection suggesting an insignificant influence of vector integration on karyotype in HeLa cells (at

least on microscopic pattern of chromosomal aberrations). In contrast to 293 cells derivatives, karyotypes of HeLa CHI3L1 clones were much more stable. We suppose that accuracy of chromosome integrity-controlling mechanisms determines the degree of genome destabilization after plasmid transfer (and other stresses). In any case, the possible indirect effects associated with the process of plasmid transfer itself and selection of antibiotic resistant clones should be verified and taken in consideration when interpreting results and making conclusions in transgene studies. Despite ectopic CHI3L1 production, we observed the opposite growth characteristics of 293 CHI3L1 and HeLa CHI3L1 cells. It was shown that 293 cells responded to the exogenous CHI3L1 stimulation [55], whereas HeLa cell line was unresponsive [23]. If it is the case, then the observed phenotype changes of HeLa CHI3L1 clones

66

A. Stepanenko et al. / Mutation Research 771 (2015) 56–69

Fig. 7. The long-term TMZ-treated HeLa CHI3L1 (clone 2) cells continued the step-wise phase of genome evolution (A) The karyographs show the degrees of clonality and variability of the chromosomes of each cell line by comparing the copy numbers of intact and marker chromosomes of 20 metaphases to each other. The karyotypic differences between cell lines are visualized by alignment and comparison of the karyographs of the designated cell lines. All chromosomes depicted on the x-axis are listed in the table on the right. (C) The table lists all CCAs and summarizes karyotypic parameters of each cell line. The most obvious differences in CCAs between cell lines are marked. NCCAs specific for each cell line are listed in Table S4.

were largely determined by karyotype changes rather than interfered in by ectopic CHI3L1 expression. In contrast, ectopic CHI3L1 production in 293 CHI3L1 cells counterbalanced the effects of high CIN on viability and significantly influenced growth characteristics. On the other hand, the cell type-dependent opposite effects

of CHI3L1 [21–24] and many other cancer-associated genes [20] were reported. We share the view that the effect and function of a (trans)gene can be opposite and versatile in cells with different genomes and are determined by genetic network, which is in turn defined by genome context [5,47]. Thereby, caution

A. Stepanenko et al. / Mutation Research 771 (2015) 56–69

67

Fig. 8. Long-term TMZ treatment promoted CNAs and more aggressive phenotype of HeLa CHI3L1 (clone 2) cells, which did not correlate with ectopic CHI3L1 production. (A) Significant differences in CNAs were identified by aCGH. Chromosomal ideograms with the drawn color bars aligned along chromosomes show the areas of genetic gain/loss. Red bars represent areas of copy number loss. Green bars represent areas of copy number gain. Positions of chromosome loci of CNAs are detailed in Table S5. (B) Long-term TMZ treatment resulted in decrease of ectopic CHI3L1 production. (C) Comparison of viability curves derived from MTT assay. Bar graph shows mean (±SD) of six individual experiments in quarduplexes. (D) The representative photographs of plates with stained colonies grown in soft agar for three weeks. (E) Comparison of CFE in soft agar colony assay. Bar graph shows mean (±SD) of four individual experiments in triplicates. ***P ≤ 0.001.

in interpretation of molecular analysis data and the relationship between cause and effect in cells with unstable genome is needed [5,10,47]. Karyotype-phenotype causality [2,3,5,6], the genome context-dependent antagonistic functional duality of cancer genes [20], potential promiscuous molecular interactions not relevant to normal cellular conditions of dosage-sensitive oncogenes with increments in concentration of protein products [56], and an introduction of a term “good oncogene” [20,57], which describes a gene that is both a marker for favorable survival probability in clinic but one with tumor promoting and transforming properties in experimental settings incline to re-evaluate the dominating concepts in cancer research. The TMZ-promoted phenotype changes of HeLa CHI3L1 cells derivatives (CL2TMZ1 and CL2TMZ2) did not correlate with ectopic CHI3L1 production. This further reinforces karyotype-phenotype causality. To the point, the CIN-driven tumorigenicity after loss of transgene expression was repeatedly documented in the

different transgenic models [33,41,58]. Generally, acute or chronic drug treatment promotes karyotype changes and intrinsic level of CIN correlates with (multi)drug resistance [2,8,59–61]. As a result, RNA/protein expression profile after long-term drug treatment differs from parental drug-naïve cells in expression of hundreds of genes [2,8]. The short or long-term TMZ-treated cells demonstrated the marked changes of transcriptome, proteome, metabolome, and kinome with the resulting versatile changes in morphology, proliferation, adhesion, migration/invasion, and tumor formation potential [62–68], consistently indicating the TMZ-promoted profound reorganization of genetic network. The confirmative evidence for TMZ-promoted genome evolution comes from the study of primary tumor samples of glioma patients before and after therapy [28]. We observed the opposite phenotype response of the longterm TMZ-treated cells in punctuated phase (reduced viability and CFE) and step-wise phase of genome evolution (increased

68

A. Stepanenko et al. / Mutation Research 771 (2015) 56–69

viability and CFE). The opposite phenotype responses of TMZtreated cells have already been reported [64,65], however, without elucidation of a driving force. Potentially, TMZ treatment of patients with tumors in different phases of genome evolution can have significantly different response and outcome. For example, correlation between excessive CIN and improved survival outcome in breast cancer was evidenced [18,19]. However, intended therapeutic promotion of tumor cell genomes to excessive instability is a double-edged sword: the primary positive objective response and reduced cell viability may be accompanied by increase in cell population genome heterogeneity and evolutionary potential of residual disease. Altogether, this work brings to light the other side of coin in plasmid transfer procedure, in artificial changes of gene expression/activity, and drug treatment, and inclines to go into a question if all the observed phenotype changes after experimental manipulations can be ascribed to a specific gene/molecular mechanism of interest? How to discriminate, which phenotype changes are caused by the manipulation of expression/activity of gene product itself and which are due to stress-induced chromosome changes? CIN-driven phenotype changes, which significantly influence growth characteristics of tumor cells, should be always taken into consideration during different experimental manipulations. Genome macroevolution, a process implying numerical and structural chromosome rearrangements with topology changes, is an engine, which allows making profound phenotype leaps and fast adaptations to microenvironmental stresses. Cancer genome evolution drives dynamic changes of transcriptome and proteome, rewires metabolic and signaling pathways, and gives rise to the phenotype variants/clonal diversification of tumor cells, which are the basis for cancer evolutionary selection. Understanding of macroevolution in carcinogenesis and its effect on tumor behavior under stress is mandatory effort in the field of cancer research. Conflict of interest statement The authors declare that there are no conflicts of interest. Acknowledgements We are grateful to Dr. I. Gout (Department of Structural and Molecular Biology, Institute of Structural and Molecular Biology, University College London), Dr. A. Iershov and Dr. O. Balynskaya (Department of Biosynthesis of Nucleic Acids, IMBG, Kyiv), and Dr. V. Grishkova (Department of Cell Signaling, IMBG, Kyiv) for provided cell lines. We thank Dr. A. Nazaruk for assistance in the preparation of manuscript figures. This work was supported in frames of the programs “Fundamental grounds of molecular and cell biotechnologies” and “Nanotechnologies and nanomaterials for 2010–2014 years” by National Academy of Sciences of Ukraine (NASU). Partial financial support was also provided by FP7 Program of European Union (Project 294932) to V. K. and Y. V. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/ j.mrfmmm.2014.12.006. References [1] P. Duesberg, R. Li, A. Fabarius, R. Hehlmann, The chromosomal basis of cancer, Cell. Oncol. 27 (2005) 293–318. [2] P. Duesberg, R. Li, R. Sachs, A. Fabarius, M.B. Upender, R. Hehlmann, Cancer drug resistance: the central role of the karyotype, Drug Resist. Updat. 10 (2007) 51–58, http://dx.doi.org/10.1016/j.drup.2007.02.003.

[3] P. Duesberg, D. Mandrioli, A. McCormack, J.M. Nicholson, Is carcinogenesis a form of speciation? Cell Cycle 10 (2011) 2100–2114. [4] H.H.Q. Heng, G. Liu, S. Bremer, K.J. Ye, J. Stevens, C.J. Ye, Clonal and non-clonal chromosome aberrations and genome variation and aberration, Genome 204 (2006) 195–204, http://dx.doi.org/10.1139/G06-023. [5] H.H.Q. Heng, J.B. Stevens, S.W. Bremer, K.J. Ye, G. Liu, C.J. Ye, The evolutionary mechanism of cancer, J. Cell. Biochem. 109 (2010) 1072–1084, http://dx.doi.org/10.1002/jcb.22497. [6] H.H. Heng, S.W. Bremer, J.B. Stevens, S.D. Horne, G. Liu, B.Y. Abdallah, et al., Chromosomal instability (CIN): what it is and why it is crucial to cancer evolution, Cancer Metastasis Rev. 32 (2013) 325–340, http://dx.doi.org/ 10.1007/s10555-013-9427-7. [7] A.A. Stepanenko, V.M. Kavsan, Immortalization and malignant transformation of eukaryotic cells, Cytol. Genet. 46 (2012) 96–129, http://dx.doi.org/ 10.3103/S0095452712020041. [8] A.A. Stepanenko, V.M. Kavsan, Evolutionary karyotypic theory of cancer versus conventional cancer gene mutation theory, Biopolym. Cell 28 (2012) 267–280, http://dx.doi.org/10.7124/bc.000059. [9] A.A. Stepanenko, V.M. Kavsan, Karyotypically distinct U251, U373, and SNB19 glioma cell lines are of the same origin but have different drug treatment sensitivities, Gene 540 (2014) 263–265, http://dx.doi.org/ 10.1016/j.gene.2014.02.053. [10] B.Y. Abdallah, S.D. Horne, J.B. Stevens, G. Liu, A.Y. Ying, B. Vanderhyden, et al., Single cell heterogeneity: why unstable genomes are incompatible with average profiles, Cell Cycle 12 (2013) 3640–3649. [11] C.J. Ye, J.B. Stevens, G. Liu, S.W. Bremer, A.S. Jaiswal, K.J. Ye, et al., Genome based cell population heterogeneity promotes tumorigenicity: the evolutionary mechanism of cancer, J. Cell. Physiol. 219 (2009) 288–300, http://dx.doi.org/10.1002/jcp.21663. [12] B.T. Ragel, W.T. Couldwell, D.L. Gillespie, M.M. Wendland, K. Whang, R.L. Jensen, A comparison of the cell lines used in meningioma research, Surg. Neurol. 70 (2008) 295–307, http://dx.doi.org/10.1016/j.surneu.2007.06.031, discussion 307. [13] D.-L. Zhang, L. Ji, L.-J. Li, G.-S. Huang, Systematically experimental investigation on carcinogenesis or tumorigenicity of VERO cell lines of different karyotypes in nude mice in vivo used for viral vaccine manufacture, Yi Chuan Xue Bao 31 (2004) 647–660. [14] N.J. Bahlis, Darwinian evolution and tiding clones in multiple myeloma, Blood 120 (2012) 927–928, http://dx.doi.org/10.1182/blood-2012-06-430645. [15] R.A. Burrell, C. Swanton, The evolution of the unstable cancer genome, Curr. Opin. Genet. Dev. 24 (2014) 61–67, http://dx.doi.org/10.1016/j.gde. 2013.11.011. [16] L. Li, A.A. McCormack, J.M. Nicholson, A. Fabarius, R. Hehlmann, R.K. Sachs, et al., Cancer-causing karyotypes: chromosomal equilibria between destabilizing aneuploidy and stabilizing selection for oncogenic function, Cancer Genet. Cytogenet. 188 (2009) 1–25, http://dx.doi.org/10.1016/ j.cancergencyto.2008.08.016. [17] A.D. Silk, L.M. Zasadil, A.J. Holland, B. Vitre, D.W. Cleveland, B.A. Weaver, Chromosome missegregation rate predicts whether aneuploidy will promote or suppress tumors, Proc. Natl. Acad. Sci. U. S. A. 110 (2013) E4134–E4141, http://dx.doi.org/10.1073/pnas.1317042110. [18] R. Roylance, D. Endesfelder, P. Gorman, R.A. Burrell, J. Sander, I. Tomlinson, et al., Relationship of extreme chromosomal instability with long-term survival in a retrospective analysis of primary breast cancer, Cancer Epidemiol. Biomarkers Prev. 20 (2011) 2183–2194, http://dx.doi.org/ 10.1158/1055-9965.EPI-11-0343. [19] N.J. Birkbak, A.C. Eklund, Q. Li, S.E. McClelland, D. Endesfelder, P. Tan, et al., Paradoxical relationship between chromosomal instability and survival outcome in cancer, Cancer Res. 71 (2011) 3447–3452, http://dx.doi.org/ 10.1158/0008-5472.CAN-10-3667. [20] A.A. Stepanenko, Y.S. Vassetzky, V.M. Kavsan, Antagonistic functional duality of cancer genes, Gene 529 (2013) 199–207, http://dx.doi.org/ 10.1016/j.gene.2013.07.047. [21] R. Shao, K. Hamel, L. Petersen, Q.J. Cao, R.B. Arenas, C. Bigelow, et al., YKL-40, a secreted glycoprotein, promotes tumor angiogenesis, Oncogene 28 (2009) 4456–4468, http://dx.doi.org/10.1038/onc.2009.292. [22] M. Kawada, H. Seno, K. Kanda, Y. Nakanishi, R. Akitake, H. Komekado, et al., Chitinase 3-like 1 promotes macrophage recruitment and angiogenesis in colorectal cancer, Oncogene 31 (2012) 3111–3123, http://dx.doi.org/ 10.1038/onc.2011.498. [23] N. Ngernyuang, R.A. Francescone, P. Jearanaikoon, J. Daduang, A. Supoken, W. Yan, et al., Chitinase 3 like 1 is associated with tumor angiogenesis in cervical cancer, Int. J. Biochem. Cell Biol. 51 (2014) 45–52, http://dx.doi.org/ 10.1016/j.biocel.2014.03.021. [24] J. Salamon, T. Hoffmann, E. Elies, K. Peldschus, J.S. Johansen, G. Lüers, et al., Antibody directed against human YKL-40 increases tumor volume in a human melanoma xenograft model in scid mice, PLOS ONE 9 (2014) e95822, http://dx.doi.org/10.1371/journal.pone.0095822. [25] M. Nakada, T. Furuta, Y. Hayashi, T. Minamoto, J.-I. Hamada, The strategy for enhancing temozolomide against malignant glioma, Front. Oncol. 2 (2012) 98, http://dx.doi.org/10.3389/fonc.2012.00098. [26] K. Yoshimoto, M. Mizoguchi, N. Hata, H. Murata, R. Hatae, T. Amano, et al., Complex DNA repair pathways as possible therapeutic targets to overcome temozolomide resistance in glioblastoma, Front. Oncol. 2 (2012) 186, http://dx.doi.org/10.3389/fonc.2012.00186.

A. Stepanenko et al. / Mutation Research 771 (2015) 56–69 [27] S. Spiegl-Kreinecker, C. Pirker, C. Marosi, J. Buchroithner, J. Pichler, R. Silye, et al., Dynamics of chemosensitivity and chromosomal instability in recurrent glioblastoma, Br. J. Cancer 96 (2007) 960–969, http://dx.doi.org/ 10.1038/sj.bjc.6603652. [28] B.E. Johnson, T. Mazor, C. Hong, M. Barnes, K. Aihara, C.Y. McLean, et al., Mutational analysis reveals the origin and therapy-driven evolution of recurrent glioma, Science 343 (2014) 189–193, http://dx.doi.org/ 10.1126/science.1239947. [29] L. Bylund, S. Kytölä, W.-O. Lui, C. Larsson, G. Weber, Analysis of the cytogenetic stability of the human embryonal kidney cell line 293 by cytogenetic and STR profiling approaches, Cytogenet. Genome Res. 106 (2004) 28–32, http://dx.doi.org/10.1159/000078556. [30] M. Macville, E. Schröck, H. Padilla-nash, E. Schro, C. Keck, B.M. Ghadimi, et al., Comprehensive and definitive molecular cytogenetic characterization of HeLa cells by spectral karyotyping, Cancer Res. (1999) 141–150. [31] A.V. Roschke, K. Stover, G. Tonon, A. a Schäffer, I.R. Kirsch, Stable karyotypes in epithelial cancer cell lines despite high rates of ongoing structural and numerical chromosomal instability, Neoplasia 4 (2002) 19–31, http://dx.doi.org/10.1038/sj/neo/7900197. [32] G. Liu, J.B. Stevens, S.D. Horne, B.Y. Abdallah, K.J. Ye, S.W. Bremer, et al., Genome chaos: survival strategy during crisis, Cell Cycle 13 (2014) 528–537, http://dx.doi.org/10.4161/cc.27378. [33] A. Stepanenko, V. Kavsan, Cancer genes and chromosome instability, in: Y. Siregar (Ed.), Oncogene and Cancer – From Bench to Clinic, InTech Publisher, Rijeka, Croatia, 2013, pp. 151–182. [34] P. Winnard, C. Glackin, V. Raman, Stable integration of an empty vector in MCF-7 cells greatly alters the karyotype, Cancer Genet. Cytogenet. 164 (2006) 174–176, http://dx.doi.org/10.1016/j.cancergencyto.2005.07.021. [35] L. Bardwell, The mutagenic and carcinogenic effects of gene transfer, Mutagenesis 4 (1989) 245–253. [36] C. Baum, O. Kustikova, U. Modlich, Z. Li, B. Fehse, Mutagenesis and oncogenesis by chromosomal insertion of gene transfer vectors, Hum. Gene Ther. 17 (2006) 253–263, http://dx.doi.org/10.1089/hum.2006.17.253. [37] W. Doerfler, Impact of foreign DNA integration on tumor biology and on evolution via epigenetic alterations, Epigenomics 4 (2012) 41–49, http://dx.doi.org/ 10.2217/epi.11.111. [38] P. Duesberg, C. Iacobuzio-donahue, J. Brosnan, A. Mccormack, D. Mandrioli, L. Chen, Origin of metastases: subspecies of cancers generated by intrinsic karyotypic variations, Cell Cycle 11 (2012) 1151–1166, http://dx.doi.org/ 10.4161/cc.11.6.19580. [39] A. McCormack, J.L. Fan, M. Duesberg, M. Bloomfield, C. Fiala, P. Duesberg, Individual karyotypes at the origins of cervical carcinomas, Mol. Cytogenet. 6 (44) (2013), http://dx.doi.org/10.1186/1755-8166-6-44. [40] J.M. Nicholson, P. Duesberg, On the karyotypic origin and evolution of cancer cells, Cancer Genet. Cytogenet. 194 (2009) 96–110, http://dx.doi.org/ 10.1016/j.cancergencyto.2009.06.008. [41] A. Klein, N. Li, J.M. Nicholson, A. aMcCormack, A. Graessmann, P. Duesberg, Transgenic oncogenes induce oncogene-independent cancers with individual karyotypes and phenotypes, Cancer Genet. Cytogenet. 200 (2010) 79–99, http://dx.doi.org/10.1016/j.cancergencyto.2010.04.008. [42] J.B. Stevens, G. Liu, B.Y. Abdallah, S.D. Horne, K.J. Ye, S.W. Bremer, et al., Unstable genomes elevate transcriptome dynamics, Int. J. Cancer 134 (2014) 2074–2087, http://dx.doi.org/10.1002/ijc.28531. [43] C. Gao, K. Furge, J. Koeman, K. Dykema, Y. Su, M. Lou Cutler, et al., Chromosome instability, chromosome transcriptome, and clonal evolution of tumor cell populations, Proc. Natl. Acad. Sci. U. S. A. 104 (2007) 8995–9000, http://dx.doi.org/10.1073/pnas.0700631104. [44] N. Donnelly, Z. Storchová, Dynamic karyotype, dynamic proteome: buffering the effects of aneuploidy, Biochim. Biophys. Acta 1843 (2014) 473–481, http://dx.doi.org/10.1016/j.bbamcr.2013.11.017. [45] J.B. Stevens, S.D. Horne, B.Y. Abdallah, C.J. Ye, H.H. Heng, Chromosomal instability and transcriptome dynamics in cancer, Cancer Metastasis Rev. 32 (2013) 391–402, http://dx.doi.org/10.1007/s10555-013-9428-6. [46] T. Ried, Y. Hu, M.J. Difilippantonio, B.M. Ghadimi, M. Grade, J. Camps, The consequences of chromosomal aneuploidy on the transcriptome of cancer cells, Biochim. Biophys. Acta 1819 (2012) 784–793, http://dx.doi.org/ 10.1016/j.bbagrm.2012.02.020. [47] H.H.Q. Heng, S.W. Bremer, J.B. Stevens, K.J. Ye, G. Liu, C.J. Ye, Genetic and epigenetic heterogeneity in cancer: a genome-centric perspective, J. Cell. Physiol. 220 (2009) 538–547, http://dx.doi.org/10.1002/jcp.21799. [48] I. Abasolo, L. Yang, R. Haleem, W. Xiao, R. Pio, F. Cuttitta, et al., Overexpression of adrenomedullin gene markedly inhibits proliferation of PC3 prostate cancer cells in vitro and in vivo, Mol. Cell. Endocrinol. 199 (2003) 179–187.

69

[49] F.M. Dautzenberg, J. Higelin, U. Teichert, Functional characterization of corticotropin-releasing factor type 1 receptor endogenously expressed in human embryonic kidney 293 cells, Eur. J. Pharmacol. 390 (2000) 51–59. [50] J.A. Gutiérrez-Pabello, D.N. McMurray, L.G. Adams, Upregulation of thymosin beta-10 by Mycobacterium bovis infection of bovine macrophages is associated with apoptosis, Infect. Immun. 70 (2002) 2121–2127. [51] T. Skoog, O. Elomaa, S.M. Pasonen-Seppänen, S. Forsberg, K. Ahokas, L. Jeskanen, et al., Matrix metalloproteinase-21 expression is associated with keratinocyte differentiation and upregulated by retinoic acid in HaCaT cells, J. Invest. Dermatol. 129 (2009) 119–130, http://dx.doi.org/10.1038/jid.2008.206. [52] C. Gourley, A.J.W. Paige, K.J. Taylor, C. Ward, B. Kuske, J. Zhang, et al., WWOX gene expression abolishes ovarian cancer tumorigenicity in vivo and decreases attachment to fibronectin via integrin alpha3, Cancer Res. 69 (2009) 4835–4842, http://dx.doi.org/10.1158/0008-5472.CAN-08-2974. [53] W.G. Jiang, D. Grimshaw, J. Lane, T.A. Martin, R. Abounader, J. Laterra, et al., A hammerhead ribozyme suppresses expression of hepatocyte growth factor/scatter factor receptor c-MET and reduces migration and invasiveness of breast cancer cells, Clin. Cancer Res. 7 (2001) 2555–2562. [54] B. Péterfia, T. Füle, K. Baghy, K. Szabadkai, A. Fullár, K. Dobos, et al., Syndecan-1 enhances proliferation, migration and metastasis of HT1080 cells in cooperation with syndecan-2, PLoS ONE 7 (2012) e39474, http://dx.doi.org/10.1371/journal.pone.0039474. [55] P.O. Areshkov, S.S. Avdieiev, O.V. Balynska, D. Leroith, V.M. Kavsan, Two closely related human members of chitinase-like family, CHI3L1 and CHI3L2, activate ERK1/2 in 293 and U373 cells but have the different influence on cell proliferation, Int. J. Biol. Sci. 8 (2012) 39–48. [56] T. Vavouri, J.I. Semple, R. Garcia-Verdugo, B. Lehner, Intrinsic protein disorder and interaction promiscuity are widely associated with dosage sensitivity, Cell 138 (2009) 198–208, http://dx.doi.org/10.1016/j.cell.2009.04.029. [57] J.M. Lee, The good oncogene: when bad genes identify good outcome in cancer, Med. Hypotheses 76 (2011) 259–263, http://dx.doi.org/10.1016/ j.mehy.2010.10.015. [58] R. Sotillo, J.-M. Schvartzman, N.D. Socci, R. Benezra, Mad2-induced chromosome instability leads to lung tumour relapse after oncogene withdrawal, Nature 464 (2010) 436–440, http://dx.doi.org/10.1038/nature08803. [59] R. Li, R. Hehlman, R. Sachs, P. Duesberg, Chromosomal alterations cause the high rates and wide ranges of drug resistance in cancer cells, Cancer Genet. Cytogenet. 163 (2005) 44–56, http://dx.doi.org/10.1016/ j.cancergencyto.2005.05.003. [60] A.J.X. Lee, D. Endesfelder, A.J. Rowan, A. Walther, N.J. Birkbak, P.A. Futreal, et al., Chromosomal instability confers intrinsic multidrug resistance, Cancer Res. 71 (2011) 1858–1870, http://dx.doi.org/10.1158/0008-5472.CAN-10-3604. [61] S.D. Horne, J.B. Stevens, B.Y. Abdallah, G. Liu, S.W. Bremer, C.J. Ye, et al., Why imatinib remains an exception of cancer research, J. Cell. Physiol. 228 (2013) 665–670, http://dx.doi.org/10.1002/jcp.24233. [62] C. Happold, P. Roth, W. Wick, N. Schmidt, A.-M. Florea, M. Silginer, et al., Distinct molecular mechanisms of acquired resistance to temozolomide in glioblastoma cells, J. Neurochem. 122 (2012) 444–455, http://dx.doi.org/ 10.1111/j.1471-4159.2012.07781.x. [63] D.M. Kumar, V. Patil, B. Ramachandran, M.V. Nila, K. Dharmalingam, K. Somasundaram, Temozolomide-modulated glioma proteome: role of interleukin-1 receptor-associated kinase-4 (IRAK4) in chemosensitivity, Proteomics 13 (2013) 2113–2124, http://dx.doi.org/10.1002/pmic.201200261. [64] D. Lamoral-Theys, M. Le Mercier, B. Le Calvé, M.A. Rynkowski, C. Bruyère, C. Decaestecker, et al., Long-term temozolomide treatment induces marked amino metabolism modifications and an increase in TMZ sensitivity in Hs683 oligodendroglioma cells, Neoplasia 12 (2010) 69–79. [65] B. Le Calvé, M. Rynkowski, M. Le Mercier, C. Bruyère, C. Lonez, T. Gras, et al., Long-term in vitro treatment of human glioblastoma cells with temozolomide increases resistance in vivo through up-regulation of GLUT transporter and aldo-keto reductase enzyme AKR1C expression, Neoplasia 12 (2010) 727–739. [66] J.C. Anderson, C.W. Duarte, K. Welaya, T.D. Rohrbach, M. Bredel, E.S. Yang, et al., Kinomic exploration of temozolomide and radiation resistance in Glioblastoma multiforme xenolines, Radiother. Oncol. (2014), http://dx.doi.org/ 10.1016/j.radonc.2014.04.010. [67] L. Hiddingh, R.S. Raktoe, J. Jeuken, E. Hulleman, D.P. Noske, G.J.L. Kaspers, et al., Identification of temozolomide resistance factors in glioblastoma via integrative miRNA/mRNA regulatory network analysis, Sci. Rep. 4 (2014) 5260, http://dx.doi.org/10.1038/srep05260. [68] V.V. Leshchenko, P.-Y. Kuo, Z. Jiang, V.K. Thirukonda, S. Parekh, Integrative genomic analysis of temozolomide resistance in diffuse large B-cell lymphoma, Clin. Cancer Res. 20 (2014) 382–392, http://dx.doi.org/10.1158/ 1078-0432.CCR-13-0669.

Step-wise and punctuated genome evolution drive phenotype changes of tumor cells.

The pattern of genome evolution can be divided into two phases: the step-wise continuous phase (step-wise clonal evolution, stable dominant clonal chr...
12MB Sizes 3 Downloads 5 Views