Cisplatin treatment of C6 rat glioma in vivo did not infl uence copy number alterations and growth pattern of tumor-derived resistant cells

A. A. Stepanenko1, V. P. Baklaushev2, 3, Y. S. Vassetzky4, V. V. Dmitrenko1 1 Institute of Molecular Biology and Genetics, NAS of Ukraine 150, Akademika Zabolotnoho Str., Kyiv, Ukraine, 03680 2 Pirogov Russian National Research Medical University, 1, Ostrovitianov Str., Moscow, Russian Federation, 117997 3 Federal Research and Clinical Centre, Federal Medical-Biological Agency 28, Orekhoviy Blvd. Moscow, Russian Federation, 115682 4 CNRS UMR 8126, Universit Paris-Sud 11, Institut Gustave Roussy 114, rue Edouard Vaillant, Villejuif, France, 94805 a.a.stepanenko@gmail.com


Introduction
Cisplatin (also called cisplatinum or cis-diam minedichloroplatinum(II) is a platinum inorganic coordination compound with a square planar geometry.It is chemically inert until one or both of its chloride atoms are displaced spontaneously by water molecules into the cytoplasm due to the relatively low concentration of chloride ions in the cytoplasm ver-sus the extracellular space.Mono-/bi-aquated forms of cisplatin are highly electrophilic and form covalent bonds with methionine and cysteine-containing peptides.Cisplatin binds with high affi nity to DNA, especially to nucleophilic N7 sites on purines, promoting the formation of protein-DNA complexes as well as intra-and inter-strand DNA adducts.Thus, cisplatin exerts its cytostatic/cytotoxic effects in both nuclear and cytoplasmic compartments with the pri-mary role of the latter as high cytotoxicity was observed in enucleated cells [1][2] The cisplatin treatment resulted in the inhibition of mitochondrial respiration and effl ux of calcium from the mitochondria, induced endoplasmic reticulum stress, reactive oxygen species formation and oxidative stress, leading to the depletion of reduced glutathione (GSH) and nicotinamide adenine dinucleotide (NADH) and peroxidation of polyunsaturated lipids and proteins (reviewed in [1,2]).
Cisplatin has been used in clinics since the 70s of the XX century for the treatment of multiple types of solid cancers.Noteworthy, cisplatin is signifi cantly effi cient only against germ cell cancers, resulting in a durable complete remission in >80 % of the patients; the biological reasons of this are still not completely understood.In contrast, the clinical response elicited by cisplatin in the patients affected by other solid tumors is very temporary, and the use of cisplatin-based chemotherapeutic regimens is restricted due to intrinsic or acquired resistance as well as considerable systemic side effects [1,2].
The glioma clinical trials also demonstrated a low effi cacy or even inferior outcome in the high-grade glioma patients after the treatment with cisplatin-based chemotherapy.On the contrary, the in vivo models showed that the treatment with cisplatin or cisplatin-loaded mAbCx43-or BSAT1-conjugated nanogels signifi cantly inhibited rat glioma C6 and 101/8 tumor growth and increased survival of animals [3,4].Similarly, a combination of cisplatin with other therapeutic approaches demonstrated a high inhibitory effect in the in vivo glioma models [5][6][7].The reason for this obvious discrepancy is that the drugmediated stress may actually foster the tumor evolution by both selecting genetic variations and generating novel variations through the induction of genome reorganization [8][9][10][11][12][13].An increase in chromosomal aberrations during and after chemotherapy was found to associate with an increased tumor aggressiveness [9].The ability of cisplatin to induce chromosome abnormalities has been known since the time of its approval for the clinic use [14].For example, the cisplatin treatment induced a signifi cant increase in the chromosome breakage at all stages of spermatogenesis in Drosophila melanogaster [15].In Saccharomyces cerevisiae, the cisplatin treatment induced chromosomal aberrations, although the degree of instability was lower than that induced by other DNA damaging agents (e.g., bleomycin, camptothecan, or γ-irradiation) [16].A comparison of the parental human ovarian cancer cells to sublines selected for resistance to cisplatin revealed that the acquired resistance to cisplatin was associated with the numerical and structural chromosome aberrations and substantial copy number alterations (CNAs) [17,18].
The intrinsic or acquired resistance is another signifi cant contributor to the therapy failure.The cell culture models revealed that cisplatin resistance is attained by several complex molecular and cellular mechanisms.They include but are not confi ned to the activation of multiple signal transduction pathways, reduction of cisplatin accumulation by either active effl ux or impaired infl ux, detoxification by cellular antioxidant systems, increase in DNA damage repair, inactivation of apoptosis pathway, alterations of the membrane protein traffi cking due to the cytoskeleton defective organization, activation of the epithelial-mesenchymal transition, overexpression of chaperones, deregulation of miRNA and transcription factors, the impact of stroma cells and components of extracellular matrix [1,2,19].An overall abnormal phenotype after the cisplatin treatment results from an increase in DNA mutation load and profound changes in the DNA methylome, transcriptome, prote ome, metabolome and kinome (kinase activity profi le) [20].Altogether, these data reveal a complex multilevel mechanism of tumor cells self-defence against the cisplatin cytotoxicity.
Here, we have investigated whether the cisplatin treatment of C6 rat glioma in vivo impacted CNAs, proliferation and colony formation effi ciency (CFE) of the tumor-derived cisplatin-resistant cells in vitro.

Cisplatin treatment of rats with intracranial C6 gliomas
The animals were kept and treated in accordance with the Guidelines on Laboratory Practices adopted by the Ministry of Health of the Russian Fede ration (Order 267, 19 June 2003), and the principles of bioethics adopted by the European Convention for the Protection of Vertebrate Animals Used for Ex perimental and Other Scientifi c Purposes (Stras bourg, 1986).A glioma modeling and tumor volume measurement were performed as described previously [3,4].After fi ve days of C6 glioma cells inoculation, rats received 20 % DMSO (n = 1, namely C6R1), which served as a solvent for drug, or cisplatin (n = 2, namely C6R4CIS and C6R5CIS) injected intraperitoneally three times per week at a dose of 5 mg/kg.Rats were sacrifi ced after 10 injections of cisplatin/ DMSO.Gliomas were harvested aseptically and disaggregated mechanically.The cell suspension was seeded on plates with growth medium.Cells were used at passage numbers 3-10 for analysis.

Cell proliferation analysis
Cells were seeded onto 6 cm dishes at the density 1 × × 10 4 and grown in the high-glucose DMEM with 10 % FBS.On the 7th day of seeding, medium was carefully removed; cells were washed with PBS, trypsinized, harvested, incubated with trypan blue, and calculated using a hemocytometer.The test was repeated at least three times.DMEM supplemented with 10 % FBS.0.35 % top agarose was poured on 1.5 ml of solidifi ed 0.5 % base agarose/10 % FBS/DMEM.Cells were seeded in triplicates in 35-mm dishes of 6-well plates and grown at 37 C for three weeks to allow colony formation.Colonies were visualized by staining with 0.005 % crystal violet, photographed, and counted using OpenCFU software [21].Test for each cell line was repeated three times.

Statistics
The Student's t-test was used to analyze the significance of variability between the results of each group and its corresponding control (Statistica 7 Software, San Diego, USA).Results with *P < 0.05, **P < < 0.01, and ***P < 0.001 were considered significant.All experimental data are reported as the mean and the error bars represent the experimental standard error (± standard deviation, SD).

Results and Discussion
Cisplatin-treated C6R4CIS, C6R5CIS and vehicletreated control C6R1 cell lines were established from the parental C6 cell line by injecting cisplatin (or a vehicle 20 % DMSO) intraperitoneally three times per week at a dose of 5 mg/kg.A morphometric analysis of C6 gliomas after two weeks of treatment showed that tumor volumes of C6R1, C6R4CIS and C6R5CIS gliomas were ≈ 80 mm 3 , ≈ 35 mm 3 and ≈ 60 mm 3 , respectively (Figure 1).After 10 injections, the vehicle-treated control and surviving cisplatin-resistant cells were tumor-derived and analyzed in vitro.
Despite the numerous reports on genotoxic effects of cisplatin [2,[15][16][17][18]22], the copy number alterations (CNAs) of parental C6 cells, control C6R1, and cisplatin-treated C6R4CIS and C6R5CIS cell lines were essentially si milar with minor variations in the genedepleted chro mosome loci.The major detected aberrations we re loss of 16q12.1-q24.3 and gain of 7p21.1-q31.1 (Fi gu re 2).It should be noted that in contrast to conventional cytogenetics/karyotyping, aCGH is based on the average profi le of genetic changes in thousands of cells; this underestimates structural chromosome com plexity and heterogeneity of tumor cells [12,23].There was no difference in proliferation rates between C6R1 and C6R4CIS; however, C6R5CIS had a moderately reduced proliferation (Figure 3).Further, no difference in CFE between C6R1 and C6 R4CIS or C6R5CIS cells was observed (Figure 4).The data that cisplatin treatment reduces tumor growth and increases survival of C6 glioma-bearing rats [4] but does not obviously infl uence chromosome stability and in vitro growth characteristics of C6 cells, which survived/resisted cisplatin treatment, suggest that cisplatin has a cytostatic rather than profound genotoxic and cytotoxic effects on C6 cells in vivo.Interestingly, we have observed reduced proliferation and CFE of the control and cisplatin-treated cells in comparison to the parental C6, suggesting that the rat brain microenvironment may select for slow-dividing C6 cells.The alternative explanation is that in brain, C6 cells are adapted to different nutrients, oxygen levels, growth-stimulating and other factors; therefore they undergo a stress when reintroduced to in vitro culture.Finally, we cannot exclude an effect of a vehicle (20 % DMSO) as DMSO treatment in vivo induced cytotoxicity at certain concentrations ( [24] and refs.herein).However, we have injected substantially lower DMSO volume and concentration than those used in these studies.
Previously, we and others have shown that cisplatin or combination of cisplatin with other therapeutic approaches demonstrated a high inhibitory effect in in vivo models, including rat glioma C6 and 101/8 tumor models [3][4][5][6][7].Unfortunately, the encouraging laboratory observations do not agree with the efficiency of treatment of the high-grade glioma patients.There were many frustrated clinical trials as exemplifi ed below that ultimately did not justify the addition of cisplatin to the chemotherapeutic regimen of the high-grade glioma patient treatment.Cisplatin and carmustine followed by radiation did not improve median survival, survival at one year, or time-to-progression of patients with newly diagnosed glioblastoma.Furthermore, this treatment was associated with more serious toxicities than standard therapy [25].Similarly, cisplatin administered concurrently with carmustine and radiotherapy resulted in a higher toxicity but provided no signifi cant im-provement in survival [26].The estimation of response rate, the rate of disease stabilization, and the probability of one-year survival of the patients with high-grade glioblastomas who received radiotherapy concurrently with cisplatin/carmustine chemotherapy did not support the routine use of concurrent cisplatin and carmustine [27].The progression-free survival and overall survival of glioblastoma patients treated with the radiotherapy, cisplatin and carmustine were comparable with those obtained with the radiotherapy and temozolomide but the toxicity was more frequent and persistent.The results argued against future use of this combination in the treatment of patients with glioblastoma [28].Although the median survival time with the cisplatin/nimustine neoadjuvant chemotherapy followed by the radiotherapy and adjuvant temozolomide was increased in patients with newly diagnosed glioblastoma, a high frequency of serious hematological toxicity limited its use [29].Further, none real advantage of neoadjuvant cisplatin and etoposide after surgery and prior to radical radiotherapy was obtained in the treatment of high-grade gliomas if compared with the standard adjuvant chemotherapy [30].Cisplatin combined with cytosine arabinoside and hydroxyurea did not improve the six-month survival rate in patients with relapsed or progressive high-grade glioblastomas.Signifi cant hematological toxicities were observed [31].Finally, the recent medulloblastoma trial demonstrated that the cumulative cisplatin dose was not associated with overall survival [32].
Recently, we have shown that the long-term treatment of tumor cell lines with a DNA damaging drug temozolomide, which is widely used for glioma patients, promoted the clonal and non-clonal chromosome aberrations, CNAs and diverse phenotype changes affecting survival, CFE, migration, and invasion ( [12], Stepanenko et al., in preparation).Furthermore, many chemotherapeutics, which have been used previously or are exploited currently in clinic for cancer treatment, were evidenced to induce/promote the chromosome instability/aneuploi-dy in rodent and human cells.These include nocodazole, a microtubule-depolymerizing agent; paclitaxel and taxol, microtubule-stabilizing agents; platinum compounds cisplatin and carboplatin; etoposide and doxorubicin, topoisomerase I and II inhibitors; bleomycin that causes breaks in DNA; actinomycin D which interferes with transcription and replication; 5-fl uorouracil, a thymidylate synthase inhibitor; imatinib and nilotinib, multiple kinase inhibitors; rapamycin, an mTOR inhibitor; tamoxifen, an estrogen receptor inhibitor, and many others (reviewed in [9]).This should be seriously taken into account as the cancer cell genome instability, promoted by drug-mediated system stress, fosters genomic, epigenetic, and non-genetic heterogeneity, intensifi es profound transcriptome and proteome changes, and rewires metabolic and signaling network, altogether, giving rise to the diverse drug-resistant phenotype variants.The genome instability signifi cantly correlates with inherent and acquired multi-drug resistance, and chemoresistance acquisition is accompa- nied by newly evolved chromosome imbalances (reviewed in [8-11, 13, 33-38]).

Fig. 1 .
Fig.1.Brain magnetic resonance imaging (MRI) using a ClinScan MRI scanner with fi eld strength of 7T was carried out for C6 rat gliomas after 2 weeks of cisplatin treatment.A morphometric analysis of C6 gliomas after this period of time showed that tumor volumes of C6R1, C6R4CIS and C6R5CIS gliomas were ≈80 mm 3 , ≈ 35 mm 3 and ≈ 60 mm 3 , respectively.Rats with C6 gliomas received cisplatin at a dose of 5 mg/kg (C6R4CIS and C6R5CIS) or 20% DMSO (control C6R1), injected intraperitoneally three times per week.The animals were sacrifi ced after 10 injections of cisplatin/DMSO

Fig. 2 .
Fig. 2. 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, whereas blue bars represent areas of copy number gain.The major detected aberrations were loss of 16q12.1-q24.3 and gain of 7p21.1-q31.1 (marked by arrows).CNAs were identifi ed by array comparative genome hybridization (aCGH)

Fig. 3 .
Fig. 3. Cisplatin treatment of C6 glioma in vivo did not consistently affect proliferation of tumor-derived C6 cells.However, signifi cantly reduced proliferation of control C6R1 and cisplatin-treated C6R4CIS and C6R5CIS cell lines in comparison to parental C6 was observed.To analyze the rate of proliferation, cells were seeded at the density 1 × 10 4 cells.The number of cells was calculated after 7 days of growth.All experiments were performed at least three times and results are presented as the means ± SD. *P < 0.05, ***P < 0.001, NS = non-signifi cant C6 C6R1 C6R4CIS C6R5CIS

Fig. 4 .
Fig. 4. Cisplatin treatment of C6 glioma in vivo did not affect colony formation effi ciency (CFE) in soft agar of tumor-derived C6 cells.However, reduced CFE of control C6R1 and cisplatin-treated C6R4CIS and C6R5CIS cell lines in comparison to parental C6 was observed.For colony formation, 5×10 3 cells were seeded in soft agar and grown for 3 weeks.Colonies were stained with crystal violet and calculated using OpenCFU software.(A) The representative photographs of plate wells with stained colonies are shown.(B) The graph compares CFE of the designated cell lines.All experiments were performed three times and results are presented as the means ± SD. *P < 0.05, NS = non-signifi cant A B