Restriction of cisplatin induction of acute apoptosis to a subpopulation of cells in a three-dimensional carcinoma culture model Walid Fayad, Slavica Brnjic, Daniel Berglind, Susanna Blixt, Maria C. Shoshan, Maria Berndtsson, Maria H€agg Olofsson and Stig Linder * Cancer Center Karolinska, Department of Oncology and Pathology, Karolinska Institute and Hospital, Stockholm, Sweden Cisplatin is a clinically important chemotherapeutical agent used to treat epithelial malignancies. High concentrations (20–100 lM) of cisplatin have been used in numerous studies to induce apopto- sis of carcinoma cells grown in monolayer culture over 24–48 hr. These conditions may not be relevant to 3-D tumor tissue in vivo and the importance of apoptosis for tumor response is controver- sial. We here studied the effects of cisplatin on a 3-D colon carcinoma in vitro model (multicellular spheroids). Cisplatin at a dose of 40 lM induced active caspase-3 preferentially in the peripheral 30 lm cell layer of spheroids, mainly during late stages (72–96 hr). The p53 response to cisplatin was also largely confined to peripheral cell layers. Despite the use of a high cisplatin concen- tration, a significant fraction of the cells in the spheroids survived treatment. A high proportion of surviving cells stained positive for b-galactosidase, a marker of premature senescence. Cells growth- arrested by cisplatin treatment showed a higher spontaneous cell death rate than untreated proliferating cells. We propose that acute apoptosis is of minor significance for the overall response of carcinoma cells to cisplatin treatment. ' 2009 UICC Key words: cisplatin; apoptosis; senescence; p21Cip1; p27Kip1 Platinum derivatives are important chemotherapeutical agents for many epithelial malignancies, including testicular, ovarian, squamous cell and colon cancers. 1,2 In the low intracellular chloride concentrations, cisplatin undergoes equation to form molecules that are reactive with cellular macromolecules. Cisplatin forms covalent bonds to the N7 positions of DNA purines leading to formation intra- or interstrand crosslinks, but also reacts with RNA and protein. 3,4 Treatment of tumor cells in vitro with cisplatin results in apoptosis within 24–48 hr. 4–6 Dysregulation of apoptosis path- ways is generally assumed to be important for resistance to cispla- tin. 6 Cisplatin-induced calcium signaling, leading to activation of calpain and cleavage of Bid, 7 and sustained activation of N-termi- nal-c-Jun kinase, 8 is important for apoptosis signaling. Recent work has shown that cisplatin is able to induce caspase activation in enucleated cells, demonstrating that cisplatin can trigger apoptosis pathways that are independent of its effects on nuclear DNA. 9,10 This cytoplasmic pathway involves reactive oxygen species (ROS) and calcium signaling. 10 According to one model, ROS induces activation of Bak, leading to VDAC1 and Bax activation. 11 Despite the very large number of studies in this area (2,970 articles on MedLine on cisplatin/apoptosis March 2009), it is controversial whether apoptosis is the primary mode of cell death of human carcinoma cells in response to DNA damaging therapeu- tic drugs, and also whether apoptosis is necessary for the therapeutic effect of such drugs. 10,12–16 The importance of apopto- sis appears to be most pronounced in short-term experiments (1–2 days) and is of lower significance when cell viability is examined after longer times or in clonogenic out-growth experiments. 12 Overexpression of Bcl-2 in ovarian cancer cells led to resistance to cisplatin when assayed at 48 hr after treatment, but to increased sensitivity at 96 hr. 17 Another important, but often over-looked issue, is that of the drug concentrations used. The mean cisplatin concentration used for induction of apoptosis was 52 lM in 100 articles examined, 10 compared to the mean GI50 concentration of 0.63 lM over all 60 cell lines in the NCI panel. Apoptosis was not induced by IC50 cisplatin doses in CHO cells or in human bladder cancer cells. 15 In a study of colon carcinoma and melanoma cells, apoptosis was not induced over 24–48 hr using IC50 cisplatin concentrations, whereas premature senescence was observed. 10 Most studies of the response of tumor cells to cisplatin and other DNA damaging agents have been carried out using cells grown as subconfluent monolayer cultures, conditions quite distinct from the situation in vivo. When tumor cells are grown as three-dimensional aggregates, or multicellular ‘‘spheroids,’’ their sensitivity to anticancer chemotherapeutic drugs generally decreases. 18,19 This is true for DNA damaging drugs as well as for microtubuli interacting agents. 20,21 We have here examined the effects of cisplatin on multicellular spheroids. Induction of apoptosis was assessed by staining for active caspase-3, and the long-term fate of cisplatin-treated cells was determined by testing for clonogenic outgrowth and staining for the senescence marker b-galactosidase. Our data show that cisplatin-induced caspase activation is confined to the peripheral cell layers of multicellular spheroids and imply that acute apoptosis is of minor importance for the response to cisplatin. Material and methods Cell culture HCT116 colon carcinoma cells were maintained in McCoy’s 5A modified medium/10% fetal calf serum at 37°C in 5% CO 2 . Tissue culture reagents were obtained from Gibco Cell Culture Products. Cisplatin was obtained from Bristol-Myers-Squibb. Treatment with cisplatin was for the duration of the experiment if not stated otherwise. Spheroids were prepared using a modifica- tion of our previously described method. 22 A cell suspension con- taining 10,000 cells was added to each well of poly-HEMA-coated 96-well plates, and the wells were overfilled by adding an additional 170 ll media to acquire a convex surface curvature. Plasticine spacers (3 mm) were placed in the corners of each plate to prevent the lids from touching the media. The plates were then inverted to allow the cells to sediment to the liquid–air interface. Plates were returned to normal position after 24-hr incubation, excess media was removed by a syringe needle connected to vacuum pump and plasticine spacers were removed. Plates were then incubated for further 4 days. Determination of S-phase fraction was performed using flow cytometry. Monolayer cells or spheroids (at different days after their formation) were dispersed by trypsinization and fixed in 1 ml 4% phosphate-buffered formaldehyde at room temperature over night, pelleted and suspended in 1 ml of 95% ethanol. Prior to analysis by flow cytometry, cells were rehydrated in distilled Additional Supporting Information may be found in the online version of this article. Grant sponsor: Chemores (EU FP6); Grant number: LSHC-CT-2007- 037665; Grant sponsors: Cancerfonden, Radiumhemmets Forsknings- fonder, Vetenskapsr ˚ adet, Jordbruksverket. *Correspondence to: Cancer Center Karolinska, R8:00, Karolinska Hospital, S-17176 Stockholm, Sweden. E-mail: Stig.Linder@ki.se Received 17 March 2009; Accepted after revision 26 May 2009 DOI 10.1002/ijc.24627 Published online 8 June 2009 in Wiley InterScience (www.interscience. wiley.com). Int. J. Cancer: 125, 2450–2455 (2009) ' 2009 UICC Publication of the International Union Against Cancer