Background
Cells of actively growing tumors are exposed to varying micro-environmental conditions due to non-homogeneous vascular supply [
1], leading to the development of localized regions in tumors having low oxygen tension, low glucose concentration and acidic extracellular pH due to accumulation of metabolic by-products such as lactic acid. As a result, cells in these regions are exposed to varying levels of hypoxia, anoxia and acidosis [
2]. From the radiobiological point of view, hypoxic cells constitute the most important cellular subpopulation because they are more radioresistant than the euoxic cells as a result of reduced generation of radiation-induced reactive oxygen species or ROS [
3,
4]. The hypoxic cell fraction increases considerably with the advancement of tumor size and grade offering significant resistance to radiotherapy. The currently used hypoxic cell sensitizers (nitroimadazole compounds) have met with limited success due to the lack of differential effect besides systemic toxicity. Therefore, there is a need to develop effective strategies to selectively enhance the radiosensitivity of these cells, which requires a detailed characterization of the nature and responses of these resistant cellular fractions. Since hypoxic cells exist in three-dimensional tumors that behave as heterogeneous systems demonstrating alterations in many vital genotypic and phenotypic characteristics regulating various biological processes, an
in vitro cellular model that closely simulates these conditions is essential for carrying out studies on hypoxic cell and tumor responses to therapeutic agents.
Monolayer cultures of established cell lines from human tumors have been widely used for studying the various molecular processes including the identification of specific molecular lesions related to the dysregulation of cell proliferation and cell death, the two important functional targets for the development of therapies. Unfortunately, investigations necessary for the development and/or evaluation of some of the therapeutic strategies cannot be carried out with this most widely used laboratory system, since complexities arising out of 3-dimensional organization of solid tumors (viz., cell-cell/cell-matrix interactions and variations in vasculature and nutrient supply) resulting in subtle changes in phenotypic expression (especially the metabolic changes) are not provided in the monolayer cultures derived from tumor cells. In contrast, multicellular spheroids of tumor cells provide an excellent three-dimensional
in vitro model in which hypoxic conditions can be generated to facilitate detailed investigations including the response to various chemical agents and radiation [
5,
6].
Spheroids are characterized by high cell-density and a closely packed, 3-D tumor like structure, which leads to severe diffusion limitations for molecules as small as glucose and oxygen, thereby creating heterogeneous cell sub-populations of actively proliferating as well as quiescent, hypoxic and necrotic cells as found in solid tumors [
7]. Available evidences suggest that low concentration of glucose and oxygen in the inner regions of spheroids may contribute to the formation of quiescent, hypoxic, anoxic and necrotic cell sub-populations [
8‐
10]. Characterization of the nature of these cell sub-populations as well as their responses to radiation and radiomodifying agents can help in the development of more effective radiosensitizing strategies. For example, hypoxic cells are known to derive a large part of their energy from glucose-dependent anaerobic metabolic pathways. Therefore, inhibitors of glycolysis could specifically modify the responses of this resistant sub-population of cancer cells. A number of
in vitro studies have indeed shown that 2-DG, a glucose analogue and inhibitor of glycolytic ATP production [
11‐
13], selectively inhibits energy-dependent DNA repair and cellular recovery processes in cancer cells [
14‐
16] resulting in enhanced cell death [
11,
13,
17,
18]. Spheroids could serve a very useful model to understand further the mechanisms of radiosensitivity induced by this glycolytic inhibitor in solid tumors, since spheroids mimic the solid tumors more closely than the monolayer culture.
Spheroids have been used for a broad spectrum of studies in cancer biology and employed extensively in radiobiological investigations [
19] as they provide a good
in vitro system to mimic the radioresistant hypoxic cell population generally found in tumors. Unfortunately however, detailed information available on the intercellular variation in mitochondrial mass, activity and ROS levels (oxidative stress mainly responsible for damage induced by low LET radiation) as a function of spheroid growth and with reference to other parameters like viability cell death etc in culture is lacking. These parameters influence the end results of
in vitro studies aimed at understanding tumour response to various cytotoxic agents and metabolic inhibitors. Therefore, the present study was undertaken to systematically characterize spheroids established from a human glioma cell line BMG-1 [
11,
20] with respect to organization, growth, viability and cell survival, cell death, metabolic and mitochondrial status, oxidative stress and radiation response with age and spheroid size using microscopy, flow cytometry and enzymatic assays.
The results of the present studies demonstrate development of S-negative cells, reduced endogenous and radiation-induced ROS besides changes in levels of anti (Bcl2) as well as pro (Bax) apoptotic regulators observed in spheroids which suggest the intricate/complex nature of endogenous as well as induced stress resistance that could exist in tumors, which contribute to the treatment resistance.
Conclusion
In summary, the present study provides a detailed characterization of multicellular spheroids with respect to a number of parameters relevant to studies aimed at evaluating efficacy of tumor therapeutic agents. Characterization of mitochondrial status and oxidative stress levels with the increasing age of multicellular spheroids, coupled with enhanced glycolysis and related variations in HIF-1α and c-Myc will prove useful in optimizing the use of this valuable cellular 3-D tumor model for predicting tumor response to chemotherapeutic/radiosensitizing agents acting on metabolic pathways. In this study, spheroids were regularly fed with fresh medium, which may resemble the in vivo conditions wherein a proportion of cells receive uninterrupted supply of nutrients and oxygen while others become hypoxic and may enter secondary necrosis. Investigations using spheroid model also highlight the inherent limitations of extrapolations made from monolayer studies for application in tumor biology and therapy. The detailed characterization presented here is aimed at obtaining a better understanding of the spheroid model for optimizing and interpreting the predictive therapy trials and drug response of various therapeutic agents and/or adjuvants, particularly metabolic modulators.
Methods
Monolayer culture
The cerebral glioma cell line (BMG-1; wild type p53) was established in Bangalore, India (20). Stock cultures were maintained in the exponentially growing state by passaging twice weekly in DMEM containing 10 mM HEPES and antibiotics supplemented with 5% fetal bovine serum.
Spheroid culture
BMG-1 spheroids were grown by inoculating 1 × 106 viable cells of BMG1 in non-adherent 90-mm petridishes in 10 ml DMEM supplemented with 5% fetal calf serum, antibiotics and 10 mM HEPES. Clusters of cells could be observed after 24 hours of initiation. However, it took nearly 4 days for these clusters to form spheroids (clusters could not be dislodged by pipetting). Approximately 10,000 spheroids could be obtained in a petridish of ~60 cm2 (90 mm diameter) which were redistributed in to 20 non-adherent 90-mm petridishes containing 10 ml of complete growth medium. The pH of the medium was monitored daily to prevent acidosis. The medium was changed on alternative days till spheroids were 14 day old, followed by redistribution of spheroids from each petridish further to 10 non-adherent petridishes and the medium was changed daily thereafter till 28 days. Unlike other spheroids reported in the literature these spheroids could not be grown for months as they disintegrated after 28 days.
Spheroid growth measurements
Spheroid growth was monitored by measuring the increase in spheroid size as a function of time. Images of spheroids were obtained using the image analysis system consisting of Olympus BX 60 fluorescence microscope and Grundig FA87 monochrome CCD camera. Volume and surface area was calculated using Optimas® image analysis software (Optimas, USA; version 5.0). The size of nearly 50 spheroids was evaluated in each group by measuring two orthogonal diameters (d1 and d2) using the line morphometry function. Volume was calculated using the formula V = 4/3Πr3, where r = 1/2√ d1d2 the geometric mean radius. Average cell number per spheroid was calculated by trypsinizing 10–20 spheroids and the total number of cells obtained was divided by the number of spheroids trypsinized.
Analysis of cell cycle kinetics using BrdU labeling
Cells were incubated with 10 μM BrdU for 30 min at 37°C at various time intervals (7–28 days). Immediately on completion of incubation, an aliquot containing 106 cells from monolayers or spheroidal cell suspension was pelleted by centrifugation (1000 g, 10 min), resuspended in PBS, fixed in 80% ethanol and stored at 4°C until use. For analysis, fixed cells were washed with 0.9% NaCl, incubated with pepsin (0.5% in 0.055 N HCl, pH 1.8) for 10 min at 37°C, washed with saline again, and treated with 2 N HCl for 30 min at room temperature. Following another wash with normal saline and a wash with PBS-Tween (0.05%), the cells were incubated with anti-BrdU antibody (mouse anti-human IgG1; BD Pharmingen, USA) in PBS-Tween at 1:40 dilution for 30 min at 4°C. Cells were then washed twice in PBS-Tween (0.05%)-BSA (1%), and incubated with rabbit antimouse IgG1-FITC conjugate (Sigma) at 1:100 dilution in PBS-Tween-BSA for 30 min at 4°C. After another wash in PBS-Tween-BSA, cells were resuspended in PBS and stained with propidium iodide (50 μg/ml) for 30 min. Fluorescence of propidium iodide (DNA) and FITC (BrdU) was measured simultaneously in the FACS Calibur Flow Cytometer (Becton Dickinson), and biparametric scatter plots were analysed using the CellQuest software. Histograms were analyzed for cell cycle phase distribution by using Modfit software.
Detection of PS-externalization using annexin-V labeling
Membrane asymmetry accompanied by translocation of the phospholipid phosphotidylserine (PS) from inner to the outer side of the plasma membrane is one of the manifestations of apoptosis. Externalization of PS was studied by Annexin V (a phospholipid binding protein) binding assay [
22]. Briefly, live cells were washed twice in PBS and resuspended in binding buffer containing 0.01 M HEPES/NaOH, pH 7.4; 0.14 mM NaCl; 2.5 mM CaCl
2. Cell suspension (1 × 10
5 cells in 100 μl) in the binding buffer was incubated with 5 μl of FITC labeled Annexin V (BD Pharmingen, USA) for 15 minutes in the dark at room temperature. Following incubation, propidium iodide was added and fluorescence of cells PI (DNA) and FITC (Annexin) was measured simultaneously in the FACS Calibur Flow Cytometer (Becton Dickinson), and biparametric scatter plots were analysed using the CellQuest software.
Cell survival
Macrocolony assay with dissociated cell population is still considered a reference standard for assessing the survival of cells. This has been extensively employed in studies with ionizing radiation and cytotoxic anticancer drugs, to predict the dose needed to cure transplanted tumors in mice as well as multi-cellular spheroids [
52,
53]. Spheroids and monolayers were washed with PBS and irradiated in HBSS using Co-60 teletherapy source (Theratron-780 C, Canada) to access the cell survival. Following irradiation, spheroids and monolayer cultures were dissociated into single cells by trypsinization and counted using haemocytometer. Appropriate number of cells was plated in 60 mm Petri dishes with complete media containing 10% FCS. Plates were incubated at 37°C in a humidified incubator (5% CO
2 and 95% air) till colonies were formed (7 days). Methanol fixed colonies was stained with 1% crystal violet and colonies containing more than 50 cells were scored.
Estimation of glucose utilization and lactate production
BMG-1 cells and spheroids were incubated in HBSS or HBSS and 2-DG for 2 h. The amount of glucose remaining unused and the lactate produced were estimated in the buffer using enzymatic assays. Glucose was measured using the reducing sugar method [
54‐
56]. Briefly a mixture of equal volumes of incubating medium (HBSS) and alkaline copper sulphate were incubated at 90°C for 10 minutes, followed by the addition of phosphmolybidic acid was added at room temperature (~27°C). O.D of the resulting chromogen was measured at 540 nm. Lactate was estimated using lactate oxidase method based kit (Randox; cat No.-LC2389). The number of viable cells in spheroids was counted and glucose consumed or lactate produced was normalized with respect to number of viable cells.
Determination of intracellular redox levels
Intracellular redox levels were measured using the fluorescent dye H2DCFDA which is a non-polar compound (Merck England) and yields a fluorescent DCF in the presence of ROS in the cells. Green fluorescence due to intracellularly trapped DCF was collected on the FL1 channel on the log scale. Spheroids and monolayer cells were just washed twice in PBS and held in PBS with Ca2+, Mg2+, H2DCFDA (10 μg/ml) and 5 mM glucose, before irradiation and subsequently incubated for half an hour at 37°C. Cells were trypsinized after a rinse with PBS, washed and resuspended in PBS with Ca2+, Mg2+ and glucose. Samples were stored on ice, and measurements were made within half an hour after trypsinization.
Flow cytometric analysis of mitochondrial mass and activity
Spheroids and monolayers were parallely stained with one of the mitochondrial fluorescent dyes NAO and Rh123, essentially according to the procedure described earlier [
57]. Briefly, the unfixed spheroids and monolayers were washed and incubated in PBS (with Ca
2+, Mg
2+), glucose and NAO (10 μM 10 min at 25°C) or Rh123 (5 μg/ml 30 min. at 37°C), followed by trypsinization washing and resuspension in PBS (with Ca
2+, Mg
2+ and glucose) and incubated on ice until analysis by the flow cytometer.
Scraped monolayer cells and spheroids were collected at 800 g for 5 minutes, washed twice with 5 ml PBS and sonicated (VC-X 500, Sonics and Materials Inc. USA) on ice in PBS for 1 minute mixed with pre chilled acetone (3 parts) and incubated overnight at -20°C. Following incubation, acetone was removed by centrifugation and pellet was dissolved in autoclaved Milli Q water and protein content was determined by Lowry's method [
58]. 2× lysis buffer [62.5 mM Tris Hcl ph 6.8, 10% (v/v) Glycerol 2% (w/v) SDS, 1 mM PMSF, 1 μg/ml pepstatin A, 1 μg/ml leupeptin and 5 μg/ml aprotinin] was added to the pellet and boiled at 100°C for 5 minutes. An equal amount of protein (10–20 μg) was loaded on 10% SDS polyacrylamide gels. Proteins were separated at a constant voltage of 100 V for 1.5 hours and then blotted to PVDF membrane (Amersham) in transfer buffer [25 mM Tris, 192 mM glycine, 15% methanol (v/v)] overnight at 14 V. The blots were blocked with 1% BSA in TTBS [15 mM Tris-HCl pH7.5, NaCl .9% and 0.1% Tween-20] for 1 hour at room temperature (25–28°C). Membranes were incubated with a mouse monoclonal antibody (1:1000) anti Bax (BD Pharmingen, USA), anti Bcl2 (BD Pharmingen, USA) anti c-Myc (BD Pharmingen, USA), anti HIF-1α (Santa Cruz Biotechnology USA) and anti Beta actin (Promega) for 1 hour in TTBS and 1% BSA. Membranes were washed with TTBS four times for 15 minutes each and incubated with anti mouse peroxidase conjugated secondary antibody (1:1000 dilutions; Banglore Genei India) for 1 hour. Blots were washed and developed using ECL chemiluminescence detection reagent (Pierce). Membranes were stripped in stripping buffer (25 mM Glycine, 1% SDS, pH 2 adjusted with HCl) for 1 hour washed twice in TTBS for 10 minutes each and re-probed.
Authors' contributions
DK contributed 50% in all aspects of this work. SC helped in the microscopy and helped to draft the manuscript. MBA helped in the enzymatic assays and statistical analysis of the data. BSD participated in design and coordination of the study and helped to draft the manuscript. All authors read and approved the final manuscript.