Three standard cell lines were used to investigate the performance of this method:
• MCF-7, a human breast cancer cell line (European Collection of Cell Culture).
• U-343, a human glioma cell line (Westermark et al 1973).
• BON, a human neuroendocrine tumor cell line (a kind gift from Dr. C.M Townsend, University of Texas, Galvestone, USA) derived from a lymph node metastasis of a pancreatic carcinoid.
The MCF-7 cells were grown in MEM-Eagle medium supplemented with 10% FCS, 1 mM sodium pyruvate, 2 mM L-glutamine, 1% non-essential amino acids and 5% Penicillin (Tamro). U-343 cells were cultured in Ham's F-10 medium supplemented with 10% FCS, 1 mM sodium pyruvate and 2 mM L-glutamine (Tamro). BON cells were grown in Ham F-12 K medium (NordCell, Sweden) mixed with DMEM medium supplemented with 10% FCS, 1 mM sodium pyruvate and 2 mM L-glutamine (Tamro).
The medium was changed twice a week and the cells were maintained in exponential growth phase.
Multicellular tumor culture
The tumor cells were trypsinized from the stem monolayer culture, then cell suspensions were seeded in 24-well 1% agarose coated culture plates, with approximately 50,000 cells per well for U-343 and MCF-7 and 15,000 for BON. The cultures were kept at 37°C with 5% CO2
, and grown for a period that depended on duplication rate for each cell line [6
MTS medium for MCF-7 was DMEM supplemented with 10% FCS, 1 mM sodium pyruvate, 2 mM L-glutamine, 1% non-essential amino acids, 5% Penicillin (Tamro), 0.01 mg/ml Insulin and 1 nM β-Estradiol (Sigma Aldrich).
The aggregates were photographed daily using a Nikon Colorpix 4500 http://www.nikon.com
digital camera mounted on a Zeiss Axiovert 135 Microscope http://www.zeiss.com
. The digital camera was set to use up to 4.0 million effective pixels, with an image resolution of 640 × 480, auto-focus, auto-shooting modes, with applied "Hi" image quality. A 5× magnification objective, 5x/0,12; 44 01 20 on Zeiss Axiovert 135 Microscope was used with adjustment of the strength of background light depending on the color of the medium that was used for growing the desired MTS.
Images were saved as sequences of JPEG files in a 128 MB, Scan Disc Compact Flash card and transferred into the hard disc of a personal computer with installed Windows 2000 for further image and statistical analysis.
A software program in Matlab (The Mathworks, Natick, Massachusetts) with user-friendly interface was developed to perform the image analysis using routines from the "image processing" toolbox. For absolute calibration of the area, a spherical object with a specified diameter of 0.79375 μm, as shown in Fig. , was used as reference object.
Reference object used for absolute calibration of area. The units on the axis are number of pixels.
After feeding some input parameters such as name of starting image, number of images to analyze, project name, the software automatically delineated a ROI (Region of Interest) discriminating the necrosis area from the area where the cells were still alive and a ROI around the whole MCT. This was done automatically and sequentially for all images. The total area of the MTS and total area of necrosis were calculated as the total number of pixels inside the ROIs and was converted into μm2 by using the radius of the reference object. To calculate the volume of the MTS the following equation was used.
was used to calculate the radius of the sphere. The total volume of the necrosis was calculated and subtracted from the total volume of the MTS to obtain the total volume of the living part. Moreover the fraction of total volume constituted by living cells was calculated.
All images with outlined ROIs discriminating the different parts of the MTSs were visualized and automatically saved as TIFF (Tagged Image File Format) for further observations. All desired statistical measurements including the total volume of the whole MTSs and total volume of the necrosis part were saved in an EXCEL file with specified file name for the full group of MTSs.
Validation of method was performed by analyzing pictures from different angles of the same aggregate. The growth curves for the three selected cell lines were generated to illustrate the use of the program. During the first few days of culture after seeding, the aggregates settle, which means they do not follow an exponential growth pattern. The growth curves after this period of 5–9 days were plotted and data were fitted to a monoexponential to indicate the volume doubling time.
Frozen Section Autoradiography and Staining
At a selected time point some of the MTSs were incubated with 18
F-labeled Fluoro-Deoxy-Glucose (FDG) to confirm the extent of the necrosis part. After 40 min incubation with 100 MBq/ml FDG and washing for 3*5 min, the MTSs were frozen quickly at -20°C and sectioned for autoradiography [7
The slices with a thickness of 25 μm were exposed on a phosphor-imaging plate for 20 h. Scanning and imaging were performed using software Image Quant (Molecular Dynamics, Sunnyvale, CA).
For further verification Hematoxylin & Eosin staining of the slices was used. This complementary part was to clarify some histological features in the MTSs.