June 1, 2007 (Vol. 27, No. 11)

Steven C. Miller Ph.D.
Evan F. F. Cromwell, Ph.D. Director of Assay Development Molecular Devices

Image Acquisition and Analysis of 3-D Cultures with the IsoCyte™ Laser-Scanning Platform

Despite a growing awareness of the value of three-dimensional (3-D) cell culture models for drug discovery, cell-based assays for high-throughput screening (HTS) are typically limited to monolayers of adherent cells. 3-D cell culture models are clearly important for understanding the processes involved in converting a noninvasive phenotype into an invasive one and for evaluating therapeutic agents.

Although there is considerable evidence demonstrating that the extracellular matrix and tissue architecture play an important role in the normal functioning of cells and in tumorigenesis, the formation of 3-D colonies in semi-solid media presents many challenges associated with performing and imaging this assay in a form that meets HTS requirements.

3-D cell culture models are especially challenging for microscope-based systems because of the need to tile multiple regions to get sufficient field-of-view and the need to stack multiple z-heights to adequately sample thick media. Moreover, fast and robust computational tools for image analysis of such samples are lacking for the HTS environment.

At Blueshift Biotechnologies (www.blueshiftbiotech.com), a solution to these problems has been achieved with an optimized assay system that utilizes the IsoCyte™ scanning cytometer. The colony assay is set up in 96-well, black-walled plates with square wells and a glass bottom (MatriCal(www.matrical.com)) that results in better control of the media layers as well as improved optical access and image quality.

A simple one-step colony-staining protocol has been developed using LavaCell™ (Active Motif(www.activemotif.com)). LavaCell stains cellular membranes of live or dead cells by reacting with free amine groups on proteins. It produces a bright fluorescent signal upon 488-nm laser excitation with a peak emission at 610 nm. The IsoCyte is then used to scan over full plates, analyzing whole-well areas in a single pass in approximately five minutes.

The IsoCyte has a proprietary optical system with a large depth of field (400 µm) and binocular setup making it well-suited for 3-D colony measurements for HTS and drug discovery efforts (Figure 1). The system can image cell colonies at several micron resolution over whole plates without need of focus adjustment or acquisition of multiple z-heights.


Figure 1

Cell Colony Assay Plate Preparation

Human lung carcinoma cells, A549 (ATCC CCL-185), were cultured in RPMI-1640 medium supplemented with 10% FBS, 25 mM Hepes, 2 mM Glutamax-1, 1 mM sodium pyruvate, 1 mM nonessential amino acids, 100 units/mL penicillin, and 100 µg/mL streptomycin. A 0.8% (w/v) agarose base layer solution was prepared by adding sterile 1.5% agarose solution (Chemicon(www.chemicon.com)) to complete RPMI-1640 medium. The base layer was formed in sample wells by adding 100 µL/well of the 0.8% agarose solution. The A549 cell concentration was adjusted to 1 x 104 cells/mL in a solution of complete RMPI-1640 medium and 0.3% agarose solution. The cell suspension (100 µL containing 1,000 cells per well) was added to the top of the base layer and allowed to solidify for one hour at 37°C. Next, 100 µL/well of RPMI-1640 media was added to sample wells, the outer wells of the MatriCal plate were filled with sterile water to aid in evaporation control, and the plate was incubated overnight in a humidified incubator at 37°C and 5% CO2. The following morning, 100 µL of 4x dilutions of staurosporine in media were added to appropriate wells (n=6). The cultures were microscopically monitored for colony formation. The day 14 cultures were stained with LavaCell by removing the growth media and adding 100 µL of 2.4 µM LavaCell dye in HBSS. Plates were incubated for 1–2 hours and analyzed using the IsoCyte.


Figure 2

Analysis of Cell Colonies and Staurosporine-induced Growth Inhibition

The IsoCyte laser-scanning platform with 488-nm laser excitation was set for whole-well, two-channel image acquisition with 600-nm long pass (LP). Fluorescence images were acquired at 5 x 5 micron sampling with a total scan plate cycle time of 5 minutes. Whole-well image results and the analysis process are shown in Figure 2. The selected well (white circle) is from the column treated with 0.1-nM staurosporine.

A dilution series of staurosporine from 1×10-6 to 1×103 nM concentrations was evaluated for A549 colony growth inhibition after 14 days. The colonies were stained with LavaCell as described and analyzed using the IsoCyte.

A common problem with imaging colonies in thick media is shadowing and meniscus effects from the walls (see loss of image quality on left well wall for Ch3 and right well wall for Ch4 images). To overcome this, Ch3 and Ch4 images were added, and the resulting composite image was used for object thresholding from a whole-well region of interest (ROI). All objects within the ROI identified from the thresholding operation were enumerated and the classified object data was saved in a tabulated form (.csv), providing object-by-object as well as average object data for each well.

Cell colonies were analyzed automatically for size, number, and total colony area. Fluorescence images for 0.01-, 0.1-, 1-, and 10-nM staurosporine concentrations are shown in Figure 3. A clear change in cell-colony size and morphology is observed. In this assay it was found that the biology was well-represented by a simple calculation of the total colony area per well. A plot of total colony area versus staurosporine concentration is shown in Figure 4. The results demonstrated that A549 colony formation was inhibited in a concentration dependent manner by staurosporine with an IC50 of 0.62 nM for colony-growth inhibition.


Figure 3

Conclusion

3-D colony assays present many imaging and computational challenges with traditional microscope systems. Nonuniform staining, meniscus effects, and the requirement of multiple z-sections makes microscopes ill-suited for HTS applications. These problems have been overcome with the use of layered media in MatriCal 96-well plates, the LavaCell dye, and the IsoCyte scanning cytometer. This system produces a cell-colony growth assay protocol that is ideal for HTS and drug discovery efforts. An entire plate can be scanned and analyzed in 5 minutes, and integration with a plate handler (Twister II) allows for walk-away operation. The rapid image acquisition and robust image analysis process for quantitation of cell colony growth makes it possible to gather information about signals involved in normal cell and tumor architecture that could only be obtained with great effort using conventional approaches. This new information holds much promise for new insights into more effective cancer therapeutics.


Figure 4

Steven C. Miller, Ph.D., is vp, biology, and Evan F. Cromwell, Ph.D., is president at Blueshift Biotechnologies.
Web: www.blueshiftbiotech.com. Phone: (408) 773-1050.
E-mail: [email protected].

Previous articleFirst Targeted Therapy for Renal Cell Carcinoma Greenlighted
Next article454 Life Sciences and Baylor College of Medicine Sequence Genome for Less than $1M