Here, we document the use of the soft agar colony formation assay to test the effects of a peptidylarginine deiminase (PADI) enzyme inhibitor, BB-Cl-amidine, on breast cancer tumorigenicity in vitro.
Given the inherent difficulties in investigating the mechanisms of tumor progression in vivo, cell-based assays such as the soft agar colony formation assay (hereafter called soft agar assay), which measures the ability of cells to proliferate in semi-solid matrices, remain a hallmark of cancer research. A key advantage of this technique over conventional 2D monolayer or 3D spheroid cell culture assays is the close mimicry of the 3D cellular environment to that seen in vivo. Importantly, the soft agar assay also provides an ideal tool to rigorously test the effects of novel compounds or treatment conditions on cell proliferation and migration. Additionally, this assay enables the quantitative assessment of cell transformation potential within the context of genetic perturbations. We recently identified peptidylarginine deiminase 2 (PADI2) as a potential breast cancer biomarker and therapeutic target. Here we highlight the utility of the soft agar assay for preclinical anti-cancer studies by testing the effects of the PADI inhibitor, BB-Cl-amidine (BB-CLA), on the tumorigenicity of human ductal carcinoma in situ (MCF10DCIS) cells.
Both non-transformed (normal) and transformed cells can readily proliferate in a 2D monolayer culture. This form of adherent cell growth is quite dissimilar from that which occurs in vivo where, in the absence of mitogenic stimulation, cells do not often rapidly divide within their microenvironment. The soft agar assay on the other hand is distinct from 2D culture systems because it quantifies tumorigenicity by measuring a cell’s ability to proliferate and form colonies in suspension within a semi-solid agarose gel1. In this setting, non-transformed cells are unable to rapidly propagate in the absence of anchorage to the extracellular matrix (ECM) and undergo apoptosis, a process known as anoikis. In contrast, cells that have undergone malignant transformation lose their anchorage dependence due to activation of signaling pathways such as phosphatidylinositol 3-kinase (PI3K)/Akt and Rac/Cdc42/PAK. Therefore, these cells are able to grow and form colonies within the semi-solid soft agar matrix2.
A common use of the soft agar assay is to test whether specific compounds, such as PADI inhibitors, are able to suppress tumor growth in vitro. In general, colony count or colony sizes are quantitative read-outs from the assay that can be compared between control and treatment groups to assess differences in cellular tumorigenicity. Therefore, if one finds that colony formation is inversely correlated with increasing drug concentration, then a conclusion could be drawn that the drug is an effective inhibitor of tumorigenicity in vitro. On the other hand, if the drug does not affect colony formation, the drug is either not at the appropriate dosage or it is not an effective tumorigenic inhibitor. Aside from using a soft agar assay to test the anti-tumor effect of a drug, this assay can also be used to probe the relationship between a specific gene and tumorigenesis. For example, the effect of suppressing PADI2 expression on tumorigenicity can be addressed by PADI2-specific siRNA treatment.
PADIs are calcium-dependent enzymes that post-translationally modify proteins by converting positively charged arginine residues into neutrally charged citrulline in a process known as citrullination or deimination3-5. We have recently found that peptidylarginine deiminase 2 (PADI2) may function as a novel breast cancer biomarker and that PADI inhibitors represent candidate therapies for early stage breast cancers6. For example, we have previously demonstrated that a “pan-PADI” inhibitor, Cl-amidine, suppresses the proliferation of breast cancer cells using 2D monolayers and that the inhibitor suppressed the growth of 3D tumor spheroids6. In this report, we extend these studies, and highlight the utility of the soft agar assay, by testing the efficacy of a new PADI inhibitor, BB-CLA, in suppressing the growth of MCF10DCIS breast cancer colonies7. We note that we used MCF10DCIS cells for this experiment because they are oncogenic derivatives of non-transformed human MCF10A cells and because they contain high steady state levels of PADI2 protein8. We hypothesize that PADI2 enzymatic activity plays a key role in the tumorigenicity of this cell line and that BB-CLA-mediated inhibition of PADI2 activity will suppress cancer progression.
1. Preparation of 3% 2-Hydroxyethyl Agarose
2. Preparation of the Bottom Layer: 0.6% Agarose Gel
3. Preparation of the Cell-containing Layer: 0.3% Agarose Gel
4. Preparation of the Feeder Layer: 0.3% Agarose Gel
5. Data Collection
The soft agar colony formation assay can be used for a broad range of applications documenting the tumorigenicity of cancer cells. A major advantage of this technique is that the semi-solid matrix selectively favors the growth of cells that can proliferate in an anchorage-independent manner. This trait is mainly exhibited by cancer cells but not by normal cells. We primarily use this technique to test the efficacy of tumor growth inhibition by drugs and to test for the effect of overexpression or depletion of our genes of interest, including PADI genes, on the tumorigenicity of breast cancer cells. Here, we assessed the effect of BB-CLA on tumorigenic inhibition of PADI2-overexpressing MCF10DCIS cells (Figure 1 and 2).
The results demonstrate that BB-CLA significantly inhibits the formation of MCF10DCIS cell-derived colonies. Figure 3 shows that, in the presence of a PADI inhibitor, there was a reduction in both colony formation and colony size when compared to the DMSO control. [Note: BB-CLA was dissolved in DMSO and thus, DMSO was used as a control.] The size of colonies for BB-CLA treated MCF10DCIS cells were predominantly within the range of 20 to 100 µm while the size of colonies for the DMSO control exhibited a greater range of 70 to 150 µm after 2.5 weeks of growth.
Colonies larger than 70 µm were counted and analyzed (Figure 4). There was an average of 3,536 colonies in the DMSO control whereas only 1,967 colonies were seen in the BB-CLA treated group after 2.5 weeks of soft agar culture. This represents a 44% decrease in the average colony formation in the presence of 1 µM BB-CLA, indicating a significant tumorigenic inhibition of breast cancer cells (MCF10DCIS cells) by the PADI inhibitor.
Figure 1. Chemical Structure of BB-CLA.
Figure 2. Schematic Overview of the Protocol for the Soft Agar Colony Formation Assay. Each well of the 6-well culture plates was first coated with 0.6% agarose gel (bottom layer). A 0.3% agarose gel mixture containing the MCF10DCIS cells and either the BB-CLA inhibitor (1 µM) or DMSO (control) was then layered on top of the 0.6% gel. Once a week, the 0.3% agarose gel mixture (containing BB-CLA) was added on top of the soft layer. After 2 to 4 weeks, colony formation was observed and counted for data analysis. Please click here to view a larger version of this figure.
Figure 3. Images of Colony Formation in BB-CLA-treated MCF10DCIS Cells. MCF10DCIS cells were grown in soft agar in the presence (1 µM BB-CLA) or absence of BB-CLA (0 µM DMSO). After 2.5 weeks, colonies were imaged using an inverted microscope for low magnification images and a light microscope for high magnification images. Please click here to view a larger version of this figure.
Figure 4. Quantification of MCF10DCIS Colony Number after BB-CLA Treatment. MCF10DCIS cells were grown in soft agar at different concentrations of BB-CLA (0 µM (DMSO), or 1 µM BB-CLA). After 2.5 weeks, individual colonies larger than 70 µm were counted. Experiments were repeated 4 times (n = 4). The statistical significance of total cell count difference between the control and treatment samples was determined by two-sample Student’s t-test (*p < 0.005).
The rate of colony formation in soft agar varies depending on the cell type9. Therefore, the number of cells to start with should be optimized and adjusted accordingly. A suggested starting range is between 5 x 102 to 1 x 104 cells per well using a 6-well plate. In addition, colony size varies depending on the growth rate of each cell. Therefore, a predefined a cut-off for colony size is needed to annotate individual colonies for downstream quantitative analyses. Here, colonies larger than 70 µm were quantified to avoid inclusion of non-proliferating cells derived from the initial plating.
For optimal growth, when making the 3D agarose gels, it is advisable to use the same cell culture medium normally used for 2D cell cultures. This is because changing medium often times alters the cells growth rate, thus necessitating additional passaging to allow the cells to adapt to the new medium. For example, MCF10DCIS cells can be grown optimally in RPMI-1640, DMEM, or DMEM/F12 media. However, when MCF10DCIS culture media is changed from RPMI-1640 to DMEM/F12, there is a noticeable reduction in cell proliferation rate. Additionally, care should be taken to maintain serum concentrations at a constant level because, without serum, colony formation will be inhibited10. Furthermore, after the agarose is microwaved, allow the bottle to equilibrate in a 45 °C water bath before mixing with media containing cells to prevent overheating the cells. Given that agarose solidifies rapidly at room temperature, we also recommend that all pipettes and 6-well culture plates are preheated in a 37 °C incubator before adding agarose solution to prevent premature solidification while handling.
While the soft agar colony formation technique is a valuable tool for measuring tumorigenicity in a range of cancer cell lines, some lines do not grow in soft agar. Another potential drawback of the method is that viable cells cannot be recovered once they are plated into the soft agar. Additionally, if a compound has a shorter bioavailability period, the feeding step will need to be performed more frequently (i.e. every other day) and the samples will need to be collected in less than seven days. The threshold for the colony size will also need to be reduced while still allowing for potential differences between the samples to be observed. In these instances, controls would need to be incorporated to optimize experimental parameters to minimize false-positive and false-negative results. Furthermore, this technique can be time consuming and difficult when testing a large number of samples. However, technological advances have helped to overcome some of these limitations. The conventional soft agar assay (as documented in this manuscript) typically uses 6-well culture plates or 6mm culture dishes. With automated plate readers, however, multiple samples can be processed in 384-well plates11. For example, investigators pre-loaded tumor cells with dyes such as alamarBlue and tetrazolium dyes and colonies are quantified using plate reader, thus obviating the need to manually count colony number11. This high-throughput capability is, therefore, amenable for large scale cancer drug screens.
In sum, the soft agar assay is a valuable pre-clinical technique that can be used to assess the tumorigenicity of a wide-range of cancer cells (breast, prostate, ovarian, and others) with regards to their sensitivity to drugs, hormones, heat, hypoxia, and a multitude of other treatment conditions. The assay continues to provide a straightforward and informative tool for cancer researchers who wish to better understand the mechanisms of cancer progression and test the anti-tumor potential of new cancer therapies.
The authors have nothing to disclose.
We are thankful to Dr. Richard Cerione, Dr. Marc Antonyak, and Kelly Sullivan, Cornell University, for providing technical advice, and to Dr. Gerlinde Van de Walle, Cornell University, for sharing their Olympus CKX41 inverted microscope.
Zeiss Axiopot | Carl Zeiss Microscopy | 1021859251 | |
Inverted Microscope | Olympus | CKX41 | |
DMEM/F-12 | Lonza BioWhittaker | 12-719F | |
HyClone Donor Equine Serum | Fisher Scientific | SH30074.03 | |
Penicillin Streptomycin | Life Technologies | 15140-122 | |
2-Hydroxyethylagarose: Type VII, low gelling temperature | Sigma-Aldrich | 39346-81-1 |