The Development of an Automated, Fluorescent, Quantitative Three-Dimensional In Vitro Organotypic Invasion Assay

Abstract

Tumour metastasis is a complex multi-step biological process, which involves dispersion of cancer cells to various organ sites within the body. The tumour-stroma interface plays an important role within cancer progression and is a major area of focus within cancer research. It is imperative that we are able to understand what mechanisms the tumour cells use in order to achieve maximum invasive potential. Fibroblasts are one of the major host cell types that are usurped by the tumour cells to promote invasion. Thus, we must understand the molecular mechanisms by which fibroblasts are transformed by the tumours into tumour-promoting fibroblasts, thus enabling cancer progression. Over 10 years ago our laboratory developed a quantitative image analysis methodology for analysing tumour and stromal fibroblast-interactions using 1ml collagen type I based three-dimensional (3D) organotypic gels. The assay is expensive, slow and can take 3-4 weeks before results are available. In this study I have developed two smaller versions of this assay that enable drug/gene screening in 3D, that are quicker to analyse and affordable; mini-organotypics (100μl) and micro-organotypics (50μl). Using fluorescently labelled tumour (red) and fibroblast (green) populations, it is possible to view the invasive process live and in situ, using confocal microscopy. Analysis of fluorescent organotypic gels after 72 hours, showed that the fibroblasts are the principal invading cell type that commence invasion and are followed by the tumour cells. For a proof-of-principle study I transfected human dermal fibroblasts with a customised SMARTpool siRNA library targeting proteases. Data collection was automated, thus allowing results within three days instead of three weeks. Relative to siRNA control treated fibroblasts, 16 out of 60 targeted proteases significantly reduced fibroblast and tumour cell invasion. Moreover, re-analysis of five lead proteases using individual siRNA for each gene identified that fibroblast MMP-17 and TMPRSS-13 is required by dermal fibroblasts to support invasion of cancer cells within this 3D assay. I believe this novel 24-well 3D fluorescent mini-organotypic invasion assay will allow drug/gene screens to be conducted affordably on multiple different tumour-stroma cell combinations. Moreover, the 96-well 3D fluorescent micro-organotypic invasion assay is almost ready for use in high-throughput screening and thus will be of great value in both academia and industry.