Keratins are intermediate filaments (IF) found in abundance in epithelial cells. Depolymerisation of these filaments causes the cell to collapse and become more plastic. We have previously shown that SCFC may trigger depolymerisation of keratins through altered protein acetylation(Reference Leech, Evans and Shaw1, Reference Drake, Griffiths and Shaw2). The aim of this study was to develop a high-throughput method to quantify IF polymerisation and to apply as a screen for IF-perturbing nutrients and drugs. Three treatments were used in a proof-of-principle study: the anti-fungal drugs griseofulvin and cordycepin (the former a c-mitotic drug known to suppress microtubule growth(Reference Rasthinasamy, Jindal and Asthana3), the latter inducing abnormal mitosis by suppressing microtubule dynamics(Reference Zieve and Roemer4)) and sodium butyrate, a histone deacetylase inhibitor that disrupts IF formation in cancer cells via post-translational modification of keratins(Reference Leech, Evans and Shaw1).
A high content analysis (HCA) approach was developed. HCA is a tool allowing quantification of staining intensities by use of an Arrayscan II platform linked to a fluorescent microscope that can read ninety-six-well plates(Reference Drake, Griffiths and Shaw2). Sixty wells of a ninety-six-well plate were seeded with 2.5×103 MCF-7 cells in 100 μl RPMI media. Plates were incubated for 24 h at 37°C, after which, media were replaced with either griseofulvin treatment (2–200 μm, 48 h), cordycepin (0–60 μm, 15 min) or sodium butyrate (0–20 mm, 16 h). Treatment was removed and plates were fixed with ice-cold methanol for 5 min.
Immuno cytochemistry was used to visualise K8. An anti-K8 antibody and Hoechst stain were used to stain cells. Indicators of depolymerisation include K8 fluorescence, texture measurement, spot fibre count and spot fibre total area (arbitrary units). Filamentousness was measured using an algorithm based on co-occurrence of adjacent pixel intensity(Reference Drake, Griffiths and Shaw2). *Z prime values between 0.5 and 1 indicate suitability as a high-throughput assay. A paired student's t test was used to compare control and test values. †P<0.05 is statistically significant.
An HCA assay for intermediate filament integrity has been demonstrated, establishing a good proof of principle with griseofulvin. We are currently producing a rat monoclonal K8 antibody to be used in conjunction with a mouse tubulin antibody and phalloidin stain for actin, allowing development of a novel triple cytoskeletal stain assay with commercial potential.
