The intermediate filament cytoskeleton is a target of statins and statin‐induced cancer cell death

Intermediate filaments (IFs), together with microfilaments and microtubules, form the cell cytoskeleton. IF proteins in humans are encoded by 73 genes and expressed in a tissue‐specific manner. IFs are critical for mechanical protection, stress sensing, growth, and survival, among other cellular functions. Missense mutations in IF genes that compromise filament assembly cause, or predispose to, over 70 human diseases. Despite their functional and disease importance, IF proteins lack pharmacological agents for selective targeting. The objective of the present study was to identify IF‐targeting drugs from a library of highly pure and biologically‐active compounds. Methods Each of 1120 compounds in the Tocriscreen TM library was applied at an initial concentration of 10μM for 1hr to SW13 human adrenal carcinoma cells expressing vimentin IFs. An image‐based screen of the vimentin IF network was performed by indirect immunofluorescence, and the effects on vimentin filament bundling or depolymerization were quantified using an image processing program (NIH ImageJ). Results Out of 22 initial positive hits, 7 were expected (they included compounds targeting kinases, phosphatases, and microtubules), validating the screening method. The 15 remaining novel unexpected hits were subjected to a second screen. The compound that elicited the most dramatic effects on vimentin IFs was simvastatin, a 3‐hydroxy‐3‐methyl‐glutaryl‐CoA (HMG‐CoA) reductase inhibitor used for cholesterol lowering. Simvastatin caused a rapid and profound perinuclear bundling of vimentin, and a 50% reduction in the area of the vimentin IF network relative to vehicle treatment, without significant changes in vimentin protein solubility. The effects of simvastatin on vimentin IFs were observed at 15min, plateaued after 60min, and exhibited dose‐dependence between 10nM‐10μM. Similar dose‐ and time‐dependent effects of simvastatin were observed on epithelial keratin IFs expressed in human hepatocellular carcinoma HepG2 cells. The IF‐targeting effects were statin‐specific, such that mevastatin acted similarly to simvastatin, whereas lovastatin and pravastatin lacked vimentin bundling activity. Importantly, triple immunostaining revealed that the early cytoskeleton‐modifying effects of simvastatin were selective for IFs, since the organization of microfilaments and microtubules was unaffected at up to 4 hours of drug treatment. Statin use is associated with reduced cancer‐related mortality, and statins are known to inhibit cancer cell proliferation and to promote cancer cell death. Therefore, the role of vimentin in simvastatin‐induced cell death was investigated by comparing responses of vimentin‐expressing and vimentin‐lacking SW13 cells (the latter also lack all other cytoplasmic IFs). A dose‐dependent (10nM‐10μM) decrease in cell number, concomitant with an increase in cell death was observed at 8hr and 24hr of simvastatin treatment in SW13 vimentin‐expressing cells. In contrast, the viability of SW13 vimentin‐lacking cells was esentially unaffected by simvastatin treatment. Conclusions This study demonstrates that the IF cytoskeleton is a target of statins, and suggests that this mechanism may underlie the sensitivity of cancer cells to statin‐induced death, independent of cholesterol metabolism. Support or Funding Information Supported by NIH grants DK093776 and DK102450.