Live-cell imaging of drug-treated cells – challenges beyond routine microscopy

Session

Poster Session

Authors

Gerald Timelthaler (1)

Affiliations

1. Medical University of Vienna - Center for Cancer Research

Keywords

live cell imaging, spinning disk confocal, fluorescent drugs, phototoxicity, photobleaching

Abstract text

Live-cell imaging - as routinely performed in cancer research - usually involves trusted staining protocols with high fidelity dyes and robust fluorophores. This is especially true when observing intracellular mechanisms that require the highest possible resolution on conventional confocal microscopes. However, when analyzing drug uptake and/or intracellular drug dynamics of intrinsically fluorescent compounds, unknown and unexpected dynamics/effects might occur, requiring high flexibility in acquisition settings. 

Several of our studies use live-cell microscopy, with the aim of following over time the intracellular distribution of experimental or already clinically approved drugs for treatment of cancer patients. These substances sometimes harbor intrinsic fluorescence properties. This implies, on one hand, that commercially available fluorophores with comparable fluorescence properties as the analyzed drug cannot be used in these experiments. On the other hand, the intrinsic fluorescence of the respective substance itself can represent an ideal tool for following its intracellular behavior. This bears tremendous potential to reveal new scientific insights, that were not even considered before. However, as the characteristics of the drug-intrinsic fluorescence properties are a priori not established or characterized in detail, unexpected, technically challenging situations can arise during data acquisition. 

In a current study, we treated cancer cells with the clinically approved and intrinsically fluorescent tyrosine kinase inhibitor (TKI) nintedanib, a small molecule RTK inhibitor that also exerts lysosomotropic behavior and shows strong green fluorescence upon intercalation into acidic intracellular vesicles. During these experiments, we were facing the following microscopy challenges: 1) Live-cell microscopy became imperative as method, as upon cell fixation (regardless of the method) the green fluorescent signal of the drug immediately vanished. 2) Since lysosomes are one of the smallest membrane-enclosed organelles, maximum resolution on the confocal microscope was needed in addition to optical sectioning and subsequent image processing to visualize the exact localization of the compound in live cells. In our setup we worked with a spinning disk confocal microscope with incubation chamber and a 100X objective with an additional 3.2X magnification lens. For improved image data quality and resolution, raw images were processed with deconvolution algorithms and - when necessary- image plane projections were performed. 3) Photobleaching and phototoxicity severely limited the time frame for live-cell observation. This poses an extra challenge, when e.g. z-stacking or several consecutive images at different time points from the identical position were needed. Moreover, laserlight-induced damage of cellular organelles due to distinct photosensitivity can obscure the real nature of an observed phenomenon/structure. For example, many lysosomotropic drugs are described to induce phospholipidosis, a pathological condition where – due to lysosomal damage – lipids cannot be processed properly and un-degraded remnants accumulate in the lysosomes, forming so called multi-lamellar bodies (MLBs). We observed in live cell microscopy, that rapid shrinking of large drug-positive vacuoles due to photosensitivity occurred, leading to structures resembling MLBs but de facto represented collapsed vesicles originally not filled with lipid membranes. Hence, it can be challenging to correctly interpret these data. 

Together, our experience with live-cell microscopy of cells treated with intrinsically fluorescent drugs underlines the importance of high performance microscope setups, with high resolution optics, fast z-stacking and refined image post processing. Finally, and most importantly, a special focus must be put on accurate experiment planning. This involves the necessity for (interdisciplinary) discussions plus an extra time scheduling. However, to gain reliable, new, and exciting results, this is - especially in this context -, inevitable.