Elsevier

Methods in Enzymology

Volume 506, 2012, Pages 291-309
Methods in Enzymology

Chapter fifteen - A Quantitative Method for Measuring Phototoxicity of a Live Cell Imaging Microscope

https://doi.org/10.1016/B978-0-12-391856-7.00039-1Get rights and content

Abstract

Fluorescence-based imaging regimes require exposure of living samples under study to high intensities of focused incident illumination. An often underestimated, overlooked, or simply ignored fact in the design of any experimental imaging protocol is that exposure of the specimen to these excitation light sources must itself always be considered a potential source of phototoxicity. This can be problematic, not just in terms of cell viability, but much more worrisome in its more subtle manifestation where phototoxicity causes anomalous behaviors that risk to be interpreted as significant, whereas they are mere artifacts. This is especially true in the case of microbial pathogenesis, where host–pathogen interactions can prove especially fragile to light exposure in a manner that can obscure the very processes we are trying to observe. For these reasons, it is important to be able to bring the parameter of phototoxicity into the equation that brings us to choose one fluorescent imaging modality, or setup, over another. Further, we need to be able to assess the risk that phototoxicity may occur during any specific imaging experiment. To achieve this, we describe here a methodological approach that allows meaningful measurement, and therefore relative comparison of phototoxicity, in most any variety of different imaging microscopes. In short, we propose a quantitative approach that uses microorganisms themselves to reveal the range over which any given fluorescent imaging microscope will yield valid results, providing a metrology of phototoxic damage, distinct from photobleaching, where a clear threshold for phototoxicity is identified. Our method is widely applicable and we show that it can be adapted to other paradigms, including mammalian cell models.

Introduction

In our experience, running the imaging facilities at the Institut Pasteur in Paris (www.imagopole.org) managing phototoxicity is critical to long-term live cell imaging studies on infectious processes. Host cell–pathogen interactions are especially challenging when it comes to their reconstitution within the context of meaningful experimental imaging paradigms, and phototoxic effects are an abundant source of problems. For example, approaches using multidimensional live cell imaging for studies on infection have become near routine as recourse to analyze subcellular dynamics (Frischknecht and Shorte, 2009, Shorte and Frischknecht, 2008). The difficulty of such approaches comes from the need to maintain spatial and temporal resolution using protocols that assure over-sampling x, y, z, and t. To achieve this, automated, high-speed acquisition aims to sample x, y “stacks” as rapidly as possible to satisfy the requirement that 3D volumes are acquired in a snapshot. Further, the 3D stack must be repeatedly sampled over time at a frequency determined by the temporal dynamics of the process being recorded. Add to this the need for multiple wavelength channels acquired at any given moment, allowing to distinguish distinct targets, which must then be colocalized, it is not uncommon to require 50–150 images to be acquired at each time point. This can amount to the need for an elevated light budget. Overcoming the limitations imposed by this light budget can often be the key to successful imaging. In the context of a real study, for example, following fluorescently labeled HIV virus (Arhel et al., 2006), bacteria (Enninga et al., 2005), parasites (Amino et al., 2006, Amino et al., 2007, Thiberge et al., 2007), or prion protein (Gousset et al., 2009), the need for extensive light exposure can have substantial impact on the quality of the data, due to photobleaching that comprises the signal-to-noise ratio of the detectable signal, and phototoxicity that risks to perturb the processes under study.

A common misconception is to equate photobleaching with phototoxicity. Photobleaching is specific to fluorescence microscopy and arises due to the loss of fluorescent signal that occurs when fluorophores are excited into a state leading to an irreversible loss of signal. Phototoxicity, on the other hand, is a related phenomenon inasmuch as it may be precipitated by photobleaching of fluorophores, but not necessarily. It may also occur in the absence of fluorophore. Phototoxicity is a generalized term used as a catch-all to describe how exogenous light energy may interact with the tissue/cell metabolism (for detailed and extensive review, see Diaspro et al., 2006). The term certainly refers to all those diverse processes resulting in light-induced free-radical generation, for example, from fluorescent labels, and/or light-sensitive metabolites. However, it also describes indirect effects such as localized thermal flux generation (undesired light-induced heating effects); light-induced ionizing, polarizing, and/or trapping effects; and of course unintended light-induced activation of membrane conductances. In turn, phototoxicity may result in extreme phenotypes such as cataclysmic cell death by, for example, free radicals rupturing cell membranes, and collapsing chemical and ionic compartmentalization. Such behaviors are easy to detect and reject before further analysis is performed. On the other hand, and much more problematic, phototoxicity may cause subtle effects, which are difficult to detect, or even distinguish because they do not kill the tissue, but rather subvert its functions. In the case of quantitative light microscopy, there is no ground truth inasmuch as it is the experimental device itself, the microscope, which induces these effects meaning that even careful control experiments may not be sufficient. Thus, our only remaining recourse to managing phototoxicity is to measure it and minimize risks by experimental design. Unfortunately, this is not a trivial task.

Section snippets

The need for a quantitative, generic, and convenient measure of phototoxicity

While it is rather well known, and somewhat implicit, that the impact of phototoxicity will vary with the amount of light delivered to the sample, much less clear are its underlying mechanisms. For example, while every biologist using live cell microscopy techniques will have an opinion on the subject, it is hard to know how a single-point-scanning confocal microscopy may be better or worse than, say, a multi-point-array-scanning confocal or a wide-field microscope. This uncertainty is due

A live specimen-based metrology

In Life Sciences, microscopes are mainly used to gather information on live specimen. It is not surprising then that assessing the performance of these microscopes involve using a live sample, particularly when it comes to quantifying the impact of imaging on live specimen. Measuring other performance criteria may not require stepping in the complexity of dealing with live samples. The rate of photobleaching, for instance, focuses on a fluorophore and its environment. To measure it, one can

Conclusion

We present a method and standard providing a non-ambiguous and singular readout reporting the phototoxicity of any imaging microscope system. The method yields a phototoxicity threshold value, which can be used to compare configuration settings in the same modality or even other. The protocol was engineered to be portable and convenient, so that it can be used in any lab. Using this approach, we demonstrate spinning-disk confocal imaging to be characterized by a lower phototoxicity threshold

Acknowledgments

We thank Johan Henriksson for a stimulating discussion. This work was funded by the European Commission FP7 Health (project “LEISHDRUG”, www.leishdrug.org, SLS) and ICT (project “MEMI”, www.memi-fp7.org, SLS), the Conny-Maeva Foundation (USA), and the Institut Pasteur Paris. J.D. received a fellowship from the Pasteur-Foundation (New York).

References (29)

  • A. Diaspro

    Photobleaching

  • R. Dixit et al.

    Cell damage and reactive oxygen species production induced by fluorescence microscopy: Effect on mitosis and guidelines for non-invasive fluorescence microscopy

    Plant J.

    (2003)
  • J. Enninga

    Secretion of type III effectors into host cells in real time

    Nat. Methods

    (2005)
  • Imaging host-pathogen interactions

    Biotechnol. J. (Special Edition)

    (2009)
  • Cited by (65)

    • Dose-independent threshold illumination for non-invasive time-lapse fluorescence imaging of live cells

      2021, Extreme Mechanics Letters
      Citation Excerpt :

      In addition, it is possible to have effects associated to photo-toxicity that are not necessarily noticeable. Such external effects can compromise data accuracy and experimental validity [18,19]. Hence, for live-cell imaging, it is imperative to eliminate the artifacts of light by choosing a suitable illumination protocol, which remained elusive for a long time.

    View all citing articles on Scopus
    View full text