Investigations into the phototoxic effects of multiphoton microscopy

Fluorescence microscopy is a technique that can reveal the organization and physiological state of living tissue probably better than any other. The main reasons for the pre-eminence of this technique are that it can provide high detection sensitivity, a high signal-to-background ratio, and a huge range of highly-specific fluorescent probes are currently available. These probes can be used to identify individual molecular species or to measure physiological parameters such as ion concentration. Unfortunately, fluorescence microscopy can all too often have detrimental effects on living tissue. Toxic effects to the specimen under study may arise due to a number of reasons:

  • Toxic effects of the exogenously-introduced probe that are independent of excitation.
  • Damage caused by the absorption of the excitation light by endogenous chromophores such as flavins (Sanford et al., 1986. In "Free Radicals, Aging, and Degenerative Diseases," p. 373. Wiley-Liss, Inc).
  • Damage caused by absorption of the excitation light by exogenous probes that have been introduced into the tissue.

Toxic effects inducted by the excitation illumination generally arise by non-radiative transitions to the ground-state following excitation that lead to the production of singlet oxygen or other toxic products. Such phototoxic effects can be harnessed for specific purposes. Probe-enhanced photoxicity is the basis for photodynamic therapy, an important, new clinical technique. Also, an experimental method for inactivating a target protein by the localized generation of free radicals has been developed. Our goal, however, is to observe specimens while preserving biological function by minimizing photodamage thereby facilitating extended observations.

Multiphoton excitation fluorescence imaging is an optical sectioning imaging technique that minimizes bulk fluorophore excitation (Denk et al. 1990 Science 248: 73). As was suggested in this original publication, we have seen dramatic viability improvements in 3-D and 4-D fluorescence microscopy of developing embryo imaging when using MPLSM compared to LSCM (Squirrell, et. al., 1999 Nature Biotechnology, in press). We have also witnessed greatly enhanced viewing times while imaging thin motile Ascaris spermatids with 1047nm MPLSM compared to 488nm LSCM (Wokosin, et. al., 1996. Proc.SPIE 2678: 38). This is particularly noteworthy for thin samples (less than 10 microns) since the excitation volume and effective excitation wavelength are roughly equivalent for MPLSM and LSCM. This success is not generally the case for all MPLSM excitation wavelengths, as phototoxic effects have been implicated via two-photon excitation of endogenous chromopores (Konig et al., 1997 Optics Letters 22(2): 135.)

The use of pulsed illumination adds other variables to study the phototoxicity problem. The characteristics of the pulse can be varied while maintaining fixed levels of two-photon excitation to assess the relative impact of three- or higher photon interactions with proteins and fluorophores as traded off with the potential for one-photon water absorption with subsequent sample heating

We are currently studying some of the mechanisms and thresholds of photodamage in both tissue culture cells and whole organisms. Our goal is to identify strategies such as choice of optimal excitation wavelength, illumination pulse width, and pulse repetition rate to minimize photodamage in practical experimental situations. In addition, we are investigating whether it is possible to improve an organism's resistance to phototoxic effects by over-expressing certain genes.