Development of quantitative optical microscopy methods for single cell analysis

Doktorsavhandling, 2009

Optical microscopy has brought powerful methods to biological research, in particular the possibility to assay dynamic molecular processes within living individual cells. This thesis deals with various aspects of fluorescence and Raman microscopy, such as the possibility of avoiding cellular damage or stress caused by illumination, how to manipulate single cells for spectroscopic analysis and how to quantify images automatically. Taken together, it is hoped that the results will help to establish guidelines for non-invasive imaging and spectroscopy and contribute towards the development of quantitative optical microscopy measurements. The question of how to perform non-invasive measurements was considered in three of the appended papers. In Paper I and II, a novel dual-beam Raman microscopy/optical tweezers set-up was evaluated by analyzing the oxygenation of hemoglobin in single, optically trapped, red blood cells. Photodamage was significant, but could be decreased by trapping at 830 nm rather than at 1064 nm and by exciting Raman spectra using 568 nm instead of 488-514 nm laser light. In Paper V, light induced stress caused by time lapse fluorescence microscopy of budding yeast was evaluated by studying the nuclear localization of a stress activated transcription factor, Msn2p-GFP. The method indicated stress at significantly lower illumination levels than what was revealed by growth defects or morphological changes, but no stress was found for light doses below ~0.16 J/cm2. Above this level, a given light dose caused more stress if it was distributed over a short than over a long exposure time. Paper III and VI dealt mostly with stress mechanisms in budding yeast. In Paper III, fluorescence microscopy was used to investigate the salt sensitivity of deletion mutants in the golgi/endosome/vacuole transport process (vps mutants). With the hypothesis that the salt sensitivity is related to an improper production, processing and membrane integration of the sodium pump Ena1p, selected deletion mutants were transformed with a plasmid to express GFP-tagged Ena1p. Fluorescence imaging indicated that the salt sensitivity of some of the vps mutants is correlated to a delay of Ena1p arrival to the plasma membrane. In Paper VI, the mechanism behind tellurite and tellurium (Te) toxicity was investigated in budding and fission yeasts. Raman and optical microscopy revealed that Te plaques formed in tellurite stressed cells. A genome-wide screen in the S. cerevisiae gene knockout collection indicated that Te accumulation and tellurite detoxification is mediated via the sulphur assimilation and L-methionine biosynthesis pathways. The image analysis in Paper III was done manually, which proved to be both time consuming and difficult to perform in an unbiased manner. This provoked the development of an image analysis software package, which is presented in Paper IV. The algorithms presented can be used for cell contour recognition of non-stained cells and statistical analysis of fluorescence intensity and localization. The algorithms proved instrumental for the quantification of Msn2p localization in Paper V.