Résumé :
|
Communication n° 308. Expression of fluorescent proteins (GFP, DsRed) after gene transfer can be conveniently followed in living animals (mouse) directly by fluorescence video imaging. We routinely used this approach for a relative quantitative assay of plasmid electrotransfer in muscles. Several limits are affecting the quality of the assay. Different ways to overcome these difficulties were found. Skin and hair autofluorescences were affecting the detection. Colour detection appeared as the most effective approach. A liquid crystal filter driven by the software was a very convenient method. Hair root signals could be removed by image analysis. Tissue and skin are responsible for the blurring of the image. Light scattering brings a loss in transmitted light, both on the excitation and emission. It depends on the organisation and on the thickness of the sample. It brings blurring on the images of fluorescent objects when they are deep in the tissue as in the case of muscle fibers. An experimental phantom approach did show that turbidity of the tissue was not the major limit. Fluorescein capillaries (to mimic the fluorescent fibers)were detected in milk (a scattering buffer) with a good spatial definition as long as the signal to noise ratio was sufficient. This was in agreement with a much sharper definition of the GFP muscle fibers when the skin was removed. Detection would be improved with a high level of expression, a strong excitation beam, longer integration time of the signal or binning on the CCD chip. Detection was improved when the signals were red shifted where both troubles were reduced. Due to its anisotropic organization and the absorption by melanin, the skin appeared as a main barrier for detection.Deconvolution appeared limited by the difficulties to obtain a PSF on thick samples. Support of the CNRS CEA Imagerie du petit animal program should be acknowledged
|