Abstract : Detection and identification of proteins either in 2D electrophoresis gels or in biological media depends largely on the quality of protein sensitive dye used (generally a fluorescent dye) as well on the efficiency of their fluorescent properties (quantum yield, Stokes' shift and overlap of excitation frequency with common laser sources), their stability (chemical and photophysical) and their accessibility. To be able to create a range of new fluorescent dyes whose profiles can be modulated starting from a proven scaffold is advantageous in terms of predicting the properties of the final products. Even if many fluorescent dyes are available on the market, they suffer from a prohibitive price, a lack of sensitivity or a restricted wavelength range (absorption/emission) or limited stability. Access to a family of fluorophores displaying similar physicochemical properties, but shifted excitation and emission wavelengths constitutes an invaluable tool Epicocconone, isolated from the fungus Epicoccum nigrum, covalently binds to primary amines (e.g. lysine of proteins), leading to a protein linked conjugate which is strongly fluorescent, emitting light in the red (610 nm) when submitted to UV-visible light (395 or 520 nm). This is a novel pro-fluorophore that allows protein quantification with unprecedented sensitivity and with an excellent linearity on a broad range of concentrations. A program of synthesis of epicocconone and analogues in order to provide a library of complementary fluorophores has been undertaken. The results concerning either the better efficiency of the analogs and the structure-fluorescence relation are presented in this dissertation.