Array of fluorophores using a high throughput, silica gel-based combinatorial platform for analyte fingerprinting and discrimination

Jarne Lardiés, Carmen
1st International Caparica Conference on Chromagenic and Emissive Materials, IC3EM 2014
Participation type: 
Costa de Caparica (Portugal)

Since most of the theoretical models of the effect of non-covalent interactions on fluorescence only consider wavelength shifts and no changes in quantum yields, no systematic analytical use of these interactions for quantitative measurements has been reported, with the exception of the quenching effect [1].We have previously demonstrated that practically almost all molecules (potential analytes) produce a response in the emission of fluorophores (either increase or quenching) when interacting with them in a non-covalent way (either non-specific or specific interactions, respectively) at room temperature [2,3]. Sensitivity of this response is particularly high in silica gel medium [3]. Depending on the nature of fluorophore and analyte, changes in emission can additionally be associated to changes in emission wavelength. All this makes it possible to develop fingerprints or profiles of analytes based on their response against a collection of different types of fluorophores [4].
Development of an array of fluorophores for screening their interaction with analytes in a high-throughput format, using a silica gel-based combinatorial platform is presented here. This system may take advantage of the combined response of its components to create specific pattern responses or fingerprints for analyte discrimination.
Arrays were prepared using 49 commercial or freshly-synthesized fluorophores with different chemical properties. They were immobilized by spray-on spotting on HPTLC silica gel plates through non-covalent adsorption. Fluorescence measurements of sprayed bands were then performed at five different wavelengths using scanning densitometry. Subsequent spray spotting of the corresponding target analyte on fluorophore bands was performed, and further fluorescence measurements recorded. Cholesterol and sphingomielin, which are not fluorescent molecules and have poor spectroscopical properties, were used as analytes.
Parameters affecting repeatability of the procedure were studied concerning sample application, scanning densitometry, intra- and inter-plate runs, as well as the effect of time on fluorophore signal. Resulting 254 signals were treated by chemometric techniques, applying different univariate and multivariate calibration models. A profil per analyte was obtained. The use of profils as analytical signal will be discussed