MDA PHOTOMETRY IN SOLUTIONS: COMPARISON OF METHODS OF DIRECT OPTICAL DENSITY DATA AND DERIVATIVE SPECTROSCOPY

In this study, we analyzed the absorption spectra of the reaction products of aqueous extracts of mouse tissues with thiobarbituric acid, with the aim to determine the concentration of malonic dialdehyde (MDA) in them. The concentration of MDA is an important part of the analysis of the redox status of tissues, which is important in the study of inflammatory reactions, for example, after various stressful effects, as well as in the study of aging. In normal practice, they calculate the concentration of MDA in a solution by its optical density at 532 nm, then these data are related with similar solutions of the tetramethoxypropane (TMP) reaction with known concentration. We have shown that in cases of submicromolar MDA content, fluctuations in the nonspecific background level of the optical density of solutions can be commensurate to the magnitude of the actual absorption signal of the resulting colored adduct. Solutions of biological tissue extracts, due to the additional content of proteins, lipids and carbohydrates, are characterized by turbidity, which distorts the absorption spectrum non-linearly. The second derivatives of the absorption spectra deprived of background scattering distortions and can be used for automatic software calculation of the pigment content. Obtaining correct derivatives is complicated by the need to smooth the original spectra. We used two smoothing methods: the moving average method and the Savitsky–Goley filter with a polynomial of the third degree. We compared the data obtained on the basis of measuring the optical densities of solutions at 532 nm with those based on the analysis of the second derivatives of their absorption spectra, and also on the basis of integral sums of the second derivatives in the range of 520-550 nm. The results of calculations using the second derivatives gave 2-5 times lower concentrations of MDA than those obtained from optical densities at the maximum absorption of the adduct. At the same time, the convergence of the data, especially when using integral sums of the second derivatives, turned out to be significantly better than for the zero order, and the resulting errors were 2-3 times smaller.