Refinement of two-dimensional electrophoresis for vitreous proteome profiling using an artificial neural network

Abstract

Despite technological advances, two-dimensional electrophoresis (2DE) of biological fluids, such as vitreous, remains a major challenge. In this study, artificial neural network was applied to optimize the recovery of vitreous proteins and its detection by 2DE analysis through the combination of several solubilizing agents (CHAPS, Genapol, DTT, IPG buffer), temperature, and total voltage. The highest protein recovery (94.9% ± 4.5) was achieved using 4% (w/v) CHAPS, 0.1% (v/v) Genapol, 20 mM DTT, and 2% (v/v) IPG buffer. Two iterations were required to achieve an optimized response (580 spots) using 4% (w/v) CHAPS, 0.2% (v/v) Genapol, 60 mM DTT, and 0.5% (v/v) IPG buffer at 35 kVh and 25 °C, representing a 2.4-fold improvement over the standard initial conditions of the experimental design. The analysis of depleted vitreous using the optimized protocol resulted in an additional 1.3-fold increment in protein detection over the optimal output, with an average of 761 spots detected in vitreous from different vitreoretinopathies. Our results clearly indicate the importance of combining the appropriate amount of solubilizing agents with a suitable control of the temperature and voltage to obtain high-quality gels. The high-throughput of this model provides an effective starting point for the optimization of 2DE protocols. This experimental design can be adapted to other types of matrices.