Benchmarking of protein carbonylation analysis in Caenorhabditis elegans: specific considerations and general advice

S Pyr Dit Ruys,J-M Bonzom,S Frelon
DOI: https://doi.org/10.1016/j.freeradbiomed.2016.08.014
Abstract:Oxidative stress has been extensively studied due to its correlation with cellular disorders and aging. In proteins, one biomarker of oxidative stress is the presence of carbonyl groups, such as aldehyde and ketone, in specific amino acid side chains such as lysine, proline, arginine and threonine, so-called protein carbonylation (PC). PC study is now a growing field in general and medical science since PC accumulation is associated with various pathologies and disorders. At present, enzyme-linked immunosorbent assays (ELISA) seem to be the most robust method of quantifying the presence of carbonyl groups in proteins, despite having some recognised caveats. In parallel, gel-based approaches present cross-comparison difficulties, along with other technical problems. As generic PC analyses still suffer from poor homogeneity, leading to cross-data analysis difficulties and poor results overlap, the need for harmonisation in the field of carbonyl detection is now widely accepted. This study aims to highlight some of the technical challenges in proteomic gel-based multiplexing experiments when dealing with PC in difficult samples like those from Caenorhabditis elegans, from protein extraction to carbonyl detection. We demonstrate that some critical technical parameters, such as labelling time, probe concentration, and total and carbonylated protein recovery rates, should be re-addressed in a sample-specific way. We also defined a procedure to cost-effectively adapt CyDye™-hydrazide-based protocols to specific samples, especially when the experimental interest is focused on studying differences between stimulating conditions with a maximised signal-to-noise ratio. Moreover, we have improved an already-existing powerful solubilisation buffer, making it potentially useful for hard-to-solubilise protein pellets. Lastly, the depicted methodology exemplifies a simple way of normalising carbonyl-related signal to total protein in SDS-PAGE multiplexing experiments. Within that scope, we also proposed a simple way to quantify carbonyl groups by on-gel spotting diluted dye-containing labelling buffer. Proof of the robustness of the procedure was also highlighted by the high linear correlation between the level of carbonyls and the ultraviolet exposure duration of whole worms (R2=0.993). Altogether, these results will help to standardise existing protocols in the growing field of proteomic carbonylation studies.
What problem does this paper attempt to address?