Normalization of protein at different stages in SILAC subcellular proteomics affects functional analysis

DOI: 10.5584/jiomics.v2i2.108

Abstract

Quantitative subcellular proteomics is a powerful method to interrogate spatial dynamics of cells or tissues. Stable isotope labeling by amino acids in cell culture (SILAC) is a popular quantitative approach that is ideally suited to subcellular proteomics because samples can be combined very early to reduce technical variability in the subcellular fractionation and downstream processing. However, validation of results using orthogonal methods such as immunoblotting do not allow mixing of samples prior to fractionation, leading to potentially different outcomes. Here we have investigated the impact protein normalization before or after subcellular fractionation has on the functional analysis and experimental conclusions. As a model system, we compared the detergent-resistant membrane (DRM) fraction of mouse embryonic fibroblasts (MEF) from caveolin-1-null mice with wildtype controls. Caveolin-1 is cholesterol-binding protein which is essential for formation of plasma membrane caveolae, a subtype of lipid raft membrane microdomains. Surprisingly, we found that the relative protein content of DRM as a percentage of total protein content is 1.6 fold higher for Cav1-/- MEF compared to wild type MEF, leading to different SILAC ratios in pre fractionation mix and post fractionation mix experiments. Most of the observed differences were replicated by mathematical modeling of the normalization effect, with the striking exception for mitochondrial DRM proteins. Interestingly, caveolin-1 affected DRM proteins in the post fractionation mix data showed a significant enrichment of the mitochondrial oxidative phosphorylation pathway, which was not observed in the pre fractionation mix experiment. The observed quantitative changes in mitochondrial DRM proteins using different analyses suggest a caveolin-1 induced change rather than simple contamination, and may support recent reports of caveolin-1-dependent mitochondrial cholesterol changes. Based on these results, we recommend a thorough understanding of how experimental conditions impact relative subcellular fraction in order to make an informed decision on the most appropriate point to combine SILAC samples for quantitative subcellular proteomic analysis.