A robust method for proteomic characterization of mouse cartilage using solubility-based sequential fractionation and 2-D electrophoresis

Purpose: There is increasing interest in proteomic techniques for global profiling of normal and pathological cartilage samples, to elucidate underlying disease mechanisms identify novel biomarkers. Identification of protein expression differences using two-dimensional electrophoresis (2-DE) and liquid chromatographic (LC)-based proteomics depends critically on reproducibility throughout sample preparation and analysis. This applies particularly where sample fractionation is used to remove high abundance or interfering components to facilitate deeper mining of the proteome. Here we have developed and validated a procedure for solubility-based cartilage sample fractionation and reproducible resolution by 2-DE. Methods: Triplicate independent sequential extractions were performed on mouse femoral head cartilage. Pulverized explants were digested with chondroitinase ABC then extracted in Tris acetate containing 1M NaCl for 24 hrs (E0 fraction). The NaCl-insoluble fraction was then further extracted for 18 hrs in sodium acetate containing 4M GuHCl (E1 fraction). Finally, to prepare samples for 2-DE, high molecular weight components that prohibit isoelectric focusing (hyaluronan and aggrecan) were removed by 100 kDa cut-off centrifugal filtration (E2 fraction). Triplicate E0, E1 and E2 fractions were first profiled by SDS-PAGE. E0 and E1 were then resolved by 2-DE in triplicate to compensate for variation in 2-DE and silver staining and the 18 gel images were analyzed in ImageMaster. The E0 and E1 extracts were characterized by identification of protein spots by tandem mass spectrometry (MS). Results: The 1-D profiles of E0, E1 and E2 fractions were highly consistent between extractions (see figure below). Centrifugal filtration caused partitioning of some proteins <100 kDa into the E2 fraction but this was also consistent between samples. E0 and E1 fractions produced distinct protein 2-DE spot patterns, with greater complexity in E0. Automated spot analysis reported 70% spot matching in E0 gel triplicates and 75% matching in E1 gel triplicates, representing approximately 600 and 500 matched spots, respectively. E0-specific spots were mostly cellular proteins, e.g. BiP, triosephosphate isomerase and gelsolin, whereas E1-specific spots were abundant matrix proteins, e.g. collagen VI, matrilins 1 and 3, and lactadherin. Interestingly, some specific proteins such as link protein and beta-actin partitioned almost equally between E0 and E1 extracts. MS results were validated by immunoblotting. Conclusions: This study has, using the minimal amounts of tissue available from mouse tissue, established a new approach to 2-DE based analysis of cartilage extracts. This method can be used to enrich one or both protein fractions for deeper mining of the cartilage proteome and to investigate cellular mechanisms and matrix components involved in developmental and degenerative cartilage disease.