Currently, microarrays reveal global transcription patterns; but to fully understand active biological pathways, global protein contents and their modification status in cells and tissues must be determined. To accomplish this on a genome-wide scale, we are evaluating adaptation of two-dimensional micro- high-pressure liquid chromatography (2D micro-HPLC) followed by tandem mass spectrometric (MS/MS) identification of peptides. As a pilot project peptides were derived from 20-30 micrograms of total proteins from mouse neonatal kidney. They were digested with trypsin and the resulting peptides were separated based on charge through a strong cation-exchange column; collected, concentrated and further fractionated based on hydrophobicity on a reverse phase column, and their sequences were determined by MS/MS. In these preliminary studies we assessed protein extraction procedures and optimization of peptide fractionation steps. We found that several fractionation approaches and multiple runs are necessary to identify the full complement of proteins in a complex tissue. Approximately one million spectra generated from the three runs were analyzed. For the analyses, publicly available software for analysis of the mass spectral data, DTAselect, was implemented in the laboratory and we generated scripts for database comparisons. Public database information was locally curated to merge multiple identitites into single representations and then searched. Identified proteins were subclassified according to information about their biological, molecular functions, or according to their cellular localization when known. With stringent analysis parameters we have thus far been able to identify 7,153 peptides and assign them to 2,749 proteins; and when analysis parameters were relaxed to standard accepted levels, 9,208 peptides were identified and assigned to 3,855 proteins. In the most stringent analysis we were able to identify peptides for 27 of the 28 small ribosomal subunit and 35 of the 38 large ribosomal subunit proteins, a coverage of 93% for these usually abundant cellular proteins. To test further the validity of the analyses, the assigned proteins were then compared to databases of RNA expression in the kidney. Based on this comparison we can identify 1951 (71%) of the proteins in the collection as having EST support for expression in kidney. All ranges of net charges and proteins from multiple organelles are represented, including Ornithine transcarbamylase and Carbamoyl-phosphate syntestase1, involved in the urea cycle, a critical function in kidney. ? Based on these encouraging results, we now plan to extend these studies to analyze proteins from mouse placental tissue and oocytes, in collaborative efforts with programs on Placental and Ovary-specific Genes and Mouse Pre-implantation Embryos.
