Li et al.  analyzed the PPP Cl-serum specimen with five different proteomics technology combinations after immunoaffinity depletion of the top-6 proteins. In all, 560 unique proteins were identified, 165 with two or more peptides. Only 32 proteins were identified by all five approaches and 37 by 2-DE, 2-D HPLC, and shotgun approaches, primarily due to finding only 78 unique proteins among 1128 spots excised, digested, and analyzed with method 1, WAX-2-DE-MALDI-TOF-MS-MS. Protein 2-D-HPLC fractionation 1 RP-HPLC/microESI-MS-MS gave 179 proteins; an online SCX shotgun strategy ("bottom-up") gave 131, an offline SCX shotgun strategy gave 224, and an offline shotgun-nanospray strategy yielded 330 proteins. High and medium abundance proteins are found by all methods, while low abundance proteins are complementary, reflecting both different methods and inherent incompleteness of sampling and identifying peptide ions. Different technology combinations give different useful information; for example, the 2-DE method 1 provided more information about pi-altered isoforms and relative abundance of identified proteins. The offline strategies sharpen the peaks and improve separation of peptides, submit more fractions to the MS instrument, and allow the MS enough time to acquire the qualified spectra of more eluting peptides. Nanoflow accentuates the same advantages, permitting ultrahigh sensitivity. Overall, electrophoresis and chromatography, coupled respectively with MALDI-TOF/ TOF-MS and ESI-MS/MS, identified complementary sets of serum proteins. Like Aebersold and Mann , they conclude that no single analytical approach will identify all the major proteins in any proteome. Others have recently used similar 2-D separation of peptides offline, intact protein fractionation prior to MS, or sensitive ESI-MS/MS analysis of fractionated peptides [36-39]. As far as cost-effectiveness, the 2-D HPLC approach required much more time and labor and was much less suited to automation than the other strategies; it has the advantage of being able to process large volumes of sample, when that is available and desired. Handling fractions also introduces more evidence of contamination; epidermal keratins are seldom found with the shotgun methods. Low abundance proteins are not only masked by medium abundance proteins on gels, but inefficient extraction of pep-tides from gels is a limitation for low abundance proteins.
Barnea et al.  expanded on their original submission as Lab 1 (Tables 1 and 2) with an analysis of several protein fractionation and several MS/MS methods on PPP reference specimen B2-serum. Albumin and IgG were depleted with are BioRad mini-kit based on Affi-Gel Blue and Affi-Gel protein A, respectively. The aim was to increase the concentrations of individual proteins and then their tryptic peptides in each fraction submitted for MS/MS analysis, seeking to reach the threshold for detection. Combining pre-proteolysis fractionation with post-digestion fractionation was more effective than more extensive fractionation of the peptides. Each method has some advantages of avoiding loss of proteins with particular characteristics (pi, Mr other). The base case was MudPITanalysis of unfractio-nated, digested proteins; then SDS-PAGE, SCX, and Rotofor fractionations were coupled with LC-MS/MS or with MudPIT. In each pair, MudPIT gave more protein IDs than LC-MS/MS. SCX gave the most IDs among the fractionation methods.
He et al.  analyzed ten pooled male and ten pooled female C1-sera, using top-6 depletion, tryptic digestion, then RP-HPLC, ESI-MS/MS shotgun analysis. They reported 944 non-redundant proteins under stringent PPP criteria based on , combining separate analyses of male (594) and female (622) sera; there were 206 with two or more peptides. Some lower abundance proteins were detected, including complement C5 and CA125. Instead of one analysis of serum, here there are eight analyses: male and female, bound and unbound, and a duplicate of each. The reproducibility of the duplicates is 40-50%; the overlap of bound and unbound is 16-18%, and of male and female 40-50% (i.e., same as duplicates). They used four databases: IPI 2.20 (June 2003), IPI 2.32 (May 2004), Swiss-Prot 43 (March 2004), and NCBI (Dec 2003) and obtained quite similar protein groups for the first three and also for NCBI, though the pre-grouping numbers of proteins were 2.5 times larger for NCBI, demonstrating the known redundancy in the NCBI database.
Was this article helpful?