Global Proteome Quantification
Rob Beynon, Ph.D., chair of proteomics at the University of Liverpool, put forth what he refers to as “a rather modest ‘grand challenge’. For a simple proteome, first quantify, in copies per cell, the abundances of all proteins. Secondly, determine the rate at which those proteins are turned over in the cell. Finally, determine both sets of parameters with high confidence and quantify the inherent biological variance in the data.”
Dr. Beynon described his experiences with the Waters ExpressionE system, which employs a high bandwidth UPLC/MSE data-acquisition strategy to consistently oversample complex protein digests, thereby delivering datasets containing evidence for all peptides above the limits of detection.
He discussed his research group's experience with label-free proteomics as a solution to quantification problems. This approach has some difficult issues; isoform resolution and quantification remain significant problems, as do post-translational variants. However, label-free methods, particularly those based on summed peptide intensities, are remarkably valuable for many proteomics studies.
“Identification is not the same as quantification,” Dr. Beynon cautioned. “Label-free approaches start to struggle for peptides or ions at about 0.1–1 fmol on column on the better instruments—at 1 fmol, this equates to about 3,000 copies per cell in yeast.” The situation is more pronounced in mammalian cells, as a typical load of a digest of HeLa cells, for example, equates to only 4,000 cells, or 150,000 copies per cell.
To get a handle on the problem, Dr. Beynon noted that the sensitivity required to meet the “grand challenge” would be a detection limit in yeast of around 10 copies of a given protein molecule per cell.
“On current instruments, we can routinely apply a digest derived from 200,000 cells, and in principle we can therefore reach between 30 and 300 copies per cell,” Dr. Beynon stated. “We anticipate instrument and informatics developments that should bring such methods to the required depth.”
He described his invention of the artificial QconCAT proteins, concatamers of tryptic peptides for several proteins, which when expressed heterologously in bacteria create stoichiometric equivalent sets of standard peptides. “The QconCAT approach is robust and the limiting factors are quantotypic peptide nomination and the development of appropriate selected reaction monitoring assays.”