What is Proteomics?
The Center
Rationale of the Proteome Center
Technology and Application Development
Post Doctoral Fellow Program
Definitions
Technology and Application Development
The availability of complete genome sequences for organisms has generated opportunities for the development of new technologies
for the functional analysis of gene products. These opportunities extend beyond DNA and RNA methods to include proteins, the predominant
functional components of the cell. For the past decade new gene sequences have been generated at rates greatly outstripping the ability of
researchers to characterize gene products using classical methods of purification and analysis. An enormous gap now exists between gene
structure and our understanding of the functions of gene products. This gap must be closed if we are to fully realize the benefits of the large
amounts of structural data which have been acquired from the genome project and begin to make inroads into understanding complex biological systems. This is a massive
undertaking when the system is as complex as a living cell, tissue, or even whole organism. New instrumental and computational tools are
needed to determine the amount, identity, and form of each protein in the cell, their direct and indirect interactions, and to determine which
gene encodes each of these proteins.
Classical approaches to analysis of proteins from 2-D gels
Classically, alterations in patterns of cell proteins have been followed using two-dimensional (2-D) gel electrophoresis. The unparalleled
resolution of this method readily distinguishes small changes in protein patterns and relative quantification via radiolabeling is relatively
straightforward. One of the problems encountered with 2-D gels is that identification of a single protein spot by Edman sequence analysis
can involve a lot of time, effort, and expense. In addition, landmark mapping is frequently ambiguous because gel reproducibility is quite
variable from lab to lab or even with different lots of ampholines. Another problem has been that the protein loadings on 2-D gels are
relatively low so that only high abundance proteins can be analyzed from a single gel. Moderate abundance proteins require that several
2-D gels be run to collect enough material for sequence analysis and low abundance proteins are inaccessible without an enrichment process.
This same problem applies to existing nano-electrospray-based approaches, leading some investigators to begin evaluating two-dimensional
liquid chromatography mass spectrometry (LC/LC/MS). However, electrophoretic methods generally provide better resolution than
chromatography methods.
Although methodologies for constructing protein maps continue to improve, allowing more proteins to be linked to their cognate genes,
protein characterization approaches must evolve to higher throughputs than can be provided by classical methods if we are to keep pace
with the avalanche of information produced by DNA sequencing efforts. Previous methods available for protein analysis (Edman degradation,
amino acid analysis, etc.) have undergone relatively small incremental improvements. Rate limiting steps in these processes include not just instrumental
analysis, but more significantly, sample purification and preparation, as well as data processing. These are issues that affect existing mass
spectrometric methods as well.
More recently developed approaches to protein analysis
Recently developed MALDI and electrospray mass spectrometric methods for mapping proteins by analysis of proteolytically generated
fragments eluted from 2-D gels have somewhat accelerated the process of protein identification and characterization. Technological
advances in mass spectrometry have lowered detection limits for proteolytic peptides dramatically and have raised considerable interest
for computer database searching based on peptide mass and sequence information [1-10].
Proteins are identified from genome databases
by matching experimental mass maps of protein spots to the theoretical masses calculated from the known gene sequences. Short
stretches of sequence information can confirm assignments or identify proteins from as little as one or two tryptic peptides or pull out an
identification from an expressed sequence tag database. Popular approaches for protein identification employ on-membrane or in-gel
proteolysis of excised bands followed by elution of the product peptides and optional on- or off-line cleanup prior to mass analysis (4, 6-8,
11-20).
A recent application of this methodology sequenced 4 tryptic peptides in identifying bovine serum albumin from an initial gel loading of 80
fmol (8). This application also highlights how far nanospray technology has been
pushed in identifying proteins from 2-D gels--none of the four tryptic peptides that were sequenced by tandem mass spectrometry were
observed in scanning Q1 of the triple quadrupole. Digest-related ions were extracted from the instrument noise with parent
ion scanning (21) by monitoring the immonium ions of Ile/Leu.
Several other mapping-based methods explored in the past few years rely on post-electrophoretic elution of intact proteins followed by
enzymatic cleavage and optional clean-up steps prior to mass analysis (11,
12, 19,
22). Other variations involve blotting proteins from
gels to a polymeric membrane support for matrix-assisted laser desorption/ionization (MALDI)-MS; they are noteworthy in their ability
to provide masses of intact proteins and may be more amenable to automation (17,
23-28). MALDI approaches desorbing intact proteins from membranes with infrared
lasers have obtained good spectra by loading less than 1 µg of standard proteins onto 1-D gels (29,
30). Desorption with 337 nm, the wavelength of the inexpensive nitrogen lasers
almost universally supplied with commercial MALDI instruments, was reported to be much less successful than with infrared lasers
(29), however, recent successes have been reported at gel loadings down to 10 pmol/mm2
(25) and 50 pmol (17). Masses of intact
proteins have also been obtained by electroelution or extraction of proteins from gels, although usually at loadings on the order of a microgram.
A need for automation
These methods are elegant but automated throughputs of even 4 proteins identified per hour are still insufficient to have a major impact on
research. We need a revolutionary change. While some labs are automating the mass spectrometry and database searching portions of
proteome analysis, the bottleneck continues to lie with the steps upstream: excising gel spots one by one, performing digests, and sample
clean-up. There is a need to develop an approach that is easily automated and that will serve for the long term--an approach that will go
beyond the most intensely stained spots on a 2-D gel to the spots that cannot be visualized using protein staining techniques
and which will not only identify the proteins but characterize the structures by identifying changes in post-translational modifications.
