Practical Proteomics

Lokesh Joshi, PhD and Colleen M. Brophy, MD

The mantra of molecular biology centers on the concept of "DNA to RNA to protein."  This central paradigm lead to the emergence of the research field of genomics. This technology makes extensive use of complementary DNA and oligonucleotide chips, to study the differential expression of genes. However, it is the protein that is the final functional unit of biological reactions and genomic chip array technology fails to give an understanding at this level.

A study which compared the levels of 19 gene products by DNA microarrays and proteomics revealed a correlation coefficient of 0.48 between mRNA and protein abundance (this value is half way between perfect and no correlation!) (1).  In the recent years there have been significant advances to extend the emphasis of genome based studies from nucleotide to proteins and functions, referred to as, "post-genomics" and/or "functional genomics;" of which proteomics is an integral part. The newly emerging field of proteomics comprises methodology to identify and characterize the function of specific proteins. Proteomics is the study of proteome which is defined as the PROTEins expressed by a genOME (Williams and Hochstrasser).

Typically, the first aspect of proteomics involves the identification of a set of candidate proteins.  This is classically performed by 2-dimensional gel electrophoresis.  The sample of interest is homogenized and the proteins are separated based on the isoelectric focusing (pI) of the protein in the presence of selective detergents and carrier molecules.  This was traditionally performed in "tube gels" containing small particles (ampholytes) of varying pI. 

There are currently a number of commercially available systems that have made this process simpler, including the Pharmacia MultiPhor II and the BioRad Protein IEF Cell.  With either of these systems, precast gels containing immobilized pH gradients provide more consistent focusing patterns.  The sample is equilibrated with the gel and subjected to a current.  The gel is then removed and placed on the top of a standard SDS page gel and the proteins are separated based on the relative mobility (Mr ) of the proteins through the acrylamide matrix.  The gels are then stained with either Coomassie Blue, silver nitrate, autoradiography (if radiolabelled proteins were used) and fluorography.  Using standards of known pI and relative mobility, the pI and Mr of the candidate protein can be determined.  The gels can be compared to data bases such as Swiss-Prot (http://expasy.hcuge.ch/sprot/sprot-top.html) and PDB (http://embl-heidelberg.de/pdb/).  When altered forms of a cellular or protein samples are analyzed, changes in the protein patterns on the 2-D gels can reveal protein modifications.

Once desired protein forms are localized on the gels, these samples are then processed for more specific identification of the protein.  This can be determined with mass spectrometry.  The main driving force for this development has been ionization techniques, which transfer the biomolecules from the liquid or solid phase to the gas phase. 

The two dominant methods to ionize the molecules are electrospray (ESI) and matrix assisted laser desorption ionization (MALDI).  In general, the protein of interest is first digested into peptide fragments using a protease (eg. trypsin).   The ionized peptides are subjected to a strong electric field in a vacuum and a measurement of the "time of flight" (TOF) of the peptides through the field gives a set of highly accurate peptide masses.  Data base searching is then used to correlate the mass spectra of the candidate molecule against the calculated mass spectra of all of the peptides in the data bases such as protein data banks available at http://www.rcsb.org/pdb/ and http://www.expasy.ch/.

Identification of proteins is just one small part of the field of proteomics.  The specific function of a protein is often altered by "post translational modifications."  The human genome project anticipated identifying over 100,000 genes from the human genome.  However only 30,000-40,000 genes were identified and the most logical explanation for this finding is the post-translational modifications that occur on over 80% of the proteins. 

This creates multiple forms of functional units from any one protein. Translation is a process whereby messenger RNA is converted to peptide chains. Enzymatic reactions occurring during and after the translation are called co- and post-translational modification. These modifications include phosphorylation, glycosylation, prenylation, demethylation, acetylation, myristoylation, palmitoylation, sulfation, etc.  Predicted sites of protein modification can be determined using data bases such ase EXPASY proteomic server (http://www.expasy.ch/www/tools).  In addition the subcellular localization and interaction with other proteins often correlates with functional aspects. Most mammalian proteins and lipids (specifically cell surface and secreted molecules) are post-translationally modified. These modifications serve multiple roles in conferring different structure and functions to the same proteins under different conditions.

Two recent discoveries have paved the way for the future of proteomics.  The first is that small peptide domains of proteins can mimic the biologic effect of the entire molecule.  For example, we have determined that the small heat shock related protein, HSP20 is an important effector protein for vasorelaxation (2).  A small (13 amino acid) domain of this protein, which contains the phosphorylation site, can effect relaxation in transiently permeabilized smooth muscle cells.  The major challenge to protein based therapeutics has been the introduction of the protein, or biologically active peptide fragment of the protein, into cells.  It has recently been determined that small, negatively charged peptides can "transduce" macromolecules into cells through an interaction with cell surface heparan glycoproteins (3).

The ultimate goal of gene therapy is to produce a biologically active protein.  However, limitations including targeting, immunogenicity of vectors, and regulation of expression, have prevented gene therapy from becoming an effective approach to human disease.  Proteomics, the identification of biologically active proteins, and protein transduction, the introduction of these proteins into cells, provides a new, innovative field for cardiovascular therapeutics.

1. Anderson L, Seilhamer J. Electrophoresis 18:533-537, 1997.
2. Beall A, Woodrum D, Stoming TA, Kato K, Suzuki A, Rasmussen H, Brophy CM. Heat shock-related protein, HSP20, is phosphorylated on serine 16 during cyclic nucleotide-dependent relaxation. J Biol Chem, 274:11344-11351, 1999.
3. Tyagi M, Rusnati M, Presta M, Giacca M. Internalization of HIV-1 Tat requires cell surface heparan sulfate proteoglycans. J Biol Chem 276:3254-3261, 2001.
   

FIGURE 1:

Silver stained two dimensional gel of proteins from intact rat aortic smooth muscle (RAT AORTA) and from cultured aortic smooth muscle cells (AORTIC SMC).  The proteins were separated by isoelectric focusing (IEF) followed by relative mobility (Mr).   Note the marked differences in proteins expressed in intact aortic smooth muscle cells compared to cultured cells.