Proteomics

views updated

Proteomics

Proteomics is the science of studying the multitude of proteomes found in living organisms. A proteome is the entire collection of proteins expressed by a genome or in a tissue. The contents of a proteome can differ in various tissue types, and it can change as a result of aging, disease, drug treatment, or environmental effects.

This is contrary to the concept of a genome, which is an organism's complete collection of DNA. A genome's composition remains more or less constant from tissue to tissue, except for mutations and polymorphisms that can occur.

The word "proteome" was first coined in late 1994. By 1997 there were a number of research conferences focusing on proteomics.

According to the first draft of the human genome, based on the work by the Human Genome Project and by Celera Inc., there are only between thirty thousand and seventy thousand genes in the human genome, many fewer than had been estimated previously. However, as of 2002 there were still groups that believed that there are at least 120,000 genes. Regardless of which of these estimates proves more accurate, the number of potential proteins in the human proteome is quite large. Although the first draft of the human genome reduced the estimates for the total number of human genes, it also predicted a greater amount of alternative splicing of genes, and therefore more distinct protein products per gene, than had been anticipated.

At its simplest level, proteomics is the study of protein expression in a proteome, or trying to understand the relative levels (amounts) of each protein within the mixture. Proteomics attempts to characterize proteins, compare variations in their expression levels in normal and disease states, study their interactions with other proteins, and identify their functional roles.

Unlike the traditional approach of studying individual proteins one at a time, proteomics uses an automated, high-throughput approach. High-throughput refers to the number of items (in this case, proteins) that can be analyzed or studied per unit of time. New technologies and substantial bioinformatics tools are required to compare entire proteomes. Expansion of the field of proteomics into the realm of "big science" (meaning many dollars invested by a large number of companies and universities) is several years behind the expansion of genomics. This is primarily because proteins are more difficult to work with in a laboratory setting than are nucleic acids such as DNA.

The development of protein analysis technologies is more difficult than the development of DNA analysis technologies for three reasons. First, the basic alphabet for encoding proteins consists of twenty amino acids, whereas there are only four different nucleotides, the alphabet of DNA. Second, the messenger RNA (mRNA) for some genes can be differentially spliced, meaning that multiple messages can be made from a single gene, resulting in multiple, distinct protein products. Finally, many proteins are modified once they have been synthesized. This is known as post-translational modification. There are a number of types of post-translational modifications, such as the addition of sugar, phosphate, sulfate, lipid, acetyl, or methyl groups. Each of these modifications has the ability to change the functional activity of a protein.

The above issues have made the elucidation of reliable, high-throughput techniques for characterizing proteins, including their expression levels, on a proteome-wide level a major challenge. Hence, techniques for doing, for example, high-throughput DNA sequencing and gene expression studies have been developed and commercialized on a large scale sooner than similar protein analysis techniques. This is not to imply that all of the techniques involved in proteomics are new. Some, such as two-dimensional gel electrophoresis , have been around since the 1970s. However, the need to adapt these techniques to a large "proteome" scale brings with it a unique set of challenges.

For researchers involved in areas such as drug discovery, proteomics approaches will need to be used to obtain a greater understanding of disease mechanisms and drugs' mechanisms of action. Large-scale studies looking at gene expression via quantification of mRNA abundance are already possible and well commercialized. These technologies are very powerful, and the highest throughput approaches are capable of analyzing tens of thousands of genes per experiment. Sophisticated bioinformatics systems have been, and continue to be, developed to analyze these vast amounts of data. However, studies have shown that mRNA levels do not necessarily correlate well with protein levels.

Researchers must understand proteins and their roles, since proteins are the functional units within cells. As of 2002, the vast majority of drug targets were proteins. There are a handful of drugs, including some chemotherapeutic agents, that bind to DNA, but most drugs bind to specific protein targets. In the cases where the target is a protein, the drugs themselves are primarily small inorganic molecules or, in some cases, small proteins, such as hormones , that bind to a larger protein target in the body. Some drugs are actually therapeutic proteins that are delivered to the site of the disease.

Laboratory Techniques

The primary attributes used to identify proteins include the protein's mass and apparent mass, its isoelectric point, and its N-and C-terminal sequence tags. A protein's mass and its apparent mass are probably the most common characteristics used. Protein mass is determined by adding the total mass of all the amino acids in the protein to the mass of any molecules added through post-translational modification. A protein's isoelectric point is the pH at which it is neutrally charged. A protein's N-and C-terminal sequence tags are short sequences of amino acids on either end of the protein. Since there are twenty different possible amino acids at each position in a protein, a peptide of only four or five amino acids in length is likely to be unique to a specific protein. There are 160,000 (20⁴) combinations of sequences that are four amino acids long.

The most commonly used laboratory techniques in proteomics are two-dimensional polyacrylamide gel electrophoresis (2-D PAGE) and mass spec-trometry. These techniques have been modified for use in proteomics. Both can be used in combination with more traditional protein separation techniques, including column chromatography.

Starting in the late 1990s, several companies also started developing "protein chips," another strategy for studying proteomes and other complex protein mixtures. These chips allow a researcher to collect minute quantities of proteins that bind to specific molecules on their surface. By 2001, some companies announced they were developing "antibody chips" onto which antibodies will be attached. The antibodies can then be used as probes to capture and quantify specific proteins found in complex mixtures.

The use of 2-D PAGE allows the simultaneous separation of thousands of proteins, and the technique is still a key tool in proteomics technologies. The first dimension of protein separation on the gel is by isoelectric focusing, in which proteins are separated along a pH gradient until they reach a stationary position, where their net charge is zero.

The second dimension of separation on the gel is by molecular mass. Sodium dodecyl sulphate (SDS) is applied, and it binds to all the proteins. This provides the proteins with a uniform charge along their length, so that they will migrate across the gel according to their molecular mass when a current is applied. After the 2-D PAGE is run, the gel is stained. The result is a two-dimensional map consisting of hundreds or thousands of protein spots.

Since the early use of 2-D PAGE in the early 1970s, a number of modifications have been made to make gels more reproducible and more amenable to the higher-throughput use necessary for proteomics applications. However, 2-D PAGE is still something of an art form, and high-quality, reproducible results are difficult to obtain except in the hands of very experienced users. The technology needs to be further simplified to allow casual and novice users to obtain reproducible, quality results.

Mass spectrometry is an analytical technique that very accurately measures the mass of proteins and peptides . There are two common types of mass spectrometry. The first type, matrix-assisted laser desorption/ionization time-of-flight mass spectrometry, can be used to analyze proteins that are embedded in solid samples and measures their mass in a flight tube. The second type, electrospray ionization mass spectrometry, can be used to analyze proteins that are in a liquid solution and measures their mass in either a flight tube or in a device known as a quadrupole. There are also other variations on these techniques.

Mass spectrometry is commonly used for peptide mass fingerprinting. In this process, a protein sample is isolated by 2-D PAGE and cut with an enzyme that specifically targets particular amino acids. Mass spectrometry is used to measure the masses of the resulting cut pieces, or peptides. These masses can be thought of as a fingerprint that can be compared to the fingerprints of proteins whose amino acid sequences have already been analyzed and stored in a database.

To determine the fingerprints of proteins that have already been sequenced, a computer program determines the amino acid composition, and thus the masses, of the pieces that would result if those proteins were also cut by the same enzyme. A list of proteins is generated from the database, sorted by how many peptides they share with the unknown experimental protein.

There are also technologies, including the yeast two-hybrid system, that can be used to study interactions between proteins. These approaches complement 2-D PAGE and mass spectrometry data by helping to elucidate functional cellular pathways.

Databases and Computational Approaches

There is an ever-increasing number of protein and proteome databases being developed. The most comprehensive information about specific proteins is found in databases that store protein sequences. One of the first and probably the best known such database is SWISS-PROT, which was created in 1986.

SWISS-PROT is a curated database that provides not only protein sequences but also such information as descriptions of a protein's function, its domain structure, and post-translational modifications, as well as links to other related databases. Other sequence-based protein databases include the Yeast Proteome Database and Human PSD.

There are also a number of widely used pattern and profile databases that are used to reveal relationships among proteins based on the presence of particular groups of amino acids in the proteins' sequences. Such groups, known as patterns, motifs, domains, signatures, or fingerprints, are found in specific regions of proteins that are important to some function of the protein. They could be in an area that performs some type of enzymatic activity or that is the site of a certain post-translational modification. Both their sequence and structure are typically well conserved. Some of the best known pattern and profile databases are: PROSITE, Pfam, PRINTS, and BLOCKS.

see also Alternative Splicing; Bioinformatics; Gel Electrophoresis; Genome; Human Genome Project; Mass Spectrometry; Post-translational Control; Proteins.

Anthony J. Recupero