Biology Reference
In-Depth Information
now used to determine the expression of all genes at the level of mRNA, at
accuracies beyond 30%.
No similar hybridization chemistry exists at the level of a chain of amino
acids (yet). Using immunological techniques however, antibody-like molecules
are now spotted onto arrays, and the abundance of proteins in extracts from cells
is determined (Walter et al., 2000). Alternative modes of genome-wide detection
of protein abundances include a methodology in which all proteins are separated
in a highly reproducible way through two-dimensional (2D) gel electrophoresis,
such that each location in 2D corresponds to a specific protein. The mapping of
spot location to the identity of the gene is a slow process, but for smaller genomes
this methodology is getting close to the possibility of genome wide detection of
gene expression at the level of protein. This methodology is inherently limited in
three important ways. First, the resolution of 2D gel electrophoresis is insufficient
to separate all proteins of genomes larger than a few thousand genes; though
useful for bacteria, the methodology is still of more limited value for human
biology. Second, the method is not quantitative yet, and indeed many proteins,
especially membrane proteins, are missed entirely. And third, it is difficult
to identify the individual proteins. The latter problem is now being alleviated
by the implementation of mass spectrometry. By extracting protein from a
specific location on the 2D gel, subjecting that to limited proteolytic digestion,
determining the precise mass and/or sequence of the resulting peptides and
combining the resulting information with the known sequence of the genome,
the protein spots can now often be attributed to specific proteins.
Mass spectrometry also offers methods that may analyse genome-wide expres-
sion at the protein level. The gel-electrophoresis step can be replaced by capillary
chromatography, a separation by mass spectrometry on the basis of the total
mass of the protein (or fragments thereof), fission of the protein/peptide in the
gas phase and then a second mass spectrometry step to determine what the
resulting fractions are. Again the availability of the genome sequence enables
one to identify the protein. For mass spectrometry, molecules have to be brought
into the gas phase as electrically charged molecules. However, existence in the
gas phase is far from the thermodynamically most favourable mode of existence
for most of the molecules that constitute the living cell. The effectiveness at
which the entry into the gas phase is achieved is low therefore more importantly,
it depends much on the presence and properties of the other molecules in the
mixture. Other molecules with electric charge can affect the tendency of a given
molecule to enter the gas phase. Consequently, the mass spectrometry method
is inherently irreproducible in the quantitative sense; it is hard to determine
expression levels accurately with this method (although this is improving both
by changing conditions in the mass spectrometer (Vaidyanathan et al., 2003)
and by isotope-based quantification. This is because isotopes behave essentially
identically with respect to the above problems, yet can be discriminated readily
