There are several desiderata that should be understood about transcription in order to better understand the limitations of the gene profiling techniques.
• The set of protein-coding RNA, otherwise known as the transcriptome, should be viewed as a pool of mRNA. Even at equilibrium, each type of mRNA is degraded at its specific rate. New transcription provides additional mRNA to replace those being degraded, maintaining the levels. Thus, a transcriptional program should more accurately be viewed as a change in the transcriptome. In addition, just because a particular mRNA is seen in the transcriptome, it does not necessarily imply that the mRNA is about to be translated. Similarly, presence of a particular mRNA in the transcriptome does not mean that mRNA was recently transcribed. To determine what genes may be in the process of being translated, one can specifically look at polyribosomal mRNA, or that mRNA found in the presence of ribosomes. To determine which genes are currently being transcribed and processed, one can construct studies looking for pre-mRNA levels instead of mRNA levels (i.e., by finding intronic splice sites, etc.).
• mRNA can be transcribed at up to several hundred nucleotides per minute but may be transcribed at much slower rates. In eukaryotic organisms, genes can take many hours to transcribe. For instance, a gene found in muscle, dystrophin, can take up to 20 hours to be transcribed. The lifespan of an mRNA molecule from initiation of transcription to ultimate degradation is a complex function of the rate of transcription, the stability of the particular mRNA molecule, and changes in its processing due to other cellular events, whether internally or externally initiated. Gene expression measurements, as performed by microarrays, measure the concentration of each mRNA as a snapshot at a single point of time or relative to another sample of mRNA. If the transcription of two genes is simultaneously stimulated and proceeds at the same rate, but the mRNA samples have different lifespans, then the expression measurements (and downstream protein production) from these two genes can be quite divergent.
• Proteins perform most cellular functions. The lifespan of proteins is at least as variable as that of mRNA. Consequently, measurements of gene expression (i.e., mRNA measurement) may not accurately correspond to the concentration or activity of the protein for which it codes. This should be a caution to any investigator imputing function to a gene based solely on gene expression patterns.
• The mRNA generated from the transcription of a gene may differ depending on which exons and introns are or are not spliced into the mRNA molecule. These alternate splicing products are mRNA molecules with divergent sequences. Therefore, measurements of mRNA that are designed to measure only one of these alternate splicing products will provide incomplete if not misleading information.
• The genetic basis for organismal diversity is due in large part to differences in sequences, also known as polymorphisms, of each gene. Most of these polymorphisms differ from one another by one nucleotide and are known as single nucleotide polymorphisms (SNPs). Due to the small portion of the genome coding for proteins and to redundancy in the mRNA code, described on page 27, only some SNPs will result in differently constructed proteins. If a gene expression measurement technology is highly specific for a particular SNP, then other variants will not be measured. When we consider that the Human Genome Project to date only includes the sequence of handfuls of individuals, the implications of such measurement specificity become apparent. Nonetheless, as common SNPs become documented for each gene, it is likely that successful expression measurement techniques will measure each of them.
Despite these caveats, it remains that gene profiling studies have proven to be remarkably robust in describing the functional grouping and coordinated behavior of genes. The reasons for this are discussed in section 2.2, page 60.
Was this article helpful?