Serial Analysis of Gene Expression SAGE

SAGE is a technique designed to take advantage of high-throughput sequencing technology to obtain a quantitative profile of cellular gene expression [183]. Rather than directly measuring the expression level, the technique "counts" the number of transcripts or "tags" for each gene. A tag is a specific oligonucleotide of length 9 to 11 nucleotides corresponding to the transcript of interest which is selected on the 32 end of a transcript, relative to a site recognized by a restriction enzyme (usually NlaIII). First, double stranded cDNA is made from mRNA, using a biotinylated oligo(dT) primer. The cDNA is then cut with the anchoring enzyme (NlaIII), expected to cut every at every 256 base pairs. The resultant double-stranded cDNA is bound to streptavidin beads, divided into two pools, and ligated to one of two linkers. The tags are then spliced out using a tagging enzyme that cuts at a location away from its recognition site (such as FokI). The tags are joined together and the resultant string of tags is sequenced. This sequencing allows each SAGE tag to be counted. Therefore, the number of tags for each gene that the tag represents should be proportional to the expression level of that gene. Figure 7.5 illustrates this process.

Figure 7.5: The Serial Analysis of Gene Expression. (Derived from Madden et al. [124]. SAGE suffers from the potential lack of specificity of these oligonucleotide tags, much in the same way as do the oligonucleotide and cDNA microarrays. It can be argued that the small length of the tags makes them even more susceptible to nonspecific interactions (i.e., multiple genes or gene products with the same sequence as the oligonucleotide). In fairness, the most recent gene profiling investigations use longer nucleotide sequences. Furthermore, and more problematically, sequencing technology has an error rate that might substantially degrade the correct assignment of a tag count to a gene. Although there have been some early, optimistic results for small numbers of genes, showing the correlation of sage results with microarray expression values [100], more comprehensive testing is necessary. Of course, as with all these technologies, the choice of a "gold standard" to rate such comparisons is hard to come by.

Figure 7.5: The Serial Analysis of Gene Expression. (Derived from Madden et al. [124]. SAGE suffers from the potential lack of specificity of these oligonucleotide tags, much in the same way as do the oligonucleotide and cDNA microarrays. It can be argued that the small length of the tags makes them even more susceptible to nonspecific interactions (i.e., multiple genes or gene products with the same sequence as the oligonucleotide). In fairness, the most recent gene profiling investigations use longer nucleotide sequences. Furthermore, and more problematically, sequencing technology has an error rate that might substantially degrade the correct assignment of a tag count to a gene. Although there have been some early, optimistic results for small numbers of genes, showing the correlation of sage results with microarray expression values [100], more comprehensive testing is necessary. Of course, as with all these technologies, the choice of a "gold standard" to rate such comparisons is hard to come by.

Was this article helpful?

0 0

Post a comment