Differential display, subtractive hybridization, amplification suppression and SAGE techniques for measuring gene expression
I have discussed at length methods for determining relative amounts of a specific and global RNA species in different preparations of total RNA from cells or tissues treated in different ways. However, all the techniques described so far require varying degrees of prior knowledge. In the case of northern hybridization or RT-PCR, where the abundance of an RNA with a specific sequence is being measured, you need to know at least part of the sequence of the gene encoding the RNA in the organism of study, or of a highly homologous gene in a closely related organism. This is because sequence specific probes require specific oligonucleotide primers in order to effect their production by DNA polymerase I in vitro, so at the very least, the sequence of the primer binding site (perhaps only within a homologous gene) must be known beforehand. In the case of RT-PCR, two, sequencespecific primers are needed, each to define one end of the sequence to be amplified and so quantified. In the other technique described, that of arraybased hybridization, the sequences of a number of genes, and in many cases, of the entire genome, must be known in order to synthesize DNA probes for immobilization onto the array. But what happens if you want to find out which genes are differentially expressed in cells growing in one comparator physiological state compared with those growing in another without knowing which genes to look for, and without having the entire genome sequence, or indeed, any sequence information? The techniques described in this chapter all have the common aim of finding RNA molecules that are present in one total RNA preparation, but absent in another. The total RNA preparations might each come from a different cell type, or from the same cell type, but growing in a different physiological state. These approaches
therefore do not really allow quantification of the amounts of an RNA species in the two comparator total RNA preparations (though they can help in this regard – e.g. SAGE, see Section 5.6) and are not really devised for measuring gene expression per se. Instead, they might be considered as methods which ask a binary question; is the RNA there, or is it not? Of course, a truly binary gene expression result is very hard to find because there is not really any such thing as a promoter that is absolutely turned off. However, in many cases, the difference in expression of a given gene between two comparator physiological growth conditions is large enough that any quantification of RNA produced from the gene might as well be binary. Hence these techniques are routinely and successfully used to characterize large gene expression differences between comparator physiological conditions. Absolute differences in RNA level can be determined using RT-PCR or northern hybridization in order to learn more about the actual expression differences once a significant difference has been highlighted.