The use of DNA microarrays has revolutionized the manner in which mRNA populations are analyzed. One limitation of the current technology is that mRNAs are often purified on the basis of their 3 poly(A) ends, which can be extremely short or absent in some mRNAs. To circumvent this limitation, we have developed a procedure for the purification of eukaryotic mRNAs using a mutant version of the mRNA 5 cap-binding protein (eIF4E) with increased affinity for the m 7 GTP moiety of the cap. By using this procedure, we have compared the populations of mammalian mRNAs purified by oligo(dT) and 5 cap selection with oligonucleotide microarrays. This analysis has identified a subpopulation of mRNAs that are present with short 3 poly(A) ends at steady state and are missed or underrepresented after purification by oligo(dT). These mRNAs may respond to specific posttranscriptional control mechanisms such as cytoplasmic polyadenylation.
Initiation of eukaryotic mRNA translation requires the recruitment of many proteins, the initiator tRNA, and a 40S ribosomal subunit to the 5Ј cap of mRNA. Eukaryotic initiation factor 4E (eIF4E or 4E) specifically recognizes the m 7 GDP moiety of the m 7 G(5Ј)ppp(5Ј)N cap present at the 5Ј end of eukaryotic mRNAs (1-3). In addition to binding the 5Ј cap of eukaryotic mRNAs, 4E also binds the ribosome adaptor protein (4G) and translational repressor proteins (4EBPs or PHAS I/II) and associates indirectly with poly(A)-binding protein to stimulate translation (4-9). 4E is a common point of growth regulation in both untransformed and cancer cells (10-13). Both structural and functional studies provide evidence that 4E binds the m 7 G moiety of the 5Ј cap of mRNAs by a -stacking interaction between two tryptophan residues, as well as hydrogen bonds between m 7 G and acidic side chains of 4E (14-17). We have recently described mutants in the S4-H2 loop of 4E (N118A, K119A, and Q120A) that had an affinity for m 7 GTP several-fold increased relative to wild-type 4E (18).DNA microarray technology has enabled new questions to be identified or answered in a wide variety of biological systems (19). Although total RNA can be analyzed by DNA microarrays to quantitate gene expression, purification of mRNA with oligo(dT) can frequently decrease problems with signal-to-noise ratios, and the analysis of specific subsets of mRNAs can provide useful information (20). Wild-type 4E fused to protein A has been used previously to isolate eukaryotic mRNAs by binding their 5Ј caps (21). We tested the use of a high-affinity mutant of 4E to purify mRNA and determine how it compared with oligo(dT) and whether it might more effectively isolate a subset of eukaryotic mRNAs. One of these 4E mutants (4E K119A ) enabled the high-yield 5Ј cap-dependent purification of functional mRNAs from total RNA. Moreover, an unexpectedly large number of mRNAs were more efficiently isolated by using high-affinity 4E compared with oligo(dT), suggesting that they had short 3Ј poly(A) ends that diminished their binding to oligo(dT). This mRNA p...