The importance of the 3'-untranslated region in the translational control of ferritin mRNA.

Ferritin synthesis provides a dramatic example of translational control; stored ferritin mRNA is translated at relatively low rates which can increase 40-50 times when cellular iron levels increase. Although it is not known if agents other than cellular iron levels can release the repression of ferritin mRNA in vivo, the repression appears to be eliminated during the isolation of poly(A+) RNA, judged by translation in wheat germ lysates (WG). Using the bullfrog tadpole as a model, because of the abundance of ferritin-rich embryonic red cells, we now show specific repression of ferritin mRNA in the isolated poly(A+) RNA translated in rabbit reticulocyte lysates (RR) (RR/WG = 25%). Repression of ferritin mRNA was associated with the inability to form polyribosomes in analogy to iron-poor cells in vivo. The addition of various complexes of iron did not relieve the repression, suggesting that in vivo at least part of the effect of iron may be indirect and mediated by factors absent in the cell-free system; all three ferritin subunit mRNAs (H, M, and L) appeared to be regulated coordinately in vitro and in vivo as well. Comparison of transcripts of DNA encoding the M subunit of ferritin, but containing deletions in the 3'-untranslated (UT) region, showed that a region 70 nucleotides long was important for repression. Comparison of secondary structures predicted for the eight known ferritin subunit mRNAs from humans, rats, chickens, and frogs indicates that a region involved in base pairing common to all the mRNAs is eliminated when the 3'-UT region is shortened to 24 nucleotides. Although regions in the 5'-UT of mRNAs, including ferritin, have been shown to be involved in translational regulation, it is clear that complete regulation can involve both the 3'-UT and the 5'-UT regions, mediated, presumably, by secondary and tertiary interactions along the mRNA molecule.