The Escherichia coli translational initiation factor IF3 is encoded by an mRNA that has an unusual ribosome binding site. We have explored a mechanism that may account for the translation of IF3 and that provides regulation of the quantity of IF3 relative to ribosomes.Escherichia coli has many small regulatory loops that tightly control the levels of various important proteins. Such loops often involve regulation of gene expression at the post-transcriptional level (1, 2). In this paper, we propose a regulatory circuit that maintains appropriate levels of the E. coli translational initiation factor IF3 (3). The proposal simultaneously predicts a precise orientation of four different domains of 16S RNA within the 30S ribosomal particle.
RESULTS AND DISCUSSIONThe Problem. IF3 is responsible (at least) for dissociating 70S ribosomes that are not making protein, so as to allow the 30S subunit to reengage an mRNA and initiate translation (3). Purified IF3 is required for 70S ribosome dissociation in vitro and for high-level translation of natural mRNAs (3). When a 30S particle, bound to IF3, initiates translation, the normal mechanism is as shown in Fig. 1. Although the nucleotides surrounding initiating codons in E. coli are not random, we do not understand the contribution of that information to the 1000-fold differences in rates of translation between different mRNAs (1, 5, 6). Specifically, we do not know whether any 16S RNA domains (besides the 3' end) are used to achieve mRNA selection. Nevertheless, we have always believed that 16S RNA contains other domains that potentiate mRNA binding, at least for some mRNAs.The sequence of the mRNA encoding the initiation factor IF3 was recently published (7); the translational initiation domain breaks many "rules" followed by most other E. coli mRNAs (1,5,6). The sequence of the IF3 mRNA around the initiation codon is (7) (9). We have attempted to account for IF3 expression by seeking to understand its translation. IF3 expression is known to be regulated. IF3 levels increase, along with the other components of the translational machinery, when growth rates increase (9). That is, E. coli must respond to transiently decreased levels of IF3 by increasing the relative rate of IF3 expression, which is itself needed for the translation of all other mRNAs in the cell. Similarly, excess IF3 should selectively diminish the relative rate of IF3 production. However, IF3 is not feedback regulated by the same mechanism as are ribosomal proteins (2). Cells carrying plasmids encoding IF3 yield more IF3 protein in response to increased plasmid copy number (10).A constant IF3/ribosome ratio could be maintained if IF3 were translated in a mode independent of IF3 function; cells could adjust the relative amount of IF3 made as a result of competition between the IF3 mRNA and all others in the cell. High IF3 levels would stimulate all other mRNAs, leaving few ribosomes for IF3 translation; low IF3 levels would make available many ribosomes for IF3-independent translation of IF3 itsel...