Identification of an intragenic ribosome binding site that affects expression of the uncB gene of the Escherichia coli proton-translocating ATPase (unc) operon

Abstract

The uncB gene codes for the a subunit of the Fo proton channel sector of the Escherichia coli F1 Fo ATPase. Control of expression of uncB appears to be exerted at some step after translational initiation. Sequence analysis by the perceptron matrices (G. D. Stormo, T. D. Schneider, L. Gold, and A. Ehrenfeucht, Nucleic Acids Res. 10:2997-3011, 1982) identified a potential ribosome binding site within the uncB reading frame preceding a five-codon reading frame which is shifted one base relative to the uncB reading frame. Elimination of this binding site by mutagenesis resulted in a four- to fivefold increase in expression of an uncB'-'lacZ fusion gene containing most of uncB. Primer extension inhibition (toeprint) analysis to measure ribosome binding demonstrated that ribosomes could form an initiation complex at this alternative start site. Two fusions of lacZ to the alternative reading frame demonstrated that this site is recognized by ribosomes in vivo. The results suggest that expression of uncB is reduced by translational frameshifting and/or a translational false start at this site within the uncB reading frame.

FIG. 1

Location of translation false start within uncB. All three perceptron weight matrices identified this region as a translation initiation region. The bases between positions 271 and 303 of the uncB reading frame are shown. The amino acids of the uncB product coded for by those bases are shown above them. The false start translation initiation site identified by the perceptron matrices, followed by a short reading frame, is indicated below the bases.

FIG. 2

β-Galactosidase activities produced by in-frame uncB′-′lacZ fusions. Locations of the unc promoter (P_unc), uncI, uncB, and uncE are shown at the top of the figure. The amounts of unc DNA fused in frame to lacZ in plasmids pDKWH103, pKS104, and pKS105 are indicated by the horizontal lines. Plasmid pSRM106 is identical to pKS104 except for the T→C mutation at position 285 of uncB, which is indicated by an asterisk. β-Galactosidase activities produced by each fusion gene in single-copy lysogens were assayed as described by Miller (17). The values for the β-galactosidase activities produced by single copies of the fusions in pDKWH103 and pKS105 are from the work of Solomon et al. (25).

FIG. 3

Putative mRNA secondary structure around translational false start. The sequence between bases 247 and 334 of uncB are shown folded into a secondary structure predicted by the MFOLD program, version 2.0 (10, 11, 30). The false-start AUG and the primer used for sequencing and for primer extension in the toeprint experiment are indicated. To obtain a toeprint of this site, the region indicated by the arrows was replaced by 11 adenosine residues, and a single-base deletion and two single-base mutations were constructed in the small loop, as indicated.

FIG. 4

Primer extension inhibition (toeprint) analysis of the intragenic false start site within uncB. The extent of ribosome binding in the initiation region of protein synthesis was determined by primer extension inhibition as described in Materials and Methods. Sequencing lanes (G, A, T, and C) are indicated. Lanes labeled (+) and (−) represent reactions run in the presence or absence of ribosomes, respectively. Next to lane (+), an arrow indicates the toeprint band which forms at +15 from the false start initiation codon. Below the autoradiogram is shown the location of the toeprint within the initiation domain of the false start, 15 nucleotides 3′ of the first base of the initiation codon AUG (boxed). The Shine-Dalgarno sequence (SD) is underlined. The band seven bases above the toeprint could be a minor ribosome binding region predicted by the w51 matrix (see Results).

FIG. 5

Plasmid constructions which fuse lacZ to the false start reading frame. The top line shows the unmutagenized region of the false start in the uncB′-′lacZ fusion gene found in pKS104 (see Fig. 1 and 2 for more detail). In plasmid pSRM114, the termination codon of the false start reading frame is eliminated, and the false start reading frame is fused in frame to the uncB reading frame. Plasmid pWSB52 consists of lacZ fused directly to the false start initiation codon. The details of these constructions are given in Materials and Methods. Ribosomes initiating at the true uncB initiation codon terminate 31 codons downstream of the deletion in pSRM114 and 6 codons after the indicated sequence in pWSB52.

FIG. 6

Immunoblots of fusion proteins. Lysates were prepared from E. coli MC1000 Δ (uncI-uncC) (1) carrying either pKS104, pSRM114, or pWSB52 as described in Materials and Methods. The pKS104 culture was grown in a 25-ml volume; the other two were grown in 250 ml. Unlysed cells were removed by centrifugation, and the supernatant fractions were loaded onto an SDS–7% polyacrylamide gel. Electrophoresis and immunoblotting, with anti-β-galactosidase (Promega), were carried out as described previously (3). The samples and amounts loaded were as follows: pKS104, 12 μg (lane 1); pSRM114, 250 μg (lane 2); pWSB52, 250 μg (lane 3). Lanes 2 and 3 therefore contained 20 times as much total protein as lane 1. The top band in lane 1 is the full-length fusion protein coded for by pKS104; it consists of most of the a subunit fused in frame to β-galactosidase. The top band in lane 2 is the same-size protein, which could result only from a frameshift at the false start of ribosomes which had initiated at the true uncB translation start site. The second band (arrow) in lane 2 represents a lac fusion protein initiated at the false start site. The top band (arrow) in lane 3 is the proper size for a protein initiated at the true uncB translational start site and fused to lacZ by frameshifting at the false start. The second bands in lanes 1 and 2 represent the β-galactosidase moiety (BG) derived from proteolysis of each fusion protein at the fusion joint. The equivalent band in lane 3 probably represents the fusion protein derived from translational initiation at the false start together with the β-galactosidase moiety derived from proteolysis of the higher-molecular-weight fusion protein at the fusion joint.