Autoregulation of gene expression: Quantitative evaluation of the expression and function of the bacteriophage T4 gene 32 (single-stranded DNA binding) protein system

The free concentration of bacteriophage T4-coded gene 32 (single-stranded DNA binding) protein in the cell is autoregulated at the translational level during T4 infection of Escherichia coli. The control of the synthesis of this protein reflects the following progression of net (co-operative) binding affinities for the various potential nucleic acid binding targets present: single-stranded DNA > gene 32 mRNA > other T4 mRNAs ? double-stranded DNA. In this paper we show that the free concentration of gene 32 protein is maintained at 2 to 3 μm, and use the measured binding parameters for gene 32 protein, extrapolated to intracellular conditions, to provide a quantitative molecular interpretation of this system of control of gene expression. These results are then further utilized to define the specific autoregulatory binding sequence (translational operator site) on the gene 32 mRNA as a uniquely unstructured finite binding lattice terminated by elements of secondary structure not subject to melting by gene 32 protein at the autoregulated concentration, and to predict how this site must differ from those found on other T4 messenger RNAs. It is shown that these predictions are fully consistent with available T4 DNA sequence data. The control of free protein concentration as a method of genome regulation is discussed in terms of other systems to which these approaches may apply.     

Non-canonical RNA arrangement in T4-even phages: accommodated ribosome binding site at the gene 26-25 intercistronic junction

Translational initiation region of bacteriophage T4 gene 25 contains three potential Shine and Dalgarno sequences: SD1, SD2 and SD3. Mutational analysis has predicted that an mRNA stem-loop structure may include SD1 and SD2, bringing the most typical sequence SD3, GAGG, to the initiation codon. Here, we report physical evidence demonstrating that previously predicted mRNA stem-loop structure indeed exists in vivo during gene 25 expression in T4-infected Escherichia coli cells. The second mRNA stem-loop structure is identified 14 nucleotides upstream of the stem-loop I, while the SD3 sequence, as well as the start codon of the gene, are proved to be within an unfolded stretch of mRNA. Phylogenetic comparison of 38 T4-like phages reveals that the T-even and some pseudoT-even phages evolve a similar structural strategy for the translation initiation of 25 , while pseudoT-even, schizoT-even and exoT-even phages use an alternative mRNA arrangement. Taken together, the results indicate that a specific mRNA fold forms the split ribosome binding site at the gene 26-25 intercistronic junction, which is highly competent in the translational initiation. We conclude that this ribosome binding site has evolved after T-even diverged from other T4-like phages. Additionally, we determine that the SD sequence GAGG is most widespread in T4.     

Kinetics and mechanism of the association of the bacteriophage T4 gene 32 (helix destabilizing) protein with single-stranded nucleic acids. Evidence for protein translocation

We have investigated the association kinetics of the co-operatively binding T4-coded gene 32 (helix destabilizing) protein with a variety of single-stranded homopolynucleotides (both RNA and DNA). Stopped-flow mixing experiments were performed by monitoring the partial quenching of the intrinsic tryptophan fluorescence of the protein upon binding to the nucleic acid under conditions where the nucleic acid concentration is in great excess over the protein concentration. Investigations of the association rate (and rate constants) as a function of solution variables has demonstrated quite different behavior at the extremes of “low” and “high” salt concentration. Under low salt (high binding constant) conditions the non-co-operative association is rate-limiting and we measure a bimolecular rate constant of 3 × 10⁶ to 4 × 10⁶ m^?1 (nucleotide)s^?1 (0·1 m-NaCl, 25·0 °C). However, at higher salt concentrations (lower binding constant) a pre-equilibrium involving non-co-operatively bound protein is established, followed by the rate-limiting formation of co-operatively bound protein clusters.Based on these observations we have proposed a mechanism for the formation of co-operatively bound T4 gene 32 protein clusters, under conditions of low binding density, which consists of three steps: (1) pre-equilibrium formation of non-co-operatively bound protein (nucleation); followed by (2) association of free protein to the singly contiguous sites established in the nucleation step, hence forming the first co-operative interactions (growth step); and (3) a redistribution of the growing protein clusters to form the final equilibrium distribution. From comparisons of our experimental values of the forward rate constant for the second step (growth of clusters) with theoretical estimates based on the work of Berg &; Blomberg (1976,1978) we infer that the T4 gene 32 protein is able to translocate along singlestranded polynucleotides. The implications of these results for the in vivo action of the T4 gene 32 protein are discussed.     

Post-transcriptional control of bacteriophage T4 gene 25 expression: mRNA secondary structure that enhances translational initiation.

Secondary structure of the mRNA in the translational initiation region is an important determinant of translation efficiency. However, the secondary structures that enhance or facilitate translation initiation are rare. We have previously proposed that such structure may exist in the case of bacteriophage T4 gene 25 translational initiation region, which contains three potential Shine-Dalgarno sequences (SD1, SD2, and SD3) with a spacing of 8, 17, and 27 nucleotides from the initiation codon of this gene, respectively. We now present results that clearly demonstrate the existence of a hairpin structure that includes SD1 and SD2 sequences and brings the SD3, the most typical of these Shine-Dalgarno sequences, to a favourable spacing with the initiation codon of gene 25.Using a phage T7 expression system, we show that mutations that prevent the formation of hairpin structure or eliminate the SD3 sequence result in a decreased level of gp25 synthesis. Double mutation in base-pair V restores the level of gene 25 expression that was decreased by either of the two mutations (C-to-G and G-to-C) alone, as predicted by an effect attributable to mRNA secondary structure. We introduced the mutations into the bacteriophage T4 by plasmid-phage recombination. Changes in the plaque and burst sizes of T4 mutants, carrying single and double mutations in the translational initiation region of gene 25, strongly suggest that the predicted mRNA secondary structure controls (enhances) the level of gene 25 expression in vivo. Hybridization of total cellular RNA with a gene 25 specific probe indicated that secondary structure or mutations in the translational initiation region do not notably affect the 25 mRNA stability. Immunoblot analysis of gp25 in Escherichia coli cells infected by T4 mutants showed that mRNA secondary structure increases the level of gp25 synthesis by three- to fourfold. Since the secondary structure increases the level of gp25 synthesis and does not affect mRNA stability, we conclude that this structure enhances translation initiation. We discuss some features of two secondary structures in the translational initiation regions of T4 genes 25 and 38.     

The downstream box: an efficient and independent translation initiation signal in Escherichia coli.

The downstream box (DB) was originally described as a translational enhancer of several Escherichia coli and bacteriophage mRNAs located just downstream of the initiation codon. Here, we introduced nucleotide substitutions into the DB and Shine-Dalgarno (SD) region of the highly active bacteriophage T7 gene 10 ribosome binding site (RBS) to examine the possibility that the DB has an independent and functionally important role. Eradication of the SD sequence in the absence of a DB abolished the translational activity of RBS fragments that were fused to a dihydrofolate reductase reporter gene. In contrast, an optimized DB at various positions downstream of the initiation codon promoted highly efficient protein synthesis despite the lack of a SD region. The DB was not functional when shifted upstream of the initiation codon to the position of the SD sequence. Nucleotides 1469-1483 of 16S rRNA ('anti-downstream box') are complementary to the DB, and optimizing this complementarity strongly enhanced translation in the absence and presence of a SD region. We propose that the stimulatory interaction between the DB and the anti-DB places the start codon in close contact with the decoding region of 16S rRNA, thereby mediating independent and efficient initiation of translation.     

An unstructured mRNA region and a 5' hairpin represent important elements of the E. coli translation initiation signal determined by using the bacteriophage T7 gene 1 translation start site.

Gene 1 of bacteriophage T7 early region--the RNA polymerase gene--is very actively translated during the infectious cycle of this phage. A 29 base pair fragment of its ribosome binding site containing the initiation triplet, the Shine-Dalgarno sequence (S-D), 10 nucleotides (nt) upstream and 6 nt downstream of these central elements was cloned into a vector to control the expression of the mouse dihydrofolate reductase gene (dhfr). Although all essential parts of this translation initiation region (TIR) should be present, this fragment showed only very low activity. Computer analysis revealed a potentially inhibitory hairpin binding the S-D sequence into its stem base paired to vector-derived upstream sequences. Mutational alterations demonstrated that this hairpin was not responsible for the low activity. However, addition of 21 nt of the T7 gene 1 upstream sequence to the 29 base pair fragment were capable of increasing the translational efficiency by one order of magnitude. Computer analysis of this sequence, including nucleotide shuffling, revealed that it contains a highly unstructured region lacking mRNA secondary structures but with a hairpin at its 5' end, here formed solely by T7 sequences. There was not much difference in activity whether the mRNA included or lacked vector-derived sequences upstream of the hairpin. Such highly unstructured mRNA regions were found in all very efficiently expressed T7 genes without any obvious sequence homologies. The delta G values of these regions were higher, i.e. potential secondary structural elements were fewer, than in TIR of genes from E. coli. This is likely due to the fact that T7 as a lytic phage is relying for successful infection on much stronger signals which a cell cannot afford because of the indispensable balanced equilibria of its interdependent biochemical processes. When the 5' ends of efficient T7 gene mRNA are formed by the action of RNase III they generally start with an unstructured region. Efficiently expressed T7 genes within a polycistronic mRNA, however, always contain a hairpin preceding the structure free sequence. We suggest that the formation of this 5' hairpin is releasing enough energy to keep the unstructured regions free of secondary RNA structures for sufficient time to give ribosomes and factors a good chance for binding to the TIR. In addition, sequences further downstream of the start codon give rise to an additional increase in efficiency of the TIR by almost two orders of magnitude.     

A retrovirus-like zinc domain is essential for translational repression of bacteriophage T4 gene 32

Gene 32 protein (gp32), a single-stranded DNA-binding protein from bacteriophage T4, contains a zinc-binding subdomain with sequence homologies to the 3-cysteine/1-histidine zinc-binding motif found in a variety of retroviruses and plant viruses. In vitro studies suggest that autoregulation of gp32 occurs at the level of translation by gp32 specifically binding gene 32 mRNA at an unusual stem-loop structure that can be modeled as an RNA pseudoknot. Nucleation of gp32 binding via this pseudoknot is thought to be needed to facilitate cooperative binding of gp32 through a largely unstructured region that overlaps the ribosome binding site (McPheeters, D. S., Stormo, G. D., and Gold, L. (1988) J. Mol. Biol. 201, 517-535). Removal of Zn(II) from gp32 results in a protein that retains the ability to bind single-stranded RNA with high affinity but is unable to specifically autoregulate itself at the level of translation. Deletion of the pseudoknot sequences from the gene 32 autoregulatory region results in an mRNA that cannot be repressed by gp32. These results suggest that the zinc-binding subdomain of gp32 plays an essential role in autoregulation by providing a critical element necessary for nucleating cooperative binding at the gene 32 mRNA pseudoknot.     

Binding of the bacteriophage T4 regA protein to mRNA targets: an initiator AUG is required.

Bacteriophage T4 regA protein translationally represses the synthesis of a subset of early phage-induced proteins. The protein binds to the translation initiation site of at least two mRNAs and prevents formation of the initiation complex. We show here that the protein binds to the translation initiation sites of other regA-sensitive mRNAs. Analysis of mRNA binding by filtration and nuclease protection assays shows that AUG is necessary but not sufficient for specific binding of regA protein to its mRNA targets. Anticipating the need for large quantities of regA protein for structural studies to further define the regA protein-RNA ligand interaction, we also report cloning the regA gene into a T4 overexpression system. The expression of regA protein in uninfected E. coli is lethal, so in our system regA driven by a strong T7 promoter is sequestered in a T4 phage until 'induction' by phage infection is desired. We have replaced the regA sensitive wild-type ribosome binding site with a strong insensitive ribosome binding site at an optimal distance from the regA initiation codon for maximizing expression. We have obtained large amounts of regA protein.     

Translational stimulation: RNA sequence and structure requirements for binding of Com protein.

Translation of the bacteriophage Mu mom gene is positively regulated by the phage Com protein. We report here that purified Com protein specifically stimulates mom gene expression in vitro. Furthermore, Com is shown to bind a site in the mom translational initiation region (TIR) in a sequence-specific manner. In vitro RNA footprint experiments have been used to define the Com-binding site and to study mRNA secondary structure in the mom TIR. Com binding is shown to correlate with a conformational change in the mom TIR both in vivo and in vitro. The role of secondary structure was further examined by testing the effects of mutations in the TIR on translation and stimulation. The results support a model for translational stimulation in which Com binding induces a conformational change in the mom mRNA, thereby enhancing ribosome binding.     

Specificity of translational regulation by two DNA-binding proteins

The gene V protein of the filamentous bacteriophages f1, fd and M13, and the gene 32 protein of bacteriophage T4 share the property of binding strongly and co-operatively to single-stranded nucleic acids, especially DNA. Moreover, both are capable of repressing the translation of specific mRNAs (gene 32 protein its own, and gene V protein that of the filamentous phage gene II), both in vivo and in vitro. If the mechanism of repression by either of these proteins were based solely on its ability to bind single strands co-operatively, then the other would be expected to mimic or interfere with its effect in vitro. We have found no such mimicry or interference, even at protein concentrations high enough to have substantial non-specific effects on translation. This suggests that the sites of repression on the mRNAs must offer something other than simple “unstructuredness” for binding and repression to occur.     

Translational repression: biological activity of plasmid-encoded bacteriophage T4 RegA protein

The RegA protein of bacteriophage T4 is a translational repressor that regulates expression of several phage early mRNAs. We have cloned wild-type and mutant alleles of the T4 regA gene under control of the heat-inducible, plasmid-borne leftward promoter (PL) of phage lambda. Expression of the cloned regA+ gene resulted in the synthesis of a protein that closely resembled phage-encoded RegA protein in biological properties. It repressed its own synthesis (autogenous translational control) as well as the synthesis of specific T4-encoded proteins that are known from other studies to be under RegA-mediated translational control. Cloned mutant alleles of regA exhibited derepressed synthesis of the mutant regA gene products and were ineffective in trans against RegA-sensitive mRNA targets. The effects of plasmid-encoded RegA proteins were also demonstrated in experiments using two compatible plasmids in uninfected Escherichia coli. The two-plasmid assays confirm the sensitivities of several cloned T4 genes to RegA-mediated translational repression and are well-suited for genetic analysis of RegA target sites. Repression specificity in this system was demonstrated by using wild-type and operator-constitutive translational initiation sites of T4 rIIB fused to lacZ. The results show that no additional T4 products are required for RegA-mediated translational repression. Additional evidence is provided for the proposal that uridine-rich mRNA sequences are preferred targets for the repressor. Surprisingly, plasmid-generated RegA protein represses the synthesis of some E. coli proteins and appears to enhance selectively the synthesis of others. The RegA protein may have multiple functions, and its binding sites are not restricted to phage mRNAs.     

Characterization of bacteriophage T4 regA protein-nucleic acid interactions

The bacteriophage T4 regA protein is a translational repressor of a group of T4 early mRNAs. We have characterized the binding of regA protein to polynucleotides and to specific RNAs. Binding to nucleic acids was monitored by the quenching of the intrinsic tryptophan fluorescence of regA protein. regA protein exhibited differential affinities for the polynucleotides examined, with the order of affinity being poly(rU) greater than poly(dT) greater than poly(dU) = poly(rG) greater than poly(rC) = poly(rA). The binding site size calculated for regA protein binding to poly(rU) was n = 9 +/- 1 nucleotides. Cooperativity was observed in binding to multiple-site oligonucleotides, with a cooperativity parameter (omega) value of 10-22. To study the specific interaction between regA protein and T4 gene 44 mRNA, the affinity of regA protein for synthetic gene 44 RNA fragments was measured. The association constant (Ka) for regA protein binding to gene 44 RNA fragments was 100-fold higher than for binding to nontarget RNA. Study of variant gene 44 RNA fragments indicated that the nucleotides required for specific binding are contained within a 12-nucleotide sequence spanning -12 to -1, relative to the AUG codon. The bases of five nucleotides (indicated in upper case type) are critical for specific regA protein interaction with the gene 44 recognition element, 5'-aaUGAGgAaauu-3'. These studies further showed that formation of a regA protein-RNA complex involves a maximum of 2-3 ionic interactions and is primarily an enthalpy-driven process.     

Autogenous translational operator recognized by bacteriophage T4 DNA polymerase

The synthesis of the DNA polymerase of bacteriophage T4 is autogenously regulated. This protein (gp43), the product of gene 43, binds to a segment of its mRNA that overlaps its ribosome binding site, and thereby blocks translation. We have determined the Kd of the gp43-operator interaction to be 1.0 x 10(-9) M. The minimum operator sequence to which gp43 binds consists of 36 nucleotides that include a hairpin (containing a 5 base-pair helix and an 8 nucleotide loop) and a single-stranded segment that contains the Shine-Dalgarno sequence of the ribosome binding site. In the distantly related bacteriophage RB69 there is a remarkable conservation of this hairpin and loop sequence at the ribosome binding site of its DNA polymerase gene. We have constructed phage operator mutants that overproduce gp43 in vivo, yet are unchanged for in vivo replication rates and phage yield. We present data that show that the replicative and autoregulatory functions are mutually exclusive activities of this polymerase, and suggest a model for gp43 synthesis that links autoregulation to replicative demand.     

Translational autocontrol of the Escherichia coli ribosomal protein S15

When rpsO, the gene encoding the ribosomal protein S15 in Escherichia coli, is carried by a multicopy plasmid, the mRNA synthesis rate of S15 increases with the gene dosage but the rate of synthesis of S15 does not rise. A translational fusion between S15 and beta-galactosidase was introduced on the chromosome in a delta lac strain and the expression of beta-galactosidase studied under different conditions. The presence of S15 in trans represses the beta-galactosidase level five- to sixfold, while the synthesis rate of the S15-beta-galactosidase mRNA decreases by only 30 to 50%. These data indicate that S15 is subject to autogenous translational control. Derepressed mutants were isolated and sequenced. All the point mutations map in the second codon of S15, suggesting a location for the operator site that is very near to the translation initiation codon. However, the creation of deletion mutations shows that the operator extends into the 5' non-coding part of the message, thus overlapping the ribosome loading site.     

The bacteriophage T4 regA gene: primary sequence of a translational repressor

The regA gene product of bacteriophage T4 is an autogenously controlled translational regulatory protein that plays a role in differential inhibition (translational repression) of a subpopulation of T4-encoded "early" mRNA species. The structural gene for this polypeptide maps within a cluster of phage DNA replication genes, (genes 45-44-62-regA-43-42), all but one of which (gene 43) are under regA-mediated translational control. We have cloned the T4 regA gene, determined its nucleotide sequence, and identified the amino-terminal residues of a plasmid-encoded, hyperproduced regA protein. The results suggest that the T4 regA gene product is a 122 amino acid polypeptide that is mildly basic and hydrophilic in character; these features are consistent with known properties of regA protein derived from T4-infected cells. Computer-assisted analyses of the nucleotide sequences of the regA gene and its three upstream neighbors (genes 45, 44, and 62) suggest the existence of three translational initiation units in this four-gene cluster; one for gene 45, one for genes 44, 62 and regA, and one that serves only the regA gene. The analyses also suggest that the gene 44-62 translational unit harbors a stable RNA structure that obligates translational coupling of these two genes.     

Quantitative correlation between mRNA secondary structure around the region downstream of the initiation codon and translational efficiency in Escherichia coli

Translational efficiency in Escherichia coli is known to be strongly influenced by the secondary structure around the ribosome‐binding site and the initiation codon in the translational‐initiation region of the mRNA. Several quantitative studies have reported that translational efficiency is attributable to effects on ribosome accessibility predominantly caused by the secondary structure surrounding the ribosome‐binding site. However, the influence of mRNA secondary structure around regions downstream of the initiation codon on translational efficiency after ribosome‐binding step has not been quantitatively studied. Here, we quantitatively analyzed the relationship between secondary structure of mRNA surrounding the region downstream of the initiation codon, referred to as the downstream region (DR), and protein expression levels. Modified hairpin structures containing the initiation codon were constructed by site‐directed mutagenesis, and their effects on expression were analyzed in vivo. The minimal folding free energy (ΔG) of a local hairpin structure was found to be linearly correlated with the relative expression level over a range of fourfold change. These results demonstrate that expression level can be quantitatively controlled by changing the stability of the secondary structure surrounding the DR. Biotechnol. Bioeng. 2009; 104: 611–616 © 2009 Wiley Periodicals, Inc.