Isolation of the rstA Gene as a Multicopy Suppressor of YjeE, an Essential ATPase of Unknown Function in Escherichia coli

Isolation of multicopy yjeE suppressors.

An E. coli chromosomal library was made by partial digestion of E. coli MG1655 DNA with Sau3AI to generate 3- to 4-kb fragments, which were ligated with BamHI-digested pGEM (10). This library was transformed into a previously characterized yjeE deletion strain (EB444; MG1655 araBAD::yjeE kan yjeE::cat) (1). This strain contains a deletion of the yjeE gene at its native locus and has a conditionally expressed complementing copy of yjeE on the chromosome at the araBAD locus, under the control of the P_BAD promoter (1). Transformants were plated on LB medium containing ampicillin (50 μg/ml; for pGEM) and kanamycin (50 μg/ml; a marker for the complementing chromosomal yjeE copy) at 37°C. No colonies were observed in a control experiment where pGEM containing no insert was transformed and plated in the absence of arabinose. Thus, transformants grown in the absence of arabinose were likely to be suppressors of the lethal phenotype. A number of colonies that suppressed the lethal yjeE phenotype were isolated after transformation of the pGEM E. coli library. To confirm that these single colonies were due to suppression from a gene carried on the plasmid and not from a spontaneous chromosomal mutation, plasmid DNA was isolated, transformed into fresh EB444 cells, and screened for a loss of arabinose dependence. Plasmids that passed this secondary screening step were isolated for the identification of cloned sequences.

Identification of clones responsible for the suppression of the lethal yjeE phenotype.

From a screening of approximately 700,000 transformants, 84 suppressing clones were identified. The plasmids isolated from these clones were initially screened by PCR with plasmid-specific T7 and SP6 primers and a yjeE-F primer (5′-ACCATGATGAATCGAGTAATTCCGCTC-3′). Three primers were used in this PCR, since the orientation of the plasmid inserts was random. This PCR identified 74 suppressor clones that contained yjeE as part of the plasmid insert. The remaining plasmids, which did not contain yjeE, segregated into two different insert sizes, as estimated by the T7-SP6 PCR product. These plasmids were designated pGEM-long (3.5-kb insert; 7 suppressors) and pGEM-short (1-kb insert; 3 suppressors). Both plasmids were sequenced and compared against GenBank to identify the open reading frames present. The pGEM-long plasmid contained three complete open reading frames (ydgB, ydgC, and rstA) and two partial gene fragments (ydgI and rstB) (Fig. 1). The pGEM-short plasmid contained only the rstA gene with a very small fragment from the rstB gene (Fig. 1), implicating rstA as the factor responsible for the suppression of the lethal yjeE phenotype. This was further confirmed by cloning rstA into plasmid pBF9, a protein overexpression vector containing a strong tac promoter, derived from pKK223-3 and described in reference 4 (Fig. 1). EB444 cells harboring the pBF9-rstA plasmid also became capable of growth in the absence of arabinose, confirming rstA as a multicopy suppressor of the lethal phenotype associated with YjeE depletion. High-copy expression subclones containing the ydgB and ydgC genes were also constructed, but were incapable of the suppression (data not shown).

• Open in new tab
• Download powerpoint

FIG. 1.

Identification of the gene responsible for the suppression of the yjeE lethal phenotype. The shaded areas indicate sequences that are present on the plasmids. The pGEM-long and pGEM-short plasmids were isolated from a random genomic library. pBF9-rstA was constructed by cloning rstA between the NdeI and HindIII sites on the pBF9 vector.

Complementation of YjeE depletion by overexpression of RstA.

The three rstA multicopy suppression clones (Fig. 1) were grown on solid medium and compared to the conditionally complemented yjeE deletion strain containing the empty pGEM vector (Fig. 2A). EB444(pGEM) was incapable of growth in the absence of arabinose. EB444(pGEM-short) grew in both the presence and absence of arabinose but produced smaller colonies than the EB444(pGEM-long) strain on the plate with no arabinose. The most robust of the three multicopy suppression plasmids was pBF9-rstA, which exhibited almost identical growth in the presence and absence of arabinose. The inability of the pGEM-short clone to fully complement the YjeE-depleted strain implies that perhaps some upstream or downstream regulatory elements are missing that control rstA expression and that these elements are present in the pGEM-long plasmid.

• Open in new tab
• Download powerpoint

FIG. 2.

Arabinose dependence of the conditional mutant strain EB444 compared with the multicopy suppressor strains. (A) Strains were grown on solid medium overnight at 37°C. The left plate shows EB444, with each plasmid grown in the presence of arabinose, and the right plate shows the same strains in the absence of arabinose. (B) Strains were grown in M9 liquid medium at 37°C with shaking at 250 rpm. ⧫, EB444; ▪, EB444 with 0.2% arabinose; ▴, EB444(pBF9-rstA); ×, EB444(pBF9-rstA) with 0.2% arabinose.

These suppression results were confirmed in liquid culture using M9 minimal medium (Fig. 2B). Overnight cultures of each strain were grown in M9 minimal medium and diluted 1,000-fold in fresh medium in the morning. Independent of arabinose, EB444(pBF9-rstA) was capable of growth similar to that of the fully complemented yjeE deletion strain (EB444 with arabinose). EB444(pGEM-long) showed a slightly decreased growth rate compared to that of EB444(pBF9-rstA), and the growth rate of EB444(pGEM-short) was further diminished (data not shown). All three of the suppressing plasmids showed greatly improved growth compared with EB444(pGEM), which exhibited almost no growth during the time course of the experiment. These results suggest that high levels of expression, such as that expected from the pKK223-derived protein overexpression plasmid pBF9, are necessary for complete complementation of the YjeE-depleted strain.

Effects of overexpression of RstA on yjeE expression.

RstA is believed to be a response regulator in E. coli (18, 20), but the downstream impact of this effector molecule has not been determined. It is likely that a number of processes could be affected, including an increase in the expression of yjeE from the P_BAD promoter. Therefore, it was necessary to test if overproduction of RstA caused overproduction of YjeE. This possibility was tested by cloning a luciferase reporter gene under the control of the P_BAD promoter and testing for induction by pBF9-rstA. Two strains were used in this experiment, EB1789 [MG1655(pBAD33-luc)] and EB1790 [MG1655(pBAD33-luc-pBF9-rstA)]. Both strains were grown overnight in LB at 37°C and diluted 100-fold in the morning in fresh medium containing 1 mM isopropyl-β-d-thiogalactopyranoside and either 0.2% arabinose or no arabinose. Strains were allowed to grow until the mid-exponential phase, when samples were removed for luminescence analysis. The Promega luciferase assay system (Promega Corporation, Madison, WI) protocol was followed, and the luminescence of each sample was determined and is shown in Table 1. Both EB1789 and EB1790 showed good luminescence in the presence of arabinose, a positive control since arabinose will turn on the P_BAD promoter which controls the expression of the luminescence reporter gene. In the absence of arabinose, overexpression of rstA had no effect on luminescence, indicating that RstA does not turn on the P_BAD promoter. The luminescence of strain EB1789 in the absence of arabinose is likely due to low levels of leakage from the P_BAD promoter, and EB1790, which overexpresses RstA, has nearly identical levels of background luminescence. Similar results were seen with Western blotting for YjeE levels in the presence and absence of RstA overexpression (data not shown). From these results, we conclude that the ability of RstA overexpression to suppress the lethal phenotype of YjeE depletion is not the result of turning on YjeE expression.

RstA alone is not sufficient to overcome the loss of YjeE.

Given the result that pBF9-rstA is capable of fully complementing the lethal phenotype of a YjeE-depleted strain, the question of whether it would be possible to replace YjeE function completely by overexpression of RstA arose. To test this possibility, pBF9-rstA was transformed into wild-type MG1655 E. coli cells, followed by attempts to replace yjeE with a chloramphenicol resistance cassette by using established methods (5) and the helper plasmid pKD119. All attempts at creating this yjeE deletion strain failed, and to confirm that the procedure was working, a similar experiment was conducted in a yjeE diploid strain (MG1655 araBAD::yjeE kan^r). In this diploid, background replacement of yjeE at the native locus was easily obtained, confirming that RstA overexpression was not capable of replacing the function of YjeE. We propose that at least a small amount of YjeE is essential for suppression of the lethal phenotype of depletion with high expression of RstA.

Possible roles of RstA in the suppression of the YjeE-depleted phenotype.

RstA is believed to be a response regulator in the RstAB two-component system (8). Two-component systems are signal transduction protein pairs in bacteria that typically communicate external stimuli to the inside of the cell (6). Two-component systems are composed of a sensor histidine protein kinase that is usually membrane associated and a cytoplasmic response regulator that functions as a transcription factor (6). The sensor kinase responds to external stimuli by autophosphorylating a histidine residue, which is then transferred to its cognate response regulator at a conserved aspartate, followed by activation or repression of gene expression (6). To date, very little is known about the function of RstAB. RstB has been shown to be capable of autophosphorylation, and this phosphate can be transferred to RstA (18). Additionally, none of 20 other E. coli histidine kinases were capable of phosphorylating RstA (18), indicating that it has a very specific functional role that is linked to RstB. Other two-component systems have been observed to mediate drug resistance when overexpressed, but the RstAB system was not found to possess this capability (8). In experiments where both rstA and rstB have been deleted, the double-deletion strain was hypersensitive towards the anticapsule/anti-inflammatory agent ketoprofen, the cholinergic antagonist pridinol, and troleandomycin, a protein synthesis inhibitor (20). In separate work looking at the general E. coli stress regulator σ^s (RpoS), an rstAB double-deletion strain was one of nine two-component deletion strains that showed increased expression of RpoS, implying that RstAB is a negative regulator of RpoS expression (13). Additionally, the rstAB operon has a PhoP box region in its promoter that PhoP has been shown to bind to (12). PhoP is stimulated in response to external Mg²⁺ concentrations and is activated under conditions of Mg²⁺ starvation (12). Nevertheless, the downstream functional targets of activated RstA remain obscure. Until these targets are elucidated, it will be difficult to understand the method by which RstA overexpression is capable of suppressing the lethal phenotype of YjeE depletion. It is unknown if the overexpression of RstA results in a very specific set of genes being activated that are related to YjeE function or if it is a general cellular response that enables the cell to better survive under conditions of YjeE depletion. The stimulus for RstA activation is also unknown, and so, it is possible that depletion of YjeE results in the activation of RstA and that, somehow, increasing the copy number of RstA allows the cells to overcome the lethal depletion phenotype associated with the loss of YjeE by turning on a number of genes required for this function. Future studies involving the function of the RstAB two-component system will help to identify its mechanism of yjeE suppression and provide information on YjeE functions.