Browse Subject Areas

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Variation of Intragenic Tandem Repeat Tract of tolA Modulates Escherichia coli Stress Tolerance

  • Kai Zhou,

    Affiliation Laboratory of Food Microbiology and Leuven Food Science and Nutrition Research Centre (LFoRCe), Department of Microbial and Molecular Systems (M2S), Faculty of Bioscience Engineering, KU Leuven, Leuven, Belgium

  • Chris W. Michiels,

    Affiliation Laboratory of Food Microbiology and Leuven Food Science and Nutrition Research Centre (LFoRCe), Department of Microbial and Molecular Systems (M2S), Faculty of Bioscience Engineering, KU Leuven, Leuven, Belgium

  • Abram Aertsen

    Affiliation Laboratory of Food Microbiology and Leuven Food Science and Nutrition Research Centre (LFoRCe), Department of Microbial and Molecular Systems (M2S), Faculty of Bioscience Engineering, KU Leuven, Leuven, Belgium

Variation of Intragenic Tandem Repeat Tract of tolA Modulates Escherichia coli Stress Tolerance

  • Kai Zhou, 
  • Chris W. Michiels, 
  • Abram Aertsen


In recent work we discovered that the intragenic tandem repeat (TR) region of the tolA gene is highly variable among different Escherichia coli strains. The aim of this study was therefore to investigate the biological function and dynamics of TR variation in E. coli tolA. The biological impact of TR variation was examined by comparing the ability of a set of synthetic tolA variants with in frame repeat copies varying from 2 to 39 to rescue the altered susceptibility of an E. coli ΔtolA mutant to deoxycholic acid, sodium dodecyl sulfate, hyperosmolarity, and infection with filamentous bacteriophage. Interestingly, although each of the TolA variants was able to at least partly rescue the ΔtolA mutant, the extent was clearly dependent on both the repeat number and the type of stress imposed, indicating the existence of opposing selective forces with regard to the optimal TR copy number. Subsequently, TR dynamics in a clonal population were assayed, and we could demonstrate that TR contractions are RecA dependent and enhanced in a DNA repair deficient uvrD background, and can occur at a frequency of 6.9×10−5.


DNA sequences harboring tandem repeats (TRs) exist in both prokaryotic and eukaryotic genomes, and are considered to be hypermutable loci in which the TR copy number can increase or decrease as a result of strand-slippage replication or recombination (reviewed in [1][3]). The frequency of TR expansions or contractions depends on intrinsic features of the TR tract (such as the length, copy number and sequence conservation of the TR unit) as well as extrinsic environmental conditions [4][6]. Obviously, TR rearrangements occurring within promoter or coding regions can affect the transcription and translation of the corresponding genes, or even the functionality of the gene products [7][13]. In microorganisms, TR variations are therefore often forwarded as a bet-hedging strategy, from which a population could phenotypically benefit on a short evolutionary time scale [14].

In silico analysis of the E. coli MG1655 genome readily reveals about 30 genes with an intragenic in frame TR region, in which TR copy number variations thus might affect the functionality of the corresponding protein (unpublished results). However, the effect of TR variation has been studied only in few of these genes. One study showed that in frame expansion of a trimeric (TCT) TR tract from four to five copies in the peroxiredoxin gene ahpC converted the enzyme into a disulfide reductase that suppressed loss of fitness in mutants defective in the reduction of protein disulfide bonds [15]. Another study showed that gain or loss of one unit from a three-unit hexameric (CTGGCG) TR tract in the mismatch repair gene mutL caused an increased mutation rate. Since the TR region is part of the ATP-binding pocket of MutL, a defective ATPase activity was suggested to have caused the mutator phenotype [16].

The current work focuses on the tolA gene, which has a TR region consisting of 13 imperfect repeats of 15 or 18 bp each, and encoding a lysine and alanine rich segment in the TolA membrane protein [17], [18]. As part of the Tol-Pal envelope complex which spans the periplasmic space from the outer membrane to the cytoplasmic membrane and which is important for cell integrity [19], TolA has been implicated in group A colicin uptake [20], [21], filamentous phage infection [22], [23] and detergent tolerance [24]. Notably, the TR region is located within the C-terminal region of domain II of TolA, which comprises a long α-helical domain that connects the cytoplasmic membrane anchor domain I with the periplasmic domain III.

Recent work of our group showed the copy number of tolA TR units to vary from 8 to 16 among 234 analyzed E. coli isolates [25], but the phenotypical impact of this variation remains unknown. In this study, we therefore aimed to investigate the function and dynamics of TolA TR variation in E. coli.

Materials and Methods

Bacterial Strains, Plasmids and Growth Conditions

The strains and plasmids used in this study are listed in Table S1. Bacterial strains were grown overnight in well aerated 4-ml cultures of Luria-Bertani (LB) or M9 medium [26] at 37°C, unless mentioned otherwise. For high osmolarity assays, the NaCl concentration of LB was increased to 0.6 M. Antibiotics (Applichem, Darmstadt, Germany) were used at the following concentrations: 100 µg/ml ampicillin (Ap); 25 µg/ml of chloramphenicol (Cm); 10 µg/ml tetracycline (Tc); 50 µg/ml kanamycin (Km); 50 µg/ml streptomycin (St).

DNA Techniques

Plasmids were isolated using a mini-prep kit (Fermentas, St. Leon-Rot, Germany), and DNA fragments were purified from agarose gels using a gel extraction kit (Fermentas). PCR was performed with DreamTaq polymerase (Fermentas) for examination and Phusion polymerase (Finnzymes, Vantaa, Finland) for cloning and sequencing. Samples for sequencing were prepared with the BigDye terminator V3.1 cycle sequencing kit (ABI, Foster City, CA, U.S.), and sequenced at the division of gene technology (KU Leuven, Belgium).

Construction of tolA Knockouts and Chromosomal TR Variants

A deletion mutant MG1655 ΔtolA was constructed using the Datsenko and Wanner method [27]. First, a ΔtolA::kan fragment containing a kanamycin resistance cassette replacing the tolA gene was amplified from strain EVV54 [28] by primer pair tolA-Fw (5′-ACTTGAATTCGTAACAGGCGAACAGTTTTT-3′) and tolA-Rev (5′-TCGTGGATCCTACCAGAACCCCGTGGCAA-3′), and used to exchange the wild-type tolA allele in the chromosome of E. coli MG1655, resulting in MG1655 ΔtolA::kan. Subsequently, MG1655 ΔtolA was derived from MG1655 ΔtolA::kan by flipping out the FRT-flanked kan gene.

TR variants of the tolA gene were constructed first on a plasmid and subsequently introduced in the chromosome by the following stepwise procedure. First, the wild-type tolA allele of MG1655 (i.e. with 13 TR units and further referred to as tolA13TR) was PCR-amplified with the tolA-Fw and tolA-Rev primers, digested with EcoRI and BamHI (Fermentas), and cloned into pTrc99A digested with the same enzymes, to yield pTrc99A-tolA13TR. The allele referred to as tolA harbors two consecutive stop codons in the third TR (TR3), and was constructed by opening pTrc99A-tolA13TR with the back-to-back primer pair tolA-pTrc stop Fw (5′-GCTGAAAAGGCTGCAGCTGATTAATAAGCGGCAGCAGAGAAAGC-3′) and tolA-pTrc stop Rev (5′-AGCCGCTTTCTTCTCAGCTTCTGCTTTGGCT-3′), of which the former contains the stop codons (underlined). After phosphorylation by PNK T4 kinase (Fermentas), this amplicon was closed again by self-ligation to yield pTrc99A-tolA. To create a tolA variant with two repeat units (tolA2TR), primer pair 2-repeats-Fw (5′-GCAGAGGCAGATGATATTTTCGGTG-3′) and 2-repeats-Rev (5′-TGCTGCTTTTTCAGCTGCTGCTTTTTCAGCCTTCTCAGCTTCTGC-3′) was used to open pTrc99A-tolA13TR by PCR and at the same time replace the entire 13TR region by two consensus repeat units (underlined), as defined by the Tandem Repeat Finder program [29]. This amplicon was phosphorylated and self-ligated to yield pTrc99A-tolA2TR. The tolA6TR and tolA8TR alleles contain 6 and 8 repeats, respectively, were derived from pTrc99A-tolA by selecting for repeat deletion. Briefly, MG1655 ΔtolA containing pTrc99A-tolA was streaked on LB plates with 0.2% (w/v) deoxycholic acid (DOC, Fisher-scientific, Erembodegem, Belgium). Several DOC tolerant revertants which based on PCR analysis had incurred deletions in the tolA allele were identified, and a specific pTrc99A-tolA6TR and pTrc99A-tolA8TR construct was retained after confirmation by sequencing. To create a tolA26TR and tolA39TR allele, back-to-back primer pair 2-repeat Fw and 26-repeat Rev (5′-GGCCGCTTTTGCTGCAGCGGCT-3′) was used to open pTrc99A-tolA13TR at the end of the TR region by PCR. Subsequently, this amplicon was ligated with an amplicon of the wild-type tolA TR region (containing 13 repeats) obtained with primer pair Pure_TRs Fw (5′-GCTGAGAAGAAAGCGGCTGC-3′) and 26- repeat Rev, to yield pTrc99A-tolA26TR (i.e. single insert) and pTrc99A-tolA39TR (i.e. double insert).

Finally, the tolA13TR wild-type allele and each of the above constructed plasmid-borne tolA alleles (except tolA; see below) were amplified with tolA-Fw and tolA-Rev and individually exchanged with ΔtolA::kan in the MG1655 ΔtolA::kan mutant expressing the λ-Red system from pKD46 [27]. Transformants were selected on LB +1% (w/v) sodium dodecyl sulfate (SDS, Applichem, Omaha, U.S.) plates, to which MG1655 ΔtolA::kan is highly sensitive. As such, MG1655 variants with different chromosomal tolA alleles (tolA2TR, tolA6TR, tolA8TR, tolA13TR, tolA26TR or tolA39TR) were constructed. Loss of the kanamycin resistance cassette and acquisition of the correct allele was corroborated by PCR and sequencing. Please note that the reconstructed MG1655 TolA13TR variant showed no phenotypic differences with the original wild-type MG1655, suggesting that the entire procedure of generating chromosomal tolA variants (including a selection on SDS) did not lead to a selection of undesired spontaneous mutants.

The construction of the MG1655 tolA mutant could not be done by SDS selection because tolA is a null allele that does not restore SDS tolerance, and was therefore done by using rpsL based counterselection [30]. Briefly, the rpsl-neo cassette was amplified from MG1655 rpsL150 kdpA4::rpsL-neo by primer pair rpsL-neo_Fw (5′-GCGAACAGTTTTTGGAAACCGAGAGTGTCAAAGGCAACCGGCCTGGTGATGATGGCGGGATCG-3′) and rpsL-neo_Rev (5′-TGCCTGATGTTGACCGTCCGAACAGTCAACATCGCGATTATCAGAAGAACTCGTCAAGAAGGCG-3′), and then transformed to MG1655 rpsL150 (StR) expressing the λ-Red system from pKD46 to replace tolA, generating MG1655 rpsL150 ΔtolA::rpsL-neo (StS). MG1655 rpsL150 tolA was then obtained by λ-Red based exchange of the ΔtolA::rpsL-neo allele with the tolA amplicon, and selecting for St resistance.

Strains ZK1, ZK2 and ZK3 were constructed by P1vir transduction of antibiotic resistance markers from donor strains E. coli TH446 recA::cat [31], E. coli AB1157 mutS::Tn10 [32] and E. coli JJC40 uvrD::Tn5 [33], respectively, into recipient strain MG1655 rpsL150 tolA.

Extraction of Membrane Proteins and Western Blot

Membrane proteins of E. coli MG1655 were extracted as previously described [34]. Briefly, cells from 200 ml stationary phase cultures were harvested by centrifugation for 10 min at 2,900×g, resuspended in 10 ml 10 mM Tris-HCl pH 8.0, and lysed by three cycles of freezing and thawing followed by sonication. These samples were centrifuged for 1 hour at 100,000×g. The resulting pellet was washed with 10 ml 10 mM Tris-HCl buffer (pH 6.8) supplemented with 1.0 M NaCl and centrifuged again for 1 hour at 100,000×g. The pellet was then resuspended in 3 ml of a 10 mM Tris-HCl buffer (pH 6.8) supplemented with 2% Triton X-100, 10 mM MgCl2 and 150 mM NaCl to dissolve the membrane proteins. Finally, the remaining cell debris was pelleted for 1 hour at 100.000×g and discarded. All steps were performed at 4°C.

Protein samples of about 50 µg as quantified with the Novagen BCA Protein Assay Kit (Merck, Darmstadt, Germany) were boiled for 10 min with loading buffer (50 mM Tris-HCl pH 6.8, 2% SDS, 10% glycerol, 1% β-mercaptoethanol, 12.5 mM EDTA, 0.02% bromophenol blue) and separated on 10% polyacrylamide gel by SDS-PAGE. Western blotting was essentially done as previously described [35], and afterwards the membrane was incubated with a 1∶1000 dilution of the anti-TolAIII polyclonal antibodies (PAbs) (a generous gift of Dr. Lloubes, CNRS, France; [36]) in phosphate buffered saline supplement with 0.1% Tween 20 (PBS-T) for 1 hour at 4°C. Subsequently, the membrane was washed with PBS-T for 3×10 min and incubated with horseradish peroxidase (HRP)-conjugated goat anti-rabbit PAbs (1∶30000 dilution) for 1 hour at 4°C. Finally, the blot was washed again in PBS-T before being developed with enhanced chemiluminescence (ECL) substrate solution (Thermo, Rockford, U.S.).

Determination of Stress Tolerance and Luria-Delbrück Fluctuation Assays

Stationary phase cultures of MG1655 and its derived mutants were serially diluted and plated on LB, LB with 1% DOC (w/v) or 4% SDS (w/v), LB with 0.6 M NaCl (LBS), and LB and LBS adjusted to pH 5.0 with HCl. After overnight incubation at 37°C, colonies were counted and the relative plating efficiency was calculated as (Nd/N0)×100%, with Nd = colony-forming units (CFU)/ml on LB with additive(s) and N0 =  CFU/ml on LB. MG1655 ΔtolA and MG1655 tolA13TR (corresponding to the wild-type) were always included as reference strains.

The Luria-Delbrück fluctuation assay was performed as previously described [37]. Briefly, 22 independent cultures of MG1655 rpsL150 tolA were grown overnight in LB-broth, after which 10 µl of each culture was separately plated on LB with 1% SDS (referred to as biological replicates). In addition, for one culture, this plating was repeated 22 times (referred to as technical replicates). After overnight incubation, the number of colonies on each plate was counted and the coefficient of variation was subsequently calculated for both the biological and technical replicates as [standard deviation]/[the average number of colonies].

Tolerance to Infection with Filamentous Phage fd

Phage fd-tet-DOG1 was propagated in E. coli TG1 in 2YT medium (16 g Bacto-tryptone, 10 g Bacto-yeast extract, 5 g NaCl in 1L) with 7.5 µg/ml tetracycline at 37°C overnight. After centrifugation (8.000 rpm, 15 min), supernatant containing the phage particles was collected and passed through a 0.2 µm pore-size filter to remove bacteria (Fisher-scientific). Phage titration was done as described previously [26]. Since the primary receptor of phage fd is the tip of the F pilus, the F’ plasmid from E. coli XL1-Blue was introduced into each of the MG1655 tolA variant strains by conjugation. Hundred µl of a log phase LB culture (OD600 ≈ 0.3) of these strains was then mixed with 100 µl fd-tet-DOG1 phage suspension titrated at 100–200 plaque-forming units (PFU), incubated for 30 min at 37°C, mixed with 3 ml LB soft-agar (0.35%) and poured onto an eosin methylene blue (EMB) +1% glycerol agar plate [38]. The plaques, which are dark red on this medium, were counted after overnight growth, and their size was scored. The efficiency of plating phage fd on different mutant backgrounds was expressed relative to that obtained on MG1655 tolA13TR, which was arbitrarily set as 100%.

Statistical Analysis

Plating efficiency experiments were conducted in threefold and results expressed as mean ± standard deviation. Two-tailed unpaired Student’s t-test or analysis of variance (ANOVA) followed by Least Significant Different (LSD) post hoc test was used to determine statistical significance of differences between strains, using log-transformed data when necessary.


Construction of Different tolA TR Variants in E. coli MG1655

In order to investigate the function of the TR region of E. coli TolA, a set of isogenic MG1655 mutants was constructed differing only in TR copy number in the chromosomal tolA locus (Fig. 1). More specifically, TolA variants with 2 (TolA2TR), 6 (TolA6TR), 8 (TolA8TR), 13 (TolA13TR, i.e. corresponding to the wild-type TolA protein in MG1655), 26 (TolA26TR) or 39 (TolA39TR) TR units were generated. While PCR (Fig. 2A) and sequencing (data not shown) confirmed the correct size and sequence of each TR region, Western blotting revealed that TolA6TR, TolA8TR, TolA26TR and TolA39TR were equally well expressed as the parental TolA13TR (Fig. 2B). In contrast, TolA2TR could hardly be detected by Western blot, indicating that it is either poorly expressed, unstable or unable to react with the applied antibodies that are targeted to domain III of TolA. In fact, since domain II and III have previously been shown to interact [39], the shortened domain II in TolA2TR might impose structural alterations in domain III that preclude its detection with the antibodies used in this study. Whatever the correct explanation, at least some functional TolA2TR is produced in MG1655 tolA2TRs since some of the phenotypes caused by a tolA deletion are at least partially reverted in this strain (see further).

Figure 1. Schematic diagram of tolA variants with different TR copy number.

All TR variants constructed in this study and introduced in the E. coli MG1655 chromosome are shown. Numbers refer to base positions, delineating the start and end points of the tolA open reading frame and TR region (vertical black bars).

Figure 2. Examination of tolA TR variants by PCR and Western blot.

(A) PCR analysis of the TR region of tolA variants. (B) Western blot of membrane proteins of TolA variants with polyclonal antibody against TolA domain III. Size markers in bp (A) or kDa (B) are shown on the left. M: Marker; Lane 1: TolA2TR; Lane 2: TolA6TR; Lane 3: TolA8TR; Lane 4: wild type (TolA13TR); Lane 5: reconstructed wild type (TolA13TR); Lane 6: TolA26TR; Lane 7: TolA39TR.

Effect of TolA Repeat Variation on Susceptibility to Infection with Phage fd

Since the TolA protein is structurally involved and essential for entry of filamentous bacteriophages (fd, f1 and M13) in E. coli [22], [39], the susceptibility of the ΔtolA mutant and the different TolA TR variants to infection with fd phage was compared by determining the phage plating efficiency (Table 1). While the ΔtolA mutant displayed resistance to fd infection as expected, all constructed TolA TR copy number variants partially restored phage susceptibility. The fd plating efficiencies on the TolA2TR, TolA6TR, TolA8TR, TolA26TR and TolA39TR variants were not significantly different from each other, but remained significantly lower (28–43%) than those obtained with the control strain expressing the parental TolA13TR (p<0.05; ANOVA). Furthermore, the plaque size seemed to be related to plating efficiency, with the TolA2TR variant hosting the smallest plaques. As a control, we also examined the plating efficiency of λ phage on the different TolA TR variants and the ΔtolA mutant in the same way, but no significant differences were observed (data not shown). This indicates that the observed differences in fd plating efficiency are probably related to the specific function of TolA in fd infection rather than to an indirect effect on the cell surface properties.

Table 1. Plating efficiency of phage fd-tet-DOG1 on E. coli MG1655 tolATR variants.

Effect of TolA Repeat Variation on Tolerance to DOC and SDS

Since TolA is also involved in the maintenance of outer membrane integrity, tolerance to membrane-disrupting agents such as DOC and SDS was determined (Fig. 3). As expected, the ΔtolA mutant exhibited hypersensitivity to both DOC (1%) and SDS (4%), with relative plating efficiencies (colony counts on detergent-containing versus detergent-free LB agar) of <10−5% (“<” indicates that no colonies were formed on the detergent plates). The TolA TR variants all showed considerably enhanced DOC tolerance compared to the ΔtolA mutant and, interestingly, the degree of tolerance increased with TR copy number (Fig. 3). This correlation was even valid for the variants with increased TR number, which had 3.2-fold (TolA26TR) (p<0.001; ANOVA) and 15.5-fold (TolA39TR) (p<0.001; ANOVA) higher plating efficiencies than the reconstructed parental strain (TolA13TR).

Figure 3. Relative plating efficiencies of MG1655 TolA TR variants on DOC 1% and SDS 4%.

The relative plating efficiencies of the ΔtolA mutant were <10−5% on both media, and are not shown. 13TR refers to the reconstructed MG1655 TolA13TR strain which is identical to and produced the same results as the wild-type strain MG1655. Error bars represent standard deviations of three independent biological replicates. Bars of the same series marked with a different letter are significantly different at 0.05 level by Analysis of Variance (ANOVA) of log-transformed data followed by LSD post hoc test.

When colonies of the TolA2TR, TolA6TR and TolA8TR variants picked from 1% DOC plates were regrown, they showed comparable plating efficiency with that of TolA39TR on 1% DOC (data not shown), suggesting they had incurred mutations conferring stable DOC resistance. To further analyze these mutants, 150 colonies from 1% DOC were analyzed by PCR to search for possible TR expansion events that could explain their enhanced tolerance. However, no such events were found, which may indicate that the frequency of TR expansions falls below that of TR-independent resistance mechanisms.

The TolA variants also enhanced tolerance to 4% SDS, but in this case tolerance was fully restored to the level of the parental strain by all variants (p = 0.458) (Fig. 3). Please note that the construction of the TR variants from the ΔtolA strain was based on selection for SDS tolerance, and that this procedure might have allowed the selection of additional SDS-tolerance mutations that could trivially eliminate possible differences between TR variants. To exclude this possibility, the set of TR variants was reconstructed in the absence of any prior SDS selection using rpsL based counterselection [30] as described for the construction of MG1655 tolA, and similarly examined for tolerance to 4% SDS and 1% DOC. The corresponding plating efficiencies of this set of mutants were indistinguishable from those shown in Fig. 3 (data not shown), indicating that the observed differences in SDS and DOC resistance can be fully ascribed to the TR variations.

Influence of TolA Repeat Variation on Sensitivity to High Osmolarity and Low pH

Another phenotype associated with knock-out of TolA in E. coli is reduced growth in LB broth with NaCl at high osmolarity [40]. In agreement with these findings, we observed that the plating efficiency of MG1655 ΔtolA on hyperosmotic medium (LBS; 0.6 M NaCl) was around 40%, significantly less than the 80% of the reconstructed wild-type strain expressing TolA13TR (p<0.001; ANOVA). All the variant TolA strains had wild-type plating efficiencies (Fig. 4). Interestingly, however, the subsequent combination of hyperosmotic conditions with low pH (LBS pH 5.0) caused more outspoken differences and a differentiation among the TolATR variants. First of all, the ΔtolA strain became hypersensitive, with a plating efficiency of only 0.0004% compared to 38% for the wild-type (p<0.001; ANOVA). The strains expressing TolA variants separated into three groups, one with wild-type plating efficiency (TolA6TR and TolA8TR), and two with intermediate (1.4–7.3%) plating efficiency (TolA2TR, and TolA26TR and TolA39TR) (p<0.001; ANOVA) (Fig. 4).

Figure 4. Relative plating efficiencies of MG1655 TolA TR variants on LBS (0.6 M NaCl), LB pH 5 and LBS pH 5 medium.

13TR refers to the reconstructed MG1655 TolA13TR strain which is identical to and produced the same results as the wild-type strain MG1655. Error bars represent standard deviations of three independent biological replicates. Columns of the same series marked with a different letter are significantly different at 0.05 level by Analysis of Variance (ANOVA) of log-transformed data followed by LSD post hoc test.

Spontaneous Repeat Variation in TolA

To examine the possible occurrence of variations in repeat copy number within clonal populations, a compromised tolA allele (i.e. tolA) harboring two consecutive stop codons in the third repeat was constructed and crossed into MG1655. Just like tolA deletion, the tolA null allele conferred hypersensitivity to DOC and SDS. However, while no colonies with stably reverted SDS or DOC tolerance could be retrieved from the ΔtolA mutant, a tolA mutant population typically yielded such phenotypical revertants at a frequency of 1.9×10−5 when plated on 1% SDS and 8×10−6 when plated on 0.2% DOC. PCR analysis of the tolA locus of 181 such colonies revealed all of them to have incurred contractions in the TR region, with a 5-TR deletion (as evaluated by electrophoretic sizing of the amplified TR region; [25]) being most predominantly retrieved under both DOC (101/104) band SDS (75/77) stress. Furthermore, TolA revertants obtained on LBS pH 5.0 also typically incurred a 5-TR deletion (17/18).

In addition, a Luria-Delbrück fluctuation assay revealed a much higher coefficient of variation in the mutation frequencies obtained from independent cultures (i.e. 1.4) than that obtained from replicates of a single culture (i.e. 0.2), indicating that TolA revertants were pre-existing in the population and not induced by selection on SDS.

Finally, since each TR in tolA has a unique sequence, a number of these contracted tolA loci were sequenced to exactly pinpoint the start and end of the TR deletion (Fig. 5). As expected, the TR unit containing the two stop codons (i.e. TR3) was always deleted, and the deletions were always contiguous and centered around TR3. Moreover, in some of the recovered alleles (e.g. allele 2; Fig. 5) the deletions extended upstream of the TR region as it was previously delineated by us and others. Upon closer inspection, these upstream sequences did also qualify as TRs although they showed a lower degree of sequence conservation, and they could also be identified as TRs when employing different alignment parameter settings (match  = 2, mismatch  = 3, indels  = 5) in Tandem Repeat Finder. Interestingly, a tolA gene with exactly the same deletion as allele 2 exists in a natural isolate of E. coli (STEC_94C; [25]), confirming that such deletion events can also occur in nature.

Figure 5. Schematic diagram of tolA TR deletions.

Different tolA TR deletions obtained from MG1655 tolA† which carries a nonfunctional tolA gene with a TR unit with two stop codons are shown. A total of 39 clones with TR deletions were sequenced, among which seven different TR deletions were found. The 13 TR units that have been recognized in previous studies [17], [18] are numbered. In some cases, the deletions extended in the region upstream of TR1 which contains some less well conserved TR units (dotted lines) (see main text). The differences of unit size were shown in proportion. Allele 1 was found 33 times, both under SDS and DOC selection, and in the recA, uvrD and mutS backgrounds. All other alleles were found only once. Alleles 2 and 3 were recovered from the strain without additional mutations in recombinational pathways under 0.6 M NaCl pH 5.0 and DOC stress, respectively. Alleles 4 and 5 were recovered from the recA background under SDS stress, and alleles 6 and 7 from the uvrD background under SDS stress.

Cellular Functions Involved in TR Contraction in tolA

To investigate potential cellular functions affecting TR rearrangements within the tolA locus, the impact of key DNA repair (uvrD and mutS) and recombination (recA) proteins on the contraction frequency of the tolA allele was determined. Since a recA mutant is hypersensitive to the DNA damage induced by DOC [41], exposure to 1% SDS was chosen as a selection pressure in these experiments, and the uvrD, mutS and recA mutations themselves did not confer increased sensitivity to this growth condition (data not shown). The TR contraction frequency in the recA mutant was only 15% (1.0×10−5) of that of the parental strain (6.9×10−5) (p<0.01; Student’s t-test), indicating these rearrangements to be RecA-dependent. In contrast, an uvrD mutation stimulated TR contractions 3.3-fold (2.3×10−4) compared to parental levels (p<0.01; Student’s t-test). Finally, mutS (3.8×10−5) had no significant effect on repeat stability in tolA (p>0.05; Student’s t-test).

As in the wild-type background, a 5-TR deletion also was predominantly retrieved under both DOC (64/66) and SDS (37/40) stress in the various DNA repair and recombination deficient backgrounds. Furthermore, upon sequencing, some novel rearranged tolA alleles were recovered from recA and uvrD mutants under SDS stress, some of which also involved the upstream sequences of the originally assigned TR region (Fig. 5).


In this study, we investigated the possible biological function and dynamics of TR variation in the E. coli tolA gene. Comparison of a constructed set of isogenic mutants varying only in the copy number of in frame TR units in the tolA gene, revealed that each of these TolA TR variants was able to rescue the aberrant phenotypes incurred by a ΔtolA mutant in response to various biological and chemical stresses, although the extent of this complementation was dependent on both the TR copy number and the type of stress imposed.

The most outspoken TR-dependent phenotype was DOC tolerance, for which plating efficiencies increased with an increasing number of TR units from TolA2TR to TolA39TR over a range of four orders of magnitude. DOC is the major component of bile salts, which constitute a major stress factor for E. coli and other bacteria in the mammalian gut. In fact, bile salts have recently been recognized as an important evolutionary selection force, contributing to the diversification of enteric species such as E. coli and Salmonella enterica [42]. As a result, a number of bile resistance mechanisms have already been identified and documented, mainly involving efflux pumps (AcrAB and EmrAB), outer membrane proteins (OmpF and OmpC), SOS response, and two-component systems (i.e. PhoPQ) ([43]; also reviewed in [41], [44]). Nevertheless, this study is the first to demonstrate that variation of TolA TR copy numbers can modulate DOC tolerance in E. coli.

In contrast to DOC sensitivity, all TolA TR variants complemented sensitivity to SDS and hyperosmolarity equally well and up to wild-type level. However, when hyperosmolarity was combined with low pH, the TolA6TR, TolA8TR, and TolA13TR strains outperformed the other variants carrying either lower or higher TR copy numbers. Although the exact molecular mechanisms behind such differences remains to be elucidated, these findings underscore the intricate phenotypical changes brought about by TolA TR variation.

Finally, all TolA variants were significantly less susceptible to filamentous phage fd than the strain expressing wild-type TolA (i.e. TolA13TR). Since entry of fd requires specific interaction of the phage minor coat gene 3 protein (G3p) with domain III of the TolA protein, the reduced fd sensitivity of the TR variants may be due to an allosteric effect of the TR-dependent variations in the length of domain II on the proper presentation of domain III. This hypothesis is further supported by the fact that domain II and III have previously been shown to physically interact [39].

From an ecological perspective, the different stresses mentioned above represent a number of opposing selective forces with regard to the optimal TR copy number in the tolA gene. Exposure to DOC, for example, is anticipated to be a strong selective force for increasing TR copy numbers, which would in turn attenuate tolerance to high osmolarity combined with low pH. We previously reported tolA alleles with TR copy numbers varying from 8 to 16 among 234 natural E. coli strains, with tolA13TR occurring in 66% of the strains, although the frequency distribution seemed to be different for some pathogens [25]. Our current findings suggest that 13 TRs may indeed provide an optimal tolerance to the different chemical stresses investigated in this study (DOC, SDS, high NaCl concentration, and high NaCl concentration at low pH).

The tolA allele carrying two stop codons in one of the TR units that was constructed in this work allowed us to demonstrate that TolA TR variations occur in a clonal wild-type population at a frequency of at least 6.9×10−5, thereby proving that TolA TRs can dynamically change on short evolutionary time scales. Moreover, these TR rearrangements were supported by RecA but suppressed by UvrD. Since RecA and UvrD are known to support and suppress homologous recombination [33], [45], respectively, these findings suggest that recombination is the primary mechanism affecting instability of the tolA TRs in E. coli. In contrast, although MutS has been shown to stimulate the rearrangement frequency of dimeric TRs [46], knocking-out mutS had no effect on contractions of the 15-mer TRs in our experiments. This observation is likely explained by the fact that DNA mismatch repair mainly targets nucleotide mismatches and insertion/deletion bulges of only 1–4 bp in length [47]. A similar conclusion was drawn from a previous study, which showed that mutS deficiency did not affect the mutation frequency at any of the 28 variable-number tandem repeats (VNTRs) with TR unit sizes >5 bp in E. coli O157:H7 [5].

Finally, it is noteworthy that rearrangement of the tolA allele typically resulted in 5-TR deletions. Moreover, neither the 12th nor 13th TR was ever shown to take part in contraction events. Possibly, these two repeats are essential for TolA function, and it was indeed suggested in a recent study that 31 residues at the C-terminal end of domain II of TolA (including the 12th and 13th TR) are required for binding the tetratricopeptide repeat domain of YbgF in the Tol-Pal complex, thereby controlling oligomeric state of YbgF [48].

In conclusion, this study demonstrates the pleiotropic phenotypic effects of TR copy number variations in the E. coli tolA gene, thereby revealing some possible selective forces able to drive TR rearrangements. Moreover, recombination-dependent TR rearrangements in tolA could be detected in clonal populations, further supporting a role of TR regions as hypermutable contingency loci that allow rapid and flexible adaptation to complex environmental conditions.

Supporting Information

Table S1.

Strains and plasmids used in this study.



We thank Dr. Heermann (Center for integrated Protein Science at the Ludwig-Maximilians-Universität München, Germany) for providing strains for rpsL counter-selection, Dr. Lloubes (Laboratoire d' Ingénierie des Systèmes Macromoleculaires UPR9027 at CNRS, France) for TolAIII antibody, Dr. Soumillion (University Catholique de Louvain, Belgium) for phage F-tet-DOG1, Dr. Delcour (KU Leuven, Belgium) for HRP-conjugated goat anti-rabbit PAbs, and Pauly A. and Vanoirbeek K. for providing technical assistance.

Author Contributions

Conceived and designed the experiments: KZ CM AA. Performed the experiments: KZ. Analyzed the data: KZ CM AA. Wrote the paper: KZ CM AA.


  1. 1. Bichara M, Wagner J, Lambert IB (2006) Mechanisms of tandem repeat instability in bacteria. Mutat Res 598: 144–163.
  2. 2. Gemayel R, Vinces MD, Legendre M, Verstrepen KJ (2010) Variable tandem repeats accelerate evolution of coding and regulatory sequences. Annu Rev Genet 44: 445–477.
  3. 3. Moxon R, Bayliss C, Hood D (2006) Bacterial contingency loci: the role of simple sequence DNA repeats in bacterial adaptation. Annu Rev Genet 40: 307–333.
  4. 4. Legendre M, Pochet N, Pak T, Verstrepen KJ (2007) Sequence-based estimation of minisatellite and microsatellite repeat variability. Genome Res 17: 1787–1796.
  5. 5. Vogler AJ, Keys C, Nemoto Y, Colman RE, Jay Z, et al. (2006) Effect of repeat copy number on variable-number tandem repeat mutations in Escherichia coli O157:H7. J Bacteriol 188: 4253–4263.
  6. 6. Cooley MB, Carychao D, Nguyen K, Whitehand L, Mandrell R (2010) Effects of environmental stress on stability of tandem repeats in Escherichia coli O157:H7. Appl Environ Microbiol 76: 3398–3400.
  7. 7. Dawid S, Barenkamp SJ, St Geme JW 3rd (1999) Variation in expression of the Haemophilus influenzae HMW adhesins: a prokaryotic system reminiscent of eukaryotes. Proc Natl Acad Sci U S A 96: 1077–1082.
  8. 8. Martin P, Makepeace K, Hill SA, Hood DW, Moxon ER (2005) Microsatellite instability regulates transcription factor binding and gene expression. Proc Natl Acad Sci U S A 102: 3800–3804.
  9. 9. van Ham SM, van Alphen L, Mooi FR, van Putten JP (1993) Phase variation of H. influenzae fimbriae: transcriptional control of two divergent genes through a variable combined promoter region. Cell 73: 1187–1196.
  10. 10. Vinces MD, Legendre M, Caldara M, Hagihara M, Verstrepen KJ (2009) Unstable tandem repeats in promoters confer transcriptional evolvability. Science 324: 1213–1216.
  11. 11. Deszo EL, Steenbergen SM, Freedberg DI, Vimr ER (2005) Escherichia coli K1 polysialic acid O-acetyltransferase gene, neuO, and the mechanism of capsule form variation involving a mobile contingency locus. Proc Natl Acad Sci U S A 102: 5564–5569.
  12. 12. Madoff LC, Michel JL, Gong EW, Kling DE, Kasper DL (1996) Group B streptococci escape host immunity by deletion of tandem repeat elements of the alpha C protein. Proc Natl Acad Sci U S A 93: 4131–4136.
  13. 13. Sheets AJ, St Geme 3rd JW (2011) Adhesive activity of the Haemophilus cryptic genospecies Cha autotransporter is modulated by variation in tandem peptide repeats. J Bacteriol 193: 329–339.
  14. 14. Rando OJ, Verstrepen KJ (2007) Timescales of Genetic and Epigenetic Inheritance. Cell 128: 655–668.
  15. 15. Ritz D, Lim J, Reynolds CM, Poole LB, Beckwith J (2001) Conversion of a peroxiredoxin into a disulfide reductase by a triplet repeat expansion. Science 294: 158–160.
  16. 16. Shaver AC, Sniegowski PD (2003) Spontaneously arising mutL mutators in evolving Escherichia coli populations are the result of changes in repeat length. J Bacteriol 185: 6076–6082.
  17. 17. Levengood-Freyermuth SK, Beyer WF, Webster RE (1991) TolA: a membrane protein involved in colicin uptake contains an extended helical region. Proc Natl Acad Sci U S A 88: 5939–5943.
  18. 18. Schendel SL, Click EM, Webster RE, Cramer WA (1997) The TolA protein interacts with colicin E1 differently than with other group A colicins. J Bacteriol 179: 3683–3690.
  19. 19. Gerding MA, Ogata Y, Pecora ND, Niki H, de Boer PA (2007) The trans-envelope Tol-Pal complex is part of the cell division machinery and required for proper outer-membrane invagination during cell constriction in E. coli. Mol Microbiol 63: 1008–1025.
  20. 20. Bouveret E, Rigal A, Lazdunski C, Bénédetti H (1998) Distinct regions of the colicin A translocation domain are involved in the interaction with TolA and TolB proteins upon import into Escherichia coli. Mol Microbiol 27: 143–157.
  21. 21. Raggett EM, Bainbridge G, Evans LJ, Cooper A, Lakey JH (1998) Discovery of critical Tol A-binding residues in the bactericidal toxin colicin N: a biophysical approach. Mol Microbiol 28: 1335–1343.
  22. 22. Riechmann L, Holliger P (1997) The C-terminal domain of TolA is the coreceptor for filamentous phage infection of E. coli. Cell 90: 351–360.
  23. 23. Lubkowski J, Hennecke F, Plückthun A, Wlodawer A (1999) Filamentous phage infection: crystal structure of g3p in complex with its coreceptor, the C-terminal domain of TolA. Structure 7: 711–722.
  24. 24. Levengood-Freyermuth SK, Click EM, Webster RE (1993) Role of the carboxyl-terminal domain of TolA in protein import and integrity of the outer membrane. J Bacteriol 175: 222–228.
  25. 25. Zhou K, Vanoirbeek K, Aertsen A, Michiels CW (2012) Variability of the tandem repeat region of the Escherichia coli tolA gene. Res Microbiol 163: 316–322.
  26. 26. Sambrook J, Fritsch EF, Maniatis T (1989) Molecular cloning : a laboratory manual, 3rd ed. New York: Cold Spring Harbor Laboratory Press.
  27. 27. Datsenko KA, Wanner BL (2000) One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc Natl Acad Sci U S A 97: 6640–6645.
  28. 28. Vinés ED, Marolda CL, Balachandran A, Valvano MA (2005) Defective O-antigen polymerization in tolA and pal mutants of Escherichia coli in response to extracytoplasmic stress. J Bacteriol 187: 3359–3368.
  29. 29. Benson G (1999) Tandem repeats finder: a program to analyze DNA sequences. Nucleic Acids Res 27: 573–580.
  30. 30. Heermann R, Zeppenfeld T, Jung K (2008) Simple generation of site-directed point mutations in the Escherichia coli chromosome using Red(R)/ET(R) Recombination. Microb Cell Fact 7: 14.
  31. 31. Hill TM, Sharma B, Valjavec-Gratian M, Smith J (1997) sfi-independent filamentation in Escherichia coli Is lexA dependent and requires DNA damage for induction. J Bacteriol 179: 1931–1939.
  32. 32. Wagner J, Nohmi T (2000) Escherichia coli DNA polymerase IV mutator activity: genetic requirements and mutational specificity. J Bacteriol 182: 4587–4595.
  33. 33. Bierne H, Seigneur M, Ehrlich SD, Michel B (1997) uvrD mutations enhance tandem repeat deletion in the Escherichia coli chromosome via SOS induction of the RecF recombination pathway. Mol Microbiol 26: 557–567.
  34. 34. Callewaert L, Aertsen A, Deckers D, Vanoirbeek KG, Vanderkelen L, et al. (2008) A new family of lysozyme inhibitors contributing to lysozyme tolerance in gram-negative bacteria. PLoS Pathog 4: e1000019.
  35. 35. Croes E, Gebruers K, Luyten N, Delcour JA, Courtin CM (2009) Immunoblot quantification of three classes of proteinaceous xylanase inhibitors in different wheat (Triticum aestivum) cultivars and milling fractions. J Agric Food Chem 57: 1029–1035.
  36. 36. Derouiche R, Benedetti H, Lazzaroni JC, Lazdunski C, Lloubes R (1995) Protein complex within E. coli inner membrane: TolA N terminal domain interacts with TolQ and TolR proteins. J Biol Chem 270: 11078–11084.
  37. 37. Luria SE, Delbrück M (1943) Mutations of bacteria from virus sensitivity to virus resistance. Genetics 28: 491–511.
  38. 38. Marvin DA, Hohn B (1969) Filamentous bacterial viruses. Bacteriol Rev 33: 172–209.
  39. 39. Karlsson F, Borrebaeck CA, Nilsson N, Malmborg-Hager AC (2003) The mechanism of bacterial infection by filamentous phages involves molecular interactions between TolA and phage protein 3 domains. J Bacteriol 185: 2628–2634.
  40. 40. Meury J, Devilliers G (1999) Impairment of cell division in tolA mutants of Escherichia coli at low and high medium osmolarities. Biol Cell 91: 67–75.
  41. 41. Merritt ME, Donaldson JR (2009) Effect of bile salts on the DNA and membrane integrity of enteric bacteria. J Med Microbiol 58: 1533–1541.
  42. 42. De Paepe M, Gaboriau-Routhiau V, Rainteau D, Rakotobe S, Taddei F, et al. (2011) Trade-off between bile resistance and nutritional competence drives Escherichia coli diversification in the mouse gut. PLoS Genet 7: e1002107.
  43. 43. Hernández SB, Cota I, Ducret A, Aussel L, Casadesús J (2012) Adaptation and preadaptation of Salmonella enterica to Bile. PLoS Genet 8: e1002459.
  44. 44. Begley M, Gahan CG, Hill C (2005) The interaction between bacteria and bile. FEMS Microbiol Rev 29: 625–651.
  45. 45. Hashem VI, Rosche WA, Sinden RR (2004) Genetic recombination destabilizes (CTG)n.(CAG)n repeats in E. coli. Mutat Res 554: 95–109.
  46. 46. Morel P, Reverdy C, Michel B, Ehrlich SD, Cassuto E (1998) The role of SOS and flap processing in microsatellite instability in Escherichia coli. Proc Natl Acad Sci U S A 95: 10003–10008.
  47. 47. Schofield MJ, Hsieh P (2003) DNA mismatch repair: molecular mechanisms and biological function. Annu Rev Microbiol 57: 579–608.
  48. 48. Krachler, A M, A Sharma, Cauldwell A, Papadakos G, Kleanthous C (2010) TolA modulates the oligomeric status of YbgF in the bacterial periplasm. J Mol Biol 403: 270–285.