E. coli strains (Table 1) were grown aerobically in Luria-Bertani (LB) medium 23 at 37°C with shaking (150 rpm). When required, antibiotics (ampicillin, 100 μg/ml or kanamycin, 100 μg/ml) were added to the medium. Cell density of broth cultures was monitored at OD₆₀₀ using a spectrophotometer (Beckman Coulter). The plasmids and primers used in this study are listed in Tables 1 and 2, respectively.

Acid challenges were conducted at pH 2.0 as described previously 10. Briefly, an overnight culture grown in LB (10 g Tryptone, 5 g yeast extract, 10 g NaCl) under appropriate antibiotic selection was diluted 1:10,000 into 50 ml of fresh LB medium and allowed to grow until early stationary phase (OD₆₀₀ between 1.2 and 1.4). Cells were diluted in fresh LB (1:10) and used to inoculate (1:100) 50 ml of acid challenge medium (LB adjusted to pH 2.0 using 6 N HCl, then autoclaved for 15 min) in a 250 ml flask. At specific time points, samples were removed and plated on LB agar following serial dilution in sterile phosphate buffered saline (PBS). The number of colony forming units (CFU) was determined after 24 h of incubation at 37°C. The limit of detection for this assay is 5 CFU/ml, based on colony growth on 2 duplicate plates each inoculated with 100 μl of undiluted sample. The growth of cells and acid challenges were performed at 37°C with shaking at 110 rpm.

The suicide vector, pCVD442, was used to generate a recA mutant and a dps recA double mutant in E. coli O157:H7 ATCC 43895 by homologous recombination 24. In brief, using primers shown in Table 2, the 2.7-kb PCR fragment alaS::mltB (ΔrecA), was amplified using the joint PCR method previously described 25. The fragment was cloned into XbaI and SphI-digested pCVD442 to form pKCJ0328. The constructed vector contains the RP4 origin of transfer (oriT) and is conjugally mobilized from donor cells containing the tra gene. The recipient strains, ATCC 43895 or FRIK 47992 (dps), and donor strain, E. coli SM10 λpir with the constructed vector, were grown separately to mid-log phase in LB and conjugated to generate recA and dps recA mutants strains, respectively 2426. The mutants were selected and confirmed as previously described 24.

Cells were grown 5 h in LB and then challenged at pH 2.0. Samples were removed after 0 min, 1, 2, 3, and 4 h of acid challenge and chromosomal DNA extracted using a genomic DNA isolation column (Qiagen) following the manufacturer's instructions. When testing for nicked-DNA, the isolated chromosomal DNA (1 μg) was subjected to Bal31 nuclease (0.2 units) digestion at 30°C for 30 min in a final volume of 20 μl and reactions were inactivated at 75°C for 10 min, and then chilled on ice. DNA was analyzed following electrophoresis in a 0.8% (w/v) agarose gel with ethidium bromide staining.

The dps gene was PCR amplified with primer pair kc0305 and kc0306. These primers contain an NdeI or XhoI restriction enzyme site, respectively, for cloning into pET21b, resulting in pKCJ0325. The DNA sequence of the cloned insert was confirmed by sequencing. C-terminal His6-tagged Dps recombinant protein was purified from E. coli BL21(DE3). To induce expression of Dps-His6 protein, isopropyl β-D-thiogalactoside (IPTG, Sigma-Aldrich) (1 mM) was added when the OD₆₀₀ of the culture reached 0.6, followed by a two-hour incubation. The cells were harvested by centrifugation (10 min at 4000 × g) and the pellet was stored at -70°C. The cell pellet was thawed for 15 min at room temperature and resuspended in 1 ml of lysis buffer (50 mM NaH₂PO₄, 300 mM NaCl, 10 mM imidazole, pH 8.0). Lysozyme was added to a final concentration of 1 mg/ml and incubated on ice for 30 min. Cells were lysed by sonication (300 watts; 6 × 10 sec with 10 sec pauses between pulses). Cellular debris was removed by centrifugation (10,000 × g, 4°C for 20 min) and the supernatant was decanted into a clean tube. Dps-His6 was isolated from the supernatant using Ni-NTA chromatography following the manufacturer's protocol (Qiagen). The final buffer was replaced with 50 mM Tris-Cl buffer (pH 7.0) containing 50 mM NaCl using PD-10 columns (Amersham Biosciences).

An in vitro DNA-binding assay was conducted as previously described 1727. pUC18 plasmid DNA (300 ng) was mixed with varying concentrations of Dps in 10 mM Tris (pH 6.8), 20 mM NaCl, 0.5 mM EDTA in a total volume of 20 μl and incubated 1 h at room temperature. The Dps and DNA were mixed at the following ratios (w/w): 3:1, 17:1, and 33:1. The Dps-DNA mixtures were extracted with phenol:chloroform (1:1), the DNA precipitated with ethanol, and analyzed in 1.0% agarose gels. The DNA damage assays were performed with 300 ng of pUC18 and 10 μg of Dps. Dps-DNA mixtures were incubated 1 h at room temperature to allow association and then acidified with 0.1 N HCl to pH 4.6, 3.6, 2.6, and 2.2. The acidified Dps-DNA mixtures were incubated for 2 h at 37°C and then 10 mM Tris (pH 8.5) was added. DNA was recovered by extraction with chloroform:isoamyl alcohol (24:1), followed by 10 min at 55°C in the presence of 2% SDS, precipitated with ethanol, and resolved by agarose gel electrophoresis. For experiments using heat inactivated Dps, 10 μg of Dps was heated to 75°C for 15 min.

CD spectra were recorded with an AVIV model 202SF CD spectrometer (Lakewood, NJ) at the Biophysics Instrumentation Facility (University of Wisconsin – Madison). Spectra were collected using 0.3 mg of Dps/ml in a solution of 10 mM Tris-Cl and 10 mM NaCl (pH 6.8) in a cuvette with a 1-mm-path length at 25°C. The signal was averaged for 7 s during wavelength scans. The percent of secondary structure was calculated by the method of Chen et al. 28.

Except where noted, data presented are representative of at least three independent trials. Images of ethidium bromide stained agarose gels were captured using a Kodak digital camera attached to the imaging system. The relative intensity of DNA staining in each lane was quantified using Kodak 1D Image Analysis software. Briefly, the lanes on each gel were divided into an arbitrary number of discrete steps, with intensity at each step being measured. Relative peak intensity data were exported to Excel (Microsoft, Redmond, WA) for further analysis and plotting. Averages, standard deviations, and student t-tests were performed using Excel.

To determine if acid stress results in chromosomal DNA damage, E. coli O157:H7 cells were exposed to acidic LB (pH 2.0) for up to 4 h, followed by extraction and examination of chromosomal DNA. At the 0 min sample point (15 min of exposure due to centrifugation step), a single, condensed band of chromosomal DNA was recovered from cells (Fig. 1). Chromosomal DNA extracted from cells that were not exposed to acidic LB also resulted in a single condensed band similar to results from the first sample point (data not shown). No significant differences were detected between cells incubated for 60 min or 120 min when compared to cells harvested at the 0-min sample point. However, with longer incubation times (180 and 240 min), the primary chromosomal DNA bands became broader, less intense, and exhibited tailing which is indicative of DNA degradation and fragmentation. The visual examination of the DNA patterns was corroborated by the densitometric measurements (Fig. 1B), where the peaks for cells incubated for 180 min and 240 min had broader bands with lower relative intensities than DNA extracted from cells incubated for shorter exposure times. Also, the decrease in the relative quantity of high-molecular weight DNA retained in the loading wells with increased time of exposure of cells to acid suggested that acid stress resulted in the deterioration of chromosomal integrity.

The nature of the DNA damage from cells exposed to acid was examined further using Bal31, which cleaves DNA at nicks, gaps, single-stranded regions or other lesions of duplex DNA. The presence of strand breaks was detected in chromosomal DNA from cells exposed to acid stress. Specifically, when chromosomal DNA from parent and dps mutant strains were compared, the DNA from the mutant strain showed a more pronounced degradation with Bal31 (Fig. 2), suggesting that chromosomal DNA from dps mutants contained more DNA damage than the parental counterpart.

Since chromosomal DNA from cells exposed to acid stress contained strand breaks, the role of RecA, which is involved in DNA repair, was examined. Using agarose gel patterns of DNA digestions with Bal31, the integrity of chromosomal DNA from parent, dps, recA, and dps recA strains that had been acid challenged were compared (Fig. 2). After exposing cells to acidic LB (pH 2.0) for two hours, the Bal31 cleavage of DNA from both recA and dps recA mutant strains was more extensive than either the parental or the dps strains, demonstrating that recA and dps recA strains had more DNA damage resulting from acid stress.

To investigate whether chromosomal DNA damage in recA and dps recA strains correlated with the survival of strains during acid stress, acid challenges were performed (Fig. 3). Since the limit of detection of the assay was 5 CFU/ml, this value was assigned to experimental conditions (recA and dps recA mutants at 90 min, and dps mutant at 240 min) where no survivors were detected. The rate of reduction in survival, as indicated by the slope of the best-fit line (not shown), was significantly higher in recA when compared to either the parental strain (p = 1.12 × 10^-3) or the dps mutant strain (p = 1.34 × 10^-5). No statistically significant difference was detected in survival during acid challenge between recA and dps recA strains. When all the time points were considered together, the rates of decline in survivability of the parent and dps strains did not differ significantly (p = 4.37 × 10^-1). However, there were differences in the number of survivors after one hour of acid challenge (p = 2.15 × 10^-3) and after 40, 60, and 120 min of acid challenge, with p < 0.05 at these time points. These findings link DNA damage and/or repair with a decrease in an organism's ability to survive acid stress.

In vitro experiments were conducted to demonstrate conclusively that the physical association of Dps with DNA provided DNA protection from acid damage. First, the binding of Dps to DNA with varying concentrations of Dps was examined to determine the Dps:DNA ratio that bound all plasmid DNA under the assay conditions employed. The results from gel mobility shift assays showed that Dps completely bound pUC18 DNA (300 ng) at a ratio of approximately 33:1 (w/w) (Fig. 4A). Using this ratio, purified Dps and plasmid DNA were incubated in solutions of varying pH. After incubation, protein was extracted using chloroform:isoamyl alcohol (24:1) in the presence of 2% SDS and the plasmid integrity was determined by the relative abundance of super-coiled, nicked, and linearized forms of the plasmid. Results showed that plasmid incubated with Dps remained in the supercoiled form at all challenge pH values, with little detectable nicking (Fig 4B). In contrast, when Dps was omitted from the assay, the plasmid was nicked and/or linearized at pH 3.6 and at pH 2.6. At pH 2.2, the plasmid DNA was degraded when Dps was absent.

The amount of DNA extracted from in vitro DNA assays was influenced by pH and the extraction method employed. When denaturing extraction [chloroform:isoamyl alcohol (24:1) with 2% SDS] was used, the amount of DNA recovered was comparable regardless of the challenge pH (Fig. 5A). In contrast, extraction using non-denaturing conditions [phenol:chloroform (1:1)] yielded less DNA when Dps and plasmid DNA were incubated at pH ≤ 3.6 (Fig. 5B), and the amount of extractable DNA decreased with lower pH. At pH 2.2, DNA was barely detectable. These data suggested that the binding of Dps to DNA was influenced and enhanced at lower pH or Dps interferes with DNA extraction in other ways.

Additional evidence that Dps-DNA complex formation may be affected by pH was gained from a negative control. When heat-denatured (75°C, 15 min) Dps was used as a control in experiments aimed at determining the role of Dps in plasmid DNA protection from acid damage, heat-denatured Dps did not bind to DNA at pH 7.0, as evidenced by the migration of pUC18 plasmid DNA in agarose gels (Fig. 6A). Similar results were observed with Dps-plasmid mixtures incubated at pH 4.6; however, the incubation of heat-denatured Dps at pH 3.6 or below restored Dps binding to DNA (Fig. 6A). This binding also protected DNA from acid damage (Fig. 6B). These results showed that in low-pH conditions, heat-denatured Dps regained its ability to bind and protect DNA from acid damage.

The previous results suggested that Dps-DNA complex formation or stability was influenced by pH. Therefore, the secondary structure of Dps in low pH conditions was examined using circular dichroism spectroscopy (Fig. 7). The spectra showed that at pH 2.2 and 3.6, Dps secondary structure contained an alpha-helix conformation 29, approximately 54% at pH 3.6 and 23% at pH 2.2. An unrecognized spectrum was generated at pH 7.0 that was likely due to Dps aggregation 30. Taken together, these three lines of evidence showed that pH influenced Dps-DNA interactions, at least between pH 3.6 and 2.2, and low pH appeared to enhance Dps association with DNA that contributed to its ability to physically protect and/or buffer DNA from acid damage.