H-NS controls metabolism and stress tolerance in Escherichia coli O157:H7 that influence mouse passage

Generation and confirmation of the hns::nptI mutant

A 695-bp fragment that includes the hns structural gene and upstream regulatory region was amplified by PCR from genomic DNA of E. coli O157:H7 ATCC 43895. The nucleotide sequence of the Eco RI-Bam HI fragment was determined and was identical to the sequenced gene in E. coli O157:H7 EDL933 [20]. Transconjugants resulting from the conjugation of pSC0063 from E. coli SM10 λpir to E. coli O157:H7 strain ATCC 43895 were kanamycin resistant, ampicillin sensitive, sucrose positive, and O157 positive. Confirmation of a double crossover in which wild-type hns was replaced with hns::nptI was confirmed by PCR (data not shown). PCR analysis of genomic DNA from the parental strain (ATCC 43895) with primers choi9917 and choi9918 generated a 375-bp fragment; whereas, genomic DNA from the hns::nptI mutant (FRIK47001P) resulted in a DNA fragment approximately 1.6-kb in length. The 1.6-kb fragment is in agreement with the projected size of the DNA fragment containing wild-type hns (375 bp) and the nptI gene (1.2 kb).

Growth and serological characteristics

The growth of the hns mutant (FRIK47001P) in LB at 37°C was slower and reached a lower maximum optical density than the parent strain or FRIK47001P containing pSC0061 (data not shown). The doubling times of the parent, complemented, and mutant strains were 27, 34, and 36 minutes, respectively. The maximum OD₆₀₀ nm achieved by the parent and complemented strains was 1.3 to 1.4 after 18 h of incubation while the hns mutant reached a maximum OD_{600 nm} = 1.0. Similar to the results in LB, the growth of the hns mutant in LB supplemented with mucin or in fecal slurries was slower and reached a lower final density than the parent strain. In addition, the colony size of the hns mutant on agar media was smaller than the parent and complemented strains. The presence of kanamycin in the growth media did not influence the growth of the mutant. These results are in agreement with results with hns mutants of non-pathogenic E. coli which had greater doubling times than the parent strains [11, 21]. One contributor to the longer doubling time is the increased σ^s level in the hns mutant, which is partially suppressed in hns rpoS double mutants [11]. The hns mutant tested negative for the H7 antigen and was non-motile in motility medium (data not shown); however, the O157 antigen was still detected. The absence of H7 is consistent with previous data showing that H-NS is involved in the control of flagella gene expression and that hns mutants are non-motile [22, 23].

Growth inhibition by pH, NaCl, and bile

In addition to slower growth, the hns mutant was more susceptible to inhibition by low pH, NaCl, bile (Table 1) and cefixime (data not shown) compared to the parent and complemented strains. Low temperature further slows the growth of hns mutants [24]. Thus, the growth of the mutant was further restricted when low temperature was combined with another inhibitor. The minimum growth pH of the hns mutant at the different temperatures was 4.5 to 5.0 which was 0.5 to 1.0 pH unit higher than the parent and complemented strains with the greatest difference observed at the lowest incubation temperature (16.5°C). Both the parent and complemented strains grew at pH 4.0 at each incubation temperature. Similar to the results with pH, the hns mutant grew in LB supplemented with a maximum of 4–6% NaCl compared to 6–7% for the parent and complemented strains. The test strains tolerated the highest concentration of NaCl when incubated at 25°C. There was one exception in which the hns mutant complemented with pSC0061 grew in a maximum of 4% NaCl at 16.5°C which was equivalent to the maximum level tolerated by the hns mutant. It is unclear why the complemented strain differed from the parent strain in this case where in all other growth conditions the results were the same. An unexpected finding was the ability of the E. coli O157:H7 to grow in LB containing bile compounds at 14–15%. The hns mutant was significantly more sensitive to bile salts and bovine bile preparations and grew in the presence of a maximum of 6.5% and 9.0% of these compounds, respectively. The growth of the strains in the presence of bile was further studied in LB containing 5% bile salts. Although the hns mutant grew, it had a greatly extended lag period (approximately 40 h) in comparison to the parent and complemented strains (<12 h) (Figure 1).

	Growth in LB
Growth conditions	ATCC 43895 (parent strain)	FRIK47001P (hns mutant)	FRIK47001P (pSC0061)
pHa
37°C	4.0	4.5	4.0
25°C	4.0	4.5	4.0
16.5°C	4.0	5.0	4.0
NaCl (%)b
37°C	6.0	5.0	6.0
25°C	7.0	6.0	7.0
16.5°C	6.0	4.0	4.0
Bile salts (%)c	14.0	6.5	14.0
Bile (%)d	15.0	9.0	15.0

^aMinimum pH at which growth occurred. LB adjusted to appropriate pH with 5 N HCl.

^bMaximum percent NaCl in LB at which growth occurred.

^cBile salts No. 3 (Difco). Maximum percent in LB that growth occurred at 37°C.

^dBovine bile (Sigma). Maximum percent in LB that growth occurred at 37°C.

Acid tolerance

The acid tolerance of the parent, hns mutant, and complemented strains was tested using both log- and stationary-phase cells. The survival of log-phase cells of the hns mutant was significantly greater (p < 0.05) than the parent strain when challenged in synthetic gastric fluid (pH 2.0) (Figure 2A). These findings are consistent with the derepression of σ^s (rpoS)-regulated genes in the absence of H-NS. Stationary phase increased the acid tolerance compared to log phase in all three strains and the numbers decreased by less than 1 log₁₀ CFU/ml after 3 h of acid challenge (Figure 2B). However, the parent and hns-complemented strains exhibited similar acid tolerance that was significantly greater (p < 0.05) than the hns mutant.

In contrast to the findings with acid challenges conducted in synthetic gastric fluid that is acidified with hydrochloric acid, there was no difference in the survival of the parent and hns mutant strains challenged in LB acidified with acetic or lactic acids (data not shown).

Glutamate decarboxylase activity

In an attempt to discern the reduced acid tolerance in stationary-phase cells of the hns mutant, glutamate decarboxylase (GAD) activity was measured (ΔpH/h/10⁸ CFU). The hns mutant had significantly less (p < 0.01) GAD activity than the parent strain and strain FRIK47001P complemented with pSC0061 (p < 0.001) when examining stationary-phase cells (data not shown). These findings provide one possible explanation for the reduced acid tolerance in comparison to the parent strain of the hns mutant in stationary phase.

Sole carbon and nitrogen sources

Gastrointestinal passage

Bacterial strains and culture conditions

Generation and confirmation of an hns::nptI mutant

A DNA fragment containing the hns structural gene and flanking regions was amplified from genomic DNA of E. coli O157:H7 ATCC 43895 by PCR using oligonucleotide primers (choi0005, 5'-TTTGGATCCAAAGCCTGGCTTGAAGAAGAGATG-3' and choi0006, 5'-TTTGAATTCTTTTGAATTCCTTACATTCCTGGC-3') containing an Eco RI (choi0005) and Bam HI (choi0006) site on the respective 5'-ends. After PCR amplification, the PCR fragment was digested with Eco RI and Bam H1 and ligated into pBR322, previously digested with the same enzymes, to produce pSC0061. The nucleotide sequence of the cloned fragment was determined (University of Wisconsin-Biotechnology Center). The suicide vector, pCVD442, was used to generate hns mutant in E. coli O157:H7 ATCC 43895 by homologous recombination. The hns gene in pSC0061 was inactivated by insertion of nptI encoding for aminoglycoside 3'-phosphotransferase (confers resistance to kanamycin) digested from pUC4K [44] using Pst I and gap filled with Klenow fragment. npt I was ligated with a partial Pst I digest of pSC0061 carrying hns which has a Pst I site present within its open reading frame (ORF). The construct (pSC0062) was digested with Eco R1 and Bam H1 to liberate the 1.9-kb hns::nptI fragment. The fragment was blunt-ended with Klenow and ligated with Sma I-digested pCVD442 [45] to produce pSC0063. The constructed vector was conjugally mobilized from E. coli SM10 λ pir [46] carrying the tra gene to recipient strain E. coli O157:H7 43895. Conjugation and mutant selection were conducted as described previously [47]. Transconjugants that were kanamycin resistant, ampicillin sensitive and tested positive for the O157 antigen by latex agglutination (Oxoid, Basingstoke, England) were selected for further study. The primers choi9917 (5'-ATGAGCGAAGCACTTAAAATTCTGAAC-3') and choi9918 (5'-TTCTTCCATTGCTTTTTTGATTACAGC-3') were used to amplify a portion of the hns ORF and confirm the presence of nptI in hns (1.6-kb fragment) (data not shown). FRIK 47001 (hns::npt I) was complemented by transforming FRIK 47001 with pSC0061.

Phenotype MicroArrays

The analysis of sole carbon and nitrogen source utilization was conducted with Phenotype MicroArrays™ (Biolog, Hayward, Calif.). This method is used to identify the function and global interactions of a gene on cellular metabolism [48, 49]. The parent E. coli O157:H7 strain ATCC 43895 and hns::npt I mutant (FRIK47001) were evaluated for their ability to use carbon and nitrogen sources using PM1, PM2, and a PM3 plates, respectively. Strains were grown on R2A agar (Biolog) overnight at 37°C and sufficient cells were transferred to inoculating fluid to equal the 85%T turbidity standard (Biolog). FRIK47001 was resuspended in inoculating fluid containing glucose rather than succinic acid when inoculating the PM3 plate because it was unable to use succinic acid as a carbon source. Each well was inoculated with 100 μl of the standardized cell suspension, incubated at 37°C for 48 h, and the results recorded.

Growth studies

Growth studies were conducted in LB adjusted to the appropriate pH or supplemented with NaCl, bile, bile salts, or cefixime using the parent, hns mutant, and hns mutant complemented with pSC0061. Cells were used to inoculate the adjusted LB medium at a final concentration of 10⁵ CFU/ml and incubated at 16.5, 25, and/or 37°C without shaking. The minimum pH for growth was determined in the pH range of 3.5 to 7.0 at 0.5 increments. The following were added to LB at the following ranges: NaCl, 0 to 10% (0.5% increments); bile salts no.3 (Difco), 0 to 16%; bovine bile (Sigma), 0 to 16%; porcine stomach mucin (Sigma), 2.5 mg/ml; bovine submaxillary gland mucin (Sigma), 2.5 mg/ml; and cefixime (Lederle), 0.2 to 0.8 μg/ml. Inoculated tubes were incubated at the designated temperature and observed daily for turbidity for up to 21 days. The growth of the strains in the presence of bile salts and mucin were studied in more detail. Growth in LB supplemented with 5% bile salts or 2.5 mg mucin/ml was monitored spectrophotometrically at 600 nm using a Bioscreen Analyzer at 37°C for 48 h (Labsystems, Helsinki, Finland). Plates were shaken for 30s prior to each optical density reading. The growth of strains was also monitored in fecal slurries (mouse feces diluted 1:10 in sterile deionized water). The fecal slurries were inoculated at a final concentration of approximately 10⁵ CFU/ml from overnight cultures grown in LB, and then incubated at 37°C. Samples were removed and the number of CFU/ml determined by plating on MacConkey sorbitol agar (MSA, Difco) supplemented with 2.5 μg potassium tellurite/ml (Fisher) and 0.05 μg cefixime/ml (Lederle Laboratories, Pearl River, NY) [50] and enumerating the number of sorbitol-negative colonies.

Acid challenge

Acid tolerance was assessed in synthetic gastric fluid adjusted to pH 1.5 and LB acidified with acetate or lactate (pH 4.4) as previously described [51]. Log (OD₆₀₀ = 0.5) and stationary (OD₆₀₀ = 1.1) phase cells were used to inoculate flasks containing 50 ml of synthetic gastric fluid or acidified LB to achieve a final concentration of ca. 10⁵ CFU/ml. Flasks were incubated at 37°C with shaking (150 rpm). Samples were removed periodically and then plated on tryptic soy agar (TSA, Difco) using a Model D Spiral Plater (Spiral Systems Inc., Cincinnati, OH). The number of colony forming units (CFU) was determined after 24 h of incubation at 37°C. The survival curves shown are the mean values from at least three trials.

GAD activity

The cells were examined for GAD activity as previously described [52, 53] with the following modifications. Strains were cultured overnight in LB at 37°C for 21 h. The cells from 1.0 ml were pelleted by centrifugation and washed with 0.9% NaCl. The cells were resuspended in 3.0 ml of GAD reagent (1 g of L-glutamic acid, 0.05 g of bromocresol green, 90 g of NaCl, and 3 ml of Triton X-100 per liter) and incubated for 3 h at 35°C. The color and pH change of the reaction mixture was recorded every 30 minutes. GAD activity was reported as the ΔpH/h/10⁸ cells. The mean values were calculated from at least three trials.

Mouse inoculation studies

The passage studies were conducted as previously described [32]. In brief, strains were grown in LB at 37°C to log and stationary phase. After harvesting and washing with PBS, the cells were resuspended in 10% sucrose or phosphate buffer (0.01 M, pH 7.4) to a concentration of 10⁴, 10³, or 10² CFU/ml and administrated to individually housed ICR mice (ca. 20 g, Harlan Sprague Dawley Inc., Madison, Wis). Three trials were conducted, and in each trial three mice were inoculated with each concentration of the test strain. Fecal samples were collected each day for 3 days after inoculation and tested for the presence or absence of E. coli O157:H7. Fecal samples were diluted (1:10) in modified EC broth supplemented with novobiocin (20 μg/ml) and incubated at 37°C with shaking (150 rpm) for 18–24 h. Following enrichment, 20 μl of Dynabeads^® anti-E. coli O157 (Dynal Biotech ASA, Oslo, Norway) was added to 0.5 ml of the enriched fecal culture and wash buffer (0.5 ml), incubated, and washed as described by the manufacturer. The resuspended beads were streaked on MacConkey sorbitol agar supplemented with potassium tellurite/ml and cefixime/ml [50]. The plates were incubated at 42°C for 24 h and examined for sorbitol-negative E. coli O157:H7 colonies. Suspect colonies were confirmed as O157 by agglutination (E. coli O157:H7 test kit, Oxoid). The minimum detection level of this procedure was approximately 100 CFU/g.

Statistical analyses

The data presented are the mean values from at least three trials. The acid tolerance of strains was compared using the inverse of slopes from linear regression plots of survivor curves. The data from acid challenges were analyzed for statistical differences by one-way analysis of variance. A z-test for was used to evaluate significant differences in the proportions of mice testing positive between inoculation strains [54]. All calculations were conducted using SigmaStat software (Jandel Scientific, San Rafael, Calif.).