Differential Expression of the Escherichia coli Autoaggregation Factor Antigen 43

Bacterial strains, plasmids, and growth conditions.

The strains and plasmids used in this study are listed in Table 1. An oxyR-null derivative of E. coli K-12 MG1655 (3) (MGJ1) was constructed by P1 transduction from E. coli N9716 (oxyR::Sp) (a kind gift from G. Storz). MS428 is a fim derivative of MG1655 (16), and MS528 is a fim flu derivative of MG1655 (16). The killing effects of H₂O₂ were assessed with both MS528 and E. coli reference strain BD1428. Ag43 expression in BD1428 is phase variable; Ag43 phase-on and phase-off derivatives are referred to as BD1512 and BD1511, respectively (7). Cells were grown at 37°C on solid or in liquid Luria-Bertani (LB) medium supplemented with the appropriate antibiotics unless otherwise stated. For induction of the araBAD promoter, 0.2% arabinose was used.

Recombinant DNA techniques.

The oxyR gene was amplified by PCR from MG1655 chromosomal DNA with primers KK71 (5-CGCGCTCGAGATAAGATGGAGGATGGATAATG) and KK72 (5′-CGCGAAGCTTGGAAGCCTATCGGGTAGCTG) containing XhoI and HindIII sites, respectively. The resulting PCR product was digested with XhoI and HindIII and inserted into the XhoI/HindIII site of plasmid pBAD/myc-HisA (8) to produce plasmid pMGJ1. In this construct, the expression of the oxyR gene is under the control of the arabinose-inducible araBAD promoter. Plasmids containing versions of the oxyR gene locked in the oxidized (pMAS141) and reduced (pMAS142) forms were constructed by using overlapping PCR with primers KK71 (upstream primer) and KK72 (downstream primer), together with primers 281 and 282 (locked reduced form; 5′-GGAAGATGGTCACTCTTTGCGCGATCAGG and 5′-CCTGATCGCGCAAAGAGTGACCATCTTCC) or primers 283 and 284 (locked oxidized form; 5′-GCTGGAAGATGGTCGCTGTTTGCGCGATC and 5′-GATCGCGCAAACAGCGACCATCTTCCAGC). These primers incorporate a specific amino acid change into the OxyR protein, C199S for OxyR^Red (OxyR^Red,C199S) and H198R for OxyR^Ox (OxyR^Ox,H198R). Plasmids containing each of the oxyR genes were introduced into oxyR mutant strain MGJ1.

DNA manipulations.

Plasmid DNA was isolated by using a QIAprep spin miniprep kit (Qiagen). Restriction endonucleases were used according to the manufacturer's specifications (Biolabs). PCRs were carried out as previously described (25). Amplified products were sequenced by using an ABI PRISM BigDye terminator cycle sequencing ready reaction kit (PE Applied Biosystems) to ensure the fidelity of the PCRs. Samples were run on a Perkin-Elmer ABI PRISM 310 genetic analyzer (PE Applied Biosystems) as described in the manufacturer's specifications.

Colony morphology.

Colony morphology was assayed by using a Carl Zeiss Axioplan epifluorescence microscope, and digital images were captured with a 12-bit cooled slow-scan charge-coupled device camera (KAF 1400 chip; Photometrics, Tucson, Ariz.) controlled by PMIS software (Photometrics).

Western immunoblotting.

Samples were subjected to sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis and transferred to polyvinylidene difluoride microporous membrane filters as described previously (25). Serum raised against the α subunit of Ag43 was used as the primary antibody, and peroxidase-conjugated anti-rabbit serum was used as the secondary antibody.

Hydrogen peroxide sensitivity.

Resistance to hydrogen peroxide was assayed by measurement of zones of inhibition. A suspension of cells from an overnight culture was added to 20 ml of overlay agar, and the mixture was poured into a petri dish and allowed to dry. Five microliters of 30% H₂O₂ was added to the center of the plate and allowed to adsorb. Zones of growth inhibition were measured after overnight growth at 37°C.

The ability of Ag43-mediated cell aggregation to protect against hydrogen peroxide killing was assessed according to the following protocol. Cells from overnight cultures were washed and resuspended in phosphate-buffered saline to a concentration of approximately 10⁹ cells per ml. For each strain, 100-μl aliquots were incubated at room temperature for 15 min in a microtiter plate. Then, 100 μl of an H₂O₂ solution (starting at 2.2 M and serially diluted to 0.03 M) was added, and the mixture was incubated for 15 min at room temperature. The cells were washed and then plated in a dilution series to determine the total number of surviving cells.

DTT-induced inhibition of Ag43 expression.

An overnight culture of E. coli strain MS428 was diluted and grown in 100 ml of LB medium with gentle shaking at 37°C. Aliquots (100 μl) of either 1 or 2 M dithiothreitol (DTT) were added every 15 min starting at an optical density at 600 nm (OD₆₀₀) of 0.1. At an OD₆₀₀ of 1.0, 10 ml of culture was harvested and washed in phosphate-buffered saline. The α subunit of Ag43 was detached by incubating the cells for 4 min at 60°C. Cells were removed by centrifugation, and the protein in the supernatant was precipitated with 75% acetone. Samples were subjected to Western blotting.

Biofilm assay.

Biofilm formation on plastic surfaces was monitored by using 96-well microtiter plates as previously described (17), the only difference being that arabinose was added for promoter induction.

Construction of locked OxyR variants.

Cysteine residue C199 is pivotal in determining the redox state of OxyR. Thus, out of six cysteine residues, only C199 seems to be essential for activity (19). It is solvent accessible and subject to redox changes (14, 19). The shift between the reduced and the oxidized forms of OxyR involves C199. A mutation, C199S, locks OxyR in the reduced form, OxyR^Red,C199S; conversely, a mutation involving position 198, H198R, causes locking of OxyR in the oxidized form, OxyR^Ox,H198R (18). Accordingly, we introduced these changes into the wild-type oxyR gene to create two locked forms: pMAS141 (C199S, locked reduced form) and pMAS142 (H198R, locked oxidized form). The tightly regulated arabinose-inducible araBAD promoter was used for expression of the genes (8). The wild-type oxyR gene, also controlled by the inducible araBAD promoter (pMGJ1), was used for comparison. All three constructs were made in the same manner to ensure that there were no differences in gene expression. As a host we constructed an oxyR derivative of the E. coli K-12 reference strain MG1655, i.e., MGJ1. The oxyR constructs were monitored for their responses to hydrogen peroxide (Fig. 1). As expected, the MGJ1 host harboring the locked reduced form of OxyR showed a response similar to that of reference strain MGJ1 and was unable to activate the oxidative stress response of the OxyR regulon. Conversely, MGJ1 harboring the locked oxidized form of OxyR was slightly more resistant than MGJ1 harboring the wild-type protein, probably due to priming of the oxidative stress response regulon.

To further characterize the oxyR derivatives, the protein profiles of the strains were investigated by SDS-polyacrylamide gel electrophoresis. As evidenced in Fig. 2, clear differences could be seen in the protein profiles, which could be ascribed to the oxyR status of the strains, i.e., ΔoxyR, wild-type oxyR, oxyR^Red,C199S, and oxyR^Ox,H198R. In addition, bands of equal intensities and migrating at the predicted molecular mass of OxyR (34 kDa) were present in all OxyR-expressing strains but not in the ΔoxyR mutant strain, suggesting that the mutations did not influence the level of OxyR. These observations led us to believe that our oxyR variants encoded type-specific functional gene products and could be used as tools to monitor Ag43 expression.

Only the reduced form of OxyR represses flu.

Western immunoblotting with Ag43 α-fragment-specific antiserum revealed a band migrating at the expected apparent molecular weight for the Ag43 α fragment in lysates from ΔoxyR and oxyR^Ox,H198R cells but not from oxyR^Red,C199S cells; a faint band was observed in a wild-type background (Fig. 2). One of the hallmarks of Ag43 expression is the ability to cause autoaggregation of cells with subsequent settling in liquid media. When the settling profiles of the four oxyR variants were monitored, it appeared that only hosts with oxyR^Red,C199S or the wild-type oxyR allele stayed in suspension, whereas the ΔoxyR and oxyR^Ox,H198R hosts flocculated (Fig. 3). These results suggest that only the reduced form of OxyR is able to repress the flu gene.

Another phenotypic hallmark associated with Ag43 expression is a characteristic frizzy colony morphology (form 1), contrasting the flat opaque colony morphology (form 3) of Ag43-negative cells (11, 16). When MGJ1 cells containing plasmid pMAS141 (oxyR^Red,C199S) or pMAS142 (oxyR^Ox,H198R) were grown on arabinose-containing solid media, distinct frizzy colony morphotypes were observed for the oxyR^Ox,H198R strain. On the contrary, only form 3 colony types were seen for the oxyR^Red,C199S strain, indicating that the flu gene was tightly repressed (Fig. 4A).

A simple reducing agent, DTT, is able to affect Ag43 expression.

Since the locked reduced form, OxyR^Red,C199S, was found to repress flu, we investigated whether a reducing environment would cause the repression of flu by driving the wild-type OxyR pool into a more reduced state. For this purpose, we used a Δfim strain, MS428. Since large doses of DTT inhibit cell growth, we modified our experimental conditions to not be growth inhibitory by pulse-wise addition of DTT over several generations (Fig. 5). The DTT-treated cultures and an untreated control were subsequently analyzed by Western blotting (Fig. 5). The data suggest that the addition of DTT reduces the expression of Ag43, likely by driving the cellular pool of OxyR into a more reduced state. These data corroborate the data obtained with the mutant form of OxyR and support the notion that OxyR control of flu is redox dependent.

Ag43-mediated autoaggregation protects against oxidizing agents.

OxyR essentially monitors the redox status of the cell and, together with its associated regulon, constitutes the primary response or defense of the cell against H₂O₂-induced oxidative stress. The above results suggested a link between the oxidative status of OxyR and Ag43 expression. This suggestion prompted us to investigate more directly the effects of Ag43 expression and associated cell aggregation on the survival of cells exposed to H₂O₂. When Ag43-expressing and non-Ag43-expressing variants of two different E. coli K-12 strains were exposed to lethal concentrations of H₂O₂ and subsequently plated to determine viable counts, clear differences in survival (approximately 10-fold) were observed between the two host strain sets (Fig. 4C). The real difference in survival was probably larger because the colony counts of Ag43-expressing cells were underestimated due to the fact that Ag43-positive cells aggregate and thus are not plated singly (Fig. 4B).

Biofilm-forming ability correlates with the redox status of OxyR.

By virtue of its excellent cell-aggregating properties, Ag43 expression enhances bacterial biofilm formation (6, 16, 17). In keeping with the notion that the redox status of OxyR determines its ability to repress Ag43 expression, we investigated the biofilm-forming ability of MGJ1 cells harboring different versions of OxyR. From the results, it became obvious that OxyR^Red,C199S hosts were significantly impeded in biofilm formation compared to their OxyR^Ox,H198R counterparts (Fig. 6). Clearly, these results should be attributed to the fact that in OxyR^Red,C199S cells, Ag43 expression was virtually abolished (Fig. 2). It appears that Ag43-assisted biofilm formation depends on cellular redox status.