The role of rpoS in Escherichia coli O157 manure-amended soil survival and distribution of allelic variations among bovine, food and clinical isolates
Correspondence: Eelco Franz, National Institute for Public Health and the Environment, Centre for Infectious Disease Control, Laboratory for Zoonoses and Environmental Microbiology, PO Box 1, 3720 AB, Bilthoven, The Netherlands. Tel.: +31 (0)30 2747063; fax: +31 (0) 30 2744434; e-mail: firstname.lastname@example.org
Although it is known that Escherichia coli O157 is capable of long-term soil survival, little is known about the mechanisms involved. This study investigated the role of the general stress response system RpoS in E. coli soil survival. The results showed that E. coli O157 isolates capable of long-term survival (longer than 200 days) in manure-amended soil were characterized by the absence of mutations in their rpoS gene. In contrast, the strains not capable of long-term survival all possessed mutations in their rpoS gene. In addition, the long-term surviving strains showed significantly higher levels of acid resistance in simulated gastric fluid (pH 2.5). Sequencing of the rpoS gene of bovine, food and clinical isolates revealed a skewed distribution of rpoS wild-type and mutant strains among the different sources. Bovine and food isolates had low numbers of mutants (< 1.4 and 6.9%, respectively), while a relatively high number of mutants was observed among human isolates (32.9%). The results indicate that a fully functional RpoS system is an advantage for survival in the manure-amended soil environment. Further deletion and complementation studies should provide more evidence on the role of RpoS in the long-term survival of E. coli O157 in diverse environments.
Shiga toxin-producing Escherichia coli (STEC) O157 is considered a serious pathogen due to its low infectious dose, severe clinical consequences, and the potential for food- and waterborne outbreaks (Caprioli et al., 2005). Its long-term survival in manure and soil can be considered a significant risk factor for the (re)contamination of cattle, food crops and ultimately human infection (Franz & van Bruggen, 2008; Fremaux et al., 2008). Escherichia coli O157 may respond to unfavourable conditions by expressing adaptive responses. Stationary-phase and almost any environmental stress that slows the growth rate of E. coli induce the RpoS-controlled general stress response (Battesti et al., 2011). Escherichia coli strains with attenuated RpoS levels have lower levels of resistance to external stress but have broader nutritional abilities and increased competitive abilities with low nutrient concentrations, and vice versa (King et al., 2004). The appearance of rpoS mutants seems to be driven by the increased ability of such mutants to scavenge for scarce nutrients (King et al., 2004; Ferenci, 2005). This RpoS regulatory trade-off between stress resistance and metabolic capacity provides a means of broadening the ecological and phenotypic properties which might especially be advantageous to E. coli as this bacterium generally experiences a biphasic lifestyle with a relatively constant and optimal host-associated phase and a fluctuating non-optimal host-independent phase (Van Elsas et al., 2011). Variation in rpoS alleles has also been observed among pathogenic E. coli strains (e.g. O157) and have been linked to variation in the level of stress resistance and metabolic capacity (Waterman & Small, 1996; Robey et al., 2001; Parker et al., 2012). However, very little is known about the role of RpoS in the long-term survival of E. coli O157 in the environment and where and under which conditions rpoS mutants arise is currently not known (Parker et al., 2012).
Recently, the variation in manure-amended soil survival capability among 18 E. coli O157 isolates was studied and a strong relationship between the individual metabolic capacity and long-term survival of the strains was observed (Franz et al., 2011). In particular, oxidative capacity on propionic acid, α-ketobutyric acid and α-hydroxybutyric acid was strongly correlated with enhanced survival. Recent gene expression studies showed that rpoS mutants of E. coli O157 demonstrated an impaired ability to oxidize these three fatty acids (Dong et al., 2009). Intrigued by this observation, the isolates used in the soil survival experiment (Franz et al., 2011) were screened for rpoS allelic variations. It was hypothesized that the conditions in manure-amended soil favour a functional RpoS system. Consequently, the manure-amended soil environment would be an unlikely source of rpoS mutants. As the bovine intestine forms the principal reservoir of E. coli O157 and humans can be considered a transient host with distinct conditions in the gastrointestinal tract, it was hypothesized that the human gut could provide a niche for the rise and selection of rpoS mutants. Therefore, the prevalence of rpoS allelic variations among a set of 187 E. coli O157 isolates of bovine, food and human origin (Franz et al., 2012) was determined.
Materials and methods
The detailed characteristics of the E. coli O157 strains used in the manure-amended soil survival study as well as the set of 187 strains (73 bovine, 29 food and 85 human clinical isolates) have been described in detail previously (Franz et al., 2011, 2012). Most of the strains were isolated and stored, and have no history of prolonged laboratory use.
The complete rpoS gene was amplified using the following primers: rpoS_−130F, 5′-CTTGCATTTTGAAATTCGTTAC-3′; and rpoS_+125R, 5′-GATGATGAACACATAGGATGC-3′ in a 50-μL PCR mixture containing 1 × PCR buffer (Invitrogen BV, Breda, the Netherlands), 2.5 mM MgCl2, 0.2 mM dNTPs, 0.2 μM of each primer, 1 U Taq DNA polymerase (Invitrogen BV) and 2 μL DNA template (± 20 ng). The following PCR programme was used: one cycle of 95 °C for 5 min; 35 cycles of 95 °C for 30 s, 56 °C for 30 s and 72 °C for 60 s; one cycle 72 °C for 10 min. The PCR product was treated with ExoSAP-IT (GE Healthcare, Diegem, the Netherlands) to remove unwanted deoxynucleotides and primers. The sequence of the generated PCR product was determined using the ABI Big Dye Terminator kit and an ABI 3730 DNA Analyzer (Applied Biosystems, Bleiswijk, the Netherlands). The PCR primers were used for sequencing as well two others: rpoS_−4F, 5′-CCTTATGAGTCAGAATACGC-3′; rpoS_773R, 5′-CTCTGCTTCATATCGTCATC-3′. The functioning of the RpoS general stress resistance system was determined phenotypically by growth on succinate minimal medium (Chiang et al., 2011). In addition, the acid resistance for all 18 isolates of the soil survival study was assessed. Briefly, simulated gastric fluid was made as described (Oliveira et al., 2011). Cells were cultured overnight in LBG medium (pH 7, 37 °C). Subsequently, 30 μL of culture was added to 30 mL of simulated gastric fluid which was adjusted to pH 2.5 with 1 M HCl. Cells were enumerated after 3 and 6 h of incubation at 37 °C by plating serial dilutions on trypticase soy agar (TSA) and overnight incubation at 37 °C.
Results and discussion
All 11 E. coli O157 strains earlier identified as short to medium–long survivors (i.e. population decline to the detection limit taking < 200 days) in manure-amended soil (Franz et al., 2011) possessed mutations within the rpoS gene, that is deletions, insertions and single nucleotide polymorphisms (SNPs; Table 1). In contrast, the seven E. coli O157 strains earlier identified as long-term survivors (i.e. population decline to the detection limit taking more than 200 days) in manure-amended soil (Franz et al., 2011) all showed absence of mutations in the rpoS gene. The seven strains showing long-term survival with absence of mutations in the rpoS gene had also been characterized before based on an impaired ability to oxidize l-rhamnose, l-glutamic acid and l-threonine and by an enhanced ability to oxidize propionic acid, α-ketobutyric acid, α-hydroxybutyric acid, methyl β-d-glucoside and l-arabinose (Franz et al., 2011). This is in complete agreement with gene expression studies with rpoS mutants of E. coli O157 showing that these cells have impaired expression regarding fatty acid oxidation (Dong & Schellhorn, 2009) and that these bacteria have decreased abilities to oxidize propionic acid, α-ketobutyric acid and α-hydroxybutyric acid but an increased ability to oxidize l-threonine (Dong et al., 2009). Recently it was shown that expression of the rpoS gene in E. coli O157 cells in sterile soil was 2.68-fold higher when compared with cells cultured in broth (Duffitt et al., 2011) and that RpoS plays a significant role in the cold stress response of E. coli O157 (Vidovic et al., 2011). Phenotypically, 10/11 short surviving strains with rpoS mutations showed growth on succinate minimal medium (demonstrating increased nutritional capability). In contrast, 6/7 long-term survivors showed absence of growth on succinate minimal medium. Clearly, the relationship between rpoS status and growth on succinate is not unambiguous, which also has been observed by others (Dong & Schellhorn, 2010). It is likely that some strains use alternative mechanisms to balance stress resistance and metabolic capacity. The acid resistance of the long-term surviving strains without mutations in rpoS was significantly higher than that of the short- to medium-term persisting strains with mutations in rpoS (96.6 vs. 63.5% survival, respectively after 6 h; Student's t-test, P = 0.0034; Table 1).
Table 1. Features of the Escherichia coli O157 strains used in the manure-amended soil survival study
|M629||Bovine||O157:H-||I||47||A||654, deletion 13 bp||+||+||59.4|
|M626||Ovine||O157:H7||I/II||67||A||488, insert 1 bp||+||+||72.5|
|M627||Horse||O157:H7||I/II||68||C||396, insert 12 bp||+||+||97.6|
|M631||Bovine||O157:H7||II||96||C||75, deletion 1 bp||+||+||58.5|
|M630||Human||O157:H-||I/II||119||C||97, deletion 4 bp||+||+||47.3|
The results of the current study suggest that E. coli O157 benefits from a functional RpoS system during its presence in manure-amended soil and that this environment is not likely to give rise to and select for rpoS mutants. A functional general stress response is probably needed in the manure-amended soil environment and nutrient availability is not the most limiting factor. This has also been shown for the plant pathogenic bacterium Pseudomonas putida (Ramos-González & Molin, 1998). However, deletion and complementation studies should provide more evidence for the role of RpoS in the survival of E. coli O157 in manure-amended soil. In addition, the conditions for survival in non-amended soil might be completely different and the role of RpoS should be considered accordingly.
Variation in rpoS alleles has been observed among E. coli O157 isolates and it remains unclear which environment gives rise to and selects for rpoS mutants (Waterman & Small, 1996; Parker et al., 2012). None of the E. coli O157 strains isolated from the environment (from feral pig, river water, cow and pasture soil) and linked to the 2006 spinach-associated outbreak in the United States showed mutations in the rpoS gene (Parker et al., 2012). In contrast, 3/3 clinical and 2/5 spinach isolates possessed mutations in the rpoS gene, produced lower levels of RpoS, showed decreased levels of rpoS-regulated genes, and showed impaired phenotypic resistance to low pH, osmotic stress and oxidative stress. Parker et al. (2012) suggested that bagged spinach could provide a niche for the rise and selection of rpoS mutants and that these mutants are merely maintained during passage through the human host. However, this suggestion is challenged by gene expression studies showing clear up-regulation of rpoS when associated with (injured) lettuce tissue, implying the need for a functional general stress response (Carey et al., 2009; Kyle et al., 2010; Fink et al., 2012).
As the bovine intestine forms the principal reservoir of E. coli O157 and humans can be considered a transient host with distinct gastrointestinal conditions, it was hypothesized that the human gut could provide a niche for the selection of rpoS mutants. Sequencing the structural part of the rpoS gene of 73 bovine, 29 food (23 meat, one lettuce and five endive isolates) revealed a skewed distribution of mutants among the different isolation sources. Bovine and food isolates had low numbers of mutants (0/73 and 2/29, respectively), while a relatively high number of mutants was observed among human isolates (28/85) (Table 2).
Table 2. Characteristics of the 85 clinical Escherichia coli O157 isolatesa
|H01||O157:H7||I/II||C||−55, deletion 5 bp|
|H02||O157:H–||II||No product||No product|
|H03||O157:H7||I/II||C||129, deletion 96 bp|
|H05||O157:H7||I||A||188, deletion 1 bp|
|H15||O157:H7||I||A||742, insert 2 bp|
|H17||O157:H7||II||C||396, insert 9 bp|
|H19||O157:H7||II||C||508, insert 1 bp|
|H20||O157:H7||I/II||C||928, deletion 1 bp|
|H21||O157:H7||I/II||–||248, deletion 654 bp|
|H34||O157:H7||I/II||A||259, insert 1 bp|
|H35||O157:H7||I/II||A||377, deletion 68 bp|
|H39||O157:H7||I/II||A||255, deletion 1 bp|
|H48||O157:H7||I/II||A||881, insert 1 bp|
|H49||O157:H7||I/II||C||94, insert 1 bp|
|H74||O157:H7||I/II||C||396, insert 1 bp|
|H78||O157:H7||I/II||C||641, deletion 1 bp|
A strong LSPA-6-specific distribution of rpoS(A543C) among the isolates was observed, with 100% of lineage I possessing rpoS(543A) whereas 100% of the lineage II strains had rpoS(543C). Lineage I/II were either rpoS(543A) or rpoS(543C): bovine strains, 44.8% rpoS(543A) and 55.2% rpoS(543C); food strains, 26.7% rpoS(543A) and 73.3% rpoS(543C); human clinical strains, 49.2% rpoS(543A) and 50.8% rpoS(543C). This is in agreement with earlier findings that lineage I/II is genetically more diverse than lineage I and II (Bono et al., 2012; Eppinger et al., 2012).
The results of the present study and published literature suggest that the bovine gastrointestinal tract, the environment (soil and manure) and food are less likely sources for the rise of and selection for E. coli O157 rpoS mutants. Apparently, these environments require a functional RpoS general stress resistance system over the need for increased nutrient scavenging abilities. Calves inoculated with equal numbers of wild-type enterohaemorrhagic E. coli and an rpoS mutant strain shed the rpoS mutant significantly less frequently than the wild-type, indicating an important role for RpoS and the glucose-repressed AR system in passage through the gastrointestinal tract of cattle (Price et al., 2000). The requirement for a functional rpoS system in the bovine gastrointestinal tract is further highlighted by the observation that bovine isolates are more resistant to adverse environmental conditions (including acid stress) than human isolates (Vanaja et al., 2010). Several studies report that RpoS negatively regulates the expression of locus of enterocyte effacement (LEE)-encoded virulence genes in E. coli O157 and that consequently rpoS mutants show higher expression of virulence genes (Dong & Schellhorn, 2010). The rpoS gene function was shown to be a disadvantage for E. coli during competitive colonization of the mouse large intestine (Krogfelt et al., 2000). Using a mouse model it was demonstrated that E. coli O157 uses sugars that are not used by commensal E. coli to colonize the intestine (Fabich et al., 2008). Fabich et al. (2008) suggested that commensal E. coli which successfully colonized the mouse intestine are at an competitive advantage over invading E. coli O157 due to a higher substrate affinity for the sugars that are used by both strains, which would force E. coli O157 to use less abundant nutrients. Subsequently, E. coli O157 gains advantage by simultaneously consuming several sugars that may be available because they are not consumed by the commensal intestinal microbiota (Fabich et al., 2008). This system could select for rpoS mutations as these mutants are characterized by increased nutrient scavenging abilities at the expense of stress-resistance (King et al., 2004). Further deletion and complementation studies ideally using in vivo systems (human and animal gut, and soil systems) should provide more insight into the role of RpoS in the adaptation of E. coli O157 to diverse environments.