Research article | Open | Published:
The uncertain consequences of transferring bacterial strains between laboratories - rpoSinstability as an example
BMC Microbiologyvolume 11, Article number: 248 (2011)
Microbiological studies frequently involve exchanges of strains between laboratories and/or stock centers. The integrity of exchanged strains is vital for archival reasons and to ensure reproducible experimental results. For at least 50 years, one of the most common means of shipping bacteria was by inoculating bacterial samples in agar stabs. Long-term cultures in stabs exhibit genetic instabilities and one common instability is in rpoS. The sigma factor RpoS accumulates in response to several stresses and in the stationary phase. One consequence of RpoS accumulation is the competition with the vegetative sigma factor σ70. Under nutrient limiting conditions mutations in rpoS or in genes that regulate its expression tend to accumulate. Here, we investigate whether short-term storage and mailing of cultures in stabs results in genetic heterogeneity.
We found that samples of the E. coli K-12 strain MC4100TF exchanged on three separate occasions by mail between our laboratories became heterogeneous. Reconstruction studies indicated that LB-stabs exhibited mutations previously found in GASP studies in stationary phase LB broth. At least 40% of reconstructed stocks and an equivalent proportion of actually mailed stock contained these mutations. Mutants with low RpoS levels emerged within 7 days of incubation in the stabs. Sequence analysis of ten of these segregants revealed that they harboured each of three different rpoS mutations. These mutants displayed the classical phenotypes of bacteria lacking rpoS. The genetic stability of MC4100TF was also tested in filter disks embedded in glycerol. Under these conditions, GASP mutants emerge only after a 3-week period. We also confirm that the intrinsic high RpoS level in MC4100TF is mainly due to the presence of an IS1 insertion in rssB.
Given that many E. coli strains contain high RpoS levels similar to MC4100TF, the integrity of such strains during transfers and storage is questionable. Variations in important collections may be due to storage-transfer related issues. These results raise important questions on the integrity of bacterial archives and transferred strains, explain variation like in the ECOR collection between laboratories and indicate a need for the development of better methods of strain transfer.
Well-resourced culture collections distribute bacteria mostly as freeze-dried ampoules [1, 2]. On the other hand, most research labs generally do not exchange lyophilized cultures and over the past 50 years a good proportion of bacterial exchanges were either in agar stabs or on impregnated glycerolized discs, as also used by the Coli Genetic Stock Center (CGSC). Generally, comparison of storage and shipping conditions test for viability and all of the above methods work well in this regard for Escherichia coli. Recently however, we became concerned about heterogeneity arising during storage and exchange of cultures for two reasons. Firstly, our recent studies with the ECOR collection  indicated a number of phenotypes had changed from those reported earlier (unpublished results). Others have also noted discrepancies in results with the ECOR collection between laboratories . Secondly, in recently exchanged stock cultures of E. coli K-12 between the Ferenci and Spira laboratories, we noted heterogeneities in some of the phenotypes we routinely assay. In this communication, we investigated the source of this heterogeneity and the role of storage conditions during shippage.
The instability of cultures and possible heterogeneities have been noted in several settings. Bacteria in long term stab cultures were found to change in a number of respects [5–8]. Extended incubation over several days in rich media also gives rise to population heterogeneity . In both cases, one of the targets of change was the rpoS gene. The sigma factor RpoS is the master regulator of the general stress response in E. coli . RpoS coordinates the transcription of genes associated with the protection of bacteria against different types of stress, such as high osmolarity, oxygen free radicals, low temperature and others [10, 11]. Bacteria that lack RpoS are more sensitive to environmental stresses, thus though rpoS is not an essential gene, its presence strongly increases bacteria survival rates in stressful environments.
RpoS levels are also shifted up under nutritional stress, namely carbon and phosphate starvation . In stationary phase or in nutrient-limited chemostats, the accumulation of RpoS in the cytosol reduces the expression of growth-related genes due to the competition between RpoS and the vegetative sigma factor σ70 for a limited amount of RNA polymerase core units . This characterizes a trade-off in which the bacterium sacrifices growth in favour of expressing protection-related genes. Under prolonged starvation periods a genetic adjustment follows when mutations in rpoS or in genes that control rpoS expression occur, resetting the SPANC (Self Preservation and Nutritional Competence) balance . The rpoS gene is highly polymorphic and many different alleles are found in both natural isolates and laboratory strains of E. coli [15–18]. This strong variation is expected given the pivotal role of RpoS in determining the SPANC balance  and is central to the instabilities we observe in mailed cultures.
The strain we exchanged was a derivative of MC4100, a widely used E. coli strain spread in many laboratories around the world. MC4100 stored at Ferenci's laboratory in Australia  was shown to express high levels of both RpoS and ppGpp [17, 20]. This version of MC4100 (hereafter called MC4100TF) efficiently exhibits protection-related phenotypes, such as resistance to stresses and glycogen production but is less competent in metabolising alternative carbon sources. It also tends to accumulate mutations in rpoS following 2-3 days of growth in a chemostat under carbon or phosphate limitation [17, 18]. It has been shown that a pair of point mutations at the N-terminus of the ppGpp-hydrolase SpoT is responsible for the high levels of ppGpp displayed by MC4100TF . Because ppGpp has a positive effect on RpoS , the high level of ppGpp partially explains the strong RpoS-related phenotypes in MC4100TF. In addition, genome sequencing of this strain revealed the presence of an IS1 insertion in the rssB locus . RssB acts as a chaperone that presents RpoS to the protease ClpXP, enhancing RpoS proteolyis . Thus, it was postulated that disruption of rssB contributes to the high-RpoS level in this strain, but no direct evidence has been presented.
Here we show that MC4100 readily accumulates mutations that suppress the high-RpoS phenotype when incubated in LB-stabs either during shipping or under laboratory conditions. Ten of these segregants were analysed and shown to carry null mutations in the rpoS ORF. It is also demonstrated that the IS1 insertion in rssB is the main factor that upregulates rpoS in MC4100.
Results and Discussion
Segregation of rpoSmutants in LB stabs
LB stabs of MC4100TF have been sent from Ferenci's laboratory in Australia to Spira's laboratory in Brazil by air mail on three different occasions. Upon arrival bacteria were streaked on LB agar and isolated colonies were checked for their RpoS status by iodine staining (glycogen accumulation is enhanced by RpoS ). MC4100TF stains darkly with iodine, but many colonies from these shipments displayed heterogeneity in iodine staining; this generally means variations in RpoS levels [17, 18]. The third shipment consisted of two LB stabs, one containing MC4100TF and the other one strain BW2952 (MC4100TF carrying a mal::lacZ fusion, but otherwise identical to MC4100TF). Bacteria were removed from each stab, suspended in 0.9% NaCl, diluted and plated on LB agar and stained with iodine. The proportion of low-staining colonies in these stabs was between 29% and 61%.
This prompted us to ask whether the shipping conditions to which the bacteria were exposed during the transcontinental flights selected mutations that caused the loss of the high-staining phenotype. To mimic the conditions during transport, a single fresh colony of MC4100TF was inoculated into an LB stab, and incubated at room temperature for 7 days. Following the incubation period bacteria were streaked on minimal medium plates supplemented with the alkaline phosphatase (AP) substrate X-P (TGP + X-P); AP expression is inversely correlated with RpoS level . Several colonies were light blue, but others showed a more intense blue colour, indicating a low-RpoS status (Figure 1A). Ten of these low-RpoS segregants were isolated and further analysed.
The segregants were tested for RpoS-dependent phenotypes (iodine staining and AP basal activity), for RpoS concentration by immunoblotting and for the presence of the IS1 insertion in rssB (MC4100TF is rssB::IS1). Amplification of rssB by PCR showed that all segregants retained the IS1 insertion and were therefore, RssB deficient. Despite that, all segregants stained lightly with iodine and showed a strong blue colour on TGP+X-P plates, suggesting that RpoS is very low or lacking in these strains (Figures 1B and 1C). A western-blot analysis revealed that with the exception of segregant number 6, a band corresponding to RpoS could not be detected in the nine other strains, suggesting that they carry null mutations in rpoS (Figure 1D).
To identify the mutations present in the 10 low-RpoS segregants, the rpoS ORF of each strain was sequenced. The results are summarised in Table 1. Six strains (nos. 1, 2, 5, 8, 9, 10) carry an adenine deletion at position 668 of rpoS ORF, which results in a frameshift and the formation of premature stop codons. Segregants 3, 4 and 7 have a TAAAG deletion (Δ515-519), which also causes a frameshift. Finally, segregant 6 carries an I128N substitution in the RpoS protein. This strain displayed high levels of RpoS (Figure 2C), but behaved as an rpoS null mutant, suggesting that RpoS activity was severely undermined by the I128N mutation. Residue 128 is located in region 2.2 of the RpoS protein. The exact function of region 2.2 is unknown, but a tentative tertiary structure of this region showed that it is formed by a helix whose polar surface constitutes one of the primary interfaces with RNA polymerase . Replacement of a hydrophobic by a polar amino acid at this position is likely to impair RpoS interaction with the core RNA polymerase, strongly inhibiting the formation of EσS holoenzyme and consequently the transcription of RpoS-dependent genes, such as glgS, involved in glycogen synthesis . As predicted by the trade-off hypothesis, once RpoS loses the ability to compete with σ70 for the binding to core RNA polymerase, the expression of σ70-dependent genes, such as phoA would increase, explaining the high level of AP showed by this mutant [13, 17, 25].
To further measure the frequency of emergence of rpoS mutations in LB stabs, a set of 15 stabs were inoculated each with a single MC4100TF fresh colony. The stabs were incubated for several weeks at room temperature and every week three stabs were tested for the rpoS status in each population. Each time bacteria were scraped off two different stabs, resuspended in saline, serially diluted and plated on LB agar. Bacteria from a third stab were streaked directly onto an LB plate for a qualitative analysis of the rpoS status. The colonies were then stained with iodine. Figure 2 shows the evolution of rpoS segregation in the stabs. At day 1, all tested bacteria were rpoS+, but by day 7 onwards, the presence of many low-RpoS colonies became apparent both in the quantitative (CFU count) and qualitative (streaks) plates. The exact proportion of these mutants varied from week to week, but was never lower than 40%.
A common and inexpensive alternative to LB-stabs is a bacterial suspension in filter disks in the presence of glycerol. To test this transporting method, a culture of MC4100TF was resuspended in 15% glycerol (v/v) and 0.1 ml of the suspension was applied onto a filter disk, which was placed in a small plastic bag and sealed. Glycerol filter disks were prepared along with the stabs reported in Figure 2 and stored at room temperature. Every week a pair of disks was removed from their plastic bags suspended in a small volume of saline and streaked on LB agar. Until day 21 all colonies recovered from the filter disks displayed a high-RpoS phenotype (stained dark brown with iodine). From day 31 onward a significant proportion (approx. 50%) of the bacteria recovered from the filter disks were low-RpoS. Furthermore, there was an increasing reduction in the number of colonies recovered every week, possibly due to prolonged starvation and dehydration of the filter disks (despite the sealing of the plastic bags). It is clear, though, that the glycerol filter disks preserved the genetic integrity of the bacteria for a longer period of time than the LB-stabs. Therefore, the use of glycerol filter disks for bacterial shipment is preferable.
The data presented here indicate that the use of LB-stabs for the exchange of bacteria between laboratories is undermined by genetic instability. Alternative storage and shipment forms, such as freeze-drying, glycerol filter disks or dry ice must be considered. Some of them are costly (shipment of glycerol stocks in dry ice) or dependent on specific equipments (lyophiliser) and none is free of drawbacks. As a matter of fact, induction of mutations during the freeze-drying process has been reported [26, 27]. Glycerol filter disks provide an inexpensive and easy alternative for bacterial shipping. Since the filters lack essential nutrients we expect very little or no bacterial growth and hence a significant reduction in mutant segregation.
Ever since the pioneer work of the Kolter group , several papers have reported the occurrence of rpoS mutations that confer selective advantage in stationary phase (the GASP phenotype) [8, 9, 29]. Accordingly, sequence variation of rpoS in E. coli natural isolates is extensively well documented [3, 3, 16, 30–32]. Even in archival cultures, a high proportion of rpoS mutants has been observed . This is not surprising because genomic rearrangements in stab cultures stored for long periods of time are common [4–6, 33], implying that long-term storage in stabs produces an environment that selects for a variety of mutations. Indeed, discrepancies among archival strain collections, like the ones found in the ECOR collection  and in bacterial strains used in compendial microbiological tests  are not uncommon.
To the best of our knowledge, this is the first report on rapid evolution in LB-stabs. This has implications not only for the storage of bacteria in the laboratory, which is less significant today because most bacteria collections are kept at -70°C freezers, but mainly to bacterial exchange by the scientific community. We demonstrated the emergence of mixed populations of rpoS+ and rpoS attenuated bacteria in LB-stabs of MC4100TF (a widely used E. coli strain) even after one-week incubation. This strain exhibits high levels of RpoS. High-RpoS strains tend to accumulate mutations in rpoS in order to reset the SPANC balance, i.e., to eliminate the competition between σ70 and RpoS enhancing the transcription of growth-related genes. However, RpoS loss is not restricted only to MC4100TF as other E. coli strains, even some with not particularly high RpoS (such as MG1655) have shown to accumulate mutations in rpoS under nutrient limiting conditions [3, 17, 18]. Although in the present study, the only tested variation was regarding rpoS, it is clear that other genes, such as lrp (a GASP allele of lrp has been isolated under prolonged starvation ) may also be affected during short-term incubation in LB-stabs, and this caveat should be taken into account when posting bacteria via air mail. It should also be noted that evolution in LB stabs are likely to occurr in other bacteria species, even in the ones regarded as more stable (such as Gram positive bacteria). This possibility can be empirically tested in the future.
The relation between RssB and RpoS in MC4100 derivatives
Sequence analysis of MC4100TF showed that it carries an IS1 insertion at the 5'-end of the rssB ORF , whose product facilitates the proteolysis of RpoS [22, 36]. Disruption of rssB is likely to contribute to the intrinsic high level of RpoS in MC4100TF because stocks of MC4100 maintained in other laboratories around the world do not carry the IS1 insertion in rssB and do not exhibit high levels of RpoS , but a direct evidence is still lacking. Furthermore, none of the tested segregants have reverted the rssB mutation, though a rssB+ allele would supposedly diminish the level of RpoS in this strain.
Therefore, to test if the IS1 insertion in rssB is related to the intrinsic high-RpoS level in MC4100TF, a series of experiments was conducted. The rssAB+ operon was cloned in a low-copy plasmid (pBS28) and transformed into MC4100TF to complement the rssB::IS1 mutation. In parallel, the rssB gene of MC4100BS (a stock of MC4100 maintained in our laboratory, which is low-RpoS and rssB+ [18, 19]) was knocked-out by the introduction of a KmR cassete. RpoS-dependent phenotypes, such as glycogen accumulation, alkaline phosphatase activity and growth on acetate as the sole carbon source (inhibited by high RpoS)  were assayed in strains MC4100TF, MC4100BS, MC4100BS rssB::KmR and MC4100TF carrying pBS28. Figure 3A shows that MC4100TF and MC4100BS rssB::KmR stained dark with iodine, showed low AP activity and poor growth on acetate, as expected for high-RpoS strains. Conversely, MC4100BS and MC4100TF carrying pBS28 accumulated less glycogen, more AP and grew stronger on acetate, all characteristics of low-RpoS strains. An immunoblot with an anti-RpoS serum confirmed that both MC4100TF and MC4100BS rssB::KmR displayed high levels of RpoS when compared to MC4100BS (rssB+) (Figure 3B). Altogether, these results demonstrate that the intrinsic high level of RpoS in strain MC4100TF is indeed caused by the IS1 insertion in rssB. There is a priori no reason why the rssB+ genotype would not be restored in MC4100TF, but we did not detect it in any of the tested low-RpoS segregants. On top of its role in RpoS proteolysis rssB helps modulating polyadenylation and degradation of specific mRNAs, and its overexpression is toxic for the cell . Thus, it is possible that under the conditions tested (incubation in LB stabs), mutations in rpoS would be less costly than reversion of the rssB::IS1 mutation. Indeed, many evolution experiments in chemostats with MC4100TF have been carried out, but reversion of the rssB::IS1 mutation has not been observed in any of them [38, 39].
Another potential input that might be upregulating rpoS in MC4100TF is ppGpp [20, 21]. ppGpp synthesis and degradation are driven by the products of the relA and spoT genes [40, 41]. MC4100TF carries two mutations in spoT (a H255Y substitution and a +QD insertion between residues 82 and 83) . When transferred to another E. coli strain, this spoT allele increased both ppGpp and RpoS levels. However, high levels of ppGpp and the same spoT mutations are present in MC4100BS as well . Therefore, ppGpp cannot explain the high RpoS level in MC4100TF, which is mainly due to the knock-out of rssB.
Mutations in the rpoS locus or in genes that regulate rpoS expression are powerful drivers of genetic instability or adaptability both in the laboratory and in nature. Growth and storage of bacteria in LB-stabs for short periods, such as the time it takes to mail a letter between different continents, is sufficient for the accumulation of rpoS mutations in high proportions. Mutations that inactivate or attenuate RpoS confer on the bacteria the GASP phenotype, explaining why they are so common across the species E. coli. A better alternative for the shipment of bacterial strains is the use of glycerol filter disks, in which a small volume of a bacteria culture resuspended in 15% glycerol is applied to a filter disk in a sealed plastic bag. Finally, of the many inputs that regulate rpoS, it was demonstrated that the high level of RpoS in strain MC4100TF is mainly due to an IS1 insertion in rssB.
Bacterial strains, plasmids and media
The strains used in this study were MC4100 (F- araD139 (argF-lac)U169 rpsL150 deoC1 relA1 thiA ptsF25 flbB5301 rbsR) stored in TF and BS laboratories; KM32 (lac Δ(recC ptr recB recD)::Ptac-gam-red cat) that carries a chromosomal copy of the λ-Red recombination system  and DH10B (F- mcrA Δ(mrr-hsdRMS-mcrBC 80dlacZΔM15 ΔlacX74 endA1 recA1 deoR (ara leu) 7697 araD39 galU galK nupG rpsL), used as recipent for plasmid transformation.
Plasmid pUC4K is a pUC19 derivative that carries a KmR cassete . pGEM T-easy is a cloning vector (Promega). pWKS130 is a low-copy cloning vector . pBS23 is a pGEM T-easy derivative that carries rssB+. pBS25 is as pBS23 except that a KmR cassete (from pUC4K) was inserted into rssB. pBS28 is a pWKS130 derivative that carries the rssAB+ operon.
TGP  plates contained 0.2% glucose, 1 mM KH2PO4 and 40 μg/ml X-P. LB plates and stabs were as described . Cells were grown overnight in either LB broth or in liquid T-salts supplemented with 0.2% glucose and 1 mM KH2PO4 at 37°C.
Bacterial storage and sampling
LB-stabs were innoculated with a single colony and immediately sealed by screwing down the tube lid. Following incubation at room temperature for different time lengths, bacteria samples were removed from the stabs either with a sterile glass rod (and subsequently streaked on a plate) or by scraping off the upper layer of the stab with a sterile metal stick.
Bacteria were then transferred to a microtube filled with 1 ml 0.9% NaCl and the turbidity of the sample was measured in a spectrophotometer. Bacteria were further diluted in 0.9% NaCl (usually 106 fold) and 0.1-0.2 ml were plated onto LB or TGP plates in duplicates. Glycerol filter disks were prepared by suspending a fresh colony in 100 μl 15% glycerol, A filter disk embedded with the bacteria suspension was sealed in a plastic bag. At appropriate time intervals, the plastic bag was opened and the disk transferred to a microtube filled with 200 μl 0.9% NaCl. 20 μl of this suspension was applied to the surface of a LB plate and streaked.
Determination of RpoS status
The level of RpoS was qualitatively assessed by staining intracellular glycogen with an iodine solution as described  or by visualising AP basal level with the chromogenic substrate X-P (5-bromo-4-chloro-3-indolyl-phosphate) . For the iodine staining, patches of bacteria or diluted samples were grown overnight on LB plates, stored at 4°C for 24 h and then flooded with iodine. The intensity of the brown colour varies according to glycogen concentration in the cell and indirectly reveals the level of RpoS [17, 18]. rpoS+ strains stain brown to dark brown.
Western-blot of RpoS
Western-blot analyses were performed essentially as described . Briefly, 2 × 109 bacteria grown overnight in LB-broth were resuspended in 200 μl application buffer (0.5 M Tris/HCl, 2% SDS, 5% 2-mercaptoethanol, 10%, v/v, glycerol and 0.01% bromophenol blue) and boiled for 5 min. Proteins were resolved in a 12.5% denaturing polyacrylamide gel and transferred to a nitrocelullose membrane (GE HealthCare) by capillary action. Following blocking with 5% skim milk, the membrane was incubated with 2, 000-fold diluted monoclonal anti-RpoS antibodies (Santa Cruz) and 20, 000-fold diluted peroxidase conjugated anti-mouse IgG (Pierce). The Super Signal West Pico kit (Pierce) was used to detect the RpoS bands as recommended by the manufacturer and the membrane was exposed to X-ray films.
Knock-out of rssB
A KmR cassete was inserted into rssB ORF by homologous recombination using the λ-Red system as described . The rssB gene was PCR amplified from E. coli chromosome with primers rssB94F (5'-CGCACCAACATTTGACCAG) and rssB1368R (5'-GTATCGCATCCCAGTATATCAG) and ligated into pGEM T-easy (Promega), resulting in plasmid pBS23. The KmR gene was excised from pUC4K by digesting with EcoRI and ligated into the MunI site of rssB in pBS23. The resulting plasmid (pBS25) was used as a template for the PCR amplification of the rssB-KmR fragment. The PCR product was resolved by electrophoresis, extracted from the gel and purified using the Wizard SV gel and PCR clean-up system (Promega). The linear DNA carrying rssB-KmR was electrotransformed into strain KM32 and plated on Km plates. One out of three colonies was KmR and AmpS, suggesting that the resistance to Km was due to insertion of KmR into the chromosome and not due to transfomation of pBS25 leftovers. The KmR insertion in rssB was verified by PCR. The rssB::KmR mutation was transferred to strain MC4100BS by P1 transduction .
Cloning of rssAB
A DNA fragment containing the entire rssAB operon was obtained by PCR amplification with primers rssA231F (5'-CCATCAATTCGGCACGTAAC) and rssB1368R (5'-GTATCGCATCCCAGTATATCAG) and cloned in pGEM T-easy (Promega) following the manufacturer instructions. The resulting plasmid was then digested with EcoRI and the rssAB fragment was ligated to the low-copy vector pWKS130  previously linearised with EcoRI, resulting in plasmid pBS28. Strain DH10B was used as a recipient for DNA transformation.
PCR and sequencing
The rpoS ORF from each strain was amplified by PCR using the sets of primers 429F (5'-GGAACAACAAGAAGTTAAGG)/9274R (5'-CAGACCACGATTGCCATAAC), C600F (5'-CAGGGATCCACACAGCGTGT)/9700R (5'-GTCATCTTGCGTGGTATCTTCC), 9363F (5'-CATACGCAACCTGGTGGATT)/10086R (5'-GTGTTAACGACCATTCTCGG). The PCR products were confirmed by electrophoresis in a 1.5% agarose gel and purified with the Concert Rapid PCR Purification System kit (Life Technologies, Bethesda, MD). Sequencing reactions were directly performed from purified PCR products using the same primers for both strands and Big Dye Terminator v3.1 (Life Technologies, Foster City, CA). Sequencing was carried out on an automated sequencer (ABI Prism 3130XL DNA Analyzer, Applied Biosystems, Foster City), according to the manufacturer recommendations. The rpoS sequences from the LB stabs isolates were deposited in the GenBank database under the accession numbers JN813535-JN813544.
Colony Forming Units
Coli Genetic Stock Center
Growth Advantage in the Stationary Phase
Self Preservation and Nutritional Competence
Lapage S, Shelton JE, Mitchell T, Mackenzie A: Chapter II Culture Collections and the Preservation of Bacteria. Part 1 of Methods in Microbiology. 1970, Academic Press, 3: 135-228. [http://www.sciencedirect.com/science/article/pii/S0580951708705403]
Sanderson KE, Zeigler DR: Storing, shipping, and maintaining records on bacterial strains. Methods Enzymol. 1991, 204: 248-264.
Ferenci T, Galbiati HF, Betteridge T, Phan K, Spira B: The constancy of global regulation across a species: the concentrations of ppGpp and RpoS are strain-specific in Escherichiacoli. BMC Microbiol. 2011, 11: 62-10.1186/1471-2180-11-62.
Johnson J, Delavari P, Stell A, Prats G, Carlino U, Russo T: Integrity of archival strain collections: the ECOR collection. ASM NEWS. 2001, 67 (6): 288-289.
Faure D, Frederick R, Wloch D, Portier P, Blot M, Adams J: Genomic changes arising in long-term stab cultures of Escherichia coli. J Bacteriol. 2004, 186 (19): 6437-6442. 10.1128/JB.186.19.6437-6442.2004.
Naas T, Blot M, Fitch WM, Arber W: Dynamics of IS-related genetic rearrangements in resting Escherichia coli K-12. Mol Biol Evol. 1995, 12 (2): 198-207.
Edwards K, Linetsky I, Hueser C, Eisenstark A: Genetic variability among archival cultures of Salmonella typhimurium. FEMS Microbiol Lett. 2001, 199 (2): 215-219. 10.1111/j.1574-6968.2001.tb10677.x.
Sutton A, Buencamino R, Eisenstark A: rpoS mutants in archival cultures of Salmonella enterica serovar typhimurium. J Bacteriol. 2000, 182 (16): 4375-4379. 10.1128/JB.182.16.4375-4379.2000.
Finkel SE, Kolter R: Evolution of microbial diversity during prolonged starvation. Proc Natl Acad Sci USA. 1999, 96 (7): 4023-4027. 10.1073/pnas.96.7.4023.
Weber H, Polen T, Heuveling J, Wendisch VF, Hengge R: Genome-wide analysis of the general stress response network in Escherichia coli σS-dependent genes, promoters, and sigma factor selectivity. J Bacteriol. 2005, 187 (5): 1591-603. 10.1128/JB.187.5.1591-1603.2005.
Patten CL, Kirchhof MG, Schertzberg MR, Morton RA, Schellhorn HE: Microarray analysis of RpoS-mediated gene expression in Escherichia coli K-12. Mol Genet Genomics. 2004, 272 (5): 580-591. 10.1007/s00438-004-1089-2.
Mandel MJ, Silhavy TJ: Starvation for different nutrients in Escherichiacoli results in differential modulation of RpoS levels and stability. J Bacteriol. 2005, 187 (2): 434-442. 10.1128/JB.187.2.434-442.2005.
Farewell A, Kvint K, Nyström T: Negative regulation by RpoS: a case of sigma factor competition. Mol Microbiol. 1998, 29 (4): 1039-1051. 10.1046/j.1365-2958.1998.00990.x.
Ferenci T: Maintaining a healthy SPANC balance through regulatory and mutational adaptation. Mol Microbiol. 2005, 57: 1-8. 10.1111/j.1365-2958.2005.04649.x.
Dong T, Chiang SM, Joyce C, Yu R, Schellhorn HE: Polymorphism and selection of rpoS in pathogenic Escherichiacoli. BMC Microbiol. 2009, 9: 118-10.1186/1471-2180-9-118.
Atlung T, Nielsen HV, Hansen FG: Characterisation of the allelic variation in the rpoS gene in thirteen K12 and six other non-pathogenic Escherichiacoli strains. Mol Genet Genomics. 2002, 266 (5): 873-81. 10.1007/s00438-001-0610-0.
King T, Ishihama A, Kori A, Ferenci T: A regulatory trade-off as a source of strain variation in the species Escherichiacoli. J Bacteriol. 2004, 186 (17): 5614-20. 10.1128/JB.186.17.5614-5620.2004.
Spira B, Ferenci T: Alkaline phosphatase as a reporter of σS levels and rpoS polymorphisms in different E.coli strains. Arch Microbiol. 2008, 189: 43-47.
Ferenci T, Zhou Z, Betteridge T, Ren Y, Liu Y, Feng L, Reeves PR, Wang L: Genomic sequencing reveals regulatory mutations and recombinational events in the widely used MC4100 lineage of Escherichiacoli K-12. J Bacteriol. 2009, 191 (12): 4025-9. 10.1128/JB.00118-09.
Spira B, Hu X, Ferenci T: Strain variation in ppGpp concentration and RpoS levels in laboratory strains of Escherichiacoli K-12. Microbiology. 2008, 154 (Pt 9): 2887-2895.
Gentry DR, Hernandez VJ, Nguyen LH, Jensen DB, Cashel M: Synthesis of the stationary-phase sigma factor sigma s is positively regulated by ppGpp. J Bacteriol. 1993, 175 (24): 7982-9.
Becker G, Klauck E, Hengge-Aronis R: Regulation of RpoS proteolysis in Escherichiacoli: the response regulator RssB is a recognition factor that interacts with the turnover element in RpoS. Proc Natl Acad Sci USA. 1999, 96 (11): 6439-44. 10.1073/pnas.96.11.6439.
Hengge-Aronis R, Fischer D: Identification and molecular analysis of glgS, a novel growth-phase-regulated and rpoS-dependent gene involved in glycogen synthesis in Escherichia coli. Mol Microbiol. 1992, 6 (14): 1877-1886. 10.1111/j.1365-2958.1992.tb01360.x.
Gruber TM, Gross CA: Multiple sigma subunits and the partitioning of bacterial transcription space. Annu Rev Microbiol. 2003, 57: 441-466. 10.1146/annurev.micro.57.030502.090913.
Taschner NP, Yagil E, Spira B: A differential effect of σS on the expression of the PHO regulon genes of Escherichiacoli. Microbiology. 2004, 150 (Pt 9): 2985-2992.
Ashwood-Smith MJ, Grant E: Mutation induction in bacteria by freeze-drying. Cryobiology. 1976, 13 (2): 206-213. 10.1016/0011-2240(76)90134-6.
Tanaka Y, Yoh M, Takeda Y, Miwatani T: Induction of mutation in Escherichiacoli by freeze-drying. Appl Environ Microbiol. 1979, 37 (3): 369-372.
Zambrano MM, Siegele DA, Almirón M, Tormo A, Kolter R: Microbial competition Escherichiacoli mutants that take over stationary phase cultures. Science. 1993, 259 (5102): 1757-60. 10.1126/science.7681219.
Vulic M, Kolter R: Evolutionary cheating in Escherichiacoli stationary phase cultures. Genetics. 2001, 158 (2): 519-526.
Waterman SR, Small PL: Characterization of the acid resistance phenotype and rpoS alleles of shiga-like toxin-producing Escherichiacoli. Infect Immun. 1996, 64 (7): 2808-2811.
Benito A, Ventoura G, Casadei M, Robinson T, Mackey B: Variation in resistance of natural isolates of Escherichiacoli O157 to high hydrostatic pressure, mild heat, and other stresses. Appl Environ Microbiol. 1999, 65 (4): 1564-1569.
Visick JE, Clarke S: RpoS- and OxyR-independent induction of HPI catalase at stationary phase in Escherichiacoli and identification of rpoS mutations in common laboratory strains. J Bacteriol. 1997, 179 (13): 4158-63.
Porwollik S, Wong RMY, Helm RA, Edwards KK, Calcutt M, Eisenstark A, McClelland M: DNA amplification and rearrangements in archival Salmonella enterica serovar Typhimurium LT2 cultures. J Bacteriol. 2004, 186 (6): 1678-1682. 10.1128/JB.186.6.1678-1682.2004.
Cundell AM, Chatellier S, Schumann P, Lilischkis R: Equivalence of quality control strains of microorganisms used in the compendial microbiological tests: are national culture collection strains identical?. PDA J Pharm Sci Technol. 2010, 64 (2): 137-155.
Zinser ER, Kolter R: Prolonged stationary-phase incubation selects for lrp mutations in Escherichia coli K-12. J Bacteriol. 2000, 182 (15): 4361-4365. 10.1128/JB.182.15.4361-4365.2000.
Loewen PC, Hu B, Strutinsky J, Sparling R: Regulation in the rpoS regulon of Escherichiacoli. Can J Microbiol. 1998, 44 (8): 707-17.
Carabetta VJ, Mohanty BK, Kushner SR, Silhavy TJ: The response regulator SprE (RssB) modulates polyadenylation and mRNA stability in Escherichia coli. J Bacteriol. 2009, 191 (22): 6812-6821. 10.1128/JB.00870-09.
Maharjan R, Seeto S, Notley-McRobb L, Ferenci T: Clonal adaptive radiation in a constant environment. Science. 2006, 313 (5786): 514-517. 10.1126/science.1129865.
Wang L, Spira B, Zhou Z, Feng L, Maharjan RP, Li X, Li F, McKenzie C, Reeves PR, Ferenci T: Divergence involving global regulatory gene mutations in an Escherichia coli population evolving under phosphate limitation. Genome Biol Evol. 2010, 2: 478-487. 10.1093/gbe/evq035.
Wu J, Xie J: Magic spot: (p) ppGpp. J Cell Physiol. 2009, 220 (2): 297-302. 10.1002/jcp.21797.
Potrykus K, Cashel M: (p)ppGpp: still magical?. Annu Rev Microbiol. 2008, 62: 35-51. 10.1146/annurev.micro.62.081307.162903.
Poteete AR, Fenton AC, Murphy KC: Roles of RuvC and RecG in phage lambda red-mediated recombination. Journal of Bacteriology. 1999, 181 (17): 5402-5408. [PMID: 10464213], [http://www.ncbi.nlm.nih.gov/pubmed/10464213]
Taylor LA, Rose RE: A correction in the nucleotide sequence of the Tn903 kanamycin resistance determinant in pUC4K. Nucleic Acids Research. 1988, 16: 358-10.1093/nar/16.1.358. [PMID: 3340535], [http://www.ncbi.nlm.nih.gov/pubmed/3340535]
Wang RF, Kushner SR: Construction of versatile low-copy-number vectors for cloning, sequencing and gene expression in Escherichiacoli. Gene. 1991, 100: 195-9.
Echols H, Garen A, Garen S, Torriani A: Genetic control of repression of alkaline phosphatase in E.coli. J Mol Biol. 1961, 3: 425-38. 10.1016/S0022-2836(61)80055-7.
Miller JH: A Short Course In Bacterial Genetics: A Laboratory Manual And Handbook For Escherichiacoli And Related Bacteria. 1992, Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y
Sambrook J, Russel D: Molecular Cloning: A Laboratory Manual. 2001, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York
Murphy KC, Campellone KG, Poteete AR: PCR-mediated gene replacement in Escherichiacoli. Gene. 2000, 246 (1-2): 321-330. 10.1016/S0378-1119(00)00071-8.
We are grateful to Fundação de Amparo á Pesquisa do Estado de São Paulo (FAPESP-Brazil), who supported this study and provided a travel allowance for TF. TF was also supported by the the Australian Research Council and the US Army Research Office. We also thank K. C. Murphy and S. Kushner for respectively providing strain KM32 and plasmid pWKS130.
BS conceived and desgined the study, performed most experiments and wrote the manuscript. RAT sequenced the rpoS mutants. TF suggested experiments, wrote and corrected the manuscript. RPM prepared cultures for transportation. All authors have read and approved the final manuscript.