Swiss Federal Institute for Environmental Science and Technology, PO Box 611, Überlandstrasse 133, CH-8600 Dübendorf, Switzerland
Correspondence
Thomas Egli
egli{at}eawag.ch
![]() |
ABSTRACT |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
INTRODUCTION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
In batch cultures, intracellular levels of RpoS typically increase during the transition from the exponential phase to the stationary phase (Gentry et al., 1993; Jishage et al., 1996
; Lange & Hengge-Aronis, 1994
), making stationary-phase cells broadly stress-resistant. What triggers this increase has been the subject of much debate, but it is clear that RpoS is controlled in a complex manner at the transcriptional, translational and protein-stability levels (Lange & Hengge-Aronis, 1994
; Loewen et al., 1993
; McCann et al., 1993
; for an overview see Hengge-Aronis, 2002
). Different environmental signals have been shown to induce elevated intracellular RpoS levels, among them reduction in growth rate (Notley & Ferenci, 1996
; Teich et al., 1999
), high osmolarity (Muffler et al., 1996
), heat shock (Muffler et al., 1997
), temperature downshift (Sledjeski et al., 1996
) and acidic pH (Bearson et al., 1996
).
Cell density has also been suggested to be an important signal for RpoS expression (Lange & Hengge-Aronis, 1994). Cell-density effects are generally thought to be mediated by external signal molecules that affect gene expression when they reach a certain threshold concentration, a phenomenon called quorum sensing (Fuqua et al., 1994
). This was first described for Vibrio fischeri, where homoserine lactone serves as signal molecule (autoinducer) for luminescence (Nealson, 1977
). Under certain conditions, E. coli excretes into the medium a compound, referred to as autoinducer 2, which is able to induce luminescence in Vibrio harveyi (Surette & Bassler, 1998
; Surette et al., 1999
). Therefore, quorum sensing has been postulated as a regulatory mechanism in E. coli and Salmonella typhimurium (Surette & Bassler, 1998
; Surette et al., 1999
; Sperandio et al., 2001
).
Conflicting data exist with regard to a quorum-sensing-mediated regulation of RpoS (reviewed by Hengge-Aronis, 2002). Intracellular RpoS concentrations increased when cell density rose from 0·09 to 0·5 g dry cell weight per litre (g DCW l1) in chemostat cultures of E. coli strain MC4100 at a constant dilution rate (D) of 0·3 h1 (Liu et al., 2000
). In contrast, rpoS mRNA levels were constant during an increase in cell density from 10 to 90 g DCW l1 in fed batch cultures of E. coli strain W3110 (Yoon et al., 2003
). RpoS transcription in Pseudomonas aeruginosa PAO1 was dependent on functional LasR and RhlR, activators that respond to N-acylhomoserine lactone (AHL) signal molecules accumulating at high cell density (Latifi et al., 1996
). However, overexpression of RelA allowed AHL production and RpoS expression at low cell density also (van Delden et al., 2001
).
In standard batch cultures, it is difficult to attribute an observed response exclusively to quorum sensing. During the transition from unrestricted growth to the stationary phase, not only does cell density change, but also specific growth-rate, extracellular carbon-source availability, metabolite concentrations (Wanner & Egli, 1990) and, in the case of LB, amino-acid availability. These complex changes in turn affect the level of intracellular signal molecules like cAMP and (p)ppGpp, and RpoS expression has been reported to be influenced by (p)ppGpp (Gentry et al., 1993
; Lange et al., 1995
), cAMP (Lange & Hengge-Aronis, 1991
; Lange & Hengge-Aronis, 1994
) and acetate (Schellhorn & Stones, 1992
), which is the main metabolite excreted by E. coli in aerobic batch culture. Chemostat cultivation offers an elegant way out of the complex situation in batch culture (Liu et al., 2000
) because the specific growth rate (which equals D under steady-state conditions) and physico-chemical factors can be kept constant, while single parameters like cell density or the nature of the limiting nutrient are varied (Pirt, 1975
).
The aim of the present study was to assess the contribution of individual environmental and physiological factors to the increase in RpoS seen in batch and certain chemostat cultures of E. coli. For this, the effects of different steady-state cell densities and different types of limitation on RpoS were studied in chemostat cultures growing at similar specific growth rates. In another set of experiments, cell density was kept low while specific growth rate was varied.
We demonstrate that neither cell density nor the type of limitation have an effect on RpoS expression, whereas specific growth rate exerts a strong control. Furthermore, we have found that the standard E. coli minimal medium A (MMA) is unsuitable for cultures with cell densities higher than 0·5 g DCW l1 because it lacks trace elements. Hence, earlier observations of cell-density effects in E. coli are most likely due to a hidden shift in the growth-limiting nutrient.
![]() |
METHODS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Growth media and their design.
Mineral medium allowing carbon-limited cultivation up to a glucose concentration of 4·0 g l1 was designed, based on the average composition of bacterial biomass as previously described (Egli, 2000; Pirt, 1975
). Basal medium for carbon-limited batch cultivation (medium ClimBa) consisted of 12·8 g Na2HPO4.2H2O, 3·0 g KH2PO4 and 1·77 g (NH4)2SO4 per litre. The following salts and trace elements were added as a 100-fold concentrated solution after autoclaving (final concentrations are given): 130 mg MgCl2.6H2O l1; 80 mg CaCO3 l1, 77 mg FeCl3.6H2O l1, 11 mg MnCl2.4H2O l1, 1·5 mg CuSO4.5H2O l1, 1·3 mg CoCl2.6H2O l1, 4 mg ZnO l1, 1·2 mg H3BO3 l1, 10 mg NaMoO4.2H2O l1 and 790 mg EDTA Na4.2H2O l1 (equimolar to di- and trivalent cations). Glucose was always autoclaved separately and added after cooling. For high-growth-rate batch culture, 2 g l1 Casamino Acids and a mixture of 11 vitamins were added to medium ClimBa at the following final concentrations: biotin and folic acid 20 µg l1; pyridoxine 100 µg l1; thiamin, riboflavin, nicotinic acid, vitamin B12, calcium D-pantothenate, p-aminobenzoic acid, lipoic acid and nicotinamide 50 µg l1.
Medium for carbon-limited chemostat cultivation (medium ClimCh) consisted of 2·72 g KH2PO4, 2·3 g NH4Cl, 1·4 g (NH4)2SO4, 0·1 ml concentrated H2SO4 and 0·15 ml silicone anti-foam (Fluka) per litre. Trace elements in the concentrations mentioned above could be added before autoclaving because the medium pH was around 4·0, which prevents salt precipitation.
For iron-limited chemostat cultivation (medium FelimCh), iron was omitted from the trace element solution. For nitrogen-limited chemostat cultivation (medium NlimCh), the ammonium supply was reduced to 0·21 g NH4Cl l1 (ammonium sulfate was omitted) and concentrated H2SO4 was increased to 0·15 ml l1 to guarantee surplus sulphur supply. LuriaBertani broth (LB) contained 10 g tryptone, 5 g yeast extract and 10 g NaCl per litre (Miller, 1972). Minimal medium A (MMA) consisted of 10·5 g K2HPO4, 4·5 g KH2PO4, 0·5 g sodium citrate and 0·25 g MgSO4.7H2O per litre (Miller, 1972
). To avoid nitrogen limitation at a high-glucose-feed concentration, (NH4)2SO4 concentration was increased from 1·0 to 4·3 g l1. Thiamin was added at a concentration of 1 mg l1, if required (e.g. for strain BW2952). For direct comparison to medium ClimCh, modified MMA was used that contained (per litre): 2·72 g KH2PO4, 2·3 g NH4Cl, 1·4 g (NH4)2SO4, 0·1 ml H2SO4, 0·42 g citric acid monohydrate, 0·25 g MgSO4.7H2O and 0·15 ml silicone anti-foam. Viable cell numbers were determined by plating on tryptone glucose yeast extract agar (TGY; Biolife). Nanopure water and high-purity chemicals were used for all solutions and media.
Cultivation conditions.
For batch cultures, bacteria were incubated at 37 °C in electromagnetically stirred Erlenmeyer flasks. For continuous culture, computer-controlled glass and stainless steel bioreactors (MBR, Wetzikon, Switzerland) were used. pH was kept at 7·0±0·1 by automatic addition of a solution of 0·5 M NaOH and 0·5 M KOH and the temperature was set at 37±0·1 °C. Oxygen concentration was kept between 95 and 100 % of air saturation at 37 °C, and the stirrer speed was set to 800 r.p.m.
Some E. coli strains quickly acquire mutations in rpoS when cultivated in carbon-limited chemostats (Notley-McRobb et al., 2002). To avoid the accumulation of mutations in the populations investigated, a new preculture was prepared for each individual batch or chemostat experiment from the same cryo-vial stored at 80 °C. Colonies grown on TGY agar for 1518 h at 37 °C were transferred to 5 ml of either complex or mineral medium used later in the experiments. After 3 h incubation at 37 °C, 12 ml of the culture was transferred into 100 ml medium in Erlenmeyer flasks. This culture was used as inoculum for the bioreactors. Batch experiments were conducted in three parallel Erlenmeyer flasks, containing pre-warmed medium, that were inoculated with exponentially growing cells from 5 ml cultures. For batch experiments with glucose mineral medium, cultures were grown overnight before using them as inoculum. Dilutions were repeated as often as necessary to keep cultures in exponential phase. In chemostat cultures, steady-state is defined as constant optical density, and this was always reached after 510 volume changes. Chemostats were sampled 3050 h after inoculation, which corresponds to 1322 generations at a dilution rate of 0·3 h1. Purity of chemostat cultures was controlled at the end of each experiment by plating on TGY agar and transferring all colonies to E. coli diagnostic agar plates (ECD-MUG; Biolife).
Specific growth rates in batch cultures were calculated by linear regression of growth curves; in chemostats, D equals specific growth rate.
Hydroperoxidase assay and -galactosidase activity.
The two E. coli hydroperoxidases (catalases) were used as reporter genes' for RpoS-dependent transcription because both hydroperoxidase I (katG gene product, cytoplasmic membrane associated) and hydroperoxidase II (katE gene product, present in the cytosol) were shown to be regulated by s (Ivanova et al., 1994
), with hydroperoxidase II expression being entirely dependent on RpoS (Visick & Clarke, 1997
). Hydroperoxidase I (HPI) and hydroperoxidase II (HPII) specific activities were measured as described by Visick & Clarke (1997)
. Briefly, E. coli cells were sampled directly onto ice. De novo protein synthesis was stopped by adding 25 µg chloramphenicol ml1. After centrifugation at 6000 g, the cell pellet was resuspended in 200 µl lysis buffer and cells were lysed by sonication (Sonifier 450; Branson; power setting of 3), while being cooled on ice. The extract was centrifuged for 10 min at 6000 g to remove cell debris. H2O2-degrading activity was determined by diluting 50 µl of the supernatant (crude extract) in 2 ml of 50 mM sodium phosphate buffer (pH 7·0, 37 °C), adding 4 µl H2O2 (37 %) and then following the decrease in absorbance at 240 nm over time. The protein concentration in the crude extract was determined with a commercial bichinoic acid assay (Bio-Rad). The specific activity of hydroperoxidase was calculated following Visick & Clarke (1997)
.
HPI and HPII were differentiated by heating an aliquot of the crude extract for 15 min at 55 °C. Visick & Clarke (1997) demostrated that HPI is heat labile at 55 °C, whereas HPII is stable. We could verify the strict RpoS-dependency of HPII expression by comparing an rpoS0 allele of our strain with the wild-type (Table 1
, Fig. 4
).
|
|
Detection of rpoS mutants.
RpoS null and partial mutants were distinguished from the wild-type by iodine staining of colonies grown on TGY agar plates and incubated overnight at 4 °C (Notley-McRobb et al., 2002). Dark-brown colonies were counted as rpoS positive, whereas pale-brown colonies were assumed to be rpoS-attenuated or null mutants.
Protein extraction and Western blotting analysis of RpoS.
Cells for immunoblot analysis were sampled directly onto ice and centrifuged at 4 °C. After carefully removing the supernatant, the pellet was immediately frozen at 20 °C. The remaining water was removed by freeze-drying. Before loading onto a 10 % acrylamide gel, the pellet was resuspended in SDS-PAGE sample buffer and boiled for 5 min. The amount of sample buffer added was normalized to the optical density of the initial culture fluid/ice mixture. Total band intensity on Coomassie-stained SDS-PAGE control gels showed that this procedure led to approximately the same protein concentrations in samples. Protein extraction, by heating samples supplemented with 0·1 M KOH to 80 °C for 40 min, also led to similar protein concentrations after normalization of optical density (as determined with a bichinoic acid assay). To verify results obtained with the above sampling method, protein was also extracted by direct precipitation with 10 % TCA (Lange & Hengge-Aronis, 1994). Because of loss of protein during washing steps, loading volumes had to be adjusted by normalizing to total band intensity on Coomassie-stained gels. Relative band intensities were quantified with the publicly available software ImageJ (http://rsb.info.nih.gov/ij/download.html).
Following SDS-PAGE, proteins were blotted overnight (4 °C, 30 V) onto nitrocellulose membranes. The transfer buffer contained 25 mM Tris, 192 mM glycine and 20 % methanol. After blotting, the membrane was blocked for 1 h in TBSTM (20 mM Tris/HCl, pH 7·5, 500 mM NaCl, 0·05 %, v/v, Tween 20, 5 %, w/v, skimmed milk powder), washed for 10 min with TBSTM and incubated for 2 h with mouse monoclonal antibodies against RpoS or RpoD (Neoclone, Madison, WI, USA) which were both diluted in TBSTM. Following two further TBSTM washing steps, membranes were hybridized for 1 h with goat anti-mouse antibodies coupled to alkaline phosphatase (Bio-Rad). RpoS bands were visualized by reaction with BCIP/NBT (Sigma-Aldrich) and identified by comparison to a rpoS-negative control strain (Fig. 2).
|
![]() |
RESULTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
|
Type of limitation does not influence RpoS levels
Due to high external glucose concentrations, cAMP levels in nitrogen- (i.e. non-carbon-) limited growing cells are very low compared to those in carbon-limited growing cells (Notley-McRobb et al., 1997). Furthermore, acetate concentrations in non-carbon-limited chemostat cultures are high, due to overflow metabolism (Neijssel et al., 1996
; Table 2
). Although both cAMP and acetate have been reported to influence RpoS levels, we found no difference in intracellular RpoS concentrations in nitrogen- and iron-limited chemostats, compared to carbon-limited chemostats operated at the same dilution rate (Fig. 2b
).
|
A direct test of the effects of cAMP and acetate on RpoS levels in chemostat cultures also indicated that differences in the nature of the limiting nutrient have no effect on RpoS expression. Neither a pulse of 0·45 mM cAMP into a nitrogen-limited chemostat culture, nor the addition of 0·4 g l1 (24 mM) actetate to the highest cell density carbon-limited chemostat culture had an effect on RpoS immunoblot band intensities (data not shown).
RpoS levels and hydroperoxidase specific activity are strongly affected by specific growth rate
RpoS band-intensity on immunoblots inversely correlated with specific growth rate in exponential-phase batch cultures and steady-state chemostat cultures, whereas the level of housekeeping sigma factor 70 (RpoD) was not much affected (Fig. 2c
). Intracellular RpoS concentration was also influenced by amino acid availability, with a higher level in amino-acid-free mineral medium cultures compared to an LB chemostat culture growing at the same specific growth rate (Fig. 2c
). All samples were taken at low cell density (i.e. below an OD546 of 0·25, corresponding to 0·1 g DCW l1), which demonstrates that regulation by specific growth rate proceeds independently of cell density.
The specific activities of HPI and HPII were also strongly regulated by specific growth rate (Fig. 4). Expression ranged from almost complete repression at µ=1·94 h1 to maximum specific activities of 54 (HPI) and 81 (HPII) µmol H2O2 min1 (mg protein)1 at the lowest specific growth rate tested (0·03 h1). Interestingly, HPII specific activity was much higher in amino-acid-free mineral medium culture than in an amino-acid-containing LB culture, in spite of similar specific growth rates (Fig. 4
), whereas HPI specific activity was much higher in the LB culture, leading to approximately similar HPI and HPII combined activities (Fig. 4
). Cultivation to stationary phase in mineral medium resulted in RpoS and hydroperoxidase levels similar to those of chemostat cultures with a dilution rate of 0·3 h1, whereas stationary phase in LB culture induced twofold higher HPI and HPII levels, in the same range as those of chemostat cultures with a dilution rate of 0·03 h1 (Figs 2 and 4
, Table 1
). This difference is not due to the presence of sodium chloride in LB because overnight LB cultures with and without 10 g sodium chloride l1 had similar hydroperoxidase specific activities (Table 1
).
In spite of loss of RpoS function, HPI specific activity was strongly elevated in overnight batch and chemostat cultures of the rpoS0 strain, leading to total hydroperoxidase specific activities in the same range as those of the wild-type strain (Fig. 4, Table 1
).
High-cell-density MMA chemostat cultures become non-carbon limited
In an earlier chemostat study in which cell-density effects on E. coli were reported, MMA was used as growth medium (Liu et al., 2000). However, MMA lacks trace elements (Miller, 1972
). Thus, we tested whether the observed effects might be due to a hidden secondary limitation in high glucose-feed chemostat cultures.
Nitrogen- and iron-limited (i.e. non-carbon-limited) chemostat cultures of E. coli MG1655 were characterized by a 4050 % lower growth-yield compared to glucose-limited cultures with medium ClimCh (Table 2). Furthermore, 20 % of the glucose carbon consumed was transformed to acetate (calculated from values in Table 2
). This is in agreement with data reported for nitrogen-, phosphorus- and potassium-limited chemostat cultures of E. coli strain PC-1000 (Neijssel et al., 1996
). Judged from growth yield, steady-state glucose concentration and acetate excretion, MMA and modified MMA sustained purely carbon-limited growth in chemostats fed with 0·1 g glucose l1 (Table 2
), whereas high glucose feed resulted in non-carbon limitation (Table 2
). In spite of 6 g l1 glucose feed, the maximum OD546 obtainable with MMA in chemostat culture was 1·7 (Table 2
), which corresponds to 0·5 g DCW l1. Addition of 77 mg l1 FeCl3.6H2O to a high-glucose-feed MMA chemostat culture of strain BW2952 led to an increase in OD546 from 1·7 to 2·9. These findings suggest that MMA is unsuitable for growing bacterial cells to densities above 0·5 g DCW l1 because the cultures become iron-limited.
Hydroperoxidase specific activity in a high-glucose-feed chemostat culture of MG1655 with modified MMA closely matched that of an iron-limited chemostat (Fig. 3b), but again, RpoS levels were similar in carbon- and non-carbon-limited modified MMA cultures (Fig. 2b
).
Medium ClimCh, designed to guarantee surplus supply of all essential nutrients, except carbon and energy sources, allowed carbon-limited growth up to 3 g l1 glucose feed, corresponding to OD546 of 3·0 and 1·1 g DCW l1 (Table 2, Fig. 1
). At all cell densities tested, residual glucose concentration remained below 0·5 mg l1, and acetate was not excreted into the medium in detectable amounts (Table 2
).
Indirect effect of the type of limitation on RpoS levels in E. coli BW2952
For the analysis of cell-density effects on MC4100, Liu et al. (2000) sampled their continuous cultures after 30 generations. However, recently it was reported that glucose-limited chemostat cultivation strongly selects for loss or attenuation of RpoS function in E. coli BW2952, an MC4100 derivative (Notley-McRobb et al., 2002
), so that 96 % of the BW2952 population was mutated in rpoS after 30 generations (Notley-McRobb et al., 2002
). Hence, at the time point of sampling for cell-density effects, low-cell-density, carbon-limited cultures must have already been dominated by rpoS mutants. The same authors reported that population takeover by rpoS mutants proceeds much more slowly in nitrogen- (i.e. non-carbon-) limited chemostats (Notley-McRobb et al., 2002
). Thus, in chemostat experiments with strain MC4100 and MMA as growth medium, a hidden non-carbon limitation at high glucose feed probably slowed down population takeover by rpoS mutants and, therefore, mimicked a cell-density effect on RpoS.
The following experimental data support this interpretation. Cells sampled after 30 generations from a low-cell-density, carbon-limited MMA chemostat culture of strain BW2952 exhibited a five-times lower HPII specific activity than cells taken from a high-cell-density, non-carbon-limited MMA chemostat culture (Table 3). Furthermore, a larger proportion of cells had already lost rpoS in the low-cell-density chemostat, as judged from dark-brown colonies after iodine staining and from malGlacZ activity (Table 3
), which was in the same range as values reported for rpoS mutants of BW2952 (Notley-McRobb et al., 2002
).
|
![]() |
DISCUSSION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
However, the evidence from chemostat and batch experiments presented in this paper strongly argues against any type of quorum-sensing regulation of RpoS in E. coli. In neither chemostat nor batch cultures did increasing cell density lead to higher RpoS levels or elevated RpoS-dependent HPII expression. The cell densities tested in our study cover the range in which quorum-sensing-dependent induction is usually reported to take place (Latifi et al., 1996; Surette & Bassler, 1998
). Low cell density is supposed to prevent the expression of quorum-sensing-dependent genes, which is clearly not the case for RpoS in our experiments.
Our conclusion that RpoS is not regulated by quorum sensing is backed by the majority of literature data. Several studies showed that supernatant fluid from stationary-phase batch cultures, which should contain quorum-sensing signal molecules, had little or no effect on rpoS : : lacZ transcriptional and translational fusions (Garcia-Lara et al., 1996; Sitnikov et al., 1996
; Hengge-Aronis, 2002
). Furthermore, intracellular RpoS concentrations, as well as rpoS mRNA levels, did not increase in parallel with cell density in fed-batch culture (Teich et al., 1999
; Yoon et al., 2003
). Finally, stationary-phase induction of a rpoS : : lacZ transcriptional fusion occurred equally well in 100-fold diluted LB and in normal-strength LB (Huisman & Kolter, 1994
).
Quorum-sensing in E. coli and Salmonella typhimurium is said to be mediated by so-called autoinducer 2 (AI-2), the production of which depends on the luxS gene product (Surette et al., 1999). However, LuxS plays an important role in central metabolism, i.e. in the recycling of S-adenosylmethionine (Winzer et al., 2002
). AI-2 may in fact not represent a signal molecule but a metabolite that is released in exponential phase and used as a carbon and energy source in later (carbon-limited) stages of growth (Winzer et al., 2002
). Even if one accepts the role of AI-2 as a signal molecule, it seems not to be important for rpoS regulation, because rpoS transcription was not induced significantly in a luxS mutant by the addition of AI-2-containing culture supernatant (DeLisa et al., 2001
). Likewise, a transcriptome analysis with E. coli EHEC O157 : H7 showed that rpoS mRNA levels are not affected by a luxS knockout mutation (Sperandio et al., 2001
).
There is one study which reported, at first sight, convincing evidence for cell-density effects on gene expression in E. coli, including an inducing effect on RpoS (Liu et al., 2000). Results obtained in our chemostat experiments suggest that these effects are an artefact caused by a hidden, secondary nutritional limitation occurring at high glucose feed in the standard E. coli mineral medium MMA (discussed below). The observed effect of cell density on RpoS levels in the strain employed by Liu et al. (2000)
is probably caused by differential loss of RpoS function in low- and high-cell-density MMA chemostats (see also Results, indirect effect of the type of limitation on RpoS levels in E. coli BW2952).
From an ecological point of view, dependence on signals from other cells for starvation survival seems to be quite risky. An egoistic approach, where each bacterial cell reacts on its own, efficiently and quickly to stresses and changes in its environment, might ensure a better survival of the species. The fact that E. coli is usually present in its primary habitat at rather low cell densities of 106 per g colon content (Smith, 1965) also speaks against intraspecies quorum sensing being important for this bacterium.
We speculate that a similar set of experiments to those described in this study might reveal that RpoS expression proceeds independently of cell density in other microorganisms as well, and particularly in P. aeruginosa, considering the conflicting data published for that organism (Latifi et al., 1996; van Delden et al., 2001
).
The type of limitation is irrelevant to RpoS expression
In chemostats operated at D=0·3 h1, RpoS was expressed at approximately the same intracellular concentrations, regardless of whether the cultures were carbon-, nitrogen- or iron-limited (Fig. 2b). This is in agreement with batch experiments showing that glucose and phosphate limitation lead to a similar increase in RpoS during growth arrest (Gentry et al., 1993
). Strictly rpoS-dependent stiA and stiC genes were also induced to about the same level during carbon, nitrogen and phosphate starvation (O'Neal et al., 1994
). Higher RpoS levels in nitrogen-limited compared to glucose-limited chemostat cultures of strain MC4100 (Liu & Ferenci, 2001
) are most likely due to a rapid loss of rpoS function under glucose limitation, in contrast to a slow loss under nitrogen limitation (Notley-McRobb et al., 2002
), and thus are not valid as counter-evidence.
Specific growth rate plays a prominent role in the regulation of the general stress response
Data presented in this study indicate a central role for specific growth rate in RpoS regulation in the absence of stresses other than nutritional limitation. Both RpoS itself and RpoS-dependent hydroperoxidase expression exhibited a strong negative correlation to specific growth rate, irrespective of cell density. This suggests that the increase in intracellular RpoS concentrations in LB batch cultures (Gentry et al., 1993; Lange & Hengge-Aronis, 1994
; Jishage et al., 1996
) is mainly triggered by a gradual decline in specific growth rate. A strong negative correlation to specific growth rate was also shown for RpoS-dependent osmYlacZ activity and trehalase expression in glucose-limited chemostats (Notley & Ferenci, 1996
). However, the tendency of MC4100 derivative strains quickly to lose rpoS function in chemostats (Notley-McRobb et al., 2002
), and the relA1 mutation in MC4100, which might affect RpoS expression, add some uncertainty to these results. Reduction in specific growth rate from 0·8 to 0·1 h1 in chemostat cultures of strain RB791 (F INrrnDrrnE1
lacIqL8) led to a twofold increase in steady-state RpoS levels (Teich et al., 1999
), which is in the same range as a 2·4-fold increase from µ=0·7 h1 to µ=0·1 h1 observed in our study (calculated from background-corrected band intensities in Fig. 2c
).
A stronger induction of the general stress response in stationary-phase LB cultures compared to mineral glucose medium (observed by Lange & Hengge-Aronis, 1994 and confirmed in this study) might be due to the extended period of very slow growth when cultures approach stationary phase in LB. Contrary to this, cultivation in mineral glucose medium is characterized by a sudden drop in specific growth rate to zero (Lange & Hengge-Aronis, 1994
) and, presumably, cells entering stationary phase in mineral medium do not have enough time and resources to induce a strong general stress response. The conditions that bacteria encounter when leaving the colon probably rather resemble those found in late stationary-phase LB, with a period of slow growth on hard-to-use substrates left in the faeces.
Which signals link specific growth rate and stress defence?
A possible intracellular signal that links RpoS and specific growth rate is (p)ppGpp, because RpoS is positively regulated by this factor (Gentry et al., 1993; Lange et al., 1995
) and (p)ppGpp levels rise in response to amino-acid deprivation (Cashel et al., 1996
), carbon starvation (Cashel et al., 1996
) and reduction in specific growth rate (Teich et al., 1999
). Not only does (p)ppGpp induce RpoS itself, but it also improves the ability of
s to compete for core RNA polymerase (Jishage et al., 2002
) and thus enhances rpoS-dependent transcription. Therefore, the strong effect of the presence or absence of amino acids in the medium on strictly RpoS-dependent HPII specific activity (Fig. 4
) might be due to different intracellular levels of (p)ppGpp.
Although HPI also correlated strongly with specific growth rate (Fig. 4, difference between white and black bars), its expression was not dependent on functional RpoS in a glucose-limited chemostat (Fig. 4
). Similarly, HPI was expressed to quite high levels in stationary-phase LB cultures of our rpoS knockout strain (Table 1
), which confirms results reported by Visick & Clarke (1997)
. Interestingly, the pattern of hydroperoxidase expression in a wild-type chemostat culture with LB complex medium was quite similar to rpoS mutant cultures, with relatively high HPI specific activities and a virtual absence of HPII (Fig. 4
). Due to the presence of amino acids, LB chemostat cultures are likely to be low in intracellular (p)ppGpp, which in turn could suppress RpoS itself and RpoS-dependent transcription initiation. Thus, it seems that HPI can take over as H2O2-degrading enzyme under low (p)ppGpp conditions or when functional RpoS is absent, which implies that additional mechanisms apart from (p)ppGppRpoS link specific growth rate and stress defence in E. coli. Possible mechanisms include changes in DNA topology (Drlica, 1992
), and a system that senses oxidative damage resulting from the cell's own respiratory activity, which increasingly becomes a problem with decreasing specific growth rate (Nyström, 1998
).
The importance of medium design for high-cell-density cultivation
Bacterial growth is dependent on adequate supply of all essential nutrients that cannot be synthesized by the cells themselves. MMA (Miller, 1972) and medium M9 are the two most-used mineral media for the cultivation of E. coli. Despite their widespread application, both media lack trace elements, particularly iron. Iron is necessary in considerable amounts for aerobic growth, so that the term trace element is actually not fully appropriate (Egli, 2000
). Not surprisingly, the highest cell density obtainable with MMA in chemostat culture was considerably lower than cell densities reached in trace-element-containing medium ClimCh (Table 2
). In spite of the use of stainless steel bioreactors, iron seemed to be the limiting nutrient in MMA. A theoretical analysis of medium MMA, based on elemental growth yields (Egli, 2000
), excludes a possible N-, P-, K- or S-limitation and leaves iron as the most probable growth-limiting nutrient.
Hence, a shift from glucose limitation at low cell density to dual glucose/iron limitation (Egli, 1997) at high cell density in experiments performed with medium MMA (Liu et al., 2000
) is very likely. Such an interpretation fits well with the increase in intracellular glucose, UDP-glucose and UDP-N-acetylglucosamine concentration with cell density (Liu et al., 2000
), indicating improved glucose availability and thus a shift to non-carbon limitation. Furthermore, the change in porin expression from predominantly OmpF to OmpC (Liu et al., 2000
) also argues for high-cell-density MMA cultures becoming non-carbon limited, because the same authors reported earlier that OmpF is much more strongly expressed than OmpC in glucose-limited chemostats (Liu & Ferenci, 1998
), whereas OmpC dominated over OmpF in nitrogen-limited chemostats (Liu & Ferenci, 1998
).
A thorough investigation of limitations affecting growth in MMA and M9 media requires further research.
Are hydroperoxidases in E. coli regulated by iron?
Hydroperoxidase specific activity was strongly decreased in iron-limited chemostats (Fig. 2b). A negative effect of iron starvation on hydroperoxidase activity was also observed for the opportunistic pathogen P. aeruginosa (Frederick et al., 2001
). The lowered hydroperoxidase specific activity could reflect either an iron deficiency-mediated repression of hydroperoxidase genes or a shortage of functional haem groups resulting in many non-functional hydroperoxidase proteins. It has been observed that iron-free cofactors are excreted into the medium under iron limitation (Townsley & Neilands, 1956
), and that total haem iron is significantly decreased in iron-limited growing yeast cells (Light, 1972
). Furthermore, expression of the iron regulator Fur is linked to the oxidative stress response through OxyR (Zheng et al., 1999
). However, although the iron-dependent superoxide dismutase SodB is regulated by Fur, this is not the case for KatG (HPI) and KatE (HPII) (Hantke, 2001
).
The effect of iron deficiency on hydroperoxidase specific activity offers an explanation for the 40 % decrease in HPI and HPII activity seen at high cell density (Fig. 3a). Although culture parameters do not indicate iron or dual carbon/iron limitation during growth in medium ClimCh (Table 2
), high cell density nevertheless might lead to difficulties in iron aquisition because of a lowered Fe3+ : [Fe3+ EDTA] ratio compared to that at low cell density. A reduced supply of iron then would cause a reduction in hydroperoxidase specific activity.
![]() |
ACKNOWLEDGEMENTS |
---|
![]() |
REFERENCES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Cashel, M., Gentry, D. R., Hernandez, V. J. & Vinella, D. (1996). The stringent response. In Escherichia coli and Salmonella: Cellular and Molecular Biology, pp. 14581496. Edited by F. C. Neidhardt and others. Washington, DC: American Society for Microbiology.
DeLisa, M. P., Wu, C. F., Wang, L., Valdes, J. J. & Bentley, W. E. (2001). DNA microarray-based identification of genes controlled by autoinducer 2-stimulated quorum sensing in Escherichia coli. J Bacteriol 183, 52395247.
Drlica, K. (1992). Control of bacterial DNA supercoiling. Mol Microbiol 6, 425433.[Medline]
Egli, T. (1997). Multiple-nutrient-limited growth of microorganisms: What are the consequences for bioremediation processes? In Environmental Biotechnology, Part I, pp. 189193. Edited by H. Verachtert & W. Verstraete. Oostende, Belgium: International Symposium on Environmental Biotechnology.
Egli, T. (2000). Nutrition of microorganisms. In Encyclopedia of Microbiology, 2nd edn, pp. 431447. San Diego: Academic Press.
Frederick, J. R., Elkins, J. G., Bollinger, N., Hassett, D. J. & McDermott, T. R. (2001). Factors affecting catalase expression in Pseudomonas aeruginosa biofilms and planktonic cells. Appl Environ Microbiol 67, 13751379.
Fuqua, W. C., Winans, S. C. & Greenberg, E. P. (1994). Quorum sensing in bacteria: the LuxR-LuxI family of cell density-responsive transcriptional regulators. J Bacteriol 176, 269275.[Medline]
Garcia-Lara, J., Shang, L. H. & Rothfield, L. I. (1996). An extracellular factor regulates expression of sdiA, a transcriptional activator of cell division genes in Escherichia coli. J Bacteriol 178, 27422748.[Abstract]
Gentry, D. R., Hernandez, V. J., Nguyen, L. H., Jensen, D. B. & Cashel, M. (1993). Synthesis of the stationary-phase sigma factor sigma s is positively regulated by ppGpp. J Bacteriol 175, 79827989.[Abstract]
Hantke, K. (2001). Iron and metal regulation in bacteria. Curr Opin Microbiol 4, 172177.[CrossRef][Medline]
Hengge-Aronis, R. (1996). Regulation of gene expression during entry into stationary phase. In Escherichia coli and Salmonella: Cellular and Molecular Biology, pp. 14971512. Edited by F. C. Neidhardt and others. Washington, DC: American Society for Microbiology.
Hengge-Aronis, R. (2000). The general stress response in E. coli. In Bacterial Stress Responses, pp. 161178. Edited by G. Storz & R. Hengge-Aronis. Washington, DC: American Society for Microbiology.
Hengge-Aronis, R. (2002). Signal transduction and regulatory mechanisms involved in control of the sigma(S) (RpoS) subunit of RNA polymerase. Microbiol Mol Biol Rev 66, 373395.
Huisman, G. W. & Kolter, R. (1994). Sensing starvation: a homoserine lactone-dependent signaling pathway in Escherichia coli. Science 265, 537539.[Medline]
Ivanova, A., Miller, C., Glinsky, G. & Eisenstark, A. (1994). Role of rpoS (katF) in oxyR-independent regulation of hydroperoxidase I in Escherichia coli. Mol Microbiol 12, 571578.[Medline]
Jishage, M., Iwata, A., Ueda, S. & Ishihama, A. (1996). Regulation of RNA polymerase sigma subunit synthesis in Escherichia coli: intracellular levels of four species of sigma subunit under various growth conditions. J Bacteriol 178, 54475451.[Abstract]
Jishage, M., Kvint, K., Shingler, V. & Nystrom, T. (2002). Regulation of sigma factor competition by the alarmone ppGpp. Genes Dev 16, 12601270.
Lange, R. & Hengge-Aronis, R. (1991). Identification of a central regulator of stationary-phase gene expression in Escherichia coli. Mol Microbiol 5, 4959.[Medline]
Lange, R. & Hengge-Aronis, R. (1994). The cellular concentration of the S subunit of RNA polymerase in Escherichia coli is controlled at the levels of transcription, translation, and protein stability. Genes Dev 8, 16001612.[Abstract]
Lange, R., Fischer, D. & Hengge-Aronis, R. (1995). Identification of transcriptional start sites and the role of ppGpp in the expression of rpoS, the structural gene for the S subunit of RNA polymerase in Escherichia coli. J Bacteriol 177, 46764680.[Abstract]
Latifi, A., Foglino, M., Tanaka, K., Williams, P. & Lazdunski, A. (1996). A hierarchical quorum-sensing cascade in Pseudomonas aeruginosa links the transcriptional activators LasR and RhIR (VsmR) to expression of the stationary-phase sigma factor RpoS. Mol Microbiol 21, 11371146.[Medline]
Light, P. A. (1972). Influence of environment on mitochondrial function in yeast. J Appl Chem Biotechnol 22, 509526.
Liu, X. & Ferenci, T. (1998). Regulation of porin-mediated outer membrane permeability by nutrient limitation in Escherichia coli. J Bacteriol 180, 39173922.
Liu, X. & Ferenci, T. (2001). An analysis of multifactorial influences on the transcriptional control of ompF and ompC porin expression under nutrient limitation. Microbiology 147, 29812989.
Liu, X., Ng, C. & Ferenci, T. (2000). Global adaptations resulting from high population densities in Escherichia coli cultures. J Bacteriol 182, 41584164.
Loewen, P. C., von Ossowski, I., Switala, J. & Mulvey, M. R. (1993). KatF (S) synthesis in Escherichia coli is subject to posttranscriptional regulation. J Bacteriol 175, 21502153.[Abstract]
McCann, M. P., Fraley, C. D. & Matin, A. (1993). The putative factor KatF is regulated posttranscriptionally during carbon starvation. J Bacteriol 175, 21432149.[Abstract]
Miller, J. (1972). Experiments in Molecular Genetics. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.
Muffler, A., Traulsen, D. D., Lange, R. & Hengge-Aronis, R. (1996). Post-transcriptional regulation of the sigma S subunit of RNA polymerase in Escherichia coli. J Bacteriol 178, 16071613.[Abstract]
Muffler, A., Barth, M., Marschall, C. & Hengge-Aronis, R. (1997). Heat shock regulation of S turnover: a role for DnaK and relationship between stress responses mediated by
S and
32 in Escherichia coli. J Bacteriol 179, 445452.[Abstract]
Mulvey, M. R. & Loewen, P. C. (1989). Nucleotide sequence of katF of Escherichia coli suggests KatF protein is a novel sigma transcription factor. Nucleic Acids Res 17, 99799991.[Medline]
Nealson, K. H. (1977). Autoinduction of bacterial luciferase. Arch Microbiol 112, 7379.[Medline]
Neijssel, O. M., De Mattos, M. J. T. & Tempest, D. W. (1996). Growth yield and energy distribution. In Escherichia coli and Salmonella: Cellular and Molecular Biology, pp. 16831692. Edited by F. C. Neidhardt and others. Washington, DC: American Society for Microbiology.
Notley, L. & Ferenci, T. (1996). Induction of RpoS-dependent functions in glucose-limited continuous culture: what level of nutrient limitation induces the stationary phase of Escherichia coli? J Bacteriol 178, 14651468.[Abstract]
Notley-McRobb, L., Death, A. & Ferenci, T. (1997). The relationship between external glucose concentration and cAMP levels inside Escherichia coli: implications for models of phosphotransferase-mediated regulation of adenylate cyclase. Microbiology 143, 19091918.[Abstract]
Notley-McRobb, L., King, T. & Ferenci, T. (2002). rpoS mutations and loss of general stress resistance in Escherichia coli populations as a consequence of conflict between competing stress responses. J Bacteriol 184, 806811.
Nyström, T. (1998). To be or not to be: the ultimate decision of the growth-arrested bacterial cell. FEMS Microbiol Rev 21, 283290.[CrossRef]
O'Neal, C. R., Gabriel, W. M., Turk, A. K., Libby, S. J., Fang, F. C. & Spector, M. P. (1994). RpoS is necessary for both the positive and negative regulation of starvation survival genes during phosphate, carbon, and nitrogen starvation in Salmonella typhimurium. J Bacteriol 176, 46104616.[Abstract]
Peters, J. E., Thate, T. E. & Craig, N. L. (2003). Definition of the Escherichia coli MC4100 genome by use of a DNA array. J Bacteriol 185, 20172021.
Pirt, S. J. (1975). Principles of Microbe and Cell Cultivation. Oxford: Blackwell Scientific Publications.
Schellhorn, H. E. & Stones, V. L. (1992). Regulation of katF and katE in Escherichia coli K-12 by weak acids. J Bacteriol 174, 47694777.[Abstract]
Sitnikov, D. M., Schineller, J. B. & Baldwin, T. O. (1996). Control of cell division in Escherichia coli: regulation of transcription of ftsQA involves both RpoS- and SdiA-mediated autoinduction. Proc Natl Acad Sci U S A 93, 336341.
Sledjeski, D. D., Gupta, A. & Gottesman, S. (1996). The small RNA, DsrA, is essential for the low temperature expression of RpoS during exponential growth in Escherichia coli. EMBO J 15, 39934000.[Abstract]
Smith, H. W. (1965). Observations on the flora of the alimentary tract of animals and factors affecting its composition. J Pathol Bacteriol 89, 95122.
Sperandio, V., Torres, A. G., Giron, J. A. & Kaper, J. B. (2001). Quorum sensing is a global regulatory mechanism in enterohemorrhagic Escherichia coli O157 : H7. J Bacteriol 183, 51875197.
Surette, M. G. & Bassler, B. L. (1998). Quorum sensing in Escherichia coli and Salmonella typhimurium. Proc Natl Acad Sci U S A 95, 70467050.
Surette, M. G., Miller, M. B. & Bassler, B. L. (1999). Quorum sensing in Escherichia coli and Salmonella typhimurium and Vibrio harveyi: a new family of genes responsible for autoinducer production. Proc Natl Acad Sci U S A 96, 16391644.
Teich, A., Meyer, S., Lin, H. Y., Andersson, L., Enfors, S. & Neubauer, P. (1999). Growth rate related concentration changes of the starvation response regulators S and ppGpp in glucose-limited fed-batch and continuous cultures of Escherichia coli. Biotechnol Prog 15, 123129.[CrossRef][Medline]
Townsley, P. M. & Neilands, J. B. (1956). The iron and porphyrin metabolism of Micrococcus lysodeikticus. J Biol Chem 224, 695705.
van Delden, C., Comte, R. & Bally, A. M. (2001). Stringent response activates quorum sensing and modulates cell density-dependent gene expression in Pseudomonas aeruginosa. J Bacteriol 183, 53765384.
Visick, J. E. & Clarke, S. (1997). RpoS- and OxyR-independent induction of HPI catalase at stationary phase in Escherichia coli and identification of rpoS mutations in common laboratory strains. J Bacteriol 179, 41584163.[Abstract]
Wanner, U. & Egli, T. (1990). Microbial growth dynamics in batch culture. FEMS Microbiol Rev 75, 1944.[CrossRef]
Wick, L. M., Weilenmann, H. & Egli, T. (2002). The apparent clock-like evolution of Escherichia coli in glucose-limited chemostats is reproducible at large but not at small population sizes and can be explained with Monod kinetics. Microbiology 148, 28892902.
Winzer, K., Hardie, K. R., Burgess, N. & 8 other authors (2002). LuxS: its role in central metabolism and the in vitro synthesis of 4-hydroxy-5-methyl-3(2H)-furanone. Microbiology 148, 909922.
Yoon, S. H., Han, M. J., Lee, S. Y., Jeong, K. J. & Yoo, J. S. (2003). Combined transcriptome and proteome analysis of Escherichia coli during high cell density culture. Biotechnol Bioeng 81, 753767.[CrossRef][Medline]
Zheng, M., Doan, B., Schneider, T. D. & Storz, G. (1999). OxyR and SoxRS regulation of fur. J Bacteriol 181, 46394643.
Received 20 October 2003;
revised 14 January 2004;
accepted 10 February 2004.
HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
INT J SYST EVOL MICROBIOL | MICROBIOLOGY | J GEN VIROL |
J MED MICROBIOL | ALL SGM JOURNALS |