Environmental Microbiology (2001) (9), 588–599 · 2012. 11. 28. · Environmental Microbiology (2001) 3(9), 588–599 Short- and long-term changes in proteome composition and kinetic

Environmental Microbiology (2001) 3(9), 588±599

Short- and long-term changes in proteome compositionand kinetic properties in a culture of Escherichia coliduring transition from glucose-excess to glucose-limited growth conditions in continuous culture andvice versa

Lukas M. Wick,1 Manfredo Quadroni2² and

Thomas Egli1*1Swiss Federal Institute for Environmental Science and

Technology (EAWAG), PO Box 611, UÈ berlandstrasse

133, CH-8600 DuÈbendorf, Switzerland.2Protein Chemistry Laboratory, Swiss Federal Institute of

Technology ZuÈrich, CH-8092 ZuÈrich, Switzerland.

Summary

To investigate the ability of Escherichia coli K12

MG1655 to cope with excess and limitation of a

carbon and energy source, we studied the changes in

kinetic properties and two-dimensional (2D) gel

protein patterns of an E. coli culture. The population

was transferred from glucose-excess batch to glu-

cose-limited continuous culture (D � 0.3 h21), in

which it was cultivated for 500 h (217 generations)

and then transferred back to glucose-excess batch

culture. Two different stages to glucose-limitation

were recognized: a short-term physiological adapta-

tion characterized by a general effort in enhancing

the cell's substrate scavenging ability and mutations

resulting in a population exhibiting increased glu-

cose affinity. Physiological short-term adaptation

to glucose-limitation was achieved by upregulation

of 12 proteins, namely MglB, MalE, ArgT, DppA,

RbsB, YdcS, LivJ (precursor), UgpB (precursor),

AceA, AldA, AtpA and GatY. Eight of these proteins

are periplasmic binding proteins of ABC transporters.

Most of them are not involved in glucose transport

regulons, but rather in chemotaxis and transport of

other substrates, whereas MalE and MglB have

previously been shown to belong to transport

systems important in glucose transport under glu-

cose-limited conditions. Evolution under low glucose

concentration led to an up to 10-fold increase in

glucose affinity (from a Ks of 366 ^ 36 mg l21 at the

beginning to 44 ^ 7 mg l21). The protein pattern of a

`500-h-old' continuous culture showed a highly

increased expression of MglB and MalE as well as

of the regulator protein MalI. When adapted cells

taken from the `500-h-old' continuous culture were

transferred to batch culture, an increased expression

of MalE was observed, compared with cells from un-

adapted batch-grown cells. Otherwise, no significant

changes were observed in the protein pattern of

batch-grown populations before and after 500 h of

evolution in the glucose-limited continuous culture.

Introduction

Escherichia coli is usually not considered to be a typical

environmental bacterium. However, this bacterium is

present in considerable number in many ecosystems

(Savageau, 1983), particularly in warm climate countries,

in which it not only survives but can also grow (Hazen and

Toranzos, 1990). Throughout the world, the presence of

E. coli is used as an indication for faecal contamination of

surface water and drinking water, and therefore knowl-

edge of its behaviour under nutrient-limited conditions in

the aqueous environment is crucial.

In principle, non-sporulating bacteria such as E. coli can

encounter three growth situations. Growth at saturating

substrate concentrations (probably exclusively in the labora-

tory), growth at subsaturating substrate concentrations and

non-growth. Growth at saturating substrate concentrations

and non-growing bacteria have been studied extensively,

particularly in batch and starvation cultures (Groat et al.,

1986; Schultz et al., 1988; Kolter et al., 1993; Weichart et al.,

1993; Hengge-Aronis, 1996; 2000). However, starvation

experiments have strong drawbacks: they are difficult to

perform reproducibly and ill-defined with respect to the

cellular status (see Discussion in Ferenci, 1999). Moreover,

for a considerable part of their lifetime, bacteria in the

environment are neither growing at their maximum rate nor

are they completely starved, but they grow at reduced rates

under nutrient-limited conditions. In most ecosystems the

limiting factor is the availability of organic compounds, which

Q 2001 Blackwell Science Ltd

Received 18 June, 2001; revised 14 August, 2001; accepted 17August, 2001. *For correspondence. E-mail [email protected]; Tel.(141) 1 823 5158; Fax (141) 1 823 5547. ²Present address: Instituteof Biochemistry, University of Lausanne, CH-1066 Epalinges,Switzerland.

results in a limitation in carbon and energy sources. This led

to the proposition that a carbon-limited continuous culture is

the most appropriate experimental tool to investigate the

behaviour of bacterial growth under environmental condi-

tions (Matin, 1979; Moriarty, 1993; Morita, 1993).

Studies concerning the global response of cells to

carbon-limited conditions have revealed a number of

different physiological phenomena; however, the molecular

mechanisms involved here are still poorly investigated.

Certain carbon sources that lead to a two-step growth

pattern under carbon-excess conditions can be used

simultaneously by cells grown under carbon-limited condi-

tions (e.g. in batch cultures containing lactose and glucose,

E. coli first uses glucose, and lactose is consumed only

after the depletion of glucose, whereas in carbon-limited

continuous cultures the two sugars are used simulta-

neously (Silver and Mateles, 1969)). Such cells also exhibit

a higher activity of several catabolic enzymes, even in the

absence of the appropriate carbon sources and are able to

utilize a wide range of sugars without an induction lag

(Matin, 1979; Harder and Dijkhuizen, 1983; Sepers, 1984;

Lendenmann and Egli, 1995).

Indications for a general improvement of glucose affinity as

a result of exposure to low glucose concentrations were

reported for a number of bacteria (marine isolates: Jannasch,

1968; E. coli: Shehata and Marr, 1971; Koch and Wang,

1982; Klebsiella pneumoniae: Rutgers et al., 1987; Cyto-

phaga johnsonii: HoÈfle, 1983). For Cytophaga johnsonii

(HoÈfle, 1983)and forE.coliML30(Sennetal., 1994), the time

course of this improvement has been documented.

Several aspects of the glucose-specific evolution of E.

coli in glucose-limited continuous cultures have been

studied (Helling et al., 1987; Kurlandzka et al., 1991;

Rosenzweig et al., 1994; Senn et al., 1994; Ferenci, 1996;

Notley-McRobb and Ferenci, 1999a;b). Ferenci (1996) has

shown two regulons to be important in glucose uptake: mal/

lam and mgl/gal. Under glucose-limited conditions, E. coli

enhances the outer membrane permeability for glucose

(and other sugars) by higher expression of lamB and uses

the high affinity ABC transport system mgl for glucose

uptake through the inner membrane. As an immediate

physiological response to glucose-limited conditions, these

systems are upregulated by endo-induction (e.g. the

induction of the mal operon by the endogenously synthe-

sized inducer maltotriose during growth on glucose) and

cAMP, whose levels are increased in continuous culture-

grown E. coli compared with cells grown under glucose-

excess conditions (Ferenci, 1996; Notley-McRobb et al.,

1997). During long-term cultivation in continuous culture,

the population evolves and mutations in three regulatory

loci (mlc, mglD/O and malT) are selected that lead to

higher expression of the mal/lam regulon and the mgl

regulon. These mutations result in a better affinity for

glucose (Notley-McRobb and Ferenci, 1999a;b).

In this work we investigated the kinetic and physiolo-

gical behaviour of E. coli during the transition from

glucose-excess to glucose-limited growth conditions in

the continuous culture. In particular, the changes occur-

ring at the proteome level were studied using two-

dimensional (2-D) gel electrophoresis.

Results

Evolution of kinetic properties of E. coli in glucose-limited

continuous culture

During long-term cultivation of E. coli in a glucose-limited

continuous culture at a fixed dilution rate, a continuous

improvement of the culture's glucose scavenging ability

with time was observed, resulting in a steadily decreasing

residual glucose concentration. The residual glucose

concentration only became (apparently) stable after

some 400 h and this improvement of the Monod-Ks was

highly reproducible between different runs (Fig. 1).

In addition to the decrease of the residual glucose

concentration in the continuous culture, cells that were

taken from the `500-h-old' (adapted) continuous culture and

transferred to batch culture with an initial glucose concentra-

tion of 500 mg l21 showed a higher mmax (0.73 ^ 0.01 h21)

than cells from the `40-h-old' continuous culture

(0.63 ^ 0.01 h21). The apparent Ks of the culture can be

calculated based on the modified (including a smin value)

Monod model of microbial growth: m � mmax � (s 2 smin)/

(Ks 1 s 2 smin), in which mmax is the maximum specific

growth rate determined in batch culture, m (� D � 0.3 ^

0.015 h21) is the specific growth rate given at the substrate

concentration s, Ks is the substrate concentration that allows

growth at half the maximum specific growth rate, and smin

(12 ^ 2 mg l21) is the minimally needed substrate concen-

tration for growth (Kovarova et al., 1996). The data in Fig. 1

suggest that, during the 500 h of cultivation (corresponding

to 217 generations), the Ks of the culture improved from the

Fig. 1. Time-course of the residual glucose concentration inglucose-limited continuous cultures of E. coli K12 operated atD � 0.3 h21. Data for four independent runs are shown.

Adaptation of E. coli to glucose limitation 589

Q 2001 Blackwell Science Ltd, Environmental Microbiology, 3, 588±599

initially 366 ^ 36 mg l21 down to 44 ^ 7 mg l21 (average

s-values in the initial 72 h of four independent continuous

cultivations were 345 ^ 21 mg l21, those between 455 and

535 h of three independent continuous cultivations were

43 ^ 4 mg l21). These Ks values correlate very well with the

10-fold increase in glucose affinity reported by Notley-

McRobb and Ferenci (1999b) for derivatives of E. coli K12

after long-term exposure to glucose-limited growth in

continuous culture.

Proteome analysis of cells grown under different growth

conditions using 2-D gel analysis

Cells that had been growing in the glucose-limited con-

tinuous culture for different periods of time (40, 156 and

500 h) and from exponentially growing batch cultures,

before and after exposure to glucose-limited continuous

culture conditions, were sampled. From each condition three

to four individual samples were collected from two different

continuous culture runs. Cells were harvested and subjected

to proteome analysis using 2-D gel electrophoresis. Protein

spots whose intensitywas found tovaryby a factor of three or

more between these different conditions were cut off and the

protein material was analysed by mass spectrometry after

in-gel digestion. The masses of proteolytic peptides were

used to perform Peptide Mass Fingerprinting (Henzel et al.,

1993; James et al., 1993; Mann et al., 1993; Pappin et al.,

1993; Cottrell, 1994). In several cases the identification was

confirmed by tandem mass spectrometry followed by

database searches with raw mass spectrometry MS/MS

spectra (Yates et al., 1995). The results for the protein

identification are summarized in Table 1.

Proteome changes during the transfer from glucose-

excess batch to glucose-limited continuous culture

conditions. Figure 2 shows the 2-D gels of cells harvested

in the log phase of the initial batch culture and cells after

40 h in continuous culture. A comparison of the gels in

Fig. 2 reveals a number of differently expressed proteins.

The spots that differed between the two growth conditions

are numbered in Fig. 2 and listed in Table 1. It is striking

that the majority of the increasingly expressed proteins

were periplasmic binding proteins of different transport

systems. Although glucose was the only energy and

carbon source in the medium, also binding proteins that

are not involved in glucose transport were upregulated

under the glucose-limited conditions compared with the

glucose-excess batch conditions. No spots could be

detected with significantly lower intensities in cells taken

from the continuous culture compared with the batch-

grown cells (note that integrated quantities for every spot

were normalized over the total optical density (OD) of

every gel and are expressed as percentages).

Proteome changes during evolution in glucose-limited

continuous cultivation. The proteins that were found to be

up/downregulated during prolonged cultivation in glucose-

limited conditions are shown in Fig. 3. Most impressive is the

amount to which the two binding proteins (MalE and MglB,

spot no. 8 and 9, respectively, Figs 2 and 3) contributed to

the total gel protein (together <30%). Another impressive

upregulation was observed for MalI (spot no. 13, Fig. 3), a

putative regulator protein, a member of a class of proteins

that are normally expressed at very low levels.

Batch growth of cells cultivated for 500 h in a glucose-

limited continuous culture

Cells taken from the `500-h-old' continuous culture (1 ml)

were used to inoculate a shaking flask containing

Table 1. Proteins analyzed in this study.a

Spot no. pIb mwb Protein (gene) Function

1 5.07 52.1 Aldehyde dehydrogenase A (aldA) Central metabolism2 5.75 57.4 Periplasmic dipeptide transport protein (dppA) Peptide transport3A 5.16 47.5 Isocitrate lyase (aceA) Central metabolism3B 5.16 47.5 Isocitrate lyase (aceA) Central metabolism4 5.98 46.1 Glycerol-3-phosphate-binding periplasmic protein (ugpB) Transport5 6.02 55.2 ATP Synthase alpha chain (atpA) Proton transport/energy6 6.27 39.9 Hypothetical ABC transporter periplasmic binding protein (ydcS) Transport7 5.54 39 Leu/Ile/Val-binding protein (precursor) (livJ) Amino acid transport8 5.22 40.7 Maltose-binding periplasmic protein (malE) Sugar transport9 5.25 33.6 D-Galactose-binding periplasmic protein (mglB) Sugar transport10 5.98 31.1 Tagatose-bisphosphate aldolase (gatY) Catabolism11 6.0 28.5 D-Ribose binding periplasmic protein (rbsB) Sugar transport12 5.22 25.8 Lysine-arginine-ornithine-binding periplasmic protein (argT) Amino acid transport13 6.46 36.6 Maltose regulon, regulatory protein (mal I) Sugar transport

a. The cellular amount of proteins differed at a factor of three or more between cells grown in a glucose-excess-batch and in a glucose-limitedcontinuous culture (1±12, see also Fig. 2), and between cells from a `40-h-old' and a `500-h-old' continuous culture (3B, 7, 8, 9, 11, 12, 13, see alsoFig. 3).b. Molecular weight and pI as calculated from the sequence. Values were found to be consistent (^ 0.15 pI units) with migration on 2-D gels.

590 L. M. Wick, M. Quadroni and T. Egli


0.5 g l21 of glucose. These cells were kept at mmax for

about 30 generations and were then used to inoculate

again a bioreactor. This population exhibited a higher

mmax (0.73 ^ 0.01 h21) than the population taken from

the stock and initially used (0.63 ^ 0.01 h21). Again cells

were harvested for 2-D gel analysis. %Vol. values of MalE

were found to increase from 0.101 ^ 0.025 to

0.928 ^ 0.232 in cells collected from the `500-h-old'

continuous culture compared with the population at the

start (Fig. 4, lanes 1 and 2).

To determine the apparent Ks of these cells, the

bioreactor was switched to continuous cultivation mode,

as described previously, at D � 0.3 h21 and the residual

glucose concentration was determined. Two hours after

the biomass had reached steady-state concentration the

residual glucose concentration was 122 mg l21, indicating

a Ks of 158 mg l21. This demonstrates that the population of

long-term glucose-adapted E. coli ± when re-grown in batch

culture ± still exhibited an improved affinity for glucose.

The percentages of total gel protein of each of the

proteins listed in Table 1, obtained from cells grown under

the different conditions, are shown in Fig. 4. After 500 h of

cultivation in the continuous culture, most proteins that

are not involved in glucose transport were found in lower

cellular relative amounts compared with cells from a 40 h-

old chemostat. This, however, is mainly a consequence of

the enormous increase of the two proteins MalE and MglB

(from about 7% to 30%), resulting in a dilution of all other

proteins. As the spots of MalE and MglB are saturated,

this value is probably still an underestimation.

DppA and GatY are exceptions, because their amount

was also increased in cells cultivated for an extended time

Fig. 2. Proteins expressed in a culture of E. coli K12 during growth under glucose-excess batch (left) and glucose-limited continuous cultureconditions (right) (D � 0.3 h21). The horizontal axes represent pHs of the isoelectric focusing gradients; the vertical axes represent molecularweights in thousands of Da, based on migration in the sodium dodecyl sulphate electrophoretic gel. The proteins that were induced in thecontinuous culture compared with the batch culture by more than a factor of three are labelled. These proteins are listed in Table 1.

Fig. 3. A gel-region of analysed samples from a `40-h-old' continuous culture (40) and a `500-h-old' continuous culture (500). The regionshows proteins the induction levels of which differed by more than a factor of 3 (labelled spots). These proteins are listed in Table 1.



in the glucose-limited continuous culture. In the case of

DppA, one can speculate that higher amounts of this

dipeptide-binding protein are used in recycling peptides

otherwise lost into the medium.

Identification of mutants

As MalE itself has only a negligible affinity to glucose

(Kellermann and Szmelcman, 1974), an overexpression

of malE does not seem to be advantageous under

glucose-limited conditions. Therefore, it was appealing

to assume that the higher MalE levels are an indication of

mal/lamB constituitivity, which was previously shown to be

of advantage in glucose-limited cultures (Notley-McRobb

and Ferenci, 1999a). Therefore, a sample from a `500-h-

old' continuous culture was diluted appropriately, plated

on nutrient agar and nine colonies were picked randomly.

These clones were analysed for their sensitivity towards

methyl-a-glucoside and 2-deoxyglucose to test for mlc

mutations. Two clones showed higher sensitivity towards

2-deoxyglucose, one of these also to methyl-a-glucoside.

In addition, sequencing of the malT gene (coding for a

regulatory protein of the mal operon) was performed.

Three of the other seven clones, which did not show

increased sensitivity to the glucose analogues, had point

mutations in malT: M311I ATG!ATA (in two clones) and

R268H CGC!CAC. The former mutation was also found

in populations evolved under glucose-limited conditions

analysed by Notley-McRobb and Ferenci (1999a). Hence,

five of the nine clones showed regulatory mutations

affecting the mal operon including malE.

Discussion

We have investigated here the response of E. coli to the

transition between two environments that are distinctly

different with respect to the availability of the carbon/

energy source. The best approach to analyse this

response experimentally seems to be the comparison of

cells grown under glucose-excess conditions in batch

Fig. 4. Quantification of protein expression incells harvested from: a batch culture the cellsof which were used as an inoculum for acontinuous culture (1); a batch cultureinoculated with cells from a `500-h-old'continuous culture (2); a `40-h-old' continuousculture (3); a `156-h-old' continuous culture(4); a `500-h-old' continuous culture (5). Thevertical axis represents %Vol. valuescalculated as stated in Experimentalprocedures.



culture and glucose-limited conditions in a continuous

culture at a set growth (dilution) rate. Thereby, it is

inevitable that other physiological parameters will also

change, the most obvious being the growth rate. Except in

the case of AtpA, there were no hints for growth rate-

dependent differences, such as e.g. the variation of the

ribosomal protein concentrations. However, the data of

Maaloe (1979) suggest that distinct physiological changes

resulting from a variation in growth rate occur only at

elevated growth rates (1 h21 to 3 h21) and should not be

observed in our study, in which the cells experienced a

change in specific growth rate from 0.6 h21 to 0.7 h21 in

batch to 0.3 h21 in the continuous culture. Our results

indicate that the response of the cells to glucose-limited

conditions can be divided into two distinct phases, namely

a short-term physiological adaptation and a long-term

selection of a population with increased affinity to glucose.

The following discussion will focus on the difference

between these short- and long-term answers. Although

the 2-D gel technique allows us to get a global picture of a

cell's response to distinct environmental stimuli, this

method allows, at best, detection of approximately 50%

of the proteins expressed. A rough estimate for this study

indicates that we were able to detect about 20% of the

total proteins. This means that it is very unlikely that

complete sets of proteins belonging to a pathway or

regulon will be found. For example, two proteins that have

been shown to be upregulated under glucose-limited

conditions, namely LamB and OmpF (Death et al., 1993;

Liu and Ferenci, 1998), were not detected in this study. In

the following discussion we will therefore take the results

obtained as hints to pathways and regulons activated by

the conditions we applied to the cells.

Effects during short-term adaptation

Eight of the 12 proteins whose expression was signifi-

cantly different in cells grown under glucose-excess and

in cells under glucose-limited conditions are binding

proteins, and, hence, are transport related. Two proteins

are involved in catabolism (AldA, GatY), one (AceA) in the

central metabolism (citric acid cycle/glyoxylate bypass)

and the other (AtpA) in energy generation.

Binding proteins. We observed enhanced expression of

seven out of the approximately 20 binding proteins

belonging to uptake systems for C- and C/N-sources

(carbohydrates and amino acids) (Linton and Higgins,

1998; Paulsen, 1999). Included are all the binding

proteins involved in chemotaxis, i.e. MglB, MalE, RbsB,

DppA (Oliver, 1996). Additionally, one putative ABC

binding protein was found (YdcS). The 2-D gel electro-

phoresis technique does not allow us to draw conclusions

about the behaviour of the remaining binding proteins

involved in uptake of alternative carbon sources that were

not detected in this study. One should keep in mind that

the expression of ABC transport systems is regulated by

different and rather complex mechanisms (Boos and

Lucht, 1996; Ehrmann et al., 1998), and carbon-limitation

is probably only one of several elements involved in the

control of binding protein expression.

Our results clearly underline the importance of ABC

uptake systems for growth in the environment and this is

also supported by the fact that, in E. coli, the family of

ABC transporter genes occupies about 5% of the whole

genome (Linton and Higgins, 1998). It was suggested that

the binding protein is largely responsible for the apparent

affinity of the transport system for its substrate (Miller

et al., 1983). The fact that the increase in the apparent

affinity for glucose from 366 mg l21 to 44 mg l21 is

paralleled by an increase in the binding protein respon-

sible for glucose transport (MglB) is additional strong

evidence for this hypothesis.

From the two binding proteins LivJ and UgpB, only the

precursors and not the mature forms were found on the 2-

D gels. This may be as a result of the overproduction of

other secreted periplasmic proteins (especially MalE and

MglB) leading to competition for signal peptidase.

Enzymes involved in catabolism and central metabolism.

The two enzymes GatY and AldA, the expression of which

was enhanced in the early phase of adaptation, are

enzymes of catabolic pathways. GatY catalyses the

cleavage of tagatose-1,6-bisphosphate to glyceraldehyde-

3-phosphate and dihydroxyacetone-phosphate; hence, it

connects the galactitol degradation to the glycolytic path-

way. It was also found to be induced at low pH-values

(Blankenhorn et al., 1999). AldA is an aldehyde dehydro-

genase known to function on a broad spectrum of

substrates and to respond to multiple regulatory signals,

one of them being cAMP-CRP (Limon et al., 1997).

The isocitrate lyase as the first and most abundant

enzyme in the glyoxylate bypass is normally considered to

be induced only when E. coli is growing on acetate or fatty

acids, and its function is to replenish the dicarboxylic

acids consumed in amino acid biosynthesis. During

growth on glucose, however, the glyoxylate pathway is

not required as the anaplerotic reaction is performed

through the PEP-pathway (Cronan and LaPorte, 1996). It

cannot be deduced from our data whether or not there is

any flux of carbon through the glyoxylate bypass, as this

depends on phosphorylation of IDH by IDH kinase/

phosphatase (Koshland et al., 1985). However, the net-

reaction of both pathways (2 pyruvate 1 oxaloacetate!2

oxaloacetate 1 2CO2) yields 4 NADH, 1 FADH2, but in the

PEP-pathway one molecule of ATP is consumed in

addition. Thus, the balance of the two anaplerotic

sequences suggests the glyoxylate pathway to be more



economical, a fact that might become important under

carbon-limited conditions. An additional reason for the

enhanced level of isocitrate lyase could be that, in this

way, the cell is able to respond more quickly to a sudden

availability of C2-generating substrates such as acetate

and fatty acids.

Two spots on the gels (Fig. 2. spot no. 3 A and 3B)

were identified as AceA. A possible explanation could be

that the protein occurs in different covalently modified

forms, although no evidence was found earlier for

covalent modification of AceA (Koshland et al., 1985). In

addition, proteolysis cannot be ruled out.

AtpA. The atpIBEFHAGDC operon encodes the genes for

the F0F1 proton-translocating ATPase, which couples the

energy derived from oxygen respiration to ATP synthesis

and the ATPase alpha chain constitutes a regulatory

subunit of this enzyme complex. Kasimoglu et al. (1996)

showed that atp gene expression is inversely related to

the growth rate. This agrees well with our finding of an

increased expression of AtpA under the glucose-limited

conditions performed at a m� D of 0.3 h21, whereas

growth under glucose-excess conditions results in a mmax

more than twice as high.

Regulatory network aspects. At least six of the genes

found upregulated under glucose-limited conditions in this

work show either a dependency on cAMP-CRP or have a

putative CRP binding site. These are malE (Ullmann and

Danchin, 1983), mglB, rbsB (Alexander et al., 1993),

ugpB (Su et al., 1991), argT (Botsford and Harman, 1992)

and aldA (Limon et al., 1997). In E. coli, stimulation of the

cAMP-CRP regulon is well known to be an immediate

response to glucose limitation aiming at the use of

alternative carbon sources (Saier et al., 1996). Further-

more, Schultz et al. (1988) found that, from the 30

proteins induced within the first 4±5 h of carbon starva-

tion, 20 of them were not induced in a Dcya or Dcrp strain.

It has recently been confirmed that cAMP levels of E. coli

cells grown in continuous culture at mM glucose concen-

trations are enhanced compared with those grown under

glucose-excess conditions (Notley-McRobb et al., 1997).

However, for many genes controlled by cAMP-CRP,

cAMP is not the only regulating signal (Blum et al.,

1990). For example, in the case of the ABC maltose

transporter, the regulation of the expression of its genes is

far more complex than originally expected (Ehrmann et al.,

1998; Boos and BoÈhm, 2000).

Long-term evolution

MalE and MglB. These two binding proteins, belonging to

the mgl and mal regulon, were the two most abundant

proteins on the 2-D gels of cells from the `500-h-old'

glucose-limited cultures. The binding protein MglB has a

high affinity for glucose (Ferenci, 1996), higher than that

of the PTS-system (Postma et al., 1993), and, at low

glucose concentrations, glucose is taken up by the cells

mainly via this transport system. The fact that this sugar

enters the cells mainly via the (high affinity) galactose

transport system during growth at low glucose concentra-

tions was visualized earlier by pulsing galactose into a

steady-state glucose-limited continuous culture of E. coli

(Egli et al., 1993). A pulse of 1.8 mg l21 of galactose

resulted in an immediate reduction of the glucose

utilization rate of the cells and led to a transient

accumulation of glucose in the continuous culture.

Hence, the proteome analysis performed in this work

suggests that the same changes take place during the

evolution of the population as those reported by Notley-

McRobb and Ferenci (1999a;b) who approached the

question from the genetic side.

Mal I. A strikingly high amount of the regulatory protein

MalI was found in cells harvested from the `500-h-old'

continuous culture. It should be pointed out that such an

increased expression of a regulatory protein might have a

much broader effect than a similarly increased expression

of a catabolic enzyme like GatY. MalI is a repressor of an

operon containing malX and malY. MalI is not dependent

on MalT (Boos and Shuman, 1998), thus excluding the

possibility that the increased levels of MalI might be an

effect of malT-con, as in the case of MalE. The gene malX

is an enzyme II of the phosphotransferase system

recognizing an as yet unidentified substrate. This operon

might thus become important under glucose-limited

conditions, in which transport systems of alternative

carbon sources are generally expressed at higher levels.

The physiological role of MalY is still obscure. It is

supposed to either act on the substrate transported by

MalX or to interact with MalT, when the carbon source

transported by MalX is available. In any case, MalY binds

MalT reversibly, stabilizing the inactive monomeric form of

MalT and in this way interferes with its transcriptional

activator function (Clausen et al., 2000). Therefore, MalY

is not desired when high MalT activity is required.

However, malT-con mutants are largely insensitive to

the effect of MalY (Boos and Shuman, 1998), making the

role of MalI in this context quite puzzling.

Pleiotropic effects. In our experiments, pleiotropic effects

on the protein expression pattern were detected only

during the short-term adaptation but not as a result of

long-term evolution of the culture. The long-term effects

observed were all geared to an improved uptake of

glucose. This is in contrast to the findings of Kurlandzka

et al. (1991) who reported that, for cells kept in glucose-

limited continuous culture for a period of 773 generations,



changes in the protein expression pattern included a

broad spectrum of cellular functions. There are several

reasons to explain these differences. In our study, the

culture was analysed as a whole, whereas Kurlandzka

et al. (1991) analysed single clones isolated from their

culture. Furthermore, after sampling they included a

labelling step with radioactive methionine. In addition,

dilution rate (0.2 vs. 0.3 h21), temperature (30 vs. 378C)

and time of evolution (773 vs. 217 generations) were

different. Therefore, it is probably difficult to compare the

two sets of data. Growth at lower growth rates and lower

temperatures is paralleled with higher RpoS and CspA

levels, respectively (Notley and Ferenci, 1996; Phadtare

et al., 2000), thus enhancing the probabilities for muta-

tions with pleiotropic effects. The most crucial difference,

however, is the one in cultivation time. The study by

Ferenci (1996) and this work have shown that, in the first

200±300 generations of glucose-limited growth, the

bacteria gain strong fitness advantages by improving

their glucose-scavenging ability through some easily

occurring mutations. After this (the largest) potential has

been exploited, further evolution probably involves genes

with pleiotropic effects, as here considerable improve-

ments can also be achieved with few mutations.

Effect of long-term evolution under glucose-limited

conditions to growth during glucose-excess conditions

There are several studies that show that mutants with

increased mmax can be isolated from cultures that had

been exposed to prolonged continuous cultivation

(Dykhuizen and Hartl, 1981; Helling et al., 1987; Weikert

et al., 1997). In addition, in this study a population with a

higher mmax than the starting culture was enriched from

the glucose-limited cultivation. This culture exhibited a

higher amount of MalE protein under glucose-excess

conditions than the starting population (lane 1 and 2 in

Fig. 4), which can be explained by assuming the mal

operon to still be under less tight control than in wild-type

cells. A higher vmax would be the result of such a process.

However, a higher transport rate should not be the reason

for the higher mmax observed for these cells, as transport

rate is not considered to be the limiting process for

determining mmax (Holmes, 1996).

Conclusion

In the short term, i.e. within the first hours of limited

glucose availability, the cells respond to glucose limitation

by making a general effort to increase the ability to

scavenge and utilize different carbon/energy substrates. It

should be pointed out that the synthesis of high-affinity

uptake systems and increased amounts of metabolic

enzymes are not only restricted to enzymes involved in

the metabolism of the supplied growth-limiting carbon

source glucose, but were also observed for a range of

other carbon/energy substrates that were not supplied in

the growth medium. This ability to utilize mixtures of

carbon sources simultaneously and the readiness to use

carbon sources immediately when they become available

is of high importance for survival and competition ability

under environmental conditions, in which the spectrum of

available carbon/energy sources is large and not

restricted to a single one (Egli, 1995). Using this strategy,

the cells not only gain an advantage in a better

exploitation of the diversity of nutrients in an ecosystem

but are also prepared for the immediate utilization of

nutrients that might become suddenly available in their

environment, thus getting a head start compared with

unprepared organisms. Hence, our findings at the

proteome level are a good basis to explain the phenom-

enon of mixed substrate growth at the molecular level

observed earlier under carbon-limited growth conditions

(Matin, 1979; Harder and Dijkhuizen, 1983; Sepers, 1984;

Lendenmann and Egli, 1995).

In the long-term, the population evolved specifically as

a response to the nutrient stress applied (glucose-limited

conditions in this case). Compared with the observed

short-term adaptation, which would make sense in nature,

this long-term adaptation with its massive upregulation of

only a few proteins important in glucose transport is

probably unlikely to be found under the growth conditions

prevailing in ecosystems, as the selection pressure

applied here (growth with only glucose for 217 genera-

tions) appears to be rather artificial.

Experimental procedures

Organism and medium

Wild-type Escherichia coli K12 MG 1655 (from the culturecollection of P. Postma, Amsterdam) was grown at 378C inmineral medium containing the following components: 20 mMKH2PO4, 14 mM NH4Cl, 0.22 mM EDTA, 0.23 mM MgSO4,0.01 mM Na2MoO4, 0.1 mM CaCO3, 0.0075 mM FeCl3,0.025 mM MnCl2, 0.0125 mM ZnO, 0.005 mM CuCl2,0.005 mM CoCl2 and 0.005 mM H3BO3. Glucose was usedas sole source of carbon and energy at concentrations of0.5 g l21 and 1 g l21 for batch conditions and at a concen-tration of 0.1 g l21 for continuous cultures, resulting in aglucose-limited medium supporting a mmax of 0.63 ^ 0.01 h21

for the wild-type strain. Nutrient agar and ECD agar MUG, anagar allowing the identification of E. coli, were obtained fromBiolife.

Chemicals and biochemicals

Hen egg lysozyme was purchased from Fluka, while RNaseA, DNase and Pefabloc SC were obtained from BoehringerMannheim. Resolytes 4±8 were purchased from BDH



Laboratory Supplies. IPG strips were obtained from Amer-sham Pharmacia Biotech. All other chemicals were of thehighest purity available from Fluka. Sequencing grade-modified trypsin was from Promega.

Shake flask cultures

Stock cultures were plated on nutrient agar twice and, fromthe second plate, a single colony was used to inoculate ashake flask containing minimal medium with 0.5 g l21 ofglucose and Na2HPO4/KH2PO4 (72 mM/22 mM) to buffer thesystem at pH 7. The cells were pregrown for approximately50 generations under excess glucose conditions. To keep thebacteria constantly growing at mmax, the cells were alwaystransferred into fresh medium before they reached the latelog phase. The resulting culture was used to inoculate abioreactor.

Cultivation in the bioreactor

Batch and continuous culturing of cells was performed in labscale bioreactors (MBR) with a working volume of 1.5 l. Theaeration was at a rate sufficient to ensure always . 90% airsaturation, impeller speed was set at 1000 r.p.m., and the pHwas maintained at 7 ^ 0.05 by automatic addition of 0.5 MKOH/0.5 M NaOH.

Batch cultivation. Using cells pregrown in shake flasks, thebioreactor was inoculated to an OD546 of 0.015, in thepresence of 1 g l21 glucose. When the cells reached anOD546 of 0.4 (early mid-log phase), 1.2 l of the culture washarvested for the subsequent 2-D gel analysis by pouring thecultivation liquid directly in centrifuge tubes containingapproximately 30% (v/v) of ice. The reactor was refilled withfresh medium containing 0.1 g l21 glucose, drained andrefilled again. The cells remaining in the reactor served as theinoculum to start the subsequent glucose-limited continuousculture. The resulting initial OD546 was approximately 0.08.

Cultivation in glucose-limited continuous culture. When thecells inoculated in a reactor as described reached an OD546

of 0.09, the medium flow (containing 0.1 g l21 of glucose)was started at a dilution rate of 0.3 h21. The culture was thenoperated in the continuous mode for 500 h. The purity of thecontinuous culture was tested by plating onto nutrient agarand ECD agar MUG. Additionally, the culture was checkeddirectly using fluorescence in situ hybridization (FISH) with aCy3-labelled probe targetting 16S rRNA (Microsynth). Basedon the report of Fuchs et al. (1998), potential sequences withlow evolutionary conservation and high fluorescence inten-sities were checked for highest specificity at the webpage ofthe Ribosomal Database Project from Michigan StateUniversity (Maidak et al., 2001) and the following sequenceproofed as the most suitable: 5 0-actttactcccttcctccc-3 0.

Sampling for 2-D gel analysis

Samples were poured directly from the continuous cultureinto 50 ml Falcon tubes containing 15±20 ml ice, to quicklycool the samples. Cells were harvested by centrifugation for

10 min at 3500 g at 48C. The cell pellets were suspended andtransferred to Eppendorf tubes using ice-cold low-salt buffer(3 mM KCl, 68 mM NaCl, 10 mM Na2HPO4/KH2PO4, pH 7).Cells were pelleted again in a cooled microfuge and stored at2208C until gel electrophoresis was performed.

Glucose analysis

Glucose concentrations in continuous cultures were deter-mined as described previously (Senn et al., 1994).

Sample preparation for 2-D gel electrophoresis

Aliquots of frozen E. coli cells, containing approx. 1010 cellseach, were thawed slowly on ice and resuspended in 250 mlof low-salt buffer (3 mM KCl, 68 mM NaCl, 10 mM Na2HPO4/KH2PO4, pH 7). After centrifugation for 5 min at 5000 g, thepellet was resuspended in 80 ml of lysis buffer (10 mM Tris/HCl pH 7.5, 1.5 mM MgCl2, 10 mM KCl, 0.5 mM 1,4-dithio-DL-threitol, 0.5 mM Pefabloc SC) and 5 ml of a hen egglysozyme solution (10 mg ml21) were added. The sampleswere incubated for 20 min at 378C and subsequentlysonicated for 2 � 10 s with a tip sonicator. To each sample3 ml of a RNase A solution (10 mg ml21) and 5 ml of a DNasesolution (10 mg ml21) were added and the samples wereincubated for 15 min at room temperature to performdigestion of nucleic acids. To each sample 500 ml ofsolubilization buffer (containing 0.9 M Urea, 4% (w/v) 3-((3-cholamidopropyl)dimethyl-ammonio)-1-propanesulphonate(CHAPS), 65 mM 1,4-dithio-DL-threitol, 0.8% Resolytes 4±8,0.01% (w/v) bromophenol blue) were then added. Aftermixing for 1 min, solubilization was allowed to proceed for15 min at room temperature. Solid urea was added to eachtube until saturation was reached, the samples were thencentrifuged at 13 000 g for 30 min and the supernatants wereused for 2-D gel electrophoresis.

2-D gel electrophoresis

Gels were run in batches of 20 to ensure reproducibility.Samples (500 ml) were used to rehydrate overnight1 � 18 cm immobiline strips pI 4±7. Isoelectric focusingwas performed at a maximum voltage of 3500 V, until aV � h count of 70 000 was reached. Equilibration andtransfer to the second dimension were done as described(Hochstrasser et al., 1988). SDS±polyacrylamide gel electro-phoresis (PAGE) was performed on 12% acrylamide gelsusing the Iso-Dalt system (Large Scale Biology Corp.), andCoomassie blue protein staining was according to Schaggerand von Jagow (1987).

Image analysis

Gels were scanned on a Personal Densitometer (MolecularDynamics). Image files were analysed using MELANIE II 2-DPAGE, Version 2.2 (Bio Rad). The spot's density normalizedto the overall density of the gel proteins (referred to as %Vol.values) was taken to compare corresponding protein spotsbetween different gels. Measurement errors were determined



the following way: the intensities of 34 protein spots, whichwere regarded as constant throughout all five conditions,were quantified and the standard deviation for each wascalculated. The spots were then classified into threecategories: those with %Vol. values between 0.03 and 0.1,those between 0.1 and 1, and those higher than 1. Theaverage percentage standard deviations for each categorywere 30%, 25% and 21% respectively.

In-gel proteolysis and extraction

Coomassie blue-stained protein spots of interest wereexcised and washed for 20 min in double-distilled waterfollowed by destaining in 30% acetonitrile, 100 mM ammo-nium bicarbonate pH 8.0 for 30 min at 378C under agitation.The excised gel slices were dried in a rotary evaporator andre-hydrated with 20 ml of 100 mM ammonium bicarbonatepH 8.0 containing 0.3 mg of sequencing-grade modifiedtrypsin. Digestion was carried out overnight at roomtemperature. Peptides were extracted by washing the gelsonce with 150 ml of double distilled water and twice with250 ml of 40% acetonitrile, 0.1% trifluoroacetic acid underagitation.

MALDI-TOF mass spectrometry (MS) analysis

Peptide extracts from protein spots were pooled, dried andredissolved in 4 ml of 0.2% trifluoroacetic acid. Aliqots of there-dissolved extracts (0.3 ml) were mixed with an equalvolume of a-hydroxycinnamic acid (20 mg ml21 in 50%acetonitrile) and spotted on the MALDI target plate. Afterdrying, the sample crystals were washed twice with 5 ml ofice-cold 0.1% trifluoroacetic acid and dried again. MALDI-TOF spectra were acquired on a Perseptive BiosystemsVoyager Elite mass spectrometer operated in reflector mode.Samples were analysed in delayed extraction reflector modeusing an accelerating voltage of 20 kV, a pulse delay time of150 ns, a grid voltage of 60% and a guide wire voltage of0.05%. Spectra were accumulated for 32 laser shots.

ESI-MS/MS analysis

Samples for electrospray MS analysis were desalted on acapillary column packed with 0.2 ml of POROS R2 resin asdescribed (Wilm et al., 1996). MS/MS sequencing wasperformed on a Finnigan MAT LC-Q ion trap mass spectro-meter. The desalted digest in 70% methanol and 1% aceticacid was loaded into a homemade nanospray tip andelectrosprayed into the mass spectrometer at a flow rate of0.2 ml min21 using a syringe pump. Peptide ions weremanually selected for fragmentation, which was carried outusing a relative collision energy of 35±60 units for MH1 ionsand 20±35 units for MH21 ions.

Database searches

MALDI-TOF spectra were examined manually to extract themonoisotopic molecular weights and the obtained sets ofmasses were used to search E. coli protein sequences in the

SWISSPROT database using the program PEPTIDENT (avail-able on-line on the site http://www.expasy.ch). Onlysequences matching for at least three peptides resultingfrom complete digestion were taken into account, togetherwith information on the molecular weight and the pI of theprotein as derived from the gel.

Raw MS/MS spectra were used to search with the programSEQUEST (Yates et al., 1995), an E. coli protein sequencedatabase compiled in-house. Only sequences that werematched by at least two distinct peptide searches with DCnvalues greater than 0.15 were considered as a positiveidentification.

Identification of mlc and malT mutations

Sequencing of the malT gene and sensitivity tests towardsthe PtsG and PtsM substrates, methyl-a-glucoside and 2-deoxyglucose, respectively, were performed as described inNotley-McRobb and Ferenci (1999a;b).

Acknowledgements

The authors are indebted to Paolo Landini for careful reading of

the manuscript and to A. J. B. Zehnder for support throughout this

study. Furthermore, we kindly acknowledge the financial support

of this research by a grant from the Swiss National ScienceFoundation (Grant number 31±50885.97).

References

Alexander, D.M., Damerau, K., and St. John, A.C. (1993)

Carbohydrate uptake genes in Escherichia coli are inducedby carbon starvation. Curr Microbiol 27: 335±340.

Blankenhorn, D., Phillips, J., and Slonczewski, J.L. (1999) Acid-

and base-induced proteins during aerobic and anaerobic

growth of Escherichia coli revealed by two-dimensional gel

electrophoresis. J Bacteriol 181: 2209±2216.

Blum, P.H., Jovanovich, S.B., McCann, M.P., Schultz, J.E.,Lesley, S.A., Burgess, R.R., and Matin, A. (1990) Cloning and

in vivo and in vitro regulation of cyclic AMP-dependent carbon

starvation genes from Escherichia coli. J Bacteriol 172: 3813±3820.

Boos, W., and Lucht, J.M. (1996) Periplasmic binding protein-

dependent ABC transporters. In Escherichia coli and Salmo-

nella Neidhardt, F.C. (ed.) Washington, DC: American Society

for Microbiology Press, pp. 1175±1209.

Boos, W., and Shuman, H. (1998) Maltose/maltodextrin systemof Escherichia coli: transport, metabolism, and regulation.

Microbiol Mol Biol Rev 62: 204±229.

Boos, W., and BoÈhm, A. (2000) Learning new tricks from an old

dog: MalT of the Escherichia coli maltose system is part of acomplex regulatory network. Trends Genet 16: 404±409.

Botsford, J.L., and Harman, J.G. (1992) Cyclic AMP in

prokaryotes. Microbiol Rev 56: 100±122.

Clausen, T., Schlegel, A., Peist, R., et al. (2000) X-ray structure

of MalY from Escherichia coli: a pyridoxal 5 0-phosphate-

dependent enzyme acting as a modulator in mal geneexpression. EMBO J 19: 831±842.

Cottrell, J.S. (1994) Protein identification by peptide mass

fingerprinting. Pept Res 7: 115±124.

Cronan, J.E., Jr, and LaPorte, D. (1996) Tricarboxylic acid cycle

and glyoxylate bypass. In Escherichia coli and Salmonella



Neidhardt, F.C. (ed.) Washington, DC: American Society forMicrobiology Press, pp. 206±216.

Death, A., Notley, L., and Ferenci, T. (1993) Derepression of

LamB protein faciliates outer membrane permeation of

carbohydrates into Escherichia coli under conditions of nutrientstress. J Bacteriol 175: 1475±1483.

Dykhuizen, D., and Hartl, D. (1981) Evolution of competitive

ability in Escherichia coli. Evolution 35: 581±594.

Egli, T. (1995) The ecological and physiological significance of

the growth of heterotrophic microorganisms with mixtures ofsubstrates. In Advances in Microbial Ecology Jones, J.G. (ed.)

New York: Plenum Press, pp. 305±386.

Egli, T., Lendenmann, U., and Snozzi, M. (1993) Kinetics of

microbial growth with mixtures of carbon sources. Antonie van

Leeuwenhoek 63: 289±298.

Ehrmann, M., Ehrle, R., Hofmann, E., Boos, W., and SchloÈsser,

A. (1998) The ABC maltose transporter. Mol Microbiol 29:

685±694.

Ferenci, T. (1996) Adaptation to life at micromolar nutrient

levels: the regulation of Escherichia coli glucose transportby endoinduction and cAMP. FEMS Microbiol Rev 18: 301±

317.

Ferenci, T. (1999) Regulation by nutrient limitation. Curr Opin

Microbiol 2: 208±213.

Fuchs, B.M., Wallner, G., Beisker, W., Schwippl, I., Ludwig, W.,

and Amann, R. (1998) Flow cytometric analysis of the in situ

accessibility of Escherichia coli 16SrRNA for fluorescently

labeled oligonucleotide probes. Appl Environ Microbiol 64:4973±4982.

Groat, R.G., Schultz, J.E., Zychlinsky, E., Bockman, A., and

Matin, A. (1986) Starvation proteins in Escherichia coli: kinetics

of synthesis and role in starvation survival. J Bacteriol 168:486±493.

Harder, W., and Dijkhuizen, L. (1983) Physiological response to

nutrient limitation. Ann Rev Microbiol 37: 1±23.

Hazen, T.C., and Toranzos, G.A. (1990) Tropical source water. In

Drinking Water Microbiology McFeters, G.A. (ed.) New York:Springer Verlag, pp. 32±53.

Helling, R., Vargas, C., and Adams, J. (1987) Evolution of

Escherichia coli during growth in a constant environment.

Genetics 116: 349±358.

Hengge-Aronis, R. (1996) Regulation of gene expression duringentry into stationary phase. In Escherichia coli and Salmonella

Neidhardt, F.C. (ed.) Washington, DC: American Society for

Microbiology Press, pp. 1497±1512.

Hengge-Aronis, R. (2000) The general stress response in

Escherichia coli. In Bacterial Stress Response Storz, G., andHengge-Aronis, R. (eds) Washington, DC: American Society


Henzel, W.J., Billeci, T.M., Stults, J.T., Wong, S.C., Grimley, C.,

and Watanabe, C. (1993) Identifying proteins from two-dimensional gels by molecular mass searching of peptide

fragments in protein sequence databases. Proc Natl Acad Sci

USA 90: 5011±5015.

Hochstrasser, D.F., Harrington, M.G., Hochstrasser, A.C., Miller,M.J., and Merril, C.R. (1988) Methods for increasing the

resolution of two-dimensional protein electrophoresis. Anal

Biochem 173: 424±435.

HoÈfle, M.G. (1983) Long-term changes in chemostat cultures of

Cytophaga johnsonae. Appl Environ Microbiol 46: 1045±1053.

Holmes, H. (1996) Flux analysis and control of the central

metabolic pathways in Escherichia coli. FEMS Microbiol Rev

19: 85±116.

James, P., Quadroni, M., Carafoli, E., and Gonnet, G. (1993)

Protein identification by mass profile fingerprinting. BiochemBiophys Res Commun 195: 58±64.

Jannasch, H.W. (1968) Growth characteristics of heterotrophic

bacteria in seawater. J Bacteriol 95: 722±723.

Kasimoglu, E., Park, S.J., Malek, J., Tseng, C.P., and Gunsalus,

R. (1996) Transcriptional regulation of the proton-translocatingATPase (atpIBEFHAGDC) operon of Escherichia coli: control

by cell growth rate. J Bacteriol 178: 5563±5567.

Kellermann, O., and Szmelcman, S. (1974) Active transport of

maltose in Escherichia coli K12. Involvement of a `periplasmic'maltose binding protein. Eur J Biochem 47: 139±149.

Koch, A.L., and Wang, C.H. (1982) How close to the theoretical

diffusion limit do bacterial uptake systems function? Arch

Microbiol 131: 36±42.

Kolter, R., Siegele, D.A., and Tormo, A. (1993) The stationaryphase of the bacterial life cycle. Annu Rev Microbiol 47: 855±

874.

Koshland, D.E., Jr, Walsh, K., and LaPorte, D.C. (1985)

Sensitivity of metabolic fluxes to covalent control. Curr Top

Cell Regul 27: 13±22.

Kovarova, K., Zehnder, A.J.B., and Egli, T. (1996) Temperature

dependent growth kinetics of Escherichia coli ML 30 in

glucose-limited continuous culture. J Bacteriol 178: 4530±

4539.

Kurlandzka, A., Rosenzweig, R.F., and Adams, J. (1991)

Identification of adaptive changes in an evolving population of

Escherichia coli: the role of changes with regulatory and highly

pleiotropic effects. Mol Biol Evol 8: 261±281.

Lendenmann, U., and Egli, T. (1995) Is Escherichia coli growingin glucose-limited chemostat culture able to utilize other sugars

without lag? Microbiology 141: 71±78.

Limon, A., Hidalgo, E., and Aguilar, J. (1997) The aldA gene of

Escherichia coli is under the control of at least threetranscriptional regulators. Microbiology 143: 2085±2095.

Linton, K.J., and Higgins, C.F. (1998) The Escherichia coli ATP-

binding cassette (ABC) proteins. Mol Microbiol 28: 5±13.

Liu, X., and Ferenci, T. (1998) Regulation of porin-mediated outer

membrane permeability by nutrient limitation in Escherichiacoli. J Bacteriol 180: 3917±3922.

Maaloe, O. (1979) Regulation of the protein-synthesizing

machinery ± ribosomes, tRNA, factors and so on. In Biological

Regulation and Development Goldberger, R.F. (ed.) New York:Plenum Press, pp. 487±542.

Maidak, B.L., Cole, J.R., Lilburn, T.G., et al. (2001) The RDP-II

(Ribosomal Database Project). Nucleic Acids Res 29: 173±

174.

Mann, M., Hojrup, P., and Roepstorff, P. (1993) Use of massspectrometric molecular weight information to identify proteins

in sequence databases. Biol Mass Spectrom 22: 338±345.

Matin, A. (1979) Microbial regulatory mechanisms at low nutrient

concentrations as studied in chemostat. In Strategies of

Microbial Life in Extreme Environments: Report of the DahlemWorkshop on Strategy of Life in Extreme Environments Shilo,

M. (ed.) Weinheim: Verlag Chemie, pp. 323±339.

Miller, D.M.I., Olson, J.S., Pflugrath, J.W., and Quiocho, F.A.

(1983) Rates of ligand binding to periplasmic proteins involvedin bacterial transport and chemotaxis. J Biol Chem 258:

13665±13672.

Moriarty, D.J.W. (1993) Bacterial growth and starvation in aquatic

environments. In Starvation in Bacteria Kjelleberg, S. (ed.) NewYork: Plenum Press, pp. 25±53.

Morita, R.Y. (1993) Bioavailability of energy and the starvation

state. In Starvation in Bacteria Kjelleberg, S. (ed.) New York:

Plenum Press, pp. 1±23.



Notley, L., and Ferenci, T. (1996) Induction of RpoS-dependentfunctions in glucose-limited continous culture: what level of

nutrient limitation induces the stationary phase of Escherichia

coli. J Bacteriol 178: 1465±1468.

Notley-McRobb, L., and Ferenci, T. (1999a) The generation ofmultiple co-existing mal-regulatory mutations through poly-

genic evolution in glucose-limited populations of Escherichia

coli. Env Microbiol 1: 45±52.Notley-McRobb, L., and Ferenci, T. (1999b) Adaptive mgl-

regulatory mutations and genetic diversity evolving in glu-

cose-limited Escherichia coli populations. Env Microbiol 1: 33±

43.Notley-McRobb, L., Death, A., and Ferenci, T. (1997) The

relationship between external glucose concentration and

cAMP levels inside Escherichia coli: implications for models

of phosphotransferase-mediated regulation of adenylatecyclase. Microbiology 143: 1909±1918.

Oliver, D.B. (1996) Periplasm. In Escherichia coli and Salmonella

Neidhardt, F.C. (ed.) Washington DC: American Society for

Microbiology Press, pp. 88±103.Pappin, D.J.C., Hojrup, P., and Bleasby, A.J. (1993) Rapid

identification of proteins by peptide-mass fingerprinting. Curr

Biol 3: 327±332.Paulsen, I.T. (1999) Transport Proteins in Escherichia coli

K12. URL http://www.biology.ucsd.edu/,ipaulsen/transport/

index.html.

Phadtare, S., Yamanaka, K., and Inouye, M. (2000) The coldshock response. In Bacterial Stress Responses Storz, G., and

Hengge-Aronis, R. (eds) Washington, DC: American Society


Postma, P.W., Lengeler, J.W., and Jacobson, G.R. (1993)Phosphoenolpyruvate: carbohydrate phosphotransferase sys-

tems of bacteria. Microbiol Rev 57: 543±594.

Rosenzweig, R., Sharp, R., Treves, D., and Adams, J. (1994)Microbial evolution in a simple unstructured environment: genetic

differentiation in Escherichia coli. Genetics 137: 903±917.

Rutgers, M., Teixeira de Mattos, M.J., Postma, P.W., and van

Dam, K. (1987) Establishment of the steady state in glucose-limited chemostat cultures of Klebsiella pneumoniae. J Gen

Microbiol 133: 445±453.

Saier, M.H., Jr, Ramseier, T.M., and Reizer, J. (1996) Regulation

of carbon utilization. In Escherichia coli and SalomnellaNeidhardt, F.C. (ed.) Washington, DC: American Society for

Microbiology Press, pp. 1325±1343.

Savageau, M.A. (1983) Escherichia coli habitats, cell types,and molecular mechanisms of gene control. Am Nat 122:

732±744.

Schagger, H., and von Jagow, G. (1987) Tricine-sodium dodecyl

sulfate-polyacrylamide gel electrophoresis for the separation ofproteins in the range from 1 to 100 kDa. Anal Biochem 166:

368±379.

Schultz, J.E., Latter, G.I., and Matin, A. (1988) Differentialregulation by cyclic AMP of starvation protein synthesis in

Escherichia coli. J Bacteriol 170: 3903±3909.

Senn, H., Lendenmann, U., Snozzi, M., Hamer, G., and Egli, T.

(1994) The growth of Escherichia coli in glucose-limitedchemostat cultures: a re-examination of the kinetics. Biochim

Biophys Acta 1201: 424±436.

Sepers, A.B.J. (1984) The uptake capacity for organic com-

pounds of two heterotrophic bacterial strains at carbon-limitedgrowth. Zeitschrift allgem Mikrobiol 24: 261±267.

Shehata, T.E., and Marr, A.G. (1971) Effect of nutrient

concentration on the growth of Escherichia coli. J Bacteriol

107: 210±216.Silver, R.S., and Mateles, R.I. (1969) Control of mixed-substrate

utilization in continuous cultures of Escherichia coli. J Bacteriol

97: 535±543.Su, T.Z., Schweizer, H.P., and Oxender, D.L. (1991) Carbon-

starvation induction of the ugp operon, encoding the binding

protein-dependent sn-glycerol-3-phosphate transport system in

Escherichia coli. Mol Gen Genet 230: 28±32.Ullmann, A., and Danchin, A. (1983) Role of cyclic AMP in

bacteria. Adv Cyclic Nucleotide Res 15: 1±53.

Weichart, D., Lange, R., Henneberg, N., and Hengge-Aronis, R.

(1993) Identification and characterization of stationary phase-inducible genes in Escherichia coli. Mol Microbiol 10: 407±420.

Weikert, C., Sauer, U., and Bailey, J.E. (1997) Use of glycerol-

limited, long-term chemostat for isolation of Escherichia colimutants with improved physiological properties. Microbiology

143: 1567±1574.

Wilm, M., Shevchenko, A., Houthaeve, T., Breit, S., Schweigerer,

L., Fotsis, T., and Mann, M. (1996) Femtomole sequencing ofproteins from polyacrylamide gels by nano- electrospray mass

spectrometry. Nature 379: 466±469.

Yates, J.R., III, Eng, J.K., and McCormack, A.L. (1995) Mining

genomes: correlating tandem mass spectra of modified andunmodified peptides to sequences in nucleotide databases.

Anal Chem 67: 3202±3210.



Documents

Environmental Microbiology (2001) (9), 588–599 · 2012. 11. 28. · Environmental Microbiology (2001) 3(9), 588–599 Short- and long-term changes in proteome composition and kinetic