Fitness and proteome changes accompanying the development of erythromycin resistance in a population of Escherichia coli grown in continuous culture

Abstract

We studied the impact of a sublethal concentration of erythromycin on the fitness and proteome of a continuously cultivated population of Escherichia coli. The development of resistance to erythromycin in the population was followed over time by the gradient plate method and minimum inhibitory concentration (MIC) measurements. We measured the growth rate, standardized efficiency of synthesis of radiolabeled proteins, and translation accuracy of the system. The proteome changes were followed over time in two parallel experiments that differed in the presence or absence of erythromycin. A comparison of the proteomes at each time point (43, 68, and 103 h) revealed a group of unique proteins differing in expression. From all 35 proteins differing throughout the cultivation, only three were common to more than one time point. In the final population, a significant proportion of upregulated proteins was localized to the outer or inner cytoplasmic membranes or to the periplasmic space. In a population growing for more than 100 generations in the presence of antibiotic, erythromycin-resistant bacterial clones with improved fitness in comparison to early resistant culture predominated. This phenomenon was accompanied by distinct changes in protein expression during a stepwise, population-based development of erythromycin resistance.

Keywords: Continuous cultivation system; Escherichia coli; erythromycin; fitness; proteome; resistance.

Figure 1

Comparison of changes in doubling time for Escherichia coli cultures continuously grown in the absence (-◊-) or presence (-□-) of erythromycin (10 μg mL⁻¹).

Figure 2

Erythromycin (0–50 μg mL⁻¹)-gradient plates showing the development of erythromycin resistance in Escherichia coli cultures continuously grown in the presence of the antibiotic (10 μg mL⁻¹). The upper line of plates (Control) shows the samples withdrawn from the culture without antibiotic at 43, 68, and 103 h. The lower line (Ery) shows the samples withdrawn from the culture grown in the presence of erythromycin.

Figure 3

(A) 2-DE proteome map from a 43 h sample of an Escherichia coli culture continuously grown in the presence of erythromycin (10 μg mL⁻¹). The gels with proteins labeled with ³⁵S methionine were exposed to phosphor screens for 4 days and scanned at a 100 μm resolution using a Molecular Imager FX (Bio-Rad). The proteins with expression levels that differ from the control levels are marked with red circles and spot numbers and are listed in Table 1. (B) Graphical presentation showing differences in protein expression in arbitrary units, representing the densities of spots for selected proteins from the 43 h samples of E. coli cultures continuously grown in the absence (green) or presence (red) of erythromycin (10 μg mL⁻¹). The spot numbers correspond to the spots marked on the 2-DE gel (A).

Figure 4

(A) 2-DE proteome map from the 68 h sample of an Escherichia coli culture continuously grown in the presence of erythromycin (10 μg mL⁻¹). The gels with proteins labeled with ³⁵S methionine were exposed to phosphor screens for 4 days and scanned at a 100 μm resolution using Molecular Imager FX (Bio-Rad). The proteins with expression levels that differ from the control are marked with red circles and spot numbers and are listed in Table 2. (B) Graphical presentation showing differences in protein expression in arbitrary units representing the densities of spots for selected proteins from the 68 h samples of E. coli cultures continuously grown in the absence (green) or presence (red) of erythromycin (10 μg mL⁻¹). The spot numbers correspond to those marked on the 2-DE gel (A).

Figure 5

(A) 2-DE proteome map from a 103 h sample of Escherichia coli culture continuously grown in the presence of erythromycin (10 μg mL⁻¹). The gels with proteins labeled with ³⁵S methionine were exposed to phosphor screens for 4 days and scanned at a 100 μm resolution using a Molecular Imager FX (Bio-Rad). The proteins with expression levels that differ from the control levels are marked with red circles and spot numbers and are listed in Table 3. The LuxA subunit of the luciferase reporter system is marked with a blue circle. (B) Graphical presentation showing differences in protein expression in arbitrary units, representing the densities of spots for selected proteins from the 103 h samples of E. coli cultures continuously grown in the absence (green) or presence (red) of erythromycin (10 μg mL⁻¹). The spot numbers correspond to the spots marked on the 2-DE gel (A).

Figure 6

(A) Functional classification of the proteins identified in the 43 h and 103 h samples from Escherichia coli cultures continuously grown in the absence or presence of erythromycin (10 μg mL⁻¹) that differ in expression. (B) Cellular distribution of proteins (cytoplasmic/cell barrier) identified in the 43 h and 103 h samples from E. coli cultures continuously grown in the absence or presence of erythromycin (10 μg mL⁻¹) that differ in expression.

Figure 7

(A) Comparison of changes in the EF-Tu concentration in Escherichia coli cultures continuously grown in the absence (-◊-) or presence (-□-) of erythromycin (10 μg mL⁻¹). (B) Comparison of changes in translation accuracy for E. coli cultures continuously grown in the absence (-◊-) or presence (-□-) of erythromycin (10 μg mL⁻¹). Accuracy is represented as the ratio of relative light units to the culture OD. The light units are generated by the active luciferase enzyme when a stop codon inserted into the proximal portion of the B subunit is accidentally read as a sense codon.