Limits...
Genomic analysis reveals distinct concentration-dependent evolutionary trajectories for antibiotic resistance in Escherichia coli.

Mogre A, Sengupta T, Veetil RT, Ravi P, Seshasayee AS - DNA Res. (2014)

Bottom Line: A second class of mutations, recovered only during evolution in higher sublethal concentrations of the antibiotic, deleted the C-terminal end of the ATP synthase shaft.This mutation confers basal-level resistance to kanamycin while showing a strong growth defect in the absence of the antibiotic.In conclusion, the early dynamics of the development of resistance to an aminoglycoside antibiotic is dependent on the levels of stress (concentration) imposed by the antibiotic, with the evolution of less costly variants only a matter of time.

View Article: PubMed Central - PubMed

Affiliation: National Centre for Biological Sciences, Tata Institute of Fundamental Research, GKVK, Bellary Road, Bangalore, Karnataka 560065, India aswin@ncbs.res.in aalapbm@ncbs.res.in.

Show MeSH

Related in: MedlinePlus

Population structure of evolving populations. (A) E. coli populations grown in 4-kan are more homogeneous in composition and comprised mostly of fit variants at 24 h. Isolates were classified into several groups based on the stationary-phase OD600 reached in 4-kan (tolerance to kanamycin) and 0-kan (growth in the absence of kanamycin; see inset in B). Plotted are counts of different types of isolates. Wild-type-like isolates (black bars) have been arbitrarily defined for the analysis as non-tolerant but fit isolates (falling in the OD range indicated in the inset of B) and predominate at the start of the experiment (0 h.). Group 1 (dark grey) isolates are tolerant but sick variants. Group 2 (grey) and 3 (light grey) are tolerant and fit variants (see inset of Figure 1B for definition of groups based on stationary-phase OD600 in plain LB and 4-kan and Supplementary Figure S5 for all plots related to this figure as well as plots for 0-kan control populations in which the wild-type-like colonies predominate throughout the course of the experiment). The counts change with time during the course of the batch evolution experiment (0 h, 24 h and P1). Results of the first replicate are displayed in the left panel and the second replicate in the right panel. (B) Populations grown in 8-kan are composed of multiple subpopulations, with the sick variants dominating at 24 h. Thus, tolerant but sick variants predominate in the first batch growth in 8-kan, and subsequent passaging results in an increase in number of the fit variants. (C) 1,000 independent growth simulations were performed for three growth regimes. This plot shows the distributions of the total population size (a reflection of the number of iterations or time) at which the ratio of the sick mutant to the fit mutant falls below 10%. Distributions are represented by box plots. The box represents the inter-quartile range and the line within the box indicates the median value. The maximum and the minimum value (1.5 times the inter-quartile range), excluding the outliers, is represented by the whiskers. The points represent the outliers. Probabilities of division per iteration are as follows. For the slow growth regime (green), µwt = 0.01, µsick = 0.03 and µfit = 0.1; for the intermediate growth regime (blue), µwt = 0.03, µsick = 0.1 and µfit = 0.3; for the fast growth regime (red), µwt = 0.1, µsick = 0.3 and µfit = 1. (D) A scatter plot of the ratio of the number of sick to fit mutants as a function of the population size (equivalent to the number of iterations or time) for an example simulation from (C).
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4263303&req=5

DSU032F2: Population structure of evolving populations. (A) E. coli populations grown in 4-kan are more homogeneous in composition and comprised mostly of fit variants at 24 h. Isolates were classified into several groups based on the stationary-phase OD600 reached in 4-kan (tolerance to kanamycin) and 0-kan (growth in the absence of kanamycin; see inset in B). Plotted are counts of different types of isolates. Wild-type-like isolates (black bars) have been arbitrarily defined for the analysis as non-tolerant but fit isolates (falling in the OD range indicated in the inset of B) and predominate at the start of the experiment (0 h.). Group 1 (dark grey) isolates are tolerant but sick variants. Group 2 (grey) and 3 (light grey) are tolerant and fit variants (see inset of Figure 1B for definition of groups based on stationary-phase OD600 in plain LB and 4-kan and Supplementary Figure S5 for all plots related to this figure as well as plots for 0-kan control populations in which the wild-type-like colonies predominate throughout the course of the experiment). The counts change with time during the course of the batch evolution experiment (0 h, 24 h and P1). Results of the first replicate are displayed in the left panel and the second replicate in the right panel. (B) Populations grown in 8-kan are composed of multiple subpopulations, with the sick variants dominating at 24 h. Thus, tolerant but sick variants predominate in the first batch growth in 8-kan, and subsequent passaging results in an increase in number of the fit variants. (C) 1,000 independent growth simulations were performed for three growth regimes. This plot shows the distributions of the total population size (a reflection of the number of iterations or time) at which the ratio of the sick mutant to the fit mutant falls below 10%. Distributions are represented by box plots. The box represents the inter-quartile range and the line within the box indicates the median value. The maximum and the minimum value (1.5 times the inter-quartile range), excluding the outliers, is represented by the whiskers. The points represent the outliers. Probabilities of division per iteration are as follows. For the slow growth regime (green), µwt = 0.01, µsick = 0.03 and µfit = 0.1; for the intermediate growth regime (blue), µwt = 0.03, µsick = 0.1 and µfit = 0.3; for the fast growth regime (red), µwt = 0.1, µsick = 0.3 and µfit = 1. (D) A scatter plot of the ratio of the number of sick to fit mutants as a function of the population size (equivalent to the number of iterations or time) for an example simulation from (C).

Mentions: What are the characteristics of the bacterial population resulting from the above-described growth in the antibiotic, in terms of its ability to grow in the presence or absence of kanamycin? For this, we first grew cells—produced by P0 cultures in antibiotic-free, 4-kan or 8-kan media—on LB agar plates. Individual colonies so obtained were then checked for their ability to grow in the absence of kanamycin (plain LB, 0-kan), moderately sublethal levels of kanamycin (4-kan) and lethal levels of kanamycin (20-kan). We note here that this experimental strategy is limited in the sense that the growth experiments performed to assess fitness also contribute to further evolution and selection. However, these steps were performed in the same manner for 4-kan and 8-kan populations, and therefore, the results thus obtained enable a direct comparison between the two antibiotic concentrations. Control colonies derived from all the populations immediately after inoculation displayed wild-type characteristics: good growth in 0-kan, poor growth at 4-kan and no growth at 20-kan (Supplementary Fig. S5A–F, 0 h.). Similarly, colonies derived from populations not treated with kanamycin retained this response throughout the course of the experiment (Supplementary Fig. S5A and S5D). Most colonies derived from 4-kan populations grew well in LB and in 4-kan; in addition, they were more resistant to 20-kan than the wild type (Fig. 2A, and Supplementary Fig. S5B and S5E). In contrast, a proportion of colonies derived from 8-kan populations, though more resistant to the antibiotic than the wild type, showed severe growth defects (Fig. 2B, and Supplementary Fig. S5C and S5F). The proportion of colonies with a growth defect varied between trials (Fig. 2B, compare group-1 and -2 colonies in 8-kan 1 and 2 at 24 h, between the two replicates). After another passage in the same concentration of the antibiotic (P1), both 4-kan and 8-kan cultures produced antibiotic-resistant colonies with little growth defect (Fig. 2A and 2B, group-3 colonies increase at P1).Figure 2.


Genomic analysis reveals distinct concentration-dependent evolutionary trajectories for antibiotic resistance in Escherichia coli.

Mogre A, Sengupta T, Veetil RT, Ravi P, Seshasayee AS - DNA Res. (2014)

Population structure of evolving populations. (A) E. coli populations grown in 4-kan are more homogeneous in composition and comprised mostly of fit variants at 24 h. Isolates were classified into several groups based on the stationary-phase OD600 reached in 4-kan (tolerance to kanamycin) and 0-kan (growth in the absence of kanamycin; see inset in B). Plotted are counts of different types of isolates. Wild-type-like isolates (black bars) have been arbitrarily defined for the analysis as non-tolerant but fit isolates (falling in the OD range indicated in the inset of B) and predominate at the start of the experiment (0 h.). Group 1 (dark grey) isolates are tolerant but sick variants. Group 2 (grey) and 3 (light grey) are tolerant and fit variants (see inset of Figure 1B for definition of groups based on stationary-phase OD600 in plain LB and 4-kan and Supplementary Figure S5 for all plots related to this figure as well as plots for 0-kan control populations in which the wild-type-like colonies predominate throughout the course of the experiment). The counts change with time during the course of the batch evolution experiment (0 h, 24 h and P1). Results of the first replicate are displayed in the left panel and the second replicate in the right panel. (B) Populations grown in 8-kan are composed of multiple subpopulations, with the sick variants dominating at 24 h. Thus, tolerant but sick variants predominate in the first batch growth in 8-kan, and subsequent passaging results in an increase in number of the fit variants. (C) 1,000 independent growth simulations were performed for three growth regimes. This plot shows the distributions of the total population size (a reflection of the number of iterations or time) at which the ratio of the sick mutant to the fit mutant falls below 10%. Distributions are represented by box plots. The box represents the inter-quartile range and the line within the box indicates the median value. The maximum and the minimum value (1.5 times the inter-quartile range), excluding the outliers, is represented by the whiskers. The points represent the outliers. Probabilities of division per iteration are as follows. For the slow growth regime (green), µwt = 0.01, µsick = 0.03 and µfit = 0.1; for the intermediate growth regime (blue), µwt = 0.03, µsick = 0.1 and µfit = 0.3; for the fast growth regime (red), µwt = 0.1, µsick = 0.3 and µfit = 1. (D) A scatter plot of the ratio of the number of sick to fit mutants as a function of the population size (equivalent to the number of iterations or time) for an example simulation from (C).
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4263303&req=5

DSU032F2: Population structure of evolving populations. (A) E. coli populations grown in 4-kan are more homogeneous in composition and comprised mostly of fit variants at 24 h. Isolates were classified into several groups based on the stationary-phase OD600 reached in 4-kan (tolerance to kanamycin) and 0-kan (growth in the absence of kanamycin; see inset in B). Plotted are counts of different types of isolates. Wild-type-like isolates (black bars) have been arbitrarily defined for the analysis as non-tolerant but fit isolates (falling in the OD range indicated in the inset of B) and predominate at the start of the experiment (0 h.). Group 1 (dark grey) isolates are tolerant but sick variants. Group 2 (grey) and 3 (light grey) are tolerant and fit variants (see inset of Figure 1B for definition of groups based on stationary-phase OD600 in plain LB and 4-kan and Supplementary Figure S5 for all plots related to this figure as well as plots for 0-kan control populations in which the wild-type-like colonies predominate throughout the course of the experiment). The counts change with time during the course of the batch evolution experiment (0 h, 24 h and P1). Results of the first replicate are displayed in the left panel and the second replicate in the right panel. (B) Populations grown in 8-kan are composed of multiple subpopulations, with the sick variants dominating at 24 h. Thus, tolerant but sick variants predominate in the first batch growth in 8-kan, and subsequent passaging results in an increase in number of the fit variants. (C) 1,000 independent growth simulations were performed for three growth regimes. This plot shows the distributions of the total population size (a reflection of the number of iterations or time) at which the ratio of the sick mutant to the fit mutant falls below 10%. Distributions are represented by box plots. The box represents the inter-quartile range and the line within the box indicates the median value. The maximum and the minimum value (1.5 times the inter-quartile range), excluding the outliers, is represented by the whiskers. The points represent the outliers. Probabilities of division per iteration are as follows. For the slow growth regime (green), µwt = 0.01, µsick = 0.03 and µfit = 0.1; for the intermediate growth regime (blue), µwt = 0.03, µsick = 0.1 and µfit = 0.3; for the fast growth regime (red), µwt = 0.1, µsick = 0.3 and µfit = 1. (D) A scatter plot of the ratio of the number of sick to fit mutants as a function of the population size (equivalent to the number of iterations or time) for an example simulation from (C).
Mentions: What are the characteristics of the bacterial population resulting from the above-described growth in the antibiotic, in terms of its ability to grow in the presence or absence of kanamycin? For this, we first grew cells—produced by P0 cultures in antibiotic-free, 4-kan or 8-kan media—on LB agar plates. Individual colonies so obtained were then checked for their ability to grow in the absence of kanamycin (plain LB, 0-kan), moderately sublethal levels of kanamycin (4-kan) and lethal levels of kanamycin (20-kan). We note here that this experimental strategy is limited in the sense that the growth experiments performed to assess fitness also contribute to further evolution and selection. However, these steps were performed in the same manner for 4-kan and 8-kan populations, and therefore, the results thus obtained enable a direct comparison between the two antibiotic concentrations. Control colonies derived from all the populations immediately after inoculation displayed wild-type characteristics: good growth in 0-kan, poor growth at 4-kan and no growth at 20-kan (Supplementary Fig. S5A–F, 0 h.). Similarly, colonies derived from populations not treated with kanamycin retained this response throughout the course of the experiment (Supplementary Fig. S5A and S5D). Most colonies derived from 4-kan populations grew well in LB and in 4-kan; in addition, they were more resistant to 20-kan than the wild type (Fig. 2A, and Supplementary Fig. S5B and S5E). In contrast, a proportion of colonies derived from 8-kan populations, though more resistant to the antibiotic than the wild type, showed severe growth defects (Fig. 2B, and Supplementary Fig. S5C and S5F). The proportion of colonies with a growth defect varied between trials (Fig. 2B, compare group-1 and -2 colonies in 8-kan 1 and 2 at 24 h, between the two replicates). After another passage in the same concentration of the antibiotic (P1), both 4-kan and 8-kan cultures produced antibiotic-resistant colonies with little growth defect (Fig. 2A and 2B, group-3 colonies increase at P1).Figure 2.

Bottom Line: A second class of mutations, recovered only during evolution in higher sublethal concentrations of the antibiotic, deleted the C-terminal end of the ATP synthase shaft.This mutation confers basal-level resistance to kanamycin while showing a strong growth defect in the absence of the antibiotic.In conclusion, the early dynamics of the development of resistance to an aminoglycoside antibiotic is dependent on the levels of stress (concentration) imposed by the antibiotic, with the evolution of less costly variants only a matter of time.

View Article: PubMed Central - PubMed

Affiliation: National Centre for Biological Sciences, Tata Institute of Fundamental Research, GKVK, Bellary Road, Bangalore, Karnataka 560065, India aswin@ncbs.res.in aalapbm@ncbs.res.in.

Show MeSH
Related in: MedlinePlus