E. coli CM2555 has a genetic defect that leads to its chloramphenicol sensitivity in the presence of cat.

Strain CM2555 has been described as one of the series of ilv⁺ dnaA mutants (CM2555 bears the dnaA508 allele) otherwise isogenic to strain CM732 (4). However, using a set of bacterial strains and plasmids (Table ​(Table1),1), we found that the chloramphenicol sensitivity of strain CM2555 bearing the cat gene is not caused by the dnaA508 allele or the combination of ilv⁺ and dnaA508 alleles (data not shown). Thus, we conclude that strain CM2555 must contain another, previously unidentified mutation(s) which makes it sensitive to chloramphenicol, despite the presence of the cat gene.

Expression of cat gene in strain CM2555.

Densitometric analyses of electropherograms after SDS-PAGE and of Western blots with anti-CAT serum revealed that CM2555/pBR328 produces at least three times more CAT than the parental strain, CM732/pBR328 (data not shown). To test whether an increased level of CAT plays a role in the mechanisms of the chloramphenicol sensitivity of CM2555, we used plasmid pJKP1 bearing the cat gene under control of an inducible promoter, p_lac, together with another plasmid, pJMH1, harboring the lacI^q allele. Addition of isopropyl-β-d-thiogalactopyranoside (IPTG) to a final concentration of 1 mM resulted in CAT levels (in both CM2555 and CM732 hosts) that corresponded to about 90% of that found in CM732/pBR328 (data not shown). Contrary to the results obtained in control experiments, the increased efficiency of cat expression correlated with the faster growth of strain CM2555/pJKP1/pJMH1 in a chloramphenicol-containing medium only up to IPTG concentrations of 0.1 mM, and a further increase in the efficiency of cat expression resulted in the inhibition of bacterial growth (Fig. ​(Fig.1).1). Therefore, we suggest that the efficiency of expression of the cat gene may be important for the chloramphenicol sensitivity of strain CM2555.

Open in a separate window

FIG. 1

Effect of intracellular amount of CAT on bacterial growth in the presence of chloramphenicol. E. coli CM732 and CM2555 harboring pJKP1 and pJMH1 were grown in Luria-Bertani liquid medium. cat gene expression from the p_lac-cat fusion was induced overnight, and then in the course of the experiment expression was maintained by the presence of various amounts of IPTG. Chloramphenicol was added when the optical density at 575 nm was 0.1. Doubling times were determined by monitoring bacterial growth (optical density at 575 nm) in the presence and absence of chloramphenicol (34 μg/ml). Symbols: open circles, CM732/pJKP1/pJMH1 grown without chloramphenicol; closed circles, CM732/pJKP1/pJMH1 grown in the presence of chloramphenicol; open rectangles, CM2555/pJKP1/pJMH1 grown without chloramphenicol; closed rectangles, CM2555/pJKP1/pJMH1 grown in the presence of chloramphenicol.

Activity of CAT protein in strain CM2555.

We measured the acetylation of chloramphenicol and calculated that the total CAT activity per 1 mg of cellular proteins in strain CM2555/pBR328 was higher than that in analogous samples prepared from the parental strain, CM732/pBR328 (1.20 CAT activity units in CM2555/pBR328 versus 0.73 CAT activity units in CM732/pBR328). Moreover, CAT reveals biological activity in both CM2555/pBR328 and CM732/pBR328, as the addition of chloramphenicol (up to 34 μg/ml) to cultures of these strains had no significant effect on protein synthesis efficiency (data not shown).

Sensitivity to chloramphenicol of strain CM2555 expressing cat results from decreased levels of acetyl-CoA.

Since CAT catalyzes the transfer of the acetyl group from acetyl-CoA to a chloramphenicol molecule, we measured the levels of acetyl-CoA in strain CM2555 and strain CM732 bearing plasmid pBR328 in the absence and presence of chloramphenicol. We found that the residual level of acetyl-CoA, without the addition of the antibiotic, was slightly lower in CM2555/pBR328 relative to that in CM732/pBR328 (the level of acetyl-CoA in CM2555/pBR328 was about 90% of that found in CM732/pBR328). However, after the addition of chloramphenicol, the level of acetyl-CoA was roughly constant in CM732/pBR328, while it gradually decreased in CM2555/pBR328 (Fig. ​(Fig.2A2A and B).

Open in a separate window

FIG. 2

Intracellular amount of acetyl-CoA as affected by chloramphenicol in E. coli CM732/pBR328 (A) and CM2555/pBR328 (B) and the effect of sodium acetate on bacterial growth in the presence of chloramphenicol (C). In the experiments whose results are depicted in panels A and B, strains harboring pBR328 were grown in Luria-Bertani liquid medium. Chloramphenicol (34 μg/ml) was added when the optical density at 575 nm was 0.1 (time zero). Samples were withdrawn every 5 min after addition of the antibiotic and were then assayed for acetyl-CoA. For each strain, the amount of acetyl-CoA at time zero was taken to be 1, and the relative concentration was calculated. In the experiments whose results are depicted in panel C, E. coli CM732 and CM2555 harboring pBR328 were grown in Luria-Bertani liquid medium supplemented with different amounts of sodium acetate. Chloramphenicol was added when the optical density at 575 nm was 0.1. The doubling times of the cultures were determined by monitoring bacterial growth (optical density at 575 nm) in the presence and absence of chloramphenicol (34 μg/ml). The values presented are means from three experiments. In all cases the standard deviation was below 10%. Symbols: open circles, CM732/pBR328 grown without chloramphenicol; closed circles, CM732/pBR328 grown in the presence of chloramphenicol; open rectangles, CM2555/pBR328 grown without chloramphenicol; closed rectangles, CM2555/pBR328 grown in the presence of chloramphenicol.

To test whether a decrease in the level of acetyl-CoA in strain CM2555/pBR328 is responsible for the sensitivity of this strain to chloramphenicol, we measured the growth rates of bacterial cultures in the presence of sodium acetate. Sodium acetate can be used as an alternative source of acetyl-CoA production in cells (6). It was reported previously (12) that cat-expressing E. coli mutants with mutations in aceE or aceF, which cannot make acetyl-CoA from pyruvate but which can make acetyl-CoA from sodium acetate, are resistant to chloramphenicol only when sodium acetate is supplied. Thus, if the hypothesis stated above were true, we should have observed improved growth of CM2555/pBR328 in the presence of sodium acetate. Indeed, we found that the presence of sodium acetate dramatically improved the growth of CM2555/pBR328 in the medium containing chloramphenicol (Fig. ​(Fig.2C).2C). We also found that strain CM2555 could be transformed by different plasmids bearing the cat gene when selection was performed on plates containing both chloramphenicol (34 μg/ml) and 0.15% sodium acetate (data not shown). Neither the pBR328 copy number nor the amount of CAT in CM2555 cells was affected by the presence of sodium acetate in the medium (data not shown). These results strongly support the hypothesis that decreased levels of acetyl-CoA in cat-expressing CM2555 cells in the presence of chloramphenicol result in the sensitivity of the bacterium to this antibiotic.