Journal of Antimicrobial Chemotherapy (1990) 26, 307-317
A comparative study on the inhibitory actions of chloramphenicol, thiamphenicol and some fluorinated derivatives M. Cannon', S. Harford* and J. Darks'
Chloramphenicol, thiamphenicol and three fluorinated derivatives, Sen 24893, Sch 25298 and Sen 25393, were studied with respect to inhibition of the growth of selected bacterial strains and cell-free translation systems. Thiamphenicol was the least potent inhibitor in the former experiments, but behaved similarly to chloramphenicol and Sch 25298 in the latter, thereby displaying selective inhibition of prokaryotic protein synthesis. Thiamphenicol and Sch 25298 were shown to be like chloramphenicol in inhibiting peptidyl transferase activity specifically on 70 S ribosomes, but the antibiotics bound to their common ribosomal-receptor site with different efficiencies in the order chloramphenicol > thiamphenicol > Sch 25298. Selected bacterial strains highly resistant to chloramphenicol and thiamphenicol because of chloramphenicol acetyltransferase production were, in contrast, highly sensitive to inhibition by the fluorinated antibiotics. Thus Sch 24893, Sch 25298 and Sch 25393 may have important uses in veterinary and clinical medicine.
Introduction
Chloramphenicol prevents bacterial protein synthesis by inhibiting peptidyl transferase specifically on 70 S ribosomes. It shows no activity against eukaryotic (80 S) ribosomes (Gale et al., 1981). Although its precise mode of action remains to be established, chloramphenicol, along with its analogues thiamphenicol and tevenel, apparently decreases the catalytic rate constant of peptidyl transferase (Drainas, Kalpaxis & Coutsogeorgopoulos, 1987). Chloramphenicol is an effective broad-spectrum antibiotic, but its clinical use has been curtailed, since it can cause serious toxic side-effects in humans and resistance to the compound readily develops in bacteria (Shaw, 1983). In an attempt to combat these problems many derivatives of chloramphenicol have been synthesized, some of which have displayed significant antibacterial activity (Pongs, 1979). Particularly promising are three fluorinated derivatives of chloramphenicol and thiamphenicol designated Sch 24893, Sch 25298 and Sch 25393 (Nagabhushan et al., 1980). These compounds have favourable MICs for a range of bacterial isolates (Schafer et al., 1980; Syriopoulou et al., 1981) and, in many such cases, are more effective inhibitors than either chloramphenicol or thiamphenicol. More importantly, these fluorinated antibiotics have considerable activity against bacteria that are resistant to chloramphenicol and thiamphenicol, thus indicating that they are not substrates for chloramphenicol acetyltransferase (CAT) (Neu, Fu & Kung, 1980; Schafer et al., 1980; Syriopoulou et al., 1981). 307 0305-7453/90/090307 +11 $02.00/0
© 1990 The British Society for Antimicrobial Chemotherapy
Downloaded from http://jac.oxfordjournals.org/ at Russian Archive on January 2, 2014
'Department of Biochemistry, King's College London, London WC2R2LS, UK; Department of Biochemistry, University of Wisconsin, Madison, Wisconsin 53706, USA; 'Unite de Ginie Microbiologique, Institut Pasteur, 75724 Paris, France
b
308
M. Cannon tf aL
Because of the potential importance of the fluorinated derivatives as antibacterial agents we wished to characterize the inhibitory action(s) of these antibiotics more precisely. Accordingly, this paper describes a detailed comparison of chloramphenicol, thiamphenicol, Sch 24893, Sch 25298 and Sch 25393 with respect to their inhibition of the growth of selected Escherichia coli and Staphylococcus aureus strains, together with the effects of these compounds on the kinetic properties of selected CAT-variants encoded by R plasmids. In addition, Sch 25298 was characterized as an inhibitor of 70 S ribosomes of E. coli and its inhibitory actions were compared with those of chloramphenicol and thiamphenicol in cell-free translation systems.
Antibiotics The fluorinated antibiotics Sch 24893, Sch 25298 and Sch 25393 were supplied by Schering Corp., Bloomficld, New Jersey, USA. Chloramphenicol and thiamphenicol were obtained from Sigma Chemical Company, London, UK. Bacteria Bacterial strains were standard isolates of E. coli and derived mutants from our laboratory collection. The yeast strain was Saccharomyces cerevisiae strain A224A (a, leu~2, can"1). Strains of E. coli were maintained on nutrient agar plates (Difco) and grown at 37°C in either nutrient broth no. 2 (Difco) or minimal medium (Cannon, 1967) supplemented with 1 g/1 vitamin-free casamino acids (Difco). Sacch. cerevisiae was maintained on YEPD plates (Gritz et al., 1982) and cultured at 30°C in either liquid YEPD medium or complete synthetic medium (Udem & Warner, 1972). Preparation of rabbit reticulocytes These were prepared as described by Smith, Cannon & Cundliffe (1975). Determination of MICs MICs were determined as described by Ericsson & Sherris (1971). Measurement of RNA and protein synthesis in growing bacterial cultures RNA and protein synthesis were measured in growing bacterial cultures as described previously (Cannon, Davies & Jimenez, 1973) by determining, respectively, uptake of [S,^3!!] undine (specific activity 35 Ci/mmol) and L-[4,5-3H] leucine (specific activity 120Ci/mmol) (Amersham International, Amersham, Bucks, UK) into intact cells of bacteria. Cell-free translation systems Cell-free protein synthesis, as directed by either endogenous messenger RNA or polyuridylic acid (polyU) in extracts of E. coli, was assayed as described by Perzynski et al. (1979). Conditions for cell-free protein synthesis using reticulocyte extracts were as
Downloaded from http://jac.oxfordjournals.org/ at Russian Archive on January 2, 2014
Materials and methods
Inhibition by cMoramphenkoI derivatives
309
described previously (Cannon, Smith & Carter, 1976; Carter, Cannon & Smith, 1976). Conditions using yeast extracts were described by Carter, Cannon & Jimenez (1980). The puromycin reaction This reaction was assayed as described by Cannon (1968), using 70 S or 80 S ribosomes derived, respectively, from E. coli and reticulocyte extracts programmed in protein synthesis by endogenous messenger RNA (Carter et al., 1976; Perzynski et al., 1979). Ribosomc suspensions were assayed for radioactivity using a LKfi mini-/? liquidscintillation counter with aquasol-2 (NEN Research Products, Boston, Massachusetts, USA) as scintillant (70% counting efficiency for I 4 Q .
This was carried out as described by Fernandez-Mufioz & Vazquez (1973), using 70 S or 80 S ribosomes derived, respectively, from E. coli and Sacch. cerevisiae or rabbit reticulocytes. The reaction takes place between puromycin and a 3'-terminal fragment of transfer RNA bearing N-acetyl leucine (CACCA-Leu-Ac) under the catalytic action of peptidyl transferase. Binding of chloramphenicol to ribosomes Binding of chloramphenicol to 70 S ribosomes was assayed using a modification of the method of Contreras, Barbacid & Vazquez (1974). Ribosomes from E. coli were radiolabelled in a cell-free extract (Perzynski et al., 1979) by L-[4,5-3H] leucine and were used to assay the binding of D-threo[dichloroacetyl-l-l*C\ chloramphenicol (specific activity 54 mCi/mmol) (Amersham International). The binding reactions were carried out in the presence of increasing concentrations of chloramphenicol, thiamphenicol or Sch25298. The 70S particles were sedimented by ultracentrifugation for 2 h at 105,000 g, resuspended and assayed for their content of I4C- and 3H-label. Kinetic studies on CAT Induction of CAT activity in Staph. aureus was carried out as described elsewhere (Shaw, 1975). Kinetic properties of plasmid-encoded CAT- activities against chloramphenicol, thiamphenicol and the three fluorinated derivatives were determined by standard methods.
Results Antibacterial action of the antibiotics The MICs of the five antibiotics were first determined for selected bacterial strains sensitive or resistant to chloramphenicol (Table I). All the drugs inhibited bacterial growth, with thiamphenicol being the least potent of the antibiotics. Reduced MICs were observed for an E. coli mutant that had reduced permeability towards chloramphenicol and an E. coli strain with increased resistance to chloramphenicol at the ribosome level (M. Nomura, personal communication). The strains incorporating
Downloaded from http://jac.oxfordjournals.org/ at Russian Archive on January 2, 2014
The fragment reaction
310
M. Cannon et aL
Table I. Inhibition of selected strains of E. coli and Staph. aweus by chloramphenicol (CM) thiamphenicol (TM) and the fluorinated derivatives Sch 24893, Sch 25298 and Sch 25393
Strain
Resistance type
Plasmid — — — JR66 S-a R387
— permeability ribosomal CAT type I CAT type II CAT type HI
— — pGG CAT-Staph
Sch 25393
CM
TM
2 32 32 256 256 512
32 128 512 1024 512 1024
4 64 32 8 4 8
4 32 32 4 4 8
8 32 64 8 8 16
2 64
8 512
2 2
2 2
4 8
different CAT activities were highly resistant to both chloramphenicol and thiamphenicol, but remained highly sensitive to the three fluorinated antibiotics. The inhibitory potencies of chloramphenicol, thiamphenicol and Sch 25298 were next compared against a sensitive E. coli strain. In two different liquid media, both chloramphenicol and Sch 25298 allowed only 20-30% residual growth at a drug concentration of 2 mg/1. In contrast, a thiamphenicol concentration of 25 mg/1 was required to produce a similar effect (Table II), thus confirming the trends indicated by the MICs (Table I). The antibiotics did not inhibit Sacch. cerevisiae growing in liquid culture (results not shown). Sch 25298, like chloramphenicol and thiamphenicol, is therefore a selective inhibitor of prokaryotic cells. Table II. Inhibition of E. coli growth by chloramphenicol, Sch 25298 and thiamphenicol Drug Chloramphenicol
Sch 25298
Thiamphenicol
Final concentration (mg/1) 0-10 0-25 0-50 1-00 2-00 0-10
% inhibition of growth" rich medium minimal medium
14 17 36 61 79 8
025
12
0-50 1-00 2-00 5-00 1
A comparative study on the inhibitory actions of chloramphenicol, thiamphenicol and some fluorinated derivatives M. Cannon', S. Harford* and J. Darks'
Chloramphenicol, thiamphenicol and three fluorinated derivatives, Sen 24893, Sch 25298 and Sen 25393, were studied with respect to inhibition of the growth of selected bacterial strains and cell-free translation systems. Thiamphenicol was the least potent inhibitor in the former experiments, but behaved similarly to chloramphenicol and Sch 25298 in the latter, thereby displaying selective inhibition of prokaryotic protein synthesis. Thiamphenicol and Sch 25298 were shown to be like chloramphenicol in inhibiting peptidyl transferase activity specifically on 70 S ribosomes, but the antibiotics bound to their common ribosomal-receptor site with different efficiencies in the order chloramphenicol > thiamphenicol > Sch 25298. Selected bacterial strains highly resistant to chloramphenicol and thiamphenicol because of chloramphenicol acetyltransferase production were, in contrast, highly sensitive to inhibition by the fluorinated antibiotics. Thus Sch 24893, Sch 25298 and Sch 25393 may have important uses in veterinary and clinical medicine.
Introduction
Chloramphenicol prevents bacterial protein synthesis by inhibiting peptidyl transferase specifically on 70 S ribosomes. It shows no activity against eukaryotic (80 S) ribosomes (Gale et al., 1981). Although its precise mode of action remains to be established, chloramphenicol, along with its analogues thiamphenicol and tevenel, apparently decreases the catalytic rate constant of peptidyl transferase (Drainas, Kalpaxis & Coutsogeorgopoulos, 1987). Chloramphenicol is an effective broad-spectrum antibiotic, but its clinical use has been curtailed, since it can cause serious toxic side-effects in humans and resistance to the compound readily develops in bacteria (Shaw, 1983). In an attempt to combat these problems many derivatives of chloramphenicol have been synthesized, some of which have displayed significant antibacterial activity (Pongs, 1979). Particularly promising are three fluorinated derivatives of chloramphenicol and thiamphenicol designated Sch 24893, Sch 25298 and Sch 25393 (Nagabhushan et al., 1980). These compounds have favourable MICs for a range of bacterial isolates (Schafer et al., 1980; Syriopoulou et al., 1981) and, in many such cases, are more effective inhibitors than either chloramphenicol or thiamphenicol. More importantly, these fluorinated antibiotics have considerable activity against bacteria that are resistant to chloramphenicol and thiamphenicol, thus indicating that they are not substrates for chloramphenicol acetyltransferase (CAT) (Neu, Fu & Kung, 1980; Schafer et al., 1980; Syriopoulou et al., 1981). 307 0305-7453/90/090307 +11 $02.00/0
© 1990 The British Society for Antimicrobial Chemotherapy
Downloaded from http://jac.oxfordjournals.org/ at Russian Archive on January 2, 2014
'Department of Biochemistry, King's College London, London WC2R2LS, UK; Department of Biochemistry, University of Wisconsin, Madison, Wisconsin 53706, USA; 'Unite de Ginie Microbiologique, Institut Pasteur, 75724 Paris, France
b
308
M. Cannon tf aL
Because of the potential importance of the fluorinated derivatives as antibacterial agents we wished to characterize the inhibitory action(s) of these antibiotics more precisely. Accordingly, this paper describes a detailed comparison of chloramphenicol, thiamphenicol, Sch 24893, Sch 25298 and Sch 25393 with respect to their inhibition of the growth of selected Escherichia coli and Staphylococcus aureus strains, together with the effects of these compounds on the kinetic properties of selected CAT-variants encoded by R plasmids. In addition, Sch 25298 was characterized as an inhibitor of 70 S ribosomes of E. coli and its inhibitory actions were compared with those of chloramphenicol and thiamphenicol in cell-free translation systems.
Antibiotics The fluorinated antibiotics Sch 24893, Sch 25298 and Sch 25393 were supplied by Schering Corp., Bloomficld, New Jersey, USA. Chloramphenicol and thiamphenicol were obtained from Sigma Chemical Company, London, UK. Bacteria Bacterial strains were standard isolates of E. coli and derived mutants from our laboratory collection. The yeast strain was Saccharomyces cerevisiae strain A224A (a, leu~2, can"1). Strains of E. coli were maintained on nutrient agar plates (Difco) and grown at 37°C in either nutrient broth no. 2 (Difco) or minimal medium (Cannon, 1967) supplemented with 1 g/1 vitamin-free casamino acids (Difco). Sacch. cerevisiae was maintained on YEPD plates (Gritz et al., 1982) and cultured at 30°C in either liquid YEPD medium or complete synthetic medium (Udem & Warner, 1972). Preparation of rabbit reticulocytes These were prepared as described by Smith, Cannon & Cundliffe (1975). Determination of MICs MICs were determined as described by Ericsson & Sherris (1971). Measurement of RNA and protein synthesis in growing bacterial cultures RNA and protein synthesis were measured in growing bacterial cultures as described previously (Cannon, Davies & Jimenez, 1973) by determining, respectively, uptake of [S,^3!!] undine (specific activity 35 Ci/mmol) and L-[4,5-3H] leucine (specific activity 120Ci/mmol) (Amersham International, Amersham, Bucks, UK) into intact cells of bacteria. Cell-free translation systems Cell-free protein synthesis, as directed by either endogenous messenger RNA or polyuridylic acid (polyU) in extracts of E. coli, was assayed as described by Perzynski et al. (1979). Conditions for cell-free protein synthesis using reticulocyte extracts were as
Downloaded from http://jac.oxfordjournals.org/ at Russian Archive on January 2, 2014
Materials and methods
Inhibition by cMoramphenkoI derivatives
309
described previously (Cannon, Smith & Carter, 1976; Carter, Cannon & Smith, 1976). Conditions using yeast extracts were described by Carter, Cannon & Jimenez (1980). The puromycin reaction This reaction was assayed as described by Cannon (1968), using 70 S or 80 S ribosomes derived, respectively, from E. coli and reticulocyte extracts programmed in protein synthesis by endogenous messenger RNA (Carter et al., 1976; Perzynski et al., 1979). Ribosomc suspensions were assayed for radioactivity using a LKfi mini-/? liquidscintillation counter with aquasol-2 (NEN Research Products, Boston, Massachusetts, USA) as scintillant (70% counting efficiency for I 4 Q .
This was carried out as described by Fernandez-Mufioz & Vazquez (1973), using 70 S or 80 S ribosomes derived, respectively, from E. coli and Sacch. cerevisiae or rabbit reticulocytes. The reaction takes place between puromycin and a 3'-terminal fragment of transfer RNA bearing N-acetyl leucine (CACCA-Leu-Ac) under the catalytic action of peptidyl transferase. Binding of chloramphenicol to ribosomes Binding of chloramphenicol to 70 S ribosomes was assayed using a modification of the method of Contreras, Barbacid & Vazquez (1974). Ribosomes from E. coli were radiolabelled in a cell-free extract (Perzynski et al., 1979) by L-[4,5-3H] leucine and were used to assay the binding of D-threo[dichloroacetyl-l-l*C\ chloramphenicol (specific activity 54 mCi/mmol) (Amersham International). The binding reactions were carried out in the presence of increasing concentrations of chloramphenicol, thiamphenicol or Sch25298. The 70S particles were sedimented by ultracentrifugation for 2 h at 105,000 g, resuspended and assayed for their content of I4C- and 3H-label. Kinetic studies on CAT Induction of CAT activity in Staph. aureus was carried out as described elsewhere (Shaw, 1975). Kinetic properties of plasmid-encoded CAT- activities against chloramphenicol, thiamphenicol and the three fluorinated derivatives were determined by standard methods.
Results Antibacterial action of the antibiotics The MICs of the five antibiotics were first determined for selected bacterial strains sensitive or resistant to chloramphenicol (Table I). All the drugs inhibited bacterial growth, with thiamphenicol being the least potent of the antibiotics. Reduced MICs were observed for an E. coli mutant that had reduced permeability towards chloramphenicol and an E. coli strain with increased resistance to chloramphenicol at the ribosome level (M. Nomura, personal communication). The strains incorporating
Downloaded from http://jac.oxfordjournals.org/ at Russian Archive on January 2, 2014
The fragment reaction
310
M. Cannon et aL
Table I. Inhibition of selected strains of E. coli and Staph. aweus by chloramphenicol (CM) thiamphenicol (TM) and the fluorinated derivatives Sch 24893, Sch 25298 and Sch 25393
Strain
Resistance type
Plasmid — — — JR66 S-a R387
— permeability ribosomal CAT type I CAT type II CAT type HI
— — pGG CAT-Staph
Sch 25393
CM
TM
2 32 32 256 256 512
32 128 512 1024 512 1024
4 64 32 8 4 8
4 32 32 4 4 8
8 32 64 8 8 16
2 64
8 512
2 2
2 2
4 8
different CAT activities were highly resistant to both chloramphenicol and thiamphenicol, but remained highly sensitive to the three fluorinated antibiotics. The inhibitory potencies of chloramphenicol, thiamphenicol and Sch 25298 were next compared against a sensitive E. coli strain. In two different liquid media, both chloramphenicol and Sch 25298 allowed only 20-30% residual growth at a drug concentration of 2 mg/1. In contrast, a thiamphenicol concentration of 25 mg/1 was required to produce a similar effect (Table II), thus confirming the trends indicated by the MICs (Table I). The antibiotics did not inhibit Sacch. cerevisiae growing in liquid culture (results not shown). Sch 25298, like chloramphenicol and thiamphenicol, is therefore a selective inhibitor of prokaryotic cells. Table II. Inhibition of E. coli growth by chloramphenicol, Sch 25298 and thiamphenicol Drug Chloramphenicol
Sch 25298
Thiamphenicol
Final concentration (mg/1) 0-10 0-25 0-50 1-00 2-00 0-10
% inhibition of growth" rich medium minimal medium
14 17 36 61 79 8
025
12
0-50 1-00 2-00 5-00 1
