ANTIMICROBiAL AGENTs AND CHEMOTHERAPY, Sept. 1978, p. 391-397 0066-4804/78/0014-0391$02.00/0 Copyright X) 1978 American Society for Microbiology
Vol. 14, No. 3
Printedmi U.S.A.
In Vitro Activity of 5-Episisomicin in Bacteria Resistant to Other Aminoglycoside Antibiotics SHERWIN A. KABINS' AND CATHERINE NATHAN
Department of Medicine, Michael Reese Hospital and Medical Center, affiliated with the Pritzker School of Medicine, University of Chicago, Chicago, Illinois 60616 Received for publication 19 April 1978
Eighty-seven isolates of Pseudomonas, Enterobacteriaceae, and Staphylococchosen because of their resistance to other aminoglycosides, were tested for susceptibility to 5-episisomicin. Tests were performed in Mueller-Hinton agar and also, with 38 of these isolates, in Mueller-Hinton broth. Of Enterobacteriaceae, 85 and 95.5% were inhibited by 5 and 10 Ag of 5-episisomicin per ml, respectively. Amikacin inhibited 74 and 91% of the strains at 10 and 20 ,ug/ml, respectively. Fifty-four percent of P. aeruginosa were inhibited by 5-episisomicin and amikacin. Eigh*f-three percent of S. aureus were inhibited by netilmicin and amikacin, whereas only 50% were inhibited by 5-episisomicin. Isolates resistant to 5-episisomicin were most often resistant to the other aminoglycosides and occurred in gram-negative bacilli that did not carry aminoglycoside-modifying enzymes. Five of 23 isolates that carried a 6'-N-acetyltransferase (AAC-6') and one of two that carried an aminoglycoside 3-acetyltransferase were resistant to and acetylated 5episisomicin. Strains carrying other aminoglycoside-modifying enzymes were inhibited by 5-episisomicin. Thus, 5-episisomicin is a promising aminoglycoside not attacked by most aminoglycoside-modifying enzymes. Resistance will probably most often be based upon nonenzymatic mechanisms which will also affect other aminoglycosides. cus,
5-Episisomicin (SCH 22591), the 5-hydroxy stereoisomer of the deoxystreptamine ring of sisomicin, possesses enhanced antibacterial activity against a variety of gram-negative bacilli resistant to gentamicin (7). The purpose of this study was to compare susceptibility to 5-episisomicin in a group of bacteria originally collected because they were resistant to gentamicin, tobramycin, or an-ikacin. We also determined the susceptibility of each of the isolates to gentamicin, tobramycin, sisomicin, netilmicin, and amikacin.
Antibiotics. Aminoglycoside preparations of known specific activity used in this study included gentamicin, sisomicin, netilmicin, and 5-episisomicin, supplied by J. A. Waitz of the Schering Corp.; tobramycin, supplied by R. S. Griffith, Eli Lilly & Co.; and amikacin, obtained from L. A. Farcione, Bristol Laboratories. Quantitative determination of aminoglycoside susceptibility. The standard assay for microbial inhibition was performed with each of the six aminoglycosides by an agar dilution technique (5). Duplicate Mueller-Hinton agar (BBL) plates were inoculated with approximately 104 organisms by means of a Steers-Foltz apparatus. The minimal inhibitory concentration (MIC) was defined as the lowest concentraMATERIALS AND METHODS tion of antibiotic that prevented any visible growth of Organisms. One hundred of the 119 strains se- organisms after overnight incubation at 37°C. The 22 isolates of P. aeruginosa, 21 isolates of lected for study were isolated in our diagnostic microbiology laboratory between January, 1971 and Decem- Enterobacteriaceae, and 4 isolates of staphylococci ber, 1976. Eighty-seven were resistant either to gen- representative of the different susceptibility patterns tamicin, tobramycin, or amikacin. Included among the shown in the agar dilution test were assayed with 19 organisms obtained from others were 3 ATCC gentamicin and 5-episisomicin for MIC and for ministrains, Pseudomonas aeruginosa 27853, Staphylo- mal bactericidal concentration (MBC) by a broth dicoccus aureus 25923, and Escherichia coli 25922. Also lution procedure using an inoculum of approximately included were organisms carrying known aminoglyco- 105 organisms per ml (5). The MIC was defined as the side-modifying enzymes, obtained from K. E. Price, lowest concentration of antibiotic in those tubes showBristol Laboratories; J. E. Davies, University of Wis- ing no turbidity after 18 h of incubation at 37°C. After consin; Ian Phillips, St. Thomas Hospital, London, 18 h of incubation, the contents of nonturbid tubes England; F. H. Kayser, University of Zurich; S. Lerner, were subcultured to antibiotic-free Mueller-Hinton University of Chicago; and P. Stiffier, Grant Hospital, agar. The MBC was defined as the lowest concentration of antibiotic that permitted the growth of 10 or Chicago, Ill. 391
392
KABINS AND NATHAN
fewer colonies. Single lots of Mueller-Hinton broth and agar were used for these studies and were the same as used in a prior study (5). To compare the effects of different antibiotics, the results of each test were recorded separately as susceptible if the organisms were inhibited by an MIC of 5 ,ug or less of gentamicin, tobramycin, sisomicin, netilmicin, or 5-episisomicin per ml and by 10 ,ug or less of amikacin per ml. Strains that required 20 jg or more of the former antibiotics per ml or 40 pg of amikacin per ml for inhibition or killing were considered resistant. Organisms requiring 10l,ug of one of the former agents per ml or 20 pg of amikacin per ml for inhibition or killing were considered intermediate. These values were selected on the basis of the therapeutic effect of drug concentrations obtained in serum by administration of gentamicin, tobramycin, sisomicin, netilmicin, and amikacin. The resistant, susceptible, and intermediate values selected for 5-episisomicin were the same as those for sisomicin in view of the close chemical relation between these agents. Enzymatic assays. Adenylylation, phosphorylation, and acetylation assays against aminoglycosides were performed according to published methods (2). Bacteria whose extracts gave at least five times the counts per minute of a similarly processed extract of a comparable antibiotic-susceptible strain were considered to acetylate, phosphorylate, or adenylate the substrate significantly. The enzymes carried by the isolates used in this study had substrate specificities similar to or the same as the following: aminoglycoside
ANTimICROB. AGIZNTS CHEMOTHER.
5-episisomicin. The MICs for approximately 20% of the isolates were two to fourfold higher by the agar dilution method. The MBCs of gentamicin and 5-episisomicin averaged 8- to 16-fold higher than the MICs for the 22 isolates of P. aeruginosa. The MBCs of 5-episisomicin were at least 20 jg/ml for the 13 gentamicin-resistant isolates and averaged (MBC50) 40 jg/ml (Table 2). Enterobacteriaceae: MICs and MBCs. Overall, 5-episisomicin was the most active of the aminoglycosides against 66 isolates of Enterobacteriaceae selected for testing on the basis of resistance to gentamicin, tobramyoin, and/or amikacin. The MIC50 was 1.2 ug/ml and was 2fold less than that of amikacin, 4-fold less than that of netilmicin, 8-fold less than that of sisomicin, and 16-fold less than that of gentamicin and tobramycin. Judged on the basis of potential attainable therapeutic serum levels, 5-episisomicin was the most active agent, inhibiting 85 and 95.5% of the strains at 5 and 10 ,ug/ml, respectively. The next most active agent, amikacin, inhibited 74 and 91% of the strains at 10 and 20 pg/ml, respectively. The third most active agent, netilmicin, inhibited 50 and 67% of the isolates at 5 and 10 pg/ml, respectively. 5-Episisomicin was the most active of the ami2"-nucleotidyltransferase [gentamicin adenylyltrans- noglycosides tested against the 16 Proteus miferase ANT(2")]; aminoglycoside 4'-nucleotidyltrans- rabilis, indole-positive Proteus, and Providenferase [ANT(4')]; aminoglycoside 6'-acetyltransferase- cia strains (Table 1). The MIC50 for 5-episisom4 (6'-N-gentamicin acetyltransferase or kanamycin icin was two to fourfold less than amikacin. acetyltransferase) [AAC(6')]; aminoglycoside 3-acetyltransferase [AAC(3)]; aminoglycoside 2'-acetyltrans- Judged on the basis of potential attainable therferase-2 [AAC(2')-2]; aminoglycoside phosphotran- apeutic serum levels, 5-episisomicin and amikacin were equivalent for the indole-positive Proferase (1). Our clinical isolate, S. aureus Bell, was earlier teus (86% inhibited) and Providencia strains shown to produce AAC(6') and an aminoglycoside (100% inhibited). 5-Episisomicin inhibited all phosphotransferase (6), whereas isolates P. aerugi- three P. mirabilis isolates, whereas amikacin nosa POW and Citrobacter frelndii WIL (previously inhibited only two of the strains. erroneously identified as E. coli) were earlier shown 5-Episisomicin and netilmicin were the most to produce ANT enzymes (3, 4). From other unpub- active aminoglycosides tested against the 39 E. lished studies, information was available from 75 additional sonic extracts of isolates of gentamicin-, to- coli, Klebsiella, Enterobacter, Citrobacter, and Serratia liquefaciens strains (Table 1). Judged bramycin-, or amikacin-resistant strains. on the basis of potential attainable serum levels, RESULTS all seven Enterobacter, all 4 Citrobacter, and P. aeruginosa: MICs and MBCs. 5-Episi- five of the six S. liquefaciens isolates were sussomicin was the most active of the aminoglyco- ceptible to 5-episisomicin, netilmicin, and amisides tested against the 13 gentamicin-resistant kacin. 5-Epii8somicin inhibited 73% of the 22 P. aeruginosa isolates (Table 1). The MIC that total strains of E. coli and Klebsiella, whereas was inhibitory to at least 50% of isolates (MIC50) netilmicin inhibited 68% and amikacin 64% of was 5 ,ug/ml. Judged on the basis of potential the strains. attainable therapeutic serum levels, 5-episisom5-Episisomicin and gentamicin were the most icin and amikacin were of equivalent activity, active aminoglycosides tested against the 11 Serboth inhibiting 54% of the gentamicin-resistant ratia marcescens strains (Table 1). 5-Episisomstrains. icin inhibited 82% and gentamicin 64% of the Twenty-two isolates of gentamicin-resistant strains at 5 ,ug/ml. and susceptible P. aeruginosa were tested by The MICs of gentamicin and 5-episisomicin the broth dilution method with gentamicin and for the 21 isolates of gentamicin-resistant Enter-
VOL. 14, 1978
5-EPISISOMICIN FOR RESISTANT BACTERIA
obacteriaceae, when tested by the broth dilution method, were virtually identical to those obtained by the agar dilution method. The MBCs for these organisms were two- to eightfold higher than the MICs. The MBC of 5episisomicin was 5 to 10 ,ug/ml for only six of the strains, and was 40 ,ug/ml for most of the strains (Table 2). However, the MIC50 for the 21 strains tested was 5 to 10lg/ml, in contrast to 1.2 ug/ml for the entire group. S. aureus: MICs and MBCs. Netilmicin, amikacin, and 5-episisomicin were the most active aminoglycosides tested against the six gentamicin-resistant S. aureus strains (Table 1). However, based on potential attainable serum levels, netilmicin and amikacin were the most active, inhibiting 83% of the strains in contrast to 50% by 5-episisomicin. The MICs of gentamicin and 5-episisomicin for the four isolates of gentamicin-resistant S. aureus tested by the broth dilution method were virtually identical to those obtained with the agar dilution method. The MBCs for the four strains were four- to 64-fold higher than the MICs. The MBC of 5-episisomicin was greater than 10 pug/ml for three of the strains (Table 2). Relation of aminoglycoside-modifying enzymes to the MICs of 5-episisomicin. In other unpublished work carried out in our laboratory, we had tested 75 of the 87 gentamicin-, tobramycin-, and/or amikacin-resistant organisms for the presence of aminoglycoside-modifying enzymes. Sonic extracts from 57 of the 75 organisms were found to contain enzymes capable of modifying gentamicin, tobramycin, and/or amikacin. The relation of MICs to enzymes carried by these 75 organisms is summarized in Table 3. Susceptibility to 5-episisomicin was found in all organisms carrying an ANT(2") enzyme (found in P. aeruginosa and Enterobacteriaceae, mediating resistance to gentamicin, tobramycin, and sisomicin), in the Staphylococcus epidermidis carrying ANT(4') (mediating resistance to tobramycin), or in organisms carrying an AAC(2') enzyme (found in Proteus and Providencia, mediating resistance to gentamicin, tobramycin, sisomicin, and netilmicin). Five of the 23 organisms carrying an AAC(6') enzyme [10 of which simultaneously carried an ANT(2") and 4 of which which simultaneously carried an aminoglycoside phosphotransferase] required 10 or more ,ug of 5-episisomicin per ml for inhibition. The five included one of the two P. aeruginosa strains, two of the six S. marcescens that carried only an AAC(6') and two of the four S. aureus isolates that carried both an AAC(6') and aminoglycoside phosphotransferase. The AAC(6')
393
group of enzymes usually mediate resistance to amikacin, tobramycin, sisomicin, and netilmicin when found in Enterobacteriaceae, and mediate resistance to gentamicin, sisomicin, and tobramycin when found in S. aureus and P. aeruginosa. Sonic extracts from the one P. aeruginosa, two S. marcescens, and two S. aureus strains were tested and strongly acetylated 5-episisomicin. One of the two P. aeruginosa isolates carrying an AAC(3) enzyme mediating resistance to gentamicin, sisomicin, and netilmicin was resistant to 5-episisomicin and strongly acetylated it. Twelve of the 18 aminoglycoside-resistant organisms that did not carry any detectable aminoglycoside-modifying enzymes required 10 pg or greater of 5-episisomicin per ml for inhibition. Among the six isolates that were inhibited by 5 pug or less of 5-episisomicin per ml, two E. coli were selectively resistant to tobramycin, and one S. liquefaciens strain was resistant only to gentamicin and tobramycin. These strains were inhibited by 0.6 to 2.5 pg of 5-episisomicin per ml. The two remaining E. coli and one S. marcescens strains were of intermediate susceptibility to all the aminoglycosides tested, having MICs ranging for the most part from 5 to 20 ,ug/ml. In each of these organisms, the MIC of 5-episisomicin was 5 jig/ml. Patterns of aminoglycoside resistance among organisms resistant to 5-episisomicin. Twenty (23%) of the 87 isolates of gentamicin-, tobramycin-, and amikacin-resistant organisms required 10 or more ug of 5-episisomicin per ml for inhibition. Only 9 or 10% required 20 or more pg of 5-episisomicin per ml for inhibition. Overall, 11 of the 20 organisms were susceptible to one or another of the other aminoglycosides. Six were susceptible to amikacin, three to gentamicin, three to netilmicin (two of which were also susceptible to amikacin), and one to tobramycin.
Uniforn resistance to all tested aminoglycosides was the most common pattern. It was found in nine isolates, eight of which carried no detectable aminoglycoside-modifying enzymes and one of which was not tested for carriage of enzymes. In each of these nine strains, 5-episisomicin gave the lowest MIC (although all were 10 ,ug/ml or greater). There appears to be a continuum of levels of aminoglycoside resistance among these nine strains and also among the three strains of intermediate susceptibility referred to in the section above. For each organism the level of resistance was fairly uniform from aminoglycoside to aminoglycoside, but the levels varied from organism to organism, suggesting different degrees of "impaired permeability" to the aminoglycosides. It appeared that for each
TABLz 1. Agar dilution MICs for organisms eeected because of resistance to aminoglycosides MIC Cumulative no. of strains with MIC (Wml) of: againstepibe 540 20 88ceptibl con- (tl) Organism (no.)' A10
P. aerugnosa (13) Gent Siso 5-Epi Tobr Neti Amik
1.2-5 0.6-5 80 5 40 20 10
0
0
0
0 6 1 0 0
0 7 4 1 3
2 9 4 5 7
P. mirabilis (3) Gent Siso 5-Epi Tobr Neti Amik
0.6-1.2
Vol. 14, No. 3
Printedmi U.S.A.
In Vitro Activity of 5-Episisomicin in Bacteria Resistant to Other Aminoglycoside Antibiotics SHERWIN A. KABINS' AND CATHERINE NATHAN
Department of Medicine, Michael Reese Hospital and Medical Center, affiliated with the Pritzker School of Medicine, University of Chicago, Chicago, Illinois 60616 Received for publication 19 April 1978
Eighty-seven isolates of Pseudomonas, Enterobacteriaceae, and Staphylococchosen because of their resistance to other aminoglycosides, were tested for susceptibility to 5-episisomicin. Tests were performed in Mueller-Hinton agar and also, with 38 of these isolates, in Mueller-Hinton broth. Of Enterobacteriaceae, 85 and 95.5% were inhibited by 5 and 10 Ag of 5-episisomicin per ml, respectively. Amikacin inhibited 74 and 91% of the strains at 10 and 20 ,ug/ml, respectively. Fifty-four percent of P. aeruginosa were inhibited by 5-episisomicin and amikacin. Eigh*f-three percent of S. aureus were inhibited by netilmicin and amikacin, whereas only 50% were inhibited by 5-episisomicin. Isolates resistant to 5-episisomicin were most often resistant to the other aminoglycosides and occurred in gram-negative bacilli that did not carry aminoglycoside-modifying enzymes. Five of 23 isolates that carried a 6'-N-acetyltransferase (AAC-6') and one of two that carried an aminoglycoside 3-acetyltransferase were resistant to and acetylated 5episisomicin. Strains carrying other aminoglycoside-modifying enzymes were inhibited by 5-episisomicin. Thus, 5-episisomicin is a promising aminoglycoside not attacked by most aminoglycoside-modifying enzymes. Resistance will probably most often be based upon nonenzymatic mechanisms which will also affect other aminoglycosides. cus,
5-Episisomicin (SCH 22591), the 5-hydroxy stereoisomer of the deoxystreptamine ring of sisomicin, possesses enhanced antibacterial activity against a variety of gram-negative bacilli resistant to gentamicin (7). The purpose of this study was to compare susceptibility to 5-episisomicin in a group of bacteria originally collected because they were resistant to gentamicin, tobramycin, or an-ikacin. We also determined the susceptibility of each of the isolates to gentamicin, tobramycin, sisomicin, netilmicin, and amikacin.
Antibiotics. Aminoglycoside preparations of known specific activity used in this study included gentamicin, sisomicin, netilmicin, and 5-episisomicin, supplied by J. A. Waitz of the Schering Corp.; tobramycin, supplied by R. S. Griffith, Eli Lilly & Co.; and amikacin, obtained from L. A. Farcione, Bristol Laboratories. Quantitative determination of aminoglycoside susceptibility. The standard assay for microbial inhibition was performed with each of the six aminoglycosides by an agar dilution technique (5). Duplicate Mueller-Hinton agar (BBL) plates were inoculated with approximately 104 organisms by means of a Steers-Foltz apparatus. The minimal inhibitory concentration (MIC) was defined as the lowest concentraMATERIALS AND METHODS tion of antibiotic that prevented any visible growth of Organisms. One hundred of the 119 strains se- organisms after overnight incubation at 37°C. The 22 isolates of P. aeruginosa, 21 isolates of lected for study were isolated in our diagnostic microbiology laboratory between January, 1971 and Decem- Enterobacteriaceae, and 4 isolates of staphylococci ber, 1976. Eighty-seven were resistant either to gen- representative of the different susceptibility patterns tamicin, tobramycin, or amikacin. Included among the shown in the agar dilution test were assayed with 19 organisms obtained from others were 3 ATCC gentamicin and 5-episisomicin for MIC and for ministrains, Pseudomonas aeruginosa 27853, Staphylo- mal bactericidal concentration (MBC) by a broth dicoccus aureus 25923, and Escherichia coli 25922. Also lution procedure using an inoculum of approximately included were organisms carrying known aminoglyco- 105 organisms per ml (5). The MIC was defined as the side-modifying enzymes, obtained from K. E. Price, lowest concentration of antibiotic in those tubes showBristol Laboratories; J. E. Davies, University of Wis- ing no turbidity after 18 h of incubation at 37°C. After consin; Ian Phillips, St. Thomas Hospital, London, 18 h of incubation, the contents of nonturbid tubes England; F. H. Kayser, University of Zurich; S. Lerner, were subcultured to antibiotic-free Mueller-Hinton University of Chicago; and P. Stiffier, Grant Hospital, agar. The MBC was defined as the lowest concentration of antibiotic that permitted the growth of 10 or Chicago, Ill. 391
392
KABINS AND NATHAN
fewer colonies. Single lots of Mueller-Hinton broth and agar were used for these studies and were the same as used in a prior study (5). To compare the effects of different antibiotics, the results of each test were recorded separately as susceptible if the organisms were inhibited by an MIC of 5 ,ug or less of gentamicin, tobramycin, sisomicin, netilmicin, or 5-episisomicin per ml and by 10 ,ug or less of amikacin per ml. Strains that required 20 jg or more of the former antibiotics per ml or 40 pg of amikacin per ml for inhibition or killing were considered resistant. Organisms requiring 10l,ug of one of the former agents per ml or 20 pg of amikacin per ml for inhibition or killing were considered intermediate. These values were selected on the basis of the therapeutic effect of drug concentrations obtained in serum by administration of gentamicin, tobramycin, sisomicin, netilmicin, and amikacin. The resistant, susceptible, and intermediate values selected for 5-episisomicin were the same as those for sisomicin in view of the close chemical relation between these agents. Enzymatic assays. Adenylylation, phosphorylation, and acetylation assays against aminoglycosides were performed according to published methods (2). Bacteria whose extracts gave at least five times the counts per minute of a similarly processed extract of a comparable antibiotic-susceptible strain were considered to acetylate, phosphorylate, or adenylate the substrate significantly. The enzymes carried by the isolates used in this study had substrate specificities similar to or the same as the following: aminoglycoside
ANTimICROB. AGIZNTS CHEMOTHER.
5-episisomicin. The MICs for approximately 20% of the isolates were two to fourfold higher by the agar dilution method. The MBCs of gentamicin and 5-episisomicin averaged 8- to 16-fold higher than the MICs for the 22 isolates of P. aeruginosa. The MBCs of 5-episisomicin were at least 20 jg/ml for the 13 gentamicin-resistant isolates and averaged (MBC50) 40 jg/ml (Table 2). Enterobacteriaceae: MICs and MBCs. Overall, 5-episisomicin was the most active of the aminoglycosides against 66 isolates of Enterobacteriaceae selected for testing on the basis of resistance to gentamicin, tobramyoin, and/or amikacin. The MIC50 was 1.2 ug/ml and was 2fold less than that of amikacin, 4-fold less than that of netilmicin, 8-fold less than that of sisomicin, and 16-fold less than that of gentamicin and tobramycin. Judged on the basis of potential attainable therapeutic serum levels, 5-episisomicin was the most active agent, inhibiting 85 and 95.5% of the strains at 5 and 10 ,ug/ml, respectively. The next most active agent, amikacin, inhibited 74 and 91% of the strains at 10 and 20 pg/ml, respectively. The third most active agent, netilmicin, inhibited 50 and 67% of the isolates at 5 and 10 pg/ml, respectively. 5-Episisomicin was the most active of the ami2"-nucleotidyltransferase [gentamicin adenylyltrans- noglycosides tested against the 16 Proteus miferase ANT(2")]; aminoglycoside 4'-nucleotidyltrans- rabilis, indole-positive Proteus, and Providenferase [ANT(4')]; aminoglycoside 6'-acetyltransferase- cia strains (Table 1). The MIC50 for 5-episisom4 (6'-N-gentamicin acetyltransferase or kanamycin icin was two to fourfold less than amikacin. acetyltransferase) [AAC(6')]; aminoglycoside 3-acetyltransferase [AAC(3)]; aminoglycoside 2'-acetyltrans- Judged on the basis of potential attainable therferase-2 [AAC(2')-2]; aminoglycoside phosphotran- apeutic serum levels, 5-episisomicin and amikacin were equivalent for the indole-positive Proferase (1). Our clinical isolate, S. aureus Bell, was earlier teus (86% inhibited) and Providencia strains shown to produce AAC(6') and an aminoglycoside (100% inhibited). 5-Episisomicin inhibited all phosphotransferase (6), whereas isolates P. aerugi- three P. mirabilis isolates, whereas amikacin nosa POW and Citrobacter frelndii WIL (previously inhibited only two of the strains. erroneously identified as E. coli) were earlier shown 5-Episisomicin and netilmicin were the most to produce ANT enzymes (3, 4). From other unpub- active aminoglycosides tested against the 39 E. lished studies, information was available from 75 additional sonic extracts of isolates of gentamicin-, to- coli, Klebsiella, Enterobacter, Citrobacter, and Serratia liquefaciens strains (Table 1). Judged bramycin-, or amikacin-resistant strains. on the basis of potential attainable serum levels, RESULTS all seven Enterobacter, all 4 Citrobacter, and P. aeruginosa: MICs and MBCs. 5-Episi- five of the six S. liquefaciens isolates were sussomicin was the most active of the aminoglyco- ceptible to 5-episisomicin, netilmicin, and amisides tested against the 13 gentamicin-resistant kacin. 5-Epii8somicin inhibited 73% of the 22 P. aeruginosa isolates (Table 1). The MIC that total strains of E. coli and Klebsiella, whereas was inhibitory to at least 50% of isolates (MIC50) netilmicin inhibited 68% and amikacin 64% of was 5 ,ug/ml. Judged on the basis of potential the strains. attainable therapeutic serum levels, 5-episisom5-Episisomicin and gentamicin were the most icin and amikacin were of equivalent activity, active aminoglycosides tested against the 11 Serboth inhibiting 54% of the gentamicin-resistant ratia marcescens strains (Table 1). 5-Episisomstrains. icin inhibited 82% and gentamicin 64% of the Twenty-two isolates of gentamicin-resistant strains at 5 ,ug/ml. and susceptible P. aeruginosa were tested by The MICs of gentamicin and 5-episisomicin the broth dilution method with gentamicin and for the 21 isolates of gentamicin-resistant Enter-
VOL. 14, 1978
5-EPISISOMICIN FOR RESISTANT BACTERIA
obacteriaceae, when tested by the broth dilution method, were virtually identical to those obtained by the agar dilution method. The MBCs for these organisms were two- to eightfold higher than the MICs. The MBC of 5episisomicin was 5 to 10 ,ug/ml for only six of the strains, and was 40 ,ug/ml for most of the strains (Table 2). However, the MIC50 for the 21 strains tested was 5 to 10lg/ml, in contrast to 1.2 ug/ml for the entire group. S. aureus: MICs and MBCs. Netilmicin, amikacin, and 5-episisomicin were the most active aminoglycosides tested against the six gentamicin-resistant S. aureus strains (Table 1). However, based on potential attainable serum levels, netilmicin and amikacin were the most active, inhibiting 83% of the strains in contrast to 50% by 5-episisomicin. The MICs of gentamicin and 5-episisomicin for the four isolates of gentamicin-resistant S. aureus tested by the broth dilution method were virtually identical to those obtained with the agar dilution method. The MBCs for the four strains were four- to 64-fold higher than the MICs. The MBC of 5-episisomicin was greater than 10 pug/ml for three of the strains (Table 2). Relation of aminoglycoside-modifying enzymes to the MICs of 5-episisomicin. In other unpublished work carried out in our laboratory, we had tested 75 of the 87 gentamicin-, tobramycin-, and/or amikacin-resistant organisms for the presence of aminoglycoside-modifying enzymes. Sonic extracts from 57 of the 75 organisms were found to contain enzymes capable of modifying gentamicin, tobramycin, and/or amikacin. The relation of MICs to enzymes carried by these 75 organisms is summarized in Table 3. Susceptibility to 5-episisomicin was found in all organisms carrying an ANT(2") enzyme (found in P. aeruginosa and Enterobacteriaceae, mediating resistance to gentamicin, tobramycin, and sisomicin), in the Staphylococcus epidermidis carrying ANT(4') (mediating resistance to tobramycin), or in organisms carrying an AAC(2') enzyme (found in Proteus and Providencia, mediating resistance to gentamicin, tobramycin, sisomicin, and netilmicin). Five of the 23 organisms carrying an AAC(6') enzyme [10 of which simultaneously carried an ANT(2") and 4 of which which simultaneously carried an aminoglycoside phosphotransferase] required 10 or more ,ug of 5-episisomicin per ml for inhibition. The five included one of the two P. aeruginosa strains, two of the six S. marcescens that carried only an AAC(6') and two of the four S. aureus isolates that carried both an AAC(6') and aminoglycoside phosphotransferase. The AAC(6')
393
group of enzymes usually mediate resistance to amikacin, tobramycin, sisomicin, and netilmicin when found in Enterobacteriaceae, and mediate resistance to gentamicin, sisomicin, and tobramycin when found in S. aureus and P. aeruginosa. Sonic extracts from the one P. aeruginosa, two S. marcescens, and two S. aureus strains were tested and strongly acetylated 5-episisomicin. One of the two P. aeruginosa isolates carrying an AAC(3) enzyme mediating resistance to gentamicin, sisomicin, and netilmicin was resistant to 5-episisomicin and strongly acetylated it. Twelve of the 18 aminoglycoside-resistant organisms that did not carry any detectable aminoglycoside-modifying enzymes required 10 pg or greater of 5-episisomicin per ml for inhibition. Among the six isolates that were inhibited by 5 pug or less of 5-episisomicin per ml, two E. coli were selectively resistant to tobramycin, and one S. liquefaciens strain was resistant only to gentamicin and tobramycin. These strains were inhibited by 0.6 to 2.5 pg of 5-episisomicin per ml. The two remaining E. coli and one S. marcescens strains were of intermediate susceptibility to all the aminoglycosides tested, having MICs ranging for the most part from 5 to 20 ,ug/ml. In each of these organisms, the MIC of 5-episisomicin was 5 jig/ml. Patterns of aminoglycoside resistance among organisms resistant to 5-episisomicin. Twenty (23%) of the 87 isolates of gentamicin-, tobramycin-, and amikacin-resistant organisms required 10 or more ug of 5-episisomicin per ml for inhibition. Only 9 or 10% required 20 or more pg of 5-episisomicin per ml for inhibition. Overall, 11 of the 20 organisms were susceptible to one or another of the other aminoglycosides. Six were susceptible to amikacin, three to gentamicin, three to netilmicin (two of which were also susceptible to amikacin), and one to tobramycin.
Uniforn resistance to all tested aminoglycosides was the most common pattern. It was found in nine isolates, eight of which carried no detectable aminoglycoside-modifying enzymes and one of which was not tested for carriage of enzymes. In each of these nine strains, 5-episisomicin gave the lowest MIC (although all were 10 ,ug/ml or greater). There appears to be a continuum of levels of aminoglycoside resistance among these nine strains and also among the three strains of intermediate susceptibility referred to in the section above. For each organism the level of resistance was fairly uniform from aminoglycoside to aminoglycoside, but the levels varied from organism to organism, suggesting different degrees of "impaired permeability" to the aminoglycosides. It appeared that for each
TABLz 1. Agar dilution MICs for organisms eeected because of resistance to aminoglycosides MIC Cumulative no. of strains with MIC (Wml) of: againstepibe 540 20 88ceptibl con- (tl) Organism (no.)' A10
P. aerugnosa (13) Gent Siso 5-Epi Tobr Neti Amik
1.2-5 0.6-5 80 5 40 20 10
0
0
0
0 6 1 0 0
0 7 4 1 3
2 9 4 5 7
P. mirabilis (3) Gent Siso 5-Epi Tobr Neti Amik
0.6-1.2
