624
Correspondence
Gentamicin assay in the presence of chloramphenicol Sir, Enzymatic assays of aminoglycoside concentrations are often thought to be unaffected by the presence of other antimicrobial agents. This has been shown to be true for a number of antibiotics with both nucleotidyltransferase and acetyltransferase assays (Holmes & Sandford, 1974; Broughall & Reeves, 1975). It has recently come to our attention that chloramphenicol, which was not tested by
Broughall & Reeves (1975), may interfere with acetyltransferase assays. The strain of Escherichia coli from which aminoglycoside 6'-N acetyltransferase [AAC(6')1 is prepared is resistant to chloramphenicol (Benveniste & Davies, 1971) as is the strain of Providencia stuartii we used for the preparation of aminoglycoside 2'-N acetyltransferase [AAC(20] (Shannon & Phillips, 1977). Table I shows that with acetyltransferase assays performed with these enzymes the apparent concentration of gentamicin decreased as the concentration of chloramphenicol increased. However, when the AAC(20 assay was performed with enzyme obtained from a chloramphenicol-sensitive strain, chloramphenicol was without effect. We also tested the effect of other antimicrobial agents on acetyltransferase assays of gentamicin. For these studies AAC(20 prepared from the chloramphenicol-sensitive strain was used. Ampicillin, cloxacillin, cephaloridine, lincomycin, clindamycin, erythromycin, tetracycline, colistin, sulphamethate, sulphamethoxazole, trimethoprim, rifampicin, nalidixic acid, nitrofurantoin, fusidic acid, vancomycin or metronidazole present at 50 mg/1 did not affect the assay of 5 mg/1 gentamicin with AAC(20 or AAC(60Both acetyltransferase assays detected the inactivation of gentamicin by carbenicillin (McLaughlin & Reeves, 1971). Mixtures of gentamicin (5 mg/l)and carbenicillin(200mg/l) were kept at —20°, 37° and 56°C for 16 h then assayed for gentamicin. With the AAQ20 assay the estimated concentrations of gentamicin were 4-8 mg/1 in the sample at —20°C, 3-6 mg/1 in the sample at 37°C and 2-9 mg/1 in
Table I. The effect of chloramphenicol on the AAC(2') and AAC(6') assays for gentamicin Gentamicin (mg/1)
Acetyltransferase Chloramphenicol (mg/I) AAC(6T AAC(2')» AAC(2')t Estimated concentration of gentamicin (mg/1)
2 2 2 2 5 5 5 5
0 10 50 200
0
10 50 200
2-3 2-1 1-5 0-5 51 4-5 3-2 1-4
21 1-7 0-7 0-5 51 41 3-4 30
•Enzyme prepared from a strain resistant to chloramphenicol. tEnzyme prepared from a strain sensitive to chloramphenicol.
1-9 1-9 21 1-9 4-8 4-8 51 4-8
Downloaded from http://jac.oxfordjournals.org/ at East Carolina University on April 19, 2015
Selwyn, S. Bacterial vaginitis and the rationale of clotrimazole therapy. Munchen Medizinische Wochenschrift 118 {Suppl. 1) S49-S52 (1976). Sherris, J. C. A discussion of susceptibility testing in anaerobic bacteria: role in disease. (Balows, A., Dehann, R. M., Dowell, V. R. Jr. & Guze, L. B., Eds). Charles C. Thomas, Springfield, Illinois (1974). Steingrimsson, O., Ryan, R. W. & Tilton, R. C. The microdilution antibiotic susceptibility test. Bacteroides fragilis. American Journal of Clinical Pathology 65:1010-5 (1976). Thornsberry, C. Factors affecting susceptibility tests and the need for standardized procedures. In Anaerobic bacteria: role in disease. (Balows, A., Dehann, R. M., Dowell, V. R. Jr. & Guze, L. B., Eds). Charles C. Thomas, Springfield, Illinois (1974). Washington, J. A. II. Methods for testing antibiotic susceptibility of anaerobic bacteria. In Progress in Chemotherapy (Daikos, G. K., Ed.). Hellenic Chemotherapy Society, Athens (1974). Wilkins, T. D. & Chaloren, S. Medium for use in antibiotic susceptibility testing of anaerobic bacteria. Antimicrobial Agents and Chemotherapy 10: 926-8 (1976).
Correspondence
625
the sample at 56°C; with the AAC(6') assay the estimated concentrations were 4-7, 3-3 and 2-7 mg/1. IAN PHILLIPS KEVIN SHANNON Department of Microbiology, St. Thomas's Hospital Medical School, London SEl 7EH, England
Table L Per cent errors of urease and agar plate diffusion methods of gentamicin assay Gentamicin content range (mg/I) 1 to 4 >4to 8 >8 to 16 >16 All specimens
Agar plate diffusion method
Urease method No. tested
Mean % error
% error
No. tested
Mean % error
10 9 9 6 34
-20 -111 + 8-5 -101 - 8-4
±10-9 ±13-6 ±14-5 ±14-5 ±16-4
7 1 8 5 21
+6-3 (-20) -5-4 + 6-8 +0-4
S.D. Of
S.D. Of
% error + 150 ±21-6 ±10-4 ±23-2
Downloaded from http://jac.oxfordjournals.org/ at East Carolina University on April 19, 2015
(P.N.E.) added known amounts of gentamicin to 34 different human sera, which were then tested, blind, by another (A.C.H.). Some of these were also examined, blind, by a third (C.C.F.), using the agar plate diffusion method (Phillips, Warren & Smith, 1974). The results are shown in Table I. Four categories of gentamicin content were represented, namely 1 to 4, > 4 to 8, > 8 to 16 and > 16 mg/1. Mean errors and standard References deviation (s.D.) of % error were consistently Benveniste, R. & Davies, J. Enzymatic acetylation 20% or less with the urease method. The plate of aminoglycoside antibiotics by Escherichia coli carrying an R factor. Biochemistry 10: 1787-96 diffusion method had a lower mean error but a higher s.D. for total specimens, and there was (1971). Broughall, J. M. & Reeves, D. S. Properties of the little to choose between the two methods. The gentamicin acetyltransferase enzyme and appli- overall mean error for the urease method was cation to the assay of aminoglycoside anti- - 8 - 4 % (s.D. ±16-4%). This compares favourbiotics. Antimicrobial Agents and Chemotherapy ably with results circulated from time to time 8: 222-3 (1975). by the Microbiology Quality Control Holmes, R. K. & Sandford, J. P. Enzymatic assay laboratory. for gentamicin and related aminoglycoside antiMost workers have found that the agar biotics. Journal of Infectious Diseases 129: 519— plate diffusion method, given sufficient care 27 (1974). Mclaughlin, J. E. & Reeves, D. S. Clinical and and time, can be more accurate than the laboratory evidence for inactivation of genta- urease method, but what little difference there micin by carbenicillin. Lancet i: 261-4 (1971). is is outweighed by the rapidity and conShannon, K. P. & Phillips, I. The use of amino- venience of the urease method as modified by glycosides 2'-N acetyltransferase for the assay us. We summarize our technique below. of gentamicin in serum, plasma and urine. Doubling dilutions (0-4-ml volumes) are Journal of Antimicrobial Chemotherapy 3: made, from left to right, of test serum (from 25-33 (1977). undiluted to •&-) in pooled human serum, plus serum control, in polystyrene disposable bijou bottles (Sterilin). To each bottle is then added Rapid gentamicin assay; 3 years experience (from right to left) 3 ml 4 % urea electrolytewith urease method deficient broth, made up as below. A second row of bottles is prepared similarly, using a Sir, We have used our modification (Edmunds & standard of pooled human serum containing Heddle, 1974) of Noone's urease method 4 mg/1 gentamicin instead of the test serum. (Noone, Pattison & Samson, 1971) for All bottles are inoculated as in Noone's gentamicin assay on 905 specimens of serum method with Proteus mirabilis BS 711, and from 486 patients. We have found it practical incubated for approximately 75 min at 37°C in and convenient to use on 1 to 3 specimens at a a water bath, the actual time chosen being the time. In addition to taking part in the Micro- minimum necessary to develop a good range biological Quality Control Scheme, one of us of colours from yellow to pink in both test and
Correspondence
Gentamicin assay in the presence of chloramphenicol Sir, Enzymatic assays of aminoglycoside concentrations are often thought to be unaffected by the presence of other antimicrobial agents. This has been shown to be true for a number of antibiotics with both nucleotidyltransferase and acetyltransferase assays (Holmes & Sandford, 1974; Broughall & Reeves, 1975). It has recently come to our attention that chloramphenicol, which was not tested by
Broughall & Reeves (1975), may interfere with acetyltransferase assays. The strain of Escherichia coli from which aminoglycoside 6'-N acetyltransferase [AAC(6')1 is prepared is resistant to chloramphenicol (Benveniste & Davies, 1971) as is the strain of Providencia stuartii we used for the preparation of aminoglycoside 2'-N acetyltransferase [AAC(20] (Shannon & Phillips, 1977). Table I shows that with acetyltransferase assays performed with these enzymes the apparent concentration of gentamicin decreased as the concentration of chloramphenicol increased. However, when the AAC(20 assay was performed with enzyme obtained from a chloramphenicol-sensitive strain, chloramphenicol was without effect. We also tested the effect of other antimicrobial agents on acetyltransferase assays of gentamicin. For these studies AAC(20 prepared from the chloramphenicol-sensitive strain was used. Ampicillin, cloxacillin, cephaloridine, lincomycin, clindamycin, erythromycin, tetracycline, colistin, sulphamethate, sulphamethoxazole, trimethoprim, rifampicin, nalidixic acid, nitrofurantoin, fusidic acid, vancomycin or metronidazole present at 50 mg/1 did not affect the assay of 5 mg/1 gentamicin with AAC(20 or AAC(60Both acetyltransferase assays detected the inactivation of gentamicin by carbenicillin (McLaughlin & Reeves, 1971). Mixtures of gentamicin (5 mg/l)and carbenicillin(200mg/l) were kept at —20°, 37° and 56°C for 16 h then assayed for gentamicin. With the AAQ20 assay the estimated concentrations of gentamicin were 4-8 mg/1 in the sample at —20°C, 3-6 mg/1 in the sample at 37°C and 2-9 mg/1 in
Table I. The effect of chloramphenicol on the AAC(2') and AAC(6') assays for gentamicin Gentamicin (mg/1)
Acetyltransferase Chloramphenicol (mg/I) AAC(6T AAC(2')» AAC(2')t Estimated concentration of gentamicin (mg/1)
2 2 2 2 5 5 5 5
0 10 50 200
0
10 50 200
2-3 2-1 1-5 0-5 51 4-5 3-2 1-4
21 1-7 0-7 0-5 51 41 3-4 30
•Enzyme prepared from a strain resistant to chloramphenicol. tEnzyme prepared from a strain sensitive to chloramphenicol.
1-9 1-9 21 1-9 4-8 4-8 51 4-8
Downloaded from http://jac.oxfordjournals.org/ at East Carolina University on April 19, 2015
Selwyn, S. Bacterial vaginitis and the rationale of clotrimazole therapy. Munchen Medizinische Wochenschrift 118 {Suppl. 1) S49-S52 (1976). Sherris, J. C. A discussion of susceptibility testing in anaerobic bacteria: role in disease. (Balows, A., Dehann, R. M., Dowell, V. R. Jr. & Guze, L. B., Eds). Charles C. Thomas, Springfield, Illinois (1974). Steingrimsson, O., Ryan, R. W. & Tilton, R. C. The microdilution antibiotic susceptibility test. Bacteroides fragilis. American Journal of Clinical Pathology 65:1010-5 (1976). Thornsberry, C. Factors affecting susceptibility tests and the need for standardized procedures. In Anaerobic bacteria: role in disease. (Balows, A., Dehann, R. M., Dowell, V. R. Jr. & Guze, L. B., Eds). Charles C. Thomas, Springfield, Illinois (1974). Washington, J. A. II. Methods for testing antibiotic susceptibility of anaerobic bacteria. In Progress in Chemotherapy (Daikos, G. K., Ed.). Hellenic Chemotherapy Society, Athens (1974). Wilkins, T. D. & Chaloren, S. Medium for use in antibiotic susceptibility testing of anaerobic bacteria. Antimicrobial Agents and Chemotherapy 10: 926-8 (1976).
Correspondence
625
the sample at 56°C; with the AAC(6') assay the estimated concentrations were 4-7, 3-3 and 2-7 mg/1. IAN PHILLIPS KEVIN SHANNON Department of Microbiology, St. Thomas's Hospital Medical School, London SEl 7EH, England
Table L Per cent errors of urease and agar plate diffusion methods of gentamicin assay Gentamicin content range (mg/I) 1 to 4 >4to 8 >8 to 16 >16 All specimens
Agar plate diffusion method
Urease method No. tested
Mean % error
% error
No. tested
Mean % error
10 9 9 6 34
-20 -111 + 8-5 -101 - 8-4
±10-9 ±13-6 ±14-5 ±14-5 ±16-4
7 1 8 5 21
+6-3 (-20) -5-4 + 6-8 +0-4
S.D. Of
S.D. Of
% error + 150 ±21-6 ±10-4 ±23-2
Downloaded from http://jac.oxfordjournals.org/ at East Carolina University on April 19, 2015
(P.N.E.) added known amounts of gentamicin to 34 different human sera, which were then tested, blind, by another (A.C.H.). Some of these were also examined, blind, by a third (C.C.F.), using the agar plate diffusion method (Phillips, Warren & Smith, 1974). The results are shown in Table I. Four categories of gentamicin content were represented, namely 1 to 4, > 4 to 8, > 8 to 16 and > 16 mg/1. Mean errors and standard References deviation (s.D.) of % error were consistently Benveniste, R. & Davies, J. Enzymatic acetylation 20% or less with the urease method. The plate of aminoglycoside antibiotics by Escherichia coli carrying an R factor. Biochemistry 10: 1787-96 diffusion method had a lower mean error but a higher s.D. for total specimens, and there was (1971). Broughall, J. M. & Reeves, D. S. Properties of the little to choose between the two methods. The gentamicin acetyltransferase enzyme and appli- overall mean error for the urease method was cation to the assay of aminoglycoside anti- - 8 - 4 % (s.D. ±16-4%). This compares favourbiotics. Antimicrobial Agents and Chemotherapy ably with results circulated from time to time 8: 222-3 (1975). by the Microbiology Quality Control Holmes, R. K. & Sandford, J. P. Enzymatic assay laboratory. for gentamicin and related aminoglycoside antiMost workers have found that the agar biotics. Journal of Infectious Diseases 129: 519— plate diffusion method, given sufficient care 27 (1974). Mclaughlin, J. E. & Reeves, D. S. Clinical and and time, can be more accurate than the laboratory evidence for inactivation of genta- urease method, but what little difference there micin by carbenicillin. Lancet i: 261-4 (1971). is is outweighed by the rapidity and conShannon, K. P. & Phillips, I. The use of amino- venience of the urease method as modified by glycosides 2'-N acetyltransferase for the assay us. We summarize our technique below. of gentamicin in serum, plasma and urine. Doubling dilutions (0-4-ml volumes) are Journal of Antimicrobial Chemotherapy 3: made, from left to right, of test serum (from 25-33 (1977). undiluted to •&-) in pooled human serum, plus serum control, in polystyrene disposable bijou bottles (Sterilin). To each bottle is then added Rapid gentamicin assay; 3 years experience (from right to left) 3 ml 4 % urea electrolytewith urease method deficient broth, made up as below. A second row of bottles is prepared similarly, using a Sir, We have used our modification (Edmunds & standard of pooled human serum containing Heddle, 1974) of Noone's urease method 4 mg/1 gentamicin instead of the test serum. (Noone, Pattison & Samson, 1971) for All bottles are inoculated as in Noone's gentamicin assay on 905 specimens of serum method with Proteus mirabilis BS 711, and from 486 patients. We have found it practical incubated for approximately 75 min at 37°C in and convenient to use on 1 to 3 specimens at a a water bath, the actual time chosen being the time. In addition to taking part in the Micro- minimum necessary to develop a good range biological Quality Control Scheme, one of us of colours from yellow to pink in both test and
