THE JOURNAL OF INFECTIOUS DISEASES • VOL. 133, NO.5· © 1976 by the University of Chicago. All rights reserved.
MAY 1976
An Enzymatic Assay for Chloramphenicol with Partially Purified Chloramphenicol Acetyltransferase From the Biochemistry Service, Notre-Dame Hospital, Montreal, Canada
Rejean Daigneault and Michel Guitard
Chloramphenicol therapy remains controversial in several circumstances [1] because of its toxicity and the availability of other effective broad-spectrum antibiotics. However, the use of chloramphenicol. alone or with other antibiotics, for periods of up to six weeks has been recommended for infections of the central nervous system [2-5]. This recommendation is based on the good penetration of chloramphenicol into the cerebrospinal fluid [6-9], cerebral tissue [10], and brain abscess [11]. With the recent appearance of ampicillin-resistant strains of Haemophilus iniiuenzae type b, an increase in utilization of chloramphenicol for treatment of bacterial meningitis is predictable [12, 13]. A rapid, reliable, and precise assay for chloramphenicol in biological fluids may constitute a valuable adjunct by means of which the clinician can carefully monitor levels of chioram-
Received for publication June 30, 1975, and in revised form December 17, 1975. This investigation was supported by grant no. MA-5016 from the Medical Research Council of Canada and by a grant from the Conseil de la Recherche Medicate du Quebec. We thank Dr. Michel Brazeau for many stimulating discussions, for reviewing the manuscript, and for permitting us to analyze serum and urine specimens from his patients. We gratefully acknowledge the valuable technical assistance of Dorothee Lemieux. Please address requests for reprints to Dr. Rejean Daigneault, Service de Biochimie, Hopital Notre-Dame, 1560 Sherbrooke est, Montreal H2L 4Ml, Canada.
phenicol in the blood of patients with impaired renal and/or hepatic function. We report herein an enzymatic assay for chloramphenicol based on the use of chloramphenicol acetyltransferase (CATase) [14]. This inactivating enzyme catalyzes the acetylation of chloramphenicol in the presence of acetyl coenzyme A to yield the 3-acetyl and 1,3-diacetyl esters [15-17]. We describe some properties of the partially purified CATase preparation and correlate results of the enzymatic assay with those of a microbiological method developed in our laboratory. Materials and Methods
Bacterial strains. Escherichia coli W677/ HJR66, an R-factor-bearing mutant [18] (kindly provided by Dr. D. H. Smith, Children's Hospital Medical Center, Boston, Mass.) served as the source of CATase.! The MIC of chloramphenicol for this strain is 1 mg/ml. Sarcina lutea (ATCC 9341) was used in the microbiological assay of chloramphenicol. Culture method and preparation of the cell-free extract. An overnight culture of E. coli in trypticase soy broth (Baltimore Biological Laboratories, Baltimore, Md.) was diluted 100-fold in 2,800 ml of the same nutrient broth and incubated at 37 C with magnetic agitation until late logarith-
1 We will gladly supply E. coli W677 /HJR66 and the HP-65 program listing to interested investigators upon request.
515
Downloaded from http://jid.oxfordjournals.org/ at University of Manitoba on September 6, 2015
In an enzymatic assay for chloramphenicol, chloramphenicol acetyltransferase partially purified by affinity chromatography was used; [3H]acetyl coenzyme A served as a substrate. The purified enzyme was sensitive to p-hydroxymercuribenzoate but insensitive to 5,5'-dithio-bis-2-nitrobenzoic acid. The Michaelis-Menten constant was 10.5 !J.M. The operating range of the enzymatic assay was 0-30 !J.gl ml. The coefficient of variation was 1.9% at a concentration of 10 !J.gl ml. The method correlated well with a microbiological agar-paper disk method for mock unknown sera (r = 0.997) and serum or urine specimens (r = 0.993). The enzymatic assay was unaffected by the presence of any of 12 other antibiotics in tested serum.
516
~g of chloramphenicol/ml. The standards were kept frozen at -70 C in 0.5-ml aliquots until use. The pool of serum was tested for its absence of antimicrobial activity. Forty-six mock unknown sera with drug concentrations ranging from 5 to 30 ug/rnl were also prepared by appropriate dilution of the stock solution (100 ug/rnl) with human pooled serum. Specimens of serum or urine from patients receiving both chloramphenicol and penicillin 0 were either processed the same day or kept frozen at - 20 C until assay. Enzymatic assay of chloramphenicol. CATase was incubated with [3H]acetyl coenzyme A under conditions such that the concentration of chloramphenicol was the rate-limiting factor of the reaction. Radioactivity of the 3-acetyl and 1,3-diacetyl esters of chloramphenicol produced was measured after extraction by benzene at slightly alkaline pH. Under these conditions, acetyl coenzyme A remained in the aqueous phase [22]. The reaction mixture contained 100 ul of 0.25 M Tris-HCI (pH 7.8), 35 f.!l of a 0.35-mM PH] acetyl coenzyme A solution (specific activity, 14.3 mCi/mmol; New England Nuclear, Boston, Mass.), 50 milliunits of purified CATase, and 50 ~I of chloramphenicol standard, mock unknown, undiluted serum specimen, or urine specimen diluted with an equal volume of distilled water to a final volume of 250 ~l. The reaction mixture, prepared at 0 C in 5-ml conical test tubes, was incubated for 20 min at 37 C. Extraction with 1 ml of benzene (Fisher, scintillation grade) stopped the reaction by precipitation of the proteins and brought the [3H]acetyl chloramphenicol and [1,3-3H]diacetyl chloramphenicol into the organic phase. This step was repeated twice. The combined organic-phase material (3 ml) was transferred to a glass scintillation vial and acidified with 0.1 ml of glacial acetic acid. The benzene was evaporated to dryness at 70 C, and 10 ml of a toluene-based scintillant was added to the vial. Radioactivity was counted in a Beckman LS 330 liquid scintillation spectrometer. All assays were performed in duplicate. For evaluation of the effect of other antibiotics on the enzymatic assay of chloramphenicol, the reaction mixture was modified as follows. The source of antibiotic was prepared by successive addition of chloramphenicol and a second anti-
Downloaded from http://jid.oxfordjournals.org/ at University of Manitoba on September 6, 2015
mic phase (A = 1.5 at 490 nm). The cells were then harvested by centrifugation at 10,000 g for 10 min at 2 C, washed twice with 0.01 M Tris-HCI (pH 7.8; Fisher Scientific, Fair Lawn, N.J.), and resuspended in 40 ml of the same buffer. The bacteria were disrupted at 0 C by continuous sonication for 3 min at 70 W with a Biosonic sonifier (Artek Systems, Farmingdale, N.Y.). Cell debris was removed by centrifugation at 30,000 g for 30 min at 2 C, and the supernatant fluid (cell-free extract) was stored in O.5-ml aliquots at - 70 C. Assay of CATase. CATase was assayed at 37 C by the colorimetric method of Foster and Shaw [19] with a Beckman Acta III spectrophotometer (Beckman Instruments, Fullerton, Ca.) [20]. A unit of enzyme activity represents 1 umol of coenzyme A liberated per min at 37 C. The same spectrophotometric method was used for determination of the Michaelis-Menten constant (Ktn ) and the effect of sodium p-hydroxymercuribenzoate (Sigma Chemical, St. Louis, Mo.) and 5,5'dithio-bis-2-nitrobenzoic acid (Sigma) on the reaction catalyzed by CATase. Purification of CA'I'ase by affinity chromatography. CATase was partially purified as described previously [20] with use of a Sepharosereduced chloramphenicol affinity column and elution by chloramphenicol competition. The eluate fractions containing the enzyme were pooled and concentrated 10-fold with a Minicon B 15 clinical ultrafiltration system (Amicon Corp., Lexington, Mass. ). Chloramphenicol was eliminated by passage of the purified CATase concentrate through a Sephadex 0-10 column (0.5 X 20 em; Pharmacia Fine Chemicals, Uppsala, Sweden) at 2 C, equilibrated with 0.1 M Tris-HCI and 0.05 M NaCI buffer at pH 7.8. The enzyme peak eluted in the void volume of the column was stored in O.5-ml aliquots at - 70 C. The amount of protein in the bacterial extract and in the purified enzyme preparations was measured according to the method of Lowry et al. [21]. Chloramphenicol standards, mock unknown sera, and human serum or urine specimens. Chloramphenicol (Parke, Davis, Detroit, Mich.) was dissolved in pooled serum obtained from healthy individuals to make a stock solution of 100 ug/rnl. The stock solution was diluted with the same pool of serum to make the following standards: 0, 2.5, 5.0, 10.0, 15.0, 20.0, and 30.0
Daigneault and Guitard
Enzymatic Assay for Chloramphenicol
517
biotic to pooled serum at 0 C; this preparation was incubated at 37 C for 30 min. Fifty microliters of this antibiotic source were processed as described previously for the assay of chloramphenicol. The amount of antibiotic in the 50-~tl fraction was expressed as a concentration (table 1) .
Effect of other antibiotics on the enzymatic assay of chloramphenicol (concentration, 10
Table 1. ug/rnl).
Antibiotic (ug/rnl ) Sulfathiazole (100) Erythromycin (5) Tetracycline . HCl (12) Minocycline . HCl (5) Kanamycin (10) Gentamicin (10) Tobramycin (10) Cephalexin (50) Cephaloridine (50) Ampicillin (25) Penicillin (100) Clindamycin (8)
Chloramphenicol measured (ug/rnl) 9.8 9.8 9.7 10.0 9.9 9.9 10.0 10.0 9.8 9.9 10.1 9.8
+
+
y
== e
[ -b+v'b 2-4ac+4cx] 2c
Results
The results of CATase purification by affinity chromatography are not presented herein since they were identical to those reported previously [20]. Concentration of the partially purified enzyme and elimination of chloramphenicol from CATase by gel filtration resulted in the total recovery of the initial level of CATase activity. The same affinity column has been used for more than one year without an appreciable loss of enzyme retention capability. More than 100 affinity chromatography separations have been achieved with this same column. All experiments were done with preparations that had a specific activity of > 125 units /mg of protein. 2
See footnote 1.
Downloaded from http://jid.oxfordjournals.org/ at University of Manitoba on September 6, 2015
The specificity of the enzymatic assay was tested against chloramphenicol palmitate (Parke, Davis) , chloramphenicol succinate (Pentagone Laboratories, Montreal, Canada) , and thiamphenicol (Sterling Winthrop Research Institute, Rensselaer, N.Y.) at a concentration of 20 ug/rnl prepared in human pooled serum. The specimens were processed as described previously for the assay of chloramphenicol. The operating range for the enzymatic assay was 0-30 ug/rnl. The following modification in the composition of the reaction mixture allowed measurements within the range 0-40 ng/rnl: 25 III of 1 M Tris-HCI (pH 7.8), 20 III of a 5-IlM [3H]acetyl coenzyme A solution (specific activity, 900 mCi/mmol), 5 III of purified CATase (5 milliunits), and 200 III of chloramphenicol standard. Microbiological assay of chloramphenicol. Each specimen or chloramphenicol standard (250 Ill) was incubated at 25 C for 30 min with 50 III of penicillinase (Bacto-penase concentrate penicillinase, 500,000 units/rnl; Difco Laboratories, Detroit, Mich.). The assay agar was prepared by mixture of 40 g of trypticase soy agar (Baltimore Biological Laboratories) with 1 liter of distilled
water. The mixture was brought to the boiling point until the medium cleared and was then autoclaved. A suspension of S. lutea (5 ml) was prepared from a blood agar slant in trypticase soy broth and adjusted to an absorbance of 24 at 490 nm. This amount was added to 100 ml of melted agar kept at 56 C. The seeded agar (12 ml) was immediately poured into 150- X 15-mm standard disposable plastic Petri dishes and allowed to solidify on a level surface. The assay, a modification of the agar-paper disk diffusion method of Sabath et al. [23], was performed by pipetting of 50 III of each penicillinase-treated specimen onto a 0.25-inch Bacto concentration sterile blank disk (Difco Laboratories), which was on the surface of the seeded agar plate. Each sample was tested in duplicate. The diameter of the zone of growth inhibition was measured after incubation for 20 hr at 37 C. Results were computed with a Hewlett-Packard 65 pocket calculator programmed- for resolving the multiple regression equation x == a b (In y) c (In y) 2, where x is the zone diameter in millimeters and y the concentration of chloramphenicol in ug/rnl. After x and y values were entered for each standard, the calculator instantly generated coefficients a, b, c, the correlation coefficient r, and (given the zone diameter) the antibiotic concentration derived from the equation:
Daigneault and Guitard
518
-
100
100
e-
';
'';:;
u
ca CD
en
....ca et:
50
Co)
::IE 50
~
ti
u ....::c
::I
'"CI
';;
= CD
25
0 0
5 pHMB (11M)
10
Figure 1. Effect of p-hydroxymercuribenzoate (pHMB) on the activity of chloramphenicol acetyltransferase (CATase). Each point represents a mean of two determinations of residual CATase activity after inhibition by pHMB under the conditions described in Materials and Methods.
5
10
15
20
25
30
CHLORAMPHENICOL IN 1l./ml
Figure 3. Typical standard curve for chloramphenicol assayed by the enzymatic method. Each standard was analyzed in duplicate, and each point represents the mean of the radioactivity measurements.
Downloaded from http://jid.oxfordjournals.org/ at University of Manitoba on September 6, 2015
Properties of CATase. The enzyme preparations were shown to be stable for at least three months when kept at -70 C. Stability was not measured for preparations kept for a longer period. Repeated freezing and thawing resulted in drastic losses of enzyme activity. The osmotic lysate of E. coli W677/HJR66 used in our laboratory for the quantitation of gentamicin [24] did not contain any CATase activity, a fact suggesting an intracellular rather than a periplasmic location for this enzyme. The K m of the purified CATase for chloramphenicol was 10.5 ~M. CATase was not sensitive to 5,5'-dithio-bis-2-nitrobenzoic acid; incubation of the enzyme with this acid at 37 C for 15 min resulted in a loss of no more than 10% of the initial activity. The enzyme was found to be sensitive to sodium p-hydroxymercuribenzoate; CATase activity was inhibited by 50% at a concentration of 3.3 f.1M (figure 1). The addition of dithiothreitol or 2-mercaptoethanol at a concentration of up to 10 mM did not reactivate the enzyme. Assay of chloramphenicol in human serum. Standard curves. The operating range of the agar-paper disk diffusion technique was 5-30 ug/rnl. A typical standard curve is shown in fig-
519
Enzymatic Assay for Chloramphenicol
amphenicol in the ng/rnl range. A linear relation was observed between radioactivity in the slightly alkaline benzene extract and concentrations of chloramphenicol of 0-40 ng/rnl (figure 4). Precision. Within-run repeatability was estimated by measurement of the same mock unknown 10 times at a concentration of 10 ug/rnl, The coefficient of variation was calculated as 100
_I '2. (x -
x l
x)2
n-l
'
100
-
75
:E a.
50
0 0
>
-
20
"e :L ~
n=58
D
- = D 15 Z ".... :J:" U
A. D
~ E 10 ~ c C ~
o ..
~
:J:
5
U
0 0
5
10
15
20
25
30
CHLORAMPHENICOL (UG/ML)
microbiological assay Discussion
The described affinity procedure for partial purification of CATase proved to be convenient, rapid, and repeatable and yielded stable preparations. The procedure facilitated the use of this inactivating enzyme for measurement of chloramphenicol. Worthy of note is the observation that the synthesis of this chloramphenicol-inactivating enzyme is a constitutive property of the strain, as was observed by Shaw [26] for similar strains. This finding contrasts with the cases of Staphylococcus aureus and Staphylococcus epidermidis, for which rates of synthesis are maximal only when the bacteria are grown in the presence of an inducer [27, 28]. The enzyme appears to be wholly intracellular, since it is not released by means of techniques that specifically_ extract enzymes believed to exist at a periplasmic site [29]. The apparent affinity of CATase for chloramphenicol is comparable to that found in class I strains [19], but the MIC of chloramphenicol is fivefold higher in E. coli W677/HJR66. The higher MIC for this strain may be due to an in-
creased synthesis of CATase. No data are available for comparison of specific CATase activity per cell in different strains. As is true for other class I CATases [19], our preparation was not inhibited by 5,5'-dithio-bis2-nitrobenzoic acid. However, the enzyme was strikingly inhibited by sodium-p-hydroxymercuribenzoate, a fact indicating that at least one sulfhydryl group is necessary for enzymatic activity. The sensitivity of E. coli CATase to sulfhydryl inhibitors would thus be comparable to that of S. aureus and S. epidermidis [30]. The microbiological assay devised allowed more precise measurement of chloramphenicol than the method published by Mikhail et al. [31]. Derived from the work of Bennett et al. [32], a logarithmic multiple regression equation was established for the curvilinear relation between zone diameters and antibiotic concentrations. The use of a calculator that could be programmed eliminated the tedious manipulation and transformation of data and allowed the choice of any concentrations for production of a standard curve. It would certainly be of interest to see if the same
Downloaded from http://jid.oxfordjournals.org/ at University of Manitoba on September 6, 2015
Figure 5. Correlation of concentrations of chloramphenicol in serum (0) and urine (Do) from patients treated with chloramphenicol and penicillin G. Concentrations of chloramphenicol were measured by the microbiological technique (abscissa) after inactivation by penicillinase and by the enzymatic assay (ordinate). Dashed line indicates theoretical curve; solid line indicates experimental curve. The actual concentration of chloramphenicol in urine specimens was twice that indicated in the figure, since no correction was made for the onefold dilution with distilled water.
Enzymatic Assay for Chloramphenicol
References 1. Goodman, L. S., Gilman, A. The pharmacological basis of therapeutics. 4th ed. Macmillan, New York, 1970. 1,704 p. 2. Dunne, M., Herxheimer, A., Newman, M., Ridley, H. Indications and warnings about chloramphenicol. Lancet 2:781-783, 1973. 3. Heineman, H. S., Braude, A. I., Osterholm, J. L. Intracranial suppurative disease. Early presumptive diagnosis and successful treatment without surgery. I.A.M.A. 218: 1542-1547, 1971. 4. Samson, D. S., Clark, K. A current review of brain abscess. Am. J. Med. 54:201-210, 1973.
5. Shackelford, P. G., Bobinski, J. E., Feigin, R. D., Cherry, J. D. Therapy of Haemophilus infiuenzae meningitis reconsidered. N. Engl. J. Med. 287: 634-638, 1972. 6. Roy, T. E., Krieger, E., Craig, G., Cohen, D., McNaughton, G. A., Silverthorne, N. Studies on the absorption of chloramphenicol in normal children in relation to the treatment of meningitis. Antibiot. Chemother. (Basel) 2: 505-516, 1952. 7. Kelly, R. S., Hunt, A. D., Jr., Tashman, S. G. Studies on the absorption and distribution of chloramphenicol. Pediatrics 8:362-367, 1951. 8. McCrumb, F. R., Hall, H. E., Merideth, A. M., Deane, G. E., Minor, J. V., Woodward, T. E. Chloramphenicol in the treatment of meningococcal meningitis. Am. J. Med. 10:696-703, 1951. 9. Schoenbach, E. B., Spencer, H. c., Monnier, J. Treatment of Haemophilus tnfiuenzae meningitis with aureomycin and chloramphenicol. Am. J. Med. 12:263-276, 1952. 10. Kramer, P. W., Griffith, R. S., Campbell, R. L. Antibiotic penetration of the brain: a comparative study. J. Neurosurg. 31 :295-302, 1969. 11. Black, P., Graybill, J. R., Charache, P. Penetration of brain abscess by systemically administered antibiotics. J. Neurosurg. 38:705-709, 1973. 12. Khan, W., Ross, S., Rodriguez, W., Cotroni, G., Saz, A. K. Haemophilus influenzae type B resistant to ampicillin. J.A.M.A. 229:298-301, 1974. 13. Nelson, J. D. Should ampicillin be abandoned for treatment of Haemophilus infiuenzae disease? I.A.M.A. 229:322-324, 1974. 14. Lietman, P. S., White, T. J., Shaw, W. V. Chloramphenicol: an enzymologic assay. Abstracts of the Annual Meeting of the American Society for for Microbiology, 1973, p. 107. 15. Shaw, W. V. Enzymatic chloramphenicol acetylation and R factor induced antibiotic resistance in Enterobacteriaceae. Antimicrob. Agents Chemother. 1966:221-226, 1967. 16. Shaw, W. V. The enzymatic acetylation of chloramphenicol by extracts of R factor-resistant Escherichia coli. J. BioI. Chern. 242:687-693, 1967. 17. Suzuki, Y., Okamoto, S. The enzymatic acetylation of chloramphenicol by the multiple drug-resistant Escherichia coli carrying R factor. J. BioI. Chern. 242:4722-4730, 1967. 18. Smith, A. L., Smith, D. H. Gentamicin:adenine mononucleotide transferase: partial purification, characterization, and use in the clinical quantitation of gentamicin. J. Infect. Dis. 129: 391-401, 1974. 19. Foster, T. J., Shaw, W. V. Chloramphenicol acetyltransferases specified by fi-: R factors. Antimicrob. Agents Chemother. 3:99-104, 1973. 20. Guitard, M., Daigneault, R. Purification of Escherichia coli chloramphenicol acetyltransferase by
Downloaded from http://jid.oxfordjournals.org/ at University of Manitoba on September 6, 2015
regression equation applies to all growth-inhibition-type assays. As has already been demonstrated for assays of aminoglycosides [18, 24], the enzymatic assay for chloramphenicol compared favorably with a microbiological method. The utilization of CATase offers rapid, precise, and specific measurement of chloramphenicol in human serum or urine even in the presence of other antibiotics. The method was unaffected by reagents used as preservatives of blood or as anticoagulants and by abnormally high levels of bilirubin, blood urea nitrogen, or triglycerides. The sensitivity of the method was even greater than that of the gasliquid chromatography procedure [33], since measurements in the ng/rnl range could be obtained when a high-specific-activity substrate (acetyl coenzyme A) was used. This modification might be used for measurement of chloramphenicol in as little as 1 111 of biological fluid. The enzymatic method may prove equally useful for measurement of thiamphenicol since CATase acetylates this compound readily. The enzymatic assay correlated well with the microbiological assay, for both urine and serum. The assay is thus specific for biologically active chloramphenicol and does not measure chloramphenicol palmitate, chloramphenicol succinate, or the 3-monoglucuronide derivative of chloramphenicol, the major excretory product of chloramphenicol in human urine [34]. These findings were expected since the 3-hydroxyl group is essential for acetylation by CATase [28, 35]. After two years of good results with the enzymatic assay, we certainly recommend the utilization of this procedure during therapy with chloramphenicol.
521
Daigneault and Guitard
522
21.
22.
23.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
chloramphenicol acetyltransferase. Antimicrob. Agents Chemother. 1968:7-10, 1969. Nossal, N. G., Heppel, L. A. The release of enzymes by osmotic shock from Escherichia coli in exponential phase. J. BioI. Chern. 241: 3055-3062, 1966. Shaw, W. V., Bentley, D. W., Sands, L. Mechanism of chloramphenicol resistance in Staphylococcus epidermis. J. Bacteriol. 104: 1095-1105, 1970. Mikhail, I. A., Kent, D. c., Sorensen, K., Sanborn, W. R., Smith, J. Concentrations of ampicillin and chloramphenicol in the serum of patients with acute Salmonella enteric fever. Antimicrob. Agents Chemother. 2:336-339, 1972. Bennett, J. V., Brodie, J. L., Benner, E. J., Kirby, W. M. M. Simplified, accurate method for antibiotic assay of clinical specimens. Appl. MicrobioI. 14: 170-177, 1966. Resnick, G. L., Corbin, D., Sanberg, D. H. Determination of serum chloramphenicol utilizing gasliquid chromatography and electron capture spectrometry. Anal. Chern. 38:582-585, 1966. Glazko, A. J., Dill, W. A., Rebstock, M. C. Biochemical studies on chloramphenicol (Chlorornycetin) . III. Isolation and identification of metabolic products in urine. J. BioI. Chern. 183: 679-691, 1950. Shaw, W. V., Brodsky, R F. Chloramphenicol resistance by enzymatic acetylation: comparative aspects. Antimicrob. Agents Chemother. 1967: 257-263, 1968.
Downloaded from http://jid.oxfordjournals.org/ at University of Manitoba on September 6, 2015
24.
affinity chromatography. Can. J. Biochem. 52: 1087-1090. 1974. Lowry, O. H., Rosebrough, N. J., Farr, A. L., Randall, R J. Protein measurement with the Folin phenol reagent. J. Biol. Chern. 193: 265275, 1951. De Crombrugghe, B., Pastan, I., Shaw, W. V., Rosner, J. L. Stimulation by cyclic AMP and ppGpp of chloramphenicol acetyl transferase synthesis. Nature [New Biol.] 241 :237-239, 1973. Sabath, L. D., Casey, J. I., Ruch, P. A., Stumpf, L. L., Finland, M. Rapid microassay of gentamicin, kanamycin, neomycin, streptomycin, and vancomycin in serum or plasma. J. Lab. Clin, Med. 78:457-463, 1971. Daigneault, R, Gagne, M., Brazeau, M. A comparison of two methods of gentamicin assay: an enzymatic procedure and an agar diffusion technique. J. Infect. Dis. 130:642-645, 1974. O'Gorman Hughes, D. W., Diamond, L. K. Chloramphenicol in blood: simple chemical estimations in patients receiving multiple antibiotics. Science 144:296-297, 1964. Shaw, W. V. Comparative enzymology of chloramphenicol resistance. Ann. N.Y. Acad. Sci. 182:234-242, 1971. Winshell, E., Shaw, W. V. Kinetics of induction and purification of chloramphenicol acetyltransferase from chloramphenicol-resistant Staphylococcus aureus. J. Bacteriol. 98: 1248-1257, 1969. Shaw, W. V., Winshell, E. Chloramphenicol resistance of Staphylococcus aureus: inducers of
MAY 1976
An Enzymatic Assay for Chloramphenicol with Partially Purified Chloramphenicol Acetyltransferase From the Biochemistry Service, Notre-Dame Hospital, Montreal, Canada
Rejean Daigneault and Michel Guitard
Chloramphenicol therapy remains controversial in several circumstances [1] because of its toxicity and the availability of other effective broad-spectrum antibiotics. However, the use of chloramphenicol. alone or with other antibiotics, for periods of up to six weeks has been recommended for infections of the central nervous system [2-5]. This recommendation is based on the good penetration of chloramphenicol into the cerebrospinal fluid [6-9], cerebral tissue [10], and brain abscess [11]. With the recent appearance of ampicillin-resistant strains of Haemophilus iniiuenzae type b, an increase in utilization of chloramphenicol for treatment of bacterial meningitis is predictable [12, 13]. A rapid, reliable, and precise assay for chloramphenicol in biological fluids may constitute a valuable adjunct by means of which the clinician can carefully monitor levels of chioram-
Received for publication June 30, 1975, and in revised form December 17, 1975. This investigation was supported by grant no. MA-5016 from the Medical Research Council of Canada and by a grant from the Conseil de la Recherche Medicate du Quebec. We thank Dr. Michel Brazeau for many stimulating discussions, for reviewing the manuscript, and for permitting us to analyze serum and urine specimens from his patients. We gratefully acknowledge the valuable technical assistance of Dorothee Lemieux. Please address requests for reprints to Dr. Rejean Daigneault, Service de Biochimie, Hopital Notre-Dame, 1560 Sherbrooke est, Montreal H2L 4Ml, Canada.
phenicol in the blood of patients with impaired renal and/or hepatic function. We report herein an enzymatic assay for chloramphenicol based on the use of chloramphenicol acetyltransferase (CATase) [14]. This inactivating enzyme catalyzes the acetylation of chloramphenicol in the presence of acetyl coenzyme A to yield the 3-acetyl and 1,3-diacetyl esters [15-17]. We describe some properties of the partially purified CATase preparation and correlate results of the enzymatic assay with those of a microbiological method developed in our laboratory. Materials and Methods
Bacterial strains. Escherichia coli W677/ HJR66, an R-factor-bearing mutant [18] (kindly provided by Dr. D. H. Smith, Children's Hospital Medical Center, Boston, Mass.) served as the source of CATase.! The MIC of chloramphenicol for this strain is 1 mg/ml. Sarcina lutea (ATCC 9341) was used in the microbiological assay of chloramphenicol. Culture method and preparation of the cell-free extract. An overnight culture of E. coli in trypticase soy broth (Baltimore Biological Laboratories, Baltimore, Md.) was diluted 100-fold in 2,800 ml of the same nutrient broth and incubated at 37 C with magnetic agitation until late logarith-
1 We will gladly supply E. coli W677 /HJR66 and the HP-65 program listing to interested investigators upon request.
515
Downloaded from http://jid.oxfordjournals.org/ at University of Manitoba on September 6, 2015
In an enzymatic assay for chloramphenicol, chloramphenicol acetyltransferase partially purified by affinity chromatography was used; [3H]acetyl coenzyme A served as a substrate. The purified enzyme was sensitive to p-hydroxymercuribenzoate but insensitive to 5,5'-dithio-bis-2-nitrobenzoic acid. The Michaelis-Menten constant was 10.5 !J.M. The operating range of the enzymatic assay was 0-30 !J.gl ml. The coefficient of variation was 1.9% at a concentration of 10 !J.gl ml. The method correlated well with a microbiological agar-paper disk method for mock unknown sera (r = 0.997) and serum or urine specimens (r = 0.993). The enzymatic assay was unaffected by the presence of any of 12 other antibiotics in tested serum.
516
~g of chloramphenicol/ml. The standards were kept frozen at -70 C in 0.5-ml aliquots until use. The pool of serum was tested for its absence of antimicrobial activity. Forty-six mock unknown sera with drug concentrations ranging from 5 to 30 ug/rnl were also prepared by appropriate dilution of the stock solution (100 ug/rnl) with human pooled serum. Specimens of serum or urine from patients receiving both chloramphenicol and penicillin 0 were either processed the same day or kept frozen at - 20 C until assay. Enzymatic assay of chloramphenicol. CATase was incubated with [3H]acetyl coenzyme A under conditions such that the concentration of chloramphenicol was the rate-limiting factor of the reaction. Radioactivity of the 3-acetyl and 1,3-diacetyl esters of chloramphenicol produced was measured after extraction by benzene at slightly alkaline pH. Under these conditions, acetyl coenzyme A remained in the aqueous phase [22]. The reaction mixture contained 100 ul of 0.25 M Tris-HCI (pH 7.8), 35 f.!l of a 0.35-mM PH] acetyl coenzyme A solution (specific activity, 14.3 mCi/mmol; New England Nuclear, Boston, Mass.), 50 milliunits of purified CATase, and 50 ~I of chloramphenicol standard, mock unknown, undiluted serum specimen, or urine specimen diluted with an equal volume of distilled water to a final volume of 250 ~l. The reaction mixture, prepared at 0 C in 5-ml conical test tubes, was incubated for 20 min at 37 C. Extraction with 1 ml of benzene (Fisher, scintillation grade) stopped the reaction by precipitation of the proteins and brought the [3H]acetyl chloramphenicol and [1,3-3H]diacetyl chloramphenicol into the organic phase. This step was repeated twice. The combined organic-phase material (3 ml) was transferred to a glass scintillation vial and acidified with 0.1 ml of glacial acetic acid. The benzene was evaporated to dryness at 70 C, and 10 ml of a toluene-based scintillant was added to the vial. Radioactivity was counted in a Beckman LS 330 liquid scintillation spectrometer. All assays were performed in duplicate. For evaluation of the effect of other antibiotics on the enzymatic assay of chloramphenicol, the reaction mixture was modified as follows. The source of antibiotic was prepared by successive addition of chloramphenicol and a second anti-
Downloaded from http://jid.oxfordjournals.org/ at University of Manitoba on September 6, 2015
mic phase (A = 1.5 at 490 nm). The cells were then harvested by centrifugation at 10,000 g for 10 min at 2 C, washed twice with 0.01 M Tris-HCI (pH 7.8; Fisher Scientific, Fair Lawn, N.J.), and resuspended in 40 ml of the same buffer. The bacteria were disrupted at 0 C by continuous sonication for 3 min at 70 W with a Biosonic sonifier (Artek Systems, Farmingdale, N.Y.). Cell debris was removed by centrifugation at 30,000 g for 30 min at 2 C, and the supernatant fluid (cell-free extract) was stored in O.5-ml aliquots at - 70 C. Assay of CATase. CATase was assayed at 37 C by the colorimetric method of Foster and Shaw [19] with a Beckman Acta III spectrophotometer (Beckman Instruments, Fullerton, Ca.) [20]. A unit of enzyme activity represents 1 umol of coenzyme A liberated per min at 37 C. The same spectrophotometric method was used for determination of the Michaelis-Menten constant (Ktn ) and the effect of sodium p-hydroxymercuribenzoate (Sigma Chemical, St. Louis, Mo.) and 5,5'dithio-bis-2-nitrobenzoic acid (Sigma) on the reaction catalyzed by CATase. Purification of CA'I'ase by affinity chromatography. CATase was partially purified as described previously [20] with use of a Sepharosereduced chloramphenicol affinity column and elution by chloramphenicol competition. The eluate fractions containing the enzyme were pooled and concentrated 10-fold with a Minicon B 15 clinical ultrafiltration system (Amicon Corp., Lexington, Mass. ). Chloramphenicol was eliminated by passage of the purified CATase concentrate through a Sephadex 0-10 column (0.5 X 20 em; Pharmacia Fine Chemicals, Uppsala, Sweden) at 2 C, equilibrated with 0.1 M Tris-HCI and 0.05 M NaCI buffer at pH 7.8. The enzyme peak eluted in the void volume of the column was stored in O.5-ml aliquots at - 70 C. The amount of protein in the bacterial extract and in the purified enzyme preparations was measured according to the method of Lowry et al. [21]. Chloramphenicol standards, mock unknown sera, and human serum or urine specimens. Chloramphenicol (Parke, Davis, Detroit, Mich.) was dissolved in pooled serum obtained from healthy individuals to make a stock solution of 100 ug/rnl. The stock solution was diluted with the same pool of serum to make the following standards: 0, 2.5, 5.0, 10.0, 15.0, 20.0, and 30.0
Daigneault and Guitard
Enzymatic Assay for Chloramphenicol
517
biotic to pooled serum at 0 C; this preparation was incubated at 37 C for 30 min. Fifty microliters of this antibiotic source were processed as described previously for the assay of chloramphenicol. The amount of antibiotic in the 50-~tl fraction was expressed as a concentration (table 1) .
Effect of other antibiotics on the enzymatic assay of chloramphenicol (concentration, 10
Table 1. ug/rnl).
Antibiotic (ug/rnl ) Sulfathiazole (100) Erythromycin (5) Tetracycline . HCl (12) Minocycline . HCl (5) Kanamycin (10) Gentamicin (10) Tobramycin (10) Cephalexin (50) Cephaloridine (50) Ampicillin (25) Penicillin (100) Clindamycin (8)
Chloramphenicol measured (ug/rnl) 9.8 9.8 9.7 10.0 9.9 9.9 10.0 10.0 9.8 9.9 10.1 9.8
+
+
y
== e
[ -b+v'b 2-4ac+4cx] 2c
Results
The results of CATase purification by affinity chromatography are not presented herein since they were identical to those reported previously [20]. Concentration of the partially purified enzyme and elimination of chloramphenicol from CATase by gel filtration resulted in the total recovery of the initial level of CATase activity. The same affinity column has been used for more than one year without an appreciable loss of enzyme retention capability. More than 100 affinity chromatography separations have been achieved with this same column. All experiments were done with preparations that had a specific activity of > 125 units /mg of protein. 2
See footnote 1.
Downloaded from http://jid.oxfordjournals.org/ at University of Manitoba on September 6, 2015
The specificity of the enzymatic assay was tested against chloramphenicol palmitate (Parke, Davis) , chloramphenicol succinate (Pentagone Laboratories, Montreal, Canada) , and thiamphenicol (Sterling Winthrop Research Institute, Rensselaer, N.Y.) at a concentration of 20 ug/rnl prepared in human pooled serum. The specimens were processed as described previously for the assay of chloramphenicol. The operating range for the enzymatic assay was 0-30 ug/rnl. The following modification in the composition of the reaction mixture allowed measurements within the range 0-40 ng/rnl: 25 III of 1 M Tris-HCI (pH 7.8), 20 III of a 5-IlM [3H]acetyl coenzyme A solution (specific activity, 900 mCi/mmol), 5 III of purified CATase (5 milliunits), and 200 III of chloramphenicol standard. Microbiological assay of chloramphenicol. Each specimen or chloramphenicol standard (250 Ill) was incubated at 25 C for 30 min with 50 III of penicillinase (Bacto-penase concentrate penicillinase, 500,000 units/rnl; Difco Laboratories, Detroit, Mich.). The assay agar was prepared by mixture of 40 g of trypticase soy agar (Baltimore Biological Laboratories) with 1 liter of distilled
water. The mixture was brought to the boiling point until the medium cleared and was then autoclaved. A suspension of S. lutea (5 ml) was prepared from a blood agar slant in trypticase soy broth and adjusted to an absorbance of 24 at 490 nm. This amount was added to 100 ml of melted agar kept at 56 C. The seeded agar (12 ml) was immediately poured into 150- X 15-mm standard disposable plastic Petri dishes and allowed to solidify on a level surface. The assay, a modification of the agar-paper disk diffusion method of Sabath et al. [23], was performed by pipetting of 50 III of each penicillinase-treated specimen onto a 0.25-inch Bacto concentration sterile blank disk (Difco Laboratories), which was on the surface of the seeded agar plate. Each sample was tested in duplicate. The diameter of the zone of growth inhibition was measured after incubation for 20 hr at 37 C. Results were computed with a Hewlett-Packard 65 pocket calculator programmed- for resolving the multiple regression equation x == a b (In y) c (In y) 2, where x is the zone diameter in millimeters and y the concentration of chloramphenicol in ug/rnl. After x and y values were entered for each standard, the calculator instantly generated coefficients a, b, c, the correlation coefficient r, and (given the zone diameter) the antibiotic concentration derived from the equation:
Daigneault and Guitard
518
-
100
100
e-
';
'';:;
u
ca CD
en
....ca et:
50
Co)
::IE 50
~
ti
u ....::c
::I
'"CI
';;
= CD
25
0 0
5 pHMB (11M)
10
Figure 1. Effect of p-hydroxymercuribenzoate (pHMB) on the activity of chloramphenicol acetyltransferase (CATase). Each point represents a mean of two determinations of residual CATase activity after inhibition by pHMB under the conditions described in Materials and Methods.
5
10
15
20
25
30
CHLORAMPHENICOL IN 1l./ml
Figure 3. Typical standard curve for chloramphenicol assayed by the enzymatic method. Each standard was analyzed in duplicate, and each point represents the mean of the radioactivity measurements.
Downloaded from http://jid.oxfordjournals.org/ at University of Manitoba on September 6, 2015
Properties of CATase. The enzyme preparations were shown to be stable for at least three months when kept at -70 C. Stability was not measured for preparations kept for a longer period. Repeated freezing and thawing resulted in drastic losses of enzyme activity. The osmotic lysate of E. coli W677/HJR66 used in our laboratory for the quantitation of gentamicin [24] did not contain any CATase activity, a fact suggesting an intracellular rather than a periplasmic location for this enzyme. The K m of the purified CATase for chloramphenicol was 10.5 ~M. CATase was not sensitive to 5,5'-dithio-bis-2-nitrobenzoic acid; incubation of the enzyme with this acid at 37 C for 15 min resulted in a loss of no more than 10% of the initial activity. The enzyme was found to be sensitive to sodium p-hydroxymercuribenzoate; CATase activity was inhibited by 50% at a concentration of 3.3 f.1M (figure 1). The addition of dithiothreitol or 2-mercaptoethanol at a concentration of up to 10 mM did not reactivate the enzyme. Assay of chloramphenicol in human serum. Standard curves. The operating range of the agar-paper disk diffusion technique was 5-30 ug/rnl. A typical standard curve is shown in fig-
519
Enzymatic Assay for Chloramphenicol
amphenicol in the ng/rnl range. A linear relation was observed between radioactivity in the slightly alkaline benzene extract and concentrations of chloramphenicol of 0-40 ng/rnl (figure 4). Precision. Within-run repeatability was estimated by measurement of the same mock unknown 10 times at a concentration of 10 ug/rnl, The coefficient of variation was calculated as 100
_I '2. (x -
x l
x)2
n-l
'
100
-
75
:E a.
50
0 0
>
-
20
"e :L ~
n=58
D
- = D 15 Z ".... :J:" U
A. D
~ E 10 ~ c C ~
o ..
~
:J:
5
U
0 0
5
10
15
20
25
30
CHLORAMPHENICOL (UG/ML)
microbiological assay Discussion
The described affinity procedure for partial purification of CATase proved to be convenient, rapid, and repeatable and yielded stable preparations. The procedure facilitated the use of this inactivating enzyme for measurement of chloramphenicol. Worthy of note is the observation that the synthesis of this chloramphenicol-inactivating enzyme is a constitutive property of the strain, as was observed by Shaw [26] for similar strains. This finding contrasts with the cases of Staphylococcus aureus and Staphylococcus epidermidis, for which rates of synthesis are maximal only when the bacteria are grown in the presence of an inducer [27, 28]. The enzyme appears to be wholly intracellular, since it is not released by means of techniques that specifically_ extract enzymes believed to exist at a periplasmic site [29]. The apparent affinity of CATase for chloramphenicol is comparable to that found in class I strains [19], but the MIC of chloramphenicol is fivefold higher in E. coli W677/HJR66. The higher MIC for this strain may be due to an in-
creased synthesis of CATase. No data are available for comparison of specific CATase activity per cell in different strains. As is true for other class I CATases [19], our preparation was not inhibited by 5,5'-dithio-bis2-nitrobenzoic acid. However, the enzyme was strikingly inhibited by sodium-p-hydroxymercuribenzoate, a fact indicating that at least one sulfhydryl group is necessary for enzymatic activity. The sensitivity of E. coli CATase to sulfhydryl inhibitors would thus be comparable to that of S. aureus and S. epidermidis [30]. The microbiological assay devised allowed more precise measurement of chloramphenicol than the method published by Mikhail et al. [31]. Derived from the work of Bennett et al. [32], a logarithmic multiple regression equation was established for the curvilinear relation between zone diameters and antibiotic concentrations. The use of a calculator that could be programmed eliminated the tedious manipulation and transformation of data and allowed the choice of any concentrations for production of a standard curve. It would certainly be of interest to see if the same
Downloaded from http://jid.oxfordjournals.org/ at University of Manitoba on September 6, 2015
Figure 5. Correlation of concentrations of chloramphenicol in serum (0) and urine (Do) from patients treated with chloramphenicol and penicillin G. Concentrations of chloramphenicol were measured by the microbiological technique (abscissa) after inactivation by penicillinase and by the enzymatic assay (ordinate). Dashed line indicates theoretical curve; solid line indicates experimental curve. The actual concentration of chloramphenicol in urine specimens was twice that indicated in the figure, since no correction was made for the onefold dilution with distilled water.
Enzymatic Assay for Chloramphenicol
References 1. Goodman, L. S., Gilman, A. The pharmacological basis of therapeutics. 4th ed. Macmillan, New York, 1970. 1,704 p. 2. Dunne, M., Herxheimer, A., Newman, M., Ridley, H. Indications and warnings about chloramphenicol. Lancet 2:781-783, 1973. 3. Heineman, H. S., Braude, A. I., Osterholm, J. L. Intracranial suppurative disease. Early presumptive diagnosis and successful treatment without surgery. I.A.M.A. 218: 1542-1547, 1971. 4. Samson, D. S., Clark, K. A current review of brain abscess. Am. J. Med. 54:201-210, 1973.
5. Shackelford, P. G., Bobinski, J. E., Feigin, R. D., Cherry, J. D. Therapy of Haemophilus infiuenzae meningitis reconsidered. N. Engl. J. Med. 287: 634-638, 1972. 6. Roy, T. E., Krieger, E., Craig, G., Cohen, D., McNaughton, G. A., Silverthorne, N. Studies on the absorption of chloramphenicol in normal children in relation to the treatment of meningitis. Antibiot. Chemother. (Basel) 2: 505-516, 1952. 7. Kelly, R. S., Hunt, A. D., Jr., Tashman, S. G. Studies on the absorption and distribution of chloramphenicol. Pediatrics 8:362-367, 1951. 8. McCrumb, F. R., Hall, H. E., Merideth, A. M., Deane, G. E., Minor, J. V., Woodward, T. E. Chloramphenicol in the treatment of meningococcal meningitis. Am. J. Med. 10:696-703, 1951. 9. Schoenbach, E. B., Spencer, H. c., Monnier, J. Treatment of Haemophilus tnfiuenzae meningitis with aureomycin and chloramphenicol. Am. J. Med. 12:263-276, 1952. 10. Kramer, P. W., Griffith, R. S., Campbell, R. L. Antibiotic penetration of the brain: a comparative study. J. Neurosurg. 31 :295-302, 1969. 11. Black, P., Graybill, J. R., Charache, P. Penetration of brain abscess by systemically administered antibiotics. J. Neurosurg. 38:705-709, 1973. 12. Khan, W., Ross, S., Rodriguez, W., Cotroni, G., Saz, A. K. Haemophilus influenzae type B resistant to ampicillin. J.A.M.A. 229:298-301, 1974. 13. Nelson, J. D. Should ampicillin be abandoned for treatment of Haemophilus infiuenzae disease? I.A.M.A. 229:322-324, 1974. 14. Lietman, P. S., White, T. J., Shaw, W. V. Chloramphenicol: an enzymologic assay. Abstracts of the Annual Meeting of the American Society for for Microbiology, 1973, p. 107. 15. Shaw, W. V. Enzymatic chloramphenicol acetylation and R factor induced antibiotic resistance in Enterobacteriaceae. Antimicrob. Agents Chemother. 1966:221-226, 1967. 16. Shaw, W. V. The enzymatic acetylation of chloramphenicol by extracts of R factor-resistant Escherichia coli. J. BioI. Chern. 242:687-693, 1967. 17. Suzuki, Y., Okamoto, S. The enzymatic acetylation of chloramphenicol by the multiple drug-resistant Escherichia coli carrying R factor. J. BioI. Chern. 242:4722-4730, 1967. 18. Smith, A. L., Smith, D. H. Gentamicin:adenine mononucleotide transferase: partial purification, characterization, and use in the clinical quantitation of gentamicin. J. Infect. Dis. 129: 391-401, 1974. 19. Foster, T. J., Shaw, W. V. Chloramphenicol acetyltransferases specified by fi-: R factors. Antimicrob. Agents Chemother. 3:99-104, 1973. 20. Guitard, M., Daigneault, R. Purification of Escherichia coli chloramphenicol acetyltransferase by
Downloaded from http://jid.oxfordjournals.org/ at University of Manitoba on September 6, 2015
regression equation applies to all growth-inhibition-type assays. As has already been demonstrated for assays of aminoglycosides [18, 24], the enzymatic assay for chloramphenicol compared favorably with a microbiological method. The utilization of CATase offers rapid, precise, and specific measurement of chloramphenicol in human serum or urine even in the presence of other antibiotics. The method was unaffected by reagents used as preservatives of blood or as anticoagulants and by abnormally high levels of bilirubin, blood urea nitrogen, or triglycerides. The sensitivity of the method was even greater than that of the gasliquid chromatography procedure [33], since measurements in the ng/rnl range could be obtained when a high-specific-activity substrate (acetyl coenzyme A) was used. This modification might be used for measurement of chloramphenicol in as little as 1 111 of biological fluid. The enzymatic method may prove equally useful for measurement of thiamphenicol since CATase acetylates this compound readily. The enzymatic assay correlated well with the microbiological assay, for both urine and serum. The assay is thus specific for biologically active chloramphenicol and does not measure chloramphenicol palmitate, chloramphenicol succinate, or the 3-monoglucuronide derivative of chloramphenicol, the major excretory product of chloramphenicol in human urine [34]. These findings were expected since the 3-hydroxyl group is essential for acetylation by CATase [28, 35]. After two years of good results with the enzymatic assay, we certainly recommend the utilization of this procedure during therapy with chloramphenicol.
521
Daigneault and Guitard
522
21.
22.
23.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
chloramphenicol acetyltransferase. Antimicrob. Agents Chemother. 1968:7-10, 1969. Nossal, N. G., Heppel, L. A. The release of enzymes by osmotic shock from Escherichia coli in exponential phase. J. BioI. Chern. 241: 3055-3062, 1966. Shaw, W. V., Bentley, D. W., Sands, L. Mechanism of chloramphenicol resistance in Staphylococcus epidermis. J. Bacteriol. 104: 1095-1105, 1970. Mikhail, I. A., Kent, D. c., Sorensen, K., Sanborn, W. R., Smith, J. Concentrations of ampicillin and chloramphenicol in the serum of patients with acute Salmonella enteric fever. Antimicrob. Agents Chemother. 2:336-339, 1972. Bennett, J. V., Brodie, J. L., Benner, E. J., Kirby, W. M. M. Simplified, accurate method for antibiotic assay of clinical specimens. Appl. MicrobioI. 14: 170-177, 1966. Resnick, G. L., Corbin, D., Sanberg, D. H. Determination of serum chloramphenicol utilizing gasliquid chromatography and electron capture spectrometry. Anal. Chern. 38:582-585, 1966. Glazko, A. J., Dill, W. A., Rebstock, M. C. Biochemical studies on chloramphenicol (Chlorornycetin) . III. Isolation and identification of metabolic products in urine. J. BioI. Chern. 183: 679-691, 1950. Shaw, W. V., Brodsky, R F. Chloramphenicol resistance by enzymatic acetylation: comparative aspects. Antimicrob. Agents Chemother. 1967: 257-263, 1968.
Downloaded from http://jid.oxfordjournals.org/ at University of Manitoba on September 6, 2015
24.
affinity chromatography. Can. J. Biochem. 52: 1087-1090. 1974. Lowry, O. H., Rosebrough, N. J., Farr, A. L., Randall, R J. Protein measurement with the Folin phenol reagent. J. Biol. Chern. 193: 265275, 1951. De Crombrugghe, B., Pastan, I., Shaw, W. V., Rosner, J. L. Stimulation by cyclic AMP and ppGpp of chloramphenicol acetyl transferase synthesis. Nature [New Biol.] 241 :237-239, 1973. Sabath, L. D., Casey, J. I., Ruch, P. A., Stumpf, L. L., Finland, M. Rapid microassay of gentamicin, kanamycin, neomycin, streptomycin, and vancomycin in serum or plasma. J. Lab. Clin, Med. 78:457-463, 1971. Daigneault, R, Gagne, M., Brazeau, M. A comparison of two methods of gentamicin assay: an enzymatic procedure and an agar diffusion technique. J. Infect. Dis. 130:642-645, 1974. O'Gorman Hughes, D. W., Diamond, L. K. Chloramphenicol in blood: simple chemical estimations in patients receiving multiple antibiotics. Science 144:296-297, 1964. Shaw, W. V. Comparative enzymology of chloramphenicol resistance. Ann. N.Y. Acad. Sci. 182:234-242, 1971. Winshell, E., Shaw, W. V. Kinetics of induction and purification of chloramphenicol acetyltransferase from chloramphenicol-resistant Staphylococcus aureus. J. Bacteriol. 98: 1248-1257, 1969. Shaw, W. V., Winshell, E. Chloramphenicol resistance of Staphylococcus aureus: inducers of
