THE JOURNAL OF INFECTIOUS DISEASES. VOL. 136, NO.2. AUGUST 1977 © 1977 by the University of Chicago. All rights reserved.
Antifungal Activity of Four Tetracycline Analogues against Candida albicans in Vitro: Potentiation by Amphotericin B From the Division of Clinical Microbiology, Sidney Farber Cancer Institute, Boston, Massachusetts
Michael A. Lew, Kevin M. Beckett, and Myron J. Levin
Amphotericin B (AmB), in spite of a number of serious limitations, is the most widely employed antimicrobial agent for the treatment of systemic fungal infections [1, 2]. The low therapeutic ratio of AmB and its poor diffusion into body compartments (i.e., cerebrospinal fluid) impose severe limitations on its usefulness [3, 4]. Attempts have been made to improve the effectiveness of AmB by combining it with other antimicrobial agents and exploiting the permeability changes induced in the fungal cell membrane by polyene antimicrobial agents [5]. By analogy with one mechanism of antimicrobial synergism established in bacteria [6], it was anticipated that AmB would potentiate the antifungal activity of other antimicrobial agents by facilitating their entry into the fungal cell. Medoff et
al. [7, 8] and other workers [9-11] demonstrated that several genera of fungi were killed more readily by the combination of AmB and 5-fluorocytosine than by comparable levels of either drug alone. Rifampin, an antibacterial agent that by itself has no useful activity against fungi [8], also interacted synergistically with AmB in vitro against a number of medically important fungi [8, 11-14]. Both 5-fluorocytosine and rifampin have shown additive or synergistic effects when combined with AmB in the treatment of experimental fungal infections [15-18]. Tetracycline, like rifampin, has no useful activity against intact yeast cells, although it inhibits protein synthesis by cell-free yeast ribosomes [19]. Tetracycline and AmB acted synergistically against a strain of SacchaTomyces ceTevisiae, but the concentration of tetracycline employed (100 ~g/ml) was in excess of levels that are safely attainable in human serum [20, 21]. Nevertheless, tetracycline potentiated the effect of AmB in treatment of experimental coccidioidomycosis in mice [22]. These findings led us to investigate the ability of AmB to potentiate the antifungal activity of several tetracycline analogues. In this report we describe the fungicidal and fungistatic activity that resulted when four tetracycline analogues (tetracycline, demeclocycline, doxycycline, and minocycline) were combined with AmB and tested against 20 strains of Candida albicans.
Received for publication November 9, 1976, and in revised form February 28,1977. This study was supported by Infectious Disease Career Training Program grant no. 5 TOI AI 00350-09 from the National Institute of Allergy and Infectious Diseases; Clinical Cancer Research Center grant no. I POI CA 19589-01; and Cancer Center (Comprehensive) Support grant no. 5-P30 CA 06516-13 from the National Cancer Institute. We thank Dr. Steven Zinner and Mr. Richard Provonchee for their help with some of the techniques employed in these studies. Please address requests for reprints to Dr. Michael A. Lew, Division of Clinical Microbiology, Sidney Farber Cancer Institute, 44 Binney Street, Boston, Massachusetts 02115.
263
Downloaded from http://jid.oxfordjournals.org/ at East Carolina University on September 17, 2015
The antifungal activities of four tetracycline analogues in combination with amphotericin B (AmB) were determined against 20 strains of Candida albicans. When a microtiter checkerboard technique was used, minocycline (~10 ,ugjml) acted synergistically with AmB against all strains, whereas doxycycline had a reduced effect, and demeclocycline and tetracycline had no potentiating effect at this concentration. Killing-curve experiments with two strains of C. albicans demonstrated that the combination of minocycline and AmB produced a decrease in number of colonyforming units (du) of >2 logs in 4 hr and a 4-log decrease in du in 24 hr at concentrations (minocycline, 0.64 ,ugjml; AmB, 0.1 ,ugjml) that were subinhibitory when each agent was used alone and that are readily achieved in human serum and body fluids with conventional doses. The killing-curve technique indicated that doxycycline had an intermediate degree of synergistic activity, whereas tetracycline had no synergistic activity at clinically relevant concentrations.
264
Materials and Methods
YNB, adjusted to an OD of 0.2 at 520 nm, and diluted 1:10 in YNB to yield an inoculum of 1-5 X 105 cfu/ml. The precise inoculum was determined by dilution and plating on Sabouraud's dextrose agar. The MIC was defined as the lowest concentration of antimicrobial agent that inhibited macroscopic growth after incubation at 35 C for 18 hr [26]. Aliquots (5 Itl) were removed from wells showing no macroscopic growth and were cultured on Sabouraud's dextrose agar for 48 hr. The MFC was defined as the lowest concentration of antimicrobial agent that produced a 3-log or greater decrease in number of organisms in the inoculum [27]. Determinations of the MIC and lVIFC for tetracycline analogues were performed in duplicate; one set of wells contained added DMSO (0.3%). Synergism studies. The interaction between AmB and tetracycline analogues was studied by a checkerboard microtiter dilution method [2729] with use of the end points described above. Serial twofold dilutions of AmB prepared with 25-1t1 microdilutors were placed in horizontal rows; the lefthand vertical row was left free for measurement· of tetracycline analogue activity. Twofold dilutions of tetracycline analogue were performed in test tubes, and 25 Itl of each dilution was then transferred to horizontal rows; the uppermost row was left free for measurement of AmB activity. Standardized inocula were then added to each well in 50-1t1 volumes. Antimicrobial synergism was defined as a fourfold or greater reduction in the MIC and/or MFC of each drug in the combination as compared with the corresponding values for each agent employed singly [27-29]. Killing-curve studies. Cultures (20 ml) were incubated in 50-ml Erlenmeyer flasks in a gyrorotary shaker (New Brunswick Scientific Co., New Brunswick, N.J.) at 35 C. Antimicrobial agents were added to flasks in 0.5-ml volumes. At working concentrations of AmB (0.1 Itg/ml), the concentration of DMSO was 0.003%. Three control flasks were included with each experiment; one contained no drug, a second flask contained tetracycline analogue (20 p.,g/ml) with 0.003% DMSO, and a third contained AmB (0.1 Itg/ml). The other flasks contained AmB (O.lltg/ml) plus graded increases of tetracycline analogue (0.0120 Itg/ml). At zero-time, I ml of inoculum (5 X 105 cfu/ ml) was added to each flask, and, after
Downloaded from http://jid.oxfordjournals.org/ at East Carolina University on September 17, 2015
Organisms. Eighteen strains of C. albicans were isolated from clinical sources in the Laboratory of Clinical :Microbiology at the Sidney Farber Cancer Institute (SFCI). All strains formed germ tubes when incubated at 35 C for 4 hr in human serum and assimilated sucrose [23]. Reference strains NIH 792 and NIH 20 (originally from Dr. H. F. Hasenclever; National Institutes of Health, BJ:thesda, ~'Id.) were provided by Dr. Helen Buckley (Harvard School of Public Health, Boston, Mass.). Antimicrobial agents. AmB was supplied as a powdered reference standard by E. R. Squibb and Sons (Princeton, N.J.). The powder was dissolved (25.6 mg/ml) in dimethylsulfoxide (DMSO) and stored at -20 C [24]. The drug remained fully active for a mInImUm of eight weeks, as indicated by determinations of MICs and minimal fungicidal concentrations (MFCs) against a reference strain of S. cerevisiae (ATCC 9763; American Type Culture Collection, Rockville, Md.). Working concentrations of AmB were prepared on the day of use by thawing the stock solution, diluting it 1: lOin 0.1 M phosphate buffer (pH 8.0) containing 60% DMSO, and making a further 1: 100 dilution in phosphate buffer to yield a concentration of 25.6 Itg/ml [24]. The maximal concentration of DMSO was 0.3% at working concentrations of ArnE. Tetracycline, demeclocycline, and minocycline were supplied as powdered reference standards by Lederle Laboratories (Pearl River, N.Y.). Doxycycline hyclate (926 Itg/mg) was obtained from Rachelle Laboratories (Long Beach, Calif.). Tetracycline (2,560 Itg/ml) was dissolved in 0.01 N HCl for microtiter experiments and in 0.1 M phosphate buffer (pH 6.0, 800 Itg/ml) for killing-curve experiments [25]. The remaining three analogues were dissolved at the above concentrations in phosphate buffer. Analogue solutions were stored at -20 C until use. Determinations of MICs and MFCs. All experiments were performed in yeast nitrogen base (YNB; Difco, Detroit, Mich.) supplemented with 0.5% glucose. MICs and MFCs of tetracycline analogues and AmB were determined against all isolates by a modified microtiter technique [25]. Organisms were grown overnight in
Lew, Beckett, and Levin
265
Antifungal Activity of Tetracyclines
Results
Sensitivity of C. albicans to individual antimicrobial agents. None of the tetracycline analogues tested inhibited or killed any strains of C. albicans at clinically achievable concentrations [21]. Tetracycline, with a median MIC of 320 JLg/ml (range, 320-2,560 JLg/ml) and a median MFC of 640 JLg/ml (range, 320-2,560 JLg/ml), had the greatest activity of the analogues tested. Demeclocycline, at a concentration of 640 JLg/ml, inhibited and killed one strain of C. albicans. Doxycycline and minocycline had intermediate levels of activity, with MICs confined largely to the range of 640-1,280 JLg/ml. The addition of DMSO did not influence activity of the analogues. AmB inhibited 14 strains at 0.8 JLg/ml and the other strains at 1.6 JLg/ml. Four strains were killed by concentrations of 0.8 JLg/ml, 14 by 1.6 JLg/ml, and two by 3.2 JLg/ml. Synergism between tetracycline analogues and AmB: checkerboard experiments. Minocycline was the only analogue that acted synergistically
with AmB against all strains when both the MIC and MFC were used as end points. Demeclocycline, despite having no individual activity against 19 of 20 strains, acted synergistically to inhibit 100% of strains (MIC) and to kill 85% of strains (MFC). Tetracycline and AmB displayed synergism against the smallest number of strains with respect to inhibition (60C;;o) and killing (65%). The minimal concentrations of tetracycline analogues' required to reduce'the -MIC and/or MFC of AmB at least fourfold are given in table 1. Minocycline clearly emerged as the most effective agent in potentiating the activity of AmB. This analogue produced a fourfold or greater reduction in the MIC of AmB against all strains at concentrations of ~10 JLg/ml. Doxycycline produced this effect against only three strains at 10 JLg/ml, whereas demeclocycline and tetracycline had no potentiating activity at levels of
Antifungal Activity of Four Tetracycline Analogues against Candida albicans in Vitro: Potentiation by Amphotericin B From the Division of Clinical Microbiology, Sidney Farber Cancer Institute, Boston, Massachusetts
Michael A. Lew, Kevin M. Beckett, and Myron J. Levin
Amphotericin B (AmB), in spite of a number of serious limitations, is the most widely employed antimicrobial agent for the treatment of systemic fungal infections [1, 2]. The low therapeutic ratio of AmB and its poor diffusion into body compartments (i.e., cerebrospinal fluid) impose severe limitations on its usefulness [3, 4]. Attempts have been made to improve the effectiveness of AmB by combining it with other antimicrobial agents and exploiting the permeability changes induced in the fungal cell membrane by polyene antimicrobial agents [5]. By analogy with one mechanism of antimicrobial synergism established in bacteria [6], it was anticipated that AmB would potentiate the antifungal activity of other antimicrobial agents by facilitating their entry into the fungal cell. Medoff et
al. [7, 8] and other workers [9-11] demonstrated that several genera of fungi were killed more readily by the combination of AmB and 5-fluorocytosine than by comparable levels of either drug alone. Rifampin, an antibacterial agent that by itself has no useful activity against fungi [8], also interacted synergistically with AmB in vitro against a number of medically important fungi [8, 11-14]. Both 5-fluorocytosine and rifampin have shown additive or synergistic effects when combined with AmB in the treatment of experimental fungal infections [15-18]. Tetracycline, like rifampin, has no useful activity against intact yeast cells, although it inhibits protein synthesis by cell-free yeast ribosomes [19]. Tetracycline and AmB acted synergistically against a strain of SacchaTomyces ceTevisiae, but the concentration of tetracycline employed (100 ~g/ml) was in excess of levels that are safely attainable in human serum [20, 21]. Nevertheless, tetracycline potentiated the effect of AmB in treatment of experimental coccidioidomycosis in mice [22]. These findings led us to investigate the ability of AmB to potentiate the antifungal activity of several tetracycline analogues. In this report we describe the fungicidal and fungistatic activity that resulted when four tetracycline analogues (tetracycline, demeclocycline, doxycycline, and minocycline) were combined with AmB and tested against 20 strains of Candida albicans.
Received for publication November 9, 1976, and in revised form February 28,1977. This study was supported by Infectious Disease Career Training Program grant no. 5 TOI AI 00350-09 from the National Institute of Allergy and Infectious Diseases; Clinical Cancer Research Center grant no. I POI CA 19589-01; and Cancer Center (Comprehensive) Support grant no. 5-P30 CA 06516-13 from the National Cancer Institute. We thank Dr. Steven Zinner and Mr. Richard Provonchee for their help with some of the techniques employed in these studies. Please address requests for reprints to Dr. Michael A. Lew, Division of Clinical Microbiology, Sidney Farber Cancer Institute, 44 Binney Street, Boston, Massachusetts 02115.
263
Downloaded from http://jid.oxfordjournals.org/ at East Carolina University on September 17, 2015
The antifungal activities of four tetracycline analogues in combination with amphotericin B (AmB) were determined against 20 strains of Candida albicans. When a microtiter checkerboard technique was used, minocycline (~10 ,ugjml) acted synergistically with AmB against all strains, whereas doxycycline had a reduced effect, and demeclocycline and tetracycline had no potentiating effect at this concentration. Killing-curve experiments with two strains of C. albicans demonstrated that the combination of minocycline and AmB produced a decrease in number of colonyforming units (du) of >2 logs in 4 hr and a 4-log decrease in du in 24 hr at concentrations (minocycline, 0.64 ,ugjml; AmB, 0.1 ,ugjml) that were subinhibitory when each agent was used alone and that are readily achieved in human serum and body fluids with conventional doses. The killing-curve technique indicated that doxycycline had an intermediate degree of synergistic activity, whereas tetracycline had no synergistic activity at clinically relevant concentrations.
264
Materials and Methods
YNB, adjusted to an OD of 0.2 at 520 nm, and diluted 1:10 in YNB to yield an inoculum of 1-5 X 105 cfu/ml. The precise inoculum was determined by dilution and plating on Sabouraud's dextrose agar. The MIC was defined as the lowest concentration of antimicrobial agent that inhibited macroscopic growth after incubation at 35 C for 18 hr [26]. Aliquots (5 Itl) were removed from wells showing no macroscopic growth and were cultured on Sabouraud's dextrose agar for 48 hr. The MFC was defined as the lowest concentration of antimicrobial agent that produced a 3-log or greater decrease in number of organisms in the inoculum [27]. Determinations of the MIC and lVIFC for tetracycline analogues were performed in duplicate; one set of wells contained added DMSO (0.3%). Synergism studies. The interaction between AmB and tetracycline analogues was studied by a checkerboard microtiter dilution method [2729] with use of the end points described above. Serial twofold dilutions of AmB prepared with 25-1t1 microdilutors were placed in horizontal rows; the lefthand vertical row was left free for measurement· of tetracycline analogue activity. Twofold dilutions of tetracycline analogue were performed in test tubes, and 25 Itl of each dilution was then transferred to horizontal rows; the uppermost row was left free for measurement of AmB activity. Standardized inocula were then added to each well in 50-1t1 volumes. Antimicrobial synergism was defined as a fourfold or greater reduction in the MIC and/or MFC of each drug in the combination as compared with the corresponding values for each agent employed singly [27-29]. Killing-curve studies. Cultures (20 ml) were incubated in 50-ml Erlenmeyer flasks in a gyrorotary shaker (New Brunswick Scientific Co., New Brunswick, N.J.) at 35 C. Antimicrobial agents were added to flasks in 0.5-ml volumes. At working concentrations of AmB (0.1 Itg/ml), the concentration of DMSO was 0.003%. Three control flasks were included with each experiment; one contained no drug, a second flask contained tetracycline analogue (20 p.,g/ml) with 0.003% DMSO, and a third contained AmB (0.1 Itg/ml). The other flasks contained AmB (O.lltg/ml) plus graded increases of tetracycline analogue (0.0120 Itg/ml). At zero-time, I ml of inoculum (5 X 105 cfu/ ml) was added to each flask, and, after
Downloaded from http://jid.oxfordjournals.org/ at East Carolina University on September 17, 2015
Organisms. Eighteen strains of C. albicans were isolated from clinical sources in the Laboratory of Clinical :Microbiology at the Sidney Farber Cancer Institute (SFCI). All strains formed germ tubes when incubated at 35 C for 4 hr in human serum and assimilated sucrose [23]. Reference strains NIH 792 and NIH 20 (originally from Dr. H. F. Hasenclever; National Institutes of Health, BJ:thesda, ~'Id.) were provided by Dr. Helen Buckley (Harvard School of Public Health, Boston, Mass.). Antimicrobial agents. AmB was supplied as a powdered reference standard by E. R. Squibb and Sons (Princeton, N.J.). The powder was dissolved (25.6 mg/ml) in dimethylsulfoxide (DMSO) and stored at -20 C [24]. The drug remained fully active for a mInImUm of eight weeks, as indicated by determinations of MICs and minimal fungicidal concentrations (MFCs) against a reference strain of S. cerevisiae (ATCC 9763; American Type Culture Collection, Rockville, Md.). Working concentrations of AmB were prepared on the day of use by thawing the stock solution, diluting it 1: lOin 0.1 M phosphate buffer (pH 8.0) containing 60% DMSO, and making a further 1: 100 dilution in phosphate buffer to yield a concentration of 25.6 Itg/ml [24]. The maximal concentration of DMSO was 0.3% at working concentrations of ArnE. Tetracycline, demeclocycline, and minocycline were supplied as powdered reference standards by Lederle Laboratories (Pearl River, N.Y.). Doxycycline hyclate (926 Itg/mg) was obtained from Rachelle Laboratories (Long Beach, Calif.). Tetracycline (2,560 Itg/ml) was dissolved in 0.01 N HCl for microtiter experiments and in 0.1 M phosphate buffer (pH 6.0, 800 Itg/ml) for killing-curve experiments [25]. The remaining three analogues were dissolved at the above concentrations in phosphate buffer. Analogue solutions were stored at -20 C until use. Determinations of MICs and MFCs. All experiments were performed in yeast nitrogen base (YNB; Difco, Detroit, Mich.) supplemented with 0.5% glucose. MICs and MFCs of tetracycline analogues and AmB were determined against all isolates by a modified microtiter technique [25]. Organisms were grown overnight in
Lew, Beckett, and Levin
265
Antifungal Activity of Tetracyclines
Results
Sensitivity of C. albicans to individual antimicrobial agents. None of the tetracycline analogues tested inhibited or killed any strains of C. albicans at clinically achievable concentrations [21]. Tetracycline, with a median MIC of 320 JLg/ml (range, 320-2,560 JLg/ml) and a median MFC of 640 JLg/ml (range, 320-2,560 JLg/ml), had the greatest activity of the analogues tested. Demeclocycline, at a concentration of 640 JLg/ml, inhibited and killed one strain of C. albicans. Doxycycline and minocycline had intermediate levels of activity, with MICs confined largely to the range of 640-1,280 JLg/ml. The addition of DMSO did not influence activity of the analogues. AmB inhibited 14 strains at 0.8 JLg/ml and the other strains at 1.6 JLg/ml. Four strains were killed by concentrations of 0.8 JLg/ml, 14 by 1.6 JLg/ml, and two by 3.2 JLg/ml. Synergism between tetracycline analogues and AmB: checkerboard experiments. Minocycline was the only analogue that acted synergistically
with AmB against all strains when both the MIC and MFC were used as end points. Demeclocycline, despite having no individual activity against 19 of 20 strains, acted synergistically to inhibit 100% of strains (MIC) and to kill 85% of strains (MFC). Tetracycline and AmB displayed synergism against the smallest number of strains with respect to inhibition (60C;;o) and killing (65%). The minimal concentrations of tetracycline analogues' required to reduce'the -MIC and/or MFC of AmB at least fourfold are given in table 1. Minocycline clearly emerged as the most effective agent in potentiating the activity of AmB. This analogue produced a fourfold or greater reduction in the MIC of AmB against all strains at concentrations of ~10 JLg/ml. Doxycycline produced this effect against only three strains at 10 JLg/ml, whereas demeclocycline and tetracycline had no potentiating activity at levels of
