Previous Article | Next Article ![]()
Antimicrobial Agents and Chemotherapy, June 2007, p. 1956-1961, Vol. 51, No. 6
0066-4804/07/$08.00+0 doi:10.1128/AAC.00062-07
Copyright © 2007, American Society for Microbiology. All Rights Reserved.

Basilea Pharmaceutica AG, Basel, Switzerland
Received 16 January 2007/ Returned for modification 18 February 2007/ Accepted 15 March 2007
|
|
|---|
|
|
|---|
Mild to moderate inflammatory acne sometimes responds well to topical treatment with antibacterial agents, usually erythromycin or clindamycin and less often tetracycline (33). Over the past two and a half decades the rate of resistance to erythromycin, clindamycin, and/or tetracycline has increased among propionibacterial populations associated with inflammatory acne and presents a worldwide problem for the treatment of this condition (12). Within Europe at least 50% of patients seen in dermatology clinics are colonized by strains resistant to erythromycin and/or clindamycin, whereas a lower prevalence of resistance (in up to 20% of patients, depending upon the location) has been reported for tetracycline (11).
As part of our ongoing drug discovery programs, attention was directed toward the identification of antibiotics that combine a relatively narrow spectrum of activity against propionibacteria, including erythromycin- and clindamycin-resistant strains, with good anti-inflammatory activity for projected topical application in patients with mild to moderate inflammatory acne. Here we report on the antibacterial activity of an experimental macrolide, BAL19403 (Fig. 1), against clinical isolates of propionibacteria, commensal skin microflora (corynebacteria, members of the family Micrococcaceae), and staphylococci and streptococci associated with skin and skin structure infections.
![]() View larger version (15K): [in a new window] |
FIG. 1. Chemical structure of BAL19403.
|
|
|
|---|
Most bacterial strains were from the culture collection of Basilea Pharmaceutica; some propionibacteria were generous gifts of J. H. Cove (Leeds, United Kingdom) and C. E. Nord (Stockholm, Sweden). The propionibacteria were biotyped, according to the criteria of Kishishita et al. (17), by using the following base medium (per liter of distilled water): Bacto heart infusion broth (Difco Laboratories, Detroit, MI), 7.5 g; Bacto tryptone (Difco), 7 g; Bacto yeast extract (Difco), 3 g; NaCl, 0.5 g; L-cysteine·HCl monohydrate (Fluka), 0.3 g; Noble agar (Difco), 1 g; and Tween 80, 0.25 ml. The pH of the medium was adjusted to 7.0 before it was autoclaved. meso-Erythritol, D-ribose, and D-sorbitol, purchased from Acr
s Organics (Geel, Belgium), were added as filter-sterilized solutions to the autoclaved base medium. Tubes containing 10 ml of complete medium were inoculated with loopfuls of propionibacterial culture harvested from the growth on plates of Wilkins-Chalgren agar (Oxoid, Inc., Basingstoke, United Kingdom) that had been allowed to grow for 72 h and were incubated at 37°C for 7 days in a GasPak 150 system (Becton, Dickinson & Co., Sparks, MD); under these conditions a CO2 concentration in the range of 4 to 10% was achieved by 60 min, and an O2 concentration of <0.2% was achieved by 10 min. Biotype 2 was defined as meso-erythritol positive, D-ribose positive, and D-sorbitol negative.
The taxonomy of the corynebacteria surveyed was confirmed by cellular fatty acid analysis, as described by Van den Velde et al. (37). Propionibacterial, staphylococcal, streptococcal, and Kocuria strains were typed by using the appropriate API galleries [bioMérieux (Suisse) SA, Geneva, Switzerland]; Lancefield groups were verified by using Slidex Strepto kits (bioMérieux). When the identification values for purported S. aureus strains were <90% with the API Staph galleries, the strains were checked for the presence of the nucA gene (3).
Determination of MICs and MBCs. The MICs for the propionibacteria were obtained by broth microdilution with Wilkins-Chalgren broth (Anaerobe Broth MIC; Difco), according to CLSI (formerly NCCLS) guidelines (23). The microtiter plates were loaded into 7-liter GENbox anaerobic incubation jars fitted with anaerobic atmosphere generators (bioMérieux) and a dry anaerobic indicator strip (BBL, Becton, Dickinson & Co.); under these conditions an O2 concentration of <0.1% was achieved by 2.5 h and a CO2 concentration of >15% was achieved by 24 h. MICs were read after incubation at 35 to 37°C for 48 h by using an illuminated microtiter plate reader fitted with a magnifying mirror (MIC 2000; Cooke Laboratory Products, Alexandria, VA). The breakpoints were those recommended by C. Oprica and C. E. Nord on behalf of the ESCMID Study Group on Antimicrobial Resistance in Anaerobic Bacteria (26, 27).
The minimal bactericidal concentrations (MBCs) for five strains of P. acnes were determined by MIC assays by using standard methodologies (9, 22). Inoculated petri plates of Wilkins-Chalgren agar were incubated in GENbox jars and were examined for growth after 72 h at 35 to 37°C. MBCs were determined only for strain-drug combinations for which MICs were <512 µg/ml.
MICs for nonpropionibacteria were obtained by broth microdilution with cation-adjusted Mueller-Hinton broth (CAMHB; BBL), according to CLSI guidelines (24), with the following modifications: (i) for streptococci CAMHB was supplemented with 5% (vol/vol) horse serum, and (ii) for corynebacteria CAMHB was supplemented with 0.5% (vol/vol) Tween 80 (20). The microtiter plates were incubated at 35°C in ambient air for 20 to 24 h (for Kocuria, staphylococci, and streptococci) or 44 to 48 h (for the corynebacteria) and then inspected visually. The geometric mean MICs for the nonpropionibacteria (8) were calculated by using Microsoft Office Excel 2003 software.
Time-kill studies. Deaerated medium was prepared by boiling 200 ml of swirled Wilkins-Chalgren broth in a stoppered 500-ml round-bottom flask under a steady stream of N2. Aliquots (12 ml) of hot broth were pipetted into Wolin-Miller tubes (21) under a stream of N2, and deaeration of the filled tubes continued for ca. 30 s, after which each tube was sealed with a gas-impermeable butyl rubber septum and crimped with an aluminum ring. The tubes were autoclaved, and filter-sterilized medium (0.8 ml) or filter-sterilized antibiotic (BAL19403, erythromycin, clindamycin, tetracycline, or cefoxitin) dissolved in Wilkins-Chalgren broth (0.8 ml) was added as needed shortly before inoculation.
P. acnes was cultured anaerobically on Wilkins-Chalgren agar (60 h, 37°C), and liquid cultures were prepared by inoculating 300 µl of a no. 3 McFarland unit suspension of cells harvested from plates (
3.5 x 107 CFU) into a bottle containing 55 ml of Wilkins-Chalgren broth. The bottle was incubated at 37°C under anaerobic conditions for
20 h; cells were recovered by centrifugation (4°C, 30 min, 8,000 x g), and the pellet was suspended in 5 ml of sterile Wilkins-Chalgren broth.
Wolin-Miller tubes containing
12.8 ml of deaerated Wilkins-Chalgren broth or
12.8 ml of deaerated Wilkins-Chalgren broth supplemented with antibiotic (2x, 4x, or 8x MIC) were inoculated with 1.2 ml of cell suspension to yield a final P. acnes viable count of
108 CFU/ml, and the tubes were set to shaking (140 rpm, 37°C). Aliquots (
0.5 ml), sampled by use of a syringe at 0, 6, 12, and 24, were diluted with sterile Wilkins-Chalgren broth and spiral plated onto Wilkins-Chalgren agar with 5% (vol/vol) laked horse blood (36). The plates were incubated anaerobically at 37°C for 48 h, after which the viable counts were determined.
For population analyses, multiple plates of reinforced clostridial agar (Oxoid) containing 0.1% (wt/vol) Tween 80 (2) and either no antibiotic (control plate) or BAL19403, erythromycin, clindamycin, or tetracycline at 8x MIC for a particular antibiotic-strain pair were inoculated with ca. 1 x 108 CFU of P. acnes in 100 µl by using a spiral plater. The plates were incubated anaerobically for 96 h at 37°C, and the numbers of colonies were averaged for all like plates. The resistance selection frequency was defined as the number of CFU growing on plates in the presence of antibiotic divided by the number of CFU growing on control plates.
|
|
|---|
In propionibacteria a G2057A transition (Escherichia coli numbering) in domain V of 23S rRNA is associated with low-level resistance to erythromycin and no resistance to clindamycin, an A2058G transition confers high-level resistance to both erythromycin and clindamycin, and an A2059G transition leads to high-level resistance to erythromycin and variable (zero to high-level) resistance to clindamycin (28, 30, 31). The MICs presented in Table 1 generally conformed to these expectations. BAL19403 had low to moderate MICs for P. acnes harboring mutations at positions 2057 and 2058 (MIC range,
0.06 to 4 µg/ml); it had poor activity (MIC, 256 µg/ml) against one P. acnes strain with a mutation at position 2059 but good activity (MIC, 0.5 µg/ml) against another strain with an A2059G mutation.
|
View this table: [in a new window] |
TABLE 1. Distribution of MICs of BAL19403 and comparator drugs for propionibacterial clinical isolates as a function of resistance phenotype
|
512 µg/ml) to erythromycin and clindamycin. Not unexpectedly, high MICs (64 to
512 µg/ml) of BAL19403 were found for three P. acnes strains carrying erm(X) (Table 1).
In propionibacteria the Tetr phenotype usually follows from a G1058C (E. coli numbering) transition in helix 34 of 16S rRNA (29); all seven P. acnes strains surveyed which had tetracycline MICs of
16 µg/ml contained this mutation (Table 1). Susceptibility or resistance to tetracycline was independent of susceptibility or resistance to BAL19403, erythromycin, or clindamycin.
The MBCs and MBC/MIC ratios (1, 34) for BAL19403 and the comparator drugs for one fully susceptible strain of P. acnes and four P. acnes strains with different antibiotic resistance mechanisms are presented in Table 2. BAL19403 had an MBC of
0.06 µg/ml for P. acnes EG7NS, P. acnes SW3CD, and P. acnes P95, whereas the MBC of BAL19403 for Tetr strain P. acnes SW101T and the dimethylase producer P. acnes GE4E was >32 µg/ml; these results compare favorably to the MBCs of erythromycin, clindamycin, and tetracycline found for these P. acnes strains.
|
View this table: [in a new window] |
TABLE 2. MBCs and MBC/MIC ratios of BAL19403 and comparator drugs for some P. acnes strains
|
3) by 24 h and "bacteriostatic" as <99.9% killing by 24 h, BAL19403 at 4x to 8x MIC was bacteriostatic for five of six of the P. acnes strains surveyed, whereas erythromycin, clindamycin, and tetracycline were bacteriostatic for all test strains. Tetracycline "froze" the viability of cultures, precluding both growth and deterioration during 24 h (cf. the findings presented in reference 5). Cefoxitin was bactericidal for two of six strains at 4x MIC and three of six strains at 8x MIC. The killing rates for cefoxitin at 4x to 8x MIC tended to be higher than those at 2x MIC, whereas for BAL19403, erythromycin, and clindamycin the killing rates between 2x and 8x MIC varied only within a narrow range. |
View this table: [in a new window] |
TABLE 3. Summary of time-kill test results for BAL19403 and comparator drugs for P. acnes
|
|
View this table: [in a new window] |
TABLE 4. Population analyses for resistance selection frequencies of BAL19403 and comparator drugs (8x MIC) for selected P. acnes strains
|
4 µg/ml), medium (MICs, 8 to 16 µg/ml), or low (MICs,
32 µg/ml), the MIC distribution for BAL19403 was bimodal (high > low) for the corynebacteria and staphylococci and unimodal (high) for streptococci. Less clear-cut MIC distributions were observed for the comparator drugs, particularly tetracycline, to which the strains tended to show low levels of susceptibility more often than they did to the other antibiotics tested. The MIC distributions for BAL19403 and the comparator drugs for the corynebacteria, staphylococci, and streptococci were reflected in the geometric mean MICs presented in Table 5. |
View this table: [in a new window] |
TABLE 5. Geometric mean MICs of BAL19403 and comparators for selected nonpropionibacteria
|
0.06 to 0.25 µg/ml; erythromycin,
0.06 to 0.125 µg/ml; clindamycin,
0.06 to 0.125 µg/ml; and tetracycline, 1 µg/ml. |
|
|---|
The relationship between the prognosis and the presence of antibiotic-resistant propionibacteria, particularly erythromycin- and/or clindamycin-resistant strains, mandates the discovery and development of new narrow-spectrum antibiotics that have potent activities against these resistant propionibacteria and that are suitable for topical application. The novel macrolide BAL19403 was identified as being highly active against cutaneous propionibacteria resistant to erythromycin and/or clindamycin due to base transitions in the region from positions 2057 to 2059 of domain V of 23S rRNA. The time-kill profiles and MBC/MIC ratios confirmed the bacteriostatic action of BAL19403, as well as of those of erythromycin, clindamycin, and tetracycline (11), against most strains of P. acnes examined, although there was a trend toward lower viable counts at 24 h in the presence of BAL19403 than in the presence of erythromycin, clindamycin, or tetracycline (Basilea Pharmaceutica AG, data on file). Unlike the macrolides, lincosamides, or tetracyclines, ß-lactams are usually bactericidal; and at 8x MIC, cefoxitin, a cephamycin with good antianaerobe activity, was bactericidal for three of six of the P. acnes strains tested. The observed resistance selection frequencies for BAL19403, erythromycin, and clindamycin, which were in the range of 108 to 109, matched those reported for erythromycin, clindamycin, and the 18-membered macrolide antibiotic tiacumicin B at 4x to 8x MIC (6, 39).
Normal human skin is colonized by a relatively stable microflora; the nonpropionibacteria encountered most frequently are corynebacteria and coagulase-negative staphylococci, with Micrococcus and Kocuria spp. occurring less often (4, 25, 35). Since application of BAL19403 for the treatment of acne lesions would lead to drug contact with cutaneous nonpropionibacteria, the in vitro activities of this macrolide and the comparator drugs against panels that comprised strains indigenous to facial skin, as well as against microflora often implicated in skin and skin structure infections (group A, C, and G streptococci; Staphylococcus aureus) were also examined. Overall, the antimicrobial activity of BAL19403 against nonpropionibacteria resembled that of erythromycin more than that of either clindamycin or tetracycline (Table 5).
It has been opined that physicians give little weight to resistance when prescribing antiacne products for their patients (10), so it is important that an antiacne antibiotic act not only against susceptible strains but also against strains with the resistance phenotypes most likely to be encountered in a clinical setting. On the basis of its activity against antibiotic-resistant propionibacteria and its potent anti-inflammatory activity (20a), BAL19403 represents a promising topical treatment for mild to moderate inflammatory acne vulgaris.
Published ahead of print on 26 March 2007. ![]()
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»