Combination therapy with ampicillin, vancomycin, and gentamicin in vitro against several clinical isolates of vancomycin-resistant, highly ampicillin-resistant Enterococcusfaecium, including VanA and VanB strains, was evaluated. The MICs of ampicillin were not significantly decreased by induction with vancomycin, and the combination of ampicillin and vancomycin was not inhibitory for any strain. Triple-combination therapy was least active against highly resistant VanA isolates, achieving a reduction of less than 1 log CFU at 24 h, but demonstrated slightly more activity against VanB strains.Since 1986, clinical isolates of enterococci with inducible resistance to the glycopeptide antibiotic vancomycin have emerged as a major clinical problem in Europe and the United States (6, 7, 13). During 1991, several hospitals in the Philadelphia, Pa., area experienced outbreaks of colonization and infection due to vancomycin-resistant Enterococcus faecium (11). These isolates demonstrate high-level inducible resistance to vancomycin and other glycopeptides (VanA phenotype), as well as high-level resistance to ampicillin (MIC, 256 ,ug/ml). Currently available alternatives for therapy of infections caused by these strains are limited, particularly for bloodstream and deep-seated tissue infections requiring bactericidal therapy.Several investigators have reported that antibiotic combinations employing concentrations of vancomycin and penicillin achievable in serum are inhibitory in vitro for enterococcal strains exhibiting both glycopeptide and beta-lactam resistance (3,10,12) and that the combination of vancomycin, penicillin, and an aminoglycoside could be bactericidal in the absence of high-level aminoglycoside resistance (10, 12). These studies have recently been verified in a rabbit model of enterococcal endocarditis (2). However MICs were determined by broth macrodilution in cationadjusted Mueller-Hinton broth (BBL) with an initial inoculum of 106 organisms. MICs were determined during vancomycin induction by preincubation of the test strain overnight in the constant presence of 8 ,ug of vancomycin per ml or by addition of vancomycin simultaneously with the other antibiotics. There were no differences in the results with either method of vancomycin induction. MBCs, defined as killing of 99% of the initial inoculum, were determined from an initial inoculum of 106 organisms with subculture volumes of 0.010 ml. Vancomycin, ampicillin, and gentamicin powder were purchased from Sigma. Killing curves were determined for overnight cultures which had been diluted 1/100 and allowed to regrow to the logarithmic phase of growth. Inocula were adjusted to approximately 5 x 105 and 5 x 107 organisms per ml for low-and high-inoculum experiments, respectively. Killing curves were determined for cultures grown in Mueller-Hinton broth. When tested in brain heart infusion (BBL), antibiotic combinations demonstrated activity equal to or less than that in Mueller-Hinton broth for all strains (data not shown). For antibiotic combination kill...

Sixty strains of vancomycin-and ampicillin-resistant Enterococcus faecium were evaluated for their susceptibilities to novobiocin in vitro. In Mueller-Hinton broth, novobiocin inhibited all strains when it was used at a concentration of 52 ,ug/ml and 40 of 60 strains when it was used at a concentration of <0.5 ,ug/ml. MICs were 8-to 16-fold higher in 50%o serum. Novobiocin alone resulted in 2-log-unit killing at 24 h. Combinations of novobiocin and a fluoroquinolone (either ciprofloxacin or ofloxacin) were additive and bactericidal for quinolone-susceptible strains in either broth or 50%o serum. Gentamicin did not affect novobiocin activity, and rifampin and doxycycline were antagonistic.

A simple defined basal medium is presented for the study of proteolytic activity, induction and repression, and protease purification with Serratia marcescens. Since the medium contains no protein, it does not interfere with or present artifact to protein assays, column chromatography, or electrophoresis. The medium consists of the basal salts and buffer medium of Bromke and Hammel (1979) plus a carbon-energy source such as glycerol, calcium chloride for the cation requirement for protease activity, and an amino acid, preferably leucine. Growth parameters and proteolytic activities are presented for unsupplemented medium and for the medium supplemented with each of 18 amino acids. Unsupplemented medium completes the logarithmic phase in 12.5 h of incubation and has a constitutive level of proteolytic activity. Supplementation with any amino acid, except cysteine and tryptophan, increases significantly the proteolytic activity, but has a varied effect on growth parameters.

A simple defined basal medium is presented for the study of proteolytic activity, induction and repression, and protease purification with Serratia marcescens. Since the medium contains no protein, it does not interfere with or present artifact to protein assays, column chromatography, or electrophoresis. The medium consists of the basal salts and buffer medium of Bromke and Hammel (1979) plus a carbon-energy source such as glycerol, calcium chloride for the cation requirement for protease activity, and an amino acid, preferably leucine. Growth parameters and proteolytic activities are presented for unsupplemented medium and for the medium supplemented with each of 18 amino acids. Unsupplemented medium completes the logarithmic phase in 12.5 h of incubation and has a constitutive level of proteolytic activity. Supplementation with any amino acid, except cysteine and tryptophan, increases significantly the proteolytic activity, but has a varied effect on growth parameters.