Search

Enterococci are intrinsically resistant to low levels of aminoglycosides. We previously selected in vitro and in vivo Enterococcus faecalis with intermediate-level resistance to gentamicin that did not abolish synergism with a cell-wall-active agent (E. Aslangul et al., Antimicrob. Agents Chemother. 49:4144-4148, 2005). The aim of this study was to investigate the mechanism of resistance to gentamicin in the 1688-G3 third-step mutant (MIC, 512 μg/ml) of E. faecalis JH2-2. No mutations were found in the genes for L6 ribosomal protein and the four copies of 16S rRNA. Production of a known aminoglycoside-modifying enzyme was unlikely due to the distinct resistance phenotype and absence of the corresponding genes. Efflux was also unlikely since ethidium bromide MICs were similar for JH2-2 and 1688-G3 and since the pump inhibitors reserpine and verapamil had no effect on gentamicin resistance in both strains. To study gentamicin accumulation, we developed a nonisotopic method based on a fluorescent polarization immunoassay. Impaired gentamicin accumulation was observed in 1688-G3 compared to JH2-2 and was only partially reversible by the N , N ′-dicyclohexylcarbodiimide (DCCD) uncoupler agent. The lower sensitivity of 1688-G3 to DCCD suggested alteration of the F o F 1 -ATPase. However, no mutations were detected in the structural genes ( atp ) for the F o channel and no difference in transcript levels of atpB and atpE was found between 1688-G3 and JH2-2. Our data are compatible with acquisition of intermediate-level gentamicin resistance by uptake impairment in E. faecalis .

The apparent diffusion coefficient of a bound drug, cefoperazone, was studied. The protein binding of cefoperazone was studied by voltammetry, a technique which permitted instant measurements. The apparent diffusion coefficients were similar in agar and fibrin and lower in rat brain tissue. The influence of protein on the value of the apparent diffusion coefficient was negligible. The hypothesis that only the free drug diffuses was supported. The percentage of binding determined by voltammetry corresponded to the true concentration of drug which diffuses and is much lower than the percentage of binding determined by the ultrafiltration centrifugation method. This discrepancy could be explained by the rate of dissociation of the protein-drug complex.

Nonpregnant and pregnant guinea pigs in the last third of gestation were injected intramuscularly with 4 mg of gentamicin per kg, and drug concentrations in plasma were determined by radioimmunoassay at several intervals after injection. The maximum gentamicin concentration was lower in pregnant than in nonpregnant animals (14.6 +/- 0.7 micrograms/ml versus 21.6 +/- 0.7 micrograms/ml), and the peak time occurred significantly later (0.57 +/- 0.12 h versus 0.13 +/- 0.02 h). Four hours after gentamicin injection, drug concentrations in plasma were 2.1 +/- 0.8 and 0.3 +/- 0.1 micrograms/ml in pregnant and nonpregnant animals, respectively. Pregnant animals therefore eliminated the drug from their plasma more slowly. These data provide good evidence that the kinetics of plasma gentamicin varies in pregnant females because its volume of distribution was larger in pregnant than in nonpregnant animals. Detectable but small amounts of gentamicin (less than or equal to 0.50 microgram/ml) were found in the plasma of 46 of 57 fetuses. However, no net variations in these concentrations were observed during the period between 15 min and 6 h after injection to the mother. Gentamicin concentrations were also determined in the kidneys, liver, lungs, heart, and brain of fetal guinea pigs after administration to their mothers of one daily injection of 4 mg/kg for 7 days. Gentamicin was present in all these fetal organs; however, as in the adult organs, the kidneys contained far more than any of the others. Gentamicin concentrations were not significantly different in the kidney cortex and medulla (1.79 +/- 0.16 versus 1.48 +/- 0.92 micrograms/g), indicating that, contrary to what is observed for adults, renal accumulation of gentamicin in the fetus does not occur preferentially in the cortex.

Clearance experiments were performed with young pups born of guinea pigs given a daily injection of 4 mg of gentamicin per kg (body weight) from days 48 to 54 of gestation (term, 68 days). For 3-day-old animals, the glomerular filtration rate was similar to that measured in control guinea pigs of the same age whose mothers were given saline during the same period of gestation. The same applied to fractional excretion of water, urea, total solutes, Na, K, Ca, and Mg but not to fractional phosphate excretion, which increased significantly in the gentamicin group when compared with the controls (mean +/- standard error, 21.7 +/- 4.9 versus 7.3 +/- 1.8%; P less than 0.05; n = 6 for both). The glomerular volume of the juxtamedullary nephrons diminished by about 40%, and their proximal tubule length decreased by about 20%. The glomerular volume of the superficial nephrons also diminished, by about 30%, but their proximal tubule length did not change. The gentamicin concentration was higher in the renal cortex than in the medulla (13.1 +/- 2.6 versus 5.7 +/- 2.2 micrograms/g [dry wt]; P less than 0.01; n = 6 for each). It decreased significantly from days 3 to 20 in both tissues. No functional impairment of the kidney was found in 10-day-old animals, and normal or even supranormal morphometry of the nephrons was observed in the 20-day-old animals. It is concluded that fetal exposure to gentamicin impairs proximal tubular function in the developing animal and might also adversely affect glomerular and tubular growth. However, both the functional and morphometric impairments of nephrons are transitory.

The kinetics of gentamicin in the inner ear fluids of rats were studied up to 15 days after cessation of a 2-day constant infusion of 10 micrograms/min. In endolymph, the concentration of gentamicin persisted at about 1 microgram/ml for up to 15 days, precluding the determination of the half-life of the drug in this fluid. In perilymph, gentamicin cleared more slowly than after a shorter period of infusion. These results suggest that the tissues of the inner ear could bind the aminoglycoside and then slowly release it into the surrounding fluids.

Four 2-g doses of ceftazidime were infused intravenously over 30 min at 8-h intervals, first between days 2 and 4 and again between days 11 and 20, in 11 patients with bacterial meningitis undergoing treatment with other antibiotics. Concentrations of ceftazidime in serum and cerebrospinal fluid samples obtained 120 or 180 min after dose 4 were measured by high-pressure liquid chromatography. Concentrations in cerebrospinal fluid ranged from 2 to 30 micrograms/ml, depending on the sampling time and the time elapsed since the onset of the disease.

The penetration of aztreonam into the cerebrospinal fluid was determined in 16 patients with bacterial meningitis undergoing treatment with other antibiotics. Three aztreonam doses of 30 mg/kg were infused intravenously over 30 to 45 min at 8-h intervals, first between days 2 and 4 and again between days 11 and 20 after onset of the disease. Concentrations of aztreonam in serum and cerebrospinal fluid samples obtained at 60, 90, 120, and 240 min after the third aztreonam dose were measured by high-pressure liquid chromatography. The concentrations of aztreonam in cerebrospinal fluid ranged from 3.5 to 62 micrograms/ml, depending on the sampling time and the time elapsed since the onset of the disease. These concentrations were equal to or higher than the MICs for most of the gram-negative bacilli (including Pseudomonas aeruginosa).

A method for determining drug concentration relationships between plasma and cerebrospinal fluid (CSF) in rats is described. Continuous CSF samples were collected directly from the third anterior ventricle with an indwelling cannula inserted through the bregma point, and drug concentrations were determined by high-pressure liquid chromatography and radioimmunoassay micromethods. Three antibiotics with different abilities to cross the blood-CSF barrier (chloramphenicol, piperacillin, and gentamicin) were tested. This method was found to be reproducible for each drug even if the antibiotic levels were low and the sample volumes very small. Peak CSF concentrations occurred between 0.75 and 1.25 h after injection for all three antibiotics. Percent penetration values at 1 h were 50, 1.2, and 5.4% for chloramphenicol, piperacillin, and gentamicin, respectively.

Penetration of moxalactam into the cerebrospinal fluid was studied in 11 patients with bacterial meningitis undergoing treatment with other antibiotics. Moxalactam at a dose of 20 mg/kg was administered as three 30- to 45- min infusions at 8-h intervals, once between days 2 and 4 and a second time between days 11 and 20 of treatment with the other antibiotics. Serum and cerebrospinal fluid were sampled 60, 90, or 120 min after the third moxalactam dose for measurement of the concentration of this drug by high-performance liquid chromatography. The concentration of moxalactam in cerebrospinal fluid ranged from 1.5 to 11 micrograms/ml, depending on the sampling time and the time elapsed since the onset of the disease. These concentrations in cerebrospinal fluid were equal to or higher than the minimum inhibitor concentrations for Neisseria meningitidis, Streptococcus pneumoniae, Haemophilus influenzae (including ampicillin-resistant strains), and most of the gram-negative bacilli except for Pseudomonas aeruginosa. These results show that moxalactam has good penetrability when the meninges are inflamed and that it might be considered in cases of bacterial meningitis when the susceptibility of the pathogen indicates its usefulness.

A method for determining diffusion coefficients of four antibiotics in extracellular tissue space according to Fick's law is described. This new method was applied to rat brain tissue and to agar. After diffusion of the antibiotic in one axis, the gradient concentration was determined with microvoltammetric electrodes. These microelectrodes (1 micron at the extreme tip) measured the free diffusible form of electroactive antibiotics in the extracellular brain space. Metronidazole, chloramphenicol succinate, cefsulodin, and piperacillin gave diffusion coefficients ranging from 0.1 x 10(-6) to 0.2 x 10(-6) cm2 . s-1 in tissue; chloramphenicol base, which is positively charged, gave a coefficient of 0.04 x 10(-6) cm2 . s-1. The coefficient ranged from 0.6 x 10(-6) to 1.2 x 10(-6) cm2 . s-1 in agar. These coefficients were used to simulate antibiotic concentrations in infectious sites and between capillaries by using a simple model of plane diffusion.