Modification of penicillin-binding proteins of penicillin-resistant mutants of different species of enterococci

Mutants resistant to penicillin G were selected in a stepwise manner from nine different species of enterococci. Mutants with the highest level of resistance showed cross-resistance to all beta-lactams tested. For eight of the nine species, resistance correlated with increased production of a low molecular weight penicillin-binding protein (PBP). Two of these species produced a new PBP of low molecular weight, while two other species produced an additional PBP of high molecular weight. With the exception of Enterococcus faecium, no difference was observed in terms of lysis or bactericidal effect when the sensitive strains and their resistant mutants were tested at ten times their respective MICs of penicillin G. With E. faecium an increased lytic and bactericidal effect was observed for the resistant mutant.     

Action of penicillin on Borrelia hermsii.

Borrelia hermsii, a spirochete and an etiological agent of relapsing fever, was cultivated in modified Kelly medium. Studies of the action of penicillin on B. hermsii strain HS1 revealed the following: (i) the in vitro minimum inhibitory concentration and minimum bactericidal concentration of benzylpenicillin for this strain were 0.4 and 3.1 nmol/ml (0.15 and 1.1 micrograms/ml), respectively; (ii) the primary morphological responses at the minimum bactericidal concentration of benzylpenicillin were the formation of spheroplast-like structures and an increased number of small, membranous blebs; (iii) radioactive benzylpenicillin bound to five penicillin-binding proteins in the whole cells of B. hermsii. The 50% binding concentrations of labeled penicillin for the five penicillin-binding proteins were within a factor of five of the minimum inhibitory concentration. More than one-half of the total bound labeled penicillin was associated with penicillin-binding protein 1, the penicillin-binding protein with the largest apparent molecular weight (90,000).     

pH-dependent penicillin tolerance of group B streptococci.

Group B streptococci lose viability without apparent lysis during treatment with beta-lactam antibiotics and vancomycin. Rapid loss of viability was observed in early-exponential-phase cultures. Cultures in the mid-exponential growth phase exhibited various degrees of resistance to the bactericidal effect of the antibiotics, whereas their susceptibilities to the growth-inhibitory effect remained unchanged. This growth-phase-dependent tolerance was caused by the gradual increase in acidity of the cultures as the cell concentration increased. Retitration of the pH to neutrality made the formerly tolerant bacteria again fully susceptible to the killing effect of penicillin. Conversely, lowering the pH value of the medium resulted in antibiotic tolerance throughout culture growth. The penicillin-binding proteins of whole bacteria and their labeling pattern were found to be independent of culture pH. It is suggested that the mechanism of Ph-dependent tolerance is indirect and may be mediated by an autolysin. The tolerance of group B streptococci for penicillin could be clinically relevant in view of the relatively low pH values known to prevail in the natural host environments colonized by these bacteria.     

Characterization of clinical isolates of beta-lactamase-negative, highly ampicillin-resistant Enterococcus faecalis.

We analyzed the penicillin-binding protein (PBP) profiles of two clinical isolates of Enterococcus faecalis for which ampicillin MICs were 32 and 64 micrograms/ml. Six PBPs were detected in both isolates, demonstrating an apparently increased amount of PBP 5 and decreased penicillin binding of PBPs 1 and 6. These results suggest that ampicillin resistance in the clinical isolates of E. faecalis described could be associated with alterations in different PBPs.     

Bactericidal activity of the fluoroquinolone WIN 57273 against high-level gentamicin-resistant Enterococcus faecalis.

The fluoroquinolone WIN 57273 showed identical bactericidal activities (MBC for 90% of the strains = 0.25 micrograms/ml) for bacteremic strains of Enterococcus faecalis with and without high-level gentamicin resistance. WIN 57273 was bactericidal in time-kill measurements with highly gentamicin-resistant, ciprofloxacin-susceptible strains of E. faecalis. However, WIN 57273 was indifferent with penicillin for gentamicin-resistant E. faecalis and was not bactericidal for ciprofloxacin-resistant E. faecalis.     

Antagonistic effect of penicillin-amikacin combinations against enterococci.

Amikacin has been shown to antagonize the bactericidal effect of penicillin against strains of Streptococcus faecalis which produce aminoglycoside 3'-phosphotransferase. The mechanism by which this phenomenon occurs was studied with an enzyme-producing strain (8436) and an enzyme-negative strain (8436c) derived by curing the former with novobiocin. Combinations of amikacin with beta-lactam antibiotics were antagonistic against strain 8436 but synergistic against strain 8436c. Against strain 8436 penicillin-amikacin combinations resulted in levels of killing comparable to those seen with high concentrations of penicillin (500 micrograms/ml), which were less bactericidal than lower concentrations of penicillin. No antagonism was observed between amikacin and non-beta-lactam cell wall-active drugs or between penicillin and kanamycin or neomycin, both of which are substrates for the enzyme. At concentrations near the MIC, amikacin was bactericidal against strain 8436c but bacteriostatic against strain 8436 (MIC, 250 micrograms/ml; MBC, 2,000 micrograms/ml). Neither penicillin nor phosphorylated amikacin affected the inhibition of ribosomal protein synthesis by amikacin in a cell-free system. Although antagonism of killing by amikacin in enzyme-positive strains was specific for combinations which included beta-lactam antibiotics, amikacin did not influence the binding of [3H]penicillin to penicillin-binding proteins in isolated bacterial cell membranes or in intact cells and did not detectably affect the autolytic system of cells exposed to penicillin. Antagonism of beta-lactam activity by a bacteriostatic effect of amikacin against the enzyme-producing strain is the most likely explanation for this phenomenon.     

Paradoxical response of Enterococcus faecalis to the bactericidal activity of penicillin is associated with reduced activity of one autolysin.

Ten clinical isolates of Enterococcus faecalis were examined for susceptibility to the bactericidal activity of penicillin. Four of these had MBCs of penicillin equal to 2 to 4 x the MIC, and six exhibited a paradoxical response to penicillin, i.e., the bactericidal activity of the antibiotic had a concentration optimum at 2 to 4 x the MIC and decreased significantly at concentrations above this. We found that the paradoxical response to penicillin was an intrinsic and stable property of a strain, but that its phenotypic expression was not homogeneous; only a fraction of the cell population that died at low concentrations was able to survive at high penicillin concentrations. The size of this fraction increased with increasing antibiotic concentration and reached a maximum in the late-log phase of growth. All 10 strains produced a lytic enzyme that was active on Micrococcus luteus heat-killed cells, whereas only some strains lysed E. faecalis heat-killed cells. Strains producing large amounts of the latter enzyme did not show the paradoxical response to penicillin, whereas mutants of these strains that lacked this enzymatic activity paradoxically responded to the antibiotic activity. In addition, from strains that showed paradoxical response to penicillin and produced only the enzyme that was active on M. luteus, it was possible to isolate mutants that were also capable of lysing E. faecalis cells and that were killed with similar efficiency by all concentrations above the MBC. On the basis of these findings, the paradoxical response to penicillin is explained as a property of certain strains of E. faecalis; this property is genetically characterized by alterations in synthesis or activity of one autolysin but phenotypically expressed only by a few cells that are in a particular physiological condition when exposed to high concentrations of antibiotics.     

Bactericidal activity of netilmicin compared with gentamicin and streptomycin, alone and in combination with penicillin, against penicillin tolerant viridans streptococci and enterococci

Netilmicin was compared with gentamicin and streptomycin for in-vitro activity against 30 strains of penicillin-tolerant streptococci including 16 strains of enterococci. Both netilmicin and gentamicin tested alone at 4 mg/l caused 99.9% kill of more than half of the 13 strains of viridans streptococci tested, whereas streptomycin, 4 mg/l, had no bactericidal effect against these strains. Netilmicin, gentamicin and streptomycin tested alone at 8.0 mg/l against 10 strains of Streptococcus faecalis resulted in 99.9% kill of six, one and zero strains respectively. Combinations of penicillin with 2 mg/l of either netilmicin or gentamicin resulted in bactericidal synergy against 12 of 13 strains of viridans streptococci and all 10 strains of S. faecalis after 18 to 24 h incubation. Parallel experiments showed that higher concentration of penicillin were required to obtain 99.9% kill of 10 streptococcal strains when 4 mg/l streptomycin was compared with 2 mg/l of the other aminoglycosides. Killing curves showed similar bactericidal synergy for netilmicin-penicillin and gentamicin-penicillin combinations against most streptococci tested after 24 h incubation but there was sometimes a greater bactericidal effect noted with netilmicin after only 6 h incubation of the broth or after 48 h incubation. The results of this in-vitro study suggest that netilmicin is at least as effective as gentamicin as a bactericidal synergic agent with penicillin against penicillin-tolerant viridans streptococci and S. faecalis strains isolated from patients with endocarditis. Neither gentamicin or netilmicin were effective as bactericidal synergic agents with penicillin against 4 of 6 strains of S. faecium tested.     

Proinflammatory activation of Toll-like receptor-2 during exposure of penicillin-resistant Streptococcus pneumoniae to beta-lactam antibiotics

OBJECTIVES: Disease caused by penicillin-resistant Streptococcus pneumoniae (PRSP) is associated with more suppurative complications than disease caused by penicillin-susceptible S. pneumoniae (PSSP). Exposure of S. pneumoniae to beta-lactam antibiotics enhances the proinflammatory activation of human cells by pneumococci via Toll-like receptor-2 (TLR2). To test the hypothesis that penicillin resistance influences cellular TLR2 activation by beta-lactam-exposed pneumococci, we compared TLR2 induction by PSSP (MIC 0.06 mg/L) and a high-level PRSP clinical isolate (159; MIC 16 mg/L) following exposure to penicillin and cefotaxime. METHODS: Both organisms were treated with penicillin or cefotaxime at and around the MIC. TLR2 signalling was measured as relative IL-8 promoter activation in transfected HeLa cells. RESULTS: On exposure to penicillin, log-phase PSSP and PRSP induced TLR2-proinflammatory activation at levels significantly higher than unexposed bacteria, and maximal in each case at the MIC. Transformants containing low-affinity penicillin-binding proteins (PBP) 2x, 1a and 2b exhibited stepwise resistance to cefotaxime and penicillin. TLR2 activation following penicillin treatment was dependent on an abnormal cell wall (PBP1a and 2x) and autolysis (PBP2b). High affinity PBP2x was required for this effect to be observed in log-phase pneumococci exposed to cefotaxime at the MIC. Cefotaxime-mediated TLR2 activation was not observed in lag-phase transformants exposed to sub-lethal concentrations. CONCLUSIONS: These data show that PRSP have similar TLR2-proinflammatory effects to PSSP when exposed to beta-lactam antibiotics but the antibiotic concentration relative to the MIC is critical. This has implications for treatment of pneumococcal disease when tissue concentrations of antibiotic are close to the MIC.     

In vitro response to bactericidal activity of cell wall-active antibiotics does not support the general opinion that enterococci are naturally tolerant to these antibiotics.

The incidence of tolerance and paradoxical response to bactericidal activity of penicillin was investigated in 50 clinical isolates of Enterococcus faecalis. Of the isolates tested, 86% exhibited the paradoxical phenomenon whereby there were more survivors at high than at low concentrations above the MIC. Low penicillin concentrations caused decreases equal to or higher than 99.9% in 11 strains, from 99.9 to 99.5% in 23 strains, and lower than 99.5% in 9 strains. Of the total strains, 14% were killed to the same extent by all concentrations above the MIC. The bactericidal activities of other beta-lactams (ampicillin and piperacillin) and other cell wall inhibitors (vancomycin and daptomycin) were also tested against some of these strains. In general, beta-lactams exhibited the best bactericidal activity at 2 x MIC. Piperacillin was the most active, as at 2 x MIC it reduced the original inoculum by 99.9% or more in most of the strains. No concentration of vancomycin above the MIC caused 99.9% killing of the strains, whereas daptomycin was bactericidal at 8 x MIC in most cases. Paradoxical response to bactericidal activity of beta-lactams was abolished by incubation of the inoculum with 2 x MIC before exposure to higher antibiotic concentrations. These findings suggest that enterococci are not always tolerant to cell wall-active antibiotics and that accurate in vitro bactericidal tests may be useful for the choice of appropriate therapy for infections caused by these microorganisms.     

Postantibiotic effect of penicillin plus gentamicin versus Enterococcus faecalis in vitro and in vivo.

A persistent suppression of bacterial growth following a brief exposure to an antibiotic (postantibiotic effect [( PAE]) has been described for a variety of antibiotics and microorganisms. If a PAE is present in vivo, antibiotic levels in tissue at the site of infection may decrease below the MIC without bacterial regrowth in the latter portion of the dosing interval. In the present studies, a PAE was sought in vitro and in vivo for penicillin G plus gentamicin versus Enterococcus faecalis. The studies demonstrated that increasing concentrations of gentamicin caused an increased rate of bactericidal action and an increasingly prolonged PAE in vitro. The combination of penicillin and gentamicin, in addition to more rapid killing, exhibited a more prolonged PAE than did penicillin alone. However, unlike these in vitro findings, no PAE could be demonstrated in vivo in rats with experimental left-sided enterococcal endocarditis treated with penicillin plus gentamicin. This suggests that antibiotic vegetation levels should be maintained above the MIC throughout the dosing interval to prevent loss of efficacy as a result of bacterial regrowth.     

Importance of Medium in Demonstrating Penicillin Tolerance by Group B Streptococci

A total of 30 clinical isolates of group B streptococci were studied for penicillin tolerance in vitro. Minimal inhibitory and bactericidal concentrations of penicillin were determined simultaneously in three test media which have been used for group B streptococci, tryptose phosphate, Mueller-Hinton, and Todd-Hewitt broths, using a logarithmic-phase inoculum of 10(5) colony-forming units per ml. Minimal inhibitory concentrations in the three media did not differ significantly. However, minimal bactericidal concentrations were significantly higher in tryptose phosphate broth (mean, 1.04 mug/ml) than in Mueller-Hinton broth (0.22 mug/ml) or Todd-Hewitt broth (0.15 mug/ml). Similarly, ratios of minimal bactericidal to minimal inhibitory concentrations were significantly greater in tryptose phosphate broth than in Mueller-Hinton or Todd-Hewitt broth. After incubation in tryptose phosphate broth for an additional 24 h, the minimal bactericidal concentration consistently fell to levels which were only twice or equal to the minimal inhibitory concentration. This study illustrates the importance of the medium in the demonstration of penicillin tolerance and of controlling laboratory variables in the susceptibility testing of group B streptococci with penicillin.     

Multiple changes of penicillin-binding proteins in penicillin-resistant clinical isolates of Streptococcus pneumoniae.

Penicillin-binding properties and characteristics of penicillin-binding proteins (PBPs) were investigated in several clinical isolates of Streptococcus pneumoniae differing in their susceptibilities to penicillin (minimal inhibitory concentration [MIC], 0.03 to 0.5 microgram/ml) and compared with the penicillin-susceptible strain R36A (MIC, 0.07 microgram/ml). Several changes accompanied the development of resistance: the relative affinity to penicillin of whole cells, isolated membranes, and two major PBPs after in vivo or in vitro labeling decreased (with increasing resistance). Furthermore, one additional PBP (2') appeared in four of five relatively resistant strains with an MIC of 0.25 microgram/ml and higher. PBP 3 maintained the same high affinity toward penicillin in all strains under all labeling conditions.     

Antimicrobial Therapy of Experimental Enterococcal Endocarditis

The successful therapy of enterococcal endocarditis requires prolonged administration of synergistic antibiotic combinations. Controversy has arisen regarding optimal therapy (i) when the organism possesses high-level streptomycin resistance, and (ii) when the patient is allergic to penicillin. This study examines these questions in vitro and in a rabbit model of enterococcal endocarditis. The combination of penicillin with either streptomycin or gentamicin increased the rate of bacterial killing in vitro and in vivo when compared with penicillin alone (P < 0.05) when the test strain was relatively susceptible to streptomycin (minimal inhibitory concentration, 128 mug/ml). Only the combination of penicillin and gentamicin was consistently more effective than penicillin alone (P < 0.01) when the test strain was highly resistant to streptomycin (minimal inhibitory concentration > 150,000 mug/ml). The combination of vancomycin and streptomycin was more rapidly bactericidal than vancomycin alone in vitro and in the animal model against the streptomycin-susceptible strain (P < 0.01). The relative rate of in vitro bacterial killing by various antibiotics and combinations was predictive of the efficacy of these drugs in eradicating enterococci from cardiac vegetation in experimental endocarditis.     

Importance of the aminoglycoside dosing regimen in the penicillin-netilmicin combination for treatment of Enterococcus faecalis-induced experimental endocarditis.

The penicillin-aminoglycoside combination is recommended for the treatment of systemic enterococcal infections. However, the optimal dosing regimen of the aminoglycoside remains to be elucidated. We evaluated the efficacy of penicillin, alone or in combination with various dosing regimens of netilmicin, for the treatment of experimental left-sided Enterococcus faecalis endocarditis in rabbits. Animals were injected intramuscularly for 4 days with penicillin alone or in combination with netilmicin in one of the following regimens: netilmicin at a low dose (2 mg/kg of body weight every 8 h), netilmicin at a high dose (4 mg/kg every 8 h), or netilmicin at a single daily high dose (12 mg/kg every 24 h). MICs and MBCs were 3.1 and 6.2 micrograms/ml and 8 and 8 micrograms/ml for penicillin and netilmicin, respectively. A netilmicin concentration of 4 micrograms/ml was the lowest concentration that achieved synergism with penicillin, as shown by the kill-curve method. Mean peak levels of netilmicin in serum were 5.6 (netilmicin at 2 mg/kg), 9.8 (netilmicin at 4 mg/kg), and 20.6 (netilmicin at 12 mg/kg) micrograms/ml. Mean penicillin levels in serum were constantly above the MIC. Penicillin plus netilmicin at a high dose given three times daily was more effective (P less than 0.05) than any other regimen in reducing bacterial titers in vegetations and was the only treatment that induced a significant bactericidal activity in rabbit serum during the trough. We concluded that divided doses of aminoglycoside are more effective than the same total dose given once daily in combination with penicillin. Our data suggest that prolonged levels of aminoglycoside in serum might be important to exhibit the greatest in vivo efficacy of the combination against E. faecalis. They also indicate that use of a reduced total daily dose of aminoglycoside or an increase in the interval between each dose might reduce the efficacy of therapy in animals with this type of infection.     

Penicillin tolerance of human isolates of group C streptococci.

Seventeen clinical isolates of group C streptococci were tested for penicillin tolerance. Sixteen of the strains showed penicillin tolerance with a 32-fold or greater difference between the minimal inhibitory concentration and the minimal bactericidal concentration. Synergism was demonstrated with a combination of penicillin and gentamicin for all 17 strains tested. The rate of antibiotic killing was measured for five of the streptococcal strains by using the combination of penicillin and gentamicin. All isolates were killed within 5 h with the combination, but viable organisms were recovered after 48 h when either drug was used alone. Our study suggests that penicillin tolerance with group C streptococci may occur frequently and may account for the poor outcome of serious group C streptococcal infections tested with penicillin alone.     

In vitro activity and mechanism of action of A21978C1, a novel cyclic lipopeptide antibiotic.

The in vitro activity of A21978C1, a novel cyclic polypeptide antibiotic, was compared with those of vancomycin, teichomycin, and several beta-lactam antibiotics against gram-positive bacteria. The new drug was at least as active as vancomycin against all species of streptococci and staphylococci tested, including methicillin-resistant Staphylococcus aureus and penicillin-resistant pneumococci. Activity of the drug was found to be strongly correlated with the calcium concentration in test media. Against enterococci, A21978C1 was bactericidal at concentrations near the MIC (MIC for 100% of the strains, 2 micrograms/ml), but combining that drug with gentamicin resulted in bactericidal synergism by time-kill methods. Studies were undertaken to examine the mechanism of action of the drug. A21978C1 did not interact with penicillin-binding proteins of bacterial cell membranes. No direct effect of the drug on the synthesis of DNA, RNA, or protein by a susceptible strain of Streptococcus faecalis could be demonstrated. However, A21978C1 inhibited peptidoglycan synthesis in early-log-phase cultures of both Streptococcus faecalis and Staphylococcus aureus.     

Penicillinase-producing Neisseria gonorrhoeae in the Netherlands: epidemiology and genetic and molecular characterization of their plasmids.

An ampicillin-resistant strain of Neisseria denitrificans was produced by serial passage of the organisms in media containing increased concentrations of antibiotic. The 400-fold increase in resistance obtained was a relatively stable characteristic. Ampicillin resistance in this organism was apparently related to a loss or modification of the penicillin-binding proteins associated with the cytoplasmic membranes. Membranes isolated from the ampicillin-resistant strain bound significantly less radioactive penicillin than those isolated from the parent strain and revealed one major and three minor penicillin-binding proteins. All four penicillin-binding proteins were present in reduced amounts or had a decreased capacity for penicillin binding in the ampicillin-resistant cells. The increased resistance did not involve enzymic degradation of the antibiotic or a general reduction in the permeability of the outer layers of the cell. No difference in the amount of peptidoglycan present in the parent and ampicillin-resistant cells or in the gross chemical structure of the peptidoglycans of the two strains was observed.     

Binding of beta-lactam antibiotics to penicillin-binding proteins of Staphylococcus aureus and Streptococcus faecalis: relation to antibacterial activity.

The binding of 14 structurally diverse beta-lactam antibiotics to penicillin-binding proteins of Staphylococcus aureus and Streptococcus faecalis was studied, and the results were examined in the context of the antibacterial activity of the compounds. Penicillin-binding proteins 1 (molecular weight, 87,000) and 3 (molecular weight, 75,000) of S. aureus and penicillin-binding proteins 1 (molecular weight, 105,000) and 3 (molecular weight, 79,000) of S. faecalis bound beta-lactam antibiotics at concentrations comparable to minimum inhibitory concentrations and might therefore be essential. The low affinity of S. faecalis penicillin-binding proteins, relative to that of S. aureus penicillin-binding proteins, toward most beta-lactam antibiotics is probably responsible for the resistance of the former organism to most of these compounds.     

Aminoglycoside-Inactivating Enzymes in Clinical Isolates of Streptococcus Faecalis: AN EXPLANATION FOR RESISTANCE TO ANTIBIOTIC SYNERGISM

Clinical isolates of enterococci (Streptococcus faecalis) with high-level resistance to both streptomycin and kanamycin (minimal inhibitory concentration >2,000 mug/ml), and resistant to synergism with penicillin and streptomycin or kanamycin were examined for aminoglycoside-inactivating enzymes. All of the 10 strains studied had streptomycin adenylyltransferase and neomycin phosphotransferase activities; the latter enzyme phosphorylated amikacin as well as its normal substrates, such as kanamycin. Substrate profiles of the neomycin phosphotransferase activity suggested that phosphorylation occurred at the 3'-hydroxyl position, i.e., aminoglycoside 3'-phosphotransferase. A transconjugant strain, which acquired high-level aminoglycoside resistance and resistance to antibiotic synergism after mating with a resistant clinical isolate, also acquired both enzyme activities. Quantitative phosphorylation of amikacin in vitro by a sonicate of the transconjugant strain inactivated the antibiotic, as measured by bioassay, and the phosphorylated drug failed to produce synergism when combined with penicillin against a strain sensitive to penicillin-amikacin synergism.No differences were found in the sensitivity of ribosomes from a sensitive and resistant strain when examined in vitro using polyuridylic acid directed [(14)C]-phenylalanine incorporation in the presence of streptomycin, kanamycin, or amikacin. Therefore, we conclude that aminoglycoside-inactivating enzymes are responsible for the aminoglycoside resistance, and resistance to antibiotic synergism observed in these strains.