CP-45,899 in combination with penicillin or ampicillin against penicillin-resistant Staphylococcus, Haemophilus influenzae, and Bacteroides.

CP-45,899 is a new, semisynthetic beta-lactamase inhibitor. When tested alone, CP-45,899 displayed only weak antibacterial activity, with the notable exception of its potent action against penicillin-susceptible and -resistant Neisseria gonorrhoeae. A combination of 3.12 microgram of CP-45,899 per ml with 3.12 microgram of ampicillin per ml, tested in broth cultures, inhibited ca. 90% of resistant Staphylococcus and Haemophilus influenzae strains; similar data were obtained in a variety of media. The same combination of CP-45,899 with ampicillin or penicillin G inhibited 90% of Bacteroides fragilis as interpreted from agar dilution minimal inhibitory concentrations. Inhibitory concentrations of CP-45,899-ampicillin were bactericidal against H. influenzae strains and were as bactericidal as nafcillin or cephalothin against S. aureus. Ampicillin-resistant S. aureus, H. influenzae, and B. fragilis strains did not develop resistance to CP-45,899-ampicillin when transferred as many as six passages in the presence of a sublethal concentration of the combination.

chemically stable fl-lactamase inhibitor, which synergistically increases the activity of ,B-lactam antibiotics against,B-lactamase-producing grampositive and gram-negative microorganisms (2). We have previously shown that ampicillin or penicillin G combinations with CP-45,899 were particularly potent against penicillin-resistant Staphylococcus aureus, S. epidermidis, Haemophilus influenzae, and Bacteroides fragilis.
Good activity was observed with CP-45,899 as a single agent only against Neisseria gonorrhoeae (2). The purpose of the study reported below was to extend these findings with additional clinical isolates, measure bactericidal activity, assess the rate of resistance emergence, and examine the effect of growth media on the potency of CP-45,899-antibiotic combinations.
MATERLALS AND METHODS Materials. CP-45,899 was prepared in Pfizer Central Research Laboratories; ampicillin.3H20 and sodium penicillin G were supplied by the Quality Control Dept., Pfizer Inc. Amoxicillin.3H20 came from Carrera C. S. a. S., Milano. Sodium cephalothin and cephalexin-H20 were products of Eli illy & Co.; sodium cefoxitin was a product of Merck Sharp & Dohme. Sodium methicillin H20 was a product of Bristol Laboratories, whereas sodium nafcillin.H20 came from Wyeth Laboratories.
Microorganisms used were recent clinical isolates obtained from hospitals in several geographical areas within the eastern United States. All (3). The overnight culture was diluted 1,000fold in BHI broth, and 0.5 ml of this dilution was added to each cup containing 0.5 ml of BHI broth plus antibiotics. Thus the final inoculum was c1 x 106 cells per ml. Incubations were at 37°C for 18 h with S. aureus, S. epidermidis, and H. influenzae and for 24 h with Bacteroides species. The N. gonorrhoeae inoculum was prepared by scraping growth from an agar plate and emulsifying it in 1 ml of BHI broth. Turbidity was adjusted to a no. 2 MacFarland standard. A loopful of inoculum was streaked over the agar surface followed by incubation at 370C in a GasPak jar in the presence of 5% C02, 10% H2, and 85% N2 for 24 to 48 h. Synergy was defined as occurring when the MIC of each component in the combination was one-fourth (two log2 dilutions) or less its MIC as single agent. Antagonism was not observed.
MBC. Values for CP-45,899 in combination with ampicillin were measured by first determining MICs in broth medium. Cups containing broth lacking visible growth in tests for MICs were subcultured by streaking a loop calibrated to deliver 0.01 ml over the surface of antibiotic-free agar medium. The agar plates were incubated overnight (anaerobic techniques were used with B. fragilis). The minimal bactericidal concentration (MBC) was defined as the lowest concentration of antibiotic permitting growth of '5 colonies on subculture. Thus, the MBC indicates '500 colony-forming units (CFU) per ml or, as recommended, 99.95% kill (1). Bactericidal values for H. influenzae isolates were determined in enriched BHI broth as described above.
MIC inoculum consisted of -5 x 10' CFU/ml; incubation was aerobic at 370C for 24 h, and 0.01 ml was subcultured as described above. Thus, the MBCs for H. influenzae represent a 99.99% kill of the original MIC inoculum.
Killing rates. Experiments with CP-45,899 in combination with ampicillin or penicillin G were initiated by diluting an overnight culture in BHI broth (enriched for B. fragilis) containing an appropriate concentration of antibiotic. During incubation (on a shaker for S. aureus), aliquots were removed and diluted at 2-log intervals, and triplicate 0.1-ml portions of the appropriate dilutions were spread on BHI agar plates. After overnight incubation, colonies were counted and recorded as CFU/milliliter. All experimental procedures with B. fragilis were carried out in an anaerobic chamber.
Resistance emergence studies: S. aureus After an initial MIC determination in BHI broth, the highest concentration supporting growth comparable to the antibiotic-free control was diluted 1,000-fold and used as inoculum in the next MIC determination. This procedure was repeated five times. The initial MIC of H. influenzae was measured in enriched BHI broth.
The first dilution with good growth was diluted 10fold and incubated overnight. This culture was diluted 10-fold for use as the inoculum in the next MIC determination. The average H. influenzae MIC inoculum was -3.7 x 10' CFU/ml. This procedure was repeated six times. Except for anaerobic techniques and the use of freshly prepared enriched BHI, the resistance emergence pattern of B. fragilis was studied in the same manner as was S. aureus.
f8-Lactamase studies. Cell-free 6-lactamase preparations of S. aureus, S. epidermidis, and H. influenzae were prepared as described previously (2). The H. influenzae and B. fragilis strains produced a constitutive enzyme. The B. fragilis culture, initially containing -10' cells per ml, was grown for 5 h anaerobically, at which time 1 mM dithiothreitol and 0.01 M ANTIMICROB. AGENTS CHEMOTHER.
/3-mercaptoethanol were added. The cells were then harvested, and the cell-free ,t-lactamase was prepared as described previously (2) with the addition of the above sulfhydryl reagents to all solutions. All cell-free extracts were stored at -76°C. The rate of hydrolysis was determined by the micro-iodometric assay of Zimmermann and Rosselet (11). Ampicillin at 30 ItM (10.5 Ag/ml) was the substrate, and incubation was at 37°C for 10 min. The B. fragilis enzyme was assayed with potassium penicillin G at 30 uM (12.5 ug/ml) as substrate, and incubation was for 20 min at 37°C. The amount of CP-45,899 required to produce 50% inhibition was estimated from plots of inhibition versus CP-45,899 concentration over a range of 30 iM (7.6 .g/ ml) to Tables  2 and 3. The CP-45,899-pencillin combinations inhibited 90% or more of the resistant strains at 3.12 plus 3.12 ,tg per ml, respectively. The antibiotic combinations were as potent as cephalexin against resistant S. aureus and more potent than cephalexin against S. epidermidisand H. influenzae-resistant strains ( Table 2). Against Bacteroides species (Table 3), the antibiotic combinations were more potent than cefoxitin.
The mechanism of methicillin resistance in Staphylococcus is not due to /8-lactamase hydrolysis of methicillin; however, most of these strains do harbor a f8-lactamase (7). Consequently, the pencillin-CP-45,899 combinations generally produced a pronounced synergistic effect and were significantly more potent than cephalexin ( Table 4). Ratios of penicillin to CP-45,899 of 2:1 and 4:1 were more potent than a 1: 1 ratio only with methicillin-resistant staphylococci (Table 4). This probably reflects the reduced importance of,-lactamase as a resistance determinant in these strains, i.e., approximately the same amount of 8-lactamase inhibitor was required to inactivate the ft-lactamase of a methicillin-resistant Staphylococcus as was required for a normally resistant Staphylococcus, Ampicillin-CP45,899 (1:1) has been shown to be effec-tive in protecting mice infected with a methicillin-resistant S. aureus strain (2). The type of growth media employed significantly affected the potency of amoxicillin when tested against 25 resistant S. aureus strains. In three different broth and agar media, the median MICs of amoxicillin varied from a low of 3.12  1). The MICs of CP-45,899, as a single agent, ranged from 0.15 to 2.5 jig per ml against susceptible and resistant N. gonorrhoeae (Table 5). This suggests that CP-45,899 might be considered as a single agent in the therapy of N. gonorrhoeae infections. The combination ampicillin-CP-45,899 was twoto four-fold more active than CP-45,899 alone against the /8-lactamase-producing strains tested (Table 5).
Bactericidal activity. The MBCs of the ampicillin-CP-45,899 antibiotic combination against S. aureus were generally only one dilution higher than the MIC except for the f8-lactam-tolerant strains (Table 6A). (Five of the S. aureus strains met the definition of 8-lactam tolerance [8], i.e., had an MBC 264-fold higher than the MIC for nafcillin and cephalothin.) The MBCs of the combination antibiotic were equal to or lower than those observed with either nafcillin or cephalothin for many of the strains. In a killing-rate experiment, the combination of 3.12-3.12 jig per ml killed 99.9% of S. aureus cells     (Fig. 1). The combination of 6.25 ,ug of CP-45,899 per ml plus 0.78 ,tg of ampicillin per ml was also bactericidal, but the reverse ratio was not (Fig. 1). This indicates that sufficient  (Table 6B). The MBC for all of the 17 strains tested was -3.12-3.12 Ag/ml, employing a criterion of a 99.99% kill of the initial inoculum.
The MBC of ampicillin-CP-45,899 versus six strains of B. fragilis highly resistant to ampicillin was the same as the MIC (Table 60). In a killing-rate experiment starting with a high inoculum (5 x 107 cells per ml), penicillin G-CP-45,899 (3.12-3.12 ,ug/ml) killed >99.9% of the penicillin G-resistant B. fragilis cells within 6 h (Fig. 2).
Resistance emergence studies. When S. aureus strains resistant to ampicillin, but rather susceptible to the ampicillin-CP-45,899 antibiotic combination (MIC -0.78-0.78 ,ug/ml), were cultured in the presence of a sublethal concentration of the combination antibiotic, the MICs increased about two dilutions after three trans- Killing rates with the penicillin G-CP-45,899 combination against a penicillin-resistant B. fragilis strain. Symbols: 0, control, B. fragilis 78C004 (MIC forpenicillin G is 200 p,g/ml, that for CP-45,899 is 25 pg/ml, and that for pencillin G-CP-45,899 is 1.56-1.56 pg/ml); Ol, 3.12 pg of CP-45,899 per ml; A, 25 pg ofpencillin G per ml; 0, pencillin G-CP-45,899 at 3.12-3.12 pg/ml. fers. The MIC then remained constant at 3.12-3.12 jig/ml for three additional transfers (Table   7A). With S. aureus strains that had an initial MIC of 1.56-1.56 or 3.12-3.12 jig/ml for the combination, the MIC increased only one dilution after six transfers. The same type of relationship seemed to apply to the small sample of methicillin-resistant strains that were tested (Table  7A). The MIC of ampicillin-CP-45,899 remained the same after seven transfers of resistant H. influenzae strains in the presence of sublethal amounts of the combination antibiotic (Table   7B). The high inoculum used (-4 x 107 CFU/ ml) should have increased the probability of resistance emerging.
The median MICs of antibiotic combinations increased only one dilution after seven transfers with B. fragilis strains highly resistant to ampicillin (Table 70).
Thus, penicillin-resistant S. aureus, H. influenzae, and B. fragilis strains could not readily adapt (synthesizing more ,B-lactamase, increasing cell wall perneability barriers, etc.) to the ampicillin-CP-45,899 combination and counteract it. Also, if a few highly resistant cells existed in the initial population of the cultures tested, these cells were not able to become dominant in the culture after six or seven transfers in the presence of the antibiotic combination.
The 1:1 combinations of ampicillin-CP-45,899 and penicillin G-CP-45,899 demonstrated potent bactericidal activity against batteries of clinical isolates of ampicillin-resistant S. aureus, H. influenzae, and B. fragilis. CP-45,899, as a single agent, was active against susceptible and resistant N. gonorrhoeae. The potency of the antibiotic combinations is such that >90% of the resistant strains of S. aureus, S. epidermidis, H. influenzae, and Bacteroides species were in-Intersci. Conf. Antimicrob. Agents Chemother. 18th, Atlanta, Ga., 1978).