ANTIMICROBIAL AGENTS AND CHEMOTHERAPY, July 1983, p. 48-5300664804/83/070048-06$02.OO/OCopyright 0 1983, American Society for Microbiology

Vol. 24, No. 1

Influence of Clindamycin on Derepression of I-Lactamases inEnterobacter spp. and Pseudomonas aeruginosa

CHRISTINO C. SANDERS,* W. EUGENE SANDERS, JR., AND RICHARD V. GOERINGDepartment ofMedical Microbiology, Creighton University School ofMedicine, Omaha, Nebraska 68178

Received 18 November 1982/Accepted 2 May 1983

Previous studies in this and other laboratories have shown that derepression ofI-lactamases in strains of Enterobacter and Pseudomonas spp. is responsible forthe rapid development of resistance to a variety of '-lactam antibioties. Thepurpose of the current study was to evahuate the efects of clindamycin onderepression of ,B-lactamases in these two genera. In tests with four strains of eachgenus, clindamycin'diminished derepression in one isolate of each genus andcompletely pyevented derepression in a second Enterobacter isolate (strain 55).Additional tests with stirin 55 revealed that other inhibitors of macromolecularsynthesis did not completely prevent derepression of ,B-lactamase when tested atconcentrations that did not inhibit replication. However, clindamycin did notaffect synthesis of P-lactamase that was constitutively produced in a mutant ofthis strain (55M). It also did not inhibit derepression of ,3-galactosidase in eitherstrain 55 or 55M. Clindamycin did not diminish the bactericidal effects of A-lactamantibiotics against Enterobacter or -Pseudomonas spp. However, it enhanced thebaericidal activity of cefamandole against strain 55. These in vitro effects ofclindmycin on strain 55 that were related to prevention of derepression of -latamase were confirmed in vivo with an animal model of infection. These resultsindicate that in some strains, clindamycin can specifically prevent derepression of4-lactamases without inhibiting growth. Such a selective effect may provide a new

approQach for the enhancement of the antibacteril activity'of certain 3-lactamantibiotics.

Certain genera of inonfastidious, gram-nega-tive bacilli possess inducible P-lactamaseswhich, when derepressed, are responsible forresistance to a variety of 1-lactam antibiotics(19). This derepression may occur via oneoftwomechanisms: (i) spontaneous mutation of thewild type to a stably derepresaed -state or (ii)reversible derepression of the wild type by a 1-lactamase inducer (9, 10, 12, 19). The formermechanism has been shown to be responsible forthe, emergence of multiple ,B-lactam resistanceduring therapy with new cephalosporins (3, 17,20),> whereas the latter mechanism has beenshown to be responsible for the antagonism of 1-lactai drugs by cefo,.itin (8,, 11, '2, 27). Sincederepression, ofv -lactmases by either mecha-nism ultimately involves protein synthesis, wewished to determine whether antibiotics thatinhibit protein, biosynthesis would specijcallyaffect this process. Therefore, tle purpose ofthis investigation was to determine whether clin-damyein would influence -the derepression of 1-lactamases in strains of Enterpter spp. andPseudomonas aeruginosa, two organisms thatcharacteristically possess inducible enzymes.Clindamycin was chosen for study for several

reasons. First, it has been shown to inhibitsynthesis of certain proteins in gram-positivebacteria without affecting others (5, 6, 16, 23).Second, it possesses little growth-inhibitory ac-tivity against Enterobacter or Pseudomonasspp. These two features of clindamycin suggestthat it may preferentially inhibit synthesis of 13-lactamases in these gram-negative bacteria with-out produping.bacteriostasis, a condition knownto antagonize the lethal effect of 3-lactam antibi-otics.

MATERIALS AND METHODSStins. All bacterial strains were clinical isolates

possessing cefoxitin-inducible P-lactamases (10, 17,20, 22). The stable derepressed mutant of Enterobac-ter cloacae 55 was selected in the laboratory and hasbeen described in detail previously (10).Tesb wh b . Disk diffusion susceptibility

tests were performed by the method ofBauer et al. (2).Minimal inhibitory concentrations (MICs) of clinda-mycin and other antibiotics in the presence of 25 pg ofcefoxitin per ml were dermind by the agar dilutionmethod with Mueer- ton r and an inocuum of105 CFU per spot. The quantitative effect of eachantibiotic alone and in various combinations was de-termined in Mueller-Hinton broth. Each strain was

48

on January 3, 2019 by guesthttp://aac.asm

.org/D

ownloaded from

VOL. 24, 1983

suspended in broth (10' CFU/ml) containing the antibi-otics and incubated at 37°C in air for 24 h. At varioustime points, samples were removed and the number ofviable CFU per milliliter was determined by agardilution plate counts performed in duplicate. Theability of cefoxitin to antagonize cefamandole wasevaluated in a disk approximation test previouslydescribed in detail (21, 22). A low-concentration cef-oxitin disk was prepared for this test by dipping steriledisks into a solution of 30 ,ug of cefoxitin per ml.

-Lactmasc assays. 1-Lactamase assays were per-formed with a spectrophotometric assay in whichwhole cells were used as described previously (21).Enzymes were induced (derepressed) in wild-typecells by growing them overnight on Mueller-Hintonagar containing 25 Fg of cefoxitin per ml. The influ-ence of other antibiotics on the derepression of 1B-lactamases by cefoxitin was determined by incorporat-ing the antibiotics into the agar with cefoxitin duringthe induction period. The concentrations of antibioticsused in combination with cefoxitin did not inhibit thegrowth of the organisms (as determined by the agardilution susceptibility tests described above). Differ-ences in ,B-lactamase activity of >2 standard devi-ations (10 nmol/109 cells) were considered significant.f-Galacs assays. Strains were grown over-

night on Mueller-Hinton agar containing 20 ,g ofclindamycin per ml with and without 0.5 mM isopro-pylthio-1-galactoside. Cells were harvested from theplates after incubation, and crude cell-free sonicatedextracts were prepared as described before (9). Theprotein content of these extracts was determined bythe method of Lowry et al. (14). 3-Galactosidaseactivity of the extracts was determined by the methodof Wallenfels (26), in which the cleavage of o-nitro-

TABLE 1. Effect of clindamycin on cefoxitinderepression of 3-lactamases in Enterobacter spp.

and P. aeruginosa strainsf-Lactamase activity of cells grown

onaStrain

MHA MHA + CX (MHA + CX) +(25 Fg/ml) [CL (20 p&g/ml)]bE. cloacae

55 0 59 0 (100)84 22 71 7791 0 61 40(34)

E. aerogenes 76 0 38 36

P. aeruginosaL21 0 20C 22R15 0 17 14164 0 20 19166 0 43 30 (30)

aResults are expressed as nanomoles of cefaman-dole inactivated per 109 cells. MHA, Mueller-Hintonagar; CX, cefoxitin; CL, clindamycin.

b Numbers in parentheses represent percent reduc-tion in cefoxitin-induced ,B-lactamase activity pro-duced by clindamycin.

c Derepression of ,B-lactamase required 100 ,ug ofcefoxitin per ml.

CLINDAMYCIN AND ,3-LACTAMASES 49

phenyl-o-D-galactopyranoside substrate is followedspectrophotometrically at 405 nm.

Animal studies. The ability of clindamydin to alterthe in vivo efficacy of cefamandole was determinedwith an animal infection model described previously(8). Male CF-1 mice weighing 20 to 25 g each wereinfected intraperitoneally with one 100% lethal dose ofE. cloacae 55. Cefamandole nafate was administeredintramuscularly at five different dosages to groups ofmice (10 mice per group) at 1 and 3.5 h postchallenge.Clindamycin phosphate (200 mg/kg of body weight)was administered intramuscularly to some animals at0.5 h before, at the time of, and 1 and 3.5 h afterchallenge. This dose schedule for clindamycin waschosen to produce consistently high serum levels (18to 25 Fg/ml; data on ifie, The Upjohn Co.) of thebiologically active drug throughout the test period.Although this regimen did produce diarrhea in allanimals, it was not fatal and lasted only 36 h. Sodiumcefoxitin (3.12 mg/kg) was also administered at thetime of challenge to some animals. This amoint waschosen because it was the lowest, nonprotective doseof cefoxitin found in preliminary tests to be capable ofantagonizing the efficacy of cefamandole. The influ-ence of clindamycin on the efficacy of cefamandoleand on the ability of cefoxitin to antagonize cefaman-dole was determined by calculating the cefamandoledose required to protect 50%1o ofthe animals from deathat 48 h postchallenge. The 95% confidence levels werealso determined (13).

RESULTSDerepression (induction) of 1-lactamases in

wild-type strains. Initial studies were performedwith E. cloacae 55 to determine the minimalconcentration of clindamycin required to signifi-cantly inhibit the derepression of ,B-lactamase bycefoxitin. Clindamycin at 2 ,ug/ml did not affectderepression; 10 ,ug/ml diminished derepressionby 17%, and 20 ,ug/ml completely preventedderepression. Therefore, the ability of clindamy-cin (20 pug/ml) to affect cefoxitin derepression of,-lactamases in three other strains ofEnterobac-ter spp. and four P. aeruginosa strains wasdetermined. Clindamycin diminished the dere-pression of ,B-lactamase in one other Enterobac-ter strain (91) and one of the four Pseudotnonasstrains (166) (Table 1). Cefoxitin at 25 ,ug/ml didnot derepress 3-lactamase in P. aeruginosa L21.However, derepression did occur when 100 p,gof cefoxitin per ml was utilized. Clindamycin (20,ug/ml) did not affect this derepression.

Since E. cloacae 55 was most susceptbile tothe effects of clindamycin, this strain was usedin subsequent studies designed to further assessthe influence of clindamycin on derepression of,B-lactamases. In the first set of tests, the abilityof clindamycin (MIC, >64 Fg/ml) to diminishderepression of 1-lactamase by cefoxitin wascompared with those of (i) pirlimycin (U-57930E, a clindamycin derivative) and erythro-mycin, two other inhibitors of protein synthesislacking intrinsic activity against E. cloacae 55

on January 3, 2019 by guesthttp://aac.asm

.org/D

ownloaded from

50 SANDERS, SANDERS, AND GOERING

(MIC, >64 ,ug/ml); (ii) chloramphenicol and tet-racycline, two inhibitors of protein synthesispossessing intrinsic activity against E. cloacae55 (MICs, 8 ,ug/ml and 2 ,ug/ml, respectively);(iii) nalidixic acid, an inhibitor ofDNA synthesis(MIC, 2 tLg/ml); and (iv) rifampin, an inhibitor ofDNA-dependent RNA synthesis (MIC, 32Sxg/ml). In these tests, all antibiotics were usedat a concentration which, when combined with25 ,ug of cefoxitin per ml, did not inhibit thegrowth of E. cloacae 55. For antibiotics activeagainst E. cloacae 55, this was the highestconcentration that did not inhibit growth. Noneof the other antibiotics diminished derepressionof p-lactamase in E., cloacae 55 to the extent thatclindamycin did (Table 2). Of the other sixantibiotics tested, only tetracycline and rifampindiminished derepression of ,B-lactamase. How-ever, it was diminished only slightly by thesedrugs, in contrast to clindamycin, which com-pletely prevented derepression.

Since clindamycin appeared to prevent dere-pression of ,B-lactamase by cefoxitin in E. clo-acae 55, tests were performed to determinewhether it would also prevent antagonism ofcefamandole by cefoxitin, as was demonstrablein disk approximation tests. These tests wereperformed exactly as described previously, ex-cept clindamycin (20 /ml) was incorporatedinto the test agar. In .these disk approximationtests, in which a low-concentration cefoxitindisk was used, ciindamycin did prevent theantagonism of cefamandole by cefoxitin. In ad-dition, incorporation of clindamycin into theagar during standard disk diffusion susceptibility

TABLE 2. Effect of various antibiotics onderepression of ,B-lactamase by cefoxitin in E.

cloacae 55Medium (concn [6ig/mjl)' P-Lactamase activityb

MHA 0

MHA + CX (25) 72

MHA + CX (25) +:CL (20) 0 (100)PL (20) 74ER (20) 67CP (2) 80TC (0.5) 48 (33)NA (1) 74RF (4) 48 (33)

a MHA, Mueller-Hinton agar; CX, cefoxitin; CL,clindamycin; PL, pirlimycin; ER, erythromycin; CP,chloramphenicol; TC, tetracycline; NA, nalidixic acid;RF, rifampin.

b kesultg are expressed as nanomoles of cefaman-dole inactivated per 109 cells. Numbers in parenthesesrepresent percent reduction in derepression.

tests enhanced the zone of inhibition producedby the commercially available high-concentra-tion cefoxitin disk. In routine tests, there was nozone surrounding the cefoxitin disk, whereas a17-mm zone was obtained in tests performedwith 20 ,ug of clindamycin per ml in the agar.Clindamycin did not significantly alter the zonesproduced by eight other 1-lactam antibiotics(cefamandole, moxalactam, cefotaxime, ceftri-axone, carbenicillin, azlocillin, mezlocillin, andpiperacillin).

Effect of clindamycn on constitutive 0-1acta-mase. The studies described above clearly dem-onstrated the ability of clindamycin to preventderepression of P-lactamases by cefoxitin in anE. cloacae strain. Therefore, additional testswere designed to determine whether, in thissame organism, clindamycin would affect -

lactamases that were constitutively produced.These tests involved the 1-lactamase of E. clo-acae 55M, the fully derepressed mutant of strain55. The 1-lactatnase activity of the fully dere-pressed strain 55M was unaffected by clindamy-cin (Table 3). i

Effect of clindamycin on,B- Sinceclindamycin appeared to specifically preventderepression of 1-lactamase in E. cloacae 55,studies were performed to determine its effecton f3-galactosidase, another enzyme under re-pressor control iti this organism, in this samestrain. The influence of clindamycin on ,B-galac-tosidase in strain 55M was also determinedsince, unlike 1-lactamase, the 3-galactosidase inthis mutant is still under repressor control. Clin-damycin did not affect derepression of 3-galac-tosidase by isopropylthio-1-galactoside in eitherstrain when tested under the same conditionsthat had been found to prevent derepression of,B-lactamase by cefoxitin in strain 55 (Table 4).

Effect of cindamycin on the bactericidal activi-ty of 1-lactam antibiotics. The influence of clin-damycin (20 pg/ml) on the bactericidal activityof cefamandole against the four strains of En-terobacter spp. and moxalactam against the four

TABLE 3. Influence of clindamycin on ,-lactamaseunder repressor control (strain 55) and on fully

derepressed P-lactamase (strain 55M)f~-Lactamase

Medium (concn LgIpDJ)a activity in strain':55 55M

MHA 0 74MHA + CX (25) 63 75(MHA + CX) + [CL (20)] 0 75

a MHA, Mueller-Hinton agar; CX, cefoxitin; CL,clindamycin.

b Results are expressed as nanomoles of cefaman-dole inactivated per 109 cells.

ANTIMICROB. AGENTS CHEMOTHER.

on January 3, 2019 by guesthttp://aac.asm

.org/D

ownloaded from

VOL. 24, 1983

TABLE 4. Influence of clindamycin on 1-

galactosidase under repressor control in E. cloacae55 and 55M

Activity in E. cloacae strainb:Medium (concn [Lg/m1])a 55 55M

MHA + CL (20) 1.90 ± 0.40 2.10 ± 0.90MHA + CL (20) + IPTG 12.39 ± 6.66 16.20 ± 4.38

(0.5 mM)a MHA, Mueller-Hinton agar; CL, clindamycin;

IPTG, isopropylthio-f3-galactoside.b Results are expressed as nanomoles of o-nitro-

phenyl-o-D-galactopyranoside hydrolyzed per minuteper milligram of protein (mean ± 2 standard devi-ations).

P. aeruginosa strains was determined in quanti-tative tests. In none of the eight tests did clinda-mycin diminish the bactericidal activity of the 1B-lactam antibiotics. In tests with E. cloacae 55,complete killing was observed with the combina-tion of cefamandole' and clindamycin (Fig. 1).Clindamycin alone possessed little activity,whereas cefamandole alone produced an initialbactericidal effect which was followed by re-

growth of the strain.Animal studies. Since in vitro tests had shown

that clindamycin could prevent derepression ofg-lactamase in E. cloacae 55 and enhancedkilling of this strain by cefamandole, additionaltests were performed to determine whether simi-lar effects could be demonstrated in vivo. In oneseries of tests, the ability of clindamycin toprevent antagonism of cefamandole by cefoxitinin vivo was evaluated. Previous studies haveshown this antagonism to be due to the in vivoderepression of 1-lactamase by cefoxitin (8). In

10

o ,o84

2 4 6 8 ' 24TIME OF INCUBATION (hr.)

FIG. 1. Influence of 20 ,ug of clindamycin (CL) per

ml on bactericidal effect of 30 ,ug of cefamandole (CM)per ml in tests with E. cloacae 55. CO, Drug-freecontrol.

CLINDAMYCIN AND ,B-LACTAMASES 51

TABLE 5. Influence of clindamycin on in vivoefficacy of cefamandole and antagonism of

cefamandole by cefoxitin

Treat- Time (h)ment DUaof administration PD,~(Mg/kg)bsched- g postchallenge Poule

.0.5 0 1 3.5

I CM + + 83.5 (73.3-95.2)

II CMc + + 112.8 (84.1-151.3)Cx +

III CM + + 89.2 (79.4-100.2)Cx +CL + + + +

IV CM + + <25.0CL I+ I+ _+ +I

a CM, Cefamandole; CX, cefoxitin; CL, clindamy-cin.

b PD , Cefamandole dose required to protect 50oof the animals from death. Numbers in parenthesesrepresent 95% confidence levels.

this test, cefoxitin was administered at the timeof challenge; clindamycin was administered 0.5h before challenge, at the time of challenge, and1 and 3.5 h postchallenge; and cefamandole wasadministered at 1 and 3.5 h postchallenge (Table5). As was seen in the in vitro tests, clindamycinprevented the moderate in vivo antagonism ofcefamandole by cefoxitin. In a second series oftests, the direct influence of clindamycin on thein vivo efficacy of cefamandole was determined.In these tests, clindamycin and cefamandolkwere administered as before. There was nocefoxitin administered. Clindamycin significant-ly enhanced the in vivo efficacy of cefamandole(Table 5).

DISCUSSIONThe results of this study indicate that clinda-

mycin is capable of diminishing derepression ofP-lactamases in some gram-negative bacterialstrains. In general, this characteristic was -notshared by other inhibitors of protein, RNA, orDNA synthesis when tested at concentrationsthat did not inhibit replication. In fact, pirlimy-cin, a closely related analog of clindamycin (1),had no effect on derepression of P-lactamases.

Several previous studies have focused on theeffects of protein synthesis inhibitors on 3-lacta-mase synthesis in bacteria. However, most ofthese studies involved strains intrinsically sus-ceptible to the protein synthesis inhibitor. Mi-chel and co-workers have shown that chloram-phenicol and cephaloridine were synergisticagainst some Enterobacteriaceae strains (15).

on January 3, 2019 by guesthttp://aac.asm

.org/D

ownloaded from

52 SANDERS, SANDERS, AND GOERING

This synergy was demonstrable only in testswith cephaloridine-resistant, chloramphenicol-susceptible strains. Antagonism usually oc-curred in tests with strains susceptible to bothdrugs. This synergy was due to the inhibition ofA-lactamase synthesis by chloramphenicol andhas been shown to be operative both in vitro andin vivo (15, 18). In similar studies with Staphylo-coccus aureus, Michel et al. demonstrated thatclindamycin diminished penicillinase productionby this organism (16). This effect, however,occurred only in tests with clindamycin-suscep-tible strains. Our investigation is the first toreport an effect of clindamycin on 1-lactamasesynthesis in nonfastidious, gram-negative bacte-ria.The results obtained with E. cloacae 55 indi-

cated that certain processes may be highly sus-ceptible to inhibition by clindamycin, whereasothers are totally unaffected. In this strain, clin-damycin prevented derepression of P-lactamaseunder conditions that had no effect on derepres-sion of f-galactosidase. In tests with the dere-pressed mutant of this strain, 55M, clindamycinhad no effect on either constitutive P-lactamasesynthesis or P-galactosidase derepression. Hy-persusceptibility of inducible enzymes to inhibi-tion has been previously reported by Sypherdand DeMoss (24) and Sypherd et al. (25) instudies with chloramphenicol. Although P-lacta-mase was not evaluated, these authors reportedthat the synthesis of a number of enzymes underrepressor control in Escherichia coli was prefer-entially inhibited by chloramphenicol. Synthesisof enzymes constitutively produced were notinhibited by the drug. This preferential inhibitionof inducible enzyme synthesis reported by Sy-pherd et al. was a much broader effect than thatobserved in our study, as not all inducible en-zymes in E. cloacae 55 were affected by clinda-mycin. The molecular basis for this highly selec-tive effect of clindamycin is unknown. However,it may involve some direct effect of clindamycinon the repressor involved with the P-lactamaseoperon. Identification of the repressor will berequired before the clindamycin effect can bemore completely explained.The prevention of derepression of P-lacta-

mase by clindamycin in E. cloacae 55 wasassociated with the inability of cefoxitin to an-tagonize the activity of cefamandole against thisstrain both in vitro and in vivo. It was alsoassociated with an enhanced in vitro susceptibil-ity of the strain to cefoxitin itself. This latteraspect suggests that, for Enterobacter strainswith inducible P-lactamases, cefoxitin may be aself-antagonizing drug; i.e., cefoxitin derepress-es a p-lactamase which is partly responsible forthe high level of resistance of the organism to thedrug. Since cefoxitin is a poor substrate for this

enzyme, resistance may be due to the nonhydro-lytic barrier mechanism that mediates resistanceto nonsubstrate drugs (19). Prevention of dere-pression partially restores susceptibility to thedrug. The relatively low intrinsic potency andhigh inducer activity of cefoxitin for Enterobac-ter spp. may be the reasons that this self-antagonizing characteristic was so readily dem-onstrated.Clindamycin enhanced the in vitro bactericid-

al effect of cefamandole against E. cloacae 55.Previous studies have shown that the failure ofcefamandole to produce a complete bactericidaleffect in vitro is due to the appearance of thestably derepressed mutant among wild-type cells(10). Thus, the enhancement of the in vitrobactericidal effects of cefamandole may havebeen due to the prevention of P-lactamase dere-pression via mutation by clindamycin. This pos-sibility was supported by animal studies whichshowed a similar enhancement of the in vivoefficacy of cefamandole by cindamycin. Previ-ous studies have shown that the relatively low invivo efficacy of cefamandole in the treatment ofexperimental infections due to Enterobacterspp. is also due to the emergence of resistantmutants (7, 8). Although the precise mechanismresponsible for this direct enhancement of cefa-mandole activity by clindamycin is unknown, itprobably also involves some interaction with theP-lactamase operon via the repressor protein.The ability of clindamycin to prevent dere-

pression of P-lactamases in some bacterialstrains without inhibiting their growth suggests anew approach for enhancement of the antibac-terial effects of ,-lactam antibiotics. In the past,attempts to enhance the activity of 3-lactamantibiotics have been focused upon the use ofinhibitors of P-lactamase function such as clavu-lanic acid (4). The new approach would involvethe use of specific inhibitors of ,B-lactamasesynthesis. These agents, having little effect onbacterial replication, would not have the poten-tial for antagonism. The results obtained in thisstudy with clindamycin indicate that such anapproach is feasible. Clearly, clindamycin doesnot represent the ideal inhibitor of P-lactamasesynthesis. It inhibited derepression of ,-lacta-mase in only a few strains. Also, the concentra-tion required to obtain this limited effect repre-sented levels achievable in human serum only atone of the highest recommended doses (600 mgintravenously). A search for protein synthesisinhibitors with a broader range of activity asinhibitors of derepression is clearly needed.However, as the results of this study indicate,the search should be performed among inhibi-tors that do not affect bacterial replication. Anagent without any antibacterial activity thatwould selectively inhibit synthesis of inducible

ANTIMICROB. AGENTS CHEMOTHER.

on January 3, 2019 by guesthttp://aac.asm

.org/D

ownloaded from

VOL. 24, 1983

,B-lactamases in gram-negative bacteria wouldcertainly enhance the antibacterial activity ofthe new P-lactam antibiotics. It would also mini-mize the problems that have been recently asso-ciated with derepression of ,B-lactamases (3, 7, 8,10-12, 17, 19-22, 27).

ACKNOWLEDGMENTSThis study was supported in part by a grant-in-aid from the

Upjohn Co. Kalamazoo, Mich.We thank L. P. Couture, S. L. Genova, and E. A. Ruff for

their technical assistance.

LITERATURE CITED

1. Abonkhai, V. I., C. E. Cherubin, M. A. Shulman, M.Jhagroo, and U. Bancroft. 1982. In vitro activity of U-57930E, a new clindamycin analog, against aerobic gram-positive bacteria. Antimicrob. Agents Chemother.21:902-905.

2. Bauer, A. W., W. M. M. Kirby, J. C. Sherris, and M.Turck. 1966. Antibiotic susceptibility testing by a stan-dardized single disk method. Am. J. Clin. Pathol. 45:493-496.

3. Beckwlth, D. G., and J. A. Jahre. 1980. Role of cefoxitin-inducible beta-lactamase in a case of breakthrough bacte-remia. J. Clin. Microbiol. 12:517-520.

4. Cole, M. 1980. f1-Lactams as ,B-lactamase inhibitors. Phil.Trans. R. Soc. London Sec. B 289:207-223.

5. Geminmell, C. G. 1978. Effect of subinhibitory concentra-tions of antibiotics on experimental pyogenic infections inmice, pp. 512-514. In W. Siegenthaler and R. Luthy (ed.),Current chemotherapy, vol. 1. American Society forMicrobiology, Washington, D.C.

6. Gemmell, C. G, P. K. Peterson, D. Schmeling, Y. Kim, J.Matthews, L. Wannamaker, and P. G. Quie. 1981. Potenti-ation of opsonization and phagocytosis of Streptococcuspyogenes following growth in the presence of clindamy-cin. J. Clin. Invest. 67:1249-1256.

7. Goering, R. V., C. C. Sanders, and W. E. Sanders, Jr.1978. Comparison of BL-S786 with cephalothin, cefaman-dole and cefoxitin in vitro and in treatment of experimen-tal infections in mice. J. Antibiot. 31:363-372.

8. Goering, R. V., C. C. Sanders, and W. E. Sanders, Jr.1982. Antagonism of beta-lactam antibiotics by cefoxitinin treatment of experimental infections in mice. Antimi-crob. Agents Chemother. 21:961-967.

9. Gootz, T. D., and C. C. Sanders. 1983. Characterizationof 1-lactamase induction in Enterobacter cloacae. Antimi-crob. Agents Chemother. 23:91-97.

10. Gootz, T. D., C. C. Sanders, and R. V. Goering. 1982.Resistance to cefamandole: derepression of beta-lacta-mase by cefoxitin and mutation in Enterobacter cloacae.J. Infect. Dis. 146:34-42.

11. Kuck, N. A., R. T. Testa, sad M. Forbes. 1981. In vitroand in vivo antibacterial effects of combinations of beta-lactam antibiotics. Antimicrob. Agents Chemother.

CLINDAMYCIN AND 1-LACTAMASES 53

19:634-638.12. Lampe, M. F., B. J. Allan, B. H. Minshew, and J. C.

Sherris. 1982. Mutational enzymatic resistance of Entero-bacter species to beta-lactam antibiotics. Antimicrob.Agents Chemother. 21:655-660.

13. Litchfield, J. T., Jr., and F. Wilcoxon. 1949. A simplifiedmethod of evaluating dose-effect experiments. J. Pharma-col. Exp. Ther. 96:99-113.

14. Lowry, 0. H., N. J. Rosebrough, A. L. Farr, and R. J.Randall. 1951. Protein measurement with the Folin phenolreagent. J. Biol. Chem. 193:265-275.

15. MIchel, J., H. Bornstein, R. Luboshitzky, and T. Sacks.1975. Mechanism of chloramphenicol-cephaloridine syn-ergism on Enterobacteriaceae. Antimicrob. Agents Che-mother. 7:845-849.

16. MIchel, J., P. Stessman, and J. Stessman. 1980. Effects ofsubminimal inhibitory concentrations of erythromycin,clindamycin, and pristinamycin on the penicillinase pro-duction of Staphylococcus aureus. Antimicrob. AgentsChemother. 17:13-15.

17. Preheim, L. C., R. G. Penn, C. C. Sanders, R. V. Goe-ring, and D. K. Giger. 1982. Emergence of resistance to 1-lactam and aminoglycoside antibiotics during moxalactamtherapy of Pseudomonas aeruginosa infections. Antimi-crob. Agents Chemother. 22:1037-1041.

18. Sacks, T., J. Mchel, A. Durst, and J. Stessman. 1977.Inhibition of J-lactamase: synergistic bactericidal effect ofcombination of chloramphenicol-cephaloridine. J. Antimi-crob. Chemother. 3:525-526.

19. Sanders, C. C. 1983. Novel resistance selected by the newexpanded-spectrum cephalosporins: a concern. J. Infect.Dis. 147:585-589.

20. Sanders, C. C., R. C. MoeHering, R. R. Martin, Jr., R. L.Perkins, D. G. Strike, T. D. Gootz, and W. E. Sanders, Jr.1982. Resistance to cefamandole: a collaborative study ofemerging clinical problems. J. Infect. Dis. 145:118-125.

21. Sanders, C. C., and W. E. Sanders, Jr. 1979. Emergenceof resistance to cefamandole: possible role of cefoxitin-inducible beta-lactamases. Antimicrob. Agents Chemo-ther. 15:792-797.

22. Sanders, C. C., W. E. Sanders, Jr., and R. V. Goering.1982. In vitro antagonism of beta-lactam antibiotics bycefoxitin. Antimicrob. Agents Chemother. 21:968-975.

23. Shibl, A. M., and I. A. Al-Sowaygh. 1979. Differentialinhibition of bacterial growth and hemolysin productionby lincosamide antibiotics. J. Bacteriol. 137:1022-1023.

24. Sypherd, P. S., and J. A. DeMoss. 1963. The stimulationby chloramphenicol of "repressor" formation in Esche-richia coli. Biochim. Biophys. Acta 76:589-599.

25. Sypherd, P. S., N. Strauss, and H. P. Treffers. 1962. Thepreferential inhibition by chloramphenicol of induced en-zyme synthesis. Biochem. Biophys. Res. Commun.7:477-481.

26. Wallenfels, K. 1962. Beta-galactosidase assays. MethodsEnzymol. 5:212-213.

27. Waterworth, P. M., and A. M. Emmerson. 1979. Dissoci-ated resistance among cephalosporins. Antimicrob.Agents Chemother. 15:497-503.

on January 3, 2019 by guesthttp://aac.asm

.org/D

ownloaded from