Zbl. Bakt. Hyg., I. Abt. Orig. A 252, 222-229 (1982)

Department of Medical Microbiology, University of Kiel, Fed. Rep. of Germany

Bacteriological Investigations with Beta-Lactamase Inhibitors*

Bakteriologische Untersuchungen mit Beta-Lakramase-Inhibitoren

UWE ULLMANN

With 2 Figures' Received February 13, 1982

Abstract

MIC's of beta-lactamase producing strains of P. aeruginosa, E. coli, P. vulgaris, P. rett­geri, E. cloacae, K. pneumoniae, S. marcescens are determined against piperacillin and inthe case of S. aureus against amoxycillin and penicillin G. All the strains studied wereresistant to these antibiotics. Their MIC's were determined in the combination with dif­ferent concentrations (2, 5, 10 mgjl) of clavulanic acid and penicillanic acid sulfone re­spectively. The combination of piperacillin and clavulanic acid acts synergisticly againstKlebsiella sp., indolpositive Proteus sp., and E. coli. Penicillanic acid sulfone was lessactive than clavulanic acid apart from indolpositive Proteus sp. Against S. aureus howeverboth inhibitors showed the same synergistic efficacy. In detail both inhibitors are veryactive against constitutive penicillinase, they also exhibited a good activity against TypeOXA-2, OXA-3 and SHV-1 beta-lactamase ; they were less active to Type TEM-1 andOXA-1 enzyme and did not work against Type Ib, TEM-2 and Type IV beta-Iactamase,

In investigations using the biophotometer it could be demonstrated that the beta­lactamase inhibitor and the penicillin should be given simultaneously in order to obtainan optimal synergistic effect.

Zusammenfassung

Die minimalen Hemmkonzentrationen bera-Laktamase-produzierender Starnme vonP. aeruginosa, E. coli, P. vulgaris, P. rettgeri, En. cloacae, K. pneumoniae, S. marcescenswurden gegeniiber Piperacillin und irn Faile von S. aureus gegeniiber Arnoxycillin undPenicillin G bestimmt. Aile untersuchten Starnme waren resistent gegeniiber diesen Anti­biotika, Zur Untersuchung der synergistischen Wirkung wurde die minimale Hernrnkon­zentration mit verschiedenen Konzentrationen (2, 5, 10 mg/I) der bera-Laktarnaseinhibiro­ren Clavulansaure und Penicillansaure ermittelt. Die Kombination von Piperacillin undClavulansaure zeigt synergistische Aktivitat gegeniiber Klebsiella sp., indolpositiven Pro­teus sp, und E. coli. Penicillansaure war weniger aktiv als Clavulansaure, abgesehen vonindolpositiven Proteus sp. Gegeniiber S. aureus besafsen beide Inhibitoren synergistische

". Herrn Prof. Dr. med. Richard-Ernst Bader zum 70. Geburtstag gewidmet.

Beta-Lactamase Inhibitors 223

Wirkung. Im einzelnen besalsen sie inhibierende Aktivitat gegenuber konstitutiver Penicil­linase, gegenuber T yp OXA-2, OXA-3 und SHV-1 beta-Lakt amase. Weniger akt iv warensie gegeniiber dem TEM -1 und OXA-1 Enzym. Keine Aktivitat dagegen wurde beobachtetgegeniiber Typ Ib, TEM-2 und T yp IV beta-Laktamase.

Bei Untersuchungen mit dern Bioph otometer konnte gezeigt werden, daiS bera-Lak­tamaseinhibitoren und Penicilline gleichzeirig den Kulturen zugegeben werden rniissen, urneinen optimalen synergistischen Effekt zu erhalt en.

Introduction

An important factor in the resistance of Gram -positive and Gram-negativebacteria to betalactarn antibi otics is their capacity to produce inactivating enzymes.Penetration barriers for antibiotics in the cell wall and reduced affinity to bindingproteins are further resistance mechanisms.

In recent years the search for beta-lactamase inhibitors has been the subject ofintensive scientific activity. This would enable therapeutic use of the less toxic butfrequently ineffective penicillins. Clavulanic acid (Brown et al., 1976; Reading andCole, 1977) and penicillanic acid sulfone (English et aI., 1978) are two substanceswhich only possess a low ant ibacter ial activity, but which selectively inhibit certainbeta-lactamases. Penicillanic acid sulfone is manufactured semisyntheticallywhereas clavulanic acid is a metabolic product of Streptomyces clavuligerus. Chem­ically, both compounds have great similarity with the nucleus of the penicillins.

Th e aim of the following study was to analyze the synergistic effect of beta­lactamase inhibitors and penicillins against beta-l actamase producing bacteria.The action of beta-lactamase inhibitors was investigated in relation to their con­centr ation against 127 microo rganisms. T hese had been recently isolated fromclinical material and were resistant to penicillins.

In a basic study, the bactericidal kinetics of the comb ination of piperacillin andpenicillanic acid sulfone was analyzed with a beta-lactamase producing K. pneu­moniaestrain.

Materials and Methods

Bacterial strains: Strains recently isolated from clinical specimens were included in thestudy ; they were identified according to Cowan and Steel (Cowan, 1979).

Antibiotics used: Piperacill in, batch no. 1907 (Cyanamid, Munich), arnoxycillin, batchno. CT 196/1786 (Beecham-Wiilfing, Neuss), mezlocillin, batch no. 3813 (Bayer, Lever­kusen), penicillin G, batch no. 174 (German Penicillin Company, Gottingen).

Beta-lactamase inhibitors : Penicillanic acid sulfone batch ED-V-001-90 (Pfizer, Munich),clavulanic acid, batch no. 13 (Beecham-Wiilfing, Neuss).

Beta-lactam ase assay : The nitrocefin test was used.Bactericidal k inetics : This was investigated as described earlier (Ullmann, 1975) by means

of the Biolog II phot ometer (Jouan, Paris).Sensitivity testin g : Minimal inhibitory concentration (MICs) were determined by the

broth dilution test ; twofold dilution steps were made : piperacillin 128 - 0.06, penicillin G,amoxycillin and mezlocillin 8 - 0.00375.

Test medium was Mueller -Hinton broth CM 405 (Oxoid, London ). The inoculum was2 x 105 bacter ia /mI. Reference strains were S. aureus AT CC 25923, E. coli ATCC 25922and P. aeruginosa AT CC 27853.

15 Zbl. Bakr. Hyg., l. Abt. Orig. A 252

224 U.Ullmann

Synergism study: To investigate the synergistic efficacy, 10, 5 and 2 mgjl penicillanicacid sulfone or clavulanic acid were added to each antibiotic dilution used.

Checking of the substrate profile: To investigate which inhibitor inactivates which typeof beta-Iacrarnase, paper discs of 6 mrn diameter containing 10 flg amoxycillin werecharged with 20 Ilg of the respective inhibitor. The sizes of the inhibition zone comparedto those of amoxycillin alone were registered using the following bera-lacramase producers :

S. aureus : constitutive penicillinase, E. coli: type 1b chromosomal cephalosporinase,E.coli: OXA-l beta-lacramase, E.co li: OXA-2 beta- lactarnase, E.coli : OXA-3 bera-lacra­rnase, E. coli : TEM-1 beta-lactamase, E. coli : TEM-2 beta-lactamase (Sykes and Matthew,1976), K. aerogenes : type IV chromoso mal broad-spectrum bera-lactarnase (Richmond andSykes, 1973), K. aerogenes: SHV-1 bera-lacrarnase (Matthew, 1979).

Results

The results of the synergism investigation are summarized in Table 1 for theGram- negative and in Table 2 for the Gram-positive bacteria. The P. aeruginosa

stra ins investigated were gentami cin-resistant. Their MIC's were distinctly in­fluenced by addition of penicillanic acid sulfone and cIavulanc acid only in theconcentration 10 mg/l, H owever, altoge ther there is no synergistic effect. With theE. coli strains synergistic efficacy was still observ ed even in the concentra tio n of2 mg/l cIavulanic acid, while for the same effect 10 mg/I pen icillanic acid sulfonewere needed . Penicillani c acid sulfone has a somew hat better efficacy than clav­ulan ic acid against ind ole-posit ive Proteus species. Using K. pneumoniae theaction of cIavulani c acid is almos t the same over the range 10-2 mg. All strainscame into the susceptib ility range. Both inhibito rs showed a restr icted actionagai nst E. cloacae beta-lacramases. The same applies to S. marcescens apa rt fromat the high concentra tion of 10 mg/l c1avulanic acid.

Table 1. Geometric mean of piperacillin MIC's (mg!l) in the presence of different concen­trations of c1avulanic acid and penicillanic acid sulfone

Strains (No.) Pipera­ciIlin

Clavulanic acid10 mg!l 5 mg!l 2 mg!l

Penicillanic acid sulfone10 mg!1 5 mg!l 2 mg/l

P. aeruginosa (12)E. coli (21)Proteus indole pas. (16)K. pneumo niae (28)E. cloacae (15)S. marcescens (13)

107.6 3286.06 4.879.5 11.824.3 3.275.5 36.854.6 9.9

45.26.2

12.93.9

42.116.7

60.47.8

14.75.4

48.623.2

45.212.03.58.6

46.333.6

60.320.75.69.3

64.044.1

80.639.39.9

13.464.049.0

T he inhibito rs showed syne rgistic efficacy agai nst the constit utive penicillinaseof S. au reus in combination with amoxycillin at each concen tra tion between 2 to10 mg (T able 2). The same ap plies to the penicillin G-resista nt strains, which comeinto the sensitivity range due to the two inh ibitor s in all concentra tions investi­gated.

Beta-Lactamase Inhibitors 225

Table 2. Geometric mean of penicillin G and amoxycillin MIC's (mgjl) in the presence ofdifferent concentrations of clavulanic acid and penicillanic acid sulfone

Strains (No.) Penicillin G Clavulanic acid Penicillanic acid sulfonelOmg/l 5 mg!l 2 mgjl 10mg/! 5 mg/! 2mg!!

S. aureus (22) 12.1 0.4 0.5 0.8 0.6 0.8 1.5Amoxycillin

S. aureus (22) 9.4 0.8 1.1 1.1 0.9 1.3 1.8

The growth curves of K. pneumoniae ATCC 4352 were continuously registeredin the biophotometer after addition of 4--64 mg!l piperacillin at the beginning ofthe lag phase. The regrowth of the cultures occurred in consequence of the degra­dation of piperacillin by beta-lactamases after 3.5 h under the influence of 4 mg/l,after 5 h with 8 mg/I, after 6 h with 16 mg/l, after 7 h with 32 mg/l and after 8.5 hwith 64 mg/l (Fig.1).When the antibiotic was added to the cultures at 70 % trans­mission during the logarithmic growth phase, no influence was observed on bac­terial cultures even at concentrations of 128 mg/l. By addition of 10 mg/l peni­cillanic acid together with piperacillin at the beginning of the bacterial multipli­cation (Fig. 2), a bactericidal effect occurred after a latency period of 2.5 h, whichremains permanent even at the lowest concentration of 4 rng/l piperacillin. Onthe other hand no effect was observed, if the antibiotic and the inhibitor wereadded during the logarithmic growth phase at 70 Ofo transmission; the same wasfound out adding the inhibitor in the lag phase and the antibiotic in the log phase(70 % transmission).

The substrate profile of the inhibitors was investigated by the agar diffusionmethod. Amoxycillin alone showed an inhibition zone (IZ) of 12 mm against S.aureus (constitutive penicillinase), and one of 10 mm against K. aerogenes (SHV-1beta-lactamase). All other reference strains were resistant. By addition of clav­ulanic acid/penicillanic acid sulfone, a favorable effect against constitutive peni­cillinase was observed (IZ: 31/27 mm); the two inhibitors were ineffective againsttype 1 b beta-lactamase and the TEM-2 enzyme efficacy was again present againstOXA-2 beta-lactamase (IZ: 25/21 mm), against OXA-3 beta-lactamase (IZ: 24/18 mm), against TEM-1 beta-lactamase (IZ: 18/15 mm) as well as against SHV-1(IZ: 27/23 mm). The efficacy against OXA-1 beta-lactamase was restricted (IZ:13/13 mm).

Discussion

If the classification of Richmond and Sykes (1973) is used, clavulanic and peni­cillanic acid sulfone have no effectiveness against bera-lactamase of class I (chromo­somally mediated) cephalosporinases. In agreement with the results of Fu and Neu(1981) activity was demonstrated for clavulanic acid against class II beta-lac­tamases; according to our own results this also applies to penicillanic acid sulfone.The efficacy of the two inhibitors against OXA-1 beta-lactamase (Wise et al., 1980)is greatly restricted, and there is no activity against TEM-2 beta-lactamase andtype IV beta-lactamase, The inactivation of the beta-lactamases by the two inhibi-

N N 0\

Kle

:bs

.p

ne

um

.A

TC

C4

35

2

~ S 8" ~ ::l ::l

..-....

.....'

I I. '.-

'._t;t

::.·

····

····

····

·-I:

·~·..·_

_····

,~..··

·.·

.;::

::;:

:J..~/"

.·~:::~~~-..,.";~

:!-;t.\'rt,,~~:.;G~~~~

rai!

.~..

.....

.:~~

-.,.

.~_.-

1=:<;

;>...

....

\-j

_.-'

S<;,

..~

••.-

---

•....•

..,.~

_..r'

"','

.,~..

,_,

."/

"..,

,',.,,·-

C,..-.,.

--

.-"/

.....1

~././"

I~~

~::-....

C~_.:

:.:.,

·:·..···l

:'/'<

;,,!~:

~?:';"/

':

:../

....:

C4

-....~

./C

,:g

row

lhCq~"

rOI

.'..

C2

:4

mgl/

~perlJclIl/n

C3

:B

mgll

.;C

4:Ie

mg

ll.:·

Cs:

32

mg

l1'

Ce

:e4

mrj

l

I :c;:;

~'I.

T

1.2

. ~1

0

." i-,.-

~T~;~~

,:4l

~~-·i··-

~-r"'

-/1

...'/"

",

C!:-

:("I~i

!"'f

Fig.

1.G

row

thcu

rves

ofK

.pn

eum

onia

eu

nd

erth

ein

flue

nce

ofv

ario

us

pip

erac

illi

nco

nce

ntr

atio

ns

(C,-

C.:

Cu

vet

te1

-6).

Beta-Lactamase Inhibit ors 227

.1

Plpe//lc ll/ in

Pen lc ll//ln lc/lc ld sultone

C2 : 4 mg/IC3 : 8 ­C4 ; 16 ­C 5 : 32 ­C6 : 64 -

C2-C6 ; fO -

F ,- .-..._..-,;::;.........-···-0- ···········~.....

I ....:.J:;.:,/.'

t ' C I : glowlh centro!,.,.'

AS'"I n h l b lt o l

7. T1= I"'I I

Klebs . pneum.ATCC 4352

I

oFig. 2. Growth curves of K. pneumoniae under the influence of different pipcr acillin con­centrations and constant doses of penicillanic acid sulfone.

tors is dose dependent. A greater dependence was observed for penicillanic acidsulfone , especially regard ing the piperacill in-resistant E. coli strains. Possibly thiscan be att ributed to the weaker efficacy of penicillanic acid sulfone against TEM-lenzymes. In agreement with other investigators (Wise et al., 1980; Hunter et al.,1980) restr icted activity to the enzymes of S. marcescens, E. cloacae and P. aerugi­nosa was demonstrated for clavulanic acid and now also for penicillanic acidsulfone. These strains produce chromo somally mediated bera-lacrarnases of class I,but also plasmid-mediated beta-lactamases of class II. The latter are in some casesalso inhibited so that the synergistic effect (e. g. piperacillin + clavulanic acidagainst S. marcescens) observed especially at high concecntrations can be readilyexplained. However, this is not sufficient to bring these bacterial strains into thesensitivity range. Wise et al. (1980) demonstrated that both inhibitors are activeagainst the beta-lactarnase of B. tragi/is.

Van Klingeren and Wi;ngaarden (1981) showed that in bera-lacrarnase-positivegonococcal strains 4 mg/l clavulani c acid are necessary to bring them into thesusceptibility range for amoxycillin. The SHV-1 beta-lactamase produced by K.aerogenes is blocked by both inhibitors, which was also demonstrated by Ball et al.(1980) for c1avulanic acid (13).

Altogether c1avulanic acid possesses a better activity against the beta-lactamasesproduced by Gram-negative bacteria (apart from indole -postive Proteus sp.) thanpenicillanic acid sulfone.

In the investigations with the biophotometer it has been demonstrated that thebeta-lactamase inhibitor and the penicillin should be given simultaneously in orderto obtain an optimal synergistic effect. If the inhibitor is added first to the cultures,the bera-lactamase which has been produced is blocked. However, the growingbacteria produce new enzyme which inactivates the penicillin added later. Theconcentration of the inhibitor should be enormously increased in order still toget an inhibitory effect under these experimental conditions. The situation is

228 U.Ullmann

similar in the case of simultaneous addition of penicillin and inhibito r at highbacterial densities (70 010 transmission correspond to about 5 X 107 CFU/ml ) atwhich a high beta-lacramase concentration can be expected.

The therapeutic combination of beta-lactamase inhibitors with penicillins is alogi cal consequence, especiall y because clavulanic acid for example possessessimi lar pharmacokinetic properties to those of ampicill in and amoxycillin (Baller aI., 1980). Thus a combinat ion of amoxycillin and c1avul anic acid (125 mgc1avulanic acid + 250 mg am oxyciIIin) is being clin icall y tested at present as ano rally avai lable antibiotic (Ball er aI., 1980; Munch et aI., 1981).

Because of the low toxicity of the penicillins it would be desirable to bringother antibiotics into the sensitivity range also, e. g. the ureidopenicillins. For thispurpose the beta-lactamase inhibitors should be available therapeutically as mono­substances. It would be even more a scientific goal to link the molecule of the beta­lactamase inhibitor to the beta-lacram ring of the penicillins. The first step to over­come the resistance of microorganisms was made by linking of 6-iodo-penicillanicacid derivatives to mecillinam (Grunberg and Cleeland, 1977), as well as thelinking of penicillanic acid sulfone to ampicillin (Baltzer et aI., 1980).

Ackn owledgement. The author thanks Mrs. Gabi Reese for her careful and excellenttechnical assistance.

References

Ball, A. P., A .M. Geddes, P. G. Davey, J. D. Farrell, and G.R. Brookes: Clavulanic acid andamoxycillin : a clinical, bacteriological and pharmacological study. Lancet I (1980) 620­623

Baltzer, B., E. Binderup, W. v. Daehne, W. O. Godtjredsen, K.Hansen, B.Nie lsen, H.Seren­sen, and S. Vengedel : Mutual pro-drugs of fJ-lactam ant ibiotics and fJ-Iactamase inhi­biters. J. Antibior, 33 (1980) 1183-1192

Brown , A. G., D. Butterworth, M. Cole, G. Hanscomb, J. D. Hood, and C. Reading : Na­rurally-occuring fJ-lactamase inhibitors with antibacterial activity. 1. Anribior, 29 (1976)668- 669

Cowa n, S. T. : Cowan and Steel's Manual farth identification of medical bacteria. 2 nd ed.Cambridge University Press, Cambridge-London-New York-Melbourne (1979)

English, H. R ., ].A.Retsema, H.E.Girard, J.E.Lynch, and W. E.Barth : CP 45899, a beta­lactarnase inhibitor that extend s the antibacterial spectrum of beta-lactam s: Initial bac­teriological characterization. Antimicrob. Agents Chemoth er. 14 (1978) 414-419

Fu, K. P. and H. C. Neu: Synergistic activity of cefoperazone in combination with fJ-lac­tarnase inhibitors. J. Antimicrob. Chemother. 7 (1981) 287-292

Grunb erg, R. and J. R. Cleeland : In vitro activity of the 6 fJ-amidino penicillanic acidderivative, mecillinam, chemically linked or combined in varying ratios with 6-aminopenicillanic acid derivatives. J. Ant imicrob. Chemother. Suppl, B 3 (1977) 59- 68

Hunter, P. A., K. Coleman , ]. Fisher, and D. Taylor : In vitro synergistic prop erties of cla­vulanic acid with ampicillin, amoxycillin and ticarcillin. J. Antimicrob. Chemother. 6(1980) 455-470

Matthew, M .: Plasmid-mediated fJ-Iactamases of Gram-negative bacteria : properties anddistr ibution. J. Antimicrob. Chemother. 5 (1979) 349-358

Munch, R., R. Liithy, ]. Blaser, and W.Siegenthaler: Hum an pharmacokinetics and CSFpenetration of c1avulanic acid. ] . Antimicrob. Chemother. 8 (1981) 29-37

Reading, C. and M. Cole: Clavulanic acid : a beta-lacrarnase inhib iting bera-Iactarn fromStreptomy ces clavuligerus. Antimicrob. Agents Chemother. 11 (1977) 852-857

Beta-Lactamase Inhibitors 229

Richmond, M.H. and R.B.Sykes: The fi-lactamases of Gram-negative bacteria and theirpossible physiological role. Advanc. Microbial Physiol, 9 (1973) 31-85

Sykes, R. B. and M. Matthew: The fi-lactamases of Gram-negative bacteria and their rolein resistance to fi-lactam antibiotics. ]. Antimicrob. Chemother. 2 (1976) 115-157

Ullmann, U.: In-vitro-Untersuchungen tiber die Wirkung der Kombination von Genta­mycin und Beta-Lactarnantibiotika auf Pseudomonas aeruginosa. Immun. Infekt.3(1975) 79-85

Van Klingeren, B. and L.]. van Wi;ngaarden: Inhibition of fi-lactamase in penicillinaseproducing gonococci by clavulanic acid. ]. Antimicrob. Chemother. 8 (1981) 79-83

Wise, R., }.M.Andrews, and K.A.Bedford: Clavulanic acid and CP-45899: a comparisonof their in vitro activity in combination with penicillins. J.Antimicrob. Chemother. 6(1980) 197-206

Prof. Dr. U.Ullmann, University Department of Medical Microbiology, BrunswikerStr. 2-6, D-2300 Kiel, Fed. Rep. of Germany