Inhibition of the RTEM. beta.-lactamase from Escherichia coli

Jun 8, 1982 - ß-Lactamases: Molecular Studies. Andrew Coulson. Biotechnology and Genetic Engineering Reviews 1985 3 (1), 219-254 ...
0 downloads 0 Views 709KB Size
Biochemistry 1982, 21, 2857-2862

2857

Inhibition of the RTEM P-Lactamase from Escherichia coli. Interaction o f the Enzyme with Derivatives of Olivanic Acidt Christopher J. Easton and Jeremy R. Knowles*

ABSTRACT:

From chemical and kinetic studies of the interaction of the RTEM @-lactamasefrom Escherichia coli with three derivatives of olivanic acid, MM22382 ( l ) , MM13902 (2), and MM4550 (3), a mechanism for the inhibition of the enzyme by these compounds is proposed: the interaction proceeds by formation of an acyl-enzyme, the 42-pyrroline,

which may either deacylate or undergo tautomerization to the more tightly bound 4'-pyrroline. The ability of olivanic acids to inhibit the enzyme thus depends on the partitioning of the acyl-enzyme to the 4'-pyrroline (a process that competes with the normal hydrolytic pathway) and on the rate of regeneration of free enzyme from this complex.

M a n y members of the carbapenem family of @-lactamsare effective as antibiotics against @-lactamase-producingbacteria (e.g., Basker et al., 1980). To be effective, these species must be resistant to the enzyme-catalyzed hydrolysis reaction. We report here studies on the interaction of one group of carbapenems, the olivanic acids (Butterworth et al., 1979; Brown et al., 1977; Hood et al., 1979), with the purified plasmidencoded RTEM' @-lactamase,to evaluate how these antibiotics protect themselves against the hydrolytic action of the enzyme. In our earlier report on the interaction of MM22382 (1)

Betchworth, Surrey, United Kingdom. MM13902, as the sodium salt, was a pale yellow crystalline solid. MM4550, as the sodium salt, was an off-white powder. MM22382, as the sodium salt, was an orange powder. These materials were stored desiccated at -20 OC. Solutions were prepared by weight by using a Cahn 25 electrobalance. Ultraviolet spectroscopic measurements were made in 0.1 M sodium phosphate buffer, pH 7.0, at 30 "C by using a Perkin-Elmer 554 or 575 spectrophotometer. MM13902 has A,, at 223 and 304 nm (t = 14 100 and 15 200 M-' cm-', respectively), MM4550 has A,, at 237 and 282 nm ( t = 13500 and 12000 M-I cm-l, respectively), and MM22382 has,,A at 225 and 304 nm ( E = 13 500 and 13 300 M-' cm-I, respectively). HPLC' was carried out on a Waters Associates chromatograph equipped with a differential ultraviolet detector operating at 254 nm, using a pBondapak CI8 reverse-phase column (0.39 X 30 cm) and eluting with water (0.4 mL m i d ) . The retention times of MM13902, MM4550, and MM22382 were 35, 29, and 52.5 min, respectively. TLC was carried out on silica (Analtech) eluted with 1butanol/methanol/water (4: 1:2) and on cellulose (EastmanKodak) eluted with 2-propanol/water (7:3). The RJvalues of MM13902, MM4550, and MM22382 were 0.6,0.45, and 0.75, respectively, on silica and 0.85,0.75, and 0.9, respectively, on cellulose. Minor decomposition of each of the olivanic acid derivatives occurred on storage, giving rise to two products in each case, as determined by HPLC and TLC. The olivanic acids were purified by HPLC so that, except where noted, they were >95% pure when used. 'H and I3C NMR spectra were recorded on a Varian CFT20, a Varian XLlOO, a Jeol FX270, or a Bruker WM300 instrument. The purification of the TEM-2' 0-lactamase that was used has been described previously (Charnas & Knowles, 198 1).

coo-

10

M M 22382

U

1

I

cooMM 13902

2

CHx

-03SH 0x*N ' HCOCH3

-

cooM M 4550

3

and MM13902 (2) with the @-lactamase,Schemes I and I1 (A or B) were shown to be the minimal kinetic schemes necessary to accommodate the behavior of these two olivanic acid derivatives, respectively (Charnas & Knowles, 1981). In the present work we probe the nature of these interactions in more detail and propose a scheme that accommodates the observed behavior. We also report studies on the interaction of the @-lactamase with a third olivanic acid derivative, MM4550 (3). Experimental Procedures Materials Olivanic acid derivatives, M M 13902, MM4550, and MM22382, were generous gifts of Beecham Pharmaceuticals, From the Department of Chemistry, Harvard University, Cambridge, Massachusetts 02138. Received November 18, 1981. This work was supported by the National Institutes of Health and the National Science Foundation.

0006-2960/82/0421-2857$01.25/0

Methods The procedures used to investigate the kinetics of the interaction of MM4550 with the @-lactamase have been described previously in the report on the interaction of MM 13902 with the enzyme (Charnas & Knowles, 1981). Reactions were studied by following the absorbance change at 282 nm, and calculations were made on the basis of the measured Ac at 282 nm of 10 100 M-' cm-I. The buffer used in experiments in] Abbreviations: RTEM specifies the source of the plasmid [see Datta & Kontomichalou (1965)] and TEM-2 specifies the enzyme [see Sut-

cliffe (1978)l; DEAE, diethylaminoethyl; HPLC, high-performance liquid chromatography.

0 1982 American Chemical Society

2858

BIOCHEMISTRY

volving product isolation was 10 mM Et3NH-HC03,pH 7.0. Enzyme-Catalyzed Hydrolysis of MM22382 (1).A solution of enzyme (100 pL, 1 pM) in buffer was added to a solution of 1 (0.25 mg, 0.75 pmol) in buffer (20 mL). The mixture was incubated at 30 "C for 0.5 h, then cooled, and concentrated to -2 mL. For removal of the enzyme, the residue was applied to a Bio-Gel P-2 gel filtration column (1.5 X 20 cm) at 4 'C, eluting with buffer. Fractions eluting between 0.7 and 2.0 column volumes were pooled, aqueous NaOH (15 pL, 0.1 M) was added to prevent acid-catalyzed decomposition of the product, and the mixture was freeze-dried to give 4a as a white powder: UV max 262 nm ( E 10600 M-'

"5

;h$R, E

~~H

R

0

0-

0.05-

N

m

0.025

-

0

11

(u

R'

coo-

4

0

0-

HN

/ coo-

5

0 0

E

0

? -S

o

R = -OH

R''

b

R = -0SOf

R ' Z -S

UI

a_NHCOCH3

n

a

NHCOCH~

0

I

I

1

I

5

10

I5

PO

time ( m i d

cm-l); TLC, Rfvalues of 0.8 on cellulose and 0.65 on silica; HPLC, retention time 13.5 min. Enzyme-Catalyzed Hydrolysis of MMl3902 ( 2 ) and MM4550 ( 3 ) . To a solution of enzyme (200 mL, 2.5 pM) in buffer, incubated at 30 "C, were added portions of 2 or 3 (0.3 mg, -0.7 pmol) at hourly intervals for 10 h. After a further 2 h at 30 OC, the mixtures were worked up as described for the hydrolysis of 1 by the enzyme. Hydrolysis of 2 gave 4b as a white powder: UV max 262 nm (t 12 000 M-' cm-I); TLC, Rfvalues of 0.7 on cellulose and 0.55 on silica; HPLC, retention time 9 min; 'H NMR, see Table I. Hydrolysis of 3 gave 4c as a white powder: UV max 248 nm ( 6 16 300 M-' cm-I); TLC, Rfvalues of 0.65 on cellulose and 0.4 on silica; HPLC, retention time 8 min; 'H NMR, see Table I. Base-Catalyzed Hydrolysis of MM22382 (l),MMl3902 (2), and MA44550 ( 3 ) . An aqueous solution of 1, 2, or 3 (0.2-0.3 mg, 0.3 mM) and NaOH (0.6 mM) was incubated at 30 OC for 12 h, and the solution was then freeze-dried. The products were identified as 4a from 1,4bfrom 2,and 4c from 3, by comparison with the products obtained by enzymecatalyzed hydrolysis. So that the product samples for I3C N M R could be obtained, the same procedure was followed, except that 15 mg of 2 or 3 was used. The samples used in this experiment each contained 15% of the two impurities formed on storage, one of which was shown by HPLC to be the normal hydrolysis product. The freeze-dried reaction products were chromatographed on a column (1 X 20 cm) of DEAE-cellulose (DE-52, Whatman) equilibrated with 10 mM Et,NH-HC03 at 4 OC, eluting with a linear gradient (10-300 mM) of,aqueous Et3NH.HC03 (200 mL), pH 7.0. Fractions containing 4 (b or c), as determined by HPLC, were pooled, NaOH (2 molar equiv) was added, and the mixtures were then freeze-dried. Analysis by ultraviolet, TLC, HPLC, and 'H N M R showed that the isolated material was 4 (b or c), and l3C N M R spectra were recorded (see Table 11).

-

Results and Discussion In our earlier investigation (Charnas & Knowles, 1981) the interaction of the RTEM j3-lactamase with 1 and 2 was studied. Compound 1 behaves simply as a good substrate of

1: Changes in absorbance a t 282 nm accompanying the hydrolysis of MM4550 (3) by @-lactamase. MM4550 (3.9 WM)was incubated with @-lactamase(2 pM) in 0.1 M sodium phosphate buffer, pH 7.0, a t 30 OC. FIGURE

the enzyme, and the kinetic scheme shown in Scheme I suffices to describe the observed behavior. In contrast, 2 is a poor substrate and an excellent inhibitor of the enzyme. The inhibition derives from a branching of the normal hydrolytic pathway in which the acyl-enzyme intermediate partitions to a transiently stable, inactive form of the enzyme. In the experiments described below we have investigated the interaction of MM4550 (3) with the p-lactamase and extended our studies on the interaction of 1 and 2 with the enzyme. Interaction of the j3-Lactamase with MM4550 ( 3 ) . Incubation of 0-lactamase with 3 results in the disappearance of the chromophore at 282 nm and the appearance of a new chromophore at 248 nm. The kinetic characteristics of this interaction are complicated by a second-order reaction between 3 and the enzyme that is significant only at concentrations of 3 above about 5 pM. This reaction does not involve the enzyme's active site, since 0-lactamase that has been inactivated by clavulanic acid (Fisher et al., 1978) still causes the loss of A282nm with a rate constant of about 25 M-' s-l. At lower concentrations of 3 (