Profiling of β-Lactam Selectivity for Penicillin-Binding Proteins in Escherichia coli Strain DC2

Abstract

Penicillin-binding proteins (PBPs) are integral players in bacterial cell division, and their catalytic activities can be monitored with β-lactam-containing chemical probes. Compounds that target a single PBP could provide important information about the specific role(s) of each enzyme, making identification of such molecules important. We evaluated 22 commercially available β-lactams for inhibition of the PBPs in live Escherichia coli strain DC2. Whole cells were titrated with β-lactam antibiotics and subsequently incubated with a fluorescent penicillin derivative, Bocillin-FL (Boc-FL), to label uninhibited PBPs. Protein visualization was accomplished by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) separation and fluorescent scanning. The examined β-lactams exhibited diverse PBP selectivities, with amdinocillin (mecillinam) showing selectivity for PBP2, aztreonam, piperacillin, cefuroxime, cefotaxime, and ceftriaxone for PBP3, and amoxicillin and cephalexin for PBP4. The remaining β-lactams did not block any PBPs in the DC2 strain of E. coli or inhibited more than one PBP at all examined concentrations in this Gram-negative organism.

INTRODUCTION

Penicillin-binding proteins (PBPs) are enzymes involved in bacterial cell wall biosynthesis and are the target of β-lactam antibiotics (1). Escherichia coli possesses five high-molecular-weight (HMW) PBPs, with three (PBP1a, PBP1b, and PBP1c) in class A and two (PBP2 and PBP3) in class B (2). This organism has seven low-molecular-weight (LMW; class C) d,d-carboxypeptidases or d,d-endopeptidases (PBP4, PBP5, PBP6, PBP6b, PBP7, PBP4b, and AmpH) (3, 4). Additionally, proteolytic cleavage of PBP7 yields PBP8 (5, 6). In E. coli, the class A and B PBPs are essential for cell growth and division and are therefore critical targets for β-lactam antibiotics. The specific tasks of each PBP have been difficult to discern, largely due to the redundant functions of these enzymes in most organisms (7). However, recent work has illuminated much about the substrate specificity (8–11), activators (12), and association of several PBPs with cell elongation or septal synthesis (13–15). Because the PBPs are crucial for cell wall biosynthesis, cell division, growth, and resistance to β-lactams, study of the roles of individual family members may be critical to devising ways to overcome and/or evade bacterial resistance.

β-Lactams have been used not only to treat bacterial infections but also as chemical probes to study the PBPs (13, 16–19). Various β-lactams target different PBPs and can therefore be used to probe the discrete functions of those proteins (18, 20–22). Radiolabeled β-lactams have been used as a detection tool to study the relationship between the concentration of β-lactams and their selectivity for PBPs in E. coli and other organisms. More recently, nonradioactive β-lactams have been utilized to examine PBP inhibition and activity (23, 24). In particular, fluorescent β-lactams, such as Bocillin-FL (Boc-FL; penicillin V conjugated to BODIPY FL dye), have enabled visualization of PBPs in live cells (25, 26). With this tool in hand, we sought to examine 22 commercially available β-lactams in E. coli to assess their PBP selectivity in live cells. Here, we report the selectivity profiles of these molecules and compare our results with previously reported in vitro data.

MATERIALS AND METHODS

Growth media and antibiotics.

Escherichia coli strain DC2, an antibiotic-hypersusceptible mutant (27), was grown in Luria-Bertani (LB) broth at 37°C to an optical density at 600 nm (OD₆₀₀) of 0.5.

Faropenem, doripenem, meropenem, (+)-6-aminopenicillanic acid (6-APA), penicillin V, penicillin G, ampicillin, methicillin, amdinocillin (mecillinam), oxacillin, cloxacillin, dicloxacillin, piperacillin, cephalexin, cefsulodin, cefoxitin, cephalothin, cefuroxime, and ceftriaxone were purchased from Sigma-Aldrich (St. Louis, MO). Cefotaxime was purchased from Calbiochem (Billerica, MA). Aztreonam and amoxicillin were obtained from MP Biomedicals (Solon, OH). Boc-FL was purchased from Life Technologies (Grand Island, NY).

All antibiotics were stored as solids and dissolved in Milli-Q purified water (unless noted otherwise) at 10 mg/ml immediately before each experiment. 6-APA, amoxicillin, and cefuroxime were not soluble in water. Thus, 6-APA was dissolved in phosphate-buffered saline (PBS) (pH 7.4) at a 5 mg/ml concentration. Amoxicillin and cefuroxime were dissolved in dimethyl sulfoxide (DMSO) at a 100 mg/ml stock concentration (further dilution in PBS decreases the final DMSO concentration). All stock solutions were serially diluted (10-fold) in PBS to yield 0.0001 to 1,000 μg/ml working solutions.

β-Lactam titration and detection of PBPs.

E. coli cells from 1.5 ml of an exponential culture at an OD₆₀₀ of 0.5 were harvested by centrifugation (8,000 × g for 2 min at room temperature). The cell pellets were washed in 1 ml of PBS (pH 7.4) by pipetting up and down a few times, and then the cells were pelleted using the same centrifugation conditions. The cells were resuspended in 50 μl PBS containing 0.0001 to 1,000 μg/ml β-lactam antibiotics, and a reference sample was resuspended in 50 μl PBS without antibiotics. After 30 min of incubation at room temperature, the cells were pelleted, washed in 1 ml PBS, and resuspended in 50 μl PBS containing 5 μg/ml Boc-FL (0.1% DMSO). After 10 min of incubation at room temperature, the cells were pelleted and washed in 1 ml PBS. Next, the cells were resuspended in 100 μl PBS and sonicated using a Branson Sonifier 250 instrument (Branson Ultrasonic, Danbury, CT) on ice (power setting 3, 30% duty cycle for three 10-s intervals with 10 s of cooling time between rounds). The membrane proteome was isolated by centrifugation at 21,000 × g for 15 min at 4°C. The supernatant was discarded, the membrane was resuspended in 100 μl PBS, and the samples were homogenized by sonication (power setting 1, 10% duty cycle for 1 s). The protein concentration was measured by the use of a NanoDrop 1000 Spectrophotometer (Thermo Scientific, Wilmington, DE). The protein concentration was adjusted to 2.5 mg/ml by dilution with PBS. A 51-μl volume of proteome sample was dispensed into a clean 1.5-ml microcentrifuge tube, and 17 μl of 4× SDS-PAGE loading buffer was added to each sample. The samples were heated for 5 min at 90°C to denature the proteins and then cooled to room temperature. A 10-μl volume of sample was loaded onto a 10% SDS-PAGE gel. The protein bands were separated by gel electrophoresis for 1.5 h at 180 V, 400 mA, and 60 W. The gel was rinsed with distilled water three times and scanned using a Typhoon 9210 gel scanner (Amersham Biosciences, Pittsburgh, PA) with a 526-nm-wavelength short-pass filter at 50-μm resolution. The gel images were analyzed using ImageJ software (National Institutes of Health, Bethesda, MD). The background signal of the gel images was subtracted, and the brightness and contrast were adjusted to optimize the signal-to-noise ratio (all operations were uniformly performed over the entire gel). Integrated density values were measured for gel band quantitation, and Boc-FL labeling of each PBP in antibiotic-treated samples is shown relative to samples not treated with antibiotics. Integrated density values were inserted into GraphPad Prism (GraphPad Software, La Jolla, CA) to create graphs showing relative percentiles of Boc-FL labeling versus inhibitory concentrations (ICs). Determinations of 50% IC values (IC₅₀s) were performed using the same software. For all dose-response curves, data were fitted to a four-parameter logistic equation.

RESULTS AND DISCUSSION

Traditional techniques for PBP detection.

The ability of various β-lactams to inhibit the PBPs in a given organism has traditionally been measured in cell membrane samples with radiolabeled penicillin (18). Most E. coli PBP studies have utilized this assay (13, 18, 19). Although in vitro experiments have provided important information about the affinity and selectivity of β-lactams for the PBPs, the activity of these proteins can be altered during membrane preparation. For instance, the effects of antibiotics on PBP7 and PBP8 in E. coli have historically not been reported due to lack of reproducibility, as these proteins appear to lose their activity during cell lysis and membrane isolation (28). There also appears to be strain-to-strain variability. More recently, fluorescent β-lactams have been used to detect PBPs in living cells (25, 26). Taking advantage of this advance, we investigated the selectivity profiles of numerous β-lactams for the PBPs in live E. coli strains using fluorescent penicillin (Boc-FL) as the readout probe.

Selectivity profile of known β-lactams for PBPs in live E. coli strain DC2.

In this study, 22 β-lactam antibiotics from several compound subclasses were tested for E. coli PBP selectivity in live cells (DC2 strain) to contrast previous in vitro work (Table 1 and Table 2) (13, 18, 19). A compound was deemed selective for a PBP if its IC₅₀ was at least 4-fold lower than that of the next most inhibited PBP. When a second or third PBP fell below this 4-fold threshold, the compound was considered to be coselective for two or three PBPs. Representative gel images and gel band quantitation graphs are provided in Fig. 1 (see also Fig. S2 and S3 in the supplemental material).

TABLE 1.

Inhibition of Boc-FL binding to PBPs in live E. coli by β-lactam antibiotics

β-Lactam	Concna (μg/ml)	Relative % Boc-FL labelingb
		PBP1a	PBP1b	PBP2	PBP3	PBP4	PBP5/PBP6	PBP7	PBP8
Monobactam
Aztreonam	0.1	197	193	208	45	105	211	109	123
Penem
Faropenem	0.1	103	93	36	102	52	97	36	48
Carbapenem
Doripenem	0.0001	81	84	52	87	28	95	49	74
Meropenem	0.0001	97	100	50	94	47	99	60	82
Penicillin
6-APA	1,000	132	90	61	58	59	97	100	93
Amdinocillin	1	91	104	43	104	99	102	88	101
Penicillin V	1	75	68	87	72	28	86	46	57
Penicillin G	1	86	77	76	64	40	100	47	75
Ampicillin	0.1	97	90	102	75	30	83	40	58
Amoxicillin	0.1	90	94	98	90	26	98	69	75
Piperacillin	0.1	100	97	95	30	96	99	75	90
Methicillin	10	94	68	85	29	28	101	45	58
Oxacillin	1	103	90	123	36	34	131	93	89
Cloxacillin	10	65	56	154	107	60	47	64	75
Dicloxacillin	10	105	71	158	95	48	118	57	66
Cephalosporin
Cephalexin	10	109	92	110	86	28	103	91	98
Cefuroxime	0.1	74	69	94	50	88	89	92	88
Cefotaxime	0.01	96	102	101	42	108	109	99	103
Ceftriaxone	0.01	110	109	114	47	110	119	113	130
Cephalothin	10	68	54	79	37	72	79	44	51
Cefsulodin	1	75	52	98	95	98	85	47	69
Cefoxitin	0.01	88	92	94	93	60	87	36	45

^aData represent concentrations of antibiotic required to reduce the subsequent labeling of Boc-FL to ≤50% for at least one PBP.

^bInhibition of ≥50% of Boc-FL labeling is indicated by underlined text. Data represent averages of the results of two independent experiments.

TABLE 2.

IC₅₀ and MICs of β-lactams in E. coli DC2

β-Lactam	MIC (μg/ml)	IC50(s) (μg/ml)a
		PBP1a	PBP1b	PBP2	PBP3	PBP4	PBP5/PBP6	PBP7	PBP8	PBP selectivity
Aztreonam	0.063	>1,000b	>1,000b	>1,000b	0.02	30	>1,000b	60	100	3
Faropenem	0.125	>1,000b	>1,000	0.06c	>1,000	0.08c	>1,000b	0.02	0.02	NS
Doripenem	0.016	>1,000	>1,000	0.0001	0.3	0.0001	2	0.0001	0.0001	NS
Meropenem	0.008	>1,000	>1,000	0.0001	0.07	0.0001	4	0.01	0.02	2, 4
6-APA	32	>1,000b	>1,000	>1,000	>1,000	>1,000	>1,000	>1,000b	>1,000b	NS
Amdinocillin	0.016	>1,000	>1,000	0.02	>1,000	10	>1,000	>1,000	>1,000	2
Penicillin V	4	8	2	50	2	0.1c	>1,000b	0.2c	0.08	4, 7, 8
Penicillin G	4	20	6	>1,000	1	0.3c	>1,000	0.09	0.3c	4, 7, 8
Ampicillin	0.25	6	2	2	0.3	0.01	>1,000	0.01	0.01	4, 7, 8
Amoxicillin	1	10	4	2	3	0.01	>1,000	0.1	0.05	4
Piperacillin	0.016	9	10	1	0.01	7	>1,000	0.1	0.2	3
Methicillin	4	>1,000	>1,000	>1,000	1	4c	>1,000b	3c	4c	NS
Oxacillin	0.5	>1,000b	>1,000b	>1,000b	0.09	0.1c	>1,000b	4	0.7	3, 4
Cloxacillin	1	>1,000b	>1,000b	>1,000b	>1,000b	3	>1,000b	7c	9c	4, 7, 8
Dicloxacillin	4	>1,000b	>1,000b	>1,000b	>1,000	2	>1,000b	10	8c	4, 8
Cephalexin	8	>1,000b	>1,000	>1,000b	20	2	>1,000b	70	100	4
Cefuroxime	0.125	1	0.2	>1,000	0.04	0.7	>1,000b	1	1	3
Cefotaxime	0.008	0.9	0.2	3	0.01	3	>1,000b	1	1	3
Ceftriaxone	0.016	0.3	0.1	0.2	0.01	1	>1,000	0.5	1	3
Cephalothind	2	>1,000	5	>1,000	3	20	>1,000	ND	ND	ND
Cefsulodin	32	2	0.8c	>1,000	20	>1,000	>1,000b	0.6c	0.4	1b, 7, 8
Cefoxitin	2	0.9	0.07	>1,000b	0.6	0.01	0.07	0.01	0.01	4, 7, 8

^aIC₅₀s were determined based on 10-fold concentration steps and are estimate values. Underlining indicates the lowest IC₅₀ for each compound. NS, not selective; ND, not determined.

^bCompounds causing increased PBP gel band intensity prevented reasonable fitting of the IC₅₀ curves and are reported as >1,000 μg/ml.

^cIC₅₀s were within 4-fold of the lowest value.

^dGraphPad Prism reported ambiguous numbers for several PBPs and precluded assignment of selectivity for this compound.

FIG 1.

Ceftriaxone titration of PBPs in E. coli. Whole cells were treated with various concentrations of ceftriaxone and subsequently labeled with Boc-FL. (A) Representative SDS-PAGE gel image for ceftriaxone titration of PBPs in E. coli. (B) Gel band quantitation for the gel shown in panel A, with the standard deviations of the results from two independent experiments plotted.

A significant number of the antibiotics that we tested, including aztreonam, specifically blocked PBP3, a class B HMW protein, confirming a previous report on E. coli strain W7 (29). Piperacillin, cefotaxime, ceftriaxone, and cefuroxime were selective for PBP3 in a dose-dependent manner (see ceftriaxone data in Fig. 1). In a previous study, cefotaxime had a lower IC₅₀ for PBP1a than for PBP3 in E. coli strain W7, indicating preferential inhibition of the former (30). In another report (31), cefotaxime selectively inhibited PBP3 in E. coli strain K-12, consistent with our result. Previously, ceftriaxone was shown to have a lower IC₅₀ for PBP1a and PBP2 than for PBP3 in E. coli strain W7 (30). We observed preferential inhibition of PBP3 in this study, which is consistent with another report for PBP3 in E. coli strain MC4100 (32). The HMW PBPs, including PBP3, are thought to be important targets in E. coli killing (33). Our data are consistent with this observation in that the IC₅₀(s) for one or several HMW PBPs was similar to or below the MIC values for a given compound (Table 2). Cloxacillin, dicloxacillin, and 6-APA were exceptions to this trend, suggesting that their mechanism of killing is not fully understood.

The selectivities of compounds in the penicillin class differed greatly. (+)-6-Aminopenicillanic acid (6-APA), the core of penicillin, did not appreciably inhibit (>50%) any of the PBPs at concentrations of up to 1 mg/ml. On the other hand, cells treated with penicillin V and G showed concentration-dependent selectivity for PBP4, PBP7, and PBP8. Ampicillin targets the same PBPs and differs from penicillin G only by the presence of an amino group (see Fig. S1 in the supplemental material). Interestingly, ampicillin was reported to have a lower IC₅₀ for PBP2 and PBP3 than for PBP4 in membrane samples when [¹⁴C]-benzylpenicillin was used as the readout probe for PBPs in E. coli strain K-12 (31). Amoxicillin, also an aminopenicillin, differs from ampicillin by the addition of a hydroxyl group on the benzyl side chain. This structural alteration renders amoxicillin selective for PBP4. In previous studies, cephalexin was shown to be selective for PBP3 in E. coli strains K-12 and MG1655 (16, 31, 34). In our work, titration of whole E. coli cells with cephalexin resulted in selective inhibition of PBP4 at 10 μg/ml. Cephalexin and ampicillin contain the same side chain in the C7 and C6 positions, respectively (see Fig. S1), which may contribute to their similar inhibition profiles.

Consistent with a previous study of E. coli KN126, amdinocillin specifically inhibited PBP2 (18). Meropenem, a carbapenem, inhibited PBP2 and PBP4. Doripenem and faropenem inhibited PBP2, PBP4, PBP7, and PBP8 and are therefore reported as not selective. Significant inhibition of PBP2 and PBP4 by doripenem and meropenem was observed in a previous study in E. coli strain MC4100 as well (35). The cephamycin cefoxitin inhibited PBP4, PBP7, and PBP8 and was least efficient at blocking PBP2, consistent with a previous report on PBPs in E. coli KN126 (19). Cefsulodin preferentially inhibited PBP1b, PBP7, and PBP8 but was formerly found to be selective for PBP1a alone in E. coli strain W7 (30). Cephalothin inhibited PBP3, PBP7, and PBP8 at 10 μg/ml, but we report the selectivity as not determined (ND) because the IC₅₀s of PBP7 and PBP8 could not be established from our data. In an earlier in vitro labeling assay, this compound inhibited only PBP3 in E. coli KN126, as PBP7 and PBP8 were not detected (19). In 2013, we reported that methicillin selectively inhibits PBP2x, a protein known to be important in the cell division machinery, in Streptococcus pneumoniae strain D39 (36). Consequently, we reasoned that the ortholog (PBP3) may be involved in the E. coli cell division process and would be specifically targeted by methicillin. In this study, methicillin inhibited the E. coli PBP3 as well as PBP4, PBP7, and PBP8.

The gel band signal of PBP5 and PBP6 (PBP5/6) increases with treatment with oxacillin at a high concentration in live E. coli strain DC2.

The penicillin compound oxacillin showed the greatest selectivity for PBP3 and PBP4 at 1 μg/ml. Surprisingly, a 10- or 100-fold increase in the oxacillin concentration resulted in higher observed gel band intensities for PBP5/6 while still inhibiting PBP3 and PBP4 (Fig. 2A and B). To further explore this change in protein labeling intensity, we performed an oxacillin titration experiment on growing cells in LB broth. This experiment yielded the same inhibition and increased band intensity profiles, confirming that it was not specific to nongrowing cells (e.g., those treated in PBS; see Fig. S4 in the supplemental material). Next, we examined cell morphology by light microscopy after treatment using oxacillin-PBS (100 μg/ml; 30 min). Cells were viable, with no discernible morphological change, implying that augmented protein labeling is unlikely to be due to breakdown of the cell wall structure and increased accessibility of the PBPs (Fig. 2C). We found that oxacillin analogs cloxacillin and dicloxacillin, which differ only by the addition of a chlorine group(s) to the benzyl side chain, yielded inhibition profiles different from that of the parent compound (see Fig. S2 and S3). Alternatively, treatment of E. coli cells with aztreonam, a dissimilar monocyclic β-lactam, caused increased Boc-FL labeling of several proteins, namely, PBP1a, PBP1b, PBP2, and PBP5/6. We confirmed that enhanced signal was not observed when oxacillin and aztreonam titrations were performed in vitro on E. coli strain DC2 membrane preparations, indicating that it occurs only on live cells and suggesting that it is not simply a function of protein accessibility (see Fig. S5 and S6). Oxacillin treatment was also performed on live E. coli strain K-12 cells, a less permeable wild-type strain, producing results similar to those from strain DC2 and demonstrating that this phenomenon is not strain dependent (see Fig. S7). Examination of the antibiotic titration gels by Coomassie blue staining revealed that the protein concentration did not change (see Fig. S8). Intriguingly, this outcome appears to be compound dependent because, although cephalexin also inhibits PBP3 and PBP4, no gel band intensity increases were observed for PBP5/6 upon treatment with this molecule (see Fig. S2). Further study will be required to determine why inhibition of PBP3 and PBP4 with high concentrations of oxacillin or aztreonam results in increased PBP5/6 gel band intensity.

FIG 2.

Oxacillin titration of PBPs in E. coli. Whole cells were treated with various concentrations of oxacillin and subsequently labeled with Boc-FL. (A) Representative SDS-PAGE gel image for oxacillin inhibition of the PBPs in E. coli over a range of antibiotic concentrations. (B) Gel band quantitation for the gel shown in panel A, with the standard deviations of the results from two independent experiments plotted. (C) Micrographs of E. coli cells not exposed to oxacillin and after 100 μg/ml oxacillin treatment for 30 min at room temperature. Scale bar, 1 μm.

This report provides an extensive analysis of the PBP selectivity for a library of β-lactams using Boc-FL as the readout tool. Eight of the 22 tested antibiotics were found to exhibit dose-dependent inhibition of a single PBP, while other molecules inhibited multiple PBPs. Five of these compounds targeted PBP3, while none of the tested compounds exhibited selectivity for class A PBPs PBP1a and PBP1b. Several of the compounds in the library (e.g., ampicillin and amoxicillin) display only small structural differences, but their PBP inhibition profiles were not similar. On the other hand, molecules that belong to different subclasses of the β-lactam family (e.g., aztreonam, piperacillin, and cefotaxime) showed similar selectivity profiles in some cases. The results of this study indicate that it is difficult to predict the PBP selectivity of the β-lactams on the basis of structural similarities alone and that inhibition profiles should be evaluated using both a direct assay and a wide range of compound concentrations for a particular strain. The data reported here will be valuable for the design of PBP-specific fluorescent β-lactam probes for rod-shaped organisms.