Multistep Resistance Selection and Postantibiotic-Effect Studies of the Antipneumococcal Activity of LBM415 Compared to Other Agents

Abstract

LBM415 is a peptide deformylase inhibitor active against gram-positive bacterial species and some gram-negative species. In multiselection studies, LBM415 had low MICs against all Streptococcus pneumoniae strains tested, regardless of their genotype, and selected resistant clones after 14 to 50 days. MIC increases correlated with changes mostly in the ₇₀GXGXAAXQ₇₇ motif in peptide deformylase. The postantibiotic effect of LBM415 ranged from 0.3 to 1.4 h.

LBM415 is a peptide deformylase (PDF) with predominant activity against gram-positive bacteria and good potency against pneumococci (4, 5, 8, 9, 18). The current study extends the evaluations of the antipneumococcal activity of LBM415 by (i) testing by a multistep methodology the capability of LBM415 to select for resistance in 12 strains compared to those of different classes of antibiotics which show clinical significance for the treatment of community-required pneumococcal respiratory tract infections, including amoxicillin-clavulanate, cefuroxime, cefdinir, ceftriaxone, imipenem, azithromycin, clarithromycin, levofloxacin, gatifloxacin, and moxifloxacin; (ii) testing the mechanisms of resistance of selected resistant clones; and (iii) examining the postantibiotic effects (PAEs) of the compounds listed above against nine pneumococci.

The following strains were studied for multistep resistance selection: four erythromycin-susceptible strains; three erm(B)-positive strains; two mef(A)-positive strains; and one strain each with erm(B) plus mef(A), an L4 mutation, and a 23S rRNA mutation. Of these, three were penicillin susceptible, four were penicillin intermediate, and five were penicillin resistant; three strains were quinolone resistant; and nine strains were quinolone susceptible. The strains used for the PAE studies comprised three erythromycin-susceptible strains (MICs, ≤0.06 μg/ml), three erm(B)-positive strains, and three strains with mef(A). Of these strains, three were penicillin susceptible, two were penicillin intermediate, four were penicillin resistant, and all were quinolone susceptible. LBM415 laboratory-grade powder was obtained from Novartis Laboratories, Hanover, NJ. The other antimicrobials were obtained from their respective manufacturers. Prior MIC testing of all strains was by the CLSI (formerly NCCLS) macrodilution method (14). Multipassage resistance selection was done as described previously (11). Daily passages were continued until a greater than fourfold increase in the MIC was found (minimum number of passages, 14; maximum number of passages, 50). The stability of the acquired resistance was determined from the MIC after 10 passages on drug-free agar. The identities of all resistant strains and the parent isolates were tested by pulse-field gel electrophoresis (11). All LBM415-resistant mutants and their parents were analyzed for the presence of mutations in defB gene, which encodes PDF. The defB gene was amplified with the following primers: defB6up (5′-GAGAAAATGTCTGCAATAGAACGT-3′) and defB6dn (5′-CCCCGTCTTGCAACGGGA-3′). These primers were used for sequencing analysis as well. Transformation of Streptococcus pneumoniae was performed as described previously (6, 17). In the transformation experiments, S. pneumoniae strain R6 was used as the DNA recipient and strain A9 served as the control. Purified DNA (final concentration, 1 μg/ml) from isolates with defined defB mutations was used (Table 1). Selection was done on brain heart infusion agar plates with 5% sheep blood and appropriate antibiotics (LBM415, 1 or 2 μg/ml; streptomycin, 100 μg/ml).

TABLE 1.

S. pneumoniae multistep resistance selection, mechanism of resistance detection, and transformation results for LBM415

Strain	Initial MIC (μg/ml)	Resistance selected		Retest MIC (μg/ml) of LBM415 after 10 antibiotic-free subcultures	Mutation(s) in PDF	Transformation expt result/MIC (μg/ml)a
		MIC (μg/ml)	No. of passages
3009	1	>16	14	>16	G69V	+/16
1076	0.25	8	39	2	D176N	−
1077	0.5	8	24	4	D176N	−
3665	0.25	4	27	1	—b	NT
3243	0.5	4	20	1	G69C, Q172H	+/8
3676	0.125	2	20	2	R135C	−
19	0.25	4	14	2	—	NT
24	0.5	4	27	2	G139F	+/4
2527	0.25	4	27	0.5	—	NT
37	1	16	50	2	G69R, P76R	+/4
3413	0.25	2	14	1	—	NT

^aMICs were determined by macrodilution. Transformation experiment results: +, positive selection on LBM415 plates (LBM415, 2 μg/ml); −, no mutants obtained on LBM415 plates (1 and 2 μg/ml LBM415); NT, not tested.

^b—, mechanism not detected.

All quinolone-resistant clones and their parental strains were analyzed for the presence of alterations in the quinolone resistance-determining regions (QRDRs) encoded by the parC, parE, gyrA, and gyrB genes with the primers and the cycling conditions described previously (13, 15). The genes encoding ribosomal proteins L4, L22, and domains II and V of 23S rRNA of the macrolide-resistant parents and the multistep-selected macrolide-resistant clones were amplified and sequenced (16, 17).

The PAE was determined as described previously (3, 11) by exposure to 10× the MIC for 1 h.

The results of multistep resistance selection tests are presented in Tables 1 and 2. LBM415 yielded resistant clones of 11/12 strains after 14 to 50 days, with the LBM415 MICs increasing from 0.125 to 1.0 μg/ml (parents) to 2.0 to >16.0 μg/ml (mutants); the moxifloxacin MICs rose from 0.125 to 1.0 μg/ml (parents) to 1.0 to 8.0 μg/ml (mutants) after 19 to 50 days for 7/12 strains. Gatifloxacin MICs rose from 1 to 2 μg/ml to 16 μg/ml for 3/12 strains after 14 to 20 days; levofloxacin MICs increased from 1 μg/ml (parents) to 16 μg/ml (mutants) for 2/12 strains after 18 to 31 days; azithromycin MICs rose from 0.016 to 0.03 μg/ml to 0.25 to >64 for 3/6 strains after 20 to 33 days; and clarithromycin MICs rose from 0.016 to 16 μg/ml to 0.25 to >128 μg/ml for 4/8 strains after 25 to 44 days. Amoxicillin-clavulanate, cefuroxime, cefdinir, ceftriaxone, and imipenem did not yield resistant clones after 50 days.

TABLE 2.

S. pneumoniae multistep resistance selection and mechanism of resistance detection for macrolides and quinolones used in the study

Strain	Antibiotic	Initial MIC (μg/ml)	Resistance selected		Retest MIC (μg/ml) after 10 antibiotic-free subculturesa											Resistance mechanism detected
			MIC (μg/ml)	No. of passages	LBM415	AMC	IMP	CXM	CDR	CRO	AZM	CLR	LVX	GAT	MXF
3009	LVX	1	16	18	0.5	≤0.008	≤0.004	≤0.016	0.06	0.016	128	0.5	32	8	4	Mutations GyrA (S81Y) and ParC (S79Y)
3009	GAT	1	16	14	2	0.016	0.008	0.03	0.125	0.016	>128	2	0.5	8	0.125	—b
3009	MXF	0.25	2	19	0.5	0.03	0.008	0.03	0.06	0.03	>128	4	2	2	4	Mutations GyrA (S81F) and ParC (D83N)
3009	CLR	16	>128	25	1	≤0.008	≤0.004	0.03	0.125	0.016	128	64	0.5	0.25	0.125	—
1076	MXF	1	8	35	0.125	0.03	0.008	0.03	0.06	0.016	>128	>128	32	8	4	ParC mutation S79Y
1076	GAT	2	16	20	0.25	0.016	0.008	0.125	0.125	0.016	>128	>128	32	8	4	ParC mutation S79Y
1077	AZM	0.03	>64	29	0.25	0.03	0.008	0.03	0.06	0.03	>64	>128	32	8	4	23S rRNA substitution A2058Gc
3665	MXF	0.125	1	38	0.5	0.25	0.03	0.5	0.125	0.06	8	2	4	0.5	1	GyrA mutation S81Y
3665	CLR	2	16	26	0.25	0.25	0.03	0.125	0.25	0.03	8	8	0.5	0.25	0.06	—
3676	MXF	0.25	8	50	0.125	0.06	0.03	0.125	0.125	0.03	4	4	16	2	8	Mutations GyrA (E85G), GyrB (S478I), and ParC (S79Y)
3676	CLR	4	32	44	0.125	0.016	0.016	0.125	0.06	0.03	16	16	1	0.5	0.25	—
3243	MXF	0.125	1	20	0.03	0.06	0.004	0.25	0.06	≤0.008	0.016	0.004	1	1	1	—
3243	AZM	0.03	0.25	33	0.5	0.125	0.03	0.5	0.125	0.125	0.5	0.5	0.5	0.25	0.25	L22 insertion 93SFRPRA94
3243	CLR	0.016	0.25	41	0.5	0.06	0.03	0.25	0.25	0.125	1	0.25	0.5	0.25	0.125	23S rRNA substitution A2059Tc
19	MXF	0.125	1	32	0.5	1	0.5	2	2	2	>128	>128	2	0.5	2	GyrA mutation: S81F
19	LVX	1	16	31	1	1	0.5	2	4	2	>128	>128	32	4	4	Mutations GyrA (S81Y) and ParC (D83Y)
24	MXF	0.25	4	50	0.125	1	0.5	2	2	1	>128	>128	32	4	4	Mutations GyrA (S81Y), GyrB (P413S) and ParC (S79F)
2527	AZM	0.016	4	20	0.5	2	0.5	16	16	2	4	16	32	8	4	—
37	GAT	1	16	14	2	1	0.5	2	4	2	>128	>128	64	16	16	Mutations GyrA (S81Y), GyrB (E474K), and ParC (S79Y)

^aCross-resistance is in boldface. Abbreviations: AMC, amoxicillin-clavulanic acid; CRO, ceftriaxone; CXM, cefuroxime; CDR, cefdinir; IPM, imipenem; AZM, azithromycin; CLR, clarithromycin; LVX, levofloxacin; GAT, gatifloxacin; MXF, moxifloxacin.

^b—, mechanism not detected.

^cEscherichia coli numbering.

Among the clones selected with LBM415, none showed cross-resistance (defined as a fourfold or greater increase in MIC). Cross-reactivity was observed only within the same group of antibiotics: quinolones (eight clones) and macrolide (four clones) (Table 2).

All resistant clones selected with quinolones (except clones 3243 and 3009, for which the moxifloxacin and gatifloxacin MICs, respectively, were increased) had changes in their QRDRs (Table 2). Among the resistant clones selected with macrolides, three had changes in 23S rRNA or the L22 protein (Table 2). Six of 11 clones selected with LBM415 had changes in the PDF protein (Table 1).

The results of the PAE studies are presented in Table 3. As can be seen, LBM415 had PAEs that ranged from 0.3 to 1.4 h. The LBM415 PAEs were not affected by the strains ' susceptibilities to β-lactams or macrolides. β-Lactam PAEs ranged from 0.2 to 3.2 h. The macrolide and the quinolone PAEs ranged from 1.8 to 6 h and from 0.5 to 3.8 h, respectively. Macrolides and quinolones had PAEs longer than those of LBM415 and β-lactams.

TABLE 3.

MICs and PAEs of nine S. pneumoniae strains

Drug	MIC range (μg/ml)	PAE (h)a
LBM415	0.25-2.0	1.0 (0.3-1.4)
Amoxicillin-clavulanate	0.03-4.0	1.0 (0.2-1.8)
Cefuroxime	0.03-8.0	1.5 (1.0-2.5)
Cefdinir	0.06-16	1.6 (1.0-3.2)
Ceftriaxone	0.016-4.0	1.3 (0.5-2.0)
Imipenem	0.008-0.25	1.7 (1.1-3.0)
Azithromycinb	0.03-4.0	3.1 (1.8-5.0)
Clarithromycinb	0.016-2.0	3.8 (2.3-6.0)
Levofloxacin	1.0	1.3 (0.8-2.2)
Gatifloxacin	0.25-0.5	1.7 (0.8-2.6)
Moxifloxacin	0.125-0.25	1.5 (0.5-3.8)

^aData represent the means (ranges) of two experiments. The PAE was induced by 1 h of exposure at 10× MIC.

^bThree erythromycin-sensitive and three erythromycin-resistant mef(A) strains were tested.

LBM415 yielded resistant clones, with defined and stepwise mutations in the PDF protein detected in most strains tested after 14 to 50 days. Among 11 clones that were selected with LBM415 and that had raised MICs, only 2 clones (Table 1) retained the raised MIC after 10 days of drug-free subculture, suggesting that this resistance that was selected may be unstable in the absence of the drug. LBM415 MICs of ≥16.0 and 4 μg/ml were associated with substitutions in the ₆₉GGVGLAAPQ₇₇ motif (G69V/R/C substitutions) (Table 1). The GXGXAAXQ motif is highly conserved among many bacterial species (10, 12) and is involved in the structural stability and catalytic mechanism of PDFs. Transfer of the defB gene with the substitutions G69V/R/C, Q172H, and/or P76R from strains 3009, 3243, and 37 to susceptible strain S. pneumoniae R6 resulted in increases of LBM415 MICs from 0.25 for strain R6 to 16, 8, and 4 μg/ml for the three strains, respectively.

A previous study has described that strain S. pneumoniae R6 expressing a mutated defB gene encoding a single substitution (Q172H) displayed reduced susceptibility to actinonin, a member of the PDF inhibitor drug class (10). In our study, a change within the zinc binding motif ₁₇₃HEIDH₁₇₇ resulted in MICs of 8 μg/ml, but transformation of defB with the substitution D176N (donor strains, 1076 and 1077) failed and no R6 mutant was selected, even on selection plates with 1 μg/ml of LBM415. Alterations G139F and R135C, near the ₁₂₈EGCLS₁₃₂ motif, which is essential for enzyme activity (10, 12), resulted in MICs of 2 to 4 μg/ml, but only mutation G139F (donor strain 24) was essential for selection of an R6 mutant with an MIC of 4 μg/ml.

Changes in PDF close to or within the ₇₀GXGXAAXQ₇₇ conserved motif were crucial to increasing the LBM415 resistance levels in the mutated clones. Also, the frequency of introduction of a mutated gene (defined as the number of transformants grown on selective medium to the number of clones grown on medium without a selective agent) with two substitutions (G69R and P76R) was 34 to 100 times higher than the frequency of introduction of other mutated forms of the defB gene tested. Mutations near or within the ₁₇₃HEIDH₁₇₇ and ₁₂₈EGCLS₁₃₂ motifs do not appear to have influenced the low level of resistance in the mutants tested. The mechanisms of resistance to LBM415 in the other three mutants with no alteration in the PDF protein remain unknown and require further investigation.

The mechanisms of resistance to the quinolones and macrolides (Table 2) have been described before (1, 2, 7, 11). However, substitutions S478I and P413S in GyrB have not, to our knowledge, been reported before.

LBM415 had short but detectable PAEs against all nine strains tested. The PAEs did not depend on the strains ' susceptibilities to other, unrelated agents. The PAEs of the macrolides and the quinolones were usually higher than those of LBM415 and all β-lactams tested.

Our results suggest that LBM415 might be of potential use for therapy (including potentially short-course regimens) for pneumococcal infections. This hypothesis will need to be clarified by pharmacokinetic/pharmacodynamic and animal experimentation studies.