In Vitro and Intracellular Activities of Peptide Deformylase Inhibitor GSK1322322 against Legionella pneumophila Isolates

Abstract

GSK1322322, a novel peptide deformylase inhibitor currently in development as an oral and intravenous agent for the treatment of hospitalized community-acquired bacterial pneumonia, showed poor in vitro activity against a panel of 50 Legionella pneumophila strains, with MICs ranging from 1 to 16 μg/ml and an MIC₉₀ of 16 μg/ml, but very potent intracellular activity, with the minimum extracellular concentrations capable of inhibiting intracellular proliferation (MIECs) ranging from 0.12 to 2 μg/ml and 98% of the strains being inhibited by concentrations of ≤1 μg/ml.

TEXT

The opportunistic Gram-negative bacterium Legionella pneumophila, which can grow free living or as a facultative intracellular organism in amoebae or human alveolar macrophages, is the major causative agent of Legionnaires' disease, a severe form of pneumonia (1). Although at least 15 serogroups have been identified, L. pneumophila serogroup 1 accounts for more than 80% of the cases worldwide (2). Except in the event of a Legionnaires' disease outbreak, L. pneumophila is treated empirically as part of the therapy used for hospitalized pneumonia, and given the severity of the disease, it is important to ensure that the antibacterial agent to be utilized shows good activity against this pathogen. GSK1322322 is a novel peptide deformylase (PDF) inhibitor with good safety and pharmacokinetic properties (3–5) and is currently in phase II development as an oral and intravenous agent for the treatment of hospitalized community-acquired bacterial pneumonia and acute bacterial skin and skin structure infections. GSK1322322 has demonstrated good antibacterial activity against other causative agents of bacterial pneumonia, such as Streptococcus pneumoniae, Staphylococcus aureus, and Haemophilus influenzae (6), so it was therefore important to investigate its potency against this atypical pathogen.

In order to assess the in vitro activity of GSK1322322, we performed susceptibility studies against 50 L. pneumophila strains: 25 from serogroup 1 and 5 each from serogroups 2, 3, 4, 5, and 6. L. pneumophila isolates were collected from 1992 to 2013, mostly from the human respiratory tract, and grown on buffered charcoal yeast extract (BCYE) agar to produce pure cultures. MIC endpoints were determined by broth microdilution according to Clinical and Laboratory Standards Institute (CLSI) guidelines (7). MIC plates, containing approximately 5 × 10⁵ CFU/ml in buffered yeast extract (BYE) broth (with Legionella BCYE growth supplement), were incubated at 35°C in aerobic conditions for 48 h. The MIC was defined as the lowest concentration of antimicrobial agent that completely inhibited visible growth after the appropriate incubation time.

GSK1322322 showed variable activities against this panel of 50 L. pneumophila strains, with MICs ranging from 1 to 16 μg/ml and MIC₅₀ and MIC₉₀ values of 8 and 16 μg/ml, respectively (Table 1). No significant differences were observed in the activities of GSK1322322 against strains from different serogroups; 8 μg/ml was the most frequent MIC in all sets (Table 1). Azithromycin and levofloxacin showed more potent extracellular activity, with MIC₅₀/MIC₉₀ values of 0.06/0.5 μg/ml and 0.008/0.016 μg/ml, respectively (Table 1). The results of this study indicate that GSK1322322 is substantially less active in vitro against L. pneumophila serogroups 1 to 6 than azithromycin or levofloxacin, the drugs normally used for the treatment of legionellosis. The GSK1322322 MIC for S. aureus ATCC 29213 tested in Mueller-Hinton broth was identical to or within a 2-fold dilution of the MIC obtained using BYE, indicating that there is no drug inactivation by the test medium.

TABLE 1.

Extracellular activities of GSK1322322, azithromycin, and levofloxacin against 50 L. pneumophila strains from serogroups 1 to 6

Antibiotic	L. pneumophila serogroup (no. of strains)	No. of strains inhibited by the indicated antibiotic at an MIC (μg/ml) ofa:
		≤0.004	0.008	0.016	0.03	0.06	0.12	0.25	0.5	1	2	4	8	16
GSK1322322	All (50)									1	9	6	28	6
	1 (25)									1	2	4	14	4
	2 (5)										3		2
	3 (5)												4	1
	4 (5)										1		3	1
	5 (5)										2	1	2
	6 (5)										1	1	3
Azithromycin	All (50)				12	16	14	3	5
	1 (25)				4	10	7		4
	2 (5)				3	1	1
	3 (5)				1		2	2
	4 (5)					1	3	1
	5 (5)				4				1
	6 (5)					4	1
Levofloxacin	All (50)	1	36	13
	1 (25)		15	10
	2 (5)		5
	3 (5)	1	4
	4 (5)		4	1
	5 (5)		3	2
	6 (5)		5

^aAll experiments were performed in duplicate or triplicate. MIC₅₀s are in italic type, and MIC₉₀s are in bold type.

As L. pneumophila is a facultative intracellular bacterium that multiplies within phagocytic cells, the ability of GSK1322322 to effectively inhibit growth of L. pneumophila in human monocytes was investigated. In these studies, human mononuclear cells (U-937) (8), cultured at a final concentration of approximately 10⁴ cells/well in Dulbecco's modified Eagle's medium (DMEM) containing 10% heat-inactivated fetal bovine serum, were infected in a suspension with 10⁴ to 10⁵ of each of the 50 L. pneumophila strains grown to logarithmic phase by incubating them together for 1 h in a shaking incubator. Infected cultures were transferred to flat-bottomed wells and maintained thereafter without shaking at 37°C in 5% CO₂ and 95% air for 7 days. After 24 h, the infected cultures were washed three times with drug-free broth to remove extracellular bacteria, and GSK1322322 (1 to 16 μg/ml), azithromycin (0.03 to 0.05 μg/ml), or erythromycin (0.06 to 1 μg/ml) was added to the wells at a concentration equal to the corresponding MIC-strain combination. At day 3, the cultures were washed, split into two groups, and incubated for an additional 4 days with or without antibiotic. The numbers of CFU were determined in duplicate using BCYE agar at time zero and then every 24 h until day 7. The results indicate that the MICs of GSK1322322 not only prevented the growth of L. pneumophila inside macrophages but also reduced intracellular bacterial counts by an average of 1.2 1og₁₀ with respect to the original inocula (Fig. 1A). Although some growth was observed after day 5, the bacterial counts never reached those present in the culture prior to the addition of the antibiotic. Azithromycin also reduced bacterial counts by an average of 1.5 log₁₀ by day 3, but significantly less reduction of growth was observed with erythromycin (Fig. 1A). Neither of the two macrolides prevented bacterial growth during the last 4 days of exposure, perhaps due to metabolism or changes in uptake by either the bacterium or the monocytes. None of the three drugs demonstrated a strong postantibiotic effect, and regrowth was observed in all cases after the removal of the antibiotic at day 3 (Fig. 1B), although exposure to GSK1322322 seemed to delay regrowth slightly more.

FIG 1.

Intracellular activities of GSK1322322, azithromycin, and erythromycin against L. pneumophila strains. At day 1, drug was added to the infected cultures at the appropriate concentration for each MIC-strain combination (GSK1322322 at 1 to 16 μg/ml, azithromycin at 0.03 to 0.05 μg/ml, and erythromycin at 0.06 to 1 μg/ml). At day 3, cultures were washed, split into two groups, and incubated until day 7 with (A) or without (B) antibiotic. Each point represents the mean of the results obtained from duplicate or triplicate experiments (and duplicate quantitative plating) with each of the 50 L. pneumophila strains. Black arrows indicate the points at which the drugs were added.

Given the clear intracellular activity demonstrated by GSK1322322, the minimum extracellular concentrations capable of inhibiting intracellular proliferation (MIECs) of these 50 L. pneumophila strains were determined in human mononuclear cells (U-937) (8, 9) using the conditions described above, except that in this case, the infected cultures were maintained without shaking for 5 days. After 24 h, extracellular bacteria were removed, and GSK1322322, azithromycin, or levofloxacin was added at fractions of their MIC to determine the precise MIEC. The MIEC was defined as the lowest extracellular concentration of compound capable of preventing L. pneumophila intracellular proliferation and was calculated at days 3 and 5 of antibiotic exposure.

Against the 50 strains of L. pneumophila with extracellular MICs ranging from 1 to 16 μg/ml, the GSK1322322 MIEC at day 3 ranged from 0.12 to 2 μg/ml, with 98% of the strains being inhibited by concentrations of ≤1 μg/ml (Table 2). MIECs of 2, 1, 0.5, 0.25, and 0.12 μg/ml, with a mean reduction of ≥73% in bacterial counts, were demonstrated against 1 (2%), 29 (58%), 9 (18%), 10 (20%), and 1 (2%) of the 50 strains, respectively (Table 2). In all cases, the compound was at least 4-fold more potent intracellularly than extracellularly. Most of the strains showed GSK1322322 MIECs 8-fold (35/50; 70%) or 16-fold (11/50; 22%) below their extracellular MIC, with the remaining 3 and 1 of the 50 strains demonstrating 4-fold and 32-fold differences between the two values, respectively (Table 2). No significant differences were observed between the serogroups tested, and in all cases, the MIECs were maintained at day 5 of exposure. GSK1322322 was clearly much more effective at inhibiting the proliferation of L. pneumophila within macrophages than in extracellular conditions. The azithromycin MIEC at day 3 of exposure was identical to its extracellular MIC in 94% of the strains (47/50) and was 2-fold lower in the other 3 strains (Table 2). The levofloxacin MIEC was 2-fold below its extracellular MIC in most of the strains (41/50; 82%), with 1/50 strains showing an MIEC 4-fold lower than the MIC and the other 7/50 strains (14%) demonstrating identical MIEC and MIC values (Table 2).

TABLE 2.

Intracellular activities of GSK1322322, azithromycin, and levofloxacin at day 3 of exposure against 50 L. pneumophila strains with different MICs

Antibiotic	No. of L. pneumophila strains	MIC (μg/ml)	No. of strains inhibited by the indicated antibiotic at an MIEC (μg/ml) ofa:
			0.002	0.004	0.08	0.016	0.03	0.06	0.12	0.25	0.5	1	2
GSK1322322	50	1–16							1	10	9	29	1
	6	16										5	1
	28	8								1	4	23
	6	4								2	3	1
	9	2								7	2
	1	1							1
Azithromycin	50	≤0.03–0.5					15	13	14	3	5
	5	0.5									5
	3	0.25								3
	14	0.12							14
	16	0.06					3	13
	12	0.03					12
Levofloxacin	50	≤0.004–0.016	1	32	16	1
	13	0.016		1	11	1
	36	0.008	1	30	5
	1	≤0.004		1

^aAll experiments were performed in duplicate or triplicate. MIC₅₀s are in italic type, and MIC₉₀s are in bold type.

This is not the first time that the extracellular activity of an antimicrobial agent against L. pneumophila did not correlate with its intracellular potency. β-Lactams, for example, are highly active in vitro but do not penetrate the intracellular compartment and are therefore inactive in vivo. On the other hand, tetracyclines have poor in vitro activity against L. pneumophila, as they are inactivated by iron in BYEα growth medium but are active in animal models of Legionnaires's disease and have been proven as an effective treatment in humans (10). However, in the case of PDF inhibitors, the difference in extracellular and intracellular activities may be due to a change in the primary target of the antibacterial agent when the bacterium grows inside macrophages.

L. pneumophila has three functional PDF enzymes that show different substrate specificities and catalytic efficiencies (11). Recently, it was shown that PdfA plays an essential role in the extracellular growth of L. pneumophila and that the activity against this enzyme is the main determinant of the compound's in vitro antibacterial activity. On the other hand, PdfC seems to play a critical role in L. pneumophila replication within macrophages and is indispensable for the bacterium to cause a productive infection in a guinea pig model of legionellosis (J. Huang, J. B. Wright, J.-D. Sauer, P. DeMarsh, G. S. Van Aller, J. West, T. S. Collingwood, K. Aubart, D. J. Holmes, and M. Zalacain, submitted for publication). Therefore, the antibacterial activity shown by GSK1322322 against intracellular L. pneumophila may be due, at least in part, to its potent inhibition of PdfC (G. Van Aller, unpublished results), although its accumulation inside macrophages may also contribute to it (12).

In conclusion, GSK1322322 demonstrated good intracellular activities against 50 L. pneumophila strains from serogroups 1 to 6, with an MIEC₉₀ of 1 μg/ml. In fact, its intracellular activities are similar to those demonstrated against other typical causative agents of pneumonia, which have GSK1322322 MIC₉₀s of 2 μg/ml (S. pneumoniae) or 4 μg/ml (S. aureus and H. influenzae) (6). Moreover, plasma and intrapulmonary pharmacokinetic studies with GSK1322322 dosed intravenously twice daily for 4 days showed that its time courses in epithelial lining fluid (ELF) and alveolar macrophages (AM) mirror the plasma concentration-time profiles. In fact, the area under the concentration-time curve from time 0 to τ (AUC_0–τ) ratios of ELF and AM to total plasma were 1.2 and 2.5, respectively (12). Therefore, GSK1322322 should be active against intracellular L. pneumophila at the concentrations necessary to achieve efficacy versus the other bacterial respiratory tract pathogens and may be an effective treatment for lower respiratory tract infections caused by this atypical pathogen.