In Vivo Efficacy of the Novel Aminoglycoside ACHN-490 in Murine Infection Models

Abstract

Aminoglycosides are broad-spectrum antibiotics with particular clinical utility against life-threatening infections. As resistance to antibiotics, including aminoglycosides, continues to grow, there is a need for new and effective antimicrobial agents. ACHN-490 is a novel aminoglycoside in clinical development with activity against multidrug-resistant Gram-negative and select Gram-positive pathogens. Here we assess the in vivo efficacy of ACHN-490 against a variety of common pathogens in two murine models: the septicemia and neutropenic thigh models. When its activity against a gentamicin-susceptible strain of Escherichia coli was tested in the septicemia model, ACHN-490 improved 7-day survival with a dose-response profile similar to that of gentamicin, with 100% survival seen at doses of 1.6 mg/kg of body weight and above. In animals infected with a gentamicin-susceptible strain of Pseudomonas aeruginosa, treatment with either ACHN-490 or gentamicin led to 100% survival at doses of 16 mg/kg and above in the septicemia model. ACHN-490 was also effective in the neutropenic thigh model, reducing multidrug-resistant Enterobacteriaceae family and methicillin-resistant Staphylococcus aureus strains, as well as broadly susceptible strains, to static levels with dose-dependent activity. Against gentamicin-sensitive Enterobacteriaceae and methicillin-resistant S. aureus, the efficacy of ACHN-490 was comparable to that of gentamicin. However, gentamicin-resistant Enterobacteriaceae strains and those harboring the Klebsiella pneumoniae carbapenemase responded to ACHN-490 but not gentamicin, with static doses ranging from 12 mg/kg to 64 mg/kg for ACHN-490. These results suggest that ACHN-490 has the potential to become a clinically useful agent against drug-resistant pathogens, including Enterobacteriaceae, P. aeruginosa, and methicillin-resistant S. aureus, and support further development of this promising novel aminoglycoside.

As antibiotic use has increased over the last half century, so has the ability of bacteria to evade the activity of antibiotics. Reports that penicillin use in humans resulted in the emergence of penicillin-resistant Staphylococcus aureus were already appearing in the medical literature in 1945, only 1 year after commercial production of penicillin began (28). Although the introduction of other classes of antibiotics has helped alleviate the problem, antibacterial resistance continues to grow rapidly as bacteria develop new mechanisms to render antibiotics ineffective and existing mechanisms of resistance spread. A novel antibiotic class active against Gram-negative bacteria has not been introduced for almost 50 years. This has led to an escalating need for new antibiotics; indeed, the Infectious Diseases Society of America has developed the 10×'20 Initiative to encourage the introduction of at least 10 new, safe, and effective antibacterial drugs by the year 2020 (13).

The first aminoglycoside, streptomycin, was introduced almost 70 years ago (16, 25). Aminoglycosides have broad-spectrum activity and have historically been useful in serious, life-threatening infections (3). As with other classes of antibiotics, many mechanisms of resistance to aminoglycosides have developed in once-susceptible pathogens. These primarily consist of the aminoglycoside-modifying enzymes (AMEs) N-acetyltransferases, AAC (N-acetylation), O-nucleotidyltransferases, ANT (O-adenylylation), O-phosphotransferases, and APH (O-phosphorylation) (15), which inactivate aminoglycosides by covalently modifying specific amino or hydroxyl moieties on the drugs. Because aminoglycosides are hydrophilic and thus require transport across cell membranes, bacteria that develop mechanisms to affect this process can also be resistant to aminoglycosides. Such mechanisms include the upregulation of efflux pumps and reductions in membrane permeability. Recently, methyltransferases that modify bacterial rRNA, the molecular target of aminoglycosides, and that confer high-level resistance to all widely used aminoglycosides have been shown to occur at a low incidence in clinical isolates (8).

ACHN-490 (Fig. 1) is a novel aminoglycoside in development by Achaogen, Inc. (South San Francisco, CA). ACHN-490 is derived from sisomicin in an 8-step process by which 2 of its amino groups are modified to maintain strong antibacterial potency and broad-spectrum activity while blocking modification by bacterial AMEs that confer resistance to other aminoglycosides. We showed previously that ACHN-490 is active in vitro against Gram-negative and Gram-positive pathogens and has potent activity against gentamicin-susceptible and -resistant bacteria (1). Members of the Enterobacteriaceae family and Acinetobacter, Pseudomonas, and Staphylococcus strains having a wide variety of single or multiple aminoglycoside resistance mechanisms were inhibited by ACHN-490 in vitro. Here we assess whether the in vitro activity of ACHN-490 translates to activity in vivo in murine infection models.

FIG. 1.

Structure of ACHN-490.

MATERIALS AND METHODS

ACHN-490 was provided by Achaogen, Inc. All aspects of this work, including housing, experimentation, and disposal of animals, were performed in accordance with the Guide for the Care and Use of Laboratory Animals (14). Animal manipulations and CFU determinations for the in vivo infection models were carried out at the contract research organization MDS Pharma Services (now Ricerca; Taipei, Taiwan), and all mice were obtained from Charles River Laboratories (Wilmington, MA). Achaogen provided non-American Type Culture Collection (ATCC) bacterial strains and determined all MICs using the broth microdilution method recommended by the Clinical and Laboratory Standards Institute (5). The MICs of ACHN-490 and comparator antibiotics against the bacterial strains used are presented in Table 1.

TABLE 1.

In vitro activities of ACHN-490 and comparator antibiotics against strains used in in vivo efficacy studies

Organism	Phenotype	Strain no.	MIC (mg/liter)a
			ACHN-490	AMK	ARB	CIP	GEN	IMP	VAN
Escherichia coli	Susceptible	ATCC 25922	1	4	2	0.008	0.5	0.125	ND
	MDR	AECO1003	1	32	8	>32	>64	0.125	ND
Klebsiella pneumoniae	Susceptible	ATCC 43816	0.5	2	ND	0.03	0.5	0.5	ND
	MDR	AKPN1073	0.5	>32	16	>16	>64	0.25	ND
	KPC	AKPN1109	0.25-0.5	ND	ND	2	64	32	ND
Pseudomonas aeruginosa	Susceptible	ATCC 27853	2	4	0.5	0.125-0.5	1	1	ND
Serratia marcescens	KPC	ASMA1030	1	ND	ND	>8	>64	>32	ND
Staphylococcus aureus	Methicillin resistant	ATCC 33591	4	16-32	4	0.5	2-4	>16	1

^aAMK, amikacin; ARB, arbekacin; CIP, ciprofloxacin; GEN, gentamicin; IMP, imipenem; VAN, vancomycin; ND, not determined.

Septicemia model.

CD-1 strain-derived male mice weighing 24 ± 2 g were inoculated with Escherichia coli ATCC 25922 or Pseudomonas aeruginosa ATCC 27853. Twofold the 90 to 100% lethal dose (LD_90-100; total inocula given, 4.5 × 10⁵ CFU/mouse for E. coli and 5.8 × 10⁴ CFU/mouse for P. aeruginosa) in 0.5 ml brain heart infusion broth containing 5% mucin was injected intraperitoneally to produce septicemia. At 1 h postinoculation, antibiotic or vehicle alone was administered by subcutaneous injection. Each treatment or control group had 10 mice. Mortality was recorded daily for 7 days postinfection. Nonlinear regression (GraphPad Prism software) was used to determine the 50% effective dose (ED₅₀) for each antibiotic from its dose-response.

Neutropenic thigh model.

Specific pathogen-free male CD-1 mice (6 mice per dosing group) weighing 24 ± 2 g were rendered neutropenic with 2 intraperitoneal injections of cyclophosphamide (150 mg/kg of body weight 4 days prior to bacterial inoculation and 100 mg/kg 1 day before inoculation) (31). Bacteria were injected into the right thigh of each mouse at time zero. Inocula were selected on the basis of pilot studies with vehicle-treated animals that determined the maximum number of CFU that could be inoculated without substantial mortality during the course of the experiment. Aminoglycoside-susceptible Enterobacteriaceae strains included Klebsiella pneumoniae ATCC 43816 (1.6 × 10³ CFU/mouse) and E. coli ATCC 25922 (1.8 × 10³ CFU/mouse). Multidrug-resistant (MDR) strains included K. pneumoniae AKPN1073 (1.3 × 10⁴ CFU/mouse) and E. coli AECO1003 (1.5 × 10³ CFU/mouse). Two strains expressing K. pneumoniae carbapenemase (KPC) were tested: K. pneumoniae AKPN1109 (8.3 × 10⁵ CFU/mouse) and Serratia marcescens ASMA1030 (3.2 × 10³ CFU/mouse). In addition, the model was also tested with a strain of methicillin-resistant S. aureus (MRSA; ATCC 33591; 1.2 × 10³ CFU/mouse). In vitro susceptibilities of the studied strains are presented in Table 1.

The antibiotic or vehicle alone was administered twice by subcutaneous injection at 2 and 14 h postinoculation. Doses are reported as total dose (mg/kg/day). To assess in vivo antibiotic efficacy, the infected thighs were removed, weighed, and homogenized with a ceramic mortar in 3 to 4 ml of phosphate-buffered saline (pH 7.4). This homogenate was serially diluted, plated on Mueller-Hinton broth (BD Biosciences, San Jose, CA) with 1.5% Bacto agar (BD Biosciences), and then quantified. The mean log number of CFU/g thigh at 26 h postinfection (24 h after the first antibiotic dose) was compared to the mean log number of CFU/g thigh for untreated mice assessed at 2 h postinfection (static level). Static doses were estimated using nonlinear parametric curve fitting (GraphPad Prism software).

RESULTS

Experimental septicemia.

Figure 2 shows the survival of mice with experimental septicemia. ACHN-490 demonstrated dose-dependent effects on survival in this model against both E. coli ATCC 25922 and P. aeruginosa ATCC 27853, with 100% survival over 7 days achieved in mice given the highest doses. Mice injected with vehicle alone showed 100% mortality in this model within 48 h postinfection. Treatment with ACHN-490 led to the survival of mice infected with E. coli ATCC 25922, with an ED₅₀ of 0.6 mg/kg (95% confidence interval [CI] = 0.5 to 0.8 mg/kg), compared to ED₅₀s of 0.7 mg/kg (95% CI = 0.5 to 0.9 mg/kg) for gentamicin and 2.5 mg/kg (95% CI = 1.9 to 3.1 mg/kg) for amikacin.

FIG. 2.

Efficacy of ACHN-490 and comparator aminoglycosides in septicemia model.

Septicemia caused by P. aeruginosa ATCC 27853 led to 100% mortality within 72 h in mice treated with vehicle alone. Amikacin increased survival up to 80% at the highest dose tested, with a calculated ED₅₀ of 22.4 mg/kg (95% CI = 15 to 39 mg/kg). ACHN-490 (ED₅₀ = 8.3 mg/kg [95% CI = 5.4 to 15 mg/kg]) and gentamicin (ED₅₀ = 5.2 mg/kg [95% CI = 3.9 to 7.1 mg/kg]) were more effective against P. aeruginosa than amikacin, achieving 100% survival and lower ED₅₀s.

Neutropenic thigh model.

Results for all compounds and strains tested in the mouse neutropenic thigh model are summarized in Table 2. ACHN-490 demonstrated dose-dependent efficacy in the neutropenic thigh model against the aminoglycoside-susceptible Enterobacteriaceae isolates, K. pneumoniae ATCC 43816, and E. coli ATCC 25922 (Fig. 3a and c). ACHN-490 doses of 16 or 64 mg/kg/day were sufficient to reduce the bacterial load to below the static level for both strains. The efficacy of ACHN-490 (static dose = 7.8 mg/kg/day for K. pneumoniae ATCC 43816, static dose = 10.6 mg/kg/day for E. coli ATCC 25922) was comparable to that of gentamicin and ciprofloxacin against these strains, whereas treatment with imipenem using the same dosing regimen was less effective against K. pneumoniae ATCC 43816 (static dose > 200 mg/kg/day). ACHN-490 was also efficacious against MDR Enterobacteriaceae, including K. pneumoniae AKPN1073 (static dose = 12 mg/kg/day) and E. coli AECO1003 (static dose = 25 mg/kg/day) (Fig. 3b and d). Against both organisms, ACHN-490 at 16 mg/kg/day reduced the bacterial load to near or slightly below the static level, while 64 mg/kg/day reduced the bacterial load to approximately 1 log below the static level. Gentamicin, on the other hand, was completely ineffective against K. pneumoniae AKPN1073; doses as high as 64 mg/kg/day failed to reduce the bacterial load relative to that in untreated animals. Likewise, gentamicin at doses of up to 64 mg/kg/day did not impede the growth of E. coli AECO1003. The highest dose of imipenem led to a bacterial load at or below the static level for both MDR strains tested.

TABLE 2.

Summary of estimated static doses of tested compounds in the mouse neutropenic thigh model

Compound	Estimated static dose (mg/kg/day)
	ATCC 25922	AECO1003	ATCC 43816	AKPN1073	AKPN1109	ASMA1030	ATCC 33591
ACHN-490	11	25	7.8	12	64	37	54
Gentamicin	7.8	>64	15	>64	>64	>64	52
Imipenem	<10	23	>200	97	>200	>200	NDa
Ciprofloxacin	<4	>64	22	>64	>64	>64	ND
Arbekacin	ND	ND	ND	ND	ND	ND	148
Daptomycin	ND	ND	ND	ND	ND	ND	<4
Vancomycin	ND	ND	ND	ND	ND	ND	20

^aND, not determined.

FIG. 3.

Efficacy of ACHN-490 and comparator antibiotics against Enterobacteriaceae in neutropenic thigh model.

Two KPC-positive (β-lactam-resistant) strains were tested, and imipenem was ineffective against these strains, as expected (Fig. 3e and f). Likewise, gentamicin and ciprofloxacin did not inhibit either strain (static dose > 64 mg/kg/day), an expected finding, since these strains also harbor AMEs and mutations in the quinolone resistance-determining region. ACHN-490, on the other hand, reduced the K. pneumoniae AKPN1109 bacterial load by 2 logs at the lowest dosage and by 3 logs at the highest dosage relative to those in tissues from untreated animals at 26 h postinfection (static dose = 64 mg/kg/day). Against S. marcescens ASMA1030, the highest dose of ACHN-490 reduced the bacterial load to below the static level (calculated static dose = 37 mg/kg/day).

ACHN-490 showed efficacy comparable to that of gentamicin in thighs infected with S. aureus ATCC 33591 (calculated static doses = 54 mg/kg/day for ACHN-490 and 52 mg/kg/day for gentamicin; Fig. 4). For ACHN-490, 64 mg/kg/day reduced the bacterial load to the static level, while a dose of 256 mg/kg/day reduced the bacterial load 2 logs below the static level, equal to the effect of the same dose of gentamicin and equal to the effect of treatment with daptomycin at 64 mg/kg/day.

FIG. 4.

Efficacy of ACHN-490 and comparator antibiotics against S. aureus in neutropenic thigh model.

DISCUSSION

The novel aminoglycoside ACHN-490 showed in vivo activity against common pathogens in two different standard infection models: experimental septicemia and the neutropenic thigh model. Bacterial strains against which ACHN-490 was active included P. aeruginosa, aminoglycoside-susceptible, carbapenemase-positive, and MDR strains of the Enterobacteriaceae, as well as MRSA. The in vivo efficacy of ACHN-490 was comparable to or better than that of legacy aminoglycosides in all cases, particularly in strains known to be aminoglycoside resistant (Tables 3 and 4 ). Against aminoglycoside-susceptible strains, for which the MICs were similar, the efficacy of ACHN-490 in the present study is as good as the efficacy shown by legacy aminoglycosides in comparable models (7, 11, 19, 20). In addition, these results confirm the results obtained in earlier experiments, in which ACHN-490 demonstrated potent in vitro activity against a wide variety of Gram-negative organisms and MRSA (1).

TABLE 3.

Relationship between in vitro activity and in vivo efficacy for ACHN-490 and gentamicin in the mouse septicemia model

Organisma	Strain no.	Gentamicin		ACHN-490
		ED50 (mg/kg)	ED50/MIC	ED50 (mg/kg)	ED50/MIC
Escherichia coli	ATCC 25922	0.7 (0.5-0.9)b	1.4	0.6 (0.5-0.8)	0.6
Pseudomonas aeruginosa	ATCC 27853	5.2 (3.9-7.1)	5.2	8.3 (5.4-15)	4.2

^aBoth isolates had the gentamicin-susceptible phenotype.

^bValues in parentheses are 95% CIs.

TABLE 4.

Relationship between in vitro activity and in vivo efficacy for ACHN-490 and gentamicin in the mouse neutropenic thigh model

Organism	Phenotype	Strain no.	Gentamicin		ACHN-490
			Static dose (mg/kg)	Static dose/MIC	Static dose (mg/kg)	Static dose/MIC
Escherichia coli	Susceptible	ATCC 25922	7.8 (5.9-14)a	16	11 (8.5-13)	11
	MDR	AECO1003	>64	NAb	25 (18-34)	25
Klebsiella pneumoniae	Susceptible	ATCC 43816	15 (13-17)	30	7.8 (6.2-9.7)	16
	MDR	AKPN1073	>64	NA	12 (8.4-16)	24
	KPC	AKPN1109	>64	NA	64 (23-130)	110
Serratia marcescens	KPC	ASMA1030	>64	NA	37 (31-44)	37
Staphylococcus aureus	Methicillin resistant	ATCC 33591	52 (41-67)	13	54 (44-56)	14

^aValues in parentheses are 95% CIs.

^bNA, not applicable.

As with all animal models, the results presented here do not ensure efficacy of ACHN-490 in clinical use, but the demonstration of efficacy, in particular against highly resistant clinical isolates, by ACHN-490 in these mouse infection models illustrates the potential utility of this novel aminoglycoside against serious infections (6). The mouse septicemia model and neutropenic thigh model presented here are broadly utilized and accepted to be reasonable models to measure the efficacy of antibiotics nonclinically against systemic infections (10, 12). The efficacy of ACHN-490 in these models is encouraging and supports the ongoing clinical evaluation of ACHN-490 injection for efficacy and safety in humans. In recent phase 1 trials with healthy volunteers, ACHN-490 injection was well tolerated at doses up to 15 mg/kg/day for 5 days, with no signs of nephrotoxicity, ototoxicity, or other drug-related serious adverse events.

The need for new antibiotics to combat antibacterial-resistant microbes is clear, given recent trends in the rates of resistance to broad-spectrum antibiotics. In the Meropenem Yearly Susceptibility Test Information Collection (MYSTIC) Program, between 9% and 13% of isolated E. coli, K. pneumoniae, and other Enterobacteriaceae isolates were resistant to tobramycin in 2008, with even greater rates of resistance to fluoroquinolones (between 13% and 32%) observed in the same isolates (24). Higher levels of resistance have been observed in other areas within Europe, Asia, and South America (22, 27, 29). In comparison, rates of resistance among all Enterobacteriaceae isolates collected in 2003 were 3.1% for tobramycin and between 7% and 9% for the fluoroquinolones, meaning that observed levels of resistance more than doubled in those 5 years. The emergence and spread of the newly recognized carbapenem-resistant Enterobacteriaceae have added further urgency to the situation, given that carbapenem antibiotics have been the last resort for treating infections caused by MDR Enterobacteriaceae (4, 26, 30). Infections caused by carbapenem-resistant K. pneumoniae have a particularly grim prognosis (23). Because many of the genes underlying carbapenem resistance are found on plasmids that carry additional resistance factors, extensively drug-resistant Enterobacteriaceae are likely to become more common as these plasmids spread. In the studies reported here, ACHN-490 demonstrated efficacy against MDR Enterobacteriaceae, including those harboring KPC, suggesting that this compound may have utility as empirical therapy for serious infections.

Inadequate empirical treatment can have catastrophic consequences for patients infected with an unexpected or resistant bacterial strain. In hospital-acquired pneumonia or ventilator-associated pneumonia, inadequate empirical treatment is associated with high costs, morbidity, and mortality (2, 9, 17, 18). In one study of adult trauma patients who experienced more than one ventilator-associated pneumonia episode for which inadequate empirical treatment was received, almost half of patients died (21). Clearly, the emergence and continued spread of antibiotic resistance require the introduction of new broad-spectrum antimicrobials that can improve outcomes, reduce the economic and medical burdens of MDR strains, and limit further development of resistance. Here we have shown that the novel aminoglycoside ACHN-490 demonstrates potent in vivo efficacy against a variety of drug-resistant and -susceptible pathogens in two different murine models. These promising results support the further evaluation of the safety and efficacy of this compound in clinical studies.