Synthesis and Spectrum of the Neoglycoside ACHN-490 James B. Aggen, Eliana S. Armstrong ...

29 Pages·2010·0.5 MB·English

AAC Accepts, published online ahead of print on 30 August 2010
Antimicrob. Agents Chemother. doi:10.1128/AAC.00572-10
Copyright © 2010, American Society for Microbiology and/or the Listed Authors/Institutions. All Rights Reserved.
1 Synthesis and Spectrum of the Neoglycoside ACHN-490
2
3 James B. Aggen, Eliana S. Armstrong, Adam A. Goldblum, Paola Dozzo, Martin S. Linsell,
4 Micah J. Gliedt, Darin J. Hildebrandt, Lee Ann Feeney, Aya Kubo, Rowena D. Matias, Sara
5 Lopez, Marcela Gomez, Kenneth B. Wlasichuk, Raymond Diokno, George H. Miller, Heinz E.
D
6 Moser* o
w
n
7 Achaogen, Inc., 7000 Shoreline Court, Suite 371, South San Francisco, CA 94080 lo
a
d
e
8 d

f
r
9 Heinz E. Moser om

h
10 Achaogen, Inc. t
t
p
:
/
11 7000 Shoreline Court, Suite 371 /a
a
c
12 South San Francisco, CA 94080 .a
s
m
13 Tel: [phone] .o
r
g
14 Fax: [phone] /
o
n

15 Email: [email protected] A
p
r
16 il 3
,

2
17 *Corresponding author 0
1
9

18 b
y

g
19 RUNNING TITLE: Synthesis and Spectrum of the Neoglycoside ACHN-490 u
e
s
t
20 ABSTRACT
21 ACHN-490 is a neoglycoside, or next-generation aminoglycoside (AG), that has been
22 identified as a potentially useful agent to combat drug-resistant bacteria emerging in hospitals
23 and health care facilities around the world. A focused medicinal chemistry campaign produced a
24 collection of over 400 sisomicin analogs from which ACHN-490 was selected. We tested
D
25 ACHN-490 against two panels of gram-negative and gram-positive pathogens, many of which o
w
n
26 harbored AG-resistance mechanisms. Unlike legacy AGs, ACHN-490 was active against strains lo
a
d
e
27 expressing known AG-modifying enzymes, including the three most common found in d

f
r
28 Enterobacteriaceae. ACHN-490 inhibited the growth of AG-resistant Enterobacteriaceae (MIC90 om

h
29 ≤4 µg/mL), with the exception of Proteus mirabilis and indole-positive Proteae (MIC =8 t
90 tp
:
/
30 µg/mL, and 16 µg/mL, respectively). ACHN-490 was more active in vitro against isolates of /a
a
c
31 Pseudomonas aeruginosa and Acinetobacter baumannii with AG-modifying enzymes alone than .a
s
m
32 against those with altered permeability/efflux. The MIC90 of ACHN-490 against AG-resistant .o
r
g
33 staphylococci was 2 µg/mL. Due to its promising in vitro and in vivo profile, ACHN-490 has /
o
n

34 been advanced into clinical development as a new antibacterial agent. A
p
r
il
3
,

2
0
1
9

b
y

g
u
e
s
t
2
35 INTRODUCTION
36 Aminoglycosides (AGs) are highly potent, broad-spectrum antibiotics used for treating
37 serious bacterial infections. They are a well-established class of antibacterials that act by binding
38 to the A-site of the bacterial ribosome and interfering with normal protein synthesis. They are
39 frequently used empirically for treating complicated urinary tract infections, nosocomial
D
40 respiratory tract infections, complicated intra-abdominal infections, and septicemia (4). AGs and o
w
n
41 β-lactam antibiotics are sometimes used in combination for treating suspected or confirmed lo
a
d
e
42 Pseudomonas aeruginosa infections, as well as empirically in endocarditis. d

f
r
o
43 As seen with other antibiotic classes, resistance to AGs has emerged in the 50 years since m

h
44 their introduction. For example, gentamicin (GEN) resistance (MIC >4 µg/mL) was seen in tt
p
:
/
/
45 31.4% of P. aeruginosa isolates from urinary tract infections collected in Europe and the a
a
c
.
46 Americas in 2000 (SENTRY Antimicrobial Surveillance Program) (12). Amikacin (AMK) a
s
m
47 resistance (MIC >16 µg/mL) was seen in 16.7% of the same isolates. Enterobacteriaceae have .o
r
g
/
48 also shown resistance to AMK (9). Among isolates collected from intra-abdominal infections o
n

A
49 worldwide in 2004 (Study for Monitoring Antibiotic Resistance Trends [SMART]), 9.0% of
p
r
il
50 extended spectrum β-lactamase–positive (ESBL+) Escherichia coli and 31.5% of ESBL+ 3
,

2
0
51 Klebsiella pneumoniae were resistant to AMK. Other AGs were not examined in this study. 1
9

b
52 At present, the most significant contribution to clinical resistance is represented by y

g
u
53 diverse AG-modifying enzymes (AMEs; Figure 1) that inactivate AGs by N-acetylation (AG e
s
t
54 acetyltransferases [AAC]), O-adenylylation (AG nucleotidyltransferases [ANT]), or O-
55 phosphorylation (AG phosphotransferases [APH]) (11). Numerous such enzymes have been
56 identified and often occur in combinations that can impart broad AG resistance (15). Less
57 common AG-resistance mechanisms (AGRMs) among the Enterobacteriaceae involve the
3
58 regulation of intracellular drug concentration by overexpression of efflux pumps and the
59 expression of enzymes that modify the ribosomal target (10). As broad-spectrum antibiotics –
60 including the existing AGs, cephalosporins, carbapenems, and fluoroquinolones – are rendered
61 ineffective by increasing resistance, new AGs are an attractive option for the treatment of serious
62 infections, especially if developed with optimized dosing regimens to maximize safety and
D
63 efficacy (7). o
w
n
64 In the present study, we modified sisomicin (SIS) in search of substituents that would lo
a
d
e
65 improve in vitro potency against AG-resistant (AG-R) bacterial strains. Among the SIS d

f
r
66 derivatives we synthesized, the novel compound ACHN-490 showed promise in a screening om

h
67 panel, and thus was tested against a larger panel that included 461 recent clinical isolates of t
t
p
:
/
68 gram-negative and gram-positive organisms. The majority of the pathogens in these panels /a
a
c
69 carried AGRMs, and many were also resistant to other classes of antibiotics. The broad spectrum .a
s
m
70 in vitro activity demonstrated in these experiments supports the continued development of .o
r
g
71 ACHN-490. /
o
n

72 A
p
r
73 (These data were presented at the 49th Interscience Conference on Antimicrobial Agents and il 3
,

2
74 Chemotherapy, San Francisco, CA, September 2009 [1].) 0
1
9

75 b
y

g
76 MATERIALS AND METHODS u
e
s
t
77 Synthesis of ACHN-490
78 SIS was used to derive over 400 neoglycoside compounds. The lead compound, ACHN-
79 490, was synthesized in an eight-step process as follows (Scheme 1). SIS sulfate (1) was
80 rendered basic by treatment with ion-exchange resin, followed by reaction with ethyl
4
81 trifluorothioacetate to selectively form the 6’-trifluoroacetamide (17). Treatment of the 6’-
82 trifluoroacetamide with Zn (II) acetate and Z-succinimide resulted in selective Z-blocking of
83 both the 2’- and 3’-positions to form intermediate 2 (14, 18).
84 Reaction of intermediate 2 with the active ester of N-Z-(S)-HABA (hydroxy-
85 aminobutyric acid) resulted in selective N-1 acylation (Scheme 2). Subsequent treatment with
D
86 concentrated ammonia liberated the 6’-amino functionality to produce compound 3. Reductive o
w
n
87 alkylation with O-benzoylglycolaldehyde appended the masked hydroxyethyl group to the 6’- lo
a
d
e
88 position to give compound 4. Finally, basic hydrolysis of the O-benzoyl group, followed by d

f
r
89 catalytic hydrogenolysis of the Z-groups and high performance liquid chromatography om

h
90 purification gave ACHN-490. Key structural assumptions were confirmed by liquid t
t
p
:
/
91 chromatography-mass spectrometry fragmentation and two-dimensional nuclear magnetic /a
a
c
92 resonance analysis. .a
s
m
93 .o
r
g
94 In vitro activity of ACHN-490 /
o
n

95 The antibacterial activity of ACHN-490 and comparator antibiotics was determined A
p
r
96 according to the broth microdilution method recommended by the Clinical and Laboratory il 3
,

2
97 Standards Institute (5). A screening panel composed of 25 strains with diverse AGRMs was used 0
1
9

98 to compare the activity of the SIS derivatives generated during the discovery program. This b
y

g
99 panel included strains with key AMEs found in both gram-negative bacteria and Staphylococcus u
e
s
t
100 aureus, as well as strains with changes in permeability or efflux, and a representative ribosomal
101 methyltransferase (armA).
102 Compounds that demonstrated increased activity against strains with AGRMs in the
103 screening panel were tested against a larger panel of 461 clinical isolates from diverse
5
104 geographic regions isolated in the years 2004-2006. The panel consisted primarily of gram-
105 negative bacteria, including multiple species of Enterobacteriaceae, P. aeruginosa, and
106 Acinetobacter spp. Gram-positive bacteria included S. aureus and Staphylococcus epidermidis.
107 The collection was segregated and analyzed according to the resistance status against
108 three AGs: AMK, GEN, and tobramycin (TOB). AG-susceptible (AG-S) isolates had AMK
D
109 MICs of ≤16 µg/mL and GEN and TOB MICs of ≤4 µg/mL. AG-R isolates were those with o
w
n
110 AMK MICs >16 µg/mL and/or GEN MICs >4 µg/mL and/or TOB MICs >4 µg/mL. To gauge lo
a
d
e
111 the extent of antibiotic resistance present in AG-R isolates, we used the broth microdilution d

f
r
o
112 method to determine MICs for a panel of antibiotics from other classes (5). These included the β- m

h
113 lactams imipenem (IMI) and ceftobiprole (BPR), the fluoroquinolone ciprofloxacin (CIP), and tt
p
:
/
/
114 the glycylcycline tigecycline (TGC). a
a
c
.
115 We then identified the specific AGRMs present in AG-R isolates in an effort to establish a
s
m
116 the mechanisms to which ACHN-490 is immune and also those responsible for any observed .o
r
g
/
117 decrease in ACHN-490 activity. The AMEs in the larger collection were initially assigned on the o
n

A
118 basis of antibiograms using standard AGs (21). Colony PCR was used to confirm the genes for
p
r
119 six of the most common AMEs in gram-negative bacteria and seven in gram-positive bacteria, il 3
,

2
120 using the primers listed in Table 1. 0
1
9

121 b
y

g
122 RESULTS u
e
s
t
123 The in vitro activity of ACHN-490 and comparator AGs against 25 strains having diverse
124 AGRMs is shown in Table 2. ACHN-490 showed potent in vitro activity (MIC ≤8 µg/mL)
125 against all but three strains tested; for 16 of 25 strains, the MIC was ≤2 µg/mL. ACHN-490 was
126 more potent than comparator AGs against E. coli, A. baumannii, A. calcoaceticus, and several
6
127 strains of P. aeruginosa and S. aureus carrying one or more AMEs. However, ACHN-490 was
128 not active against a strain expressing armA methylase nor against Providencia stuartii carrying
129 the AAC(2’)-I enzyme.
130 The activity of ACHN-490 was further assessed against a collection of 461 clinical
131 isolates consisting of representative nosocomial and community pathogens. ACHN-490
D
132 displayed potent activity against a broad range of gram-negative and Staphylococcus spp. o
w
n
133 isolates, including those showing AG resistance (Table 3). It is notable that many AG-R isolates lo
a
d
e
134 were also resistant to other classes of antibiotics in vitro (Table 4). Although ACHN-490 was d

f
r
135 broadly active against these bacteria, the modal MICs of the AG-R isolates of Acinetobacter spp. om

h
136 and P. aeruginosa were four- to eight-fold higher than the corresponding values for the AG-S t
t
p
:
/
137 populations of these species. No AMEs were identified that correlated with the increased MICs /a
a
c
138 observed in some isolates of these species. The presence of AMEs alone did not alter the activity .a
s
m
139 of ACHN-490 in several species backgrounds (Table 5). .o
r
g
140 /
o
n

141 DISCUSSION A
p
r
142 Compared to the activity of SIS, AMK, and GEN, the specific structural modifications il 3
,

2
143 incorporated in ACHN-490 resulted in equivalent or improved activity against bacterial strains 0
1
9

144 carrying common AGRMs. In Enterobacteriaceae, resistance to AGs is primarily caused by b
y

g
145 AMEs, either alone or in combination with other AMEs (16). These AMEs are diverse with u
e
s
t
146 respect to the positions on the AG scaffold that they attack and in the number of different genes
147 that encode them. The enzymes may be present on transferable elements or chromosomally, as in
148 the AAC(6’)-Ic enzyme found in Serratia marcescens (24). Many of these enzymes cause the
149 MICs of widely used AGs such as AMK and GEN to increase beyond a clinically useful level.
7
150 As shown in Tables 2 and 5, virtually all AMEs that modify AMK or GEN have no
151 impact on the activity of ACHN-490 across different species backgrounds. ACHN-490 and its
152 parent molecule, SIS, naturally lack the 3’ and 4’-OH groups which protects them from the
153 APH(3’) and ANT(4’) enzymes that generate resistance to AMK (Figure 1). The introduction of
154 the HABA substituent provides protection from the AAC(3), ANT(2”) and APH(2”) AMEs. The
D
155 hydroxyethyl substituent at the 6’ position blocks the multitude of AAC(6’) AMEs without o
w
n
156 reducing potency as had been found with other efforts to shield this position (17). The prevalence lo
a
d
e
157 of each AME varies geographically, but by far the three most prevalent enzymes that contribute d

f
r
158 to AG resistance in Enterobacteriaceae are AAC(3)-II, AAC(6’)-I and ANT(2”)-I (2, 3). As om

h
159 observed in the studies we report here, ACHN-490 remains highly potent against isolates with t
t
p
:
/
160 each of these three resistance mechanisms. /a
a
c
161 There are specific resistance mechanisms in Enterobacteriaceae that remained effective .a
s
m
162 against ACHN-490, but are of limited clinical significance. ACHN-490 was found to be inactive .o
r
g
163 (MIC >64 µg/mL) against a P. stuartii isolate with the AAC(2’)-I enzyme that also inactivates /
o
n

164 GEN. This is a chromosomal AME restricted to P. stuartii and is therefore unlikely to spread to A
p
r
165 organisms of broader clinical importance. The Proteae utilize AMEs to provide AG resistance il 3
,

2
166 but, in addition, have intrinsic outer membrane characteristics that form a barrier to AG 0
1
9

167 uptake.(23) Changes in membrane permeability that decrease AG uptake are only rarely b
y

g
168 observed in other Enterobacteriaceae. However, one such isolate was included in the expanded u
e
s
t
169 panel, and ACHN-490 and the two legacy AGs were similarly affected. One example of a
170 ribosomal methyltransferase, armA, was included in the screening panel, and was associated
171 with high MICs for ACHN-490 and each of the legacy AGs. ArmA methylates a key guanine at
172 the AG ribosomal target site and prevents binding of 4,6-disubstituted AGs. Although this
8
173 mechanism is very rare at the present time, it is being monitored closely for changes in
174 prevalence (6).
175 In staphylococci, there are fewer types of AMEs, but these tend to be widespread and
176 clonal (16). None of them altered the activity of ACHN-490 in our experiments, whereas the
177 MICs of the legacy AGs were increased. For ACHN-490, the MIC distribution against
D
178 Staphylococcus spp. isolates was found to be the same regardless of AG resistance status. We o
w
n
179 therefore concluded that the presence of resistance to AMK, GEN, and/or TOB had no impact on lo
a
d
e
180 the activity of ACHN-490 against clinical isolates in these species regardless of the type or d

f
r
181 number of AGRMs present. om

h
182 AG-R P. aeruginosa and Acinetobacter isolates were found to have higher ACHN-490 t
t
p
:
/
183 MICs than their AG-S counterparts. Increased efflux is a common mechanism of resistance to /a
a
c
184 the legacy AGs in P. aeruginosa, usually through the upregulation of the MexXY efflux pump .a
s
m
185 (22) and may have a similar impact on ACHN-490. MICs in A. baumannii isolates that appear to .o
r
g
186 have increased efflux based on their AG-R phenotype were similarly elevated for ACHN-490. It /
o
n

187 is important to note that efflux is rarely the sole AGRM present in either species (16). A
p
r
188 Underlying AMEs are frequently present in strains with increased efflux. It is clear from our il 3
,

2
189 results that the activity of ACHN-490 is not altered by the AMEs in P. aeruginosa and A. 0
1
9

190 baumannii even though the MICs of AMK and GEN are elevated in the presence of such AMEs. b
y

g
191 AGs are often used in combination with other agents in an effort to prevent the u
e
s
t
192 emergence of resistance, to treat polymicrobial infections empirically, and to benefit from
193 synergistic activity against organisms that are otherwise nonsusceptible (13, 20). We plan to
194 explore the in vitro synergy between ACHN-490 and other antipseudomonal agents against
195 clinical isolates of Pseudomonas. We anticipate that many of these isolates will have the MexXY
9
196 pump upregulated in addition to AMEs that cause resistance to existing AGs. The
197 imperviousness of ACHN-490 to the AMEs present in P. aeruginosa may serve to bolster its
198 utility in this setting.
199 The broader collection of clinical isolates was compiled in an effort to ascertain the
200 activity of ACHN-490 against the diversity of AGRMs present in multiple combinations and
D
201 genetic backgrounds. However, AGRMs are commonly found in combination with resistance o
w
n
202 mechanisms such as β-lactamases and gyrase mutations that generate resistance to other lo
a
d
e
203 antibiotic classes (8). This was evident from an analysis of the activity of comparator antibiotics d

f
r
204 against the AG-R collection. The MIC90 for CIP is above the breakpoint in each species group. om

h
205 Newer and investigational agents such as TGC and BPR display elevated MIC s against many t
90 tp
:
/
206 of the gram-negative organisms. IMI remains a reliable drug of last resort; however IMI activity /a
a
c
207 is diminished against some isolates of Klebsiella spp., P. aeruginosa, and Acinetobacter spp. .a
s
m
208 Given that this collection was assembled to represent the AG-R population, it is not surprising .o
r
g
209 that AMK and GEN show limited activity, but it is disturbing that the activity of unrelated drugs /
o
n

210 is so altered. This observation makes it all the more striking that ACHN-490 was not similarly A
p
r
211 affected. il 3
,

2
212 ACHN-490 has emerged as a promising new antibacterial agent with the potential to 0
1
9

213 rejuvenate the AG class of antibiotics, and thus has been advanced into clinical development. b
y

g
214 u
e
s
t
215 ACKNOWLEDGMENTS
216 All authors are employees of Achaogen, Inc. Melanie Watson, PhD, at AlphaBioCom, LLC
217 provided editorial assistance, which was funded by Achaogen, Inc.
10