Toxicity modulation, resistance enzyme evasion, and A-site X-ray structure of broad-spectrum antibacterial neomycin analogs

Article abstract of DOI:10.1021/cb5003416

Aminoglycoside antibiotics are pseudosaccharides decorated with ammonium groups that are critical for their potent broad-spectrum antibacterial activity. Despite over three decades of speculation whether or not modulation of pK_a is a viable str

Full text of DOI:10.1021/cb5003416

Articles
pubs.acs.org/acschemicalbiology
Toxicity Modulation, Resistance Enzyme Evasion, and A‑Site X‑ray
Structure of Broad-Spectrum Antibacterial Neomycin Analogs
Juan Pablo Maianti,† Hiroki Kanazawa,‡ Paola Dozzo,§ Rowena D. Matias,§ Lee Ann Feeney,§
Eliana S. Armstrong,§ Darin J. Hildebrandt,§ Timothy R. Kane,§ Micah J. Gliedt,§ Adam A. Goldblum,§
Martin S. Linsell,§ James B. Aggen,§ Jiro Kondo,‡ and Stephen Hanessian*,†
†Department of Chemistry, Universite
́ ́ ́ ́
de Montreal, CP 6128 Succ. Centre-Ville, Montreal, Quebec H3C3J7, Canada
‡Department of Materials and Life Sciences, Sophia University, 7-1 Kioi-cho, Chiyoda-ku, 102-8554 Tokyo, Japan
§Achaogen Inc., 7000 Shoreline Court, Suite 371, South San Francisco, California 94080, United States
S
* Supporting Information
ABSTRACT: Aminoglycoside antibiotics are pseudosaccharides decorated with
ammonium groups that are critical for their potent broad-spectrum antibacterial
activity. Despite over three decades of speculation whether or not modulation of
pKa is a viable strategy to curtail aminoglycoside kidney toxicity, there is a lack of
methods to systematically probe amine-RNA interactions and resultant
cytotoxicity trends. This study reports the first series of potent aminoglycoside
antibiotics harboring fluorinated N1-hydroxyaminobutyryl acyl (HABA)
appendages for which fluorine-RNA contacts are revealed through an X-ray
cocrystal structure within the RNA A-site. Cytotoxicity in kidney-derived cells
was significantly reduced for the derivative featuring our novel β,β-difluoro-
HABA group, which masks one net charge by lowering the pKa without
compromising antibacterial potency. This novel side-chain assists in evasion of
aminoglycoside-modifying enzymes, and it can be easily transferred to impart
these properties onto any number of novel analogs.
highly conserved A-site rRNA helix H44, at the core of the
Aminoglycosides (AGs) comprise a diverse group of potent
broad-spectrum natural and semisynthetic bactericidal anti-
biotics used in clinical practice worldwide.1 AGs exploit the
mRNA-tRNA decoding center of the bacterial 30S ribosomal
subunit.2−5 The binding of an AG to the A-site influences the
bacterial ribosome to indiscriminately accept mismatched
tRNAs while translating mRNA codons, which compromises
the accuracy of protein synthesis.2−6 Among the antibiotics that
target the ribosome, only AGs possess broad-spectrum
bactericidal activity against Gram-positive and Gram-negative
bacteria,7 and they are, in particular, highly effective against
pathogens that cause life-threatening infections such as the
ESKAPE group (Escherichia coli, Staphylococcus aureus,
Klebsiella sp., Acinetobacter sp., Pseudomonas aeruginosa, and
Enterobacteriaceae sp.).1,8 AG antibiotics are subdivided into
two classes based on the substitution pattern of the 2-
deoxystreptamine core ring B (Figure 1).1 The subclass of 4,5-
disubstituted AGs includes the natural products neomycin (1),
paromomycin (2), and butirosin (3).1 Neomycin (1) is used in
topical ointments,9 while its 6′-hydroxyl congener paromomy-
cin (2) is effective against leishmaniasis and dysenteric
amoebiosis.10 Butirosin (3) is a naturally occurring AG bearing
a N1-(S)-hydroxy-aminobutyric acyl group (L-HABA),11
a
potentiating side-chain that has found wide application in
Received: May 2, 2014
Accepted: July 1, 2014
Published: July 14, 2014
Figure 1. Representative AG antibiotics of 4,5- and 4,6-disubstited
subclasses, and N1-substituted analogs.
© 2014 American Chemical Society
2067
dx.doi.org/10.1021/cb5003416 | ACS Chem. Biol. 2014, 9, 2067−2073

ACS Chemical Biology
Articles
Scheme 1. Asymmetric Synthesis of β-Substituted HABA Derivatives and Synthesis of Tetradeoxy-Neomycin Analogs Featuring
N1−HABA Substituentsc Ranging Over 3 Orders of Magnitude in γ-N pKa Values29,30
semisynthetic AG antibiotics (Figure 1).1,12 The 4,6-disubsti-
tuted AG class includes clinically relevant natural products such
as the gentamicin complex (4), tobramycin, and semisynthetic
analogs such as amikacin (5), a kanamycin derivative with an
appended N1-L-HABA chain (Figure 1).1
antibacterial potency.1,12,22,23 Kidney toxicity correlates with
the number of protonated and nonprotonated amino groups on
the AG sugar scaffolds,24 yet these features are also a
requirement for potent antibacterial action.1,5 This prevailing
model raises the possibility that modulation of the net charge of
AG analogs at physiological pH could lead to improved
therapeutic indexes.1,22−24 However, despite many decades of
speculation, this idea remains untested because of our lack of
systematic methods to modulate AG net charge without
concomitantly disrupting key target rRNA interactions
(Supporting Information Figure S2).1,12 These observations
highlight the need to understand the relationship among
antibacterial potency, cellular toxicity, and functionally relevant
RNA contacts. Furthermore, structure−activity differentiation
between the eukaryotic and prokaryotic ribosomal A-sites may
be useful to curtail inner ear toxicity of AGs.1,25−28 Herein, we
report a generalizable strategy to shift the protonation state of
the most basic amino group of AG analogs over 3 orders of
magnitude of the pH scale (10.5 to 7.4), with broad
implications for the development of potent and potentially
less toxic next-generation AG antibiotics.
Selective pressure from medical and veterinary use of
antibiotics has led to the evolution and dispersion of numerous
aminoglycoside-modifying enzymes (AMEs) that undermine
the effectiveness of clinically deployed antibiotics (Supporting
Information Figure S1).13−16 These enzymes target amine and
hydroxyl groups in AGs and are accordingly classified N-
acetyltransferases (AAC), O-phosphotransferases (APH), and
O-adenyltransferases (ANT) (Supporting Information Table
S1).13−16 Other resistance mechanisms include efflux pumps
and rRNA methyltransferases (ArmA) that abrogate target
binding by all 4,6-disubstituted type AGs such as 4 and 5.1,17
Extensive efforts have been made over the years to advance to
the clinic novel AG antibiotics that evade AME resistance
mechanisms by semisynthetic deoxygenation of susceptible
functionalities, such as the 3′ and 4′-hydroxyl groups, and also
by appendage of the N1-L-HABA substituent that clashes with
the active site pockets of multiple AMEs (Supporting
Information Figure S1).18−21 Because clinical AG antibiotics
1−5 and congeners of the 4,6-disubstituted AG class do not
evade an important subset of AMEs (Supporting Information
Figure S1),13−16 our group has investigated a series of potent
broad-spectrum semisynthetic neomycin and paromomycin
analogs,18−21 for example the novel antibiotic 3′,4′,3‴,4‴-
tetradeoxy-neomycin 6 and its N1-L-HABA analog 7
(Scheme 1) capable of evading multiple AG-resistance
mechanisms and modifying enzymes (Table 1, Supporting
Information Figure S1).20
RESULTS AND DISCUSSION
■
We devised a divergent synthetic strategy to access a complete
series of enantiopure β-substituted D- and L-HABA derivatives
suitable for examination of potency and cytotoxicity trends.
Precedent routes for the synthesis of HABA groups were
unsuitable to access a difluoro substitution pattern,29 which is
required to lower the pKa of ammonium groups to the range of
physiological pH.30 Our strategy required functionalities and
protecting groups compatible with the fluorination reagent
diethylaminosulfur trifluoride (DAST),31 such as O-benzyl and
azide groups, while the carboxylic acid was latently masked as
an olefin until a later stage (Scheme 1). Our retrosynthetic
analysis (Supporting Information Figure S3) brought us to
A long-standing challenge for medicinal chemists is to
systematically improve the toxicity profiles of next-generation
AG antibiotics without compromising their broad-spectrum
2068
dx.doi.org/10.1021/cb5003416 | ACS Chem. Biol. 2014, 9, 2067−2073

ACS Chemical Biology
Articles
Table 1. Minimum Inhibitory Concentration (MIC, μg/mL) Against ESKAPE Strains Expressing AMEs
NeoB
(1)
Par
(2)
Gent
(4)
Amk
(5)
entry
MIC values (μg/mL) bacterium/enzyme
6
7
18
19
20
21
22
23
a
1
E. coli
S. aureus
2
0.5
0.5
1
4
2
0.5
0.5
0.25
2
2
2
0.5
0.5
0.5
0.5
1
0.5
0.5
0.5
0.5
0.5
2
0.5
0.5
0.5
0.5
1
1
1
4
1
2
b
2
0.5
0.5
0.5
0.5
0.5
2
0.5
0.5
0.5
1
0.5
1
c
3
K. pnuemoniae
2
0.5
2
0.5
2
4
A. baumannii
d
4
0.25
0.25
2
1
5
P. aeruginosa APH(3′)
32
>64
64
>64
>64
>64
8
>64
>64
>64
>64
>64
>64
>64
4
0.5
32
2
0.5
4
4
1
6
E. cloacae, APH(3′)-I, ANT(2′′)-I, AAC(6′)
E. coli APH(3′)-Ib
S. aureus APH(3′/5′′)-III
A. baumannii, AAC(3)-I, APH(3′)-VI, ANT(2′′)-I
S. aureus ANT(4′)-I
P. aeruginosa ANT(4′)-II
E. coli ANT(2′′)-I
P. aeruginosa AAC(6′)-II
S. aureus AAC(6′)/APH(2′′)
E. coli AAC(3)-IV
64
0.5
8
>64
2
4
32
4
16
7
0.25
0.5
>64
0.5
2
32
32
32
0.25
1
0.25
1
0.25
1
0.5
1
1
2
2
2
2
2
8
2
8
2
2
8
1
4
9
>64
64
32
4
2
1
2
2
2
8
10
11
12
13
14
15
16
0.5
1
1
2
1
1
4
1
2
1
1
2
1
2
16
8
2
64
2
2
2
2
2
8
>64
>64
8
32
4
2
1
1
2
1
2
16
16
8
>64
2
>64
16
64
2
>64
2
2
2
4
2
8
2
1
1
1
4
E. coli rRNA methyltransferase ArmA
1
4
>64
>64
1
1
0.5
1
1
2
4
a
b
c
d
Strains were obtained from ATCC: ATCC 25922, ATCC 29213, ATCC 10031, ATCC 27853. Abbreviations: NeoB, neomycin B (1); Par,
paromomycin (2); Gent, gentamicin-C complex (4); Amk, amikacin (5).
divinyl-carbinol (8) as an ideal precursor for divergent
asymmetric synthesis. In practice, 8 was desymmetrized using
the Sharpless epoxidation procedure with the ligand
(−)-diisopropyl-tartrate (DIPT),32 followed by benzylation to
yield enantiopure epoxide 9 (99% ee, chiral gas chromatog-
raphy). Similarly, we used (+)-DIPT to access an “unnatural”
series of D-HABA chains for comparative cytotoxicity studies
(Scheme 1). Subsequently, epoxide 9 was opened using
ammonium azide to afford 10 as a single diastereomer. The
azido alcohol intermediate 10 was divergently elaborated to
produce the β-substituted and β,β-disubstituted γ-azido-L-
HABA derivatives 12, 14, and 16 in 3−4 steps and 15−51%
overall yield (Scheme 1). The relative and absolute stereo-
chemistries for (R,R)-12 and the enantiomer (S,S)-ent-12 were
verified by X-ray crystallography (Supporting Information
Figure S4). To our knowledge, this is the first successful
fluorination strategy to access the challenging β,β-difluorinated
L-HABA motif.29
With a series of seven HABA derivatives in hand, comprising
N-Cbz-HABA and azide-protected natural and unnatural
enantiomers of 12, 14, and 16, we were poised for amide
coupling to a suitable AG scaffold. We selected tetradeoxy-
neomycin 6 selectively protected as intermediate 17, which we
previously reported in the synthesis of antibiotic 7.20 This
intermediate can be produced in four steps and in excellent
yield from neomycin (1), employing a diol-selective deoxyge-
nation procedure followed by selective N1 deprotection via a
base-labile N1,O6-oxazolidinone (Supporting Information
Figure S3).20 For the coupling reaction, the acyl chains were
activated as N-hydroxysuccinimide esters29 (Scheme 1). The
corresponding N1-acyl products were obtained in 52−91%
yield after chromatographic purification, then subjected to
global hydrogenolytic deprotection with concomitant reduction
of azide and olefins using Pearlman’s catalyst (20% Pd(OH)2/
C) in yields ranging from 52 to 87% (Scheme 1). The final
analogs 18 to 23 were purified by silica gel column
chromatography eluted with ammonia/MeOH/CHCl3,18−21
and they were evaluated using NMR spectroscopy (1H, 13C,
and 19F), and using Charged Aerosol Detection HPLC, which
indicated purities >90 to 95% (Supporting Information).
Potency and Evasion of Resistance Mechanisms. To
examine the effects of γ-N pKa modulation on antibacterial
potency we compared the minimum inhibitory concentration
(MIC) of clinical antibiotics with the novel analogs against a
broad panel of susceptible and resistant strains of ESKAPE
pathogens (Table 1). The resulting MIC values for the β-
substituted L-HABA analogs 18 to 20 against wild-type strains
were in the range 0.5−2.0 μg/mL (Table 1, entries 1−5),
essentially indistinguishable from the parent analog 7.
Remarkably, excellent antibacterial potency was also maintained
throughout the panel of AG-resistant strains (Table 1, entries
6−16, and Supporting Information Figure S1).
Analogs 18 to 20 display improved MICs values over
neomycin B (1), gentamicin (4), amikacin (5), and tetradeoxy-
neomycin (6) against the collection of ESKAPE strains (Table
1). In particular, the clinical AG antibiotics are ineffective
against corresponding subsets of strains expressing AMEs (MIC
ranges 16 to >64 μg/mL, Table 1, see Supporting Information
Figure S1 for summary).13−16 Comparing the MIC values for
the unsubstituted tetradeoxy-neomycin analog 6 (Scheme 1,
Table 1) reveals the broad-spectrum potentiation and enhanced
evasion of AMEs conferred by the combination of 3′,4′-
dideoxygenation and N1-acylation displayed by the analogs 7,
18, 19, and 20 (Table 1).
By contrast, the series of control D-HABA analogs (23−25)
produced relatively weaker antibacterial activities, displaying a
strong correlation with γ-N pKa (Table 1). For example, the
pair of monofluoro analogs 18 and 21 produced comparable
MIC values, whereas the difluoro congener 22 revealed a major
discrepancy in potency with analog 19. These data suggest both
D- or L-β,β-difluoro-HABA substituents effectively influence the
γ-amine net charge at physiological pH,30 a point at which
isosteric modifications become essential to maintain antibiotic
potency.
X-ray Cocrystal Structural Studies. The potentiating
effect of N1-HABA substituents on AG antibiotics has been
extensively documented but remains poorly understood.18,33
We speculated on the possibility that introduction of a threo-
fluorine atom on HABA groups (12 and 14) could potentially
alter the binding mode of these side-chains on our novel AG
analogs. To investigate these interactions we solved the X-ray
2069
dx.doi.org/10.1021/cb5003416 | ACS Chem. Biol. 2014, 9, 2067−2073

ACS Chemical Biology
Articles
Figure 2. (A) Oligonucleotide comprising two palindromic A-site models (boxes), key functional residues are color-coded, numbering is for E. coli
rRNA. (B) Co-crystal of 18 within the A-site RNA (see also Supporting Information Figure S5). (C) H-bond interactions of 18 with RNA (dashed
lines). The threo-fluorine of 18 displays interactions with the Hoogsteen faces of C1404 and C1496 (red dashes).
Figure 3. Conserved A-site RNA cocrystal interactions with the HABA groups of (A) analog 18, (B) amikacin (5),33 and (C) predicted placement of
the second fluorine of 19 (see Supporting Information Figures S6 and S7 for more comparisons).
cocrystal structure of 18 within the A-site RNA binding site
(Figure 2, Supporting Information Tables S2−3). We observed
the placement of the AG sugar backbone is essentially
superimposable to that of paromomycin within the 30S
ribosome X-ray structure (Supporting Information Figure S5)
indicating that the A-site model (Figure 2B) is faithful to the
larger ribosome structure.2 Furthermore, the placement of the
fluorinated N1-HABA group and the H-bond interaction
network (Figure 2C) are essentially identical to those observed
for amikacin (5) in the A-site (Figure 3A-B), as well as a doubly
functionalized paromomycin analog reported by our group
(Supporting Information Figure S6).18,33 Importantly, the
threo-fluorine atom is accommodated by making halogen-H-
bond contacts with the Hoogsteen faces of N4(C1404) and
N4(C1496). This model predicts that the second erythro-
fluorine substitution of 19 can be accommodated by pointing
away from the RNA (Figure 3C). A detailed study of the X-ray
structure (Supporting Information Figure S7) suggests that
only perfectly isosteric L-HABA appendages are able to preserve
all the rRNA interactions that endow optimum potency,
consistent with the comparable antibiotic activity of analogs 7,
18, 19, and 20 (Table 1). The novel β,β-difluoro-L-HABA chain
of 19 enables unprecedented modulation of the γ-N pKa,
effectively neutralizing its charge at physiological pH, and
displays excellent broad-spectrum potency by precluding
inactivation by AMEs as effectively as cationic HABA groups
(Table 1).
Modulation of Kidney Cell Toxicity. Next, we examined
the cytotoxic potential of our analogs by measuring induction
of cell death in the human kidney cell line HK2.34 Kidney cells
are the standard model for studying cytotoxicity mechanisms,
because they actively accumulate AGs and subsequently elicit
concentration-dependent activation of apoptosis through the
caspase-3/7 intrinsic pathway.24,35,36 The effective AG concen-
tration for 50% (EC50) activation of caspases 3/7 was measured
using a luminescence based assay (Figure 4).34 Our results for
both the L- and D-stereochemical series suggest that the HABA
analogs of highest γ-N pKa elicited apoptosis with an average
2070
dx.doi.org/10.1021/cb5003416 | ACS Chem. Biol. 2014, 9, 2067−2073

ACS Chemical Biology
Articles
Figure 4. Apoptosis assay of human kidney cell line HK2 exposed to β-substituted N1-HABA analogs of (A) natural L-series or (B) unnatural D-
series. The parent N1-HABA-tetradeoxy-neomycin 7 was used for comparison. Gentamicin (4) was included as assay control. Dotted lines indicate
the EC50 value of the parent analog 7 and the β,β-difluoro-N1−HABA analogs 19 and 22.
for ∼1 min, and dispensed into a clear-bottom 96-well polystyrene
tissue-culture plate (EK-25098, E&K Scientific) at a density of 1.6 ×
104 cells/well in a final volume of 100 μL. The plates were incubated at
37 °C with 5% CO2 for 3 days. Plates were removed from the
incubator and the media was removed using a multichannel aspirator.
Test compounds were serially diluted in KSFM medium containing 20
mM HEPES buffer. The compound solutions were prewarmed when
added to plates, and then returned to incubation at 37 °C for 31 h.
Blanks without cells and negative controls without compound were
included. A serial dilution curve for gentamicin was included as an
assay control.
For caspase-3/7-activation assays, following incubation, the plates
were equilibrated at RT for 30 min, and added 50 μL of Caspase-Glo
reagent (Promega) to each well. The plates were briefly agitated on a
shaker platform, and incubated at RT for 30 min. Luminescence was
measured using the luminometer Analyst HT (LJL Biosystems, 482
nm, 0.100 s integration). Area under the curve was calculated using the
trapezoid area equation.
EC50 = 26 μg/mL and were indistinguishable with the parent
analog 7 within assay error (EC50 = 28 μg/mL). By contrast,
the β,β-difluoro HABA analogs 19 and 22 required
approximately 2-fold higher concentrations to elicit apoptosis
(EC50 = 47 and 58 μg/mL, respectively). The disconnect
between MIC potency and apoptosis trends are consistent with
the reported mechanism of AG cytotoxicity that is independent
of cellular ribosomes.24,35,36 These results suggest that the novel
β,β-difluoro-L-HABA derivatives with reduced γ-N pKa were
better tolerated by kidney cells compared to the parent analog
7 or isosteric congeners of higher pKa (Figure 4). These
observations are also supported by comparing calculated area
under the curves (AUC) for the concentration-dependent
apoptotic responses (Supporting Information Figure S8), as
well as complementary cell viability assays (Supporting
Information Figure S9).
Conclusions. Taken together, the broad-spectrum anti-
bacterial in vitro activity coupled with the lower kidney cell
toxicity bodes well for neomycin analogs such as 19 as new
generation aminoglycosides with high therapeutic potential.
This study provides the first demonstration of a systematic
reduction in kidney cell toxicity using difluorination as a means
to modulate basicity in the AG antibiotics. The novel β,β-
difluoro-L-HABA side-chain is designed to maximize and
preserve stabilizing A-site interactions, endowing potent
broad-spectrum bactericidal activity against Gram-positive and
Gram-negative bacteria, including ESKAPE group pathogens
expressing multiple resistance mechanisms. Our findings
suggest the antibacterial potentiation and toxicity-lowering
features of the β,β-difluoro-L-HABA derivative may translate to
numerous semisynthetic analogs, which has implications for the
development of a safer next generation of AG antibiotics.
For cell viability assays, following incubation, the plates were
equilibrated at RT, and 100 μL of CellTiter-Glo reagent (Promega)
was added to each well. The plates were agitated on a shaker platform
for 2 min, incubated at RT for 10 min, and followed by luminescence
measurements.
Antibacterial Assays. MICs were determined using the standard
serial dilution method, as previously described.18−21 AME genes and
resistance mechanisms in the Achaogen Strain Collection have been
confirmed by PCR.18−21
Crystallization and Structure Determination. The A-site
model RNA oligomer was chemically synthesized (Dharmacon,
Boulder, CO), purified by 20% denaturing polyacrylamide gel
electrophoresis and desalted by reversed-phase chromatography.
Plate-shaped crystals of the RNA in complex with antibiotic 18 (A-
site•18) were obtained using condition described in Supporting
Information Table S2. An X-ray data set of A-site•18 was collected at
100 K with synchrotron radiation at the structural biology beamlines
BL-5A in the Photon Factory (Tsukuba, Japan). The data set was
processed with the program CrystalClear (Rigaku/MSC). The
obtained intensity data were further converted to structure-factor
amplitudes using TRUNCATE from the CCP4 suite. The statistics of
data collection and the crystal data are summarized in Supporting
Information Table S3. The initial phase of A-site•18 was solved with
the molecular replacement program AutoMR from the Phenix suite
using the coordinate of the bacterial A-site in complex with tobramycin
(PDB code: 1LC4). The molecular structure was constructed and
manipulated with the programs Coot. The atomic parameters of the
structure were refined with the program CNS through a combination
of simulated-annealing, crystallographic conjugate gradient minimiza-
METHODS
■
Extended experimental and chemical synthesis procedures are found in
the Supporting Information file.
Apoptosis and Viability Assays. Human kidney proximal tubule
HK-2 cells (ATCC CRL-2190) were cultured in keratinocyte serum-
free medium (KSFM) supplemented with 5 ng/mL epidermal growth
factor (EGF), and 50 μg/mL bovine pituitary extract (BPE). Cells
were maintained at subconfluence, and when they reached 80%
confluence, they were washed with Dulbecco’s PBS buffer
(Invitrogen), harvested by treatment with trypsin-EDTA (Invitrogen)
2071
dx.doi.org/10.1021/cb5003416 | ACS Chem. Biol. 2014, 9, 2067−2073

ACS Chemical Biology
Articles
tion refinements and B-factor refinements. The statistics of structure
refinement are summarized in Supporting Information Table S3.
Molecular drawings were made using PyMOL.
(11) Howells, J. D., Anderson, L. E., Coffey, G. L., Senos, G. D.,
Underhill, M. A., Vogler, D. L., and Ehrlich, J. (1972) Butirosin, a new
aminoglycosidic antibiotic complex: Bacterial origin and some
microbiological studies. Antimicrob. Agents Chemother. 2, 79−83.
(12) Price, K. E. (1986) Aminoglycoside research 1975−1985:
Prospects for development of improved agents. Antimicrob. Agents
Chemother. 29, 543−548.
(13) Shaw, K. J., Rather, P. N., Hare, R. S., and Miller, G. H. (1993)
Molecular genetics of aminoglycoside resistance genes and familial
relationships of the aminoglycoside-modifying enzymes. Microbiol. Rev.
57, 138−163.
(14) Mingeot-Leclercq, M. P., Glupczynski, Y., and Tulkens, P. M.
(1999) Aminoglycosides: Activity and resistance. Antimicrob. Agents
Chemother. 43, 727−737.
(15) Azucena, E., and Mobashery, S. (2001) Aminoglycoside-
modifying enzymes: Mechanisms of catalytic processes and inhibition.
Drug Resist. Updates 4, 106−117.
ASSOCIATED CONTENT
* Supporting Information
■
S
Figures S1−S8, Tables S1−S3, experimental procedures, X-ray
coordinate files, full spectroscopic data, GC and HPLC reports.
This material is available free of charge via the Internet at
http://pubs.acs.org.
Accession Codes
The atomic coordinates for the A-site•18 complex have been
deposited in the Protein Data Bank (PDB) with the ID code
3WRU.
AUTHOR INFORMATION
Corresponding Author
*Email: stephen.hanessian@umontreal.ca.
■
(16) Magnet, S., and Blanchard, J. S. (2005) Molecular insights into
aminoglycoside action and resistance. Chem. Rev. 105, 477−498.
(17) Paulsen, I. T., Brown, M. H., and Skurray, R. A. (1996) Proton-
dependent multidrug efflux systems. Microbiol. Rev. 60, 575−608.
(18) Kondo, J., Pachamuthu, K., Francois, B., Szychowski, J.,
Hanessian, S., and Westhof, E. (2007) Crystal structure of the
bacterial ribosomal decoding site complexed with a synthetic doubly
functionalized paromomycin derivative: A new specific binding mode
to an A-minor motif enhances in vitro antibacterial activity.
ChemMedChem 2, 1631−1638.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported by the Achaogen Research Chair and
the Natural Sciences and Engineering Research Council of
Canada (NSERC). J.P.M. received fellowship support from
(19) Hanessian, S., Pachamuthu, K., Szychowski, J., Giguer
Swayze, E. E., Migawa, M. T., Francois, B., Kondo, J., and Westhof, E.
(2010) Bioorg. Med. Chem. Lett. 20, 7097−7101.
(20) Hanessian, S., Giguere, A., Grzyb, J., Maianti, J. P., Saavedra, O.
̀
e, A.,
́ ́
Fonds de Recherche en Sante du Quebec (FRSQ). J.K. was
̧
supported by a Grant-in-Aid for Young Scientists (23790054)
from the Ministry of Education, Culture, Sports, Science and
Technology, Japan. We acknowledge the Photon Factory
synchrotron radiation facilities (No. 2012G519) and the BL-
̀
M., Aggen, J. B., Linsell, M. S., Goldblum, A. A., Hildebrandt, D. J.,
Kane, T. R., Dozzo, P., Gliedt, M. J., Matias, R. D., Feeney, L. A., and
Armstrong, E. S. (2011) Toward overcoming Staphylococcus aureus
aminoglycoside resistance mechanisms with a functionally designed
neomycin analogue. ACS Med. Chem. Lett. 2, 924−928.
̂
5A beamline staff. We thank Dr. Deschenes-Simard for X-ray
structure analysis of 12 and ent-12.
(21) Hanessian, S., Maianti, J. P., Matias, R. D., Feeney, L. A., and
Armstrong, E. S. (2011) Hybrid aminoglycoside antibiotics via Tsuji
palladium-catalyzed allylic deoxygenation. Org. Lett. 13, 6476−6479.
(22) Kahlmeter, G., and Dahlager, J. I. (1984) Aminoglycoside
toxicityA review of clinical studies published between 1975 and
1982. J. Antimicrob. Chemother. 13 (Suppl A), 9−22.
(23) Chen, L., Hainrichson, M., Bourdetsky, D., Mor, A., Yaron, S.,
and Baasov, T. (2008) Structure-toxicity relationship of aminoglyco-
sides: correlation of 2′-amine basicity with acute toxicity in pseudo-
disaccharide scaffolds. Bioorg. Med. Chem. 16, 8940−8951.
(24) Mingeot-Leclercq, M. P., and Tulkens, P. M. (1999)
Aminoglycosides: nephrotoxicity. Antimicrob. Agents Chemother. 43,
1003−1012.
(25) Nudelman, I., Rebibo-Sabbah, A., Cherniavsky, M., Belakhov, V.,
Hainrichson, M., Chen, F., Schacht, J., Pilch, D. S., Ben-Yosef, T., and
Baasov, T. (2009) Development of novel aminoglycoside (NB54) with
reduced toxicity and enhanced suppression of disease-causing
premature stop mutations. J. Med. Chem. 52, 2836−2845.
(26) Kondo, J. (2012) A structural basis for the antibiotic resistance
conferred by an A1408G mutation in 16S rRNA and for the
antiprotozoal activity of aminoglycosides. Angew. Chem., Int. Ed. Engl.
51, 465−468.
(27) Kondo, J., Koganei, M., Maianti, J. P., Ly, V. L., and Hanessian,
S. (2013) Crystal structures of a bioactive 6′-hydroxy variant of
sisomicin bound to the bacterial and protozoal ribosomal decoding
sites. ChemMedChem 8, 733−739.
REFERENCES
■
(1) Arya, D. P. (2007) Aminoglycoside Antibiotics: From Chemical
Biology to Drug Discovery, Wiley-Interscience, Hoboken, NJ.
(2) Carter, A. P., Clemons, W. M., Brodersen, D. E., Morgan-Warren,
R. J., Wimberly, B. T., and Ramakrishnan, V. (2000) Functional
insights from the structure of the 30S ribosomal subunit and its
interactions with antibiotics. Nature 407, 340−348.
(3) Wimberly, B. T., Brodersen, D. E., Clemons, W. M., Jr., Morgan-
Warren, R. J., Carter, A. P., Vonrhein, C., Hartsch, T., and
Ramakrishnan, V. (2000) Structure of the 30S ribosomal subunit.
Nature 407, 327−339.
(4) Ogle, J. M., and Ramakrishnan, V. (2005) Structural insights into
translational fidelity. Annu. Rev. Biochem. 74, 129−177.
(5) Francois, B., Russell, R. J., Murray, J. B., Aboul-ela, F., Masquida,
B., Vicens, Q., and Westhof, E. (2005) Crystal structures of complexes
between aminoglycosides and decoding A-site oligonucleotides: Role
of the number of rings and positive charges in the specific binding
leading to miscoding. Nucleic Acids Res. 33, 5677−5690.
(6) Wilson, D. N. (2009) The A−Z of bacterial translation inhibitors.
Crit. Rev. Biochem. Mol. Biol. 44, 393−433.
(7) Kohanski, M. A., Dwyer, D. J., and Collins, J. J. (2010) How
antibiotics kill bacteria: From targets to networks. Nat. Rev. Microbiol.
8, 423−435.
(8) Boucher, H. W., Talbot, G. H., Bradley, J. S., Edwards, J. E.,
Gilbert, D., Rice, L. B., Scheld, M., Spellberg, B., and Bartlett, J. (2009)
Bad bugs, no drugs: no ESKAPE! An update from the Infectious
Diseases Society of America. Clin. Infect. Dis. 48, 1−12.
(9) Waksman, S. A., Lechevalier, H. A., and Harris, D. A. (1949)
Neomycin: Production and antibiotic properties. J. Clin. Invest. 28,
934−939.
(28) Perez-Fernandez, D., Shcherbakov, D., Matt, T., Leong, N. C.,
Kudyba, I., Duscha, S., Boukari, H., Patak, R., Dubbaka, S. R., Lang, K.,
Meyer, M., Akbergenov, R., Freihofer, P., Vaddi, S., Thommes, P.,
Ramakrishnan, V., Vasella, A., and Bottger, E. C. (2014) 4′-O-
substitutions determine selectivity of aminoglycoside antibiotics. Nat.
Commun. 5, 3112.
(10) Davidson, R. N., den Boer, M., and Ritmeijer, K. (2009)
Paromomycin. Trans. R. Soc. Trop. Med. Hyg. 103, 653−660.
2072
dx.doi.org/10.1021/cb5003416 | ACS Chem. Biol. 2014, 9, 2067−2073

ACS Chemical Biology
Articles
(29) Takahashi, Y., Ueda, C., Tsuchiya, T., and Kobayashi, Y. (1993)
Study on fluorination-toxicity relationships. Syntheses of 1-N-
[(2R,3R)- and (2R,3S)-4-amino-3-fluoro-2-hydroxybutanoyl] deriva-
tives of kanamycins. Carbohydr. Res. 249, 57−76.
(30) Morgenthaler, M., Schweizer, E., Hoffmann-Roder, A., Benini,
F., Martin, R. E., Jaeschke, G., Wagner, B., Fischer, H., Bendels, S.,
Zimmerli, D., Schneider, J., Diederich, F., Kansy, M., and Muller, K.
(2007) Predicting and tuning physicochemical properties in lead
optimization: Amine basicities. ChemMedChem. 2, 1100−1115.
(31) Hudlicky, M. (1988) Fluorination with diethylaminosulfur
́
rrifluoride and related aminofluorosulfuranes. In Organic Reactions,
Vol. 35, p 513, John Wiley & Sons, Inc., New York.
(32) Jager, V., Schroter, D., and Koppenhoefer, B. (1991)
̈
̈
Asymmetric sharpless epoxidation of divinylcarbinol. Erythro-D- and
-L-4-pentenitols by hydrolysis of regioisomeric epoxy-4-pentenols.
Tetrahedron 47, 2195−2210.
(33) Kondo, J., Francois, B., Russell, R. J., Murray, J. B., and Westhof,
E. (2006) Crystal structure of the bacterial ribosomal decoding site
complexed with amikacin containing the γ-amino-α-hydroxybutyryl
(haba) group. Biochimie 88, 1027−1031.
(34) Armstrong, E. S., Feeney, L. A., Kostrub, C. F. (Achaogen Inc.)
Compositions and methods for determining nephrotoxicity, Sept. 3,
2009, US20090220982.
(35) Denamur, S., Van Bambeke, F., Mingeot-Leclercq, M. P., and
Tulkens, P. M. (2008) Apoptosis induced by aminoglycosides in LLC-
PK1 cells: Comparative study of neomycin, gentamicin, amikacin, and
isepamicin using electroporation. Antimicrob. Agents Chemother. 52,
2236−2238.
(36) Servais, H., Ortiz, A., Devuyst, O., Denamur, S., Tulkens, P. M.,
and Mingeot-Leclercq, M. P. (2008) Renal cell apoptosis induced by
nephrotoxic drugs: Cellular and molecular mechanisms and potential
approaches to modulation. Apoptosis 13, 11−32.
NOTE ADDED AFTER ASAP PUBLICATION
This paper was published ASAP on June 14, 2014 with errors in
Table 1. The corrected version was reposted on June 17, 2014.
■
2073
dx.doi.org/10.1021/cb5003416 | ACS Chem. Biol. 2014, 9, 2067−2073