High-throughput assay for quantification of aminoglycoside-Ribosome interaction

Article abstract of DOI:10.1246/cl.160508

The serious side effects of aminoglycosides and the spread of aminoglycoside-resistant strains have restricted the clinical application of aminoglycosides. A compound with the identical mode of action with aminoglycosides and with different molecular skeleton would be an alternative drug for aminoglycosides. In this study, we constructed an SPR-based high-throughput screening system for the discovery of such compounds.

Full text of DOI:10.1246/cl.160508

CL-160508
Received: May 20, 2016 | Accepted: June 6, 2016 | Web Released: June 17, 2016
High-throughput Assay for Quantification of Aminoglycoside-Ribosome Interaction
Shun Yamashita,1,2 Dominik Bergmann,3 Ayato Sato,1 Mika Nomoto,2 Yasuomi Tada,2
Hans-Ulrich Humpf,3 Kenichiro Itami,1,2 and Shinya Hagihara*1,2,4
1Institute of Transformative Bio-Molecules (WPI-ITbM), Nagoya University, Chikusa, Nagoya, Aichi 464-8602
2Graduate School of Science, Nagoya University, Chikusa, Nagoya, Aichi 464-8602
3Institute of Food Chemistry, Westfälische Wilhelms-Universität Münster, Corrensstr. 45, 48149 Münster, Germany
4JST, PRESTO, Chikusa, Nagoya, Aichi 464-8602
(E-mail: hagi@itbm.nagoya-u.ac.jp)
The serious side effects of aminoglycosides and the spread
of aminoglycoside-resistant strains have restricted the clinical
application of aminoglycosides. A compound with the identical
mode of action with aminoglycosides and with different molec-
ular skeleton would be an alternative drug for aminoglycosides.
In this study, we constructed an SPR-based high-throughput
screening system for the discovery of such compounds.
in residues A1408 and G1491 of 16S rRNA.6 The aminoglyco-
sides that are less active to mammals than bacteria can be used as
antibiotics.
The concerns with the clinical use of aminoglycosides are
nephrotoxicity and ototoxicity, which are the major reasons for
therapeutic limitation of aminoglycosides.7 Diverse mechanisms
are proposed for the exertion of side effects. For instance,
mutations in the A-site rRNA of the mitochondrial ribosome
induces hypersensitivity to aminoglycoside-derived side effects,
indicating that the binding of aminoglycosides to the mitoribo-
some can be a cause of the side effects.8 Despite their serious
side effects, aminoglycosides have commonly been used to treat
infectious diseases because the clinical efficacy of aminoglyco-
side is irreplaceable. In particular, aminoglycosides have been
used to treat chronic lung infections associated with cystic-
fibrosis.9
Another concern with aminoglycosides is the emergence of
drug-resistant strains. The major mechanism of resistance is
reduced drug uptake, accumulation in bacteria, and expression
of aminoglycoside-modifying enzymes, such as acetyltrans-
ferases, phosphotransferases, and adenyltransferases.10 To over-
come these issues, various aminoglycoside derivatives have
been synthesized and their antibiotic activities to the resistant
bacteria have been tested.11 For example, amikacin is a semi-
synthetic aminoglycoside, of which the 1II-position of kana-
mycin A is chemically modified, is used to treat multidrug-
resistant tuberculosis. However, the newly developed amino-
glycosides often cause new drug-resistant strains like other
antibiotics. Especially, multidrug-resistant strains are causing
serious problems for hospitalized patients.
One of the possible strategies to overcome the concerns with
aminoglycosides is to develop non-aminoglycoside compounds
that bind to the aminoglycoside-binding site in ribosome. Such
compounds should be potent antibiotic drugs against existing
aminoglycoside-resistant strains. In addition, a compound with
a higher selectivity to the bacterial A-site should have fewer
side effects than aminoglycosides. However, lack of an efficient
method for evaluating the ribosome-ligand interaction has
hampered the screening of such compounds. We herein describe
a high-throughput screening system for ligands that bind to the
16S-rRNA A-site.
Keywords: Aminoglycoside
| Antibiotics | Ribosome
Aminoglycosides are highly potent and broad-spectrum
antibiotics, which are active against most Gram-negative aerobic
bacteria such as Pseudomonas, Acinetobacter, and Enterobacter
species.1 The first aminoglycoside, streptomycin, was isolated
from Streptomyces griseus in 1944, which was used as the first
effective therapeutic agent for the treatment of tuberculosis.2 So
far, several aminoglycosides such as neomycin B, tobramycin,
gentamicin, and paromomycin (PAR), have been identified and
have been used to combat severe infectious diseases (Figure 1a).
Aminoglycosides bind to the decoding A-site, where the
ribosome accurately selects aminoacyl tRNA (Figure 1b).3 The
binding of aminoglycosides to helix 44 of 16S-rRNA forces
two adenine bases (A1492 and A1943) in the bulged-out
conformation, which plays a critical role for codon-anticodon
recognition.4 Consequently, aminoglycosides perturb the protein
synthesis by impairing the proofreading process, leading to the
production of nonfunctional proteins.5 The structure of the
ribosome differs between prokaryotes and eukaryotes, especially
HO
a)
H2
N
O
HO
HO
O
HO
H2N
HO
NH2
HO
H2N
O
OH
O
NH2
HO
NH2
HO
HO
O
NH2
HO
N
HO
O
NHR
1II
O
OH
NH2
NH2
O
HN
HO
OH
O
N
O
N
O
OHC
H2
N
O
OH
O
OH
OH
O
HO
OH
OH
OH
H2
NH2
OH
Streptomycin
Kanamycin A: R = H
Amikacin: R =
Paromomycin
NH2
O
b)
rRNA
Adenine
As shown in Figure 2a, we designed a competition assay for
the evaluation of ribosome-binding molecules based on surface
plasmon resonance (SPR). The injection of A-site model hairpin
RNA (Figure 2b) to the aminoglycoside-immobilized SPR
sensor should result in the increase of the response signal,
whereas the co-injection of a competitive ligand to the A-site
should impede the increase in response.3,12 Since the SPR
Aminoglycoside
3’
mRNA
5’
tRNA
Figure 1. Aminoglycoside antibiotics. a) Structures of aminoglyco-
side antibiotics. b) The mode of action for aminoglycosides. Amino-
glycosides binds to the A-site rRNA constituting decoding center
leading to the production of inaccurately structured proteins.
1048 | Chem. Lett. 2016, 45, 1048–1050 | doi:10.1246/cl.160508
© 2016 The Chemical Society of Japan

HO
HO
HO
HO
HO
HO
a)
b)
5'
G
O
O
O
3'
5'
3'
Binding response
Hairpin RNA
C
G
C
HO
HO
HO
NHR
NHBoc
NHBoc
Sample
RHN
BocHN
BocHN
O
G
C
G
U
C
C
G
C
U
G
A
G
C
G
C
U
G
A
O
O
HO
HO
HO
O
NHR
O
NHBoc
O
NHBoc
C
b
c
O
O
O
OH
OH
OH
G
U
Binding site
Hairpin RNA
RHN
OH
BocHN
O
BocHN
O
OH
OH
O
O
OH
OH
C
O
O
O
1408 A
1709 G
3IV
OH
3IV
NHR
NHBoc
NHBoc
A
A
C
A
C
C
G1491
C
U
A
C
U
A1823
S
O
U
G
G
G
A
U
G
G
2
R= H : Paromomycin (PAR)
R= Boc: 1
PAR-biotin
Hairpin RNA + sample
Time
OH
a
U
U
C
U
C
3
SPR sensor
E. coli
H. sapiens
HO
c)
d)
O
HO
HO
NHR
1.2
RHN
O
HO
O
NHR
mRNA
1
0.8
0.6
0.4
0.2
0
O
d
OH
RHN
O
OH
OH
O
O
IC50
NHR
HN
H
NH
H
S
PAR
0.97 nM
O
O
PAR
O
PAR–biotin 330 nM
N
N
S
2
H
H
3IV
R= Boc : 4
R= H : PAR-biotin
e
rRNA
–12
–9
–6
–3
log( [sample]/M )
Scheme 1. Preparation of paromomycin derivative PAR-biotin.
Reagents and conditions: a) Boc2O, Et3N, 1,4-dioxane/water, rt, 1 d,
58%; b) DIAD, PPh3, 4-nitrobenzoic acid, THF/toluene, 0 °C to rt,
3 h; 28% NH3 aq MeOH, rt, 5 h 75%; c) 4-mercaptophenylacetic acid,
NaH, DMF, 0 °C to rt, 65%; d) NHS, EDC¢HCl, DMF, rt, 30 min;
EZ-Link· Amine-PEG2-Biotin, DMF, rt, 15%; e) TFA/CH2Cl2, rt,
30 min., quant. Boc: tert-butoxycarbonyl; DIAD: diisopropyl azodi-
carboxylate; TFA: trifluoroacetic acid.
Figure 2. a) Illustration of competition assay for analyzing the
binding of ligand to A-site hairpin RNA. Sample binding can be
detected by the decrease in response. b) Secondary structures of
hairpin RNAs. Solid lines during the strands represent Watson-Crick
base pairs, while dashed lines represent mismatched base pairs. c) The
binding structure of paromomycin to bacterial ribosome (PDBid:
4V51). d) Translation inhibition activity of paromomycin derivatives
on E. coli ribosome. The results are the averages of at least three
independent experiments and the mean « s.d. is plotted.
a)
b)
1.2
1
1000 nM
300 nM
400
300
200
100
0
response is proportional to the change in mass concentration on
the surface, the competition assay should be more sensitive than
direct detection of small molecular ligand with immobilized
RNA. According to the crystal structure of paromomycin in
complex with bacterial ribosome, we assumed that modification
of the 3IV position of paromomycin has less effect on the
paromomycin-ribosome interaction compared to the other
positions (Figure 2c).3,12 Therefore, we introduced biotin to
the 3IV position of paromomycin. Synthesis of the compound
was conducted by following the previously reported method for
the modification of neomycin B (Scheme 1).13 The talo-epoxide
was synthesized by Mitsunobu reaction of Boc-protected
paromomycin and was subjected to an addition reaction with
4-mercaptophenylacetic acid. After the condensation with the
biotin containing an amino-linker, the compound was treated
under acidic conditions to afford the biotinylated paromomycin
(PAR-biotin). The inhibitory activity of PAR-biotin on trans-
lation was analyzed using bacterial ribosome. An mRNA-coding
firefly luciferase was translated in the presence of paromomycin
or PAR-biotin, and the luciferase expression was quantified by
a standard luciferase assay.14 The inhibitory activity of PAR-
biotin on the ribosome was considerably weaker than that of
paromomycin but still strong with an IC50 value below the
micromolar range (Figure 2d). These data indicate that PAR-
biotin retains strong affinity to the bacterial A-site.
100 nM
30 nM
0.8
0.6
0.4
0.2
0
10 nM
3 nM
Hairpin RNA
Kd
E. coli
37 nM
1 nM
H. sapiens 130 nM
0.3 nM
0
100
Time/s
200
300
0.1
1
10 100 1000 10000
[Hairpin RNA]/nM
IC50
c)
1.2
1
Competitor
0.8
0.6
0.4
0.2
0
Spectinomycin
Hygromycin B
Capreomycin
Kanamycin A
Gentamicin
>3 mM
1.1 mM
410 µM
200 µM
14 µM
Paromomycin
3.2 µM
0.1
1
10 100 1000 10000
Conc./µM
Figure 3. Evaluation of ligand binding to model hairpin RNA of
ribosome A-site. a) Concentration dependency of the response. The
binding response at 200 s was plotted against the concentration of
hairpin RNA. The dissociation constant of each RNA to paromomycin
was obtained by curve fitting. b) SPR sensorgram for the binding of
model hairpin RNA to the paromomycin-immobilized SPR sensor
surface. c) Results of competition assay for known A-site ligands as
well as aminoglycosides that do not bind to A-site helix 44 of 16S-
rRNA. Each compound (0-3 mM) was co-injected with hairpin RNA
(50 nM). The response at 200 s was plotted against the concentration of
hairpin RNA.
PAR-biotin was then immobilized to the streptavidin-coated
surface of an SPR sensor. A solution of hairpin RNA containing
the aminoglycoside-binding site in E. coli was injected to the
sensor. The response increased over 200 s and reached a plateau.
The increase of response was dependent on the concentration
of hairpin RNA, as shown in Figures 3a and 3b and the
dissociation constant was calculated to be 37 nM, which is
comparable to the previously reported value of 27 nM obtained
from isothermal titration calorimetry experiment.6a The binding
of human ribosome hairpin RNA was evaluated as well, and
turned out to be 130 nM, which is of the same order as the
previously reported value of 420 nM. These data suggest that
the SPR-based analysis of RNA-aminoglycoside interaction
Chem. Lett. 2016, 45, 1048–1050 | doi:10.1246/cl.160508
© 2016 The Chemical Society of Japan | 1049

correctly reflects the affinity in solution. The reduced activity of
PAR-biotin on translational inhibition may indicate that the
complex structure of ribosome hinders the binding of paromo-
mycin derivative containing a large substituent.
Ramakrishnan, Science 2006, 313, 1935.
4
5
a) T. Pape, W. Wintermeyer, M. V. Rodnina, Nat. Struct.
Mol. Biol. 2000, 7, 104. b) X. Zeng, J. Chugh, A. Casiano-
Negroni, H. M. Al-Hashimi, C. L. Brooks, III, J. Mol. Biol.
2014, 426, 3201.
a) J. Davies, L. Gorini, B. D. Davis, Mol. Pharmacol. 1965,
1, 93. b) A. Tsai, S. Uemura, M. Johansson, E. V. Puglisi,
R. A. Marshall, C. E. Aitken, J. Korlach, M. Ehrenberg, J. D.
Puglisi, Cell Rep. 2013, 3, 497.
a) M. Kaul, C. M. Barbieri, D. S. Pilch, J. Mol. Biol. 2005,
346, 119. b) M. I. Recht, S. Douthwaite, J. D. Puglisi, EMBO
J. 1999, 18, 3133.
C. A. Peloquin, S. E. Berning, A. T. Nitta, P. M. Simone, M.
Goble, G. A. Huitt, M. D. Iseman, J. L. Cook, D. Curran-
Everett, Clin. Infect. Dis. 2004, 38, 1538.
The competition assay using this system was demonstrated
with well-defined aminoglycosides. The hairpin RNA was co-
injected with spectinomycin, hygromycin B, kanamycin A,
gentamicin, or paromomycin and the response at 200 s was
plotted against the concentration of aminoglycosides (Figure 3c).
The known ligands to the A-site rRNA, such as paromomycin,
gentamicin, and kanamycin A, reduced the response in a
concentration-dependent manner. The non-aminoglycoside A-
site ligands, capreomycin, also reduced the response. In contrast,
spectinomycin and hygromycin B that bind to sites other than
the helix 44 of 16S-rRNA barely affected the response. These
data suggest that this system well-reflects the binding affinity of
ligands to the ribosome A-site.
6
7
8
9
S. N. Hobbie, S. Akshay, S. K. Kalapala, C. M. Bruell, D.
Shcherbakov, E. C. Böttger, Proc. Natl. Acad. Sci. U.S.A.
2008, 105, 20888.
A. Prayle, A. R. Smyth, Curr. Opin. Pulm. Med. 2010, 16,
604.
In summary, we have established an SPR-based method for
evaluating ribosome-ligand interaction. Our method is appli-
cable to the screening of A-site binding molecules, which may
enable the discovery of novel antibiotic agents.
10 a) D. N. Wilson, Nat. Rev. Microbiol. 2014, 12, 35.
b) M.-P. Mingeot-Leclercq, Y. Glupczynski, P. M. Tulkens,
Antimicrob. Agents Chemother. 1999, 43, 727. c) M. S.
Ramirez, M. E. Tolmasky, Sci. York 2010, 13, 151.
11 M. Y. Fosso, Y. Li, S. Garneau-Tsodikova, MedChemComm
2014, 5, 1075.
This work was supported by the PRESTO program from JST
(S.H.) and a Grant-in-Aid from JSPS (No. 15H05491 to S.H.)
and the German Research Foundation DFG (IRTG 1143
Münster-Nagoya). We thank Dr. Keiko Kuwata for HRMS
measurements. ITbM is supported by the World Premier
International Research Center (WPI) Initiative, Japan.
12 D. Fourmy, M. I. Recht, S. C. Blanchard, J. D. Puglisi,
Science 1996, 274, 1367.
13 S. Quader, S. E. Boyd, I. D. Jenkins, T. A. Houston, J. Org.
Chem. 2007, 72, 1962.
References
1
2
3
T. Hermann, Cell. Mol. Life Sci. 2007, 64, 1841.
14 a) S. Hagihara, S. Kusano, W.-C. Lin, X. Chao, T. Hori, S.
Imoto, F. Nagatsugi, Bioorg. Med. Chem. Lett. 2012, 22,
3870. b) S. Hagihara, W.-C. Lin, S. Kusano, X. Chao, T.
Hori, S. Imoto, F. Nagatsugi, ChemBioChem 2013, 14, 1427.
A. Schatz, S. A. Waksman, Exp. Biol. Med. 1944, 57, 244.
M. Selmer, C. M. Dunham, F. V. Murphy, IV, A.
Weixlbaumer, S. Petry, A. C. Kelley, J. R. Weir, V.
1050 | Chem. Lett. 2016, 45, 1048–1050 | doi:10.1246/cl.160508
© 2016 The Chemical Society of Japan