A proof of the specificity of kanamycin-ribosomal RNA interaction with designed synthetic analogs and the antibacterial activity

Article abstract of DOI:10.1016/j.bmcl.2005.02.043

The binding specificity of designed synthetic kanamycins with model RNA sequences (wild-type and point-mutated type) derived from the 16S ribosomal A-site was evaluated using surface plasmon resonance imaging. It was observed that kanamycins have nonspecific and multiple interactions with RNA hairpins and that the binding potency is not always proportional to the antimicrobial activity.

Full text of DOI:10.1016/j.bmcl.2005.02.043

Bioorganic & Medicinal Chemistry Letters 15 (2005) 2159–2162
A proof of the specificity of kanamycin-ribosomal RNA interaction
with designed synthetic analogs and the antibacterial activity
Yoshio Nishimura,a,* Hayamitsu Adachi,a Motoki Kyo,b Shoichi Murakami,c
Seiko Hattoria and Keiichi Ajitoc
aMicrobial Chemistry Research Center, 3-14-23, Kamiosaki, Shinagawa-ku, Tokyo, Japan
bToyobo Co. Ltd., 10-24, Toyocho, Tsuruga, Fukui, Japan
cMeiji Seika Kaisha, Ltd., 760, Morooka-cho, Kohoku-ku, Yokohama, Japan
Received 16 December 2004; accepted 4 February 2005
Available online 19 March 2005
Abstract—The binding specificity of designed synthetic kanamycins with model RNA sequences (wild-type and point-mutated type)
derived from the 16S ribosomal A-site was evaluated using surface plasmon resonance imaging. It was observed that kanamycins
have nonspecific and multiple interactions with RNA hairpins and that the binding potency is not always proportional to the anti-
microbial activity.
Ó 2005 Elsevier Ltd. All rights reserved.
Since streptomycin was discovered by Waksman and co-
workers in 1944,1 aminoglycoside antibiotics have been
recognized as a clinically important antimicrobial agent
with broad antibacterial activities.2 However, the emer-
gence of resistance3 and the toxicities such as oto- and
nephrotoxicity decrease the usefulness of this class of
antibiotics. To overcome resistance as well as side ef-
fects, an enormous number of synthetic analogs have
empirically been developed for five decades.4 However,
little is understood about the mode of action at
the molecular level until Moazed and Noller5 have re-
cently clarified the principles governing aminoglyco-
side–RNA recognition. At present it is believed that
aminoglycoside antibiotics cause the bactericidal action
by interaction with the A-site of the decoding region of
ribosomal RNA. Recent crystrallographic study has
clarified that most aminoglycosides of neomycin and
kanamycin families bind to the ribosomal decoding
Asite.6 The enzymatic N-acetylation, O-phosphoryl-
ation, and O-nucleotidylation of aminoglycosides by
resistant bacteria decrease drastically their affinity to
the RNA targets, and cause the inactivation of amino-
glycosides.7 Recently, Wong and Tor groups8,9 have
demonstrated the mode of action of aminoglycosides
on a molecular level. They have used surface plasmon
resonance (SPR) to study the interactions of a number
of structurally diverse aminoglycoside antibiotics by
use of commercial ones with RNAs. In contrast to other
methods, SPR is able to detect interactions in real time
without derivatization of aminoglycosides. It is also well
known that the slight structural changes effects antimi-
crobial activity of aminoglycoside antibiotics. Under
OH
H N
2
O
O
HO
H N
HO
RHN
2
HO
O
NH
2
O
NH
2
Dibekacin (1): R = H
DKB-G (2): R = -C(=NH)NH
2
s
Arbekacin (3): R = -C(=O)C(OH)CH CH NH
2
2
2
s
ABK-G (4): R = -C(=O)C(OH)CH CH NHC(=NH)NH
2
2
2
DKB-ButNH (5): R = -C(=O)CH CH CH NH
2
2
2
2
2
DKB-ButG (6): R = -C(=O)CH CH CH NHC(=NH)NH
2
2
2
2
s
DKB-Arg (7): R = -C(=O)CH(NH )CH CH CH NHC(=NH)NH
2
2
2
2
2
s
DKB-Arg(NO ) (8): R = -C(=O)CH(NH )CH CH CH NHC(=NH)NHNO
2
2
2
2
2
2
*
Figure 1. Commercialized dibekacin (1) and arbekacin (3) and their
derivatives (2, 4–8).
Corresponding author. Tel.: +81 3 3441 4173; fax: +81 3 3441
7589; e-mail: nishimuray@bikaken.or.jp
0960-894X/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.02.043

2160
Y. Nishimura et al. / Bioorg. Med. Chem. Lett. 15 (2005) 2159–2162
these circumstances, we are interested in the relationship
between RNA binding and antimicrobial activity among
slightly modified kanamycins to better understand. In
this study, an alternative SPR method was applied to
examine the binding and the dissociation conditions
between aminoglycoside antibiotics (kanamycins) and
RNA.10
It is well known that the amino groups of aminoglyco-
side antibiotics play critical role in antibacterial activ-
ity,11 and that acylation or alkylation of the 1-amino
group in 2-deoxystreptamine-contaning aminoglyco-
sides as shown in butirosins,12 amikacin,13 netilmicin,14
isepamicin,15 and arbekacin16 improve often activity
against sensitive and resistant bacteria. On the other
hand, the guanidine moiety, a strong basic function is
an important feature in streptidinecontaining aminogly-
cosides such as streptomycins.1 With the aim of studying
the effects of RNA–aminoglycoside interactions on the
antibacterial activities in more detail, the only one ami-
no group at C-1 position of dibekacin (DKB, 1)17 was
replaced by the various types of amino subsituents
(Fig. 1). DKB-G (2)18 with a guanidine group was ob-
tained upon treatment of 9 with N,N0-bis(tert-butyloxy-
carbonyl)thiourea in the presence of HgCl2 in DMF to
give 11 which was removed protecting groups (concd
NH4OH; 4 M HCl/dioxane). ABK-G (4)18 was prepared
by coupling of 10 with -(p-methoxybenzyloxycarbony)
OH
O
BocHN
HO
O
HO
R'HN
O
NHBoc
O
R"HN
NHBoc
9: R' = -C(=O)CF , R" = H
3
10: R' =Boc, R" = H
11: R' = -C(=O)CF , R" = -C(=NBoc)NHBoc
3
s
12: R' = Boc, R" = -C(=O)C(OH)CH CH NHPMZ
2
2
s
13: R' = Boc, R" = -C(=O)C(OH)CH CH NHPMZ
2
2
s
14: R' = Boc, R" = -C(=O)C(OH)CH CH NHC(=NBoc)NHBoc
2
2
15: R' = Boc, R" = -C(=O)CH CH CH NHZ
2
2
2
16: R' = Boc, R" = -C(=O)CH CH CH NHC(=NBoc)NHBoc
2
2
2
s
17: R' = Boc, R" = -C(=O)CH(NHBoc)CH CH NHC(=NH)NHNO
2
2
2
s
G
G
C
G
C
C
G
C
G
G
C
G
C
C
G
C
18: R' = Boc, R" = -C(=O)CH(NHBoc)CH CH NHC(=NH)NH
2
2
2
a
f
c
d
b
a
e
11
2
5
12
16
13
6
14
9
10
10
4
1406U A1495
1406U U1495
g
e
d
d,e
h
e
G
18
G
10
17
15
8
17
7
10
C
A
C
A
e
A
C
A
A
C
A
G
G
1410A U1490
1410A U1490
Scheme 1. Reagents and conditions: (a) (BocNH)2CS, HgCl2, Et3N,
DMF, rt (67%); (b) concd NH4OH/MeOH, rt; 4 M HCl/dioxane,
C
C
G
G
C
C
G
G
S
rt (74%); (c) HO2C C OHÞCH2CH2NHPMZ, BOP, Et3N, DMF, rt
(54%); (d) H2, Pd/C, MeOH/EtOAc, rt (81%); (e) 4 M HCl/dioxane
(98%); (f) HO2CCH2CH2CH2NHZ, BOP, Et3N, DMF, rt (53%);
U
G
U
G
U
C
U
C
(g) HO2CCH2CH2CH2ÆNHC(@NBoc)NHBoc, BOP, Et3N, DMF,
AS-U1495A
AS-wt
S
rt (59%); (h) HO2C C HðNHBocÞCH2CH2CH2NHCð@NHÞNHNO2,
BOP, Et3N, DMF, rt (70%).
Figure 2. Sequences of the RNA molecules used in this study.
Table 1. Antibacterial activity (MIC lg/ml) of dibekacin (1), arbekacin (3) and analogs (2, 4–8)
Test organisms
1
23
4
5
6
7
8
Staphylococcus aureus FDA209P
Micrococcus luteus FDA16
B. subtilis PCI219
0.20
6.25
0.20
1.56
0.78
0.78
25
0.78
>100
0.78
25
0.05
0.39
0.05
0.39
0.39
0.78
1.56
0.10
0.78
6.25
0.20
6.25
6.25
1.56
6.25
0.20
1.56
100
1.56
100
6.25
>100
3.13
25
12.5
>100
12.5
50
12.5
0.10
0.78
0.78
1.56
3.13
0.39
1.56
0.78
6.25
12.5
25
1.56
6.25
12.5
25
B. cereus ATCC10702
E. coli K-12
E. coli K-12 ML1410
50
100
50
100
>100
>100
3.13
>100
>100
25
>100
>100
25
E. coli JR66/W677
25
50
>100
100
50
Mycobacterium smegmatis ATCC607ꢀ
Klebsiella pneumoniae PCI 602
Serratia marcescens
1.56
1.56
12.5
0.39
>100
3.13
50
6.25
12.5
>100
3.13
>100
>100
>100
>100
>100
6.25
12.5
>100
6.25
>100
>100
>100
>100
>100
100
>100
100
25
>100
25
Pseudomonas aeruginosa A3
P. aeruginosa GN315a
P. aeruginosa 21–75a
S. aureus MRSA No.5a
S. aureus MRSA No.17a
S. aureus MS 16526 (MRSA)a
0.39
>100
12.5
1.56
25
>100
>100
>100
>100
>100
>100
>100
50
>100
>100
>100
>100
>100
>100
>100
>100
>100
12.5
25
MICs were determined by 2-fold agar dilution streak method at 37 °C for 18 and 42h. ꢀ
a Clinically isolated resistant strains.

Y. Nishimura et al. / Bioorg. Med. Chem. Lett. 15 (2005) 2159–2162
2161
amino-hydroxybutylic acid (BOP, Et3N, DMF) to give
12, followed by successive sequences of deprotection of
p-methoxybenzyloxycarbonyl group (H2, 10% Pd/C,
MeOH/AcOEt = 4:1;13), functionalization with a guani-
dine group [(BocNH)2CS, HgCl2, Et3N, DMF;14], and
removal of protecting groups (4 M HCl/dioxane).
DKB–ButNH2 (5)18 was generated by coupling of 10
with –(benzyloxycarbonyl)aminobutylic acid to give 15
which was removed protecting groups by the similar
methods mentioned above. DKB-ButG (6)18 was
similarly prepare via 16 from 10 by coupling with-
[N,N0-bis(tert-butyloxycarbonyl)guanidino]butylic acid
and removal protecting groups. DKB-Arg (7)18 was ob-
tained by coupling of 10 with -(tert-butyloxycarbonyl-
amino) nitro-L-arginine to give 17, followed by
reduction of a nitro group (H2, 10% Pd/C, AcOH,
MeOH/AcOEt = 1:1) to give 18 which was removed
Boc groups. DKB-Arg (NO2) (8)18 was also obtained
from 17 by removal of protecting groups. On the other
hand, DKB (1) and arbekacin (ABK, 3)16 are clinically
useful and commercially available antibiotics (Scheme 1).
Antimicrobial activities of thus obtained compounds (1–
8) against various kinds of bacteria containing resistant
ones were evaluated (Table 1). As shown in Table 1, the
order of activity is ABK (3) > ABK-G (4) > DKB (1) ꢀ
DKB-ButNH2 (5) = DKB-ButG (6) > DKB-G (2) ꢀ
DKB-Arg (7) = DKB-ArgNO2 (8). Of these com-
pounds, 3 shows the most potent activity against all of
the tested bacteria. Activities of 4 and 1 except for those
against the specific resistant bacteria are almost 1/2and
1/2–1/4 of those of 3, respectively. On the other hand, 5
14
14
12
10
8
ABKG
ABK
12
10
8
6
6
4
4
2
2
0
0
-2
-2
0
300
600
900
1200
0
300
600
900
1200
Time (sec)
Time (sec)
14
14
12
10
8
DKBArg
DKB
12
10
8
6
6
4
4
2
2
0
0
-2
-2
0
300
600
900
1200
0
300
600
900
1200
Time (sec)
Time (sec)
14
12
10
8
14
12
10
8
DKBArgNO2
DKBBuG
6
6
4
4
2
2
0
0
-2
-2
0
300
600
900
1200
0
300
600
900
1200
Time (sec)
Time (sec)
14
12
10
8
14
12
10
8
DKBBut
DKBG
6
6
4
4
2
2
0
0
-2
-2
0
300
600
900
1200
0
300
600
900
1200
Time (sec)
Time (sec)
Figure 3. SPR analyses by exposure of aminoglycoside solution on the RNA arrays: SPR signal changes in AS-wt (black line), AS-U1495A (dark
gray line) and Blank (light gray line).

2162
Y. Nishimura et al. / Bioorg. Med. Chem. Lett. 15 (2005) 2159–2162
Chemother. 1999, 43, 727; (d) Daigle, D. M.; Hughes, D.
W.; Wright, G. D. Chem. Biol. 1999, 6, 99.
and 6 show the very weak activities of 1/10–1/20 of those
of 3, and especially no activities against resistant Pseu-
domonas aeruginosa and MRSA. Compounds 2, 7, and
8 lost almost all the bactericidal activities. These results
indicate that the slight structural changes lead the big
alteration in the activity, and suggest more importantly
that the basicity of aminoglycosides is not straightly
proportional to the strong activity (see Fig. 1).
8. (a) Weizman, H.; Tor, Y. In Carbohydrate-based Drug
Discovery; Wong, C.-H., Ed.; Wiley-VCH, 2003; Vol. 2,
pp 661–668; (b) Luedtke, N. W.; Tor, Y. In Small
Molecule DNA and RNA Binders: From synthesis to
nucleic acid complexes; Demeunynck, M., Bailly, C.,
Wilson, D., Eds.; Wiley-VCH, 2003; pp 18–40; (c) Tor,
Y. ChemBioChem. 2003, 4, 998.
9. (a) Hendrix, M.; Scott Priestly, E.; Joyce, G. F.; Wong,
C.-H. J. Am. Chem. Soc. 1997, 119, 3641; (b) Sucheck, S.
J.; Wong, C.-H. Curr. Opin. Chem. Biol. 2000, 4, 678; (c)
Agnelli, F.; Sucheck, S. J.; Marby, K. A.; Rabuka, D.;
Yao, S. L.; Sears, P. S.; Liang, F.-S.; Wong, C.-H. Angew.
Chem., Int. Ed. 2004, 43, 1562.
10. (a) Kyo, M.; Yamamoto, T.; Motohashi, H.; Kamiya, T.;
Kuroita, T.; Tanaka, T.; Engel, J. D.; Kawakami, B.;
Yamamoto, M. Genes Cells 2004, 9, 153; (b) Nelson, B. P.;
Frutos, A. G.; Brockman, J. M.; Corn, R. M. Anal. Chem.
1999, 71, 3928; (c) Brockman, J. M.; Nelson, B. P.; Corn,
R. M. Annu. Rev. Chem. 2000, 51, 41.
11. Kondo, S.; Hotta, K. J. Infect. Chemother. 1999, 5, 1.
12. Woo, P. W. K.; Dion, H. W.; Bartz, Q. R. Tetrahedron
Lett. 1971, 2625.
13. Kawaguchi, H.; Naito, T.; Nakagawa, S.; Fujisawa, K.
J. Antibiot. 1972, 25, 695.
Next, we have examined the recognition of RNA kana-
mycin derivatives (1–8) and the RNA selectivity of these
derivatives. Binding of compounds (1-8) to the wild-type
(AS-wt)19 and point-mutated A-site (ASU1495)19 of the
ribosomal decoding region of RNA (Fig. 2) was evalu-
ated using SPR imaging technique.10,20 As shown in Fig-
ure 3, increasing a number of amino groups tends to
increase binding. These results indicate that the proton-
ated amino groups at physiological pH play a significant
role in RNA-aminoglycoside binding by electrostatic
interactions. However, there are no differences between
the wildtype and the point-mutated A-site in the binding
potency, suggesting that these binding fashion are non-
specific and that aminoglycosides may have multiple
interactions with RNA hairpins. The present results
show that the increasing amino groups may produce
the more nonspecific and multiple bindings to RNA.
Although permeability of the compounds into bacterial
cell-membrane is not clear at this stage, these bindings
seem to reflect not always strong activity. These facts
suggest that the antibiotic activity may be affected in
cooperation with possible tertiary interaction with ribo-
somal proteins that can not be accounted in this ribo-
some-free model system. While the importance of
electrostatic interactions of the aminoglycosides with
the RNA target is generally recognized,8,9,21 the observed
absence of binding specificity implies that the design of
novel aminoglycosides targeted to the defined ribosomal
decoding region A-site may need other approaches.
14. Wright, J. J. J. Chem. Soc., Chem. Commun. 1976,
206.
15. Nagabhushan, T. L.; Cooper, A. B.; Tsai, H.; Daniels,
P. J. L.; Miller, G. H. J. Antibiot. 1978, 31, 681.
16. Kondo, S.; Iinuma, K.; Yamamoto, H.; Maeda, K.;
Umezawa, H. J. Antibiot. 1973, 26, 412.
17. Umezawa, H.; Umezawa, S.; Tsuchiya, T.; Okazaki, Y. J.
Antibiot. 1971, 24, 485.
23
21
18. 2: ½aꢁD +57.3 (c 0.42, H2O) 4: ½aꢁD +62.2 (c 0.73, H2O); 5:
21
21
½aꢁD +77.7 (c 0.76, H2O); 6: ½aꢁD +74.6 (c 0.8,
21
21
H2O); 7: ½aꢁD +74.9 (c 0.85, H2O); 8: ½aꢁD +78.8 (c
0.85, H2O).
19. AS-wt and AS = U1495A RNAs were purchased from
Intergrated DNA Technologies, Inc., 1710 Commercial
Park Coralville, IA, U.S.A.
20. Fabrication of RNA arrays: RNA array was obtained by
the modified multi-step procedure, which was reported on
DNA array.9 A gold-coated chip (Toyobo), which has
amino groups in 96 areas with 500 lm square and poly
ethylene glycol (PEG) background was reacted for 60 min
with 300 lL of 10 mg/mL MAL-PEG12-NHS ester
(Quanta Biodesign) to create a maleimido-modified sur-
face. 10 nL drops of 10 lM thiol terminated RNA (IDT)
were delivered automatically on the maleimido surface
using an automated spotter (Toyobo), and the maleimido-
thiol reaction was carried out overnight. 300 lL of 10 mg/
mL PEG-thiol was reacted on the RNA array for 60 min
to block the unreacted maleimido group. Then, the surface
was rinsed with phosphate buffer and water. A phosphate
buffer (10 mM phosphate: pH 7.2and 150 mM NaCl) was
used for all reactions in array fabrications. SPR imaging
analysis: The RNA array was placed into SPR imaging
instrument (Toyobo). The aminoglycosides were applied
to the array surface with 100 lL/min in the running buffer
(10 mM HEPES: pH 7.4, 150 mM NaCl, 3.4 mM EDTA).
The array was exposed with 40 lM of aminoglycoside
solution for 10 min and rinsed with the running buffer for
10 min. The array was regenerated and used repeatedly.
The bound aminoglycosides were desorbed by exposure to
0.2X SSC/0.1% SDS solution for 5 min. The signal data
were collected with an analysis program (Toyobo). All
SPR experiments were performed at 30 °C.
References and notes
1. Schatz, A.; Bagie, E.; Waksman, S. A. Proc. Soc. Exp.
Biol. Med. 1944, 55, 66.
2. Aminoglycoside Antibiotics; Umezawa, H., Hooper, I. R.,
Eds.; Springer: New York, Heidelberg, 1982.
3. (a) Kotra, L. P.; Golemi, D.; Vakulenko, S.; Mobashery,
S. Chem. Ind. 2000, 10, 341; (b) Doi, Y.; Yokoyama, K.;
Yamane, K.; Wachino, J.; Shibata, N.; Yagi, T.; Shiba-
yama, K.; Kato, H.; Arakawa, Y. Antimicrob. Agents
Chemother. 2004, 48, 491; (c) Doi, Y.; Wachino, J.;
Yamane, K.; Shibata, N.; Yagi, T.; Shibayama, K.; Kato,
H.; Arakawa, Y. Antimicrob. Agents Chemother. 2004, 48,
2075; (d) Vogne, C.; Aires, J. R.; Bailly, C.; Hocquet, D.;
Plesiat, P. Antimicrob. Agents Chemother. 2004, 48, 1676.
4. Kumazawa, J.; Yagisawa, M. J. Infect. Chemother. 2002,
8, 125.
5. Moazed, D.; Noller, H. F. Nature 1987, 327, 389.
6. Carter, A. P.; Clemons, W. M.; Broderson, D. E.;
Morgan-Warren, R. J.; Wimberly, B. T.; Ramakrishnan,
V. Nature 2000, 407, 340.
7. (a) Wright, G. D.; Berghuis, A. M.; Mobashery, A. Adv.
Exp. Med. Biol. 1998, 456, 27; (b) Kondo, S.; Hotta, K. J.
Infect. Chemother. 1999, 5, 1; (c) Mingeot-Leclerco, M.-P.;
Glupczynski, Y.; Tulkens, P. M. Antimicrob. Agents
21. Herman, T. Biopolymers 2003, 70, 4.