Design and synthesis of novel ring-expanded arbekacin analogues

Article abstract of DOI:10.3987/COM-07-11081

Novel homologated aminoglycosides having a seven-membered ring were designed and synthesized by treatment of 5-keto arbekacin with diazomethane in CH₂Cl₂-Et₂O. The ring-expanded arbekacin analogue showed good antibacterial

Full text of DOI:10.3987/COM-07-11081

HETEROCYCLES, Vol. 71, No. 8, 2007
1715
HETEROCYCLES, Vol. 71, No. 8, 2007, pp. 1715 - 1721. © The Japan Institute of Heterocyclic Chemistry
Received, 16th April, 2007, Accepted, 18th May, 2007, Published online, 22nd May, 2007. COM-07-11081
DESIGN AND SYNTHESIS OF NOVEL RING-EXPANDED ARBEKACIN
ANALOGUES
Nobuto Minowa,a,* Yukiko Hiraiwa,a Yoshihisa Akiyama,a Kazunori
Maebashi,a Takayuki Usui,a and Daishiro Ikeda b,†
aPharmaceutical Research Center, Meiji Seika Kaisha, Ltd., 760 Morooka-cho,
b
Kohoku-ku, Yokohama, 222-8567, Japan; Institute of Microbial Chemistry,
3-14-23
Kamiosaki,
Shinagawa-ku,
Tokyo,
141-0021,
Japan.
e-mail: nobuto_minowa@meiji.co.jp
Abstract – Novel homologated aminoglycosides having a seven-membered ring
were designed and synthesized by treatment of 5-keto arbekacin with
diazomethane in CH2Cl2-Et2O. The ring-expanded arbekacin analogue showed
good antibacterial activity against Staphylococcus aureus and Escherichia coli
including aminoglycoside-resistant bacterial strain.
Aminoglycoside antibiotics are a clinically important class of natural products with antibacterial activity
against Gram-positive and Gram-negative bacteria.1 The aminoglycosides bind specifically to the A-site
of the decoding region of the 16S bacterial ribosome RNA (rRNA) and interfere with protein biosynthesis,
leading to bacterial cell death.2 The recent studies of X-ray crystal structures of several aminoglycosides
complexed to the rRNA suggest that the central 2-deoxystreptamine ring of aminoglycoside structures
play a crucial role in the interactions with 16S rRNA.3
Recently ring-expanded analogues of natural pyranose sugars have been reported as new efficient tools
for the investigation of protein-carbohydrate interactions.4 Further, the ring-expanded carbohydrate
analogues are useful for the development of glycosidase inhibitors and the investigation of glycosidase
mechanisms.4c Based on these findings, we have been interested in the ring-expansion of the
2-deoxystreptamine ring of aminoglycosides in our continuous search for novel aminoglycosides with
efficient antibacterial activity.5 However, to the best of our knowledge, ring-expanded aminoglycosides
have not been reported so far. Herein, we disclose the first example of the synthesis of
2-deoxystreptamine ring-expanded aminoglycosides and its antibacterial activity.
†Present Address: Microbial Chemistry Research Center, 18-24 Miyamoto, Numazu, Shizuoka, 410-0301, Japan.

1716
HETEROCYCLES, Vol. 71, No. 8, 2007
OR2
OBz
R2O
R1HN
O
O
MozHN
BzO
MozHN
NHR1
O
R1HN
O
NHMoz
HO
HN
O
R2O
R2O
c
BzO
O
NHR1
O
O
O
NHMoz
HN
O
BzO
O
3
R1HN
MozHN
arbekacin (1): R1 = R2 =H
2: R1 = Moz, R2 = PhCO
a, b
OH
O
OBz
O
H2N
HO
H2N
MozHN
BzO
O
O
NH2
MozHN
BzO
NHMoz
HO
O
HN
O
HO
d
e or f, g
or h
O
O
NH2
O
NHMoz
HO
HN
O
BzO
X
O
4
5: X = OMe
6: X = OH
H2N
MozHN
7: X = NMe2
Scheme 1. Reagents and conditions: a) Moz-S, Et3N, H2O, IPA, THF, 60 ℃, 79%; b) BzCl, Py, rt, 97%;
c) DMSO, Ac2O, rt, 99%; d) excess CH2N2, 1:1 MeOH/Et2O, 0 ℃, 86%; e) (1) 1M NaOMe, MeOH, rt;
(2) 90%TFA, 5 41%; f) 1N NaOH, MeOH, CHCl3, rt, 73%; g) (1) CsOAc, DMF, 80 ℃; (2) 1N NaOH,
H2O, rt; (3) 90% TFA, rt, 6 76%; (h) (1) excess Me2NH, H2O, EtOH, 80 ℃; (2) 1M NaOMe, MeOH, rt;
(3) 90% TFA, 7 79%.
We focused on arbekacin6 (ABK, 1) as a starting material and attempted expansion of the
2-deoxystreptamine ring of 5-keto ABK 3 by treatment with diazomethane.7 Compound 3, having a
5-keto group, was prepared from 1 in three steps (Scheme 1). p-Methoxybenzyloxycarbonyl8 (Moz)
protection
of
the
amino
groups
of
1
with
S-p-methoxybenzyloxycarbonyl-4,6-dimethyl-2-mercaptopyrimidine (Moz-S)9 followed by treatment of
the resulting penta-N-Moz arbekacin with benzoyl chloride (BzCl) afforded 2, which was treated with
DMSO-Ac2O10 to provide the 5-keto 3. The first attempt to expand the deoxystreptamine ring of 3 by
treatment with diazomethane7 in MeOH-Et2O11 provided only the undesired epoxide 4 in good yield as a
single stereoisomer; no ring-expanded product was obtained (Scheme 1). We assume that the selective
formation of epoxide in this reaction is attributable to the pyranosyl substituents at positions C4 and C6 of
OH
O
H N
HO
H N
2
O
2
H
H
OH
NH
2
4
HO
5 H
6
O
O
2
HO
HN
NH
2
1
H
7
3
H
H
O
O
8
H N
2
Figure 1. Selective ROESY correlations of 5.

HETEROCYCLES, Vol. 71, No. 8, 2007
1717
12
ketone 3. To confirm the configuration of 4 at position C5, the epoxide was converted into the
ring-opened derivative. Thus, treatment of 4 with 1M NaOMe followed by deprotection of the Moz
groups with trifluoroacetic acid (TFA) afforded 5-methoxymethyl ABK 5.13 The results of ROE
experiments on 5 indicated that the 5-OH of 5 was equatorial in orientation and the 5-methoxymethyl
group was axial in configuration (Figure 1). Therefore, the configuration of 4 at position C5 was
determined to be as illustrated in Scheme 1. Presumably, nucleophilic attack of diazomethane on the
ketone at position C5 would take place stereoselectively from the less hindered α face to afford 4. Further,
we attempted several ring opening reactions to investigate the reactivity of the epoxide 4. Deprotection of
the Bz groups of 4 with 1N NaOH followed by treatment with CsOAc at 80 ℃ in DMF gave
penta-N-Moz-5-epiacetoxymethyl ABK. Then, stepwise deprotections of the Ac and Moz groups gave the
5-hydroxymethyl ABK 6.13 While, cleavage of epoxy group of 4 with excess dimethylamine at 80 ℃
followed by stepwise deprotections of Bz and Moz groups provided the 5-dimethylaminomethyl ABK
7.13 These results suggested that the epoxide 4 would be a useful intermediate for preparation of various
5α-substituted aminoglycoside derivatives.
OBz
O
MozHN
BzO
MozHN
O
NHMoz
O
BzO
a
O
O
NHMoz
HN
O
BzO
MozHN
3
OBz
O
OBz
O
MozHN
O
BzO
MozHN
BzO
MozHN
MozHN
O
O
NHMoz
NHMoz
O
BzO
BzO
BzO
BzO
O
O
O
O
HN
O
NHMoz
+
HN
O
NHMoz
+
4
MozHN
MozHN
9
8
OH
OH
O
O
OMe
11
H2N
H2N
N
HO
HO
O
O
MeO
10
H2N
NH2
H2N
NH2
N
b, c
HO
HO
HO
HO
O
O
O
8 + 9
O
+
HN
O
HN
O
NH2
NH2
H2N
H2N
Scheme 2. Reagents and conditions: a) excess CH2N2, 1:1 CH2Cl2/Et2O, rt, 8 + 9 36%, 4 40%; b)
MeONH2 ·HCl, Py, 60 ℃, a mixture of oximes 74%; c) (1) 1M NaOCH3, MeOH, CHCl3, rt; (2) 90% TFA,
rt, 10 65%, 11 16%.

1718
HETEROCYCLES, Vol. 71, No. 8, 2007
After considerable experimentation, it was found that the product ratio of the ring-expanded compounds
to the undesired epoxide was significantly influenced by the choice of solvent. Thus, ring expansion was
achieved by treatment of 3 with diazomethane in the aprotic solvent mixture, CH2Cl2-Et2O, to give the
desired 8 and 9 as an inseparable regioisomeric mixture in 36% yield, and the epoxide 4 in 40% yield
(Scheme 2). It is assumed that use of CH2Cl2 in place of MeOH resulted in a decrease in the relative rate
of the reaction to form the thermodynamically stable ring-expanded product.11,12 After the keto group of
the mixture of 8 and 9 was converted into the apparently stable oxime in 74% yield, by treatment with
hydroxylamine, deprotection of the Moz and Bz groups afforded the ring-expanded 1013 and 1113 in 65%
and 16% yield respectively.14 These results suggested that the ring expansion of 3 with diazomethane in
CH2Cl2-Et2O proceeds stereoselectively to afford 8 as the major isomer. Presumably, the selective
formation of 8 is explained by considering the relative stabilities of the conformations of the
intermediates A and B (Figure 2).15
R3
Bz
NH
O
MozHN
O
O
O
O-
NHMoz
8
MozHN
O
H
+N2
H
OBz
OBz
NH
A
Moz
R3
Bz
NH
O
MozHN
O
O
O
O-
NHMoz
9
MozHN
O
+
H
N2
OBz
OBz
H
NH
B
Moz
3
R = (S)-4-amino-2-hydroxybutyryl
Figure 2. Pathways to ring-expanded products via intermediates A and B.
The antibacterial activity of 10 was tested against several strains. Interestingly, the 2-deoxystreptamine
ring-expanded arbekacin 10 showed good antibacterial activity against Staphylococcus aureus 209P JC-1,
S. aureus RN4220 and Escherichia coli JM109 (minimum inhibitory concentrations (MICs) 2, 8, and 4
μg/mL, respectively).16 Further, 10 retained activity against S. aureus RN4220/pCR1948, which is a
aminoglycoside-resistant strain, expressing aminoglycoside-modifying enzyme AAC(6’)-APH(2”)1,6,17
(MIC 16 μg/mL). These results seem to indicate that 2-deoxystreptamine ring-expanded arbekacin would
be an attractive lead compound for development of novel aminoglycosides.

HETEROCYCLES, Vol. 71, No. 8, 2007
1719
In summary, the novel ring-expanded aminoglycoside synthesized in this study represents a new class of
aminoglycoside antibiotic, which is active against E. coli and S. aureus including
aminoglycoside-resistant bacterial strain. The design would be useful for future development of novel
aminoglycosides with good antibacterial activity, while, the undesired epoxide 4 would be a useful
intermediate for preparation of various 5α-substituted aminoglycoside derivatives. The structure-activity
relationship studies of the 2-deoxystreptamine ring-expanded aminoglycosides are currently in progress.
ACKNOWLEDGEMENTS
We thank Dr. M. Oyama and Miss. S. Miki (Meiji Seika Kaisha Ltd.) for NMR spectroscopic and mass
spectrometric analyses, respectively.
REFERENCES AND NOTES
1. S. Magnet and J. S. Blanchard, Chem. Rev., 2005, 105, 477.
2. A. P. Carter, W. M. Clemons, D. E. Brodersen, R. J. Morgan-Warren, B. T. Wimberly, and V.
Ramakrishnan, Nature, 2000, 407, 340; J. M. Ogle, D. E. Brodersen, W. M. Clemons Jr., M. J. Tarry,
A. P. Carter, and V. Ramakrishnan, Science, 2001, 292, 897; Q. Vicens and E. Westhof, Structure,
2001, 9, 647.
3. G. F. Busscher, F. P. J. T. Rutjes, and F. L. Delft, Chem. Rev., 2005, 105, 775.
4. S. Castro, W. S. Fyvie, S. A. Hatcher, and M. W. Peczuh, Org. Lett., 2005, 7, 4709; L. A. Paquette
and Y. Zhang, Org. Lett., 2005, 7, 511; H. Li, Y. Bleriot, J. M. Mallet, E. R. Garcia, P. Vogel, Y.
Zhang, and P. Sinay, Tetrahedron: Asymmetry, 2005, 16, 313.
5. N. Minowa, Y. Akiyama, Y. Hiraiwa, K. Maebashi, T. Usui, and D. Ikeda, Bioorg. Med. Chem. Lett.,
2006, 16, 6351.
6. S. Kondo, K. Iinuma, H. Yamamoto, K. Maeda, and H. Umezawa, J. Antibiotics, 1973, 26, 412; K.
Ubukata, N. Yamashita, A. Gotoh, and M. Konno, Antimicrob. Agents Chemother., 1984, 25, 754.
7. Diazomethane was prepared by treatment of 1-nitroso-1-methyl-3-nitroguanidine in Et2O-H2O with
KOH. S. R. Sandler and W. Karo, Organic Functional Group Preparations, Academic Press: New
York, 1968; p395.
8. F. Weygand and K. Hunger, Chem. Ber., 1962, 95, 1.
9. T. Nagasawa, K. Kuroiwa, K. Narita, and Y. Isowa, Bull. Chem. Soc. Jpn, 1973, 46, 1269.
10. J. D. Albright and L. Goldman, J. Am. Chem. Soc., 1965, 87, 4214.
11. MeOH accelerated the reaction of cyclohexanone with diazomethane: H. O. House, E. J. Grubbs,
and W. F. Gannon, J. Am. Chem. Soc., 1960, 82, 4099; J. N. Bradley, G. W. Cowell, and A. Ledwith,
J.Chem. Soc., 1964, 4334.

1720
HETEROCYCLES, Vol. 71, No. 8, 2007
12. Electron-withdrawing or bulky groups attached to the α-positions of ketone, in addition to
accelerating the reaction, usually increase the amount of oxide formed. C. D. Gutsche, Org.
Reactions; Wiley and Sons: New York, 1954; Vol. 8, pp. 364-429.
13. Compound 5: 1H NMR (400MHz, D2O + DCl): δ 1.69-1.75 (m, 1H, H-4’), 1.90 (q, J = 12.7 Hz, 1H,
H-2a), 1.99-2.06 (m, 2H, H-4’, H-3’’’a), 2.08-2.16 (m, 2H, H-3’), 2.24 (ddt, J = 14.4, 7.3, 3.9 Hz,
1H, H-3’’’b), 2.39 (dt, J = 12.9, 4.6 Hz, 1H, H-2b), 3.20-3.24 (m, 1H, H-6’a), 3.24 (t, J = 7.4 Hz, 2H,
H-4’’’) 3.35 (dd, J = 13.6, 3.4 Hz, 1H, H-6’b), 3.44 (t, J = 10.5 Hz, 1H, H-3’’), 3.50 (s, 3H, H-8),
3.63-3.68 (m, 1H, H-2’), 3.77 (t, J = 10.0 Hz, 1H, H-4’’), 3.82 (d, J = 10.7 Hz, 1H, H-7a), 3.86 (dd,
J = 10.0, 3.7 Hz, 1H, H-2’’), 3.87 (d, J = 2.9 Hz, 2H, H-6’’), 3.87-3.92 (m, 1H, H-3), 3.94 (d, J =
10.8 Hz, 1H, H-7b), 4.02 (d, J = 10.7 Hz, 1H, H-6), 4.06 (dt, J = 10.0, 2.9 Hz, 1H, H-5’’), 4.12 (d, J
= 11.0 Hz, 1H, H-4), 4.23-4.29 (m, 1H, H-5’’), 4.33 (dd, J = 9.0, 3.6 Hz, 1H, H-2’’’), 4.35-4.40 (m,
1H, H-1), 5.19 (d, J = 3.6 Hz, 1H, H-1’’), 5.76 (d, J = 3.2 Hz, 1H, H-1’); 13C NMR (100MHz, D2O +
DCl): δ 21.5 (C-3’), 26.0 (C-4’), 31.2 (C-2), 31.7 (C-11), 37.9 (C-12), 43.3 (C-6’), 48.4 (C-1), 48.9
(C-3), 49.9 (C-2’), 56.3 (C-3’’), 60.0 (C-8), 60.6 (C-6’’), 66.4 (C-4’’), 67.2 (C-5’), 69.0 (C-2’’), 70.5
(C-10), 71.1 (C-7), 73.5 (C-5’’), 77.7 (C-5), 81.8 (C-4), 84.0 (C-6), 96.6 (C-1’), 100.4 (C-1’’), 176.4
1
(C-9). Compound 6: H NMR (400MHz, D2O + ND3): δ 1.32-1.50 (m, 2H, H-2a, H-4’a), 1.65-1.80
(m, 4H, H-3’, H-4’b, H-3’’’a), 1.85-1.95 (m, 1H, H-3’’’b), 1.97-2.05 (m, 1H, H-2b), 2.60-2.70 (m,
2H, H-6’), 2.70-2.80 (m, 2H, H-4’’’), 2.85-2.92 (m, 1H, H-2’), 2.98 (t, J = 10.0 Hz, 1H, H-3’’),
3.15-3.25 (m, 1H, H-3), 3.32 (t, J = 10.0 Hz, 1H, H-4’’), 3.39 (dd, J = 10.5, 3.9 Hz, 1H, H-2’’), 3.45
(d, J = 10.3 Hz, 1H, H-4), 3.76-3.78 (m, 2H, H-6’’), 3.81 (d, J = 10.7 Hz, 1H, H-6), 3.85-3.94 (m,
1H, H-5’), 3.90 (d, J = 11.7 Hz, 1H, H-7a), 3.99 (d, J = 11.7 Hz, 1H, H-7b), 3.99-4.05 (m, 1H,
H-5’’), 4.18 (dd, J = 9.2, 3.6 Hz, 1H, H-2’’’), 4.32-4.40 (m, 1H, H-1), 5.05 (d, J = 3.9 Hz, 1H, H-1’’),
1
5.11 (d, J = 3.2 Hz, 1H, H-1’). Compound 7: H NMR (400MHz, D2O + ND3): δ 1.36 (q, J = 12.5
Hz, 1H, H-2a), 1.42-1.52 (m, 1H, H-4’a), 1.66-1.81 (m, 4H, H-3’, H-4’b, H-3’’’a), 1.87-1.96 (m, 1H,
H-3’’’b), 2.02 (dt, J = 13.2, 4.6 Hz, 1H, H-2b), 2.60 (s, 6H, H-8, H-9), 2.63-2.70 (m, 2H, H-6’),
2.73-2.82 (m, 2H, H-4’’’), 2.97 (t, J = 9.8 Hz, 1H, H-3’’), 3.02 (d, J = 14.4 Hz, 1H, H-7a), 3.03-3.10
(m, 2H, H-3, H-2’), 3.15 (d, J = 14.2 Hz, 1H, H-7b), 3.33 (t, J = 9.7 Hz, 1H, H-4’’), 3.40 (d, J = 10.5
Hz, 1H, H-2’’), 3.41 (d, J = 10.5 Hz, 1H, H-4), 3.75-3.82 (m, 3H, H-6, H-6’’), 3.90-3.98 (m, 2H,
H-1, H-5’), 4.00-4.04 (m, 1H, H-5’’), 4.18 (dd, J = 9.5, 3.6 Hz, 1H, H-2’’’), 5.05 (s, 1H, H-1’ or
H-1’’), 5.06 (s, 1H, H-1’ or H-1’’). Compound 10: 1H NMR (400MHz, D2O + ND3): δ 1.40-1.54 (m,
1H, H-4’a), 1.60-1.88 (m, 6H, H-2, H-3’, H-4’b, H-3’’’a), 1.96-2.05 (m, 1H, H-3’’’b), 2.46 (t, J =
11.9 Hz, 1H, H-6a), 2.71 (dd, J = 13.4, 7.8 Hz, 1H, H-6’a), 2.79 (dd, J = 13.4, 3.9 Hz, 1H, H-6’b),
2.82-2.92 (m, 3H, H-2’, H-4’’’), 3.04 (t, J = 9.8 Hz, 1H, H-3’’), 3.34-3.38 (m, 1H, H-6b), 3.38 (t, J =
9.8 Hz, 1H, H-4’’), 3.42-3.45 (m, 1H, H-3), 3.46 (dd, J = 10.2, 3.9 Hz, 1H, H-2’’), 3.81-3.94 (m, 4H,

HETEROCYCLES, Vol. 71, No. 8, 2007
1721
H-5’, H-5’’, H-6’’), 3.94-3.96 (m, 1H, H-7), 3.99 (d, J = 8.1 Hz, 1H, H-4), 4.07 (s, 3H, OCH3),
4.08-4.14 (m, 1H, H-1), 4.27 (dd, J = 9.2, 3.9 Hz, 1H, H-2’’’), 4.86 (d, J = 3.4 Hz, 1H, H-1’), 5.10 (d,
13
J = 3.9 Hz, 1H, H-1’’); C NMR (100MHz, D2O + ND3): δ 29.0 (C-3’), 30.5 (C-4’), 31.7 (C-6),
38.7 (C-2), 38.9 (C-3’’’), 40.2 (C-4’’’), 48.0 (C-6’), 52.1 (C-2’), 54.6 (C-3), 57.1 (C-3’’), 58.4 (C-1),
63.5 (C-6’’), 64.7 (OCH3), 72.5 (C-4’’), 72.8 (C-2’’’), 73.5 (C-5’), 74.6 (C-2’’), 75.4 (C-5’’), 81.0
(C-7), 83.2 (C-4), 98.1 (C-1’), 102.4 (C-1’’), 157.2 (C-5), 179.1 (C-1’’’). NMR assignments were
made by interpretation of COSY experiments.
+
14. The mixture of 10 and 11 were separated by column chromatography on CM Sephadex (NH4 form,
gradient elution with 0.05-1.1N NH4OH).
15. J. A. Marshall and J. J. Partridge, J. Org. Chem., 1968, 33, 4090; N. J. Turro and R. B. Gagosian, J.
Am. Chem. Soc., 1970, 92, 2036; M. A. Mckinney and P. P. Patel, J. Org. Chem., 1973, 38, 4064.
16. MICs were detemined by the two-fold agar dilution method according to NCCLS.
17. N. Ardic, B. Sareyyupoglu, M. Ozyurt, T. Hazunedaroglu, and U. IIga, Microbiological Research,
2006, 161, 49.