J. G. da Silva et al. / Bioorg. Med. Chem. 15 (2007) 3624–3634
3631
as neomycin sulfate (7), it was possible to obtain a frag-
mentation product, methyl neobiosaminide B (17), with
the desired vestibular selectivity and lacking cochleotox-
icity. Preliminary results showed different patterns of co-
chlear and vestibular in vivo activities indicating that
this may be an interesting approach to develop new
drugs. Furthermore, previous studies in this area per-
formed by Owada15 only explored monomers of neomy-
cin units and did not give information about cochlear
and vestibular toxicity of compounds 8 and 17. In this
report we showed the high activity of neamine (8)
against cochlear and vestibular tissue, the later being
higher than neomycin (7).
used to separate the solid product which was further
dried under high vacuum (715 mg). The solution of the
solid material in the smallest amount of methanol was
titrated with ethyl ether until it became turbid. The
insoluble fraction was separated by the same filtration
system. The process was repeated until TLC (MeOH/
acetone) showed complete removal of impurities in the
remaining solution. Thus, the solution was concentrated
to dryness in a rotatory evaporator in high vacuum giv-
ing compound 17 as a highly hygroscopic white powder
(335 mg, 1.03 mmol, 42%) composed of a mixture of
1:4.5 a/b mixture at C-1. The melting point was higher
than 300 ꢁC. Rf 0.61 (MeOH/acetone 2:1, v/v), while
the main impurity neamine hydrochloride was much
slower (Rf, 0.40). Data for the NMR 1H (400 MHz/
A comparative investigation between the cochlear and
vestibular activities and a qualitative antibacterial anal-
ysis pointed out that neomycin sulfate (7) and its frag-
mentation products, neamine as a free base (8) and
methyl-neobiosaminide B dihydrochloride (17) showed
a broader range of inhibition than compounds 9, 18,
and 22. The hydrophobic properties of compounds 19–
21 prevented their use in microbiological assays due to
the low water solubility and the use of DMSO was det-
rimental since the solvent has displayed some influence
on all tested strains.
0
0
0
0
D2O) major isomer: d 3.19 (1H, dd, J5 ;6 a 3.3, J6 a;6 b
13.4, H-60a); 3.29 (1H, dd, J5 ;6 b 6.3, J6 a;6 b 13.4, H-
60b); 3.33 (3H, s, OCH3); 3.48–3.51 (1H, m, H-20);
3.62 (1H, dd, J4,5a 6.0, J5a,5b 12.1, H-5a); 3.67–3.70
(1H, m, H-40); 3.72 (1H, t, J2,3 3.8, H-2); 3.75 (1H, dd,
J4,5b 3.3, J5a,5b 12.1, H-5b); 3.91 (1H, ddd, J3,4 3.0,
0
0
0
0
0
0
0
0
J4,5b 3.3, J4,5a 6.0, H-4); 4.11 (1H, dd, J3 ;4 3.3, J2 ;3
2.5, H-30), 4.14–4.20 (2H, m, H-50, H-3); 4.63 (1H, d,
J 4.5, H-1); 5.27 (1H, d, J = 1.7, H-10). NMR 13C
(100 MHz/D2O): d 40.74 (C-60), 51.1 (C-20), 56.24
(OCH3), 63.46 (C-5), 67.59 (C-40), 67.70 (C-4), 68.05
(C-30), 68.95 (C-2), 70.66 (C-50), 74.84 (C-3), 95.41 (C-
10), 101.55 (C-1), ESIMS found m/z 325.1725 [M+H+].
Calcd for C12H24N2O8 324.1533.
4. Experimental
1
1H, 13C, H–1H COSY, and 13C–1H COSY NMR spec-
tra were measured on a Bruker DPX-400 (400 and
100 MHz) using CDCl3 (Aldrich) as solvent and TMS
as internal standard. Chemical shifts were reported in
d units (ppm) and coupling constants (J) in Hz. IR spec-
tra were recorded on a Nicolete Protege FT460 spectro-
photometer. ESI-EM spectra data were obtained from
ULTROTOFQ Bruker Daltonics, Billerica. TLC was
performed on silica gel (E. Merck) SIL G-25UV254
and detection of the separated compounds was either
by viewing with a UV lamp (254 nm), or by spraying
with a 10% sulfuric acid, 1.5% molybdic acid, 1% ceric
sulfate spray followed by heating to 150 ꢁC. Column
chromatography was performed on a Kieselgel 60 (70–
230 mm mesh, E. Merck) column. Neomycin sulfate
was purchased from GIBCOTM. Neamine tetrahydro-
chloride (8),50,51 2-deoxyestreptamine (9),52 tetra-N-
acetylneamine (18),53 tetra-azidoneamine (22),41
tetra-N-carboxy-methylneamine (19)36,37 and tetra-N-
carboxy-benzylneamine (20)38 were prepared in the
same manner as previously described.
4.2. Tetra-p-methoxy-benzyliminoneamine (21)
Neamine tetrahydrochloride (8) (93 mg, 0.2 mmol) was
sequentially treated with a 1 M NaOH (1 mL) and
p-anisaldehyde (0.15 mL). The mixture was stirred until
the spontaneous precipitation of product 21. After
remaining at 4 ꢁC for 2 h, it was filtered through a med-
ium-porosity sintered glass-funnel. The white solid was
washed several times with cold water and portions of
ethyl ether/ethanol (1:1, v/v) and then dried under high
vacuum to give compound 21 as a pure product
(127 mg, 0.17 mmol, 85%). Mp 255 ꢁC; NMR 1H
(400 MHz/DMSO-d6): d 1.47–1.57 (1H, m, H-20a), 1.97
(1H, q, J = 13.6, H-20b), 2.82 (1H, dd, J = 12.6, H-6a),
3.02 (1H, dd, J = 2.5, 9.5, H-3), 3.16–3.25 (1H, m, H-
10 or H-30), 3.32–3.49 (4H, m, H-10 or H-30, H-20, H-
40, H-4), 3.53–3.67 (4H, m, H-2, H-6b, H-50, H-60),
3.70–3.89 (13H, m, H-5, 4 · OCH3), 4.74 (1H, d,
J = 5.0, OH), 4.85–4.90 (2H, m, H-1, OH), 5.05 (1H,
d, J = 5.0, OH), 6.34 (1H, s, OH), 7.00 (4H, d, J = 8.8
ArH), 7.05 (4H, d, J = 8.8, ArH), 7.47 (1H, s, @CH),
7.63 (2H, d, J = 8.8, H), 7.68 (2H, d, J 8.6, ArH), 7.73
(2H, d, J = 8.6, ArH), 7.77 (2H, d, J = 8.6, ArH), 8.22
(1H, s, @CH), 8.24 (1H, s, @CH), 8.33 (1H, s, @CH);
4.1. Methyl neobisamidine B dihydrochloride (17)41
The compound was prepared from neomycin sulfate (7)
(2 g) as previously described by Ford28 and Botto,29
with minor modifications. The remaining mother liquor
from the acid methanolysis of neomycin sulfate (7)
(240 mL) was concentrated in a rotatory evaporator to
approximately 50 mL and treated with anhydrous ethyl
ether (150 mL). The flocculent white precipitate started
to form immediately but the process was completed after
refrigeration for 12 h. Filtration under nitrogen pressure
through a medium-porosity sintered glass-funnel was
13
NMR C (100 MHz/DMSO-d6) d 38.44 (C-20), 55.65
(OCH3), 60.1 (C-6), 70.18, 70.55, 70.89, 71.00, 72.11,
75.03, 76.86, 88.46 (C-10, C-30, C-40, C-50, C-60, C-2,
C-3, C-4, C-5), 101.79 (C-1), 114.45, 114.32, 114.28,
114.24, 114.87, 129.62, 129.95, 130.25, 130.79 (ArC),
132.18 (ArCq), 160.28 (C@N), 160.42 (C@N), 161.12
(C@N), 164.88 (C@N); IRpMax: 3360 (OH), 1641
(C@N).