Synthesis of (+),(-)-neamine and their positional isomers as potential antibiotics

Article abstract of DOI:10.1016/S0960-894X(02)01073-9

The syntheses of (+)-neamine 1, (-)-neamine ent-1 and their positional isomers 2, 3, ent-2 and ent-3 are reported as potential new scaffolds for novel aminoglycoside antibiotics. These isomers exhibit similar inhibitory activities, as shown using an in vitro translation assay. A simple model is proposed to explain this lack of stereospecific binding to the ribosomal RNA.

Full text of DOI:10.1016/S0960-894X(02)01073-9

Bioorganic & Medicinal Chemistry Letters 13 (2003) 901–903
Synthesis of (+),(ꢀ)-Neamine and Their Positional Isomers
as Potential Antibiotics
Do Hyun Ryu,y Choon-Hong Tan and Robert R. Rando*
DepartmentofBiological ChemistryandMolecularPharmacology, HarvardMedicalSchool, 45ShattuckStreet, Boston, MA02115, USA
Received 4 October 2002; revised 10 December 2002; accepted 11 December 2002
Abstract—The syntheses of (+)-neamine 1, (ꢀ)-neamine ent-1 and their positional isomers 2, 3, ent-2 and ent-3 are reported as
potential new scaffolds for novel aminoglycoside antibiotics. These isomers exhibit similar inhibitory activities, as shown using an in
vitro translation assay. A simple model is proposed to explain this lack of stereospecific binding to the ribosomal RNA.
# 2003 Elsevier Science Ltd. All rights reserved.
Aminoglycoside antibiotics are clinically useful drugs
known to function by binding to the A-site decoding
region on bacterial 16S rRNA.1 This binding causes
mistranslation and premature termination during pro-
tein synthesis in bacteria, leading to the bacteriocidal
effects of this class of drug.2 However, the high toxicity
and rapid emergence of resistance has limited their use,
and resulted in the need for novel aminoglycosides
devoid of these limitations.3 Since several aminoglyco-
sides such as neomycin B, ribostamycin and kanamycin
B share a common pseudodisaccharide, (+)-neamine 1,
it has been used as a scaffold for the synthesis of diverse
small molecules for developing new aminoglycoside
antibiotics.3c,3d,4
of (+)-neamine and their enantiomers ent-2, ent-3 to
investigate the origin of the weak stereospecific inter-
actions, and to search for new scaffolds for novel anti-
biotics. We describe here the synthesis of (+)-neamine
1, (ꢀ)-neamine ent-1, all positional isomers 2, 3, ent-2,
ent-3 (Fig. 1) and a working model to explain the source
of weak stereospecificity.
The synthesis of pseudodisaccharide 1–3 and ent-1–ent-
3 was accomplished by glycosidation of protected
2-deoxystreptamine acceptors 4 and 6 with thioglyco-
side donors 12 and ent-12 (Scheme 1). Methoxymethyl
protected 2-deoxystreptamine acceptor 6 was prepared
fromthe known diacetylated 2-deoxystreptamine deri-
vative 4.4a The d-thioglycoside donor 12 was synthe-
sized fromthe protected azidoglycoside 7, derived from
d-glucosamine.7 Azidoglycoside 7 (a/b, 1:3.5) was con-
verted to the phenylthioglycoside 8 (a/b, 1.6:1) by
treatment with phenylthiotrimethylsilane and zinc
iodide.8 Acetate deprotection of 8 afforded triol 9,
which was monotosylated to give 10. The tosylate 10
was displaced with sodiumazide providing 11 in quan-
titative yield. Dibenzylation of 11 gave d-thioglycoside
donor 12 (a/b, 1:1). The enantiomer of 12, ent-12 (a/b,
1:1) was obtained from ent-7,9 a l-glucosamine deriva-
tive, using the same protocol just described.
In order to design a more potent antibiotics, the ques-
tion of stereospecificity of this class of drug needs to be
addressed. Recently, we synthesized (+)-neamine 1 and
(ꢀ)-neamine ent-1 and binding measurements with the
A-site rRNA constructs demonstrated weak stereo-
specific binding.5 Functional studies, including in vitro
translation assays, antibiotic disk assays and minimum
inhibitory concentration studies, also reveal a low level
of stereospecificity. Interestingly, (ꢀ)-neamine ent-1,
inhibits the growth of aminoglycoside resistant
Escherichia coli.5,6
These results led us to synthesize positional isomers 2, 3
Glycosylation of the 2-deoxystreptamine acceptor 4
with d-thioglycoside donor 12 (Scheme 2) and acetate
deprotection provided the pseudodisaccharide 13 in
74% yield (a/b, 10:1). Coupling of donor 12 and accep-
tor 6 gave pseudodisaccharide 14 and 15 after depro-
tection of the MOM group and chromatographic
*Corresponding author. Tel.:+1-617-432-1794; fax:+1-617-432-0471;
e-mail: robert_rando@hms.harvard.edu
yCurrent Address: Department of Chemistry and Chemical Biology,
12 Oxford Street, Cambridge, MA 02138, USA.
0960-894X/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0960-894X(02)01073-9

902
D. H. Ryu et al. / Bioorg. Med. Chem. Lett. 13 (2003) 901–903
Using the same coupling conditions, ent-12 and 4 gave
ent-15 in 73% yield with a better selectivity (a/b, 50:1).
Similarly, 6 and ent-12 coupled to provide ent-14 and
ent-13 in 14 and 46% yield respectively, with 15% of the
b isomeric mixture after deprotection and chromato-
graphic separation. Azidoglycosides ent-13–ent-15 were
also individually deprotected to give the 4-positional
isomer of (ꢀ)-neamine ent-3, its 5-positional isomer ent-
2 and (ꢀ)-neamine ent-1 in 92, 78, and 90% yields,
respectively.
Binding assays of the aminoglycosides with either A-site
decoding region rRNA constructs or ribosomes were
found not to be quantitatively predictive of the bacter-
iocidal activity of these antibiotics, compared to in vitro
translation assays.5 In vitro translation studies of (+)-
neamine 1, (ꢀ)-neamine ent-1, and their positional iso-
mers 2, 3, ent-2, ent-3 were determined using Ambion
PROTEINscriptTM-PRO kit (Table 1). The translation
assay utilized highly active E. coli S30 extracts and a
plasmid containing a gene expressing truncated lecithin
retinol acyltransferase (tLRAT).10 Various concen-
trations of the antibiotics were added to inhibit translation.
Figure 1. Structures of (+)-neamine 1, (ꢀ)-neamine ent-1 and their
positional isomers (2, 3, ent-2 and ent-3).
The translation assay revealed that (+)-neamine is
approximately a 5-fold better inhibitor than (ꢀ)-nea-
mine. The assay also revealed two trends. The inhibitory
ability of the 4-positional isomers 1 and ent-3 >5-posi-
tional isomers 2 and ent-2 >6-positional isomers 3 and
ent-1. The (+)-neamine and its isomers are more potent
inhibitors than (ꢀ)-neamine and its isomers.
Footprinting studies of (+)-neamine 1 and (ꢀ)-neamine
ent-1 with the A-site rRNA construct, revealed that they
bind to the same site.5 NMR studies of the A-site of 16S
rRNA and (+)-neamine by Puglisi11 showed the pres-
ence of intermolecular NOEs between 30OH, 40OH,
1NH2 and 3NH2 of (+)-neamine 1 and the rRNA
(Orientation I, Fig. 2). An alternative binding mode was
also proposed (Orientation II, Fig. 2). In both binding
modes, 1NH2 and 3NH2 are essential for binding. The
5-positional isomers, 2 and ent-2, has its deoxy-
streptamine arranged such that one of its two amino
groups point away fromthe active site, resulting in
much less effective binding. The least potent 6-posi-
tional isomers, 3 and (ꢀ)-neamine ent-1, have both
amino groups pointing away. The most effective inhi-
bitor in the enantiomeric series, ent-3, adopted con-
formations that mimics (+)-neamine and allows 1NH2
and 3NH2 to face the active site (Fig. 2). These alter-
Scheme 1. Synthesis of glycosyl acceptors 4, 6 and donor 12 and ent-
12. Reagents and conditions: (a) CH2(OCH3)2, CHCl3, P2O5; 96% (b)
NaOMe, MeOH, 100%; (c) PhSSi(CH3)3, ZnI2, ClCH2CH2Cl, 50 ꢁC,
90%; (d) NaOMe, MeOH, 98%; (e) toluenesulfonyl chloride, pyridine,
94%; (f) NaN3, DMF, 80 ꢁC, 100%; (g) BnBr, NaH, DMF, 86%.
Table 1. IC50 (50% inhibition of translation, mM) of the in vitro
translation assay of (+)-neamine, (ꢀ)-neamine and their positional
isomers
Figure 2. A simple model of (+)-neamine 1 and ent-3’s interaction
with the A-site of rRNA.
Aminoglycoside
IC50, (mM)
separation. Pseudodisaccharides 14 and 15 were
obtained in 30 and 37% yields, respectively, along with
a 15% yield of a mixture of b isomers. Azidoglycosides
13–15 were individually subjected to Staudinger reaction4a
conditions followed by benzyl deprotection to give (+)-
neamine 1, its 5-positional isomer 2 and its 6-positional
isomer 3 in 90, 77 and 91% yields, respectively.
1
2
38.6ꢂ4.6
132ꢂ9.0
197ꢂ10
198ꢂ17
152ꢂ12
102ꢂ6
3
ent-1
ent-2
ent-3
Neomycin B
14.9ꢂ0.1

D. H. Ryu et al. / Bioorg. Med. Chem. Lett. 13 (2003) 901–903
903
Acknowledgements
This work was supported by the US Public Health Ser-
vice NIH Grant EY-12375.
References and Notes
1. (a) Woodcock, J.; Moazed, D.; Cannon, M.; Davies, J.;
Noller, H. F. EMBO J. 1991, 10, 3099. (b) Fourmy, D.;
Recht, M. I.; Blanchard, S. C.; Puglisi, J. D. Science 1996, 274,
1367.
2. Cundliffe, E.; Dahlberg, A. E.; Garret, R. A.; Moore, P. B.;
Schlessinger, D.; Warner, J. R. In The Ribosome: Structure,
Function & Evolution; Hill, W. E., Ed.; American Society for
Microbiology: Washington DC, 1990; p 479.
3. (a) Davies, J. Science 1994, 264, 375. (b) Daigle, D. M.;
Hughes, D. W.; Wright, G. D. Chem. Biol. 1999, 6, 99. (c)
Sucheck, S. J.; Wong, A. L.; Koeller, K. M.; Boehr, D. D.;
Draker, K.; Sears, P.; Wright, G. D.; Wong, C.-H. J. Am.
Chem. Soc. 2000, 122, 5230. (d) Haddad, J.; Kotra, L. P.;
Llano-Sotelo, B.; Kim, C.; Azucena, E. F.; Liu, M.; Vaku-
lenko, S. B.; Chow, C. S.; Mobashery, S. J. Am. Chem. Soc.
2002, 124, 3229.
Scheme 2. Synthesis of (+)-neamine 1, (ꢀ)-neamine ent-1 and their
positional isomers (2, 3, ent-2 and ent-3). Reagents and conditions: (a)
N-iodosuccinimide, silver trifluoromethanesulfonate, Et2O/CH2Cl2
(3:1), ꢀ20 ꢁC; (b) NaOMe, MeOH; (c) 1 N HCl, MeOH, 75 ꢁC; (d)
P(CH3)3, THF/0.1 N NaOH (9:1); (e) H2, 20% Pd(OH)2/C, AcOH/
H2O (1:1).
4. (a) Greenberg, W. A.; Priestley, E. S.; Sears, S. P.; Alper,
P. B.; Rosenbohm, C.; Hendrix, M.; Hung, S.-C.; Wong, C.-
H. J. Am. Chem. Soc. 1999, 121, 6527. (b) Nunns, C. L.;
Spence, L. A.; Slater, M. J.; Berrisford, D. J. Tetrahedron Lett.
1999, 40, 9341.
5. Ryu, D. H.; Litovchick, A.; Rando, R. R. Stereospecificity
of aminoglycoside-ribosomal interactions. Biochemistry 2002,
41, 10499.
6. (ꢀ)-Neamine ent-1 and ent-3, showed similar antibiotic
activity against both E. coli and neomycin/kanamycin resis-
tant E. coli, ref 5 and unpublished results.
7. Alper, P. B.; Hung, S.-C.; Wong, C.-H. Tetrahedron Lett.
1996, 34, 6029.
native conformations could explain the relative inhibi-
tory power of these antibiotics and also the low level of
stereospecificity observed.
In conclusion, the syntheses of (+)-neamine 1, (ꢀ)-
neamine ent-1 and their positional isomers 2, 3, ent-2
and ent-3 are reported. These isomers exhibit similar
inhibitory activities as shown using the translation
assay. We also proposed that these low level of stereo-
specificity could be due to possible alternative con-
formations that ent-1, ent-2 and ent-3 may adopt. These
results may serve as the basis for the preparation of
novel unnatural aminoglycosides.
8. Buskas, T.; Garegg, P. J.; Konradsson, P.; Maloisel, J.-L.
Tetrahedron: Asymmetry 1994, 5, 2187.
9. Lin, C.-H.; Sugai, T.; Haicomb, R. L.; Ichikawa, Y.; Wong,
C.-H. J. Am. Chem. Soc. 1992, 114, 10138.
10. Jahng, W. J.; Cheung, E.; Rando, R. R. Biochemistry
2002, 41, 6311.
11. Fourmy, D.; Recht, M. I.; Puglisi, J. D. J. Mol. Biol. 1998,
277, 347.
Supporting information available: Experimental proce-
dure for the preparation of 1, 2, 3, ent-1, ent-2, and ent-3,
their spectroscopic data, and in vitro translation assay.