Tuning the regioselectivity of the staudinger reaction for the facile synthesis of kanamycin and neomycin class antibiotics with N-1 modification

Article abstract of DOI:10.1021/ol051045d

(Chemical Equation Presented) A novel method for achieving the desired regioselective reduction of the N-1 azido group on a tetraazidoneamine has been developed that leads to the synthesis of both kanamycin and neomycin class antibiotics bearing N-1 modif

Full text of DOI:10.1021/ol051045d

ORGANIC
LETTERS
2005
Vol. 7, No. 14
3061-3064
Tuning the Regioselectivity of the
Staudinger Reaction for the Facile
Synthesis of Kanamycin and Neomycin
Class Antibiotics with N-1 Modification
Jie Li, Hsiao-Nung Chen, Huiwen Chang, Jinhua Wang, and
Cheng-Wei Tom Chang*
Department of Chemistry and Biochemistry, Utah State UniVersity,
0300 Old Main Hill, Logan, Utah 84322-0300
chang@cc.usu.edu
Received May 6, 2005
ABSTRACT
A novel method for achieving the desired regioselective reduction of the N-1 azido group on a tetraazidoneamine has been developed that
leads to the synthesis of both kanamycin and neomycin class antibiotics bearing N-1 modification. Both classes of aminoglycosides are
active against aminoglycoside-resistant bacteria carrying APH(3′)-I and AAC(6′)/APH(2′′).
For over sixty years, aminoglycoside antibiotics have been
a valuable resource against infectious diseases.1 Nevertheless,
the prevalence of aminoglycoside-resistant bacteria has
significantly reduced their effectiveness.2 With the advance-
ments in studies of resistant mechanisms and the structural
information from the binding of aminoglycosides toward the
target, the A-site decoding region of 16S rRNA,3 new
strategies have been developed that aim to revive the
antibacterial activity against aminoglycoside-resistant bac-
teria.4 Among these strategies, attaching functionalities at
the N-1 position of the 2-deoxystreptamine among kanamy-
cin or neomycin class antibiotics is one of the most effective
methods. This strategy has led to the development of
semisynthetic amikacin that has an (S)-4-amino-2-hydroxy-
butyryl (AHB) group at the N-1 position (Figure 1).
The synthesis of kanamycin and neomycin class ami-
noglycosides with N-1 modification can be achieved via
enzymatic5 or chemical methods.6 However, several aspects
in the synthesis of kanamycin and neomycin class aminogly-
(1) (a) Hooper, I. R. In Aminoglycoside Antibiotics; Springer-Verlag:
New York, 1982. (b) Haddad, J.; Kotra, L. P.; Mobashery, S. In
Glycochemistry Principles, Synthesis, and Applications; Wang, P. G.,
Bertozzi, C. R., Ed, Marcel Dekker: New York, 2001; pp 307-424.
(2) (a) Mingeot-Leclercq, M.-P.; Glupczynski, Y.; Tulkens, P. M.
Antimicrob. Agents Chemother. 1997, 43, 727-737. (b) Kotra, L. P.;
Haddad, J.; Mobashery, S. Antimicrob. Agents Chemother. 2000, 44, 3249-
3256. (c) Wright, G. D. Curr. Opin. Microbiol. 1999, 2, 499-503.
(3) (a) Fourmy, D.; Recht, M. I.; Blanchard, S. C.; Puglisi, J. D. Science
1996, 274, 1367-1371. (b) Ma, C.; Baker, N. A.; Joseph, S.; McCammon,
J. A. J. Am. Chem. Soc. 2002, 124, 1438-1442. (c) Fong, D. H.; Berghuis,
A. M. EMBO J. 2002, 21, 2323-2331. (d) Vicens, Q.; Westhof, E. Structure
2001, 9, 647-668.
(4) (a) Hanessian, S.; Kornienko, A.; Swayze, E. E. Tetrahedron 2003,
59, 995-1007. (b) Sharma, M. N.; Kumar, V.; Remers, W. A. J. Antibiot.
1982, 35, 905-910. (c) Iwasawa, H.; Ikeda, D.; Kondo, S.; Umezawa, H.
J. Antibiot. 1982, 35, 1715-1718. (d) Shitara, T.; Umemura, E.; Tsuchiya,
T.; Matsuno, T. Carbohydr. Res. 1995, 276, 75-89.
(5) (a) Cappelletti, L. M.; Spagnoli, R.; Spa, P. J. Antibiot. 1983, 36,
328-330. (b) Woo, P. W. K.; Haskell, T. H. J. Antibiot. 1982, 35, 692-
702. (c) Umezawa, S.; Tsuchiya, T.; Torii, T. J. Antibiot. 1982, 35, 58-
61.
10.1021/ol051045d CCC: $30.25
© 2005 American Chemical Society
Published on Web 06/11/2005

Figure 1. Structures of kanamycin and neomycin class aminogly-
cosides bearing N-1 modification.
Figure 2. Strategy for regioselective reduction of adizo group.
introduction of electron-withdrawing protecting groups such
as acyl groups at the O-5 and/or O-6 positions. It is expected
that the acyl groups will increase the reactivity of the N-1
azido group (Figure 2).
Various acyl protecting groups were examined, including
trifluoroacetyl, 2-chlorobenzoyl, 4-chlorobenzoyl, benzoyl,
and acetyl groups, which have pKa values of -0.25, 2.92,
3.98, 4.19, and 4.76 for the corresponding carboxylic acids,
respectively (Scheme 1).10 The electron-withdrawing capabil-
cosides with N-1 modification have not been satisfactorily
addressed. First, only a few examples of neomycin class
antibiotics bearing N-1 modification have been reported.5b,c,6b
Second, the synthetic works involved in N-1 modification
on kanamycin class antibiotics are commonly achieved by
using various metal chelation methods, which require certain
configurations of hydroxyl and amino groups. Introduction
of additional functional groups may disrupt the chelation,
resulting in the loss of regioselective differentiation of the
amino groups. Therefore, metal chelation methods are often
limited to modifications of neamine or kanamycin. Third,
many prior works employed carbamate-type protecting
groups for the protection of amino groups on the aminogly-
coside, resulting in the formation of polycarbamate com-
pounds that pose difficulties in their purification and
characterization. To alleviate the solubility and purification
problems, Wong’s group7 and our group8 have been using
azido groups as the surrogate of amino groups for the
synthesis of novel aminoglycosides. Nevertheless, there is
no precedent of regioselective conversion of the N-1 azido
group to an amino group.
Scheme 1. Synthesis of O-5 and/or O-6 Acylated Neamine
Derivatives
Wong and others have reported that electron-withdrawing
protecting groups will enhance the reactivity of their vicinal
azido group toward the Staudinger reaction.9a Such an
electron-withdrawing effect can be correlated with chemical
shifts of the corresponding protons (H-1, H-3, H-2′, and H-6′)
adjacent to the azido groups. However, the N-2′ azido group
appears to be more reactive than the N-1 azido group from
the experimental and spectroscopic data (Figure 2). To
selectively reduce the N-1 azido group, we began to examine
the possibility of tuning the stereoelectronic effect by
ity of acyl groups is expected to be proportional to the pKa
of the corresponding acids. The attempt to employ trifluo-
roacetyl group at the O-5 and/or O-6 positions was, however,
unsuccessful.
By comparison to compound 2, incorporation of acyl
groups results in various degrees of deshielding of protons
(Table 1). One important finding is the effect from the acyl
group at the O-5 position. Although the O-6 hydroxyl group
can be selectively protected, we noticed that only the
attachment of the O-5 acyl group causes significant upfield
shift for the H-2′ proton. The result suggests that a diacyl-
protected azidoneamine is essential in offering desired
regioselective Staudinger reaction. A possible anisotropic
effect from the acyl group at the O-5 position was proposed
(6) (a) Akita, E.; Horiuchi, Y.; Miyazawa, T.; Umezawa, H. J. Antibiot.
1983, 36, 745-748. (b) Takagi, Y.; Komuro, C.; Tsuchiya, T.; Umezawa,
S.; Hamada, M.; Umezawa, H. J. Antibiot. 1981, 34, 1-4. (c) Tsuchiya,
T.; Takagi, Y.; Umezawa, S. Tetrahedron Lett. 1979, 51, 4951-4954.
(7) Liang, F. S.; Wang, S.-K.; Nakatani, T.; Wong, C.-H. Angew. Chem.,
Int. Ed. 2004, 43, 6496-6500.
(8) (a) Elchert, B.; Li, J.; Wang, J.; Hui, Y.; Rai, R.; Ptak, R.; Ward, P.;
Takemoto, J. Y.; Bensaci, M.; Chang, C.-W. T. J. Org. Chem. 2004, 69,
1513-1523. (b) Li, J.; Wang, J.; Czyryca, P. G.; Chang, H.; Orsak, T. W.;
Evanson, R.; Chang, C.-W. T. Org. Lett. 2004, 6, 1381-1384.
(9) (a) Nyffeler, P. T.; Liang, C.-H.; Koeller, K. M.; Wong, C.-H. J.
Am. Chem. Soc. 2002, 124, 10773-10778. (b) Ariza, X.; Vrpi, F.;
Viladomat, C.; Vilarrasa, J. Tetrahedron Lett. 1998, 39, 9101-9102.
(10) CRC Handbook of Chemistry and Physics, 76th ed.; CRC Press:
Boca Raton, FL, 1995; pp 8-45-8-51.
3062
Org. Lett., Vol. 7, No. 14, 2005

Scheme 3. Synthesis of N-1-Modified Neamine
Table 1. Chemical Shifts of Protons on Acylated Neamine
Derivatives
entry
compounds
H-1
H-3
H-2′
H-6′
1
2
3
4
5
6
7
8
9
2
3
8a
4
5
8c
6
8d
7
3.28
3.65
3.81
3.68
3.81
4.05
3.78
4.01
3.80
4.05
3.43
3.46
3.47
3.38
3.61
3.62
3.61
3.59
3.60
3.61
3.62
3.35
3.36
3.61
3.18
3.18
3.25
3.23
3.20
3.19
3.57/3.41
3.62/3.42
3.60/3.42
3.58/3.42
3.54/3.42
3.54/3.42
3.54/3.42
3.55/3.39
3.54/3.42
3.54/3.42
10
8e
After one-pot Staudinger reaction/Boc protection and then
hydrolysis of the acyl groups, the desired N-1-Boc-protected
neamine, 10, was synthesized in an overall yield of 33%
along with 3-N-Boc-protected neamine, 11, (5%) as the
minor product (Scheme 3). Compound 10 can be directly
glycosylated with the selected glycosyl donors, 19 and 20
(Scheme 4).12 Deprotection of Boc followed by coupling with
to explain the dramatic upfield shifting of the H-2′ proton
signal.11 After the attachment of diacyl groups, a more
deshielded H-1 emerges, indicating that the N-1 azido group
will be more reactive toward the Staudinger reaction than
N-2′ azido group (entries 2, 5, 7, and 9 in Table 1).
A one-pot azido reduction/amine protection was then
employed (Scheme 2).9b A characteristic downfield shift (0.2
Scheme 4
Scheme 2. Selective Reduction and Protection of the N-1
Azido Group
the Cbz-protected AHB side chain yielded compounds, 15
and 16. Staudinger reduction of the remaining azido groups
followed by hydrogenation provided the final amikacin
analogues in modest to excellent yields. Compound 10 can
be selectively benzoylated at O-6 and then glycosylated at
O-5 (Scheme 3). The glycosyl donor was selected on the
basis of leads from our previous work (Scheme 5).8 The
glycosylated product was subjected to similar treatment to
furnish the desired pyranmycin with an AHB side chain at
N-1.
ppm) of the proton resulting from the Boc-protected amino
group and information from COSY experiment confirm the
regioselectivity (entries 3, 6, 8, and 10 in Table 1).
(11) We have used molecular modeling (HyperChem) to reveal the
possible anisotropic effect around H-2′ caused by the O-5 carbonyl group.
Please refer to Supporting Information for details.
(12) Glycosyl donors were selected on the basis of our previous work
described in ref 8 and unpublished results.
Org. Lett., Vol. 7, No. 14, 2005
3063

Table 2. Minimum Inhibitory Concentrations (MICs)a
Scheme 5
strains
E. coli TG1
E. coli TG1
entry compounds E. coli TG1 (APH(3′)-I) (AAC(6′)/APH(2′′))
1
2
3
4
5
6
7
8
9
amikacin
kanamycin
JLN005
JL005b
JLN027
JL027b
1
4
4
8
1
2
1
2
4
8
0.5
inactive
2
inactive
0.25
inactive
0.5
inactive
4
inactive
1
inactive
2
inactive
1
inactive
0.25
8
4
8
butirosin
ribostamycin
JT005
10 TC005c
a Unit: µg/mL. b Related kanamycin analogues without an AHB group
at N-1. c Related pyranmycin without an AHB group at N-1.
be valuable for designing new aminoglycosides against a
broad spectrum of aminoglycoside-resistant bacteria.
In conclusion, we have demonstrated that the reactivity
of azido groups can be tuned by the stereoelectronic effect
of protecting groups. We have developed a convenient
synthesis that can be used for modification at N-1 of both
kanamycin and neomycin classes of aminoglycosides. With-
out relying on the chelating method and carbamate-type of
protecting group, novel structural entities can be easily
introduced via a glycodiversification strategy. One synthe-
sized compound, JLN027, manifests even better activity than
the clinically used amikacin against aminoglycoside-resistant
bacteria. Ongoing works that focus on further modifications
based on the leads are being performed.
The constructed amikacin analogues and N-1-modified
pyranmycin were assayed against three strains of Escherichia
coli (one susceptible and two resistant strains) using ami-
kacin, kanamycin, ribostamycin, and butirosin as the controls.
One resistant strain is equipped with the pTZ19U-3 plasmid
encoded for APH(3′)-I. The other resistant strain is equipped
with the pSF815 plasmid encoded for a bifunctional enzyme
(AAC(6′)/APH(2′′)). These enzymes are among the most
prevalent modes of resistance found in aminoglycosides
resistant strains. From the minimum inhibitory concentrations
(MICs), all the synthesized aminoglycoside with the attach-
ment of an AHB group at N-1 regain their activity against
both resistant strains of bacteria (Table 2). One of the
synthesized kanamycin analogues, JLN027, is even more
active than amikacin (entry 5, Table 2) against APH(3′)-I.
Without the attachment of an AHB group at N-1, kanamycin
class antibiotics are ineffective against AAC(6′) and APH-
(2′′), while the neomycin class antibiotics (ribostamycin and
pyranmycin8 (TC005, entry 10, Table 2)) seem to be
modestly active against the same enzyme. Therefore, neo-
mycin class antibiotics with the AHB group at the N-1
position may represent a better template for further modifica-
tion. Nevertheless, the problem of the acid-labile glycosidic
bond between rings II and III in neomycin or butirosin is an
obstacle that remains to be overcome. Therefore, our design
of pyranmycin that has better stability in acidic media could
Acknowledgment. We acknowledge the National Insti-
tutes of Health (AI053138) and DARPA for financial
support. We thank Prof. Mobashery from Notre Dame
University for providing the pTZ19U-3 and pSF815 plas-
mids.
Supporting Information Available: Experimental pro-
1
cedures for the preparation of compounds and H, 13C, and
COSY spectra of selected compounds. This material is
available free of charge via the Internet at http://pubs.acs.org.
OL051045D
3064
Org. Lett., Vol. 7, No. 14, 2005