A method for preparing N-alkylated kanosamines from diacetone D-glucose

Article abstract of DOI:10.1016/j.tetlet.2019.06.012

The aminoglycoside (AG) antibiotics have seen a resurgence in their clinical use given the increase in multi drug resistant bacterial infections. Campaigns to generate novel analogs show promise that structural modification can lead to compounds with improved pharmacological properties. The results described herein include a new method to synthesize mono-, di-, and mixed N-alkylated kanosamine sugars and their elaboration into novel glycosides that inhibit bacterial protein synthesis in vitro.

Full text of DOI:10.1016/j.tetlet.2019.06.012

Accepted Manuscript
A method for preparing N-alkylated kanosamines from diacetone D-glucose
Zachary P. Cannone, Mark W. Peczuh
PII:
S0040-4039(19)30553-2
https://doi.org/10.1016/j.tetlet.2019.06.012
TETL 50853
DOI:
Reference:
Tetrahedron Letters
To appear in:
Received Date:
Revised Date:
Accepted Date:
8 April 2019
3 June 2019
6 June 2019
Please cite this article as: Cannone, Z.P., Peczuh, M.W., A method for preparing N-alkylated kanosamines from
diacetone D-glucose, Tetrahedron Letters (2019), doi: https://doi.org/10.1016/j.tetlet.2019.06.012
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

A method for preparing N-alkylated kanosamines from diacetone D-
glucose
Zachary P. Cannone and Mark W. Peczuh*
Department of Chemistry, University of Connecticut, 55 N. Eagleville Road, U3060, Storrs, CT, 06269 USA
*Corresponding author. E-mail: mark.peczuh@uconn.edu
ABSTRACT:
The aminoglycoside (AG) antibiotics have seen a resurgence in their clinical use given the increase in multi drug resistant
bacterial infections. Campaigns to generate novel analogs show promise that structural modification can lead to compounds
with improved pharmacological properties. The results described herein include a new method to synthesize mono-, di-, and
mixed N-alkylated kanosamine sugars and their elaboration into novel glycosides that inhibit bacterial protein synthesis in
vitro.
Keywords: amino sugar, synthesis, reductive amination, diversity
Recent years have witnessed a dramatic increase in multi
drug resistant (MDR) Gram-negative bacterial infections.1,2
These infections pose an ever-increasing threat to human
health, and a search for potential therapeutics has prompted
research into the underlying resistance mechanisms, and new
chemical entities to combat them.3,4 In clinical practice,
aminoglycoside (AG) antibiotics, which had not been
considered an option for some time, have seen a resurgence in
their utilization given the emergence of infections caused by
MDR bacteria. Efforts to develop derivatives of AGs that
improve their pharmacological properties have also been
given new life in light of MDR infections and clinical AG use.5
Issues with current treatments include the frequent
occurrence of nephrotoxicity and ototoxicity.6 While AGs
remain active against many clinical isolates of MDR bacteria,
toxicity limits their utility. The observed retention of activity
gives promise that these compounds hold potential for new
generations with improved pharmacodynamic properties to
alleviate such issues.6
kanosamines has yet to be reported. Here we present a
methodology for accessing various mono, di-, and mixed N-
alkyl kanosamine derivatives, and their incorporation into
kanamycin analogs.
H2N
H2N
HO
O
A
NH2
B
O
HO
R
O
HO
O
OH
HO
C
Kanamycin A R = OH
Kanamycin B R = NH2
H2N
OH
Figure 1. Kanamycins A and B.
Efforts to develop new AG analogs have focused on
circumventing resistance mechanisms. Resistance primarily
involves aminoglycoside modifying enzymes (AMEs)4 that
render AGs inactive by acetylation, phosphorylation, or
nucleotide addition reactions. In one study, designed
kanamycin B derivatives bearing N-alkylated and N-acylated
modifications at the 4’ position of the neamine A ring replaced
its 4’ hydroxyl group.9 Antibiotic activity was observed for the
series including improved activity against some MDR bacteria
Aminoglycosides, including kanamycins A and B (Figure 1),
are a potent class of antibiotics that act via protein synthesis
inhibition. This mechanism of action requires binding to the
ribosome as the key factor to their potency. Just as the
structures of AGs vary, so too do the specifics of their binding
interactions. In general, AGs derive their activity by
interacting with 16s rRNA of the decoding A-site of the
bacterial ribosome.7 Crystal structures of various AGs in
complex with oligonucleotides mimicking the A-site topology
show that AG binding forces A1492 and A1493 to take up a
“bulged-out” conformation that leads to a loss in translational
fidelity. Due to their polycationic nature, AGs also owe much
of their binding properties to electrostatic interactions.8
While several studies on aminoglycoside derivatives have
been conducted (vide infra), one using N-alkylated
from clinical isolates. In
a
different approach,
conformationally constrained kanamycin A derivatives were
studied.10 Bridging the 2’ hydroxyl of the neamine ring (A)
with the 5’ hydroxyl of the central deoxystreptamine ring (B)
constrained the rotation between the two rings. The
modification also masked these sites from deactivation by
AMEs. MIC values for the series varied with the length of the
linker between the two rings, though no analog performed
better than kanamycin A itself. In addition to the neamine
core, the kanamycins bear a 3-deoxy-3-amino glucose unit
(kanosamine, ring C in Figure 1). This ring aides in RNA

recognition via electrostatic and hydrogen bonding
interactions,8 is integral to binding the
excess of dimethylamine in a pressure vessel, followed by acid-
catalyzed acetonide deprotection, gave dimethyl kanosamine
4 in 96% yield
Scheme 1. Synthesis of N-alkyl kanosamines from
diacetone D-glucose
Table 1. Reductive aminations with amine 5
O
O
O
O
1) PDC, Ac2O
O
O
O
O
O
2) NaBH4, EtOH/H2O
O
NaBH3CN
O
O
O
R1
R2
H
H
H
H
O
O
O
O
3) Tf2O, Pyr./DCM
R3
HO
N
MeOH, AcOH
TfO
H2N
O
O
O
O
R4
2
1
3, 7 - 15
6
HN(CH3)2, THF
or
yield
(%)
R1
R2
R3
R4
prod.
NaN3, DMF
O
1
H
H
H
H
H
H
H
Me
Me
Et
H
3
97
95
93
92
33a
96
O
2
3
4
5
6
Me
Et
Et
7
O
OH
OH
O
HO
HO
2M HCl
H
O
n-Pr
n-Bu
n-Bu
Bn
8
X
O
n-Pr
n-Pr
Ph
H
9
N
X = N(CH3)2
X = N3
3
5
6
n-But
10
11
12
4
LAH,
Et2O
H
X = NH2
(3-OMe,4-
OH)Ph
(3-OMe,4-
OH)Bn
7
H
H
94
ribosome, and is also acetylated by AMEs.4 Modification of
this ring was the central aspect of work employing a
glycodiversification approach.11 A library of kanamycin B
derivatives varying the placement and number of the amino
groups on the C ring reinforced the placement of the amino
group at the natural C3 position. The preparation of
glycosylated small molecule compound libraries was essential
to these studies. For antibiotics in particular, where the
presence of glycosylated compounds is prevalent, synthesis
presents unique challenges.
Leading methods for the preparation of glycosylated small
molecule library include enzymatic glycorandomization and
chemoselective neo-glycorandomization.12,13 Both methods
have been useful, though onerous in their execution. Previous
work in our lab on glycosylated small molecule library
synthesis leveraged a strategy where simple glycosides
containing functionalized aglycones were elaborated in
chemoselective reactions. We dubbed the approach post
glycosylation diversification, or PGD.14 We used desosamine,
the amino sugar that anchors erythromycin in the ribosome
exit tunnel, to prepare glycosides containing aglycones with
amino, azido, or aryl halide moieties. Small libraries prepared
from the PGD precursors were shown to inhibit protein
synthesis in an in vitro assay and similarly inhibit bacterial
growth.15 In the course of that study, we became interested in
changing the sugar itself, which led us to the synthesis of
alkylated kanosamines and their glycosides described herein.
Kanosamine analogs were attractive targets because of their
additional functionality relative to desosamine. We
envisioned starting with N,N-dimethyl kanosamine, whose
structure formally adds a hydroxyl group at C6 relative to
desosamine. The C6 hydroxyl group could ultimately be an
additional site of diversification. The initial objective,
however, was a synthetic route that provided access to the
N,N-dimethyl compound, as well as other N-substituted
kanosamines. To that end, we recognized the utility of known
triflate 2, itself obtained from the epimerized product of 1
(Scheme 1) as a useful intermediate. Treatment of 2 with an
8
Me
Ph
Me
Me
i-Pr
H
H
H
13
14
15
91
96b
9
1-PhEt
Cylohexyl
10
Cyclohexanone
92
aIsolated as sole product.
bCombined yield of separable diastereomers
over two steps. Optimization of the amine equivalents, triflate
concentration, and volume of the pressure vessel were all
required to obtain this yield. The challenges encountered
when using other amine partners required a different
synthetic route if additional analogs were to be obtained.
We turned, therefore, to reductive amination in place of
triflate displacement. Reductive amination using 6 as the
amine presented itself as a robust and facile method. Synthesis
of 6 followed a previously established synthetic route to
kanosamine.16 Epimerization of diacetone D-glucose 1,
through an oxidation-reduction sequence, gave the C3-
epimerized allo-configured product in 86% yield over two
steps. Treatment with triflic anhydride and pyridine then
converted it to triflate 2. Without further purification, this
compound was then subjected to nucleophilic displacement
using sodium azide to give intermediate 5 (96% over two
steps) followed by reduction with LiAlH4 yielding amine 6,
diacetone D-kanosamine, as a single diastereomer in 86%
yield. To test the reductive amination, amine 6 was reacted
with three equivalents of both formaldehyde and sodium
cyanoborohydride. TLC indicated complete consumption of
the starting material after two hours. Workup and purification
of the crude product via column chromatography gave N,N-
dimethyl kanosamine derivative 3 in near quantitative yield.
Following on the heels of the successful preparation of 3 by
reductive amination, we aimed to extend the methodology to
other N-alkylated derivatives. The scope was tested using as
series of aldehydes and ketones (Table 1). To our delight, the
reactions proceeded in very good yields (91-97%), similar to
the initial reaction with formaldehyde. Di-alkylation products
were observed in the case of small aldehydes such as

formaldehyde and acetaldehyde (entries 1-2), which gave
products 3 and 7. Longer chain aldehydes primarily gave
mono-alkylation products. Propionaldehyde (entry 3) gave 8
whereas
sequence with formaldehyde to deliver products 17-21 in 91-
97% yield (entries 2-6). Deprotection of purified alkylated
acetonide intermediates (3, 7-21) was accomplished by
treatment with 2M HCl. Evaporation of the acidic aqueous
solution upon completion gave rise to the corresponding
hydrochloride salts 4 and 22-37 in quantitative yields (Eq. 1).
Table 2. Formation of N,N-dialkyl kanosamines via
reductive aminations
A telescoped sequence would eliminate the need for
isolating the protected acetonide intermediate, and facilitate
the formation of a pyranose suitable for further manipulation,
starting from our key intermediate 6. To illustrate the
strategy, 6 was alkylated with vanillin, then acetaldehyde, to
form intermediate 38 (Scheme 2). Addition of HCl to the
reaction
O
O
O
i)
, NaBH3CN
R1
R2
O
O
O
O
H
H
MeOH, AcOH
O
O
Me
N
H2N
ii) formaldehyde
O
O
R3
6
16 - 21
Scheme 2. One pot modification to N,N-dialkyl
kanosamine peracetates
yield
(%)
R1
R2
R3
prod.
1
Et
H
H
H
n-Prop
n-But
Bn
16
17
18
19
97
95
93
O
2
3
n-Prop
O
O
O
Ph
O
i) vanillin, Na(CN)BH3
ii) acetaldehyde
H
O
O
H
3-OMe,4-
OH)Ph
(3-OMe,4-
OH)Bn
O
N
4
H
92
O
H2N
O
5
Me
Me
i-Prop
20
21
91
OH
6
6
Cyclohexanone
Cylohex
92
OMe
38
O
2M
HCl
O
O
O
OH
OH
HO
HO
R1
2M HCl
H
O
(Eq. 1)
R1
N
O
OH
OH
O
OAc
OAc
O
HO
HO
N
AcO
AcO
.HCl
R2
R2
N
N
.HCl
7 - 21
4, 22 - 37
Ac2O, Pyr.
80% (4 steps)
butyraldehyde (entries 4-5) gave both mono- (9) and di-
alkylated (10) products in separate runs, where di-alkylated
product 10 was the only exception to the trend. The
methodology was further extended to aromatic aldehydes
such as benzaldehyde or vanillin, yielding products 11 and 12.
Reactivity with ketones proceeded to give mono-alkylation
products (entries 8-10), giving products 13-15. The observed
mono alkylations are most likely due to steric interactions
between the substrate and aldehydes/ketones that prevents
di-alkylation.
O
O
OH
OAc
40
39
Table 3. Glycosylations using compound 41 as donor
O
Br
O
OR
ROH
Conditions
AcO
AcO
AcO
AcO
OAc
OAc
N
N
42 - 45
41
prod.
We next considered exploiting the mono-alkylation selectivity
to prepare disubstituted analogs containing two different
alkyl groups. The logic was to do serial alkylations in one pot
- first with a large aldehyde or ketone, and then a small one to
give the mixed di-alkyl product. To test the approach, 6 was
reacted with 1.5 equivalents of propionaldehyde under the
established reaction conditions, monitored by TLC. Upon
consumption of 6 and formation of a less polar spot, excess
formaldehyde was added. After a short period, a new spot
appeared on TLC, accompanied with the disappearance of the
mono-alkylated product. Work up and purification gave N-
methyl,N-propyl derivative 16 in high yield (97%) with no
evidence of N,N-dimethyl or N-propyl side products (Table 2).
The scope of this methodology to form mixed alkyl derivatives
was then extended as summarized in Table 2. Other carbonyl
partners such as butyraldehyde, benzladehyde, vanillin,
cyclohexanone, and acetone were utilized in the reaction
alcohol
(ROH)
yield
(%)
conditions
()
1
allyl
Ag2CO3, 4Å MS
Ag2CO3, 4Å MS
Ag2CO3, I2, 4Å MS
42 (6:1)
43 (3:1)
49
46
2
3
Cl-ethyl
4-Br-benzyl
44 (>20:1) 55
4
3-Br-benzyl
Ag2CO3, I2, 4Å MS
45 (20:1) 48
mixture at this point served two purposes. It quenched the
residual borohydride from the reductive amination and it also
facilitated acetonide deprotection, giving lactol 39, which was
not isolated. Workup to remove unreacted aldehyde was
followed by acetylation and purification by chromatography
to deliver 40 in 80% yield. Remarkably, the sequence included
four reactions with only one chromatography.

Having secured a route to N-alkylated kanosamines, we
endeavored to incorporate one into a Kanamycin B analog.
Initial efforts were focused on glycosylation onto a neomycin-
derived acceptor.11 A per-O-acetylated thiopyrimidine donor
was prepared starting from N,N-dimethyl kanosamine 4 and
conditions were screened for its attachment onto a Neamine
acceptor, but without success.17 It was likely that the
combination of basic amine functionality and acetate
protecting
Scheme 3. Post glycosylation diversification of kanosamine analogs
1) CuSO4, ascorbate
Ph
1) Pd(0), Mes-B(OH)2
base
Br
H
O
N
O
O
N
O
O
O
N
O
N
O
HO
HO
N
AcO
AcO
AcO
AcO
N3
HO
HO
N
n-butanol, H2O
2) NaOCH3, HOCH3
OH
OAc
OAc
OH
2) NaOCH3, HOCH3
N
73% (2 steps)
p-mesityl 46 82% (2 steps)
m-mesityl 47 72% (2 steps)
p-Br 44
m-Br 45
48
49
We thank A.M. Shaqra and V.L. Robinson (U. Connecticut)
for assistance with the transcription-translation assay.
groups deactivate the donor, preventing glycosylation.
These effects were compounded by a weakly nucleophilic
acceptor. N,N- dimethyl kanosamine 4 was alternatively
converted to anomeric bromide 41using HBr in acetic acid.
As shown in Table 3, 41 served as a viable donor, delivering
glycosides 42-45 in moderate yields (46-55%). The new
glycosides contained latent reactivity handles that were
exploited in post glycosylation diversification reactions.
As a proof of concept, we conducted Suzuki couplings on
bromobenzyl glycosides 44 and 45 to prepare, after de-
acetylation, compounds 46 and 47 in 82 and 72% yields,
respectively (Scheme 3). Chloroethyl glycoside 43 was
converted to the azido ethyl compound 48 by substitution,
and was then subjected to a click reaction with phenyl
acetylene to give, after de-acetylation, compound 49 in
72% yield over two steps. A cell free, in vitro transcription-
translation assay that monitors the fluorescence of EmGFP
was used to evaluate new compounds 46, 47, and 49.14
Compounds 46 and 49 exhibited inhibition of
approximately 70% relative to negative control of protein
synthesis in the absence of inhibitor. The m-mesityl
compound 47 reduced protein synthesis by ~40%.17 These
results show that the PGD methodology identified new
protein synthesis inhibitors that are amenable to
modification to improve pharmacological properties.
References
1. C.L. Ventola P & T 40 (2015) 277.
2. S. Zaman, S Bin; Hussain, M. A.; Nye, R.; Mehta, V.; Mamun,
K. T.; Hossain, N. A Cureus 9 (2017) e1403.
3. J.M.A. Blair, M.A. Webber, A.J. Baylay, D.O. Ogbolu, L.J.V.
Piddock, Nat. Rev. Microbiol. 13 (2015) 42.
4. K.J. Labby, S. Garneau-Tsodikova, Future Med. Chem. 5 (2013)
1285.
5. M.E. Falagas, A.P. Grammatikos, X. Michalopoulos, Exp. Rev.
Anti. Infect. Ther. 6 (2008) 593.
6. E. Durante-Mangoni, A. Grammatikos, R. Utili, M.E. Falagas,
Int. J. Antimicrob. Agents 33 (2009) 201.
7. B. Francois, R.J.M. Russell, J.B. Murray, F. Aboul-ela, B.
Masquida, Q. Vicens, E. Westhof, Nucleic Acids Res. 33 (2005)
5677.
8. M. Kulik, A.M. Goral, M. Jasiński, P.M. Dominiak, MJ. Trylska,
Biophys. J. 108 (2015) 655.
In conclusion, we have presented a method for the
preparation of alkylated kanosamine sugars from
diacetone D-glucose. The robust and versatile method
allows mono, di- and mixed alkyl patterns to be accessed
in a single pot fashion. In addition, the method allows
taking the diacetonide starting material to a peracetylated
pyranose, with a single purification, over a four-step
process. While efforts to incorporate the sugars yielded
from this process into Kanamycin B derivatives fell short,
the PGD platform successfully generated compounds that
inhibited bacterial protein synthesis.
9. R.-B.Yan, M. Yuan, Y. Wu, X. You, X.-S. Ye, Bioorg. Med.
Chem. 19 (2011) 30.
10. W. Zhang, Y. Chen, Q. Liang, H. Li, H. Jin, L. Zhang, X. Meng,
Z. Li, J. Org. Chem. 78 (2013) 400.
11. J. Li, J. Wang, P.G. Czyryca, H. Chang, W.W. Orsak, R.
Evanson, C.-W.T. Chang, Org. Lett. 6 (2004) 1381.
12. R. W. Gantt, P. Peltier-Pain, J.S. Thorson, Nat. Prod. Rep. 28
(2011) 1811.
13. R.D. Goff, J.S. Thorson, Med. Chem. Comm. 5 (2014) 1036.
14. Z. Cannone, A.M. Shaqra, C. Lorenc, L. Henowitz, S.
Keshipeddy, V.L. Robinson, A. Zweifach, D. Wright, M.W.
Peczuh, ACS Comb. Sci. 21 (2019) 192.
Supplementary data
Supplementary data for this article is available free of charge
at URL. The data include experimental details,
characterization data, and spectra for new compounds.
15. J.A. Dunkle, L. Xiong, A.S. Mankin, J.H.D. Cate, Proc. Natl.
Acad. Sci. 2010, 107, 17152.
16. J. Guo, J.W. Frost, J. Am. Chem. Soc. 124 (2002) 10642.
Acknowledgements

17. See Supplementary data for details.
5

A method for preparing N-alkylated kanosamines from
diacetone D-glucose
Zachary P. Cannone and Mark W. Peczuh*
Department of Chemistry, University of Connecticut, 55 N. Eagleville Road, U3060, Storrs, CT, 06269 USA
TOC graphic:
O
N
O
N
HO
HO
N
N
O
O
OH
O
N
OH
OH
PGD
HO
HO
R‘
O
R
H
O
H
O
H2N
O
O
N
O
R’
HO
HO
OH