Synthesis of 2,4,5-triaminocyclohexanecarboxylic acid as a novel 2-deoxystreptamine mimic

Article abstract of DOI:10.1016/j.tetlet.2010.01.111

Since aminoglycosides have been known to exert their antibacterial activities by binding to various RNA targets, 2-deoxystreptamine served as a particularly important model in understanding RNA-small molecule interaction. Herein, 2,4,5-triaminocyclohexanecarboxylic acid was prepared as a novel 2-deoxystreptamine (2-DOS) mimic, which replaces highly flexible glycosidic bonds of aminoglycosides with amide bonds and may improve binding selectivity toward RNA targets through increased rigidity and additional hydrogen-bonding capability. This unnatural g-amino acid can also be used as a potential component for synthetic foldamers.

Full text of DOI:10.1016/j.tetlet.2010.01.111

Tetrahedron Letters 51 (2010) 1779–1781
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Synthesis of 2,4,5-triaminocyclohexanecarboxylic acid as a novel
2-deoxystreptamine mimic
Sarah Roberts, Maruthi Chittapragada, Krishnaiah Pendem, Brandon J. Leavitt, John W. Mahler,
*
Young Wan Ham
Department of Chemistry and Biochemistry, Brigham Young University, Provo, UT 84057, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Since aminoglycosides have been known to exert their antibacterial activities by binding to various RNA
targets, 2-deoxystreptamine served as a particularly important model in understanding RNA-small mol-
ecule interaction. Herein, 2,4,5-triaminocyclohexanecarboxylic acid was prepared as a novel 2-deoxy-
streptamine (2-DOS) mimic, which replaces highly flexible glycosidic bonds of aminoglycosides with
amide bonds and may improve binding selectivity toward RNA targets through increased rigidity and
additional hydrogen-bonding capability. This unnatural g-amino acid can also be used as a potential com-
ponent for synthetic foldamers.
Received 21 December 2009
Revised 25 January 2010
Accepted 28 January 2010
Available online 2 February 2010
Ó 2010 Elsevier Ltd. All rights reserved.
Aminoglycosides, a large class of clinically significant antibiot-
ics, are known to selectively bind to various RNA targets. They
are comprised of 2-deoxystreptamine (2-DOS), a highly conserved
central aminocyclitol scaffold, and up to four aminosugar subunits.
While established as a particularly important model in under-
standing the principles that decree RNA recognition processes by
small molecules, aminoglycosides have shown highly promiscuous
binding patterns. They bind to many different RNA secondary
structural elements, such as bulges, internal loops, and stem junc-
tions. For instance, neomycin has been shown to bind to bulge re-
gions of unrelated RNA sequences from the 16S ribosome,1 HIV
TAR,2 HIV RRE,3 and Group I intron4 with affinities in the low
micromolar range.
This lack of RNA target selectivity of aminoglycosides is largely
attributed to their conformational adaptability rendered by flexible
glycosidic bonds, which allows aminoglycosides to adjust to fit into
different shapes of RNA cavities. Tor5 and Mobashery6 groups re-
ported conformationally constrained aminoglycoside derivatives
to reduce the number of available conformations of these amino-
glycosides and therefore increase the target selectivity.
RNA recognition and biological activity and its omni existence in
all clinically important aminoglycosides, mimicking the 2-DOS
structure has been of great interest in preparing novel aminoglyco-
side analogs.9
Here in this Letter, 2,4,5-triaminocyclohexanecarboxylic acid 1
was prepared as a novel 2-DOS mimetic by replacing 4-OH and
6-OH of 2-DOS with amino and carboxylic acid functionalities,
respectively (Fig. 1b).
This mimic will enable the introduction of rigid amide backbone
in the place of the highly flexible glycosidic bonds, which may help
improve sequence specificity in binding by (1) promoting addi-
tional hydrogen bonding and (2) limiting structural adaptation of
the ligands to different RNA targets. The 4-OH and 6-OH are chosen
for the conversion into an amine and a carboxylic acid to facilitate
Although aminoglycosides as a whole demonstrate poor bind-
ing selectivity, it is well established that 2-DOS, the main compo-
nent of aminoglycoside, recognizes 50-GU-30 sequence steps in a
sequence-specific manner regardless of the number of aminosugar
substituents and substitution positions (Fig. 1a).7,8 Puglisi and co-
workers demonstrated that the 2-DOS itself, with no aminosugar
attached, recognized 50-GU-30 base steps, although its binding
affinity was low (ꢀ1 mM).8 Due to the crucial role of 2-DOS in
Figure 1. (a) Schematic diagram of 5’-GU-3’ recognition of RNA targets by 2-DOS.
(b) Rational design of 2,4,5-triaminocyclohexanes-carboxylic acid derivative (1) as a
novel 2-DOS mimic.
* Corresponding author. Tel.: +1 801 422 1939; fax: +1 801 422 0153.
E-mail address: yham@chem.byu.edu (Y.W. Ham).
0040-4039/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2010.01.111

1780
S. Roberts et al. / Tetrahedron Letters 51 (2010) 1779–1781
NH
CCl3
(1)
O
MeO
cat. CSA, CH2Cl2, rt
N3
N3
N3
N3
(1) TPP, Pyr., NH4OH
(2) Boc2O, NaHCO3,
H2O, THF
(2) LiOH, H2O, MeOH, 0 ºC
HO
OAc
PMBO
OH
2
3
BocHN
PMBO
NHBoc
BocHN
PMBO
NHBoc
OMe
BocHN
PMBO
NHBoc
OH
(1) Swern oxid.
X
H
(2) Ph3PCH2OMe,
LiHMDS, THF
O
6
4
5
Scheme 1. Initial attempt to introduce a carboxylic acid group on the 6-OH position of the 2-DOS.
amide backbone as they point in the same direction as the RNA he-
lix axis. The 5-OH is omitted in this mimic for the sake of simplicity
in the synthesis because it points out of the helix axis and does not
extend aminosugar substituents in the same direction as the RNA
helix axis.
tion (70 °C). Thus, the reaction was quenched after 24 h and signif-
icant amount of the starting material was not fully converted into
methyl ester 14b. As a result, MOM-deprotected product 14a was
isolated usually in 5–15% yield, which was esterified in the same
reaction condition or in cat. concd H2SO4 in MeOH. The latter reac-
tion condition required prolonged reaction time (3–4 days) at 70 °C
for a comparable result.
We have chosen the asymmetrically protected precursor 2 as
the starting compound and it was prepared following the previ-
ously reported literature procedure.10 Introduction of the carbox-
ylic acid was initially performed after the procedures reported by
Hermann and co-workers11 (Scheme 1). The 4-OH of compound 2
was protected with PMB in acidic condition followed by hydrolysis
of the acetyl group. The azides were then transformed to Boc-pro-
tected amine 4 by Staudinger reaction and subsequent treatment
with Boc2O. After oxidation of the 6-OH, the resulting ketone was
treated with triphenylphosphinemethoxymethyl chloride/LiHMDS
to prepare vinyl ether 5, which was intended for the transforma-
tion to an aldehyde 6 as a precursor to a carboxylic acid. However,
the Wittig reaction failed to yield the desired vinyl ether 5 even in
the presence of large excess of ylide (7–14 equiv).
An alternative approach was attempted to introduce a nitrile as
a precursor to an acid (Scheme 2). The equatorial hydroxyl group of
compound 3 was converted in various reaction conditions to com-
pound 7 that has a leaving group at the axial position. However,
SN2 reaction to an equatorial nitrile resulted in the decomposition
of the starting material.
Then, an attempt was made to introduce the nitrile group at the
axial position from which an axially positioned aldehyde (13) or
ester (14b) can be prepared followed by epimerization to the de-
sired equatorial position (Scheme 3). The axial nitrile 11 was pre-
pared in high yield through MOM protection of the hydroxyl
group of compound 2, saponification of the ester 9, activation of
the alcohol (10) to a triflate, and SN2 reaction with NaCN. Then,
conversion of nitrile (11) to an aldehyde (13) was attempted using
DIBAL-H after reducing two azides of compound 9 and protecting
the subsequent amines with Boc. However, transformation to alde-
hyde 13 was not successful using various reaction conditions.
Application of the same reaction condition to azide 11 led to the
decomposition of the starting material. Therefore, an esterification
route was pursued. Nitrile 11 was converted to methyl ester 14b
using HCl/MeOH solution (5–7 M) in about 80% yield. Slow decom-
position of the starting compound was observed when the reaction
was allowed to run for a period beyond 24 h at the reaction condi-
Then, the two azides of compound 14b were transformed to
Boc-protected amine 15 before epimerization with strong bases
to avoid the elimination reaction of the azide located at b-position
to the methyl ester. The best epimerization result to the more
stable epimer 16 was obtained with NaOMe at 40 °C. No epimer-
ization product was found at room temperature and the Boc-pro-
tecting groups became decomposed at higher temperatures
(>50 °C). After 6 h at 40 °C (30 equiv of NaOMe in MeOH), equilib-
rium was established between epimers 15 and 16 with 1:15
respective ratio as determined by 1H NMR. Epimer 16 was isolated
in 75–85% yield after silica gel column chromatography. Initially,
several different literature-reported conditions were employed
for the epimerization using strong bases such LDA, LiHMDS, and
KHMDS.12 These strong bases either yielded no product at low
temperature or led to decomposition of the starting material at
an elevated temperature. NaH (5–10 equiv) has shown the forma-
tion of the product at 40 °C, but only as a minor product. Mixtures
of undesired byproducts dominated while no change in TLC was
found at room temperature.
The configurations of the two epimers 15 and 16 were con-
firmed by high-resolution NMR studies (Fig. 2). The coupling con-
stants between the neighboring Hs provide clear evidence for the
proper stereochemistry for both 15 and 16. In the case of 15, the
coupling constants were found to be 4, 5, and 8 Hz, respectively,
for J(Ha–Hb), J(Ha–Hc), and J(Ha–Hg), consistent with the axial
conformation of the methyl ester. Similarly, the coupling constants
for epimer 16 were found to be 4, 12, and 12 Hz, respectively, for
J(Ha0–Hb0), J(Ha0–Hc0), and J(Ha0–Hg0) confirming the equatorial
conformation of the methyl ester. The axial/equatorial conforma-
tions of 15 and 16 were further confirmed by NOE study among
the axial protons. While 16 showed NOEs among three axial pro-
tons (Hc0, He0, and Hg), 15 showed NOEs between two protons
(Hd and Hf).
After the successful introduction of an ester functional group
onto the ring system, two consecutive Mitsunobu reactions were
Scheme 2. Attempts to introduce a nitrile in an equatorial position as the acid precursor.

S. Roberts et al. / Tetrahedron Letters 51 (2010) 1779–1781
1781
MOMCl, NEt3
N3
N3
N3
N3
LiOH
N3
N3
TBAI, DMF
70 ºC, 20 min
96%
MeOH/H2O
0 ºC, 90%
HO
OAc
MOMO
OH
MOMO
OAc
2
10
9
(1) Tf2O, pyr.,
CH2Cl2, -20 ºC
N3
N3
BocHN
MOMO
NHBoc
CN
(1) PPh3, H2O/MeOH
(2) Boc2O, NEt3
MeOH, 82%
(2) NaCN, DMF;
93% in two steps
MOMO
CN
12
11
HCl/MeOH
overnight
70 ºC
DIBAL-H
X
BocHN
MOMO
NHBoc
CHO
N3
N3
N3
N3
+
HO
CN
HO
CO2Me
13
14a
5-15%
14b
80%
HCl/MeOH or conc.
H2SO4/MeOH, 70 ºC
Pd/C, H2,
Boc2O, NEt3
MeOH, 93%
BocHN
HO
NHBoc
CO2Me
BocHN
HO
NHBoc
CO2Me
NaOMe
MeOH
40 ºC, 82%
15
16
Scheme 3. Synthesis of an ester derivative 16 as a precursor to 1.
to prepare aminoglycoside analogs that have a more rigid amide
backbone than the native glycosidic linkage, which may increase
the binding selectivity for any given RNA target. The amide back-
bone may help differentiate distinctive RNA targets through extra
hydrogen-bonding capability. It may also be used to prepare novel
synthetic aminoglycosides that are less prone to bacterial resis-
tance, which is a very common resistance mechanism for amino-
glycoside-based antibiotics. In addition to being a novel 2-DOS
Figure 2. The coupling constants and NOE’s among the neighboring H’s of the two
epimers 15 and 16.
mimic, compound 1 belongs to c-amino acid and can be used to
build synthetic foldamers mimicking the structures and/or func-
tions of natural proteins. The two amines are available for further
functionalization of the foldamers.
Acknowledgments
The authors recognize BYU ORCA grant for the financial support
for Brandon Leavitt. They greatly appreciate Dr. Scott Burt for his
help with 2D NMR experiments.
Supplementary data
Scheme 4. Introduction of –NH2 in the place of –OH of 16 using two consecutive
Mitsunobu reactions.
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.tetlet.2010.01.111.
performed to introduce –NH2 in the place of the –OH of 16
(Scheme 4). Compound 17 with an axial-OH was prepared after
Mitsunobu reaction and subsequent hydrolysis of the ester in over-
all 91% yield. Another Mitsunobu reaction was performed with
PPh3, DEAD, and DPPA to prepare azide 18, a precursor to the de-
sired compound 1. In this reaction condition, the desired com-
pound (18) was isolated in moderate yield and typically 5–23% of
the starting compound (17) was recovered. When less bulky azide
sources such as hydrazoic acid and nicotinyl azide were used in the
place of DPPA in an effort to increase the reaction yield, the desired
azide 18 was obtained in lower yield due to the formation of the
undesired diastereomer. Finally, the desired 2-DOS mimic (1)
was successfully obtained after hydrogenation of azide 18.
References and notes
1. Moazed, D.; Noller, H. F. Nature 1987, 327, 389–394.
2. Wang, S.; Huber, P. W.; Cui, M.; Czarnik, A. W.; Mei, H.-Y. Biochemistry 1998, 37,
5549–5557.
3. Kwon, Y. J. Microbiol. Biotechnol. 2006, 16, 109–117.
4. von Ahsen, U.; Noller, H. F. Science 1993, 260, 1500–1503.
5. Blount, K. F.; Zhao, F.; Hermann, T.; Tor, Y. J. Am. Chem. Soc. 2005, 127, 9818–
9829.
6. Kling, D.; Hesek, D.; Shi, Q.; Mobashery, S. J. Org. Chem. 2007, 72, 5450–5453.
7. Francois, B.; Russell, R. J. M.; Murray, J. B.; Aboul-ela, F.; Masquida, B.; Vicens,
Q.; Westhof, E. Nucleic Acids Res. 2005, 33, 5677–5690.
8. Yoshizawa, S.; Fourmy, D.; Eason, R. G.; Puglisi, J. D. Biochemistry 2002, 41,
6263–6270.
9. Chittapragada, M.; Roberts, S.; Ham, Y.-W. Perspect. Med. Chem. 2009, 21–37.
10. Chenevert, R.; Jacques, F. Tetrahedron: Asymmetry 2006, 17, 1017–1021.
11. Vourloumis, D.; Takahashi, M.; Winters, G. C.; Simonsen, K. B.; Ayida, B. K.;
Barluenga, S.; Qamar, S.; Shandrick, S.; Zhao, Q.; Hermann, T. Bioorg. Med. Chem.
Lett. 2002, 12, 3367–3372.
In conclusion, 2,4,5-triaminocyclohexanecarboxylic acid (1)
was designed as a novel 2-DOS mimetic based on a structure-based
rational approach and successfully prepared in nine steps with
overall 28% yield from the known compound 2. It may be utilized
12. Klotz, P.; Mann, A. Tetrahedron Lett. 2003, 44, 1927–1930.