A broad spectrum anticancer nucleoside with selective toxicity against human colon cells in vitro

Article abstract of DOI:10.1016/j.bmcl.2011.01.003

2′,3′-Bis-O-tert-butyldimethylsilyl-5′-deoxy-5′-[N- (methylcarbamoyl)amino]-N⁶-(N-phenylcarbamoyl)adenosine, a new member of the N⁶,5′-bis-ureidoadenosine class of anticancer nucleosides, is found to exhibit broad spectrum antiproliferative activity. A majority of the cell lines in the NCI-60 are inhibited with an average GI ₅₀ = 3.13 μM. Selective toxicity against human colon cancer cell lines (COLO 205, HCC-2998, HCT-116, HT29, KM12) was also exhibited (LC ₅₀'s = 6-10 μM).

Full text of DOI:10.1016/j.bmcl.2011.01.003

Bioorganic & Medicinal Chemistry Letters 21 (2011) 1484–1487
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
A broad spectrum anticancer nucleoside with selective toxicity against
human colon cells in vitro
⇑
Jadd R. Shelton, Scott R. Burt, Matt A. Peterson
Department of Chemistry and Biochemistry, Brigham Young University, Provo, UT 84602, United States
a r t i c l e i n f o
a b s t r a c t
20,30-Bis-O-tert-butyldimethylsilyl-50-deoxy-50-[N-(methylcarbamoyl)amino]-N6-(N-phenylcarbamoyl)-
adenosine, a new member of the N6,50-bis-ureidoadenosine class of anticancer nucleosides, is found to
exhibit broad spectrum antiproliferative activity. A majority of the cell lines in the NCI-60 are inhibited
Article history:
Received 19 November 2010
Revised 29 December 2010
Accepted 3 January 2011
Available online 12 January 2011
with an average GI50 = 3.13
l
M. Selective toxicity against human colon cancer cell lines (COLO 205, HCC-
M).
2998, HCT-116, HT29, KM12) was also exhibited (LC50’s = 6–10
l
Ó 2011 Elsevier Ltd. All rights reserved.
Keywords:
Purine nucleosides
Bio-active adenosine derivatives
Antiproliferative nucleosides
Recently, we reported the synthesis and biological evaluation of
a new class of antiproliferative bis-ureidoadenosine derivatives
typified by compounds 1–3 (Fig. 1).1 A preliminary SAR study re-
vealed that substitution in the NE, NW, and SE quadrants of these
O
N
O
Ph
Ph
HN
N
N
H
HN
N
N
H
N
N
N
N
N
O
O
CH3
CH3
molecules was essential for optimal (low
lM) antiproliferative
N
H
N
H
N
H
N
H
O
O
2
activities (Fig. 2). Left in doubt from these preliminary studies
was the essential character of the 30-C functionalization found in
the SW quadrant of compounds 1 and 2. A single-dose screen of
compound 3 suggested that the 30-C carboxymethyl moieties of 1
and 2 might not be essential for activity in this class of compounds.
Here, we report the multi-dose data from the NCI-60 screen of
compound 3, which confirms that substitution at the 30-position
with the unnatural (and synthetically challenging) 30-C carboxy-
methyl moieties of 1 and 2 is not necessary for retention of anti-
proliferative activity of the parent compounds (Table 1).2 Indeed,
compound 3 generally exhibited greater potency than either com-
pounds 1 or 2, and also exhibited selective toxicity toward several
of the cells in the NCI-60 panel (Table 2). In contrast, compounds 1
and 2 were not as potent and did not exhibit any significant degree
of selective toxicity (Table 2). With only a few exceptions, LC50 val-
O
OTBS
1
O
OTBS
OEt
N
CH3O
CH3
O
N
O
N
Ph
Ph
HN
N
H
HN
N
H
N
N
N
N
O
O
CH3
CH3
N
N
N
H
N
H
N
H
O
O
O
TBSO
OTBS
3
TBSO
OTBS
4
Figure 1.
The synthesis of compound 3 required six steps from commer-
cially available adenosine (Scheme 1).1 In an effort to reduce the
number of steps and to test the antiproliferative activity of the
isosteric, yet synthetically more tractable, 50-N-methylcarbamoyl
derivative, compound 4 was also prepared and tested. Its synthesis
required four steps and gave compound 4 in 61% overall yield from
adenosine.4
Unfortunately, and perhaps not too surprisingly given that com-
pound 4 has one less potential hydrogen bond donor relative to
compound 3, the single-dose NCI-60 screening data indicated that
compound 4 had substantially inferior antiproliferative activity in
the single-dose growth inhibition assay (Table 1).
ues for compounds 1 and 2 were >100
selective toxicity was apparent. Compound 3 on the other hand
had LC50’s in the 6–10 M range for five of the six human colon
cancer cell lines, and three of the renal cell lines exhibited similar
levels of toxicity. The average GI50 for compound 3 was 3.13 M,
lM, and no clear trend for
l
l
compared to 7.57 and 17.2 lM for compounds 1 and 2, respectively
(Table 1).3 Thus compound 3 exhibited both greater potency and
selective toxicity than either lead compounds 1 or 2.
⇑
Corresponding author.
E-mail address: matt_peterson@byu.edu (M.A. Peterson).
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.01.003

J. R. Shelton et al. / Bioorg. Med. Chem. Lett. 21 (2011) 1484–1487
1485
Table 1
N
Results of growth inhibition assays (GI50, l
M)a (growth percent)b
NE
O
N
1a
2a
3a
3b
4b
Ph
HN
N
Leukemia
CCRF-CEM
HL-60(TB)
K-562
MOLT-4
RPMI-8226
H
N
N
6.69
3.01
3.59
2.39
1.09
6.37
3.59
—
—
—
3.27
43
24
34
24
20
—
NW
CH3
O
1.81
3.12
2.13
1.58
113
75
75
88
N
N
H
N
H
O
E
W
SE
O
OTBS
Non-small cell lung cancer
A549/ATCC
EKVX
HOP-62
HOP-92
NCI-H226
NCI-H23
NCI-H322M
NCI-H460
NCI-H522
SW
OEt
4.18
17.7
8.96
< 0.01
>100
33.3
>100
5.54
4.36
9.35
26.4
24.9
2.71
41.9
57.2
>100
7.49
11.1
2.90
3.70
4.98
1.56
3.10
3.17
6.86
1.72
2.39
23
23
62
ꢀ3
49
36
87
ꢀ7
46
81
1
104
104
55
93
98
118
103
53
S
Figure 2.
Colon cancer
COLO 205
HCC-2998
HCT-116
HCT-15
HT29
KM12
SW620
Although differences in pharmacological and/or biological
parameters associated with the different cell lines could be impor-
tant, the difference in antiproliferative activities between 3 and 4
could be at least partially attributable to the loss of favorable hydro-
gen bonding in compound 4 relative to compound 3 and their (as-
sumed) mutual biological receptor(s). A classic example of the
effect of an NH to O substitution is seen in the development of Van-
3.84
>100
3.20
8.50
4.20
3.95
4.80
12.3
30.6
4.20
6.47
5.37
23.9
>100
1.89
2.02
2.11
3.62
2.02
2.13
—
ꢀ100
ꢀ1
0
106
—
67
103
81
110
98
21
ꢀ68
ꢀ4
36
Melanoma
LOX IMVI
MALME-3M
M14
SK-MEL-2
SK-MEL-28
SK-MEL-5
UACC-257
UACC-62
comycin resistance when
(O) in the peptidoglycan cell-wall precursor of Gram-positive bacte-
ria.5 The NH to O substitution in the
-lys- -ala- -ala to -lys- -ala-D-
D-alanine (NH) is substituted by D-lactate
5.46
10.3
2.51
5.42
6.85
4.34
5.68
>100
7.30
11.4
15.2
14.9
7.77
5.81
22.6
41.9
1.82
2.98
4.33
2.45
—
1.98
3.78
3.98
13
49
33
57
60
48
43
42
87
114
102
83
102
101
—
L
D
D
L
D
lac mutation causes a loss of binding affinity of at least 1000-fold be-
tween Vancomycin and its wild-type ligand. In a similar, but less
profound manner, loss of an NH moiety in compound 4 relative to
3 could contribute to decreased binding between 4 and its biological
receptor. Another factor that could influence decreased binding of 4
is the loss of conformational rigidity resulting from the absence of an
intramolecular hydrogen bond between N3 and the 50-NH of the N-
methylureido moiety. Such loss of conformational rigidity would
be expressed as an increased syn/anti glycosyl fluctionality that
could lead to decreased binding of compound 4 relative to the more
conformationally rigid compound 3.6
106
CNS cancer
SF-268
SF-295
SF-539
SNB-19
SNB-75
U251
6.53
5.73
5.19
29.0
4.56
4.69
8.29
9.09
22.3
>100
12.7
5.66
3.58
3.84
3.07
4.57
2.43
2.96
36
45
33
76
46
17
85
108
97
114
77
86
Ovarian cancer
IGROV1
Intramolecular hydrogen bonds between N3 and 50-hydrogen-
bearing substituents have been reported to impart significant con-
formational rigidity to adenosine derivatives that otherwise exist
in dynamic syn/anti glycosyl equilibria.7 In one such notable case,
intramolecular hydrogen bonding between the 50-OH and N3 of
1-deaza-30-O-methyladenosine locks the nucleoside in the syn con-
formation in the solid state, and 1H–1H-NOE difference spectra
suggested a high population of the syn-conformer in solution.8 In
the case of compound 3, NOESY and 3J(1H–1H) coupling data sug-
gest that compound 3 exists primarily in the syn conformation in
solution. In contrast, these data suggest that compound 4 is much
more conformationally labile (Fig. 3). It has been well documented
that for purine nucleosides, a correlation exists between the syn/
anti equilibrium about the glycosidic bond and the conformational
bias (C20-endo vs C30-endo) of the sugar. The C20-endo (S) conforma-
tion correlates with a predominant syn glycosyl conformation
3.85
4.59
12.3
31.1
4.92
21.0
3.72
7.11
53.0
38.2
9.02
52.7
2.74
2.60
3.44
5.24
4.01
6.88
53
111
107
86
120
89
OVCAR-3
OVCAR-4
OVCAR-5
OVCAR-8
SK-OV-3
ꢀ29
30
69
33
78
106
Renal cancer
786-0
A498
2.00
3.34
8.55
29.7
2.01
9.10
12.4
12.1
9.01
3.87
14.4
53.8
9.74
85.3
20.6
7.79
2.04
2.03
3.83
2.70
2.14
1.81
4.29
2.50
17
—
21
60
ꢀ37
ꢀ79
34
95
98
95
106
94
101
96
106
ACHN
CAKI-1
RXF393
SN12C
TK-10
UO-31
48
Breast cancer
BT-549
HS578T
>100
3.60
3.42
3.96
—
29.0
5.79
5.59
12.3
—
2.94
—
2.69
2.61
3.35
2.87
40
61
0
38
36
14
—
98
88
110
104
116
MCF7
(v
= 10–30°), whereas the C30-endo (N) conformation correlates
with the anti glycosyl conformation (
= 200°–210°).7 To a first
approximation, the equilibrium constant (Keq) can be calculated
MDA-MB-231/ATCC
MDA-MB-468
T-47D
v
2.55
13.9
0
0
0
0
from the observed J1 ,2 and J3 ,4 according to Eq. 1, where XS and
XN correspond to the mole fraction of S conformer and N con-
former, respectively. Equilibrium constants calculated with this
expression compare very favorably to those obtained by a full
pseudorotational analysis.7,9
Prostate cancer
DU-145
PC-3
4.97
2.25
16.6
—
2.58
2.69
52
8
106
59
a
Multi-dose growth inhibition, GI50 determined from dose–response curve.
b
Single-dose growth inhibition percent calculated as: [(Ti ꢀ Tz)/C ꢀ Tz)] ꢁ 100 for
Ti P Tz; [(Ti ꢀ Tz)/Tz)] ꢁ 100 for Ti < Tz; where Tz = absorbance at t = 0; Ti = absor-
0
0
0
0
ðKeqÞ ¼ XS=XN ¼ J1 ;2 =J3 ;4
ð1Þ
bance at t = 48 h (10 lM test compound); C = absorbance of control at t = 48 h.

1486
J. R. Shelton et al. / Bioorg. Med. Chem. Lett. 21 (2011) 1484–1487
Table 2
NH2
N
NH2
Results of multi dose growth inhibition assay (LC50, l
M)a
N
N
N
N
N
Cell line
1
2
3
N
N
N3
HO
d f
a c
O
O
Leukemia
CCRF-CEM
HL-60(TB)
K-562
MOLT-4
RPMI-8226
3
>100
>100
>100
>100
59
>100
>100
>100
>100
>100
>100
—
—
—
>100
73 %
TBSO
HO
OTBS
OH
5
66 %
NH2
N
NH2
N
Non-small cell lung cancer
A549/ATCC
EKVX
HOP-62
HOP-92
NCI-H226
NCI-H23
NCI-H322M
NCI-H460
NCI-H522
N
N
N
79.1
>100
73.1
41.2
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
48
97.6
>100
86
>100
7.11
>100
N
N
N
HO
HO
f,d
c,g
O
O
4
76%
HO
OH
TBSO
OTBS
6
80 %
O
O
N
Ph
Ph
HN
N
N
HN
N
H
H
Colon cancer
COLO 205
HCC-2998
HCT-116
HCT-15
HT29
KM12
SW620
N
N
N
N
O
O
>100
>100
45.6
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
5.98
6.99
8.8
>100
6.91
8.05
—
CH3
CH3
N
N
N
O
N
N
H
O
O
H
H
TBSO
OTBS
TBSO
OTBS
4
3
Melanoma
LOX IMVI
MALME-3M
M14
SK-MEL-2
SK-MEL-28
SK-MEL-5
UACC-257
UACC-62
>100
>100
78.6
>100
48.7
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
6.21
>100
93.8
>100
28.3
13.6
>100
>100
Scheme 1. Reagents and conditions: (a) SOCl2/CH3CN, rt, 15 h; (b) NaN3/DMF,
150 °C, 1 h; (c) TBSCl/DMF, rt, 16 h; (d) PhN@C@O/CH2Cl2, rt, 5d; (e) H2/Pd–C/
EtOAc; (f) p-NO2–C6H4OCONHCH3/base/EtOAc, rt; (g) TFA/H2O (9:1), rt, 3 h.
H
H
H
H
6''
5''
O
CNS cancer
SF-268
SF-295
SF-539
SNB-19
SNB-75
U251
m
92.5
>100
>100
>100
>100
76.0
>100
16.4
>100
>100
>100
>100
>100
>100
94.3
>100
>100
>100
2''
N
H
5
6
1''
3''
N
4''
w
7
6
H
H
H
N
1
N
2
H
H
H
8
N
2'''
N
3'''
O
N
5'
4
Ovarian cancer
IGROV1
N
H
4'''
9
CH3
1'''
3
O
79.2
91.3
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
98.2
>100
>100
>100
>100
H
H
1'
4'
s
OVCAR-3
OVCAR-4
OVCAR-5
OVCAR-8
SK-OV-3
H
H
2'
3'
3
TBSO
OTBS
H
H
H
6''
5''
w
O
Renal cancer
786-0
A498
17.9
>100
>100
>100
19.5
>100
>100
71.7
>100
>100
>100
>100
>100
>100
>100
>100
6.75
19.5
>100
>100
7.5
6.29
>100
>100
2''
N
H
4
6
3''
1''
m
N
4''
ACHN
6
7
CAKI-1
RXF393
SN12C
TK-10
H
H
5
N
N
1
N
2
H
N
H
w
8
O
3'''
O
m
N
3
H
9
CH3 2'''
UO-31
5'
1'''
O
H
H
1'
m
4'
Breast cancer
BT-549
HS578T
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
—
>100
28.3
>100
>100
H
H
3'
2'
4
TBSO
OTBS
MCF7
MDA-MB-231/ATCC
MDA-MB-468
T-47D
Figure 3. NOESY correlation data for compounds 3 and 4: (s = strong, m = medium,
w = weak).
Prostate cancer
DU-145
PC-3
12.5
78.4
>100
>100
18.7
>100
0
0
0
0
Applying the observed J1 ,2 and J3 ,4 to Eq. 1 gives Keq = 11 and
1.9 for compounds 3 and 4, respectively.10 These values correspond
to the following N/S sugar conformational preferences for
compounds 3 and 4: 3 (8% N, 92% S), 4 (34% N, 66% S). The above
a
LC50 = concentration required to reduce total cell count by 50%. Calculated as
[(Ti ꢀ Tz)/Tz)] ꢁ 100 = ꢀ50, where Tz = absorbance at t = 0; Ti = absorbance at
t = 48 h.

J. R. Shelton et al. / Bioorg. Med. Chem. Lett. 21 (2011) 1484–1487
1487
noted correlation between sugar conformations and the preferred
Supplementary data
glycosyl conformation (S correlates with syn; N correlates with
anti),7 indicates that compound 3 exists in a predominantly syn
glycosyl conformation, while compound 4 is much more conform-
ationally labile. The strong interactions between H8 and H10 as
well as between H2 and H2000 in the NOESY spectrum for compound
3 also support the conclusion that the syn conformer is preferred.
NOESY data for compound 4 show relatively weaker interactions
between H8 and H10 and stronger interactions between H8 and
H2000 and between H8 and H20, all of which point to the anti confor-
mation being more populated in compound 4 than in compound 3.
Taken together, these observations suggest that compound 3
enjoys a greater degree of conformational rigidity about the glyco-
sidic linkage. The difference in conformational bias for compounds
3 and 4 could contribute to the difference in biological activities, as
has been reported for conformationally biased nucleosides in a
number of instances.6
In summary, we have discovered a broad spectrum anticancer
nucleoside that exhibits low micromolar antiproliferative effects
against a majority of the cell lines in the NCI-60. Selective toxicity
toward colon cell lines was also exhibited. Compound 4, a synthet-
ically simpler, 50-N-methylcarbamoyl analog of 50-N-methylureido
derivative 3, was less active. The difference in activities may be due
to the loss of one potential hydrogen bond donor in compound 4
relative to 3, and/or may also be influenced by the greater confor-
mational flexibility of compound 4, as supported by NOESY and
3J(1H–1H) coupling data. The selective toxicity of compound 3
may make it useful for future clinical applications. Appropriate
derivatization of 3 could result in compounds with enhanced po-
tency and selectivity. We are currently pursuing this line of
investigation.
Supplementary data (detailed experimental procedures for
compound 4. NMR spectral data for compounds 3 and 4) associated
with this article can be found, in the online version, at doi:10.1016/
j.bmcl.2011.01.003.
References and notes
1. Peterson, M. A.; Oliveira, M.; Christiansen, M. A.; Cutler, C. E. Bioorg. Med. Chem.
Lett. 2009, 19, 6775.
2. Data from the NCI-60 screens is provided in the Supplementary data.
3. Average GI50 values are determined from NCI-60 testing results included in the
Supplementary data. They do not take into account the cell lines with GI50
>100 lM.
4. Full experimental details for all new compounds can be found in the
Supplementary data. 1H and 13C NMR and HRMS data for 4 are as follows: 1H
NMR (CDCl3, 500 MHz) d 12.25 (br s, 1H), 9.86 (br s, 1H), 8.80 (s, 1H), 8.66 (s,
1H), 7.60 (d, J = 7.5 Hz, 2H), 7.38 (t, J = 7.8 Hz, 2H), 7.16 (t, J = 7.3 Hz, 1H), 6.21
(d, J = 5.4 Hz, 1H), 5.82 (d, J = 4.0 Hz, 1H), 4.64 (t, J = 4.8 Hz, 1H), 4.50 (dd, J = 3.8,
12.8 Hz, 1H), 4.34 (t, J = 3.89 Hz, 1H), 4.31–4.29 (m, 2H), 2.47 (d, J = 5.0 Hz, 3H),
0.95 (s, 9H), 0.82 (s, 9H), 0.12 (s, 6H), 0.00 (s, 3H), ꢀ0.20 (s, 3H); 13C NMR
(CDCl3, 125 MHz) d 157.0, 153.4, 151.4, 150.8, 150.4, 143.9, 137.8, 129.3, 124.7,
121.6, 120.6, 87.8, 84.5, 77.4, 72.9, 63.6, 29.9, 27.1, 26.0, 25.9, 18.2, 18.0, ꢀ4.29,
ꢀ4.61,
ꢀ4.72,
ꢀ5.15;
MS
(FAB)
m/z
672.3356
(MH+
[C31H50N7O6Si2]) = 672.3353.
5. Dancer, R. J.; Try, A. C.; Sharman, G. J.; Williams, D. H. J. Chem. Soc., Chem.
Commun. 1996, 1445.
6. For excellent discussions of the impact of nucleoside conformation on
biological activities, see: (a) Mathé, C.; Périgard, C. Eur. J. Org. Chem. 2008,
1489; (b) Ford, H., Jr.; Dai, F.; Mu, L.; Siddiqui, M. A.; Niklaus, M. C.; Anderson,
L.; Marquez, V. E.; Barchi, J. J., Jr. Biochemistry 2000, 39, 2581; (c) Mu, L.;
Sarafianos, S. G.; Nicklaus, M. C.; Russ, P.; Siddiqui, M. A.; Ford, H., Jr.; Mitsuya,
H.; Le, R.; Kodama, E.; Meier, C.; Knispel, T.; Anderson, L.; Barchi, J. J., Jr.;
Marquez, V. E. Biochemistry 2000, 39, 11205; (d) Eoff, R. L.; McGrath, C. E.;
Maddukuri, L.; Salamanca-Pinzon, S. G.; Marquez, V. E.; Marnett, L. J.;
Guengerich, F. P.; Egli, M. Angew. Chem., Int. Ed. 2010, 49, 7481.
7. Davies, D. B. Nucl. Magn. Reson. Spectrosc. 1978, 12, 135.
8. Seela, F.; Debelak, H.; Reuter, H.; Kastner, G.; Mikhailopulo, I. A. Nucleosides
Nucleotides 1998, 17, 729.
9. Davies, D. B.; Danyluk, S. S. Biochemistry 1974, 13, 4417.
Acknowledgment
10. Coupling constants for compounds 3 and 4 were as follows: compound 3
0
0
0
0
0
0
0
0
(J1 ,2 = 7.8 Hz,
J3 ,4 = 0.7 Hz);
4
(J1 ,2 = 5.43 Hz,
J3 ,4 = 2.83 Hz).
See
The NCI Developmental Therapeutics program is thanked for
performance of antiproliferative and cytotoxicity evaluations.
Supplementary data for complete list of coupling constants for compounds 3
and 4.