Synthesis and evaluation of 8-oxoadenine derivatives as potent Toll-like receptor 7 agonists with high water solubility

Article abstract of DOI:10.1016/j.bmcl.2012.11.114

We report the discovery of novel series of highly potent TLR7 agonists based on 8-oxoadenines, 1 and 2 by introducing and optimizing various tertiary amines onto the N(9)-position of the adenine moiety. The introduction of the amino group resulted in not only improved water solubility but also enhanced TLR7 agonistic activity. In particular compound 20 (DSR-6434) indicated an optimal balance between the agonistic potency and high water solubility. It also demonstrated a strong antitumor effect in vivo by intravenous administration in a tumor bearing mice model.

Full text of DOI:10.1016/j.bmcl.2012.11.114

Bioorganic & Medicinal Chemistry Letters 23 (2013) 669–672
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Synthesis and evaluation of 8-oxoadenine derivatives as potent Toll-like
receptor 7 agonists with high water solubility
Tomoaki Nakamura a, Hiroki Wada b, Hirotaka Kurebayashi a, Tom McInally b, Roger Bonnert b,
Yoshiaki Isobe a,
⇑
a Chemistry Research Laboratories, Dainippon Sumitomo Pharma. Co., Ltd, 3-1-98 Kasagade-naka, Konohana-ku, Osaka-city, Osaka 554-0022, Japan
b Medicinal Chemistry, AstraZeneca R&D Charnwood, Bakewell Road, Loughborough, Leicestershire LE11 5RH, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
We report the discovery of novel series of highly potent TLR7 agonists based on 8-oxoadenines, 1 and 2
by introducing and optimizing various tertiary amines onto the N(9)-position of the adenine moiety. The
introduction of the amino group resulted in not only improved water solubility but also enhanced TLR7
agonistic activity. In particular compound 20 (DSR-6434) indicated an optimal balance between the
agonistic potency and high water solubility. It also demonstrated a strong antitumor effect in vivo by
intravenous administration in a tumor bearing mice model.
Received 15 October 2012
Accepted 27 November 2012
Available online 5 December 2012
Keywords:
TLR7
Interferon
Innate immunology
Solubility
Ó 2012 Elsevier Ltd. All rights reserved.
Antitumor
The immune system is comprised of innate and acquired immu-
nity, both of which work cooperatively to protect the host from
microbial infections. Toll-like receptors (TLRs) are a family of type
I transmembrane receptors and are the central component of the
innate immune system. TLRs stimulate immune cells via the
MyD88-dependent interleukin-1 receptor signaling pathway. So
far, 13 subtypes of TLRs have been reported and each is known
to be fundamental in the recognition of pathogen-associated
molecular patterns.1,2 Ligation of TLRs on antigen-presenting cells,
such as dendritic cells (DCs), leads to production of proinflamma-
tory cytokines, DC maturation and priming of the adaptive immune
system.3 TLR7 is expressed by plasmacytoid dendritic cells (pDCs)
and ligand recognition leads to the secretion of type I interferon
(IFN).4
Imiquimod has been reported as the first synthetic TLR7 agonist
and preclinical studies indicate that it is likely to function through
the induction of type I IFN and IFN-inducible genes, which in turn
can have direct effects on tumor cell growth and/or harness com-
ponents of the adaptive immune system. TLR7 activation can also
enhance antitumor effects through inhibition of angiogenesis, NK
cell-mediated cytotoxicity and direct apoptosis of tumor cells.5–7
Although Imiquimod has been used to treat several dermatological
cancers, such as superficial basal cell carcinoma by topical applica-
tion, it is not applicable for other types of cancer by systemic
administration.8 852A, an Imiquimod analogue, has also been
evaluated in a phase II clinical trial for the treatment of several
types of cancer, such as ovarian cancer, lung cancer and metastatic
melanoma.9,10 Despite the positive clinical responses to 852A
observed in some patients, further development has been sus-
pended. We hypothesized that a narrow therapeutic window due
to its weak TLR7 agonistic activity (EC50 2657 nM in our screen)
may have resulted in its discontinuation of the clinical trial.
We have previously reported several 8-oxoadenines 1 and 2 as
novel structural class IFN inducers, a mechanism of action of them
was TLR7 agonism and EC50 toward TLR7 of 166 and 32 nM, respec-
tively (Fig. 1).11,12 However, these compounds were poorly soluble
in water (<1 lg/mL at pH 7.4 buffer), which would require the
co-injection of a solubilising agent for intravenous administration.
In this manuscript we describe our studies is to identify potent
TLR7 agonists with improved water solubility by the molecular
modification of the 8-oxoadenine.
In general, the aqueous solubility of small molecules depends
on their hydrophobicity. We have prepared and evaluated several
adenines which have hydrophilic group at N(9)-position, that is,
hydroxyl 3,11 carboxylic acid 4,13 dimethylamino 5,11 and methoxy
6,11 however, improved solubility was not seen. Thus, we investi-
gated the use of basic amines as a solubilising functional agents.
We prepared 8-oxobenzyladenine 7 with a dimethylamino group
on a C(2)-thio linked side chain. But the IFN inducing activity of
it was decreased significantly.14 Since an unsubstituted amino
group at the C(6)-position and an oxo group at C(8)-position of
the adenine contributed to their activity, we initiated a program
of work to introduce an amino group to the adenine N(9)-position.
⇑
Corresponding author.
E-mail address: yoshiaki-isobe@ds-pharma.co.jp (Y. Isobe).
0960-894X/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2012.11.114

670
T. Nakamura et al. / Bioorg. Med. Chem. Lett. 23 (2013) 669–672
NH2
NH2
soluble than 1 (<1 lg/mL). In contrast to the case of 7, the introduc-
NH2
N
H
N
N
N
N
N
tion of a basic amino group to the adenine N(9)-position enhanced
its TLR7 agonistic activity and this result encouraged us to opti-
mize the amino group. The piperidine 14 showed comparable
activity with 13, whereas the morpholine 15 gave a decreased
activity (EC50 = 58 nM), suggesting that stronger basicity is
required for the higher potency. Compound 14 was less soluble
than 13 probably due to its increased hydrophobicity. To increase
hydrophilicity we prepared di-amino typed compound, N-methyl-
piperazine 16 which showed a similar level of activity with 13, but
the solubility was not improved.
N
N
N
O
N
N
H
R
X
1
X=CH, R=Me
NHSO2Me
Imiquimod
852A
2 X=N, R=Me
3
4
X=CH, R=OH
X=CH, R=CO2H
NH2
5 X=N, R=NMe2
X=N, R=OMe
H
6
N
N
N
O
N
S
N
The increase of aqueous solubility might be caused by disrup-
tion of molecular symmetry.16 As shown in Table 1, the amino
group was placed onto the meta-position of the benzyl group.
meta-Substituted compounds 17 and 18 showed weaker potency
than the corresponding para-substituted compounds 14 and 16,
respectively. Addition of amino group to the N(9)-substituent of
1 enhanced both potency and solubility. We, therefore, investi-
gated this further to optimize the balance between the aqueous
solubility and the potency.
Previously we reported the structure–activity relationship of
9-pyridylmethyladenines. The order of potency was 3-pyri-
dyl > 4-pyridyl > 2-pyridyl and especially, 6-substituted-3-pyridyl
compound such as 2, 5, and 6 demonstrated highly potent IFN
inducing activity in vitro.11 Hence, we carried out the synthesis
of 6-amino-substituted-3-pyridyl compounds in an attempt to at-
tain further improvements in their profiles.
7
Figure 1. Structures of lmiquimod 852A, and compounds 1–7.
The synthetic route for N(9)-benzyladenines is shown in
Scheme 1. The reaction of 8 with 4-hydroxymethylbenzyl chloride
in the presence of potassium carbonate gave 9. Bromination at
C(8)-position of 9 was carried out with bromine in the presence
of sodium acetate and 8-oxoadenine 11 was obtained by heating
the bromide 10 in concd HCl. Chlorination of the meth-
ylbenzylhydroxyl group at the N(9)-position of the 11 with thionyl
chloride, followed by displacement of the chloro group with
corresponding amines afforded the target compounds 13–16. me-
ta-Substituted compounds 17 and 18 were also prepared by using
a similar method described above.
The results of pyridyl compound are summarized in Table 2.
We were delighted to see that, as in the case of N(9)-benzyl com-
pounds, the introduction of an amino group was useful, dimethyl-
aminoethoxy analogue 2017 gave single figure nano-molar potency
with improved solubility compared to methoxy compound 6. We
examined the methylene chain length and number of methyl
groups based on 20. Both C3 and C4 linked compounds, 21 and
22 showed approximately twofold weaker potencies with similar
solubility as 20, respectively, whereas the mono-methyl analogue
23 afforded a significantly decreased level of its activity. Potency of
compound 24, in which the dimethylamino group in 20 was re-
placed with morpholino, showed weaker potency than 20. And
the result was similar with that of compound 13 versus 15, sug-
gesting that basicity of the amino group have effects on the
potency.
The synthetic route of pyridyl compounds is shown in Scheme
2. Compound 19 was prepared according to the reported method.11
Target compounds 20–29 were prepared through the displacement
of the chloro group in 19 with the corresponding amino alcohols in
the presence of sodium hydride, or heating with the corresponding
amines. All target compounds were recrystallized from alcoholic
solvent and provided for the evaluation.
The in vitro human TLR7 agonistic activities of the prepared
compounds were evaluated by
HEK293 cells, which were stably transfected with human TLR7
along with the NF-
B SEAP reporter.15 The results of benzyl com-
a reporter gene assay using
j
pounds are summarized in Table 1. The EC50 value of dimethyl-
amine analogue 13 was evaluated as 9.3 nM, approximately
20-fold greater than that of 1. In addition, the water solubility of
13 was determined to be 98
lg/mL, and 13 was found to be more
NH2
NH2
NH2
c
a
N
b
N
N
N
N
N
N
Br
N
H
N
H
N
N
N
H
N
N
H
N
OH
OH
10
9
8
NH2
NH2
N
NH2
H
N
H
N
H
N
d
e
N
N
N
O
O
O
N
N
N
N
H
N
N
H
N
H
N
Cl
R
OH
12
11
13
: R=NMe2
14
15
: R=piperidino
: R=molpholino
16: R=N-Me-piperazino
NH2
NH2
N
H
N
f,b,c,d,e
N
N
N
N
O
N
H
N
N
R
N
H
H
17
18
8
: R=piperidino
: R=molpholino
Scheme 1. Reagents and conditions: (a) 4-hydroxybenzylchloride, K2CO3, DMF (quant); (b) Br2, NaOAc, CHCl3 (77%); (c) c-HCl (68%); (d) SOCl2ÁCH2Cl2; (e) corresponding
amines, DIEA (20–75%), CHCl3; (f) 3-hydroxybenzylchloride, K2CO3, DMF (quant).

T. Nakamura et al. / Bioorg. Med. Chem. Lett. 23 (2013) 669–672
671
Table 2
NH2
NH2
H
N
H
N
Human TLR7 agonistic activity and solubility of N(9)-pyridylmethyl compounds
g
N
N
O
O
NH2
H
N
N
N
H
N
N
H
N
N
N
O
N
Cl
R
19
N
N
N
H
N
20
: R=O(CH2)2NMe2
21: R=O(CH2)3NMe2
22
R
N
: R=O(CH2)4NMe2
23: R=O(CH2)2NHMe
Solubilityb (mg/mL)
a
Compd
R
hTLR7 EC50 (nM)
24
25
26
27
28
29
: R=O(CH2)2-molpholino
: R=NMe(CH2)3NMe2
: R=4-NMe2-piperidino
: R=3-NMe2-pyrrolidino
: R=N-Me-piperazino
: R=piperazino
2
6
Me
OMe
32
28.1
<1
<1
NMe2
20
21
22
23
7.2
372
O
O
NMe2
NMe2
NHMe
19.6
15.2
165.7
288
350
O
O
Scheme 2. Reagents and conditions: (g) corresponding alcohols, NaH, DMF, or
corresponding amines (25–65%).
Not tested
O
24
25
26
50.2
72.8
9.4
34
N
O
Table 1
Human TLR7 agonistic activity and solubility of N(9)-benzyl compounds
N
Me
NMe2
725
71
NH2
H
NMe2
NMe2
N
N
O
N
N
N
N
H
N
CH2R
27
28
7.4
5.4
53
7
NMe
NH
Compd
Position
R
hTLR7 EC50 (nM)
Solubilityb (mg/mL)
a
N
N
1
13
para
para
H
NMe2
166
9.3
<1
98
29
48.5
63
852A
2657
Not tested
14
15
16
17
para
para
para
meta
meta
7.6
58
19
N
a
Mean values of two independent experiments.
Solubility at pH 7.4 buffer.
b
O
8
N
N
N
N
NMe
7
108
Table 3
Antitumor activity of 20 on HM-1 metastasis model
30.1
Not tested
Compd
Dose (mg/kg)
% Inhibition of lung metastasis
NMe
20
0.1
1
1
78 21*
18
130
Not tested
Not tested
100 0**
852A
2657
852A
6
20
a
10
55 19
Mean values of two independent experiments.
Solubility at pH 7.4 buffer.
b
The data are expressed as the mean SD of 6 mice.
P <0.05 versus vehicle (Wilcoxon test).
*
**
P <0.01.
We also prepared and evaluated nitrogen atom linked com-
Compound 20 showed potent agonistic activity toward mice
TLR7 (EC50 = 4.6 nM), whereas 852A afforded weak activity
(EC50 = 6224 nM) which is consistent with human TLR7. Com-
pound 20 was dosed to Balb/c female mice intravenously at a dose
of 1 mg/kg to confirm a PK/PD relationship. The quantity of IFN in
plasma was measured by bio-assay using L929 cells with vesicular
stomatitis virus.18,19 It had notably high volumes of distribution
(Vdss = 6.6 L/kg) and the T1/2 was calculated as 1.3 h, however, a sig-
nificant amount of IFN was induced into plasma (840 32 IU/mL)
at 6 h post dosing. In comparison, 852A induced IFN at a dose of
10 mg/kg, but to a lower extent (19 3 IU/mL) than that of the
compound 20.
pounds. N-Methyl analogue 25 showed a fourfold loss of potency
compared to the oxygen atom linked compound 20. Meanwhile,
the cyclic amino compounds 26–28 demonstrated high potency
with single figure nano-molar EC50 values but with reduced solu-
bility compared to 25, suggesting that appropriate structural flex-
ibility may be necessary for improved aqueous solubility. As same
as the cases of 20 versus 23, compound 29 gave ninefold loss of the
activity compared to 28, indicating the secondary amino group was
not preferable to show potent activity.
Among the compounds prepared, compound 20 exhibited the
best balance of potency (EC50 = 7.2 nM) with high aqueous solubil-
ity (372 lg/mL). A small tertiary amino group with a short methy-
We evaluated the in vivo antitumor activity of 20 using a mice
syngeneic melanoma lung metastasis tumor model. HM-1 ovarian
cancer cells were injected into B6C3F1 mice subcutaneously on day
0, the primary tumor was surgically removed under anesthesia
condition with isofluorane on day 10, test compounds were intra-
venously administered biweekly commencing on day 11 and the
number of lung metastasis was counted on day 35. As shown in
Table 3, the compound 20 suppressed the lung metastasis
lene linker is considered to be crucial to satisfy our criteria for
compounds progression, that is, potent activity with high water
solubility. It is of note that TLR7 agonistic potency of 852A,
EC50 = 2672 nM is 370-fold weaker than that of the compound 20.
Thus, we selected compound 20 for further evaluation. Before
conducting an in vivo antitumor activity, we evaluated the
activities of 20 and 852A towards mice TLR7 and established the
pharmacokinetics/pharmacodynamics (PK/PD) profile in mice.

672
T. Nakamura et al. / Bioorg. Med. Chem. Lett. 23 (2013) 669–672
6. Dumitru, C. D.; Antonysamy, M. A.; Gorski, K. S.; Johnson, D. D.; Reddy, L. G.;
Lutterman, J. L.; Piri, M. M.; Prokesch, J.; McGurran, S. M.; Egging, E. A.; Cochran,
F. R.; Lipson, K. E.; Tomai, M. A.; Gullikson, G. W. Cancer Immunol. Immunother.
2009, 58, 575.
7. Majewski, S.; Marczak, M.; Mlynarczyk, B.; Benninghoff, B.; Jablonska, S. Int. J.
Dermatol. 2005, 44, 14.
8. Schon, M.; Bong, A. B.; Drewniok, C.; Herz, J.; Gellen, C. C.; Reifenberger, J.;
Benninghoff, B.; Slade, H. B.; Golinick, H.; Schon, M. P. J. Natl. Cancer Inst. 2003,
95, 1138.
9. Reinhard, D.; Axel, H.; Juergen, C. B.; Jean-Jacques, G.; Dirk, S.; Veronica, T.;
Jeannine, S.; Katharina, C. K.; Stephanie, M.; Ruth, C.; Tze-Chiang, M.; Mirjana,
U. Clin. Cancer Res. 2008, 14, 856.
10. Geller, M. A.; Cooley, S.; Argenta, P. A.; Downs, L. S.; Carson, L. F.; Judson, P. L.;
Ghebre, R.; Weigel, B.; Panoskaltsis-Mortari, A.; Curtsinger, J.; Miller, J. S.
Cancer Immunol. Immunother. 1877, 2010, 59.
11. Isobe, Y.; Kurimoto, A.; Tobe, M.; Hashimoto, K.; Nakamura, T.; Norimura, K.;
Ogita, H.; Takaku, H. J. Med. Chem. 2006, 49, 2088.
significantly, 78% inhibition was seen at 0.1 mg/kg dosing (with no
tumor metastasis at the 1 mg/kg group). However, the efficacy of
852A was limited, that is, 55% suppression at 10 mg/kg. These re-
sults correlate well with their in vitro activities and indicates that
the in vivo efficacy of the compound 20 is >100-fold higher than
that of 852A.
In summary, a study on the structure–activity relationship of
8-oxoadenines for TLR7 agonistic activity and solubility was con-
ducted based upon the structures of 1 and 2. The introduction of
an alkylamino group on the N(9)-position of the adenine resulted
in significantly improved aqueous solubility without loss of
potency. In particular, the compound 20 (DSR-6434) was identified
as a potent human TLR7 agonist, EC50 = 7.2 nM with high water sol-
12. Hirota, K.; Kazaoka, K.; Niimoto, I.; Kumihara, H.; Sajiki, H.; Isobe, Y.; Takaku,
H.; Tobe, M.; Ogita, H.; Ogino, T.; Ichii, S.; Kurimoto, A.; Kawakami, H. J. Med.
Chem. 2002, 45, 5419.
ubility 372 lg/mL. In addition, 20 demonstrated a strong antitu-
mor effect in an in vivo model by intravenous administration.
13. Kurimoto, A.; Hashimoto, K.; Nakamura, T.; Norimura, Kei; Ogita, H.; Takaku,
H.; Bonnert, R.; McInally, T.; Wada, H.; Isobe, Y. J. Med. Chem. 2010, 53, 2964.
14. Kurimoto, A.; Ogino, T.; Ichii, S.; Isobe, Y.; Tobe, M.; Ogita, H.; Takaku, H.; Sajiki,
H.; Hirota, K.; Kawakami, H. Bioorg. Med. Chem. 2004, 12, 1091.
15. Gibson, S. J.; Lindh, J. M.; Riter, T. R.; Gleason, R. M.; Rogers, L. M.; Fuller, A. E.;
Oesterich, J. L.; Gorden, K. B.; Qiu, X.; McKane, S.; Noelle, R. J.; Miller, R. L.; Kedl,
R. M.; Fitzgerald-Bocarsly, P.; Tomai, M. A. Cell. Immunol. 2002, 218, 74.
16. Ichikawa, M.; Hashimoto, Y. J. Med. Chem. 2011, 54, 1539.
17. Spectral data for compound 20: mp = 248–249 °C; 1H NMR (DMSO-d6, 300 MHz)
d 9.64 (1H, br s), 8.13 (1H, d, J = 2.2 Hz), 7.65 (1H, dd, J = 2.4, 8.4 Hz), 6.76 (1H,
d, J = 8.4 Hz), 6.26 (1H, t, J = 5.6 Hz), 6.02 (2H, s), 4.74 (2H, s), 4.33 (2H, t,
J = 5.8 Hz), 3.17 (2H, q, J = 6.8 Hz), 2.56 (2H, t, J = 5.8 Hz), 2.17 (6H, s), 1.45 (2H,
m), 1.30 (2H, m), 0.88 (3H, t, J = 7.2 Hz); MS (ESI) m/z 401 (M+1); Anal. Calcd for
Acknowledgments
The authors would like to thank Dr. Masashi Murata and Ms.
Erina Koga for providing various biological data.
References and notes
1. Tabeta, K.; Georgel, P.; Janssen, E.; Du, X.; Hoebe, K.; Crozat, K.; Mudd, S.;
Shamel, L.; Sovath, S.; Goode, J.; Alexopoulou, L.; Flavell, R. A. Proc. Natl. Acad.
Sci. U.S.A. 2004, 101, 3516.
2. Takeda, K.; Akira, S. Int. Immunol. 2005, 17, 1.
C
19H28N8O2Á0.5H2O: C, 55.73; H, 7.14; N, 27.36. Found: C, 55.50; H, 7.12; N,
3. Iwasaki, A.; Medzhitov, R. Nat. Immunol. 2004, 5, 987.
4. Jarrossay, D.; Napolitani, G.; Colonna, M.; Sallusto, F.; Lanzavecchia, A. Eur. J.
Immunol. 2001, 31, 3388.
5. Urosevic, M.; Dummer, R.; Conrad, C.; Beyeler, M.; Laine, E.; Burg, G.; Gilliet, M.
J. Natl. Cancer Inst. 2005, 97, 1143.
27.51 (recrystallized from MeOH).
18. Watanabe, Y.; Kawade, Y. In Lymphokines and Interferons: A Practical Approach;
Clemens, M. J., Morris, A. G., Gearing, A. J. H., Eds.; IRL Press: Oxford, 1987; p 6.
19. Pestka, S. In Methods in Enzymology; Pestka, S., Ed.; Academic Press: New York,
1986; Vol. 119, p 16.