Regioselective synthesis of multiply halogenated azaxanthones

Article abstract of DOI:10.1016/j.tetlet.2014.11.032

Four multiply halogenated azaxanthones 3, 4b, 5, and 6 were synthesized as novel core building blocks of β-site amyloid precursor protein cleaving enzyme 1 (BACE1) inhibitors. Each of these heterocycles requires a specific synthetic strategy to control not only the aza-positions, but also the regiochemistry of the fully differentiated and highly reactive halogen substituents.

Full text of DOI:10.1016/j.tetlet.2014.11.032

Tetrahedron Letters 55 (2014) 7229–7232
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Regioselective synthesis of multiply halogenated azaxanthones
⇑
Wenyuan Qian , James Brown, Jian Jeffrey Chen, Yuan Cheng
Department of Medicinal Chemistry, Amgen Inc., One Amgen Center Drive, Thousand Oaks, CA 91320-1799, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
Four multiply halogenated azaxanthones 3, 4b, 5, and 6 were synthesized as novel core building blocks of
b-site amyloid precursor protein cleaving enzyme 1 (BACE1) inhibitors. Each of these heterocycles
requires a specific synthetic strategy to control not only the aza-positions, but also the regiochemistry
of the fully differentiated and highly reactive halogen substituents.
Received 23 October 2014
Revised 5 November 2014
Accepted 6 November 2014
Available online 13 November 2014
Ó 2014 Elsevier Ltd. All rights reserved.
Keywords:
Azaxanthones
Multiple halogen substitution
Regiochemistry
Amide-directed metalation
Cyclization
Heterocycle synthesis
The tricyclic xanthone is a major scaffold for a range of natural
and synthetic products that exhibit various interesting biological
activities depending on the nature and pattern of the substituents.1
Efficient assembly strategies that allow for accurate control of the
substitution regiochemistry on this ‘privileged structure’ are there-
fore of great interest in medicinal chemistry. Many syntheses of
the xanthone core are available, which typically require a biaryl
ether or a benzophenone as the key intermediate.2,3 Azaxanthones
are also of interest as they modulate the physicochemical proper-
ties of xanthones. However, reports of azaxanthones are relatively
sparse in the literature.4
4-positions (3-aza-4-F or 4-aza-3-F) on the xanthene skeleton
would further optimize the overall properties as more efficacious
BACE 1 inhibitors. In addition, we sought to explore the effect of
nitrogen insertion at the 1-position. Consequently, rapid access
to the four halogenated azaxanthones 3, 4, 5, and 6 were highly
desirable in our program.
These densely halogenated azaxanthones were unprecedented
in the literature and presented unusual synthetic challenges. First
of all, the annulation of a pyridine with four differentiated substit-
uents was nontrivial in terms of the regiochemistry of not only the
aza-positions but also the different halogens. Especially, the fluo-
ride group on this subunit was expected to be highly activated
due to the strong electron-withdrawing effect of the carbonyl
group and the pyridine nitrogen. Moreover, multiply substituted
fluoropyridines were not readily available as starting materials.
Herein we describe several different synthetic strategies to address
the unique issues in each of these novel targets.
The synthesis of 3-aza-4-F-xanthone 3 began with a copper cat-
alyzed, regioselective Ullmann coupling between 2,5-dibromoben-
zoic acid (7) and 2-fluoro-3-hydroxypyridine (8) under conditions
developed by Buchwald and co-workers (Scheme 1).7 After
workup, the resulting crude biaryl ether 9 was directly treated
with diethylamine and TBTU to give amide 10 in 50% yield over
two steps. Then, an amide-directed lithiation on the pyridine and
the subsequent in situ cyclization provided the azaxanthone
11.4b,8 N-oxidation, which was necessary for the final installation
of a chlorine atom, failed presumably due to the strong electron-
withdrawing effects of the para-carbonyl and the ortho-F groups
on the pyridine moiety in 11. To circumvent this issue, N-oxidation
Recently, we have developed a unique aminooxazoline xan-
thene series (1) as potent and CNS penetrant b-site amyloid pre-
cursor protein (APP) cleaving enzyme 1 (BACE1) inhibitors for
the potential treatment of Alzheimer’s disease (Fig. 1).5 In this
class, xanthone 2 served as the key intermediate in which the
two differentiated halogen groups at the 2- and 7-positions offered
flexibility in the structure–activity relationship (SAR) study to
install a variety of R1 and R2 groups that can bind into the S2
0
and S3 pockets of the BACE 1 enzyme, respectively. Interestingly,
through early core modifications, we found that a single N-inser-
tion (azaxanthone) or fluorine-substitution at either 3- or 4-posi-
tions of the xanthene core 2 can improve in vitro potency,
modulate CNS penetration, PKDM properties and/or cardiovascular
safety profiles.6 Based on these results, we reasoned that combina-
tions of both the aza and fluorine modifications at the 3- and
⇑
Corresponding author. Tel.: +1 805 447 0078; fax: +1 805 480 1337.
E-mail address: wqian@amgen.com (W. Qian).
http://dx.doi.org/10.1016/j.tetlet.2014.11.032
0040-4039/Ó 2014 Elsevier Ltd. All rights reserved.

7230
W. Qian et al. / Tetrahedron Letters 55 (2014) 7229–7232
H2N
O
1
O
O
1
4
N
R1
2
R2
7
Br
I (Cl)
S2'
S3
3
O
3
4
S1
1
2
key intermediate
Potent BACE inhibitors
This work: Novel F-azaxanthone cores for lead optimization:
O
O
O
O
Br
Cl
Br
X
Br
N
Cl
N
O
N
F
O
F
R
4
3
5, R = H
6, R = F
X = I or Bu3Sn
Figure 1. Multiply halogenated azaxanthones as key intermediates toward the development of novel BACE inhibitors.
Br
CO2H
HO
F
CO2H
Br
Br
Br
b
a
+
N
O
N
F
7
8
9
O
NEt2
O
O
O
Br
Br
d
X
c
N
N
N
O
O
O
F
F
F
10 (50%, two steps)
11 (60 %)
d
NEt2
NEt2
O
O
Br
Br
Cl
Br
Cl
e
O
c
N
N
N
O
O
O
O
F
F
F
12 (50%)
13 (55%)
3 (67 %)
Scheme 1. Reagents and conditions: (a) Cu(OTf)2, Cs2CO3, toluene, 110 °C; (b) Et2NH, TBTU, DCM, rt; (c) LDA, THF, ꢀ78 °C; (d) UHP, TFAA, DCM, rt; (e) POCl3, DCM, DMF, rt.
was performed successfully on the more electron-rich biaryl ether
10 to give 12, which upon treatment of POCl3 afforded chloride 13.
Finally, a tandem lithiation/cyclization yielded the desired tris-hal-
ogenated 3-azaxanthone 3.
fluorine-directed ortho-metalation and electrophile quenching
failed to install the desired iodine onto this heterocycle.
After many unsuccessful trials, a totally different route was
developed (Scheme 3). Mono-lithiation of 15, followed by trapping
with aldehyde 19 provided the benzhydryl alcohol 20 which was
then protected as TBS ether 21. Next, instead of the highly reactive
iodide, we chose to install a more stable Bu3Sn group after the sec-
ond fluorine-directed ortho-lithiation of the pyridine. This group
can also serve as a handle for further functionalizations. Removal
of the TBS group, followed by oxidation yielded the biaryl ketone
23. Finally, cleavage of the methyl group under mild conditions
unveiled the hydroxyl group, which cyclized in an intramolecular
SNAr fashion with the highly activated difluoropyridine moiety to
furnish the trisubstituted 4-azaxanthone 4b.
To prepare the 3-F-4-azaxanthone 4, a ‘reversed’ first step cou-
pling was conveniently conducted between 5-bromo-2-hydroxy-
benzoic acid (14) and the symmetric 2,6-difluoropyridine (15)
under catalyst-free conditions (Scheme 2). Following the amide
formation (16), the tandem lithiation/cyclization process provided
the azaxanthone core 17. Interestingly, compared to the regioisom-
er 11 in Scheme 1, the fluoride on 17 is much more reactive
because of the double activation by the para-carbonyl group and
the ortho-nitrogen atom. As a result, a diethylamine substitution
byproduct 18 was also isolated. To our disappointment, efforts of

W. Qian et al. / Tetrahedron Letters 55 (2014) 7229–7232
7231
NEt2
O
CO2H
Br
Br
a, b
+
OH
F
F
N
O
F
N
14
16 (40%, two steps)
15
O
O
O
Br
c
Br
+
F
N
N
O
N
17 (48%)
18 (16%)
d
X
O
Br
I
F
O
N
4a
Scheme 2. Reagents and conditions: (a) Cs2CO3, DMSO, 110 °C; (b) Et2NH, TBTU, DCM, rt; (c) LDA, THF, ꢀ78 °C; (d) LDA, then I2, THF, ꢀ78 °C.
Cl
Br
NC
O
N
Br
a
b
Br
+
HO
Br
a
b
N
CN
O
+
O
24
25
26 (97%)
F
N
F
HO
O
15
19
F
N
(76%)
F
O
O
O
20
c, d
Br
N
Br
N
Cl
Br
Br
O
O
O
N
c
d, e
27 (100%)
5 (64%, two steps)
TBS
SnBu3
F
TBS
Br
O
O
Scheme 4. Reagents and conditions: (a) Cs2CO3, DMSO, 85 °C; (b) PPA, 180 °C; (c)
UHP, TFAA, DCM, rt; (d) POCl3, DCM, DMF, rt.
F
N
F
F
21 (88%)
22 (89%)
much more difficult task (6, Fig. 1). The four different strategies
to access 3, 17, 4b, and 5 cannot be simply applied to this target
because the corresponding multiply functionalized 4-F-pyridine
starting materials are not easily available and/or could cause
selectivity issues in the intermediate steps. In addition, the 4-F on
pyridine is highly reactive. Based on these considerations, we chose
3-bromo-4-nitropyridine 1-oxide (28) as a surrogate to couple
with the sodium salt of methyl 5-bromo-2-hydroxybenzoate (29,
Scheme 5). Both the N-oxide and the nitro group on 28 directed
the substitution to the bromide and no nitro group displacement
was detected in this regioselective SNAr reaction.9 Treatment of
the resulting biaryl ether 30 with LDA gave a mixture of the desired
azaxanthone 31 and the deoxygenated product 32, together with
40% of recovered starting material. The N-oxide 31 served as the
direct precursor to chloride 33. Gratifyingly, the pyridine-activated
nitro group was selectively displaced in the presence of the chloride
to give fluoride 6 at the final stage.10
O
O
O
f, g
Br
SnBu3
F
SnBu3
F
O
N
F
N
4b (89%, two steps)
23 (66%, two steps)
Scheme 3. Reagents and conditions: (a) LDA, THF, ꢀ78 °C; (b) TBSCl, Imidazole,
DMF, 50 °C; (c) LDA, then Bu3SnCl, THF, ꢀ78 °C; (d) TBAF, THF, rt; (e) TPAP, NMO,
DCM, rt; (f) 9-Br-BBN, DCM, rt; (g) K2CO3, MeCN, rt.
The preparation of the disubstituted 1-azaxanthone 5 was
straightforward (Scheme 4). SNAr reaction between an activated
chloride 24 and phenol 25 resulted in the biaryl ether 26 in nearly
quantitative yield after simple aqueous workup and filtration.
Then, an intramolecular Friedel–Crafts reaction in PPA, followed
by in situ hydrolysis provided the 1-azaxanthone core 27. N-oxida-
tion and subsequent treatment with POCl3 produced the desired
chloride 5. This four-step procedure does not need any column
purification and achieved high overall yield at 100 g scale.
In summary, we have showcased the rapid construction of sev-
eral isomeric azaxanthones with fully differentiated halogen sub-
stitutions as novel and flexible core structures for drug discovery.
In some cases, a well-established assembly strategy for one
azaxanthone is of limited value for a very similar structure bearing
just one extra halogen due to the new chemo- and regioselectivity
Compared to the robust synthesis of 5, introducing an extra
fluorine at the 4-position of the 1-azaxanthone core presented a

7232
W. Qian et al. / Tetrahedron Letters 55 (2014) 7229–7232
O
Br
NO2
N
O
HO
Br
a
O
O
b
O
+
O2N
Br
O
N
29
28
O
30 (62%)
c
(76%)
O
O
O
O
O
N
Br
N
Br
d
+
NO2
NO2
32 (30%)
31 (8%)
O
N
O
O
Br
Cl
Br
N
Cl
e
O
NO2
F
33 (87%)
6 (62%)
Scheme 5. Reagents and conditions: (a) NaH, DMF, 50 °C; (b) LDA, THF, ꢀ78 °C; 40% of starting material recovered; (c) UHP, TFAA, DCM, rt; (d) POCl3, DMF, DCM, rt; (e) TBAF,
THF/DMF, 0 °C.
2. For selected reviews on the synthesis of xanthones, see (a) Sousa, M. E.; Pinto,
issues associated with this ‘small’ change. Consequently, alterna-
tive approaches needed to be developed, as highlighted by the
M. M. M. Curr. Med. Chem. 2005, 12, 2447–2479; (b) Azevedo, C. M. G.; Afonso,
C. M. M.; Pinto, M. M. M. Curr. Org. Chem. 2012, 16, 2818–2867; (c) Masters, K.-
comparison of the syntheses between 17 and 4b, as well as 5 and
6, respectively. The SAR study and identification of high-quality
BACE inhibitors based on these new scaffolds will be disclosed in
due course.
S.; Brase, S. Chem. Rev. 2012, 112, 3717–3776.
3. For recent examples of non-classical syntheses of xanthones, see: (a)
Menendez, C. A.; Nador, F.; Radivoy, G.; Gerbino, D. C. Org. Lett. 2014, 16,
2846–2849; (b) Wang, P.; Rao, H.; Hua, R.; Li, C.-J. Org. Lett. 2012, 14, 902–905;
(c) Zhang, H.; Shi, R.; Gan, P.; Liu, C.; Ding, A.; Wang, Q.; Lei, A. Angew. Chem.,
Int. Ed. 2012, 51, 5204–5207; (d) Zhao, J.; Larock, R. C. Org. Lett. 2005, 7, 4273–
4275.
4. (a) Zhang, X.; Yang, L.; Wu, Y.; Du, J.; Mao, Y.; Wang, X.; Luan, S.; Lei, Y.; Li, X.;
Sun, H.; You, Q. Tetrahedron Lett. 2014, 55, 4883–4887; (b) Familoni, O. B.;
Ionica, I.; Bower, J. F.; Snieckus, V. Synlett 1997, 1081–1083; (c) Trecourt, F.;
Queguiner, G. J. Chem. Res. 1982, 76–77.
Acknowledgements
The authors thank Dr. Jennifer Allen and Dr. Margaret
Chu-Moyer for discussion.
5. Huang, H.; La, D. S.; Cheng, A. C.; Whittington, D. A.; Patel, V. F.; Chen, K.;
Dineen, T. A.; Epstein, O.; Graceffa, R.; Hickman, D.; Kiang, Y.-H.; Louie, S.; Luo,
Y.; Wahl, R. C.; Wen, P. H.; Wood, S.; Fremeau, R. T., Jr. J. Med. Chem. 2012, 55,
9156–9169.
Supplementary data
6. (a) Epstein, O.; Bryan, M. C.; Cheng, A. C.; Derakhchan, K.; Dineen, T. A.;
Hickman, D.; Hua, Z.; Human, J. B.; Kreiman, C. E.; Marx, I.; Weiss, M. M.; Qu, Y.;
Wahl, R. C.; Wen, P. H.; Whittington, D. A.; Wood, S.; Zheng, X.; Fremeau, R. T.,
Jr.; White, R. D.; Patel, V. F. J. Med. Chem. submitted for publication, http://
dx.doi.org/10.1021/jm501266w.; (b) Dineen, T. A.; Chen, K.; Cheng, A. C.;
Epstein, O.; Esmay, J.; Hickman, D.; Kreiman, C. E.; Marx, I. E.; Wahl, R. C.; Wen,
P. H.; Weiss, M. M.; Whittington, D. A.; Wood, S.; Fremeau, R. T., Jr.; White, R.
D.; Patel, V. F. J. Med. Chem. submitted for publication, http://dx.doi.org/
10.1021/jm5012676.
7. Marcoux, J.-F.; Doye, S.; Buchwald, S. L. J. Am. Chem. Soc. 1997, 119, 10539–
10540.
8. Snieckus, V. Chem. Rev. 1990, 90, 879–933.
9. Eggers, L.; Grahn, W. Synthesis 1996, 763–766.
10. Kuduk, S. D.; DiPardo, R. M.; Bock, M. G. Org. Lett. 2005, 7, 577–579.
Supplementary data associated with this article can be found,
in the online version, at http://dx.doi.org/10.1016/j.tetlet.2014.
11.032.
References and notes
1. For selected recent reviews on xanthones with biological activities, see (a)
Pinto, M. M. M.; Sousa, M. E.; Nascimento, M. S. J. Curr. Med. Chem. 2005, 12,
2517–2583; (b) El-Seedi, H. R.; El-Barbary, M. A.; El-Ghorab, D. M. H.; Bohlin, L.;
Borg-Karlson, A.-K.; Goeransson, U.; Verpoorte, R. Curr. Med. Chem. 2010, 17,
854–901; (c) Winter, D. K.; Sloman, D. L.; Porco, J. A., Jr. Nat. Prod. Rep. 2013, 30,
382–391.