Concise Synthesis of Dictyoquinazol A via a Dimerisation-Cyclocondensation Sequence

Article abstract of DOI:10.1055/s-0035-1561569

A two-step total synthesis of the neuroprotective alkaloid, dictyoquinazol A, has been achieved. The brevity of this synthesis was enabled by exploiting the hidden symmetry of the target molecule. Several structural analogues were also prepared using a similar strategy. These results provide a platform for future structure-activity relationship studies in the quest for a novel treatment for stroke.

Full text of DOI:10.1055/s-0035-1561569

SYNLETT
0
9
3
6
-
5
2
1
4
1
4
3
7
-
2
0
9
6
© Georg Thieme Verlag Stuttgart · New York
2016, 27, 1237–1240
letter
1237
J. Wangsahardja et al.
Letter
Synlett
Concise Synthesis of Dictyoquinazol A via a Dimerisation–Cyclo-
condensation Sequence
Jonatan Wangsahardja
Gabriella M. Marcolin
Yuvixza Lizarme
OMe
O
MeO
N
Jonathan C. Morris
Luke Hunter*
N
OH
Two-step total synthesis
exploiting "hidden symmetry"
School of Chemistry, UNSW Australia, Kensington NSW 2052,
Australia
l.hunter@unsw.edu.au
Received: 02.08.2015
Accepted after revision: 19.01.2016
Published online: 23.02.2016
O
N
S
F
O
N
N
N
N
DOI: 10.1055/s-0035-1561569; Art ID: st-2015-b0608-l
N
H
S
Abstract A two-step total synthesis of the neuroprotective alkaloid,
dictyoquinazol A, has been achieved. The brevity of this synthesis was
enabled by exploiting the hidden symmetry of the target molecule.
Several structural analogues were also prepared using a similar strate-
gy. These results provide a platform for future structure–activity rela-
tionship studies in the quest for a novel treatment for stroke.
HN
HN
N
1
2
3
N
OMe
O
N
N
MeO
N
OH
Key words quinazolinone, natural products, neuroprotective activity,
4
total synthesis, stroke
Figure 1 Quinazolinone derivatives that exhibit CNS activity include
methaqualone (1),2 the sulfur-containing analogue 2,3 idelalisib (3),4
and dictyoquinazol A (4)5
The quinazolinone moiety (Figure 1) is found within a
large number of alkaloids of plant, animal, and microbial
origin.1 A number of drugs containing this motif have also
been commercialised for the treatment of a variety of dis-
eases.1 An interesting subset of the bioactive quinazoli-
nones are those derivatives that have activity towards tar-
gets in the central nervous system (CNS). For example,
methaqualone (1, Figure 1) is a clinical sedative and has
also gained notoriety for its abuse as a hypnotic agent.2
Other examples of CNS-active quinazolinone derivatives in-
clude compounds 2 and 3, which exhibit neuroprotective
activity through distinct mechanisms,3,4 and which there-
fore have the potential to be developed into new drugs to
treat stroke.
The importance of the latter application prompted our
interest in another quinazolinone derivative, dictyo-
quinazol A (4, Figure 1).5 This natural product was isolated
in 2002 from the mushroom Dictyophora indusiata, which
has been employed as a traditional remedy for various ail-
ments including nervous disorders. The purified alkaloid 4
was subsequently shown to have neuroprotective activity
against glutamate-induced excitotoxicity, one of the ad-
verse biochemical events caused by stroke.5 Thus, it ap-
pears that compound 4 is a promising candidate for devel-
opment into a treatment for stroke.
Many different approaches have been developed over
the decades1 for the synthesis of substituted quinazoli-
nones. Contributing to this body of work, Oh and Song re-
ported in 2007 a total synthesis of 4 requiring six steps and
proceeding in 36% overall yield.6 In 2012, Ma and co-work-
ers reported an alternative (formal) synthesis of 4 in an ap-
proach requiring only five steps.7 Building on this literature
precedent, we were motivated to attempt a new total syn-
thesis of 4, and at the outset we had two goals in mind:
first, to identify an even more concise synthetic route; and
second, to avoid if possible a radical bromination reaction
that featured in the previous synthesis6 and which required
toxic reagents and precise experimental conditions.
The structure of 4 contains an element of pseudosym-
metry (Scheme 1). Therefore, a dimerisation strategy might
allow a more concise total synthesis to be achieved. If the
benzylic alcohol group of 4 is conceptually converted into
an ester (5, Scheme 1), then compound 5 could in turn be
© Georg Thieme Verlag Stuttgart · New York — Synlett 2016, 27, 1237–1240

1238
J. Wangsahardja et al.
Letter
Synlett
derived from two molecules of anthranilic ester 6 along
with a C1 source such as formic acid (7). In the forward
sense, the three fragments 6, 6 and 7 could potentially be
united through a simple, thermodynamically controlled di-
merisation–cyclocondensation reaction8–10 to enable a rap-
id total synthesis of dictyoquinazol A (4).
difficult to achieve. When ester 5 was treated with three
equivalents of lithium aluminium hydride at 0 °C, the de-
sired product 4 was obtained in low yield alongside 77% of
the over-reduced compound 9 (Scheme 2).14 When the
loading of lithium aluminium hydride was reduced, only
unreacted 5 and the unwanted compound 9 were obtained.
The use of milder reducing agents (e.g., sodium borohy-
dride, diisobutylaluminium hydride, sodium borohy-
dride/zinc chloride) did not improve the yield of 4, nor did
the investigation of a hydrolysis/borane reduction se-
quence. Finally, the optimal conditions were identified
when lithium borohydride in refluxing THF was found to
deliver 4 in 24% yield alongside 44% recovered 5.15 Overall,
despite the low yield in this final reaction (Scheme 2), the
synthetic sequence did successfully deliver the natural
product 4 in only two steps, with a reasonable overall yield
of 22%.
It became of interest to probe the mechanism of the cy-
clocondensation reaction (i.e., 6 → 5, Scheme 2), and more
specifically to ascertain whether the amide 8 was a true in-
termediate in this process or simply a shunt product.12 If
the amide 8 was an intermediate and could be carried for-
ward to a quinazolinone derivative, then this might broad-
en the variety of structural analogues of 4 that could be cre-
ated, because the substitution pattern on the quinazolinone
moiety could be different from that on the pendant aryl
ring. Accordingly, the purified amide 8 was re-subjected to
the cyclocondensation reaction conditions (Table 1, entry
OMe
O
MeO
4
N
CO2Me
N
5
O
MeO
OMe
NH2
6 (2 equiv)
+
HCO2H
7
Scheme 1 Potential strategy for a concise synthesis of 4
To investigate this synthetic plan, the commercially
available11 ester 6 was heated in neat formic acid or
trimethyl orthoformate (Scheme 2). When formic acid was
used as the C1 reagent, the desired quinazolinone 5 was ob-
tained in moderate yield, along with a variable quantity of
the amide 8 which was presumed to be an intermediate in
the dimerisation–cyclocondensation process.12 When
trimethyl orthoformate was used as the C1 source and
strong heating was applied (Scheme 2), the desired
quinazolinone 5 was obtained in an excellent yield of 91%.13
Having secured the quinazolinone 5, the only task re-
maining to complete the total synthesis of 4 was to reduce
the ester group of 5 (Scheme 2). Unfortunately, this seem-
ingly straightforward transformation proved surprisingly
1). Under these conditions
a small quantity of the
quinazolinone 5 was obtained, but it was not possible to as-
certain whether this product was derived from 8 or from
two molecules of 6. The reaction was repeated with less or
no formic acid (Table 1, entries 2 and 3), but this necessitat-
ed the use of acetonitrile as the solvent and none of the
quinazolinone 5 was formed in these instances.
HCO2H or (MeO)3CH,
MeO
CO2Me
MeO
MeO
MeO
CO2Me
NH2
μW
N
O
H
6
8
6
(one pot)
OMe
O
N
N
OH
OMe
various reducing
conditions
O
N
4 (up to 24%)
MeO
N
OMe
CO2Me
O
N
5 (up to 91%)
N
H
9 (up to 77%)
Scheme 2 Two-step total synthesis of dictyoquinazol A (4)
© Georg Thieme Verlag Stuttgart · New York — Synlett 2016, 27, 1237–1240

1239
J. Wangsahardja et al.
Letter
Synlett
Table 1 Probing the Mechanism of the Cyclocondensation Reaction
O
O
NH
O
MeO
R
conditions
MeO
or
8
N
OMe
N
N
H
10
(R = p-OMe-o-CH2OH)
(R = p-OMe-o-CO2Me)
(R = H)
4
5
11
Entry
Conditions
Product, yield (%)
1
2
3
4
5
6
6 (1 equiv), HCO2H (excess), MW, 4 h
6 (1 equiv), HCO2H (1 equiv), MeCN, MW, 4 h
6 (1 equiv), MeCN, MW, 4 h
5 5
8 100
NRa
6 (1 equiv), KOt-Bu (2 equiv), THF, air, 3 h
10 87
4 17
11 17
2-amino-5-methoxybenzyl alcohol (1 equiv), KOt-Bu (2 equiv), THF, air
aniline (1 equiv), KOt-Bu (2 equiv), THF, air
a NR = no reaction.
Since it was not possible to directly demonstrate that 8
was an intermediate in the dimerisation–cyclocondensa-
tion reaction (6 → 5, Scheme 2), an alternative method was
investigated for elaborating 8 into quinazolinone deriva-
tives. Ester 8 was treated with a variety of aniline deriva-
tives under conditions that selectively convert esters into
amides (Table 1, entries 4–6).16 When the aniline 6 was em-
ployed in this process (Table 1, entry 4), it unexpectedly di-
merised to give 10 instead of reacting with 8. However,
when other anilines were employed (Table 1, entries 5 and
6), the expected quinazolinones 4 and 11 were successfully
obtained, presumably via bisamide intermediates which
were not isolated or observed.12 Incidentally, this consti-
tutes an alternative two-step synthesis of the natural prod-
uct 4, but more importantly it also confirms that ‘nonsym-
metrically’ substituted analogues (e.g., 11) can be created
through this approach, albeit in low yields.
ole or thiocarbonyldiimidazole as the C1 source instead of
formic acid (Scheme 3). The dimerisation–cyclocondensa-
tion reactions successfully delivered the heterocycles 12
and 13 in good yield (Scheme 3). Gratifyingly, the subse-
quent reduction reactions (Scheme 3) were also found to be
straightforward and efficient in the absence of an easily re-
duceable heterocyclic moiety, and this provided two new
structural analogues of the natural product (14 and 15).
Finally, attention was turned to another method for pro-
ducing structural analogues of 4 in which the substitution
patterns on the quinazolinone and phenyl moieties can be
different (Scheme 4). Related cyclocondensation approach-
es have previously been developed by Wang9 and Khajavi,10
starting from anthranilic acids instead of anthranilic esters;
these methods allow ‘nonsymmetrical’ targets to be created
in high yields and without requiring any intermediates to
be isolated. In the current work, the anthranilic acid meth-
od was successfully employed to generate several structural
analogues of 4 in which the substitution on the pendant
aryl ring was varied (11, 17–20, Scheme 4). The modest
yields were mostly attributable to losses during chromato-
graphic separation from close-running side products. The
The next objective was to determine whether the origi-
nal dimerisation–cyclocondensation approach could be
modified to produce structural analogues of 4 in which the
heterocyclic moiety was varied. Accordingly, the dimerisa-
tion reaction of 6 was repeated but using carbonyldiimidaz-
X
N
N
N
N
OMe
OMe
O
O
1,4-dioxane or
LiAlH4, HMPA
MeO
CO2Me
NH2
MeCN, Δ
THF, −84 °C
MeO
MeO
N
N
CO2Me
N
H
X
OH
N
H
X
6
12 (X = O): 64%
13 (X = S): 66%
14 (X = O): 92%
15 (X = S): 62%
Scheme 3 Synthesis of structural analogues of 4 in which the heterocycle is modified
© Georg Thieme Verlag Stuttgart · New York — Synlett 2016, 27, 1237–1240

1240
J. Wangsahardja et al.
Letter
Synlett
anthranilic acid method was also investigated for an at-
tempted one-step synthesis of the natural product (16 → 4,
Scheme 4), but this was unsuccessful; an intractable mix-
ture of products was formed, and the expected benzylic hy-
(5) Lee, I.-K.; Yun, B.-S.; Han, G.; Cho, D.-H.; Kim, Y.-H.; Yoo, I.-D.
J. Nat. Prod. 2002, 65, 1769.
(6) Oh, C. H.; Song, C. H. Synth. Commun. 2007, 37, 3311.
(7) Xu, L.; Jiang, Y.; Ma, D. Org. Lett. 2012, 14, 1150.
(8) Leiby, R. W. J. Org. Chem. 1985, 50, 2926.
1
drogen signals were absent from the H NMR spectrum of
(9) Wang, L.; Xia, J.; Qin, F.; Qian, C.; Sun, J. Synthesis 2003, 1241.
(10) Rad-Moghadam, K.; Khajavi, M. S. J. Chem. Res., Synop. 1998,
702.
(11) Also available in two synthetic steps from 2-nitro-4-methoxy-
benzoic acid, see Supporting Information.
(12) Various intermediates in related cyclocondensation processes
have been proposed: (a) Alexandre, F.-R.; Berecibar, A.; Besson,
T. Tetrahedron Lett. 2002, 43, 3911. (b) Mhaske, S. B.; Argade, N.
P. J. Org. Chem. 2001, 66, 9038. (c) Mhaske, S. B.; Argade, N. P.
Tetrahedron 2004, 60, 3417. (d) Wang, H.; Sim, M. M. J. Nat. Prod.
2001, 64, 1497.
the crude product mixture. This outcome is perhaps unsur-
prising given the known propensity of ortho-aminobenzyl
alcohols to undergo iminoquinone methide chemistry,17
which in this context would be expected to lead to po-
lymerisation or decomposition.
R
H2N
HCO2H, μW
O
MeO
CO2H
NH2
R
(13) A microwave reaction vessel was charged with compound 6
(355 mg, 1.96 mmol) then cooled to –20 °C. Trimethyl orthofor-
mate (110 μL, 1.01 mmol) was added, and the mixture was irra-
diated at 170 W (155 ± 4 °C) for 4 h. The mixture was cooled
and concentrated under a stream of nitrogen. The residue was
diluted with CH2Cl2 (25 mL), dried over MgSO4, filtered, and
concentrated in vacuo. The crude product was purified by flash
chromatography using EtOAc–CH2Cl2 (1:6) as eluent to give
compound 5 as a white solid (303 mg, 91%); mp 155–156 °C. IR
(neat): νmax = 3081, 2626, 2105, 1938, 1847, 1719, 1668, 1610,
MeO
N
N
16
11 (R = H): 25%
17 (R = p-Me): 37%
18 (R = p-Cl): 34%
19 (R = p-OMe): 30%
20 (R = m,p-Cl2): 29%
4 (R = p-OMe-o-CH2OH): 0%
Scheme 4 Synthesis of nonsymmetrically substituted analogues
1482 cm–1 1H NMR (300 MHz, CDCl3): δ = 7.92 (s, 1 H, ArH),
.
7.72 (d, J = 9.0 Hz, 1 H, ArH), 7.70 (d, J = 3.0 Hz, 1 H, ArH), 7.67
(d, J = 2.9 Hz, 1 H, ArH), 7.38 (dd, J = 8.9, 3.0 Hz, 1 H, ArH), 7.29
(d, J = 8.6 Hz, 1 H, ArH), 7.20 (dd, J = 8.7, 2.9 Hz, 1 H, ArH), 3.92
(s, 6 H, OMe), 3.70 (s, 3 H, OMe). 13C{1H} NMR (75 MHz, CDCl3):
δ = 165.0, 161.4, 160.2, 159.0, 144.4, 142.6, 130.6, 130.2, 129.3,
129.2, 124.7, 123.2, 119.4, 116.8, 106.8, 55.6, 56.0, 52.7. ESI-
HRMS (+ve): m/z calcd for C18H16N2O5Na+ [MNa+]: 363.0951;
found: 363.0948.
In conclusion, a two-step total synthesis of the neuro-
protective alkaloid dictyoquinazol A (4) has been achieved,
proceeding in 19% overall yield. The key step was a dimeri-
sation–cyclocondensation reaction, the design of which
was guided by the insight that the target molecule possess-
es hidden symmetry. Related cyclocondensation reactions
were also successfully employed in this work for the pro-
duction of structural analogues of 4. These results should
facilitate the medicinal development of 4 towards a possi-
ble future treatment for stroke.
(14) For spectral data of similar compounds, see: (a) Yoo, C. L.;
Fettinger, J. C.; Kurth, M. J. J. Org. Chem. 2006, 70, 6941.
(b) Khurana, J. M.; Kukreja, G. J. Heterocycl. Chem. 2003, 40, 677.
(15) A solution of compound 5 (57.7 mg, 0.170 mmol) and LiBH4 (5.0
mg, 0.23 mmol) in anhydrous THF (1 mL) was heated at reflux
for 2 h, then quenched by the addition of sat. aq NH4Cl solution
at 0 °C. The mixture was diluted with CH2Cl2 (20 mL) then
washed with H2O (2 × 20 mL) and brine (2 × 20 mL). The organic
layer was dried over MgSO4, filtered, and concentrated in vacuo.
The crude product was purified by flash chromatography (1:6
→ 1:2 EtOAc–CH2Cl2) to give compound 4 as a white solid (12.7
mg, 24%); mp 199–202 °C. IR (neat): νmax = 3163, 3058, 2909,
Supporting Information
Supporting information for this article is available online at
http://dx.doi.org/10.1055/s-0035-1561569.
S
u
p
p
ortioInfgrmoaitn
S
u
p
p
ortiInfogrmoaitn
References and Notes
2837, 2685, 2111, 2081, 1684, 1611, 1493 cm–1 1H NMR (400
.
MHz, CDCl3): δ = 7.92 (s, 1 H, ArH), 7.71–7.69 (m, 2 H, ArH), 7.41
(dd, J = 8.9, 2.9 Hz, 1 H, ArH), 7.18 (d, J = 2.8 Hz, 1 H, ArH), 7.16
(d, J = 8.6 Hz, 1 H, ArH), 6.99 (dd, J = 8.6, 2.9 Hz, 1 H, ArH), 4.45
(d, J = 12.6 Hz, 1 H, CHH), 4.39 (d, J = 12.6 Hz, 1 H, CHH), 3.93 (s,
3 H, OMe), 3.88 (s, 3 H, OMe). 13C{1H} NMR (100 MHz, CDCl3):
δ = 162.0, 160.8, 159.3, 144.6, 142.8, 139.7, 129.3, 128.9, 128.6,
125.3, 122.9, 115.6, 115.0, 106.7, 61.7, 56.1, 55.8. ESI-HRMS
(+ve): m/z calcd for C17H16N2O4Na+ [MNa+]: 335.1002; found:
335.0948.
(1) (a) Johne, S. Prog. Chem. Org. Nat. Prod. 1984, 46, 159.
(b) Mhaske, S. B.; Argade, N. P. Tetrahedron 2006, 62, 9787.
(c) Khan, I.; Ibrar, A.; Ahmed, W.; Saeed, A. Eur. J. Med. Chem.
2015, 90, 124.
(2) van Zyl, E. F. Forensic Sci. Int. 2001, 122, 142.
(3) Redondo, M.; Zarruk, J. G.; Ceballos, P.; Pérez, D. I.; Pérez, C.;
Perez-Castillo, A.; Moro, M. A.; Brea, J.; Val, C.; Cadavid, M. I.;
Loza, M. I.; Campillo, N. E.; Martínez, A.; Gil, C. Eur. J. Med. Chem.
2012, 47, 175.
(4) Low, P. C.; Manzanero, S.; Mohannak, N.; Narayana, V. K.;
Nguyen, T. H.; Kvaskoff, D.; Brennan, F. H.; Ruitenberg, M. J.;
Gelderblom, M.; Magnus, T.; Kim, H. A.; Broughton, B. R. S.;
Sobey, C. G.; Vanhaesebroeck, B.; Stow, J. L.; Arumugam, T. V.;
Meunier, F. A. Nat. Commun. 2014, 5, 3450.
(16) Kim, B. R.; Lee, H.-G.; Kang, S.-B.; Sung, G. H.; Kim, J.-J.; Perk, J.
K.; Lee, S.-G.; Yoon, Y.-J. Synthesis 2011, 42.
(17) Sander, W.; Morawietz, J. Tetrahedron Lett. 1993, 34, 1913.
© Georg Thieme Verlag Stuttgart · New York — Synlett 2016, 27, 1237–1240