3,3a-dihydropyrano[4,3,2-de]quinazolin-2(1H)-ones are potent non-nucleoside reverse transcriptase inhibitors

Article abstract of DOI:10.1016/S0960-894X(00)00624-7

A series of unique 3,3a-dihydropyrano[4,3,2-de]quinazolin-2(1H)-ones and a 2a,5-dihydro-2H-thieno[4,3,2-de]quinazoline-4(3H)-thione were found to be HIV-1 non-nucleoside reverse transcriptase inhibitors. One of these compounds, as the racemate, possessed

Full text of DOI:10.1016/S0960-894X(00)00624-7

Bioorganic & Medicinal Chemistry Letters 11 (2001) 211±214
3,3a-Dihydropyrano[4,3,2-de]quinazolin-2(1H)-ones are Potent
Non-Nucleoside Reverse Transcriptase Inhibitors
Je€rey W. Corbett,Ã,y Senliang Pan, Jay A. Markwalder, Beverly C. Cordova,
Ronald M. Klabe, Sena Garber, James D. Rodgers and Susan K. Erickson-Viitanen
DuPont Pharmaceuticals Company, Experimental Station, PO Box 80500, Wilmington, DE 19880-0500, USA
Received 10 October 2000; accepted 2 November 2000
AbstractÐA series of unique 3,3a-dihydropyrano[4,3,2-de]quinazolin-2(1H)-ones and a 2a,5-dihydro-2H-thieno[4,3,2-de]quinazo-
line-4(3H)-thione were found to be HIV-1 non-nucleoside reverse transcriptase inhibitors. One of these compounds, as the race-
mate, possessed an IC90=4.6 nM against wild-type virus in a whole cell antiviral assay and had an IC90=76 and 897 nM against
the clinically signi®cant K103N and K103N/L100I mutant viruses, respectively. # 2001 DuPont Pharmaceuticals Company.
Published by Elsevier Science Ltd. All rights reserved.
Human immunode®ciency virus type-1 (HIV-1) is a ret-
rovirus and is the causative agent in the development
and transmission of HIV.1 It is estimated that the
infectivity rate in the United States has leveled o€ at an
alarming 40,000 new infections per year.2 Current
treatments for HIV infection are designed to interfere
with the ability of the virus to replicate by either inhibit-
ing HIV protease or HIV reverse transcriptase (RT).
There are two types of drugs available for inhibiting
HIV-1 RT: nucleoside reverse transcriptase inhibitors
(NRTIs) and non-nucleoside reverse transcriptase inhi-
bitors (NNRTIs).3 The NRTIs are competitive inhibitors
of the nucleoside triphosphate binding site on RT and
act as substrate decoys and chain terminators, while the
NNRTIs bind at an allosteric site that regulates enzyme
activity and are noncompetitive with the NTP binding
site.4
pliance, partial pre-existing resistance or poor bioavail-
ability. Frequently outgrowth or de novo generation of
resistant viral strains occurs in these virologic failures.
This situation is compounded by the selection of some
resistance mutations which confer cross-class resistance
such as the Q151M mutation for the NRTIs or the
K103N single mutation (or the K103N/L100I double
mutation) for the NNRTIs. Our continuing e€orts to
identify NNRTIs that possess a broad spectrum of
activity against mutant viral forms of RT has resulted in
the selection of a series of quinazolinones for further
development, two of which are DPC083 and DPC961
(Fig. 1).6,7 This paper describes a series of novel tricyclic
quinazolinones that were found during these investiga-
tions to possess noteworthy biological activity.
One of the diculties encountered clinically during anti-
viral therapy has been the emergence of resistant mutant
viral forms. This was particularly true when agents were
used alone or in two drug combination therapy. This led
to the use of triple combination therapy such as two
NRTIs plus an NNRTI or a protease inhibitor.5 With
such regimens, high rates of virological control can be
achieved3a but there remains a subset of patients who do
not show durable response for reasons of poor com-
Figure 1. Structures of DPC083 and DPC961.
During the preparation of quinazolinthiones,8 an inter-
esting cyclization occurred to a€ord a unique tricyclic
structure. When 5,6-di¯uoro-3-pyridyl-ethynyl analogue
3, prepared by generating 3-pyridyl acetylide in situ from 2
and adding it to PMB-protected ketimine 1,9 was heated
ÃCorresponding author. Tel.: +1-616-833-4459; fax: +1-616-833-4240;
e-mail: je€rey.w.corbett@am.pnu.com
yCurrent address: Pharmacia, 7255-209-643, 301 Henrietta St., Kala-
mazoo, MI 49007-4940, USA.
0960-894X/01/$ - see front matter # 2001 DuPont Pharmaceuticals Company. Published by Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(00)00624-7

212
J. W. Corbett et al. / Bioorg. Med. Chem. Lett. 11 (2001) 211±214
Figure 2. Intramolecular cyclization pathways.
and the resulting aniline protected as the Boc carbamate
to a€ord 7 (Scheme 2). Metallation of 7 using n-BuLi
followed by quenching the anion with ethyl tri¯uoro-
acetate gave the desired tri¯uoromethyl ketone 8. The
protecting group was removed under standard condi-
tions to yield 9.
Scheme 1. Preparation of thiourea 5.
with hot phosphorus oxychloride (POCl3), a mixture of
imidoyl chlorides 4 were obtained after concentration
(Scheme 1).10 Heating the crude imidoyl chloride mix-
ture 4 in a solution of thiourea in ethanol a€orded (2Z)-
8-¯uoro-2-(3-pyridinylmethylene)-2a-(tri¯uoromethyl)-
2a,5-dihydro-2H-thieno[4,3,2-de]-quinazoline-4(3H)-
thione (5). The structure of 5 was determined by single
crystal X-ray analysis11 and by long range H±13C cou-
1
Scheme 2. Preparation of keto-aniline 9.
pling between H-9 and C-10 and C-14. The ole®n stereo-
chemistry was spectroscopically con®rmed by an NOE
from the CF3 to H-9. The ability to con®rm the structure
through NMR techniques became important when ana-
lyzing other tricyclic analogues (i.e., compounds 14±16
in Scheme 3). Thiourea 5 possessed good activity in the
antiviral assays, exhibiting an IC90=7 and 84 nM
against wild-type and the K103N mutant, respectively.12
The tetrasubstituted tri¯uoromethyl ketone 9 was trea-
ted with potassium cyanate to yield aminol 10, which was
subsequently dehydrated in xylenes at re¯ux containing
4 A molecular sieves to give ketimine 11 (Scheme 3).
Alkylation of the ketimine was achieved using lithium
.
acetylides in THF in the presence of BF3 OEt2 to a€ord
12a±c. Treating 12a±c with boron tribromide in CH2Cl2
delivered phenols 13a,b. The facility with which the
intermediate phenol 13c underwent cyclization under the
reaction conditions to a€ord the 3,3a-dihydropyrano-
[4,3,2-de]quinazolin-2(1H)-one 16 precluded its isolation
and characterization. The phenols 13a,b were cyclized onto
the alkyne using either sodium methoxide in methanol or
1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) to a€ord 14
and 15. All of the cyclization processes a€orded exclu-
sively the 6-endo-dig products.
We postulated that 5 represented a constrained ole®n
mimic, and it was already apparent from the quinazoli-
none SAR that ole®n analogues possessed good potency
against wild-type and mutant virus forms.6,7 Based
upon the potency of 5, the e€ect on the resistance pro®le
by con®ning ole®ns such as DPC083 (Fig. 1) to a
de®ned region of space needed further exploration.
We were intrigued by the possibility of substituting the
alkyne moiety present in the quinazolinones with a
nucleophile via an intramolecular cyclization process as
depicted in Figure 2. We are unaware of any reports in
the literature describing an intramolecular 6-endo-dig
ring closure of a phenol onto an alkyne, while two
reports exist describing the 5-exo-dig closure.13
The structural assignments were con®rmed by NMR
1
analysis. For example, long range three bond H±13C
coupling was observed between H-4 and carbon atoms
on the pendant phenyl ring of 15. Also, the ability to
compare the 19F and 1H NMR and NOE di€erence
spectra of the 5-exo-dig product 5 with the 6-endo-dig
product 15 assisted in the structural assignment of the
products. In particular, a strong NOE was observed
between the CF3 and H-9 in 5 which could not occur in
The approach used to synthesize these molecules is
shown in Schemes 2 and 3. Commercially available
2-chloro-5-nitroanisole (6) was reduced with tin chloride

J. W. Corbett et al. / Bioorg. Med. Chem. Lett. 11 (2001) 211±214
213
against the K103N mutant was generally poorer than
the second generation NNRTIs DPC083 and DPC961,
but better than efavirenz. For example, 15 had an
IC90=46 nM against the K103N mutant while DPC083,
DPC961 and efavirenz had IC90s of 27, 10, and 64 nM,
respectively.
Previous SAR on the quinazolinones had shown that
when large R1 groups (i.e., compounds related to 12 in
Scheme 3) were incorporated into the alkyne, the resis-
tance pro®le became substantially worse.7 In contrast to
this observation, the larger R1 is in the tricyclics 14±16,
the better the resistance pro®le is. For instance, the 3-
pyridyl derivative 16 is marginally better against K103N
and K103N/L100I when compared to 12c and DPC961,
respectively. It is also interesting to note the decrease in
activity against the K103N/L100I double mutant for
compound 14, which has R1=cyclopropyl: methyl ether
12a had an IC90=367 nM, while the tricycle 14 had an
IC90=15,000 nM.
Scheme 3. Synthesis of tricyclic quinazolinones.
A possible explanation for the divergence in the SAR
between the quinazolinones and the 3,3a-dihydropyr-
ano[4,3,2-de]quinazolin-2(1H)-ones in their ability to
inhibit the K103N/L100I double mutant form of HIV-1
RT can be posited through an examination of Figure 3.
In this ®gure, the aromatic ring of DPC96115 was
oriented using MacroModel with the aromatic ring of
the energy minimized structure of 15.16 It is evident that
the phenyl group (in yellow) would be expected to
extend further into the NNRTI binding pocket than the
cyclopropyl acetylene (in red), but not as far as a phe-
nylacetylene group would. Therefore, the assumption is
that the larger alkynes in the quinazolinones17 interact
adversely with amino acid residue(s) present in the
NNRTI binding pocket of the K103N/L100I mutant
form of RT, while the phenyl and 3-pyridyl R1 groups in
compounds 15 and 16 do not. Hence, the mutated viral
forms appear better able to accommodate the larger R1
groups in compounds 15 and 16.
15. The structural assignment of 14 and 16 was also
con®rmed through extensive NMR experiments. Struc-
tural assignments could also be determined by the dis-
tinctive chemical shift of the ole®nic protons in 5 and
14±16: the chemical shift of the ole®nic proton in 5 was
7.82 ppm, while in 14±16 the ole®nic signal was between
5.2 and 6.2 ppm.
Interestingly, methyl ether 12a had an excellent resistance
pro®le: IC90=7.5 nM and 377 nM against the K103N and
K103N/L100I mutants, respectively (Table 1).12 The most
potent quinazolinones previously identi®ed had an
IC90=15 nM against K103N mutant14 and 550 nM for
the pure enantiomer (DPC082) against the K103N/
L100I double mutant.6 The activity of phenols 13a,b
against wild-type virus remained in the low nanomolar
range. However, the resistance pro®le of phenol 13a is
not as impressive as the corresponding methyl ether 12a:
IC90=28 nM and 3024 nM against the K103N and
K103N/L100I mutants, respectively.
It remains unclear why the cyclopropyl R1 group present
in 14 su€ered such a dramatic loss in activity toward the
K103N/L100I RT mutant when compared to 12a. A pos-
sible explanation is that the smaller cyclopropyl group of
14 is unable to form critical nonbonding interactions with
residue(s) in the mutated RT binding pocket.
The 3,3a-dihydropyrano[4,3,2-de]quinazolin-2(1H)-ones
14±16 also had IC90 values in the low nanomolar range
against wild-type virus. The activity of these compounds
Table 1. Biological data for compounds 12a±c, 13a,b, and 14±1612
Compounds
RNA IC90
(nM)
Enzyme IC50
(nM)a
K103N IC90
(nM)
K103N/L100I IC90
(nM)
DPC083
DPC961
Efavirenz
12a
12b
12c
13a
13b
14
15
2.1 (Æ0.8)a
27 (Æ11)a
1690 (Æ160)a
1100 (Æ160)a
7300 (Æ5000)a
377
5515
2541
3024
1936
15000
1663
897
2.0 (Æ0.7)a
31 (Æ8)
47 (Æ25)
70 (Æ39)
144 (Æ60)
139 (Æ23)
54 (Æ8)
10 (Æ3.2)a
1.7 (Æ0.5)a
64 (Æ24)a
3.5
7.9
7.5
113
147
28
7.1
3.6
6.8
3.6
8.2
4.6
174 (Æ78)
90 (Æ14)
244 (Æ109)
122 (Æ15)
30
76
46
76
16
aValues are means of three experiments, standard deviation is given in parentheses.

214
J. W. Corbett et al. / Bioorg. Med. Chem. Lett. 11 (2001) 211±214
8. Corbett, J. W. U.S. patent application WO9950253, 1999;
Chem. Abstr. 131:257577.
9. Compound 1 was prepared by minor modi®cation of a
procedure given in ref 7. Namely, formation of the inter-
mediate aminol required warming to 80 ꢀC.
10. The PMB-group was presumably cleaved by adventitious
acid generated during the formation of the iminoyl chlorides.
11. Data for single crystal X-ray of 5: T= 100 ꢀC, space group
P21/n, unit cell: a=9.956(1) A, b=10.953(1) A, c=13.882(1) A,
b=96.88(1)ꢀ, volume=1502.9 A3, Z=4, 1714 unique re¯ec-
tions collected with Iꢁ3.0s(I).
Figure 3. Overlay of DPC961 (red highlights) crystal structure and
minimized structures of 15 (yellow highlights).
In summary, several compounds were identi®ed which
have (or would be expected to have as the single enan-
tiomer) a better K103N/L100I IC90 than either DPC083,
DPC961 or efavirenz. A unique SAR was also identi®ed
for the 3,3a-dihydropyrano[4,3,2-de]quinazolin-2(1H)-
ones which di€ers from that previously reported for the
quinazolinones. Future studies will explore the SAR of
this unique series of compounds.
12. The biological assays were performed as described in ref 6.
13. (a) Dupuy, C.; Crozet, M.-P.; Surzur, J.-M. Bull. Soc.
Chim. Fr. 1980, 2, 361. (b) P¯ieger, D.; Muckenstrum, B. Tet-
rahedron 1989, 45, 2031.
14. 5-Chloro-3,4-dihydro-6-¯uoro-4-(3-methylbutyn-1-yl)-4-
(tri¯uoromethyl)-2(1H)-quinazolinone, compound 22, from
ref 7.
Acknowledgements
15. The structure of DPC961 was obtained from a single
crystal X-ray: T= 100 ꢀC, space group P212121, unit cell: a=
13.851(1) A, b=26.283(2) A, c=8.850(1) A, volume=3221.8 A3,
Z=8, 3786 unique re¯ections collected with Iꢁ3.0s(I).
We thank Will Marshal of DuPont Central Research &
Development Department for X-ray crystallography
determinations and Carolyn Weigelt for assistance with
the molecular modeling.
References and Notes
1. Levy, J. A. Microbiol. Rev. 1993, 57, 183.
2. Fauci, A. S. N. Engl. J. Med. 1999, 1046.
3. (a) Jonckheere, H.; Anne, J.; De Clercq, E. Med. Res. Rev.
2000, 20, 129. (b) Stazewski, S.; Morales-Ramirez, J.;
Tashima, K. T.; Rachlis, A.; Skiest, D.; Stanford, J.; Stryker,
R.; Johnson, P.; Labriola, D. F.; Farina, D.; Marnion, D. J.;
Ruiz, N. M. N. Engl. J. Med. 1999, 34, 1865.
4. Esnouf, R.; Ran, J.; Ross, C.; Jones, Y.; Stammers, D.;
Stuart, D. Nat. Struct. Biol. 1995, 2, 303.
5. Department of Health and Human Services guidelines,
January 28, 2000.
6. Corbett, J. W.; Ko, S. S.; Rodgers, J. D.; Je€rey, S.;
Bacheler, L. T.; Klabe, R. M.; Diamond, S.; Lai, C.-M.;
Rabel, S. R.; Saye, J. A.; Adams, S. P.; Trainor, G. L.;
Anderson, P. S.; Erickson-Viitanen, S. K. Antimicrob. Agents
Chemother. 1999, 43, 2893.
7. Corbett, J. W.; Ko, S. S.; Rodgers, J. D.; Gearhart, L. A.;
Magnus, N. A.; Bacheler, L. T.; Diamond, S.; Je€rey, S.;
Klabe, R. M.; Cordova, B. C.; Garber, S.; Logue, K.; Trainor,
G. L.; Anderson, P. S.; Erickson-Viitanen, S. K. J. Med.
Chem. 2000, 43, 2019.
16. Mohamadi, F.; Richards, N. G. J.; Guida, W. C.; Lis-
kamp, R.; Lipton, M.; Cau®eld, C.; Chang, G.; Hendrickson,
T.; Still, W. C. J. Comp. Chem. 1990, 11, 440.
17. For example, 6-chloro-3,4-dihydro-4-(phenyl-ethynyl)-4-
(tri¯uoromethyl)-2(1H)-quinazolinone, compound 34 from ref
7, has an IC90=7 nM and 250 nM against the wild-type and
K103N viruses, respectively.