Construction of multifunctional modules for drug discovery: Synthesis of novel thia/oxa-azaspiro[3.4]octanes

Article abstract of DOI:10.1021/ol402127b

New classes of thia/oxa-azaspiro[3.4]octanes are synthesized through the implementation of robust and step-economic routes. The targeted spirocycles have been designed to act as novel, multifunctional, and structurally diverse modules for drug discovery.

Full text of DOI:10.1021/ol402127b

ORGANIC
LETTERS
XXXX
Vol. XX, No. XX
000–000
Construction of Multifunctional Modules
for Drug Discovery: Synthesis of Novel
Thia/Oxa-Azaspiro[3.4]octanes
Dong Bo Li,† Mark Rogers-Evans,*,‡ and Erick M. Carreira*,†
Laboratorium fu€r Organische Chemie, ETH Zu€rich, CH-8093 Zu€rich, Switzerland, and
Discovery Research, Chemistry Technologies & Innovation, F. Hoffmann-La Roche AG,
CH-4070 Basel, Switzerland
mark.rogers-evans@roche.com; carreira@org.chem.ethz.ch
Received July 26, 2013
ABSTRACT
New classes of thia/oxa-azaspiro[3.4]octanes are synthesized through the implementation of robust and step-economic routes. The targeted
spirocycles have been designedtoactas novel, multifunctional, andstructurallydiverse modules fordrugdiscovery. Furthermore, enantioselective
approaches to the spirocycles are reported.
Recently we have documented research efforts aimed at
new classes of spirocyclic ring systems designed to expand
the palette of tailored module scaffolds available to medic-
inal chemists, which constitute an important role for syn-
thetic chemistry in the drug discovery process. An essential
component for this process1 is to provide access to specific
molecular topologies with functional group diversity, essen-
tial for generating leads that discriminate among biological
targets, therefore promoting the selectivity and enhancing
the safety profile of the final candidates. In recent endeavors
involving spirocyclic modules for drug discovery, we have
described the generation of angular azaspiro[3.3]heptanes
and the first collection of thia-azaspiro[3.4]octanes (Figure 1,
structures I and II);2 herein we report the expansion of the
azaspiro[3.4]octane family to include complementary struc-
tures (Figure 1, structures III, IV, and V).
Figure 1. Evolution of azaspiro[3.4]octanes incorporating sul-
We have previously reported 6-thia-2-azaspiro[3.4]octanes
fones and oxygens shown with variable exit vectors.
incorporating alcohols.2b The utility of these scaffolds
† ETH Z€urich.
would be significantly enhanced by the design of the
spirocyclic scaffold with additional functional groups,
‡ F. Hoffmann-La Roche AG.
(1) Galloway, W. R. J. D.; Isidro-Llobet, A.; Spring, D. R. Nat.
leading to a collection of molecules with variable exit
vectors. By virtue of the 1,3 relationship between sulfur
and FG in III (Figure 1), it would be possible to utilize 1,3-
dipolar cycloadditions as a convergent means to prepare
Commun. 2010, 1, 80.
(2) (a) For the new azaspiro[3.3]heptanes, see: Burkhard, J. A.;
ꢀ
Guerot, C.; Knust, H.; Carreira, E. M. Org. Lett. 2012, 14, 66–69. (b)
For the thia-azaspiro[3.4]octanes, see: Li, D. B.; Rogers-Evans, M.;
Carreira, E. M. Org. Lett. 2011, 13, 6134–6136.
r
10.1021/ol402127b
XXXX American Chemical Society

an ester, as a variable domain to access other groups, such
as carboxylic acids and amines.
Table 1. Optimization of 1,3-Dipolar Cycloadditions
Conjugated ester 1 was prepared quantitatively by
Wittig olefination3 of N-Boc azetidin-2-one on gram scale
(Table 1). The azetidine product obtained 1 was then
subjected to the 1,3-dipolar cycloaddition reactions with
thiocarbonyl ylid generated in situ through either of two
ways involving the use of sulfoxide or sulfide precursors.
In the first series of attempts, heating bis(trimethylsilyl-
methyl)sulfoxide resulted in a sila-Pummerer rearrange-
ment4 (Table 1, entry 1), which in the presence of 1 yielded
1,3-dipolar cycloadduct 1a as an inseparable mixture,
including largely unreacted 1 (ca. 30% conversion). After
subsequent oxidation with mCPBA, sulfone ester 2 was
then isolated and fully characterized (29% yield over two
steps). Examination of various solvents for this transfor-
mation did not prove helpful. For example, although
HMPA has been recommended as optimal in certain
related cases (Table 1, entry 2), we did not record an
improvement.4b We next scrutinized the use of thiocarbo-
nyl ylid formed in situ from chloromethyl trimethylsilyl-
methyl sulfide. Treatment of the latter with CsF in
CH3CN/HMPA at 80 °C for 30 h (Table 1, entry 3)5 led
to ca. 80% conversion and a 49% overall yield following
subsequent oxidation of the cycloadduct with mCPBA
(Table 1, entry 3). It is worth noting that the results of
entry 3 constitute the first example of a β,β-disubstituted
conjugated ester as a dipolarophile, reacting with a thio-
carbonyl ylid to form the corresponding β,β-disubstituted
thiolane in a useful yield.
conversion
yield
entry
path
conditions
(1H NMR)
of 2a
1
2
a
a
CH3CN, 80 °C, 4 h
HMPA, 100 °C,
overnight
ca. 30%
ca. 15%
29%
13%
3
b
CH3CN/HMPA (10:1),
80 °C, 30 h
ca. 80%
49%
a Total yield after the oxidation with mCPBA.
Following construction of the 6-thia-2-azaspiro[3.4]octane,
we proceeded to elaborate the scaffold (Scheme 1). Saponifi-
cation of ester 2 proceeded smoothly to furnish the cor-
responding acid 3, as the first of the targeted modules.
Subsequent implementation of a Curtius rearrangement
provided orthogonally protected bisamine 4, which was
then unmasked to produce amine 5 in 78% yield after
reductive deprotection.
Scheme 1. Synthesis of Spirocyclic Modules 3 and 5
Further modification in the thiolane of the 5-thia-2-
azaspiro[3.4]octanes2b (Figure 1, structures II f IV) is a
fascinating way to generate the new type modules with two
exit vectors in a different tilted direction. Thus the new
modules11and 12weredesigned and synthesizedasshown
in Scheme 2.
Ester 1 underwent modified Woodward thiolane cycli-
zation (thia-Michael additionÀDieckmann cyclization)6
to provide an inseparable (and inconsequential) mixture
of the two spirocyclic ketoesters 6a and 6b, which were
directly subjected to Krapcho decarboxylation to sulfenyl
ketone 7 in 62% yield over two steps. Reduction of ketone
7 with NaBH4 proceeded well to provide the correspond-
ing alcohol 8 (89%). Transformationof8 to9 was achieved
employing Mitsunobu inversion,7 giving 9 as an insepar-
able mixture, whose subsequent oxidation with mCPBA
furnished the corresponding sulfonyl imide 10in 38% yield
over two steps. Following protecting group removal by
treatment of 10 with hydrazine, targeted amine 11 was
(3) (a) Collier, P. N. Tetrahedron Lett. 2009, 50, 3909–3911. (b) For a
similar example of 1 prepared by the Wittig olefination, see: Burkhard,
ꢀ
J. A.; Guerot, C.; Knust, H.; Rogers-Evans, M.; Carreira, E. M. Org.
Lett. 2010, 12, 1944–1947.
(4) (a) Aono, M.; Hyodo, C.; Terao, Y.; Achiwa, K. Tetrahedron
Lett. 1986, 27, 4039–4042. (b) Aono, M.; Terao, Y.; Achiwa, K.
Heterocycles 1995, 40, 249–260.
(5) (a) For 1,3-dipolar cycloaddition with thiocarbonyl ylid in acet-
€
onitrile under heating conditions, see: Karlsson, S.; Hogberg, H.-E. Org.
Lett. 1999, 1, 1667–1669. (b) For the first report on 1,3-dipolar cycloaddi-
tion with thiocarbonyl ylid in acetonitrile, see: Hosomi, A.; Matsuyama,
Y.; Sakurai, H. J. Chem. Soc., Chem. Commun. 1986, 1073–1074.
(6) (a) Liu, H.-J.; Ngooi, T. K. Can. J. Chem. 1982, 60, 437–439.
(b) Woodward, R. B.; Eastman, R. H. J. Am. Chem. Soc. 1946, 68, 2229–
2235.
(7) Sen, S. E.; Roach, S. L. Synthesis 1995, 756–758.
Org. Lett., Vol. XX, No. XX, XXXX
B

Scheme 2. Synthesis of Spirocyclic Modules 11 and 12
Scheme 3. Synthesis of Spirocyclic Modules 14À16
identification of a promising subunit in a targetted project
generates incentive for studies aimed at developing enantio-
selective approaches. We describe one of these below.
In order to examine whether scaffold construction was
amenable to a more precise spatial vectorization, we also
examined a route to provide 16 enantioselectively. In this
respect, the addition of a vinylmagnesium bromide solu-
tion to N-Boc-3-azetidinone quantitatively afforded the
vinyl alcohol 17, which was subjected to O-allylation to
give diene 18 in 93% yield (Scheme 4). Subsequent pre-
paration of the corresponding dihydrofuran proceeded
smoothly to furnish enol ether 19 (89%) via the cascade
involving [Ru]-catalyzed ring-closing metathesis and
olefin isomerization.12 Asymmetric hydroboration with
(À)-Ipc2BH13 at low temperature followed by oxidative
workup converted spirocyclic dihydrofuran 1914 into alcohol
(À)-(R)-15 in 74% yield with 90% ee.15 Mitsunobu inversion
of the secondary alcohol with phthalimide and subsequent
exposure to hydrazine were utilized for the production of the
corresponding amine (À)-(S)-16 in 74% overall yield and
without any loss in configuration.15 The absolute configura-
tion was unambiguously confirmed by an X-ray diffraction
analysis after the derivatization of chiral amine 16 to its
carbamide 21 (ORTEP format with ellipsoids at 50%
probability).
isolated in quantitative yield. The oxidation of 8 with
mCPBA produced the sulfonyl alcohol 12 in 89% yield.
The tandem conjugate additionÀDieckmann cyclization
protocol described in the preceding paragraph8 offered
opportunities for the construction of oxa-azaspiro[3.4]-
octanes. The sequence starting with 1 and methyl glycolate
smoothly provided the desired spirocyclic ketoester 13
(Scheme 3), which, without purification, was subjected
to Krapcho decarboxylation to form oxa-azospirocyclic
ketone 14 in 61% yield over two steps.9 Its reduction with
NaBH4 furnished the corresponding alcohol 15. Com-
pound 14 could also be subjected to reductive amination
conditions (7 M NH3 in MeOH and titanium(IV) isoprop-
oxide, followed by NaBH4)10 to afford amine 1611 in 53%
yield. The concise approach thus provides access to three
5-oxa-2-azaspiro[3.4]octane modules 14À16 in two to three
steps. The synthetic routes to the building blocks described
above provide access to racemates. Since, at the outset,
it is not clear which enantiomer will be optimal in a given
drug discovery program, we have found that it is most
practical to carry out chromatographic resolution, thereby
providing access to the two enantiomers. The subsequent
Finally, the last modules targeted for construction were
8-functionized 5-oxa-2-azaspiro[3.4]octanes (Scheme 5).
(8) Bunnage, M. E.; Davies, S. G.; Roberts, P. M.; Smith, A. D.;
Withey, J. M. Org. Biomol. Chem. 2004, 2, 2763–2776 and references
therein.
(9) The synthesis of spirocyclic module 14 has been recently reported
by using one [Au]-catalytic spirocyclization protocol; see: Painter, T. O.;
Bunn, J. R.; Schoenen, F. J.; Douglas, J. T.; Day, V. W.; Santini, C.
J. Org. Chem. 2013, 78, 3720–3730.
(12) Schmidt, B. J. Org. Chem. 2004, 69, 7672–7687.
(13) (a) For preparation of (À)-Ipc2BH, see: Brown, H. C.; Singaram,
B. J. Org. Chem. 1984, 49, 945–947. (b) For asymmertric hydroboroation of
2,3-dihydrofuran, see: Brown, H. C.; Gupta, A. K.; Rangaishenvi, M. V.;
Prasad, J. V. N. V. Heterocycles 1989, 28, 283–294.
(14) Intermediate 19 has been recently prepared by using one differ-
ent synthetic approach; see: Kumar, S.; Thornton, P. D.; Painter, T. O.;
Jain, P.; Downard, J.; Douglas, J. T.; Santini, C. J. Org. Chem. 2013, 78,
6529–6539.
(10) Miriyala, B.; Bhattacharyya, S.; Williamson, J. S. Tetrahedron
2004, 60, 1463–1471.
(11) This spirocyclic module has been registered by WuXi AppTec
Co., Ltd. with CAS no.1250998-24-3; However no information is
currently available for its synthesis.
(15) The determinations of enantiomeric excess of (À)-(R)-15 and
(À)-(S)-16 were performed on a Jasco2080Plus SFC after UV-detectable
derivatization to the corresponding benzoates or benzamides.
Org. Lett., Vol. XX, No. XX, XXXX
C

Scheme 4. Enantioselective Synthesis of Spirocyclic Modules 15
and 16
Scheme 5. Synthesis of Spirocyclic Modules 24À26
steps. Next, its treatment with TsOH in acetone proceeded
in 94% yield to give ketone 24. Following reduction with
NaBH4 and reductive amination,10 alcohol 25 and amine
26 were produced.
In summary, we have disclosed approaches for the
construction of novel azaspiro[3.4]octanes as multifunc-
tional modules amenable for further elaboration in drug
discovery. Ten designed modules were prepared in up to
6 steps, and access to two of these in enantioenriched form
is reported. The compact nature of these novel modules
coupled with inherent tunable physicochemical properties
will enable their potential targeting both inside and outside
of the brain for drug discovery projects, as well as assure a
novel intellectualpropertyposition. Our current endeavors
in this direction will be revealed in due course.
Acknowledgment. Dr. W. B. Schweizer and Mr. M.
€
Through the use of Magnus’ allene cyclization method,16
construction of the spirocyclic system is effected as shown
in Scheme 5. Addition of lithiated methoxyallene17 to
N-Boc-3-azetidinone cleanly produced adduct 22 on gram
scale, which was sufficiently pure for subsequent spirocy-
clization (tBuOK/tBuOH, cat. dicyclohexyl-18-crown-6)
to form the spirocyclic enol ether 23 in 63% yield over two
Solar at ETH-Zurich are acknowledged for X-ray struc-
ture determination. We are grateful to F. Hoffmann-La
Roche, the Swiss National Science Foundation, and
€
ETH-Zurich for the generous support of the research
program.
Supporting Information Available. Experimental proce-
dures, compound characterization data, and crystallographic
information file (CIF) for 21. This material is available free
of charge via the Internet at http://pubs.acs.org.
(16) Gange, D.; Magnus, P. J. Am. Chem. Soc. 1978, 100, 7746–7747.
(17) Forpreparationofmethoxyallene, see:Anderson, K. R.;Atkinson,
S. L. G.; Fujiwara, T.; Giles, M. E.; Matsumoto, T.; Merifield, E.;
Singleton, J. T.; Saito, T.; Sotoguchi, T.; Tornos, J. A.; Way, E. L. Org.
Process Res. Dev. 2010, 14, 58–71.
The authors declare no competing financial interest.
D
Org. Lett., Vol. XX, No. XX, XXXX