Spiro-fused pyrrolidine, piperidine, and oxindole scaffolds from lactams

Article abstract of DOI:10.1021/ol302176z

Expedient routes to three classes of novel spiro-fused pyrrolidine, piperidine, and indoline heterocycle scaffolds are described. These threedimensional frameworks, which conform to the "rule of three", are suitably protected to allow for site-selective m

Full text of DOI:10.1021/ol302176z

ORGANIC
LETTERS
XXXX
Vol. XX, No. XX
000–000
Spiro-fused Pyrrolidine, Piperidine, and
Oxindole Scaffolds from Lactams
†
‡
§
†
€
Christoph Hirschhauser, Jeremy S. Parker, Matthew W. D. Perry, Mairi F. Haddow,
and Timothy Gallagher*,†
School of Chemistry, University of Bristol, Bristol, BS8 1TS, U.K., AstraZeneca,
Pharmaceutical Development, Silk Road Business Park, Macclesfield, SK10 2NA,
U.K., and AstraZeneca, Innovative Medicines, R&I Chemistry, AstraZeneca R&D
€
Molndal, Peparedsleden, SE-431 83, Molndal, Sweden
€
T.Gallagher@bristol.ac.uk
Received August 5, 2012
ABSTRACT
Expedient routes to three classes of novel spiro-fused pyrrolidine, piperidine, and indoline heterocycle scaffolds are described. These three-
dimensional frameworks, which conform to the “rule of three”, are suitably protected to allow for site-selective manipulation and functionalization.
Different modes of substrate control were explored, which allow for good to excellent levels of diastereoselectivity and dispense with the need for
additional chiral reagents or catalysts. The concepts developed were applied in short, formal syntheses of (()-coerulescine and (()-horsfiline.
The “rule of three” guides the design of library candi-
dates for fragment-based drug discovery: this rule requires
molecules with a molecular weight of e300 g/mol, a C log
P e 0.3, and a number of hydrogen bond donors and
acceptors of e3.1 Given our general interest in providing
new routes to small, chiral, and “sp3-rich” drug-like het-
erocycles, we have exploited the reactivity of cyclic
sulfamidates2 to generate a set of rigid heterocyclic scaf-
folds, which fit these requirements of molecular weight and
hydrogen bond donor/acceptors. Key considerations were
the generation of an effective three-dimensional environment,
as well as the presence of orthogonally protected func-
tional groups, that allows the ability to “evolve” eventual
hits. Spirobased azacycles have found extensive applica-
tion in medicinal chemistry,3 and this motif is also present
(3) (a) Yamashita, T.; Kamata, M.; Endo, S.; Yamamoto, M.;
Kakegawa, K.; Watanabe, H.; Miwa, K.; Yamano, T.; Funata, M.;
Sakamoto, J.; Tani, A.; Mol, C. D.; Zou, H.; Dougan, D. R.; Sang, B.;
Snell, G.; Fukatsu, K. Bioorg. Med. Chem. Lett. 2011, 21, 6314–6318. (b)
Converso, A.; Hartingh, T.; Garbaccio, R. M.; Tasber, E.; Rickert, K.;
Fraley, M. E.; Yan, Y.; Kreatsoulas, C.; Stirdivant, S.; Drakas, B.;
Walsh, E. S.; Hamilton, K.; Buser, C. A.; Mao, X.; Abrams, M. T.; Beck,
S. C.; Tao, W.; Lobell, R.; Sepp-Lorenzino, L.; Zugay-Murphy, J.;
Sardana, V.; Munshi, S. K.; Jezequel-Sur, S. M.; Zuck, P. D.; Hartman,
G. D. Bioorg. Med. Chem. Lett. 2009, 19, 1240–1244. (c) Kaptein, A.;
Oubrie, A.; de Zwart, E.; Hoogenboom, N.; de Wit, J.; van de Kar, B.;
van Hoek, M.; Vogel, G.; de Kimpe, V.; Schultz-Fademrecht, C.;
Borsboom, J.; van Zeeland, M.; Versteegh, J.; Kazemier, B.; de Roos,
J.; Wijnands, F.; Dulos, J.; Jaeger, M.; Leandro-Garcia, P.; Barf, T.
Bioorg. Med. Chem. Lett. 2011, 21, 3823–3827. (d) Sippy, K. B.;
Anderson, D. J.; Bunnelle, W. H.; Hutchins, C. W.; Schrimpf, M. R.
Bioorg. Med. Chem. Lett. 2009, 19, 1682–1685. (e) Hansen, J. D.; Newhouse,
B. J.; Allen, S.; Anderson, A.; Eary, T.; Schiro, J.; Gaudino, J.; Laird, E.;
Allen, A. C.; Chantry, D.; Eberhardt, C.; Burgess, L. E. Tetrahedron Lett.
2006, 47, 69–72.
† University of Bristol.
‡ AstraZeneca, Pharmaceutical Development.
§ AstraZeneca, Innovative Medicines.
(1) Congreve, M.; Carr, R.; Murray, C.; Jhoti, H. Drug Discov. Today
2003, 8, 876–877.
ꢀ
(2) (a) Bower, J. F.; Svenda, J.; Williams, A. J.; Charmant, J. P. H.;
Lawrence, R. M.; Szeto, P.; Gallagher, T. Org. Lett. 2004, 6, 4727–4730.
ꢁ
For reviews, see: (b) Melendez, R. E.; Lubell, W. D. Tetrahedron 2003,
59, 2581–2616. (c) Bower, J. F.; Rujirawanich, J.; Gallagher, T. Org.
Biomol. Chem. 2010, 8, 1505–1519.
r
10.1021/ol302176z
XXXX American Chemical Society

in a variety of biologically active natural products.4 Con-
sequently, spirocycles A, B, and C emerged as attractive
targets incorporating sites (Figure 1) for functionalization/
derivatization and for which efficient stereocontrol is
available directly.
(essentially complete) diastereoselectivity (Scheme 1).7
While alkylation of 2 with the five-membered cyclic sulfa-
midate 3a proceeded efficiently at room temperature,
reaction of 2 with homologue 3b required heating at
60 °C for 13 h. Completion of the spirocyclic scaffold
was achieved by Cbz cleavage and subsequent base-
mediated cyclization. To facilitate lactamization onto the
tert-butyl ester, strong bases were necessary: the crude
product, after Cbz removal, was heated in THF at 60 °C
with LiHMDS to afford spiro-lactams 5a and 5b, in 69%
and 60% yield, respectively.8 Borane reduction (carried
out in THF at 60 °C) provided the target scaffolds
(of type A) 6a and 6b in 61% and 34% yield, respectively,
where the constituent N- and O-sites are differentiated.9
Scheme 2. Tandem AcylationꢀAlkylation and Spiro-Lactam
Formation; Piperidine-Based Targets A (m = 0, 1; n = 1)
Figure 1. Target spirocycles of types AꢀC.
Scheme 1. Tandem AcylationꢀAlkylation and Spiro-Lactam
Formation; Pyrrolidine-Based Targets A (m = 0, 1; n = 0)
While diastereoselective acylation/alkylation of 1b was
clearly directed by the adjacent silyoxymethyl substituent,
the effect of a more remote 1,3-interaction was more
challenging and hence of interest. Acylation/alkylation of
7 with cyclic sulfamidates 3a and 3b each delivered an
inseparable mixture of diastereomers 8 and 9 (Scheme 2).
1H NMR indicated only a modest diastereomeric excess
(de = 33% with 3a at rt; 20% with 3b at 60 °C). Based on
subsequent conversions, it is likely that the formation of
diastereomers 8a and 8b is favored over 9a and 9b respec-
tively, but no conclusive stereochemical evidence was
obtained at this stage. Amine deprotection and cyclization
was performed on diastereomeric mixtures using sodium
methoxide, with concomitant desilylation, which facilitated
separation of diastereomers 10 and 1110 by chromatogra-
phy. Interestingly, products 10 predominated and only
Initial studies focused on readily available lactams 1b
(prepared from 1a by silylation) and 7.5 Enolization of 1b
with an excess of LiHMDS and subsequent acylation with
Boc2O6 delivered the stabilized enolate 2 which reacted
with cyclic sulfamidates 3a and 3b, generating acylation/
alkylation products of type 4 in one step and with high
(4) For example: (Manzamines) (a) Magnier, E.; Langlois, Y.
Tetrahedron 1998, 54, 6201–6258. (Horsfiline) (b) Jossang, A.; Jossang,
P.; Hadi, H. A.; Sevenet, T.; Bodo, B. J. Org. Chem. 1991, 56, 6527–6530.
(5) Compounds derived from 1a and 7 are racemic, and in Schemes 1
and 2 the relative stereochemistry is depicted. For preparation of 1a, see:
(a) Stanetty, P.; Turner, M.; Mihovilovic, M. Molecules 2005, 10, 367–
375. (b) Gallagher, T.; Derrick, I.; Durkin, P. M.; Haseler, C. A.;
€
Hirschhauser, C.; Magrone, P. J. Org. Chem. 2010, 75, 3766–3774.
€
For preparation of 7, see: (c) Hirschhauser, C.; Haseler, C. A.; Galla-
gher, T. Angew. Chem., Int. Ed. 2011, 50, 5162–5165. For access to
enantimerically pure precursors, see: (d) Lima, E. C.; Domingos,
J. L. O.; Dias, A. G.; Costa, P. R. R. Tetrahedron: Asymmetry 2008,
19, 1161–1165. (e) Galeazzi, R.; Martelli, G.Mobbili, G.; Orena, M.;
Panagiotaki, M. Heterocycles 2003, 60, 2485–2498. (f) Gray, D.; Gallagher,
T. Angew. Chem., Int. Ed. 2006, 45, 2419–2423.
(6) Attempts to substitute Boc2O in this type of sequence with
dimethyl carbonate did not allow for conducting a subsequent alkyla-
tion in the same pot. Since the latter reagent releases 1 equiv of highly
nucleophilic methoxide, side reactions that consume the cyclic sulfami-
date are likely involved (cf. Supporting Information, p 10).
(7) For a reagent-controlled aproach, see: Moss, T. A.; Alonso, B.;
Fenwick, D. R.; Dixon, D. J. Angew. Chem., Int. Ed. 2009, 49, 568–571.
(8) When conducted with NaOMe, the reaction was cleaner and
slightly more efficient; however partial cleavage of the TBS-ether also
occurred.
(9) The relative configurations of 5a and 6b were confirmed by NOE.
(10) The relative configurations of 10a, 11a, and 10b were established
unambiguously by X-ray crystallography. The relative configuration of
11b was confirmed by NOE.
B
Org. Lett., Vol. XX, No. XX, XXXX

small amounts of the corresponding diastereomers 11 were
isolated (10/11 dr ≈ 5:1; de ≈ 66%). The question then
arises as to whether products of type 11 epimerize under
the reaction conditions. This was examined by heating 11a
with sodium methoxide (10 equiv, THF 60 °C, 3 d), but no
conversion to 10a was observed by TLC. Instead, slow
decomposition of 11a occurred, both of which observations
Scheme 4. Synthesis of Scaffolds of Type B
1
were confirmed by H NMR. These results suggest that
although products of type 11 do not epimerize (to 10),
these intermediates nevertheless decompose at a faster rate,
thereby providing a mechanism for diastereomer enrich-
ment. Since these spirolactams are generally highly crystal-
line and compounds of type 10 are significantly less soluble
in chloroform than the epimers (i.e., 11), chromatography-
free purification should be possible for purposes of scale-
up. Finally, borane reduction of 10a and 10b furnished
scaffolds 12a and 12b in 56% and 44% yield, respectively.
alkenes 17a and 17b, respectively, which underwent ring
closing metathesis to give 18a/b.13 Piperidine formation
was achieved with the ozonolysis/reductive amination
conditions recently published by Dussault,14 which deliv-
ered 19a and 19b in 40% and 39% yield. Chemoselective
amine N-debenzylation and subsequent lactam reduction
completed the desired scaffolds 20a and 20b in 59% and
60% overall yield, respectively.15
Scheme 3. Attempted Extension into Scaffolds of Type B
Spiro-fused oxindoles have previously been approached
using in situ generated R-acylated lactam enolates as
vehicles for nucleophilic aromatic substitution (SNAr).16
However, our attempt to apply this strategy to the synthe-
sis of scaffolds of type C failed as the less activated
2-fluoro-1-nitrobenzene did not react with enolate 2 at rt
or 60 °C. Consequently, malonate 21 was chosen as an
alternative starting material.
Alkylation of the enolate of 21 with 1,2-cyclic sulfami-
dates 3a and 3c at 60 °C produced 22a and 22c in 72% and
68% yield, respectively (Scheme 5). In the case of 22c,
unidentified byproducts were formed, which were insepar-
able from the product (purity of 22c ≈ 90%), but these
were easily removed at a later stage. The less reactive 1,3-
cyclic sulfamidate 3b (six-membered and less strained) and
An effort to apply this chemistry in the synthesis of
target scaffolds of type B is outlined in Scheme 3 with
alkylated lactams 14a/b as the focus. Initial attempts to
alkylate directly lactam enolates (generated from 1b or 7
and LiHMDS in THF at ꢀ78 °C) with cyclic sulfamidate
3a, at either rt or 60 °C, did not deliver the desired products
14a/b. We also examined the feasibility of direct (in situ)
decarboxylation of the initial alkylation adducts (i.e.,
N-sulfated 13a/b) derived by the carboxyl-directed alkyla-
tion exploited in Schemes 1 and 2. However, this led to
desilylation without decarboxylation, after heating at
60 °C for 5 h.11 In the case of the pyrrolidine variant 13a,
desilylation was followed by lactonization to give 15 in
89% yield. After amine deprotection, in situ lactone-to-
lactam conversion provided 16 (the desilylated variant of 5a)
in 90% yield.
(13) While this reaction worked well with the GrubbsꢀHoveyda
second generation catalyst in DCM, significant amounts of byproducts
resulting from olefin migration were observed when Grubbs’ second
generation catalyst was employed.
(14) Kyasa, S.; Fisher, T.; Dussault, P. Synthesis 2011, 3475–3481.
(15) Dialdehyde derivatives obtained from 17a, 17b, 18a, and 18b (by
either ozonolysis or bishydroxylation and subsequent diol cleavage)
proved to be unstable and were best transformed into the corresponding
piperidines (19) by immediate reductive amination with benzylamine.
Unfortunately, this sequence was capricious, and although a variety of
conditions were explored, yields remained low and the reaction was
plagued by several unidentified byproducts.
(16) (a) Bella, M.; Kobbelgaard, S.; Jørgensen, K. A. J. Am. Chem.
Soc. 2005, 127, 3670–3671. (b) Cowley, A. R.; Hill, T. J.; Kocis, P.;
Moloney, M. G.; Stevenson, R. D.; Thompson, A. L. Org. Biomol.
Chem. 2011, 9, 7042. (c) Sen, S.; Potti, V. R.; Surakanti, R.; Murthy,
Y. L. N.; Pallepogu, R. Org. Biomol. Chem. 2011, 9, 358. (d) Kobbelgaard,
S.; Bella, M.; Jørgensen, K. A. J. Org. Chem. 2006, 71, 4980–4987.
(17) Kulkarni, M. G.; Dhondge, A. P.; Chavhan, S. W.; Borhade,
A. S.; Shaikh, Y. B.; Birhade, D. R.; Desai, M. P.; Dhatrak, N. R.
Beilstein J. Org. Chem. 2010, 6, 876–879.
(18) An attempt to reduce and cyclize lactam 23a under standard
conditions (Pd/C, H2, THF, rt, o/n) was not successful. Judging by
1H NMR no cyclization had occurred, although the nitro group had
been (at least partially) reduced.
The synthesis of scaffolds of type B was achieved using
an alternative solution, albeit still based on lactam enolate
reactivity (Scheme 4).12 Bisallylation of 1b and 7 provided
(11) Refluxing 8a/9a (but following isolation) with p-TsOH H2O in
3
THF for 14 h (as opposed to 5 h) did deliver 14b in 55% yield, but given
the facile formation of 15 from 1b (via 13a) this approach was discarded
in favor of the more general strategy illustrated in Scheme 4.
(12) For a related aproach, also see ref 3e.
Org. Lett., Vol. XX, No. XX, XXXX
C

the 5-substituted variant 3d did not react with 21 at rt and
underwent decomposition at 60 °C; this may reflect the
extensive delocalization/lower reactivity of the enolate of
21. Cyclization of 22a and 22c proceeded readily under
basic conditions (LiHMDS, THF, ꢀ78 °C to rt), delivering
lactams 23a/c. Since Cbz cleavage occurred under these
conditions, quenching with reactive alkyl halides (BnBr or
MeI) provided a convenient means of further N-substitution.
substituent R1, and diastereomer 23c was formed exclu-
sively in 67% yield.
Scaffolds of type C were completed by nitro group
reduction of 23 and lactamization. The N-substituted
derivatives 23a-1 and 23a-2 cyclized readily upon reduc-
tion (Pd/C, THF, H2, 1 atm, rt), delivering 24a-1 and 24a-2
in 100% and 89% yield, respectively. Selectivereduction of
the N-substituted lactam of 24a-1 to deliver (()-coerules-
cine and subsequent functionalization at C-50 of the aro-
matic ring to complete (()-horsfiline have been described
previously.17 Formation of the spirocycle from the pyrro-
lidone 23a was best achieved by nitro group reduction with
iron in acetic acid.18 The 5-benzyl substituted congener 23c
formed a 4:1 mixture of 24c and its C-3 epimer under these
conditions, indicating lactam opening and reclosure. Re-
duction and cyclization of 23c-1 were conducted in THF
with Pd/C and ammonium formate at 60 °C.16c This
reaction initially delivered the corresponding lactam N-
hydroxide (which was isolated and characterized), and
resubmission of this material to the reaction conditions
delivered desired lactam 24c-1 in 63% yield from 23c-1.
In conclusion, three types of lead-like, spiro-fused scaf-
folds have been synthesized with core structures which are
in agreement with the “rule of three”. These frameworks
also incorporate residues that are equipped with functionality
suitable for selective further manipulation. By exploiting the
inherent chirality of a range of different starting materials
(i.e., cyclic sulfamidates), additional chiral reagents and
catalysts were avoided. Finally, short, formal syntheses of
(()-coerulescine and (()-horsfiline have been achieved.
Scheme 5. Synthesis of Scaffolds of Type C Including Formal
Syntheses of (()-Coerulescine and (()-Horsfiline
Acknowledgment. Funding by AstraZeneca is ack-
nowledged.
Supporting Information Available. Experimental pro-
cedures; crystallographic details for 10a/b, 11a, and 24a-
1; and 1H and 13C NMR spectra for all new compounds.
This material is available free of charge via the Internet at
http://pubs.acs.org.
Cyclization of 22c was of particular interest since dia-
stereoselectivity is dependent on the preference for one of
two diastereotopic esters, induced by the relatively remote
The authors declare no competing financial interest.
D
Org. Lett., Vol. XX, No. XX, XXXX