Synthesis of four novel natural product inspired scaffolds for drug discovery

Article abstract of DOI:10.1021/jo802456w

Inspired by the novel spiro structures of a number of bioactive natural products such as the histrionicotoxins, a series of novel spiro scaffolds have been designed and robust syntheses developed. The scaffolds are ready-to-use building blocks and can be easily prepared on a 5-20 g scale. They contain two amino groups (one Boc-protected) and have been designed for ease of conversion to a lead generation library, using either amide formation or reductive amination procedures. The synthesis of the 1,9-diazaspiro[5.5]undecane and 3,7-diazaspiro[5.6]dodecane ring systems was achieved using RCM as the key step. A simple workup procedure is reported for the removal of highly colored ruthenium residues. The synthesis of the 1,8-diazaspiro[4.5]decane scaffold has been achieved using a bromine-mediated 5-endo cyclization of the corresponding 4-aminobutene intermediate under acidic conditions. This is the first example of this type of cyclization to be reported. A novel mechanism involving a bromine transfer reaction from an initially formed bromonium ion to a neighboring nitrogen atom is suggested as the reason for the failure of this type of reaction under "normal" bromination conditions. An unusual rearrangement of a 1-acyl-1,9-diazaspiro[5.5]undecane to the corresponding 9-acyl-1,9-diazaspiro[5.5]undecane is reported.

Full text of DOI:10.1021/jo802456w

Synthesis of Four Novel Natural Product Inspired Scaffolds for
Drug Discovery
Ian D. Jenkins,*,† Fabienne Lacrampe,† Justin Ripper,† Lilian Alcaraz,‡ Phuc Van Le,†
George Nikolakopoulos,† Priscila de Almeida Leone,† Rodney H. White,† and
Ronald J. Quinn†
Eskitis Institute, Griffith UniVersity, Brisbane, QLD, 4111, Australia, and Department of Medicinal
Chemistry, AstraZeneca R&D Charnwood, Bakewell Road, LE11 5RH, U.K.
i.jenkins@griffith.edu.au
ReceiVed NoVember 3, 2008
Inspired by the novel spiro structures of a number of bioactive natural products such as the histrionicotoxins,
a series of novel spiro scaffolds have been designed and robust syntheses developed. The scaffolds are
ready-to-use building blocks and can be easily prepared on a 5-20 g scale. They contain two amino
groups (one Boc-protected) and have been designed for ease of conversion to a lead generation library,
using either amide formation or reductive amination procedures. The synthesis of the 1,9-
diazaspiro[5.5]undecane and 3,7-diazaspiro[5.6]dodecane ring systems was achieved using RCM as the
key step. A simple workup procedure is reported for the removal of highly colored ruthenium residues.
The synthesis of the 1,8-diazaspiro[4.5]decane scaffold has been achieved using a bromine-mediated
5-endo cyclization of the corresponding 4-aminobutene intermediate under acidic conditions. This is the
first example of this type of cyclization to be reported. A novel mechanism involving a bromine transfer
reaction from an initially formed bromonium ion to a neighboring nitrogen atom is suggested as the
reason for the failure of this type of reaction under “normal” bromination conditions. An unusual
rearrangement of a 1-acyl-1,9-diazaspiro[5.5]undecane to the corresponding 9-acyl-1,9-diazaspiro[5.5]undecane
is reported.
Introduction
The major hurdle encountered in the development of a newly
discovered NP lead is often its structural complexity. The lead
compound may have adequate potency and selectivity but poor
pharmacokinetics, so the challenge for the medicinal chemist
is to turn it into a drug by appropriate modification. In many
cases, this is not trivial for reasons of molecular complexity
and synthetic accessibility. One possible approach is to use a
NP scaffold, a mono- or polycyclic structure having several
functional groups available for elaboration.5 This approach
retains some of the features of the original bioactive natural
product structure (such as a novel ring system and conformation)
but allows the medicinal chemist to introduce more druglike6,7
functional groups. Moreover, as the natural products database
contains many more scaffolds than the drugs database, such
Finding small molecules that selectively modulate protein
function remains one of the major challenges in drug discovery.
Natural products (NPs), which have been an invaluable source
of drug leads for the pharmaceutical industry over many years,1,2
interact with proteins during their biosynthesis, and there is good
reason to believe that this binding to biosynthetic enzymes may
correlate with binding to certain drug targets.3 NPs can therefore
be considered as “privileged structures” or biologically validated
starting points for the design of compound libraries.4
* To whom correspondence should be addressed. Tel: + (07) 3735 6025.
Fax: + (07) 3735 6001.
† Griffith University.
‡ AstraZeneca R&D Charnwood.
(4) Arve, L.; Voigt, T.; Waldmann, H. QSAR Comb. Sci. 2006, 25, 449–
456.
(5) Khersonsky, S. M.; Chang, Y.-T. Comb. Chem. High Throughput Screen
2004, 7, 645–652.
(1) Lam, K. S. Trends Microbiol. 2007, 15, 279–289.
(2) Newman, D. J.; Cragg, G. M. J. Nat. Prod. 2007, 70, 461–477.
(3) McArdle, B. M.; Quinn, R. J. ChemBioChem 2007, 8, 788–798.
1304 J. Org. Chem. 2009, 74, 1304–1313
10.1021/jo802456w CCC: $40.75 2009 American Chemical Society
Published on Web 12/23/2008

Natural Product Inspired Scaffolds for Drug DiscoVery
SCHEME 1. Design of Scaffold 7
for the reasons noted above, we preferred to target orthogonally
protected diamines, so we decided to replace the carbonyl by a
nitrogen atom as in scaffold 7. Scaffold 7 has the added
advantage over scaffold 6 that it would not give rise to
diastereoisomeric mixtures.
FIGURE 1. Natural products containing spiro rings.
unexplored scaffolds may be promising new starting points in
drug discovery.8 We were interested in scaffolds that were of
low molecular weight (<350), “three-dimensional” rather than
flat (most commercially available scaffolds are planar mol-
ecules), and contained functional groups spread around the
molecule. Compounds containing two amino groups (orthogo-
nally protected) were preferred because amide formation and
reductive amination are robust reactions that are amenable to
automation and library synthesis. We were particularly attracted
to the novel spiro scaffolds present in a number of natural
products such as histrionicotoxin (1),9 schelhammericine (2),10
manzamine A (3),11 and halichlorine (4),12 which contain the
1-azaspiro[5.5]undecane, 7-azaspiro[5.6]dodecane, 2,7-diaza-
spiro[4.5]decane, and 6-azaspiro[4.5]decane ring systems, re-
spectively (Figure 1). For example, the histrionicotoxins (such
as 1), which are isolated from the skin of the Colombian poison
dart frog, Dendrobates histrionicus, are noncompetitive inhibi-
tors of nicotinic acetylcholine receptors and inspired the novel
scaffold 7 (the histrionicotoxins are actually not toxic; the family
was originally misnamed because of the source; the main toxins
in poison dart frogs are pumiliotoxins and batrachotoxin).13 In
arriving at scaffold 7, we first considered the scaffold 5, which
reflects the main functional groups present in 1 (Scheme 1).
For reasons of chemical tractability (5 could undergo ring-
opening via a retro-Michael reaction), we moved the carbonyl
to the adjacent 9-position (6). In principle, scaffold 6 could be
readily converted into a library by reductive amination with a
series of secondary amines, followed by removal of the Boc
group and acylation with a series of acid chlorides. However,
Results and Discussion
Spiro[5.5] Compounds. In developing a synthesis of the
scaffold 7 and related spirocyclic diamino compounds, we
required a method that was robust and could be carried out on
a 10 g scale with the minimum number of chromatography steps.
The starting material was the relatively inexpensive 1-benzyl-
4-piperidone (8). Treatment of 8 with allylamine (4 equiv) in
toluene overnight at 60 °C in the presence of anhydrous
potassium carbonate gave the imine 9 (Scheme 2) in high yield
(97%). The crude product was then treated with allylmagnesium
bromide to give the aminodiene 10 in good yield (82%). An
attempted ring-closing metathesis reaction (RCM)14 on 10 using
Grubbs first-generation catalyst (12) gave very low yields of
the ring-closed product, so the secondary amino group was
protected as the trifluoroacetamide 11 (attempted Boc protection
of 10 gave only low yields, presumably because of the very
hindered nature of the secondary nitrogen adjacent to a tertiary
center). The trifluoroacetamide 11 was obtained in good yield
(70%) following purification by chromatography on silica gel.
Treatment of 11 with 12 (7-9 mol %) in DCM at room
temperature under an atmosphere of nitrogen for 24 h followed
by filtration through a plug of amino-bonded silica gel to remove
the residues of ruthenium gave the spiro compound 13 in
quantitative yield. The yields in the RCM reaction varied
between 70-100% depending on the mol % of Grubbs catalyst
12 used. With 7 mol % of 12, the reaction generally didn’t go
to completion; close to 9 mol % was required for quantitative
conversion. It might be possible to reduce the amount of catalyst
by employing Grubbs second-generation catalyst,15 but this
catalyst is much more expensive than 12. To remove the highly
colored ruthenium byproduct from the reaction, we first tried
the “stirring with DMSO overnight” method.16 However, we
subsequently found that a more efficient procedure was to filter
the reaction mixture through a plug of amino-bonded silica. We
would recommend this extremely simple and fast procedure for
(6) Walters, W. P.; Murcko, M. A. Curr. Opin. Chem. Biol. 1999, 3, 384–
387.
(7) Leeson, P. D.; Davis, A. M.; Steele, J. Drug DiscoVery Today: Technol.
2004, 1, 189–195.
(8) Ortholand, J.-Y.; Ganesan, A. Curr. Opin. Chem. Biol. 2004, 8, 271–
280.
(9) For a review, see: Sinclair, A.; Stockman, R. A. Nat. Prod. Rep. 2007,
24, 298–326.
(10) Johns, S. R.; Lamberton, J. A.; Sioumis, A. A. Aust. J. Chem. 1969,
22, 2219–31.
(11) (a) Nishida, A.; Nagata, T.; Nakagawa, M. Top. Heterocycl. Chem. 2006,
5, 255–280. (b) Magnier, E.; Langlois, Y. Tetrahedron 1998, 54, 6201–6258.
(12) (a) Kuramoto, M.; Tong, C.; Yamada, K.; Chiba, T.; Hayashi, Y.;
Uemura, D. Tetrahedron Lett. 1996, 37, 3867–3870. (b) Arimoto, H.; Hayakawa,
I.; Kuramoto, M.; Uemura, D. Tetrahedron Lett. 1998, 39, 861–862.
(13) Daly, J. W.; Spande, T. F.; Garraffo, H. M. J. Nat. Prod. 2005, 68,
1556–1575.
(14) Trnka, T. M.; Grubbs, R. H. Acc. Chem. Res. 2001, 34, 18–29.
(15) Wright, D. L.; Schulte, J. P., II; Page, M. A. Org. Lett. 2000, 2, 1847–
1850.
(16) Ahn, Y. M.; Yang, K.; Georg, G. I. Org. Lett. 2001, 3, 1411–1413.
J. Org. Chem. Vol. 74, No. 3, 2009 1305

Jenkins et al.
SCHEME 3. Synthesis of Scaffolds 17, 19, and 21
SCHEME 2. Synthesis of Scaffold 7
the clean up of RCM reactions involving basic substrates (we
have not tried it with nonbasic substrates).17
SCHEME 4. Synthesis of Scaffolds 22 and 23
The trifluoroacetyl protecting group was removed by refluxing
in methanol with potassium carbonate for 3 h, followed by
installation of the Boc group and purification by chromatography
on silica gel to give 14 in high yield (91%). Presumably, the
secondary nitrogen is less hindered in the spiro compound
(where the two allyl groups have been shortened and “tied
back”) than in compound 10, but reaction with Boc anhydride
was still sluggish and took 3 days at room temperature.
The final step was a combined hydrogenation/hydrogenolysis
in the presence of 1 equiv of HCl in methanol to give the desired
scaffold 7 as its white crystalline hydrochloride salt 15 in high
yield (90%). In summary, the scaffold 7 was synthesized in six
steps in an overall yield of 47%. Each of the steps is robust
and can be carried out easily in the laboratory on a 10-20 g
scale. Only two chromatography steps are required. The
intermediate 13 can be obtained in four steps and 58% overall
yield and can be easily converted into the complementary
scaffolds 17, 19, and 21 (Scheme 3). None of these compounds
has been described previously (the intermediate 18 has been
prepared from the more expensive 1-Boc-4-piperidone, but scant
experimental details were provided).18
Two other closely related scaffolds prepared were the
benzamides 22 and 23. Benzamide 22 was prepared on a 20 g
scale in three steps from the intermediate 10 (Scheme 4) using
a procedure analogous to that in Scheme 2 (five steps from
N-benzylpiperidone 8, overall yield 49%). During the synthesis
of 22, it was observed that a small amount of the isomeric
benzamide 23 was formed on standing in methanol. Upon further
investigation, it was found that 22, a viscous oil, rearranged to
23 even in the absence of solvent (after 1 week at 70 °C,
virtually none of the more polar 22 remained). In methanol,
(17) See also: McEleney, K.; Allen, D. P.; Holiday, A. E.; Crudden, C. M.
Org. Lett. 2006, 8, 2663–2666.
(18) Gracias, V.; Gasiecki, A. F.; More, J. D.; Akritopoulou-Zanze, I.; Djuric,
S. W. Tetrahedron Lett. 2006, 47, 8977–8980.
1306 J. Org. Chem. Vol. 74, No. 3, 2009

Natural Product Inspired Scaffolds for Drug DiscoVery
SCHEME 5. Synthesis of Scaffold 29
the reaction appears to be bimolecular rather than unimolecular,
as it is concentration-dependent. Presumably, the mechanism
of this unusual and unexpected rearrangement involves attack
by the free NH group of one molecule on the amide carbonyl
of another, although it is conceivable that the process is
intramolecular in the absence of solventsa study of molecular
models suggests that if the “upper” piperidine ring adopts a boat
conformation, and the amide nitrogen is axial, the free NH is
in very close proximity to the amide carbonyl. We believe that
the unusual lability of the acyl group on the 1-nitrogen is a
result of steric assistance.19 Normally, the amide group would
be expected to be flat, but in these spiro compounds, the benzoyl
group is subject to steric compression by hydrogens on the
adjacent ring (the hydrogens on C-8 or C-10 if the nitrogen is
axial, and the hydrogens on C-7 or C-11 if equatorial). We
suggest that this causes the bond between the carbonyl and the
nitrogen atom to twist out of the plane, thereby reducing the
resonance interaction between the nitrogen lone pair and
the carbonyl, and thus activating the carbonyl toward nucleo-
philic attack. This suggestion is supported by the infrared
stretching frequency of the carbonyl group in 22 (ν ) 1636
cm-1 for the HCl salt, 24), which is considerably higher than
that in the isomeric benzamide 23 (ν ) 1611 cm-1 for the HCl
salt, 25).
It is important to be aware of the potential lability of acyl
groups on the 1-nitrogen of these scaffolds when preparing an
amide library from scaffolds 17 or 19, for example. Once the
protecting group (Boc or benzyl respectively) is removed, a free
NH group is generated and there is the potential for acyl
migration (as in 22 to 23) on standing or on heating. We have
found that this migration does not occur if the free amine is
converted to its hydrochloride salt. Thus, the hydrochloride salts
of 22 (24) and 23 (25) are stable compounds and can be stored
for more than a year without change.
Spiro[5.6] Compounds. The spirocyclic diamino scaffold 29
was inspired by the spiro system present in the cephalotaxus
alkaloid schelhammericine10 (2). Our approach to this scaffold
was the same as that outlined in Scheme 2 for the synthesis of
the spiro[5.5] intermediate 13: addition of 3-butenylmagnesium
bromide to the imine 9, followed by protection of the nitrogen
and RCM, should give the corresponding spiro[5.6] intermediate.
Unfortunately, the 3-butenyl Grignard reagent did not undergo
addition to the imine, a result consistent with the findings of
others.18 The imine 9 presumably undergoes proton abstraction
by the Grignard rather than addition, as the only product isolated
after workup is the N-benzylpiperidone 8. Since allylmagnesium
bromide adds readily to these imines, we were able to prepare
the desired scaffold 29 by interchanging the allyl and butenyl
groups, i.e., by first preparing the imine from 3-butenylamine
and then addition of allylmagnesium bromide. Protection of the
amino group as its trifluoroacetate gave the intermediate 26
(68% yield from 8 after purification by chromatography).
Treatment of 26 with Grubbs catalyst 12 gave the spiro[5.6]
intermediate 27 in high yield (89%). It was found that the RCM
reaction to form the 7-membered ring (Scheme 5) was not as
efficient as that to form the 6-membered spiro compound 13.
To effect complete conversion of 26 to 27 required 15 mol %
of catalyst (added in two portions) and stirring in DCM for 3
days. The intermediate 27 was then converted into the desired
3-Boc-3,7-diazaspiro[5.6] scaffold 29 by standard protection/
deprotection procedures. As in the case of scaffold 7, we
required a synthesis of 29 that would produce sufficient material
for preparation of a small library. We were able to achieve this
with only two chromatography steps, but the overall (unopti-
mised) yield was rather low (5%, seven steps from N-benzyl-
4-piperidone).
Spiro[4.5] Compounds. The spirocyclic scaffold 34 was
inspired by the spiro system present in halichlorine (4)12 and
manzamine A (3),11 the latter containing a diazaspiro system
within its complex polycyclic skeleton. However, the synthesis
of the spiro[4.5] scaffold 34 presented more of a challenge. As
was found with the butenyl Grignard reagent, treatment of the
imine 9 with vinylmagnesium bromide resulted only in the
isolation of N-benzylpiperidone 8. With no simple way of
converting the piperidone 5 into an imine bearing a vinyl group
on the nitrogen, it was not possible to employ the RCM strategy
used for the synthesis of 7 or 29. As the allyl Grignard addition
to an imine chemistry worked so well and could be scaled up
to at least 50 g, we considered a bromine-mediated cyclization
of the 4-aminobutene intermediate 31 (prepared in 89% yield
from N-benzylpiperidone 8 and benzylamine using the same
conditions as those used for the preparation of intermediate 10).
However, treatment of 31 with bromine in DCM failed to give
the cyclized product 32 (Scheme 6). We also investigated
N-bromosuccinimide, iodine, N-iodosuccinimide, phenylsele-
nium chloride, and phenylselenium bromide as alternative
electrophilic reagents. Only phenylselenium bromide resulted
in a cyclized product, but the yield was low (18%) and this
reaction was not investigated further.
We thought that in the case of attempted bromine-mediated
cyclization, 31 might be undergoing preferential bromination
on nitrogen, thereby preventing a cyclization reaction. It was
decided, therefore, to carry out the bromination on the HBr salt
of 31. This strategy was successful. Thus, 31 was treated with
HBr (2.5 equiv) in methanol (generated by addition of acetyl
bromide to methanol) and evaporated to give a foam. DCM
/methanol (10:1) was added and the mixture cooled to -78 °C,
followed by the addition of bromine (1.5 equiv). The mixture
was stirred at 0 °C for 3 h, potassium carbonate and water were
added, and stirring was continued for a further 16 h at rt. The
bromo spiro[4.5] product 32 was isolated from the DCM layer
in excellent yield (90%) after purification by chromatography.
Conversion of 32 to the required scaffold 34 required
reduction of the bromo group, removal of the two benzyl
(19) ElielE. L. WilenS. H. ManderL. N. Stereochemistry of Organic
Compounds; John Wiley & Sons: New York, 1994; p 721.
J. Org. Chem. Vol. 74, No. 3, 2009 1307

Jenkins et al.
SCHEME 7. Cyclization Mediated by Bromonium Ions
SCHEME 6. Synthesis of Scaffold 34
owing to the difficulties connected with this kind of reaction.22
We suggest that the main reason for the failure of this type of
reaction is competitive bromination on nitrogen (to give an
unstable N-bromo derivative),23 possibly because the 5-endo-
tet process 38 is disfavored by Baldwin’s rules.24 Although the
formation of a N-bromo derivative could occur directly, we
believe it is more likely to occur via a bromine transfer reaction
from an initially formed bromonium ion 39 (to give 40). In
support of this proposed mechanism, a key step in the recent
synthesis of (-)-epibatidine was the bromination of a 4-ami-
nocyclohexene. The dibromide could only be obtained when
the bromination was carried out in the presence of a large (10-
fold) excess of Et4N+Br-.25 In the present work, the bromination
was carried out in the presence of HBr. This achieves the same
purpose as addition of Et4N+Br-; it increases the concentration
of bromide ion (which favors interception of the bromonium
ion 39, to give the dibromide 35), but also, by protonating the
nitrogen atom, formation of an N-bromo compound is prevented.
In summary, the simple expedient of carrying out the
bromination on the protonated amino alkene, as illustrated here,
has enabled the synthesis of a pyrolidine ring system from a
4-aminobutene. This methodology may be more generally
applicable to the synthesis of heterocycles containing nitrogen
and complements the iodine-mediated cyclization of but-1-enyl-
amines.26,27
Spiro[3.5] Compounds. Synthesis of the spiro[3.5] scaffold
41 containing a 4-membered (azetidine) ring was a logical
extension to the series of scaffolds 29, 7, and 34 (containing
7-, 6-, and 5-membered rings, respectively). However, the core
structure of this 1,7-diazaspiro[3.5] system (i.e., 41 without the
Boc protecting group) is well-known in the patent literature,28
so we turned our attention instead to the novel scaffold 43.
Surprisingly, although the ketone 42 has been prepared, scaffold
43 has not been reported. Ketone 42 is readily prepared on a
large (20 g) scale in four steps from 1-Boc-4-piperidone using
literature procedures.29 Reductive amination of 42 with ben-
zylamine and sodium triacetoxyborohydride followed by hy-
protecting groups, and introduction of the Boc group on the
8-nitrogen. This was achieved in a one-pot procedure by
hydrogenolysis in the presence of Boc anhydride at 75 °C for
2 days (the higher temperature was necessary for the reduction
of the bromo compound) to give 34 (38%). We attribute the
rather poor yield to partial loss of the Boc group either during
the reaction or during workup and possibly also the formation
of some di-Boc product (both the free diamine and the di-Boc
products would be removed during the reaction workup). We
did not optimize this reaction.
The cyclization reaction (31 to 32) almost certainly proceeds
via bromine addition to the olefin to give the dibromide 35,
followed by a 5-exo-tet cyclization.20 In support of this, the mass
spectrum (electrospray, +ve ion) of the reaction mixture after
the addition of bromine but before the addition of the potassium
carbonate, showed peaks at 479, 481, and 483, corresponding
to the protonated dibromide 35, whereas after stirring overnight
with potassium carbonate, these peaks had disappeared, being
replaced by peaks at 399 and 401, corresponding to the
protonated monobromo derivative 32.
Interestingly, there are no known examples of bromonium
ion-mediated cyclizations of the type 36 to 37 (Scheme 7) where
R ) alkyl or hydrogen,21 and the direct cyclization of amines
to heterocycles containing nitrogen has rarely been employed
(22) Cardillo, G.; Orena, M. Tetrahedron 1990, 46, 3321–3408.
(23) There is recent literature precedent for this in the conversion of
quinotoxine into N-bromoquinotoxine: Smith, A. C.; Williams, R. M. Angew.
Chem., Int. Ed. 2008, 47, 1736–1740.
(24) Smith, M. B.; March, J. March’s AdVanced Organic Chemistry; 6th
ed.; John Wiley & Sons: Hoboken, NJ, 2007; p 306.
(25) Lee, C.-L. K.; Loh, T.-P. Org. Lett. 2005, 7, 2965–2967.
(26) Diaba, F.; Puigbo´, G.; Bonjoch, J. Eur. J. Org. Chem. 2007, 303, 8–
3044.
(27) Chang, K.-T.; Jang, K. C.; Park, H.-Y.; Kim, Y.-K.; Park, K. H.; Lee,
W. S. Heterocycles 2001, 55, 1173–1179.
(28) See, for example: Adam, G.; Cesura, A.; Jenck, F.; Kolczewski, S.;
Roever, S.; Wichmann, J. Eur. Pat. Appl. EP 970957, 2000.
(29) (a) Trost, B. M.; Bogdanowicz, M. J. J. Am. Chem. Soc. 1973, 95, 5298–
5307. (b) Finke, P. E.; Loebach, J. L.; Parker, K. A.; Plummer, C. W.; Mills,
S. G. U.S. Pat. Appl. US 2005070609, 2005.
(20) There are a number of examples of cyclization of 4-bromoamines; see,
for example: (a) Minin, P. L.; Walton, J. C. J. Org. Chem. 2003, 68, 2960–
2963. (b) Booth, H.; King, F. E.; Mason, K. G.; Parrick, J.; Whitehead, R. L. St.
D. J. Chem. Soc. 1959, 1050–4. (c) Drake, N. L.; Ross, A. B. J. Org. Chem.
1958, 23, 794–6. (d) Prelog, V.; Szpilfogel, S. HelV. Chim. Acta 1945, 28, 178–
82.
(21) Cyclization is possible when R ) acyl; see, for example: (a) Movassaghi,
M.; Schimidt, M. A.; Ashenhurst, J. Angew. Chem., Int. Ed. 2008, 47, 1485–
1487. (b) Gomez-Sanchez, E.; Soriano, E.; Marco-Contelles, J. J. Org. Chem.
2007, 72, 8656–8670. (c) Kapferer, P.; Vasella, A. HelV. Chim. Acta 2004, 87,
2764–2789. (d) Corey, E. J.; Loh, T. P.; AchyuthaRao, S.; Daley, D. C.; Sarshar,
S. J. Org. Chem. 1993, 58, 5600–2.
1308 J. Org. Chem. Vol. 74, No. 3, 2009

Natural Product Inspired Scaffolds for Drug DiscoVery
SCHEME 8. Spiro[3.5] Scaffolds
2.41-2.44 (t, J ) 6.0 Hz, 2H), 3.24 (s, 2H), 3.84-3.85 (d, J )
5.5 Hz, 2H), 5.05-5.07 (dd, J ) 10.0 Hz, J ) 1.5 Hz, 1H),
5.22-5.26 (dd, J ) 2.0 Hz, J ) 17.0 Hz, 1H), 5.99-6.06 (m, 1H),
7.10-7.20 (m, 5H). Imine 9 was not purified but used directly in
the next step. To the imine 9 (18.78 g) dissolved in toluene (80
mL) was added allylmagnesium bromide (100 mL, 1 mol/L in ether)
at 0 °C under N2. The mixture was stirred overnight at room
temperature. Water (750 mL) was added to the mixture. The organic
layer was separated, evaporated to dryness, and set aside. The
aqueous layer was filtered and extracted with DCM (4 × 250 mL).
The organic layers were combined, and the solvent was concentrated
to a volume of 200 mL. The resulting solution was washed with
water (100 mL), dried over MgSO4, and evaporated to dryness to
give the diallyl derivative 10 (18.92 g, 82%). 1H NMR (500 MHz,
CDCl3): δ 1.60-1.62 (m, 4H), 2.21 (d, J ) 8.0 Hz, 2H), 2.43-2.54
(m, 4H), 3.16 (d, J ) 5.5 Hz, 2H), 3.54 (s, 2H), 5.08-5.15 (m,
3H), 5.22-5.25 (m, 1H), 5.80-5.89 (m, 1H), 5.94-6.01 (m, 1H),
7.25-7.39 (m, 5H). 13C NMR (125 MHz, CDCl3): δ 35.2 (C3/5),
42.0 (4-allyl), 44.0 (N-allyl), 49.5 (C2/6), 52.4 (C4), 63.4 (benzyl),
115.3 (4-allyl), 118.2 (N-allyl), 127.0, 128.3, 129.2 (benzyl), 134.1
(4-allyl), 138.0 (N-allyl), 139.0 (benzyl). LRMS (ESI): calcd for
(M + H+) C18H26N2 271, found 271. The diallyl derivative 10 was
not purified but used directly in the next step. To 10 (18.9 g, 70
mmol) were added DMAP (862 mg, 7.0 mmol) and DCM (90 mL)
under N2. TFA anhydride (15 mL, 108 mmol) was added at 0 °C.
The mixture was stirred overnight at room temperature. DCM (300
mL) was added. The mixture was washed with saturated NaHCO3
solution (200 mL) and water (250 mL) and dried over MgSO4.
The solvent was evaporated and the product purified by flash
chromatography (DCM, 2% MeOH, 0.1% NEt3) to give 1-benzyl-
4-allyl-4-(N-allyl-N-trifluoroacetylamino)piperidine (11, 17.9 g,
70%). 1H NMR (500 MHz, CDCl3): δ 2.10-2.14 (m, 2H),
2.31-2.37 (m, 4H), 2.69-2.71 (m, 2H), 2.85-2.86 (d, J ) 7.0
Hz, 2H), 3.52 (s, 2H), 4.03-4.04 (m, 2H), 5.08-5.14 (m, 2H),
5.20-5.24 (m, 2H), 5.64-5.73 (m, 1H), 5.82-5.88 (m, 1H),
7.27-7.36 (m, 5H). 13C NMR (125 MHz, CDCl3): δ 33.70 (C3/5),
35.3 (4-allyl), 47.1 (N-allyl), 50.2 (C2/6), 63.1 (benzyl), 64.6 (C4),
116.7 (q, J ) 288 Hz, CF3), 117.1 (4-allyl), 119.4 (N-allyl), 127.4,
128.6. 129.4 (benzyl), 133.0 (4-allyl), 136.5 (N-allyl), 138.3
(benzyl), 158.1 (q, J ) 34 Hz, CO). LRMS (ESI): calcd for (M +
H+) C20H25F3N2O 367, found 367.
drogenolysis gave the desired scaffold 43 in good yield (84%)
on a 5-10 g scale. The (racemic) product was purified by
conversion to its hydrochloride salt and recrystallization. We
did not attempt to resolve 43 into its enantiomers as we
considered that the racemic compound was adequate for the
preparation of a lead generation library.
In summary, inspired by the novel spiro structures of a
number of bioactive natural products, we have developed
methods for the synthesis of four new scaffolds for drug
discovery. Related spirocyclic scaffolds have had a relatively
high success rate in the discovery of chemical entities interfering
with biological systems, especially in the GPCR area.30 The
spiro[4.5], [5.5], and [5.6] systems described here give the
scaffolds a 3-D shape that reflects the original natural product
architecture. The scaffolds contain two amino groups (one Boc-
protected) and have been designed for ease of conversion to a
lead generation library, using either amide formation or reductive
amination procedures. The scaffolds are ready-to-use building
blocks and can be easily prepared on a 5-20 g scale. Some
interesting new chemistry has come out of this work. The first
was the discovery of an unusual rearrangement of a 1-acyl-1,9-
diazaspiro[5.5]undecane to the corresponding 9-acyl-1,9-
diazaspiro[5.5]undecane. The second was the development of
a simple methodology for effecting the bromine-mediated
5-endo cyclization of 4-aminobutenes, a reaction that has not
previously been reported. A novel mechanism involving a
bromine transfer reaction from an initially formed bromonium
ion to a neighboring nitrogen atom is suggested as the reason
for the failure of this type of reaction under “normal” bromi-
nation conditions.
tert-Butyl 1,9-Diazaspiro[5.5]undecane-1-carboxylate Hydro-
chloride (15). A solution of the diene 11 (10.0 g, 27.3 mmol) in
dichloromethane (220 mL) was evacuated and filled three times
with nitrogen before addition of Grubbs catalyst (1st generation,
2.00 g, 2.43 mmol). The reaction mixture was stirred under an
atmosphere of nitrogen for 24 h. The reaction mixture was then
filtered through a plug of amino-bonded silica, and the residues
were washed with DCM (3 × 100 mL). The solvent was removed
from the filtrate in vacuo to give the spiro compound 13 (9.23 g,
∼100%). 1H NMR (500 MHz, CDCl3): δ 1.75-1.80 (m, 2H),
2.12-2.30 (m, 4H), 2.65-2.71 (m, 4H), 3.54 (s, 2H), 4.00 (bs,
2H), 5.68-5.70 (m, 1H), 5.79-5.81 (m, 1H), 7.27-7.36 (m, 5H).
13C NMR (125 MHz, CDCl3): δ 33.8 (C7/11), 35.3 (C5), 43.0 (C2),
49.9 (C8/10), 58.3 (C6), 63.0 (benzyl), 116.4 (q, J ) 289 Hz, CF3),
123.5 (C2), 126.0 (C3), 127.5, 128.5, 129.5 (benzyl), 137.7 (benzyl),
158.8 (q, J ) 34 Hz, CO). LRMS (ESI): calcd for (M + H+)
C18H21F3N2O 339, found 339. The spiro intermediate 13 was not
purified but used directly in the next step. A solution of 13 (8.19
g, 24.20 mmol) and potassium carbonate (6.69 g, 48.41 mmol) in
MeOH (160 mL) was heated at reflux for 3 h. The reaction mixture
was cooled and the solvent removed in vacuo. Water (200 mL)
was added to the residue and the mixture extracted with chloroform
(3 × 100 mL). The organic phase was dried (MgSO4) and filtered
and the solvent removed in vacuo. The crude residue was then
dissolved in acetonitrile (160 mL), di-tert-dibutyl carbonate (10.56
g, 48.40 mmol) added, and the reaction mixture stirred for 3 d.
The solvent was then removed in vacuo, saturated sodium hydrogen
carbonate (100 mL) added, and the mixture extracted with
Experimental Section
1-Benzyl-4-allyl-4-(N-allyl-N-trifluoroacetylamino)piperidine (11).
1-Benzyl-4-piperidone 8 (16 g, 84 mmol) and anhydrous K2CO3
(24.3 g, 17.6 mmol) were stirred in toluene (160 mL) under N2.
Allylamine (26 mL, 347 mmol, 4 equiv) was added at room
temperature. The mixture was stirred overnight at 60 °C, filtered,
and evaporated to dryness to give the imine 9 (18.78 g, 97%). H
NMR (500 MHz, C6D6): δ 1.99-2.01 (t, J ) 6.0 Hz, 2H),
2.17-2.20 (t, J ) 5.7 Hz, 2H), 2.34-2.36 (t, J ) 5.7 Hz, 2H),
1
(30) (a) See, for example: Tang, F.-Y.; Qu, L.-Q.; Xu, Y.; Ma, R.-J.; Chen,
S.-H. Synth. Commun. 2007, 37, 3793–3799. (b) Gasiecki, A. F.; Cross, J. L.;
Henry, R. F.; Gracias, V.; Djuric, S. W. Tetrahedron Lett. 2008, 49, 6286–
6288. (c) Ihle, D. C.; Li, G.; Wustrow, D. J.; Hodgetts, K. PCT Int. Appl. WO
2008024692, 2008, 68 pp. (d) Schudok, M.; Wagner, M.; Bauer, A.; Kohlmann,
A. PCT Int. Appl. WO 2007137738, 2007, 153 pp. (e) Kusumi, K.; Kokubo,
M.; Ochiai, H.; Shibayama, S. PCT Int. Appl. WO 2007132846, 2007, 178 pp.
(f) Berk, S. C.; Close, J.; Hamblett, C.; Heidebrecht, R. W.; Kattar, S. D.; Kliman,
L. T.; Mampreian, D. M.; Methot, J. L.; Miller, T.; Sloman, D. L.; Stanton,
M. G.; Tempest, P.; Zabierek, A. A. U.S. Pat. Appl. US 2007117824, 2007, 96
pp. (g) Boerjesson, L.; Connolly, S.; Johansson, H.; Kristoffersson, A.; Linnanen,
T.; Shamovsky, I.; Skrinjar, M. PCT Int. Appl. WO 2007030061, 2007, 257 pp.
(h) Eriksen, B. L.; Peters, D.; Nielsen, E. O.; Redrobe, J. P. PCT Int. Appl. WO
2007000463, 2007, 27 pp. (i) Torisu, K.; Nakai, Y.; Egashira, H.; Sakai, M.
PCT Int. Appl. WO 2006006490, 2006, 161 pp. (j) Janssens, F. E.; Schoentjes,
B.; Coupa, S.; Poncelet, A. P.; Simonnet, Y. R. F. PCT Int. Appl. WO
2005097795, 2005, 107 pp. (k) Tsuchiya, Y.; Nomoto, T.; Osawa, K.; Kawakami,
K.; Owaki, K.; Nishikibe, M. Jpn. Pat. JP 2001039950, 2001, 36 pp.
J. Org. Chem. Vol. 74, No. 3, 2009 1309

Jenkins et al.
chloroform (3 × 100 mL). The organic phase was dried (MgSO4)
and filtered and the solvent removed in vacuo to give a residue
that was purified by flash chromatography using ethyl acetate/
catalyst Pd/C (1.31 g, 1.2 mmol, 10%) was added. The flask was
evacuated and flushed with N2 followed by H2. The mixture was
stirred overnight at room temperature under an atmosphere of
hydrogen (balloon). The reaction mixture was filtered through
Celite, evaporated to dryness, and purified by flash chromatography
(DCM, 1% MeOH) to give compound 18 as a white solid (2.2 g,
77%). 1H NMR (500 MHz, CDCl3): δ 1.45 (s, 9H), 1.49-1.53 (m,
2H), 1.69-1.73 (m, 4H), 1.80-1.82 (m, 2H), 2.80-2.85 (m, 2H),
3.27-3.32 (m, 2H), 3.50-3.57 (m, 4H). 13C NMR (125 MHz,
CDCl3): δ 16.9 (C4), 24.2 (C3), 28.6 (Boc), 31.5 (C5), 32.0 (C7/
11), 40.4 (C8/10), 41.5 (C2), 60.7 (C6), 79.8 (C-O), 116.8 (q, J
) 288 Hz, CF3), 155.0 (CO), 157.2 (q, J ) 34 Hz, CO). LRMS
(ESI): calcd for (M + Na+) C16H25F3N2O3 373, found 373.
Compound 18 (624 mg, 1.78 mmol) was dissolved in MeOH/H2O
(1:1, 18 mL) and K2CO3 (1.24 g, 9.0 mmol, 5 equiv) added. The
mixture was stirred for 2.5 h at 90 °C, evaporated to dryness, and
purified by chromatography on neutral alumina (DCM, Pent 90:
10, then DCM, and DCM 2% MeOH) to give tert-butyl 1,9-
diazaspiro[5,5]undecane-9-carboxylate (19) (332 mg, 83%). 1H
NMR (500 MHz, CDCl3): δ 1.35-1.52 (m, 19H), 2.72-2.74 (m,
2H), 3.27-3.31 (m, 2H), 3.38-3.41 (m, 2H). 13C NMR (125 MHz,
CDCl3): δ 20.1 (C4), 27.2 (C3), 28.5 (Boc), 35.5 (b, C7/11), 36.4
(C5), 39.5 (b, C8/10), 40.7 (C2), 49.4 (C6), 79.2 (C-O), 155.0
(CO). HRMS (ESI): calcd for (M + H+) C14H26N2O2 255.2067,
found 255.2068.
1
hexane (3/7) to give the alkene 14 (7.55 g, 91%). H NMR (500
MHz, CDCl3): δ 1.47 (s, 9H), 1.71-1.75 (m, 2H), 2.16 (m, 2H),
2.41 (t, J ) 10 Hz, 2H), 2.56-2.58 (m, 2H), 2.69 (s, 2H), 3.60 (s,
2H), 3.99 (s, 2H), 5.63-5.65 (m, 1H), 5.70-5.72 (m, 1H),
7.25-7.40 (m, 5H). 13C NMR (125 MHz, CDCl3): δ 28.8 (Boc),
34.8 (C7/11), 35.7 (C5), 44.0 (C2), 50.2 (C8/10), 55.1 (C6), 63.1
(benzyl), 80.0 (C-O), 124.9 (C3), 125.5 (C4), 127.7, 128.6, 129.8
(benzyl), 156.1 (CO). LRMS (ESI): calcd for (M + H+) C21H30N2O2
343, found 343. The alkene 14 was converted directly into the
desired scaffold 15 by careful addition of 1.0 equiv of HCl followed
by reduction/hydrogenolysis: Acetyl chloride (1.82 mL, 25.55
mmol) was dissolved in methanol (150 mL). The resulting solution
of HCl in methanol was added to a solution of alkene 14 (8.75 g,
25.55 mmol) in methanol (725 mL), and to this solution was added
10% palladium on carbon (6.15 g). The mixture was evacuated
and flushed three times with hydrogen and allowd to stir under an
atmosphere of hydrogen for 18 h. The reaction mixture was then
filtered through Celite, and the residues were washed with methanol
(3 × 200 mL). The methanol was reduced to a minimum in vacuo
and toluene added. The solvent was again reduced to a minimum
in vacuo causing the product to precipitate. The product was
collected by filtration with washing with toluene and ether and
drying to give tert-butyl 1,9-diazaspiro[5.5]undecane-1-carboxylate
1-Benzyl-1,9-diazaspiro[5.5]undecane (21). Compound 18 (3.68
g, 10.5 mmol) and K2CO3 (7.29 g, 52.8 mmol, 5 equiv) was
dissolved in methanol (50 mL) and water (50 mL). The reaction
mixture was heated for 2.5 h at 90 °C. Benzyl bromide (2 mL,
16.8 mmol) was added dropwise at room temperature. The reaction
mixture was stirred overnight at room temperature, and then the
methanol was evaporated. The aqueous layer was extracted with
DCM (3 × 50 mL). The organic layer was washed with water (2
× 50 mL), dried with MgSO4, and then evaporated to dryness. The
product was purified by flash chromatography on silica (DCM, 1%
MeOH) to give the spiro compound 20 (2.49 g, 89%) as a pale
yellow solid. 1H NMR (500 MHz, CDCl3): δ 1.48-1.70 (m, 17H),
1.87-1.91 (m, 2H), 2.55-2.57 (m, 2H), 3.13-3.18 (m, 2H), 2.65
(s, 2H), 3.80 (bs, 2H), 7.21-7.37 (m, 5H). 13C NMR (125 MHz,
CDCl3): δ 20.4 (C4), 24.0 (C3), 28.7 (Me), 31.7 (b, C7/11), 40.2
(b, C8/10), 46.0 (C2), 51.9 (benzyl CH2), 54.1 (C6), 79.5 (C-O),
126.7, 128.4 (Ph), 155.2 (CO). LRMS (ESI): calcd for (M + H+)
C21H32N2O2 345, found 345. Compound 20 (2.49 g, 7.23 mmol)
was dissolved in DCM (110 mL) under N2. TFA (6.6 mL, 88.8
mmol) was added dropwise at 0 °C. The mixture was stirred for
2 h at 0 °C and then overnight at room temperature. The reaction
mixture was evaporated to dryness. The crude product was dissolved
in toluene (50 mL) and evaporated twice, then dissolved in water
(20 mL) and evaporated to dryness. At this stage, the yellow oil
was dissolved in methanol (10 mL) and toluene (20 mL) and then
evaporated to dryness to give the trifluoroacetic acid salt of
1-benzyl-1,9-diazaspiro[5.5]undecane (21.CF3COOH, 3.02 g, 88%)
1
hydrochloride (15, 6.67 g, 90%) as a white solid. H NMR (500
MHz, CDCl3): δ 1.45 (s, 9H, Boc), 1.49 (m, 2H), 1.61 (m, 2H),
1.68 (m, 2H), 1.77 (app t, J ) 12.5 Hz, 2H), 2.96 (app d, J ) 14.5
Hz, 2H), 3.16, (m, 2H), 3.32 (app d, J 12.5 Hz, 2H), 3.43 (app t,
J 5.5 Hz, 2H, H2), 9.40 (br s, 2H, NH2). 13C NMR (125 MHz,
CDCl3): δ 19.5 (C4), 25.0 (C3), 28.7 (Boc), 32.5 (C5), 35.7 (C7/
11), 40.8 (C8/10), 42.7 (C2), 56.0 (C6), 80.4 (C-O), 155.9 (CO).
HRMS (ESI): calcd for (M + H+) C14H26N2O2 255.2067, found
255.2075.
9-Benzyl-1,9-diazaspiro[5.5]undecane (17). The spiro inter-
mediate 13 (3.85 g, 11.4 mmol) was dissolved in EtOH/AcOH (9:
1, 110 mL) under N2. The catalyst PtO2 (232 mg, 1.1 mmol, 10%)
was added. The flask was evacuated and flushed with N2 followed
by H2. The mixture was stirred 2 h at room temperature under an
atmosphere of hydrogen (balloon) and evaporated to dryness. After
dilution with CH2Cl2 (200 mL), the crude product was washed with
saturated NaHCO3 solution (200 mL) and water (200 mL). The
organic layer was dried (MgSO4) and evaporated to dryness. The
crude product was purified by flash chromatography (DCM, 1%
1
MeOH) to give compound 16 (2.9 g, 75%). H NMR (500 MHz,
CDCl3): δ 1.57-1.63 (m, 2H), 1.67 (m, 4H), 1.77-1.78 (m, 2H),
2.43-2.47 (m, 4H), 2.87-2.92 (m, 2H), 3.48-3.51 (m, 4H),
7.24-7.35 (m, 5H). 13C NMR (125 MHz, CDCl3): δ 17.3 (C4),
24.9 (C3), 31.9 (C5), 32.3 (C7/11), 41.5 (C2), 50.0 (C8/10), 61.0
(C6), 62.8 (C-O), 116.9 (q, J ) 287 Hz, CF3), 127.2, 128.4, 129.3,
138.6 (benzyl), 157.1 (q, J ) 34 Hz, CO). LRMS (ESI): calcd for
(M + H+) C18H23F3N2O 341, found 341. Compound 16 (835 mg,
2.46 mmol) was dissolved in MeOH/H2O (1:1, 20 mL), and K2CO3
(1.64 g, 11.9 mmol, 5 equiv) was added. The mixture was stirred
for 5 h at 90 °C. The crude product was evaporated to dryness and
then purified by chromatography on basic alumina (DCM then
DCM, 2% MeOH) to give 9-benzyl-1,9-diazaspiro[5.5]undecane
1
as a yellow solid. H NMR (500 MHz, DMSOd): δ 1.54 (bs, 3H),
1.73-1.82 (m, 2H), 2.09-2.25 (m, 4H), 2.44-2.48 (m, 1H), 2.90
(bs, 1H), 3.07-3.14 (m, 3H), 3.31-3.33 (bs, 2H), 4.11 (bs, 1H),
4.55-4.58 (m, 1H), 7.40-7.48 (m, 5H): 8.89 (bs, 1H), 9.17 b(s,
1H), 9.65 (bs, 1H).13C NMR (125 MHz, DMSOd): δ 17.4 (C4),
20.8 (C3), 25.0 (C5), 27.0 (C7/11), 32.0 (C8/10), 46.1 (C2), 51.6
(benzyl), 63.0 (C6), 117.4 (q, J ) 296 Hz, CF3), 129.4, 130.1, 130.8,
132.0 (benzyl), 159.1 (q, J ) 34 Hz, CO). HRMS (ESI): calcd for
(M + H+) C16H24N2 245.2012, found 245.2008.
1
(17, 440 mg, 73%). H NMR (500 MHz, CDCl3): δ 1.41-1.43
(m, 2H), 1.48-1.50 (m, 2H), 1.52-1.55 (m, 2H), 1.60-1.63 (m,
4H), 2.34-2.39 (m, 2H), 2.42-2.46 (m, 2H), 2.76-2.79 (m, 2H),
3.50 (s, 2H), 7.22-7.32 (m, 5H). 13C NMR (125 MHz, CDCl3): δ
20.3 (C4), 27.1 (C3), 35.7 (C7/11), 36.6 (C5), 40.7 (C2), 49.3 (C6),
49.5 (C8/10), 63.5, 127.2, 128.4, 129.4, 138.7 (benzyl). HRMS
(ESI): calcd for (M + H+) C16H24N2 245.2012, found 245.2001.
{The trifluoroacetic acid salt can be converted to the free base
by passage through a column of basic alumina (DCM, 2% MeOH,
0.5%NH3) to give 1-benzyl-1,9-diazaspiro[5,5]undecane (21, 89%
from 20) as a yellow solid. 1H NMR (500 MHz, CDCl3): δ
1.46-1.69 (m, 8H), 1.84-1.88 (m, 2H), 2.54-2.56 (m, 2H), 2.63
(bs, 1H), 2.78-2.82 (m, 2H), 2.98-3.02 (m, 2H), 3.69 (s, 2H),
7.20-7.23 (m, 1H), 7.19-7.22 (m, 2H), 7.27-7.28 (m, 2H). 13C
NMR (125 MHz, CDCl3): δ 20.4 (C4), 24.7 (C3), 32.0 (C5), 32.7
tert-Butyl 1,9-Diazaspiro[5.5]undecane-9-carboxylate (19).
The spiro intermediate 13 (2.75 g, 8.14 mmol) was dissolved in
MeOH (64 mL) under N2. (Boc)2O (5.62 g, 25.7 mmol, 3 equiv)
was added. The flask was evacuated and flushed with N2. The
1310 J. Org. Chem. Vol. 74, No. 3, 2009

Natural Product Inspired Scaffolds for Drug DiscoVery
(C7/11), 43.0 (C8/10), 46.2 (C2), 52.1 (benzyl), 54.3 (C6), 126.5,
128.3, 128.35, 142.2 (benzyl). LRMS (ESI): calcd for (M + H+)
C16H24N2 245, found 245.}
25.6 (C3), 30.5 (C5), 33.5 (C7/11), 40.4 (C8/10), 45.7 (C2), 56.5
(C6), 127.7, 129.2, 130.6, 139.1 (Ph), 173.8 (CO). HRMS (ESI):
calcd for (M + H+) C16H23N2O 259.1805, found 259.1809. IR: ν
1-Benzoyl-1,9-diazaspiro[5.5]undecane Hydrochloride (22). A
solution of 1-benzyl-4-piperidone (8, 18.9 g, 100 mmol), allylamine
(30.0 mL, 400 mmol), and potassium carbonate (27.6 g, 200 mmol)
in toluene (150 mL) was heated at 60 °C, under nitrogen, for 18 h.
The reaction mixture was cooled and filtered through a Buchner
funnel with fritted disk. The residues were washed with fresh
toluene (3 × 50 mL), and the solvent from the combined filtrates
was removed in vacuo. The residue was dissolved in dry toluene
(150 mL), the solution cooled to 5 °C, and 1 M allylmagnesium
bromide in diethyl ether (110 mL, 110 mmol) added, slowly, under
an atmosphere of nitrogen. The resulting mixture was allowed to
warm to room temperature and stirred for 2.5-3 h. Hydrochloric
acid (1 M, 300 mL, 300 mmol) was then added, and the phases
were separated. The aqueous phase was basified with 28%
ammonium hydroxide solution and extracted with chloroform (3
× 100 mL). The organic phase was dried (MgSO4) and filtered
and the solvent removed in vacuo to give the diallyl derivative 13
(24.80 g, 92%) of sufficient purity for the following step. To
compound 13 (24.80 g, 91.7 mmol) in toluene (225 mL) was added
benzoyl chloride (12.00 mL, 103 mmol) and the resulting solution
heated at reflux for 18 h. The reaction mixture was cooled and
filtered through a plug of amino bonded silica (100 g) and the
residue eluted with DCM (3 × 100 mL). The solvent was removed
from the filtrate in vacuo and the residue purified by flash
chromatography using methanol/dichloromethane (1/9) to give
N-benzoyl-10 (22.43 g, 65%). 1H NMR (CDCl3, 500 MHz): δ 2.26
(m, 2H), 2.36 (m, 2H), 2.45 (m, 2H), 2.67 (m, 2H), 2.97 (d, J )
7.5 Hz, 2H), 3.58 (s, 2H), 3.89 (d, J ) 5.5 Hz, 2H), 5.00-5.19 (m,
4H), 5.76 (m, 1H), 5.87 (m, 1H), 7.29-7.36 (m, 10H). 13C NMR
(CDCl3, 125 MHz): δ 33.6 (C3/5), 37.4 (4-allyl), 50.1 (N-allyl),
50.2 (C2/6), 61.6 (C4), 63.2 (benzyl), 116.5 (4-allyl), 118.5 (N-
allyl), 126.4, 127.3, 128.48, 128.51, 129.1, 129.5 (benzyl), 134.4
(4-allyl), 137.4 (N-allyl), 138.2, 139.8 (benzyl), 174.2 (CO). LRMS
(ESI): calcd for (M + H+) C25H30N2O 375, found 375.
1636 cm-1
.
9-Benzoyl-1,9-diazaspiro[5.5]undecane Hydrochloride (23). A
solution of 22 (15.24 g, 51.7 mmol) in dichloromethane (100 mL)
was treated with amino-bonded silica (150 g) and filtered, and the
silica was washed with dichloromethane (3 × 100 mL). The solvent
was removed from the filtrate in vacuo, and the residue was heated
neat at 70 °C for 1 week. The residue was purified by flash
chromatography on neutral alumina using chloroform. The purified
product was dissolved in methanol and converted to the salt by
addition of methanolic hydrogen chloride. Toluene was then added
and the solvent removed in vacuo to give the hydrochloride salt of
9-benzoyl-1,9-diazaspiro[5.5]undecane (23 ·HCl) as a light tan solid
(11.96 g, 78%). 1H NMR (500 MHz, DMSO-d6): δ 1.58 (br s, 2H),
1.68 (m, 2H), 1.79 (m, 2H), 1.88 (br s, 4H), 3.00 (br s, 2H), 3.30
(br s, 2H), 3.53 (br s, 1H), 4.11 (br s, 1H), 7.37 (m, 2H), 7.47 (m,
3H), 9.34 (br s, 2H). 13C NMR (125 MHz, DMSO-d6): δ 18.0 (C4),
22.3 (C3), 30.3 (C5), 32.4 (b, C7/11), 37.4 (C8), 39.1 (C2), 43.1
(C10), 55.7 (C6), 127.3, 129.2, 130.3, 136.6 (Ph), 169.7 (CO); (125
MHz, CDCl3): δ 18.2 (C4), 22.2 (C3), 31.4 (C5), 33.0 (C7/11),
37.4 (b, C8), 39.8 (C2), 43.0 (b, C10), 56.1 (C6), 127.2, 128.8,
130.1, 135.6 (Ph), 170.7 (CO). HRMS (ESI): calcd for (M + H+)
C16H23N2O 259.1805, found 259.1811. IR: ν 1611 cm-1
.
tert-Butyl 3,7-Diazaspiro[5.6]dodecane-3-carboxylate Hydro-
chloride (29). 1-Benzyl-4-piperidone (5, 51.48 g, 0.272 mol) was
dissolved in toluene (300 mL). K2CO3 (112.78 g, 0.816 mol) was
added followed by 3-butenylamine hydrochloride (29.104 g, 0.272
mol, prepared from 3-buten-1-ol by mesylation, displacement with
trifluoroacetamide in the presence of potassium carbonate in
refluxing toluene, followed by deacylation with potassium carbonate
in methanol). The mixture was stirred at 60 °C for 20 h. After
cooling, the mixture was filtered through diatomaceous earth and
concentrated to dryness to afford the corresponding imine as a
brown oil (64.21 g, 98%). The crude imine (64.21 g, 0.265 mol)
was dissolved in toluene (250 mL) and cooled in an ice bath under
N2. Allylmagnesium bromide in diethyl ether (265 mL, 0.265 mol)
was added and the mixture was stirred overnight while warming
to room temperature. The mixture was then cooled back down to
0 °C and acidified with 1 M HCl to ∼pH 2. The toluene layer was
separated, and the aqueous layer was treated with solid K2CO3 (∼
pH 12). The organic material was extracted into CHCl3 (7 × 200
mL), dried (MgSO4), filtered, and concentrated to afford the
butenylamine analogue of 10 as a red oil (61.55 g, 82%). Without
purification, a portion of the red oil (18.46 g, 0.065 mol) was
dissolved in CH2Cl2 (250 mL) and cooled to -78 °C. Trifluoroacetic
anhydride (13.597 mL, 0.098 mol) was added dropwise and the
mixture stirred overnight while warming to room temperature.
CH2Cl2 (200 mL) was added and the mixture washed with saturated
NaHCO3 (2 × 200 mL) and H2O (2 × 200 mL). The organic layer
was dried (MgSO4), filtered, and concentrated. The product was
chromatographed on a SiO2 column eluting with Hex/EtOAc (4:1)
to afford the diene 26 as a red/orange oil (20.70 g, 84%). 1H NMR
(500 MHz, CDCl3): δ 2.07 (m, 2H), 2.23-2.44 (m, 6H), 2.74 (m,
2H), 2.87 (m, 2H), 3.38 (m, 2H), 3.54 (s, 2H), 5.07-5.19 (m, 4H),
5.64-5.74 (m, 2H), 7.28-7.39 (m, 5H). 13C NMR (125 MHz,
CDCl3): δ 33.0, 34.0, 37.3, 43.9, 49.0, 63.1, 63.9, 116.8 (q, J )
288 Hz, CF3), 117.8, 119.2, 127.4, 128.5, 129.3, 129.4, 131.0, 132.9,
133.7, 157.5 (q, J ) 40 Hz, CO). LRMS (ESI): calcd for (M +
H+) C21H27F3N2O 381, found 381.
To a solution of N-benzoyl-10 (19.65 g, 52.5 mmol) in
dichloromethane (200 mL) was added Grubbs I catalyst (4.5 g),
and the resulting solution was stirred for 48 h. Amino-bonded silica
(45 g) was added, the mixture was then filtered through a plug of
amino-bonded silica (250 g), and the residues were eluted with
dichloromethane (6 × 200 mL). The solvent was removed from
the filtrate in vacuo and the residue purified by flash chromatography
using methanol/dichloromethane (1/9) to give the 1-N-benzoyl
analogue of 14 [1-benzoyl-9-benzyl-1,9-diazaspiro[5.5]undec-3,4-
1
ene] (16.47 g, 91%). H NMR (CDCl3, 500 MHz): δ 1.86 (br s,
2H), 2.35 (bs, 2H), 2.47 (m, 2H), 2.83 (m, 4H), 3.70 (s, 2H), 3.79
(m, 2H), 5.50 (m, 1H), 5.77 (m, 1H), 7.37 (m, 10H). 13C NMR
(CD3OD, 125 MHz): δ 33.4 (C7/11), 34.9 (C5), 47.4 (C2), 48.9
(C8/10), 55.6 (C6), 62.5 (benzyl), 124.2 (C2), 124.9 (C4), 127.3,
127.5, 128.0, 128.5, 129.9, 130.5, 136.2, 138.1 (benzyl), 174.9 (CO).
LRMS (ESI): calcd for (M + H+) C23H26N2O 347, found 347.
To solution of 1-benzoyl-9-benzyl-1,9-diazaspiro[5.5]undec-3,4-
ene (25.63 g, 73.98 mmol) in methanol (1 L) were added acetyl
chloride (5.26 mL, 73.98 mmol) and 10% palladium on carbon (12
g), and the mixture was evacuated and flushed three times with
hydrogen. The reaction mixture was stirred under an atmosphere
of hydrogen until the reaction was complete. The mixture was
filtered through Celite, and the residues were washed with methanol
(3 × 200 mL). Most of the solvent was removed from the combined
filtrate in vacuo and toluene (50 mL) added. The product was
precipitated by the addition of dry ether, collected by decanting
and washing with ether, and dried under vacuum to give the
hydrochloride salt of 1-benzoyl-1,9-diazaspiro[5,5]undecane
The diene 26 (49.94 g, 0.131 mol) was dissolved in CH2Cl2 (300
mL) and degassed. Grubbs first-generation catalyst (10.78 g, 13.1
mmol, 10 mol %) was added, and the reaction was stirred at room
temperature for 48 h. A second batch of Grubbs first-generation
catalyst (5.39 g, 6.55 mmol, 5 mol %) was added, and the reaction
was stirred for another 24 h. The solvent was then removed, and
the crude material was flushed through a plug of amino-bonded
SiO2. The crude material was concentrated to dryness to afford
1
(22·HCl, 19.65 g, 90%) as a white solid. H NMR (DMSO-d6,
500 MHz): δ 1.41 (m, 2H), 1.64 (m, 2H), 1.72 (m, 4H), 3.03 (m,
4H), 3.11 (m, 2H), 3.31 (m, 2H), 7.46 (s, 5H), 8.90 (br s, 1H),
9.31 (br s, 1H). 13C NMR (DMSO-d6, 125 MHz): δ 18.6 (C4),
J. Org. Chem. Vol. 74, No. 3, 2009 1311

Jenkins et al.
3-benzyl-7-trifluoroacetyl-3,7-diazaspiro[5.6]dodec-10-ene (27) as
a brown oil (41.04 g, 89%). H NMR (500 MHz, CDCl3): δ 1.57
The mixture was then recooled in an ice bath, and 1 M HCl was
used to adjust the pH to approximately 1. The toluene layer was
then removed, and the aqueous layer was made basic with K2CO3
(pH ∼12). The aqueous layer was extracted with CHCl3 (5 × 30
mL), which was dried (MgSO4), filtered, evaporated, and chro-
matographed on a SiO2 column, eluting with CH2Cl2/MeOH (9:1)
to afford the allyl derivative 31 as a yellow oil (22.99 g, 92%). 1H
NMR δ (500 MHz, CDCl3): 1.67 (m, 4H), 2.30 (m, 2H), 2.55 (m,
4H), 3.57 (s, 2H, benzyl), 3.68 (s, 2H, benzyl), 5.16 (m, 2H), 5.90
(m, 1H), 7.27-7.40 (m, 10H, ArH). 13C NMR (125 MHz, CDCl3):
δ 35.4 (C3/5), 42.5 (allyl), 45.6 (C4), 49.6 (C2/6), 52.6 (4-benzyl),
63.6 (1-benzyl), 118.1 (allyl), 127.1, 127.2, 128.5, 128.6, 128.62,
129.5 (benzyl), 134.4 (allyl), 139.0, 141.8 (benzyl). Compound 31
(22.99 g, 72.0 mmol) was dissolved in MeOH (30 mL) and cooled
to 0 °C. In a separate flask, AcBr (13.41 mL, 180 mmol) was added
to MeOH (30 mL). The contents were then added to the first flask.
After the mixture was stirred for 20 min, the solvent was removed
to afford an off-white foam. Dry CH2Cl2 (1038 mL) and MeOH
(112 mL) were added and the mixture was cooled to -78 °C.
Bromine (5.53 mL, 108 mmol, 1.5 eq) was added and the mixture
was stirred for 3 h (LRMS taken of solution at this point, m/z 479.1/
481.1/483.1). The mixture was then cooled to 0 °C and K2CO3
(19.90 g, 144 mmol) was added followed by H2O (50 mL). The
mixture was stirred for 16 h while warming to room temperature.
The mixture was then washed with sat. Na2S2O3 (1 × 250 mL),
sat. K2CO3 (1 × 300 mL) and H2O (1 × 300 mL), dried (MgSO4),
filtered and concentrated to dryness. The compound was purified
on a SiO2 column eluting with CH2Cl2/MeOH (95:5) to afford (()-
1,8-dibenzyl-3-bromo-1,8-diazaspiro[5.6]decane (32) as a red/
1
(m, 2H), 2.30 (m, 2H), 2.41 (bs, 2H), 2.68 (m, 4H), 3.04-3.10 (m,
2H), 3.57 (s, 2H), 3.81 (m, 2H), 5.61 (m, 1H), 5.66 (m, 1H),
7.25-7.38 (m, 5H). 13C NMR (125 MHz, CDCl3): δ 29.9 (C9),
31.7 (C12), 33.4 (C1/5), 42.1 (C8), 50.2 (C2/4), 62.4 (benzyl), 67.0
(C6), 116.9 (q, J ) 288 Hz, CF3), 124.5 (C10), 127.2 (C11), 128.4,
129.2, 129.7, 138.6 (benzyl), 157.3 (q, J ) 34 Hz, CO). LRMS
(ESI): calcd for (M + H+) C19H23F3N2O 353, found 353.
Without further purification, compound 27 (70.05 g, 0.19 mol)
was dissolved in MeOH (1 L) and H2O (200 mL) added. NaOH
(39.8 g, 0.995 mol) was added, and the mixture was refluxed for
72 h. After cooling, the solvent was removed and the solution
acidified with HCl. The mixture was then washed with CHCl3 (100
mL) and the aqueous layer separated and made basic with K2CO3.
The organic compound was extracted into CHCl3 (5 × 200 mL),
which was dried over anhydrous MgSO4, filtered, and evaporated
to give 38.16 g of a dark oil. The oil was then dissolved in MeCN
(600 mL) and stirred with (Boc)2O (65.04 g, 0.298 mol) for 16 h.
The solvent was then removed, and the residue was washed with
satd NaHCO3 (250 mL) and extracted into CHCl3 (3 × 200 mL).
The organic layers were combined, dried (MgSO4), filtered, and
concentrated to dryness to afford tert-butyl 3-benzyl-3,7-diaza-
spiro[5.6]dodecane-3-carboxylate as a brown oil (50.98 g, 75%).
Without purification, the brown oil (49.43 g, 0.139 mol) was
dissolved in MeOH (1.0 L), 10% Pd-C (49.43 g) was added, and
the mixture was stirred under a H2 atmosphere for 96 h. The mixture
was filtered through diatomaceous earth, and the solvent was
removed to afford 22.19 g of a pale yellow oil. This was suspended
in diethyl ether (100 mL). HCl in diethyl ether/dioxane was added
until a pale yellow solid dropped out. The solid was filtered and
dried to afford 28 (11.85 g, 49%) (loss of the Boc protecting group
at this stage was unexpected; clearly, too much HCl was added. It
is recommended that if the 7-Boc analogue of 29 is desired the
product obtained after hydrogenolysis be purified directly on neutral
alumina). Compound 28 (11.85 g, 49 mmol) was redissolved in
MeOH (250 mL) and cooled to 0 °C. (Boc)2O (10.77 g, 49 mmol)
was added followed by K2CO3 (6.77 g, 0.049 mol). After the
mixture was stirred for 1 h, the solvent was removed and the paste
obtained was dissolved in diethyl ether (50 mL) and washed with
H2O (50 mL). The organic layer was dried over anhydrous MgSO4,
filtered, and reduced to give 11.563 g (87%) of tert-butyl 3,7-
diazaspiro[5.6]dodecane-7-carboxylate (29) as a brown oil. A small
ample of the brown oil (1 g) was purified by chromatography on
silica gel, eluting with 5% MeOH/CHCl3, to give 29 (0.25 g, 25%;
the low yield was possibly due to partial loss of the Boc group on
the column) as a clear oil. 1H NMR (500 MHz, CDCl3): δ 1.3-1.6
(m, 21H), ∼2.0 (vbs, 1H), 2.68-2.73 (m, 2H), 3.26-3.36 (m, 2H),
3.45-3.55 (m, 2H). 13C NMR (125 MHz, CDCl3): δ 22.5 (C11),
28.7 (Boc), 30.3 (C10), 33.7 (C9), 37.6 (b, C1/5), 40.3 (b, C2/4),
41.0 (C13), 42.1 (C8), 53.3 (C6), 79.5 (C-O), 155.3 (CO). [It
1
orange oil/gum (25.81 g, 90%). H NMR δ (500 MHz, CDCl3):
1.26 (m, 1H, H6a), 1.63 (m, 1H, H10a), 1.94-2.08 (m, 4H, H6b,
H7b, H9b, H10b), 2.16 (dd, J ) 5, 14.5 Hz, 1 H, H4b), 2.35 (dd,
J ) 8.5, 14.5 Hz, 1H, H4a), 2.88 (m, 2H, H7a, 9a), 2.89 (dd, J )
6, 10.5 Hz, 1H, H2b), 3.07 (dd, J ) 6, 10.5 Hz, H2a), 3.49 (s, 2H,
CH2Ph), 3.57 (d, J ) 13.5 Hz, CHaPh), 3.66 (d, J ) 13.5 Hz,
CHbPh), 4.21 (m, J ) 5, 6, 8.5, 12 Hz, 1H, H3), 7.12-7.28 (m,
10H, ArH). 13C NMR (125 MHz, CDCl3): δ 32.4 (C6), 33.5 (C10),
43.78 (C3), 46.9 (C4), 51.6 (C2), 51.7 (C7), 51.8 (C9), 60.2 (1-
benzyl), 62.4 (8-benzyl), 63.3 (C5), 127.1, 127.6, 128.5, 128.6,
128.7, 129.6, 137.5, 140.4 (benzyl). LRMS: m/z 399/401 [M +
H]+. HRMS (ESI): calcd for (M + H+) C22H2879BrN2 399.1430,
found 399.1432.
tert-Butyl 1,8-Diazaspiro[5.6]decane-8-carboxylate (34). Com-
pound 32 (25.81 g, 65 mmol) was dissolved in isopropanol (700
mL). (Boc)2O (12.66 g, 58 mmol) was added followed by 10%
Pd-C (25.81 g). The mixture was stirred under a hydrogen
atmosphere while heating to 75 °C. After 48 h, the mixture was
cooled and filtered through diatomaceous earth. The solvent was
removed affording 18.67 g of an orange foam. This was suspended
in diethyl ether (50 mL) and acidified to ∼pH 5 with satd NH4Cl.
The diethyl ether layer was removed, and the aqueous layer was
washed with diethyl ether (2 × 50 mL) again. The aqueous layer
was then made basic with 1 M NaOH, and the organic compound
was extracted into CH2Cl2 (3 × 100 mL). The chloroform layer
was dried (MgSO4), filtered, and concentrated to dryness to afford
tert-butyl 1,8-diazaspiro[5.6]decane-8-carboxylate (34) as a brown/
1
should be noted that if the CDCl3 contains any HCl, both H and
13C chemical shifts change significantly]. HRMS (ESI): calcd for
(M + H+) C15H29N2O2 269.2223, found 269.2231.
(()-1,8-Dibenzyl-3-bromo-1,8-diazaspiro[5.6]decane(32).1-Ben-
zyl-4-piperidone (8, 15.0 g, 80.0 mmol) was dissolved in toluene
(100 mL). K2CO3 (22.1 g, 160 mmol) was added followed by
benzylamine (11.4 mL, 104 mmol). The mixture was stirred at 60
°C for 16 h. After cooling, the material was filtered through
diatomaceous earth and concentrated to dryness to afford the imine
1
yellow oil (5.85 g, 38%). H NMR (500 MHz, CDCl3): δ 1.39 (s,
9H, Boc), 1.40-1.49 (m, 4H, H6/10), 1.50 (m, 2H, H4), 1.72 (m,
2H, H3), 2.89 (m, 2H, H2), 3.28 (m, 2H, H7a/9a), 3.43 (m, 2H,
H7b/9b), 4.11 (br s, 1H, NH). 13C NMR (125 MHz, CDCl3): δ
25.5 (C3), 28.6 (Boc), 36.8 (C4), 37.6 (b, C6/10), 41.7 (b, C7/9),
45.7 (C2), 60.1 (C5), 79.4 (C-O), 155.0 (CO). HRMS (ESI): calcd
for (M + H+) C13H25N2O2 241.1911, found 241.1922.
1
30 as an orange oil (23.99 g, 97%). H NMR (500 MHz, CDCl3):
δ 2.55 (m, 4H), 2.60 (m, 2H), 2.70 (m, 2H), 3.60 (s, 2H, 4-benzyl),
4.58 (s, 2H, 1-benzyl), 7.28-7.38 (m, 10H, ArH). 13C NMR (125
MHz, CDCl3): δ 29.3 (C3), 39.3 (C5), 53.5 (4-benzyl), 54.5 (C2/
6), 62.7 (1-benzyl), 126.9, 127.3, 128.0, 128.6, 128.7, 129.2, 138.7,
140.5 (benzyl), 171.2 (CdN). The imine 30 (23.99 g, 78.0 mmol)
was dissolved in toluene (185 mL) and cooled in an ice bath.
Allylmagnesium bromide (78.0 mL, 78.0 mmol) was added, and
the mixture was stirred for 16 h while warming to room temperature.
tert-Butyl 1-Amino-7-azaspiro[3.5]nonane-7-carboxylate Hy-
drochloride (43). To a solution of tert-butyl 1-oxo-7-azaspiro[3,5]-
nonane-7-carboxylate29 (42, 8 g, 33.5 mmol) in anhydrous 1,2-
dichloroethane (230 mL) was added benzylamine (3.93 g, 36.8
mmol, 1.1 equiv) followed by sodium triacetoxyborohydride (15.5
g, 73.6 mmol, 2 equiv). The mixture was stirred at rt for 18 h and
1312 J. Org. Chem. Vol. 74, No. 3, 2009

Natural Product Inspired Scaffolds for Drug DiscoVery
then quenched with saturated aqueous sodium bicarbonate and the
product extracted with chloroform (3 × 200 mL). The combined
extracts were washed with water (2 × 200 mL) and brine (2 ×
100 mL), dried through a phase-separating paper, and concentrated
to dryness. The product was purified by silica gel chromatography
eluting with 5% MeOH/DCM to give tert-butyl 1-(benzylamino)-
7-azaspiro[3.5]nonane-7-carboxylate (10 g, 88%). 1H NMR (CDCl3,
500 MHz): δ 1.30-1.40 (m, 12H), 1.55-1.80 (m, 4H), 2.08 (m,
1H), 2.62 (m, 1H), 2.80 (m, 1H), 2.97 (m, 1H), 3.70 (s, 2H), 3.80
(m, 1H), 3.95 (m, 1H), 7.20-7.30 (m, 5H). 13C NMR (CDCl3, 125
MHz): δ 25.2 (C2), 25.4 (C3), 28.7 (Boc), 29.2 (C5), 38.8 (C9),
41.0 (b, C6/8), 43.6 (C4), 52.1 (benzyl), 60.7 (C1), 79.5 (C-O),
127.3, 128.4, 128.6, 140.0 (benzyl), 155.1 (CO). LRMS (ESI): calcd
for (M + H+) C20H31N2O2 331, found 331.
To a solution of tert-butyl 1-(benzylamino)-7-azaspiro[3.5]nonane-
7-carboxylate (8.5 g, 25.7 mmol) in methanol (200 mL) was added
20% Pd(OH)2/C (5 g) in one lot. The reaction mixture was stirred
at rt under a hydrogen atmosphere (balloon) for 20 h, and then the
catalyst was filtered off through a Celite-packed funnel. The celite
layer was washed with fresh methanol (3 × 100 mL). The combined
filtrate and washings were concentrated. The residue was diluted
with water (25 mL) and acidified with a solution of citric acid (20%,
pH 5). The acidified aqueous solution was extracted with chloroform
(3 × 100 mL) to remove impurities. The aqueous fraction was then
basified with a solution of sodium hydroxide (2 M, pH 8-9). The
free amine was extracted with chloroform (3 × 100 mL),
concentrated, and dried under reduced pressure to give tert-butyl
1-amino-7-azaspiro[3.5]nonane-7-carboxylate (5.88 g, 95%) as a
colorless oil.1H NMR (CD3OD, 500 MHz): δ 1.45 (s, 9H),
1.40-1.65 (m, 5H), 1.75 (m, 2H), 2.20 (m, 1H), 2.85 (m, 1H),
3.00 (m, 1H), 3.04 (m, 1H), 3.83 (m, 1H), 3.94 (m, 1H), 4.80 (m,
2H). 13C NMR (CD3OD, 125 MHz): δ 26.4 (C2), 27.4 (C3), 28.7
(Boc), 29.9 (C5), 38.5 (C9), 41.7 (b, C6/8), 44.4 (C4), 56.2 (C1),
80.8 (C-O), 156.6 (CO). LRMS (ESI): calculated for (M+H+)
C13H25N2O2 241, found 241. The oil was converted into the
hydrochloride salt: to an ice-cold solution of tert-butyl 1-amino-
7-azaspiro[3.5]nonane-7-carboxylate (5.88 g, 24.5 mmol) in anhy-
drous diethyl ether (50 mL) was added dropwise hydrogen chloride
in ether (12.25 mL, 2 M, 1 equiv). A white precipitate formed
quickly. After addition of the hydrogen chloride, a further amount
of dry ether (50 mL) was added. The precipitate was filtered and
washed with dry ether (3 × 100 mL), then dried under vacumm
overnight to give the crude product (5.5 g, 81%) as a white solid.
The white solid (5.5 g) was dissolved in the minimum amount of
methanol (3 mL) to give a clear solution. To this stirred solution
was added anhydrous ether until a white precipitate formed. After
5 min, more ether was added (100 mL), followed by one more
addition (50 mL). The white solid was filtered and washed with
dry ether. The solid was dried under vacuum to give tert-butyl
1-amino-7-azaspiro[3.5]nonane-7-carboxylate hydrochloride (43,
1
4.1 g, 60%). H NMR (CD3OD, 500 MHz): δ 1.45 (s, 9H), 1.50
(m, 1H), 1.63 (m, 2H), 1.78 (m, 2H), 2.04 (m, 2H), 2.35 (m, 1H),
2.78 (m, 1H), 2.92 (m, 1H), 3.30 (MeOH - used as reference), 3.45
(m, 1H), 3.90 (m, 1H), 4.01 (m, 1H). 13C NMR (D2O, 125 MHz):
d 22.2 (C2), 26.0 (C3), 28.6 (Boc - used as reference), 29.6 (C5),
36.9 (C9), 40.6 (b, C6), 41.5 (b, C8), 41.8 (C4), 53.2 (C1), 82.7
(C-O), 157.5 (CO). HRMS (ESI): calcd for (M + H+) C13H25N2O2
241.1911, found 241.1919.
Acknowledgment. We thank AstraZeneca and the Eskitis
Institute, Griffith University, for financial support.
Supporting Information Available: General experimental
1
procedures, experimental data for 42, and H and 13C NMR
spectra for the scaffolds 15, 17, 19, 21-23, 29, 32, 34, and 43.
This material is available free of charge via the Internet at
http://pubs.acs.org.
JO802456W
J. Org. Chem. Vol. 74, No. 3, 2009 1313