Two-directional synthesis as a tool for diversityoriented synthesis: Synthesis of alkaloid scaffolds

Article abstract of DOI:10.3762/bjoc.8.95

Two-directional synthesis represents an ideal strategy for the rapid elaboration of simple starting materials and their subsequent transformation into complex molecular architectures. As such, it is becoming recognised as an enabling technology for divers

Full text of DOI:10.3762/bjoc.8.95

Two-directional synthesis as a tool for diversity-
oriented synthesis: Synthesis of alkaloid scaffolds
Kieron M. G. O’Connell, Monica Díaz-Gavilán§, Warren R. J. D. Galloway
and David R. Spring*
Full Research Paper
Open Access
Address:
Beilstein J. Org. Chem. 2012, 8, 850–860.
Department of Chemistry, University of Cambridge, Lensfield Rd,
Cambridge, CB2 1EW, UK
doi:10.3762/bjoc.8.95
Received: 01 March 2012
Accepted: 14 May 2012
Published: 06 June 2012
Email:
David R. Spring* - spring@ch.cam.ac.uk
* Corresponding author
This article is part of the Thematic Series "Recent developments in
chemical diversity".
§ Present address: Departemento de Química Farmacéutica y
Orgánica, Facultad de Farmacia, Campus de Cartuja, s.n. 18071
Granada, Spain
Guest Editor: J. A. Porco Jr.
Keywords:
© 2012 O’Connell et al; licensee Beilstein-Institut.
License and terms: see end of document.
alkaloids; cascade reactions; chemical diversity; diversity-oriented
synthesis; Lewis acid catalysis; two-directional synthesis
Abstract
Two-directional synthesis represents an ideal strategy for the rapid elaboration of simple starting materials and their subsequent
transformation into complex molecular architectures. As such, it is becoming recognised as an enabling technology for diversity-
oriented synthesis. Herein, we provide a thorough account of our work combining two-directional synthesis with diversity-oriented
synthesis, with particular reference to the synthesis of polycyclic alkaloid scaffolds.
Introduction
Diversity-oriented synthesis (DOS) aims to prepare structurally linear substrates into cyclic molecular architectures [6-8]. The
diverse compound collections in an efficient manner [1-3]. Of design and synthesis of these linear substrates can, in itself,
the possible “types of diversity” that can be incorporated into a represent a significant challenge as it is desirable that these
compound collection, the most important, in terms of creating a compounds are easily accessible in a small number of synthetic
functionally (biologically) diverse collection, is generally steps. Two-directional synthesis [9-12] offers a powerful
considered to be scaffold (or skeletal) diversity, i.e., the vari- method for the synthesis of such substrates, because each syn-
ation of molecular frameworks between compounds [4,5]. thetic transformation has the potential to provide twice as much
Therefore, one of the key challenges in DOS is the develop- molecular complexity compared to standard approaches.
ment of strategies that allow the efficient generation of a range
of complex molecular scaffolds. A large number of approaches We have recently reported a strategy for DOS that combined
towards this goal have been reported, with some of the most two-directional synthesis with the use of these folding reaction
effective being based around the “folding-up” of functionalised pathways [13]. In this work, two-directional synthesis was used
850

Beilstein J. Org. Chem. 2012, 8, 850–860.
to rapidly generate a series of linear aminoalkenes, which were Lewis acid catalysed cyclisation reactions by intramolecular
then folded into bicyclic and tricyclic scaffolds through Lewis conjugate addition. Three of these compounds (1–3) were
acid mediated cascade processes. The compounds produced in obtained from the corresponding alcohols [19,20] in three steps:
this campaign were reminiscent of naturally occurring alkaloids, Mitsunobu reaction with NH-Boc-tosylate, followed by tosyl
such as the Coccinellidae natural products, which are secreted deprotection with magnesium, and finally two-directional cross
by ladybirds to deter predators [14]. The total synthesis of one metathesis with ethyl acrylate to install the desired α,β-unsatur-
of their number, myrrhine, was also achieved by the elabora- ated ester functionality. This sequence provided the desired
tion of one of the compounds produced. In this article, the work N-Boc-aminoalkenes in respectable overall yields of 38–56%.
is presented in more detail, alongside additional results from our Compound 4 was prepared in a four-step sequence from the
work combining two-directional synthesis with DOS. Treated requisite phenyldialkyl alcohol. Ritter reaction with chloro-
together, we believe these works provide useful insights into the acetonitrile followed by cleavage of the resulting chloroacet-
potential utility of two-directional synthesis as an enabling tech- amide with thiourea gave the free amine [21], which was then
nology for DOS.
protected with Boc anhydride. Finally, cross metathesis with
ethyl acrylate furnished the desired compound 4 in 24% overall
The initial DOS campaign was largely inspired by the pion- yield.
eering work of Robert Stockman on combining two-directional
synthesis with tandem reactions to create complex molecular Cyclisation reactions
architectures [15-18]. The key folding step in the DOS was the The first attempts at the tandem Boc-deprotection/bicyclisation
Lewis acid mediated pairing reaction of a nucleophilic amino of these substrates were performed by using AlCl3 as the Lewis
group with suitable electrophilic functionality, provided by acid (Scheme 2); compounds 1–4 were treated with 1.1 equiv of
Michael acceptor α,β-unsaturated ester groups. Two-directional AlCl3 in dichloromethane at room temperature. These condi-
synthesis was used to append these electrophilic groups at two tions proved effective at promoting bicyclisation for com-
positions around the linear substrates, allowing bicyclisation pounds 1, 2 and 4, for which the desired bicyclic products were
processes to be instigated. The scaffold diversity between the obtained in 67–85% yield, as a mixture of diastereomers. The
products then resulted from the different ring sizes that it was cyclisation of 1 gave pyrrolizidine 5 as a mixture of 4,10-trans-
possible to form from these linear substrates. (Figure 1 shows 7,10-trans (trans-5) (42%) and 4,10-cis-7,10-trans (cis-5)
an overview of the DOS strategy.)
(28%) isomers, which proved to be separable by flash chroma-
tography. The stereochemistry of cis-5 was confirmed by
analogy to known 1H and 13C NMR values [22] and by NOESY
spectroscopy, which showed enhancements between H-7, H-4
Results and Discussion
Synthesis of linear precursors
Our initial studies explored the synthesis and reactivity of four and H-10. In this case, it proved possible to achieve an impro-
N-Boc-aminoalkenes containing α,β-unsaturated ester groups ved yield of both diastereomers by treating 1 with an excess
(1–4, Scheme 1), as substrates for intramolecular pairing reac- (163 equiv) of trifluroacetic acid, which furnished trans-5 and
tions. The N-Boc protecting group was chosen to give the cis-5 in 53% and 34% yield, respectively. The corresponding
potential for deprotection to be carried out in tandem with the reaction of phenyl-substituted analogue 4 also gave a mixture of
Figure 1: Overview of the DOS strategy.
851

Beilstein J. Org. Chem. 2012, 8, 850–860.
Scheme 1: Synthesis of linear cyclisation precursors 1–4.
two diastereomers; the major diastereomer was 4,10-trans-7,10- In light of the difficulties encountered in producing the desired
cis isomer (trans-6), which was formed in 50% yield, and a bicyclic species, an optimisation study of the cyclisation of 3
17% yield of the 4,7,10-cis-isomer (cis-6) was also obtained. was undertaken, which resulted in a number of interesting find-
ings that are summarised in Table 1. The initial alterations made
Indolizidine 7 was produced in very good overall yield (85%). little difference to the process; increasing the reaction time up to
The unsymmetrical nature of this compound gave rise to the seven days had essentially no effect on the product ratio, with
possibility of formation of additional diastereomers when trans-8 and 9 obtained in 21% and 37% yield, respectively, and
compared to 5 and 6; however, once again only two were increasing the temperature to the point of reflux in dichloro-
formed in any appreciable amount. The compounds obtained methane also had little effect on the conversion. However, when
were both found to have a trans-fused geometry at the ring the reaction solvent was changed to toluene and the reaction
junction, as indicated by IR (strong absorbance at 2850 cm−1) mixture heated to reflux, trans-8 and 9 were still formed in
and 1H NMR (H-7 chemical shift around 2.4 ppm) spectro- similar proportions (24% and 21%), but in this case an add-
scopy [23,24]. The ester-bearing side chains were also found to itional product was also isolated in 30% yield. This compound
be trans to each other in both cases, meaning that the two was found to be tricyclic compound 10 (Scheme 3a), which was
products obtained differ from each other only in which ring has obtained as a single diastereomer with all of the ring junction
the side chain cis to the ring junction hydrogen. The favoured protons on the same face. This all–cis-stereochemistry was
product was the 4,11-trans-7,11-cis isomer (trans-7) in which surprising, as so far 8 had only been obtained with the side
the side chain of the 6-membered ring is cis to the ring junction chains in trans- configuration; however the configuration of 10
proton; this compound was isolated in 55% yield. The alter- was unambiguously confirmed by X-ray crystallography. In
native 4,11-trans-4,7-cis isomer (trans’-7) was obtained in 30% some ways the formation of a tricyclic species such as 10 was
yield. Despite the good yields obtained for these three not altogether surprising, as a similar tricyclic species was
examples, the cyclisation of 3 under these conditions proved generated in Stockman’s synthesis of hippodamine [25]. In that
disappointing, with 4,12-trans-8,12-cis-quinolizidine (trans-8) work, bicycle trans-8 was transformed into the corresponding
only obtained in 18% yield, along with 40% of monocyclic tricyclic compound (possessing cis–trans ring junction stereo-
species 9.
chemistry) by a base-mediated Dieckmann cyclisation
852

Beilstein J. Org. Chem. 2012, 8, 850–860.
Scheme 2: AlCl3 catalysed tandem Boc-removal/bicyclisation processes; the yields quoted refer to the isolated yields of single compounds.
Table 1: Overview of the Lewis acid mediated folding reactions of 3.
solvent
Lewis acid
(equiv)
temp.
% yield
trans-8
cis-8
9
10
DCMa
AlCl3 (1.1)
rt
18
21
24
43
50
—
30
29
—
—
—
—
—
9
40
37
21
13
21
—
—
10
—
—
30
—
—
72
23
—
DCMb
AlCl3 (1.1)
rt
toluenea
toluenea
toluenea
toluenea
toluenea
toluene
AlCl3 (1.1)
reflux
reflux
reflux
reflux
reflux
reflux
AlCl3 (3)
Sc(OTf)3 (3)
Sn(OTf)2 (0.5)
Sn(OTf)2 (1.1)
Sn(OTf)2 (3)
13
—
aReaction stirred overnight; breaction stirred for seven days.
853

Beilstein J. Org. Chem. 2012, 8, 850–860.
Scheme 3: (a) AlCl3 catalysed formation of tricyclic alkaloid 10 along with an X-ray crystal structure of 10; (b) base-mediated Dieckmann cyclisation
of trans-8 employed by Stockman and co-workers in the synthesis of hippodamine [25].
(Scheme 3b). It seems in our case that, under the correct condi- substrate simply by varying the amount and identity of the
tions, a Dieckmann reaction can be made to occur in tandem Lewis acid used.
with the Boc-deprotection and double-conjugate addition
processes. This one-pot, four-step reaction process is extremely Interestingly, the catalytic variant of the reaction also produced
interesting for the amount of molecular complexity generated in 9% of the bicyclic cis-8, which had not been isolated from any
a single transformation, and also because the all-cis-stereo- of the previous reactions. For completeness, one further reac-
chemistry of 10, which differs from the cis–trans-stereochem- tion was performed by using 1.1 equiv of Sn(OTf)2; careful
istry observed by Stockman for the Dieckman cyclisation of purification of this reaction gave 23% of 10, 30% of trans-8 and
trans-8. For these reasons, further investigations into the 13% of cis-8. The presence of the previously undetected cis-8 in
process were carried out.
these two final reactions was intriguing and led us to speculate
as to whether a different mechanistic pathway could be in oper-
The amount of AlCl3 was increased to 3 equiv; however, this ation depending on the amount of Lewis acid used.
led to the suppression of tricycle formation in favour of a slight
increase in the yield of trans-8 to 43% (a 13% yield of 9 was A number of factors led to this mechanistic speculation; prin-
also obtained). Switching the Lewis acid to Sc(OTf)3 and still cipal among them was the fact that cis-8 was never detected in
using 3 equiv, gave a slight improvement in the yields of the reactions in which 10 was not formed. In all of the earlier
trans-8 to 50% and 9 to 21% but no formation of 10. Switching experiments the only bicycle detected was trans-8, implying
the Lewis acid again to Sn(OTf)2 resulted in a decrease in the that the double-conjugate addition process heavily favours the
yields of trans-8 to 29% and 9 to 10% and again no formation formation of this compound. Therefore it was considered that
of 10. The amount of Lewis acid was then reduced to 0.5 equiv, cis-8 could be forming from 10; suggesting, somewhat counter
which dramatically altered the course of the reaction. intuitively, that the Dieckmann cyclisation to give 6,10-bridged
Performing the reaction under reflux in toluene with 0.5 equiv bicycle 11 could in some cases be favoured over the expected
Sn(OTf)2 produced tricycle 10 in 72% isolated yield. Thus both double-conjugate addition. Transannular conjugate addition
trans-8 and 10 could be accessed in good yields from the same across the 10-membered ring of 11 would give 10, and a retro-
854

Beilstein J. Org. Chem. 2012, 8, 850–860.
Scheme 4: (a) Optimal conditions to obtain trans-8 and 10 and the control experiments carried out to probe the mechanism of the process; (b)
proposed mechanistic pathway for the tricyclisation of 3.
Dieckmann process could then form cis-8. Several control the reactions. The direct formation of 10 from trans-8 was not
experiments were run in an attempt to validate this hypothesis thought to be possible due to their differing stereochemistry;
(Scheme 4a).
however, an alternative diastereomer of 10 may be expected to
form (as observed by Stockman for the corresponding base-
Trans-8 was treated with both stoichiometric and catalytic mediated process) [25]. Monocycle 9 was treated with catalytic
amounts of Sn(OTf)2 and there was no evidence of tricycle for- Sn(OTf)2, which led to a mixture of trans-8 and 10. As the for-
mation in either case, with only starting material recovered from mation of 10 from trans-8 had been proven not to occur, it
855

Beilstein J. Org. Chem. 2012, 8, 850–860.
seemed reasonable to assume from this that it is possible to mation of trans-8 is not observed suggesting that under these
form both species from 9. It is also noteworthy that the best conditions the simple double-conjugate addition cannot occur.
yields of 10 were obtained when the reaction was performed The identity of the Lewis acid used in the reaction seems not to
while fitted with a Dean–Stark apparatus containing pieces of overtly affect the course of the reaction in terms of the products
sodium to trap the ethanol formed in the Dieckmann cyclisation, obtained, as similar product distributions were observed for the
and thus inhibit the retro-Dieckmann reaction. Finally, 10 was reactions using three equiv of AlCl3, Sc(OTf)3 and Sn(OTf)2.
treated with catalytic Sn(OTf)2, which, as expected, gave cis-8, The formation of 10 was also not limited to the use of
suggesting that the retro-Dieckmann reaction does occur.
Sn(OTf)2, as a 30% yield of 10 was obtained for the reaction
using 1.1 equiv of AlCl3.
Taking these results into account, we tentatively suggest that the
mechanism for the tricyclisation of 3 does indeed proceed via a Between them, the catalytic and stoichiometric variants of this
Dieckmann cyclisation to form the 10-membered ring followed reaction provide a useful illustration of the use of reagent-based
by transannular conjugate addition (Scheme 4b). The fact that diversification within a predominantly substrate-based strategy.
this appears to occur when lower amounts of Lewis acid are Attempts were then made to apply this reagent-based diversific-
used (0.5 or 1.1 equiv) but not when three equiv are used indi- ation to the other linear substrates (Scheme 5). Compound 2
cates that the mechanistic path of the cascade is dependent on was treated with 0.5 equiv of Sn(OTf)2 in acetonitrile, which
the amount of the Lewis acid present. When three equiv of surprisingly did not lead to the formation of the 6-6-5-tricyclic
Lewis acid are present (the “stoichiometric” process), it is feas- species, instead producing trans’-7 in 69% yield along with
ible that during the course of the reaction both of the carbonyl 10% trans-7. While this reaction did not produce any tricyclic
groups are activated simultaneously, allowing the double- species, it was an interesting result, as the selectivity for trans’-
conjugate addition to proceed smoothly. However, when fewer 7 over trans-7 was the opposite of that observed for the original
equiv are used (the “catalytic” process) there is a relative defi- AlCl3 catalysed process. Performing the reaction in toluene
ciency of Lewis acid present, and thus, it is less probable that with 1.1 equiv of Sn(OTf)2 did produce the expected tricyclic
both carbonyl groups can be coordinated to separate Lewis acid species in 50% yield as a mixture of two diastereomers (cis-13
molecules simultaneously. If the Dieckmann reaction occurs via and trans-13) in a 4:1 ratio, proving that the folding of this
a standard 6-membered transition state, it requires both linear substrate into tricyclic species in one-pot was also
carbonyls to be either coordinated to, or bonded to, a single possible. Unfortunately, this did not prove to be the case for 1;
metal centre, and so this could go some way towards explaining all attempts to transform 1 into a tricyclic species in one pot,
the apparent course of the reaction. Assuming that the Lewis using either stoichiometric or catalytic Lewis acid were unsuc-
acid remains bonded to the carbonyl of the enol form of the cessful. However, it did prove possible to perform the Dieck-
ester group after the first conjugate addition, there may be mann reaction on the bicyclic compounds cis-5 and trans-5 to
insufficient Lewis acid free in solution to activate the second access the tricyclic architectures. The bicyclic species were
ester group separately, leaving it effectively inert to conjugate treated with LDA at −18 °C, which affected the desired cyclisa-
addition (and Dieckmann cyclisation). However, if the second tion in both cases. Strangely, the process proved far more effi-
carbonyl group becomes coordinated to the metal centre that is cient for trans-5, giving trans-14 in 91% yield, compared to
bonded to the first, the Dieckmann reaction becomes the most 38% for the cis-isomer. In both cases the 1H NMR suggested
favourable process and so occurs preferentially to the second that these compounds exist as the usually disfavoured enol
conjugate addition. For these reasons, we cautiously postulate tautomer.
that the Dieckmann reaction may occur via a chelated transition
state such as 12, in which both carbonyls are coordinated to a These reactions, along with the reactions of 1, clearly illustrate
single metal centre.
the power of this two-directional approach to DOS. Using this
methodology it was possible to access five bicyclic and tricyclic
The data in Table 1 provides further support for the suggestion scaffolds covering a range of 3D shapes, including pyrroli-
that different mechanisms can operate depending on the amount zidines, indolizidines and quinolizidines, along with 6-6-6 and
of Lewis acid used. This support is provided by the fact that the 5-6-6 azatricyclic species in a single transformation from a
formation of 10 proceeds best under truly catalytic conditions: small collection of structurally simple linear starting materials.
when 0.5 equiv Sn(OTf)2 were used, a 72% yield of 10 was One further tricyclic scaffold (5-5-6) was also accessible in one
obtained compared to 23% when 1.1 equiv were used. In fact, it further synthetic step. The effective introduction of reagent-
appears that when 1.1 equiv of Lewis acid are used both mech- based diversification into the strategy was extremely satisfying,
anisms can occur, as indicated by the formation of both trans-8 as by altering the choice of Lewis acid and reaction tempera-
and 10 in these reactions, but when 0.5 equiv are used the for- ture we were able to adjust the course and selectivity of the
856

Beilstein J. Org. Chem. 2012, 8, 850–860.
Scheme 5: DOS of 5-5-6 and 5-6-6 tricyclic alkaloids 13 and 14.
reactions to generate different scaffolds and stereochemistry. tion of the double bond with hydrogen gas and Raney-nickel.
This combination of reagent and substrate-based approaches for This was achieved in a moderate 57% yield and with good
the generation of molecular diversity can afford many interest- (~10:1) diastereoselectivity by complexing the nitrogen lone
ing possibilities not achievable by either approach alone.
pair with tosic acid, effectively blocking the undesirable face of
the tricycle during the course of the reduction. This reduction
also produced 5% of the unnatural (or as yet undiscovered)
Total synthesis of myrrhine
Inspired by Stockman’s syntheses of the related alkaloids isomer epi-myrrhine. The N-oxide of myrrhine, which has also
hippodamine and epi-hippodamine [26], tricyclic species 10 not been isolated from natural sources, was then synthesised in
was identified as a potential intermediate for the total synthesis 96% yield by treating myrrhine with mCPBA. This synthesis of
of myrrhine, which was then achieved in three steps from 10 myrrhine compares favourably with previously reported
(Scheme 6). Compound 10 was treated with Na2CO3 in a mix- syntheses [26-28], achieving the feat in eight steps and 7%
ture of EtOH and H2O under reflux to achieve ester saponifica- overall yield.
tion, which was followed by decarboxylation, proceeding
smoothly to give the corresponding ketone in 76% yield. The Alternative starting materials
ketone was then transformed to the exocyclic alkene 15 in 61% The evident efficiency of two-directional synthesis in a DOS
yield by Wittig reaction with the appropriate phosphonium salt. context, as exemplified by our synthesis of these alkaloid scaf-
The final step in the synthesis was a diastereoselective reduc- folds, has led us to continue investigations in this area and to
857

Beilstein J. Org. Chem. 2012, 8, 850–860.
phenylvinylketone. Subjecting diketone 18 to H2 gas and
Raney-nickel then reduced the nitro group and effected the
desired double reductive amination to give 17 in 30% yield. It is
likely that the scope of this sequence could be extended to
include indolizidine and quinolizidine scaffolds, and so provide
an alternative route to these frameworks, instead of the double
Michael addition strategy.
Scheme 6: Total synthesis of myrrhine, epi-myrrhine and myrrhine-N-
oxide.
Scheme 7: Use of nitromethane in DOS: Synthesis of meso-diphenyl-
pyrrolidizine 17.
explore the potential utility of this approach for a range of
different substrates. Among these substrates, two that stand out
as particularly promising are nitromethane and tris(hydroxy-
methyl)aminomethane (Tris) 16.
The three hydroxy and single amino groups of Tris give the
potential for many variations in substitution; however, our
Nitromethane is of great interest to us as a potential DOS sub- studies have so far focused on allylated derivatives, in particu-
strate, as we have a long-standing interest in developing diver- lar the triallyl derivative 19 (Scheme 8). The synthesis of 19
gent reaction pathways from small and simple starting ma- was achieved in two steps from Tris: N-Boc protection
terials [29,30]. As it represents a one-carbon unit, nitromethane proceeded in 70% yield, and was followed by alkylation with an
is an ideal substrate for investigation. For Tris, the central excess of allyl bromide to provide the desired triallyl species in
quarternary carbon centre is the key point of interest, as we 66% yield. Cross metathesis of 19 with ethyl acrylate was then
believe that, with the judicious selection of appendages, it performed. Fortunately, it proved to be possible to achieve
should allow us to access a range of bicyclic structures, some selectivity for different products by varying the catalyst
including examples of fused, bridged and spiro bicycles. used. Treating 19 with 3% Grubbs II catalyst in neat ethyl
Preliminary studies into the utility of these substrates in a DOS acrylate at room temperature gave monoester 20 in 41% yield,
context have yielded some promising results.
whereas performing the reaction with 5% Hoveyda-Grubbs II
catalyst gave a 73% yield of triester 21.
Our work with nitromethane has so far led to the synthesis of
meso-diphenylpyrrolizidine 17, which was achieved in three These two metathesis products were then subjected to the
steps (Scheme 7). Nitromethane was treated with NaOH, and tandem deprotection-cyclisation conditions, this time by using
the resulting anion was used to displace the chloride from two equiv of AlCl3 in dichloromethane at room temperature.
3-chloro-1-phenylpropan-1-one giving a 90% yield of the For monoester 20, only a single conjugate addition was
nitroketone. Two-directional synthesis of diketone 18 in this possible, leaving the pendant terminal alkene groups as handles
fashion did not prove to be feasible; however, it was achieved in for further reactivity. This cyclisation proceeded to give the
good yield by Michael addition of the nitroketone anion to expected, functionalised morpholine scaffold 22 in a moderate
858

Beilstein J. Org. Chem. 2012, 8, 850–860.
Scheme 8: Use of Tris as a substrate for DOS: Synthesis of decorated morpholines 22 and 23.
33% yield. A similar product was also obtained for 21, with a involving the use of Tris as a potential DOS substrate remain
single conjugate addition process occurring to give 23 in 73% on-going within our laboratories.
yield and no trace of the bicyclic product detected. A number of
attempts were then made to force the reaction to occur, by Conclusion
varying the Lewis acid to Sn(OTf)2 and Sc(OTf)3 and by The work presented in this article serves to illustrate the poten-
heating under reflux in toluene, all of which produced varying tial power of two-directional synthesis in DOS. The use of two-
amounts of 23, with no trace of the bicyclic product detected. directional synthesis allowed us to access a range of bicyclic
This result was disappointing and also, to some degree, and tricyclic molecular scaffolds, rapidly and efficiently, by
surprising given the relative ease of formation of the corres- following a common reaction scheme. The nature of the two-
ponding quinolizidine compounds. We speculate that the directional synthesis lends itself to the formation of bicyclic
relative difficulty in the reaction is due to the reduced availabil- compounds by the folding up of doubly substituted precursors,
ity of the lone pair of the morpholine nitrogen compared to the and it proved to be a very effective strategy for the synthesis of
piperidine nitrogen (pKa 8.36 versus 11.22) [31], retarding the natural-product-like alkaloid scaffolds. Our work so far in this
rate of nucleophilic attack. Another possibility is that the area has focused mainly on the synthesis of fused bicyclic com-
oxygens in the pendant chains can form a stable hydrogen- pounds; however, we hope in the future to be able to apply a
bonded species with the NH group that inhibits the reaction of two-directional synthesis approach to the DOS of a wide range
the nitrogen with the Michael acceptor ester groups. Studies of molecular scaffolds and structural classes.
859

Beilstein J. Org. Chem. 2012, 8, 850–860.
16.Stockman, R. A.; Sinclair, A.; Arini, L. G.; Szeto, P.; Hughes, D. L.
J. Org. Chem. 2004, 69, 1598–1602. doi:10.1021/jo035639a
17.Rejzek, M.; Stockman, R. A.; van Maarseveen, J. H.; Hughes, D. L.
Chem. Commun. 2005, 4661–4662. doi:10.1039/b508969g
18.Robbins, D.; Newton, A. F.; Gignoux, C.; Legeay, J.-C.; Sinclair, A.;
Rejzek, M.; Laxon, C. A.; Yalamanchili, S. K.; Lewis, W.;
O’Connell, M. A.; Stockman, R. A. Chem. Sci. 2011, 2, 2232–2235.
doi:10.1039/c1sc00371b
Supporting Information
Supporting Information File 1
Experimental procedures and spectral data for all
previously unreported compounds.
[http://www.beilstein-journals.org/bjoc/content/
supplementary/1860-5397-8-95-S1.pdf]
19.Boyer, F.-D.; Hanna, I. Eur. J. Org. Chem. 2006, 471–482.
doi:10.1002/ejoc.200500645
Supporting Information File 2
NMR Spectra of novel compounds.
20.Schwartz, B. D.; McErlean, C. S. P.; Fletcher, M. T.;
Mazomenos, B. E.; Konstantopoulou, M. A.; Kitching, W.;
De Voss, J. J. Org. Lett. 2005, 7, 1173–1176. doi:10.1021/ol050143w
21.Jirgensons, A.; Kauss, V.; Kalvinsh, I.; Gold, M. R. Synthesis 2000,
1709–1712. doi:10.1055/s-2000-8208
[http://www.beilstein-journals.org/bjoc/content/
supplementary/1860-5397-8-95-S2.pdf]
22.Scarpi, D.; Occhiato, E. G.; Guarna, A. J. Org. Chem. 1999, 64,
1727–1732. doi:10.1021/jo981946i
Acknowledgements
The authors thank UCB Celltech, the ERC, EU, EPSRC,
BBSRC, MRC, Wellcome Trust and Frances and Augustus
Newman Foundation for funding.
23.Crabb, T. A.; Newton, R. F.; Jackson, D. Chem. Rev. 1971, 71,
109–126. doi:10.1021/cr60269a005
24.Rader, C. P.; Young, R. L., Jr.; Aaron, H. S. J. Org. Chem. 1965, 30,
1536–1539. doi:10.1021/jo01016a048
25.Rejzek, M.; Stockman, R. A.; Hughes, D. L. Org. Biomol. Chem. 2005,
3, 73–83. doi:10.1039/b413052a
References
1. Galloway, W. R. J. D.; Isidro-Llobet, A.; Spring, D. R. Nat. Commun.
26.Ayer, W. A.; Dawe, R.; Eisner, R. A.; Furuichi, K. Can. J. Chem. 1976,
54, 473–481. doi:10.1139/v76-064
2010, 1, No. 80. doi:10.1038/ncomms1081
27.Mueller, R. H.; Thompson, M. E.; DiPardo, R. M. J. Org. Chem. 1984,
49, 2217–2231. doi:10.1021/jo00186a029
2. Nielsen, T. E.; Schreiber, S. L. Angew. Chem., Int. Ed. 2008, 47,
48–56. doi:10.1002/anie.200703073
28.Gerasyuto, A. I.; Hsung, R. P. J. Org. Chem. 2007, 72, 2476–2484.
doi:10.1021/jo062533h
3. Spandl, R. J.; Díaz-Gavilán, M.; O’Connell, K. M. G.; Thomas, G. L.;
Spring, D. R. Chem. Rec. 2008, 8, 129–142. doi:10.1002/tcr.20144
4. Galloway, W. R. J. D.; Díaz-Gavilán, M.; Isidro-Llobet, A.; Spring, D. R.
Angew. Chem., Int. Ed. 2009, 48, 1194–1196.
29.Wyatt, E. E.; Fergus, S.; Galloway, W. R. J. D.; Bender, A.; Fox, D. J.;
Plowright, A. T.; Jessiman, A. S.; Spring, D. R. Chem. Commun. 2006,
3296–3298. doi:10.1039/b607710b
doi:10.1002/anie.200805452
30.Thomas, G. L.; Spandl, R. J.; Glansdorp, F. G.; Welch, M.; Bender, A.;
Cockfield, J.; Lindsay, J. A.; Bryant, C.; Brown, D. F. J.; Loiseleur, O.;
Rudyk, H.; Ladlow, M.; Spring, D. R. Angew. Chem., Int. Ed. 2008, 47,
2808–2812. doi:10.1002/anie.200705415
5. Dow, M.; Fisher, M.; James, T.; Marchetti, F.; Nelson, A.
Org. Biomol. Chem. 2012, 10, 17–28. doi:10.1039/c1ob06098h
6. Burke, M. D.; Berger, E. M.; Schreiber, S. L. Science 2003, 302,
613–618. doi:10.1126/science.1089946
31.Hall, H. K., Jr. J. Am. Chem. Soc. 1957, 79, 5441–5444.
doi:10.1021/ja01577a030
7. Morton, D.; Leach, S.; Cordier, C.; Warriner, S.; Nelson, A.
Angew. Chem., Int. Ed. 2009, 48, 104–109.
doi:10.1002/anie.200804486
8. O’Leary-Steele, C.; Pedersen, P. J.; James, T.; Lanyon-Hogg, T.;
Leach, S.; Hayes, J.; Nelson, A. Chem.–Eur. J. 2010, 16, 9563–9571.
doi:10.1002/chem.201000707
License and Terms
9. Poss, C. S.; Schreiber, S. L. Acc. Chem. Res. 1994, 27, 9–17.
doi:10.1021/ar00037a002
This is an Open Access article under the terms of the
Creative Commons Attribution License
10.Vrettou, M.; Gray, A. A.; Brewer, A. R. E.; Barrett, A. G. M. Tetrahedron
2007, 63, 1487–1536. doi:10.1016/j.tet.2006.09.109
11.Kennedy, A.; Nelson, A.; Perry, A. Beilstein J. Org. Chem. 2005, 1,
No. 2. doi:10.1186/1860-5397-1-2
(http://creativecommons.org/licenses/by/2.0), which
permits unrestricted use, distribution, and reproduction in
any medium, provided the original work is properly cited.
12.Gignoux, C.; Newton, A. F.; Barthelme, A.; Lewis, W.; Alcaraz, M.-L.;
Stockman, R. A. Org. Biomol. Chem. 2012, 10, 67–69.
doi:10.1039/c1ob06380d
The license is subject to the Beilstein Journal of Organic
Chemistry terms and conditions:
13.Díaz-Gavilán, M.; Galloway, W. R. J. D.; O'Connell, K. M. G.;
Hodkingson, J. T.; Spring, D. R. Chem. Commun. 2010, 46, 776–778.
doi:10.1039/b917965h
(http://www.beilstein-journals.org/bjoc)
14.Happ, G. M.; Eisner, T. Science 1961, 134, 329–331.
doi:10.1126/science.134.3475.329
The definitive version of this article is the electronic one
which can be found at:
15.Stockman, R. A. Tetrahedron Lett. 2000, 41, 9163–9165.
doi:10.1016/S0040-4039(00)01640-3
doi:10.3762/bjoc.8.95
860