Synthesis of azetidines and pyrrolidines via iodocyclisation of homoallyl amines and exploration of activity in a zebrafish embryo assay

Article abstract of DOI:10.1039/c3ob41007b

Room temperature iodocyclisation of homoallylamines stereoselectively delivers functionalised 2-(iodomethyl)azetidine derivatives in high yield. Increasing reaction temperature from 20 °C to 50°C switches the reaction outcome to realise the stereoselectiv

Full text of DOI:10.1039/c3ob41007b

Organic&
BiomolecularChemistry
www.rsc.org/obc
Volume 11 | Number 31 | 21 August 2013 | Pages 5049–5196
ISSN 1477-0520
PAPER
John S. Fossey et al.
Synthesis of azetidines and pyrrolidines via iodocyclisation of homoallyl
amines and exploration of activity in a zebrafish embryo assay
1477-0520(2013)11:31;1-9

Organic &
Biomolecular Chemistry
View Article Online
View Journal | View Issue
PAPER
Synthesis of azetidines and pyrrolidines via
iodocyclisation of homoallyl amines and exploration
Cite this: Org. Biomol. Chem., 2013, 11,
5083
of activity in a zebrafish embryo assay†
Antonio Feula,a,b Sundeep S. Dhillon,c Rama Byravan,b Mandeep Sangha,b
Ronald Ebanks,b Mariwan A. Hama Salih,b Neil Spencer,d Louise Male,e
Istvan Magyary,f Wei-Ping Deng,a Ferenc Müllerc and John S. Fossey*b
Room temperature iodocyclisation of homoallylamines stereoselectively delivers functionalised 2-(iodo-
methyl)azetidine derivatives in high yield. Increasing reaction temperature from 20 °C to 50 °C switches
the reaction outcome to realise the stereoselective formation of functionalised 3-iodopyrrolidine deriva-
tives. It was shown that these pyrrolidines are formed via thermal isomerisation of the aforementioned
azetidines. Primary and secondary amines could be reacted with iodomethyl azetidine derivatives to
deliver stable methylamino azetidine derivatives. With subtle changes to the reaction sequences homo-
allyl amines could be stereoselectively converted to either cis- or trans-substituted 3-amino pyrrolidine
derivatives at will. The stereochemical divergent synthesis of cis and trans substituted pyrrolidines sup-
ports an ion part, aziridinium, isomerisation pathway for azetidine to pyrrolidine isomerisation. Six azeti-
dine derivatives were probed in a zebrafish embryo developmental assay to detect potential biological
effects through the analysis of morphology and motility behaviour phenotypes. The range of effects
across the probed molecules demonstrates the suitability of this assay for screening azetidine derivatives.
One of the probed molecules, rac-(((cis)-1-benzyl-4-phenylazetidin-2-yl)methyl)piperidine, exhibited par-
ticularly interesting effects in the developmental assay presenting with hypopigmentation and reduced
circulation amongst others. This shows that the zebrafish embryo provides a fast, sensitive and effective
way to screen new compounds and in the future in combination with existing in vivo and in vitro assays it
will become an integral part in drug discovery and development.
Received 29th April 2013,
Accepted 18th June 2013
DOI: 10.1039/c3ob41007b
www.rsc.org/obc
agrochemical and pharmacological motifs. Whilst pyrrolidines
are a mainstay of the pharmaceutical and natural products
Introduction
Four and five membered saturated nitrogen containing hetero- arenas far fewer molecules containing azetidines have been
cycles, azetidines and pyrrolidines respectively, are important reported, despite their intriguing biological activities.1 Natu-
rally occurring azetidine containing compounds include the
siderophore mugenic acid2,3 and a marine sponge extract
penaresidin A.4–6 Azetidines are often compared to pyrroli-
dines in terms of various biological assays, for example nico-
tine binds less effectively than its azetidine analogue, Fig. 1,
to acetylcholine receptors.7 This coupled with studies into the
psychotropic potency,8 potential CNS-focused leads in drug-
discovery9 and antibacterial activity10 demonstrates that azeti-
dine derivatives are attractive biologically active motifs.
aSchool of Pharmacy, East China University of Science and Technology, 130 Meilong
Road, Shanghai 200237, China
bSchool of Chemistry, University of Birmingham, Birmingham, West Midlands B15
2TT, UK. E-mail: j.s.fossey@bham.ac.uk
cSchool of Clinical and Experimental Medicine, College of Medical and Dental
Sciences, University of Birmingham, Birmingham, West Midlands B15 2TT, UK
dSchool of Chemistry, NMR Spectroscopy Facility, University of Birmingham,
Birmingham, West Midlands B15 2TT, UK
eSchool of Chemistry, X-Ray Crystallography Facility, University of Birmingham,
Birmingham, West Midlands B15 2TT, UK
fUniversity of Kaposvar, Faculty of Animal Science, Department of Nature Protection,
H7400 Kaposvar Guba S. u. 40, Hungary
†Electronic supplementary information (ESI) available: General and experi-
mental chemical procedures, biological procedures and statistics and X-ray crys-
Results and discussion
Synthesis
tallographic information are available. CCDC 936733. For ESI and
crystallographic data in CIF or other electronic format see DOI: 10.1039/
c3ob41007b
In an earlier report we detailed an iodine mediated cyclisation
and isomerisation protocol that delivered pyrrolidines via
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 5083–5093 | 5083

View Article Online
Paper
Organic & Biomolecular Chemistry
Table 1 Optimisation of reaction conditions to form cis-azetidine 2a from 1a
via iodine mediate cyclisation
Entry
Base
Solvent
1aa
2aa
44.0
75.0
61.0
35.5
3aa
4aa
5aa
1
2
3
4
5
6
7
8
—
CH3CN
CH3CN
CH3CN
CH3CN
CH3CN
CH3CN
CH3CN
CH3CN
CH3CN
CH3CN
CH3CN
CH3CN
THF
56.0
25.0
—
—
—
48.5
—
—
—
—
—
—
63.0
—
58.0
—
—
—
39.0
—
—
—
—
—
64.5
82
—
—
—
—
—
—
30.0
—
29.0
—
—
—
—
—
18
—
—
—
—
—
—
—
—
—
—
—
Fig. 1 Examples of azetidine containing molecules alongside pyrrolidine con-
Li2CO3
Na2CO3
K2CO3
Cs2CO3
LiOH
NaHCO3
NaOH
KOH
LiOAc
NaOAc
KOAc
NaHCO3
NaHCO3
NaHCO3
NaHCO3
taining nicotine.
—
33.0
>99.5
60.0
53.0
73.0
68.5
50.0
37.0
50.0
42.0
12.0
18.5
—
40.0
47.0
27.0
31.5
20.0
—
21.0
—
55.0
9
10
11
12
13
14
15
16
MeOH
CH2Cl2
DMSO
33.0
a Ratio determined by comparison of 1H NMR integrations in post
work-up NMR spectra.
Scheme 1 Temperature dependent product distribution in the iodine
mediated cyclisation of homoallyl amine 1a.
unstable azetidine intermediates.11 Herein new observations
and understanding are presented, such as streamlined access
to pyrrolidines, and exploiting the observation that simply
carefully controlling reaction temperature allows direct access
to either azetidines 2 or pyrrolidines 3. For example reaction of
N-benzyl-1-phenylbut-3-en-1-amine (1a) with iodine and
sodium hydrogen carbonate in acetonitrile at 20 °C, delivered
cis-azetidine 2a in 96% yield (which is prone to isomerisation
to cis-pyrrolidines), whereas running this cyclisation reaction
at a slightly elevated temperature (50 °C) cleanly delivers cis-
pyrrolidine 3a as the only detectable product (82% isolated
yield), Scheme 1. It should be noted that the temperature
should be closely controlled through manipulations since aze-
tidines 2 were found to readily isomerise under solvent
removal conditions above 20 °C.
16 h room temperature reaction of homoallyl amine 1a with
three equivalents of iodine in acetonitrile. Whilst initially
selected sodium hydrogen carbonate proved to be the superior
base for azetidine formation, providing 2a as the sole product
with no detectable starting material or by-products (Table 1,
entry 7), the use of no base at all (Table 1, entry 1) permits the
formation of the desired azetidine product in a modest 44%
yield. Use of lithium carbonate as base improves the conver-
sion of amine 1a to azetidine 2a (Table 1, entry 2) a little
(75%). Lithium, sodium and potassium hydroxide (Table 1,
entries 6, 8 and 9 respectively) gave 50–100% conversion of
amine 1a to mixtures of azetidine (2a) and pyrrolidine (3a), as
did lithium and sodium acetate (Table 1, entries 10 and 11
respectively). Potassium acetate (Table 1, entry 12) showed evi-
dence of oxidation of 1a with a product distribution that con-
tained 50% of the desired azetidine 2a, 20% of the
corresponding pyrrolidine 3a accompanied by imine 4a (30%).
Potassium carbonate (Table 1, entry 4) gave complete conver-
sion of 1a to a mixture of azetidine 2a and imine 4a (∼1 : 2
ratio). Somewhat surprisingly cesium carbonate gave only
imine containing products, both the aforementioned oxidation
product 4a (from oxidation of 1a) but also imine 5a (the con-
densation product of benzaldehyde and benzylamine).12,13
After confirming sodium hydrogen carbonate was the most
effective base for azetidine formation THF, methanol, dichloro-
methane and DMSO (Table 1, entries 13–16 respectively) were
tried as solvent. Whilst none of the solvents tried proved to be
superior to acetonitrile in azetidine formation, it is noteworthy
Azetidines
Having established the general synthetic protocol for the syn-
thesis of 2,4-cis-azetidines from homoallyl amines, i.e. treat-
ment with iodine and sodium hydrogen carbonate in
acetonitrile at room temperature,11 the effect of varying the
base was probed. Since iodine containing azetidines 2 are
prone to isomerisation with time the post-workup mixture
obtained from the reaction was used to determine conversion
(proton NMR spectroscopy), post work-up isolates had already
been shown by us to be excellent starting points for sub-
sequent derivatisation.11 Table 1 (entries 1 to 12) details the
effect of varying the base under the standard conditions of a
5084 | Org. Biomol. Chem., 2013, 11, 5083–5093
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
Table 2 Substrate scope for the iodine mediate cyclisation of homoallyl
amines 1
Scheme 2 Proposed mechanism for the deallylation and imine formation
observed in the reaction of 1r under typical cyclisation conditions.
Entry
R1
R2
2 + 3/%a
2 : 3b
Alkyl tert-butyl and cyclohexyl groups at R1 (Table 2, entries
8 and 11) gave complete conversion to azetidine products. An
organometallic fragment in R1 ferrocenyl (Table 2, entry 14)
also gave only azetidine product in quantitative consumption
of homoallylamine starting material. In the variation of R2
para-methoxybenzyl, para-tolylbenzyl, para-nitrobenzyl and
CH2-3-pyridyl (Table 2, entries 2, 3, 12 and 20) respectively all
gave quantitative conversion with only the para-nitrobenzyl
derivative providing anything less than complete selectivity in
favour of azetidine products. Alkyl groups in R2 proved trouble-
some cyclohexyl and n-propyl (Table 2, entries 15 and 16) gave
low conversion to pyrrolidine products (with sodium carbonate
as base), and CH2C(CH3)3 gave slightly higher conversion but
the product distribution was a 1 : 1 mixture of pyrrolidine and
azetidine (under standard conditions).
1
2
3
4
5
6
7
8
a
b
c
d
e
f
g
h
i
j
k
l
m
n
o
p
qd
r
Ph
Ph
Ph
Bn
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
38
>99 : 1
>99 : 1
>99 : 1
5 : 1
6 : 1
3 : 1
3 : 1
>99 : 1
6 : 1
>99 : 1
>99 : 1
7 : 1c
PMB
PTB
Bn
Bn
Bn
Bn
Bn
Bn
Bn
3-Py
4-Py
PNP
2-Br-Ph
t-Bu
3-OMe-Ph
3-Furyl
Cy
Ph
ONP
Fc
Ph
Ph
9
10
11
12
13
14
15
16
17
18
19
Bn
PNB
PMB
Bn
Cy
n-Pr
Bn
3 : 1
>99 : 1
1 : >99c
33
—
80
>99
1 : >99c
4-OMe-Ph
1-Naph
Ph
—
Bn
3-Py
>99 : 1
>99 : 1
s
PMB = para-methoxybenzyl. PTB = para-tolylbenzyl. PNP = para-nitro-
The observation that the iodine mediated cyclisation con-
sistently delivers cis relative stereochemistry can be ascribed to
the suggested differences in transition state geometries that
would be required to deliver cis or trans products. Addition of
iodine to the carbon–carbon double bond of a homoallyl
amine, such as 1a, could form two diastereomeric iodiranium
species (Scheme 3). In order for cyclisation to a four-
membered azetidine to occur two possible transition states are
possible. The most disfavoured of the two transition states
would increase a pseudo 1,3-diaxial interaction, whereas the
transition state reasoned to be more favoured positions
protons in the pseudo axial positions (Scheme 3: upper and
lower transition state respectively).
Since it had been already demonstrated that iodoazetidines
of the general class cis-2 were highly prone to isomerisation
and that the trick of rapidly reacting the crude, post work-up
azetidine with a nucleophile to replace iodine could deliver
stable cis-azetidine derivatives (cis-6), a range of amine deriva-
tives of cis-azetidines 2 were prepared in 50 to 87% isolated
yield over two steps from readily accessible homoallyl
amines 1, Table 3. It was vital that liquid amines were used
neat (as solvent) in the iodine displacement step, when co-
solvents were used isomerisation to the corresponding pyrroli-
dines (3) became problematic, however this feature can be
exploited (as discussed later) in pyrrolidine synthesis. Yields of
amine bearing azetidines 6 ranged from 50 to 82% over two
steps from starting homoallyl amines 1. Primary and second-
ary amines could be used in the derivatisation, linear
and branch aliphatic, benzylic, and heteroatom containing
substituents in the derivatising amine part could all be
accommodated.
phenyl. PNB
= para-nitro-benzyl. ONP = ortho-nitrophenyl. Fc =
ferrocenyl. a 2 + 3 by 1H NMR spectroscopy. b cis-(2 : 3) Ratio obtained
by inspection of the 1H NMR spectrum post aqueous work-up.
c Sodium carbonate was used as base. d 1r suffered loss of allyl group
and oxidation to corresponding imine.
that both methanol and DMSO did show evidence of amine
oxidation to imine 4a (∼30% in both cases).
Next substrate scope was investigated, Table 2 shows the
effect of varying the substituents labelled R1 and R2 in starting
amines 1a–t. Under conditions optimised for iodoazetidine
formation conversion of starting amine to cyclised products
was generally high and showed no evidence for trans substi-
tution about the azetidine (or pyrrolidine where detected)
rings, the ratio of azetidine to pyrrolidine was good to excellent
for benzylic R2 substituents. Aromatic substituents bearing
electron donating groups in R1 (e.g. 3- and 4-OMe, Table 2,
entries 9 and 17 respectively) were employed, and whilst the
3-methoxy derivative gave complete conversion to a 3 : 1 mixture
of the corresponding azetidine and pyrrolidine the 4-methoxy
derivative failed to give any cyclised product, rather an apparent
loss of the allyl group and oxidation of 1r was observed.
Perhaps, after initial activation of the alkene by iodine, allyl
iodide might be expelled assisted by the resonance stabiliz-
ation of the generated cation, as proposed in Scheme 2. Elec-
tron withdrawing groups in R1 4- and 2-nitro phenyl (Table 2,
entries 6 and 13) gave 3 : 1 mixtures of azetidine to pyrrolidine.
Heteroaromatic 3- and 4-pyridyl and 3-furyl (Table 2, entries 4,
5 and 10) gave complete conversion to cyclised product with
good to excellent azetidine to pyrrolidine ratios.
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 5083–5093 | 5085

View Article Online
Paper
Organic & Biomolecular Chemistry
Scheme 3 Proposed transition state argument for the cis-selectivity observed in the iodine mediated cyclisation of homoallyl amines.
The nucleophilic displacement of iodine from cis-iodoazeti- synthesis arena.20–23 It was shown that pyrrolidine cis-3a could
dines was employed in an azide to triazole so-called click reac- be quantitatively formed upon reaction of homoallyl amines
tion sequence. Homoallyl amines 1a and 1d were reacted with iodine and a base with mild heating. It was also pre-
under standard cyclisation conditions and worked-up as viously shown (by us in an earlier report11) that iodine contain-
before, the iodoazetidines 2a and 2d were treated with sodium ing azetidine derivative 2d stereospecifically isomerized, with
azide in DMF at room temperature and stirred overnight time, to the corresponding pyrrolidine cis-3d. In order to
(16 h), all the while ensuring the temperature did not exceed confirm isomerisation of iodoazetidines to iodopyrrolidines,
20 °C, the reactions were monitored by TLC (silica EtOAc– azetidines 2c–e were separately heated at 50 °C in acetonitrile
Hex), which indicated that although the displacement of for 4 hours (Scheme 6), in each case ≥95% of the corres-
iodide by azide was incomplete the iodides 2a and 2b were ponding pyrrolidine 3c–e was isolated. In this report evidence
showing evidence of a small level of isomerisation to the of a dissociative isomerisation mechanism is revealed, which
corresponding pyrrolidines (as judged by TLC). In order to corroborates hypotheses of Couty and others in the area (see
minimise contamination from products of reaction after azeti- later).24,25
dine isomerisation a terminal alkyne (phenyl acetylene or
Paulmier, Outurquin and co-workers reported selenium
pent-1-yne) was added at that point with copper(I) iodide and induced cyclisation of homoallyl amines to azetidines and pyr-
the reaction stirred for a further 24 h. Triazole appended azeti- rolidines.26,27 Bott et al. reported a one carbon ring expansion
dines 7aa, 7ab and 7da were obtained in approximately 50% of azetidines via ammonium ylide shifts.28 Couty and co-
yield over the three synthetic steps from homoallyl amines 1a workers have studied a number of ring enlargement reactions
or 1d (Scheme 4).
that deliver pyrrolidines from azetidines, such as fluoride,29
The cyclisation, iodide displacement sequence was applied chloride25 and bromide derivatives,30 alkene derivatives31 and
to the synthesis of single enantiomer azetidines (S,R)-6ab, hydroxy derivatives.25 Van Brabandt et al. also utilised ring
(S,R)-6tb, (S,R)-6ua and (S,R)-6ue.14 The synthesis of (S,R)-6ab expanded chloride and bromide bearing azetidines.32 Amongst
is shown in Scheme 1 which describes the use of single enan- this catalogue of work Couty and coworkers reported a compu-
tiomer 1a, no loss of enantiopurity is witnessed and (S,R)-6ab tational study where ring expansion of 2-(chloromethyl)-1-iso-
was isolated in 84% yield. Single enantiomer 1a was obtained propylazetidine was discussed,24 and noted that a polar
utilising a stereoselective allylation of the corresponding solvent (DMSO) reduced the barrier to isomerisation and the
imine15 via an Ellman auxiliary protocol (Scheme 5).16–19
reaction proceeds via two transition states through a strained
intermediate, Scheme 7.
Pyrrolidines
It is also important to point out that a five (pyrrolidine) to
Whilst azetidines represent an attractive synthetic target pyrro- six (piperidine) membered ring expansion reaction that
lidines, the side product of our azetidine formation protocol, exploits the nucleophilic displacement of a beta iodide by
are highly desirable heterocyclic constructs in the organic Davies et al. supports an ion pair aziridinium salt hypothesis
5086 | Org. Biomol. Chem., 2013, 11, 5083–5093
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
Table 3 Aminoazetidines synthesised in two steps from homoallylamines 1
Scheme 5 Non-racemic synthesis of amino azetidine (2S,4R)-6ab in 84% yield.
Scheme 4 Synthesis of triazole appended azetidines 7 in ∼50% yield over
three steps.
Using the knowledge that the isomerisation of 2,4-cis-azeti-
dines 2 to 2,4-cis-pyrrolidines 3 is facile and our knowledge
that nucleophilicly displacing iodine from azetidine products
since they were able to isolate (and characterise by single
crystal X-ray diffraction) the tetrafluoroborate salt 13 of their delivers stable amino azetidines we chose to run a similar pro-
intermediate (Scheme 8), which further reacts with acetate to tocol for generation of trans-pyrrolidine (whilst never observed
give a 55 : 45 mixture of compounds 14 : 15.33
it was speculated that iodopyrrolidines might be prone to
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 5083–5093 | 5087

View Article Online
Paper
Organic & Biomolecular Chemistry
Scheme 6 Thermal isomerisation of cis-azetidines 2c–e to cis-pyrrolidines
3c–e.
Scheme 7 Computationally determined reaction sequence via a strained inter-
mediate of Couty et al.24
Scheme 9 The synthesis of 2,4-cis- and trans-pyrrolidines (8, 9 and 10) from
homoallyl amines (1). Inset: X-ray crystal structure of trans-9da, only pyrrolidine
and triazole C–H protons shown for clarity, nitrogen atoms are coloured blue
(Ortep plot 30% probability ellipsoids), for full details see ESI.†
Scheme 8 Ring expansion of 12 through aziridinium formation (13) to piperi-
dine 14 by Davies et al.33
products (as indicated earlier in the azetidine section, use of a
solvent rather than neat amine gave some pyrrolidine product)
polymerisation via N-quaternisation or other side reactions so but these products were the cis-diastereoisomer 8ab, aj and kb
reaction of the iodide directly was deemed appropriate). It was in greater than 77% isolated yield from the corresponding
envisaged that, in one pot, a homoallyl amine (1) could be homoallylamines, Scheme 9 – right side. Thus it is possible to
cyclised to a 2,4-cis-azetidine (2), which in turn could be iso- synthesise both diastereoisomers of the 2,4-pyrrolidines from
merised by action of heat to a 2,4-cis-pyrrolidine (3). The cis- the same starting material in good yield and high diastereo-
pyrrolidine thus formed upon treatment with a nucleophile selectivity simply by adjusting the addition sequence and reac-
would deliver a trans-substituted pyrrolidine. Indeed, for the tion temperature. Alongside this finding it was also possible to
case of homoallyl amines 1a, c, d and m the corresponding cis- prepare cis-2,4-hydroxy functionalised pyrrolidines (although it
azetidines (2a, c, d and m) and then cis-pyrrolidines (3a, c, d was not possible to obtain the trans isomer) by using hydrox-
and m) were identified by proton NMR spectroscopy and ide as the nucleophile in 72–89% yield in the cases tried
treated with a primary amine to give 2,4-trans-pyrrolidines 8ab, (including a pyrrolidine bearing a 3-pyridyl functionality in R1
aj and kb (>73% yield over three steps) or with sodium azide offering access to functionalised analogues of a nicotine core).
followed by a terminal alkyne and a copper(I) catalyst to give
The observation that both cis- and trans-pyrrolidines 8 are
2,4-trans-pyrrolidines 9aa, ca and da as single diastereoisomers readily accessible from the same starting materials by simply
in good isolated yields (57–65% yield over four steps), subtly changing the reaction conditions (namely adding
Scheme 9 – left side. An X-ray crystal structure of trans-9da nucleophile before or after isomerisation of azetidine to pyrro-
helped confirm the relative trans-stereochemistry (inset lidine) leads to the conclusion that the isomerisation is likely
Scheme 9). To our surprise (and later delight) when cis-azeti- to proceed via intermediate 11, Scheme 10. In the absence of
dines 2a and 2k (confirmed by proton NMR spectroscopy) any external nucleophile the cis-azetidine isomerises to the
where stirred at room temperature in DMSO with primary corresponding cis-pyrrolidines 3, which can then undergo an
amine nucleophiles not only were pyrrolidines the major SN2 reaction with a nucleophile, such as a primary amine, to
5088 | Org. Biomol. Chem., 2013, 11, 5083–5093
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
Scheme 10 Divergent conversion of cis-azetidines 2 via intermediate 11 to both cis- and trans-pyrrolidines 8.
give trans-substituted pyrrolidines 8. However, intermediate 11
ought to be stabilised by polar solvent (according to calcu-
lations for a related system by Couty et al.), so if it forms in the
presence of an external nucleophile (e.g. primary amine) it is
intercepted to give cis-8.
Preliminary biological activity
Zebrafish embryo development can be used as an assay to
assess potential drug-like behaviour of molecules in a complex
organism; certain developmental manifestations can help to
provide early indications of possible utility in therapeutic
regimes and also toxic side-effects that could develop.34–41
Fig. 2 Azetidines selected for preliminary biological screening.
Zebrafish are a suitable vertebrate model for drug screening
due to their ease of access, low maintenance costs, optical
transparency, fast development and relatedness to humans in
terms of physiology, genetics and toxic responses.40 In com-
parison, rodent models are more expensive to maintain, their
development is slow and occurs in utero making drug testing
difficult and the number of embryos that can be obtained is
also small reducing the scale of the experiments.42 Addition-
ally, in comparison to in vitro cell culture models, the zebrafish
embryo enables the response of the organism as a whole to a
compound to be observed, such as cell–cell and cell–matrix
interactions, as well as enabling the efficiency of uptake,
metabolism and excretion of a compound to be determined;
information that cannot be obtained from in vitro cell culture
models.42 Zebrafish screens require small amounts of com-
pound, which saves money and the compound can also be
added directly to the bathing solution, so injection is not
required for drug delivery. Zebrafish screens can also be auto-
forward chemical screen in zebrafish carried out by Das et al. a
retinoid analogue which caused specific cardiovascular defects
in zebrafish (heart tube deformations) was identified, indicat-
ing potential usefulness against conditions such as acute pro-
myelocytic leukaemia.45
For our investigation of biological activity using zebrafish
embryo development as a screening technique six azetidine
derivatives were selected. The azetidines selected, racemic cis-
6ae, 6ac, 6ad, 6ja, 6ga and 6di represent a cross section of the
azetidines synthesized and provide an interesting comparison
in terms of the chemical space they explore, Fig. 2. Compari-
sons between the structures selected could be made along the
following lines tertiary amine versus secondary amine; ali-
phatic versus benzylic; heteroaromatic versus aromatic; etc.
Zebrafish embryo developmental morphology assay
mated enabling high throughput screening to be carried out Table 4 shows a summary of the morphological defects
where thousands of compounds can be tested at one time. The observed after five days following treatment of zebrafish
first chemical screen in zebrafish was carried out by Peterson embryos at 75%-epiboly stage with the six selected azetidines
and colleagues in 2000 where compounds from a library were at a concentration of 25 μM. Azetidines were added as DMSO
screened in zebrafish to see what phenotypes they evoked, solutions against a control and a DMSO blank. Most com-
some of the phenotypes observed included effects on pigmen- pounds had minimal effects on overall morphology (see
tation and ear development.43 These phenotype based screens Table S1, ESI†). However, compound cis-6ae was found to
can help to uncover the mechanism by which compounds may cause severe morphological effects over the whole exposure
be acting. For example when fertilised zebrafish embryos are period showing increased severity of phenotypes at higher
exposed to a certain concentration of probe molecule and the doses (see Fig. S1 to S8, ESI†), cis-6ga was also found to cause
morphological changes are monitored, reduced pigmentation morphological effects at later time points (see Fig. S9, ESI†).
can be an indicator that the probe molecule may be acting on
Compound cis-6ae, which contains a piperidine fragment,
thyroid hormone signaling and therefore warrants further shows effects in all of the nine morphological defect indi-
investigation as an agent for treating thyroid disorders.44 In a cators. This suggests cis-6ae has high biological activity in a
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 5083–5093 | 5089

View Article Online
Paper
Organic & Biomolecular Chemistry
Table 4 Summary of effects of chosen azetidines (25 μM) on zebrafish embryos after 5 days
Morphological defects
Reduced
circulation/blood
clots
Curved
tails
Abnormal
jaw
Delayed
growth
Lack of
pigmentation
Slowed
heart rate
Cardiac
oedema
Brain
oedema
Additive
Tail necrosis
cis-6ae
cis-6ja
cis-6ga
cis-6ac
cis-6di
cis-6ad
Control
Blank
++++
+
+++
+
+/−
+/−
+/−
—
++++
—
—
—
—
+/−
—
—
++++
+
+
+/−
—
+/−
+/−
—
++++
—
—
—
—
—
—
—
++++
++
+++
+
+/−
+
++++
—
—
—
—
—
—
—
++++
++
+++
+
+/−
+
+++
—
—
—
—
—
—
—
++++
++
+++
+
+/−
+
+/−
+/−
+/−
+/−
+/−
+/−
++++ = very severe effect (75–100%); +++ = severe effect (50–75%); ++ = moderate effect (25–50%); + = minimal effect (5–25%); +/− = either
minimal or no effect (0–5%); − = no effect (0%).
number of tissues, resembling many effects seen when zebra- benchmarking, a small ring nitrogen-containing heterocycle
fish embryos are exposed to piperidines, effects such as hypo- which has been extensively studied is nicotine, which is a natu-
pigmentation and cardiac edema.46 Next most active is rally occurring alkaloid that acts on nACh receptors. Nicotine
compound cis-6ga, which bears an ortho-bromophenyl substi- in is extremely toxic to humans and is a known human terato-
tuent in the 2-position of the of the azetidine ring. Compound gen.53 In zebrafish embryos, nicotine has been found to
cis-6ga resulted in curved tails, slowed heart rate, cardiac reduce notochord length, cause apoptotic death of brain
edema and reduced circulation in more than 50% of embryos. cells,54 lead to underdeveloped tails, reduced number of
The 3-furyl derivative 6ja showed a similar but less extensive somites, increased muscular bends in spine and cause compli-
spread of defects. Compounds cis-6ae, cis-6ga and cis-6ja all cations in jaw formation.53 Many of these effects correlate with
produced effects similar to those seen with nicotine;47 those seen with some of the azetidines screened here, particu-
however, they showed varying degrees of severity. An additional larly compound cis-6ae. The muscular bends, which are
feature observed with cis-6ae not included in the morphologi- common features of exposure to compound cis-6ae and cis-6ga
cal defect scoring was the development of an atrioventricular are thought to occur due to slow muscle fibres becoming dis-
block, specifically a 2 : 1 type (data not shown), which is often organised and the nACh receptors becoming desensitised,
observed when zebrafish embryos are treated with QT prolong- which explains the short stature of the fish and the bent tails.55
ing drugs,48 suggesting that this compound in particular may With compound cis-6ae reduced locomotion was observed when
also be affecting cardiac ion channels. Additionally, somites 6 dpf larval zebrafish were treated at a concentration of 20 μM,
were found to be u-shaped and melanophores were found to which is similar to that which occurs when zebrafish are treated
be small and granulated, as well as reduced in number. with high concentrations of nicotine.56 This is because at low
U-shaped somites are characteristic features observed when concentrations, nicotine causes stimulation of nACh receptors
there are defects in slow muscle development caused by dis- in the CNS, whereas at higher concentrations nicotine will bind
ruption to the Hedgehog signaling pathway.49 The pigmenta- more preferentially to nACh receptors in muscles (particularly
tion effects seen may have been caused by reduced cyclic those involved in movement), which can have negative depress-
adenosine monophosphate (cAMP) levels and disruption of ant effects reducing locomotion.
microtubule function, as both are required for aggregating and
Since compound cis-6ae was shown to elicit morphological
dispersing melanin.50 Additional disruption to xanthophores defects across the nine probed criteria it was selected for a
may have also contributed to the pigmentation phenotype at more detailed investigation. Specific characterisation (detailed
later stages.51 Embryos treated with compound cis-6ae were phenotypic analysis) of compound 6ae to look for further
also unable to hatch, which could have been due to the hatch- effects undetectable by morphology analysis alone was
ing enzyme chorionase being inhibited, osmotic disturbances embarked upon (Fig. 3). Analysis of locomotion by assessing
or due to weakened muscular movements.52 The other com- motility of zebrafish larvae through measurement of pixel
pounds tested in this preliminary screen did not show any change over time, overall blood vessel development through
marked defects, which could be due to the small concentration fluorescence visualisation of vascular endothelial cells using
range tested, as different compounds are likely to have the Tg(fli-1:EGFP) transgenic zebrafish line and blood circula-
different effective concentrations. However, it is clear that tion through fluorescence visualisation of red blood cells
among even this small sample set various important pharma- using Tg(gata-1:dsRed); Tg(fli-1:EGFP) transgenic zebrafish line
cological effects can be seen and small structure variation have revealed that compound cis-6ae reduced locomotion, did not
wide impact on activity, boding well for future investigations. affect gross blood vessel development and reduced circulation
Since these compounds have not been previously tested in zebra- with formation of blood clots. Compound cis-6ae at a concen-
fish, correlation to a similar molecule can provide important tration of 20 μM was found to significantly reduce movement
5090 | Org. Biomol. Chem., 2013, 11, 5083–5093
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
Fig. 3 Effects of compound cis-6ae: (A) bar chart comparing movement/locomotion of cis-6ae-treated 6 dpf zebrafish larvae compared to controls; (B) bright field
lateral view of a cis-6ae-treated 2 dpf embryo presenting with morphological defects (H = hypopigmentation, B = bent spine, C = cardiac edema) compared to
control; (C) bright field lateral view of a cis-6ae-treated 36 hpf zebrafish embryo compared to control (left) with corresponding fluorescence images of vasculature
through fli-1:EGFP transgene activity (right) where individual blood vessels can be seen (arrows) and the formation of a blood clot in the cis-6ae-treated embryo; (D)
bar chart comparing frequency of tail necrosis in cis-6ae-treated 36 hpf embryos compared to controls; (E) bright field lateral view of a cis-6ae-treated 36 hpf
zebrafish embryo compared to control (left) with corresponding fluorescence images of blood circulation through gata-1:dsRed transgene activity (right) where
erythrocytes can be seen (arrows) and the formation of a blood clot in the cis-6ae-treated embryo; (F) bar chart comparing frequency of blood clots in cis-6ae-
treated 36 hpf embryos compared to controls.
of 6 dpf zebrafish compared to controls (Fig. 3, A) (n = 60, P < not fully formed and tail necrosis was evident with no blood
0.05). Morphological defects observed after 48 hours following vessels present in the necrotic part of the tail. Tail necrosis
treatment with 25 μM cis-6ae at 75%-epiboly stage included was found to occur in all embryos treated with 50 μM cis-6ae
bent tails, hypopigmentation, severely delayed development, (Fig. 3, D) (n = 40, P < 0.05). However, overall vascular develop-
brain and cardiac edema (Fig. 3, B). Analysis of 2 dpf Tg(fli-1: ment was unperturbed. Analysis of 2 dpf Tg(gata-1:dsRed); Tg-
EGFP) zebrafish treated at 75%-epiboly with 50 μM cis-6ae (fli-1:EGFP) zebrafish treated at 75%-epiboly with 50 μM cis-6ae
(Fig. 3, C) under the green channel (which marks blood (Fig. 3, E) under the red channel (which labels circulating
vessels) showed severely delayed growth with the tip of the tail blood cells) showed reduced circulation and the formation of
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 5083–5093 | 5091

View Article Online
Paper
Organic & Biomolecular Chemistry
blood clots. The formation of blood clots was found to occur in this iodine and base mediated cyclisation and subsequent
in all embryos treated with 50 μM cis-6ae with aggregation of halide displacement that gave selective access to these three
erythrocytes evident at the end of the tail due to reduced blood groups of compound. A zebrafish embryo developmental mor-
circulation (Fig. 3, F) (n = 40; P < 0.05).
phology assay was demonstrated as a useful tool in assessing
Exposure to compound cis-6ae resulted in numerous mor- the biological activity of a sample of azetidine derivatives and a
phological defects in zebrafish embryos, suggesting that it was compound bearing an azetidine and a piperidine to elicit mor-
able to act on many tissues. Many similarities to nicotine toxi- phological responses across the panel of indicators of potential
city were also observed suggesting that it may have a similar activity in a variety of settings. This demonstrates both the
mode of action or operate via a similar route. Compound cis- utility of the fertilised zebrafish egg morphology screening tech-
6ae displayed similar effects to those which have been docu- nique and the potential for the underlying azetidine scaffold
mented in zebrafish embryos following treatment with discussed to be probed for biological activity using advanced
different piperidines, such as 2-methylpiperidine,46 such com- high throughput techniques employing zebrafish embryos.
pounds are often used as insecticides.57 Recent studies with With the increasing number of transgenic lines available and
pyrethroid insecticides also showed similar morphological the latest technological advances the zebrafish will provide a
defects, such as jaw abnormalities, curvature of body axis and platform for drug screening in the years to come.
pericardial edema.58 Additionally, the pigmentation defects
observed with cis-6ae are similar to those seen with anti-
thyroid agents such as methimazole,59 suggesting there might
be an effect on thyroid hormone signalling. However, there are
Acknowledgements
many ways in which pigmentation can be reduced (e.g. disturb- The authors thank the University of Birmingham and ERDF
ances to non-skeletal neural crest or melanophore formation), AWM II for support. AF thanks The School of Chemistry, Uni-
therefore further investigation would be required to ascertain versity of Birmingham for a studentship and the EPSRC for
the precise cause in this case. Overall, cis-6ae shows high toxi- underpinning support. MAHS thanks the HCDP - Ministry of
city at the concentrations that were used in this study and Higher Education-Kurdistan Region for financial support. JSF
would be detrimental particularly to aquatic habitats if it were is extremely grateful to the Leverhulme Trust (F/00351/P and
to be released into such an environment. Further studies to F/00094/BC) for early funding, East China University of Science
investigate mechanism of action and dose ranging may reveal and Technology for a visiting professorship, the Natural
potential therapeutic benefits (e.g. treatment for overactive Science Foundation of China for a research fellowship (no.
thyroid/lice/scabies) or industrial applications (e.g. use as an 21050110426) and the Royal Society Industry Fellowship
agricultural insecticide).
scheme (2012/R1) for on-going support. FM and SSD thank the
From the selection of compounds tested at the specified Marie Curie ITN project BOLD (Biology of liver and pancreatic
concentrations, compound cis-6ae, as mentioned previously, development and disease) and “ZF-Health” Integrating pro-
has detrimental effects on zebrafish embryo development, jects in the Framework 7 programme of the European Com-
making this compound particularly toxic to aquatic environ- mission for their financial support. The EPSRC UK National
ments. Further studies would be necessary to elucidate its Crystallography Service at the University of Southampton are
mechanism of action and metabolism in zebrafish. This study thanked for the collection of the crystallographic data.60
however shows that the zebrafish embryo is a suitable verte-
brate model to investigate lead compounds for toxic effects
due to its ease of accessibility, optical transparency, contri-
bution to 3Rs (replacement, reduction and refinement) of
References
animals in research, fast development, low cost of mainten-
ance and high similarity in toxic responses when compared to
humans and rodents. These advantages make the zebrafish
embryo an attractive model for testing during pre-clinical drug
development before tests are performed in rodents or higher
mammals, saving time and money by filtering out any lead
compounds with safety liabilities before they reach the next
step in the drug development pipeline.
1 G. Cardillo, L. Gentilucci and A. Tolomelli, in Asymmetric
Synthesis of Nitrogen Heterocycles, ed. J. Royer, Wiley-VCH
Verlag GmbH & Co. KGaA, 2009, pp. 1–50.
2 J. F. Ma and K. Nomoto, Plant Physiol., 1993, 102, 373–378.
3 T. Takemoto, K. Nomoto, S. Fushiya, R. Ouchi, G. Kusano,
H. Hikino, S. I. Takagi, Y. Matsuura and M. Kakudo, Proc.
Jpn. Acad., Ser. B, 1978, 54, 469–473.
4 S. Knapp and Y. H. Dong, Tetrahedron Lett., 1997, 38, 3813–
3816.
5 H. Takikawa, T. Maeda and K. Mori, Tetrahedron Lett.,
1995, 36, 7689–7692.
6 H. Yoda, T. Uemura and K. Takabe, Tetrahedron Lett., 2003,
44, 977–979.
7 J. D. Schmitt, Curr. Med. Chem., 2000, 7, 749–800.
8 D. X. Wang, H. Booth, N. Lerner-Marmarosh, T. S. Osdene
and L. G. Abood, Drug Dev. Res., 1998, 45, 10–16.
Conclusions
A synthetically versatile protocol for the stereospecific cyclisa-
tion of homoallyl amine derivatives ultimately delivered cis-
azetidine and cis or trans pyrrolidine derivatives. Importantly,
it was the subtle change of temperature and reaction sequence
5092 | Org. Biomol. Chem., 2013, 11, 5083–5093
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
9 J. T. Lowe, M. D. Lee, L. B. Akella, E. Davoine, 34 C. Delvecchio, J. Tiefenbach and H. M. Krause, Assay Drug
E. J. Donckele, L. Durak, J. R. Duvall, B. Gerard, Dev. Technol., 2011, 9, 354–361.
E. B. Holson, A. Joliton, S. Kesavan, B. C. Lemercier, H. Liu, 35 C. A. Macrae, Expert Opin. Drug Discovery, 2010, 5, 619–
J.-C. Marié, C. A. Mulrooney, G. Muncipinto, M. Welzel- 632.
O’Shea, L. M. Panko, A. Rowley, B.-C. Suh, M. Thomas, 36 T. V. Bowman and L. I. Zon, ACS Chem. Biol., 2010, 5, 159–
F. F. Wagner, J. Wei, M. A. Foley and L. A. Marcaurelle,
J. Org. Chem., 2012, 77, 7187–7211.
10 J. Frigola, J. Pares, J. Corbera, D. Vano, R. Merce,
161.
37 C. Chakraborty, C. H. Hsu, Z. H. Wen, C. S. Lin and
G. Agoramoorthy, Curr. Drug Metab., 2009, 10, 116–124.
A. Torrens, J. Mas and E. Valenti, J. Med. Chem., 1993, 36, 38 T. P. Barros, W. K. Alderton, H. M. Reynolds, A. G. Roach
801–810. and S. Berghmans, Br. J. Pharmacol., 2008, 154, 1400–1413.
11 A. Feula, L. Male and J. S. Fossey, Org. Lett., 2010, 12, 5044– 39 L. I. Zon and R. T. Peterson, Nat. Rev. Drug Discovery, 2005,
5047. 4, 35–44.
12 The unexpected oxidation, deallylation, sequence wit- 40 G. J. Lieschke and P. D. Currie, Nat. Rev. Genet., 2007, 8,
nessed when using cesium carbonate as base was exam-
ined in detail in a parallel report, see next reference.
13 A. Feula and J. S. Fossey, RSC Adv., 2013, 3, 5370.
14 See ESI† for details.
15 S. Kobayashi, Y. Mori, J. S. Fossey and M. M. Salter, Chem.
Rev., 2011, 111, 2626–2704.
16 G. C. Liu, D. A. Cogan and J. A. Ellman, J. Am. Chem. Soc.,
1997, 119, 9913–9914.
17 G. C. Liu, D. A. Cogan, T. D. Owens, T. P. Tang and
J. A. Ellman, J. Org. Chem., 1999, 64, 1278–1284.
18 D. A. Cogan, G. C. Liu and J. Ellman, Tetrahedron, 1999, 55,
8883–8904.
353–367.
41 R. T. Peterson and M. C. Fishman, Methods Cell Biol., 2011,
105, 525–541.
42 G. N. Wheeler and A. W. Brandli, Dev. Dyn., 2009, 238,
1287–1308.
43 R. T. Peterson, B. A. Link, J. E. Dowling and S. L. Schreiber,
Proc. Natl. Acad. Sci. U. S. A., 2000, 97, 12965–12969.
44 O. A. Elsalini and K. B. Rohr, Dev. Genes Evol., 2003, 212,
593–598.
45 B. C. Das, K. McCartin, T. C. Liu, R. T. Peterson and
T. Evans, PLoS One, 2010, 5, e10004.
46 C. Schulte and R. Nagel, ATLA, Altern. Lab. Anim., 1994, 22,
12–19.
47 L. T. Thomas, L. Welsh, F. Galvez and K. R. Svoboda,
Toxicol. Appl. Pharmacol., 2009, 239, 1–12.
19 J. A. Ellman, T. D. Owens and T. P. Tang, Acc. Chem. Res.,
2002, 35, 984–995.
20 T. Harrison, Contemp. Org. Synth., 1995, 2, 209–224.
21 M. Wang, Z. Wang, Y.-H. Shi, X.-X. Shi, J. S. Fossey and 48 U. Langheinrich, G. Vacun and T. Wagner, Toxicol. Appl.
W.-P. Deng, Angew. Chem., Int. Ed., 2011, 50, 4897–4900. Pharmacol., 2003, 193, 370–382.
22 M. P. Patil, A. K. Sharma and R. B. Sunoj, J. Org. Chem., 49 I. G. Woods and W. S. Talbot, PLoS Biol., 2005, 3, e66.
2010, 75, 7310–7321.
23 G. Barker, J. L. McGrath, A. Klapars, D. Stead, G. Zhou,
50 D. W. Logan, S. F. Burn and I. J. Jackson, Pigm. Cell Res.,
2006, 19, 206–213.
K. R. Campos and P. O’Brien, J. Org. Chem., 2011, 76, 5936– 51 J. Odenthal, K. Rossnagel, P. Haffter, R. N. Kelsh,
5953.
E. Vogelsang, M. Brand, F. J. van Eeden, M. Furutani-Seiki,
M. Granato, M. Hammerschmidt, C. P. Heisenberg,
Y. J. Jiang, D. A. Kane, M. C. Mullins and C. Nusslein-
Volhard, Development, 1996, 123, 391–398.
24 F. Couty and M. Kletskii, J. Mol. Struct. (THEOCHEM), 2009,
908, 26–30.
25 F. Durrat, M. V. Sanchez, F. Couty, G. Evano and J. Marrot,
Eur. J. Org. Chem., 2008, 3286–3297.
26 B. Berthe, F. Outurquin and C. Paulmier, Tetrahedron Lett.,
1997, 38, 1393–1396.
27 F. Outurquin, X. Pannecoucke, B. Berthe and C. Paulmier,
Eur. J. Org. Chem., 2002, 1007–1014.
52 Q. He, K. Liu, S. Wang, H. Hou, Y. Yuan and X. Wang, Drug
Chem. Toxicol., 2012, 35, 149–154.
53 E. W. Klee, J. O. Ebbert, H. Schneider, R. D. Hurt and
S. C. Ekker, Nicotine Tob. Res., 2011, 13, 301–312.
54 B. Parker and V. P. Connaughton, Zebrafish, 2007, 4, 59–68.
28 T. M. Bott, J. A. Vanecko and F. G. West, J. Org. Chem., 55 L. Welsh, R. L. Tanguay and K. R. Svoboda, Toxicol. Appl.
2009, 74, 2832–2836. Pharmacol., 2009, 237, 29–40.
29 B. Drouillat, F. Couty, O. David, G. Evano and J. Marrot, 56 A. M. Petzold, D. Balciunas, S. Sivasubbu, K. J. Clark,
Synlett, 2008, 1348.
30 F. Couty, F. Durrat and D. Prim, Tetrahedron Lett., 2003, 44,
5209–5212.
V. M. Bedell, S. E. Westcot, S. R. Myers, G. L. Moulder,
M. J. Thomas and S. C. Ekker, Proc. Natl. Acad. Sci. U. S. A.,
2009, 106, 18662–18667.
31 F. Couty, F. Durrat, G. Evano and J. Marrot, Eur. J. Org. 57 K. A. Bandara, V. Kumar, U. Jacobsson and L. P. Molleyres,
Chem., 2006, 4214–4223. Phytochemistry, 2000, 54, 29–32.
32 W. Van Brabandt, R. Van Landeghem and N. De Kimpe, 58 A. DeMicco, K. R. Cooper, J. R. Richardson and L. A. White,
Org. Lett., 2006, 8, 1105–1108. Toxicol. Sci., 2010, 113, 177–186.
33 S. G. Davies, R. L. Nicholson, P. D. Price, P. M. Roberts, 59 O. A. Elsalini and K. B. Rohr, Dev. Genes. Evol., 2003, 212,
A. J. Russell, E. D. Savory, A. D. Smith and J. E. Thomson,
593–598.
Tetrahedron: Asymmetry, 2009, 20, 758–772.
60 S. J. Coles and P. A. Gale, Chem. Sci., 2012, 3, 683–689.
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 5083–5093 | 5093