Studies toward labeling cytisine with [11C]phosgene: Rapid synthesis of a δ-lactam involving a new chemoselective lithiation-annulation method

Article abstract of DOI:10.1021/jo0498157

With the aim of the radiolabeling of cytisine, a potent agonist of nicotinic receptors, with [¹¹C]-phosgene, the rapid synthesis of a lactam model of our target has been studied. The key step of the δ-lactam formation is a new chemoselective lithiation-annulation method, under high dilution, of a suitable piperidinylcarbamoyl chloride. This precursor was obtained from (2-hydroxyethyl)-piperidine in a linear synthetic sequence involving a Corey-Fuchs olefination of the corresponding aldehyde, followed by a selective reduction, using a diimide equivalent, of an iodoalkyne into a (Z)-iodopropene piperidine. This alkene served as main precursor to study the cyclization according to several procedures using phosgene as the required carbonylating reagent.

Full text of DOI:10.1021/jo0498157

Stu d ies tow a r d La belin g Cytisin e w ith [11C]P h osgen e: Ra p id
Syn th esis of a δ-La cta m In volvin g a New Ch em oselective
Lith ia tion -An n u la tion Meth od
J acques Rouden,* Thomas Seitz, Laurent Lemoucheux, and Marie-Claire Lasne*
Laboratoire de Chimie Mole´culaire et Thio-organique, CNRS UMR 6507, ENSICAEN,
Universite´ de Caen-Basse Normandie, 6 Boulevard du Mare´chal J uin, F14050 Caen Cedex, France
rouden@ismra.fr
Received February 2, 2004
With the aim of the radiolabeling of cytisine, a potent agonist of nicotinic receptors, with [11C]-
phosgene, the rapid synthesis of a lactam model of our target has been studied. The key step of the
δ-lactam formation is a new chemoselective lithiation-annulation method, under high dilution, of
a suitable piperidinylcarbamoyl chloride. This precursor was obtained from (2-hydroxyethyl)-
piperidine in a linear synthetic sequence involving a Corey-Fuchs olefination of the corresponding
aldehyde, followed by a selective reduction, using a diimide equivalent, of an iodoalkyne into a
(Z)-iodopropene piperidine. This alkene served as main precursor to study the cyclization according
to several procedures using phosgene as the required carbonylating reagent.
In tr od u ction
Recent studies have shown that nicotinic acetylcholine
receptors (nAChRs) are involved in various physiological
effects associated with nicotine. Moreover, the nAChR
densities are altered in several neurodegenerative pa-
thologies such as Alzheimer and Parkinson’s diseases.1
As a result, the in vivo quantitation of receptors using
positron emission tomography (PET)2 has attracted tre-
mendous interest as a tool to investigate the role of the
nicotinic system in these pathologies.3 Several radio-
ligands have been prepared to visualize nAChRs using
PET (labeling with carbon-11, â+ , t1/2 ) 20 min or
fluorine-18, â+ , t1/2 ) 110 min)4 or using single photon
emission computed tomography (SPECT, labeling with
iodine 123, γ , t1/2 ) 13.1 h).5 However, due to their
toxicity or their high nonspecific binding, none of these
ligands is ideally suited for imaging studies of the human
brain.
F IGURE 1. (-)-cytisine 1 and model lactam 2.
(-)-Cytisine 1 (Figure 1) is a quinolizidine alkaloid
isolated from various plants of the Leguminosae family6
and, in particular, from seeds of Laburnum anagyroides.7
Owing to its nanomolar affinity8 and its high selectivity
toward the neuronal R4â2 receptor subtypes,9 (-)-cytisine
1 is often used as a reference ligand in the studies of the
nicotinic neurotransmission. Its long half-life in vivo,10
its ability to cross the blood-brain barrier,11 its low
nonspecific binding, and its slow clearance from the
* To whom correspondence should be addressed. Tel: 33 2 31 45 28
93. Fax: 33 2 31 45 28 77.
(5) (a) Saji, H.; Watanabe, A.; Magata, Y.; Ohmomo, Y.; Kiyono, Y.
Chem. Pharm. Bull. 1997, 45, 284-290. (b) Musachio, J . L.; Ville-
magne, V. L.; Scheffel, U.; Stathis, M.; Finley, P.; Horti, A.; London,
E. D.; Dannals, R. F. Synapse 1997, 26, 392-399. (c) Musachio, J . L.;
Scheffel, U.; Finley, P.; Zhan, Y.; Machizuki, T.; Wagner, H. N. J .;
London, E. D.; Dannals, R. F. Life Sci. 1998, 62, PL351-PL357.
(6) (a) Leonard, N. J . In The Alkaloids; Manske, R. H. F., Holmes,
H. L., Eds.; Academic Press: New York, 1953; Vol. 3, pp 119-199. (b)
El-Shazly, A.; Sarg, T.; Ateya, A.; Witte, E. A.; Wink, M. Pharmazie
1996, 51, 768-772.
(1) (a) Williams, M.; Sullivan, J . P.; Arneric, S. P. Drugs News
Perspect. 1994, 7, 205-223. (b) Perry, E. K.; Morris, C. M.; Court, J .
A.; Cheng, A.; Fairbairn, A. F.; McKeith, I. G.; Irving, D.; Brown, A.;
Perry, R. H. Neuroscience 1995, 64, 385-395.
(2) Fowler, J . S.; Wolf, A. P. Acc. Chem. Res. 1997, 30, 181-188.
(3) Mazie`re, M. Pharmacol. Ther. 1995, 66, 83-101.
(4) (a) Muzic, R. F.; Berridge, M. S.; Friedland, R. P.; Zhu, N.;
Nelson, A. D. J . Nucl. Med. 1998, 39, 2048-2054. (b) Sihver, W.; Fasth,
K. J .; O¨ gren, M.; Bivehed, H.; Bergstro¨m, M.; Nordberg, A.; Watanabe,
Y.; Langstro¨m, B. J . Neurochem. 1998, 71, 1750-1760. (c) Kassiou,
M.; Scheffel, U. A.; Ravert, H. T.; Mathews, W. B.; Musachio, J . L.;
London, E. D.; Dannals, R. F. Life Sci. 1998, 63, PL13-PL18. (d)
Valette, H.; Bottlaender, M.; Dolle, F.; Dolci, L.; Syrota, A.; Crouzel,
C. Nucl. Med. Commun. 1997, 18, 164-168. (e) Patt, J . T.; Spang, J .
E.; Westera, G.; Buck, A.; Schubiger, P. A. Nucl. Med. Biol. 1999, 26,
165-173. (f) Horti, A. G.; Scheffel, U.; Kimes, A. S.; Musachio, J . L.;
Ravert, H. T.; Mathews, W. B.; Zhan, Y.; Finley, P. A.; London, E. D.;
Dannals, R. F. J . Med. Chem. 1998, 41, 4199-4206. (g) Ding, Y. S.;
Molina, P. E.; Fowler, J . S.; Logan, J .; Volkow, N. D.; Kuhar, M. J .;
Carroll, F. I. Nucl. Med. Biol. 1999, 26, 139-148.
(7) Lecoq, H. Bull. Soc. Chim. Fr. 1943, 153-160.
(8) Anderson, D. J .; Arneric, S. P. Eur. J . Pharmacol. 1994, 253,
261-267.
(9) (a) Pabreza, L. A.; Dhawan, S.; Kellar, K. J . Mol. Pharmacol.
1991, 39, 9-12. (b) Glennon, R. A.; Dukat, M. Med. Chem. Res. 1996,
465-486. (c) Hall, M.; Zerbe, L.; Leonard, S.; Freedman, R. Brain Res.
1993, 600, 127-133.
(10) Sloan, J . W.; Martin, W. R.; Bostwick, M.; Hook, R.; Wala, E.
Pharmacol. Biochem. Behav. 1988, 30, 255-267.
(11) (a) Lippiello, P. M.; Caldwell, W. S. US 5,242,916, 1993; Chem.
Abstr. 1993, 119, 217429g. (b) Romano, C.; Goldstein, A.; J ewel, N. P.
Psychopharmacol. 1981, 74, 310-315.
10.1021/jo0498157 CCC: $27.50 © 2004 American Chemical Society
Published on Web 04/21/2004
J . Org. Chem. 2004, 69, 3787-3793
3787

Rouden et al.
brain12 in vivo compared to nicotine make cytisine a good
candidate for PET studies. Prior to our interest for this
molecule, only one analogue of cytisine, (N-11C-methyl)-
cytisine, was labeled with [11C]CH3I and studied in vivo.4d
However, substitution on the secondary amino group of
cytisine had a deleterious effect on the biological profile
of this analogue since it displayed high nonspecific
binding and a somewhat different biodistribution com-
pared to the one of cytisine in baboon brains.
SCHEME 1. Retr osyn th etic An a lysis for La belin g
1 (* ) 11C)
In our continuing interest for studying neurotransmis-
sion and in particular nAChRs, we undertook a project
for labeling cytisine or some pyridone-substituted ana-
logues for their use as radiotracers in PET studies. One
part was devoted to the preparation of 18F-labeled
fluorinated analogues of cytisine, and we recently pub-
lished some of this work.13 The other part aims at labeling
cytisine itself with carbon-11, and as pointed out several
years ago,12 this is a challenging task. The difficulty lies
in the design of a rapid synthesis of [11C]cytisine compat-
ible with the limited number of labeled precursors and
the structure of cytisine which lacks an easy-to-label
position. Moreover, due to the short half-life of carbon-
11, the introduction of the radioisotope must be rapid and
carried out at the latest stage of the synthesis using a
11C-labeled precursor directly available from the cyclotron
([11C]carbon dioxide, [11C]methane) or rapidly prepared
from those ([11C]iodomethane, [11C]phosgene, 11C-acids,
and derivatives).14 Considering these limitations, one
structural feature of cytisine which attracted our atten-
tion was the carbonyl moiety of the pyridone ring.
Therefore we focused our research on the use of phosgene
as a possible labeled precursor for the introduction of a
11C-carbonyl into cytisine.
SCHEME 2. En visa ged Retr osyn th etic An a lyses
for th e Cycliza tion (M ) Meta l, X ) Ha logen )
These approaches required either R,â-unsaturated acid
derivatives, carbon monoxide under pressure, or a pre-
formed amide or lactone. Recently, it was shown that the
carbamate protecting group of amines was reactive
enough toward a carbanion to undergo an intramolecular
cyclization.20 All these substrates require too lengthy
preparations to be usuable in short-lived isotope chem-
istry. To our knowledge, there are only two examples of
carbonyl-labeled lactam synthesis.21 However, they use
[11C]CO and a unique technology not available in our
radiochemical facilities. Development of a synthetic
strategy that could afford access to isotopically labelled
lactams and more generally to δ-lactams is therefore
highly desired.
In this context, our strategy for radiolabeling cytisine,
depicted in Scheme 1, was centered on the use of
phosgene, a precursor easily available in carbon-11. The
key step is the double condensation of a properly func-
tionalized precursor X with [11C]phosgene and cyclization
of anionic intermediate Y. The feasibility of this reaction
was evaluated by synthesizing the model compound 2
starting from precursor 3 and was envisaged according
to several procedures (Scheme 2). The first and the most
obvious route (route A) involves a dianionic structure,
This paper described our efforts toward the synthesis
of the lactam 2 (Figure 1), a model compound for cytisine
1. Different approaches based on the use of phosgene as
the carbonylating reagent for building the lactam ring
were studied. They involved the coupling of a (Z)-
halogenoalkene piperidine either directly with phosgene
or indirectly with a preformed carbamoyl chloride fol-
lowed by an annulation.
Resu lts a n d Discu ssion
There are several methods for the synthesis of δ-lac-
tams. Most of them include intramolecular aza annula-
tion of imines or enamines,15 catalyzed carbonylations,16
ring-closure metathesis,17 condensation of alkenamides
with aryl aldehydes,18 or aminolysis of a γ-lactone.19
(12) Flesher, J . E.; Scheffel, U.; London, E. D.; Frost, J . J . Life Sci.
1994, 54, 1883-1890.
(13) (a) Rouden, J .; Marrie`re, E.; Tadino, V.; Lasne, M.-C. Org. Lett.
2000, 2, 1121-1124. (b) Roger, G.; Lagnel, B.; Rouden, J .; Besret, L.;
Valette, H.; Demphel, S.; Gopisetti, J . M.; Coulon, C.; Ottaviani, M.;
Wrenn, L.; Letchworth, S.; Bohme, G. A.; Benavides, J .; Lasne, M.-C.;
Bottlaender, M.; Dolle´, F. Bioorg. Med. Chem. 2003, 11, 5333-5343.
(14) Ferrieri, R. A.; Antoni, G.; Kihlberg, T.; Långstro¨m, B. In
Handbook of Radiopharmaceuticals. Radiochemistry and Applications;
Welch, M. J ., Redvanly, C. S., Eds.; Wiley: Chichester, 2003; pp 229-
282, 141-194.
(17) (a) Park, S. H.; Kang, H. J .; Ko, S.; Park, S.; Chang, S.
Tetrahedron: Asymmetry 2001, 12, 2621-2624. (b) Bieraugel, H.;
J ansen, T. P.; Schoemaker, H. E.; Hiemstra, H.; van Maarseveen, J .
H. Org. Lett. 2002, 4, 2673-2674.
(18) Marson, C. M.; Grabowska, U.; Fallah, A.; Walsgrove, T.;
Eggleston, D. S.; Baures, P. W. J . Org. Chem.1994, 59, 291-296
(19) Nishimura, Y.; Adachi, H.; Satoh, T.; Shitara, E.; Nakamura,
H.; Kojima, F.; Takeuchi, T. J . Org. Chem. 2000, 65, 4871-4882.
(20) (a) Hernandez, N. M.; Sedano, M. J .; J acobs, H. K.; Gopalan,
A. S. Tetrahedron Lett. 2003, 44, 4035-4039. (b) Rys, V.; Couture, A.;
Deniau, E.; Grandclaudon, P. Eur. J . Org. Chem. 2003, 1231-1237.
(21) (a) Karimi, F.; Långstro¨m, B. Org. Biomol. Chem. 2003, 1, 541-
546. (b) Karimi, F.; Långstro¨m, B. Eur. J . Org. Chem. 2003, 2132-
2137.
(15) (a) Aranda, V. G.; Barluenga, J .; Gotor, V. Tetrahedron Lett.
1974, 977-978. (b) Paulvannan, K.; Chen, T. J . Org. Chem. 2000, 65,
6160-6166. (c) Agami, C.; Dechoux, L. Hebbe, S. Tetrahedron Tett.
2003, 44, 5311-5313.
(16) (a) Kang, S.-K; Kim, K.-J .; Yu, C.-M.; Hwang, J .-W.; Do, Y.-K.
Org. Lett. 2001, 3, 2851-2853. (b) Dong, C.; Alper, H. Tetrahedron:
Asymmetry 2004, 15, 35-40. (c) Knight, J . G.; Tchabanenko, K.
Tetrahedron 2002, 58, 6659-6664.
3788 J . Org. Chem., Vol. 69, No. 11, 2004

Studies toward Labeling Cytisine with [11C]Phosgene
SCHEME 3. Dir ected Lith ia tion -Allyla tion of
mide 11a with tributyltin hydride catalyzed by pal-
ladium(0) afforded the (Z)-bromoalkene 8a in 82% yield.31
The same reduction conditions applied to diiodide 11b
led, under optimized conditions, to a mixture of products
from which the expected (Z)-iodoalkene 8b could be
isolated with 39% yield. The moderate yield obtained in
this case was due to over-reduction: reduced terminal
alkene 13 was produced in 43% yield. Attempts to
improve selectivity and yield for our substrate by chang-
ing the ligand of palladium [binap, dppp, dppf, P(o-tolyl)3
instead of PPh3]32 resulted mainly in the recovery of
starting material 11b. It is worth noting that we obtained
the (Z)-iodoalkene 8b from (Z)-bromoalkene 8a via a
recently developed copper-mediated substitution under
Buchwald’s conditions.33 The reaction was completely
stereoselective and could find wide application in the
future. Treatment of dibromoalkene 11a with n-butyl-
lithium and quenching the reaction with iodine afforded
the terminal iodoalkyne 12 with an average overall yield
of 95%.34 The same alkyne 12 was obtained with a
moderate 40% yield in one step from aldehyde 10 using
iodoform, triphenylphosphine, and an excess of t-BuOK
according to a described procedure.30 Moreover, purifica-
tion of the product from phosphorus-based side products
was quite difficult, and these conditions cannot be
recommended here. Reduction of iodoalkyne 12 to (Z)-
iodoalkene 8b was performed in 73% yield with diimide
using o-nitrobenzenesulfonylhydrazide (o-NBSH) in com-
bination with Et3N.35 Finally, removal of the Boc group
led to the target precursors 3a and 3b. Under the most
efficient synthetic sequence, gram quantities of pip-
eridines 3a and 3b were obtained with an overall yield
of 61% and 53%, respectively.
Boc-p ip er id in ea
a
Reagents and conditions: (a) s-BuLi, THF, -78 °C, with or
without sparteine or TMEDA; (b) 7, -78 to +20 °C, 15 h.
generated from precursor 3, which reacts with phosgene.
The second approach (route B) uses the N-benzylated
precursor 4, which undergoes an halogen-metal ex-
change reaction followed by condensation-debenzylation
with phosgene. The final strategy tested (route C) starts
from carbamoyl chloride 5, prepared by reaction of
phosgene with precursor 3, and which reacts with the
anion, formed via an halogen-metal exchange reaction,
to cyclize into lactam 2.
Syn t h esis of P r ecu r sor 3. A convergent approach
was first tried in order to prepare the common precursor
3 (Scheme 3). It was based on Beak’s methodology to
functionalize N-Boc-piperidines via a directed lithiation
R to nitrogen and subsequent alkylation with an electro-
phile.22 All attempts to metalate N-Boc-piperidine 6 with
sec-butyllithium (with or without sparteine, TMEDA,
followed or not by a transmetalation step with CuCN‚
2LiCl23) then trapping the anion with (Z)-3-bromo-1-
iodopropene24 7 failed to give the 2-alkylated piperidine
8b, direct precursor of 3b (X ) I). Boc-protected piperi-
dine 6 was quantitatively recovered with small amounts
of alkene 7.
Then we turned to a linear strategy starting from 2-(2-
hydroxyethyl)piperidine 9 (Scheme 4). Protection of the
secondary amine with a Boc group and Swern oxidation
of the primary alcohol gave aldehyde 10 (80% overall
yield). Using o-iodoxybenzoic acid (IBX),25 the oxidation
step was performed with a higher yield (95%) and under
more convenient experimental conditions.26 Thus, the
crude aldehyde 10 was used without purification in the
further steps. Wittig olefination27 with bromomethyl-
triphenylphosphonium bromide afforded an inseparable
mixture 70/30 of Z/E bromo alkenes 8a . Therefore, to
obtain selectively the (Z)-halogenoalkenes 8, the following
synthetic sequences were carried out.
Cycliza tion Stu d ies Usin g P h osgen e. To have a
reference sample of lactam 2 we focused our research first
on using a palladium cross-coupling reaction described
for synthesizing amides from carbamoyl chlorides and tin
derivatives.36 However, the intramolecular version of this
cross-coupling reaction required some adjustments. At-
tempts to transform bromoalkene piperidine 8a into a
(Z)-tributyltin alkene derivative failed [t-BuLi, THF, -78
°C followed by Bu3SnCl or (Bu3Sn)2, Pd(PPh3)4, toluene,
reflux]. The bulkiness of the tributyltin group probably
prevented its introduction on a cis position of a double
bond. Only the reduced terminal alkene 13 was detected
in the crude mixture. Therefore, bromine-tin exchange
was tried using a trimethyltin group. The palladium-
mediated conditions ([(Me3Sn)2, Pd(PPh3)4], toluene, re-
Dibromo olefination28 of aldehyde 10 furnished the
dibromoalkene 11a 29 or the diiodoalkene 11b under
modified conditions.30 Selective reduction of the dibro-
1
flux) resulted in complete conversion as shown by the H
NMR of the crude product but purification on silica gel
(22) (a) Bailey, W. F.; Beak, P.; Kerrick, S. T.; Ma, S.; Wiberg, K. B.
J . Am. Chem. Soc. 2002, 124, 1889-1895. (b) Beak, P.; Lee, W. K. J .
Org. Chem. 1993, 58, 1109-1117 and references therein.
(23) (a)Dehmel, F.; Abarbri, M.; Knochel, P. Synlett 2000, 345-346.
(b) Knochel, P.; Singer, R. P. Chem. Rev. 1993, 93, 2117-2188.
(24) (a) Piers, E.; Renaud, J .; Rettig, S. J . Synthesis 1998, 590-
602. (b) Marek, L.; Alexakis, A.; Normant, J .-F. Tetrahedron Lett. 1991,
32, 532-5332.
(25) (a) Frigerio, M.; Santagostino, M.; Sputore, S.; Palmisano, G.
J . Org. Chem. 1995, 60, 7272-7276. (b) Frigerio, M.; Santagostino,
M.; Sputore, S. J . Org. Chem. 1999, 64, 4537-4538 and references
therein.
(26) More, J . D.; Finney, N. S. Org. Lett. 2002, 4, 3001-3003.
(27) Stork, G.; Zhao, K. Tetrahedron Lett. 1989, 30, 2173-2174.
(28) Ramirez, F.; Desai, N. B.; McKelvie, N. J . Am. Chem. Soc. 1962,
84, 1745-1747.
(29) Ikeda, M.; Kugo, Y.; Sato, T. J . Chem. Soc., Perkin Trans. 1
1996, 1819-1824.
(30) Michel, P.; Rassat, A. Tetrahedron Lett. 1999, 40, 8579-8581.
(31) Uenishi, J .; Kawahama, R.; Yonemitsu, O.; Tsuji, J . J . Org.
Chem. 1998, 63, 8965-8975.
(32) binap: 2,2′-bis(diphenylphosphino)-1,1′-binaphthyl; dppp: 1,3-
bis(diphenylphosphino)propane; dppf: 1,1′-bis(diphenylphosphino)-
ferrocene.
(33) Buchwald, S. L.; Klapars, A. J . Am. Chem. Soc. 2002, 124,
14844-14845.
(34) Corey, E. J .; Fuchs, P. L. Tetrahedron Lett. 1972, 36, 3769-
3772.
(35) (a) Myers, A. G.; Zheng, B.; Movassaghi, M, J . Org. Chem. 1997,
62, 7507. (b) Wang, Y.-G.; Kobayashi, Y. Org. Lett. 2002, 4, 4615-
4618. When using disodium azodicarboxylate in acetic acid as diimide
precursor, the reduction, which gave 8b in 89% yield, went to
completion in about 1 month by slowly adding 100 equiv of reagent.
Moreover, this result was difficult to reproduce.
(36) J ousseaume, B.; Kwon, H.; Verlhac, J .-B.; Denat, F.; Dubac, J .
Synlett 1993, 117-118.
J . Org. Chem, Vol. 69, No. 11, 2004 3789

Rouden et al.
SCHEME 4. Syn th esis of La cta m P r ecu r sor s 3a
a
Reagents and conditions: (i) (Boc)2O, Et3N, CH2Cl2, rt, 99%; (ii) IBX, EtOAc, 75 °C, 95%; (iii) 11a : CBr4, PPh3, Et3N, THF, -78 to
rt, 87%. 11b: CHI3, PPh3, t-BuOK, THF, rt, 68%; (iv) n-BuLi, THF, -78 to 0 °C then I2, 95-99%; (v) n-Bu3SnH, Pd(PPh3)4, toluene, rt,
8a : 82%, 8b: 39%; (vi) CHI3, PPh3, t-BuOK, THF, rt, 40%; (vii) o-NBSH, Et3N, i-PrOH-THF, rt, 73%; (viii) NaI, CuI, (CH3NHCH2-)2,
n-BuOH, 120 °C, 47%; (ix) TFA, CH2Cl2, rt, 3a : 92%, 3b: 94%.
SCHEME 5. Stille-Kelly Typ e Cycliza tion for th e
required reagent for the carbonylation step according to
our initial strategies depicted in Scheme 2. Three differ-
ent pathways were tested depending on the starting
compound and Scheme 6 describes the most efficient
conditions for each route leading to lactam 2. The most
simple procedure, route A, utilized bromo or iodopip-
eridines 3 as starting materials. It consisted in generat-
ing a dianionic intermediate A by simultaneous amine
deprotonation and halogen-metal exchange steps with
subsequent trapping by a carbonylating reagent. Double
metalation was carried out at - 78 °C in THF or Et2O
using t-BuLi (3.5 equiv), with or without transmetalation
(MgBr2, CuCN‚2LiCl), followed by the addition of a
carbonyl equivalent (phosgene, triphosgene, carbon di-
oxide, or dimethyl carbonate). Small amounts (less than
Syn th esis of La cta m 2a
a
Reagents and conditions: (i) BTC, pyridine, CH2Cl2, rt, 86%;
(ii) Pd(PPh3)4, (Me3Sn)2, toluene, 110 °C, 0.5 h, 39%.
proved to be difficult and did not allowed us to isolate
the expected trimethyltin alkene, probably because of the
known instability of the trimethyltin group on silica. To
avoid any intermediate purification, we performed the
cyclization using a procedure similar to the Stille-Kelly
conditions37 starting from carbamoyl chloride 5a (Scheme
5): the labile trimethyltin group was generated in situ
and coupled directly with the carbamoyl chloride group
previously synthesized from precursor 3a .
Under the best conditions tried, this procedure resulted
in a mixture of two unseparable isomers in a 3/1 1H NMR
ratio, the δ-lactam 2 and the R-methylene-γ-lactam 14 ,
respectively, in 39% total yield.38 The formation of
exomethylene lactam 14 could be explained by an in-
tramolecular Heck reaction between the carbamoyl chlo-
ride group and the terminal alkene generated during the
palladium-mediated halogen-exchange step.39 Hexabu-
tylditin was not a suitable reagent for this particular
cyclization since only starting material 5a was recovered.
This confirmed our previous observation on the difficulty
of exchanging a bromide by a tributyltin group on (Z)-
bromoalkene bromopiperidine 8a .
1
10%, from H NMR) of lactam 2 were detected only from
piperidine 3b when the transmetalation step using
MgBr2 was followed by trapping the anion by BTC. In
1
the other cases, the major compound detected on the H
NMR of the crude product was carbamoyl chloride 15
bearing a dehalogenated double bond. GC-MS analyses
of crude mixtures of these unsuccessful experiments
confirmed the formation of product 15 which was inde-
pendently synthesized from allyl piperidine 13. This
result suggests that the halogen-lithium exchange reac-
tion occurred but the resulting anion was probably too
unstable and was reprotonated before reacting with the
carbonylating reagent. A further attempt to exchange
iodine of piperidine 3b with an excess of i-PrMgBr was
unsuccessful.41 Only starting material was recovered.
We then turned our efforts toward route B which takes
advantage of a reaction recently developed in our labora-
tory for the synthesis of 11C-carbamoyl chlorides from
N-benzylamines.42 According to this procedure, it should
be possible to carry out a ring closure by cleaving the
benzyl group of piperidines 4a -c with the acid chloride
generated by the reaction of the vinyl anion B with
phosgene (Scheme 6). Piperidines 4, readily available
from 3, were treated by t-BuLi (followed or not by a
We next undertook the cyclization using phosgene or
triphosgene (BTC),40 its safer solid substitute, as the
(37) (a) Kelly, T. R.; Li, Q.; Bhushan, V. Tetrahedron Lett. 1990,
31, 161-164. (b) Grigg, R.; Teasdale, A.; Sridharan, V. Tetrahedron
Lett. 1991, 32, 3859-3862.
(38) 1H NMR identification. (a) For compound 2: Takatsu, N.;
Ohmiya, S.; Otomatsu, H. Chem. Pharm. Bull. 1987, 35, 891-894. (b)
For compound 14: Mori, M.; Washioka, Y.; Urayama, T.; Yoshiura,
K.; Chiba, K.; Ban, Y. J . Org. Chem. 1983, 48, 4058-4067.
(39) Henin, F.; Muzart, J .; Pete, J .-P. Tetrahedron Lett. 1986, 27,
6339-6340.
(41) For a review on halogen-magnesium exchange using Grignard
reagents, see: Knochel, P.; Dohle, W.; Gommermann, N.; Kneisel, F.
F.; Kopp, F.; Korn, T.; Sapountzis, I.; Vu, V. A. Angew. Chem., Int.
Ed. 2003, 42, 4302-4320.
(42) Lemoucheux, L.; Rouden, J .; Ibazizene, M.; Sobrio, F.; Lasne,
M.-C. J . Org. Chem. 2003, 68, 7289-7297.
(40) BTC, bistrichloromethyl carbonate.
3790 J . Org. Chem., Vol. 69, No. 11, 2004

Studies toward Labeling Cytisine with [11C]Phosgene
SCHEME 6. Ch em oselective Lith ia tion -An n u la tion P r ovid in g La cta m 2a
a
Reagents and conditions: (i) t-BuLi, THF, -78 °C, 2 h; (ii) BTC, -78 °C to rt, 15 h; (iii) aq Na2CO3, CH2Cl2, rt, 15 h, 50-73%; (iv)
t-BuLi, THF, -78 °C, 2 h then MgBr2, 0.5 h; (v) BTC, Pyr, CH2Cl2, rt, 15 h, 86-92%; (vi) M, THF, 2 h, see Table 1 for details; (vii)
RCuLiCN, -78 °C to rt, see Table 1 for details.
transmetalation step with CuCN‚2LiCl or Lipshutz’s
thienylcyanocuprate43) and phosgene or BTC were used
to trap the anion B as a putative acid chloride B′.
Starting from iodoalkenes piperidine 4b or 4c, lactam 2
was isolated in a disappointing 10% yield following the
conditions described in Scheme 6. As previously observed,
the main compound of the reaction was the carbamoyl
chloride 15, suggesting again a competitive protodeha-
logenation. With bromoalkene piperidine 4a the reactions
led to complex mixtures of products, some of these
containing an alkynyl group (characterized by 13C NMR
spectrum of the crude mixture) resulting probably from
an elimination instead of the halogen-metal exchange
reaction. This observation was confirmed by our last set
of experiments. As expected, in all cases the debenzyla-
tion occurred when we added phosgene to trap the vinylic
anion. Other conditions tried for the iodine-metal ex-
change included the use of Rieke’s copper,44 i-PrMgBr,
and (Neophy)2CuLi,45 however, without success.
electrophilic counterpart in such a reaction. Carbamoyl
chlorides 5 were conveniently synthesized from pip-
eridines 3 (Scheme 6). Several reagents and conditions
were tested to perform the halogen-metal exchange
(lithium, magnesium, copper) and then to promote the
cyclization onto the carbamoyl chloride moiety. The most
representative results are summarized in Table 1.
The first conditions tried (entries 1 and 2) using
magnesium metal to generate the anion at room tem-
perature from the vinyl halide moiety in the presence of
a carbamoyl chloride group provided lactam 2 in rather
encouraging yields. These remarkable results (mainly
entry 2) suggested that a Grignard reagent could be
generated at room temperature in the presence of a
carbamoyl chloride group. We anticipated that at low
temperature (-78 °C or below), t-BuLi would not react
with a carbamoyl chloride group under the conditions
used. Nevertheless, treatment of bromoalkene 5a with
t-BuLi at -120 °C did not produce any lactam 2 (entry
3). Instead, alkyne 16 was isolated as the main product
in 41% yield from a complex mixture. Because of a
difficult purification, this compound was only partially
characterized.49 Moreover, addition of 2 equiv of t-BuLi
on 5a at -120 °C, as recommended for lithiation of (Z)-
bromoalkenes,50 did not perform the bromine-lithium
exchange step. An elimination reaction leading to a
terminal alkyne followed by the addition of one molecule
Last, we tried route C, which involved a cyclization
similar to a Parham46 reaction. There are only a few
reports on the cyclization between a vinyllithium and a
reactive electrophile.47 However, to our knowledge, none
of them have used a carbamoyl chloride group as the
(43) Lipshutz, B. H.; Kozlowski, J . A.; Parker, D. A.; Nguyen, S. L.;
McCarthy, K. E. J . Organomet. Chem. 1985, 285, 437-447.
(44) Ebert, G. W.; Rieke, R. D. J . Org. Chem. 1988, 53, 4482-4488.
(45) Piazza, C.; Knochel, P. Angew. Chem., Int. Ed. 2002, 41, 3263-
3265.
(46) Parham, W. E.; Bradsher, C. K. Acc. Chem. Res. 1982, 15, 300-
305.
(48) Bertz, S. H.; Eriksson, M.; Miao, G.; Snyder, J . P. J . Am. Chem.
Soc. 1996, 118, 10906-10907.
(49) See the Experimental Section and the Supporting Information
for spectral data of alkyne 16.
(47) Clayden, J . In Organolithiums: Selectivity for Synthesis; Per-
gamon: Amsterdam, 2002; pp 284-285.
(50) Neumann, H.; Seebach, D. Tetrahedron Lett. 1976, 4839-4842.
J . Org. Chem, Vol. 69, No. 11, 2004 3791

Rouden et al.
TABLE 1. Ch em oselective Lith ia tion -An n u la tion
or by using mesityllithium as a chemoselective reagent
for a halogen-lithium exchange reaction.53 Finally, an
attempt to exchange iodine of piperidine 5b with an
excess of i-PrMgBr was unsuccessful and starting mate-
rial was recovered. Evidently from these results, the
countercation did not seem to have a major effect on the
cyclization whereas high dilution was the main param-
eter to achieve this new version of the Parham reaction.
This original annulation strategy, involving the selective
generation of a lithioalkene besides a carbamoyl chloride,
allowed for a direct access to R,â-unsaturated lactams.
Con d ition s P r ovid in g La cta m 2 (Rou te C)
sub- reagent tempa
yieldd
(%)
entry strate
(°C)
transmetalationb
[M]c
0.03
1
2
3
4
5
6
7
8
9
5a
5b
5a
5b
5b
5b
5b
5b
5b
5b
5b
Mg (30 to 40) n.a.
18
Mg (30 to 40) n.a.
t-BuLi (-120) n.a.
t-BuLi (-78) n.a.
0.03
0.02
0.05
0.05
0.025
40
0f
0g,k
0h,k
30i,k
t-BuLi (-78) 2-thienylCuLiCN
t-BuLi (-78) 2-thienylCuLiCN
t-BuLi (-78) TMSCH2CuLiCN48 0.0025 47i,k
t-BuLi (-78) 2-thienylCuLiCN
t-BuLi (-78) 2-thienylCuLiCN
0.0025 59i,k
0.0012 60i,k
10
11
t-BuLi (-90) 2-thienylCuLiCNe 0.0025 65j,k
Con clu sion
t-BuLi (-90) n.a.
0.005
50j,k
a
b
Halogen-exchange step in THF for 2 h. n.a.: not applicable;
On our way to address the challenge of radiolabeling
of (-)-cytisine 1 with short-lived isotopes, we have
developed a new lactam annulation on a model compound
(lactam 2) involving the chemoselective lithiation of a
bifunctional precursor, prepared according to a seven-
step linear synthesis. Of particular interest were the
efficient conversion of an iodoalkynyl piperidine selec-
tively to a (Z)-iodoalkene using o-NBSH as a diimide
equivalent and the bromo-to-iodoalkene copper-mediated
substitution. The strategy we have developed uses phos-
gene (or an equivalent) as the reagent to introduce the
carbonyl part of the lactam. The new annulation method
disclosed here is the first example of a successful halogen-
lithium (magnesium) exchange in the presence of a
carbamoyl chloride, followed by its intramolecular con-
densation. High dilution of the carbamoyl chloride was
the key factor to achieve the cyclization, conditions
required in chemistry using carbon isotopes. This proce-
dure represents a valuable alternative to the carbon
monoxide palladium-catalyzed chemistry, especially for
six-membered ring lactams. Moreover, it provides a novel
access to the preparation of functionalized organometallic
compounds, nucleophilic intermediates which are of
increasing interest for modern organic synthesis. This
methodology will be applied to the labeling of (-)-cytisine
1 with [11C]phosgene and elucidation of the mechanism
leading to the formation of lactam 17 is currently
underway.
when applicable, the transmetalation was carried out at -78 °C
and the temperature reached rt over 15 h. c Initial concentration
of starting material. Isolated yield. e The reaction temperature
d
reached rt in 2 h after transmetalation. f Alkyne 16 was isolated
as the major product with a 41% yield. g Compound 17 was isolated
h
as the major product with a 48% yield. Compound 15 was
1
detected by H NMR and GC-MS analyses as the major product.
With 15-20% of product 17. j With 5-10% of product 17. Com-
i
k
pound 18 was detected by GC-MS analysis of the crude mixture
(less than 10%).
of t-BuLi on the carbamoyl chloride moiety generated the
amide group. When starting from iodopiperidine 5b, a
peculiar lactam 17 was isolated as the major product with
a 48% yield, but no trace of compound 2 (entry 4). We
are currently investigating the mechanism of this inter-
esting anionic carbocyclization leading to 17.51 Lithiation
of 5b followed by transmetalation with copper (entry 5)
still did not afford the expected lactam 2 but mainly
1
alkene 15 identified by H NMR and GC-MS analyses
of the crude mixture. As previously discussed, this
showed that the iodine-lithium exchange step occurred,
but for unknown reasons, the resulting anion was either
unreactive or rapidly reprotonated. However, when per-
forming the same experiment after diluting twice more,
lactam 2 was isolated with a significant 30% yield along
with 15-20% of lactam 17, unseparable from lactam 2
(entry 6). Therefore, while increasing the dilution, the
yield of the cyclization reached 60% (entry 7-9), lactam
2 was still contaminated with 15-20% of side product
17. To decrease the amount of 17, the iodine-lithium
exchange was carried out at -90 °C followed by a
transmetalation at -78 °C with Lipshutz cyanocuprate.
Thus, the highest yield (65%) with 5-10% of lactam 17,
was obtained under high dilution (2.5 mM) and rapid
warming (2 h) of the reaction temperature after the
transmetalation (entry 10). Interestingly, the expected
cyclization leading to lactam 2 occurred well with a
lithiated anion, however under high dilution (entry 11).
In all reactions starting with iodopiperidine 5b, we
detected by GC-MS analysis of the crude mixtures small
amounts (less than 10%) of a pivaloylpiperidine 18
(characterized by its MS spectrum), resulting from the
addition of t-BuLi on the carbamoyl chloride group. This
side reaction generating amide 18 could certainly be
avoided by lowering the temperature (-100 °C or below)52
Exp er im en ta l Section
1,6,7,8,9,9a -Hexa h yd r o-4H-qu in olizin -4-on e (2).38 Gen-
eral procedure for entries 1 and 2, Table 1: To a magnesium
(0.20 g, 8.3 mmol, 25 equiv) suspension in dry THF (3 mL)
under nitrogen was added an iodine crystal followed by 1,2-
dibromoethane (0.18 g, 0.9 mmol, 3 equiv) at rt. When the
temperature increased, a solution of 1,2-dibromoethane (0.12
g, 0.6 mmol, 2 equiv) and carbamoyl chloride 5b (0.10 g, 0.3
mmol, 1 equiv) in THF (5 mL) was slowly added during 10
min. After 1 h at rt, the reaction mixture was hydrolyzed with
a saturated NH4Cl solution and extracted with CH2Cl2. The
organic layer was dried over MgSO4 and concentrated. Puri-
fication by flash chromatography on silica gel (AcOEt/Pentane
5/1) afforded 20 mg of lactam 2 as a colorless oil (slightly
contaminated by impurities).
General procedure for entries 5 to 10, Table 1: To a cold
(-78 or -90 °C, see Table 1) solution of carbamoyl chloride
5b (0.156 g, 0.5 mmol, 1 equiv) in THF (see Table 1 for
concentration) under nitrogen was slowly added t-BuLi (1.7
M in pentane, 0.65 mL, 1.1 mmol, 2.2 equiv). The mixture was
(51) At the moment we do not believe that compound 17 could result
from a 1,4 addition of remaining t-BuLi on lactam 2.
(52) For an example of the effect of the temperature on a chemo-
selective lithiation, see: Parham, W. E.; J ones, L. D. J . Org. Chem.
1976, 41, 1187-1191.
(53) Kondo, Y.; Asai, M.; Miura, T.; Uchiyama, M.; Sakamoto, T.
Org. Lett. 2001, 3, 13-15.
3792 J . Org. Chem., Vol. 69, No. 11, 2004

Studies toward Labeling Cytisine with [11C]Phosgene
stirred for 2 h. When applicable, a lithium 2-thienylcyano-
cuprate solution (0.25 M in THF, 2 mL, 0.5 mmol, 1 equiv)
was added at -78 °C, and the temperature was allowed to
reach rt (overnight or in 2 h). The reaction mixture was
hydrolyzed with a saturated NH4Cl solution and extracted
with CH2Cl2. The combined organic layers were dried over
MgSO4 and concentrated to give lactam 2 after purification
by chromatography, unseparable from small amounts of amide
17. 1H NMR: δ ) 1.25-1.80 (m, 6H), 2.05-2.20 (m, 1H), 2.35-
2.50 (m, 2H), 3.25-3.50 (m, 1H), 4.42 (dd, J ) 11.6, 1.5 Hz,
1H), 5.90 (ddd, J ) 9.8, 2.3, 1.5 Hz, 1H), 6.48 (ddd, J ) 9.8,
5.0, 3.4 Hz, 1H). 13C NMR: δ ) 24.0, 24.8, 31.1, 33.4, 43.0,
54.8, 124.7, 138.0, 165.6. IR (NaCl): 3456, 2936, 2860, 1668,
1636, 1618, 1468, 1442, 1274, 824 cm-1. MS (EI) m/z: 151 (M,
94), 136 (12), 122 (33), 108 (10), 84 (100). HRMS (EI): calcd
for C9H13NO 151.0997, found 151.1002.
1-[2-(2-P r op en yl)p ip er id in -1-yl]-2,2-d im et h yl-1-p r o-
p a n on e (16). To a cold (-120 °C) solution of carbamoyl
chloride 5a (0.270 g, 1.01 mmol, 1 equiv) in a 2/2/1 mixture of
THF/Et2O/pentane (50 mL) under nitrogen was slowly added
t-BuLi (1.7 M in pentane, 1.3 mL, 2.23 mmol, 2.2 equiv). The
mixture was stirred for 2 h at -115 °C or below. Then the
temperature was allowed to reach rt overnight, and the
reaction mixture was hydrolyzed with a saturated NH4Cl
solution and extracted with CH2Cl2. Purification by flash
chromatography on silica gel (AcOEt/pentane 3/2) afforded
0.085 g of amide 16 (41% yield) as a colorless oil (slightly
contaminated by impurities). 1H NMR: δ ) 1.28 (s, 9H), 1.30-
2.20 (m, 5H), 1.80-2.00 (m, 2H), 2.30-2.70 (m, 2H), 2.85 (br
s, 1H), 4.18 (br d, J ) 12.8 Hz 1H), 4.78 (br s, 1H). 13C NMR:
δ ) 18.7, 19.5, 25.5, 26.6, 28.4, 38.8, 49.6, 52.4, 70.1, 81.0,
176.5; MS (CI) m/z: 208 (M + 1, 100), 168 (12). GC analysis:
retention time, 10.3 min (Tinj, 230 °C, Col., 50 °C for 1 min,
then +15 °C/min until 250 °C).
2-(Dim eth yleth yl)octa h yd r o-4H-qu in olizin -4-on e (17).
To a cold (-78 °C) solution of carbamoyl chloride 5b (0.156 g,
0.5 mmol, 1 equiv) in THF (10 mL) under nitrogen was slowly
added t-BuLi (1.7 M in pentane, 0.65 mL, 1.1 mmol, 2.2 equiv).
The mixture was stirred for 2 h, and then the temperature
was allowed to reach rt in 2 h. The reaction mixture was
hydrolyzed with a saturated NH4Cl solution and extracted
with CH2Cl2. The combined organic layers were dried over
MgSO4 and concentrated to give, after purification by flash
chromatography (heptane/EtOAc: 1/1) on silica gel, 0.050 g
of lactam 17 (48% yield). 1H NMR: δ ) 0.87 (s, 9H), 1.25-
1.67 (m, 8H), 1.75-2.20 (m, 2H), 2.40-2.49 (m, 2H), 3.38-
3.41 (m, 1H), 4.73 (dq, J ) 12.8, 2.0 Hz, 1H). 13C NMR: δ )
25.4, 25.5, 26.9, 30.1, 31.7, 32.8, 34.8, 39.0, 43.7, 56.6, 169.2.
IR (NaCl): 3018, 2964, 2940, 2870, 1618, 1474, 1216 cm-1. MS
(EI) m/z: 209 (M, 100), 195 (86), 152 (68), 138 (10), 124 (18),
97 (47), 84 (64).
Ack n ow led gm en t. We gratefully acknowledge the
“Ministe`re de la Recherche et des Nouvelles Technolo-
gies” for fellowships to L.L. and T.S, “PunchOrga”
Network (Poˆle Universitaire Normand de Chimie Or-
ganique), CNRS (Centre National de la Recherche
Scientifique), the “Re´gion Basse-Normandie”, and the
European Union (FEDER funding) for financial support.
Su p p or tin g In for m a tion Ava ila ble: General experimen-
tal conditions; preparation, characterization, 1H NMR, and 13C
NMR spectra for compounds 3a ,b, 4a -c, 5a ,b, 8a ,b, 11a ,b,
12, 13, and 15; 1H NMR and 13C NMR spectra for compounds
2, 16, and 17. This material is available free of charge via the
Internet at http://pubs.acs.org.
J O0498157
J . Org. Chem, Vol. 69, No. 11, 2004 3793