Gold(I)-catalyzed rearrangement of N-aryl 2-alkynylazetidines to pyrrolo[1,2-a]indoles

Article abstract of DOI:10.1021/ol303518n

Various N-aryl-2-alkynylazetidines were very efficiently converted to pyrrolo[1,2-a]indoles with gold catalysts, especially the 2-biphenyl- dicyclohexylphosphino-gold(I) hexafluoroantimonate, in dichloromethane at room temperature. Additionally, two forma

Full text of DOI:10.1021/ol303518n

ORGANIC
LETTERS
2013
Vol. 15, No. 4
836–839
Gold(I)-Catalyzed Rearrangement
of N‑Aryl 2‑Alkynylazetidines to
Pyrrolo[1,2‑a]indoles
ꢀ
Nicolas Kern, Marie Hoffmann, Aurelien Blanc,* Jean-Marc Weibel, and Patrick Pale*
ꢁ
ꢀ
ꢀ
ꢀ
ꢀ
Laboratoire de Synthese, Reactivite Organiques & Catalyse, associe au CNRS, Institut
de Chimie, Universite de Strasbourg, 4 rue Blaise Pascal, 67070 Strasbourg, France
ablanc@unistra.fr; ppale@unistra.fr
Received December 22, 2012
ABSTRACT
Various N-aryl-2-alkynylazetidines were very efficiently converted to pyrrolo[1,2-a]indoles with gold catalysts, especially the 2-biphenyl-dicyclo-
hexylphosphino-gold(I) hexafluoroantimonate, in dichloromethane at room temperature. Additionally, two formal syntheses of bioactive non-
natural compounds, i.e. 7-methoxymitosene and an 5-HT2C receptor agonist, have been achieved.
The now well-known activation of an alkene or alkyne
by π-coordination with gold salts or complexes1 and our
involvement in gold multifaceted catalysis,2 i.e. cascade
reactions implying both π and σ Lewis acidities of gold,3
led us to imagine a rapid access to pyrrolo[1,2-a]indoles 2,
based on the opening of N-aryl 2-alkynylazetidines 1
(Scheme 1). The process is based on π and σ Lewis acidities
of gold, which could favor alkynylazetidine opening and
allene formation (Scheme 1) (A and/or A0 upon π or σ Au
activation). The benzazocine intermediate B would then be
cyclized upon another gold activation, leading after pro-
todeauration of C to the pyrrolo[1,2-a]indole.
among which mitomycins and its derivatives,4 known as
mitosanes and mitosenes,5 are the most well-known due to
their antitumor activities (Figure 1).6 Yuremamine, iso-
lated from the stem bark of Mimosa hostilis, is another
recent interesting example, due to its traditional use pro-
moting hallucinogenic and psychoactive effects.7 Another
psychoactive alkaloid harmalidine and its derivatives have
been isolated from the seeds of Pegalum harmala.8 The
pyrrolo[1,2-a]indole motif can also be found in some non-
natural compounds, exhibiting useful biological activities
(Figure 1).9,10 Due to their interesting properties, these
compounds have given rise to various synthetic ap-
proaches.11,12 Most of them relied on multistep sequences,
The pyrrolo[1,2-a]indole tricyclic structure is a common
motif found in a number of naturally occurring products,
(5) (a) Webb, S.; Cosulich, D. B.; Mowat, J. H.; Patrick, J. B.;
Broschard, R. W.; Meyer, W. E.; Williams, R. P.; Wolf, C. F.; Fulmor,
W.; Pidackas, C.; Lancaster, J. E. J. Am. Chem. Soc. 1962, 84, 3185. (b)
Webb, S.; Cosulich, D. B.; Mowat, J. H.; Patrick, J. B.; Lancaster, J. E.;
Broschard, R. W.; Meyer, W. E.; Williams, R. P.; Wolf, C. F.; Fulmor,
W.; Pidackas, C. J. Am. Chem. Soc. 1962, 84, 3187.
(6) (a) Bradner, W. T. Cancer Treat. Rev. 2001, 27, 35. (b) For the
initially proposed but now proven mode of action, see: Iyer, V. N.;
Szybalski, W. Science 1964, 145, 55.
(1) For selected recent reviews on gold chemistry, see: (a) Lu, B.-L.;
Dai, L.; Shi, M. Chem. Soc. Rev. 2012, 41, 3318. (b) Corma, A.; Leyva-
Perez, A.; Sabater, M. J. Chem. Rev. 2011, 1657. (c) Krause, N.; Winter,
C. Chem. Rev. 2011, 1994. (d) Gagosz, F. Actual. Chim. 2010, 347, 12. (e)
Belmont, P.; Parker, E. Eur. J. Org. Chem. 2009, 6075. (f) Hashmi,
A. S. K.; Hutchings, G. J. Angew. Chem., Int. Ed. 2006, 45, 7896.
(2) (a) Kern, N.; Blanc, A.; Miaskiewicz, S.; Robinette, M.; Weibel,
J.-M.; Pale, P. J. Org. Chem. 2012, 77, 4323. (b) Kern, N.; Blanc, A.;
Weibel, J.-M.; Pale, P. Chem. Commun. 2011, 47, 6665. (c) Britton, J.;
Camp, J. E. Chemistry Today 2012, 30 (3, Suppl.), 6.
ꢀ
(7) Vepsalainen, J. J.; Auriola, S.; Tukiainen, M.; Ropponen, N.;
Callaway, J. Planta Med. 2005, 71, 1053.
(8) Siddiqui, S.; Khan, O. Y.; Siddiqui, B. S.; Faizi, S. Phytochem.
1987, 26, 1548.
(9) Peters, R.; Waldmeier, P.; Joncour, A. Org. Process Res. Dev.
(3) Yamamoto, Y. J. Org. Chem. 2007, 72, 7817.
(4) (a) Kudo, S.; Marumo, T.; Tomioka, T.; Kato, H.; Fujimoto, Y.
Antibiot. Chemother. 1958, 8, 228. (b) Lefemine, D. V.; Dann, M.;
Barbatschi, F.; Hausmann, W. K.; Zbinovsky, V.; Monnikendam, P.;
Adam, J.; Bohonos, N. J. Am. Chem. Soc. 1962, 84, 3184. (c) Tulinsky,
A. J. Am. Chem. Soc. 1962, 84, 3188.
2005, 9, 508.
(10) Allen, G. R.; Poletto, J. F.; Weiss, M. J. J. Am. Chem. Soc. 1964,
86, 3877.
r
10.1021/ol303518n
Published on Web 01/25/2013
2013 American Chemical Society

especially in the mitomycin area,13 and any additional con-
vergent and rapid route would be beneficial to this area.
Table 1. Screening of Reaction Conditions for the Au-Catalyzed
Transformation of N-Phenyl 2-Alkynylazetidine 1a into 2a
Scheme 1. Hypothesis for Au-Catalyzed Conversion of N-Aryl
Alkynylazetidines to Pyrrolo[1,2-a]indoles
time
(h)
yield
entry
catalyst (mol %)
AuCl (4)
conditionsa
2a (%)
1
CH2Cl2, rt
CH2Cl2, rt
CH2Cl2, rt
(CH2Cl)2, rt
THF, rt
18
18
45b
31b
80
2
AuCl3 (4)
3
PPh3AuNTf2 (4)
00
00
00
00
00
1
4
0.5
0.33
18
78
5
86
6
toluene, rt
acetone, rt
CH3CN, rt
CH2Cl2, rt
CH2Cl2, rt
CH2Cl2, rt
CH2Cl2, rt
CH2Cl2, rt
37b
65b
55b
92
7
18
8
18
9
L1AuNTf2 (4)
0.5
0.33
2
10
11
12
13
L1AuSbF6.MeCN (4)
L2AuSbF6.MeCN (4)
L1AuSbF6.MeCN (2.5)
AgSbF6 (4)
98
98
0.5
18
98
<10c
a C = 0.2 mol/L. b Conversion, the starting material accounting
for the mass balance. c Degradation occurred leading to unidentified
byproducts.
The Gagosz catalyst15 rapidly gave 2a in high yield in
chlorinated solvents (Table 1, entries 3ꢀ4) or in THF
(entry 5), but slow reactions and low conversions were
observed in less polar solvents (entry 6) as well as in more
polarandcoordinatingsolvents(entries7ꢀ8). Switching to
the bulky and stabilizing (Cy2)JohnPhos ligand L1 induced
a fast and very efficient reaction (entry 9). Combining this
ligand or the bulkiest JohnPhos ligand L2 and a more
electrophilic counterion, such as SbF6ꢀ, gave almost quan-
titatively the pyrrolo[1,2-a]indole 2a (entries 10ꢀ11) and
allowed the catalytic charge to be decreased to only 2.5%
(entry 12).16 Control experiments showed that the silver
salt used to prepare the cationic gold species was not
responsible for the rearrangement, but led to degradation
(entry 13).
With these results in hand, the scope of this gold(I)-
catalyzed rearrangement of N-aryl 2-alkynylazetidines to
pyrrolo[1,2-a]indoles was then examined. Various N-aryl
2-alkynylazetidines, 1aꢀj, were thus prepared and sub-
mitted tothe optimalconditionsdescribedabove(Table2).
Steric and electronic effects at the acetylene moiety on
one hand and electronic effects at the N-aryl group on the
other hand should influence the first opening-cyclization
step and would thus shed some light on the actual mechan-
ism (see Scheme 1). Their role was thus examined with
substrates 1aꢀc and 1dꢀg. A branched alkyl chain next
to the acetylene moiety did not significantly change the
Figure 1. Natural products and some bioactive non-natural
compounds exhibiting the pyrrolo[1,2-a]indole motif.
In the search for optimal conditions, the simple 2-(hept-
1-ynyl)-2-methyl-1-phenylazetidine 1a was readily pre-
pared in three simple steps with a good overall yield (see
Supporting Information) and submitted to various Au
catalysts (Table 1).14 Simple gold(I) or gold(III) chlorides
gave the expected compound 2a in dichloromethane in a
clean reaction but with low conversions despite long reac-
tion times (Table 1, entries 1ꢀ2). The more soluble and
more electrophilic cationic phosphinogold(I) complexes
showed a higher reactivity but a high dependence on
solvent nature (entries 3ꢀ8).
(11) For selected references on the synthesis of pyrrolo[1,2-a]indoles,
see: (a) Chen, K.; Zhang, Z.; Shi, M. Chem. Commun. 2012, 48, 7696. (b)
Zeldin, R. M.; Toste, F. D. Chem. Sci. 2011, 2, 1706. (c) Patil, D. V.;
Cavitt, M. A.; France, S. Org. Lett. 2011, 13, 5820.
(12) For a leading reference on yuremamine synthetic studies, see:
Johansen, M. B.; Kerr, M. A. Org. Lett. 2008, 10, 3497.
(13) For a review on mitomycin synthesis, see: (a) Andrez, J.-C.
Beilst. J. Org. Chem. 2009, 5, n°33.
(14) For gold-catalyzed formation of pyrrolo[1,2-a]indoles, see: (a)
Kusama, H.; Miyashita, Y.; Takaya, J.; Iwasawa, N. Org. Lett. 2006, 8,
289. (b) Yan, Z.-Y.; Xiao, Y.; Zhang, L. Angew. Chem., Int. Ed. 2012, 51,
8624.
(15) Mezailles, N.; Ricard, L.; Gagosz, F. Org. Lett. 2005, 7, 4133.
(16) A further decrease in catalyst amount led to slower reactions
(e.g., with 1%, the reaction required almost 10 days to give a 64%
conversion, but still in a spot-to-spot transformation).
Org. Lett., Vol. 15, No. 4, 2013
837

group at the meta position relative to the azetidine ni-
trogen, mixtures of rearranged products were obtained
(entries 6ꢀ7). The methoxylated derivative reacted more
rapidly than its chlorinated analog, but a better yield was
obtained with the latter, suggesting contradictory electronic
influences. However, regioisomeric ratios were different,
from 1:1 for a methoxy group to 1:4 for the chloro sub-
stituent (entry 6 vs 7). In the latter, the major product was
the 3-chloro-9-pentylpyrrolo[1,2-a]indole 2g0, the less hin-
dered isomer, suggesting that steric interactions influenced
the reaction. To conclude on these aspects, the meta-iso-
propyl derivative 1h was rearranged under standard condi-
tions (entry 8). A single regioisomer was observed and
isolated in high yield, characterized as the 3-isopropyl-9-
pentylpyrrolo[1,2-a]indole 2h, again the less hindered iso-
mer. These results thus confirmed the major role of steric
interactions in this rearrangement, probably during the first
arylation step (Scheme 1) in which the meta-aryl substituent
would experience 1,3-allylic strain with the acetylenic
substituent. We were able to ensure that β-substitution
of the azetidine did not affect the reaction. As expected,
the rearrangement of cis-1iꢀj very efficiently afforded
the corresponding tricyclic indoles in high yields, although
this reaction has to be performed at 0 °C to avoid OBn
elimination (entries 9ꢀ10).
Table 2. Scope of the Gold(I)-Catalyzed Formation of
Pyrrolo[1,2-a]indoles from N-Aryl 2-Alkynylazetidines
To go further, and with total synthesis perspectives,17 we
applied this Au-catalyzed rearrangement toward a few
pyrrolo[1,2-a]indoles, either known intermediates in the
synthesis of bioactive compounds or models of natural
products (Table 3). We selected 7-ethoxy-2,3-dihyropyrrolo-
[1,2-a]indole,9 an intermediate en route to an agonist of the
5-HT2c receptor (Figure 1), linked to the regulation of
food intake and anxiety.18 This compound could easily be
obtained through the Au-catalyzed rearrangement of the
simple N-(4-ethoxyphenyl)-2-ethynylazetidine 1k in 85%
yield (Table 3, entry 1). Yet, the psychoactive alkaloid
harmalidine19 seemed to have never been synthesized
(Figure 1). The present Au-catalyzed rearrangement would
offer the first approach to such a compound and analogs,
starting from either N-(aryl) 2-(4-hydroxybut-1-ynyl)-3,3-
dimethylazetidine 1l or 1m. The hindered tert-butyldi-
methylsilyl-protected derivative 1l gave 2l and 2l0 in good
yield but as a 1:1 mixture (entry 2). To obtain the required
regioselectivity toward the harmalidine core, we introduced
a bromo substituent (1m), blocking one ortho position. The
latter indeed furnished the expected regioisomer in good
yield (entry 3). Looking for the mitomycins skeleton for-
mation, we then engaged the free and TIPS-protected
propargyl alcohols 1n and 1o in the gold-catalyzed rearrange-
ment. We rapidly observed the formation of the same
compound in both reactions (entries 4ꢀ5), which was
identified as the 3,30-methylbispyrrolo[1,2-a]indole 3 com-
ing from the condensation of the desired pyrrolo[1,2-a]-
indole motif on itself followed by a loss of formaldehyde.20
a Reaction run at 40 °C to complete the conversion. b Reaction run at
0 °C to avoid degradation.
reaction rate compared to the simple linear alkyl group
(Table 2, entry 2 vs 1). With a phenyl group conjugated
with the acetylene moiety, the reaction proceeded as well,
although the isolated yield was lower due to product
instability (entry 3 vs 1). With an electron-donating group
at the para position relative to the azetidine nitrogen, the
rearrangement proceeded as rapidly but very efficiently
(entry 4) while an electron-withdrawing group slightly
slowed the reaction leading to a lower isolated yield
(entry 5). With an electron-donating or -withdrawing
(17) Rudolph, M.; Hashmi, A. S. K. Chem. Soc. Rev. 2012, 41, 2448.
(18) Tecot, L. H.; Sun, L. M.; Akana, S. F.; Strack, A. M.;
Lowenstein, D. H.; Dallman, M. F.; Julluis, D. Nature 1995, 374, 542.
(19) Siddiqui, S.; Khan, O. Y.; Siddiqui, B. S.; Faizi, S. Phytochem.
1987, 26, 1548.
838
Org. Lett., Vol. 15, No. 4, 2013

Rewardingly, the benzyl protected analog 1p gave the ex-
pected carbon core 2p of the mitomycin family, stable
enough to be isolated in good yield (entry 6). Finally, the
simple mitosene derivative 2q was efficiently produced from
2-ethynyl-1-(2,4,5-trimethoxy-3-methylphenyl)-azetidine
1q, although forcing conditions had to be used (entry 7).
This result achieved the formal synthesis of the bioactive
7-methoxymitosene21 (Figure 1).
Scheme 2. Deuterium Distribution in Product 2r from Labeled
Compound 1r
the expected position of the gold atom after rearrangement
(C, Scheme 1). This result corroborates the second part of
the proposed mechanism, although the allyl gold part in C
seems to not be protodemetalated through an SE0 mechan-
ism as recently reported.23
Table 3. Gold(I)-Catalyzed Formation of Pyrrolo[1,2-a]indoles
of Biological Interesta
However, the first step is still unclear. On one hand,
cyclizations of propargyl anilines initiated by π-coordina-
tion have been reported.24 Such a pathway would lead to
A0 in our case. On the other hand, the alkynyl azetidine
opening can also be viewed as a gold-catalyzed amino-
Claisen rearrangement, according to a few precedents in
the literature.25 The strain release of the four-membered
ring would thus be the driving force of this process, and
azaphilic gold interaction would initiate such a pathway
leading to A, in analogy to very recent calculations per-
formedfor gold(I)-catalyzed Claisen rearrangement.26 The
fact that β-lactams, precursors of the azetidine derivatives
(see Supporting Information), failed to give pyrrolo[1,2-a]-
indolones supports the azaphilic activation pathway (A in
Scheme 1), as in this case the carbophilic mechanism
should havebeen favored by theweaklydonating nitrogen.
In conclusion, we have developed a novel and efficient
Au-catalyzedrearrangement ofN-aryl-2-alkynylazetidines
to pyrrolo[1,2-a]indoles. Further investigations of this
rearrangement and applications in alkaloid synthesis are
ongoing in our laboratory. In the latter aspect, two formal
syntheses (5-HT2C receptor agonist and 7-methoxymitosene)
have been achieved.
a Performed with (Cy2)JohnPhosAuSbF6 MeCN (4 mol %) at rt in
3
Acknowledgment. We gratefully acknowledge the CNRS
and the Agence Nationale de la Recherche (Grant ANR-11-
JS07-001-01). N.K. and M.H. thank the French Ministry of
Research for a PhD fellowship.
CH2Cl2. b Performed with JohnPhosAuSbF6 MeCN (10 mol %) at
3
80 °C. c R = H. d R = TIPS (triisopropylsilyl). e R = Bn (benzyl).
f Isolated yields after 72 h of stirring using 10 mol % of catalyst added in
four portions.
Supporting Information Available. Complete experi-
mental procedures, characterization data, and spectral
data. This material is available free of charge via the
Internet at http://pubs.acs.org.
To gain insight into the proposed mechanism,22 we
prepared the perdeuterated compound 1r and monitored
the deuterium labeling into the rearranged product 2r
(Scheme 2). In the latter, a single deuterium label was
found at the allylic position within the pyrrolidine ring, at
(25) (a) Sherry, B. D.; Toste, F. D. J. Am. Chem. Soc. 2004, 126,
15978. (b) Tejedor, D.; Mendez-Abt, G.; Cotos, L.; Garcia-Tellado, F.
Chem. Soc. Rev. 2013, 42, 458. For a gold-catalyzed tandem Claisen-
cyclization reaction, see: (c) Sherry, B. D.; Maus, L.; Lafortza, B. N.;
Toste, F. D. J. Am. Chem. Soc. 2006, 128, 8132. (d) Saito, A.; Konishi,
T.; Hanzawa, Y. Org. Lett. 2010, 12, 372.
(20) Pawlak, J. M.; Khau, V. V.; Hutchison, D. R.; Martinelli, M. J.
J. Org. Chem. 1996, 61, 9055.
(21) Liang, L.; Wang, Y.; Zhang, L. Org. Lett. 2012, 14, 3736.
(22) Hashmi, A. S. K. Angew. Chem., Int. Ed. 2010, 49, 5232.
(23) Hasmi, A. S. K.; Schuster, A. M.; Litters, S.; Rominger, F.;
Pernpointner, M. Chem.;Eur. J. 2011, 17, 5661.
(26) Vidhani, D. V.; Cran, J. W.; Krafft, M. E.; Manoharan, M.;
Alabugin, I. V. J. Org. Chem. 2013dx.doi.org/10.1021/jo302152j.
(24) (a) Menon, R. S.; Findlay, A. D.; Bissember, A. C.; Banwell,
M. G. J. Org. Chem. 2009, 74, 8901. (b) Nevado, C.; Echavarren, A.
Chem.;Eur. J. 2005, 11, 3155.
The authors declare no competing financial interest.
Org. Lett., Vol. 15, No. 4, 2013
839