Spontaneous biomimetic formation of (±)-dictazole B under irradiation with artificial sunlight

Article abstract of DOI:10.1002/anie.201403454

Guided by biosynthetic considerations, the total synthesis of dictazole B is reported for the first time. Experimental evidence for an easy access to challenging cyclobutane alkaloids of marine origin, which are often postulated to be biosynthetic precurs

Full text of DOI:10.1002/anie.201403454

Angewandte
Chemie
DOI: 10.1002/anie.201403454
Biomimetic Synthesis
Hot Paper
Spontaneous Biomimetic Formation of (Æ)-Dictazole B under
Irradition with Artificial Sunlight**
Adam Skiredj, Mehdi A. Beniddir, Delphine Joseph, Karine Leblanc, Guillaume Bernadat,
Laurent Evanno,* and Erwan Poupon*
Abstract: Guided by biosynthetic considerations, the total
synthesis of dictazole B is reported for the first time. Exper-
imental evidence for an easy access to challenging cyclobutane
alkaloids of marine origin, which are often postulated to be
biosynthetic precursors of more complex structures, is pro-
vided.
T
he aplysinopsins, a family of indolic marine natural
products with a common biosynthetic origin, were isolated
from several sponges and stony corals.[1] Three types of
skeletons are encountered in nature (Figure 1): 1) aplysinop-
sin-type monomeric structures (see the structures of aplysi-
nopsin (1) and its brominated analogues 2 and 3 in Figure 1),[2]
2) cycloaplysinopsins, which are tetrahydrocarbazole-type
dimeric compounds (see cycloaplysinopsin A (4) and dicta-
zoline C (5)),[3,4] and 3) intriguing spiro-fused cyclobutanes
(dictazoles A (6) and B (7)).[5] These compounds mainly differ
in the bromination state of the indole moiety and the N-
methylation patterns of the creatinine heterocycle. A stereo-
chemical analysis also revealed interesting facts, as variable
E/Z ratios were observed for the monomeric aplysinopsins.[6]
Furthermore, the cycloaplysinopsins were found to be of low
enantiopurity, and the substituents in the polycyclic structures
displayed various spatial arrangements.[7] Interestingly, most
of these compounds were isolated from samples that were
collected in shallow waters (see below).[8]
Despite being highly appealing at first sight, the previ-
ously postulated Diels–Alder cycloaddition[9] was questioned
when dictazoles, which may arise from an initial [2+2]
reaction, were isolated and presented as plausible intermedi-
[*] A. Skiredj,[+] Dr. M. A. Beniddir,[+] Prof. Dr. D. Joseph, K. Leblanc,
Dr. L. Evanno, Prof. Dr. E. Poupon
Figure 1. Overview of the structural diversity of the aplysinopsin
family.[14]
Laboratoire de Pharmacognosie associꢀ au CNRS
UMR 8076 BioCIS, LabEx LERMIT, Universitꢀ Paris-Sud
5, rue Jean-Baptiste Clꢀment, 92296 Chꢁtenay-Malabry (France)
E-mail: laurent.evanno@u-psud.fr
erwan.poupon@u-psud.fr
Homepage: http://www.biocis.u-psud.fr
ates towards their tetrahydrocarbazole counterparts by ring
expansion
with
piecemeal
experimental
evidence
(Scheme 1).[10] Recently, the key [2+2] photocycloaddition
step was thoroughly discussed by Baran and co-workers.[11]
Such processes have remained a challenge in total synthesis,
in particular, when the resulting cyclobutane structures have
been postulated as biosynthetic intermediates on the way to
even more complex rearranged structures.[10]
Dr. G. Bernadat
ꢂquipe “Molꢀcules fluorꢀes et chimie mꢀdicinale”
UMR 8076 BioCIS, LabEx LERMIT, Universitꢀ Paris-Sud
5, rue Jean-Baptiste Clꢀment, 92296 Chꢁtenay-Malabry (France)
[+] These authors contributed equally to this work.
[**] LabEx LERMIT is gratefully acknowledged for funding (grant to
A.S.). We thank Jean-Christophe Jullian and Jean-FranÅois Gallard
(ICSN-CNRS) for assistance with NMR spectroscopy.
Our interest in biomimetic strategies has led us to
investigate particular cases of “molecular” self-assembly,[12]
which allowed us to observe the spontaneous formation of
natural substances from simple reactive intermediates.[13]
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201403454.
Angew. Chem. Int. Ed. 2014, 53, 1 – 7
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

.
Angewandte
Communications
[M+2H]2+ ion and a monocharged [M+H]+ ion, excluding
the possibility of a mass-source dimer (In situ generated MS
ions corresponding to [2M+H]+ not related to real dimers).[19]
This peak could correspond to a new dimeric entity, but was
barely isolable at this stage. Several months were needed to
clarify and finally propose suitable conditions to obtain
a small amount of conversion in the solid state and identify
the new compound, even though the substrate/product ratio
was hardly reproducible, especially upon scale-up.[20] Gratify-
ingly, the structure appeared to be cyclobutane 12, which was
confirmed by NMR spectroscopy (see below), and the best
conditions entailed the use of artificial light that recreates the
irradiation with intense sunlight.[21] Many experimental
details had to be resolved to provide a reliable procedure
towards 12 and, as anticipated, its congeners, including
dictazole B (7). Another fortuitous observation rapidly
turned into a crucial discovery. Highly concentrated solutions
of 1, which mimic the high concentrations that may arise when
precursors are sequestered in confined compartments or
enzyme active sites in living cells, were studied. Monomeric
1 was dissolved in DMF (5 mm), placed in a crystallizing dish
as a thin film, and irradiated with the artificial light for one
night. To our delight, one day later, the solution had
evaporated to dryness, and analysis of the solid residue
revealed a reproducible 20% conversion into 12.[22] Solvents,
concentrations, additives, and illumination time were among
the parameters that had to be optimized. DMFappeared to be
the only suitable solvent for the dimerization.[23] Among many
additives, copper(I) triflate[24] satisfyingly improved the
dimerization up to 40% conversion. The most intriguing
experimental aspect resides in the fact that as long as 1 is in
solution in DMF very little dimerization occurs, whereas it
immediately takes place in the solid state, which suggests
a crucial self-organization step upon crystallization. A finely
tuned and reproducible method was devised that also avoids
the degradation of 12 by retro-dimerization when it is
irradiated in the solid state for longer periods of time. The
process was then extended to the synthesis of brominated
analogue 13, for which a similar ratio of dimerization was
observed. Formylindoles 8 and 9 were also formed in the
course of both reactions: Interestingly, they are also known as
natural products and have already been co-isolated with
aplysinopsins from sponges[25] (see LC/MS chromatogram in
Scheme 3).
Scheme 1. Biosynthetic hypothesis.
Hence, the “aplysinopsin cascade” process also attracted our
attention.
Therefore, the three monomeric aplysinopsins 1–3 were
prepared in a single step from 3-formylindole derivatives (8 or
9) and the suitable creatinine (10 or 11)[14] to set the stage for
the dimerization studies (Scheme 2). At first, the above-
mentioned Diels–Alder cycloaddition was investigated with
monomer 1 under different conditions, including heating for
Scheme 2. Synthesis of monomeric aplysinopsin derivatives. Reaction
conditions: a) MeI, EtOH (>95%); b) Bunsen flame, neat (59%);
c) piperidine, reflux (41%).
One of the expected drawbacks of a [2+2] photocycliza-
tion[26] is the lack of selectivity, which has been extensively
discussed in the literature, for example, by Baran and co-
workers.[11] Templated processes, including topochemical,
supramolecular, and biocatalytic approaches, to overcome
this major issue therefore constitute a currently appealing
topic of research.[27] In our case, the photochemical [2+2]
cycloaddition of the N-methylcreatinine containing E mono-
mers 1 and 2 stereoselectively yielded the anti head-to-tail
photodimers 12 and 13, respectively, when high local concen-
trations were achieved by sunlamp-assisted solvent evapora-
tion. The predominant formation of anti head-to-tail diaste-
reomers is in accordance with the hypothesis that topochem-
ical control of the photodimerization occurs through the
stacking of a centrosymmetric pair of neighboring monomers.
[15]
several days in DMF, use of SbCl3 and the Ledwith–Weitz
aminium salt (SbCl6N(p-BrPh)3),[16] or ultra-high pressures,
but the desired transformation was not observed.[17,18]
Despite several words of warning from the literature, we
embarked on studying the alternative [2+2] photocycloaddi-
tion reaction. We rapidly confirmed the results of previous
investigations,[4c,6] as E/Z isomerization was observed when
a solution of 1 was exposed to irradiation. Taken together,
these results could have appeared discouraging, but a subtle
detail markedly influenced the outcome of this project.
In fact, when running a routine LC/MS analysis of a solid
sample of 1 (that had stood for several weeks under sunlight
in the laboratory), we observed for the first time a small peak
in the chromatogram that is related to both a doubly charged
2
www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1 – 7
These are not the final page numbers!

Angewandte
Chemie
observed, and unfavorable self-organization could be feared.
The heterodimerization process was first evaluated with the
fully methylated derivatives 1 and 2 with CuOTf as the
additive and gave the expected statistical distribution in favor
of 14 over 12 and 13 (Scheme 4). The relative anti config-
uration of the indole moieties that are appended to the
cyclobutane framework was clearly demonstrated in the
course of the structure determination of 14. Subsequent
Scheme 4. Heterodimerization reactions. Reaction conditions:
a) DMF, hn, 1 (1 equiv), 2 (1 equiv), CuOTf toluene complex (1 equiv),
14 h; 14: 19%; 12: 11%; 13 co-eluted with 8. b) DMF, hn, 2 (1 equiv),
3 (1 equiv), Bi(OTf)3 (1 equiv), 14 h; dictazole B (7): 14%; 13: 19%.
Scheme 3. Solid-state dimerization under irradiation with artificial sun-
light. Reaction conditions: A solution of the starting material in DMF
(5 mm) was allowed to concentrate, CuOTf toluene complex
(0.5 equiv), 14 h (12: 40%; 13: 35%). LC/MS chromatogram of the
described reaction (see the Supporting Information).
exposure of a 1:1 mixture of 2 (N,N-dimethylated) and 3 (N-
monomethylated) to light led to new challenges. At first sight,
CuOTf was found to only allow the dimerization of fully
methylated aplysinopsins, as the conversion of 2 and 3 into 7
was minute (in fact not better than for the same reaction
without CuOTf). Therefore, new reaction conditions had to
be established. An understanding of the inherent topology
and innate reactivity of 3 was needed. The pseudo-aromatic
character of the a,b-unsaturated creatinine moiety may
explain the failure of the photodimerization of 3. In fact,
a particular mesomeric form of the imidazoline moiety
illustrates that for the [2+2] cycloaddition, 3 is the least
reactive precursor of the series because of deconjugation of
the C8/C9 double bond and facile E/Z photoisomerization.[29]
Hence, some other Lewis acids were evaluated, and the weak
Lewis acid bismuth(III) triflate significantly enhanced the
conversion into 7 and was thus selected as the best additive to
initiate the regio- and stereoselective cross-dimerization.[30,31]
With these conditions in hand, condensation of 2 and 3 was
finally possible and delivered dictazole B with 14% conver-
sion along with 13. Analysis of the ROESY NMR spectrum
clearly established the relative configurations of the C8, C8’,
C9, and C9’ positions based on correlations between the H8’
and H14’ hydrogen atoms and between the H8 and H14
hydrogen atoms. No correlation was observed between the
H8 and H8’ hydrogen atoms. The ROESY spectrum therefore
An investigation of the effects of the Lewis acid complexation
upon photodimerization showed that irradiation of a 2:1
monomer CuOTf complex significantly enhanced the effi-
ciency of the [2+2] cycloaddition without modifying the
regio- and stereoselectivities. In fact, complexation with
CuOTf may move the head-to-tail stacked pair of monomers
closer together, thereby shortening the distance between the
reactive double bonds. Finally, close analysis of a previous
crystallographic study of two synthetic aplysinopsin ana-
logues not only mirrored our findings, but also opened the
way to an improved understanding of the process.[28] In short,
first, crystals were obtained from crystallization in DMF with
a molecule of DMF incorporated into the lattice; second,
a perfect head-to-tail anti arrangement between two adjacent
sheets imposes a geometry that is highly favorable for the
light-induced [2+2] reaction.
The total synthesis of dictazole B (7) was the next
challenge. Because of its pseudodimeric structure ([2+2]
cycloaddition between 2 and 3), statistical couplings will be
Angew. Chem. Int. Ed. 2014, 53, 1 – 7
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

.
Angewandte
Communications
confirmed the trans configuration, which is the same as for
natural dictazole B (7).[32]
biosynthesis, demethylation reactions may occur after the
cycloaddition step.
In an attempt to rationalize our results and to understand
the difference in reactivity between monomers 1, 2, and 3 in
terms of their N-methylation patterns, we first calculated their
p/p* orbital energies and corresponding atomic coeffi-
cients.[33] HOMO and LUMO energies of the N,N-dimethy-
lated monomers 1 and 2 are drastically lower than those of the
N-monomethylated 3. The atomic orbital coefficients of the
HOMO for 1 and 2 revealed an “allyl-like” nonbonding
orbital, which might be detrimental for the p orbital overlap
that is required for the direct formation of the six-membered
ring of the cycloaplysinopsin framework (Figure 2). This fact
may explain the failure of all of our experiments that targeted
a thermal Diels–Alder reaction.
To understand and synthetically reproduce the whole
metabolism from formylindoles to cycloaplysinopsins, the
next step, which is currently under investigation, will be to
provide a reliable procedure for the conversion of dictazoles
into the corresponding cycloaplysinopsins. Meanwhile, the
herein described first total synthesis of dictazole B once again
highlights the long-lasting value of biosynthetic considera-
tions in the design of synthetic routes to natural substances,
especially when they exhibit densely functionalized struc-
tures.
Received: March 18, 2014
Published online: && &&, &&&&
Keywords: alkaloids · aplysinopsin · biomimetic synthesis ·
.
natural products · self-assembly
[1] For reviews, see: a) D. Bialonska, J. K. Zjawiony, Mar. Drugs
2009, 7, 166 – 183; b) E. M. Boyd, J. Sperry, Chem. N. Z. 2010, 74,
109 – 112.
[2] For the first isolation and description of aplysinopsin (1), see: R.
Kazlauskas, P. T. Murphy, R. J. Quinn, R. J. Wells, Tetrahedron
Lett. 1977, 18, 61 – 64. For the isolation of its congeners, see
Ref. [1]; for a recent example, see: N. Cachet, L. Loffredo, O. O.
Vicente, O. P. Thomas, Phytochem. Lett. 2013, 6, 205 – 208.
[3] For a description of cycloaplysinopsins A and B, see: a) I.
Mancini, G. Guella, H. Zibrowius, F. Pietra, Tetrahedron 2003,
59, 8757 – 8762; for a description of cycloaplysinopsin C, see:
b) M. Meyer, F. Delberghe, F. Liron, M. Guillaume, A. Valentin,
M. Guyot, Nat. Prod. Res. 2009, 23, 178 – 182.
Figure 2. HOMO and LUMO energies and “isodensity surfaces” of
1 and 2. Not shown: HOMO of 3: À5.18 eV; LUMO of 3: À1.47 eV.
[4] For a description of tubastrindoles A–C, see: a) T. Isagawa, M.
Miyazaki, H. Okamura, M. Nakatani, M. Doe, K. Takemura,
Tetrahedron Lett. 2003, 44, 2533 – 2535; for a description of
tubastrindoles D – H, see: b) T. Isagawa, M. Miyazaki, Y.
Yokogawa, H. Okamura, M. Nakatani, M. Doe, Y. Morimoto,
K. Takemura, Heterocycles 2008, 75, 2023 – 2028; for a descrip-
tion of tubastrindole B, see also: c) W. Balansa, R. Islam, D. F.
Gilbert, F. Fontaine, X. Xiao, H. Zhang, A. M. Piggott, J. W.
Lynch, R. J. Capon, Bioorg. Med. Chem. 2013, 21, 4420 – 4425.
[5] For a description of dictazolines A and B, see: a) J. Dai, J. I.
Jimꢀnez, M. Kelly, S. Barnes, P. Lorenzo, P. G. Williams, J. Nat.
Prod. 2008, 71, 1287 – 1290; for a description of dictazolines C –
E and dictazoles A and B, see: b) J. Dai, J. I. Jimꢀnez, M. Kelly,
P. G. Williams, J. Org. Chem. 2010, 75, 2399 – 2402.
[6] Mancini et al. studied the changing E/Z ratio of aplysinopsins
especially when they are exposed to light; see: a) G. Guella, I.
Mancini, H. Zibrowius, F. Pietra, Helv. Chim. Acta 1988, 71,
773 – 782; b) G. Guella, I. Mancini, H. Zibrowius, F. Pietra, Helv.
Chim. Acta 1989, 72, 1444 – 1450. See also Ref. [4c] for sunlight-
induced E/Z interconversion.
Concerning the [2+2] photocyclization reaction, both
singlet and triplet states can be involved in the process.
Nevertheless, in the presence of triplet photosensitizers,
significant dimer formation was not observed.[34] This finding
could indicate that the reaction occurs under solid-state
conditions and involves short-lived singlet states because of
the high proximity of the monomers.[35] Furthermore, accord-
ing to Schmidt’s topochemical postulate, the precise packing
of molecules (i.e., with properly placed reactive centers) also
controls the outcome of [2+2] photodimerization reactions in
the solid state.[36] In line with this hypothesis, the aforemen-
tioned crystallographic structures[28] gave us valuable indica-
tions concerning both the spatial arrangement of structures
such as 3 in the solid state and the electronic structure of the
creatinine moiety, which may preclude their direct dimeriza-
tion.[37]
The effectiveness of Bi(OTf)3 to enhance the stereoselec-
tive cross-photodimerization between 2 and 3, which is
needed for the synthesis of dictazole B (7), may thus be
attributed to an increase in electrophilicity and singlet-state
lifetimes combined with lowered orbital energies for com-
plexed 3, which restores the properties of the N,N-dimethy-
lated monomers 1 and 2 to a certain extent.[38] The higher
propensity of dimethylated aplysinopsin derivatives to dimer-
ize is corroborated by the fact that all dictazole- or cyclo-
aplysinopsin-type natural products contain at least one such
dimethylated imidazolidinone. It is thus proposed that during
[7] The enantiomeric excess of cycloaplysinopsin A (4) was found to
be poor by using the chiral shift reagent Eu(tfc)3; the relevant
signals in the NMR spectrum were split in a ratio of approx-
imately 65:35 (see Ref. [3a]).
[8] The sponge S. cerebriformis was collected at the northwest coast
of Panama from a depth of 2 – 3 m (Ref. [5a,b]); stony-coral
Tubastraea sp. samples were collected in a reef environment
from a depth of 3 m at Comoro Islands (Ref. [3a]), from a depth
of ca. 3 m at Capsalon Island, Philippines (Ref. [6a] and [3a]),
and from
a depth of 12 m depth in the Hanish Islands
archipelago, Yemen (Ref. [3b]). The depth was not mentioned
in the description of the collection of Tubastraea sp. in the
4
www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1 – 7
These are not the final page numbers!

Angewandte
Chemie
Odomari area, Kagoshima prefecture, Japan (Ref. [3a] and [4b])
and for the sponge Ianthella cf. flabelliformis, which was
collected at Lonsdale wall, Port Phillip, Australia (38817’55’’S,
144837’76’’E). For the penetration of light into (tropical) sea
water, see for example: R. P. Dunne, B. E. Brown, Mar. Ecol.
Prog. Ser. 1996, 144, 109 – 118; E. M. Fleischmann, Limnol.
Oceanogr. 1989, 34, 1623 – 1629.
[20] The reaction, which takes place at the surface of powdered 1,
necessitated daily fine grinding for the renewal of the surface.
[21] Artificial sunlight restitutes high UV-B quanta; see the Support-
ing Information for details.
[22] See the Supporting Information for details concerning the
expression of the results in terms of conversion percentages.
[23] DMF was chosen because it appeared to be the best solvent to
solubilize 1 and 2. DMSO afforded a small dimer peak in the
HPLC chromatogram but mainly provided the retro-aldol
products 8 and 9.
[24] For the use of copper triflate, see: a) R. G. Salomon, Tetrahedron
1983, 39, 485 – 575; b) R. G. Salomon, K. Folting, W. E. Streib,
J. K. Kochi, J. Am. Chem. Soc. 1974, 96, 1145 – 1152.
[25] For the isolation of formylindoles 8 and 9, see, for example
(isolation from a Smenospongia sponge): M. J. McKay, A. R.
Carroll, R. J. Quinn, J. N. A. Hooper, J. Nat. Prod. 2002, 65, 595 –
597.
[26] a) T. Bach, J. P. Hehn, Angew. Chem. 2011, 123, 1032 – 1077;
Angew. Chem. Int. Ed. 2011, 50, 1000 – 1045; b) J. Iriondo-
Alberdi, M. F. Greaney, Eur. J. Org. Chem. 2007, 4801 – 4815.
[27] a) K. M. Hutchins, J. C. Sumrak, L. R. MacGillivray, Org. Lett.
2014, 16, 1052 – 1055; b) G. K. Kole, T. Kojima, M. Kawano, J. J.
Vittal, Angew. Chem. 2014, 126, 2175 – 2178; Angew. Chem. Int.
Ed. 2014, 53, 2143 – 2146; c) R. Medishetty, R. Tandiana, L. L.
Koh, J. J. Vittal, Chem. Eur. J. 2014, 20, 1231 – 1236.
[28] See the Supporting Information of J. E. Johnson, D. C. Canseco,
D. D. Dolliver, J. A. Schetz, F. R. Fronczek, J. Chem. Crystallogr.
2009, 39, 329 – 336.
[9] The putative Diels–Alder cycloaddition was particularly inves-
tigated and discussed by Mancini et al. (see Ref. [3a]): Sponta-
neous formation during the extraction was rapidly set aside, and
the existence of a putative Diels–Alderase was discussed.
[10] Indeed, preliminary results (LC/MS traces, see Ref. [5b])
indicated a conversion of dictazole A into dictazoline C under
microwave activation in water, following the seminal and
beautiful conversion of sceptrin into ageliferin under such
conditions; see: B. H. Northrop, D. P. OꢁMalley, A. L. Zografos,
P. S. Baran, Angew. Chem. 2006, 118, 4232 – 4236; Angew. Chem.
Int. Ed. 2006, 45, 4126 – 4130, and references therein. In the
particular case of pyrrole aminoimidazole alkaloids, Molinski
and co-workers recently disclosed a crucial step forward in
understanding the biosynthesis of this highly diverse family of
natural substances; see: E. P. Stout, Y.-G. Wang, D. Romo, T.
Molinski, Angew. Chem. 2012, 124, 4961 – 4965; Angew. Chem.
Int. Ed. 2012, 51, 4877 – 4881; E. P. Stout, B. I. Morinaka, Y.-G.
Wang, D. Romo, T. F. Molinski, J. Nat. Prod. 2012, 75, 527 – 530.
[29] The existence of particular pseudo-aromatic mesomeric forms
may explain the following:
[11] a) W. R. Gutekunst, P. S. Baran, J. Org. Chem. 2014, 79, 2430 –
2452. This paper also features a synthetic approach towards
dictazole A (6). The special [2+2] photochemical origin of
sceptrin has also been discussed, but attempts to realize this
[2+2] dimerization have previously been unsuccessful; see:
b) R. P. Walker, D. J. Faulkner, D. Van Engen, J. Clardy, J. Am.
Chem. Soc. 1981, 103, 6772 – 6773.
[12] a) E. J. Sorensen, Bioorg. Med. Chem. 2003, 11, 3225 – 3228;
b) E. Poupon, E. Gravel, Eur. J. Org. Chem. 2008, 27 – 42.
[13] See, for example: a) E. Gravel, A. Harfouche, R. Salame, K.
Leblanc, A. Maciuk, E. Poupon, Chem. Eur. J. 2013, 19, 14515 –
14519, and references therein; b) F. Senejoux, L. Evanno, E.
Poupon, Eur. J. Org. Chem. 2013, 453 – 455.
[30] Evaluated Lewis acids include La(OTf)3, Sc(OTf)3, Y(OTf)3,
Yb(OTf)3, MgCl2, and ZnCl2, which mainly provided retro-aldol
products, that is, formylindoles 8 and 9. Bi(OTf)3 was selected.
For the use of Bi(OTf)3 in organic synthesis, see: J. M. Bothwell,
S. W. Krabbe, R. S. Mohan, Chem. Soc. Rev. 2011, 40, 4649 –
4707.
[31] Activation by a mild oxophilic Lewis acid in conjunction with
activation by protonation, which arises from the catalytic
liberation of triflic acid, may explain the outcome of the
reaction. Protonation of the creatinine moiety of 3 indeed
enables the loss of its pseudo-aromatic character, which renders
the C8/C9 double bond more reactive for the [2+2] photo-
cycloaddition.
[14] Procedures adapted from Ref. [6a] and Ref. [4c], for example.
For alternative strategies, see: L. Ermolenko, H. Zhaoyu, C.
Lejeune, C. Vergne, C. Ratinaud, T. B. Nguyen, A. Al-Mourabit,
Org. Lett. 2014, 16, 872 – 875.
[15] G. Maiti, P. Kundu, Tetrahedron Lett. 2006, 47, 5733 – 5736.
[16] a) S. L. Drew, A. L. Lawrence, M. S. Sherburn, Angew. Chem.
2013, 125, 4315 – 4318; Angew. Chem. Int. Ed. 2013, 52, 4221 –
4224; b) H. N. Lim, K. A. Parker, Org. Lett. 2013, 15, 398 – 401.
[17] W. G. Dauben, C. R. Kessel, K. H. Takemura, J. Am. Chem. Soc.
1980, 102, 6893 – 6894.
[18] Diels–Alder reaction with 1: 1) heating in DMF for 5 days;
2) DMF, RT, 24 h; 3) SbCl6N(p-BrPh)3 (10 mol%), CH2Cl2 (low
solubility) or DMF, RT, 12 h; 4) DMF/THF (1:1), 1.2 GPa, RT,
15 h. Reaction conditions 1–3 resulted in no reaction, conditions
4 mainly afforded a retro-aldol reaction.
[32] See the Supporting Information for in-depth analysis of 2D
NMR data.
[33] See the Supporting Information, p. 21 for the calculation method
and a Figure showing orbital energies and atomic coefficients of
1, 2, and 3.
[34] Evaluated photosensitizers include benzophenone and salco-
mine, which fully or partially inhibited the dimerization reaction.
[35] H. GonÅalves, G. Robinet, M. Barthelat, A. Lattes, J. Phys.
Chem. A 1998, 102, 1279 – 1287.
[19] See the Supporting Information for an analysis of the mass
spectrum of 12 (RT = 8.08 min), which shows the [M+2H]2+ ion
at m/z = 255.1194 along with the [M+H]+ ion at m/z = 509.2409.
Angew. Chem. Int. Ed. 2014, 53, 1 – 7
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!

.
Angewandte
Communications
[36] a) M. D. Cohen, G. M. J. Schmidt, J. Chem. Soc. 1964, 1996 –
2000; b) M. D. Cohen, G. M. J. Schmidt, F. I. Sonntag, J. Chem.
Soc. 1964, 2000 – 2013; c) G. M. J. Schmidt, J. Chem. Soc. 1964,
2014 – 2021.
[37] Whereas a 3’-exocyclic imine and a perfect stacking between the
indole and the imidazolinone heterocycle was observed with
superposed N-1,N-10’-dimethylated monomers
Ref. [28]), an endocyclic imine and different stacking may
explain the differences in reactivity.
A
(see
[38] a) F. D. Lewis, S. L. Quillen, P. D. Hale, J. D. Oxman, J. Am.
Chem. Soc. 1988, 110, 1261 – 1267; b) F. D. Lewis, J. E. Elbert,
A. L. Upthagrove, P. D. Hale, J. Org. Chem. 1991, 56, 553 – 561.
6
www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1 – 7
These are not the final page numbers!

Angewandte
Chemie
Communications
Biomimetic Synthesis
A. Skiredj, M. A. Beniddir, D. Joseph,
K. Leblanc, G. Bernadat, L. Evanno,*
E. Poupon*
&&&& — &&&&
Spontaneous Biomimetic Formation of
(Æ)-Dictazole B under Irradition with
Artificial Sunlight
Always the sun? Biosynthetic considera-
tions guided the first total synthesis of
dictazole B. Furthermore, insights into
the biosynthetic pathway towards the
aplysinopsin family were gained, and an
easy access to challenging cyclobutane
alkaloids of marine origin, which are often
postulated to be biosynthetic precursors
of more complex structures, was devel-
oped.
Angew. Chem. Int. Ed. 2014, 53, 1 – 7
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7
These are not the final page numbers!