Photoassisted synthesis of enantiopure alkaloid mimics possessing unprecedented polyheterocyclic cores

Article abstract of DOI:10.1021/ja4042109

Enantiopure alkaloid mimics are synthesized via high yielding intramolecular cycloadditions of photogenerated azaxylylenes tethered to pyrroles, with further growth of molecular complexity via post-photochemical transformations of primary photoproducts. This expeditious access to structurally unprecedented polyheterocyclic cores is being developed in the context of diversity-oriented synthesis, as the modular design allows for rapid "pre-assembly" of diverse photoprecursors from simple building blocks/diversity inputs.

Full text of DOI:10.1021/ja4042109

Communication
pubs.acs.org/JACS
Photoassisted Synthesis of Enantiopure Alkaloid Mimics Possessing
Unprecedented Polyheterocyclic Cores
N.N. Bhuvan Kumar, Olga A. Mukhina, and Andrei G. Kutateladze*
Department of Chemistry and Biochemistry, University of Denver, Denver, Colorado 80208, United States
S
* Supporting Information
In this Communication, we report that the N-tethered
ABSTRACT: Enantiopure alkaloid mimics are synthe-
sized via high yielding intramolecular cycloadditions of
photogenerated azaxylylenes tethered to pyrroles, with
further growth of molecular complexity via post-photo-
chemical transformations of primary photoproducts. This
expeditious access to structurally unprecedented poly-
heterocyclic cores is being developed in the context of
diversity-oriented synthesis, as the modular design allows
for rapid “pre-assembly” of diverse photoprecursors from
simple building blocks/diversity inputs.
pyrroles undergo an exclusive [4 + 2] cycloaddition to
photogenerated azaxylylenes yielding polycyclic aminals 8−10
which possess a reactive enamine moiety suitable for
subsequent transformations. Chiral α-substituted N-pyrrole
acetic acids 4 are readily available on the multigram scale from
α-amino acids.5 They are readily coupled with the photoactive
moiety, aromatic amino-aldehydes and ketones 1−3, and
irradiated as shown in Scheme 1 to yield bowl-shaped
tetracyclic enantiopure aminals 8−10 with the high exo-R and
endo-OH diastereoselectivity.
Scheme 1
igh throughput synthetic methods are blamed for
H“steering discovery efforts toward achiral, aromatic
compounds” while natural products, possessing a broad
spectrum of bioactivity, look nothing like the sp2-dominated
aromatic heterocycles in the proverbial “flatland” of the lead
discovery. Lovering1 has shown convincingly that as drug
candidates progress through the stages of development their
average saturation factor fsp32 grows steadily, starting from fsp3
= 0.36 for the “discovery” stage, and increasing for compounds
which cleared Phase I through Phase III clinical trials (0.38 →
0.43 → 0.45). An average fsp3 for a sample of 1179 approved
drugs is 0.47. It appears that no one deems these findings
controversial. However, the practical difficulties of synthesizing
complex, sp3-rich natural products drive the observed shift into
the flatland, where the well-developed catalytic sp2−sp2
coupling reactions dominate.
Aminals 8−10 and the products of their postphotochemical
transformations are structurally related to the alkaloids of the
chaetominine6 and kapakahine7 families, except that in a
pseudo symmetric fashion, the benzene ring is “translocated”
from the five-membered pyrrole- to the six-membered
piperidine moiety, resulting in a new and unique polyheter-
ocyclic core which has not been previously reported.
Chaetominine (Figure 1), which is isolated from Chaetomium
species of endophytic fungi, is active against two lines of human
cancer cell lines (SW1116, 28 nM; and K562, 21 nM), which
triggered synthesis efforts in several groups, starting with
Snider’s total synthesis in 2007.6b
The novel N,N-ethanone-linked pyrrolo[2,3-b]quinoline core
in photoproducts 8−10 has a reactive enamine moiety, which
makes them excellent synthons for subsequent modifications to
further increase molecular complexity. Scheme 2 illustrates the
rapid “pre-assembly” of photoprecursor 7b from aminotetr-
alone 3 and alanine-derived pyrrole 4b and its photoinduced
Our approach to this problem, in the context of diversity-
oriented synthesis, is the modular “assembly” of photo-
precursors with subsequent intramolecular photoinduced
cyclizations as the key step delivering a rapid increase in
complexity while also increasing saturation and installing
additional stereogenic centers. The synthesis of photo-
precursors, which by design are largely unsaturated, is based
on well-developed coupling reactions, fully compatible with
high throughput combinatorial synthetic techniques.
The time-proven hetero Diels−Alder reactions are a
powerful example of synthesis of partially saturated heterocyclic
scaffolds.3 However, the scope of these ground state reactions is
limited, and the search for new cycloadditions is well-justified
and ongoing. We have recently discovered one of such
reactions which has a tremendous synthetic potential. We
found that azaxylylenes, photogenerated via the excited state
intramolecular proton transfer (ESIPT) in aromatic o-
amidoketones, could be trapped intramolecularly by tethered
unsaturated pendants.4
Received: May 1, 2013
Published: June 21, 2013
© 2013 American Chemical Society
9608
dx.doi.org/10.1021/ja4042109 | J. Am. Chem. Soc. 2013, 135, 9608−9611

Journal of the American Chemical Society
Communication
Table 1. Two-Step Photoassisted Conversion of Precursors
5−7 to Cyclic N-Sulfonyl Amidines 11−16
Figure 1. Pyrido[2,3-b]indole core of chaetominines and kapakahines
and the pyrrolo[2,3-b]quinoline core accessed in this work.
Scheme 2
conversion to pentacyclic enamine 10b. Without additional
purification, 10b is reacted with toluenesulfonyl azide yielding
cyclic N-sulfonyl amidine 11b (the ORTEP drawing of its X-ray
structure is shown).8
Initially, 5% aqueous acetonitrile was used for the irradiation
step, after which the solution was preconcentrated, diluted with
DMSO, and reacted with the azide. Quantitative conversion of
primary photoproducts, enamines 8−10 to amidines 11−16
occurs in DMSO at ambient temperature within 2 h. Table 1
summarizes the overall two-step photoassisted conversion of
precursors 5−7 to N-sulfonyl cyclic amidines 11−16, attesting
to the broad scope of this synthetic sequence.
a
Product obtained with procedure: (A) one pot, irradiated and reacted
with the azide in DMSO; (B) irradiated in aq MeCN, concentrated,
diluted with DMSO, reacted with the azide. Isolated yield over two
steps. X-ray structure is available. 43% syn- and 29% anti-.
b
c
d
Later it was found that the irradiation step can also be carried
out in DMSO, so that photoprecursor 7b is converted to
amidine 11b in a two-step, one-pot fashion. Furthermore,
sulfonyl azides do not absorb much above 350 nm, so they are
not considerably affected by the UV-LED@365nm irradiation.
This allows implementing an even simpler one-pot procedure,
where 7b is irradiated in DMSO in the presence of tosyl azide,
which is added to the solution before irradiation. Figure 2
shows the condition optimization runs and reveals that while a
two-step, one-pot procedure in DMSO is slightly superior (84%
over two steps), the one-step procedure is still reasonably
competitive: the best yield is 71%. It also reveals that, although
the azide step is bimolecular, overconcentration does not help
the yield, which at 1 M is only 38%.
Figure 2. One-step vs two-step procedure in DMSO.
Irradiation of 6h in DMSO with subsequent addition of
dimethyl acetylene dicarboxylate (DMAD) at an elevated
temperature (80 °C) yields 17h as a result of [2 + 2]
cycloaddition to enamine 9h (Ssheme 3). At higher temper-
atures, azepino-[2,3-b]-quinolinol 18h is obtained, presumably
via electrocyclic ring-opening in 17h.
DMSO is an atypical choice for photochemical reactions.
However, due to its hydrogen bonding ability, it is known to
stabilize the ESIPT species9 extending the excited state lifetimes
and therefore enhancing the reactivity of photogenerated
azaxylylenes.
Only in the case of aminobenzaldehyde-based photo-
precursor 5h, both diastereomers of the final amidine, syn,
43% (OH and the amidine cycle are on the same face of the
quinoline moiety) and anti, 29%, are isolated. The rationale for
this decrease in stereospecificity is that the small formyl group
in the aldehyde photoprecursor is more freely rotatable than in
ketones.
In the presence of acids, the enamine moiety in the primary
photoproducts 8−10 expectedly forms iminium ions, which can
be trapped by a variety of nucleophiles. Given that there are a
number of natural and non-natural amino acids possessing a
secondary (ω)-nucleophilic functionality, this offers an
opportunity to trap these iminium ions intramolecularly. We
found not only that such trapping is possible but also that in
9609
dx.doi.org/10.1021/ja4042109 | J. Am. Chem. Soc. 2013, 135, 9608−9611

Journal of the American Chemical Society
Communication
Scheme 3
Scheme 5
this sequence of post-photochemical modifications a new C−C
bond can be formed. In bis-pyrrole substituted carboxylic acids
4e and 4f, which are readily obtained from bis-amino acids
ornithine and lysine, the generated iminium ions are capable of
aromatic electrophilic substitution in the second pyrrole
moiety, leading to the formation of large eight-membered
diazocane- and nine-membered diazonane rings in moderate
yields, Scheme 4 (the ORTEP drawing of the X-ray structure of
1,4-diazonane 20 is shown).
Scheme 6
Scheme 4
trapping by quinazolone in a reaction with phenylalanine-
based 9h.
In summary, the new intramolecular cyclization of photo-
generated azaxylylenes with tethered pyrrole pendants based on
α-amino acids provides a powerful tool for building complex
enantiopure alkaloid mimics possessing unique polyheterocyclic
cores, with multiple (5−6) stereogenic centers, structurally
similar to the chaetominine and kapakahine families of
alkaloids. The photoprecursors for these compounds are readily
synthesized via 2−3 simple and well-developed steps. The key
photochemical step allows for a dramatic growth of molecular
complexity augmented with experimentally simple post-photo-
chemical steps. The resulting enantiopure polyheterocyclic
scaffolds are rich in sp3 carbons (fsp3 ≈ 0.3−0.6) and possess
multiple new stereogenic centers, including quaternary.
Having employed C-nucleophiles for trapping of the iminium
cations, we also looked into C-electrophiles for their
generation, as enamines are well-known to react with soft
carbon electrophiles. Utilization of C-electrophiles, such as
Eschenmoser’s aminomethylating reagent, allows for an addi-
tional diversity input connected to the polyheterocyclic scaffold
via a carbon−carbon bond. As shown in Scheme 5, the one-pot
procedure is run in 5% aqueous acetonitrile, which stabilizes the
primary photoproduct, enamine 9. After irradiation excess
water is removed by adding anhydrous sodium sulfate, and
Echenmoser’s salt is added to yield hemi-aminals 21. In the case
ornithine- and lysine-based photoprecursors 6e and 6f, the
terminal pyrrol moiety reacts with excess Eschenmoser’s
reagent yielding diamines 21e and 21f.
Unlike the protic acid-induced reaction shown in Scheme 4,
the bulky Eschenmoser electrophile prevents intramolecular
cyclization involving terminal pyrroles in 9f,e. Our attempts to
force the cyclization in diamines 21e,f by treating them with
trifluoroacetic acid failed. Judging by NMR, they instead
eliminate water from the pyrrolidine ring to form enamines.
The cyclic iminium ion, generated from the primary
photoproducts, enamines 8−10, with a protic acid is capable
of trapping more complex external nucleophiles as additional
diversity inputs. Scheme 6 illustrates such intermolecular
ASSOCIATED CONTENT
■
S
* Supporting Information
Detailed experimental procedures and characterization data.
This material is available free of charge via the Internet at
http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
akutatel@du.edu
Notes
The authors declare no competing financial interest.
9610
dx.doi.org/10.1021/ja4042109 | J. Am. Chem. Soc. 2013, 135, 9608−9611

Journal of the American Chemical Society
Communication
ACKNOWLEDGMENTS
This research is supported by the NIH, GM-093930 and NSF,
CHE-1057800.
■
REFERENCES
■
(1) Lovering, F.; Bikker, H.; Humblet, C. J. Med. Chem. 2009, 52,
6752.
(2) fsp3 = (number of sp3 hybridized carbons)/(total carbon count),
see ref 1.
(3) Boger, D. L.; Weinreb, S. M. Hetero Diels-Alder Methodology in
Organic Synthesis; Academic Press: San Diego, CA, 1987.
(4) (a) Mukhina, O. A.; Kumar, N. N. B.; Arisco, T. M.; Valiulin, R.
A.; Metzel, G. A.; Kutateladze, A. G. Angew. Chem., Int. Ed. 2011, 50
(40), 9423−9428. (b) Nandurkar, N. S.; Kumar, N. N. B.; Mukhina,
O. A.; Kutateladze, A. G. ACS Comb. Sci. 2013, 15 (1), 73.
(5) Jefford, C. W.; de Villedon de Naide, F.; Sienkiewicz, K.
Tetrahedron: Asymmetry 1996, 7, 1069.
(6) (a) Jiao, R. H.; Xu, S.; Liu, J. Y.; Ge, H. M.; Ding, H.; Xu, C.;
Zhu, H. L.; Tan, R. X. Org. Lett. 2006, 8, 5709. (b) Snider, B. B.; Wu,
X. Org. Lett. 2007, 9, 4913. (c) Toumi, M.; Couty, F.; Marrot, J.;
Evano, G. Org. Lett. 2008, 10, 5027. (d) Malgesini, B.; Forte, B.;
Borghi, D.; Quartieri, F.; Gennari, C.; Papeo, G. Chem.Eur. J. 2009,
15, 7922.
(7) (a) Nakao, Y.; Yeung, B. K. S.; Yoshida, W. Y.; Scheuer, P. J.;
Kelly-Borges, M. J. Am. Chem. Soc. 1995, 117, 8271. (b) Yeung, B. K.
S.; Nakao, Y.; Kinnel, R. B.; Carney, J. R.; Yoshida, W. Y.; Scheuer, P.
J.; Kelly-Borges, M. J. Org. Chem. 1996, 61, 7168. (c) Nakao, Y.; Kuo,
J.; Yoshida, W. Y.; Kelly, M.; Scheuer, P. J. Org. Lett. 2003, 5, 1387.
(d) Newhouse, T.; Lewis, C. A.; Baran, P. S. J. Am. Chem. Soc. 2009,
131, 6360. (e) Espejo, V. R.; Rainier, J. D. Org. Lett. 2010, 12, 2154.
(f) Newhouse, T.; Lewis, C. A.; Eastman, K. J.; Baran, P. S. J. Am.
Chem. Soc. 2010, 132, 7119. (g) Gaich, T.; Baran, P. S. J. Org. Chem.
2010, 75, 4657. (h) Ruiz-Sanchis, P.; Savina, S. A.; Albericio, F.;
Alvarez, M. Chem.Eur. J. 2011, 17, 1388. (i) Rainier, J. D.; Espejo, V.
R. Isr. J. Chem. 2011, 51, 473.
(8) Cycloaddition of arylsulfonyl azides to enamines is often
accompanied by extrusion of N2 from the initially formed
dihydrotriazoles and a secondary rearrangement to N-sulfonyl
amidines, which were reported to be promising pharmacophores:
(a) Ondrus, T. A.; Knaus, E. E. Can. J. Pharm. Sci. 1979, 14, 55.
(b) Warren, B. K.; Knaus, E. E. J. Heterocycl. Chem. 1982, 19, 1259.
(c) Buolamwini, J. K.; Knaus, E. E. Eur. J. Med. Chem. 1993, 28, 447.
(9) McGarry, P. F.; Jockusch, S.; Fujiwara, Y.; Kaprinidis, N. A.;
Turro, N. J. J. Phys. Chem. A 1997, 101, 764.
9611
dx.doi.org/10.1021/ja4042109 | J. Am. Chem. Soc. 2013, 135, 9608−9611