Stereoselective synthesis of complex polycyclic aziridines: Use of the Bronsted acid-catalyzed aza-Darzens reaction to prepare an orthogonally protected mitomycin C intermediate with maximal convergency

Article abstract of DOI:10.1039/c0cc05734g

A concise synthesis of a highly functionalized intermediate lacking only C10 of the mitomycin backbone is described. The key to this development is the Bronsted acid-catalyzed aza-Darzens reaction used to forge the cis-aziridine. Additionally an oxidative

Full text of DOI:10.1039/c0cc05734g

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
ChemComm
Cite this: Chem. Commun., 2011, 47, 3975–3977
www.rsc.org/chemcomm
COMMUNICATION
Stereoselective synthesis of complex polycyclic aziridines: use of the
Brønsted acid-catalyzed aza-Darzens reaction to prepare an orthogonally
protected mitomycin C intermediate with maximal convergencyw
c
Jayasree M. Srinivasan,b Priya A. Mathew,a Amie L. Williams,b John C. Huffmanz and
Jeffrey N. Johnston*a
Received 22nd December 2010, Accepted 10th February 2011
DOI: 10.1039/c0cc05734g
A concise synthesis of a highly functionalized intermediate lacking
only C10 of the mitomycin backbone is described. The key to this
development is the Brønsted acid-catalyzed aza-Darzens reaction
used to forge the cis-aziridine. Additionally an oxidative ketaliza-
tion fortuitously occurs during the quinone–enamine coupling step,
leading to an orthogonally protected hydroquinone.
to prepare bromoquinone 2a.7 The reactivity of this intermediate
was reflected by the use of cold column chromatography to purify
enamine 2a and its tendency to decompose during spectroscopic
analysis. As a result of our desire to incorporate new reactions
into this sequence to further streamline access to advanced
intermediates, we considered the ostensibly less desirable strategy
in which both nitrogens bear identical protecting groups (2b).
In this report, we detail our finding that protecting group
differentiation can be achieved during preparation of 2b by a
selective hydride transfer that reduces the quinone, and
then protects it via ketalization. Furthermore, the cis-aziridine
precursor to 2b is readily prepared using a diastereoselective,
Brønsted acid-catalyzed aza-Darzens reaction.8,9
The mitomycins are a family of potent antibacterial and
anticancer agents that have sustained broad interest since their
isolation in the late 1950s due to their biological activity and
the synthetic challenge they offer.1 Mitomycin C (1) has been
shown to crosslink DNA and has been clinically used to treat
cancer for several decades.2 The challenge of mitomycin
chemical synthesis is largely due to the compact array of
reactive functionalities, and reflected by only two total syntheses
of rac-mitomycin C and two total syntheses of rac-mitomycin K.3–5
It could be argued that the length of these preparations
(28 steps or more for 1) has limited the further development
of the mitomycins as powerful therapeutics.6
Enamine 2a was prepared using
a
base-promoted
aza-Darzens reaction of a-bromo ethyl acetate with the aldimine
derived from tert-butyl glyoxal.7 The esters were then chemo-
selectively refunctionalized to alkyne and aminomethyl
substituents of the central cis-aziridine. As an alternative, we
investigated whether a Brønsted acid-catalyzed aza-Darzens
reaction10 might produce the alkyne-substituted aziridine
more immediately. The imine substrate (4) for this was prepared
in three steps without purification from commercially available
propargyl alcohol (3) (Scheme 1). Alkyne protection of
We have outlined a strategy to convergently access the
mitomycins using a regioselective enamine–quinone coupling
a Department of Chemistry & Vanderbilt Institute of Chemical
Biology, Vanderbilt University, Nashville, TN 37235-1822, USA.
E-mail: jeffrey.n.johnston@vanderbilt.edu;
Web: http://www.johnstonchemistry.org; Fax: (+1) 615 343 1234
b Department of Chemistry, Indiana University, Bloomington,
IN 47405, USA
c IU Molecular Structure Center, Indiana University, Bloomington,
IN 47405, USA
w Electronic supplementary information (ESI) available: Complete
experimental details and analytical data for all new compounds;
X-ray data for 12. CCDC 806065. For ESI and crystallographic data
in CIF or other electronic format see DOI: 10.1039/c0cc05734g
z To whom questions regarding crystallography should be addressed.
Scheme 1 Preparation of an alkynyl amine using the Brønsted acid-
catalyzed aza-Darzens reaction.
c
This journal is The Royal Society of Chemistry 2011
Chem. Commun., 2011, 47, 3975–3977 3975

View Article Online
propargyl alcohol was followed by oxidation11 and condensation
of the resulting aldehyde with diphenyl methyl amine (DPM-
NH2) to afford imine 4 (E : Z > 95 : 5). The imine E : Z ratio is
dependent on the alkyne protecting group and ranges from
63 : 37 (trimethylsilyl) to exclusively E geometry (triphenylsilyl
group, TPS). The Brønsted acid-catalyzed aza-Darzens reaction
of 4 with ethyl diazoacetate provided cis-5 in 70% isolated
yield compared to 39–50% for the analogous TMS-imine. The
ethyl ester was then saponified and coupled with diphenyl
methyl amine to give amide 6. Reduction of the amide to
amine 7 readied this fragment for coupling.
and resulted in oxidation of the hydroquinone and production
of benzophenone (a DPM oxidation product).
Since the next planned synthesis step was to install C10, the
enamine nucleophilicity was assessed using N-chlorosuccinimide.
The chlorinated product was crystalline and an X-ray structure
was obtained (Fig. 1). This revealed that the bromine on the
hydroquinone ring had been replaced by a hydrogen and one
hydroquinone hydroxyl group had formed a seven-membered
ring with the pyrrolidine diphenylmethyl group. By retro-
spective analysis, the coupling reaction had resulted in an
oxidative ketalization to form N,O-ketal 11, thus explaining
the observation that the coupling product could only be singly
benzoylated.
Cyclization of alkynyl amine 7 with HgCl2 leads to an
intermediate enamine12 (perhaps 8) that is treated with quinone
9.13 Standard reductive workup with NaBH4 was expected to
give the hydroquinone of 2b (10, Scheme 2). Although a
colorless product was isolated from this mixture, its spectro-
scopic data was different in key areas from 2a or its analogous
hydroquinone, specifically in the downfield shift of both
enamine and pyrrolidine-DPM methine protons by B0.5 ppm.
Assuming the formation of 10, we treated this material with
benzoyl chloride and triethylamine. A single phenolic hydroxyl
group was acylated despite varying reaction conditions. Attempts
to oxidatively cleave the pyrrolidine protecting group failed,14
This unusual transformation requires two reducing equivalents
during the reaction. The oxidation of the pyrrolidine diphenyl-
methyl group suggested an intramolecular source of one
equivalent of hydride. This would also explain why the
a-methylbenzyl-protected pyrrolidine in 2a, which is a poorer
hydride donor, did not exhibit this behavior. The second
reducing equivalent may be attributed to triethylamine, as
variations that did not employ sodium borohydride still
produced the mixed ketal.
Fig. 1 X-ray crystal structure of 12.
Scheme 2 Formation of an N,O-ketal during the second convergent
Scheme 3 Deuterium labeling experiment revealing selective transfer
coupling.
of deuterium from one diphenylmethyl group.
Scheme 4 Mechanistic proposal for the RedOx ketalization subsequent to coupling.
c
This journal is The Royal Society of Chemistry 2011
3976 Chem. Commun., 2011, 47, 3975–3977

View Article Online
Our hypothesis that the diphenylmethyl group delivered
the hydride was investigated using a deuterium labeling
experiment (Scheme 3). Deuterated alkynyl amine d-7 was
prepared and subjected to the standard reaction conditions.
The coupled N,O-ketal showed an 88% deuterium incorporation
at the hydroquinone C–H. This verifies that one of the two
reducing equivalents results from the conserved transfer of the
diphenylmethyl methine.15
D. Chowdary, J. Pawlak, G. L. Verdine and K. Nakanishi, Science,
1987, 235, 1204; (d) W. Szybalski and V. N. Iyer, Antibiotics I
Mechanism of Action, ed. D. Gottlieb and P. D. Shaw, Springer-
Verlag, New York, 1967, p. 211; (e) H. Kohn, N. Zein, X. Q. Lin,
J. -Q. Ding and K. M. Kadish, J. Am. Chem. Soc., 1987, 109, 1833;
(f) J. -C. Andrez, Beilstein J. Org. Chem., 2009, 5, 33.
3 Mitomycin
C syntheses: (a) T. Fukuyama, F. Nakatsubo,
A. J. Cocuzza and Y. Kishi, Tetrahedron Lett., 1977, 49, 4295;
(b) T. Fukuyama and L. Yang, J. Am. Chem. Soc., 1987, 109, 7881;
(c) T. Fukuyama and L. Yang, J. Am. Chem. Soc., 1989, 111, 8303.
4 Mitomycin K syntheses: (a) J. W. Benbow, G. K. Schulte and
S. J. Danishefsky, Angew. Chem., Int. Ed. Engl., 1992, 31, 915;
(b) J. W. Benbow, K. F. McClure and S. J. Danishefsky, J. Am.
Chem. Soc., 1993, 115, 12305; (c) Z. Wang and L. S. Jimenez,
Tetrahedron Lett., 1996, 37, 6049.
5 Selected synthetic efforts towards mitomycin C: (a) F. E. Ziegler
and M. Y. Berlin, Tetrahedron Lett., 1998, 39, 2455;
(b) R. S. Coleman, F.-X. Felpin and W. Chen, J. Org. Chem.,
2004, 69, 7309; (c) K. J. Shaw, J. R. Luly and H. Rapoport, J. Org.
Chem., 1985, 50, 4515; (d) D. R. Bobeck, D. L. Warner and
E. Vedejs, J. Org. Chem., 2007, 72, 8506.
6 For reviews of structure–activity comparisons of mitomycin
analogues, see: (a) W. A. Remers, The Mitomycins in Anticancer
Agents from Natural Products, D. G. I. Kingston, G. M. Cragg and
D. J. Newman, ed. CRC Press LLC, Boca Raton, Fla, 2005, p. 475;
(b) W. T. Bradner, W. A. Remers and D. M. Vyas, Anticancer Res.,
1989, 9, 1095; (c) H. Arai, Y. Kanda, T. Ashizawa, M. Morimoto,
K. Gomi, M. Kono and M. Kasai, J. Med. Chem., 1994, 37, 1794;
(d) H. D. Beall and S. L. Winski, Front. Biosci., 2000, 5, D639.
7 A. L. Williams, J. M. Srinivasan and J. N. Johnston, Org. Lett.,
2006, 8, 6047.
The mechanism outlined in Scheme 4 for the formation of
11 is consistent with these observations. The initial amino-
mercuration of alkynyl amine 7 and subsequent enamine
addition to quinone 9 would form coupled intermediate 2b.
An E/Z isomerization would bring the pyrrolidine in close
proximity for the 1,6-hydride shift of the diphenylmethyl
methine, resulting in elimination of the bromide and sub-
sequent cyclization of the quinone oxygen onto the iminium
ion. A second hydride addition and tautomerization would
then give N,O-ketal 11.
The cascade that ensues after coupling of the two halves of
the mitomycin backbone was unexpected, but delivers an
advanced intermediate containing all but one of the required
carbons for mitomycin C. The oxidative ketalization provides
differential protection of the hydroquinone hydroxyls while
chemically differentiating the pyrrolidine and aziridine nitrogens.
Equally important is the maintenance of enamine nucleo-
philicity, a feature that will be used to install the final carbon.
Overall, convergency of approach and the Brønsted acid-
catalyzed aza-Darzens reaction of propargyl imine 4 combine
to deliver N,O-ketal 11 in 8 steps.
8 (a) J. C. Antilla and W. D. Wulff, J. Am. Chem. Soc., 1999, 121,
5099; (b) J. C. Antilla and W. D. Wulff, Angew. Chem., Int. Ed.,
2000, 39, 4518.
9 A. L. Williams and J. N. Johnston, J. Am. Chem. Soc., 2004, 126,
1612.
10 For a review of reports demonstrating the compatibility of
diazoalkanes with Brønsted acids, see: J. N. Johnston and
H. Muchalski, T. Troyer. Angew. Chem. Int. Ed., 2010, 49, 2290.
11 J. R. Hwu and P. S. Furth, J. Am. Chem. Soc., 1989, 111, 8834.
12 Aminomercuration lead references: (a) J. Barluenga, F. Aznar,
R. Liz and R. Rodes, J. Chem. Soc., Perkin Trans. 1, 1980, 2732;
(b) J. Barluenga, F. Aznar and R. Liz, Synthesis, 1984, 304;
(c) J. Barluenga, F. Aznar, R. Liz and M.-P. Cabal, Synthesis,
1986, 26, 960.
13 Quinone regioselectivity studies lead references: (a) A. L. Cox and
J. N. Johnston, Org. Lett., 2001, 3, 3695; (b) J. R. Luly and
H. Rapoport, J. Org. Chem., 1981, 46, 2745; (c) K. W. Stagliano,
Z. Lu, A. Emadi, J. S. Harwood and C. A. Harwood, J. Org.
Chem., 2004, 69, 5128; (d) T. Ling, E. Poupon, E. J. Rueden and
E. A. Theodorakis, Org. Lett., 2002, 4, 819; (e) T. Ling, E. Poupon,
E. J. Rueden, S. H. Kim and E. A. Theodorakis, J. Am. Chem.
Soc., 2002, 124, 12261.
Financial support provided by the NIH (GM 063557) is
gratefully acknowledged.
Notes and references
1 (a) T. Hata, R. Sano, A. Sugawara, K. Matsume, T. Shima and
T. Hoshi, J. Antibiot., 1956, 9, 141; (b) D. V. Lefemine, M. Dann,
F. Barbatschi, W. K. Hausmann, V. Zbinovsky, P. Monnikendam,
J. Adam and N. Bohonos, J. Am. Chem. Soc., 1962, 84, 3184;
(c) J. S. Webb, D. B. Cosulich, J. H. Mowat, J. B. Patrick,
R. W. Broschard, W. E. Meyer, R. P. Williams, C. F. Wolf,
W. Fulmor, C. Pidacks and J. E. Lancaster, J. Am. Chem. Soc.,
1962, 84, 3185; (d) A. Tulinsky and J. H. van den Hende, J. Am.
Chem. Soc., 1967, 89, 2905; (e) R. Yahashi and I. Matsubara,
J. Antibiot., 1976, 29, 104; (f) U. Hornemann and M. J. Heins,
J. Org. Chem., 1985, 50, 1301; (g) K. Shirahata and N. Hirayama,
J. Am. Chem. Soc., 1983, 105, 7100.
14 J. R. Luly and H. Rapoport, J. Am. Chem. Soc., 1983, 105, 2859.
15 A crossover experiment was not performed to determine whether
this transfer is strictly intramolecular.
2 (a) G. S. Kumar, R. Lipman, J. Cummings and M. Tomasz,
Biochemistry, 1997, 36, 14128; (b) A. C. Sartorelli, Biochem.
Pharmacol., 1986, 35, 67; (c) M. Tomasz, R. Lipman,
c
This journal is The Royal Society of Chemistry 2011
Chem. Commun., 2011, 47, 3975–3977 3977