Total synthesis of asperazine [13]

Full text of DOI:10.1021/ja016616l

9468
J. Am. Chem. Soc. 2001, 123, 9468-9469
Scheme 1
Total Synthesis of Asperazine
Steven P. Govek and Larry E. Overman*
Department of Chemistry, 516 Rowland Hall
UniVersity of California, IrVine, California 92697-2025
ReceiVed July 13, 2001
In 1997, Crews and co-workers reported the isolation of
asperazine (1) from a saltwater culture of the fungus Aspergillus
niger obtained from a Caribbean Hyrtios sponge.1 Asperazine is
a member of a large family of diketopiperazine alkaloids that
contain two tryptophan units.2 However, asperazine differs from
other members of this family by having the tryptophan units linked
in an unsymmetrical fashion; this bonding generates the diaryl-
substituted quaternary stereocenter C3. The relative configuration
of asperazine was proposed on the basis of extensive NMR
investigations, whereas the absolute configuration was assigned
when hydrolysis of 1 provided only (R)-phenylalanine.3 Biological
evaluation of asperazine revealed no antibacterial or antifungal
activity but did show asperazine to exhibit significant differential
cytotoxicity toward leukemia in vitro.4 Unfortunately, further
biological studies have not been possible due to the inability to
regrow asperazine-producing A. niger cultures.5 Herein, we report
the first total synthesis of asperazine by an efficient sequence
capable of providing meaningful amounts of asperazine and its
congeners.
Scheme 2a
Because there was some ambiguity concerning the relative
configuration of the diketopiperazine rings,1,3 we developed a
flexible synthetic strategy that would allow for ready access to
asperazine, its diastereomers, and derivatives thereof. The fun-
damental challenge in a stereocontrolled synthesis of asperazine
is forming the quaternary stereocenter C3. Although this challenge
could have been addressed using an asymmetric Heck reaction,6
we chose to develop an alternate approach in which the quaternary
center would evolve from an uncommon diastereoselective Heck
cyclization (3 f 2) (Scheme 1).7 Two significant advantages of
this strategy were envisaged: (1) the enamine (or enamide)
functionality of 2 would provide a good handle for introduction
of the C11 stereocenter, and (2) facial selectivity of the pivotal
Heck cyclization could be ascertained easily from the geometry
of the trisubstituted double bond of 2 because the newly formed
sp2 and sp3 stereocenters (C11 and C3, respectively) would be
correlated by the stereospecifity of the migratory insertion and
â-hydride elimination steps.6 Indole 3 was seen as arising from
(R)-serine and (S)-tryptophan. This general approach, employing
R-amino acid building blocks, would allow access to asperazine
as well as a diversity of analogues.
(1) Varoglu, M.; Corbett, T. H.; Valeriote, F. A.; Crews, P. J. Org. Chem.
1997, 62, 7078-7079.
(2) (a) Hibino, S.; Choshi, T. Nat. Prod. Rept. 2001, 18, 66-87 and earlier
reviews in this series. (b) Anthoni, U.; Christophersen C.; Nielsen, P. H. In
Alkaloids Chemical and Biological PerspectiVes; Pelletier, S. W., Ed.;
Pergamon: London, 1999; Vol. 13, pp 163-236.
a Key:8 (a) n-Bu3SnH, (Ph3P)4Pd, CH2Cl2, -10 °C, 88%; (b) Et3SiH,
TFA, 50 °C; (c) Boc2O, aq Na2CO3, dioxane; (d) NaBH4, LiCl, EtOH,
THF; (e) 2,2-dimethoxypropane, p-TsOH, PhH, 87% (over four steps);
(f) s-BuLi, TMEDA, Et2O, -78 °C; ICH2CH2I, 73%; (g) DDQ, K2CO3,
PhMe, 70 °C, 47% (48% recovered sm); (h) 5, Pd2(dba)3‚CHCl3, (2-
furyl)3P, CuI, NMP; (i) DMSO, 130 °C, 84% (over two steps); (j)
2-iodoaniline, Me3Al, K2CO3, CH2Cl2, 94%; (k) NaH, SEMCl, DMF, 0
°C, 79%.
(3) The absolute configuration of C2, C3, and C11 was assigned on the
basis of the absence of a five-bond (∼1 Hz) coupling constant between C11
and C15. The absolute configuration at C34 could not be assigned by NMR
methods and was postulated on the basis of biogenetic considerations.1
(4) Only 28 of over 20 000 purified organic compounds have shown
leukemia selectivity in the Corbett-Valeriote disk diffusion assay.1
(5) Personal communication to L.E.O. from Professor Phillip Crews.
(6) (a) Donde, Y.; Overman, L. E. In Catalytic Asymmetric Synthesis, 2nd
ed.; Ojima, I., Ed.; Wiley: New York, 2000; Chapter 8G. (b) Shibasaki, M.
In AdVances in Metal-Organic Chemistry; Liebeskind, L. S., Ed.; JAI:
Greenwich, 1996; pp 119-151.
The synthesis began with the assembly of iodoanilide 11, the
Heck cyclization precursor, utilizing a Stille cross-coupling
reaction to form the demanding C3-C24 bond (Scheme 2). The
stannane-coupling partner 5 was formed as a 16:1 mixture of
regioisomers (R:â) by palladium-catalyzed hydrostannylation9 of
serine-derived ynoate 4.10 Halide-coupling partner 7 was synthe-
(7) Heck reactions in which the stereoinducing element is external to the
ring being formed are uncommon. For two recent examples, see: (a) Overman,
L. E.; Paone, D. V.; Stearns, B. A. J. Am. Chem. Soc. 1999, 121, 7702-
7703. (b) Buezo, N. D.; Mancheno, O. G.; Carretero, J. C. Org. Lett. 2000, 2,
1451-1454.
10.1021/ja016616l CCC: $20.00 © 2001 American Chemical Society
Published on Web 08/31/2001

Communications to the Editor
J. Am. Chem. Soc., Vol. 123, No. 38, 2001 9469
Scheme 3a
a Key:8 (a) Pd2(dba)3‚CHCl3, (2-furyl)3P, PMP, DMA, 90 °C, 66%; (b) Bu4NF, 0.1 mm, 85%; (c) 1000 psi H2, 10% Pd/C, DMF, 95%; (d) 1 N HCl,
dioxane, 50 °C; Boc2O, 1 N NaOH, Bu4NHSO4, CH2Cl2, 59%; (e) LiEt3BH, THF, -78 °C f 0 °C; HCl, 0 °C f rt; (f) KOH, MeOH, 79% (over two
steps); (g) SO3‚pyr, Et3N, DMSO; (h) NaClO2, NaH2PO4, THF, H2O, t-BuOH, 2-methyl-2-butene; (i) (R)-PheOMe‚HCl, HATU, Et3N, CH2Cl2, 65%
(over three steps); (j) HCO2H; (k) 0.7 N AcOH, n-BuOH, 120 °C, 59%.
sized in six steps from (S)-TrpOMe‚HCl. The indole ring of this
starting material was reduced with Et3SiH in warm TFA,11 and
then both nitrogens were acylated with Boc2O. The ester was
subsequently reduced with LiBH4, and the resulting amino alcohol
protected to provide indoline 6. Ortho-lithiation12 of 6 with s-BuLi
and subsequent quenching with 1,2-diiodoethane installed the
iodide in 73% yield. Oxidation to restore the indole was best
accomplished with DDQ in warm toluene. At 52% conversion,
the yield of 7 was 91%, and the recovered indoline starting
material could be recycled.13 Stille coupling of iodide 7 and
stannane 5 proceeded with high efficiency to provide enoate 8.14,15
Selective removal of the bulky Boc-protecting group from the
indole nitrogen of 8 allowed 2-iodoaniline to be incorporated by
Weinreb aminolysis16 to generate anilide 10. Finally, protection
of both the indole and the anilide nitrogens with SEM groups
gave rise to 11.
To complete the synthesis of asperazine (1), we needed
to form the pyrrolidine ring and both diketopiperazines. To
this end, removal of the acetonides of 14 with 1 N HCl
and exhaustive Boc-protection of the resulting product provided
15.18 Low-temperature reduction of 15 with LiEt3BH gave
a mixture of hemiaminals which cyclized to form the desired
pyrrolidine ring upon addition of ethereal HCl. Selective
cleavage of the carbonates yielded diol 16.19 Oxidation of
16 with SO3‚pyridine, NaClO2 oxidation of the resulting
dialdehyde, and finally HATU-mediated coupling of this crude
diacid with (R)-PheOMe‚HCl provided tetrapeptide 17.
Removal of the three Boc groups of 17 with formic acid
and subsequent cyclization to form both diketopiperazines in
refluxing butanol containing acetic acid20 delivered asperazine
(1) in 59% yield from 17.21
In summary, the first total synthesis of asperazine was
accomplished in 22 steps from readily available amino acid
starting materials. This synthesis confirms the structure of
asperazine and provides yet another example of the tremendous
utility of intramolecular Heck reactions for forging highly
congested quaternary carbon centers.
With 11 in hand, we turned to the pivotal Heck reaction
(Scheme 3). After extensive experimentation, it was found that
heating of 11 with 20 mol % Pd2(dba)3‚CHCl3 and 100 mol %
(2-furyl)3P in the presence of excess 1,2,2,6,6-pentamethylpip-
eridine (PMP) provided a single hexacyclic product 12 in 66%
yield. Removal of the SEM-protecting groups17 from 12 gave 13.
1
Acknowledgment. We thank NIH (HL-25854) for financial support,
Glaxo-Wellcome for fellowship support (S.P.G) and Professor Philip
Crews for providing copies of unpublished NMR spectra of natural 1
and valuable discussion. NMR and mass spectra were determined at UCI
with instruments purchased with the assistance of NSF and NIH.
At this point, H NOE analysis revealed the E stereochemistry
of the exocyclic double bond, establishing that carbopalladation
had occurred from the desired face. Survey experiments demon-
strated that hydrogenation of the C11-C12 double bond of
protected oxindole 12 would be extremely difficult. Fortunately,
hydrogenation of 13 could be realized over Pd/C in DMF at 1000
psi H2 to provide oxindole 14 in 95% yield as a 4:1 mixture of
C11 epimers.
Supporting Information Available: Experimental procedures and
characterization data for key transformations (preparation of 8, 9, 12,
14, and 1), copies of 1H and 13C NMR spectra of these compounds, tables
comparing NMR spectra of synthetic and natural 1, and the CD spectrum
of synthetic 1 (PDF). This material is available free of charge via the
Internet at http://pubs.acs.org.
(8) For abbreviations not defined in J. Org. Chem. 2001, 66, 24A, see
Supporting Information.
(9) Zhang, H. X.; Guibe, F.; Balavoine, G. J. Org. Chem. 1990, 55, 1857-
1867.
JA016616L
(10) Reginato, G.; Mordini, A.; Caracciolo, M. J. Org. Chem. 1997, 62,
6187-6192.
(11) Lanzilotti, A. E.; Littell, R.; Fanshawe, W. J.; McKenzie, T. C.; Lovell,
F. M. J. Org. Chem. 1979, 44, 4809-4813.
(18) At this point, the diastereomers could be separated by silica gel
chromatography. The C11 epimer of 15 was isolated in 17% yield.
(19) That the major isomer produced in the hydrogenation step had the
(12) Iwao, M.; Kuraishi, T. Heterocycles 1992, 34, 1031-1038.
(13) At longer reaction times, oxidation at C1 of the indole side chain was
observed.
1
desired configuration at C11 could now be established by H NOE analysis
of 16.
(14) (a) Liebeskind, L. S.; Fengl, R. W. J. Org. Chem. 1990, 55, 5359-
5364. (b) Farina, V.; Krishnan, B. J. Am. Chem. Soc. 1991, 113, 3, 9585-
9595.
(20) Suzuki, K.; Sasaki, Y.; Endo, N.; Mihara, Y. Chem Pharm. Bull. 1981,
29, 233-237.
1
(21) Synthetic asperazine showed H NMR spectra in three solvents, 13C
(15) Stannane 5 was a 16:1 mixture of regioisomers; trace amounts of
â-coupled product were separated after removal of the indole-protecting group.
(16) Lipton, M. F.; Basha, A.; Weinreb, S. M. Org. Synth. 1978, 59, 49-
53.
NMR spectra in two solvents, and mass spectral data that compared favorably
to those of the natural isolate.1 The optical rotation of synthetic asperazine,
[R]D +95.7 (c, 0.2, MeOH) was higher than that reported for the natural
material, [R]D +52 (c, 0.2, MeOH);1 however, CD spectra were nearly
identical.
(17) Moreno, O. F.; Kishi, Y. J. Am. Chem. Soc. 1996, 118, 8180-8181.