Syntheses of xanthofulvin and vinaxanthone, natural products enabling spinal cord regeneration

Article abstract of DOI:10.1002/anie.201205837

Growth promoting: The first synthesis of the natural product xanthofulvin (SM-216289) has resolved issues regarding its structural assignment and that described for the natural product 411J. The structurally and biologically related natural product vinaxanthone was similarly prepared through a novel dimerization reaction. Treatment of C. elegans with synthetic xanthofulvin and vinaxanthone enhanced axonal branching.

Full text of DOI:10.1002/anie.201205837

Angewandte
Chemie
DOI: 10.1002/anie.201205837
Natural Product Synthesis
Syntheses of Xanthofulvin and Vinaxanthone, Natural Products
Enabling Spinal Cord Regeneration**
Abram Axelrod, Anders M. Eliasen, Matthew R. Chin, Katherine Zlotkowski, and
Dionicio Siegel*
The failure of neurons in the central nervous system (CNS) to
undergo regeneration following injury accounts for the
permanent and debilitating effects that accompany spinal
cord injury, for which there is no cure. Gene therapy,
biologics, and stem-cell-based approaches have received
considerable attention in promoting CNS regeneration,
while the use of low-molecular-weight compounds has not
been as extensively investigated.[1] However, in the context of
spinal cord injury small molecules hold considerable potential
for the accelerated development of new therapeutics. The
delivery of drugs directly into the spinal cavity through spinal
injection can expedite small-molecule-based drug develop-
ment. Moreover, a variety of hydrogels and other polymers
for continuous drug delivery, developed specifically for spinal
cord therapy, when coupled with a validated small molecule
will provide a unique and promising platform for therapeutic
development.[2]
The natural product xanthofulvin (1, also named SM-
216289) represents one of the most promising leads in the
development of treatments for spinal cord injury. Xanthoful-
vin (1) and the related compound vinaxanthone (2) were
isolated from fungal extracts of Penicillium sp. SPF-3059
(Scheme 1).[3] Both compounds strongly block the effects of
the inhibitor of axonal regeneration semaphorin3A
(Sema3A) with no observable cytotoxicity at concentrations
above 1000 times the effective dose.[3b,4] Animal studies of
xanthofulvin have demonstrated remarkable effects after
complete spinal cord transection.[5] The dramatically
improved functional recovery observed resulted from signifi-
cant axonal regeneration and myelination, reduction of the
number of apoptotic cells, and enhanced angiogenesis. While
the pronounced effects of xanthofulvin have been attributed
to the inhibition of Sema3A, removal of Sema3A function
does not enhance regeneration after spinal cord injury, thus
suggesting that the natural product functions through a more
complex mode of action than initially described.
Scheme 1. Structures of xanthofulvin (1), vinaxanthone (2), and 411J
(3).
In addition to questions surrounding the modes of action,
the atomic connectivity of xanthofulvin was previously
unclear, since there were two conflicting structures proposed.
The isolation and structural characterization data of the
natural products xanthofulvin (1) and 411J (3) reveal the
structures show nearly identical spectral properties
(Scheme 2).[6] In addition, the keto form 4 of xanthofulvin
Scheme 2. Equilibria of xanthofulvin (1) and 4 and 411J (3) and 5.
and the keto form 5 of 411J also possess overlapping spectral
properties.[7] This observation led us to the conclusion that the
assignments for xanthofulvin and 411J were based on the
same natural product. In both studies the natural product was
co-isolated with vinaxanthone (2) as well. We sought to
resolve these conflicting structural assignments through the
synthesis of xanthofulvin (1), the structure that appeared the
most plausible.
We propose the formation of xanthofulvin can proceed
through the union of the known natural product polivione
with 5,6-dehydropolivione (6), a putative unsaturated deriv-
ative of the known natural product polivione.[8,9] Similarly, the
[*] A. Axelrod, A. M. Eliasen, M. R. Chin, K. Zlotkowski, Dr. D. Siegel
Department of Chemistry and Biochemistry
Norman Hackerman Building
The University of Texas at Austin
1 University Station, Austin, TX 78712 (USA)
E-mail: dsiegel@cm.utexas.edu
Homepage: http://dsiegel.cm.utexas.edu/
[**] We thank Dr. Wrigley for helpful discussions regarding 411J.
Financial support from The University of Texas at Austin, the Welch
Foundation (F-1694), and the NSF (CHE-1151708) are gratefully
acknowledged.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201205837.
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

.
Angewandte
Communications
formation of vinaxanthone can be envi-
sioned to occur through a dimerization of
5,6-dehydropolivione (Scheme 3). These
proposals provided a template for the
development of our synthetic routes. Pre-
vious proposals for the biosyntheses of
xanthofulvin and vinaxanthone have been
postulated to occur through a Knovenagel
sequence[10] or, in the case of vinaxan-
thone, through a Diels–Alder/oxidative
aromatization sequence. The Diels–
Alder/oxidative aromatization sequence
was applied in the sole existing synthesis
of vinxanthone.[11] Previous to this report
there was no laboratory preparation of
xanthofulvin.
The synthesis of 5,6-dehydropolivione
(6) was accomplished in eight steps from
tetronic acid (12; Scheme 4). A regiose-
lective Diels–Alder reaction of furan
13,[12] prepared from tetronic acid (12),
with keto ester 14 (also prepared in as few
as two steps[13]) provided the bicyclic
adduct 15 in good yield and high regiose-
lectivity (> 20:1).[14] The high level of
selectivity was possible owing to the
polarization of the diene in concert with
the ketone functioning as the dominant
activating group. Treatment of Diels–
Alder adduct 15 with hydrochloric acid
induced aromatization, providing aceto-
phenone 17 with migration of the pivaloyl
Scheme 3. Proposed dimerization of 5,6-dehydropolivione (6) generating vinaxanthone (2).
group.[15] The first four steps of this sequence have been
performed without chromatography and on scales above 50 g.
Protection of phenol 17 was required prior to chromone-
triketone formation and was accomplished using methoxy-
methyl chloride and Hꢀnigꢁs base. Enaminone formation by
using the dimethyl acetal of dimethylformamide in dimeth-
oxyethane at elevated temperature occurred with simulta-
neous cleavage of the TBS group, thus generating enaminone
19. Conversion of enaminone 19 to triketone 21 was achieved
by treatment of 19 with 5-acetyl Meldrumꢁs acid (20) in
toluene heated to reflux, yielding protected 5,6-dehydropo-
livione (21) in 42% yield.[16] Simultaneous cleavage of the tert-
butyl ester, pivaloyl, and methoxymethyl groups with boron
trichloride in dichloromethane at 238C provided 5,6-dehy-
dropolivione (6) in 96% yield.
In agreement with our proposal, heating 5,6-dehydropo-
livione (6) in deionized water to 558C provided vinaxanthone
(2) in 61% yield (Scheme 5). Thus, vinaxanthone was
synthesized in nine steps from tetronic acid, whereas the
previous synthesis required 14 steps to provide the natural
product.
A related approach to access xanthofulvin failed owing to
the facile dimerization of 5,6-dehydropolivione (6) forming
vinaxanthone. Therefore, an ynone was envisioned as a surro-
gate for 5,6-dehydropolivione; this ynone would function only
as a Michael-acceptor (Scheme 6).[17a] Related reactivity of 3-
(1-alkynyl)chromones was previously described by Hu and
Scheme 4. Synthesis of 5,6-dehydropolivione (6) from tetronic acid
(12). Piv=pivaloyl, DMAP=4-(dimethylamino)pyridine, TBS=tert-
butyldimethylsilyl, MOM=methoxymethyl, DME=1,2-dimethoxy-
ethane.
co-workers.[17b] Michael addition into the chromone portion of
ynone 22 and bond fragmentation provides phenol 25 that is
poised to undergo conjugate addition and, after alkene
2
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
These are not the final page numbers!

Angewandte
Chemie
The major by-product of the reaction (14%) was the
derivative of 33 lacking the acetyl group. Saponification of
the methyl ester with sodium hydroxide in a 3:1 tetrahydro-
furan/water mixture yielded the corresponding carboxylic
acid 34. After optimization, coupling of carboxylic acid 34 and
enaminone 19 using HBTU and Hꢀnigꢁs base in dimethyl-
formamide at 238C generated diketone 36 in 88% yield.[21]
Existing methods for coupling ortho-hydroxy aryl enami-
nones with carboxylic acid derivatives to initiate O-to-C
carboxyl transfers employ anhydrides and other activated
carboxylic derivatives.[22] The strategy of using coupling
reagents proved more direct and tolerant of sensitive
functionality. Elimination of dimethylamine using pyridinium
chloride in acetonitrile generated endione 37 that underwent
conjugate reduction with sodium cyanoborohydride yielding
xanthofulvin in protected form.
Scheme 5. Dimerization of 5,6-dehydropolivione (6) under neutral
conditions to generate vinaxanthone (2).
Simultaneous removal of all six oxygen-bound groups on
the phenols and carboxylic acids was accomplished with
boron trichloride in dichloromethane at 238C providing
material that matched all the reported spectral values for
xanthofulvin (1) and 411J (3; Scheme 8). Synthesis has
confirmed the structural assignment as that described for
xanthofulvin (1), thus indicating the structural reassignment
of 411J (3) to that of xanthofulvin (1) is required.
Through our in vivo outgrowth assay xanthofulvin and
vinaxanthone were found to enhance neuronal outgrowth.
This assay utilizes green fluorescent protein (GFP)-labeled
Scheme 6. Coupling of ynone 22 with methyl acetoacetate (23).
isomerization, generate adduct 27. The tetracarbonyl-con-
taining portion of adduct 27 can tautomerize and react
through a 6p electrocyclization and subsequent dehydration
as described in the biosynthetic proposal (Scheme 3), provid-
ing the xanthone core of xanthofulvin (28).
Reaction of the previously prepared enaminone 19 with
iodine in chloroform at 238C generated iodochromone 29
(Scheme 7).[18] Sonogashira coupling of iodochromone 29
with 3-butyn-2-ol (30)[19] and subsequent oxidation[20] yielded
ynone 32 as a solid (m.p. 178–1798C).
As proposed in Scheme 6 addition of the sodium anion of
methylacetoacetate into ynone 32 proceeded to generate the
desired pentasubstituted arene 33 in 83% yield (Scheme 8).
Scheme 8. Completion of the synthesis of xanthofulvin (1). HBTU=O-
(1H-benzotriazoyl-1-yl)-N,N,N’,N,’-tetramethyluronium hexafluorophos-
phate.
Scheme 7. Synthesis of ynone 32. PDC=pyridinium dichromate.
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

.
Angewandte
Communications
199; b) J. Baier Leach, K. A. Bivens, C. W. Patrick, Jr., C. E.
Schmidt, Biotechnol. Bioeng. 2003, 82, 578 – 589; c) J. Struve,
P. C. Maher, Y. Q. Li, S. Kinney, M. G. Fehlings, C. t. Kuntz, L. S.
Sherman, Glia 2005, 52, 16 – 24; d) T. Gros, J. S. Sakamoto, A.
Blesch, L. A. Havton, M. H. Tuszynski, Biomaterials 2010, 31,
6719 – 6729.
cholinergic neurons in C. elegans to visualize enhanced
branching after treatment with neurotrophic compounds.[23]
Similar to other known neurotrophic compounds, xanthoful-
vin and vinaxanthone were found to increase the rate of
branching in C. elegans with 32% and 31% of the treated
animals displaying outgrowth at 2 mm respectively. This
activity is comparable to dibutyryl cAMP,[24] which promotes
branching in 36% of animals at 2 mm (Figure 1).
[3] a) M. Aoki, Y. Itezono, H. Shirai, N. Nakayama, A. Sakai, Y.
Tanaka, A. Yamaguchi, N. Shimma, K. Yokose, H. Seto,
Tetrahedron Lett. 1991, 32, 4737 – 4740; b) K. Kumagai, N.
Hosotani, K. Kikuchi, T. Kimura, I. Saji, J. Antibiot. 2003, 56,
610 – 616.
[4] K. K. Kumagai, K. Kishino, A. Hosotani, N. Ito, A. Saji, I.
Kimura, Sumitomo Kagaku 2005, 1 – 8.
[5] S. Kaneko, A. Iwanami, M. Nakamura, A. Kishino, K. Kikuchi,
S. Shibata, H. J. Okano, T. Ikegami, A. Moriya, O. Konishi, C.
Nakayama, K. Kumagai, T. Kimura, Y. Sato, Y. Goshima, M.
Taniguchi, M. Ito, Z. He, Y. Toyama, H. Okano, Nat. Med. 2006,
12, 1380 – 1389.
[6] S. K. Wrigley, M. A. Latif, T. M. Gibson, M. I. Chicarellirobin-
son, D. H. Williams, Pure Appl. Chem. 1994, 66, 2383 – 2386.
[7] See the Supporting Information for a comparison of the spectra
of xanthofulvin and 411J.
[8] a) A. K. Demetriadou, E. D. Laue, F. J. Leeper, J. Staunton, J.
Chem. Soc. Perkin Trans. 1 1988, 763 – 768; b) A. K. Demetria-
dou, E. D. Laue, J. Staunton, J. Chem. Soc. Perkin Trans. 1 1988,
769 – 772; c) A. K. Demetriadou, E. D. Laue, F. J. Leeper, J.
Staunton, J. Chem. Soc. Chem. Commun. 1985, 762 – 764;
d) A. K. Demetriadou, E. D. Laue, J. Staunton, J. Chem. Soc.
Chem. Commun. 1985, 764 – 766.
Figure 1. Outgrowth of GFP-labled cholinergic neurons in vivo in
C. elegans after treatment with dibutyryl cAMP, xanthofulvin, and
vinaxanthone. Control: 0.2% DMSO in M9 buffer.
[9] See the Supporting Information for full schemes for the
proposed dimerization of 5,6-dehydropolivione and the reaction
of polivione with dehydropolivione.
[10] S. L çsgen, O. Schlçrke, K. Meindl, R. Herbst-Irmer, A. Zeeck,
Eur. J. Org. Chem. 2007, 2191 – 2196.
[11] K. Tatsuta, S. Kasai, Y. Amano, T. Yamaguchi, M. Seki, S.
Hosokawa, Chem. Lett. 2007, 36, 10.
[12] T. M. Balthazor, E. L. Williams, Synth. Commun. 1992, 22, 1023 –
1026.
[13] L. A. M. Cornelius, R. G. A. Bone, R. H. Hastings, M. A.
Deardorff, R. A. Scharlach, B. E. Hauptmann, C. J. Stankovic,
H. W. Pinnick, J. Org. Chem. 1993, 58, 4774 – 4774.
[14] A. Gorgues, A. Simon, A. Lecoq, A. Hercouet, F. Corre,
Tetrahedron 1986, 42, 351 – 370.
[15] T. R. Kelly, S. H. Bell, N. Ohashi, R. J. Armstrong-Chong, J. Am.
Chem. Soc. 1988, 110, 6471 – 6480.
[16] a) Y. S. Oikawa, K. Sugano, O. Yonemitsu, J. Org. Chem. 1978,
43, 2087 – 2088; b) F. Xu, J. D. Armstrong, G. X. Zhou, B.
Simmons, D. Hughes, Z. H. Ge, E. J. J. Grabowski, J. Am. Chem.
Soc. 2004, 126, 13002 – 13009.
The synthesis of xanthofulvin has provided laboratory
access to this promising regenerative natural product for the
first time. The second synthesis of vinaxanthone was achieved
and provides the natural product in 2/3 the number of steps
previously achieved. The chemical transformations employed
in the syntheses included: a highly regioselective Diels–Alder
reaction of an ynone ester, tandem reaction sequences for the
formation of vinaxanthone from 5,6-dehydropolivione and
the core of xanthofulvin, and an HBTU-mediated coupling of
carboxylic acid derivative 33 and ortho-hydroxy enaminone
19 with subsequent O-to-C transfer, bringing two functional-
ized fragments together. The validation of the structural
assignment of the natural product as that described for
xanthofulvin has resolved ambiguity regarding the alternative
assignment of 411J (3). Tests of synthetic xanthofulvin and
vinaxanthone in vivo demonstrated the compounds possess
growth-promoting activity. With synthetic access future
efforts to optimize the biological performance of the natural
products and to describe the modes of action will be pursued.
[17] a) F. Eiden, H. Fenner, Chem. Ber. 1968, 101, 2894 – 2895;
b) L. Z. Zhao, F. C. Xie, G. Cheng, Y. H. Hu, Angew. Chem.
2009, 121, 6642 – 6645; Angew. Chem. Int. Ed. 2009, 48, 6520 –
6523.
[18] R. B. Gammill, Synthesis 1979, 901 – 903.
[19] K. Sonogashira, Y. Tohda, N. Hagihara, Tetrahedron Lett. 1975,
16, 4467 – 4470.
Received: July 23, 2012
Revised: October 3, 2012
Published online: && &&, &&&&
[20] E. J. Corey, G. Schmidt, Tetrahedron Lett. 1979, 20, 399 – 402.
[21] V. Dourtoglou, J. C. Ziegler, B. Gross, Tetrahedron Lett. 1978, 19,
1269 – 1272.
[22] I. Yokoe, K. Maruyama, Y. Sugita, T. Harashida, Y. Shirataki,
Chem. Pharm. Bull. 1994, 42, 1697 – 1699.
[23] K. Zlotkowski, D. Siegel, unpublished results..
[24] R. E. Rydel, L. A. Greene, Proc. Natl. Acad. Sci. USA 1988, 85,
1257 – 1261.
Keywords: natural product synthesis · neuronal outgrowth ·
regeneration · structure elucidation · tandem reaction
.
[1] R. M. Wilson, S. J. Danishefsky, Acc. Chem. Res. 2006, 39, 539 –
549.
[2] a) N. N. Madigan, S. McMahon, T. OꢁBrien, M. J. Yaszemski,
A. J. Windebank, Respir. Physiol. Neurobiol. 2009, 169, 183 –
4
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
These are not the final page numbers!

Angewandte
Chemie
Communications
Natural Product Synthesis
Growth promoting: The first synthesis of
the natural product xanthofulvin (SM-
216289) has resolved issues regarding its
structural assignment and that described
for the natural product 411J. The struc-
turally and biologically related natural
product vinaxanthone was similarly pre-
pared through a novel dimerization reac-
tion. Treatment of C. elegans with syn-
thetic xanthofulvin and vinaxanthone
enhanced axonal branching.
A. Axelrod, A. M. Eliasen, M. R. Chin,
K. Zlotkowski, D. Siegel*
&&&& — &&&&
Syntheses of Xanthofulvin and
Vinaxanthone, Natural Products Enabling
Spinal Cord Regeneration
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!