Synthesis of the C4- epi -lomaiviticin B core reveals subtle stereoelectronic effects

Article abstract of DOI:10.1021/ol400832r

An efficient synthesis of the C4-epi-lomaiviticin B core is reported. The synthesis features a diastereoselective anionic formal furan Diels-Alder reaction and a stereoselective oxidative enolate dimerization. During the investigation, subtle yet critical stereoelectronic effects imparted by the C4-stereocenter were observed.

Full text of DOI:10.1021/ol400832r

ORGANIC
LETTERS
XXXX
Vol. XX, No. XX
000–000
Synthesis of the C4-Epi-Lomaiviticin B
Core Reveals Subtle Stereoelectronic
Effects
Amy S. Lee and Matthew D. Shair*
Department of Chemistry and Chemical Biology, Harvard University, 12 Oxford Street,
Cambridge, Massachusetts 02138, United States
shair@chemistry.harvard.edu
Received March 26, 2013
ABSTRACT
An efficient synthesis of the C4-epi-lomaiviticin B core is reported. The synthesis features a diastereoselective anionic formal furan DielsꢀAlder
reaction and a stereoselective oxidative enolate dimerization. During the investigation, subtle yet critical stereoelectronic effects imparted by the
C4-stereocenter were observed.
The lomaiviticins comprise a family of type-II polyketide
natural products with remarkable C2-symmetric structures
(Figure 1). They were isolated from a strain of actinomycetes,
Micromonospora lomaivitiensis, and exhibit an array of bio-
logical activity.1 Lomaiviticin A (1) is potently cytotoxic
toward 24 cultured cancer cell lines with IC50 values ranging
from 0.007 to 72 nM. Furthermore, both 1 and lomaiviticin
B (2) are antibiotics against Gram-positive bacteria.
The structural complexity of 1 and 2 poses several
synthetic challenges. The highly oxidized carbon skeleton
includes up to four 2-deoxyglycosides, and the central
C2ꢀC20 bond links two densely functionalized halves to
generate up to eight contiguous stereocenters. Each mono-
meric half possesses an unusual diazofluorene system, a
naphthazarin, and a β-alkoxyenone subunit.2 Altogether,
these features render the lomaiviticins challenging targets.
Figure 1. Structures of lomaiviticin A (1) and B (2).
(1) (a) He, H.; Ding, W. D.; Bernan, V. S.; Richardson, A. D.;
Ireland, C. M.; Greenstein, M.; Ellestad, G. A.; Carter, G. T. J. Am.
Chem. Soc. 2001, 123, 5362–5363. (b) Woo, C. M.; Beizer, N. E.; Janso,
Indeed, despite efforts by various groups,3,4 only one synthesis
J. E.; Herzon, S. B. J. Am. Chem. Soc. 2012, 134, 15285–15288.
of the aglycon (3, Scheme 1) has been accomplished to date.5
(2) β-Alkoxyenone may undergo an eliminationꢀaromatization
Our retrosynthetic analysis of the lomaiviticin aglycon
pathway.
(3) (a) Nicolaou, K. C.; Denton, R. M.; Lenzen, A.; Edmonds, D. J.;
(3) is outlined in Scheme 1. We envisioned that the
Li, A.; Milburn, R. R.; Harrison, S. T. Angew. Chem., Int. Ed. 2006, 45,
2076–2081. (b) Nicolaou, K. C.; Nold, A. L.; Li, H. Angew. Chem., Int.
Ed. 2009, 48, 5860–5863. (c) Zhang, W.; Baranczak, A.; Sulikowski,
G. A. Org. Lett. 2008, 10, 1939–1941. (d) Feldman, K. S.; Selfridge, B. R.
Org. Lett. 2012, 14, 5484–5487.
(4) (a) Krygowski, E. S.; Murphy-Benenato, K.; Shair, M. D. Angew.
Chem., Int. Ed. 2008, 47, 1680–1684. (b) Lee, H. G.; Ahn, J. Y.; Lee,
A. S.; Shair, M. D. Chem. Eur. J. 2010, 16, 13058–13062.
r
10.1021/ol400832r
XXXX American Chemical Society

π-system with the antibonding σ*-orbital of the bridging
CꢀO bond (6, Scheme 1).9 An exocyclic enolate (C5ꢀC4a)
of 5, however, would have greater conformational flexibil-
ity and can thus achieve requisite orbital overlap for
β-elimination.10 We exploited this stereoelectronic dichotomy
of the oxanorbornanone system successfully in our initial
reported studies4a and achieved stereoselective oxidative
enolate dimerization without β-elimination. Furthermore,
we were able to subsequently fragment the oxygen
bridge of 9 (Figure 2). However, displacement of the
C4-phenylsulfone with various oxygen nucleophiles was
low-yielding and the corresponding product could not be
converted to the lomaiviticin core from the cyclic hydrate.
Unexpectedly, in our most recently published work,4b
we were unable to fragment the oxygen bridge of the
full aglycon carbon skeleton 10 (Figure 2). We attributed
the unsuccessful fragmentation to the additional strain and
rigidity that the AB/A0B0-naphthalenes introduce to the
system. Predicated on the two aforementioned observa-
tions, we embarked on a new route to synthesize a sub-
strate with the correct C4-stereochemistry (Scheme 1, 5)
and planned to introduce the naphthalene system after
oxygen bridge fragmentation.
Scheme 1. Retrosynthetic Analysis of Lomaiviticin Aglycon (3)
AB/A0B0-naphthalenes of 3 could be constructed from
lomaiviticin B core 4 via a late-stage cyclopentadienyl
anion bisannulation.6 4 could arise from oxanorbornanone
dimer 5, where the key C2ꢀC20 bond would be established
through a stereoselective oxidative enolate dimerization
of monomer 6. The oxanorbornanone system could be
constructed from an intramolecular exo-selective furan
DielsꢀAlder reaction of a suitable precursor.
Scheme 2. Intramolecular Exo-Selective Furan DielsꢀAlder
Reactiona
Figure 2. Previous work from the Shair group.7,8
Arguably, the stereoselective construction of the key
C2ꢀC20 bond presents the most formidable obstacle in
designing a route toward the lomaiviticins. Constructing
the C2ꢀC20 bond via a dimerization strategy is compli-
cated by two critical challenges: (1) a high potential for
β-elimination of the C3-alkoxy group from a C1ꢀC2
enolate and (2) difficulty in achieving stereoselective for-
mation of the C2/C20-stereocenters. To address this, our
lab has developed a strategy utilizing the oxidative enolate
dimerization of an oxanorbornanone system.4 Stereoselec-
tive formation of the C2-stereocenter is controlled by
dimerization of the enolate from the less hindered convex
exo faceof the oxanorbornanone. Intheoxanorbornanone
system, β-elimination is prevented due to the nearly ortho-
gonal orientation of the endocyclic enolate (C1ꢀC2)
a Conditions: (a) TBSCl, Imid, CH2Cl2, 0°C, 94%; (b) BnOC(NH)CCl3,
cat. TfOH, Et2O, rt, 81%; (c) LiAlH4, THF, ꢀ40 °C, 94%; (d) I2, PPh3,
Imid, CH2Cl2, rt, 88%; (e) n-BuLi, THF, ꢀ78 f 0 °C; 12, ꢀ78 f 0 °C;
(f) TBAF, THF, 0 °C f rt, 93% (2 steps); (g) DMSO, (COCl)2, CH2Cl2,
ꢀ78 °C; i-Pr2NEt, ꢀ78 f ꢀ20 °C; (h) NaI, DBU, THF, 0 °C; then 15,
ꢀ78 °C, 74% (2 steps, 12.5:1 Z:E); (i) LDA, THF, ꢀ78 °C, 64%.
The synthesis commenced with the selective TBS-
protection of the primary hydroxyl of diol 11, which can
be readily synthesized in two steps from (S)-malic acid
(Scheme 2).11 Benzyl protection of the secondary hydroxyl
(5) (a) Herzon, S. B.; Lu, L.; Woo, C. M.; Gholap, S. L. J. Am. Chem.
Soc. 2011, 133, 7260–7263. (b) Woo, C. M.; Gholap, S. L.; Lu, L.;
Kaneko, M.; Li, Z.; Ravikumar, P. C.; Herzon, S. B. J. Am. Chem. Soc.
2012, 134, 17262–17273.
(9) (a) Tojo, S.; Isobe, M. Synthesis 2005, 1237–1244. (b) Jackson,
S. R.; Johnson, M. G.; Mikami, M.; Shiokawa, S.; Carreira, E. M.
Angew. Chem., Int. Ed. 2001, 40, 2694–2697. (c) Vogel, P.; Cossy, J.;
Plumet, J.; Arjona, O. Tetrahedron 1999, 55, 13521–13642.
(6) Birman, V. B.; Zhao, Z.; Guo, L. Org. Lett. 2007, 9, 1223–1225.
(7) Lee, H. G. Ph.D. Thesis, Harvard University, 2012.
(8) Compounds 9 and 10 have stereochemistry corresponding to the
enantiomer of the natural product.
(10) (a) McMorris, T. C.; Staake, M. D.; Kelner, M. J. J. Org. Chem.
2004, 69, 619–623. (b) Jung, M. E.; Min, S. J. Tetrahedron 2007, 63,
3682–3701. (c) Krygowski, E. S. Ph.D. Thesis, Harvard University,
2008. See Supporting Information (Figure 1) for more details.
B
Org. Lett., Vol. XX, No. XX, XXXX

Scheme 3. Stereoselective Oxidative Enolate Dimerization and Attempted Oxygen Bridge Fragmentationa
a Conditions: (a) 10 mol % Pd(PPh3)4, morpholine, THF, rt; (b) (COCl)2, cat. DMF; p-NPBA, Py, cat. 4-DMAP, CH2Cl2, 0 °C; (c) PhH, 80 °C;
(d) K2CO3, MeOH, 0 °C, 38% (4 steps); (e) TBSOTf, i-Pr2NEt, CH2Cl2, 0 °C f rt, 80%; (f) LiHMDS, HMPA, THF, ꢀ78 °C; then [Cp2Fe]PF6, ꢀ78 f
ꢀ55 °C, 4 d, 44%; (g) 2-tert-butyl-1,1,3,3-tetramethylguanidine (Barton’s base), MeCN, ꢀ5 °C; (h) Pd(OH)2/C, H2, THF, rt, quant (2 steps); (i) SO3•Py,
i-Pr2NEt, DMSO, CH2Cl2, ꢀ10 °C.
group, reduction of the methyl ester, and formation of the
corresponding iodide occurred smoothly to yield alkyl
iodide 12. Lithiation of furan 13, followed by alkylation
with 12, afforded coupled product 14, which upon global
silyl deprotection with TBAF, provided the corresponding
furanone alcohol (93%, two steps). Swern oxidation then
cleanly delivered aldehyde 15.
With 17 in hand, an oxidative “carboxy-inversion”
sequence for converting the C4-ester to a hydroxyl with
retention of configuration was required (Scheme 3).15
First, the allyl ester of DielsꢀAlder product 17 was readily
deprotected to carboxylic acid 18. Next, p-nitroperbenzoic
acid (p-NPBA) was coupled to carboxylic acid 18 via the
acid chloride intermediate to afford crude diacyl peroxide
19, which underwent an ionic rearrangement (“carboxy-
inversion”) upon heating to afford the corresponding acyl
carbonate species. Methanolysis of this crude intermediate
provided the desired secondary carbinol 20 as a single
diastereomer (38%, four steps). Protection of 20 as a TBS
ether yielded the dimerization precursor, monomer 21.
Utilizing the optimal oxidative enolate dimerization con-
ditions developed in our group, ketone 21 was added to
LiHMDS and HMPA in THF at ꢀ78 °C to generate the
corresponding lithium enolate, which was then exposed
to [Cp2Fe]PF6 and allowed to stir at ꢀ55 °C for 4 days.
Contrary to our prior studies where only exo-exo dimer-
ization was observed,4 exo-endo dimer 22 was obtained
exclusively (44%). It appears that although dimerization
occurs with complete exo facial selectivity in the absence
of any substitution on the oxanorbornanone carbon fra-
mework, the C4-substituent (pseudoaxial sulfone) plays
a crucial role in reinforcing the exo-exo selectivity in our
prior more complex polycyclic systems.4 Fortunately, 22
could be selectively epimerized to the exo-exo dimer 23
by treatment with Barton’s base.
Aldehyde 15 was then converted to (Z)-enoate 16 (12.5:1
Z/E), the DielsꢀAlder substrate, via a Z-selectivemodified
HornerꢀEmmons reaction.12 We anticipated that the
stereoselectivity of the exo-selective13 intramolecular furan
DielsꢀAlder reaction would be controlled by the single
C5-stereocenter, which enforces a conformation where
1,3-allylic strain is minimized. Initial attempts to promote
the DielsꢀAlder reaction by conventional thermal and
Lewis acidic conditions failed to provide the desired
cycloadduct in synthetically useful yields. We rationalized
that tautomerization of the furanone to the requisite furan
may be slow and that basic conditions may therefore
promote the desired transformation. We discovered that
treatment of 16 with LDA14 provided the DielsꢀAlder
product 17 in 64% yield as a 10:1 mixture of separable
diastereomers, favoring the expectedcis-5,5 fusion product.
This process presumably occurs via a stepwise Michaelꢀ
Michael reaction sequence. We were able to prepare over
10 g of 17 using this protocol.
(11) (a) Gaunt, M. J.; Jessiman, A. S.; Orsini, P.; Tanner, H. R.;
Hook, D. F.; Ley, S. V. Org. Lett. 2003, 5, 4819–4822. (b) Saito, S.;
Ishikawa, T.; Kuroda, A.; Koga, K.; Moriwake, T. Tetrahedron 1992,
48, 4067–4086.
(12) Ando, K.; Oishi, T.; Hirama, M.; Ohno, H.; Ibuka, T. J. Org.
Chem. 2000, 65, 4745–4749.
(13) The exo transition state, which results in cis-5,5 fusion product
17, should be favored over the endo transition state, which would result
in a highly strained trans-5,5 fusion product.
(14) Caine, D. S.; Paige, M. A. Synlett 1999, 9, 1391–1394.
(15) (a) Denney, D. B.; Sherman, N. J. Org. Chem. 1965, 30, 3760–
3761. (b) Fujimori, K.; Shigeru, O. J. Chem. Soc., Perkin Trans. 2 1989,
1335–1348. (c) Meng, Z.; Danishefsky, S. J. Angew. Chem., Int. Ed. 2005,
44, 1511–1513.
With exo-exo dimer 23 in hand, the benzyl ethers were
cleaved and the corresponding diol 24 was oxidized to afford
1,4-diketone dimer 25. Fortuitously, 25 did not exist as
a cyclic hydrate if silica gel column chromatography was
avoided, which is in contrast to our previous systems4
(Figure 2) where the analogous C5-ketone substrates ex-
isted exclusively as the cyclic hydrates. This suggests that
the C4-stereochemistry has subtle yet far-reaching stereo-
electronic consequences on the system. Unfortunately, all
Org. Lett., Vol. XX, No. XX, XXXX
C

attempts to fragment the oxygen bridge were unsuccessful.
In all cases, either nonspecific decomposition or no reac-
tion was observed.16 Surprisingly, deuterium incorpora-
tion studies (KOD in THF) revealed 100% deuterium
incorporation at C4a of 25. Unlike the original successful
model studies where the dimer contained a pseudoaxial
C4-phenylsulfone (Figure 2, 9), we rationalized that frag-
mentation is disfavored due to a 1,3-allylic interaction
between the enolate (C5ꢀC4a) oxygen of 25 and the
pseudoequatorial C4-TBS-ether that must occur during
the transition state in order to achieve proper orbital
overlap for fragmentation to occur. Unfortunately,
moving to a sterically smaller protecting group (MOM)
and even the free hydroxyl did not remedy this problem.
In line with our hypothesis, we rationalized that the
C4-epimer of 25 would not suffer from an unfavorable
1,3-allylic-type interaction during oxygen bridge fragmen-
tation. To this end, 18 was subjected to a one-pot Barton
radical decarboxylationꢀoxidation reaction (Scheme 4).
First, the Barton ester was formed from carboxylic acid
18 by using S-(1-oxido-2-pyridinyl) 1,1,3,3-tetramethyl-
thiouronium hexafluorophosphate (HOTT).17 Upon
complete formation of the Barton ester, the reaction
was saturated with O2 and Sb(SPh)3 was added.18 This
protocol afforded alcohol 27 as a 2:1 mixture of diaster-
eomers, favoring the desired epimer in 46% isolated
yield.19 Protection of 27 as a TBS ether then yielded the
dimerization precursor, monomer 28.
Scheme 4. Successful Oxygen Bridge Fragmentation and
Conversion to C4-Epi-Lomaiviticin A (33) and B (34) Coresa
a Conditions: (a) HOTT, cat. 4-DMAP, Et3N, THF; then Sb(SPh)3,
O2, rt, 46%; (b) TBSOTf, i-Pr2NEt, CH2Cl2, 0 °C f rt, 89%; (c)
LiHMDS, HMPA, THF, ꢀ78 °C; then [Cp2Fe]PF6, ꢀ78 f ꢀ60 °C,
5 d; (d) Pd(OH)2/C, H2, THF, rt; (e) SO3•Py, i-Pr2NEt, DMSO, CH2Cl2,
0 °C f rt, 47ꢀ57% (3 steps); (f) KOH, 3:1 THF/H2O, 0 °C; (g) MgSO4,
PhH, 80 °C; (b) p-TsOH, PhH, rt, 62% (3 steps).
Oxanorbornanone 28 underwent successful oxidative
enolate dimerization to exclusively provide exo-exo di-
mer 29. Benzyl ether cleavage and subsequent oxidation
afforded cyclic hydrate 31, which gratifyingly underwent
successful fragmentation upon treatment with KOH at
0 °C, confirming oursuspicion thattheC4-stereocenterhas
far-reaching stereoelectronic effects. Bisenone product 32
was then converted to C4-epi-lomaiviticin B core 34 in two
steps, involving (1) dehydration of cyclic hydrate 32 in the
presence of MgSO4 to yield C4-epi-lomaiviticin A core
33 and (2) stirring 33 with catalytic p-TsOH to provide
C4-epi-lomaiviticin B core 34 (62%, three steps).
an intramolecular exo-selective furan DielsꢀAlder reac-
tion to construct the oxanorbornanone system, a stereo-
selective oxidative enolate dimerization to establish the key
C2ꢀC20 bond, and successful oxygen bridge fragmenta-
tiontogenerate the lomaiviticin core. Crucial tothe success
of our oxygen bridge fragmentation was the discovery of a
subtle stereoelectronic effect imparted by the C4-stereo-
center, which necessitated the synthesis of a substrate with
the opposite C4-stereochemistry to that found in the
natural product. Current efforts are now focused on the
elaboration of the lomaiviticin B core 34 to the full carbon
skeleton of the aglycon and will be reported in due course.
In summary, we have reported a synthesis of the C4-epi-
lomaiviticin A and B cores, the first time this has been
achieved in our lab. Noteworthy transformations include
Acknowledgment. A.S.L. acknowledges an NDSEG and
an NSF predoctoral fellowship for financial support. A.S.L.
thanks Brian B. Liau (Harvard University) for helpful
discussions.
(16) Even the cyclic hydrate form of 25, which can be readily accessed
by exposure to KOH in H2O/THF, did not undergo the desired
fragmentation.
(17) Garner, P.; Anderson, J. T.; Dey, S. J. Org. Chem. 1998, 63,
5732–5733.
(18) (a) Barton, D. H. R.; Bridon, D.; Zard, S. Z. Tetrahedron 1989,
45, 2615–2626. (b) Zhu, J.; Klunder, A. J. H.; Zwanenburg, B. Tetra-
hedron 1995, 51, 5099–5116.
Supporting Information Available. Experimental pro-
1
cedures, spectroscopic data, and copies of H and 13C
(19) Use of the more conventional Barton radical decarboxylationꢀ
oxidation conditions (Barton ester formation, followed by exposure to
O2, t-BuSH, and a sunlamp, then reduction with PPh3) afforded the
alcohol product in a higher 4:1 dr, but in a somewhat lower isolated yield
of the desired diastereomer (37ꢀ40%).
NMR. This material is available free of charge via the
Internet at http://pubs.acs.org.
The authors declare no competing financial interest.
D
Org. Lett., Vol. XX, No. XX, XXXX