Total synthesis, revised structure, and biological evaluation of biyouyanagin A and analogues thereof

Article abstract of DOI:10.1021/ja802805c

Isolated from Hypericum species H. Chinese L. var. salicifolium, biyouyanagin A was assigned structure 1a or 1b on the basis of NMR spectroscopic analysis. This novel natural product exhibited significant anti-HIV properties and inhibition of lipopolysacc

Full text of DOI:10.1021/ja802805c

Published on Web 07/23/2008
Total Synthesis, Revised Structure, and Biological Evaluation
of Biyouyanagin A and Analogues Thereof
K. C. Nicolaou,*,†,‡, T. Robert Wu,†,O David Sarlah,† David M. Shaw,†
Eric Rowcliffe,§ and Dennis R. Burton§,|
Department of Chemistry, The Skaggs Institute for Chemical Biology, Department of
Immunology and Microbial Science, Department of Molecular Biology, The Scripps Research
Institute, 10550 North Torrey Pines Road, La Jolla, California 92037, and the Department of
Chemistry and Biochemistry, UniVersity of California, San Diego, 9500 Gilman DriVe,
La Jolla, California 92093
Received April 16, 2008; E-mail: kcn@scripps.edu
Abstract: Isolated from Hypericum species H. chinese L. var. salicifolium, biyouyanagin A was assigned
structure 1a or 1b on the basis of NMR spectroscopic analysis. This novel natural product exhibited significant
anti-HIV properties and inhibition of lipopolysaccharide-induced cytokine production. Described herein are
the total syntheses of biyouyanagin A and several analogues (3-11), structural revision of biyouyanagin
A to 2b, and the biological properties of all synthesized compounds. The total synthesis proceeded through
cascade sequences that efficiently produced enantiomerically pure key building blocks 15b (ent-zingiberene)
and 18 (hyperolactone C) and featured a novel [2 + 2] photoinduced cycloaddition reaction which occurred
with complete regio- and stereoselectivity. Biological investigations with the synthesized biyouyangagins A
(2-11) and hyperolactones C (12-16) revealed that the activity of biyouyanagin A most likely resides in
its hyperolactone C structural domain.
Introduction
The Hypericum species H. chinese L. Var. salicifolium has
been used in Japan as a folk medicine for the treatment of female
disorders for centuries.1 Investigations with this plant led to the
discovery of biyouyanagin A, a substance which was originally
assigned structure 1a or 1b (Figure 1) on the basis of NMR
spectroscopic analysis.2 Biyouyanagin A exhibited significant
and selective inhibitory activity against HIV replication in H9
lymphocytes (EC50 ) 0.798 µg mL-1) as compared to nonin-
fected H9 lymphocytes (EC50 >25 µg mL-1), thus demonstrat-
ing a therapeutic index (TI) of greater than 31.3.2 In addition,
this substance inhibited strongly lipopolysaccharide (LPs)-
induced cytokine production at 10 µg mL-1 [IL-10 ) 0.03; IL-
12 ) 0.02; tumor necrosis factor-R (TNF-R) ) 0.48].2 In a
recent communication we reported the total synthesis and
structural revision of biyouyanagin A (2b) and its 24-epimer,
24-epi-biyouyanagin A (2a, Figure 2).3 In this article we report
our detailed investigations in the biyouyanagin field, including
Figure 1. Originally proposed structures for biyouyanagin A.
Figure 2. Revised structures of biyouyanagin A (2b) and 24-epi-
biyouyanagin A (2a).
O Present address: Chemical Synthesis Laboratory@Biopolis, ICES,
A*STAR, Singapore.
an X-ray crystallographic analysis of biyouyanagin A (2b) and
the synthesis and biological evaluation of several analogues
(3-16, Figure 3) of this novel natural product.
† Department of Chemistry, The Scripps Research Institute.
‡ The Skaggs Institute for Chemical Biology, The Scripps Research
Institute.
§ Department of Immunology and Microbial Science, The Scripps
Research Institute.
Results and Discussion
| Department of Molecular Biology, The Scripps Research Institute.
University of California, San Diego.
By virtue of its novel molecular architecture, the remaining
ambiguity shadowing its structure, and its significant biological
properties, biyouyanagin A presented itself as an attractive
synthetic target. Of particular interest was the verification of
the all-cis stereochemistry of the substituents around the
cyclobutane ring, an arrangement that looked rather odd at the
(1) Murakami, K. Tokushima-ken Yakusouzukan 1984, 102–103.
(2) Tanaka, N.; Okasaka, M.; Ishimaru, Y.; Yakaishi, Y.; Sato, M.;
Okamoto, M.; Oshikawa, T.; Ahmed, S. U.; Consentino, L. M.; Lee,
K.-H. Org. Lett. 2005, 7, 2997–2999.
(3) Nicolaou, K. C.; Sarlah, D.; Shaw, D. M. Angew. Chem., Int. Ed. 2007,
46, 4708–4711.
9
11114 J. AM. CHEM. SOC. 2008, 130, 11114–11121
10.1021/ja802805c CCC: $40.75
2008 American Chemical Society

Biyouyanagin A and Analogues
A R T I C L E S
Figure 4. Retrosynthetic analysis of biyouyanagin A and transition states
of the proposed photoinduced [2 + 2] cycloaddition reaction. For simplicity,
only the 24(R) isomers (1b and 2b) are shown.
Retrosynthetic Analysis. While there are myriad ways to
disassemble the biyouyanagin A molecule retrosynthetically, the
one made possible by a retro [2 + 2] cycloaddition reaction
(Figure 4) is both aesthetically and practically most appealing.
In the synthetic direction such a reaction can, in principle, be
realized by irradiation with UV light, although no precedent
existed at the outset of this work for the photoinduced [2 + 2]
cycloaddition of substrates such as the two components defined
by the proposed cyclobutane disconnection (i.e., triene 17b or
its 7-epimer 17a and enone 18, Figure 4). If successful, however,
this approach would consititute a highly convergent strategy
for the total synthesis of the natural product and might also
have implications in its biosynthesis. To be sure, however, this
rather obvious hypothesis had been proposed as a plausible
biosynthetic pathway toward biyouyanagin A by its discoverers.2
In considering such a scenario, inspection of the two transition
states that could lead to the biyouyanagin A molecule (I: exo
and II: endo, Figure 4) was instructive. Thus, formation of the
cis, cis, cis, cis structure (1a or 1b) originally proposed for
biyouyanagin A2 requires the endo transition state II, an
arrangement that suffers from severe steric congestion between
the γ-lactone moiety of the enone and the side chain of the
triene component as demonstrated by manual molecular models
and shown in Figure 4. On the other hand, the alternative
arrangement of the reacting components as shown in the exo
transition state I is free of such unfavorable interactions. This
realization created a suspicion in our minds with regards to the
structure of biyouyanagin A as proposed in the isolation paper.2
Specifically, we began to favor the cis, cis, trans, trans
stereochemistry as shown in structure 2b, although the cloud
of ambiguity over the configuration of the C-24 stereocenter
(see structure 2a) remained. In addition, the NOE interactions
reported for biyouyanagin A,2 in conjunction with manual
molecular models, did not exclude the cis, cis, trans, trans
structure 2b (or 2a), a fact that fueled further our skepticism
about the true structure of the natural product. It was with this
Figure 3. Synthesized biyouyanagin A (3-11) and hyperolactone C
(12-16) analogues.
outset,2 given the steric congestion associated with it. Another
impetus for undertaking the total synthesis of biyouyanagin A
was to advance further the advent of cascade reactions4 and
exploit recent developments in organocatalysis5 for total syn-
thesis purposes.
(4) For a recent review article on cascade reactions in total synthesis,
see: (a) Nicolaou, K. C.; Edmonds, D. J.; Bulger, P. G. Angew. Chem.,
Int. Ed. 2006, 45, 7134–7186.
(5) Berkessel, A.; Gro¨ger, H. Asymmetric Organocatalysis: From Bio-
mimetic Concepts to Applications in Asymmetric Synthesis; Wiley-
VCH, Weinheim, 2006; p 454.
9
J. AM. CHEM. SOC. VOL. 130, NO. 33, 2008 11115

A R T I C L E S
Nicolaou et al.
Scheme 1. Synthesis of Terpenoids 17a and 17ba
Scheme 2. Synthesis of Propargyl Alcohol 26a
a Reagents and conditions: (a) DMP (2.0 equiv), CH2Cl2, 25 °C, 5 h,
92%; (b) acetylene, n-BuLi, THF, -78 °C, 1 h, 92%, 3:1 d.r.; (c) 3,5-
dinitrobenzoyl chloride (1.2 equiv), Et3N (1.2 equiv), 4-DMAP (0.1 equiv),
CH2Cl2, 25 °C, 3 h, 98%. DMP ) Dess-Martin periodinane; 4-DMAP )
4-dimethylaminopyridine.
choice of these particular enantiomeric forms of 17a and 17b
was based on the assumption that the hyperolactone C structural
motif of biyouyanagin A would be of the same absolute
configuration as the naturally occurring hyperolactone C,6 a
postulate that, in turn, suggested the absolute configuration of
the terpenoid fragments as shown.
a Reagents and conditions: (a) 19a or 19b (1.0 equiv), MVK (1.5 equiv),
20 (5 mol%), ethyl 3,4-dihydroxybenzoate (20 mol%), 0 °C, 24 h; then
KOH (0.1 N aq., 1.0 equiv), n-Bu4NOH (40% aq., cat.), Et2O/THF/H2O
(3:1:3), reflux, 6 h, 72% yield, 93% de for 22a; 68% yield, 86% de for
22b; (b) KHMDS (1.5 equiv), THF, -78 °C, 3 h; then Comins reagent
(1.5 equiv), THF, -78 °C, 1 h; (c) MeMgI (3.0 M in Et2O, 1.5 equiv), CuI
(2 mol%), THF, 0 °C, 15 min, 80% over the two steps. MVK ) methyl
vinyl ketone; THF ) tetrahydrofuran; KHMDS ) potassium hexamethyl-
disilazanide; Tf ) trifluoromethanesulfonyl.
The synthesis of the terpenoid fragments 17a (ent-7-epi-
zingiberene) and 17b (ent-zingiberene) begun from (S)-citronel-
lal (19a) and (R)-citronellal (19b), respectively, and proceeded
according to the sequence depicted in Scheme 1. Thus, enamine-
mediated 1,4-addition of (S)-citronellal (19a) to methyl vinyl
ketone (MVK) facilitated by the proline-derived catalyst 20 (5
mol%)8 and ethyl 3,4-dihydroxybenzoate (21) as a cocatalyst
(20 mol%)9 furnished, after an intramolecular aldol condensation
(KOH, n-Bu4NOH cat.) of the resulting ketoaldehyde product,
24(S) enone 22a in 72% overall yield and 93% de. The
conversion of the latter compound to the desired building block
17a (ent-7-epi-zingiberene) proceeded smoothly and in 80%
overall yield through a Kumada coupling of the corresponding
vinyl triflate (KHMDS, Comins reagent)10 with MeMgI (CuI
cat.).11 A similar route starting from (R)-citronellal (19b) and
MVK, and proceeding through intermediate enone 22b (68%
yield, and 86% de), furnished ent-zingiberene (17b), the other
targeted terpenoid fragment (80% overall yield for the last two
steps).
background that we embarked on the synthetic journey to
biyouyanagin A, whose true structure became our immediate
puzzle to solve.
Construction of Building Blocks for Biyouyanagin A. Having
defined the required building blocks for the devised [2 + 2]
photocycloaddition strategy toward biyouyanagin A, their
construction evolved as shown in Scheme 1 (17a: ent-7-epi-
zingiberene and 17b: ent-zingiberene) and Schemes 2 and 3
(18: hyperolactone C). Interestingly, hyperolactone C (18) is
already known in the literature as a naturally occurring sub-
stance,6 while 17a and 17b are enantiomers of natural terpenes
(7-epi-zingiberene and zingiberene, respectively).7 The initial
The synthesis of hyperolactone C (18, Scheme 2)6e began
with (S)-malic acid (23) and proceeded, sequentially, through
lactone 26 (Scheme 2) and spirolactone intermediate 27 (Scheme
4). As shown in Scheme 2, (S)-malic acid was converted in
five steps, and by literature procedures,6e to hydroxy lactone
24, which was oxidized with DMP to ketolactone 25 in 92%
yield. The latter compound reacted with lithium acetylide
(generated from acetylene and n-BuLi in THF at -78 °C) to
afford, stereoselectively, propargylic alcohol 26 (92% yield, ca.
(6) For selected papers on hyperolactone C, see: (a) Isolation and
structure: Aramaki, Y.; Chiba, K.; Tada, M. Phytochemistry 1995,
38, 1419–1421. (b) X-ray structure: Crockett, S. L.; Schuhly, W.; Belaj,
F.; Khan, I. A. Acta Crystallogr., Sect. E 2004, 60, o2174-o2176.
(c) Synthesis: Ichinari, D.; Ueki, T.; Yoshihara, K.; Kinoshita, T. Chem.
Commun. 1997, 1743–1743. (d) Ueki, T.; Ichinari, D.; Yoshihara, K.;
Morimoto, Y.; Kinoshita, T. Tetrahedron Lett. 1998, 39, 667–668.
(e) Ueki, T.; Doe, M.; Tanaka, R.; Morimoto, Y.; Yoshihara, K.;
Kinoshita, T. J. Heterocycl. Chem. 2001, 38, 165–172. (f) Kraus, G. A.;
Wei, J. J. Nat. Prod. 2004, 67, 1039–1040.
(7) For selected papers on zingiberenes, see: (a) Eschenmoser, A.; Schinz,
H. HelV. Chim. Acta 1950, 33, 171–177. (b) Arigoni, D.; Jeger, O.
HelV. Chim. Acta 1954, 37, 881–883. (c) Joshi, G. D.; Kulkarni, S. N.
Indian J. Chem. 1965, 3, 91–92. (d) Uhde, G.; Ohloff, G. HelV. Chim.
Acta 1972, 55, 2621–2625. (e) Breeden, D. C.; Coates, R. M.
Tetrahedron 1994, 50, 11123–11132. (f) Bhonsle, J. B.; Deshpande,
V. H.; Ravindranathan, T. Indian J. Chem., Sect. B 1994, 33, 313–
316.
(8) Enders, D.; Kipphardt, H.; Gerdes, P.; Brena-Valle, L. J.; Bhushan,
V. Bull. Soc. Chim. Belg. 1988, 97, 691–704.
(9) Chi, Y.; Gellman, S. H. Org. Lett. 2005, 7, 4253–4256.
(10) Comins, D. L.; Dehghani, A. Tetrahedron Lett. 1992, 33, 6299–6302.
(11) Karlstro¨m, A. S. E.; Ro¨nn, M.; Thorarensen, A.; Ba¨ckvall, J.-E. J.
Org. Chem. 1998, 63, 2517–2522.
9
11116 J. AM. CHEM. SOC. VOL. 130, NO. 33, 2008

Biyouyanagin A and Analogues
A R T I C L E S
Scheme 3. Separation of Propargyl Alcohols 26 and 4-epi-26
Scheme 4. Synthesis of Hyperolactone C (18) and
(Biyouyanagin Numbering)a
4-epi-Hyperolactone C (16) through Palladium-Catalyzed Cascade
Sequences (Biyouyanagin Numbering)a
a Reagents and conditions: (a) 3,5-dinitrobenzoyl chloride (1.2 equiv),
Et3N (1.2 equiv), 4-DMAP (0.1 equiv), CH2Cl2, 25 °C, 3 h; (b) K2CO3
(5.0 equiv), MeOH, 0 f 25 °C, 30 min, 98%.
Figure 5. ORTEP drawing of compound 26a with the thermal ellipsoids
at 30% probability level.
Figure 6. ORTEP drawing of compound 26 with the thermal ellipsoids at
30% probability level.
3:1 isomeric ratio). Although this mixture could not be
conveniently resolved chromatographically, the desired stereo-
isomer could be isolated easily by fractional crystallization from
CH2Cl2/hexanes (62% yield). Alternatively, the two isomers
could be separated by flash column chromatography of their
4-nitrobenzoates (4-nitrobenzoyl chloride, Et3N, 4-DMAP, 95%
combined yield), and then the two free alcohols (26 and 4-epi-
26) were released after hydrolysis (K2CO3, MeOH, 98% yield,
Scheme 3). The stereoselectivity observed in this reaction is
attributed to the presumed chelation control exerted by the
benzyloxy group. The stereochemistry of 26 was initially
confirmed by an X-ray crystallographic analysis (see ORTEP
drawing, Figure 5) of its 2,4-dinitrobenzoate (26a, mp 159-160
°C, toluene; obtained by reaction of 26 with 2,4-dinitrobenzoyl
chloride in the presence of Et3N and 4-DMAP, 98% yield) and
subsequently by an X-ray crystallographic analysis performed
on the compound itself (26), which crystallized upon prolonged
standing in a mixture of CH2Cl2 and hexanes (mp 84-86 °C,
CH2Cl2/hexanes; see ORTEP drawing, Figure 6).
a Reagents and conditions: (a) Pd(PPh3)4 (5 mol%), PhI (1.2 equiv), CO
(200 psi), CO2 (200 psi), Et3N, 100 °C, 5 h, 79%; (b) BBr3 (1.5 equiv),
CH2Cl2, -78 °C, 30 min, 95%; (c) o-NO2C6H4SeCN (1.2 equiv), P(n-Bu)3
(1.2 equiv), THF, 25 °C, 4 h, 90%; (d) H2O2 (30% aq., 10 equiv), THF, 0
f 25 °C, 1 h, 85%, (e) same as (a), 71%; (f) same conditions as for (b);
(g) same conditions as for (c); (h) same conditions as for (d), 71% over the
three steps.
What was developed next was a highly efficient and pleasing
cascade sequence that produced spirolactone 27a from propar-
gylic alcohol 26 and phenyl iodide in one pot. This remarkable
palladium-catalyzed cascade was brought about by Pd(PPh3)4
(5 mol%) and Et3N under an atmosphere of a mixture of CO
and CO2 (200 psi) as depicted in Scheme 4 (79% yield, single
9
J. AM. CHEM. SOC. VOL. 130, NO. 33, 2008 11117

A R T I C L E S
Nicolaou et al.
Scheme 5. Completion of the Total Synthesis of
24-epi-Biyouyanagin A (2a) and Biyouyanagin A (2b)a
Figure 7. ORTEP drawing of compound 27a with the thermal ellipsoids
at 30% probability level.
diastereomer).12 An X-ray crystallographic analysis of 27a (mp
176-177 °C, toluene, see ORTEP drawing, Figure 7) allowed
its unambiguous stereochemical assignment. The retention of
stereochemistry at the spirocenter of product 27a as compared
to the starting substrate (26) of the cascade sequence is consistent
with the proposed mechanism involving intermediates I-VI
shown in Scheme 4. Specifically, it is postulated12 that a
Sonogashira-type process under the reaction conditions allows
the initial formation of acetylenic ketone I, which absorbs a
molecule of CO2 to furnish species II, an intermediate that
serves as the precursor to cyclic carbonate III, whose palladium-
catalyzed extrusion of CO2 leads to the π-palladium complex
IV T V, stereospecifically. Finally, rearrangement of V, as
shown, leads to palladacycle VI, which readily loses Pd0 to
afford the observed product (27a) with retention of stereochem-
istry. The completion of the synthesis of hyperolactone C (18)
required debenzylation (BBr3), selenenylation [o-(NO2)C6H4-
SeCN, n-Bu3P], and oxidation/syn-elimination (H2O2), a se-
quence that proceeded in 73% overall yield {[R]D25 ) -272.8
(c ) 0.69, CHCl3); lit.,6b [R]D25 ) -270.7 (c ) 0.11, CHCl3)}
as shown in Scheme 4. Similarly, 4-epi-hyperolactone C (16)
was synthesized from 4-epi-26 without loss of integrity of
stereochemistry (Scheme 4).
a Reagents and conditions: (a) 17a or 17b (4.0 equiv), 2′-acetonaphthone
(1.0 equiv), CH2Cl2 (0.5 M for 18), 320 nm filter, 5 °C, 8 h, 48% for 2a,
54% for 2b.
Total Synthesis of Biyouyanagin A and 24-epi-Biyouyanagin.
Having established enantioselective and practical routes to
fragments 17a and 17b as well as 18, the next task became
their photoinduced fusion into the biyouyanagin A framework
(Scheme 5). To this end, each of the two terpenoid fragments
17a and 17b (4.0 equiv) were separately mixed with hypero-
lactone C (18, 1.0 equiv) and irradiated with UV light in the
presence of 2′-acetonaphthone13 (1.0 equiv) as a triplet sensitizer
in a concentrated CH2Cl2 solution in a quartz cell (320 nm filter)
at 5 °C for 8 h. Under these conditions, we were pleased to
observe regio- and stereoselective formation of the desired
cyclobutane product in each case, leading to 2a (48% yield,
based on 18) and 2b (54% yield, based on 18). These
compounds exhibited NMR spectroscopic data consistent with
their structures, with 2b revealing identical signals to those
reported for the natural product, biyouyanagin A.2 Crucial NOE
data (see Figure 8) for this compound, however, supported 2b,
rather than the originally proposed structure 1b. Particularly
diagnostic for the cis, cis, trans, trans stereochemistry were the
NOEs between H-6 and H-17, H-6 and H-22, and H-17 and
H-22. Note that adjacent protons of the cyclobutane ring may
exhibit an NOE, even if they are trans to each other, as it is the
case here. In addition, the indicated NOEs between the aromatic
and C-23 methyl protons (see Figure 8) revealed the syn
orientation of these substituents. The absolute structures of 2a
Figure 8. Selected NOEs exhibited by biyouyanagin A (2b).
Figure 9. ORTEP drawing of compound 2a with the thermal ellipsoids at
30% probability level.
(mp 94-95 °C, hexanes) and 2b (mp 75-76 °C, hexanes) were
ultimately solved by X-ray crystallographic analysis (see ORTEP
drawings, Figures 9 and 10, respectively), which unambiguously
established the complete structure of biyouyanagin A as 2b. Its
absolute stereochemistry was deduced from those of the starting
materials [(R)-citronellal and (S)-malic acid] employed for the
synthesis of the natural enantiomer {[R]D25 ) -256.6 (c ) 0.80,
25
CHCl3); lit.,2 [R]D ) -240.0 (c ) 0.50, CHCl3)}.
Design and Synthesis of Biyouyanagin A Analogues. With
an efficient route to the biyouyanagin A molecule and the
establishment of its relative and absolute stereochemistry, we
then turned our attention to the application of the developed
synthetic technology to the construction of designed analogues
(12) Inoue, Y.; Ohuchi, K.; Yen, I.-F.; Imaizumi, S. Bull. Chem. Soc. Jpn.
1989, 62, 3518–3522.
(13) (a) Cantrell, T. S. J. Org. Chem. 1974, 39, 3063–3070. (b) Schenck,
G. O.; Kuhls, J.; Krauch, C. H. Justus Liebigs Ann. Chem. 1966, 693,
20–43.
9
11118 J. AM. CHEM. SOC. VOL. 130, NO. 33, 2008

Biyouyanagin A and Analogues
A R T I C L E S
Table 1. Syntheses of Biyouyanagin A Analogues by [2 + 2]
Photocycloadditiona
Figure 10. ORTEP drawing of compound 2b with the thermal ellipsoids
at 30% probability level.
Figure 11. ORTEP drawing of compound 8 with the thermal ellipsoids at
30% probability level.
Scheme 6. Synthesis of Enone 14 and Ketone 15a
a Reagents and conditions: (a) H2 (1 atm), 10 wt % Pd/C (5 mol%),
MeOH, 25 °C, 3 h, 99%; (b) L-Selectride (3.0 equiv), THF, -78 f -10
°C, 1 h, 70%.
for biological investigations. The biyouyanagin A analogues
3-11 (Figure 3) were designed in order to take advantage of
the flexibility and scope of the developed synthetic strategy and
to test the effect of substituents at C-3, C-7, C-19, and C-22,
while the truncated structures 12-16 (hyperolactone C ana-
logues) were designed in order to explore structure activity
relationships within the hyperolactone C (18) structural motif,
which was suspected to be the active pharmacophore of the
biyouyanagin A molecule.
The required building blocks were constructed by procedures
either as described above (30, 12, 13) for biyouyanagin A or as
reported previously in the literature (31,14 3215), except for enone
14 and ketone 15. Enone 14 was obtained from hydrogenation
of hyperolactone C (18) with H2 (1 atm) and 10 mol% Pd/C
a Reactions were carried out by irradiating (quartz cell, 320 nm filter)
a concentrated solution (0.5 M) of the terpenoid (4.0 equiv) and the
enone (1.0 equiv) components in CH2Cl2 at 5 °C for 8 h.
cat. in quantitative yield, while a stereoselective 1,4-reduction
of 18 with L-selectride gave ketone 15 in 70% yield (Scheme
6). Analogues 3-9 (see Table 1 and Figure 2) were synthesized
from the corresponding terpenoid dienes and enone components
through [2 + 2] photocycloaddition reactions as indicated in
Table 1. The structures of these products were consistent with
their spectroscopic data; that of 8 was further confirmed by
X-ray crystallographic analysis (mp 170-171 °C, hexanes; see
Figure 11 for ORTEP drawing).
The hexacyclic compound 10 was unexpectedly obtained
(40% yield) upon irradiation of a mixture of terpenoid compo-
nent 17b and enone 28a in the presence of 2′-acetonaphthone
through the presumed intermediacy of the initially formed adduct
33 as shown in Scheme 7. The spontaneous ring closure of the
(14) Sen, A.; Grosch, W. FlaVour Frag. J. 1990, 5, 233–234.
(15) Buttrus, N. H.; Cornforth, J.; Hitchcock, P. B.; Kumar, A.; Stuart,
A. S. J. Chem. Soc., Perkin Trans. 1 1987, 851–857.
9
J. AM. CHEM. SOC. VOL. 130, NO. 33, 2008 11119

A R T I C L E S
Nicolaou et al.
Scheme 7. Synthesis of Biyouyanagin A Analogue 10a
Scheme 8. Synthesis of Analogue 11 via Olefin Cross Metathesisa
a Reagents and conditions: Grubbs II cat. (10 mol%), 34 (5.0 equiv),
CH2Cl2, 40 °C, 24 h, 74%.
Table 2. Anti-HIV-1 Activities and Cytotoxicities of Biyouyanagin A
and Analoguesa
a Reagents and conditions: (a) 17b (4.0 equiv), 2′-acetonaphthone (1.0
equiv), CH2Cl2 (0.5 M for 28a), 320 nm filter, 5 °C, 8 h, 40%; (b) I2 (3.0
equiv), PPh3 (3.0 equiv), imidazole (4.0 equiv), CH2Cl2; (c) o-NO2C6H4SeNa
(2.0 equiv), THF; then H2O2 (30% aqueous, 10 equiv), THF, 25 °C, 5 h,
51% over the three steps.
b
c
entry
compd
HIV-1 IC50 (µM)
MT-2 CC50 (µM)
TId
1
2
3
4
5
6
7
8
AZT
2a
2b
3
4
5
6
7
9
10
11
12
13
14
15
16
18
0.078
20
26
25
25
20
19
23
31
19
27
16
12
34
23
35
29
>50
285
198
444
268
197
213
333
232
143
512
1301
417
1377
459
462
925
>641
14.4
7.5
17.5
10.8
10.0
11.1
14.4
7.5
9
10
11
12
13
14
15
16
17
7.5
18.9
80.0
35.6
40.0
20.0
13.3
32.0
a Average value from quintuplicate experiments. b IC50 is the
concentration at which 50% inhibition of HIV-1 replication is observed
as measured by a TZM-bl/luciferase assay. c CC50 is the concentration at
which 50% inhibition of cell metabolism is observed as measured by a
colorimetric XTT assay. d Therapeutic index (CC50/IC50).
Figure 12. ORTEP drawing of compound 10 with the thermal ellipsoids
at 30% probability level.
hydroxyl moiety upon the carbonyl function and the stability
of its lactol form (10) are clearly favored within the latter
molecule, as opposed to the architecture of the starting enone,
which apparently (1H and 13C NMR spectroscopic analysis)
remains open as shown. The structure of compound 10 was
initially based on its spectroscopic data and its facile conversion
to biyouyanagin A (2b) through iodination (I2, PPh3, imid.),
followed by sequential exposure of the resulting iodide to
o-NO2C6H4SeNa and H2O2 (51% overall yield for the three
steps). Ultimately, an X-ray crystallographic analysis of 10 (mp
127-128 °C, CH2Cl2/hexanes) unambiguously established its
assigned structure (see ORTEP drawing, Figure 12).
Finally, compound 11 was constructed from biyouyanagin
A (2b) by a cross metathesis with (Z)-2-butene-1,4-diol diacetate
(34, Scheme 8). This remarkably chemoselective reaction was
brought about, in 73% yield, by exposure of a mixture of 2b
and 34 (5.0 equiv) to Grubbs catalyst II (10 mol%) in a CH2Cl2
solution at 40 °C.16 Interestingly, no products arising from cross
metathesis with any of the other olefinic bonds of biyouyanagin
A (2b) were observed.
against HIV-1 replication in MT-2 lymphocytes and their
cytotoxicity against noninfected MT-2 lymphocytes using the
TZM-bl/luciferase assay and AZT as a control. As seen in Table
2, synthetic biyouyanagin A (2b) exhibited significant activity
against HIV-1 replication in this assay (IC50 ) 26 µM) and
lower cytotoxicity against the MT-2 lymphocytes, scoring a
therapeutic index (TI) of 7.5. The lower inhibitory potency
observed in these experiments compared to those reported by
Tanaka et al. (IC50 ) 1.7 µM)2 may be due to sensitivity
differences in the two assays employed. 24-epi-Biyouyanagin
A (2a) and the other biyouyanagin A analogues listed in Table
2 (and shown in Figures 2–4) displayed similar activities against
HIV-1 replication and cytotoxicities against the MT-2 lympho-
cytes (IC50 ) 19-31 µM; CC50 ) 143-512 µM) and TIs
(7.5-18.9), with analogues 6 and 10 being the most potent
against HIV-1 (IC50 ) 19 µM) and analogue 11 possessing the
most favorable TI (18.9). Interestingly, however, we discovered
that hyperolactone C (18, Figure 4) and its analogues (12-16,
Figure 3) showed comparable, and in some cases more potent,
anti-HIV activities compared to biyouyanagin A and its
analogues. Thus, hyperolactone C (18, entry 17), itself, exhibited
an IC50 value of 29 µM against HIV-1 infected cells and a CC50
value of 925 µM (TI ) 32.0). The most potent member of the
series against HIV-1 replication was found to be hyperolactone
Biological Evaluation of Biyouyanagin A and Analogues. The
synthesized compounds were tested for their inhibitory activity
(16) Chatterjee, A. K.; Choi, T. L.; Sanders, D. P.; Grubbs, R. H. J. Am.
Chem. Soc. 2003, 125, 11360–11370.
9
11120 J. AM. CHEM. SOC. VOL. 130, NO. 33, 2008

Biyouyanagin A and Analogues
A R T I C L E S
C analogue 13 (IC50 ) 12 µM), possessing a para-methoxy
group on its phenyl moiety. The most impressive compound of
the class, however, turned out to be hyperolactone C analogue
12 which exhibited a TI of 80.0 (IC50 ) 16 µM; CC50 ) 1301
µM). With a fluoride residue in its structure, this relatively small
molecule may serve as a new lead for a drug discovery program
in the anti-HIV area of pharmaceutical research.
lactone C, a plausible biogenetic precursor of biyouyanagin A,
possesses a higher and significant potency against HIV-1
replication. Based on these biological results and considering
the efficiency of the established synthetic strategy, further
investigations into chemical biology and medicinal chemistry
may be warranted.
Acknowledgment. Dedicated to Professor Duilio Arigoni on
the occasion of his 80th birthday. We thank Dr. D. H. Huang and
Dr. L. Pasterneck for NMR spectroscopic assistance and Dr. Siuzak
and Dr. R. Chadha for mass spectrometric and X-ray crystal-
lographic assistance, respectively. Financial support for this work
was provided by the National Institutes of Health (USA) and the
Skaggs Institute for Chemical Biology and A*STAR (postdoctoral
fellowship to T.R.W.).
Conclusion
Relying on a highly convergent strategy, the described total
synthesis of biyouyanagin A and its 24-epimer highlights the
importance of cascade reactions in total synthesis and the
continuing role the latter discipline plays in natural products
chemistry and biology. Thus, based on both organocatalysis and
metallocatalysis, the developed synthetic technology culminated
in the revision of the originally assigned structure of biyouy-
anagin A and determination of its absolute stereochemistry. It
also delivered a series of analogues of the natural product, whose
biological evaluation established the first structure-activity
relationships in the field and led to the discovery that hypero-
Supporting Information Available: Experimental procedures
and full compound characterization. This material is available
free of charge via the Internet at http://pubs.acs.org.
JA802805C
9
J. AM. CHEM. SOC. VOL. 130, NO. 33, 2008 11121