A redox economical synthesis of bioactive 6,12-guaianolides

Article abstract of DOI:10.1021/ol400904y

Syntheses of two 6,12-guaianolide analogs are reported within. The scope of the tandem allylboration/lactonization chemistry is expanded to provide a functionalized allene-yne-containing α-methylene butyrolactone that undergoes a Rh(I)-catalyzed cyclocarbonylation reaction to afford a 5-7-5 ring system. The resulting cycloadducts bear a structural resemblance to other NF-κB inhibitors such as cumambrin A and indeed were shown to inhibit NF-κB signaling and cancer cell growth.

Full text of DOI:10.1021/ol400904y

ORGANIC
LETTERS
2013
Vol. 15, No. 11
2644–2647
A Redox Economical Synthesis
of Bioactive 6,12-Guaianolides
Bo Wen,† Joseph K. Hexum,‡ John C. Widen,‡ Daniel A. Harki,*,‡ and
Kay M. Brummond*,†
Department of Chemistry, University of Pittsburgh, Pittsburgh, Pennsylvania 15260,
United States, and Department of Medicinal Chemistry, University of Minnesota,
Minneapolis, Minnesota 55414, United States
kbrummon@pitt.edu; daharki@umn.edu
Received April 2, 2013
ABSTRACT
Syntheses of two 6,12-guaianolide analogs are reported within. The scope of the tandem allylboration/lactonization chemistry is expanded to
provide a functionalized allene-yne-containing R-methylene butyrolactone that undergoes a Rh(I)-catalyzed cyclocarbonylation reaction to afford
a 5À7À5 ring system. The resulting cycloadducts bear a structural resemblance to other NF-κB inhibitors such as cumambrin A and indeed were
shown to inhibit NF-κB signaling and cancer cell growth.
Guaianolides are the most abundant group of sesqui-
terpene lactones (SLs), possessing a privileged natural
product status and a wide range of biological activities.1
Yet there is only one guaianolide, arglabin, available as a
marketed drug, constituting one of only 24 natural pro-
ducts approved for therapeutic use between 1974 and
2006.2,3 Reasons for the slow realization of their therapeu-
tic potential include poor bioavailability due to high
plasma protein interactions, poor toxicological profiles,
and hydrophobicity.4 Moreover, the biological activity of
these compounds is attributed to covalent bonding to the
R,β-unsaturated carbonyl groups, the same functionality
responsible for their toxicity.5 Despite potential toxicities,
3 of the top 10 drugs in the US, and one-third of all enzyme
targets for which there is an FDA approved inhibitor,
operate by a covalent mechanism of action.6 These proven
biomedical applications, combined with the finding that
irreversible binding may be an important factor against
drug resistance, have led to a reinvestment of the pharma-
ceutical community in covalent drugs.6,7
Natural products, such as guaianolides, can serve as
excellent leads for drug development, but molecular com-
plexity can pose formidable synthetic challenges.8 To date,
most synthetic approaches toward 6,12-guaianolides can
be characterized as target-oriented synthesis (TOS) strate-
gies that have not been explored for analog preparation of
these highly oxygenated skeletons,9 the synthesis of thap-
sigargin (2) being one exception (Figure 1).10 Oxidation
level [O] constitutes one measure of molecular complexity
† University of Pittsburgh.
‡ University of Minnesota.
(1) Drew, D. P.; Krichau, N.; Reichwald, K.; Simonsen, H. T.
Phytochem. Rev. 2009, 8, 581–599.
(2) (a) Newman, D. J.; Cragg, G. M. J. Nat. Prod. 2012, 75, 311–335.
(b) Newman, D. J.; Cragg, G. M. J. Nat. Prod. 2007, 70, 461–477.
(3) Science and Technology in Kazakhstan: Current Status and Future
Prospects, 2006; National Research Council, Office for Central Europe and
Eurasia, 2006.
(7) Amslinger, S. ChemMedChem 2010, 5, 351–356.
(8) Rosen, J.; Gottfries, J.; Muresan, S.; Backlund, A.; Oprea, T. I.
(4) Ghantous, A.; Gali-Muhtasib, H.; Vuorela, H.; Saliba, N. A.;
Darwiche, N. Drug Discovery Today 2010, 15, 668–678.
(5) Rishton, G. M. Drug Discovery Today 1997, 2, 382–384.
(6) (a) Potashman, M. H.; Duggan, M. E. J. Med. Chem. 2009, 52,
1231–1246. (b) Singh, J.; Petter, R. C.; Kluge, A. F. Curr. Opin. Chem.
Biol. 2010, 14, 475–480. (c) Singh, J.; Petter, R. C.; Baillie, T. A.; Whitty,
A. Nat. Rev. Drug Discovery 2011, 10, 307–317. (d) Liu, Q.; Sabnis, Y.;
Zhao, Z.; Zang, T.; Buhrlage, S. J.; Jones, L. H.; Gray, N. S. Chem. Biol.
2013, 20, 146–159.
J. Med. Chem. 2009, 52, 1953–1962.
(9) Schall, A.; Reiser, O. Eur. J. Org. Chem. 2008, 2353–2364.
(10) (a) Ley, S. V.; Antonello, A.; Balskus, E. P.; Booth, D. T.;
Christensen, S. B.; Cleator, E.; Gold, H.; Hogenauer, K.; Hunger, U.;
Myeres, R. M.; Oliver, S. F.; Simic, O.; Smith, M. D.; Sohoel, H.;
Woolford, A. J. A. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 12073–
12078. (b) Andrews, S. P.; Ball, M.; Wierschem, F.; Cleator, E.; Oliver,
S.; Hogenauer, K.; Simic, O.; Antonello, A.; Hunger, U.; Smith, M. D.;
Ley, S. V. Chem.;Eur. J. 2007, 13, 5688–5712.
r
10.1021/ol400904y
Published on Web 05/10/2013
2013 American Chemical Society

which can be directly correlated with synthetic accessibility
when performing a TOS.11 For example, the synthetic
steps required to prepare arglabin (8) and chinensiolide
(7) where [O] = 4 were fewer than 20. In contrast, more
than 40 steps were required to complete the synthesis of
thapsigargin (2).10 Given the highly oxidized nature of
6,12-guaianolides, a synthetic approach employing the prin-
ciples of redox economy would greatly alleviate the synthetic
challenges associated with the class of compounds.11
Described within is an 11-step synthesis of two guaia-
nolide analogs with oxidation levels equivalent to thapsi-
gargin and eupatochinilide VI, concise syntheses that were
realized by limiting the number of redox adjustments in
the synthetic sequence. We have previously demonstrated the
advantages of early stage incorporation of an R-methylene
butyrolactone on the Rh(I)-catalyzed allenic PausonÀ
Khand reaction (APKR).12 This study expands on the scope
of the APKR by incorporating additional functionality
into the allene-yne precursor 10. Furthermore, bio-
activity studies provide support for the preparation of
non-naturally occurring guaianolide analogs such as 11
(Scheme 1).13
Scheme 1. An APKR Approach to Highly Oxidized
Guaianolides
first examined.14 Compounds 12aÀd were prepared and
converted to the corresponding carbomethoxy allylboron-
ates 13aÀd by addition of DIBAL and subsequent trapping
of the intermediate aluminum species with ClCH2BPin
(Scheme 2). CuI was not required for the 1,4-addition
reaction of hydride to the ynoate, possibly because the
ether adjacent to the alkyne directs the addition. More-
over, Z/E ratios of allylboronates 13aÀd were dependent
upon the protecting group. For example, the reaction of
12aÀb, with silyl protecting groups, afforded 13aÀb in
Z/E ratios of 2À3:1. Whereas, the reaction of methyl- and
MOM-protected ethers, 12c and 12d, afforded the allyl-
boronates 13c and 13d with Z:E ratios of 9:1 and 4:1,
respectively. The stereochemical determining step is the
addition of the electrophile to one face over the other of
the intermediate allenoate 14. We propose that the Z/E
ratios correlate with the degree of chelation of the respec-
tive ether groups with the aluminum species of the
allenoate, where more chelation directs electrophilic
addition to the R-face.15
Scheme 2. Generation of the Allylboronates, Z/E Ratios
Figure 1. Examples of highly oxidized 6,12-guaianolides.
Synthesis of allene-yne 10 was envisioned using the
allylboration/lactonization chemistry developed by Hall
and previously used by us to access less functionalized
allene-yne precursors. Becausethereisonlyonereport with
functionality at a propargylic position, a model system was
Next, the lactonization step was examined on these
model systems (Scheme 3). Unfortunately, the E/Z isomers
of allylboronate 13 were not readily separated by column
chromatography sotheyweretakenon tothe lactonization
step as a mixture. Reaction of allylboronate 13a or 13b,
with either a TBS or TBDPS protecting group with boron
trifluoride etherate, triflic acid, or scandium triflate gave
only decomposition. However, reaction of allylboronate
(11) Burns, N. Z.; Baran, P. S.; Hoffmann, R. W. Angew. Chem., Int.
Ed. 2009, 48, 2854–2867.
(12) Grillet, F.; Huang, C.; Brummond, K. M. Org. Lett. 2011, 13,
6304–6307.
(13) Merfort, I. Curr. Drug Targets 2011, 12, 1560–1573.
(14) Kennedy, J. W. J.; Hall, D. G. J. Org. Chem. 2004, 69, 4412–
4428.
(15) Evans, D. A.; Allison, B. D.; Yang, M. G.; Masse, C. E. J. Am.
Chem. Soc. 2001, 123, 10840–10852.
Org. Lett., Vol. 15, No. 11, 2013
2645

13c with either triflic acid or scandium triflate gave an
∼75% yield of 15c in a 3À4:1 trans/cis lactone ratio. For
the MOM-protected ether 13d, purely thermal conditions
gave the best results, whereby heating 13d and phenylpro-
piolaldehyde to90°C for 48h gavean82% yield of 15d asa
trans/cis ratio of 2.7:1; acidic conditions led to decomposi-
tion of 13d. Next, the feasibility of allylation/lactonization
chemistry was tested on a more functionalized substrate.
The tert-butyldiphenylsilyl (TBDPS) group of 21a was
removed using triethylamine hydrogen fluoride to give
22a in 64% yield for the two steps. Reaction of the cis-
lactone 20b to the same sequence afforded 22b in 37% yield
(two steps).
Scheme 4. Synthesis of 6,12-Guaianolide Analogs 22aÀb
Scheme 3. A Model System for the Lactonization Protocola
a (a) BF3 OEt2; (b) TfOH; (c) Sc(OTf)3; (d) toluene, 90 °C.
3
Tothisend, allenylester 16isobtainedin82%yieldfrom
the monoprotected butyne-diol using a JohnsonÀClaisen
rearrangement (Scheme 4). Ester 16 is reacted with meth-
oxymethylamine hydrochloride and isopropyl magnesium
chloride to afford the corresponding Weinreb amide in
84% yield, which is taken on to alkynone 17 by reaction
with ethynyl magnesium bromide. Reduction of the car-
bonyl of ynone 17 is accomplished with lithium aluminum
hydride in 98% yield. The propargylic alcohol is not
purified but taken directly on to the corresponding methyl
ether in 90% yield. Deprotonation of the terminus of the
alkyne with n-butyllithium followed by addition of chloro-
methylester gives the alkynoate 18 in 84% yield. Reaction
of alkynoate 18 with DIBAL, CuI, MeLi, and ClCH2BPin
afforded allylboronate 19 in 78% yield with a Z/E ratio of
1.2:1. Performing this reaction in the absence of MeLi and
CuI gave allylboronate 19 in 80% yield with a Z/E ratio of
2.2:1 with more byproduct contamination. The Z/E iso-
mers were not separated, but taken on as a mixture to the
allylboration/lactonization step. The reaction of 19 with
phenylpropiolaldehyde using the acidic conditions de-
scribed above resulted in decomposition of the allyl-
boronate. Purely thermal conditions in toluene afforded
starting material at 50 °C and decomposition at 90 °C.
Interestingly, heating allylboronate 19 with 3-phenylpro-
piolaldehyde in chloroform for 7 days afforded some of the
desired lactones 20aÀb, but the bulk of the material
consisted of intermediate hydroxy esters.16 This complex
mixture was reacted with PTSA to afford lactone trans-20a
as 2:1 mixture of diastereomers in 40% yield. Uncyclized
material was recovered after chromatography and reacted
with NaH to afford lactone cis-20b as a single diastereomer
in 14% yield. The cis- and trans-lactones were taken on
independently to the Rh(I)-catalyzed cyclocarbonylation
reaction. Reaction of 20a with rhodium biscarbonyl chlo-
ride dimer in toluene at 90 °C afforded the cyclocar-
bonylation product 21a as a mixture of diastereomers.
Natural products bearing R-methylene butyrolactones
are well-established bioactive molecules. Inhibition of the
NF-κB signaling pathway, a hingepoint for the activation
of the cellular inflammatory response, has been demon-
strated by molecules of this class.17 In addition, it has been
shown that a natural product analog, dimethylamino-
parthenolide (DMAPT; LC-1), has the ability to simulta-
neously knockdown NF-κB levels and activate the p53
pathway, thus promoting the apoptosis of cancer cells.18
Inspired by these previous studies, we evaluated 22aÀb for
inhibition of induced NF-κB activity in cell cultures
(Figure 2). A549 cells bearing a stably transfected NF-κB
reporter construct were treated with each compound.
Figure 2. NF-κB luciferase reporter assay in A549 cells. Com-
pounds 22aÀb were dosed at 20, 10, and 1 μM, and PTL was
dosed at 10 and 1 μM. Cells were induced with TNF-R (15 ng/mL)
30 min after molecule treatment, except NI control. Shown is the
mean of triplicate data, and error bars represent propagated
standard deviation. NI = noninduced, I = induced.
(16) Huang, Y.; Rawal, V. H. J. Am. Chem. Soc. 2002, 124, 9662–63.
2646
Org. Lett., Vol. 15, No. 11, 2013

Table 1. Antiproliferative Activities of 22aÀb and Parthenolide (PTL)a
compound
DU-145
HeLa
HL-60
U-87 MG
NCI/ADR-RES
Vero
22a
22b
PTL
29.1 ( 4.7
21.6 ( 1.9
8.9 ( 4.6
20.3 ( 6.0
39.7 ( 16.4
45.1 ( 3.7
5.5 ( 0.4
7.8 ( 2.3
9.3 ( 3.8
27.1 ( 4.8
9.8 ( 1.4
8.8 ( 2.1
80.9 ( 24.0
25.4 ( 1.0
57.6 ( 8.9
32.2 ( 7.0
30.1 ( 5.5
22.4 ( 1.5
a Compounds were dosed to cells and incubated for 48 h. Viability was measured by Alamar Blue staining. Mean IC50 values ( SD (μM) are shown.
Activation of NF-κB signaling yields an increase in repor-
ter luminescence that is diminished in the presence of
NF-κB inhibitors.19 Results from our study were bench-
marked against parthenolide (PTL), a known NF-κB
inhibitor bearing an R-methylene butyrolactone. trans-22a
and cis-22b were equipotent inhibitors in this assay,
diminishing induced NF-κB activity to noninduced levels
at 20 μM. Both analogs resulted in substantial decreases in
NF-κB activity, with 57% (22a) and 59% (22b) residual
activity measured at 10 μM. PTL was found to be slightly
more potent, reducing NF-κB levels to 53% residual
activity at 10 μM.
Inhibition of NF-κB signaling is an emerging strategy
for developing novel anticancer agents.20 Additionally,
many R-methylene butyrolactone-containing natural prod-
ucts have documented antiproliferative activities.13 We
evaluated 22aÀb for growth inhibitory activity against a
panel of cancerous and noncancerous cell lines. Both
compounds were benchmarked against PTL and clinically
used drugs gemcitabine and doxorubicin (Figures S1 and
S2). Antiproliferative data for PTL has been previously
reported for HL-60, HeLa, U-87 MG, and Vero, and our
data are in close agreement to previous reports.21 In
general, 22a and 22b were similarly active when compared
to each other, and slightly less active than PTL. Notable
exceptions to this trend include HeLa breast carcinoma
and HL-60 leukemia cells, in which 22a was ∼2-fold more
active than PTL (Table 1). Conversely, cis-22b was more
active than trans-22a in U-87 MG brain tumor cells
(IC50 = 9.8 μM vs 27.1 μM) and has similar activity to
PTL (IC50 = 8.8 μM). Interestingly, 22b (IC50 = 25.4 μM)
was more active than both 22a (IC50 = 80.9 μM) and PTL
(IC50 = 57.6 μM) against the well-known NCI/ADR-RES
cell line, which is a model of drug-resistant ovarian cancer
due to overexpression of p-glycoprotein (P-gp) efflux
pump.22 NCI/ADR-RES is resistant to doxorubixin
(adriamycin) and gemcitabine (IC50’s > 500 μM, Figure
S2). These results suggest that molecules with covalent
mechanisms of activity, such as the guaianolide analogs
22aÀb, may be valuable scaffolds for targeting drug-
resistant cells. Both compounds were screened against
the noncancerous cell line Vero, and moderate
toxicity was observed for all R-methylene butyro-
lactone analogs.
In conclusion, the scope of the APKR has been extended
to the preparation of highly oxygenated guaianolide ana-
logs, 22aÀb. Bioactivity data support the potential of this
class of compounds as regulators of NF-κB and cell
proliferation and further validates the medicinal relevancy
of this region of chemical space. Our ability to modify the
structure of these compounds de novo enables the optimi-
zation of analog solubility and pharmacokinetic properties
for advanced biological applications. Furthermore, our
strategy provides ready access to uniquely functionalized
6,12-guaianolide analogs with activities on par with a
highly studied member of the SLs, parthenolide. Studies
are underway to establish structureÀactivity relationships
and the mechanism by which compounds 22aÀb inhibit
NF-κB, in addition to benchmarking their thiol reactivity.
(17) Rungeler, P.; Castro, V.; Mora, G.; Goren, N.; Vichnewski, W.;
Pahl, H. L.; Merfort, I.; Schmidt, T. J. Bioorg. Med. Chem. 1999, 7,
2343–52.
(18) (a) Guzman, M.; Rossi, R.; Neelakantan, S.; Li, X.; Corbett, C.;
Hassane, D.; Becker, M.; Bennett, J.; Sullivan, E.; Lachowicz, J.;
Vaughan, A.; Sweeney, C.; Matthews, W.; Carrol, M.; Liesveld, J.;
Crooks, P.; Jordan, C. Blood 2007, 110, 4427–4435. (b) Dey, A.;
Tergaonkar, V.; Lane, D. P. Nat. Rev. Drug Discovery 2008, 7, 1031–40.
(19) Hexum, J. K.; Tello-Aburto, R.; Struntz, N. B.; Harned, A. M.;
Harki, D. A. ACS Med. Chem. Lett. 2012, 3, 459–464.
Acknowledgment. We thank the NIH for financial sup-
port (NIGMS P50GM067082 and GM54161). D.A.H.
acknowledges Hyundai Hope on Wheels (Hope Grant
award) and The V Foundation for Cancer Research
(V Scholar award to D.A.H.) for financial support.
(20) (a) Karin, M. Nature 2006, 441, 431–436. (b) Karin, M.; Greten,
F. R. Nat. Rev. Immunol. 2005, 5, 749–759. (c) Karin, M.; Yamamoto,
Y.; Wang, M. Nat. Rev. Drug Discovery 2004, 3, 17–26. (d) Naugler,
W. E.; Karin, M. Curr. Opin. Gen. Dev. 2008, 18, 19–26.
(21) (a) Han, C.; Barrios, F. J.; Riofski, M. V.; Colby, D. A. J. Org.
Chem. 2009, 74, 7176–23. (b) Collu, F.; Bonsignore, L.; Casu, M.; Floris,
C.; Gertsch, J.; Cottiglia, F. Bioorg. Med. Chem. Lett. 2008, 18, 1559–62.
(c) Zanotto-Filho, A.; Braganhol, E.; Schroder, R.; de Souza, L. H.;
Dalmolin, R. J.; Pasquali, M. A.; Gelain, D. P.; Battastini, A. M.;
Moreira, J. C. Biochem. Pharmacol. 2011, 81, 412–24. (d) Onozato, T.;
Nakamura, C. V.; Cortez, D. A.; Dias Filho, B. P.; Ueda-Nakamura, T.
Phytother. Res. 2009, 23, 791–6.
Supporting Information Available. Detailed procedures
and data for all compounds in Schemes 2À4, biochemical
assays, and Figures S1ÀS3. This material is available free
of charge via the Internet at http://pubs.acs.org.
(22) (a) Batist, G.; Tulpule, A.; Sinha, B. K.; Katki, A. G.; Myers,
C. E.; Cowan, K. H. J. Biol. Chem. 1986, 261, 15544–15549. (b) Ke, W.;
Yu, P.; Wang, J.; Wang, R.; Guo, C.; Zhou, L.; Li, C.; Li, K. Med. Oncol.
2011, 28, S135–S141.
The authors declare no competing financial interest.
Org. Lett., Vol. 15, No. 11, 2013
2647