Synthesis and Biological Evaluation of Kibdelone C and Its Simplified Derivatives

Article abstract of DOI:10.1021/jacs.6b05484

Poylcyclic tetrahydroxanthones comprise a large class of cytototoxic natural products. No mechanism of action has been described for any member of the family. We report the synthesis of kibdelone C and several simplified analogs. Both enantiomers of kibdeleone C show low nanomolar cytotoxicity toward multiple human cancer cell lines. Moreover, several simplified derivatives with improved chemical stability display higher activity than the natural product itself. In vitro studies rule out interaction with DNA or inhibition of topoisomerase, both of which are common modes of action for polycyclic aromatic compounds. However, celluar studies reveal that kibdelone C and its simplified derivatives disrupt the actin cytoseketon without directly binding actin or affecting its polymerization in vitro.

Full text of DOI:10.1021/jacs.6b05484

Subscriber access provided by CORNELL UNIVERSITY LIBRARY
Article
Synthesis and biological evaluation of kibdelone C and simplified derivatives
Janjira Rujirawanich, Soyeon Kim, Ai-Jun Ma, John R. Butler, Yizhong Wang, Chao
Wang, Michael Rosen, Bruce Posner, Deepak Nijhawan, and Joseph M. Ready
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.6b05484 • Publication Date (Web): 26 Jul 2016
Downloaded from http://pubs.acs.org on July 26, 2016
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of the American Chemical Society is published by the American Chemical
Society. 1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 12
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Synthesis and biological evaluation of kibdelone C and simplified de-
rivatives
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Janjira Rujirawanich,a Soyeon Kim,b Ai-Jun Ma,a John R. Butler,a Yizhong Wang,a Chao Wang,a
Michael Rosen,b Bruce Posner,a Deepak Nijhawan,a,c and Joseph M. Ready*a
aDepartment of Biochemistry and bDepartment of Internal Medicine, UT Southwestern Medical Center, 5323 Harry
Hines Blvd, Dallas, Texas, 75390-9038.
ABSTRACT: Poylcyclic tetrahydroxanthones comprise a large class of cytototoxic natural products. No mechanism of
action has been described for any member of the family. We report the synthesis of kibdelone C and several simplified
analogs. Both enantiomers of kibdeleone C show low nM cytotoxicity towards multiple human cancer cell lines. Moreo-
ver, several simplified derivatives with improved chemical stability display higher activity than the natural product itself.
In vitro studies rule out interaction with DNA or inhibition of topoisomerase, both of which are common modes of action
for polycyclic aromatic compounds. However, celluar studies reveal that kibdelone C and simplified derivatives disrupt
the actin cytoseketon without directly binding actin or affecting its polymerization in vitro.
Introduction
istry shown in Scheme 1. The three congeners vary at the
oxidation state of the B ring (quinone vs. hydroquinone)
and the C ring (saturated vs. unsaturated). Furthermore,
the isolation group revealed that kibdelone C could spon-
taneously oxidize to kibdelones A and B under aerobic
conditions.
Kibdelones A-C (Scheme 1, 1-3) are members of a grow-
ing class of natural products characterized by a hexacy-
clic, polyaromatic scaffold and a tetrahydroxanthone
moiety.1,2 These microbial metabolites arise biosyntheti-
cally from multiple cyclizations from a single polyacetate
chain, accounting for the high degree of oxygenation on
the aromatic and heteroaromatic rings.3 Structurally re-
lated natural products include simaomicin α,4 actino-
planone,5 SCH-560366 and kigamicin A.7 Additionally,
natural products such as cervinomycin A28 and FD-5949
feature aromatic F-rings as part of a xanthone moiety. All
of these secondary metabolites share a common hexacy-
clic core structure and differ with regard to the substitu-
tion pattern on the aryl and heteroaryl rings. They display
potent cytotoxicity, although no mechanism of action has
been elucidated for any member of the family. In particu-
lar, they generally show broad-spectrum toxicity includ-
ing anti-cancer, antibiotic and anti-fungal activity.
Simaomicin α has shown activity preventing coccidiosis
in chickens,4b and was found to induce G1 arrest in cul-
tured cancer cells.4c Likewise, the kigamicins have
demonstrated efficacy in inhibiting tumor growth in im-
munocompromised mice.7c Despite these promising bio-
logical activities, little is understood regarding the mech-
anism by which polycyclic tetrahydroxanthone natural
products kill microbial and mammalian cells.
The kibdelones displayed potent cytotoxicity against a
panel of human cancer cell lines. All three compounds
arrested growth of cell lines derived from lung, colon,
ovarian, prostrate, breast and other tumor types with
GI50’s < 5 nM. They additionally showed robust toxicity
towards B. subtilis bacteria. It is unclear if the kibdelone
A-C are in fact equally active or if they interconvert under
the conditions of the biological assays. The NCI
COMPARE analysis did not reveal a strong correlation
between the toxicity profile of the kibdelones and that of
other known cytotoxins, indicating that they might oper-
ate through an unknown mode of action.
The combination of unique structural features and
compelling biological properties make the polycyclic xan-
thone natural products attractive targets for total synthe-
sis.2 The Kelly group described the first synthesis of the
xanthone-containing natural product cervinomycins A1
and A2,11 and reports from the Rao and Mehta groups fol-
lowed shortly thereafter.12,13 The Suzuki group made an
important contribution to the area with a synthesis of FD-
594 aglycone; theirs was the first enantioselective synthe-
sis of a member of the family.14 The Porco group reported
an elegant synthesis of the naturally occurring (+)-
kibdelone C,15 and we described a synthesis of (-)-
simaomicin α in 2013.16 More recently, the Martin group
recently described a total synthesis of the aglycone of IB-
00208.17 Additionally, several groups have reported syn-
The Capon group discovered the kibdelones in the cul-
ture media of a soil actinomycete Kibdelosporangium
sp.1,10 Extensive 1- and 2-dimensional NMR data revealed
the presence of a chlorinated isoquinolinone AB ring sys-
tem. The DEF ring tetrahydroxanthone included three
stereogenic hydroxyl groups with the relative stereochem-
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 12
thetic studies towards natural products of the hexacyclic
xanthone family.18
1
2
3
4
5
6
7
8
9
We set as an initial objective the development of a flex-
ible and enantioselective synthetic route that could pro-
vide access to the natural product and allow assignment
of the absolute stereochemistry of 1-3.19 Based on initial
biological results described below, we then pursued sim-
plified analogues of the kibdelones. Here we describe the
evolution of our synthetic strategy, the synthesis of sim-
plified analogues maintaining full biological activity, and
initial efforts to understand the mode of action of the
kibdelones.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
because multiple methods exist for the formation of biaryl
bonds including cross-coupling, oxidative coupling and
C-H functionalization reactions. Finally, we reasoned that
the isoquinolinone fragment 11 could be joined to tetra-
hydroxanthone subunit 12 using an equivalent of acety-
lene as a lynchpin.
Fragment synthesis. To prepare the isoquinolinone AB
rings, we explored cyclization of aryl aldehyde 15 (Scheme
3). To this end, amino alcohol 1320 was coupled to the acid
chloride derived from benzoic acid 14.21 Subsequent oxi-
dation generated the cyclization substrate, aldehyde 15.
Both BCl3 and BBr3 promoted electrophilic addition to
yield a benzylic alcohol (not shown), which could be de-
hydrated to yield isoquinolinone 16 under acidic condi-
tions. While BCl3 only demethylated the phenol ortho to
the amide, BBr3 yielded doubly demethylated material. It
is unclear if demethylation is required for cyclization, but
we note that we never observed cyclized products that
retained both methyl groups. Moreover, the highest yields
Results and Discussion
Synthetic strategy. In considering a synthetic ap-
proach to the kibdelones and related natural products, we
targeted the C5-C6 bond that connects the B and D rings
(Scheme 2). Strategically, retrosynthetic disconnection of
this bond would deconstruct the C-ring to suggest an in-
termediate such as diaryl alkyne 10 in which the isoquino-
line and tetrahydroxanthone moieties can be viewed as
independent ring systems. From a tactical perspective, we
were attracted to a late-stage C5-C6 bond construction
ACS Paragon Plus Environment

Page 3 of 12
Journal of the American Chemical Society
were obtained when the BCl3 was added in two portions –
1 equiv followed by 1.5 equiv. Monitoring of a reaction in
1
2
3
4
5
6
7
8
Synthesis of the tetrahydroxanthone subunit started
with preparation of the D-ring. Formylation of phenol 1824
was anticipated to introduce the xanthone carbonyl.
However, reaction with formaldehyde under Lewis acid
conditions proved unsatisfactory (Table 1, entry 1).25 Phe-
nol 18 was completely consumed, and the crude reaction
mixtures were remarkably clean. Nonetheless, isolated
yields remained under 50% under a variety of tempera-
ture and concentration profiles. We speculate that the
benzylic alcohol intermediate formed under the reaction
CDCl3 by 1H NMR revealed that the first equivalent of BCl3
completely consumed aldehyde 15 to form an intermedi-
ate we characterized as a chlorohydrin.22 Subsequent ad-
dition of the remaining 1.5 equiv BCl3 resulted in demeth-
ylation and cyclization. By contrast, the starting material
was recovered largely intact following exposure to TiCl4,
Sc(OTf)3, AlCl3 and ZnCl2. Following cyclization with
BBr3, an acetylene equivalent was introduced using a So-
nogashira coupling to provide terminal alkyne 17.23
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
conditions (21, or corresponding metal alkoxide) might
have eliminated water to form the quinone methide 24,
which might have undergone polymerization. Alternative
formlyating agents such as methoxydichloromethane (en-
try 2) or tetraazaadamantane (23, entry 3) provided poor
regioselectivity and/or low yields.26 A MOM-ether, 19,
could be cleanly lithiated, but as shown in entry 4, trap-
ping with dimethylformamide or CO2 was not successful.
Finally, we discovered that aluminum-based Lewis acids
generated the benzylic alcohol 21 cleanly and in high yield
(entry 5).27 Subsequent protection of 21 with MOM-Cl
under phase transfer conditions and oxidation gave the
desired aldehyde 25 (eq 1).
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 12
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Our synthesis of the F-ring took advantage of pseudo-
vent reaction with a second equivalent of TsCl. Next, the
C2 symmetry of the F ring that related the C11 and C13
hydroxyl groups and the C10 and C14 oxygenation. In this
endeavor, we were inspired by Myers’ use of the Shi epox-
idation to form trans-diols.28,29 Specifically, resorcinol was
silylated and reduced with sodium and ammonia to pro-
vide the known bis-enol ether 26 (Scheme 4).28 Epoxida-
tion with the Shi catalyst (27) was followed by immediate
reduction with borane to provide the doubly protected
tetraol 30 as a single observed diastereomer and enantio-
mer.30 This reaction installed the C11 and C13 alcohols
with the correct relative stereochemistry for the kibdelo-
nes. Additionally, it introduced the required oxygenation
at C10 and C14. As proposed by Myers, borane presumably
acts as a Lewis acid to promote ring-opening of the si-
lyloxyoxirane (see 29); intramolecular hydride delivery
yields the protected diol. Selective mono-tosylation of 30
C10 hydroxyl was oxidized under basic conditions, which
resulted in β-elimination of the tosyl group to yield the
enone 32. Finally, iodination and Luche reduction of the
enone completed the synthesis of the F-ring iodide 33.
With access to an appropriate F-ring vinyl iodide, we
attempted to join it to the D-ring and construct the tetra-
hydroxanthone (Scheme 5). In preliminary studies, we
were unable to effect carbonylative coupling of aryl stan-
nane 34 with variously protected F-ring precursors (35, eq
2). Likewise, fully protected F-ring subunits could be met-
alated, but we observed no addition to the hindered and
electron-rich D-ring aldehyde (eq 3). Ultimately, Li and
Nicolaou’s synthesis of diversonol proved informative.31
Thus, we speculated that the C10 protecting group might
be causing steric crowding during the addition. Accord-
ingly we deprotonated 33 with MeLi and then converted it
to the vinyl lithium with n-BuLi. Addition to aldehyde 25
then yielded the benzylic alcohol 40 as an inconsequen-
tial mixture of diastereomers. (eq 4).
1
came as a welcome surprise. Analysis of H-1H coupling
constants for 31 suggested a trans-diaxial orientation of
the large TBS and tosyl groups. This conformation places
the remaining OH in an equatorial position as shown.
Steric hindrance from the adjacent OTBS group may pre-
To synthesize the tetrahydroxanthone subunit, diol 40
was oxidized under Dess-Martin conditions to an unsta-
ACS Paragon Plus Environment

Page 5 of 12
Journal of the American Chemical Society
and prompted dehydration of the A-ring back to the iso-
ble diketone (Scheme 6). This material was immediately
1
2
3
4
5
6
7
8
exposed to perchloric acid in a mixed solvent system con-
taining acetone and tert-butanol to form the tricyclic tet-
rahydroxanthone 42.30 This cyclization may involve in-
tramolecular conjugate addition of acetone hemiketal 41
to install the acetonide and set the correct C10 stereo-
chemistry. When the cyclization was performed in the
absence of acetone, the triol 43 was formed as a mixture
of C10 diastereomers. Likewise, when tert-butanol was
excluded from the reaction mixture, the methylene ketal
44 was formed, presumably from the formaldehyde liber-
ated during MOM deprotection.
quinolinone. However, only about half of the C13 MOM
group was removed, generating similar amounts of diols
53 and 54. Efforts to increase conversion led to decompo-
sition. Moreover, some of the formaldehyde (or its
equivalent) that formed during removal of the MOM
group was trapped by the C10-C11 diol to yield the meth-
ylene ketal 55. This mixture of products was then treated
with PhI(OAc)2 to remove the B-ring methyl group. Final-
ly, exposure to BCl3 cleaved the D-ring methyl group and
any remaining MOM or methylene ketal. Reductive
workup converted the C-ring quinone to the hydroqui-
none, and completed the synthesis of kibdelone C. The
natural product was prone to oxidation, as had been re-
ported by the isolation group. Accordingly, the final
product was purified by HPLC, immediately lyophilized,
and handled under an inert atmosphere.19
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Synthesis of ent-Kibdelone C. The isoquinolinone
fragment was linked to the tetrahydroxanthone using a
Cu-free Pd coupling.32 Even in the absence of copper salts,
rigorous degassing and slow addition of the alkyne was
required to avoid oxidative coupling to form a diyne side
products. Hydrogenation and Cu-catalyzed iodination33
yielded the phenol 46, which could be treated with
(Boc)2O to form the carbonate 47. An intramolecular, Pd-
mediated C-H arylation reaction formed the C4-C5 biaryl
bond and constructed the C-ring of the natural product.34
This transformation required extensive optimization,
which revealed that the OBoc group, the Pd source, the
temperature and the sodium pivalate buffer were all im-
portant. The free phenol 46 or methyl ether 48 only gen-
erated trace amounts of desired product, while the meth-
oxyethoxy methyl (MEM) containing substrate 49 gave
around 15% of the desired product. The success of the
MEM and Boc-containing substrates indicate that these
groups may coordinate to and stabilize the Pd(II) center
within an aryl palladium iodide intermediate. In this cy-
clization, Pd(OAc)2 was effective, but more Lewis acidic
Pd(II) salts [e.g. PdCl2, Pd(TFA)2, Pd(OTs)2] caused aro-
matization of the F-ring. Likewise, at temperatures below
85 oC, only deiodination was observed whereas at temper-
The specific rotation of synthetic kibdelone (-40) was
similar in magnitude but opposite in sign to the naturally
occurring material (+47). This observation indicated that
our synthetic material was enantiomeric to the natural
product, a conclusion consistent with contemporaneous
results
from
the
Porco
group.15b
150
100
50
nat-KibC (3)
nat-MeKibC (57
ent-KibC (
ent-MeKibC (57
)
)
3
)
o
atures higher than 95 C, extensive decomposition oc-
curred. Finally, in situ generation of NaOPiv facilitated C-
H functionalization, presumably through a concerted
metalation-deprotonation as proposed by Fagnou and
coworkers.35
Access to the full kibdelone carbon skeleton allowed us
to prepare the natural product and several simple deriva-
tives. Thus, halogenation with 51 introduced the A-ring
chloride, albeit with an unexpected complication. Specifi-
cally, the desired chloroisoquinolinone was accompanied
by side products that we tentatively characterized as the
hydration products (see 52). Multiple stereoisomeric
halohydrins were thus formed, and all of the products
were generated as ~1:1 mixtures of atropisomers due to
hindered rotation around the C4-C5 bond. Accordingly,
no purification was possible at this stage. Next, treatment
with HClO4 removed the Boc group and the acetonide
0
10-10
10-8
10-6
10-4
[Kibdelone Derivative] (M)
Figure 1. Dose-response curves of enantiomeric forms of
kibdelone C (3) and methyl kibdelone C (57) against
HCT116 colon cancer cells. Cell viability was determined
with CellTiter-Glo®.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 12
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Synthesis and evaluation of intitial derivatives of
treatment with BCl3 retained both the B-ring and inner-
rim D-ring methoxy groups to yield dimethyl kibdelone C
(56) as a 1:1 mixture of atropisomers.
kibdelone C. We repeated the sequence shown above
except for the use of the enantiomeric Shi catalyst36 (ent-
27) to provide the natural enantiomer for side-by-side
comparison of their biological potencies (see below).22
Additionally, our synthetic scheme allowed us to prepare
two simple analogs to test the requirements for biological
activity. In particular, omitting the oxidative removal of
the B-ring methyl group yielded methyl-kibdelone C (57).
As anticipated, this compound was much more stable
than the natural product, with no observed oxidative con-
version to the B-ring hydroquinone under ambient condi-
tions. Similarly, intermediate 50 was chlorinated and
deprotected under acidic conditions; low-temperature
The syntheses of both enantiomers of kibdelone C and
methyl kibdelone C allowed us to answer two initial ques-
tions. Is the absolute configuration of the natural product
important for biological activity? And is a quinone re-
quired for biological activity? Concerning the latter ques-
tion, the Kapon group had observed rapid aerobic oxida-
tion of kibdelone C to kibdelones A (1) and B (2), which
feature quinone moieties. Moreover, they reported nearly
identical cytotoxicity with the 3 congeners. Since qui-
nones are reactive to nucleophiles and can act as electron
ACS Paragon Plus Environment

Page 7 of 12
Journal of the American Chemical Society
acceptors, we wondered if a quinone might be required
clization involving SeO2 provided modest yields of the
diol 64, which lacks an alcohol at C13 (Scheme 7b).38
1
for biological activity (eq 2). In this scenario, kibdelone C
would act as a pro-drug, undergoing oxidation under the
assay conditions. Of relevance to this question, we noted
the increased aerobic stability of methyl kibdelone C,
which appeared to be indefinitely stable under ambient
conditions. It provided a control compound that would be
less prone to generate quinones compared to the naturel
product. We therefore tested the cytotoxicity of both en-
antiomers of kibdelone C and methyl kibdelone C (57)
against the colon cancer cell line HCT116. As shown in Fig
1, the four compounds showed potent toxicity with over-
lapping dose-response curves. The concentration that
resulted in 50% less viability compared to DMSO control
(IC50) was calculated for each compound. The calculated
IC50’s were between 3 and 5 nM under these assay condi-
tions. From these data we conclude that a) the absolute
configuration of the natural product does not affect activ-
ity, and b) quinone formation is likely not required for
biological activity.37
2
3
4
5
6
7
8
9
The alcohols of 60 and 61 were protected, which al-
lowed separation of the cis and trans diastereomeric diols
66 (Scheme 7c). Separately, these aryl bromides were
coupled with alkyne 17 under Cu-free conditions to yield
the diaryl alkynes 67-68.31 Hydrogenation of the alkyne
proceeded in the presence of the benzyl and BOM groups.
Next, C4 was iodinated under our Cu-catalyzed condi-
tions, and the B-ring phenol was converted to a car-
bonate. Iodide 69, featuring only one alkoxy substitutent
on the F-ring, was cyclized in the presence of Pd(OAc)2,
PCy3, and NaOPiv. This C-H arylation reaction formed the
C-ring and completed the carbon skeleton to provide
hexacycle 71. Likewise, trans-70 reacted cleanly under the
same reaction conditions to provide the hexacyclic prod-
uct trans-72 in an excellent 86% yield. By contrast, the cis
diastereomer, cis-70, generated the tetrahydroxanthone
cis-72 in only 43% yield. The reaction additionally provid-
ed the xanthone 73 containing a fully aromatic F ring.
This side product arose from elimination of the benzyl
and BOM ethers. Finally, removal of the benzyl, BOM and
Boc groups and the D-ring methyl of intermediates 71-73
gave the deprotected analogs of kibdelone C. Both the cis-
72 and trans-72 diastereomers produced the same ~1:1
ratio of diol 76 under the acidic deprotection conditions.
Using the same synthetic strategy, di-benzyl ether 64 was
advanced to diol 77, which lacks the C13 hydroxyl of
kibdelone (Scheme 7d).22 Of note, each of these analogs
also differs from kibdelone by the absence of the A-ring
chloride and the presence of the B-ring methoxy group.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Several synthetic analogs were tested for biological ac-
tivity against three non-small cell lung cancer cell lines
that have been shown to display differential sensitivity to
various cytotoxins.39 Each cell line was treated with each
compound in a 12-point dose-response experiment in trip-
licate ranging from 50 ꢀM to 30 pM. Cell viability was
measured with CellTiter-Glo® (Promega, Inc.) after four
o
days in culture at 37 C. The IC50 values are shown in Ta-
ble 2. As described above, kibdelone C and methyl-
kibdelone (57) showed low-nM toxicity. By contrast, di-
methyl-kibdelone 56, which retains a methyl group on the
convex D-ring phenol, lost >100-fold potency. Excitingly,
diol 76, which features the same F-ring as simaomicin α,
showed IC50 values less than 5 nM, as did diol 77, which
lacks the C13 hydroxyl. The analog 75 retains only the C10
hydroxyl, but was as active as the natural product. Finally,
the simplest analog, 74, with an unsubstituted aryl F-ring,
was likewise as active as kibdelone C. Taken together, the
data show that hydroxylation on the F-ring, chlorination,
and a B-ring hydroquinone are not required for activity.
Therefore, simplified achiral and racemic analogs main-
tain full cytotoxic activity, which could facilitate studies
to understand their mechanism of action.
F-ring derivatives of kibdelone C. The equivalent cy-
totoxicity of (+)- and (-)-kibdelone C is consistent with
our recent discovery that the unnatural enantiomer of
simaomicin α shows potent cytotoxicity towards human
cancer cell lines (3 lines, IC50 < 100 nM). Of note, Porco
and coworkers previously found that kibdelone deriva-
tives lacking an F-ring were >1000-fold less active than the
natural product.36 These observations prompted us to
prepare analogs of kibdelone C featuring simplified F-
rings to determine the requirements for activity (Scheme
7). Towards that end, iodo-cyclohexenols 58 and 59 were
lithiated and exposed to aldehyde 25. Subsequent Dess-
Martin oxidation and treatment with perchloric acid
yielded the tetrahydroxanthones that lack oxygenation at
C11 (61) or C11 and C13 (60, Scheme 7a). As an alternative
to the acid-induced cyclization, enone 62 was treated
with HCl to remove the MOM group (63). Oxidative cy-
Biological Studies
Kibdelone does not bind DNA. Molecular models of
kibdelone C revealed a helical topology of the largely pla-
nar ring system. This architecture suggested the possibil-
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 12
ity that kibdelone C and derivative might interact with DNA fragments (lanes 3, 4), methyl kibdelone C again had
1
2
3
4
5
6
7
8
nucleic acids and/or nucleic acid-binding proteins. In-
deed, polycyclic aromatic molecules can intercalate into
DNA, and this activity often underpins their cytotoxici-
ty.40 To determine if intercalation into DNA was respon-
sible for the cell killing associated with kibdelone, we at-
tempted to rescue cells from kibdelone by adding exoge-
nous DNA. In this experiment, the added DNA would
compete with cellular DNA for binding to drug, thereby
resulting in a rightward shift in the dose-response curve.
As a positive control, H2122 non-small cell lung cancer
cells were treated with varying concentrations of actino-
mycin, an anti-cancer agent known to bind DNA and in-
hibit transcription.41 In the absence of added DNA, acti-
nomycin displayed an IC50 of 0.5 nM.42 However, the IC50
value shifted to 51 nM in the presence of 40 ug of herring
sperm DNA (Table 3). By contrast, paclitaxel, which does
not bind DNA and served as a negative control, displayed
equivalent IC50 values in the presence and absence of
added DNA (6 nM). In the key experiment, adding exoge-
nous DNA to the cell cultures had no impact on the tox-
icity of ent-kibdelone C. We interpret this result to indi-
cate that kibdelone’s mechanism of action does not in-
volve binding to DNA.
no effect on enzymatic activity or product distribution up
to 10 ꢀM (lanes 6-11). We conclude that the kibdelone
family of compounds does not inhibit either topoisomer-
ase
I
or
II.
a.
Topo I +
Sc DNA
+CPT
Topo I + Sc DNA
Sc
DNA
relax lin.
57: 0 0.01 0.03 0.1 1.0 10 ꢀM
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Lane: 1
b.
2
3
4
5
6
7
8
9
10
11
Topo II + catenated kDNA
Decat.
kDNA
Pos.
Control 57: 0 0.01 0.03 0.1 1.0 10 ꢀM
Cat.
kDNA
lin.
Table 3. Impact of exogenous DNA on cytotoxicity of
natural products.a
Lane: 1
2
3
4
5
6
7
8
9
10
11
Figures 2. a. Interaction of (-)-methyl kibdelone C (57) with
topoisomerase I (Topo I). Supercoiled plasmid DNA (Sc DNA,
lane 1) is converted to relaxed plasmid DNA (lane 2) by Topo
I. Linear (lin.) DNA is shown for comparison (lane 3). Topo I
generates nicked open circular DNA, which runs as a single
band (lane 4). Methyl kibdelone does not inhibit Topo I up to
10 ꢀM (lanes 6-11). b. Interaction of 57 with topoisomerase II.
Catenated kinetoplast DNA (Cat. kDNA, lane 1) is converted
to decatenated kDNA (lane 2) by topoisomerase II. A positive
control, etoposide, induced the formation of linear DNA (lin.,
lanes 3, 4). Methyl kibdelone does not inhibit Topo II up to 10
ꢀM (lanes 6-11).
IC50 (nM) in presence of added DNAb
compound
actinomycin 0.5
paclitaxel 6.4
kibdelone C 67
0 ꢀg
4 ꢀg
1.5
12 ꢀg
5.9
40 ꢀg
51
7.8
43
7.2
6.4
61
64
aHerring sperm DNA was added to H2122 cell cultures at the
indicated levels in 96-well plates, and cells were incubated
for 4 days prior to quantifying cell viability with CellTiter-
Glo®. bCalculated concentrations at which there were 50%
fewer viable cells compared to DMSO control.
Kibdelone and derivatives affect the cytoskeleton.
Kibdelone C and its active derivatives caused morphologi-
cal changes to human cancer cells including cell contrac-
tion followed by cell death within 24 h. To study this ac-
tivity in more detail, HeLa cells were cultured in the pres-
ence of DMSO or various cytotoxins. After 12h, cells were
fixed and stained with DAPI to mark the nuclei and with
an F-actin marker (FITC-phalloidin or Alexa448-
phalloidin) to observe the cytoskeleton. As shown in Fig
3a, DMSO-treated cells displayed an oval-shaped nucleus
and diffuse actin cytoskeleton. Cell membranes featured
cortical actin structures as expected. By contrast, ent-
kibdelone C [(-)-3] resulted in contraction of the cells and
the formation of actin fibers. Specifically, 7% of untreated
cells contained these stress fibers whereas greater than
60% of cells treated with 50 nM kibdelone C featured the-
se structures after 12h (Fig 3A, B; quantification in Fig S5).
The formation of stress fibers appears to be associated
with the tetrahydroxanthone family of natural products
and is not a general property of cytotoxins. For example,
simaomicin α caused an identical phenotype (Fig 3D), as
did the active analogs described above (Fig S6). As shown
Kibdelone does not inhibit topoisomerase. Several
polycyclic, cytotoxic natural products inhibit topoisomer-
ase, offering a potential mechanism for the toxicity ob-
served with the kibdelone family of cytotoxins.43 To ex-
plore this possibility, methyl kibdelone C (57) was incu-
bated with topoisomerase I and supercoiled DNA (Sc
DNA). Topoisomerase I relaxes supercoiled DNA by in-
troducing single-strand DNA breaks, allowing the cleaved
strand to unwind, and rejoining the DNA fragments. As
shown in Fig 2a, Sc DNA (lane 1) is readily relaxed to a
series of topoisomers in the absence of methyl kibdelone
C (lane 6), and this activity is not affected by up to 10 ꢀM
compound (lanes 7-11). While topoisomerase I induces
single strand DNA breaks, topoisomerase II creates dou-
ble-strand breaks. This activity can be assayed using cate-
nated kinetoplast DNA (kDNA), a series of interlocked
circles of DNA (Fig 2b, lane 1). This high molecular weight
complex can be decatenated with topoisomerase II to
generate circular DNA (lane 2). While etoposide was
found to interfere with this process by generating linear
ACS Paragon Plus Environment

Page 9 of 12
Journal of the American Chemical Society
in Fig S6, simplified analogs 57, 76 and 77 showed similar
Conclusions
1
2
3
4
5
6
7
8
cell contraction/rounding and stress fiber formation as
the natural products themselves. Two representative anti-
cancer drugs did not cause actin stress fibers to form. For
example, bortezomib, a proteasome inhibitor, instead led
to cell rounding without nuclear disruption. Paclitaxel, a
microtubule stabilizing agent, created disorganized nuclei
and cell rounding (Fig 3E, F). This comparison indicates
that the stress fibers and morphological effects are not
general activities associated with cell death. Rather, they
appear to be specific to the mode of action of the cytotox-
ic tetrahydroxanthones.
A convergent total synthesis of kibdelone C allowed ac-
cess to several derivatives of the natural product. Substit-
uents around the convex periphery appear to have little
impact on the biological activity. Thus, the methyl and
propyl groups of the isoquinolinones of simaomicin α and
kibdelone C both lead to potent toxicity. Similarly, the
chloride of the latter natural product is dispensable. The
B-ring can tolerate a cyclic ketal as in simaomicin α, phe-
nol as in kibdelone C, or a methoxy group as in several
synthetic derivatives. Similarly, the hydroxylation around
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
A
B
DMSO
C
50 nM (-)-3
D
10 ꢀm
100 nM
50 nM (-)-3 simaomicin α (4)
10 ꢀm
10 ꢀm
10 ꢀm
Figure 4. (-)-Kibdelone C (3) has no effect on the
polymerization rate of actin. The main chart shows the
changes in fluorescence as recombinant actin was
polymerized over time (2 ꢀM actin including 10%-pyrene
labeled protein). Single experiments are shown in the
absence or presence of the indicated concentrations of 3.
Inset shows the average rate of polymerization in the
absence or presence of 3 relative to the rate in the absence
of DMSO (n = 3).
F
E
100 nM
bortezomib
100 nM
paclitaxel
10 ꢀm
10 ꢀm
Figure 3. HeLa cells treated with A) DMSO, B) and C)
the F-ring is apparently not required for activity. Both
enantiomers of kibdelone C and simaomicin α are active,
as are analogues missing the C11 and/or C13 hydroxyls.
Most surprising, an aryl F ring lacking hydroxylation was
fully active. By contrast, the hydroxyls on the concave
portion of the natural product appear critical, as shown
by the substantial loss in activity associated with dimethyl
ether 56. Prior studies showed that the EF rings of the
tetrahydroxanthone are indispensible.36 These results
suggest that the polycyclic ring system may essentially act
as scaffolding to present the A- and E-ring carbonyls and
B- and D-ring phenols in a defined orientation. It is per-
haps noteworthy that all of the natural products shown in
Scheme 1 share this functionality. This common structure
could suggest a similar biological mechanism for these
natural products. Moreover, it seems plausible to envision
a role for metal binding by the phenols common the B
and D rings of the tetrahydroxanthone natural products,
but that possibility remains speculative at present.
kibdelone C [(-)-3], D) simaomicin α (4), E) bortezomib or F)
paclitaxel for 12h. Representative stress fibers indicated with
yellow arrow. Staining of nuclei (blue, DAPI) and actin (green,
FITC-phalloidin) reveals cell contraction and stress fiber
formation in the kibdelone and simaomicin α-treated cells.
Having observed effects on the actin cytoskeleton, we
considered whether kibdelone C might directly affect ac-
tin polymerization. To assess this possibility, actin was
polymerized in vitro in the absence or presence of kibde-
lone C. This assay involves incorporation of tracer quanti-
ties of pyrene-labeled actin into growing actin filaments,
which provides a fluorescent readout of polymerization.44
Individual time-course traces are shown in Fig 4, and they
demonstrate that actin polymerization is not affected by
up to 100 nM kibdelone C. The inset shows the average
relative rates of polymerization in the presence or ab-
sence of kibdelone C (n = 3), again indicating no effect.
Taken together, the data suggests that kibdelone and re-
lated compounds affect the actin cytoskeleton, but this
effect does not involve a direct interaction with kibdelo-
ne.45
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 12
1
2
3
4
5
Kumagai, A.; Zhou, X.-p.; Koyota, S.; Sugiyama, T. Cancer Sci.
2009, 100, 322-326.
6
7
8
9
(5) Kobayashi, K.; Nishino, C.; Ohya, J.; Sato, S.; Mikawa, T.;
Shiobara, Y.; Kodama, M. J. Antibiotics 1988, 41, 502-511.
(6) Chu, M.; Truumees, I.; Mierzwa, R.; Terracciano, J.; Patel,
M.; Das, P. R.; Puar, M. S.; Chan, T.-M. Tetrahedron Lett. 1998,
39, 7649-7652.
(7) (a) Kunimoto, S.; Lu, J.; Esumi, H.; Yamazaki, Y.; Kinoshita,
N.; Honma, Y.; Hamada, M.; Ohsono, M.; Ishizuka, M.; Takeuchi,
T. J. Antibiotics 2003, 56, 1004-1011. (b) Kunimoto, S.; Someno, T.;
Yamazaki, Y.; Lu, J.; Esumi, H.; Naganawa, H. J. Antibiotics 2003,
56, 1012-1017. (c) Lu, J.; Kunimoto, S.; Yamazaki, Y.; Kaminishi,
M.; Esumi, H. Cancer Sci. 2004, 95, 547-552.
(8) Nakagawa, A.; Omura, S.; Kushida, K.; Shimizu, H.;
Lukacs, G. J. Antibiotics 1987, 40, 301-308.
(9) Qiao, Y.-F.; Okazaki, T.; Ando, T.; Mizoue, K.; Kondo, K.;
Eguchi, T.; Kakinuma, K. J. Antibiotics 1998, 51, 282-287.
(10) Ratnayake, R.; Lacey, E.; Tennant, S.; Gill, J. H.; Capon, R.
J. Org. Lett. 2006, 8, 5267-5270.
(11) Kelly, T. R.; Jagoe, C. T.; Li, Q. J. Am. Chem. Soc. 1989, 111,
4522-4524.
(12) Rao, A. V. R.; Yadav, J. S.; Reddy, K. K.; Upender, V. Tetra-
hedron Lett. 1991, 32, 5199-5202.
(13) Mehta, G.; Shah, S. R.; Venkateswarlu, Y. Tetrahedron
1994, 50, 11729-11742.
(14) Masuo, R.; Ohmori, K.; Hintermann, L.; Yoshida, S.; Su-
zuki, K. Angew. Chem. Int. Ed. 2009, 48, 3462-3465.
(15) (a) Sloman, D. L.; Mitasev, B.; Scully, S. S.; Beutler, J. A.;
Porco, J. A. Angew. Chem. Int. Ed. 2011, 50, 2511-2515. (b) Sloman,
D. L.; Bacon, J. W.; Porco, J. A. J. Am. Chem. Soc. 2011, 133, 9952-
9955.
Our initial studies ruled out common modes of activity
– DNA binding and topoisomerase. Cell imaging dis-
played profound changes to the actin cytoskeleton, but in
vitro studies showed no direct effect on actin polymeriza-
tion. These data indicate that the kibdelones must be act-
ing upstream of actin regulation. The structural require-
ments for activity and the simplified derivatives described
here should facilitate the discovery of tools to identify the
direct binding partners for these cytotoxins.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ASSOCIATED CONTENT
Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website.
Synthetic procedures, characterization data, supplementary
figures (PDF)
AUTHOR INFORMATION
Corresponding Author
*joseph.ready@utsoutwestern.edu
(16) Wang, Y.; Wang, C.; Butler, J. R.; Ready, J. M. Angew.
Chem. Int. Ed. 2013, 52, 10796-10799.
Notes
The authors claim no competing financial interests.
(17) (a) Yang, J.; Knueppel, D.; Cheng, B.; Mans, D.; Martin, S.
F. Org. Lett. 2015, 17, 114-117. b) Knueppel, D.; Yang, J.; Cheng, B.;
Mans, D.; Martin, S. F. Tetrahedron 2015, 71, 5741-5757.
(18) (a) Duthaler, R. O.; Heuberger, C.; Wegmann, U. H. U.;
Scherrer, V. Chimia 1985, 39, 174-182. (b) Walker, E. R.; Leung, S.
Y.; Barrett, A. G. M. Tetrahedron Lett. 2005, 46, 6537-6540. (c)
Xiao, Z.; Cai, S.; Shi, Y.; Yang, B.; Gao, S. Chem. Commun. 2014,
50, 5254-5257.
(19) Butler, J. R.; Wang, C.; Bian, J.; Ready, J. M. J. Am. Chem.
Soc. 2011, 133, 9956-9959.
(20) Johansson, R.; Karlstroem, S.; Kers, A.; Nordvall, G.; Rein,
T.; Slivo, C. International Patent WO 039139 A1, 2008.
(21) Barfknecht, C. F.; Nichols, D. E. J. Med. Chem. 1971, 14,
370-372.
(22) See supporting information for details.
(23) Hundertmark, T.; Littke, A. F.; Buchwald, S. L.; Fu, G. C.
Org. Lett. 2000, 2, 1729-1731.
(24) Yonezawa, S.; Komurasaki, T.; Kawada, K.; Tsuri, T.; Fuji,
M.; Kugimiya, A.; Haga, N.; Mitsumori, S.; Inagaki, M.; Nakatani,
T.; Tamura, Y.; Takechi, S.; Taishi, T.; Ohtani, M. J. Org. Chem.
1998, 63, 5831-5837.
(25) Casiraghi, G.; Casnati, G.; Puglia, G.; Sartori, G.; Terenghi,
G. J. Chem. Soc. Perkin 1 1980, 1862-1865.
ACKNOWLEDGMENT
Financial support from the Welch Foundation (I-1612), CPRIT
(RP101016) and the NIH (R01 GM102403). We appreciate as-
sistance from Dr. Luke Rice with in vitro tubulin polymeriza-
tion assays, from Yingli Duan culturing cancer cells, and
from Lynda Doolittle with actin polymerization assays.
REFERENCES
(1) Ratnayake, R.; Lacey, E.; Tennant, S.; Gill, J. H.; Capon, R. J.
Chem. Eur. J. 2007, 13, 1610-1619.
(2) Winter, D. K.; Sloman, D. L.; Porco Jr, J. A. Nat. Prod. Rep.
2013, 30, 382-391.
(3) Carter, G. T.; Goodman, J. J.; Torrey, M. J.; Borders, D. B.;
Gould, S. J. J. Org. Chem. 1989, 54, 4321-4323.
(4) (a) Lee, T. M.; Carter, G. T.; Borders, D. B. J. Chem. Soc.
Chem. Commun. 1989, 1771-1772. (b) Maiese, W. M.; Korshalla, J.;
Goodman, J.; Torrey, M. J.; Kantor, S.; Labeda, D. P.; Greenstein,
M. J. Antibiotics 1990, 43, 1059-1063; (c) Koizumi, Y.; Tomoda, H.;
(26) Blazevic, N.; Kolbah, D.; Belin, B.; Sunjic, V.; Kajfez, F.
Synthesis 1979, 161-176.
(27) Bigi, F.; Casiraghi, G.; Casnati, G.; Sartori, G.; Gasparri
Fava, G.; Ferrari Belicchi, M. J. Org. Chem. 1985, 50, 5018-5022.
(28) Lim, S. M.; Hill, N.; Myers, A. G. J. Am. Chem. Soc. 2009,
131, 5763-5765.
ACS Paragon Plus Environment

Page 11 of 12
Journal of the American Chemical Society
1
2
3
4
(29) Shi, Y. Accts. Chem. Res. 2004, 37, 488-496.
(38) Chan, K.-F.; Zhao, Y.; Chow, L. M. C.; Chan, T. H. Tetra-
5
6
(30) The e.r. of tetraol 30 was determined from the bis-
Mosher’s ester as described in ref 28.
hedron 2005, 61, 4149-4156.
(39) (a) For cytotoxicity screening against several NSCLC lines
including the ones used here, see data generated by the Cancer
Target Discovery and Development (CTD2) Network
(https://ctd2.nci.nih.gov/dataPortal/) established by the Nation-
al Cancer Institute’s Office of Cancer Genomics. (b) The-
odoropoulos, P. C.; Gonzales, S. S.; Winterton, S. E.; Rodriguez-
Navas, C.; McKnight, J. S.; Morlock, L. K.; Hanson, J. M.; Cross,
B.; Owen, A. E.; Duan, Y.; Moreno, J. R.; Lemoff, A.; Mirzaei, H.;
Posner, B. A.; Williams, N. S.; Ready, J. M.; Nijhawan, D. Nat.
Chem. Biol. 2016, 12, 218-225.
(40) (a) K Prasad, K. S.; Nayak, R. FEBS Letters 1976, 71, 171-174.
(b) Shafer, R. H.; Burnette, R. R.; Mirau, P. A. Nucleic Acids Res.
1980, 8, 1121-1132. (c) Van Dyke, M.; Dervan, P. Science 1984, 225,
1122-1127.
(41) Sobell, H. M. Proc. Natl. Acad. Sci. 1985, 82, 5328-5331.
(42) Warnick-Pickle, D. J.; Byrne, K. M.; Pandey, R. C.; White,
R. J. J. Antibiotics 1981, 34, 1402-1407.
(43) (a) Latham, M. D.; King, C. K.; Gorycki, P.; Macdonald, T.
L.; Ross, W. E. Cancer Chemother. Pharm. 1989, 24, 167-171. (b)
Capranico, G.; Ferri, F.; Fogli, M. V.; Russo, A.; Lotito, L.;
Baranello, L. Biochimie 2007, 89, 482-489.
(44) Doolittle, L. K.; Rosen, M. K.; Padrick, S. B. Methods Mol.
Biol. 2013, 1045, 273-293.
(45) Kibdeleone C also does not affect the polymerization of
tubulin, another common target of cytotoxic natural products.
See Figure S8.
(31) Nicolaou, K. C.; Li, A. Angew. Chem. Int. Ed. 2008, 47,
6579-6582.
(32) Soheili, A.; Albaneze-Walker, J.; Murry, J. A.; Dormer, P.
G.; Hughes, D. L. Org. Lett. 2003, 5, 4191-4194.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(33) (a) This iodination was discovered serendipitously when a
sample of the reduction product of 45 containing Cu salts (from
a failed attempt at Cu-promoted phenolic coupling cyclization)
was exposed to N-iodosuccinimide. For Cu-catalyzed bromina-
tion and chlorination of arenes, see Yang, L.; Lu, Z.; Stahl, S. S.
Chem. Commun. 2009, 6460-6462. (b) Iodination as a side-
product: Li, X.; Hewgley, J. B.; Mulrooney, C. A.; Yang, J.; Ko-
zlowski, M. C. J. Org. Chem. 2003, 68, 5500-5511. (c) Cu(OH)Cl・
TMEDA: Nakajima, M.; Miyoshi, I.; Kanayama, K.; Hashimoto,
S.-i.; Noji, M.; Koga, K. J. Org. Chem. 1999, 64, 2264-2271.
(34) Seminal contributions on C-H arylation: (a) Hennings, D.
D.; Iwasa, S.; Rawal, V. H. J. Org. Chem. 1997, 62, 2-3. (b) Acker-
mann, L.; Vicente, R.; Kapdi, A. R. Angew. Chem. Int. Ed. 2009,
48, 9792-9826. (c) Gutekunst, W. R.; Baran, P. S. Chem. Soc. Rev.
2011, 40, 1976-1991.
(35) Campeau, L.-C.; Parisien, M.; Jean, A.; Fagnou, K. J. Am.
Chem. Soc. 2005, 128, 581-590.
(36) Zhao, M.-X.; Shi, Y. J. Org. Chem. 2006, 71, 5377-5379.
(37) Winter, D. K.; Endoma-Arias, M. A.; Hudlicky, T.; Beutler,
J. A.; Porco, J. A. J. Org. Chem. 2013, 78, 7617-7626.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 12 of 12
1
2
3
4
Me
N
Pr
O
5
6
7
8
OH
OH
O
O
MeO
9
OMe
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Simplified deriva ve of kibdelone C
Insert Table of Contents artwork here
12
ACS Paragon Plus Environment