A concise approach to the polycyclic scaffold of frondosin D

Article abstract of DOI:10.1021/jo800682n

(Chemical Equation Presented) In a study directed at developing a concise approach to the polycyclic core of frondosin D, a Stille-Heck sequence has been identified that gives direct access to the trimethylbicyclo [5.4.0]undecane ring system common to all frondosins.

Full text of DOI:10.1021/jo800682n

dosins have proved quite a formidable synthetic challenge, and
as of yet, there has been no synthesis of frondosin D or E. In
this report, we describe our approaches to the molecular scaffold
of frondosins D. This work has culminated in a very effective
means of producing the trimethylbicyclo[5.4.0]undecane ring
system common to all frondosins (shown in bold, Figure 1).
A Concise Approach to the Polycyclic Scaffold of
Frondosin D
Kye-Simeon Masters and Bernard L. Flynn*
Department of Medicinal Chemistry, Victorian College of
Pharmacy, Monash UniVersity, 381 Royal Parade,
ParkVille, VIC 3052, Australia
bernard.flynn@Vcp.monash.edu.au
ReceiVed April 2, 2008
FIGURE 1. (+)-Frondosins A-E.4
Our retrosynthetic analysis of frondosin D focuses on
accessing prochiral tetracycle 6 (Scheme 1), which could
feasibly deliver frondosin D through a sequence of asymmetric
double bond reduction,5 phenylmethyl ether cleavage and
oxidation.3h In this report, we will describe our assessment of
different approaches to this key tetracycle 6.
In our initial approach to 6, we anticipated performing ring
closing metathesis (RCM) on 7 similar to that used in our
frondosin B synthesis (Scheme 1).3e We envisaged that chromene
7 could be accessed through Kumada-Corriu coupling of
isopropenylmagnesium bromide with iodide 8, which in turn
could be formed by 1,6-endodigonal iodocyclization of 9, which
results from the 1,2-addition of lithium acetylide 11 to known
ketone 10,6 followed by regioselective elimination of H2O.7
A key step in our initially proposed access to frondosin D
(4) is the 1,6-endodigonal iodocyclization of 9 to give 8. At
the time of conducting this work, such iodocyclizations to form
3-iodochromenes had not been reported and we undertook a
brief model study to verify this key step (Scheme 2).8 This work
revealed that the sequence of 1,2-addition to give 12 (92%),
followed by elimination of H2O to give 13 (82%), albeit with
a small amount of chloroallene 14 (11%) byproduct, could be
readily acheived.9 Iodocyclization of 13 was best achieved using
I2 in CH3CN or THF with K2CO3 as base, affording the 1,6-
endodigonal cyclization product, 2-iodochromene 14 in good
yield (83%). However, we noted that success in this reaction
was dependent on the use of anhydrous solvent, and that a
specific byproduct was formed in wet solvents. This byproduct
was determined to be the 1,5-endodigonal cyclization product
16. This byproduct was formed as the exclusive product of the
reaction when a solvent combination of THF/H2O (99:1) was
In a study directed at developing a concise approach to the
polycyclic core of frondosin D, a Stille-Heck sequence has
been identified that gives direct access to the trimethylbicyclo
[5.4.0]undecane ring system common to all frondosins.
Frondosins A-E, 1-5 (Figure 1), are a family of related
marine sesquiterpenoids first isolated in their dextro-rotatory
form from the sponge Dysidea frondosa.1a Additionally, leVo-
rotatory frondosins A and D were isolated from an unidentified
Eurospongia species.1b Frondosins A-E are compounds of
interest due to their promising interleukin-8 (IL-8) affinity and
protein kinase C inhibition.1a IL-8 antagonists are of particular
interest in view of their antiinflammatory,2a anti-HIV,1b,2b and
antitumor2c-f properties. To date, frondosins A, B, and C have
been synthesized.3 Notwithstanding these successes, the fron-
(1) (a) Patil, A. D.; Freyer, A. J.; Killmer, L.; Offen, P.; Carte, B.; Jurewicz,
A. J.; Johnson, R. K. Tetrahedron 1997, 53, 5047–5060. (b) Hallock, Y. F.;
Cardellina, J. H., II; Boyd, M. R. Nat. Prod. Lett 1998, 11, 153–160.
(2) (a) Seitz, M.; DeWald, B.; Gerber, N.; Baggioni, M. J. Clin. InVest. 1991,
87, 463–469. (b) Lane, B. R.; Paul, K. L.; Block, J.; Andersson, J.; Coffey,
M. J.; Strieter, R. M.; Markovitz, D. M. J. Virol. 2001, 57, 8195–8202. (c)
Karashima, T.; Sweeney, P.; Kamat, A.; Huang, S.; Kim, S. J.; McConkey, D. J.;
Dinney, C. P. Clin. Cancer Res. 2003, 9, 2786–2797. (d) Zhu, Y. M.; Webster,
S. J.; Flower, D.; Woll, P. J. Br. J. Cancer. 2004, 91, 1970–1976. (e) Mian,
B. M.; Dinney, C. P.; Bermejo, C. E.; Sweeney, P.; Tellez, C.; Yang, X. D.;
Gudas, J. M.; McConkey, D. J.; Bar-Eli, M. Clin. Cancer Res. 2003, 9, 3167–
3175. (f) Brat, D. J.; Bellail, A. C.; Van-Meir, E. G. Neuro-oncol. 2005, 7, 122–
133.
(3) (a) Inoue, M.; Frontier, A. J.; Danishefsky, S. J. Angew. Chem., Int. Ed.
2000, 39, 761–764. (b) Inoue, M.; Carson, M. W.; Frontier, A. J.; Danishefsky,
S. J. J. Am. Chem. Soc. 2001, 123, 1878–1889. (c) Hughes, C. C.; Trauner, D.
Angew. Chem., Int. Ed. 2002, 41, 1569–1572. (d) Hughes, C. C.; Trauner, D.
Tetrahedron 2004, 60, 9675–9686. (e) Kerr, D. J.; Willis, A. C.; Flynn, B. L.
Org. Lett. 2004, 6, 457. (f) Martinez, I.; Alford, P. E.; Ovaska, T. V. Org. Lett.
2005, 7, 1133–1135. (g) Li, X.; Kyne, R. E.; Ovaska, T. V. Org. Lett. 2006, 8,
5153. (h) Li, X.; Kyne, R. E.; Ovaska, T. V. Tetrahedron 2007, 63, 1899–1906.
(i) Trost, B. M.; Hu, Y.; Horne, D. M. J. Am. Chem. Soc. 2007, 129, 11781–
11790. (j) Li, X.; Ovaska, T. V. Org. Lett. 2007, 9, 3837–3840.
(4) The absolute stereochemistry depicted for (+)-frondosins is tentatively
depicted as ꢀ-methyl based on previous asymmetric synthesis efforts; see ref 3i
and references cited therein.
(5) Roseblade, S. J.; Pfaltz, A. Acc. Chem. Res. 2007, 40, 1402–1411.
(6) (a) Na¨f, F.; Decorzant, R. HelV. Chim. Acta 1974, 57, 1317. (b) Magatti,
C. V.; Kaminski, J. J.; Rothberg, I. J. Org. Chem. 1991, 56, 3102.
(7) Kno¨lker, H.-J.; Ecker, A.; Neef, G. Tetrahedron 1996, 53, 91–108.
(8) (a) Barluenga, J.; Trincado, M.; Marcos.Arias, M.; Ballestros, A.; Rubio,
E.; Gonzalez, J. M. Chem. Commun. 2005, 2008–2010.
(9) Presumably, 14 is formed by SN2′ substitution of the intermediate sulfenate
ester by a chloride ion. (a) Bhatia, Y. R.; Landor, P. D.; Landor, S. R. J. Am.
Chem. Soc. 1959, 81, 24–29. (b) Mori, K.; Watanabe, H. Tetrahedron 1986, 42,
273–281. (c) Soukup, M.; Widmer, E.; Luka´cˇ, T. HelV. Chim. Acta 1990, 73,
868–873.
10.1021/jo800682n CCC: $40.75
Published on Web 09/18/2008
2008 American Chemical Society
J. Org. Chem. 2008, 73, 8081–8084 8081

SCHEME 1. Retrosynthetic Analysis of Frondosin D
SCHEME 3. Mechanism of Formation of 15 and 16
SCHEME 4. Preparation of Cyclohexanone 10
SCHEME 2. Model Iodocyclization and Coupling Study
SCHEME 5. Attempted Iodocyclization Pathway to 7
employed (70% isolated yield). At no time in this study did we
detect any 1,5-exocyclization product 17.
While the previous studies on 1,6-endodigonal iodocyclization
to 2-iodochromenes did not report the formation of the 1,5-
endodigonal product, some related cyclizations have been
recently reported.10 On the basis of these previous reports,
connection of the alkyne through an electron-releasing ether
group would be expected to direct nucleophilic attack of the
activated alkyne through the ortho position of the phenyl ring
in intermediate 19 to give 20 then 15 (Scheme 3). However,
this ortho-directing effect of this ether tether appears to be over-
ridden by the para-directing effect of the methoxy group,
effectively giving an ipso-attack of the phenyl group upon the
tethered alkyne, giving intermediate 21 in a 1,5-endodigonal
cyclization. In the presence of water, 21 is “trapped”, giving
the spirocyclic product 16.10a In the absence of water, the
initially formed intermediate 21 funnels back through 19 to give
the intermediate 20, which undergoes irreversible proton
elimination to give 15. Alternatively, kinetically formed 21
migrates directly to 20 prior to trapping.
Kumada-Corriu coupling of 15 with isopropenylmagnesium
bromide was successfully employed in giving 18 (69%) (Scheme
2).
On the basis of this successful access of 18, we prepared
cyclohexanone 10, which bears the requisite substitution pattern
for application to the synthesis of tetracycle 7 (Scheme 4).
Compound 10 was accessed from dione 22 by initial conversion
to 23,11 followed by a one-pot addition-elimination-addition
sequence with methylcuprate, giving 10 in good yield (77%).
Unfortunately, application of 10 to the preparation of 7 was
not successful (Scheme 5). Initial lithiation and 1,2-addition of
alkyne 11 to cyclohexanone 10 proceeded well to give alcohol
24 as a single diastereomer (78%), but the elimination reaction
gave an inseparable mixture of products 9, 25, and 26 (both
12
1
diastereomers) in a 5:3:2 ratio by H NMR. Nonetheless,
(10) (a) Appel, T. R.; Yehia, N. A. M.; Baumeister, U.; Hartung, H; Kluge,
R.; Stro¨hl, D.; Fangha¨nel, E. Eur. J. Org. Chem. 2003, 47–53. (b) Barluenga, J.;
Palomas, D.; Rubio, E.; Gonzalez, J. M. Org. Lett. 2007, 9, 2823–2826. (c)
Zhang, X.; Larock, R. C. J. Am. Chem. Soc. 2005, 127, 12230–12231. (d) Ren,
X.-F.; Turos, E.; Lake, C. H.; Churchill, M. R. J. Org. Chem. 1995, 60, 6468.
(11) Piers, E.; Grierson, J. R.; Lau, C. K.; Nagakura, I. Can. J. Chem. 1982,
60, 210–223.
(12) Investigation of alternative conditions included a reduction of temperature
to-78 °C, which gave similar results, and the use of alternatives to SO2Cl2 such
as POCl3, TsCl, or Tf2O led to worse outcomes.
8082 J. Org. Chem. Vol. 73, No. 20, 2008

SCHEME 6. Revised Approach to Frondosin D Core 6
SCHEME 8. Preparation of Stille-Heck Partner 29
SCHEME 9. Stille-Heck Outcomes
SCHEME 7. Preparation of r,ꢀ-Diactivated Chromenes
treatment of the mixture with iodine gave 3-iodochromene 8 in
44% yield from 24. Attempted Kumada-Corriu coupling of 8
with isopropenylmagnesium bromide did not yield the intended
coupling product 7 but gave exclusive formation of 27, which
had resulted from an intermolecular 1,6-exotrigonal Heck
cyclization of 8.
We revised our plan to take advantage of this facile Heck
reaction and to avoid the requirement to prepare 9. In the revised
plan, an R,ꢀ-diactivated chromene 28 is employed in a reaction
sequence involving a Stille then Heck reaction (Stille-Heck
reaction) with 29 to give 30, which is then gem-dimethylated
to give 6 (Scheme 6).13,14
A considerable component of this work focused on selecting
the correct arrangement of activating groups in 28. The
positioning of the functional groups, halide and triflate, in 28a-d
was based on the relative reactivity of these different groups
with palladium(0) (usually I > OTf > Br . Cl), such that a
domino or stepwise Stille-Heck sequence with 29 will proceed
with the correct regioselectivity to give 30.
We prepared 28a and 28b by halocyclization of the io-
doalkyne 31 using N-bromosuccinimde (NBS) and N-chloro-
succinimde (NCS), respectively (Scheme 7).8 1-Trifloxy-2-
bromo/chlorochromenes 28c and 28d were both prepared from
chromanone 32 by initial R-halogenation with CuBr2 or SO2Cl2
giving 33 and 34, respectively, followed by enoltriflation
(Scheme 7).15,16
Cyclohexanedione 35 was converted to bromoenone 36 (93%)
and the bromo group substituted for a stannyl group using
(Me3Sn)2Cu·LiCN, giving 29 (90%).
We next turned to evaluating the proposed Stille-Heck
sequence on the various R,ꢀ-diactivated chromene systems
28a-d (Scheme 9). Our optimal method for performing Stille
couplings involved Pd(dba)2, tri-(2-furyl)phosphine (TFP),
ZnCl2, and copper(I) thiophenecarboxylate (CuTPC) in N-
methylpyrrolidine (NMP) (Method A).18,19 Also, for reactions
where we anticipated that a subsequent Heck reaction may occur
in situ, such as domino reaction of iodobromide 28a or
bromotriflate 28c, we added the hindered base Cy2NMe to
ensure continual recycling of the catalyst (Method B).
In the case of the 1-iodo-2-bromochromene 28a, the use of
Method A or B led to rapid formation of a very complex reaction
mixture with no discernible product formation. When the
corresponding 1-iodo-2-chlorochromene 28b was reacted with
29 using Method A, only a low yield of the coupled product 38
(<5%) was acheived. Attempted coupling of 28b under slightly
modified conditions using Pd(PPh3)4, LiCl, and CuTPC in NMP
(Method C) returned a superior yet unsatisfying yield of 38
(36%). We did not further attempt to optimize coupling of either
of iodochromenes 28a or 28b with 29 but focused instead on
the more stable bromo- and chlorotriflates 28c and 28d. Using
Stannylcyclohexenone 29 was prepared by a variation of the
method previously developed by Piers et al. (Scheme 8).17 1,3-
(13) The use of doubly activated alkenes in sequential Stille-Heck processes
has been previously reported by de Meijere and co-workers: (a) Su¨nnemann,
H.-W.; de Meijere, A. Angew. Chem., Int. Ed. 2004, 13, 895–897. (b) Voigt,
K.; von Zezschwitz, P.; Rosauer, K.; Lansky, A.; Adams, A.; Reiser, O.; de
Meijere, A. Eur. J. Org. Chem. 1998, 1521–1534. (c) von Zezschwitz, P.; Petry,
F.; de Meijere, A. Chem.sEur. J. 2001, 7, 4035–4046.
(14) For related gem-dimethylations, see refs 3e and 3d.
(15) Dewald, H. A.; Heffner, T. G.; Jaen, J. C.; Lustgarten, D. M.; McPhail,
A. T.; Meltzer, L. T.; Pugsley, T. A.; Wise, L. D. J. Med. Chem. 1990, 33,
445–450.
(16) (a) Comins, D. L.; Dehghani, A. Tetrahedron Lett. 1992, 33, 6299–
6302. (b) Pal, K. Synthesis 1995, 1485–1487.
(17) (a) Piers, E.; Morton, H. E. J. Chem. Soc., Chem. Commun. 1978, 1033–
1034. (b) Piers, E.; Morton, H. E.; Chong, J. M. Can. J. Chem. 1987, 65, 78–
87.
(18) (a) Farina, V.; Baker, S. R.; Sapino, C, Jr Tetrahedron Lett. 1988, 29,
6043–6046. (b) Farina, V.; Roth, G. P. Tetrahedron Lett. 1991, 32, 4243–4246.
(c) Kapadia, S.; Krishnan, B.; Wang, C.; Liebeskind, L. S. J. Org. Chem. 1994,
59, 5905–5911.
(19) Allred, G. D.; Liebeskind, L. S. J. Am. Chem. Soc. 1996, 118, 2748–
2749.
J. Org. Chem. Vol. 73, No. 20, 2008 8083

114.0 (CH), 110.5 (CH), 68.7 (CH2), 55.8 (CH3), 38.3 (CH2), 32.7
(CH2), 29.9 (CH2), 27.0 (CH2), 22.7 (CH2); IR (cm-1) 1672; HRMS
(ESI) calcd for [C20H22ClO3]+ (M + H+) 345.1252, found 345.1244.
Method A, 28c coupled to 29 in a domino Stille-Heck sequence
to give 30 as a mixture of double bond isomers (30a/b) in a
respectable yield (50%). Unfortunately, this material was
contaminated by significant amounts of the bromo-reduced
material 37 (23%), and it was not possible to separate these
two products.
Using Method A with 2-chloroalkenyl triflate 28d, we were
able to achieve an excellent yield of the Stille product 38 (88%).
Although alkenyl chlorides are not very reactive substrates for
Heck reactions, by using Pd0(t-Bu3P)2 as a catalyst, the 1,7-
intramolecular Heck cyclization of 38 to 30a/b (mixture of
double bond isomers) was acheived in excellent yield (84%).
This mixture was quantitatively converted to the thermodynami-
cally most stable endo-isomer 30b upon heating in anhydrous
EtOH with catalytic RhCl3.
Ketone 30b was converted to gem-dimethylated tetracycle 6
using a variation of the Reetz conditions developed by Trauner
and co-workers.3c,d This involves initial reaction of the ketone
with MeMgBr to give a tertiary alchohol (not shown), followed
by reaction with Me2TiCl2 to give 6 (77%), the structural
framework of frondosins D and E (4 and 5, respectively).
2-Methoxy-7-methyl-9,11,12,13-tetrahydro-6H-5-oxabenzo[6,7]
cyclohepta[1,2-a]naphthalene-10-one (30b). Method A: (from alk-
enylchloride 38) Cs2CO3 (1.25 g, 3.83 mmol) was flame dried in a
Schlenk tube under vacuum and backfilled with N2(g). NMP (7
mL) was added, and the stirring suspension was degassed [evacuated
and backfilled with N2(g)]. Pd(t-Bu3P)2 (355 mg, 0.70 mmol) was
added and the suspension stirred at 85 °C for 0.2 h (catalyst
dissolves), then alkenylchloride 38 (1.20 g, 3.48 mmol) was added,
and the reaction was stirred for 2 h, after which time the starting
material 38 was consumed by TLC. The reaction mixture was
passed through a plug of alumina (neutral, Brockmann grade I)
rinsing with Et2O. The Et2O filtrate was washed with H2O (4 ×
100 mL) and brine (50 mL) and dried with MgSO4, and the volatiles
were removed under reduced pressure to give a brown crude
product. Column chromatography of the crude on silica gel with
Et2O-PS (9:41 then 11:39) delivered tetracycles 30a/b (901 mg,
2.92 mmol, 84%) as a pale yellow gum.
RhCl3 (3.7 mg, 5 mol %) was added to a stirred solution of
tetracycles 30a/b (57.0 mg, 0.185 mmol) in EtOH (4 mL) under
N2(g). The mixture was heated to 75 °C and stirred for 7 h, then
cooled, poured into H2O (30 mL), and extracted with Et2O (3 ×
10 mL). The combined extracts were dried over MgSO4 and
concentrated under reduced pressure to deliver tetracycle 30b (57.0
mg, 0.185 mmol, 100%) as a thick colorless gum which solidified
upon standing at -20 °C. This compound was recrystallized from
Experimental Section
2-(But-3-enyl)-3-(3-chloro-6-methoxy-2H-chromen-4-yl)cyclohex-
2-enone (38). ZnCl2 (708 mg, 5.00 mmol) was flame dried carefully
in a round-bottomed flask under vacuum and backfilled with N2(g),
then NMP (18 mL) was added, the reaction vessel was evacuated
and backfilled with N2(g), and 2-chloroalkenyl triflate 28d (500
mg, 1.45 mmol), Pd(dba)2 (83 mg, 0.145 mmol), TFP (68 mg, 0.29
mmol), CuTPC (28 mg, 0.147 mmol), and stannane 29 (654 mg,
2.03 mmol) were all added. The reaction mixture was heated to 55
°C in an oil bath under N2(g). After 2.5 h, the starting materials
were consumed (TLC analysis), the reaction was cooled to rt and
poured onto a plug of alumina (neutral, Brockmann grade I), and
the organics were washed through with Et2O (100 mL). The
organics were washed with H2O (4 × 100 mL) and dried over
MgSO4, and the volatiles were removed under reduced pressure.
Column chromatography of the brown crude on silica gel with
EtOAc-PS (1:19 then 7:93) delivered chloride 38 (438 mg, 1.27
mmol, 88%) as a colorless viscous gum: 1H NMR (300 MHz,
CDCl3) δ 6.88 (d, J ) 8.7 Hz, 1H), 6.77 (dd, J ) 8.7, 3.0 Hz, 1H),
6.45 (d, J ) 3.0 Hz, 1H), 5.82-5.65 (mc, 1H), 4.98-4.88 (m, 2H),
4.86 (d, J ) 5.4 Hz, 2H), 3.78 (s, 3H), 2.65-2.55 (m, 1H), 2.59 (t,
J ) 6.9 Hz, 2H), 2.47-2.30 (m, 2H), 2.26-2.17 (m, 3H),
2.16-2.12 (t, J ) 6.9 Hz, 2H); 13C NMR (75.5 MHz, CDCl3) δ
183.8 (CdO), 154.7 (C), 150.2 (C), 146.5 (C), 139.0 (C), 138.5
(CH), 132.0 (C), 121.6 (C), 121.4 (C), 116.8 (CH), 114.5 (CH2),
1
Et2O-PS to pale-yellow square crystals: mp 119.5-121.5 °C; H
NMR (300 MHz, CDCl3) δ 6.93 (d, J ) 8.7 Hz, 1H), 6.92 (dd, J
) 8.7, 2.7 Hz, 1H), 6.59 (d, J ) 2.7 Hz, 1H), 5.75 (dt, J ) 7.2, 1.2
Hz, 1H), 5.07 (d, J ) 14.4 Hz, 1H), 4.53 (d, J ) 14.4 Hz, 1H),
3.78 (s, 3H), 3.60 (dd, J ) 12.6, 8.1 Hz, 1H), 2.89 (m, 1H),
2.70-2.30 (m, 3H), 2.2-1.85 (m, 2H), 2.00 (s, 3H), 1.65 (dd, J )
12.6, 8.1 Hz, 1H); 13C NMR (75.5 MHz, CDCl3) δ 197.6 (CdO),
154.5 (C), 148.4 (C), 146.9 (C), 138.5 (C), 134.6 (C), 134.1 (C),
131.0 (C), 130.1 (CH), 124.5 (C), 117.1 (CH), 114.3 (CH), 113.2
(CH), 66.8 (CH2), 55.8 (CH3), 38.4 (CH2), 29.7 (CH2), 24.1 (CH2),
23.7 (CH2), 20.8 (CH3); IR (cm-1) 1671; LRMS (+ESI) calcd
308.9; HRMS (+ESI) calcd [C20H21O3]+ (M + H+) 309.1485,
found 309.1486.
Supporting Information Available: Experimental details and
copies of 1H NMR and 13C NMR spectra for all other
compounds can be found in the Supporting Information. This
material is available free of charge via the Internet at
http://pubs.acs.org.
JO800682N
8084 J. Org. Chem. Vol. 73, No. 20, 2008