Synthetic studies on Et-743. Assembly of the pentacyclic core and a formal total synthesis

Article abstract of DOI:10.1021/jo801159k

(Chemical Equation Presented) A formal total synthesis of the potent anticancer agent Et-743 is described. The tetrahydroisoquinoline core is stereoselectively constructed using a novel radical cyclization of a glyoxalimine. Further elaboration of this core rapidly accessed the pentacyclic core of Et-743, but a mixture of regiosisomers was obtained in the key Pictet-Spengler ring closure. A known advanced intermediate in the synthesis of Et-743 was intercepted, constituting a formal synthesis of the molecule.

Full text of DOI:10.1021/jo801159k

Synthetic Studies on Et-743. Assembly of the Pentacyclic Core and a
Formal Total Synthesis
Dan Fishlock† and Robert M. Williams*,†,‡
Department of Chemistry, Colorado State UniVersity, Fort Collins, Colorado 80523, and UniVersity of
Colorado Cancer Center, Aurora, Colorado 80045
rmw@lamar.colostate.edu
ReceiVed May 30, 2008
A formal total synthesis of the potent anticancer agent Et-743 is described. The tetrahydroisoquinoline
core is stereoselectively constructed using a novel radical cyclization of a glyoxalimine. Further elaboration
of this core rapidly accessed the pentacyclic core of Et-743, but a mixture of regiosisomers was obtained
in the key Pictet-Spengler ring closure. A known advanced intermediate in the synthesis of Et-743 was
intercepted, constituting a formal synthesis of the molecule.
Introduction
Members of the tetrahydroisoquinoline family of alkaloids
display a wide range of biological properties such as antitumor
and antimicrobial activities.1 Of particular significance within
this family is Ecteinascidin 743 (Et-743, 1, Figure 1,) which
has been demonstrated to possess extremely potent cytotoxic
activity with in vitro IC50 values in the 0.1-1 ng/mL range in
several cell lines (as a measure of RNA, DNA, and protein
synthesis inhibition).2 Et-743 is currently in phase II/III clinical
trials for the treatment of ovarian, endometrial, and breast
cancers and several sarcoma lines.3The scarcity of the natural
product from marine sources renders Et-743 an important target
for synthesis. Corey and co-workers reported the first total
synthesis of Et-743 in 36 steps with an overall yield of 0.72%.4a
FIGURE 1. Ecteinascidin 743 (1).
A second-generation synthesis improved the overall yield to
2.04%, but still required 36 steps.4b Fukuyama and co-workers
achieved a total synthesis of Et-743 in 50 steps and 0.56%
overall yield.5 More recently, Zhu and co-workers reported a
31 step synthesis in 1.7% overall yield.6 Most recently,
Danishefsky and co-workers reported a formal total synthesis7via
a pentacyclic compound that intercepted a late-stage intermediate
of Fukuyama’s route.5 Despite the advancements in the state-
of-the-art in total synthetic approaches to Et-743, the clinical
supply of this complex drug is semisynthetically derived from
natural cyanosafracin B, obtained by fermentation as reported
by PharmaMar.8
† Colorado State University.
‡ University of Colorado Cancer Center.
(1) Scott, J. D.; Williams, R. M. Chem. ReV. 2002, 102, 1669–1730.
(2) (a) Rinehart, K. L.; Holt, T. G.; Fregeau, N. L.; Keifer, P. A.; Wilson,
G. R.; Perun, T. J.; Sakai, R.; Thompson, A. G.; Stroh, J. G.; Shield, L. S.;
Seigler, D. S. J. Nat. Prod. 1990, 53, 771–792. (b) Rinehart, K. L.; Holt, T. G.;
Fregeau, N. L.; Stroh, J. G.; Keifer, P. A.; Sun, F.; Li, L. H.; Martin, D. G. J.
Org. Chem. 1990, 55, 4512–4515. (c) Wright, A. E.; Forleo, D. A.; Gunawardana,
G. P.; Gunasekera, S. P.; Koehn, F. E.; McConnell, O. J. J. Org. Chem. 1990,
55, 4508–4512. (d) Guan, Y.; Sakai, R.; Rinehart, K. L.; Wang, A. H.-J.
J. Biomol. Struct. Dyn. 1993, 10, 793–818. (e) Aune, G. J.; Furuta, T.; Pommier,
Y. Anti-Cancer Drugs 2002, 13, 545–555. (f) Rinehart, K. L. Med. Drug ReV.
2000, 1–27.
Our laboratory has been developing methodology for the
assembly of tetrahydroisoquinoline natural products and has
9594 J. Org. Chem. 2008, 73, 9594–9600
10.1021/jo801159k CCC: $40.75 2008 American Chemical Society
Published on Web 08/08/2008

Synthetic Studies on Et-743
(E).14 This was achieved via an intramolecular 6-endo radical
closure on a glyoxalimine, and the desired 1,3-cis-diastereomer
was obtained exclusively. The synthesis of a tetrahydroiso-
quinoline such as E can be problematic because of the acid
sensitivity of the benzylic hydroxyl, particularly because it is
ortho to the phenolic hydroxyl of the aromatic ring and thus
has a high propensity for ortho-quinonemethide formation.
Herein, we report a formal total synthesis of Et-743 as part of
our ongoing efforts to devise a practical and scalable synthesis
of this potent antitumor antibiotic that would be amenable to
the construction of analogues with anticipated potent cytotoxic
activity.
SCHEME 1. Synthetic Plan
Results and Discussion
The synthesis began with Borchardt’s catechol 315 that was
regioselectively brominated to generate 4 (92% yield) (Scheme
2.) Conversion of catechol 4 to the methylenedioxy aldehyde 5
was accomplished using bromochloromethane in a sealed vessel
(69% yield). Baeyer-Villiger oxidation using m-CPBA provided
bromophenol 6 as an off-white solid following hydrolysis of
the resulting formate intermediate (73% yield). Stereoselective
aldol condensation of the titanium phenolate of 6 with (R)-
Garner’s aldehyde (7)16 using a modification of Casiraghi’s
method17 provided the anti-product 8 followed by allyl protec-
tion of the phenolic oxygen delivering 9 (65% yield, two steps).
Subsequent hydrolysis of the oxazolidine and formation of the
trans-acetonide (84% yield, two steps) provided 10 as an oil
that cleanly underwent N-Boc deprotection using Ohfune’s
protocol18 (76% yield) to afford free amine 11 as a stable
crystalline solid. From 11, the glyoxalimine intermediate 13 (see
Scheme 3) was readily obtained by condensation with ethyl
glyoxalate. Following isolation by filtration through Celite and
concentration, the radical ring closure commenced with slow
addition of Bu3SnH and AIBN via syringe pump to a refluxing
dilute solution of the glyoxalimine (13). Concentration and KF/
silica chromatography19 of the crude reaction mixture provided
solid 12 as a single diastereomer (58% yield, two steps). The
reported syntheses of D,L-quinocarcinamide,9 (-)-tetrazomine,10
(-)-renieramycin G,11 (-)-jorumycin,11 and cribrostatin 4
(renieramycin H).12 As a part of this program, we have targeted
Et-743 by a convergent route that envisioned coupling of a
suitably functionalized tyrosine derivative13 with the complete
tetrahydroisoquinoline core (Scheme 1.) We have successfully
deployed this strategy, with the present objective of construction
of pentacycle A, in the synthesis of (-)-renieramycin G and
(-)-jorumycin.11,12
We have previously reported a concise and highly diastereo-
selective synthesis of the tetrahydroisoquinoline core of Et-743
(3) (a) Zelek, L.; Yovine, A.; Etienne, B.; Jimeno, J.; Taamma, A.; Mart´ın,
C.; Spielmann, M.; Cvitkovic, E.; Misset, J. L. Ecteinacidin-743 in Taxane/
Antracycline Pretreated AdVanced/Metastatic Breast Cancer Patients: Prelimi-
nary Results with the 24 h Continuous Infusion Q3 Week Schedule; 36th Annual
Meeting of the American Society of Clinical Oncology, New Orleans, May 20-
23, 2000; Abstract number 592. (b) Delaloge, S.; Yovine, A.; Taamma, A.; Cottu,
P.; Riofrio, M.; Raymond, E.; Brain, E.; Marty, M.; Jimeno, J.; Cvitkovic, E.;
Misset, J. L. Preliminary EVidence of ActiVity Of Ecteinacidin-743 (ET-743) in
HeaVily Pretreated Patients with Bone and Soft Tissue Sarcomas; 36th Annual
Meeting of the American Society of Clinical Oncology, New Orleans, May 20-
23, 2000; Abstract number 2181. (c) Le Cesne, A.; Judson, I.; Blay, J. Y.;
Radford, J.; an Oosterom, A.; Lorigan, P.; Rodenhuis, E.; Donato Di Paoula,
E.; Van Glabbeke, M.; Jimeno, J.; Verweij, J. Phase II of ET-743 in AdVance
Soft Tissue Sarcoma in Adult:A STBSG-EORTC Trial; 36th Annual Meeting of
the American Society of Clinical Oncology, New Orleans, May 20-23, 2000;
Abstract number 2182. (d) Aune, G. J.; Furuta, T.; Pommier, Y. Anti-Cancer
Drugs 2002, 13, 545–555.
1
relative stereochemistry of 12 was secured H NMR data and
corroborated by X-ray crystallography. Examination of the crude
1H NMR revealed the formation of a single diastereomer in the
radical closure and exclusive 6-endo regioselectivity. In addition
1
to 12 and tin impurities visible in the H NMR spectrum, an
aromatic proton arising from hydride quenching of the aryl
radical revealed a ∼6.6:1 ratio of 12 to reduced substrate. Slower
addition rates (over 18 or 36 h) did not improve the isolated
yield of 12.
(4) (a) Corey, E. J.; Gin, D. Y.; Kania, R. S. J. Am. Chem. Soc. 1996, 118,
9202–9203. (b) Martinez, E. J.; Corey, E. J. Org. Lett. 2000, 2, 993–996.
(5) Endo, A.; Yanagisawa, A.; Abe, M.; Tohma, S.; Kan, T.; Fukuyama, T.
J. Am. Chem. Soc. 2002, 124, 6552–6554.
(6) Chen, J.; Chen, X.; Bois-Choussy, M.; Zhu, J. J. Am. Chem. Soc. 2006,
128, 87–89.
(7) Zheng, S.; Chan, C.; Furuuchi, T.; Wright, B. J. D.; Zhou, B.; Guo, J.;
Danishefsky, S. J. Angew. Chem., Int. Ed. 2006, 45, 1754–1759.
(8) Cuevas, C.; Pe´rez, M.; Mart´ın, M. J.; Chicharro, J. L.; Ferna´ndez-Rivas,
C.; Flores, M.; Francesch, A.; Gallego, P.; Zarzuelo, M.; de la Calle, F.; Garc´ıa,
J.; Polanco, C.; Rodr´ıguez, I.; Manzanares, I. Org. Lett. 2000, 2, 2545–2548.
(9) Flanagan, M. E.; Williams, R. M. J. Org. Chem. 1995, 60, 6791–6797.
(10) (a) Scott, J. D.; Williams, R. M. Angew. Chem., Int. Ed. 2001, 40, 1463–
1465. (b) Scott, J. D.; Williams, R. M. J. Am. Chem. Soc. 2002, 124, 2951–
2956.
(11) (a) Lane, J. W.; Chen, Y.; Williams, R. M. J. Am. Chem. Soc. 2005,
127, 12684–12690. (b) Lane, J. W.; Estevez, A.; Mortara, K.; Callan, O.; Spencer,
J. R.; Williams, R. M. Bioorg. Med. Chem. Lett. 2006, 16, 3180–3183. (c)
Vincent, G.; Lane, J. W.; Williams, R. M. Tetrahedron Lett. 2007, 48, 3719–
3722. (d) Jin, W.; Metobo, S.; Williams, R. M. Org. Lett. 2003, 5, 2095–2098.
(12) (a) Vincent, G.; Williams, R. M. Angew. Chem., Int. Ed. 2007, 46, 1517–
1520. (b) Vincent, G.; Chen, Y.; Lane, J. W.; Williams, R. M. Heterocycles
2007, 72, 385–398.
(13) Jin, W.; Williams, R. M. Tetrahedron Lett. 2003, 44, 4635–4639.
(14) Fishlock, D.; Williams, R. M. Org. Lett. 2006, 8, 3299–3301.
(15) Prep a red from 2,3-dimethoxytoluene according to: Shinhababu, A. K.;
Ghosh, A. K.; Borchardt, R. T. J. Med. Chem. 1985, 28, 1273–1279.
(16) (R)-Garner‘s aldehyde was synthesized from D-serine according to:
Garner, P.; Park, J. M. Org. Synth. 1992, 70, 18–28.
(17) Casiraghi, G.; Cornia, M.; Rassu, G. J. Org. Chem. 1988, 53, 4919–
4922. In our case, sonication was not required as described in the original paper.
(18) Sakaitani, M.; Ohfune, Y. J. Org. Chem. 1990, 55, 870–876.
(19) Effective removal of tin impurities from Bu3SnH-mediated reactions:
Harrowven, D. C.; Guy, I. L. Chem. Commun. 2004, 1968–1969.
(20) (a) De Paolis, M.; Chiaroni, A.; Zhu, J. Chem. Commun. 2003, 2896–
2897. (b) Chen, X.; Chen, J.; De Paolis, M.; Zhu, J. J. Org. Chem. 2005, 70,
4397–4408.
(21) (a) Herberich, B.; Kinugawa, M.; Vazquez, A.; Williams, R. M.
Tetrahedron Lett. 2001, 42, 543–546. (b) Jin, W; Metobo, S.; Williams, R. M.
Org. Lett. 2003, 5, 2095–2098.
J. Org. Chem. Vol. 73, No. 24, 2008 9595

Fishlock and Williams
SCHEME 2. Tetrahydroisoquinoline Core of Et-743
SCHEME 3. Pentacycle Construction
The diastereoselectivity of this reaction stands apart from
numerous Pictet-Spengler cyclizations on related substrates that
provide tetrahydroisoquinolines exclusively as the 1,3-trans-
diastereomers.11,20,21 We qualitatively rationalize the cis-dias-
tereoselectivity of this radical process using the Beckwith-Houk
chairlike transition state model for intramolecular radical ring
closures (Figure 2).22 The lowest-energy chair conformation (A)
adopted by the trans-acetonide of the substrate (13) results in
both the glyoxalimine and aryl substituent being in an equatorial
disposition. In this conformation, 1,3-diaxial steric effects and
allylic strain interactions are minimized in the ring-forming
transition state. To further examine the stereocontrol imparted
by the acetonide ring, the cis-acetonide substrate 14 was
prepared (using Casiraghi’s method from the magnesium
phenolate of 6).17 Substrate 14 resulted in a 1:1 mixture of 1,3-
(22) (a) Beckwith, A. L. J.; Schiesser, C. H. Tetrahedron 1985, 41, 3925–
3941. (b) Spellmeyer, D. C.; Houk, K. N. J. Org. Chem. 1987, 52, 959–974.
(23) Chen, X.; Zhu, J. Angew. Chem., Int. Ed 2007, 46, 3962–3965.
9596 J. Org. Chem. Vol. 73, No. 24, 2008

Synthetic Studies on Et-743
FIGURE 2. Transition state models to rationalize the observed 1,3 relative stereochemistry in the tetrahydroisoquinoline radical ring closure.
trans- and 1,3-cis-tetrahydroisoquinolines (15 and 16, both are
known componds),20 which suggests the energy difference
between transition state conformations B and C (axial aryl group
versus axial glyoxalimine) is negligible.
diol, but oxidation of the primary alcohol (in the presence of
the free benzylic alcohol) could not, in our hands, be cleanly
accomplished. The methyl ether was thus a fortuitous selective
protection of the benzylic alcohol, ultimately simplifying the
subsequent manipulations.
As shown in Scheme 3, reduction of the tetrahydroisoquino-
line ester (12)14 with LAH, followed by immediate protection
as the benzyl ether (17), proceeded cleanly in 77% yield over
two steps. The substituted tyrosine amino acid component (18)
has been previously reported by us, utilizing the oxazinone
template technology developed in our laboratory that was
benzylated with the advanced aromatic side chain.13 Thus,
acylation of the tetrahydroisoquinoline (17) was achieved via
the N-Fmoc-protected amino acid chloride (18) to give amide
19a without epimerization. The use of the N-Boc free acid with
a variety of coupling agents (DCC, HOBt, HATU) all resulted
in very sluggish reactions with poor isolated yields, as did the
attempted use of the N-Boc acid fluoride.
Facile deprotection of the O-TBS-protected phenol using
TBAF was followed by oxidation of the primary alcohol using
Swern conditions in high yield. This oxidation product (20)
existed as an equilibrium mixture of the aldehyde and the
corresponding hemiaminal species (illustrated) as observed by
1H NMR, which was otherwise additionally complicated by
amide and carbamate rotamers. The attempted oxidation using
either Dess-Martin periodinane or TPAP/NMO both failed,
leading to extensive decomposition. Following filtration of crude
20 through a plug of silica gel, this substance was immediately
subjected to the Pictet-Spengler conditions.
The objective at this stage was to achieve the Pictet-Spengler
reaction via N-Boc deprotection, iminium ion formation, and
electrophilic aromatic substitution to provide the desired pen-
tacyclic core of Et-743. This meant that the aromatic substitution
must occur ortho to the free phenol, and the benzylic methyl
ether must survive these conditions. Unfortunately, it had already
been demonstrated above that the electron-rich aromatic ring
of the tetrahydroisoquinoline component was highly sensitive
to protic conditions, leading to ortho-quinonemethide formation.
Treatment of 19a with diethylamine provided the free amine,
which was not isolated in favor of immediate evaporation of
excess base and solvent and subsequent Boc protection of the
crude material. Isolation following chromatography provided
compound 19b in 90% yield. Removal of the acetonide from
19b was accomplished using the extremely mild, albeit slow,
method of stirring with Dowex 50W-X8 cationic resin in
methanol. Complete deprotection took 8-12 h, but the yield
was quantitative following simple filtration and concentration.
Instead of providing the usual diol product, this substrate
incorporated methanol at the benzylic position thus providing
the methyl ether as a ∼1:1 mixture of diastereomers. Not
unexpectedly, the benzylic stereogenic center loses stereochem-
ical integrity since the methanol is incorporated via the incipient
ortho-quinonemethide species arising from the acidic depro-
tection conditions.
Indeed, when substrate 20 was treated with trifluoroacetic
acid in methylene chloride, it cleanly underwent the expected
pentacycle formation furnishing 21 + 22 as a ∼0.72:1 ortho:
para mixture of regioisomers in 72% combined yield. As
anticipated, the benzylic methoxy group was eliminated presum-
ably via the incipient ortho-quinonemethide species that forms
under these conditions. In a fruitless effort to circumvent the
vexing olefin formation, pentacycle formation with TFA in dry
methanol resulted in extensive decomposition of the substrate.
Alternatively, we found that the use of water/dichloromethane
with cationic resin on 19b could provide the corresponding free
J. Org. Chem. Vol. 73, No. 24, 2008 9597

Fishlock and Williams
SCHEME 4. Pictet-Spengler Regioselectivity
As part of these synthetic investigations, the intermediate 23
was prepared (in parallel with the O-benzyl-protected synthesis)
bearing an O-allyl-protected hydroxymethyl at C1 of the THIQ
core. This substrate was used to examine the regioselectivity
of the pentacycle-forming ring closure and was utilized to
acquire detailed 1H NMR data, while the O-benzyl material 21
was carried forward in the synthesis. One interesting observation
was the behavior of compound 25 containing the O-Boc
carbonate-protected phenolic oxygen. Treatment of 25 under
the same reaction conditions provided the pentacycles 26 + 27
in a 2:3 ratio of ortho:para regioisomers. The O-Boc carbonate
would presumably be deprotected quickly under these conditions
to reveal the free phenol-containing reactive species, thus
resulting in a comparable regioselectivity as observed with
substrate 20 (beginning with a free phenol on the aryl nucleo-
phile moiety). Notably, however, when substrate 25 was treated
with K2CO3/MeOH, the O-Boc carbonate was selectively
removed (28) with apparent olefin formation prior to the
Pictet-Spengler reaction and pentacycle formation. Treatment
of 28 with TFA in dichloromethane produced the pentacycles
26 + 27 in a 1:3 ratio of ortho:para regioisomers, supporting
the hypothesis that some regioselectivity in the closure might
arise from an intramolecular H bond with a heteroatom at the
benzylic position.11c
attempt to reproduce the Zhu conditions on substrate 24 using
0.01% methanesulfonic acid in CH2Cl2 did not affect the
regioselectivity of this reaction. The substrate was consumed
to provide some material that appeared to still contain the N-Boc
1
protecting group, but the H NMR of the crude product was
prohibitively complex. Subsequent treatment of this reaction
crude with a TFA/anisole/CH2Cl2 mixture provided the penta-
cycles 26 + 27 with ∼1:1 regioselectivity. The same ratio is
obtained if the TFA/anisole conditions are used directly on
substrate 24.
In order to redeem the synthetic utility of the olefinic products
(21 or 26), our attention was captured by the recent formal
synthesis of Et-743 reported by the Danishefsky group7 in which
the olefin (29, Scheme 5) underwent facile oxidation using
DMDO and immediate hydride reduction delivering the benzylic
alcohol 30. With the availability of this methodology in the
literature, our efforts were briefly redirected to convert our
synthetic pentacycle 21 into compound 29 which would
constitute a formal total synthesis of Et-743 by relay through
the Danishefsky7 and then Fukuyama5 syntheses, respectively.
In the event, the desired pentacycle 21 (Scheme 3) was
N-protected as the trichloroethyl carbamate (Troc), and the
phenolic residue was protected as the corresponding O-benzyl
ether in 85% yield for the two steps (Scheme 5). Removal of
the O-allyl group under standard conditions followed by
reprotection as the corresponding MOM ether provided com-
pound 29 (56% yield for the two steps). Compound 29 perfectly
matched Danishefsky’s substrate by 1H, 13C NMR, and optical
rotation, confirming the structure of compound 29.
In their synthesis of renieramycin H, the Zhu group has
interestingly reported control of Pictet-Spengler regioselectivity
in a related system by variation of acid concentration (Scheme
4).23 It was found in that case that lowering the concentration
of methanesulfonic acid to 0.01% in CH2Cl2 could invert the
ortho:para selectivity from 3.4:1 to 2:3. Furthermore, the use
of acetonitrile as the solvent instead of dichloromethane favored
the undesired isomer, giving ortho:para selectivity of 1:10. Our
Since Danishefsky has previously converted7 compound 29
into a late-stage intermediate in Fukuyama’s total synthesis5
(namely, compound 30, Scheme 5), this two-stage relay of our
9598 J. Org. Chem. Vol. 73, No. 24, 2008

Synthetic Studies on Et-743
SCHEME 5. Formal Synthesis of Et-743 via 21 to 29 and Danishefsky to Fukuyama Relay
equiv). After stirring for 12 h, the reaction was concentrated and
synthetic 21 thus constitutes a formal total synthesis of Et-743
and provides firm structural corroboration of our synthetic
material and methods.
immediately purified by flash chromatography (9:1 to 5:1 hexanes:
EtOAc, silica gel) to provide 19b as a clear colorless oil (115 mg,
90% over 2 steps): Rf ) 0.43 (3:1 hexanes:EtOAc, UV, CAM);
[R]25D -26.6 (c 1.0, CH2Cl2); IR (thin film) 3319, 2930, 2858, 1711,
While the present formal synthesis reveals that our glyoxal-
imine radical cyclization technology14 holds considerable
potential for the efficient total synthesis of Et-743 and congeners,
we are currently endeavoring to improve the regioselectivity
of the key pentacycle formation (20 to 21) as well as refining
the overall synthetic efficiency of our approach. These objectives
are currently under study in our laboratory and will be reported
in due course.
1
1646 cm-1; H and 13C NMR spectra are extremely complex due
1
to amide and carbamate rotamers; see the H spectra (CDCl3, rt)
and (DMSO-d6, 373 K) and 13C spectrum (CDCl3, rt) in the
Supporting Information; HRMS(ESI/APCI+) m/z calcd for
C48H66N2O11NaSi (M + Na)+ 897.4328, found 897.4310.
Compounds 21 and 22. Boc (OBn) peptide 19b (115 mg, 0.132
mmol) was dissolved in dry MeOH (5 mL), and Dowex 50W-X8
cationic resin (100 mg) was added (the resin was first rinsed with
dry methanol and dried under a stream of argon). After 65 h, the
reaction was complete by TLC and a single streak was observed
(during the course of the reaction, two streaks initially arise due to
a mixture of diol and methyl ether/alcohol products). The reaction
was filtered through a plug of Celite, eluting with dry MeOH, and
the filtrate was combined to provide the methyl ether as clear,
colorless oil (100 mg, 90% yield): Rf ) 0 to 0.35 streak (3:1
hexanes:EtOAc, UV, CAM); HRMS(FAB+) m/z calcd for
C46H65N2O11Si (M + H)+ 849.4358, found 849.4354. The methyl
ether (100 mg) was dissolved in THF (3 mL), and TBAF (1 M in
THF, 125 µL, 1.06 equiv) was added in one portion. After 20 min,
the reaction was concentrated by rotary evaporation and passed
through a silica plug (eluting with 3:1 to 1:1 hexanes:EtOAc) to
provide the free phenol as a clear, colorless oil (82 mg, 95% yield):
Rf ) 0 to 0.43 streak (3:1 hexanes:EtOAc, UV, CAM); HRMS-
(FAB+) m/z calcd for C40H51N2O11 (M + H)+ 735.3493, found
735.3490. Oxalyl chloride (15 µL 1.5 equiv) was added carefully
to a solution of DMSO (25 µL, 3.2 equiv) in CH2Cl2 (1 mL)
previously cooled to -78 °C. A solution of the above alcohol (82
mg, 0.11 mmol) in CH2Cl2 (2 mL) was added dropwise, and the
resulting mixture was stirred at -78 °C for 40 min. The reaction
was quenched with Et3N (125 µL, 8 equiv) and then allowed to
warm to rt. The reaction was diluted with CH2Cl2 and washed with
brine, and then the combined organic fractions were dried (Na2SO4),
filtered, and concentrated. The crude material was passed through
a silica gel plug (eluting with hexanes:EtOAc 1:1) to provide a
yellow oil/foam (82 mg, quant.) of hemiaminal 20 which was used
without further purification: Rf ) 0.5 (hexanes:EtOAc 1:1, UV,
CAM). Hemiaminal 20 (232 mg, 0.32 mmol) was dissolved in
CH2Cl2 (3 mL) to which were added TFA (3 mL) and anisole (0.350
mL) at rt. The reaction was stirred for 14 h and then concentrated
to remove TFA, then redissolved in CH2Cl2 and washed with
saturated aq NaHCO3. The organic fraction was dried (Na2SO4),
filtered, and concentrated. Crude 1H NMR shows 0.72:1 ortho (21)
to para (22) regioisomers. Purification by PTLC (2% MeOH in
Experimental Section
For general methods and considerations, see Supporting Informa-
tion.
Compound 19. The Fmoc-amino acid (410 mg, 0.727 mmol,
1.2 equiv) was dissolved in dry toluene and concentrated (×2),
and then dried under high vacuum. This oil was dissolved in dry
CH2Cl2 (4 mL) to which was added oxalyl chloride (1 mL) at room
temperature, followed by dry DMF (20 µL). After stirring for 20
min, the solution was concentrated and reconcentrated from dry
toluene (×2) and then dried under high vacuum. This acid chloride
18 was dissolved in CH2Cl2 (4 mL) and cooled to 0 °C. THIQ(OBn)
17 (275 mg, 0.61 mmol, 1 equiv) was dissolved in dry CH2Cl2 (2
mL) and 2,6-lutidine (77 µL 0.67 mmol, 1.1 equiv). This solution
was transferred into the acid chloride solution slowly dropwise,
and the resulting mixture was warmed to rt and stirred 7 h (TLC
showed consumption of the THIQ(OBn) starting material). The
reaction was quenched with saturated NH4Cl (aq) and then extracted
to EtOAc (×3). The combined organic fractions were dried
(Na2SO4), filtered, and concentrated to provide a crude orange oil.
Purification by flash chromatography (hexanes:EtOAc 5:1, silica
gel) gave the peptide 19a as a pale yellow oil (426 mg, 70%); Rf
) 0.34 (3:1 hexanes:EtOAc, UV, CAM); [R]25 -22.8 (c 1.14,
D
1
CH2Cl2); IR (thin film) 3289, 2929, 2858, 1717, 1634 cm-1; H
and 13C NMR spectra are extremely complex due to amide and
carbamate rotamers. See the rt (CDCl3) and 373 K (DMSO-d6) 1H
spectra and rt (CDCl3) 13C spectra in the Supporting Information;
HRMS(ESI/APCI+) m/z calcd for C58H68N2O11NaSi (M + Na)+
1019.4485, found 1019.4499.
Compound 19b. Fmoc (OBn) peptide 19a (146 mg, 0.146 mmol)
was dissolved in a 20% v/v solution of Et2NH in CH2Cl2 [CH2Cl2
(2.5 mL) and diethylamine (0.6 mL]. After stirring for 6 h, the
solution was concentrated and then reconcentrated from toluene
and dried under high vacuum. The crude material was dissolved in
EtOH:CH2Cl2 (2:0.5 mL) to which was added Boc2O (370 mg, 10
J. Org. Chem. Vol. 73, No. 24, 2008 9599

Fishlock and Williams
EtOAc) provided the ortho (63 mg) and para products (69 mg) for
a combined yield of 72%. Data for 21: Rf ) 0.61 (EtOAc:MeOH
95:5, UV, CAM); [R]25D -18.0 (c 1.0, CH2Cl2); IR (thin film) 3295,
2932, 1672, 1632, 1455, 1428 cm-1; 1H NMR (400 MHz, CDCl3)
δ 7.14-7.30 (m, 3H), 6.98 (s, 1H), 6.97 (s, 1H), 6.24 (s, 1H), 6.19
(s, 1H), 6.12 (dddd, J ) 16.0, 11.0, 5.4, 5.4 Hz, 1H), 6.05 (dd, J
) 7.2, 5.0 Hz, 1H), 5.85 (br s, 1H), 5.82 (br s, 1H), 5.78 (v br s,
1H), 5.45 (app dd, J ) 17.1, 1.1 Hz, 1H), 5.29 (app dd, J ) 10.3,
0.8 Hz, 1H), 4.9 (s, 1H), 4.30 (app d of AB quartet, J ) 12.3, 5.4
Hz, 2H), 4.03 (d, J ) 6.1 Hz, 1H), 3.87 (AB quartet, J ) 12.1 Hz,
2H), 3.63 (s, 3H), 2.95-3.2 (m, 5H), 2.11 (s, 3H), 2.09 (s, 3H);
13C NMR (100 MHz, CDCl3) δ 168.8, 147.6, 145.6 (×2), 143.4,
139.7, 138.7, 134.5, 133.9, 129.4, 128.8, 128.0 (×2), 127.1, 126.8
(×2), 122.5, 119.3, 117.7, 117.5, 113.0, 108.7, 101.5, 100.2, 75.3,
72.6, 70.0, 60.8, 54.4, 50.0, 46.9, 33.4, 15.9, 9.4. HRMS(ESI/
APCI+) m/z calcd for C34H35N2O7 (M + H)+ 583.2439, found
583.2441. Data for 22: Rf ) 0.5 (EtOAc:MeOH 95:5, UV, CAM);
69.8, 60.5, 54.4/53.7, 50.9/50.1, 47.3/47.2, 32.8/32.4, 16.0, 9.5;
HRMS(ESI/APCI+) m/z calcd for C44H42N2O9Cl3 (M + H)+
847.1950, found 847.1949.
The allyl-protected pentacycle obtained above (20 mg, 0.024
mmol) was dissolved in CH2Cl2 (400 µL), and pyrrolidine (6 µL,
3 eq) was added, followed by Pd(PPh3)4 (2 mg, 0.002 mmol) under
Ar. After 16 h, the reaction was still not complete, so additional
portions of pyrrolidine and palladium catalyst were added. After
stirring an additional 4 h (20 h total), the dark green reaction was
applied directly to flash chromatography (silica gel, hexanes:EtOAc
3:1). The pure fractions were combined to provide the phenol as
yellow oil (11 mg 56%), used without characterization: Rf ) 0.26
(hexanes:EtOAc 3:1, UV, CAM). Phenol (11 mg, 0.014 mmol) was
dissolved in CH2Cl2 (200 µL) to which were added iPr2NEt (12
µL, 0.07 mmol, 5 equiv) and MOMBr (3.3 µL, 0.042 mmol, 3
equiv). The mixture was stirred for 30 min at rt and then quenched
with water and extracted with CH2Cl2 (×3). The combined organic
fractions were dried (Na2SO4), filtered, and concentrated. Flash
chromatography (hexanes:EtOAc 3:1) provided the protected pen-
tacycle 29 (11.5 mg, quant.): Rf ) 0.41 (hexanes:EtOAc 3:1, UV,
[R]25 +47.8 (c 1.45, CH2Cl2); IR (thin film) 3298, 2931, 1671,
D
1631, 1430, 1409 cm-1; 1H NMR (400 MHz, CDCl3) δ 7.11-7.22
(m, 3H), 6.94 (s, 1H), 6.92 (s, 1H), 6.41 (s, 1H), 6.11 (dddd, J )
16.1, 10.6, 5.5, 5.5 Hz, 1H), 6.08 (s, 1H), 6.03 (dd, J ) 6.6, 5.1
Hz, 1H), 5.86 (br s, 1H), 5.83 (br s, 1H), 5.45 (app dd, J ) 17.1,
1.1 Hz, 1H), 5.30 (app dd, J ) 10.4, 0.8 Hz, 1H), 4.65 (s, 1H),
4.36 (app d of A of AB quartet, J ) 12.5, 5.5 Hz, 1H), 4.24 (app
d of B of AB quartet, J ) 12.5, 5.5 Hz, 2H), 4.01 (d, J ) 6.0 Hz,
1H), 3.91 (AB quartet, J ) 12.2 Hz, 1H), 3.56 (s, 3H), 2.95-3.24
(m, 5H), 2.27 (s, 3H), 2.12 (s, 3H); 13C NMR (100 MHz, CDCl3)
δ 168.7, 148.0, 147.6, 145.9, 144.3, 139.7, 138.4, 134.6, 133.7,
129.6, 128.7, 128.1 (×2), 127.2, 126.9 (×2), 124.9, 117.8, 117.1,
113.8, 113.0, 108.8, 101.5, 100.5, 75.4, 72.8, 70.1, 61.0, 54.3, 52.6,
46.8, 35.4, 12.0, 9.4; HRMS (ESI/APCI+) m/z calcd for
C34H35N2O7 (M + H)+ 583.2439, found 583.2429.
CAM); [R]25 +45.4 (c 0.8, CHCl3) [lit. +50 (c 1.0, CHCl3)]; IR
D
1
(thin film) 2932, 1723, 1681, 1654, 1432, 1371 cm-1. H and 13C
NMR spectra perfectly match the data provided by the Danishefsky
group for this intermediate in their formal synthesis (copies of their
spectra included in the Supporting Information):7 1H NMR (400
MHz, CDCl3, mixture of carbamate rotamers) δ 7.56-7.31 (m, 5H),
7.13-7.23 (m, 3H), 6.96 (app br d, J ) 6.9 Hz, 2H), 6.46 (d, J )
9.4 Hz, 1H), 6.01-6.15 (m, 3H), 5.86 (app d, J ) 3.0 Hz, 2H),
5.82 (br s, 1H), 4.97-5.19 (m, 4H), 4.86 (d, J ) 11.9 Hz, 1H),
4.79 (A of AB quart, J ) 12.0 Hz, 1H), 4.68 (B of AB quart, J )
11.9 Hz, 1H), 4.49-4.60 (m, 2H), 4.43 (app d, J ) 6.1 Hz, 1H),
4.01 (d of A of AB quart, J ) 11.8, 4.4 Hz, 1H), 3.85 (B of AB
quart, J ) 12.1 Hz, 1H), 3.71 (app d, J ) 10.6 Hz, 3H), 3.38
(rotomeric s, 3H), 3.03-3.29 (m, 5H), 2.11 (s, 6H); 13C NMR (100
MHz, CDCl3, mixture of carbamate rotamers) δ 166.0/165.9, 151.6/
151.4, 149.9, 148.7/148.2, 147.3, 146.2/146.1, 139.8, 138.4/138.4,
137.8/137.7, 132.6/132.5, 131.1/131.1, 128.8, 128.2, 127.9, 127.2,
126.8, 126.6/126.5, 125.2/125.0, 117.0/116.8, 113.7/113.6, 108.5/
108.4, 103.3/102.7, 101.6, 100.4/100.4, 95.3/95.2, 75.4/75.3, 74.4/
74.0, 72.6/72.6, 69.9/69.9, 60.4, 57.6/57.5, 54.4/53.7, 50.8/50.1,
47.4/47.3, 32.7/32.3, 16.0, 9.9; HRMS(ESI/APPI+) m/z calcd for
C43H42N2O10Cl3 (M + H)+ 851.1900, found 851.1897.
Preparation of Compound 29. The desired ortho-regioisomer
21 (55 mg, 0.095 mmol) was dissolved in CH2Cl2 (2 mL) and
pyridine (11 µL, 0.14 mmol, 1.5 equiv) at 0 °C. TrocCl (13.5 µL,
0.1 mmol, 1.0 equiv) was added and the reaction maintained at 0
°C for 2 h, and then diluted with CH2Cl2 and washed with saturated
aq NH4Cl. The organic layer was dried (Na2SO4), filtered, and then
concentrated. The crude oil was passed through a plug of silica gel
eluting with EtOAc, and then concentrated and dried under vacuum.
The resulting oil was dissolved in CH2Cl2 (600 µL), and MeOH
(200 µL) and K2CO3 (52 mg, 0.38 mmol, 4 equiv) were added
followed by benzyl bromide (22 µL, 0.19 mmol, 2 equiv) and a
catalytic amount of tetrabutylammonium iodide. The resulting
mixture was stirred at rt for 13.5 h then filtered through a pad of
Celite, rinsing with CH2Cl2. Flash chromatography (5:1 hexanes:
EtOAc) provided the N-Troc/O-benzyl product as a pale yellow
oil (68 mg, 85% over 2 steps): Rf ) 0.46 (hexanes:EtOAc 3:1,
Acknowledgment. This paper is fondly dedicated to the
memory of the late Professor A.I. Meyers. We gratefully
acknowledge financial support from the National Institutes of
Health (Grant CA85419). We are grateful to Prof. Samuel J.
Danishefsky of the Memorial Sloan-Kettering Cancer Research
Institute for kindly providing characterization data for compound
29.
UV, CAM); [R]25 +58.1 (c 1.7, CH2Cl2); IR (thin film) 2927,
D
1724, 1681, 1434, 1371 cm-1; 1H NMR (400 MHz, CDCl3, mixture
of carbamate rotamers) δ 7.30-7.56 (m, 3H), 7.14-7.25 (m, 4H),
6.92-7.00 (m, 2H), 6.45 (d, J ) 9.9 Hz, 1H), 6.22 (d, J ) 4.1 Hz,
1H), 6.12 (J ) 16.1 Hz, 1H), 6.01-6.08 (m, 1H), 5.79-5.90 (m,
3H), 4.98-5.29 (m, 5H), 4.85 (d, J ) 12.0, 2.8 Hz, 1H), 4.60 (d,
J ) 11.9, 6.5 Hz, 1H), 3.97-4.11 (m, 3H), 2.85 (app d, J ) 12.1
Hz, 1H), 3.70 (s, 3H), 3.04-3.30 (m, 4H), 2.10 (s, 3H), 2.07 (s,
3H); 13C NMR (100 MHz, CDCl3, mixture of carbamate rotamers)
δ 166.1/166.0, 151.5/151.4, 149.9, 148.9, 148.4, 148.3/148.2, 146.2,
139.6, 138.5, 137.6/137.5, 133.8/133.7, 132.6/132.5, 131.3/131.2,
128.9 (×2), 128.3, 128.1 (×2), 128.0, 127.2, 126.8, 126.6/126.5,
125.4/125.0, 117.6, 117.3, 116.9, 113.3/113.3, 108.6/108.4, 103.3,
102.9, 101.6, 95.3/95.2, 75.4/75.3, 75.2/75.0, 74.6/74.4, 72.6, 69.9/
Note Added after ASAP Publication. Reference 6 contained
an incorrect publication date and the description of the condi-
tions used by Zhu et al. (below Scheme 4) was erroneous in
the version published ASAP August 8, 2008; the corrected
version was published ASAP September 17, 2008.
Supporting Information Available: Complete experimental
procedures and spectroscopic data. This material is available
free of charge via the Internet at http://pubs.acs.org.
JO801159K
9600 J. Org. Chem. Vol. 73, No. 24, 2008