Applications of Zr-catalyzed carbomagnesation and Mo-catalyzed macrocyclic ring closing metathesis in asymmetric synthesis, enantioselective total synthesis of Sch 38516 (fluvirucin B1)

Article abstract of DOI:10.1021/ja972191k

The first enantioselective total synthesis of antifungal agent Sch 38516, also known as fluvirucin B₁, is described. The synthesis includes a convergent asymmetric preparation of amine 17 and acid 18, which are then united to afford diene 62. Metal-catalyzed transformations play a crucial role in the synthesis of the latter moiety. Of particular note are the diastereo- and enantioselective Zr-catalyzed alkylations, a tandem Ti- and Ni-catalyzed process that constitutes a hydrovinylation reaction, and a Ru-catalyzed alcohol oxidation to afford carboxylic acid 18. The requisite carbohydrate 38 is synthesized in a highly diastereo- and enantioselective fashion. Optical purity of the carbohydrate moiety arises from the use of the asymmetric dihydroxylation method of Sharpless; diastereochemical control is achieved through a selective dipolar [3 + 2] cycloaddition with a readily available amine serving as the chiral auxiliary. Union of the appropriately outfitted carbohydrate 71 and diene 62 through an efficient and diastereoselective glycosylation is followed by a remarkably efficient Mo-catalyzed macrocyclization that proceeds readily at room temperature.

Full text of DOI:10.1021/ja972191k

10302
J. Am. Chem. Soc. 1997, 119, 10302-10316
Applications of Zr-Catalyzed Carbomagnesation and
Mo-Catalyzed Macrocyclic Ring Closing Metathesis in
Asymmetric Synthesis. Enantioselective Total Synthesis of
Sch 38516 (Fluvirucin B1)
Zhongmin Xu, Charles W. Johannes, Ahmad F. Houri, Daniel S. La,
Derek A. Cogan, Gloria E. Hofilena, and Amir H. Hoveyda*
Contribution from the Department of Chemistry, Merkert Chemistry Center,
Boston College, Chestnut Hill, Massachusetts 02167
ReceiVed July 1, 1997X
Abstract: The first enantioselective total synthesis of antifungal agent Sch 38516, also known as fluvirucin B1, is
described. The synthesis includes a convergent asymmetric preparation of amine 17 and acid 18, which are then
united to afford diene 62. Metal-catalyzed transformations play a crucial role in the synthesis of the latter moiety.
Of particular note are the diastereo- and enantioselective Zr-catalyzed alkylations, a tandem Ti- and Ni-catalyzed
process that constitutes a hydrovinylation reaction, and a Ru-catalyzed alcohol oxidation to afford carboxylic acid
18. The requisite carbohydrate 38 is synthesized in a highly diastereo- and enantioselective fashion. Optical purity
of the carbohydrate moiety arises from the use of the asymmetric dihydroxylation method of Sharpless;
diastereochemical control is achieved through a selective dipolar [3 + 2] cycloaddition with a readily available
amine serving as the chiral auxiliary. Union of the appropriately outfitted carbohydrate 71 and diene 62 through an
efficient and diastereoselective glycosylation is followed by a remarkably efficient Mo-catalyzed macrocyclization
that proceeds readily at room temperature.
Background
Chart 1
The identification and isolation of effective antifungal agents
is critical to human health care; in the past decade, the number
of related infections acquired in hospitals has doubled.1 It was
therefore noteworthy when, in 1990, scientists at Schering-
Plough reported a new class of antifungals. A representative
member is Sch 38516 (1),2 a compound active against Candida
sp. (MIC 0.91 µg/mL) and dermatophytes (MIC >80.6 µg/mL),
the detailed structure of which has been established through
X-ray crystallography of the derived p-bromobenzenesulfon-
amide (Chart 1). Further interest in this natural product stems
from the likelihood that Sch 38516 is identical to fluvirucin
B1,3 an agent that is effective against influenza A virus. Other
members, exemplified by Sch 38518 (2; fluvirucin B2), have
received less attention;4 the relative and absolute stereochemical
identity of the latter is yet to be rigorously established. We
thus considered a general and modular approach to the total
synthesis of this category of antifungals a worthy objective.5
Sch 38516 is an attractive target for enantioselective synthesis
because of the paucity of stereogenic sites in the macrolactam
structure, a characteristic that renders effective stereocontrol a
X Abstract published in AdVance ACS Abstracts, October 15, 1997.
(1) Marston, W. The New York Times Magazine 1996, 11, 43-45.
(2) (a) Hegde, V. R.; Patel, M. G.; Gullo, V. P.; Ganguly, A. K.; Sarre,
O.; Puar, M. S.; McPhail, A. T. J. Am. Chem. Soc. 1990, 112, 6403-6405.
(b) Hegde, V. R.; Patel, M.; Horan, A.; Gullo, V.; Marquez, J.; Gunnarsson,
I.; Gentile, F.; Loebenberg, D.; King, A. J. Antibiot. 1992, 45, 624-632.
(3) (a) Naruse, N.; Tenmyo, O.; Kawano, K.; Tomita, K.; Ohgusa, N.;
Miyaki, T.; Konishi, M.; Oki, T. J. Antibiot. 1991, 44, 733-740. (b) Naruse,
N.; Tsuno, T.; Sawada, Y.; Konishi, M.; Oki, T. J. Antibiot. 1991, 44, 741-
755. (c) Naruse, N.; Konishi, M.; Oki, T. J. Antibiot. 1991, 44, 756-761.
(d) Komita, K.; Oda, N.; Hoshino, Y.; Ohkusa, N.; Chikazawa, H. J.
Antibiot. 1991, 44, 940-948.
formidable task (C2-C6 and C6-C9 relative stereocontrol).
The target molecule bears a novel carbohydrate appendage that
makes its debut here as part of a natural product. Furthermore,
research efforts directed toward the total synthesis of Sch 38516
would allow us to challenge and establish the utility of
stereoselective Zr-catalyzed ethylmagnesation, a process that
has been developed in these laboratories.
(5) For preliminary accounts of this research, see: (a) Houri, A. F.; Xu,
Z.; Cogan, D. A.; Hoveyda, A. H. J. Am. Chem. Soc. 1995, 117, 2943-
2944. (b) Xu, Z.; Johannes, C. W.; Salman, S. S.; Hoveyda, A. H. J. Am.
Chem. Soc. 1996, 118, 10926-10927. For an alternative approach to the
macrolactam segment, see: Trost, B. M.; Ceschi, M. A.; Ko¨nig, B. Angew.
Chem., Int. Ed. Engl. 1997, 36, 1486-1489.
(4) (a) Hegde, V.; Patel, M. G.; Gullo, V. P.; Puar, M. S. J. Chem. Soc.,
Chem. Commun. 1991, 810-812. (b) Hegde, V. R.; Patel, M. G.; Gullo,
V. P.; Horan, A. C.; King, A. H.; Gentile, F.; Wagman, G. H.; Puar, M. S.;
Loebenberg, D. J. Antibiot. 1993, 46, 1109-1115.
S0002-7863(97)02191-4 CCC: $14.00 © 1997 American Chemical Society

EnantioselectiVe Total Synthesis of FluVirucin B1
J. Am. Chem. Soc., Vol. 119, No. 43, 1997 10303
Scheme 1
Chart 2
Synthesis Plan
Many of our preliminary plans involved metal-catalyzed
transformations. It was clear to us that such strategies offered
exceedingly efficient routes to the requisite fragments, highlight-
ing the effectiveness of a range of catalytic reactions, the
majority of which are not feasible in the absence of a catalyst.
What we did not expect was that the successful total synthesis
would rely more heavily on transition metal-catalysis than
originally expected.
At the outset, our overall strategy was along the established
lines. We planned first to prepare the macrocyclic and
carbohydrate moities; we expected that subsequent union of the
two fragments would yield a protected form of the antifungal.
This blueprint would eventually prove inadequate and had to
be modified. Nevertheless, chemistry developed for the con-
struction of the two segments was central to the successful
implemenation of the total synthesis. Details of the enantio-
selective syntheses of the macrolactam and carbohydrate frag-
ments are presented below.
Scheme 2
Macrolactam Synthesis
With regard to the design of an efficient synthesis for the
macrocyclic segment, we agreed on trisubstituted alkene 4
(Scheme 1) as a desirable intermediate. The reason for this
predilection was twofold: (1) Molecular mechanics modeling
(AM1 level on the TBS-protected ring structure) indicated that
catalytic hydrogenation of the resident olefin would benefit from
selective peripheral addition6 to deliver the desired stereoisomer
(see Chart 2, below), offering a convenient solution to the
question of remote C2-C6 and C6-C9 stereocontrol. (2) A
range of analogues would become readily accessible through
stereoselective functionalization of the alkene.
Macrolactam Synthesis Plan A. Our initial approach to the
synthesis of the macrocyclic amide 4 was influenced by the
more traditional macrocyclization strategies.7 As illustrated in
Scheme 1, we planned to assemble fragments 6 and 7 through
a Wittig olefination; the 14-membered ring 4 would then emerge
from a macrocyclic amide bond formation (5 f 4). This general
scheme, nevertheless, posed an important challenge in
stereocontrol: Would stereoselective hydrogenation require the
a 5 equiv of EtMgCl, 10 mol % Cp2ZrCl2, Et2O; TsN(CH2CH2), 10
mol % CuI, THF (>20:1 diastereoselectivity, 52% overall yield). b80%
aqueous HOAc, 65 °C, 60%. cNa(CH2)SO(CH3), 12, then 11, 20-55%.
presence of only one of the two olefin isomers (E-4 or Z-4)? If
so, a difficult stereoselective synthesis of an unfunctionalized
and remote trisubstituted alkene would have to be devised.
Modeling studies, however, hinted that, as illustrated in Chart
2, either olefin stereoisomer would lead to the formation of the
appropriate C6 stereochemistry, and that both E and Z macro-
cyclic alkenes are energetically within reach (AM1 level
calculations indicate ∼1 kcal/mol energy difference).
To examine the viability of the olefination-macrolactam-
ization, ketone 11 (Scheme 2) was prepared in five steps and
15% overall yield in a stereoselective manner. As shown in
Scheme 2, treatment of allylic alcohol 8 with EtMgCl in the
presence of 10 mol % Cp2ZrCl2 (Et2O, 22 °C) resulted in
diastereoselective carbomagnesation (f 9);8 the derived chiral
alkylmagnesium chloride was treated with 10 mol % CuI and
tosyl aziridine9 to afford ketal 10. Subsequent removal of the
(6) (a) Corey, E. J.; Hopkins, P. B.; Kim, S.; Yoo, S.-E.; Nambia, K. P.;
Falck, J. R. J. Am. Chem. Soc. 1979, 101, 7131-7134. (b) Still, W. C.;
Galynker, I. Tetrahedron 1981, 37, 3981-3996. (c) Still, W. C.; Novack,
V. J. J. Am. Chem. Soc. 1984, 106, 1148-1149. (d) Schreiber, S. L.; Santini,
C. J. Am. Chem. Soc. 1984, 106, 4038-4039. (e) Neeland, E. G.; Ounsworth,
J. P.; Sims, R. J.; Weiler, L. J. Org. Chem. 1994, 58, 7383-7394. (f) Evans,
D. A.; Ratz, A. M.; Huff, B. E.; Sheppard, S. G. J. Am. Chem. Soc. 1995,
117, 3448-3467.
(7) For example, see: (a) Corey, E. J.; Weigel, L. O.; Chamberlin, A.
R.; Cho, H.; Hua, D. H. J. Am. Chem. Soc. 1980, 102, 6613-6615. (b)
Meyers, A. I.; Reider, P. J.; Campbell, A. L. J. Am. Chem. Soc. 1980, 102,
6597-6598.
(8) (a) Hoveyda, A. H.; Xu, Z. J. Am. Chem. Soc. 1991, 113, 5079-
5080. (b) Hoveyda, A. H.; Xu, Z.; Morken, J. P.; Houri, A. F. J. Am. Chem.
Soc. 1991, 113, 8950-8952. (c) Houri, A. F.; Didiuk, M. T.; Xu, Z.; Horan,
N. R.; Hoveyda, A. H. J. Am. Chem. Soc. 1993, 115, 6614-6624. (d)
Morken, J. P.; Hoveyda, A. H. J. Org. Chem. 1993, 4237-4244.
(9) (a) Baldwin, J. E.; Adington, R. M.; Neil, I. A.; Schofield, C.; Spivey,
A. C.; Sweeney, J. B. J. Chem. Soc., Chem. Commun. 1989, 1852-1854.
(b) Church, N. J.; Young, D. W. Tetrahedron Lett. 1995, 36, 151-154.

10304 J. Am. Chem. Soc., Vol. 119, No. 43, 1997
Xu et al.
acetal unit released ketone 11 (containing the derived internal
hemiketal, as judged by H NMR analysis). To expedite our
Scheme 3
1
olefination studies, we selected achiral phosphonium salt 12 as
the model ylide substrate. Attempts to couple 11 and 12 under
various conditions based on procedures reported by Corey10
(Na(CH2)SO(CH3), DMSO) for similar systems afforded 13 in
low yields (typically ∼20-55%) and, not surprisingly, as a
mixture of alkene isomers. Moreover, these reactions proved
particularly irregularswe were unable to identify which factors
would reliably deliver >55% yields.
We were hesitant to use elevated temperatures to facilitate
olefination because of our concern that, particularly with the
highly polar medium (DMSO), the integrity of the stereogenic
center in the requisite phosphonium salt 7 would be compro-
mised. Furthermore, we decided that inferior yields, at a stage
where two optically pure segments are joined, would be
detrimental to the efficiency of the overall plan.
forming highly substituted alkene. To challenge the veracity
of these concerns and to examine the validity of the ring
formation strategy, we devised plans for an efficient and
enantioselective synthesis of 16, which would in turn demand
a facile route to amine 17 and carboxylic acid 18.
Macrolactam Synthesis Plan B. In seeking an alternative
and more efficient scenario to the formation of macrolactam 4,
we were intrigued by a succession of pioneering reports by
Grubbs11 and Schrock12 on metal-catalyzed ring closing me-
tathesis and thus considered the possibility of using this protocol.
At the time (1994), synthesis of an eight- and a thirteen-
membered heterocycle through the use of such catalytic
procedures was reported by Martin13 and Pandit,14 respectively.
In the former case, a disubstituted alkene was prepared with
Schrock’s Mo(CHCMe2Ph)(N(2,6-(i-Pr)2C6H3))OCMe(CF3)2)2
(14) as the catalyst, whereas in the latter instance, Grubbs’s
alkylidene complex (PCy3)2Cl2RuCHCHPh2 (15) was used. In
both cases, however, the diene precursor contained a rigid cyclic
framework. Another disclosure by Martin15 illustrated that
efficient synthesis of 12-membered rings from conformationally
mobile aliphatic starting materials could be problematic. Any
doubts about the significance of the flexibility of the precursor
structure on the ease of ring closing metathesis of medium and
large rings was quelled by Grubbs’s seminal report on the
influence of structural scaffolds on the facility of eight-
membered ring construction promoted by 15.16
Enantioselective Synthesis of Amine 17. As depicted in
Scheme 4, the asymmetric synthesis of unsaturated amine 17
began with conversion of homoallylic alcohol 19 to racemic
allylic alcohol 20 (51% overall yield), which was subsequently
resolved according to the procedure by Sharpless (>99% ee,
66% yield, based on 50% conversion).19 Treatment of optically
pure 20 with 5 equiv of EtMgCl, 5 mol % Cp2ZrCl2 (THF, 22
°C),8 and quenching of the resulting carbomagnesation product
21 with tosyl aziridine in the presence of 5 mol % CuI (or
CuBr‚Me2S) afforded 22 in 42% isolated yield, with >99% site
selectivity and 97:3 diastereochemical control. The one-pot
catalytic and stereoselective double-alkylation process was
followed by protection of the secondary carbinol as its derived
TBS ether (87% yield) and removal of the tosyl protecting group
to afford 17 (90% yield). The six-step procedure was carried
out with excellent diastereo- and enantioselectivity in 20-22%
overall yield.
Scheme 4
Because of the above myriad observations, the prospects for
an efficient and reliable catalytic ring closure of 16 to afford 4
did not appear promising. We were aware that the presence of
stereogenic centers have been reported to infer appreciable
degrees of structural rigidity on a number of occasions.17,18
However, we were also mindful of the fact that our synthesis
plan required facile formation of a trisubstituted macrocyclic
alkene. We were concerned that substrate oligomerization,
catalyst decomposition, and conformational freedom would not
be readily overcome when the desired product is a slower-
a Ph3P, I2, Et3N, CH2Cl2, 0 °C f 22 °C. bn-BuLi, H2CCHCHO, THF,
-78 °C, 51% overall. cTi(OiPr)4, tert-BuOOH, (+)-dicyclohexyl
d
tartrate, 4 Å molecular sieves, 66%. 5 equiv of EtMgCl, 5 mol %
(10) Corey, E. J.; Schaaf, T. K.; Huber, W.; Koelliker, U.; Weinshenker,
N. M. J. Am. Chem. Soc. 1970, 92, 397-380.
Cp2ZrCl2, Et2O, 22 °C, 12 h; 6 equiv of TsN(CH2CH2), 5 mol % CuI,
e
f
(11) (a) Fu, G. C.; Grubbs, R. H. J. Am. Chem. Soc. 1992, 114, 7324-
7325. (b) Fu, G. C.; Grubbs, R. H. J. Am. Chem. Soc. 1993, 115, 3800-
3801. (c) Grubbs, R. H.; Miller, S. J.; Fu, G. C. Acc. Chem. Res. 1995, 28,
446-452 and references cited therein. (d) Schmalz, H.-G. Angew. Chem.,
Int. Ed. Engl. 1995, 107, 1981-1984 and references cited therein.
(12) (a) Schrock, R. R.; Murdzek, J. S.; Bazan, G. C.; Robbins, J.;
DiMare, M.; O’Regan, M. J. Am. Chem. Soc. 1990, 112, 3875-3886. (b)
Bazan, G. C.; Schrock, R. R.; Cho, H.-N.; Gibson, V. C. Macromolecules
1991, 24, 4495-4502.
THF, 42%. TBSOTf, Et3N, CH2Cl2, -78 °C, 87%. 7 equiv of Na,
NH3, -50 °C, 95%.
Enantioselective Synthesis of Carboxylic Acid 18. To
obtain the carboxylic acid piece (18) in the optically pure form,
we first considered catalytic enantioselective ethylmagnesation
of unsaturated oxocene 23 (option 1, Scheme 5). Subsequent
(13) Martin, S. F.; Liao, Y.; Wong, Y.; Rein, T. Tetrahedron Lett. 1994,
35, 691-694. (c) Martin, S. F.; Liao, Y.; Chen, H.-J.; Patzel, M.; Ramser,
M. N. Tetrahedron Lett. 1994, 35, 6005-6008.
(18) Recent elegant studies by Grubbs on the effect of stereochemistry
on the synthesis of macrocyclic peptides (disubstituted alkenes) by ring
closing metathesis were not disclosed at the time of our planning. See: (a)
Miller, S. J.; Gubbs, R. H. J. Am. Chem. Soc. 1995, 117, 5855-5856. (b)
Miller, S. J.; Blackwell, H.; Grubbs, R. H. J. Am. Chem. Soc. 1996, 118,
9606-9614.
(19) (a) Gao, Y.; Hanson, R. M.; Klunder, J. M.; Ko, S. Y.; Masamune,
H.; Sharpless, K. B. J. Am. Chem. Soc. 1987, 109, 5765-5780. For a
comprehensive review, see: (b) Johnson, R. A.; Sharpless, K. B. In Catalytic
Asymmetric Synthesis; Ojima, I., Ed.; VCH: New York, 1993; pp 103-
158.
(14) Borer, B. C.; Deerenberg, S.; Bieraugel, H.; Pandit, U. K.
Tetrahedron Lett. 1994, 35, 3191-3194.
(15) Martin, S. F.; Liao, Y.; Chen, H.-J.; Patzel, M.; Ramser, M. N.
Tetrahedron Lett. 1994, 35, 6005-6008.
(16) Miller, S. J.; Kim, S.-H.; Chen, Z.-R.; Grubbs, R. H. J. Am. Chem.
Soc. 1995, 117, 2108-2109.
(17) For a classic example, see: Woodward, R. B. et al. J. Am. Chem.
Soc. 1981, 3213-3215.

EnantioselectiVe Total Synthesis of FluVirucin B1
J. Am. Chem. Soc., Vol. 119, No. 43, 1997 10305
Scheme 5
Scheme 7
Scheme 6
b
a 1 equiv of DCC, 1.2 equiv of HOBT, 22 °C, 12 h, 85%. 20 mol
% Mo(CHCMe2Ph)(N(2,6-(i-Pr)2C6H3))OCMe(CF3)2)2 , C6H6, 22 °C,
c
d
90%. H2 (1 atm), 10% Pd(C), 84%. HF, MeCN, 80%.
When unsaturated alcohol 28 was treated with 10 mol % (n-
Pr)4NRuO4 in the presence of 2 equiv of H2O and 3 equiv of
NMO, 18 was obtained in 77% yield (>99% ee; chiral GLC
analysis). Further studies with regard to the mechanism and
utility of this mild catalytic oxidation are in progress. It is
plausible that the purported aldehyde intermediate undergoes
hydration in the presence of H2O; further oxidation affords the
carboxylic acid product.
a 0.8 equiv of EtMgBr, 0.4 mol % (S)-(EBTHI)-Zr-binol, THF, 70
b
turnovers. 3 equiv of n-PrMgBr, 3 mol % Cp2TiCl2, THF; then 15
equiv of H2CCHBr, 3 mol % (Ph3P)2NiCl2, 72%. c10 mol % (n-
Macrocycle Synthesis through Catalytic Ring-Closing
Metathesis. With amine 17 and carboxylic acid 18 available,
these two fragments were coupled to afford diene-amide 29 in
85% yield after silica gel chromatography (Scheme 7).25
Initially, based on previous studies on similar medium- and
large-ring syntheses, we surmized that elevated temperatures
would be necessary for effective macrocyclization. We there-
fore excluded ambient temperature conditions from our pre-
liminary investigations. Accordingly, we established that, when
29 is treated with 20 mol % Mo catalyst 14 in benzene (0.01
M) and the mixture is heated to 50 °C (12 h), macrolactam 4 is
obtained in 60-65% isolated yield as a single olefin stereo-
CH3(CH2)2)4NRuO4, MeCN, 2 equiv of H2O, 3 equiv of NMO, 77%.
functionalization of the resulting alkylation product 24 would
afford 18. In practice, however, this scheme was rendered
unattractive by the inefficiency of catalytic ring closing me-
tathesis to afford 23; significant amounts of oligomeric products
were obtained instead. This outcome supported previous notions
that rigidity of the structural framework is an underpinning
aspect of successful ring synthesis by catalytic metathesis.16,20
Alternatively, we considered Zr-catalyzed enantioselective
ethylmagnesation of dihydrofuran 25,21 a commercially available
heterocycle that we had already established to be an outstanding
substrate for asymmetric catalytic alkylation. If this approach
were to be used successfully, an efficient hydrovinylation
process would be needed (Scheme 5). A synthesis scheme was
accordingly devised that made optically pure 18 available in
four catalytic steps (Scheme 6).
The sequence began with asymmetric catalytic ethylmag-
nesation of 25 (0.4 mol % (S)-(EBTHI)Zr-binol, ∼70 turnovers)
to give 26 in >99% ee (chiral GLC analysis).21 A one-pot
catalytic hydrovinylation of the terminal alkene was effected
in the following manner: Subjection of 26 to 3 equiv of
n-PrMgBr and 3 mol % Cp2TiCl2 led to hydromagnesation of
the olefin,22 affording 27. Treatment of the resulting Grignard
reagent with 3 mol % (Ph3P)2NiCl2 and 15 equiv of vinyl
bromide23 yielded the corresponding cross coupling product 28
(72% from 26). At this juncture, a Ru-catalyzed oxidation of
the primary alcohol to the derived acid was designed through a
simple modification of the well-known procedure by Ley.24
1
isomer (>98% Z, as judged by 500 MHz H NMR analysis).
As illustrated, the stereochemical identity of the olefin isomer
was established through NOE difference experiments.
We also determined that with lower catalyst loadings (15 mol
% 14), reaction efficiency suffers significantly (<50% conver-
sion). When Ru catalyst 15 was used (20 mol %, benzene, 80
°C), 5-10% dimer derived from reaction of the terminal olefins
was formed as the only product. Further investigation of the
catalytic macrocyclization indicated that, with freshly prepared
or recrystallized Mo catalyst, ring closing metathesis occurs
smoothly at 22 °C to afford 4 in 90% isolated yield after only
4 h. With 40 mol % 14, reaction yield improved to 97%. Less
than 20 mol % catalyst gave notably lower conversions and
yields.
(20) For an unusually facile eight-membered ring synthesis through
catalytic metathesis, see: (a) Visser, M. S.; Heron, N. M.; Didiuk, M. T.;
Sagal, J. F.; Hoveyda, A. H. J. Am. Chem. Soc. 1996, 118, 4291-4298.
For a more recent related study, see: (b) Linderman, R. J.; Siedlecki, J.;
O’Neill, S. A.; Sun, H. J. Am. Chem. Soc. 1997, 119, 6919-6920.
(21) (a) Morken, J. P.; Didiuk, M. T.; Hoveyda, A. H. J. Am. Chem.
Soc. 1993, 115, 6697-6698. (b) Reference 20. For a recent review, see:
Hoveyda, A. H.; Morken, J. P. Angew. Chem., Int. Ed. Engl. 1996, 35,
1262-1284.
(22) (a) Eisch, J. J.; Galle, J. E. J. Organomet. Chem. 1978, 160, C8-
C12. (b) Sato, F. J. Organomet. Chem. 1985, 185, 53-64.
(23) Tamao, K.; Sumitani, K.; Kumada, M. J. Am. Chem. Soc. 1972,
94, 4374-4376.
(24) Ley, S. V.; Norman, J.; Griffith, W. P.; Marsden, S. P. Synthesis
1994, 639-666 and references cited therein.
With Mo-catalyzed macrolactamization secured, we turned
our attention to the critical issue of C2-C6 remote stere-
ochemical control. Catalytic hydrogenation of unsaturated
cyclic amide 4, as predicated by initial modeling studies (Chart
(25) Li, W. R.; Ewing, W. R.; Harris, B. D.; Joullie, M. J. Am. Chem.
Soc. 1990, 112, 7659-7672.

10306 J. Am. Chem. Soc., Vol. 119, No. 43, 1997
Xu et al.
1
Scheme 8
MHz H NMR analysis. When 4 was subjected to the same
reaction conditions, <2% of any of the acyclic products was
detected. Although, we do not as yet have a positive proof
that 37 is formed in cyclization of 29 (f 4), this observation
suggests that if dimerization were to occur, the material can be
readily converted to the desired macrolactam, which is kineti-
cally immune to rupture.
a Conditions: 25 mol % 14, 0.01 M in C6H6, 50 °C, 18 h.
2), in the presence of 10% Pd(C) resulted in the formation of 3
in 84% yield and with >98% diastereoselectivity (400 MHz
1H NMR analysis). To ascertain that our stereochemical
assignment was valid, the TBS protecting group was removed
and the resulting aglycon 30 acylated under standard reaction
conditions (Ac2O, Et3N, 10 mol % DMAP). This procedure
delivered material identical to that obtained from degradation
of the natural product (1H NMR, IR, TLC).
The Origin of the Facile Catalytic Macrocyclization. To
gain insight into the remarkably efficient cyclization reaction
that converts 29 to 4, and to determine the extent to which
existing stereogenic centers provide diene precursor 29 with
substrate preorganization, we examined ring closing metathesis
reactions of diene-amides 31 and 34. As illustrated in Scheme
8, when 31 was subjected to 25 mol % 14 (50 °C, 18 h), <2%
32 was formed. Instead, dimer 33 was obtained in 52% yield
after chromatography (3:1 mixture of olefin isomers, judged
by 400 MHz 1H NMR; identity of major product not determined;
HRMS: Calcd, 474.4185. Found, 474.4187). When diene 34
was subjected to identical conditions, macrolactam 35 was
obtained in 41% isolated yield (HRMS: Calcd, 209.1780.
Found, 209.1777), along with 20% of the derived 28-membered
ring cyclic dimer 36 (presumably as a mixture of head-to-head
and head-to-tail isomers; HRMS: Calcd, 418.3559. Found,
418.3560); <10% of the acyclic dimer was detected. These
data clearly illustrate that with the slower forming trisubstituted
olefin, the presence of stereogenic centers is required for facile
ring closure (cf. 29 f 4, Scheme 7), whereas when reacting
alkenes are terminal olefins, formation of the fourteen-membered
ring is more favored (41% yield of 35). Our results lend further
credence to Grubbs’s contention16 that with conformationally
mobile diene structures some form of structural restraint is
needed for ring closing metathesis to occur readily.26
Carbohydrate Synthesis28
At the planning stage for the carbohydrate synthesis, we
realized that this moiety of the natural product, as shown in
Scheme 9, may arise from appropriate functionalization of
tricycle 40, fabricated through a [3 + 2] dipolar cycloaddition
involving nitrone 41 and vinylene carbonate. In this context,
we were intrigued by the possibility of using the precedence
established by DeShong.29 Below, we describe how the general
synthesis plan depicted in Scheme 9 was modified to provide
an efficient, diastereo- and enantioselective route to 38.
Enantioselective Synthesis of Nitrone 41. Preparation of
nitrone 41 requires aldehyde 45 (Scheme 10), which has been
used in the synthesis of daunosamine, the sugar region of the
notable antitumor agent daunomycin.30 Two routes for the
Scheme 9
Considering the facility with which dimerization products 33
and 36 are obtained, we reasoned that, in catalytic ring closure
of 29, the derived dimer is perhaps initially formed as well. If
the metathesis process is reversible,27 such adducts may
subsequently be converted to the desired macrocycle 4. To
examine the validity of this paradigm, diene 29 was dimerized
(f 37) by treatment with Ru catalyst 15. As eq 1 depicts, when
37 was treated with 22 mol % 14 (after pretreatment with
ethylene to ensure formation of the active complex), 50-55%
conversion to macrolactam 4 was detected within 7 h by 400
(28) A synthesis of the carbohydrate, 3-amino-3,6-deoxy-D-talopyranose,
was disclosed by Stanek and Jary in 1974. Their reported procedure affords
the amino sugar only as a byproduct, formed en route to a mixture of various
3-amino-3,6-pyranoses (2.4% of total mixture). See: Capek, K; Stanek, J.;
Jary, J. Coll. Czech. Chem. Commun. 1974, 39, 1462-1478.
(29) DeShong, P.; Dicken, C. M.; Leginus, J. M.; Whittle, R. R. J. Am.
Chem. Soc. 1984, 106, 5598-5602 and references cited therein.
(30) (a) Pedersen, S. F.; Konradi, A. W. J. Org. Chem. 1990, 55, 4506-
4508. (b) Hauser, F.; Ellenberger, S. R. J. Org. Chem. 1986, 51, 50-57.
(c) Fuganti, C.; Grasselli, P.; Pedrocchi-Fantoni, G. J. Org. Chem. 1983,
48, 909-910.
(26) It is likely that, as recently demonstrated by Furstner, higher yields
of the disubstituted olefin 35 may be obtained under strict high dilution
conditions and with longer reaction times. Our experiments clearly illustrate
that the synthesis of trisubstituted macrocyclic alkenes is significantly more
demanding than that of their disubstituted analogues. See: Furstner, A.;
Langemann, K. J. Org. Chem. 1996, 61, 3942-3943.
(27) Marsella, M. J.; Maynard, H. D.; Grubbs, R. H. Angew. Chem., Int.
Ed. Engl. 1997, 36, 1101-1103 and references cited therein.

EnantioselectiVe Total Synthesis of FluVirucin B1
J. Am. Chem. Soc., Vol. 119, No. 43, 1997 10307
Scheme 10
required for this purpose. Hence, it was clear that the cyclo-
addition reaction would have to be rendered highly stereo-
selective for effective material throughput necessary in the
course of the total synthesis.
a AD-mix-R, 1 equiv of MeSO2NH2, tert-BuOH, H2O. b 2,2-
dimethoxypropane, 5 mol % p-TsOH, 52% from 42. c 2.5 equiv of
DIBAL-H, -78 °C, d O3, 8:1 CH2Cl2:MeOH, -78 °C; Me2S, 64%
from 43. e 1.0 equiv of N-hydroxybenzylamine, 20 h, 99%.
To design a more selective variant of the cycloaddition
process, we were encouraged by an investigation reported by
Wovkulich and Uskokovic.36 These researchers were able to
enhance stereoselectivity in a similar, albeit intramolecular, [3
+ 2] cycloaddition through the use of chiral amine ((S)-R-
methylbenzylamine37 (in place of a benzylamine). We therefore
prepared nitrones 47 and 49 by treatment of aldehyde 45 with
the appropriate enantiomeric antipodes of the nonracemic
hydroxylamines. It was at this point that the minor isomer
formed in the asymmetric dihydroxylation of 42 was separated
from the desired product. As illustrated in eq 3, (R)-N-hydroxyl-
R-methylbenzylamine 47 undergoes cycloaddition with Vinylene
carbonate to afford 48 with >20:1 diastereoselectiVity, >25:1
construction of 45 have been reported. One is by Mukaiyama,31
a procedure that utilizes diethyl tartrate as starting material and
is effected in 53% overall yield in seven steps. The other is
that of Schmid,32 where the target carboxaldehyde is obtained
in five steps and 16% overall yield, starting with D-arabinose.
We decided to outline a more concise preparation scheme, where
a catalytic enantioselective process is used to introduce absolute
chirality. We avoided starting with carbohydrates or other
naturally available chiral sources, circumventing substrate
functional group alterations and, ultimately, unnecessarily
lengthy routes. A four-step protocol (Scheme 10) that delivers
45 in 33% overall yield and 80% ee is described below.
Asymmetric dihydroxylation33 of commercially available
ethyl sorbate (Scheme 10), according to the procedure by
Sharpless, gave 43 in 61% yield and 80% ee (as judged by chiral
GLC). Protection of the vicinal diol and subsequent reduction
of the ester group led to the formation of allylic alcohol 44 in
80% yield after chromatography. Ozonolysis of 44 under
standard conditions provided carboxaldehyde 45 (67%);34 treat-
ment of 45 with N-benzylhydroxylamine gave nitrone 41 in 99%
yield.
1
endo selectiVity (400 MHz H NMR) and in 69% yield after
silica gel chromatography.38 In contrast, when the S chiral
auxiliary was used, stereoselectivity reduced to 1.3:1, endo:exo
ratio diminished to 10:1 (from 15:1 with benzylamine), and the
combined yield of products was 40% (eq 4). It must be pointed
out that thermal cycloaddition must be carried out under strictly
controlled reaction conditions; temperatures above 85 °C caused
significant material decomposition, whereas with temperatures
below 85 °C, little or no reaction occurred.
Several issues with regard to the synthesis route in Scheme
10 merit comment: (1) Enantioselective dihydroxylation of 42,
depending on which stereoisomeric chiral oxidant is used,
affords different degrees of absolute facial selectivities (80%
ee with AD-mix-R vs 96% ee with AD-mix-â). (2) The minor
isomer obtained in the latter catalytic reaction can be readily
separated from the desired major enantiomer at a later stage in
the synthesis (see below).
The Diastereoselective [3 + 2] Cycloaddition. As il-
lustrated in eq 2, when nitrone 41 was treated with 5 equiv of
vinylene carbonate at 85 °C (22 h), cycloadducts 40 and 46
were obtained with similar selectivity as when the original
DeShong protocol35 was employed (exo products are not
shown). The stereochemistry of the major diastereomer is that
desired for the stereoselective synthesis of carbohydrate 38 (see
below for proof of stereochemistry). However, inferior levels
of diastereocontrol (38% isolated yield of 40) made this
approach less appealing, especially since chromatographic
separation of isomeric cycloadducts (40 and 46) proved cumber-
some; medium pressure liquid chromatography (MPLC) was
The Origin of Stereocontrol in the [3 + 2] Cycloaddition.
The following observations are critical for a better understanding
of the mechanism of the cycloaddition reaction: (1) Nuclear
(31) (a) Mukaiyama, T.; Goto, Y.; Shoda, S. Chem. Lett. 1983, 671-
674. (b) Fuganti, C.; Grasselli, P.; Pedrocchi-Fantoni, G. J. Org. Chem.
1983, 48, 910-912.
(36) (a) Wovkulich, P. M.; Uskokovic, M. R. J. Am. Chem. Soc. 1989,
111, 3956-3957 and references cited therein. For other related examples,
see: (b) Belzecki, C.; Panfil, C. J. Org. Chem. 1979, 44, 1212-1216. (c)
Vasella, A.; Voeffray, R. HelV. Chim. Acta. 1982, 65, 1134-1137 and
references cited therein. (d) Kita, Y.; Itoh, F.; Tamura, O.; Ke, Y. Y.
Tetrahedron Lett. 1987, 28, 1431-1434.
(37) For synthesis of the chiral hydroxylamine, see: Polonski, T.;
Chimiak, A. Tetrahedron Lett. 1974, 2453-2456.
(38) For the use of this chiral auxiliary in stereoselective reactions with
O-silylketene acetals, see ref 36d.
(32) Binder, W. H.; Prenner, R. H.; Schmid, W. Monat. Chem. 1994,
125, 763-771.
(33) (a) Xu, D.; Crispino, G.; Sharpless, K. B. J. Am. Chem. Soc. 1992,
114, 7570-7571. (b) Kolb, H. C.; VanNieuwehze, M. S.; Sharpless, K. B.
Chem. ReV. 1994, 94, 2483-2547.
(34) Direct ozonolytic cleavage of protected 43 affords 45 in 40% isolated
yield (versus 64% for the two-step sequence shown in Scheme 10).
(35) Conditions originally reported by DeShong are 30 equiv of vinylene
carbonate, 95 °C, 72 h (cf. ref 29).

10308 J. Am. Chem. Soc., Vol. 119, No. 43, 1997
Xu et al.
Scheme 11
Scheme 12
Overhauser difference spectroscopy indicated that both nitrones,
as expected,39 preferentially exist as their Z isomers (Scheme
11). (2) Crossover experiment shown in Scheme 11 demon-
strated that cycloaddition is under kinetic control. (3) Reactions
proceed in an identical fashion in the presence of
2,6-di-tert-butyl-4-methylphenol: radical intermediates are prob-
ably not involved.40
A plausible transition structure (I) for the cycloaddition of
nitrone 47 is shown in Scheme 12. This proposal is based on
the following principles: (1) Supported by previous studies, we
assume that the cycloaddition is concerted but asynchronus.41
Additionally, theoretical investigations suggest that formation
of the C-C bond is more advanced than that of the C-O bond
in the transition state.42 These mechanistic considerations
suggest that factors differentiating the two diastereotopic faces
of the CdN unit (rather than the entire CdNsO moiety) are
critical to the eventual stereoselective outcome. (2) The more
reactive nitrone conformer is one where the electron-withdraw-
ing C-O bond is antiperiplanar to the CdN π cloud; electron
donation from CdN π to CsO σ* promotes the electron
deficiency and thus reactivity of the nitrone. (3) Electronic
effects alone are not sufficient to give rise to significant levels
of diastereodifferentiation; this is particularly likely since all
reactions are carried out at elevated temperatures (both electronic
and steric factors need to favor one transition structure for
appreciable ∆∆Gq). Electronic factors, only in conjunction with
supporting steric effects, lead to high levels of stereocontrol.
On the basis of the above principles, the favored transition
structure I (Scheme 12) thus involves the addition of vinylene
carbonate according to the Cram-Felkin-Anh model43 (attack
of dipolarophile through the Burgi-Dunitz angle). In the
alternative mode of addition II, the incoming heterocycle adds
syn to the larger alkyl substituent (anti-Cram) and the phenyl
unit of the chiral auxiliary suffers from unfavorable steric
interactions with the nitrone oxygen.
Scheme 13
electronically preferable orientation of the C-O bond with
respect to the CdN unit (see above). Similarly, when (S)-
nitrone 49 is utilized, energetically similar modes of addition
V and VI, consistent with the Felkin-Anh model, lead to
opposite levels of diastereoselection (Scheme 14). The present
paradigm offers a mechanistically consistent picture for the
observed degrees of diastereoselectivity; the dependence of
endo:exo ratios on the nature of the amine group is, on the other
hand, more difficult to decipher.
Reduction of the N-O Bond and Proof of Stereochem-
istry. Although reduction of similar cycloadducts were already
reported in the literature,29 we found that treatment of these
polycyclic systems with Pd(OH)2 or Pd(C) through a range of
neutral and acidic conditions resulted in either recovery of the
starting material or complete decomposition.44 After extensive
experimentation, the following were determined: (1) The
reaction must be carried out with rigorous control of HCl
stoichiometry. Control experiments indicated that too little acid
(<2.5 equiv) leads to substrate recovery, and excess HCl (>2.5
equiv) promotes starting material decay. (2) Although the acidic
With benzyl nitrone 41 (without the chiral auxiliary), reaction
through conformations III and IV (Cram) compete to reduce
diastereoselection (Scheme 13). Modest selectivity arises from
(39) (a) Bjorgo, J.; Boyd, D. R.; Neill, D. C. J. Chem. Soc, Chem.
Commun. 1974, 478-479. Energy differences between E and Z nitrones
have been measured: (b) Breuer, E.; Aurich, H.; Nielsen, A. In Nitrones,
Nitronates and Nitroxides; John Wiley & Sons: New York, 1989; p 151.
(40) Firestone, R. A. J. Org. Chem. 1968, 33, 2285-2290 and references
cited therein.
(41) (a) Huisgen, R. Angew. Chem., Int. Ed. Engl. 1963, 2, 633-645.
(b) Professor K. N. Houk, personal communication.
(42) Houk, K. N. Acc. Chem. Res. 1975, 8, 361-369.
(43) (a) Anh, N. T.; Eisenstein, O. NouV. J. Chim. 1977, 1, 61-70. (b)
Houk, K. N.; Paddon-Row, M. N.; Rondan, N. G.; Wu, Y.-D.; Brown, F.
K.; Spellmeyer, D. C.; Metz, J. T.; Li, Y.; Loncharich, R. J. Science 1986,
231, 1108-1117 and references cited therein.
(44) Representative conditions examined (all catalysts used in 50% by
weight): 50 psi H2, Pd(OH)2, 10% HCl, MeOH and 50 psi H2, Pd(OH)2,
MeOH (24 h) resulted in substrate degradation; 50 psi H2, Pd(C), MeOH
and 50 psi H2, Pd(C), HOAc (14 h) resulted in starting material recovery;
200 psi H2, Raney Ni, MeOH (14 h) afforded a complex mixture of
unidentifiable products.

EnantioselectiVe Total Synthesis of FluVirucin B1
J. Am. Chem. Soc., Vol. 119, No. 43, 1997 10309
Scheme 14
Scheme 15
Scheme 16
a 10% HCl in MeOH, 70 h, 88%. b 10 equiv of Ac2O, 5 mol %
4-(dimethylamino)pyridine, pyridine, 98%.
of the natural product indicated that these compounds are
identical, verifying the stereochemical assignment of the [3 +
2] cycloaddition product 48.
a 4:1 THF:1.0 M HCl, 65 °C. b Pd(OH)2 (50% by weight), 300 psi
H2, 2.5 equiv of MeCOCl, MeOH.
Glycosylation Studies
Reactions with Acetoxy Glycosides. With a diastereo- and
enantioselective carbohydrate synthesis in hand, we focused our
efforts on diastereoselective glycosylation of the fourteen-
membered macrolactam. First, we decided to use the stereo-
selective protocol put forward by Mangusson,47 where BF3‚Et2O
is used as the Lewis acid promoter and the carbohydrate unit is
introduced as an acetoxy glycoside. Control of stereochemistry
would arise from participation of the neigboring axial acetoxy
group, properly situated to ensure axial delivery of the macro-
cyclic carbinol (Scheme 17).
conditions required for hydrogenation are sufficient to remove
an acetonide moiety, higher yields and more reliable results were
obtained when excision of the protective group was carried out
prior to reduction of the N-O bond. It is possible that the N-O
bond rupture occurs before release of the resident diol; in such
a case, the sensitive R-hydroxy, â-amino aldehyde may decom-
pose before the carbinol site required for ring closure is released.
(3) Pd catalysts from commercial sources were to be avoided;
use of freshly prepared Pearlman’s catalyst proved imperative.45
Thus, when 51 was reduced under 300 psi H2 with freshly
prepared Pearlman’s catalyst (Pd(OH)2, 50% by weight) in the
presence of 2.5 equiv of HCl, carbohydrate 38‚HCl was obtained
in 92% yield as a 7:1 mixture of R and â anomers, respectively
(Scheme 15). As further depicted in Scheme 15, when these
conditions were applied to 52, the desired sugar 38 was isolated
in 95% yield.
Scheme 17
To prove the stereochemical identity of 38 through compari-
son with the corresponding authentic sample, methoxy triacetate
derivative 53 was prepared (>98% R anomer),46 as illustrated
in Scheme 16. Analysis of 1H NMR, 13C NMR, and IR spectra
of 53 and comparison to the material obtained from degradation
To synthesize the triacetate carbohydrate derivative, we
directly subjected 38 to various standard acylating conditions;
this led to large amounts of decomposed materials. We argued
that the protected aldehyde is unmasked under the reaction
conditions and rapidly decomposes as a result of its sensitive
nature (particularly if the R-OH and â-NH units are acylated;
acylations are successful in Scheme 16, since hemiacetal OH
is protected). Accordingly, we decided to protect the unstable
(45) Pearlman, M. W. Tetrahedron Lett. 1967, 17, 1663-1664.
(46) Several additional issues in connection to the transformation depicted
in Scheme 16 merit comment: (1) In the formation of the intermediate
methoxyacetal, the initial product consists of a 3:1 mixture of R:â isomers
(after 8 h); however, after prolonged reaction times (70 h), the R anomer
becomes the sole product. (2) Whereas formation of the above mentioned
R-methylacetal requires 70 h, that of daunosamine (lacking the C2 axial
OH) is complete within 30 min. This rate difference may be due to
destabilization of the intermediate oxonium ion by the axially disposed
hydroxide.
(47) (a) Dahmen, J.; Frejd, T.; Mangusson, G.; Noori, G. Carbohydr.
Res. 1983, 114, 328-330. (b) Gurjar, M. K.; Viswanadham, G. Tetrahedron
Lett. 1991, 32, 6191-6194.

10310 J. Am. Chem. Soc., Vol. 119, No. 43, 1997
Xu et al.
Scheme 18
To circumvent the aforementioned problems, we investigated
reaction conditions originally put forth by Mukaiyama,50 where
the derived TMS ether and the milder Lewis acidic SnCl4-
AgClO4 mixture are involved, and the carbinol reaction partner
participates as its TMS ether derivative. A range of unsuccessful
attempts rapidly followed. For example, when silyl ethers 58
and 59 were subjected to 55 in the presence of 5 mol % SnCl4
and AgClO4 in Et2O at 0 °C for 10 min, desilylated substrates
were obtained as the only products. Additional modification
of the glycosylation protocol was no doubt needed.
a 10% anhydrous HCl in MeOH, 88%. b 1.5 equiv of CbzCl, 1.5
equiv of NaHCO3, MeOH, 78%. c 10 equiv of Ac2O, 4-(dimethyl-
amino)pyridine, pyridine, 82%. d 1% H2SO4 in Ac2O, 0 °C, 1 h, 85%.
Scheme 19
Reactions with Fluoroglycosides. The chemistry of the
unusually labile TMS ethers 58 and 59 suggested to us that a
glycosylation process involving the parent alcohol would be
more desirable. However, attempts to couple 22 and 55 (cf.
Scheme 19) in the presence of 5 mol % SnCl4 and AgClO4
resulted in significant carbohydrate decomposition. We there-
fore reasoned that the corresponding fluoroglycoside may
provide an attractive option in stereoselective glycosylation,
since such glycosyl derivatives are notably more stable (they
survive silica gel chromatography and high temperatures) and
have been successfully used by Mukaiyama51 and Nicolaou52
in complex molecule syntheses.
Fluoroglycoside 61 was therefore prepared, as outlined in
Scheme 20. Conversion of triacetate 55 to thioglycoside 60
(82%),52b followed by treatment of the latter with (dimethyl-
amido)sulfur trifluoride and N-bromosuccinamide delivered 61
in 85% yield after silica gel chromatography.52a
Our initial efforts utilizing 61 involved attempts at stereo-
selective glycosylation with macrolactam alcohol 30 in the
presence of 5 mol % SnCl2 and AgClO4. Repeated initiatives
along these lines resulted in complete decomposition of the
fluorinated carbohydrate (it is presumably being converted to
the derived oxonium ion) and low recovery of the macrocyclic
substrate (<15%). The principal reason for such unsatisfactory
outcomes is likely the notorious lack of solubility of 30 in Et2O,
as well as the more polar CH2Cl2sthe fourteen-membered
alcohol is sparingly soluble in methanol at ambient temperatures.
The above observations implied that a more viable strategy
would be to achieve stereoselective glycosylation of the more
readily soluble acyclic diene 62 (Scheme 21), which might then
carbonyl function by first preparing the derived methyl glycoside
(Scheme 18). Subsequent protection of the secondary amine
(f N-Cbz) delivered 54. Acylation of the remaining hydroxyl
groups and conversion of the methoxy acetal to its acetoxy
derivative, according to the procedure documented by Pogszay,48
afforded 55 (>98% R). The appropriately outfitted glycoside
55 was thus prepared in an overall yield of 51% (from 38).
As depicted in Scheme 19, our initial studies with 2-butanol
as substrate and BF3‚Et2O as the Lewis acid catalyst resulted
in stereoselective and efficient C-O bond formation (equal
mixture of 2-butanol and 55); 56 was formed in 90% isolated
yield (>95:5, as judged by 400 MHz 1H NMR analysis). When
a similar glycosylation was attempted on macrolactam 30, a
large mixture of unidentifiable products was formed. The latter
experiment was especially handicapped by the lack of solubility
of the macrocyclic alcohol. Since stereoselective glycosylation
may, in principle, be carried out prior to macrocyclization, we
became interested in related acyclic alcohols as substrates.
Within this context, we found that the use of the readily soluble
acyclic alcohol 22, as illustrated in Scheme 19, gave rise to the
facile formation of furan 57 (85% yield) and <2% of the desired
product.49
Scheme 20
a 1.2 equiv of PhSH, 0.7 equiv of SnCl4, 50 °C, 1 h, 82%. b 1.3
equiv of Et2NSF3, 1.5 equiv of NBS, 0 °C, 85%.
(50) Mukaiyama, T.; Katsurada, M.; Takashima, T. Chem. Lett. 1991,
985-988.
(51) (a) Mukaiyama, T.; Murai, Y.; Shoda, S. Chem. Lett. 1981, 431-
432. (b) Mukaiyama, T.; Hashimoto, Y.; Shoda, S. Chem. Lett. 1983, 935-
938.
(48) Pozsgay, V. J. Chem. Soc. 1995, 117, 6673-6681.
(49) Although the BF3‚Et2O used was freshly distilled, it is tenable that
HF contamination gives rise to the observed intramolecular etherification.
(52) (a) Nicolaou, K. C.; Dolle, R. E.; Papahatjis, D. P.; Randall, J. L.;
J. Am. Chem. Soc. 1984, 106, 4189-4192. (b) Nicolaou, K. C.; Randall, J.
L.; Furst, G. T. J. Am. Chem. Soc. 1985, 107, 5556-5558.

EnantioselectiVe Total Synthesis of FluVirucin B1
J. Am. Chem. Soc., Vol. 119, No. 43, 1997 10311
Scheme 22
Scheme 21
a 2.2 equiv of AgClO4, 2.2 equiv of SnCl2, 4 Å molecular sieves,
1.1 equiv of 61, -15 °C (1 h), 0 °C (1 h), 22 °C (2 h), Et2O, 92%. b 20
mol % Mo(CHCMe2Ph)N(2,6-(i-Pr)2C6H3))(OCMe(CF3)2)2, C6H6, 22
°C, 4 h, 92%. c MeOH, NH3, 96%.
be induced to undergo catalytic ring closure. The success of
the proposed route depended on whether the Lewis acidic
metathesis catalyst (14) would remain operative in the presence
of additional Lewis basic heteroatoms carried by the pendant
carbohydrate moiety.
Diene alcohol 62 was prepared by desilylation of silyl ether
29 (Scheme 7; 48% HF, MeCN, 22 °C, 85% yield).53 As
illustrated in Scheme 21, after extensive experimentaion,
glycosylation of 62 with 61 was effected in 92% isolated yield,
in the presence of only 1.1 equiv of the fluoroglycoside, to afford
carbohydrate-diene ensemble 63 as a single diastereoisomer
(>98%, as judged by 400 MHz 1H NMR analysis). Variations
in reaction temperature are critical to the efficiency of the
coupling process; when the reaction was allowed to proceed at
-15 °C and then at 0 °C, notably lower conversions were
achieved (<60% after 24 h). When glycosylation was not
initiated at -15 °C, inferior selectivities were observed.
in facile generation of the amine carbohydrate (2-3 h; in
addition to small amounts of alkene hydrogenation products).
Longer reaction times were however required for hydrogenation
of the trisubstituted olefin. Toward this end, prolonged subjec-
tion of 66 to the same reaction conditions resulted in significant
decomposition of the glycosidic moiety (unidentified products).
When methyl glycoside 54 was treated to identical hydrogena-
tion conditions, the deprotected sugar 67 was obtained cleanly
after 2 h, but could not be recovered after 8 h of reaction time
(Scheme 22).
The sensitivity of the unprotected glycoside to hydrogenoly-
sis/hydrogenation conditions was puzzling; after all, carbohy-
drate 38 was synthesized under 300 psi H2 in the presence of
Pd(OH)2 (cf. Scheme 15). Since the latter process is effected
in acidic medium (2.5 equiv of HCl), we conjectured that the
ammonium salt of the carbohydrate may prove to be stable to
the hydrogenation conditions. Control experiments, shown in
Scheme 23, were therefore carried out, indicating that the
hydrochloride salt of 54 is stable to the above conditions even
after 10 h and the macrocyclic trisubstituted olefin 68 can be
selectively hydrogenated in the presence of 3 equiv of HCl. In
spite of these positive indications, attempts to deprotect and
hydrogenate 65 under identical conditions resulted in significant
substrate decomposition. Careful examination of the 1H NMR
indicated that alkene hydrogenation only proceeded to ∼50%
conversion after 10 h, but that Cbz deprotection was complete.
Catalytic Macrocyclization
Studies with the Cbz-Protected Glycoside. We quickly
established that subjection of 63 with 20 mol % of freshly
prepared 14 after 4 h at 22 °C affords 64 in 92% yield after
silica gel chromatography (>98% Z). At this juncture, what
remained to be accomplished was the stereoselective hydrogena-
tion of the trisubstituted olefin and deprotection of the pendant
sugar.
Based on our preliminary studies, catalytic hydrogenation of
the macrocyclic olefin promised to proceed with the desired
sense and appropriate levels of stereocontrol (cf. Chart 2 and
Scheme 7). In addition, the acetate and Cbz groups of the
carbohydrate region were to be cleaved. We judged that
removal of the Cbz unit as the first step could lead to unwanted
acyl transfer to afford the derived acetamides, cleavage of which
would engender the same for the macrocycle amide. Thus, as
depicted in Scheme 21, the acetoxy groups were excised first
to afford 65 (96%). Ammonia in MeOH was used to simplify
the workup procedure; the volatile reagent and solvent were
then easily evaporated in Vacuo.
Scheme 23
Deprotection of the Cbz group of 65 under 1 atm H2 pressure
and in the presence of Pd(C), as shown in Scheme 22, resulted
(53) Diene carbinol 62 may also be prepared by direct coupling of the
amine derived from 22 (cf. Scheme 4) with carboxylic acid 18 (cf. Scheme
6). However, this procedure is significantly lower yielding than reaction
with silylated derivative 17 (85% vs 55-60%). The difference in efficiency
likely arises from difficulties in the purification of 22 prior to the coupling
procedure.
Studies with the Trifluoroacetate Glycoside. The above-
mentioned complications suggested that deprotection of the

10312 J. Am. Chem. Soc., Vol. 119, No. 43, 1997
Xu et al.
Scheme 24
a 10% anhydrous HCl in MeOH, 88%. b 1.5 equiv of CF3COSEt, 1.5 equiv of Et3N, MeOH. c 10 equiv of Ac2O, pyridine, 5 mol %
(dimethylamino)pyridine. d 1% H2SO4 in Ac2O, 0 °C, 1 h, 85% from 69. e 1.2 equiv of PhSH, 0.7 equiv of SnCl4, 50 °C, 1 h. f 1.3 equiv of Et2NSF3,
1.5 equiv of NBS, 0 °C, 65% from 70. g 2.2 equiv of AgClO4, 2.2 equiv of SnCl2, 4 Å molecular sieves, 1.1 equiv of 62, -15 °C (1 h), 0 °C (1
h), 22 °C (2 h), Et2O, 92%. h 20 mol % Mo(CHCMe2Ph)N(2,6-(i-Pr)2C6H3))(OCMe(CF3)2)2, C6H6, 22 °C, 10 h, 91%. i H2, 10% Pd(C), EtOH, 72%.
j 20 equiv of N2H4, MeOH 24 h, 96%.
carbohydrate unit would have to be accomplished subsequent
to hydrogenation of the macrolactam olefin. Such consider-
ations, in turn, implied that a different protective group, other
than a Cbz unit, would have to be employed. The most
desireable amine protection would be with a group that satisfies
at least two important requirements: (1) For the purpose of
efficiency, it would have to be excised under the same conditions
as the acetoxy groups. (2) The amine protecing group may be
removable in the same vessel as the acetoxy functions; however,
it would have to be more robust than the acetate groups. This
would ensure that premature unmasking of the secondary amine
would not lead to acyl tranfer and formation of the derived
acetamide (see above). The trifluoroacetoxy moiety satisfies
the required criteria.
remarkably efficient and facile catalytic ring closing metathesis
of dienes 29 (Scheme 7), 63 (Scheme 21), and 72 (Scheme 24).
These examples illustrate that this process offers a mild and
stereoselective route to the synthesis of highly functionalized
macrocycles bearing trisubstituted olefins. Subsequent to
preliminary reports concerning this work,5 elegant studies by
Danishefsky,54 Nicolaou,55 Fuchs,56 Furstner,57 and Schinzer58
have further demonstrated the impressive utility of this protocol
in the preparation of macrocyclic disubstituted alkenes (with
15 as catalyst).59 Finally, the macrolactam synthesis (3, Scheme
7) underlines the significance of inorganic chemistry and metal-
catalyzed processes to organic synthesis: nine of the required
thirteen steps are catalytic and every issue in stereo- and
regiocontrol is resolved by the action of an organometallic
complex. The reaction methods and strategies developed in this
study should find applications in future endeavors in complex
molecule total synthesis.
As illustrated in Scheme 24, perfluoroglycoside 71 was
prepared through similar synthesis routes as described above
(1.5:1 mixture of R:â anomers; 58% overall yield from 38‚HCl).
Diastereoselective glycosylation was carried out with 71 and
diene carbinol 62 (cf. Scheme 21) to afford 72 in 92% yield
Experimental Section60
1
2-Methyl-6-(R)-ethyl-1-nonen-5-(R)-ol-9-toluenesulfonyl Amide
(22). Allylic alcohol (R)-20 (320 mg, 2.54 mmol) was dissolved in 6
mL of anhydrous Et2O in a 50 mL round bottom flask. EtMgCl (6.3
(>98% R anomer, as judged by 400 MHz H NMR analysis).
Furthermore, Mo-catalyzed ring closure was effected in 91%
isolated yield at ambient temperature to afford 73 (>98% Z).
Stereocontrolled hydrogenation of the trisubstituted olefin (72%)
and removal of the acetate and trifluroacetate groups, effected
by subjection of the hydrogenated adduct with hydrazine in
MeOH, delivered Sch 38516 (1) in 96% yield. The synthetic
(54) (a) Bertinato, P.; Sorensen, E. J.; Meng, D.; Danishefsky, S. J. J.
Org. Chem. 1996, 61, 8000-8001. (b) Meng, D.; Su, D.-S.; Balog, A.;
Bertinato, P.; Sorensen, E. J.; Danishefsky, S. J.; Zheng, Y.-H.; Chou, T.-
C.; He, L.; Horwitz, S. B. J. Am. Chem. Soc. 1997, 119, 2733-2734.
(55) (a) Nicolaou, K. C.; He, Y.; Vourloumis, D.; Vallberg, H.; Yang,
Z. Angew. Chem., Int. Ed. Engl. 1996, 35, 2399-2401. (b) Yang, Z.; He,
Y.; Vourloumis, D.; Vallberg, H.; Nicolaou, K. C. Angew. Chem., Int. Ed.
Engl. 1997, 36, 166-168. (c) Nicolaou, K. C.; Sarabia, F.; Ninkovic, S.;
Yang, Z. Angew. Chem., Int. Ed. Engl. 1997, 36, 525-527. (d) Nicolaou,
K. C.; Winssinger, N.; Pastor, J.; Ninkovic, S.; Sarabia, F.; He, Y.;
Vourloumis, D.; Yang, Z.; Li, T.; Giannakakou, P.; Hamel, E. Nature 1997,
387, 268-272.
(56) Kim, S. H.; Figueroa, I.; Fuchs, P. L. Tetrahedron Lett. 1997, 38,
2601-2604.
(57) Furstner, A.; Kindler, N. Tetrahedron Lett. 1996, 37, 7005-7008.
(58) Schinzer, D.; Limberg, A.; Bauer, A.; Bohm, O. M.; Cordes, M.
Angew. Chem., Int. Ed. Engl. 1997, 36, 523-524.
(59) More recent accounts from the laboratories of Grubbs and Sauvage
indicate that the metathesis technology may be readily applied to the
synthesis of marcocyclic crown ethers and [2]catenanes. See: (a) Marsella,
M. J.; Maynard, H. D.; Grubbs, R. H. Angew. Chem., Int. Ed. Engl. 1997,
36, 1101-1103. (b) Mohr, B.; Weck, M.; Sauvage, J.-P.; Grubbs, R. H.
Angew. Chem., Int. Ed. Engl. 1997, 36, 1308-1310.
(60) General experimental information for select reagents and compounds
not involved in the final synthesis route and copies of spectral data for all
synthetic intermediates can be found in the Supporting Information section.
1
material proved identical to the natural sample based on H
NMR, 13C NMR, IR, HRMS, and TLC analyses.
Conclusions
The first enantioselective total synthesis of Sch 38516
(fluvirucin B1) has been accomplished. The key features of this
study include: (i) Diastereoselective catalytic ethylmagnesation
of allylic alcohol 20 and trapping of the resulting alkyl-
magnesium halide with tosyl aziridine (f 22). (ii) Zirconium-
catalyzed enantioselective alkylation of unsaturated furan 25
to afford optically pure 26. (iii) The subsequent one-pot
catalytic hydrovinylation process (26 f 28). (iv) The Ru-
catalyzed oxidation of primary alcohol to carboxylic acid (28
f 18). (v) Diastereoselective dipolar cycloaddition between
nitrone 47 and vinylene carbonate, promoted by a readily
available chiral auxiliary (R-methylphenylamine). (vi) The

EnantioselectiVe Total Synthesis of FluVirucin B1
J. Am. Chem. Soc., Vol. 119, No. 43, 1997 10313
mL, 2.02 M, 12.70 mmol) was added to the reaction flask at 0 °C and
stirred for 5 min. Zirconocene dichloride (37 mg, 0.13 mmol, 5.0 mol
%) was subsequently added at 22 °C. The solution was allowed to
stir for 18 h at 22 °C after which the reaction was determined to be
complete by thin layer chromatography analysis (4:1 hexanes-ethyl
acetate). In a separate flame-dried flask, N-(toluenesulfonyl)aziridine
(3.0 g, 15.2 mmol, 6.0 equiv) was dissolved in 15 mL of anhydrous
THF. The latter solution was quickly added to the original flask Via
cannula. The second vessel was rinsed twice with 1.0 mL portions of
anhydrous THF, which was also transferred to the original reaction
flask. CuBr‚Me2S (26 mg, 0.13 mmol; 5.0 mol %) was placed in a 10
mL round bottom flask in a dry box; to this was added 0.5 mL of
anhydrous THF and 0.5 mL of Me2S. Slight agitation facilitated
solvation of CuBr‚Me2S, at which point the solution was transferred
to the cooled reaction flask (-20 °C). The CuBr‚Me2S flask was rinsed
once with 0.5 mL of anhydrous THF and 0.5 mL of Me2S, and the
solution resulting from the wash was transferred to the original reaction
flask. The mixture was allowed to warm to 22 °C over 2 h (the yellow-
white slurry turned dark brown as the temperature rose). After the
solution was allowed to stir for 30 h in a cold room (-4 °C), the reaction
resulting mixture was allowed to stir at 22 °C for 10 min, after which
it was treated with (S)-(EBTHI)Zr-binol (184 mg, 0.36 mmol, 0.4 mol
% relative to the Grignard reagent). The mixture was allowed to stir
under N2 for 40 h. Subsequent addition of water and aqueous workup
with three 20 mL portions of CH2Cl2, followed by careful solvent
evaporation afforded the crude product as a dark oil. Vacuum
distillation afforded 2.42 g (24.2 mmol) of 26 as a colorless oil. GLC
analysis of the derived epoxides indicated the product to be >99%
enantiomerically pure. The reaction could alternatively be performed
with 10 mol % catalyst to afford 70% yield but with seven catalyst
turnovers. Enantioselectivity was determined by GLC of the derived
epoxide. IR (NaCl): 3440 (br, s), 3270 (m), 3245 (m), 3210 (m), 1640
(w) cm-1; 1H NMR (300 MHz in CDCl3): δ 5.57 (ddd, 1H, J ) 16.8,
10.5, 8.5 Hz, vinylic CH), 5.15 (m, 2H, vinylic CH2), 3.57 (dd, 1H, J
) 10.3, 5.3 Hz, CH2OH), 3.41 (dd, 1H, J ) 10.6, 8.2 Hz, CH2OH),
2.11 (1H, m, CHCH2OH), 1.45 (m, 1H, CH2CH3), 1.26 (m, 1H,
CH2CH3), 0.89 (t, 3H, J ) 7.4 Hz, CH3); 13C NMR (75 MHz in
CDCl3): δ 139.7, 117.2, 65.3, 48.6, 23.5, 11.4. (Due to substrate
volatility, the derived (S)-MTPA esters were subjected to combustion
analysis.) Anal. Calcd for C16H19F3O3: C, 60.75; H, 6.05. Found:
C, 60.63; H, 6.21.
1
was judged complete by H NMR analysis of the aliquots removed
from the reaction mixture. After the addition of a 10 mL solution of
saturated aqueous ammonium chloride, the mixture was washed three
times with 30 mL portions of CH2Cl2. The combined organic layers
were dried over anhydrous MgSO4 and concentrated in Vacuo to afford
a dark oil. Purification by silica gel chromatography (1:5:0.3 hexane:
CH2Cl2:EtOAc) afforded the desired product as a yellow oil (347 mg,
1.0 mmol, 42% yield) (although the product after chromatography in
some cases still had some impurity, it could be submitted to the next
step). IR (NaCl): 3523 (br m), 3278 (br s), 3071 (w), 2961 (s), 2934
(s), 2874 (s), 1648 (w), 1598 (w), 1449 (s), 1376 (w), 1324 (s), 1158
(R)-2-Ethyl-5-hexen-1-ol (28). Homoallylic alcohol 26 (230 mg,
2.3 mmol) was added to a solution of Cp2TiCl2 (16 mg, 0.06 mmol,
2.8 mol %) in 5.2 mL of anhydrous THF. The mixture was cooled to
5 °C (ice bath), and n-PrMgBr (2.4 mL, 2.9 M in Et2O, 6.90 mmol)
was added slowly, at which time the solution changed color from red
to dark green accompanied by gas evolution. The reaction was then
allowed to warm to 22 °C, after which the reaction vessel was placed
in an oil bath and the mixture was heated to reflux for 36 h. The
reaction flask was then cooled to -78 °C (acetone, dry ice), and (Ph3P)2-
NiCl2 (34 mg, 0.05 mmol, 2.2 mol %) was added as a solution in 3.0
mL of THF. A previously prepared solution of vinyl bromide (2.6
mL) in THF (3.0 mL) at -78 °C was then added slowly, resulting in
vigorous gas evolution. The mixture was allowed to stir for 12 h at
22 °C, after which it was treated with 20 mL of a 2.0 M solution of
HCl and washed with three 15 mL portions of Et2O to afford a dark
oil as the crude residue. Purification by silica gel (containing 10%
AgNO3) chromatography (6:1 pentanes:Et2O) afforded 213 mg of 28
as a colorless oil (1.66 mmol, 72% yield). IR (NaCl): 3342 (brs),
3077 (w), 2964 (s), 2930 (s), 2874 (s), 1640 (m), 1462 (m), 1382 (m),
(s), 1094 (s), 888 (m), 814 (s), 736 (s), 663 (s) cm-1 1H NMR (400
.
MHz in CDCl3): δ 7.74 (d, 2H, J ) 8.1 Hz, CHCSO2), 7.30 (d, 2H,
J ) 8.1 Hz, CHCCH3), 4.73 (s, 1H, CHHCCH3), 4.71 (s, 1H,
CHHCCH3), 3.58 (m, 1H, CHOH), 2.94 (q, 2H, J ) 6.6 Hz, CH2NH),
2.42 (s, 3H, CH3-Ar), 2.17 (m, 1H, CHHCCH2), 2.04 (m, 1H,
CHHCCH2), 1.73 (s, 3H, CH3CCH2), 1.6-1.1 (m, 9H, alkyl-H), 0.86
(t, 3H, J ) 7.2 Hz, CH3CH2). 13C NMR (75 MHz in CDCl3): δ 145.7,
143.1, 137.0, 129.5, 127.0, 110.0, 72.7, 44.7, 43.3, 34.5, 31.1, 27.4,
26.3, 22.3, 21.8, 21.3, 11.8. HRMS Calcd for C19H31O3SN (M + H):
354.2106. Found: 354.2103.
1044 (s), 999 (m), 909 (s) cm-1 1H NMR (300 MHz in CDCl3): δ
.
5.76 (ddt, 1H, J ) 16.8, 10.2, 6.6 Hz, CH2CH), 4.94 (dq, 1H, J )
18.0, 1.8 Hz, CHHCH), 4.85 (dm, 1H, J ) 10.2 Hz, CHHCH), 3.49
(d, 2H, J ) 4.8 Hz, CH2OH), 2.03 (q, 2H, J ) 7.8 Hz, CH2CHCH2),
1.35 (m, 5H, alkyl), 0.84 (t, 3H, J ) 7.2 Hz, CH3). 13C NMR (75
MHz in CDCl3): δ 139.0, 114.3, 64.9, 41.3, 31.0, 29.6, 23.2, 11.0.
HRMS Calcd for C8H16O: 128.1201. Found: 128.1200.
2-Methyl-6-(R)-ethyl-1-nonen-5-(R)-[(tert-butyldimethylsilyl)oxy]-
9-amine (17). Tosyl amide 22 (165.0 mg, 0.35 mmol) was dissolved
in 1.0 mL of anhydrous THF, and the reaction mixture was cooled to
-40 °C (acetone, CO2). Ammonia (∼10 mL) was condensed into the
reaction vessel, and the solution was allowed to stir at -40 °C for
approximately 1 h. Sodium (48.3 mg, 2.11 mmol), washed initially
with anhydrous pentane three times, was added to the reaction mixture,
which immediately turned blue; this color persisted for 45 min. The
reaction was quenched carefully and slowly by the addition of 5 mL
of CH2Cl2, causing the blue color to disappear. After half the volume
of the ammonia evaporated, the resulting mixture was poured in a
separatory funnel containing 10 mL of brine. Organic and aqueous
layers were separated, and the aqeuous layer was washed four times
with 15 mL portions of CH2Cl2. Organic layers were dried over
anhydrous MgSO4, filtered through a fritted glass funnel, and concen-
trated in Vacuo to afford a pale yellow oil. Purification by silica gel
chromatography (Et2O, 1:5 Et2O/MeOH) gave 107 mg (0.34 mmol) of
17 (95% yield) as a colorless oil. [R]D +14.13 (c 0.29, CH2Cl2). IR
(NaCl): 3087 (w), 3077 (w), 2955 (s), 2857 (s), 2176 (w), 1649 (m),
1583 (m), 1471 (s), 1380 (m), 1254 (s), 1082 (s), 886 (m), 835 (s)
2-(R)-Ethyl-5-hexenoic Acid (18). N-Methylmorpholine oxide
(NMO, 316 mg, 2.70 mmol) was added to a solution of 28 (90 mg,
0.70 mmol) in 7 mL of anhydrous CH3CN at 22 °C. To this mixture
was added tetrapropylammonium perruthenate (TPAP, 16 mg, 0.045
mmol) as a solution in CH3CN (2.0 mL) and the solution stirred for 5
h. The solvent was evaporated in Vacuo and the black residue passed
through a short column of silica gel (2 cm) and eluted with EtOAc to
afford 64 mg of 18 as a colorless oil (0.44 mmol, 77% yield). IR
(NaCl): 2900 (brs), 1702 (s), 1642 (m), 1460 (s), 1416 (s), 1289 (s),
1274 (s), 1228 (s), 938 (m), 913 (s) cm-1 1H NMR (300 MHz in
.
CDCl3): δ 5.78 (ddt, 1H, J ) 17.1, 10.2, 6.6 Hz, CHCH2), 5.02 (d,
1H, J ) 19.2 Hz, CHCHH), 4.96 (d, 1H, J ) 10.2 Hz, CHCHH), 2.33
(m, 1H, CHCOOH), 2.09 (m, 2H, CH2CHCH2), 1.8-1.5 (m, 4H,
CH2CH3 and CH2CHCOOH), 0.94 (t, 3H, J ) 7.5 Hz, CH3). 13C NMR
(75 MHz in CDCl3): δ 183.4, 137.7, 115.1, 46.4, 31.4, 30.7, 25.1,
11.6. HRMS Calcd for C8H14O2: 142.0993. Found: 142.0993.
N-(4-(R)-Ethyl-5-(R)-((tert-butylmethylsilyl)oxy)-8-methyl-8-
nonenyl)-2-(R)-ethyl-5-hexenamide (29). Carboxylic acid 18 (107 mg,
0.34 mmol) was dissolved in 3.0 mL of anhydrous CH2Cl2 in a 25 mL
round bottom flask. The solution was cooled to 0 °C (ice bath),
followed by addition of dicyclohexylcarbodiimide (76 mg, 0.37 mmol)
and hydroxybenzotriazole (55 mg, 0.41 mmol). After 5 min amine 17
(48.2 mg, 0.34 mmol), dissolved in 1.0 mL of CH2Cl2, was added to
the original mixture. The reaction mixture was allowed to warm to 22
°C and stirred at this temperature for 12 h. White solid formation was
cm-1 1H NMR (400 MHz in CDCl3): δ 4.66 (br d, 2H, J ) 11.4 Hz,
.
CH2CHCH2), 3.65 (m, 1H, SiOCH), 2.66 (m, 2H, CH2NH2), 2.07 (m,
1H, CHHCCH2), 1.90 (m, 1H, CHHCCH2), 1.70 (s, 3H, CCH3), 1.6-
0.8 (9H, alkyl-H), 0.86 (s, 12H, C(CH3)4), 0.03 (s, 3H, SiCH3), 0.01
(s, 3H, SiCH3). 13C NMR (100 MHz in CDCl3): δ 146.2, 109.5, 73.4,
45.3, 42.7, 34.4, 32.2, 30.6, 26.4, 25.8, 22.6. 18.1, 12.4, -4.39, -4.35.
HRMS Calcd for C18H44NOSi (M + H): 314.2879. Found: 314.2886.
(S)-2-Ethyl-3-buten-1-ol (26). 2,5-Dihydrofuran 25 (7.30 mL, 96.67
mmol) was dissolved in 30 mL of anhydrous THF in a 100 mL round
bottom flask, and the solution was cooled to 0 °C. Ethylmagnesium
chloride (35.0 mL, 2.07 M in THF, 72.8 mmol) was added, and the

10314 J. Am. Chem. Soc., Vol. 119, No. 43, 1997
Xu et al.
observed as soon as the reaction temperature reached 22 °C. The
reaction was treated with 15 mL of saturated aqeuous sodium
bicarbonate solution. The layers were separated and the aqeuous layer
was washed four times with 25 mL portions of CH2Cl2. The organic
layers were combined and dried over anhydrous MgSO4 and filtered
through a fritted funnel. Solvent removal in Vacuo afforded a yellow
oil, which was purified by silica gel (containing 5% AgNO3) chroma-
tography (3:1 hexanes:EtOAc) to afford 125 mg of the desired amide
29 (0.33 mmol, 85% yield) as a clear oil. IR (NaCl): 3291 (s), 3080
(m), 2960 (s), 2932 (s), 2858 (s), 1643 (s), 1554 (s), 1462 (s), 1378
(m), 1256 (s), 1085 (s), 1064 (s), 1006 (m), 911 (m), 886 (s), 836 (s),
to afford a dark brown oil. Purification by silica gel chromatography
with 5:1 hexanes/EtOAc afforded 96.4 mg (0.28 mmol) of desired
product 48 as a white solid (69% yield). IR (NaCl): 2986 (m), 2936
(w), 2886 (w), 1822 (s), 1463 (w), 1381 (m), 1262 (w), 1168 (m), 1067
(m), 992 (m), 859 (w), 702 (w) cm-1 1H NMR (400 MHz in CDCl3):
.
δ 7.27-7.38 (m, 5H, aromatic-H), 6.28 (dd, 1H, J ) 5.2, 0.4 Hz,
OCHO), 5.48 (d, 1H, J ) 5.2 Hz, CHCHOCO), 4.02 (q, 1H, J ) 6.0
Hz, PhCHMe), 3.45 (d, 1H, J ) 7.6 Hz, NCH), 3.29 (t, 1H, J ) 7.6
Hz, NCHCH), 3.24 (dq, 1H, J ) 7.6, 6.0 Hz, MeCHOCH), 1.57 (d,
3H, J ) 6.0 Hz, PhCHCH3), 1.31 (s, 3H, CH3CCH3), 1.23 (s, 3H, CH3-
CCH3), 1.22 (d, 3H, J ) 6.0 Hz, CH3CO). 13C NMR (100 MHz in
CDCl3): δ 152.6, 140.7, 128.9, 128.7, 128.1, 109.0, 105.1, 87.2, 79.4,
76.0, 68.4, 67.5, 27.1, 26.5, 21.2, 19.1. HMRS Calcd for C18H23NO6:
349.1525. Found: 349.1527.
774 (s) cm-1 1H NMR (400 MHz): δ 5.81 (ddt, 1H, J ) 17.1, 10.2,
.
6.8 Hz, CHCH2), 5.45 (br, s, 1H, NH), 5.03 (d, 1H, J ) 17.5 Hz,
CHCHH), 5.00 (d, 1H, J ) 11.1 Hz, CHCHH), 4.74 (s, 1H, CCHH),
4.70 (s, 1H, CCHH), 3.71 (m, 1H, CHOTBS), 3.29 (q, 2H, J ) 6.3
Hz, CH2NH), 2.13 (m, 2H, CH2CHCH2), 2.03 (p, 1H, J ) 8.0 Hz,
CHCONH), 1.94 (m, 2H, CH2C(CH3)CH2), 1.76 (s, 1H, CH3CCH2),
1.70-1.20 (m, 13H, alkyl CH), 0.92 (br, s, 15H, (CH3)3Si and CH3),
0.08 (s, 3H, CH3Si), 0.07 (s, 3H, CH3Si). 13C NMR (100 MHz, in
C6D6): δ 176.2, 146.1, 138.3, 114.9, 109.5, 94.4, 73.4, 49.0, 45.2, 39.8,
34.4 31.7, 31.7, 30.4, 28.4, 26.6, 26.0, 25.9, 22.9, 22.6, 18.1, 12.4,
12.1, -4.2, -4.4. HRMS Calcd for C26H51O2NSi (M - (t-Bu)):
380.2988. Found: 380.2984.
Diol Isoxazolidine 52. Tricycle 48 (931 mg, 2.66 mmol) was
dissolved in 13.3 mL of a 4:1 THF/1 M aqueous hydrochloric acid
solution. The resulting mixture was heated to 65 °C. Reaction progress
was monitored by TLC (1:1 hexane/EtOAc, Rf ) 0.75 for starting
material, 0.30 for product), which indicated complete consumption of
starting material after 4 h . Reaction was quenched by the addition of
50 mL of a saturated solution of sodium bicarbonate. Aqueous and
organic layers were separated, and the aqueous layer was washed four
times with 50 mL portions of EtOAc. The combined organic layers
were washed with brine (100 mL), dried over anhydrous MgSO4, filtered
through a fritted funnel, and concentrated in Vacuo to afford a pale
yellow oil, which could be used for next step without purification. When
necessary, purification by silica gel chromatography (2:1 hexanes/
EtOAc) provides 733 mg (2.50 mmol) of diol 52 as a colorless oil
(94% yield). IR (NaCl): 3459 (br), 3069 (w), 3032 (w), 2980 (m),
2936 (w), 1822 (s), 1734 (m), 1495 (m), 1451 (s), 1376 (s), 1313 (m),
Ethyl 4,5-Dihydroxy-2-trans-hexenoate (43). Ethyl sorbate (24.0
g, 0.17 mol) was placed in a 3 L round bottom flask, after which was
added a 1:1 mixture of t-butyl alcohol (860 mL) and H2O (860 mL) at
0 °C, followed by the addition of 249.0 g of AD-mix-R and
methanesulfonamide (16.3 g, 0.17 mol). This mixture was allowed to
stir at 0 °C while the reaction was monitored by TLC (Rf ) 0.05 and
1
0.95 for ethyl sorbate, 3:1 hexanes/EtOAc) and 400 MHz H NMR
analysis. After 20 h, the reaction was quenched by the addition of a
saturated aqueous solution of sodium thiosulfite pentahydrate (24.0 g,
96.7 mmol). The resulting solid was subsequently removed by
filtration. Organic and aqueous layers were separated, and the aqueous
layer was washed four times with 200 mL portions of EtOAc. The
combined organic layers were washed with 200 mL of a 2.0 M KOH
solution, followed by 400 mL of brine. Organic layers were dried over
anhydrous magnesium sulfate, filtered through a fritted funnel, and
concentrated in Vacuo to afford a dark yellow oil. The unpurified
product was passed through a short column of silica gel to remove
excess salts. This material was used directly for conversion to
acetonide. IR (NaCl): 3431 (br, s), 2980 (m), 2934 (w), 2908 (w),
1168 (s), 1080 (s), 985 (s) cm-1
.
1H NMR (400 MHz in CDCl3): δ
7.42-7.30 (m, 5H, aromatic-H), 6.62 (dd, 1H, J ) 5.6, 0.8 Hz, OCHO),
5.64 (d, 1H, J ) 5.6 Hz, CHCHOCO), 4.02 (q, 1H, J ) 6.4 Hz, PhCH),
3.78 (qd, 1H, J ) 6.4, 2.0 Hz, CH3CHOH), 3.43 (d, 1H, J ) 8.8 Hz,
NCH), 2.95 (dd, 1H, J ) 8.8, 2.0 Hz MeCHOHCHO), 1.40 (d, 3H, J
) 8.8 Hz, PhCHCH3), 1.10 (d, 3H, J ) 6.4 Hz, CH3CHO). 13C NMR
(100 MHz in CDCl3): δ 152.8, 141.0, 129.1, 128.7, 128.1, 105.3, 87.2,
72.6, 67.6, 67.57 , 65.6, 21.0, 19.0. HRMS Calcd for C14H17O6N (M
+ H): 296.1134 . Found: 296.1129.
3-Amino-3,6-dideoxy-r-L-talopyranose (38). Diol 52 (0.737 mmol)
was dissolved in 36 mL of MeOH and brought into a glove box, at
which time 50% (by weight) Pd(OH)2 was added. The reaction flask
was removed from the glove box and charged with acetyl chloride (143
mL, 2.01 mmol). The resulting mixture was placed in a high pressure
bomb which was subsequently purged with hydrogen three times.
Hydrogenation was allowed to proceed under 300 psi of hydrogen at
22 °C. After 14 h, the reaction mixture was filtered through Celite
and concentrated in Vacuo to provide a light yellow solid, which could
be directly submitted to the next step. When necessary, purification
by silica gel chromatography (4:1 CH2Cl2/MeOH) afforded 120 mg
(0.70 mmol) of 38 (7:1 ratio of R/â anomers, 95% yield), the ratio of
both anomers varied upon silica gel chromatography. The mixture is
a white solid. IR (NaCl): 3312 (brs), 2985 (s), 2938 (s), 1606 (w),
1711 (s), 1662 (m), 1447 (w), 1370 (m), 1278 (m) cm-1
.
1H NMR
(400 MHz in CDCl3): δ 6.92 (dd, 1H, J ) 15.6, 5.2 Hz, CHCHCOO),
6.15 (dd, 1H, J ) 15.6, 2.5 Hz, CHCHCOO), 4.22 (q, 2H, J ) 7.1 Hz,
OCH2CH3), 4.05 (qd, 1H, J ) 5.2, 1.6 Hz, CH3CHO), 3.74 (dq, 1H, J
) 6.0, 4.0 Hz, CHCHCHO), 2.54 (m, 1H, OH), 2.16 (m, 1H, OH),
1.29 (t, 3H, J ) 7.0 Hz, CH3CH2), 1.25 (d, 3H, J ) 6.4 Hz, CH3CHO).
13C NMR (100 MHz in CDCl3): δ 166.4, 146.5, 122.5, 75.6, 70.3,
60.5, 18.9, 14.8.
Nitrone 47. Aldehyde 45 (55.0 mg, 0.38 mmol) was dissolved in
4 mL of benzene, and (R)-N-hydroxy-R-methylbenzylamine (60.0 mg,
0.43 mmol) was added to this solution. The resulting mixture was
allowed to reflux for 5 h. The reaction mixture was cooled to 22 °C,
and solvent was removed in Vacuo to give a yellow oil. Purification
by silica gel chromatography (1:1 hexanes:EtOAc) afforded 84.0 mg
of the desired nitrone as a colorless oil (0.32 mmol, 74% yield). IR
(NaCl): 3069 (m), 2987 (s), 2924 (S), 2867 (m), 1734 (w), 1590 (s),
1457 (s), 1388 (s), 1250 (s), 1168 (s), 1086 (s), 859.4 (s). 1H NMR
(400 MHz in CDCl3): δ 7.4 (m, 5H, aromatic-H), 6.88 (d, 1H, J ) 5.6
Hz, OCHCHN), 5.02 (q, 1H, J ) 6.8 Hz, NCHCH3), 4.87 (dd, 1H, J
) 7.3, 5.6 Hz, NCHCH), 3.97 (qd, 1H, J ) 6.0, 1.2 Hz, MeCHO),
1.80 (d, 3H, J ) 6.9 Hz, PhCHCH3), 1.45 (d, 1H, J ) 6.4 Hz, CH3-
CHOCH), 1.44 (s, 3H, CH3CCH3), 1.35 (s, 3H, CHCCH3). 13C NMR
(100 MHz in CDCl3): δ 138.1, 135.4, 128.8, 128.7, 127.2, 126.9, 109.5
77.3, 76.3, 73.5, 53.6, 27.1, 26.3, 19.4, 18.9. HRMS Calcd for C15H22-
NO3 (M + H): 264.1599. Found: 264.1600.
1505 (m), 1380 (w), 1054 (s), 1016 (m), 975 (w), 860 (w) cm-1 1H
.
NMR (400 MHz in CD3OD): δ (R-anomer): 5.20 (s, 1H, C1-H), 4.17
(q, 1H, J ) 6.2 Hz, C5-H), 3.75 (br, m, 1H, C2-H), 3.69 (br, m, 1H,
C4-H), 3.56 (br, m, 1H, C3-H), 1.24 (d, 3H, J ) 6.5 Hz, C6-H3). (â-
anomer): 4.75 (s, 1H, C1-H), 3.85 (br, s, 1H, C2-H), 3.71 (br, 1H, C4-
H), 3.65 (br, m, 1H, C3-H), 1.28 (d, 3H, J ) 6.0 Hz, C6-H3), C5-H
buried underneath one peak of â-anomer. 13C NMR (100 Hz in
CD3OD): δ (R-anomer) 95.6, 70.0, 68.9, 66.8, 50.21, 17.1. HRMS
Calcd for C6H13O4N (M - Cl): 164.0928. Found: 164.0928.
Methoxy-3-amino-3,6-dideoxy-r-L-talopyranose (69). Hydroxy
glycoside 38 (128 mg, 0.76 mmol) was placed in a 100 mL round
bottom flask, to which was added 36 mL of 10% (by weight) anhydrous
methanolic hydrochloric acid solution. Stirring was allowed to continue
for 70 h at 22 °C. The resulting reaction mixture was concentrated in
Vacuo to give a light yellow solid, which could be subsequently
submitted to next step without purification. Purification by silica gel
chromatography (4:1 CH2Cl2/CH3OH) to provide 116 mg (0.63 mmol)
of the R-methoxy glycoside 69 (84% yield from the diol). Note: If
reaction time is too brief (e.g., 20 h), the product will be contaminated
Isoxazolidine 48. Nitrone 47 (106 mg, 0.40 mmol) was placed in
a flame-dried round bottom flask equipped with a reflux condenser
(under argon). Vinylene carbonate (102 mL, 1.6 mmol), freshly distilled
over calcium hydride under vacuum (15 mmHg), was added to the
nitrone substrate. The resulting mixture was heated at 85 °C for 19 h

EnantioselectiVe Total Synthesis of FluVirucin B1
J. Am. Chem. Soc., Vol. 119, No. 43, 1997 10315
with the â-anomer. IR (NaCl): 3339 (brs), 2949 (s), 2829 (s), 2363
dried over anhydrous MgSO4, filtered, and concentrated in Vacuo.
Purification by silica gel chromatography (2:1 Et2O/pentane) gave 213
mg (0.58 mmol, 80% yield) of the thiophenyl glycoside as a pale yellow
oil. IR (NaCl): 3320 (w), 3049 (w), 2987 (w), 1740 (s), 1539 (m),
(w), 1621 (w), 1507 (m), 1444 (m), 1381 (m), 1193 (w), 1117 (s),
1067 (s), 1016 (s) cm-1 1H NMR (400 MHz in CD3OD): δ 4.68 (br,
.
s, 1H, C1-H), 3.92 (q, 1H, J ) 6.5 Hz, C5-H), 3.75 (dd, 1H, J ) 3.3,
1.6 Hz C2-H), 3.66 (m, 1H, C4-H), 3.47 (t, 1H, J ) 3.3 Hz, C3-H),
3.34 (s, 3H, OCH3), 1.20 (d, 3H, J ) 6.5 Hz, C6-H3). 13C NMR (125
MHz in CD3OD): δ 102.4, 69.7, 67.9, 67.3, 55.8, 50.4, 17.0. HRMS
calcd for C7H16O4N (M - Cl): 178.1079. Found: 178.1083.
1444 (w), 1375 (m), 1218 (s), 1155 (s), 1098 (m) cm-1 1H NMR
.
(400 MHz in CDCl3): δ 7.4-7.26 (m, 5H, aromatic-H), 6.83 (br, 1H,
NH), 5.60 (s, 1H, C1-H), 5.20 (m, 1H, C2-H), 5.13 (d, 1H J ) 3.7 Hz,
C4-H), 4.60 (m, 2H, C3-H, C5-H), 2.18 (s, 3H, COCH3), 2.16 (s, 3H,
COCH3), 1.19 (d, 3H, J ) 6.6 Hz, C6-H3). 13C NMR (100 MHz in
CDCl3): δ 170.8, 170.3, 156.8 (q, JC-F ) 38 Hz), 132.5, 131.7, 129.0,
127.8, 115.2 (q, JC-F ) 286 Hz), 83.3, 69.4, 69.0, 66.2, 47.2, 20.7,
20.3, 16.4. The resulting thioglycoside (192 mg, 0.44 mmol), dried
by azeotropic distillation with benzene, was dissolved in 8.8 mL of
CH2Cl2. The resulting solution was subsequently treated with DAST
(70.2 mL, 0.58 mmol), and the resulting mixture was allowed to stir
for 2 min before NBS (118 mg, 0.67 mmol) was added at -15 °C (dry
ice and acetone bath). The reaction mixture was allowed to stir at this
temperature for 1 h and at 0 °C for 2 h, after which the reaction was
quenched by the addition of a 15 mL solution of a saturated solution
of sodium bicarbonate. Aqueous and organic layers were separated,
and the aqueous layer was washed with CH2Cl2 (3 × 20 mL portions).
Organic layers were dried over anhydrous MgSO4, filtered through a
fritted glass funnel, and concentrated in Vacuo to provide a pale yellow
oil. Purification by silica gel chromatography (2:1 Et2O/pentane)
afforded a partial separable 3:2 mixture of both R and â anomers of
71 (129 mg, 0.36 mmol, 81% yield); the product was a white foam.
For R-anomer of 71: IR (NaCl): 3339 (w), 1734 (s), 1539 (m), 1382
Acetoxy-2,4-di-O-acetyl-3-N-(trifluoroacetyl)-3,6-dideoxy-r-L-
talopyranose (70). To a solution of methoxy glycoside 69 (133 mg,
0.63 mmol) in 10 mL of methanol were added triethylamine (0.13 mL,
0.94 mmol) and S-ethyl trifluorothioacetate (117 mL, 0.93 mmol). The
resulting mixture was allowed to stir for 4 h at 22 °C and then
concentrated to dryness in Vacuo. Pyridine (1.0 mL, 10.6 mmol), acetic
anhydride (295 mL, 6.25 mmol), and a catalytic amount of DMAP
(∼10 mg) were added to the resulting solid. The reaction mixture was
allowed to stir at 22 °C for 10 h, after which the reaction was quenched
by the addition of 15 mL of a saturated aqueous copper sulfate solution.
Organic and aqueous layers were separated; the aqueous layer was
washed with 3 × 20 mL of EtOAc. Combined organic layers were
virgously stirred over 50 mL of saturated aqeuous sodium bicarbonate
for 40 min. Layers were once again separated and the aqueous layer
was washed with 3 × 20 mL portions of EtOAc. Combined organic
layers were washed with a 100 mL solution of brine, dried over
anhydrous MgSO4, filtered through a fritted funnel, and concentrated
in Vacuo to afford a light yellow solid, which could be directly used in
the next step without purification. When necessary, purification by
silica gel chromatography (2:1, hexanes/EtOAc) afforded 154 mg (0.43
mmol) of the desired peracylated methyl glycoside as a white foam
(R-anomer, 92% yield). IR (NaCl): 3446 (w), 3326 (w), 3068 (w),
2943 (m), 2841 (w), 1759 (s), 1545 (m), 1381 (s), 1230 (s), 1167 (s),
(m), 1224 (s), 1162 (s), 1036 (m), 966 (m) cm-1 1H NMR (300 MHz
.
in CDCl3): δ 6.83 (br, s, 1H, NH), 5.68 (d, 1H, JH-F ) 48.0 Hz, C1-
H), 5.20 (m, 1H, C2-H), 5.03 (m, 1H, C4-H), 4.66 (dt, 1H, J ) 7.9, 3.3
Hz, C3-H), 4.35 (qd, 1H, J ) 6.6, 1.2 Hz, C5-H), 2.18 (s, 3H, COCH3),
2.17 (s, 3H, COCH3), 1.2 (d, 3H, C5-CH3, J ) 6.6 Hz). 13C NMR
(100 MHz in CDCl3): δ 170.8, 170.2, 156.8 (q, JC-F ) 40.0 Hz), 115.5
(q, JC-F ) 280.0 Hz, 104.3 (d, JC-F ) 221.0 Hz), 68.4, 67.6, 67.59,
66.2 (d, JC-F ) 41.0 Hz), 45.9, 20.7, 20.5, 16.2. For â-anomer of 71:
IR (NaCl): 3333 (w), 2955 (w), 2351 (w), 1750 (s), 1533 (m), 1376
(m), 1220 (s), 1164 (s), 1099 (m), 1029 (m). 1H NMR (400 MHz in
CDCl3): δ 7.00 (br, d, 1H, J ) 6.8 Hz , 1H, NH), 5.50 (dd, 1H, JH-F
) 50.0, 1.6 Hz, C1-H), 5.30 (ddd, 1H, J ) 8.2, 3.8, 1.6 Hz C2-H), 5.16
(m, 1H, C4-H), 4.54 (dt, 1H, J ) 8.0, 3.8 Hz, C3-H), 4.03 (qd, 1H, J
) 6.6, 3.0 Hz,C5-H), 2.18 (s, 3H, COCH3), 2.15 (s, 3H, COCH3), 1.34
(d, 3H, J ) 6.6 Hz, C6-H). 13C NMR (100 MHz in CDCl3): δ 170.42,
170.40, 157.1 (q, JC-F ) 38.0 Hz), 115.4 (q, JC-F ) 288.0 Hz) 105.1
(d, JC-F ) 217.0 Hz), 70.8, 67.2, 66.1 (d, JC-F ) 20.0 Hz), 48.5, 48.48,
20.6, 20.4, 16.7.
1136 (s), 1073 (s), 1029 (s), 960 (m) cm-1
.
1H NMR (400 MHz,
CDCl3): δ 6.85 (bd, 1H, J ) 8.0 Hz, NH), 5.12 (m, 1H, C2-H), 4.83
(m, 1H, C4-H), 4.76 (d, 1H, J ) 1.1 Hz, C1-H), 4.60 (dt, 1H, J ) 8.0,
3.5 Hz, C3-H), 4.10 (qd, 1H, J ) 6.2, 1.1 Hz, C5-H), 3.4 (s, 3H, OCH3),
2.15 (s, 3H, COCH3), 2.148 (s, 3H, COCH3), 1.18 (d, 3H, J ) 6.6 Hz,
C6-H3). 13C NMR (100 MHz in CDCl3): δ 170.8, 170.6, 156.6 (q,
JC-F ) 38 Hz), 115.4 (q, JC-F ) 286 Hz), 97.9, 69.0, 68.2, 64.6, 55.2,
46.3, 20.8, 20.4, 16.2. 19F (in CDCl3): δ 95.4. HRMS Calcd for
C13H18F3NNaO7 (M + Na): 380.0933. Found: 380.0942. To the
peracylated amine glycoside (261 mg, 0.73 mmol) was added 2.53 mL
of freshly prepared 1% (by volume) H2SO4 acetic anhydride solution
at 0 °C. After 1 h, the reaction solution was slowly poured into 50 mL
of saturated aqeuous sodium bicarbonate solution at 0 °C and vigrously
stirred for 30 min at 22 °C. Aqueous and organic layers were separated,
and the aqueous layer was washed four times with 30 mL portions of
CH2Cl2. Combined organic layers were dried over anhydrous MgSO4,
filtered through a fritted funnel, and concentrated in Vacuo to afford a
white foam, which was virtually pure and could be submitted to the
next step without purification. When necessary, purification by silica
gel chromatography (3:1 hexanes/EtOAc) gave 240 mg (0.67 mmol)
of 70 as a white foam (92% yield). IR (NaCl) 3327 (w), 2993 (w),
1747 (s), 1539 (m), 1372 (m), 1224 (s), 1151 (m), 1021 (m) cm-1. 1H
NMR (400 MHz in CDCl3): δ 6.90 (br, d, 1H, J ) 7.3 Hz NH), 5.18
(m, 1H, C1-H), 4.88 (m, 1H, C4-H) 4.65 (dt, 1H, J ) 6.8, 3.4 Hz, C3-
H), 4.21 (qd, 1H, J ) 6.6, 1.5 Hz, C5-H), 2.17 (br, 6H, COCH3,
COCH3), 2.14 (s, 3H, COCH3), 1.17 (d, 1H, J ) 6.6 Hz, C6-H3). 13C
NMR (100 Hz in CDCl3): δ 170.8, 170.4, 168.2, 157.0 (q, JC-F ) 38
Hz), 115.3 (q, JC-F ) 286 Hz), 90.3, 68.6, 67.3, 67.1, 64.4, 20.8, 20.7,
20.4, 16.3. HRMS Calcd for C20H25O9N (M + H): 370.1114.
Found: 370.1114.
N-(4-(R)-Ethyl-5-(R)-((2′,4′-di-O-acetyl-3′-N-(trifluoroacetyl)-3′,6′-
dideoxy-r-L-talopyranosyl)oxy)-8-methyl-8-nonenyl)-2-(R)-ethyl-5-
hexenamide (72). In a 25 mL flame-dried round bottom flask diene
62 (35.0 mg, 0.11 mmol) was dried by azeotropic distillation with
benzene (3 × 5 mL). AgClO4 (49.3 mg, 0.24 mmol), SnCl4 (45.1 mg,
0.24 mmol), and flame dried 4 Å molecular sieves were then added in
a glove box. The mixture was subsequently charged with 2.0 mL of
anhydrous diethyl ether. This resulting slurry was cooled to -15 °C,
and after 2 min the glycosyl fluoride 71 (50.3 mg, 0.14 mmol dried by
azeotrope distillation with benzene) solution in 3.0 mL of Et2O was
transferred to the slurry Via cannula. The resulting mixture was allowed
to stir for 1 h at -15 °C, at 0 °C for 1 h, and finally to 22 °C for 1-2
h. When thin layer chromatography (1:1 hexane/EtOAc, Rf ) 0.54
for 62, 0.63 for 71, 0.60 for 72) indicated complete consumption of
the starting material, the reaction solution was diluted with 10 mL of
CH2Cl2, and the solid was filtered and washed with an additional 10
mL of CH2Cl2. The filtrate was washed with 15 mL of aqueous
NaHCO3 solution, and the layers were separated. The aqueous layer
was washed with 3 × 30 mL portions of CH2Cl2. The combined
organic layers were dried over anhydrous MgSO4, filtered, and
concentrated in Vacuo to give a yellow oil. Purification by silica gel
chromatography (3:1 hexanes/EtOAc) afforded 66 mg (0.10 mmol, 92%
yield) of the desired compound as a colorless oil. IR (NaCl): 3288
(m), 2961 (m), 2935 (m), 2867 (m), 2363 (w), 2332 (w), 1744 (s),
Fluoro-2,4-di-O-acetyl-3-N-(trifluoroacetyl)-3,6-dideoxy-r-L-
talopyranose (71). To a benzene (8.9 mL) solution of acetoxy
glycoside 70 (230 mg, 0.62 mmol) was added thiophenol (76.6 mL,
0.75 mmol), followed by SnCl4 (51.1 mL, 0.44 mmol). The mixture
was placed in an oil bath preheated to 50 °C; reaction progress was
monitored carefully by thin layer chromatographic analysis (1:1 hexane/
EtOAc, Rf ) 0.50 for starting material, 0.60 for the product). After
45 min, the reaction was quenched by the addition of a 20 mL portion
of a saturated aqueous sodium bicarbonate solution. Organic and
aqueous layers were separated, and the aqueous layer was washed with
CH2Cl2 (three 20 mL portions). The combined organic layers were
1646 (m), 1539 (m), 1375 (m), 1369 (m), 1224 (s) cm-1 1H NMR
.
(400 MHz in CDCl3): δ 6.95 (bd, 1H, J ) 7.3 Hz, CH2CONH), 5.76
(dddd, 1H J ) 17.2, 10.3, 7.1, 6.2 Hz, CH2CHCH2), 5.62 (br, t, J )

10316 J. Am. Chem. Soc., Vol. 119, No. 43, 1997
Xu et al.
6.3 Hz, 1H, NHCOCF), 5.12 (m, 1H,C2′-H), 4.97 (d, 1H, J ) 17.3 Hz,
CH2CHCH2), 4.96 (br, s, 1H, C1′-H), 4.94 (m, 1H, CH2CHCHH), 4.82
(m, 1H, C4′-H), 4.71 (d, 2H, J ) 14.1 Hz, CH2CH2CHCH2), 4.60 (dt,
1H, J ) 7.9, 2.5 Hz, C3′-H), 4.20 (qd, 1H, J ) 6.6, 1.4 Hz, C5′-H),
3.65 (m, 1H, EtCHCHO), 3.25 (q, 2H, J ) 5.9 Hz, NCH2), 2.17 (s,
3H, COCH3), 2.16 (s, 3H, COCH3), 2.16-1.20 (15H, alkyl-H), 1.71
(s, 3H, CH3CCH2), 1.16 (d, 2H, J ) 6.6 Hz, C6′-H3), 0.90 (t, 3H, J )
7.3 Hz, CH2CH3), 0.87 (t, 3H, J ) 7.3 Hz, CH2CH3). 13C NMR (100
MHz in CDCl3): δ 175.4, 170.9, 170.6, 157.3 (q, JC-F ) 38 Hz), 145.0,
138,3, 116.1 (q, JC-F ) 287.0 Hz), 114.7, 110.5, 97.3, 81.5, 69.0, 65.4,
48.7, 46.6, 43.6, 39.7, 33.8, 31.66, 31.64, 28.4, 27.6, 27.0, 25.9, 23.1,
22.3, 20.9, 20.5, 16.3, 12.2, 12.0. HRMS Calcd for C30H47N2O8F3Na
(M + Na): 671.3495. Found: 671.3516.
product in 5 mL of acetyl chloride in 1 mL of EtOH solution, followed
by concentration in Vacuo to dryness to afford synthetic 1. 1H, 13C
NMR and IR spectra of synthetic and natural product salts proved
superimposable. For authentic free base product:61 IR (KBr): 3310
(m), 3307 (br, m), 2954 (m), 2932 (m), 1652 (s) cm-1 1H NMR (400
.
MHz in CD3OD): δ 3.94 (dq, 1H, J ) 6.1, 1.0 Hz, C5′-H), 3.65-3.55
(2H, EtCHCHO, NCHH), 3.54 (m, 1H, C2′-H), 3.46 (m, 1H, C4′-H),
2.95 (m, 1H, NCHH), 2.91 (br, t, 1H, J ) 3.0 Hz, C3′-H), 2.06 (m, 1H,
NCOCHEt), 1.80-1.00 (m, 21H, alkyl-H), 1.21 (s, 3H, J ) 6.6 Hz,
C5′-CH3), 0.92 (d, 3H, J ) 7.0 Hz, CH2CHCH3), 0.86 (t, 3H, J ) 7.2
Hz, CH2CH3), 0.84 (t, 3H, J ) 7.2 Hz, CH2CH3). For synthentic
hydrochloric salt product: 1H NMR (400 MHz in CD3OD): δ 4.02 (q,
1H, J ) 7.0 Hz, C5′-H), 3.70 (m, 1H, C2′-H), 3.64 (m, 1H, C4′-H),
3.62-3.50 (m, 2H, EtCHCHO, NCHH), 3.41 (t, 1H, 3.3 Hz, C3′-H),
2.97 (m, 1H, NHCHH), 2.09 (m, 1H, NCOCHEt), 1.80-1.00 (m, 21H,
alkyl-H), 1.14 (d, 3H, J ) 6.4 Hz, C5′-CH), 0.93 (d, 3H, J ) 7.0 Hz,
CH2CHCH3), 0.87 (t, 3H, J ) 7.3 Hz, CH2CH3), 0.86 (t, 3H, J ) 7.2
Hz, CH2CH3). 13C NMR (100 MHz in CD3OD): δ 179.2, 98.7, 78.8,
69,7, 68.7, 68.0, 51.3, 50.3, 42.1, 40.0, 35.5, 34.7, 32.5, 28.6, 27.7,
26.5, 26.4, 25.9, 22.3, 22.1, 21.1, 16.9, 12.7, 9.4. HRMS Calcd for
C24H46N2O5Na (M + Na): 443.3485. Found: 443.3478.
2,10-(R,R)-Diethyl-6-methyl-9-(R)-((2′,4′-di-O-acetyl,3′-N-(tri-
fluoroacetyl)-3′,6′-dideoxy-r-L-talopyranosyl)oxy)-13-trideca-5-
ene Lactam (73). Freshly prepared Mo catalyst 14 (5.3 mg, 7 × 10-3
mmol) was weighed in a glove box in a 10 mL pear shaped flask
containing 15.0 mg (0.023 mmol) of diene 72 dried by azeotropic
distillation with benzene (3 × 3 mL) . The reaction flask was then
equipped with a reflux condenser and removed from the glove box;
the reaction mixture was diluted with 1.9 mL of anhydrous benzene.
The resulting mixture was allowed to stir at 22 °C under an atmosphere
of argon for 4 h. At this time, the solvent was removed in Vacuo to
leave behind a dark brown oil. Purification by silica gel chromatog-
raphy (2:3 hexanes/EtOAc) afforded 13 mg of 73 (0.022 mmol, 91%
yield) as a white solid. IR (NaCl): 3314 (w), 2961 (m), 2930 (m),
Acknowledgment. The National Institutes of Health (GM-
47480) and the National Science Foundation (NSF-9632278)
supported the research. Professor Samuel J. Danishefsky, Dr.
V. P. Gullo, and Dr. V. R. Hegde of the Schering-Plough Co.
generously supplied us with samples of the natural product and
derivatives. We are grateful to Professor Danishefsky for words
of encouragement as well. We thank Professors Kendall Houk
and Philip DeShong for helpful discussions and Dr. Alexander
Muci for patiently teaching us how to set up an efficient MPLC
system. Our colleague, Professor David G. J. Young, suggested
the dipolar cycloaddition approach to the carbohydrate synthesis.
A.H.H. is an NSF National Young Investigator, a Sloan
Research Fellow, and a Camille Dreyfus Teacher-Scholar.
D.S.L. is a Department of Education GAANN Fellow and
G.E.H. was a 1995 Pfizer Undergraduate Summer Fellow.
Steven Scully and Sarri Salman provided experimental as-
sistance.
2880 (m), 2351 (w), 1746 (s), 1532 (m), 1174 (s) cm-1
.
1H NMR
(400 Hz in CDCl3): δ 6.76 (br, d, 1H, J ) 7.0 Hz, CHCONH), 5.47
(br, t, 1H, J ) 7.0 Hz, NHCOCF3), 5.14 (m, 1H, C2′-H), 5.07 (m, 1H,
CH2CHCMe), 5.04 (s, 1H, C1′-H), 4.75 (br, d, 1H, J ) 3.7 Hz, C4′-H),
4.62 (dt, 1H, J ) 8.2, 3.5 Hz, C3′-H), 4.20 (qd, 1H, J ) 6.6, 1.1 Hz,C5′-
H), 3.60 (br d, 1H, HNCHH), 3.57 (m, 1H, EtCHCHO), 3.14 (m, 1H,
CHHNH), 2.17 (s, 3H, OCOH3), 2.16 (s, 3H, OCOCH3), 2.00-1.20
(15H alkyl-H), 1.69 (s, 3H, CHCCH3), 1.16 (d, 3H, J ) 6.4 Hz, C6′-
H3), 0.89 (t, 3H, J ) 7.3 Hz, CH2CH3), 0.84 (t, 3H, J ) 7.5 Hz,
CH2CH3). 13C NMR (100 MHz in CDCl3): δ 175.8, 170.9, 170.3,
156.6 (q, JC-F ) 38.0 Hz), 135.7, 124.2, 115.5 (q, JC-F ) 268.7), 94.1,
77.7, 69.2, 65.3, 50.6, 46.2, 40.0, 38.7, 34.0, 27.8,27.3, 27.2, 26.5, 24.9,
24.6, 23.5, 20.9, 20.5, 16.0, 12.2, 8.8. HRMS Calcd for C30H47N2O8-
Na (M + Na): 643.3182. Found: 643.3186.
2,10-(R,R)-Diethyl-6-(S)-methyl-9-(R)-((2′,4′-di-O-acetyl,3′-N-(tri-
fluoroacetyl)-3′,6′-dideoxy-â-L-talopyranosyl)oxy)-13-tridecane Lac-
tam (Sch 38516; 1). Into a solution of 2.5 mg (4 × 10-3 mmol) of
hydrogenation product 73 in 0.7 mL of CH3OH, 3.9 mL of hydrazine
was added. After 24 h, 1H NMR analysis indicated complete
consumption of the starting material. The reaction solution was
concentrated to dryness to give 1.6 mg of product as a white solid
(96%). This compound was determined to be indentical to the natural
product by the comprarison with 1H NMR of natural product. For
further proof, the product was converted to the corresponding hydro-
chloric acid salt, which was accomplished simply by dissolving the
Supporting Information Available: Experimental proce-
dures and spectral and analytical data for select reagents and
reaction products (42 pages). See any current masthead page
for ordering and Internet access instructions.
JA972191K
(61) The sample obtained from Schering-Plough Co. was determined to
be the derived hydrochloric salt of the natural product. The free base Sch
38516 (neutral amine) was obtained through treatment of this material with
excess Et3N.