Synthesis of a C(1)-C(14)-containing fragment of callipeltoside A.

Article abstract of DOI:10.1021/ol9906514

[formula: see text] A C(1)-C(14)-containing fragment of callipeltoside A (1, Scheme 1) was synthesized efficiently via a dianion aldol coupling reaction between aldehyde 2 and ketoester 3. A surprising lack of reactivity between the alkenes in 13 and the

Full text of DOI:10.1021/ol9906514

ORGANIC
LETTERS
1999
Vol. 1, No. 1
169-171
Synthesis of a C(1)−C(14)-Containing
Fragment of Callipeltoside A
Thomas R. Hoye* and Hongyu Zhao
Department of Chemistry, UniVersity of Minnesota, Minneapolis, Minnesota 55455
hoye@chem.umn.edu
Received May 10, 1999
ABSTRACT
A C(1)−C(14)-containing fragment of callipeltoside A (1, Scheme 1) was synthesized efficiently via a dianion aldol coupling reaction between
aldehyde 2 and ketoester 3. A surprising lack of reactivity between the alkenes in 13 and the Grubbs initiator 15 was encountered. An equally
surprising rate acceleration of the reaction between 15 and allylic alcohols (alk-1-en-3-ols) as well as their subsequent cleavage to methyl
ketones was discovered. In situ 1H NMR analysis has proven to be a very useful tool for monitoring RCM reactions of complex substrates
such as 13.
Callipeltoside A (1) was isolated from the shallow water
sponge, Callipelta sp., in 1996 and was found to inhibit in
Scheme 1
vitro proliferation of KB and P388 cells and to protect cells
infected with HIV virus.1 A limited amount (3.5 mg) of
callipeltoside A was obtained and its relative configuration
was deduced by extensive NMR experiments. To determine
the absolute configuration of callipeltoside A (1), to confirm
the assigned relative configurations, and to provide material
for further biological studies, we have undertaken its total
synthesis. We report here our progress toward construction
of the central macrolactone portion of 1. Our efforts have
resulted in the synthesis of a C(1)-C(14)-containing frag-
ment having the assigned relative configuration. In one
retrosynthetic plan (Scheme 1), we imagined the macrolac-
tone in 1 to arise from precursors 2 and 3 by a dianion aldol-
coupling and ring-closing metathesis strategy. While we have
not yet successfully closed the macrocyclic ring, we have
gained (surprising) insight into the reactivity of possible ring-
closing metathesis substrates.
The synthesis of aldehyde 2 is outlined in Scheme 2.
Exchange of the oxazolidinone auxiliary in the known syn-
aldol adduct 42 with N,O-dimethylhydroxylamine generated
the Weinreb amide.3,4 Protection of the secondary hydroxyl
group as its trimethylsilyl ether followed by hydroboration/
oxidation gave the primary alcohol 5 as essentially one
diastereomer.2 Protection of the new hydroxyl group as its
p-methoxybenzyl ether 6 was followed by vinyllithium
addition to give the terminal vinyl ketone.5,6 PPTS/MeOH
removal of the TMS ether quantitatively produced the
â-hydroxyenone 7. The 1,3-anti-diol 8 was then obtained as
(1) Zampella, A.; D’Auria, V.; Minale, L.; Debitus, C.; Roussakis, C. J.
Am. Chem. Soc. 1996, 118, 11085-8.
(2) Evans, D. A.; Fitch, D. M. J. Org. Chem. 1997, 62, 454-5.
10.1021/ol9906514 CCC: $18.00 © 1999 American Chemical Society
Published on Web 05/26/1999

trapping by the C(7)-OH, gave a cyclic p-methoxyben-
zylidene acetal; (ii) methylation of the C(9)-OH; and (iii)
reductive opening of the acetal with DIBAL gave the primary
alcohol 9 [a 10% yield of the regioisomeric primary PMB
ether (i.e., the C(9)-methyl ether of 8) was also formed].10
Swern oxidation of 9 completed the synthesis of fragment
2.
Scheme 2
Preparation of the â-keto ester subunit 3 (Scheme 3) was
achieved by copper-catalyzed addition of isopropenylmag-
nesium bromide to the TBS ether of (R)-glycidol (10). The
resulting homoallylic alcohol 11 was converted to its
acetoacetate derivative 3 by heating with dioxinone 12.11
Scheme 3
the major isomer (9:1; anti:syn) by a hydroxyl-directed
reduction with triacetoxyhydridoborate.7 The configuration
of the new center was determined by both acetonide8a and
Mosher ester7b analyses. Migration of the PMB ether from
the terminal C(5)-position to the more hindered C(7)-
hydroxyl group and regioselective methyl ether formation
was achieved by a three-step operation: (i) DDQ oxidation
under anhydrous conditions,9 accompanied by intramolecular
Subunits 2 and 3 were coupled via aldol reaction (Scheme
4). An excess of the dianion derived from the â-keto ester 3
with NaH/BuLi was added to aldehyde 2 at -78 °C to give
a readily separable diastereomeric mixture of epimeric C(5)-
alcohols 13 (2.5:1; 70% yield). To assign the configuration
of the new stereocenter, the major diastereomer was oxida-
tively cyclized in a remarkably clean reaction to give the
p-methoxybenzylidene acetal 14.12 The coupling constant
between H(5) and H(6) in this derivative was 5.2 Hz, a value
consistent with the C(5)/C7)-anti rather than -syn acetal.13
Attempts to form the macrocyclic lactone by ring-closing
metathesis14 of 13 have been unsuccessful. We routinely use
(3) Consistent with the report that the â-hydroxyl group is important for
clean installation of the N,O-dimethylhydroxylamine moiety (Evans, D. A.;
Gage, J. K.; Leighton, J. L. J. Am. Chem. Soc. 1992, 114, 9434-53), we
observed that if the C(7)-hydroxyl group was first protected as its TMS
ether prior to reaction with Me2AlN(Me)OMe, the main product was i. An
analogous byproduct was recently described (Romo, D.; Rzasa, R. M.; Shea,
H. A.; Park, K.; Langenhan, J. M.; Sun, L.; Akheizer, A.; Liu, J. O. J. Am.
Chem. Soc. 1998, 120, 12237-54).
(4) New compounds 2-13 and the additional isolable Scheme 2
1
intermediates gave satisfactory H and 13C NMR, IR, and MS data.
(5) Linderman, R. J.; Cutshall, N. S.; Becicka, B. T. Tetrahedron Lett.
1994, 35, 6639-42.
(6) Initial attempts to add vinylmagnesium bromide (Guanti, G.; Banfi,
L.; Riva, R. Tetrahedron. 1995, 51, 10343-60) rather than CH2dCHLi to
6 gave a complex mixture of products, which included the aldehyde ii as
the major product (∼20%). This reduction might proceed by way of complex
iii between a molecule of 6 and the magnesium salt of allyl alcohol.
Fragmentation of a Weinreb amide to formaldehyde has been recently
reported (Seong, M. R.; Kim, J. N.; Kim, H. R.; Ryu, E. K. Synth. Commun.
1998, 28, 139-45).
Scheme 4
(7) Evans, D. A.; Chapman, K. T.; Carreira, E. M. J. Am. Chem. Soc.
1988, 110, 3560-78.
170
Org. Lett., Vol. 1, No. 1, 1999

1
in situ monitoring of RCM reactions by H NMR spectros-
reaction of either alkene in 13 with Cl2(Cy3P)2RudCHPh
(15), was detected.15
copy to follow their progress. In the case of substrate 13-
maj or 13-min no styrene, the obligatory product from
In a different study, we found that a terminal vinyl group
with a free tertiary allylic hydroxyl group was much more
reactive toward the Grubbs initiator than its methyl ether
analogue.16 We therefore hypothesized that the allylic alcohol
16 (Scheme 5), the C(9)-OH analogue of 13, would be more
(8) (a) Diastereomeric acetonides were prepared from the major and
minor diastereomeric diols 8 and C(9)-epi-8. Rychnovsky/Evans 13C NMR
analysis clearly indicated the 1,3-anti vs 1,3-syn nature of each: Rych-
novsky, S. D.; Skalitzky, D. J. Tetrahedron Lett. 1990, 31, 945-8. Evans,
D. A.; Rieger, D. L.; Gage, J. R. Tetrahedron Lett. 1990, 31, 7099-100.
Rychnovsky, S. D.; Rogers, B.; Yang, G. J. Org. Chem. 1993, 58, 3511-
15. (b) Mosher ester analysis (Dale, J. A.; Mosher, H. A. J. Am. Chem.
Soc. 1973, 95, 512-9) clearly supports the assignment of C(9)-R vs C(9)-S
configuration in the major and minor diastereomers of 8.
Scheme 5
(9) (a) Oikawa, Y.; Yoshioka, T.; Yonemitsu, O. Tetrahedron Lett. 1982,
23, 889-92. (b) Wang, Z. Tetrahedron Lett. 1989, 30, 6611-4.
(10) Initially, we tried a more direct sequence to prepare 9, but attempts
to protect the hindered hydroxyl in 4 with PMB were unsuccessful. Under
acidic conditions (PMB-trichloromethylimidate/BF3‚OEt2) we isolated only
the Friedel-Crafts product iv (in 50% yield) from the mixture of products.
Under basic conditions (NaH, PMB-Cl, THF, 0 °C) retroaldol fragmenta-
tion predominated.
(11) Clemens, R. J.; Hyatt, J. A. J. Org. Chem. 1985, 50, 2431-5.
(12) This cyclization was performed using 1 mg of pure 13-maj. A
second diastereomer of 14, which we presume to be its anomeric epimer,
was also observed (∼1:10 ratio).
(13) The reported value for the vicinal (H-H) coupling constant in a
relevant 1,3-syn-acetal is much larger (J ) 9.2 Hz; Evans, D. A.; Gauchet-
Prunet, J. A. J. Org. Chem. 1993, 58, 2446-53) than that reported for a
relevant 1,3-anti-acetal (J ) 5.4 Hz; Roush, W. R.; Bannister, T. D.;
Ermolendo, M. S.; Yashunsky, D. V.; Borodkin, V. S. Tetrahedron Lett.
1992, 33, 3587-90).
(14) (a) Xu, Z.; Johannes, C. W.; Houri, A. F.; La, D. S.; Cogan, D. A.;
Hofilena, G. E.; Hoveyda, A. H. J. Am. Chem. Soc. 1997, 119, 10302-16
and earlier refs therein. (b) For a recent review that includes known examples
of macrocyclization reactions, see: Grubbs, R. H.; Chang, S. Tetrahedron
1998, 54, 4413-50.
(15) Using the Schrock initiator {[(CF3)2(Me)CO]2}ArNdModCH-
C(Me)2Ph, we observed formation of H2CdCHC(Me)2Ph at room temper-
ature in C6D6 by NMR spectroscopy, which implies initial loading of the
metal onto one of the olefins in 13. However, the reaction is very slow and
the overall conversion is low. We were unable to determine the fate of the
resulting Mo-alkylidene.
(16) Linalool rapidly underwent RCM with 15 under conditions where
linalool methyl ether was unreactive. We are continuing to study this
phenomenon and report additional details elsewhere.
(17) Prepared from 8 by the sequence: (i) DDQ; (ii) TESOTf; (iii)
DIBALH; (iv) Swern; (v) addition of dianion of 3; and (vi) HF/MeOH.
(18) A major advantage of in situ NMR monitoring is that we can quickly
assess with confidence whether a given substrate will engage in initial
reaction with a given metal-carbene initiator (by the presence or absence
of styrene and/or new MdCHR resonances). This in turn has permitted us
to prepare only ∼1 mg of numerous complex substrates and confidently
judge their reactivity.
(19) Alkenol 8 (2-3 mg) was exposed to ∼50 mol % of 15 in CDCl3 at
room temperature. Ketone 17 was the major product observed by direct
NMR and GC/MS analysis of the reaction mixture and the only product
(along with an approximately equal amount of starting 8) isolated following
HPLC purification.
reactive than 13 itself. Indeed, by monitoring the reaction
between 15 and 1617 by 1H NMR spectroscopy, we observed
rapid disappearance of the terminal vinyl group and a
comparable rate of appearance of styrene.18 However, the
desired RCM reaction did not occur; resonances for the
methylene protons at C(11′) remained intact throughout. This
prompted us to examine the reaction of a simpler secondary
alk-1-en-3-ol (i.e., 8, Scheme 5) with 15. We observed rapid
formation of the methyl ketone 17.19 Intermediates 18-20
can account for this stoichiometric cleavage reaction.20 These
observations suggest that secondary allylic alcohols represent
a liability in RCM reactions when the desired transformation
of intermediates such as 18 is slow.
In conclusion substrate 13, which bears an allylic methyl
ether [at C(9)] adjacent to its terminal vinyl group, is
unreactive with either the ruthenium or molybdenum carbene
complexes we have examined. However, since substrates
containing secondary allylic ethers are known to engage in
metathesis reactions,21 we suspect that the additional remote
branching in 13 further reduces its reaction rate. Strategies
to circumvent this problem are now under study.
(20) We speculate that an event similar to the 8 to 17 transformation
converted 16 (∼1 mg)18 to a methyl ketone, perhaps in the form of its
hemiketal to the C(5)-OH.
(21) For example, see: Kirkland, T. A.; Grubbs, R. H. J. Org. Chem.
1997, 62, 7310-7318.
Acknowledgment. We thank the National Institutes of
Health (CA-76497) for funding this project.
OL9906514
Org. Lett., Vol. 1, No. 1, 1999
171