Synthesis of Cladiellin Diterpenes
J. Am. Chem. Soc., Vol. 123, No. 37, 2001 9043
the amount of 7 produced. Spectral data for 7, however, did
not match those reported for sclerophytin A.12 At this stage we
considered the possibility that sclerophytin A was the alcohol
epimer of 7 and consequently oxidized 7 to give ketone 93; the
minor endocyclic alkene isomer could be easily removed at this
point by flash chromatography on silica gel. Reduction of 93
with NaBH4 proceeded with high selectivity from the less-
hindered â-face to generate 94 in high yield. NMR data for
this product were again distinctly different from those reported
for sclerophytin A.12
The obvious differences between 7 and 94 and sclerophytin
A prompted us to look for a sample of natural sclerophytin A,
at which point we learned of independent investigations in the
Paquette group where 7, 93, and 94 had also been prepared.14
These workers had reinvestigated the natural isolate and
proposed that sclerophytin A was cladiell-11(17)-ene-3,6,7-triol
(8).16a Since this structure was simply the double bond regio-
isomer of cladiellin triol 1, we were able to quickly confirm
this conclusion by photochemically isomerizing 1 to 8 (eq 4).70
structural assignment for this diterpene be revised. A similar
conclusion was arrived at contemporaneously by the Paquette
group, who after reexamining a sample of natural sclerophytin
A proposed structure 8 for this natural product.16a We were able
to confirm the correctness of this proposal by photoisomerization
of the endocyclic double bond of 1 (whose structure had been
rigorously established by single-crystal X-ray analysis) to
complete a total synthesis of sclerophytin A (8).
The concise total syntheses detailed in this account further
highlight the power of pinacol-terminated cationic cyclizations
for assembling complex oxacyclic natural products. These
studies also established for the first time that relatively unstable
(Z)-R,â-unsaturated aldehydes could take part in Prins-pinacol
constructions without erosion of double bond stereochemistry.
Experimental Section71
Prins-Pinacol Condensation Using BF3‚OEt2. Preparation of
(1R,3R,3aS,7R,7aR)-7-Isopropyl-4-methyl-1-[(E)-1-methyl-3-(triiso-
propylsilyloxy)propenyl]-3-[3-(trimethylsilyl)prop-2-ynyl]-1,6,7,7a-
tetrahydroisobenzofuran-3a-carbaldehyde (46). A mixture of dienyl
diol 44 (350 mg, 1.1 mmol), (E)-enal 45 (2.9 g, 11 mmol), MgSO4
(200 mg), and CH2Cl2 (42 mL) was cooled to -55 °C and treated
dropwise with BF3‚Et2O (0.15 mL, 1.2 mmol). The mixture was stirred
at -55 °C for 3 h, allowed to warm to -20 °C, stirred for 15 min, and
added to saturated aqueous NH4Cl (40 mL). The layers were separated,
the aqueous layer was extracted with CH2Cl2 (3 × 40 mL), and the
combined organic extracts were dried (Na2SO4), filtered, and concen-
trated. Excess aldehyde 45 was removed by bulb-to-bulb distillation
(150 °C, 1.0 mm), and the residue was purified by flash chromatography
on silica gel (19:1 hexanes-ethyl acetate) to provide 0.45 g (79%) of
46 as a clear colorless oil: [R]25D +23.4 (c 1.0, CHCl3); 1H NMR (500
MHz, CDCl3) δ 9.82 (s, 1 H), 5.70 (s, 1 H), 5.63 (bt, J ) 5.2 Hz, 1 H),
4.24 (d, J ) 4.8 Hz, 2 H), 4.07 (t, J ) 6.6 Hz, 1 H), 4.01 (d, J ) 9.9
Hz, 1 H), 2.73 (dd, J ) 9.9, 4.0 Hz, 1 H), 2.69 (d, J ) 6.6 Hz, 1 H),
2.12-1.92 (m, 3 H), 1.93 (s, 3 H), 1.66 (s, 3 H), 1.37 (m, 1 H), 1.14-
0.82 (m, 22 H), 0.79 (app d, J ) 7.6 Hz, 6 H), 0.14 (s, 9 H); 13C NMR
(125 MHz, CDCl3) δ 201.2, 133.5, 132.2, 131.3, 126.4, 103.4, 89.1,
88.3, 83.8, 62.2, 60.9, 46.0, 38.8, 28.3, 24.4, 23.7, 21.7, 21.2, 21.0,
Conclusions
A versatile and concise strategy for the total synthesis of
cladiellin diterpenes has been developed (Scheme 2). The
defining step in this approach is an efficient pinacol-terminated
Prins cyclization23-26 that assembles the hexahydroisobenzo-
furan core and five of the six invariant stereocenters of these
marine diterpenes with complete stereocontrol from two simple
precursors: enantiopure cyclohexadienyl diol 44 and an R,â-
unsaturated aldehyde. The final oxacyclononane ring of these
natural products is formed by a highly diastereoselective
intramolecular Nozaki-Hiyama-Kishi reaction.30 The viability
of this total synthesis strategy was first verified by a total
synthesis of 6-acetoxycladiell-7(16),11-dien-3-ol (deacetoxyal-
cyonin acetate, 6) which, when originally disclosed in 1995,18
constituted the first total synthesis of a 2,11-cyclized cembranoid
oxacyclic diterpene. This inaugural total synthesis was achieved
in 20 steps and 4.3% overall yield from (S)-(+)-carvone and
21 steps and 3% yield from (S)-glycidyl pivalate. A second-
generation total synthesis of 6 in which the C3 stereocenter was
established using substrate, rather than catalyst, control was
one step shorter and proceeded in nearly identical overall yield.
The first total synthesis of cladiell-11-ene-3,6,7-triol (1) was
realized in similar overall efficiency from advanced tricyclic
intermediate 57.
18.6, 12.6, 11.7, 0.4; IR (film) 2953, 2178, 1718, 1462, 1114 cm-1
;
HRMS (CI) m/z 544.3775 (M, 544.3768 calcd for C32H56O3Si2).
Prins-Pinacol Condensation Using SnCl4. Preparation of (1R,3R,-
3aS,7R,7aR)-7-Isopropyl-4-methyl-1-[(E)-1-methyl-4-(triisopropyl-
silyloxy)but-1-enyl]-3-[3-(trimethylsilyl)prop-2-ynyl]-1,6,7,7a-tet-
rahydroisobenzofuran-3a-carbaldehyde (70). A mixture of cyclo-
hexadienyl diol 44 (0.82 g, 2.7 mmol), (E)-enal 69 (7.2 g, 27 mmol)
and MeNO2 (27 mL) was cooled to -29 °C and then treated dropwise
with SnCl4 (0.16 mL, 1.3 mmol), and the solution was maintained at
-29 °C for 12 h. The reaction mixture then was added to saturated
aqueous NH4Cl (100 mL), the aqueous layer was extracted with hex-
anes (3 × 100 mL), and the combined organic extracts were dried
(Na2SO4), filtered, and concentrated. Excess 69 was removed by bulb-
to-bulb distillation (150 °C, 0.10 mm), and the remaining residue was
purified by medium-pressure liquid chromatography (MPLC) (Lobar
pre-packed column, LiChroprep Si 60 silica gel; 19:1 hexanes-ethyl
acetate) to provide 1.0 g (70%) of 70 as a clear pale yellow oil: [R]23
D
+32.9 (c 1.0, CHCl3); 1H NMR (500 MHz, C6D6) δ 9.77 (s, 1 H), 5.55
(t, J ) 6.6 Hz, 1 H), 5.45 (br s, 1 H), 4.02-3.98 (m, 2 H), 3.62-3.59
(m, 2 H), 2.83 (dd, J ) 9.6, 4.2 Hz, 1 H), 2.73 (d, J ) 6.4 Hz, 2 H),
2.28 (q, J ) 6.8 Hz, 2 H), 1.85-1.81 (m, 5 H), 1.74 (s, 3 H), 1.53-
1.49 (m, 1 H), 1.13-1.02 (m, 22 H), 0.84 (d, J ) 6.6 Hz, 3 H), 0.75
(d, J ) 6.6 Hz, 3 H), 0.17 (s, 9 H); 13C NMR (125 MHz, C6D6) δ
199.6, 135.2, 131.4, 127.1, 125.8, 104.0, 88.5, 87.8, 83.7, 63.1, 62.0,
45.8, 38.9, 32.0, 27.9, 24.2, 23.4, 21.3, 20.7, 20.1, 18.3, 12.3, 11.2,
-0.1; IR (neat) 2958, 2866, 2726, 2179, 1716, 1462 cm-1; HRMS
(FAB) m/z 557.3864 (M - H, 557.3846 calcd for C33H57O3Si2). Anal.
Calcd for C33H58O3Si2: C, 70.91; H, 10.46. Found: C, 70.85; H, 10.42.
As a prelude to future efforts to synthesize members of the
more complex briarellin and asbestinin diterpene families, we
developed a variant of this strategy where the R,â-unsaturated
aldehyde component of the pivotal Prins-pinacol reaction had
the Z, rather than the more stable E, configuration (Figure 5).
This approach was employed to prepare the most likely
stereoisomer of the purported structure12 of sclerophytin A,
tetracyclic ether 7. Neither 7 nor its secondary alcohol epimer
94 was identical with sclerophytin A, thus requiring that the
(70) This conversion was carried out only once on a small scale. It is
likely that the efficiency of this conversion could be optimized to be similar
to that realized in the closely related photoisomerization of 92 f 7 (64%
yield).
(71) General experimental details have been described: Minor, K. P.;
Overman, L. E. J. Org. Chem. 1997, 62, 6379-6387. For standard
abbreviations employed in this article, see: J. Org. Chem. 2001, 66, 24A.