Please cite this article in press as: Hong et al., Photoinduced Skeletal Rearrangements Reveal Radical-Mediated Synthesis of Terpenoids, Chem
2019), https://doi.org/10.1016/j.chempr.2019.04.023
(
Figure 3. Protecting-Group-free Synthesis of Isodon Diterpenoids 3 and 4
ꢀ
ꢀ
ꢀ
ꢀ
Reagents and conditions are as follows: (a) Cp
2
ZrCl
2
, AlMe
3
, CH
2
Cl
2
, À15 C, 3 h, then I
2
, THF, À15 C, 2 h, 70%; (b) 14, 9-BBN, THF, 0 C to 23 C, 5 h, then
ꢀ
ꢀ
ꢀ
Pd(dppf)
2
Cl
2
(10 mol%), 3M Cs
2
CO
3
, THF, 23 C, 10 h, 84%; (c) DMSO, (COCl
2
)
2
, Et
3
N, CH
2
Cl
2
, À78 C to 23 C, 3.5 h, 91%; (d) 17 (30 mol %), Cu(OTf)
2
,
ꢀ
NaTFA, TFA, DME/MeCN (v/v = 1:1.5), 23 C, 9 h, 72%, 85% ee, 96% ee after one recrystallized from n-hexane; (e) MeI, t-BuOK, t-BuOH/THF (v/v = 1:2),
ꢀ
ꢀ
ꢀ
ꢀ
0
C to 23 C, 3 h, 82%, 85% ee, >99% ee after one recrystallized from n-hexane; (f) N
2
H
4
$H
2
O, diglycol, 160 C, 2 h, then KOH, diglycol, 180 C, 6 h, 90%; (g)
ꢀ
ꢀ
ꢀ
CrO
dr 5/1; (j) n-Bu
THF/t-BuOH/H
3
, AcOH/Ac
2
O (v/v = 3:1), 23 C, 7 h, 85%; (h) Na, NH
3
(liq.), EtOH, THF, À78 C, 5 h, then 3M HCl, 54%; (i) 20, NaHMDS, HMPA, THF, À78 C, 1 h, 90%,
ꢀ
ꢀ
3
SnH, AIBN, PhMe/t-BuOH (v/v = 1:3), MW, 80 W, 120 C, 10 min, then PPTS, CH
2
Cl
2
, 23 C, 6 h, 50%; (k) OsO
4
4
(20 mol%), DABCO, NaIO ,
ꢀ
ꢀ
ꢀ
2
O (v/v/v = 3:2.2:1), 23 C, 6 h, 94%; (l) Me
2
NCH
2
NMe
2
, Ac
2
O, DMF, MW, 80 W, 100 C, 25 min, 96%; (m) NaBH
4
, CeCl
3
, MeOH, À78 C, 1 h,
7
1%. THF, tetrahydrofuran; 9-BBN, 9-borabicyclo[3.3.1]nonane; DMSO, dimethyl sulfoxide; NaTFA, sodium trifluoroacetate; TFA, trifluoroacetic acid;
DME, 1,2-dimethoxyethane; NaHMDS, sodium bis(trimethylsilyl)amide; HMPA, hexamethylphosphoramide; AIBN, azobisisobutyronitrile; MW,
microwave; PPTS, pyridinium p-toluenesulfonate; DABCO, 1,4-diazabicyclo[2.2.2]octane; DMF, N,N-dimethylformamide; dr, diastereomeric ratio.
5
7–59
density functional theory (DFT) calculations,
is mainly controlled by geometry reason. The transition states from radical addition
to both the C and C11 positions are significantly distorted in comparison with the
which show that the regioselectivity
9
transition state from radical addition to the C
S24–S41 and Figures S107–112).
8
position (for more details, see Tables
Upon oxidative cleavage of the exo-methylene moiety in 22, diketone 23 (Figures
S51 and S52) was obtained in 94% yield. Next, a-methylenation of 23 delivered
(
+)-ent-kauradienone (3) (Tables S6–S8; Figures S1 and S2). Selective Luche reduc-
tion of 3 furnished 24 (Figures S53–S58), which is the key proposed precursor for
the photoinduced skeletal rearrangement. Upon treatment of 24 with UV irradiation
at 254 nm (Table S4), (À)-jungermannenone C (4) (Tables S9–S11; Figures S3 and S4)
was obtained in 58% isolated yield along with 28% of recovered 24. To investigate
the equilibrium nature of the photoinduced radical rearrangement, we irradiated
(
À)-4 at 365 nm (Table S5). Pleasingly, 24 was generated in 21% yield along with
7
1% of recovered (À)-4. Finally, to further elucidate the biosynthetic relationship be-
tween ent-kaurane-type and jungermannenone-type diterpenoids, we exposed
both 24 and (À)-4 to sunlight as neat compounds. Gratifyingly, late-stage photoin-
duced skeletal rearrangement occurred smoothly in 22%‒70% yield according to
1
H NMR analysis of the crude samples. Therefore, we achieved the total synthesis
of (+)-ent-kauradienone (3) and (À)-jungermannenone C (4) in 12 and 14 steps,
respectively, via late-stage photoinduced skeletal rearrangements of the bicyclo
[
3.2.1]octene ring system, and we thus established the absolute configurations of
2
1
3
2
and 4. In addition, from the interconversion between 4 and 24, we believe that
4 is most likely a natural product.
Late-Stage Photoinduced Skeletal Rearrangements of Terpenoids
We further validated the late-stage photoinduced skeletal rearrangements by sub-
jecting other ent-kaurane-type and jungermannenone-type diterpenoids or analogs
to photochemical irradiation. Of note, substrates (vide infra) used here were either
commercially available natural products or derivatives thereof (see Tables S14–
S23 and Figures S7–S18). As exemplified in Figure 4, under UV irradiation, ent-kaur-
6
0
ane-type diterpenoids 25 (Tables S14 and S15; Figures S7 and S8) and 26 (Tables
S16 and S17; Figures S9–S12) underwent photochemical rearrangement smoothly to
generate, respectively, jungermannenone-type 27 (Figures S59–S64) and 28 (Fig-
ures S65–S70) and vice versa. Although we did not observe strict correlation
between the UV absorption data of both ent-kaurane-type and jungermannenone-
type products with the emission spectrum of the lamp used (Figures S103–S106),
our product compositions varied drastically from different wavelengths of the
lamp as well as from the different solvents used during UV irradiation. The computed
energy difference of 24 and 4 (DG = 2.7 kcal/mol) also implies the non-existence of a
thermodynamic equilibrium (Tables S42 and S43; Figure S113).
6
Chem 5, 1–11, June 13, 2019