Communication
of R5 does not have any diastereoselective effect on
the spirocyclization of ynamides.
Table 2. Substrate scope for tetracyclic pyrrolidinoindolines 9.[a]
To achieve an asymmetric synthesis, we then tested
the gold-catalyzed annulation in the presence of chiral
ligands.[19] We expected that when an Au-catalyst bear-
ing chiral ligands approached the ynamide, the Au-as-
sociated keteniminium species formed could induce
the asymmetric nucleophilic addition of indole,[20] thus
affording chiral spirocyclic pyrrolidinoindoline. To ex-
plore this possibility, several commercially available li-
gands were tested (see the Supporting Information),
and we found that when ynamide 6 was annulated
under the optimized conditions in the presence of
chiral ligand (S)-L*, product 7 was obtained in 85%
yield with 60% ee (Scheme 2).
9a: R1 =Bn, 89%
9b: R1 =PMB,
86%
9d: R2 =Ns, 88%
9e: R2 =Ms, 90%
9c: R1 =Allyl,
74%
9 f: R3 =5-OMe,
9j: R3 =5-F, 72%
85%
To demonstrate the potential of our pyrrolidinoindo-
lines in the synthesis of complex molecules, we then
selected compound 9b as a model, and transformed it
into the useful synthetic intermediate 11,[21] which
could potentially be applied to the total synthesis of
spirocyclic pyrrolidinoindoline-based natural products,
such as those shown in Figure 1. In the event, com-
pound 9b was easily reduced with Na(BH3)CN, and the
resultant allylic alcohol 10 was then oxidized to alde-
hyde 11 through a Dess–Martin periodinane (DMP)-oxi-
dation in 71% overall yield over two steps (Scheme 3).
To study the substrate scope of the ynamide annula-
tion, we also interestingly found that the ynamides 12
bearing secondary alcohols could be directly converted
into their corresponding tricyclic products 13 in moder-
ate-to-good yields when the reactions were carried out
in the presence of additional BF3·Et2O (Table 3). It is
worthwhile to mention that when ynamide 12 f with
a substituent at the C2 position of the indole was used,
product 13 f was obtained as a single diastereomer.
Scheme 4 highlights our proposed reaction mecha-
nism to account for the observed Au-catalyzed intra-
molecular tandem cyclization of indole-ynamides and
9g: R3 =5-Me,
82%
9k: R3 =5-Cl, 75%
9l: R3 =5-Br, 70%
9m: R3 =5-CN, 52%
9h: R3 =6-Me,
77%
9i: R3 =7-OMe,
76%
9n: R3 =5-CO2 Me, 63%
9o: R3 =4-CO2 Me, 56%
9p: R3 =5-CF3, 68%
9q: R4 =Me, X=H, X-ray structure of 9s and 9s’:[b] R5 =Me, 69%, d.r.=1:1
89%
9q
9r: R4 =Me, X=F,
78%
[a] The yields of the products 9a–9q, which are isolated as pairs of enantiomers.
For products 9m, 9n, 9o and 9s, the reaction time was 30 min and for other sub-
strates, the reaction was completed within 5 min. [b] Substrate 8s was used as its
optical pure form.
nulated smoothly to afford the expected products 9q and 9r
in good yields, and again the electron-rich substrate 8q gave
better results than the electron-deficient substrate 8r; (5) in
contrast to a previous report,[18] when enantiomerically pure
substrate 8s underwent annulation under the typical reaction
conditions, products 9s and 9s’ were obtained as a pair of dia-
stereomers in a 1:1 ratio, indicating that the stereogenic center
1,5-hydride shift for the formation of the a,b-unsaturated car-
bonyl compounds. Accordingly, after coordination of gold cata-
lyst to ynamide A, the resultant keteniminium B might under-
go 5-endo-trig cyclization[22] to form intermediate C (through
an anti-addition format). Thus, if the reaction proceeded with-
out an additional Lewis acid, the intermediate C would under-
go a direct 6-exo-trig nucleophilic addition to afford a tetracy-
clic product D, and the regenerated Au-catalyst could enter
into another catalytic cycle. On the other hand, if the reaction
proceeded with an additional Lewis acid, such as BF3·Et2O,
product D would coordinate with the Lewis acid to form inter-
mediate E through a hemi-aminol ring opening reaction,
which in turn could undergo a 1,5-hydride shift to afford prod-
uct F in a diastereoselective manner.[23] It is noteworthy to
mention that other potentially competing reactions, such as
the Wagner–Meerwein rearrangement, was not observed in
this reaction.[9b,c]
To further support our proposed reaction mechanism, two
additional experiments were conducted. Our first experiment
Scheme 2. Application of the chiral dinuclear gold catalyst.
[a] Isolated yield. [b] Determined by HPLC analysis with a chiral column.
Chem. Asian J. 2016, 11, 371 – 375
373
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim