Expedient synthetic approaches to the highly functionalized polycyclic alkaloids communesin F and perophoramidine are described using a unified approach featuring a key decarboxylative allylic alkylation to access a crucial and highly congested 3,3-disubstituted oxindole. Described are two distinct, stereoselective alkylations that produce structures in divergent diastereomeric series possessing the critical vicinal all-carbon quaternary centers needed for each synthesis. Synthetic studies toward these challenging core structures have revealed a number of unanticipated modes of reactivity inherent to these complex alkaloid scaffolds. Additionally, several novel and interesting intermediates en route to the target natural products, such as an intriguing propellane hexacyclic oxindole encountered in the communesin F sequence, are disclosed. Indeed, such unanticipated structures may prove to be convenient strategic intermediates in future syntheses.
Expedient synthetic approaches to the highly functionalized polycyclicalkaloidscommunesin F and perophoramidine are described using a unified approach featuring a key decarboxylative allylic alkylation to access a crucial and highly congested 3,3-disubstituted oxindole. Described are two distinct, stereoselective alkylations that produce structures in divergent diastereomeric series possessing the critical vicinal all-carbon quaternary centers needed for each synthesis. Synthetic studies toward these challenging core structures have revealed a number of unanticipated modes of reactivity inherent to these complex alkaloid scaffolds. Additionally, several novel and interesting intermediates en route to the target natural products, such as an intriguing propellane hexacyclic oxindole encountered in the communesin F sequence, are disclosed. Indeed, such unanticipated structures may prove to be convenient strategic intermediates in future syntheses.
In 1993, communesin
A (1a) was isolated along with
communesin B (1b) from a strain of Penicillium sp. found growing on a marine alga by the Numata group (Figure 1).[1] Communesins A (1a) and B (1b) exhibit antiproliferative activity
against P-388 lymphocytic leukemiacells (ED50 = 3.5 μg/mL
and 0.45 μg/mL, respectively).[1] In
addition, communesin B (1b) disrupts actin microfilaments
in cultured mammaliancells and shows cytotoxic activity against LoVo
and KB cells (MIC values of 2.0 μg/mL and 4.5 μg/mL, respectively).[2] Several other members of the comunesin family,
communesins B–H (1b–h), were
disclosed from related marine fungal strains of Penicillium sp. in the following years.[3] With the
exception of communesins G (1g) and H (1h), the communesins show insecticidal activity and antiproliferative
activity against a variety of cancercells, with communesin B (1b) being the most potent.[1−3] These indole alkaloidscontain several interesting structural features including vicinal
all-carbon quaternary centers, bis-aminal functionalities, and a complex
polycycliccore. The communesins are structurally unique when compared
against other known microfilament-disrupting agents, which are primarily
macrolides. Macrolide microfilament-disrupting agents show considerable
structural similarity, and their interactions with actin have been
crystallographically characterized, leading to hypotheses regarding
their mechanism of action.[4] The unique
structure of communesin B (1b) suggests that it may exhibit
a novel mechanism of action on the cytoskeleton relative to other
microfilament-disrupting agents.[5a] The
development of a unified synthetic route to the communesins would
enable the understanding of their effects on the cellular cytoskeleton
while addressing the scarcity of naturally occurring sources of the
compounds.
Figure 1
Communesins (1), nomofungin (2), and
perophoramidine (3).
Communesins (1), nomofungin (2), and
perophoramidine (3).In 2001, an intriguing natural product, nomofungin (2) was isolated from an unidentified fungus found on the bark
of Ficus microcarpa by the Hemscheidt group.[2] Interestingly, the only structural difference
between communesin
B (1b) and nomofungin (2) is that communesin
B has an aminal moiety instead of the N,O-acetal
moiety present in nomofungin. A combination of experimental and theoretical
exercises led to the independent discovery by our laboratory and the
Funk group that the reported structure of nomofungin was incorrect
and that it is actually that of communesin B.[5a,5b] Although the structure of nomofungin was erroneously assigned, its
isolation and structural revision to that of an older structure can
be viewed as the inception point for all synthetic efforts to the
communesin family members over the past decade. Interestingly, there
were no reports of synthetic efforts toward the communesins from 1993
up to our initial report in 2003.[5i]A structurally related compound, perophoramidine (3)
was isolated in 2002 from the ascidian Perophora namei.[6] The core is comparable to the one found
in the communesins, albeit in a higher oxidation state, with the alternate
diastereomeric relationship between the vicinal quaternary carbons
and without the azepine ring system. Perophoramidine (3) possesses modest cytotoxicity against the HCT 116 humancolon carcinomacell line (IC50 = 60 μM) and induces apoptosis.[7]These complex, polycyclic, bioactive alkaloids
have been the subject
of intense synthetic efforts over the past decade.[5] Numerous approaches have been reported in the literature,
including three from our laboratory.[5a,5c,5w] Herein, we report the evolution of an efficient,
unified approach toward the synthesis of these unique alkaloids.
Results
and Discussion
Biosynthesis-Inspired Diels–Alder
Cycloaddition
Strategy to Communesin F
Our early efforts toward the communesin
structure centered on the laboratory implementation of our proposed
biosynthesis (Scheme 1).[5a,5c,8] As the key step in the process, we envisioned
a Diels–Alder cycloaddition to unite the two indole-based fragments
by coupling of 5, an N-methylated derivative
of the ergot alkaloidaurantioclavine (4),[9,10] and an o-azaxylylene indolone 6 to
generate the bridged lactam 7. We anticipated that lactam 7 would be highly reactive due to the poor alignment of the
nitrogen lone pair with the carbonyl.[11−13] As such, the pendant
amino group would be expected to easily open the lactam, thus forming
spirocycle 8. Further tailoring would produce communesin
A (1a) and B (1b).
Scheme 1
Biosynthesis-Inspired
Approach
Toward this end, (±)-aurantioclavine
was prepared using known
methods,[5a,14] and an enantioselective synthesis of (−)-aurantioclavine
utilizing our oxidative kinetic resolution (OKR) technology was developed.[15] We proceeded to develop an efficient cycloaddition
between (±)-indole 9 as a model coupling partner
and benzyl chloride 10 using conditions previously developed
by Steinhagen and Corey[16] that resulted
in a mixture of pentacyclic diastereomers (89% yield). Removal of
the tosyl group with magnesium in methanol produced a 2:1 mixture
of diastereomers 11 and 12 in 80% combined
yield, with the desired relative stereochemistry evident in the major
diastereomer (cf. 11 and 1a) (Scheme 2).[5a]
Scheme 2
Model Studies for
a Diels–Alder Cycloaddtion Strategy To Construct
the Pentacyclic Core Structure
Despite the success of this model system, more advanced
electrophiles
(e.g., mesylate 14, cyclopropane 16, or
epoxide 17(17)) did not succumb
to cycloaddition conditions (Scheme 3). Nor
have we been successful in the oxidation of 11 and 12 at C(8), which would provide a functional handle for introduction
of the second quaternary stereocenter.
Scheme 3
Attempted Diels–Alder
Cycloadditions with Advanced Electrophiles
To obviate the difficulties encountered in our attempts
to functionalize
C(8), we next considered dienes possessing a functional handle at
C(8) that could unite diene and dienophile such as benzisoxazole 19, thereby enabling an intramolecular Diels–Alder
cycloaddition (Scheme 4). Thus, when coupled
to aurantioclavine 4, benzisoxazole 20 would
offer a stable o-methide imine that could react with
the indole moiety of compound 19 in a controlled and
intramolecular manner.
Scheme 4
Retrosynthetic Analysis of Communesin F
by an Intramolecular Diels–Alder
Cycloaddition
Fischer esterification
of commercially available carboxylic acid 21 followed
by heating in neat sulfuric acid provided the
benzisoxazole acid 20 in 44% yield over two steps (Scheme 5).[18] Treatment of benzisoxazole
acid 20 with oxalyl chloride provided the corresponding
acid chloride, which was smoothly coupled with aurantioclavine 4 to furnish carboxamide 22 (91% yield, two steps).
Similarly, 1-methylaurantioclavine 5 reacted with the
acid chloride to afford carboxamide 19 (77% yield, two
steps). Substrates 22 and 19 were subjected
to an intramolecular Diels–Alder cycloaddition under acidicconditions.[19] Unfortunately, the benzisoxazole
reacted with the butenyl side chain of the aurantioclavinecore to
generate the bridged polycycles 23 and 24. Nuclear Overhauser effect NMR spectroscopy (NOESY) studies and
X-ray analysis (Figure 2) demonstrated the
relative stereochemistry shown for 24 and that of 23 was assigned by analogy.
Scheme 5
Intramolecular Diels–Alder
Cycloaddition
Figure 2
X-ray structure of bridged
polycycle 24.
X-ray structure of bridged
polycycle 24.At this point, we turned our attention to synthesizing 3-bromooxindole 26, which would be a precursor to an o-methide
imine such as reactive intermediate 6, allowing for the
construction of the communesincore according to our original biosynthesis-inspired
model (Scheme 1). Aurantioclavine derivative 25 was reacted with bromooxindole 26 in an effort
to produce adduct 27 (Scheme 6a). Interestingly, different reactivity was observed in coordinating
and noncoordinating solvents. In THF or acetonitrile, the reaction
afforded indole 28 in 69% yield, wherein the oxindole
was introduced to position C(2) of the indole nucleus, presumably
via rearrangement of the initially formed adduct 27 at
C(3) (Scheme 6b). Sulfonylation of indole 28 with o-NsCl under basicconditions was
accompanied by unexpected chlorination of the indole moiety to afford
chloroindolenine 29 (73% yield), the structure of which
was unambiguously confirmed by X-ray crystallography (Figure 3). To the best of our knowledge, this constitutes
the first use of o-NsCl for chlorination of an indole
to provide the 3-chloroindolenine. On the other hand, the same coupling
of derivatives 25 and 26 in benzene or dichloromethane
furnished indole 28 (24% yield) and two additional undesired
products 30 (32% yield) and 31 (24% yield)
(Scheme 6c). Adduct 30 results
from nucleophilic attack at C(6) of the aurantioclavine indolecore,
while double adduct 31 is produced from both C(6) and
C(2) functionalization. The structure of 30 was unambiguously
determined following preparation of lactam 32 (Scheme 6d). Subjecting 30 to excess sodium
hydride and o-NsClconditions functionalized both
the oxindole and indolenitrogens (66% yield) and subsequent reduction
of the azide allowed for cyclization to lactam 32 in
66% yield. The structure of 32 was confirmed by single-crystal
X-ray diffraction (Figure 4).
Scheme 6
Reaction
of Aurantioclavine Derivative 25 with Bromooxindole 26
Figure 3
X-ray structure of chloroindoline 29.
Figure 4
X-ray structure of lactam 32.
X-ray structure of chloroindoline 29.X-ray structure of lactam 32.
Alkylation
Route to Communesin F
Discouraged
by the unsuccessful Diels–Alder cycloaddition-based approaches
to communesin F (1f), we considered an alternative strategy
toward the natural product. In 2007, as a direct result of our efforts
toward the communesins and perophoramidine, we developed a method
to generate 3,3-disubstituted oxindoles via the base-mediated coupling
of oxindole electrophiles with malonate-derived nucleophiles. (Scheme 7a).[20] We also developed
an asymmetric variant of this reaction utilizing copper bis(oxazoline)complexes (Scheme 7b).[21]
Scheme 7
Construction of 3,3-Disubstituted Oxindoles
With the method shown in Scheme 7, we devised
a new synthetic strategy that cast our coupling fragments in an umpolung manner, invoking an electrophilicaurantioclavine
portion and a nucleophilic right-hand fragment. We first pursued this
notion in the context of the model azepine 35 (Scheme 8). Treatment of 35 with DBU and a pronucleophile
(e.g., 36(22) and 38) produced oxindole adducts (i.e., 37 and 39) possessing the key C(7)–C(8) linkage in modest, but encouraging
yields. Importantly, adduct 37 was crystalline, and we
confirmed both the new C–C bond as well as the relative stereochemistry
of the sole diastereomeric isolate via X-ray analysis.
Scheme 8
Construction
of C(7)–C(8) Linkage by Alkylation Strategy
Having produced the key C(7)–C(8) linkage
via an umpolung strategy, we treated aurantioclavine-derived
bromooxindole 40 with malonate 41 in the
presence of DBU (Scheme 9). Smooth reactivity
under our standard conditions
led to the isolation of a single stereoisomeric adduct 42 in 74% yield. To our delight, oxindole adduct 42 was
amenable to single-crystal X-ray diffraction, however, the X-ray analysis
surprisingly revealed that the alkylation occurs with high syn selectivity relative to the existing isobutenyl substituent
(Figure 5). This result was intriguing, given
that in the Diels–Alder cycloaddition of the corresponding
indole 9 with the o-azaxylylene derived
from benzyl chloride 10, the selectivity at C(7) favored
the anti diastereomer 11 (cf. Schemes 9 and 2).[23]
Scheme 9
Alkylation of Azepine Bromooxindole 40 with Malonate 41
Figure 5
X-ray structure of oxindole adduct 42.
X-ray structure of oxindole adduct 42.Since the undesired relative stereochemistry was
obtained in adduct 42 from the alkylation of azepine 40 and malonate 41, we explored our strategy
in a model system lacking the
azepine ring of the oxindole (Scheme 10). Known
silyl ether 43(21,22) was converted into
malonate adducts 46 and 47 in 85% and 96%
yield, respectively, under our previously reported conditions in Scheme 7. Importantly, in the nonazepine system, the efficiency
of those alkylations is increased, even in these cases where vicinal
quaternary centers are generated.[24] Methylation
of oxindoles 46 and 47 produced 48 and 49 in 99% and 92% yield, respectively.
Scheme 10
Alkylation
of 3-Bromooxindole 43
Acid-catalyzed desilylation and cyclization of diester 48 proceeded smoothly to furnish lactone 50 in
85% yield
as a single diastereomer (Scheme 11a).[25] To our delight, lactone 50 underwent
decarboxylative allylic alkylation when treated with Pd(PPh3)4, yielding 51 in 90% yield as a single
diastereomer.[26,27] Single-crystal X-ray analysis
confirmed that lactone 51 possesses the relative stereochemistry
at the vicinal quaternary carboncenters C(7) and C(8) that is needed
for further elaboration to communesin F (1f). Interestingly,
direct decarboxylative allylic alkylation of diester 49 again provided an alkylated product (i.e., 52) as a
single diastereomer in 78% yield (Scheme 11b). Through X-ray analysis, we discovered that the relative stereochemistry
at the vicinal quaternary stereocenters C(20) and C(4) of 52 was complementary to that of the lactone 51 and thus
ideal for elaboration to perophoramidine (3).
Scheme 11
Model
Studies for Construction of the Vicinal Quaternary Centers
At this time, the underlying
reasons for the stereochemical relationships
observed in these two alkylation reactions are unclear. The fact that
the reactions proceed stereodivergently with high diastereocontrol
is quite remarkable. Work toward building reasonable models for stereoinduction
of β-quaternary tetrasubstituted enolates in both cyclic and
acyclic settings as well as the development of these interesting processes
in more general cases is ongoing. Nevertheless, with the promising
model systems 51 and 52 completed, we next
applied our findings to expedient formal syntheses of communesin F
(1f) and perophoramidine (3).
Formal Synthesis of Communesin F (1f)
As depicted
in our retrosynthetic strategy (Scheme 12),
communesin Fcould be completed from advanced
intermediate 53 in Qin’s synthesis.[5g] We anticipated the initial disconnection of
the aminal linkage in 53, thereby revealing oxindole
and aniline moieties in 54. Then, the lactam ring in 54 would be excised, affording lactone 55. We
envisioned that the relative stereochemical relationship at C(7) and
C(8) of lactone 55 could be established by employing
our decarboxylative allylic alkylation. The quaternary center on oxindole 56 was disassembled into 3-bromooxindole 57 and
diallyl malonate 44.
Scheme 12
Retrosynthesis of Communesin F (1f)
In the forward synthetic
sense, our efforts toward communesin Fcommenced with the elaboration of 4-bromooxindole 58 to
diallyl malonate 60 (Scheme 13). Treatment of 4-bromoindole 58 with oxalyl chloride
and methanol provided an oxoacetate (78% yield, two steps), which
was reduced to the corresponding primary alcohol 59 with
LiAlH4 in 91% yield.[5g] Silylation
of the primary alcohol with TIPSCl (98% yield) and subsequent oxidation
with pyridinium tribromide afforded dibromooxindole 57 in 89% yield.[28] Despite the extra steric
encumbrance of C(4) substitution, we were delighted to find that smooth
coupling of dibromooxindole 57 with malonate 44 produced a 3,3-disubstituted oxindole in 95% yield. Protection of
the oxindole with MeI delivered adduct 60 in 92% yield.
Microwave assisted lactonization of diester 60 with p-TsOH proceeded smoothly to furnish lactone 61 as a single diastereomer (85% yield). Gratifyingly, decarboxylative
allylic alkylation constructed the quaternary center at C(8) of compound 62 as a single diastereomer in 97% yield under Pd(PPh3)4 catalysis. The relative stereochemistry at C(7)
and C(8) of 61 and 62 was unambiguously
confirmed by X-ray analysis.
Scheme 13
Development of the Vicinal Quaternary
Center
Although ozonolysis
of alkene 62 delivered aldehyde 63 in 94%
yield, attempted reductive amination of aldehyde 63 did
not produce the desired γ–lactam 66 (Scheme 14). Upon treatment of aldehyde 63 with p-methoxybenzylammonium acetate and
sodium cyanoborohydride, amine intermediate 64 was likely
produced.[29] Instead of opening the lactone
directly (path a), nucleophilic attack by the newly generated amine
at the oxindole moiety (path b), and subsequent ring-shift tautomerization
delivered dihydroquinolinone 65 in 67% yield.
Scheme 14
Ozonolysis
and Reductive Amination of Lactone 62
Alternatively, we found that lactam 54 (an analogue
of 66) could be obtained via the reaction sequence summarized
in Scheme 15. The nitro group on compound 62 was reduced to the aniline, which resulted in concomitant
lactone ring opening to furnish a bis-oxindole 67 in
80% yield. Protection of the primary alcohol with TIPSCl (90% yield)
and protection of the oxindolenitrogen with methyl chloroformate
afforded carbamate 68 in 98% yield. Ozonolysis of alkene 68 generated aldehyde 69 (94% yield),[30] which underwent subsequent reductive amination
and selective lactamization with the electron-deficient oxindole to
afford γ-lactam 54 in 95% yield.
Scheme 15
Synthesis
of Lactam 54
With lactam 54 in hand, we envisioned that
the piperidine
D ring of 70 would be prepared by AlH3–Me2NEt mediated reductive cyclization (Scheme 16).[21,31] To our disappointment, treatment
of lactam 54 with AlH3–Me2NEt produced undesired pyrrolidinoindoline derivative 71 as a single diastereomer in 61% yield resulting from chemoselective
reduction of the N-PMB-lactam in the presence of
the oxindole. After cleavage of the TIPS group by TBAF (98% yield),[32] the PMB group was removed with DDQ[33] to provide alcohol 72. The structure
of the pentacyclic heterocycle 72 was confirmed by X-ray
analysis (Figure 6).
Scheme 16
Reductive Cyclization
of Lactam 54 with AlH3-Me2NEt
Figure 6
X-ray structure of pyrrolidinoindoline 72.
X-ray structure of pyrrolidinoindoline 72.Having failed on our
initial exploration, alternative conditions
for construction of the piperidine D ring were next explored. Treatment
of the lactam 54 with LiAlH4[34] produced debrominated compound 73 in 83% yield
(Scheme 17a). X-ray analysis of compound 73 showed a hydrogen-bonding interaction between the carbonyl
group of the PMB-protected amide and the NH group of the carbamate.
We reasoned that the undesired pyrrolidine was formed preferentially
to the piperidine due to the close proximity of the carbamate NH and
the carbonyl group of the PMB-protected amide. Next, a reductive cyclization
reaction was attempted by treatment of 54 with Tf2O and NaBH4 to construct the piperidine ring. To
our surprise, treatment of lactam 54 with Tf2O provided the PMB-protected hexacyclic oxindole 76 in
95% yield (Scheme 17b). The PMB-protected amide
of 54 was activated by Tf2O to provide 74, and nucleophilic attack by the aniline functionality furnished
pyrrolidinoindoline derivative 75. After the TIPS group
was removed under the reaction conditions, the resultant hydroxyl
group attacked the amidinium to generate the propellane structure
of hexacyclic oxindole 76. After cleavage of the PMB
group using DDQ, the propellane structure of hexacyclic oxindole 77 was confirmed using X-ray analysis.
Scheme 17
Attempted Reductive
Cyclization of Lactam 54
Despite this unexpected turn of events, we envisaged that
the desired
aminal 81 could be accessed from the propellanecompound 76 using suitable conditions, since the oxidation state at
C(9) of 76 is identical to that of the desired aminal 81 (Scheme 18). Moreover, the reactive N-PMB-pyrrolidinone in 54 is now protected
by the propellane structure of 76, thus leaving the oxindole
as the only reducible carbonyl group. Fortunately, after extensive
experimentation, we were pleased to find that reductive cyclization
of hexacyclic oxindole 76 could be accomplished with
DIBAL and Et2AlCl to furnish aminal 81 in
87% yield (Scheme 18). Presumably, the oxindole
of 76 was reduced by DIBAL to provide 78, and rearrangement of the propellane structure generated iminium 79. After the workup, water attacked the iminium moiety of 79 to afford aniline 80, and the resultant aniline
group attacked the iminium of 80 to construct aminal 81. In the last stage of the synthesis, we screened a variety
of reaction conditions to remove the PMB group on the lactam 81 (e.g., DDQ, CAN, TFA, etc.), but surprisingly, removal
of the PMB group failed under all conditions attempted. This unexpected
turn was particularly insidious since the PMB group was easily removed
from hexacyclic oxindole 76 by DDQ (Scheme 17). The cleavage of allyl or benzyl groups were
also examined, but disappointingly, cleavage of these groups on the
lactam was similarly unsuccessful under several conditions.[35]
Scheme 18
Synthesis of Aminal 81
Given the difficulty of removal
of PMB, allyl, and benzyl groups,
our attention turned to exploring the o-nitrobenzyl
group as a protecting group. However, subjecting the hexacyclic oxindole 77 to o-nitrobenzyl bromide under basicconditions
to produce the o-nitrobenzyl-protected propellanehexacyclic oxindole turned out to be challenging. Thus, we next investigated
reductive amination of aldehyde 69 and were pleased to
find that treatment of 69 with o-nitrobenzylammonium
acetate 82 furnished lactam 83 in 97% yield
(Scheme 19). Formation of the o-nitrobenzyl-protected propellane hexacyclic oxindole using Tf2O (75% yield) was followed by reductive cyclization with DIBAL
and Et2AlCl to furnish aminal 84 in 60% yield.
To our delight, we found that removal of the o-nitrobenzyl
group could be achieved by photolysis/irradiation at 350 nm in 40%
yield.[36] Surprisingly, we discovered that
removal of the o-nitrobenzyl group to produce compound 53 was also accomplished using 20% aqNaOH in methanol at
75 °C in 70% yield—a previously unknown deprotection protocol.[37] Aminal 53 has been advanced by
the Qin group to communesin F,[5g] thus completing
our formal synthesis of the natural product.
Scheme 19
Completion of Formal
Synthesis of Communesin F
Formal Synthesis of Perophoramidine (3)
Our retrosynthetic analysis of perophoramidine (3) was based on our previously established expedient synthesis
of oxindole derivative 52 (Scheme 20). We speculated that the aminal and lactam ring functionalities
of pentacycle 85, an intermediate in Funk’s synthesis,[5o] could be cleaved, thereby leading to aldehyde 86. The N–C bond of the 6-bromooxindole
moiety in 86 was excised to arrive at nitroarene 52. The construction of the contiguous quaternary centers
at C(20) and C(4) of allyl ester 52 with the proper relative
stereochemistry was accessed by decarboxylative allylic alkylation
as previously described (Scheme 11b).
Scheme 20
Retrosynthesis
of Perophoramidine (3)
Carbamate 88 was obtained by reduction of
nitroarene 52 with titanium chloride and simultaneous
oxindole formation[38] to furnish the bis-oxindole
moiety 87 in 91% yield followed by protection with Boc
anhydride in 85% yield.
Ozonolysis of olefin 88 produced aldehyde 86 in 90% yield. Reductive amination of aldehyde 86 with o-nitrobenzylammonium acetate 82 resulted in
an amine that underwent in situ lactam formation to afford oxindolelactam 89 in 91% yield (Scheme 21)
Scheme 21
Synthesis of Amide 89
Initially, we attempted to generate the o-nitrobenzylprotected propellane hexacyclic oxindole 90 under analogous
conditions to those used in our formal synthesis of communesin F on
the pseudo-diastereomeric series (vide supra). However, treatment
of lactam 89 with Tf2O yielded an unexpected
azepine 91 in 70% yield (Scheme 22). Both the Boc and the TIPS groups on amide 89 were
removed under the reaction conditions, and the resulting primary alcohol
was presumably converted to the corresponding triflate. Finally, the
aniline likely attacked the newly formed triflate to form azepine 91.
Scheme 22
Formation of Azepine 91 Using Tf2O
After extensive experimentation,
we discovered that in contrast
to the communesin system, the desired reductive cyclization in the
perophoramidine diastereomer occurred directly with AlH3–Me2NEt[31] to furnish
cyclization product 92 in 42% yield (66% yield based
on recovered starting material) (Scheme 23).
The indoline methyl group was converted to a formyl group using PDC
oxidation in 62% yield (93% yield based on recovered starting material).[39] To our delight, an attempt to remove the formyl
group with 20% aqNaOH at 75 °C resulted in removal of both the
formyl group and the o-nitrobenzyl group to produce
aminal 85 in 50% yield.[37,40] This molecule
was previously advanced by the Funk group to perophoramidine[5o] and constitutes an expedient formal synthesis
of the natural product.
Scheme 23
Completion of Formal Synthesis of Perophoramidine
Conclusion
In
conclusion, we have conducted synthetic studies toward unique
polycyclicalkaloids and completed formal syntheses of communesin
F (1f) in 9% overall yield over 17 steps and perophoramidine
(3) in 6% overall yield (13% overall yield, based on
recovered starting material) over 10 steps using a unified stereodivergent
alkylation approach. The all-carbon quaternary center on the oxindole
was established via stabilized enolate alkylation of 3-bromooxindoles,
a method previously developed by our laboratory and now shown to be
quite versatile even in particularly sterically challenging situations.
The complementary relative stereochemistry of the two contiguous quaternary
stereogeniccenters found in communesin F (1f) and perophoramidine
(3), respectively, was established by substrate controlled
diastereoselective decarboxylative allylic alkylation. A reductive
amination approach furnished the A ring, and reductive cyclization
produced the D ring for both communesin F (1f) and perophoramidine
(3). En route to the evolution of our eventual successful
strategy, we have discovered a method to convert an indole to a 3-chloroindolenine
using a mild reagent such as o-NsCl during the synthesis.
In addition, previously unknown, mild and efficient deprotection conditions
for the o-nitrobenzyl group on the lactam were discovered.
Further studies to rationalize unprecedented complementary selectivity
by Pd-catalyzed allylic alkylation reactions are currently in progress.
To a solution of 4-methyl-N-(2-vinylphenyl)benzenesulfonamide (SI-1) (6.11 g, 22.4
mmol, 1.00 equiv) in THF (140 mL) and water (70 mL) were added N-methylmorpholine N-oxide (5.96 g, 50.8
mmol, 2.30 equiv) and osmium tetroxide (11.6 mg, 43.9 mmol, 0.002
equiv). After addition, reaction was stirred for 3 days. The reaction
was concentrated to approximately 50 mL under reduced pressure and
then extracted with a mixture ether and THF (1:1) (3 × 45 mL).
The organic layers were dried over sodium sulfate, and the solvent
was removed under reduced pressure. Impurities were removed by washing
solid with dichloromethane to afford diol SI-2 (5.43
g, 80% yield) as a white solid: R = 0.13 (1:1 hexane/EtOAc); 1H NMR (300 MHz, CDCl3) δ 8.47 (s, 1H), 7.71 (d, J = 8.3
Hz, 2H), 7.39 (dd, J = 8.1, 1.1 Hz, 1H), 7.25–7.18
(m, 3H), 7.15–7.02 (m, 2H), 4.82 (t, J = 6.5
Hz, 1H), 3.66–3.57 (m, 2H), 2.97 (br, s, 1H), 2.39 (s, 3H),
1.98 (br, s, 1H); 13C NMR (75 MHz, CDCl3) δ
144.1, 137.1, 136.3, 129.9, 129.8, 129.2, 128.5, 127.4, 127.0, 122.2,
74.78, 66.0, 21.8; IR (neat film NaCl) 3271, 1318, 1150 cm–1; HRMS (MM: ESI-APCI+) m/z calcd
for C15H18NO4S [M + H]+ 308.0951, found 308.0967.To a solution of diol SI-2 (500 mg, 1.63 mmol, 1.00 equiv) in toluene (70 mL) was added dibutyltin
dimethoxide (410 μL, 1.79 mmol, 1.10 equiv). The flask was fitted
with a short path distillation apparatus, and approximately half of
the solvent was removed by distillation. To this solution were added
MOMCl (136 μL, 1.79 mmol, 1.10 equiv) and tetrabutylammonium
iodide (900 mg, 2.44 mmol, 1.50 equiv). After addition, the reaction
was stirred for 12 h, and then brine was added to this solution. The
reaction mixture was extracted with EtOAc (3 × 50 mL). The combined
organic phases were dried over MgSO4 and concentrated in
vacuo. The residue was purified by flash column chromatography (3:1
→ 1:1 hexanes/EtOAc) to afford alcohol 13 (513
mg, 90% yield, two steps) as a white solid: R= 0.27 (1:1 hexane/EtOAc); 1H NMR (300 MHz, CDCl3) δ 8.76 (s, 1H), 7.72
(d, J = 8.5 Hz, 2H), 7.50 (d, J = 8.0 Hz, 1H), 7.27–7.20 (m, 3H), 7.11–7.02 (m, 2H),
4.83–4.78 (m, 1H), 4.64 (s, 2H), 3.60 (dd, J = 10.5, 3.5 Hz, 1H), 3.48–3.41 (m, 2H), 3.39 (s, 3H), 2.39
(s, 3H); 13C NMR (75 MHz, CDCl3) δ 143.9,
137.1, 136.3, 129.7, 129.5, 128.9, 128.2, 127.2, 124.7, 122.0, 97.0,
73.3, 72.4, 55.6, 21.6; IR (neat film NaCl) 3233, 2932, 1598, 1497,
1335, 1161 cm–1; HRMS (MM: ESI-APCI+) m/z calcd for C17H22NO5S [M + H]+ 352.1213, found 352.1219.
A flame-dried flask (25 mL) equipped with
a Teflon stirbar was charged with sodium hydride (60% dispersion in
mineral oil, 22 mg, 0.55 mmol, 1.10 equiv), which was washed 3 times
with dry hexanes. Then, DMSO (5.5 mL) and trimethylsulfoxonium iodide
(119 mg, 0.58 mmol, 1.20 equiv) were added. To this solution was added
methyl (E)-2-(2-oxoindolin-3-ylidene)acetate (SI-3) (100 mg, 0.49 mmol, 1.00 equiv) in a solution of DMSO
(2.5 mL). After addition, the reaction was stirred for 2 h, and then
the temperature was raised to 50 °C. The reaction was complete
after another hour. Brine was added and then the mixture was extracted
with EtOAc (3 × 5 mL). The combined organic phases were dried
over MgSO4 and concentrated in vacuo. The residue was purified
was by flash column chromatography (3:1 → 1:1 hexanes/EtOAc)
to afford oxindole 16 as two diastereomers. Diastereomer
1: (44.7 mg, 42% yield). Diastereomer 2: (28.6 mg, 27% yield). Diastereomer
1: R= 0.52 (1:1 hexane/EtOAc); 1H NMR (300 MHz, CDCl3) δ 8.28 (br, s, 1H), 7.34 (d, J = 7.7 Hz,
1H), 7.22 (dd, J = 7.7, 1.3 Hz, 1H), 7.02 (dd, J = 7.7, 1.1 Hz, 1H), 6.99–6.93 (m, 1H), 3.69 (s,
3H), 2.72 (dd, J = 8.6, 7.4 Hz, 1H), 2.16 (dd, J = 7.4, 4.5 Hz, 1H), 2.04 (dd, J = 8.6,
4.4 Hz, 1H); 13C NMR (75 MHz, CDCl3) δ
177.5, 169.3, 141.8, 127.9, 126.4, 123.0, 122.4, 110.3, 52.4, 34.3,
32.9, 21.1; IR (neat film NaCl) 3214, 1712, 1622, 1470, 1209 cm–1; HRMS (MM: ESI-APCI+) m/z calcd for C12H12NO3 [M
+ H]+ 218.0812, found 218.0825. Diastereomer 2: R= 0.45 (1:1
hexane/EtOAc); 1H NMR (300 MHz, CDCl3) δ
7.99 (br, s, 1H), 7.25–7.19 (m, 1H), 7.02 (td, J = 7.6, 1.0 Hz, 1H), 6.93 (d, J = 7.8 Hz, 1H), 6.85–6.80
(m, 1H), 3.75 (s, 3H), 2.66 (t, J = 8.3 Hz, 1H),
2.39 (dd, J = 5.0, 8.0 Hz, 1H), 1.84 (dd, J = 5.0, 8.6 Hz, 1H); 13C NMR (75 MHz, CDCl3) δ 176.0, 167.7, 141.1, 129.5, 188.0, 122.4, 118.9,
110.3, 52.6, 33.5, 32.9, 21.3; IR (neat film NaCl) 3256, 1739, 1710
cm–1; HRMS (MM: ESI-APCI+) m/z calcd for C12H12NO3 [M
+ H]+ 218.0812, found 218.0828.
Benzo[c]isoxazole-3-carboxylic Acid 20
A flame-dried
flask (500 mL) equipped with a Teflon stirbar
was charged with 2-nitrophenylacetic acid 21 (10.0 g,
55.2 mmol, 1.00 equiv), ethanol (60 mL), sulfuric acid (200 μL),
and toluene (280 mL). The flask was fitted with a condenser, and the
solution was refluxed for 14 h. The solvent was removed under reduced
pressure and sulfuric acid (280 mL) was added. After addition, the
reaction was heated to 110 °C and stirred for 90 min. The solution
was then poured onto ice (600 g), and the mixture was extracted with
ether (3 × 200 mL). The combined organic layers were dried over
MgSO4 and concentrated in vacuo. Purification was performed
via crystallization from water to afford acid 20 (3.94
g, 44% yield, 2 steps) as an off-white solid: 1H NMR (300
MHz, acetone-d6) δ 10.82 (br, s,
1H), 7.94 (d, J = 9.0 Hz, 1H), 7.75 (d, J
= 10.0 Hz, 1H), 7.50 (dd, J = 6.5, 9.5 Hz,
1H), 7.34 (dd, J = 7.0, 8.5 Hz, 1H); 13C NMR (75 MHz, acetone-d6) δ 158.5,
158.1, 155.3, 132.4, 128.8, 121.4, 121.0, 116.7; IR (neat film NaCl)
2360, 1731, 1301, 1231, 1189, 753 cm–1; HRMS (MM:
ESI-APCI+) m/z calcd for C8H6NO3 [M + H]+ 164.0342, found 164.0341.
To a solution of (E)-2-methyl-4-(3-(2-nitroethyl)-1H-indol-4-yl)but-3-en-2-ol (SI-4) (386 mg,
1.41 mmol, 1.00 equiv) in THF (14 mL) was added methyl iodide (875
μL, 14.1 mmol, 10.0 equiv) at 0 °C. Sodium hydride (60%
dispersion in mineral oil, 562 mg, 14.5 mmol, 10.3 equiv) was then
added to the solution, and the mixture was stirred for 25 min at 23
°C. The reaction was quenched with satdrated ammonium hydroxide
solution and extracted with EtOAc (3 × 10 mL). The combined organic
layers were dried over MgSO4 and concentrated in vacuo.
The residue was purified by flash column chromatography (3:1 →
2:1 hexanes/EtOAc) to afford (E)-2-methyl-4-(1-methyl-3-(2-nitroethyl)-1H-indol-4-yl)but-3-en-2-ol (SI-5) (363.5 mg,
90% yield) as a yellow solid.To a solution of nitrocompound SI-5 (512 mg, 1.78 mmol, 1.00 equiv) in MeOH (125 mL) and
2 N HCl (40 mL) was added amalgamated zinc, which had been formed
from zinc dust (6.5 g, 98.3 mmol, 55.0 equiv) and mercuric chloride
(1.10 g, 3.55 mmol, 2.00 equiv) in 2 N HCl and subsequently rinsed
with MeOH. The mixture was stirred at reflux for 3 h. The reaction
was then decanted from the remaining amalgam and then basified to
pH >10. The solid was removed by filtration, and the resulting
solution
was extracted with dichloromethane (3 × 100 mL). The combined
organic layers were dried over MgSO4 and concentrated in
vacuo. The residue was purified by flash column chromatography (18:1
CH2Cl2/MeOH) to afford 1-methylaurantioclavine 5 (258 mg, 60% yield) as a yellow oil: R= 0.30 (18:1 CH2Cl2/MeOH); 1H NMR (300 MHz, CDCl3) δ 7.19–7.12 (m, 2H), 6.89–6.83 (m, 2H), 5.48
(d, J = 9.0 Hz, 1H), 4.92 (d, J = 9.0 Hz, 1H), 3.76 (s, 3H), 3.62–3.54 (m, 1H), 3.13–3.02
(m, 3H), 2.26 (br, s, 1H), 1.86 (s, 6H); 13C NMR (75 MHz,
CDCl3) δ 138.5, 137.8, 133.3, 127.7, 125.9, 121.1,
117.4, 114.2, 107.3, 62.6, 48.9, 32.7, 30.8, 25.9, 18.4; IR (neat
film NaCl) 3332, 2910, 1554, 1455 cm–1; HRMS (MM:
ESI-APCI+) m/z calcd for C16H21N2 [M + H]+ 241.1699, found 241.1712.
A flame-dried
vial (20 mL) equipped with a Teflon stirbar was charged with amide 22 (100 mg, 0.269 mmol, 1.00 equiv) and cooled to 0 °C.
To this reaction mixture was added a 0.5 M solution of HCl in MeOH
(2.7 mL, generated from addition of acetyl chloride to methanol at
0 °C) at 0 °C. The mixture was stirred for 1 h and then
warmed to 23 °C over 30 min. The solvent was then removed under
reduced pressure. Purification was performed by washing the solid
with dichloromethane to afford indole 23 (31.1 mg, 31%
yield) as a white solid: R= 0.22 (1:1 hexane/EtOAc); 1H
NMR (300 MHz, DMSO) δ 11.05 (d, J = 2.5 Hz,
1H), 7.35–7.17 (m, 6H), 7.09 (t, J = 7.7 Hz,
1H), 6.61 (dt, J = 7.4, 1.0 Hz, 1H), 5.47 (d, J = 6.6 Hz, 1H), 4.21 (dt, J = 13.2, 3.7
Hz, 1H), 3.45 (td, J = 13.2, 12.5, 2.4 Hz, 1H), 3.15
(dt, J = 16.0, 3.0 Hz, 1H), 3.08–2.94 (m,
1H), 2.37 (dd, J = 6.6, 0.9 Hz, 1H), 1.77 (s, 3H),
1.03 (s, 3H); 13C NMR (75 MHz, DMSO) δ 164.3, 153.0,
139.6, 137.0, 134.6, 127.3, 126.6, 123.7, 122.6, 121.4, 118.6, 118.0,
115.2, 112.7, 110.2, 96.4, 70.1, 63.1, 61.2, 44.4, 26.9, 26.6, 26.0;
IR (neat film NaCl) 3314, 1681, 753 cm–1; HRMS (MM:
ESI-APCI+) m/z calcd for C23H22N3O2 [M + H]+ 372.1707,
found 372.1710.
A flame-dried
vial (20 mL) equipped with a Teflon stirbar was charged with amide 19 (100 mg, 0.259 mmol, 1.00 equiv) and cooled to 0 °C.
To this solution was added a 0.5 M solution of HCl in MeOH (2.6 mL,
generated from addition of acetyl chloride to methanol at 0 °C)
at 0 °C. The mixture was stirred for 1 h and then warmed to 23
°C over 30 min. The solvent was then removed under reduced pressure.
Purification was performed via flash column chromatography (3:1 →
1:1 hexanes/EtOAc) to afford indole 24 (101 mg, 99% yield)
as a white solid: R= 0.29 (1:1 hexane/EtOAc); 1H NMR (300 MHz, CDCl3) δ 7.33–7.32 (m, 1H), 7.24–7.14 (m, 5H),
7.00 (s, 1H), 6.76–6.73 (m, 1H), 5.51 (d, J = 6.5 Hz, 1H), 4.57–4.50 (m, 1H), 3.80 (s, 3H), 3.41 (dd, J = 12.4, 10.0 Hz, 1H), 3.33–3.24 (m, 1H), 3.19–3.12
(m, 1H), 2.57 (d, J = 6.5 Hz, 1H), 1.95 (s, 3H),
1.18 (s, 3H); 13C NMR (75 MHz, CDCl3) δ
165.4, 153.5, 139.8, 137.9, 135.6, 127.5, 127.0, 126.9, 124.6, 121.9,
119.1, 118.4, 115.9, 113.4, 108.4, 97.2, 70.9, 64.1, 61.9, 45.2, 33.0,
27.5, 27.2, 26.6; IR (neat film NaCl) 3315, 2932, 1699, 1456, 1317,
754 cm–1; HRMS (MM: ESI-APCI+) m/z calcd for C24H24N3O2 [M + H]+ 386.1863, found 386.1867.
To a solution
of 2,3,4,6-tetrahydro-1H-azepino[5,4,3-cd]indole (SI-9) (106 mg, 0.614 mmol, 1.00 equiv) and
Et3N (0.17 mL, 1.23 mmol, 2.00 equiv) in CH2Cl2 (4 mL) cooled to 0 °C was added a solution of
TsCl (117 mg, 0.614 mmol, 1.00 equiv) in CH2Cl2 (3 mL) dropwise. The ice bath was removed, and the reaction mixture
was stirred for 5 h, diluted with EtOAc (160 mL), and washed with
0.5 N HCl (2 × 30 mL) and brine. The organic layers were combined,
dried over MgSO4, and concentrated in vacuo. The residue
was purified by column chromatography (2:1 hexanes/EtOAc) to afford
2-tosyl-2,3,4,6-tetrahydro-1H-azepino[5,4,3-cd]indole (SI-10) (164 mg, 82% yield).Indole SI-10 was dissolved in THF (10 mL), t-BuOH (10 mL), and water (1 mL). The solution was cooled to 0 °C,
and pyridinium tribromide (504 mg, 1.54 mmol, 1.02 equiv) was added.
The reaction mixture was stirred at 0 °C for 45 min and then
allowed to warm to ambient temperature. The reaction was quenched
by addition of 10 mL of 1:1 v/v 1 M Na2S2O3/satd NaHCO3. The reaction mixture was diluted
with brine (50 mL) and extracted with EtOAc (3 × 50 mL). The
combined organic extracts were dried with MgSO4 and concentrated
in vacuo. The residue was purified by silica gelchromatography (2:1
hexanes/acetone) to afford 2-tosyl-1,2,3,4,4a,6-hexahydro-5H-azepino[5,4,3-cd]indol-5-one (SI-11) (397 mg, 75% yield) as a white solid: R= 0.15 (1:1 hexanes/EtOAc); 1H NMR (300 MHz, CDCl3) δ 8.77 (br, s, 1H),
7.56 (d, J = 8.2 Hz, 2H), 7.21–7.14 (m, 3H),
6.94 (d, J = 7.7 Hz, 1H), 6.80 (d, J = 7.7 Hz, 1H), 4.80 (d, J = 15.5 Hz, 1H), 4.30
(d, J = 13.6 Hz, 1H), 4.15 (d, J = 15.5 Hz, 1H), 3.52 (dd, J = 12.5, 3.5 Hz, 1H),
3.24 (t, J = 12.6 Hz, 1H), 2.38 (s, 3H), 2.30 (m,
1H), 1.53 (m, 1H); 13C (75 MHz, CDCl3) δ
178.4, 143.3, 140.5, 137.1, 135.9, 129.6, 128.4, 128.3, 127.0, 121.5,
109.1, 53.3, 51.6, 46.2, 28.6, 21.5; IR (neat film NaCl) 3276, 2925,
2853, 1698, 1618, 1463, 1326, 1153 cm–1; HRMS (MM:
ESI-APCI+) m/z calcd for C18H19N2O3S [M + H]+ 343.1111,
found 343.1104.LiHMDS (429 mg, 2.57 mmol, 2.50 equiv) was dissolved
in THF (5
mL). The solution was cooled to −78 °C, and a solution
of oxindoleSI-11 (352 mg, 1.03 mmol, 1.00 equiv) in
THF (20 mL) was added dropwise over 20 min. The reaction mixture was
stirred at −78 °C for 20 min and transferred to a precooled
solution of N-bromosuccinimide (457 mg, 2.57 mmol,
2.50 equiv) in THF (10 mL) that was protected from light. The resulting
reaction mixture was placed in a −40 °C bath for 1 h,
while being protected from light, and then quenched with satd NH4Cl. The reaction mixture was allowed to warm to ambient temperature,
diluted with brine (100 mL), and extracted with EtOAc (3 × 70
mL). The combined organic extracts were dried over MgSO4 and concentrated to afford a yellow oil, which was purified by silica
gel chromatography (2:1 hexanes/EtOAc) to afford bromooxindole 35 (334 mg, 76% yield) as a yellow solid: R = 0.40 (1:1 hexane/EtOAc); 1H NMR (300 MHz, CDCl3) δ 8.49 (br, s, 1H), 7.58
(d, J = 8.2 Hz, 2H), 7.24–7.19 (m, 3H), 6.96
(d, J = 7.7 Hz, 1H), 6.83 (d, J =
7.8 Hz, 1H), 4.76 (d, J = 15.4 Hz, 1H), 4.32 (m,
2H), 3.80 (t, J = 13.2 Hz, 1H), 2.39 (s, 3H), 2.33
(m, 1H), 1.90 (m, 1H); 13C NMR (75 MHz, CDCl3) δ 175.1, 143.5, 139.0, 137.7, 136.9, 130.7, 129.7, 128.7,
126.9, 122.6, 110.1, 59.2, 51.6, 48.3, 35.2, 21.5; IR (neat film NaCl)
3313, 2930, 1734, 1615, 1460, 1334, 1155, 1096, 727 cm–1; HRMS (MM: ESI-APCI+) m/z calcd
for C18H18BrN2O3S [M +
H]+ 421.0216, found 421.0213.
To a solution of aurantioclavine 4 (300 mg,
0.00133 mol, 1.00 equiv) and Et3N (0.37 mL, 0.00265 mol,
2.00 equiv) in CH2Cl2 (4 mL) cooled to 0 °C
was added a solution of TsCl (254 mg, 0.00133 mmol, 1.00 equiv) in
CH2Cl2 (3 mL) dropwise. The ice bath was removed,
and the reaction mixture was stirred for 5 h, then diluted with EtOAc
(200 mL) and washed with 0.5 N HCl (2 × 35 mL) and brine. The
organic layers were combined, dried over MgSO4 and concentrated
in vacuo. The residue was purified by column chromatography (2:1 hexanes/EtOAc)
to afford 1-(2-methylprop-1-en-1-yl)-2-tosyl-2,3,4,6-tetrahydro-1H-azepino[5,4,3-cd]indole (SI-12) (405 mg, 80% yield).A solution of indoleSI-12 (902.2 mg, 2.371 mmol, 1.00 equiv) in THF/t-BuOH/H2O (10:10:1 v/v/v, 52.5 mL) was cooled to 0 °C, and pyridinium
tribromide (834.2 mg, 2.608 mmol, 1.10 equiv) was added in small portions
over 5 min. The reaction mixture was stirred at 0 °C for 15 min,
and then allowed to warm to ambient temperature. After 5 min at ambient
temperature, the reaction mixture was quenched by addition of 1:1
v/v satd NaHCO3/1 M aqNa2S2O3 (15 mL), poured into brine (150 mL), and extracted with EtOAc
(3 × 100 mL). The combined extracts were dried over Na2SO4 and concentrated under reduced pressure to afford
a brown solid, which was purified by silica gelchromatography (2:1
→ 1:1 hexanes/EtOAc) to afford 1-(2-methylprop-1-en-1-yl)-2-tosyl-1,2,3,4,4a,6-hexahydro-5H-azepino[5,4,3-cd]indol-5-one (SI-13) (747.5 mg, 80% yield): R = 0.22 (5:1 benzene/MeCN); 1H NMR (300 MHz, CDCl3) δ 8.12 (br, s, 1H), 7.33 (dd, J =
8.5, 2.1 Hz, 2H), 7.03 (m, 1H), 6.96 (d, J = 7.2
Hz, 2H), 6.82 (d, J = 7.7 Hz, 1H), 6.64 (d, J = 8.0 Hz, 1H), 5.76 (d, J = 8.0 Hz, 1H),
5.31 (m, 1H), 4.00 (dt, J = 15.7, 2.8 Hz, 1H), 3.64–3.50
(m, 2H), 2.22 (s, 3H), 2.00 (m, 1H), 1.67 (s, 3H), 1.65 (s, 3H), 1.25
(m, 1H); 13C NMR (75 MHz, CDCl3) δ 178.4,
142.9, 140.9, 140.5, 138.2, 129.2, 128.3, 128.3, 127.0, 126.7, 121.3,
119.0, 108.7, 59.0, 46.0, 43.8, 27.8, 26.0, 21.4, 18.5; IR (neat film
NaCl) 3246, 2925, 1713, 1615, 1460, 1326, 1155, 732 cm–1; HRMS (MM: ESI-APCI+) m/z calcd
for C22H25N2O3S [M + H]+ 397.1580, found 397.1586.A solution of oxindoleSI-13 (172.0 mg, 0.434 mmol,
1.00 equiv) in THF (5 mL) was added dropwise to a freshly prepared
solution of LiHMDS (217.8 mg, 1.301 mmol, 3.00 equiv) in THF (5 mL)
that had been precooled to −78 °C. After 20 min at −78
°C, the resulting solution was transferred via cannula to a solution
of N-bromosuccinimide (231.6 mg, 1.301 mmol, 3.00
equiv) in THF (5 mL) that had been precooled to −78 °C.
The resulting yellow reaction mixture was allowed to warm to −15
°C (the reaction flask was transferred to a bath composed of
ethylene glycol and dry ice) and maintained at this temperature for
2 h. The reaction mixture was then cooled to −78 °C and
quenched by addition of satd NH4Cl (5 mL). The yellow reaction
mixture was allowed to warm to ambient temperature and diluted with
H2O (80 mL), then extracted with EtOAc (3 × 70 mL).
The combined organic extracts were washed with brine (100 mL), dried
over MgSO4, and concentrated under reduced pressure to
afford a yellow oil, which was purified immediately by silica gelchromatography (2:1 hexanes/EtOAc) to afford a 3:1 mixture of bromooxindole 40 (>20:1 dr) and dehydrobromination product, 1-(2-methylprop-1-en-1-yl)-2-tosyl-1,2,3,6-tetrahydro-5H-azepino[5,4,3-cd]indol-5-one (SI-14) (131.1 mg, 64% combined yield, 50% yield of bromooxindole 40). Bromooxindole 40 was stored frozen in benzene
and used without further purification. The relative configuration
of this bromooxindole was assigned based on the stereochemistry of
the malonate adduct obtained (see below). Quenching the reaction prior
to completion afforded the bromooxindole 40 as a single
diastereomer, which showed greater stability, and could be fully characterized: R = 0.50 (1:1 hexane/EtOAc); 1H NMR (300 MHz, CDCl3) δ 8.62 (br, s, 1H),
7.49 (d, J = 8.0 Hz, 2H), 7.21 (t, J = 8.0, 1H), 7.13 (d, J = 8.0 Hz, 2H), 6.90 (d, J = 7.5 Hz, 1H), 6.79 (dd, J = 8.0, 1.0
Hz, 1H), 5.96 (td, J = 8.5, 1.5 Hz, 1H), 5.88 (d, J = 8.5 Hz, 1H), 4.18–4.02 (m, 2H), 2.35 (s, 3H),
2.27 (m, 1H), 1.97 (m, 1H), 1.76 (d, J = 1.0 Hz,
3H), 1.75 (d, J = 1.5 Hz, 3H); 13C NMR
(75 MHz, CDCl3) δ 174.8, 143.2, 142.4, 140.2, 139.2,
137.8, 130.8, 129.4, 126.9, 126.1, 122.7, 120.1, 109.7, 66.8, 59.5,
39.8, 34.4, 26.1, 21.4, 18.4; IR (neat film NaCl) 3291, 1734, 1617,
1602, 1457, 1326, 1156, 1092, 738 cm–1; HRMS (MM:
ESI-APCI+) m/z calcd for C22H24BrN2O3S [M + H]+ 475.0686,
found 475.0668.
A solution of alkene 62 (47.1
mg, 100 μmol, 1.00 equiv) in a mixture of CH2Cl2 (2.5 mL) and MeOH (2.5 mL) in a Schlenk flask hooked up to
an ozone generator was purged with oxygen gas at −78 °C
(5 min, flow 0.25). Then the ozone generator was turned on (low–medium
setting), and an ozone/oxygen gas mixture was bubbled through the
reaction. The progress of the reaction was checked via TLC (9:1 hexanes/CH2Cl2) in short time intervals (1–2 min).
Upon completion of the reaction, the mixture was purged with oxygen
gas for 5 min, and DMS (36.0 μg, 500 μmol, 5.00 equiv)
was added. The reaction mixture was slowly warmed to ambient temperature,
and stirred for 16 h. The residue was purified by column chromatography
(9:1 CH2Cl2:EtOAc) on silica gel to afford aldehyde 63 (44 mg, 94% yield): R = 0.28 (9:1 CH2Cl2/EtOAc); 1H NMR (500 MHz, DMSO-d6) δ 8.64
(d, J = 3.4 Hz, 1H), 7.83–7.79 (m, 1H), 7.65–7.59
(m, 1H), 7.43–7.39 (m, 1H), 7.31 (t, J = 8.0
Hz, 1H), 7.21 (d, J = 7.9 Hz, 1H), 7.05 (d, J = 8.1 Hz, 1H), 6.92 (d, J = 8.3 Hz, 1H),
5.15 (ddd, J = 12.5, 10.8, 4.7 Hz, 1H), 4.59 (dd, J = 11.1, 6.9 Hz, 1H), 3.94 (ddd, J = 14.5,
12.6, 7.1 Hz, 1H), 3.22 (s, 3H), 3.08 (d, J = 17.0
Hz, 1H), 2.67 (dd, J = 17.0, 3.5 Hz, 1H), 1.96 (dd, J = 14.7, 4.6 Hz, 1H); 13C NMR (125 MHz, DMSO)
δ 198.0, 174.8, 171.0, 151.5, 146.3, 133.3, 131.9, 131.3, 129.9,
129.7, 127.6, 125.5, 124.6, 122.4, 109.3, 65.4, 55.6, 52.6, 49.8,
26.4, 22.3; IR (neat film NaCl) 1695, 1600, 1528, 1458, 1354, 1294,
1222, 1118, 850, 787 cm–1; HRMS (MM: ESI-APCI+) m/z calcd for C21H18BrN2O6 [M + H]+ 473.0343, found
473.0346.
Authors: John W Blunt; Brent R Copp; Murray H G Munro; Peter T Northcote; Michèle R Prinsep Journal: Nat Prod Rep Date: 2006-01-09 Impact factor: 13.423
Authors: Seo-Jung Han; Florian Vogt; Shyam Krishnan; Jeremy A May; Michele Gatti; Scott C Virgil; Brian M Stoltz Journal: Org Lett Date: 2014-06-09 Impact factor: 6.005
Authors: Hsiao-Ching Lin; Travis C McMahon; Ashay Patel; Michael Corsello; Adam Simon; Wei Xu; Muxun Zhao; K N Houk; Neil K Garg; Yi Tang Journal: J Am Chem Soc Date: 2016-03-18 Impact factor: 15.419
Authors: Hsiao-Ching Lin; Grace Chiou; Yit-Heng Chooi; Travis C McMahon; Wei Xu; Neil K Garg; Yi Tang Journal: Angew Chem Int Ed Engl Date: 2015-01-08 Impact factor: 15.336
Figure 1
Scheme 1
Scheme 2
Scheme 3
Scheme 4
Scheme 5
Figure 2
Scheme 6
Figure 3
Figure 4
Scheme 7
Scheme 8
Scheme 9
Figure 5
Scheme 10
Scheme 11
Scheme 12
Scheme 13
Scheme 14
Scheme 15
Scheme 16
Figure 6
Scheme 17
Scheme 18
Scheme 19
Scheme 20
Scheme 21
Scheme 22
Scheme 23