Synthesis of a Spirocyclic Indoline Lactone

Article abstract of DOI:10.1021/jo0499569

Base-promoted cyclization of tert-butyl [2-(benzylideneamino)phenyl] acetate (13a) and subsequent C3-alkylation with allyl bromide affords 3-allyl-2-phenyl-2,3-dihydro-1H-indole-3-carboxylic acid, tert-butyl ester (15b) in high yield as a single diastereomer. This result is contrary to prior publications that describe failed cyclization of an analogous ethyl ester (ethyl [2-(4-methoxybenzylideneamino)phenyl] acetate) under strongly basic conditions. N-Acylation, olefin dihydroxylation, and tert-butyl ester cleavage affords the spirocyclic lactone 18 as a pair of diastereomers. Isolation and characterization of individual diastereomers 18a and 18b are described.

Full text of DOI:10.1021/jo0499569

Syn th esis of a Sp ir ocyclic In d olin e La cton e
J ohn C. Hodges,* Wei Wang, and Frank Riley
Pfizer Global Research and Development, 2800 Plymouth Road, Ann Arbor, Michigan 48105
jack.hodges@pfizer.com
Received J anuary 6, 2004
Base-promoted cyclization of tert-butyl [2-(benzylideneamino)phenyl]acetate (13a ) and subsequent
C3-alkylation with allyl bromide affords 3-allyl-2-phenyl-2,3-dihydro-1H-indole-3-carboxylic acid,
tert-butyl ester (15b) in high yield as a single diastereomer. This result is contrary to prior
publications that describe failed cyclization of an analogous ethyl ester (ethyl [2-(4-methoxybenz-
ylideneamino)phenyl]acetate) under strongly basic conditions. N-Acylation, olefin dihydroxylation,
and tert-butyl ester cleavage affords the spirocyclic lactone 18 as a pair of diastereomers. Isolation
and characterization of individual diastereomers 18a and 18b are described.
SCHEME 1
In tr od u ction
Numerous biologically active natural products contain
a spirocyclic ring system and many others contain a 2,3-
dihydroindole (indoline) structural component.1 In an
effort to add readily synthesized molecules with natural
product features to our screening collection, we investi-
gated synthetic routes to molecules with both indoline
and spirocyclic components. The facile oxidation of indo-
lines to indoles complicates many synthetic approaches
to indolines. Thus, we chose to investigate 3,3-disubsti-
tuted indolines to block such oxidation. In particular,
indolines bearing a 3,3-spirocyclic element were of inter-
est since several synthetic approaches have previously
been published.2-5
One approach that appeared to be conceptually attrac-
tive is shown in Scheme 1. This strategy employs base-
promoted intramolecular cyclization of imines prepared
from 2-aminophenylacetic acid esters such as 1. This
cyclization would potentially afford a wide variety of
2-substituted-indoline-3-carboxylates (2), which might be
subsequently alkylated at C3 as a prelude to elaboration
of spirocycles such as 4. Considerable literature precedent
is available in key papers by Speckamp et al.,3-6 who
thoroughly investigated a number of analogous reactions
as shown in Scheme 2. The complete failure of 5 to
undergo cyclization following treatment with a variety
of solvent and base combinations advised against the
approach in Scheme 1. There are two potential explana-
tions proposed by Speckamp for the failure of 5 to cyclize:6
(1) “unfavorable geometry for a 5-endo-trig ring closure
(anti-Baldwin process)” and (2) “incomplete carbanion
formation in protic solvents”. An alternative approach to
cyclization,6 utilizing a Lewis acid to promote the conver-
sion of 10 to a mixture of 11a and 11b, provides evidence
that unfavorable geometry for ring closure is not the
primary issue. Furthermore, 6 and 8 do indeed cyclize
under basic conditions and in the latter case diastereo-
selectivity is remarkably well controlled by variation of
solvent and base. Therefore, we chose to reinvestigate
the failed cyclization of 5 with the goal of finding suitable
conditions for diastereoselective cyclization.
Resu lts a n d Discu ssion
Preliminary experiments with the synthesis of the
starting imine 1 revealed that the tert-butyl esters are
typically crystalline when the imine is formed with an
aromatic aldehyde. The imine methyl and ethyl esters
were surprisingly difficult to crystallize and hence puri-
fication was considered an issue that could impact the
success of the intramolecular cyclization. Thus the initial
cyclization experiments (Scheme 3) were performed with
the tert-butyl esters, 13a and 13b, which were prepared
in three steps from 2-nitrophenylacetic acid, 12.
(1) Southon, I. W.; Buckingham, D. J . Dictionary of Alkaloids;
Cordell, G. A., Saxton, J . E., Shamma, M., Eds.; Chapman and Hall;
New York, 1989.
(2) Ibrahim-Ouali, M.; Sinibaldi, M.; Troin, Y.; Cuer, A.; Dauphin,
G.; Gramain, J . Heterocycles 1995, 41 (9), 1939-1950.
(3) Speckamp, W. N.; Veenstra, S. J .; Dijkink, J .; Fortgens, R. J .
Am. Chem. Soc. 1981, 103 (15), 4643-4644.
(4) Veenstra, S. J .; Speckamp, W. N. J . Am Chem. Soc. 1981, 103
(15), 4645-4646.
(5) Speckamp, W. N. Heterocycles 1984, 21 (1), 211-234.
(6) Dijkink, J .; Zonjee, J . N.; de J ong, B. S.; Speckamp, W. N.
Heterocycles 1983, 20 (7), 1255-1258.
10.1021/jo0499569 CCC: $27.50 © 2004 American Chemical Society
Published on Web 03/09/2004
2504
J . Org. Chem. 2004, 69, 2504-2508

Synthesis of a Spirocyclic Indoline Lactone
SCHEME 2
byproducts, presumably by air oxidation. Furthermore,
cyclization of 13a , followed by quenching with io-
domethane afforded the 3-methylated 15a . In this reac-
tion no N-methylation and no indole byproduct 14b′ were
observed; however, a small amount of 14a was isolated
due to incomplete 3-methylation. Again, only one dias-
tereomer of 15a was produced. After removal of the tert-
butyl ester, a 2D ROESY study showed a strong nOe
between the 2-proton and the 3-methyl group of 16a ,
indicating a syn relationship. A chemical shift difference
between the t-Bu ester of 14a (δ1.48 ppm) and 15a (δ0.98
ppm) was observed, consistent with inversion of the
3-position upon alkylation.
In optimizing this tandem cyclization-alkylation pro-
cedure we took advantage of the intense purple color of
the highly conjugated ester enolate as a colorimetric
indicator. Thus 1 M KOt-Bu in THF was added dropwise
to a -78 °C solution of 13a in THF until the purple color
persisted then 1.0 equiv more was added. The reaction
was warmed gradually by removal of the cold bath until
the color faded from purple to orange. The reaction flask
was returned to the cold bath and excess alkyl halide was
added. As the reaction mixture was allowed to warm to
room temperature, the orange color faded and the potas-
sium halide salt precipitated as a white solid, indicating
completion of the reaction. The excess alkyl halide was
conveniently removed with a scavenger resin.8 A variety
of alkyl halides performed well in the alkylation as is
shown in Scheme 4. All gave high yields and C-alkylation
is observed. A spot-to-spot conversion of 13a to 15a -d
1
was observed by TLC and H NMR spectra of the crude
products are consistent with a single diastereomer in
each case. The esters 15a -d were converted to the
corresponding acids 16a -d by treatment with TFA in
DCM, the latter being more crystalline and thus facilitat-
ing greater characterization. When benzyl bromide was
used as the alkylating agent, the product 16d was found
to be contaminated by a small amount of 1,3-bis-benz-
To our pleasant surprise, the first attempts to cyclize
13a and 13b by treatment with potassium tert-butoxide
in THF, warming from -78 °C to room temperature, were
both successful. Indolines 14a and 14b were the major
products, found as single diastereomers. The related
indoles 14a ′ and 14b′ were also produced as minor
SCHEME 3
J . Org. Chem, Vol. 69, No. 7, 2004 2505

Hodges et al.
SCHEME 4
SCHEME 5
ylated product, which is observed in the 1H NMR and
mass spectra. Unlike the cyclization of 13a to 14a , indole
contaminant 14a was not observed in the conversion of
13a to 15a -d . Quenching with an alkylating agent
appears to block this oxidative side reaction.
an efficient and diastereospecific method for the conver-
sion of 13a to 2,3,3-trisubsituted indolines such as 15a -
d . It is convenient to determine the correct amount of
base required and follow the progress of the cyclization
reaction by observing the appearance and disappearance
of the intensely purple enolate of 13a . Indoline 15b is
subsequently converted by routine synthetic transforma-
tions to the diastereomeric spirocyclic lactones 18a and
18b in high yield.
The remaining three steps of the synthesis are shown
in Scheme 5. Acylation of 15b with methyl chloroformate,
using polymer-supported Hunig’s base, afforded 17.
Catalytic dihydroxylation of the allyl group with OsO4-
NMO and removal of the tert-butyl ester with trifluoro-
acetic acid in dichloromethane afforded the spirocyclic
lactone 18 as a 1:1 mixture of diastereomers. The overall
yield of the 7-step process for preparation of 18 from 12
was 40%. The diastereomeric products were separated
by preparative reversed-phase chromatography, affording
the faster eluting diastereomer 18a and the slower
eluting diastereomer 18b. 2D ROESY spectra were found
to be consistent with the relative stereochemistry as
shown for 18a and 18b.
Exp er im en ta l Section
The following solvent and reagent abbreviations are em-
ployed in the experimental methods: dichloromethane (DCM),
N,N-diisopropycarbodiimide (DIC), N,N′-diisopropylurea (DIU),
N-methylmorpholine-N-oxide (NMO), tetrahydrofuran (THF),
and trifluroacetic acid (TFA).
[2-(Ben zylid en ea m in o)p h en yl]a cet ic Acid ter t-Bu t yl
Ester (13a ). Step A: 2-ter t-Bu tyl-1,3-d iisop r op ylisou r ea .7
A mixture of DIC (6.3 g, 50 mmol) and anhydrous t-BuOH (4.3
g, 58 mmol) was treated with CuCl (50 mg, 0.5 mmol) and
stirred 24 h at room temperature. Polyvinylpyridine (1 g) and
DCM (25 mL) were added and the resulting slurry was stirred
for 15 min before filtering and evaporation of DCM at reduced
pressure to afford crude 2-tert-butyl-1,3-diisopropylisourea (8.4
g, 84%) as a colorless liquid that was used without further
Con clu sion s
The results presented herein correct a long-standing
fallacy in preexisting literature. In contrast with previ-
ously published results describing the lack of base-
promoted cyclization with 5, our investigations show that
the KOt-Bu-promoted cyclization of analogous tert-butyl
ester-imines such as 13a and 13b is indeed a feasible
route to 2,3-disubstituted indolines such as 14a and 14b.
Furthermore, cyclization and tandem 3-alkylation offers
1
purification in Step B below. H NMR (CDCl3) δ 1.04 (d, 6H),
1.08(d, 6H), 1.46 (s, 9H), 3.13 (m, 1H), 3.67 (m, 3H), 3.73 (m,
1H).
Step B: (2-Nitr op h en yl)a cetic Acid , ter t-Bu tyl Ester .
A solution of the isourea from above (3.0 g, 15 mmol) and (2-
nitrophenyl)acetic acid (1.81 g, 10 mmol) in DCM (50 mL) were
stirred overnight at room temperature. Hexanes (50 mL) were
added, stirring for 15 min before filtering to remove precipi-
tated DIU. The filtrate was evaporated and the residue was
purified by flash chromatography on silica gel, eluting with
(7) Mathias, L. J . Synthesis 1979, 1979 (8), 561-576.
(8) Booth, R. J .; Hodges, J . C. J . Am. Chem. Soc. 1997, 119 (21),
4882-4886.
2506 J . Org. Chem., Vol. 69, No. 7, 2004

Synthesis of a Spirocyclic Indoline Lactone
hexanes/EtOAc (90:10) to afford (2-nitrophenyl)acetic acid tert-
butyl ester (1.42 g, 60%) as a colorless oil. H NMR (CDCl3) δ
1.44 (s, 9H), 3.94 (s, 2H), 7.34 (d, 1H), 7.45 (t, 1H), 7.58 (t,
1H), 8.09 (d, 1H).9 This procedure was repeated and the
combined products used in Step C below.
Step C: (2-Am in op h en yl)a cetic Acid , ter t-Bu tyl Ester .
A solution of (2-nitrophenyl)acetic acid, tert-butyl ester (3 g,
12.6 mmol) in a mixture of EtOH (50 mL) and THF (50 mL)
was treated with 5% Pd/C (0.3 g) and placed under H2
atmosphere (45 psi) for 4 h. The resulting mixture was filtered
through Celite and evaporated to afford a brown liquid. Flash
chromatograpy on silica gel, eluting with hexane/EtOAc (90:
10), afforded (2-aminophenyl)acetic acid tert-butyl ester as a
yellow oil (2.6 g, 99%). 1H NMR (CDCl3) δ 1.44 (s, 9H), 3.45
(s, 2H), 4.06 (br, 2H), 6.74 (m, 2H), 7.06 (m, 2H).9
2.0 mmol) was added, and the reaction was allowed to warm
to room temperature over 2 h. Trisamine resin (0.5 g, 1.68
mmol) and DCM (5 mL) were added, stirring at room temper-
ature overnight. The resulting mixture was diluted with
hexanes (30 mL) and filtered through a pad of Celite, rinsing
solids with hexanes-EtOAc (1:1). Evaporation of solvent
afforded a yellow oil. Flash chromatography on silica gel,
eluting with hexane-CHCl3 (50:50 to 40:60), provided purified
15a as a pale yellow gum upon evaporation. Drying to constant
weight at 0.25 mmHg, room temperature, afforded 118 mg
(95.5% yield). 1H NMR (CDCl3) δ 0.98 (s, 9H), 1.72 (s, 3H),
4.03 (br, 1H), 4.67 (s, 1H), 6.72 (d, 1H), 6.79 (t, 1H), 7.14 (t,
1H), 7.19 (d, 1H), 7.26 (m, 3H), 7.34 (m, 2H). MS (APCI+) 310
(m + 1). Crystallization of 15a was not achieved. Therefore,
further characterization is provided by removal of the tert-
butyl ester, affording 16a as described below.
Step B: 3-Meth yl-2-p h en yl-2,3-d ih yd r o-1H-in d ole-3-
ca r boxylic Acid (16a ). A solution of 15a (102 mg, 0.33 mmol)
in DCM (6 mL) was treated with TFA (3 mL) and stirred for
22 h at room temperature. The resulting solution was concen-
trated at reduced pressure then dissolved in CHCl3 (10 mL)
and evaporated three times to remove excess TFA. Flash
chromatography on silica gel, eluting with CHCl3-MeOH
(98.5:1.5), gave 16a as a gum upon evaporation of solvent. This
gum was dissolved in DCM, concentrated to a syrup, and
diluted dropwise with pentane to effect crystallization, 56 mg
(67%). 1H NMR (CDCl3) δ 1.72 (s, 3H), 4.84 (s, 1H), 6.79 (d,
1H), 6.84 (t, 1H), 7.11 (d, 1H), 7.17 (t, 1H), 7.28 (m, 3H), 7.41
(m, 2H). MS (APCI+) 254 (m + 1). IR (KBr; cm-1) 1698
(CdO), 3356 (NH). Calcd for C16H15NO2‚0.15H2O: C, 75.07;
H, 6.02; N, 5.47. Found: C, 75.26; H, 5.93; N, 5.45.
3-Allyl-2-p h en yl-2,3-d ih yd r o-1H -in d ole-3-ca r b oxylic
Acid (16b). Allyl bromide was substituted for iodomethane
in the typical procedure above. 1H NMR (CDCl3) δ 2.78 and
2.97 (ABX, 2H), 4.85 (s, 1H), 5.16 (m, 2H), 5.73 (m,1H), 6.75
(d, 1H), 6.82 (t, 1H), 7.18 (m, 2H), 7.17 (t, 1H), 7.25 (m, 3H),
7.30 (m, 2H). MS (APCI+) 280 (m + 1). IR (KBr; cm-1) 1719
(CdO), 3372 (NH). Calcd for C18H17NO2‚0.15H2O: C, 76.65;
H, 6.18; N, 4.97. Found: C, 76.69; H, 5.79; N, 4.80.
2-P h en yl-3-p r op yl-2,3-d ih yd r o-1H -in d ole-3-ca r b oxyl-
ic Acid , Hyd r och lor id e (16c). 1-Iodopropane was substi-
tuted for iodomethane in the general procedure given above.
The product was converted to a hydrochloride salt with
ethereal hydrogen chloride. 1H NMR (DMSO-d6) δ 0.80 (t, 3H),
1.08 (m, 1H), 1.33 (m, 1H), 1.79 (t, 1H), 2.02 (t, 1H), 4.58 (d,
1H), 6.60 (m, 2H), 6.97-7.17 (complex, 7H). MS (APCI+) 282
(m + 1). IR (KBr; cm-1) 1742 (CdO). Calcd for C18H19NO2‚
HCl: C, 68.03; H, 6.34; N, 4.41. Found: C, 67.95; H, 6.46; N,
4.35.
1
Step D: [2-(Ben zylid en ea m in o)p h en yl]a cetic Acid ,
ter t-Bu tyl Ester (13). A mixture of (2-aminophenyl)acetic
acid, tert-butyl ester (0.99 g, 4.8 mmol), benzaldehyde (0.53 g,
5.0 mmol), and trimethyl orthoformate (4 mL) was stirred at
room temperature for 24 h. The mixture was concentrated on
a rotary evaporator then further concentrated under high
vacuum (0.2 mmHg) at room temperature until it became a
wet crystalline mass. Recrystallization from hexanes gave 13
1
as off-white crystals (1.22 g, 86%). H NMR (CDCl3) δ 1.31 (s,
9H), 3.74 (s, 2H), 7.01 (d, 1H), 7.19 (t, 1H), 7.28 (m, 3H), 7.47
(m, 3H), 7.90 (m, 2H), 8.40 (s, 1H). Calcd for C19H21NO2: C,
77.26; H, 7.17; N, 4.74. Found: C, 77.55; H, 7.30; N, 4.48.
Compound 13b was also prepared by the above method. See
the Supporting Information for the 1H NMR spectrum.
2-P h en yl-2,3-dih ydr o-1H-in dole-3-car boxylic Acid, ter t-
Bu tyl Ester (14a ). A septum-sealed flask containing a solu-
tion of imine 13 (75 mg, 0.25 mmol) in dry THF (5 mL) was
evacuated and pressurized with N2 five times. The flask was
then chilled on a dry ice/acetone bath. A solution of 1 M KOt-
Bu in THF (0.25 mL, 0.25 mmol) was added by syringe. The
cold bath was removed and the resulting purple solution was
allowed to warm gradually. After a few minutes the solution
faded to orange. After 20 min, the reaction was quenched with
saturated aqueous NH4Cl (0.5 mL). The reaction mixture was
diluted with a mixture of EtOAc (10 mL) and hexanes (10 mL).
The organic layer was separated, dried over MgSO4, and
evaporated. The residue was purified by flash chromatography
on silica gel, eluting with hexane/EtOAc (97:3) to give 14a as
a colorless film (60 mg) that crystallized on standing. Tritu-
ration with cold pentane afforded a colorless solid (28 mg,
37%). TLC (silica gel, hexanes-EtOAc, 90:10, I2 stain) Rf 0.6.
1H NMR (CDCl3) δ 1.48 (s, 9H), 3.98 (d, 1H), 4.19 (br d, 1H),
5.33 (d of d, 1H), 6.69 (d, 1H), 6.76 (t, 1H), 7.14 (t, 1H), 7.22-
7.38 (m, 4H), 7.47 (d, 2H). MS (APCI+) 296 (m + 1). IR (KBr;
cm-1) 1707 (CdO), 3336 (NH). Calcd for C19H21NO2: C, 77.26;
H, 7.17; N, 4.74. Found: C, 76.84; H, 7.26; N, 4.67.
3-Ben zyl-2-p h en yl-2,3-d ih yd r o-1H -in d ole-3-ca r b oxyl-
ic Acid (16d ). Benzyl bromide was substituted for io-
1
A highly UV-active, minor byproduct (14a ′, 4 mg, 5%) eluted
after the desired product. TLC (silica gel, hexanes-EtOAc,
domethane in the typical procedure above. H NMR (CDCl3)
δ 3.18 (d, 1H), 3.63 (d, 1H), 4.75 (s, 1H), 6.53 (d, 1H), 6.65 (t,
1H), 6.70 (d, 1H), 6.84 (d, 2H), 7.15-7.28 (complex, 9H). MS
(APCI+) 330 (m + 1). IR (KBr; cm-1) 1692,1708 (CdO), 3386
(NH). Contains ∼4% 1,3-dibenzyl-2-phenyl-2,3-dihydro-1H-
1
90:10, I2 stain) Rf 0.25. H NMR lacks H(2) and H(3) seen at
δ 3.98 and 5.33 ppm. MS (APCI+) 294 (m + 1).
Compound 14b was also prepared by the above method. See
the Supporting Information for the 1H NMR spectrum.
Typ ica l Ta n d em Cycliza tion -Alk yla tion P r oced u r e:
3-Met h yl-2-p h en yl-2,3-d ih yd r o-1H -in d ole-3-ca r b oxylic
Acid (16a ). Step A: 3-Meth yl-2-p h en yl-2,3-d ih yd r o-1H-
in dole-3-car boxylic Acid, ter t-Bu tyl Ester (15a). A septum-
sealed flask containing a solution of imine 13 (118 mg, 0.4
mmol) in dry THF (8 mL) was evacuated and pressurized with
N2 five times. The flask was then chilled on a dry ice/acetone
bath. KOt-Bu (1 M) in THF was added dropwise until a purple
color persisted then 1.0 equiv more (0.4 mL) was added. The
cold bath was removed and the flask warmed gradually by
ambient air until the color faded to orange (∼7 min). The flask
was returned to the cold bath for 2 min, iodomethane (125 µL,
1
indole-3-carboxylic acid by H NMR δ 3.21 (d, 0.04H), 3.55 (d,
0.04H), 3.81 (d, 0.04H), 4.35 (d, 0.4H), 4.63 (s, 0.04H) and MS
420 (m + 1).
5-(Hyd r oxym eth yl)-2-oxo-2′-p h en yl-4,5-d ih yd r osp ir o-
[fu r a n -3,3′-in d ole]-1′(2′H)ca r boxylic Acid , Meth yl Ester
(18). Step A: 3-Allyl-2-p h en yl-2,3-d ih yd r o-n d ole-1,3-d i-
ca r boxylic Acid , 3-ter t-Bu tyl Ester , 1-Meth yl Ester (17).
A solution of 15b (201 mg, 0.60 mmol) in DCM (2.5 mL) was
added to PS-DIEA (0.5 g, 1.83 mmol). The resulting slurry was
treated with methyl chloroformate (51 µL, 0.66 mmol). The
reaction was stirred gently in a capped flask. After 3 h the
addition of methyl chloroformate was repeated. After another
3 h PS-trisamine (150 mg, 0.5 mmol) was added, stirring 1 h
further before resins were filtered and rinsed several times
with DCM. The combined filtrate and washings were evapo-
rated and the residual gum was triturated with t-BuOMe. The
(9) NMR spectra are consistent with: Flitsch, W.; Russkamp, P.
Liebigs Ann. Chem. 1985, 1985 (7), 1398-1412.
J . Org. Chem, Vol. 69, No. 7, 2004 2507

Hodges et al.
resulting slurry was evaporated to give 17 as an off-white solid
(195 mg, 82.7%). H NMR (CDCl3) δ 0.98 (s, 9H), 1.57 (s,3H),
2.59 (m, 1H), 2.88 (m, 1H), 3.62 (br s, 3H), 5.02 (m, 3H), 5.59
(m, 1H), 7.0-7.6 (complex, 9H). MS (APCI+) 393 (m - tBu +
1). This material was used directly in Step B below without
further purification.
of d, 0.5H), 2.64 (d of d, 0.5H), 2.68 (d of d, 0.5H), 2.82 (d of d,
0.5H), 3.6-3.8 (m + br, 4H), 4.04 (m, 1H), 4.74 (m, 0.5H), 4.84
(m. 0.5H), 5.43 (s, 0.5H), 5.59 (s, 0.5H), 7.11 (m, 2.5H), 7.15
(m, 0.5H), 7.29 (m, 4H), 7.39 (m, 1H), 7.87 (br s, 1H). MS
(APCI+) 354 (m + 1). Analytical HPLC shows two diastere-
omers of equal amounts: Stationary phase Xterra 150 × 4.6
mm, C18. Mobile phase MeCN-H2O buffered with 0.1%
HCO2H, linear gradient 30% to 50% MeCN over 20 min, flow
1.0 mL/min. 18a retention time 8.83 min. 18b retention time
9.78 min. They were separated by preparative HPLC: Station-
ary phase Xterra 150 × 30 mm, C18. Mobile phase MeCN-
H2O buffered with 0.1% HCO2H, linear gradient 30% to 50%
MeCN over 20 min, flow 35 mL/min. For 18a : 1H NMR
(CDCl3) δ 1.88 (br t, 1H), 2.45 (d of d, 1H), 2.68 (d of d, 1H),
3.72 (br s, 3H), 4.02 (d, 1H), 4.84 (m, 1H), 5.43 (s, 1H), 7.11
(m, 3H), 7.29 (m, 4H), 7.39 (m,1H), 7.87 (br s, 1H). HRMS:
calcd 354.1341; found (ESI MeCN/H2O) 354.1340 (m + 1).
Calcd for C20H19NO5‚0.5H2O: C, 66.19; H, 5.57; N, 3.86.
Found: C, 65.81; H, 5.29; N, 3.76. For 18b: 1H NMR (CDCl3)
δ 1.91 (t, 01H), 2.64 (d of d, 1H), 2.82 (d of d, 1H), 3.69 (br s,
3H), 3.78 (m, 1H), 4.04 (m, 1H), 4.74 (m, 1H), 5.59 (s, 1H),
7.11 (m, 2H), 7.15 (m, 2H), 7.27 (m, 4H), 7.37 (m,1H), 7.87 (br
s, 1H). HRMS: calcd 354.1341; found (ESI MeCN/H2O)
354.1339 (m + 1). Calcd for C20H19NO5‚H2O: C, 64.65; H, 5.70;
N, 3.77. Found: C, 64.27; H, 5.24; N, 3.71.
1
St ep B: 3-(2,3-Dih yd r oxyp r op yl)-2-p h en yl-2,3-d ih y-
d r oin d ole-1,3-d ica r boxylic Acid , 1-Meth yl Ester . A solu-
tion of 17 (115 mg, 0.29 mmol) in a mixture of acetone (4 mL)
and H2O (1 mL) was treated with OsO4 (0.075 M in t-BuOH,
125 µL, 9 µmol) and NMO (68 mg, 0.58 mmol). The reaction
was stirred 5 h at room temperature then concentrated at
reduced pressure. The residue was treated with 10% aqueous
NaHSO3 (6 mL) and EtOAc (25 mL), stirring vigorously for 5
min. The organic layer was separated, washed successively
with 1 N HCl, water, and brine, dried over MgSO4, and
evaporated. The residue was dissolved in CHCl3 and evapo-
rated again, drying under vacuum to constant weight to afford
the crude glycol. 1H NMR is complex due to the presence of
two diastereomers (see the Supporting Information). MS
(APCI+) 372 (m + 1), 354 (m - H2O + 1).
Step C: 5-(Hyd r oxym eth yl)-2-oxo-2′-p h en yl-4,5-d ih y-
dr ospir o[fu r an -3,3′-in dole]-1′(2′H)-car boxylic Acid, Meth -
yl Ester (18, 18a , 18b). The crude glycol from Step B above
(0.29 mmol) was dissolved in DCM (2 mL) and treated with
TFA (1 mL). The resulting solution was stirred at room
temperature for 4 h, diluted with DCM (10 mL), and evapo-
rated. The residue was dissolved in DCM (10 mL) and
evaporated twice to afford a brown gum. Filtration through a
pad of silica gel, eluting with CHCl3-MeOH (99:1), removed
a trace of TLC baseline material from the major product.
Evaporation gave 18 as a colorless foam that was dried to
constant weight under vacuum (103 mg, 100% from 17). TLC
(silica gel, CHCl3-MeOH, 99:2, I2 stain) shows a single spot,
Rf 0.3. 1H NMR (CDCl3) δ 1.88 (t, 0.5H), 1.91 (t, 0.5H), 2.45 (d
Su p p or tin g In for m a tion Ava ila ble: 1H NMR spectra for
compounds 13a ,b, 14a ,b, 16a -d , 17 and related glycol, 18,
and 18a ,b; HPLC traces for the separation of 18a and 18b;
and 2D ROESY spectra for compounds 16a , 18a , and 18b. This
material is available free of charge via the Internet at
http://pubs.acs.org.
J O0499569
2508 J . Org. Chem., Vol. 69, No. 7, 2004