Synthetic studies toward the partial ergot alkaloid LY228729, a potent 5HT(1A) receptor agonist

8640

J . Org. Chem. 1997, 62, 8640-8653

Syn th etic Stu d ies tow a r d th e P a r tia l Er got Alk a loid LY228729, a

P oten t 5HT_1ARecep tor Agon ist

M. A. Carr, P. E. Creviston, D. R. Hutchison, J . H. Kennedy, V. V. Khau, T. J . Kress,*

M. R. Leanna, J . D. Marshall, M. J . Martinelli,* B. C. Peterson, D. L. Varie,* and

J . P. Wepsiec

Chemical Process R&D, Lilly Research Laboratories, A Division of Eli Lilly and Company,

Lilly Corporate Center, Indianapolis, Indiana 46285-4813

Received J uly 11, 1997^X

Synthetic studies on LY228729 (3) afforded two innovative approaches for the construction of this

class of partial ergoline 5HT_1areceptor agonists. The first synthesis is based upon a diastereose-

lective epoxidation of racemic olefin 5, followed by ring opening and covalent resolution to furnish

the key amino alcohol 8. Aziridination of amino alcohol 8, with subsequent tandem hydrogenolysis

of the benzylic aziridine and auxiliary bonds, provided access to the optically active primary amine

13. A novel catalytic carboxamidation reaction installed the requisite side chain. Alternatively,

the chiral pool was drawn upon for the single stereogenic center by virtue of ^L-tryptophan, albeit

by a more circuitous route than expected. ^L-Tryptophan was differentially protected and reduced

to the indoline diastereomers 26a ,b which were separated by fractional crystallization. The two

indoline diastereoisomers were independently cyclized by a Friedel-Crafts protocol, which under

thermodynamic control afforded enantiomeric ketones 30a . The ketone was deoxygenated with a

two-step reduction protocol to intersect the initial route and complete the second total synthesis.

The two routes offer complementary access to this exciting class of partial ergot alkaloids.

In tr od u ction

istry, was not sufficiently efficient for larger scale pur-

suit.¹The second- and third-generation syntheses are

discussed here in detail, with particular focus on some

of the key transformations that enabled the synthesis of

LY228729 for clinical trials. Also, based upon the J a-

cobsen catalytic asymmetric epoxidation,²it was possible

to achieve kinetic resolution in the epoxidation of olefin

5 and hence afford a fourth-generation synthetic option.

However, competing aromatization proved to be a dif-

ficult problem in this asymmetric epoxidation and it will

not be discussed further.

Serotonin is an important neurotransmitter, respon-

sible for signal transduction and ultimately, for example,

muscle control. The receptors for serotonin are account-

able for a variety of functions and have thus been the

target for a number of therapeutic reagents. The sero-

tonin (5-HT, 1) receptors are divided into families and

subtypes, including 5HT₁, 5HT₂, 5HT₃, 5HT₄, 5HT₅,

5HT₆, and 5HT₇. Both agonists and antagonists for many

serotonin receptors have been identified and developed

into valuable therapeutic reagents. Lysergic acid diethyl-

amide (LSD, 2) is an agonist for serotonin receptors, but

is not very specific at any given receptor. Although this

activity may not be surprising, the lack of specificity is

intriguing on the basis of its conformational rigidity. The

minor structural similarities of an indole tethered with

an amine moiety suggested that the conformationally

more fixed compound 3 should provide unique activity.

SAR revealed the optimal structure with a di-n-propyl-

amino substituent and the 5-carbamoyl moiety in

LY228729 (3) for 5HT_1areceptor agonism. Thus, this

molecule became the target drug candidate for further

preclinical and clinical evaluations requiring synthetic

methods for its construction.

Secon d -Gen er a tion Syn th etic Rou te. Dia ster eo-

selective Ep oxid a tion . The first total synthesis of

lysergic acid was published by Kornfeld, Woodward, and

colleagues from the Lilly Research Laboratories in 1954.³

This elegant campaign was originally meant to address

the difficulty in obtaining meaningful quantities from

nature and stood the test of time for several decades

which followed as the sole total synthesis. A critical

intermediate in the Kornfeld-Woodward synthesis was

a compound which became affectionately known as

Kornfeld’s ketone (4). NaBH₄reduction of the ketone and

subsequent dehydration afforded the crystalline olefin 5

in excellent overall yield (Scheme 1). The epoxidation

of this olefin with peracids was known since the early

lysergic acid synthesis days of Kornfeld. However, the

stereochemical outcome of this epoxidation remained

unknown for the four decades that followed the first

disclosure. In any event, these molecules were success-

(1) (a) Flaugh M. E.; Murdoch G. L. Total Synthesis of LY228729, a

Novel 5-HT_1AAgonist, presented at the 20th National Meeting of the

American Chemical Society, Washington, DC; August 26, 1990. (b) For

the synthesis of racemic LY228729, see: Flaugh, M. E.; Mullen, D. L.;

Fuller, R. W.; Mason, N. R. J . Med. Chem. 1988, 31, 1746.

(2) J acobsen, E. N.; Zhang, W.; Muci, A. R.; Ecker, J . R.; Deng, L.

J . Am. Chem. Soc. 1991, 113, 7063.

(3) Kornfeld, E. C.; Fornfeld, E. J .; Kline, G. B.; Mann, M. J .;

Morrison, D. E.; J ones, R. G.; Woodward, R. B. J . Am. Chem. Soc. 1956,

78, 3087.

The first generation synthesis was designed for optimal

versatility in order to access a wide variety of analogues.

However this approach, involving amino tetralin chem-

(4) Nichols, D. E.; Robinson, J . M.; Li, G. S.; Cassady, J . M.; Floss,

H. G. Org. Prep. Proc. Int. 1985, 17, 391.

^XAbstract published in Advance ACS Abstracts, November 1, 1997.

Partial Ergot Alkaloid LY228729

J . Org. Chem., Vol. 62, No. 25, 1997 8641

Sch em e 1

fully employed as valuable intermediates in a total

synthesis of that complex natural product.

In late 1989, we had the opportunity to reinvestigate

the epoxidation of the olefin,⁴believing that a mixture

of diastereomeric epoxides (eg., 6, 7) would result. To

our surprise, we determined that the epoxidation of olefin

5 was highly diastereoselective, affording primarily the

anti-epoxides 7 with >96% de.⁵The rationale for this

result was based upon a torsional model combined with

transition state modeling.⁶The importance of this

insight into the chemical process suggested that the

epoxidation stereochemistry should be reagent indepen-

dent. In other words, m-CPBA, NBS/H₂O, MMPP (mono-

magnesium peroxyphthalate) or OsO₄should provide the

same stereofacially selective oxidation with respect to

selectivity sense and magnitude. That this was indeed

the case allowed us the flexibility to choose the optimal

oxidant with confidence that the reaction could be reliably

scaled-up.

Interestingly, m-CPBA was employed in the initial

epoxidation studies with great success in a variety of

solvents including CH₂Cl₂, toluene, and ethereal solvents.

During that time, the availability of relatively pure

m-CPBA became difficult, being replaced with a some-

what more stable but less pure grade of the reagent

(∼50% potency). Also coincident with this transition was

the introduction of monomagnesiumperoxyphthalate

(MMPP)⁷as a potential replacement of the former

oxidant. Since we believed the epoxidation was reagent

and solvent independent, our incorporation of this new

reagent into the process met only with the difficulties

inherent to using the reagent itself. Namely, the relative

insolubility of MMPP limited our solvent choice. None-

theless, this resulted in the development of a 50%

aqueous n-BuOH biphasic reaction medium. With sub-

strate and reagent soluble under these conditions, the

reaction proceeded smoothly. Upon complete reaction,

the layers were separated to effect efficient removal of

the spent oxidant. The organic phase contained the

desired epoxide with an isomer profile identical with that

derived from m-CPBA. The solvent choice resulted partly

from a survey but equally importantly from consideration

of how the epoxidation reaction would dovetail into the

subsequent process (vide infra).

was previously studied by Kornfeld and shown to be

regioselective for attack at the 4-position.³Thus, we

concluded that amine would open epoxide 7 to afford

amino alcohol 10, which, in principle, could be deoxy-

genated to provide the correct relative stereochemistry

(eq 1). However, it was disappointing (but rational)^8a

that attack occurred at the more electrophilic benzylic

position to afford the amino alcohol 8.^8bContrasted with

the earlier Kornfeld results with trisubstituted epoxides,

the amine opening of these disubstituted epoxides occurs

at the less hindered side. A wide variety of 1° and 2°

amines were reported to open this racemic epoxide in

excellent yield. These epoxide openings were best con-

ducted in n-BuOH at 110 °C, thus fitting very well with

the epoxide-forming step above in the same solvent.

Consequently, a solution of the racemic epoxide when

reacted with an optically pure amine, such as (S)-R-

phenethylamine, produced an equal mixture of diaster-

eomers 8 and 9. This mixture upon cooling provided the

single isomer 8^8bin 43% yield, thus affording an efficient

means for separation of optically enriched material.

Simple reslurry of this crude material in warm 2-pro-

panol effectively removed the more soluble undesired

isomer 9.

As noted above, it was disappointing due to the

structural attributes of the target molecule that the

epoxide opening had occurred at the benzylic position.

Transposition of the nitrogen around the ring through

the intermediacy of the aziridine 12 was then considered.

Initially, a Mitsunobu reaction⁹furnished this aziridine

in high yield, but concomitant with the production of

highly crystalline contaminants, triphenylphosphine ox-

ide, and the reduced form of DEAD (diethyl) or DIAD

(diisopropyl azodicarboxylate) (eq 2). While chromato-

The epoxide opening of trisubstituted epoxide sub-

strates (related to 7) with a variety of amine nucleophiles

(5) Leanna, M. R.; Martinelli, M. J .; Varie, D. L.; Kress, T. J .

Tetrahedron Lett. 1989, 30 3935.

(6) Martinelli, M. J .; Peterson, B. C.; Khau, V. V.; Hutchison, D. R.;

Leanna, M. R.; Audia, J . E.; Droste, J . J .; Wu, Y-D.; Houk, K. N. J .

Org. Chem. 1994, 59, 2204.

(7) Brougham, P.; Cooper, M. S.; Heaney, H.; Thompson, N. Syn-

thesis 1987, 1015.

graphic removal of these impurities was feasible on a

8642 J . Org. Chem., Vol. 62, No. 25, 1997

Carr et al.

small scale, it was less likely at the larger scale. The

aziridine could likewise be formed with Ph₃PCl₂, Ph₃PBr₂,

and other similar reagents but with the same purification

liabilities. A new procedure employing methanesulfonyl

chloride and triethylamine was serendipitously discov-

ered and then implemented to overcome this purification

requirement, securing the aziridine 12 as a viable

intermediate. That mesyl chloride could be utilized for

this transformation is remarkable since N-mesylation

was the expected primary course of reaction. It was not

possible to isolate any sulfonamide derivative, nor me-

sylate from this reaction. However, application of this

protocol to substrates without N-alkyl branching resulted

only in N-mesylation.

To introduce the requisite carboxamide moiety, several

options were available. It was expected that Friedel-

Crafts acylation followed by amide formation would be

problematic. However, the bromide moiety offered sev-

eral possibilities for aromatic substitution or metal-

catalyzed replacement. Rosenmund-von Braun reaction

of the bromide 14c with CuCN in NMP (N-methylpyr-

rolidinone) at 200 °C gave the nitrile 15a in 76% yield

(eq 4). An alternative to NMP included DMF at reflux,

but the reaction was much slower even in the presence

of CuI. A competing side reaction was the cleavage of

an N-propyl group under these extreme conditions.

Conversion of the nitrile into the corresponding amide

15b was accomplished quite smoothly with neat poly-

phosphoric acid (PPA) at 90 °C in excellent yield.

Although this reaction was run quite concentrated, large

volumes of water were required in the workup procedure

to break the intermediate complex. This latter fact

presented us with a throughput issue, suggesting a

search for alternatives (vide infra).

Tandem benzylic hydrogenolysis of the aziridine bond

and the auxiliary bond was achieved by exposure to

hydrogen in the presence of a palladium catalyst. Thus,

cleavage of the aziridine bond occurred at 0 °C under 1

atm of H₂gas. Upon complete conversion to the 2° amine,

the benzylic auxiliary bond was cleaved at 55 °C in the

same vessel. The only detectable side product from these

reactions was the desamino compound in which the

exocyclic nitrogen was completely lost. This unwanted

pathway could be minimized by careful temperature

control to conduct the endocyclic reductive ring opening

at low temperatures, followed by auxiliary cleavage at

the elevated temperature. To prevent competing sol-

volytic reactions, nucleophilic solvents such as acetic acid

or alcohols were avoided. With those solvents, aziridine

opening could be observed to afford the corresponding

acetate or methyl ether derivatives. The optimal binary

solvent medium selected for this hydrogenolysis was

therefore THF and phosphoric acid. To further prevent

hydrolysis of the benzamide protecting group, the pH was

adjusted to 3 prior to temperature elevation. Under these

conditions, both diastereomeric amino alcohols could be

converted to the enantiomers of 13 with equal but

opposite signs of rotation ([R]_D) 59° (c ) 1, THF)).

Regioselective aromatic electrophilic para-substitution

on an indoline moiety is well precedented,¹⁰although

many reaction conditions were not compatible with the

existing functionality. The mildly directing and activat-

ing effects of the p-benzamide and o-alkyl moieties were

essential for regiocontrol. The primary amine 13 thus

obtained could undergo aromatic electrophilic substitu-

tion reaction with bromine in a carefully buffered NaOAc/

HOAc¹⁰reaction medium, thereby preventing oxidation

of the amine moiety, to afford the aryl bromide 14a in

89% yield (eq 3). Alternatively, iodination with iodine

and periodic acid furnished the aryl iodide 14b in 85%

yield.¹¹Subsequent alkylation with excess iodopropane

and K₂CO₃in acetonitrile at 80 °C afforded the tertiary

amine 14d in 80% yield. Quaternization of the amine

group, rather than incomplete alkylation, was the pre-

ferred option since the quaternary ammonium salt could

be easily removed by aqueous extraction.

Deprotection of the nitrile 15a could be accomplished

with a variety of reagents. In the early phases of the

project during an attempt to transmetalate the aryl

bromide, we noted the facile amide cleavage with n-BuLi.

In fact, the cleavage was so efficient that we actually

implemented this protocol to prepare initial quantities

of the desired indoline 16a . On a small scale, however,

we also noted the indole as the major byproduct. This

could perhaps be formed from autoxidation of a dipole-

stabilized R-anion and was noted in the Rebek lysergic

acid synthesis.¹²The suspected autoxidation could not

be developed into an efficient process for net oxidation.

(8) (a) Nedelec, L; Brown, N. L.; Plassard, G. French Patent FR2522-

658-A 82 03726. (b) Martinelli, M. J .; Leanna, M. R.; Varie, D. L.;

Peterson, B. C.; Kress, T. J .; Wepsiec, J . P.; Khau, V. V. Tetrahedron

Lett. 1990, 31, 7579.

(9) Mitsunobu, O. Synthesis 1981, 1.

(10) (a) Russell, H. F.; Harris, B. J .; Hood, D. B.; Thompson, E. G.;

Watkins, A. D.; Williams, R. D. Org. Prep. Proc. Int. 1985, 17, 391. (b)

J ohnson, H. E.; Crosby, D. G. J . Org. Chem. 1963, 55, 2794. (c) Borror,

A. L.; Chinoporos, E.; Filosa, M. P.; Herchen, S. R.; Petersen, C. P.;

Stern, C. A.; Onan, K. D. J . Org. Chem. 1988, 53, 2047.

(11) (a) Suzyuki, H. Org. Synth. 1971, 51, 94. (b) Suzuki, H. Bull

Chem. Soc. J pn. 1970, 43, 481. (c) Suzuki, H.; Nakamura, K.; Goto, R.

Bull. Chem. Soc. J pn. 1966, 39, 128.

(12) (a) Rebek, J .; Tai, D. F.; Shue, Y. K. J . Am. Chem. Soc. 1984,

106, 1813-1819. (b) Rebek, J .; Shue, Y. K.; Tai, D. F. J . Org. Chem.

1984, 49, 3540-3545.

Partial Ergot Alkaloid LY228729

J . Org. Chem., Vol. 62, No. 25, 1997 8643

Ta ble 1. P a lla d iu m -Ca ta lyzed Ca r boxa m id a tion of

Iod id e 14d

protocol seemed reasonable at first, but the workup

revealed some hidden troubles. Filtration of the man-

ganese waste at the end of the reaction was problematic,

due presumably to a myriad of acetate salts. The filtrate

obtained after these lengthy separations contained a new

species (18), derived from N-depropylation. Employing

Mn(OAc)₂or Mn(OAc)₃itself as the primary oxidant also

proved successful and avoided dealkylation since it is a

much milder oxidant.¹⁶

palladium

source

added ligand

(equiv)

reaction

time (h)

% yield,

amide 15b

Pd(PPh₃)₂Cl₂

Pd(PPh₃)₂Br₂

Pd black

6

5

14

8

66

76

89

84

PPh₃(4)

Ph₂P(CH₂)₃PPh₂(2)

Pd black

Later, it was found that the more conventional KOH/

EtOH conditions were well-suited for benzamide cleav-

age, and these were the conditions ultimately imple-

mented for our large scale work. The deprotection

workup incorporated an acid-base extraction to separate

the benzoic acid and neutral impurities. Thus, depro-

tection could occur at either the nitrile or amide stage

with equal efficiency to provide a substrate compatible

with subsequent oxidation conditions.

Alternatives to the manganese-based oxidations in-

cluded DDQ,¹⁷which was less effective at completing the

oxidation. Swern oxidation¹⁸of the indoline provided the

indole very efficiently and concomitantly dehydrated the

amide to the nitrile. Palladium on carbon¹⁹in MeOH at

reflux smoothly transformed the indoline to the indole

in 75-85% yield. As with the manganese procedure,

scale-up proved that N-dealkylation was the major

byproduct. Potentially, coordination with either nitrogen

could lead to an oxidative insertion across a carbon-

nitrogen bond. Reductive elimination would then provide

the desired indole, or the undesired propyl-cleaved

secondary amine. Additional new components in the

reaction mixture were characterized as the transposed

“Uhle’s” ketone 19 and the saturated compound 20.

These were rationalized as arising from oxidative addi-

tion to form an enamine, which produced the ketone. This

ketone, when isolated and resubjected to the reaction

conditions, afforded the corresponding saturated mate-

rial. The reaction profile varied with various lots of

catalyst. Catalyst water content and age played a

significant role in securing the optimal catalytic system.

Isolation of LY228729 as the hippurate salt²⁰offered a

convenient method for purification by recrystallization

(Scheme 2).

Tr yp top h a n Rou te. ^L-Tryptophan (22) contains the

carbon framework and an appropriately positioned ste-

reocenter, with the desired absolute configuration needed

for LY228729. In 1984, Rebek reported the synthesis of

lysergic acid and its C-4 epimer (23) from ^L-tryptophan.¹²

These factors and the ready availability of L-tryptophan

lead us to pursue it as a starting material for the

synthesis of LY228729.²¹On the basis of Rebek’s work,

we believed there was a high probability that ^L-tryp-

tophan could be converted to tricyclic amine 21 and thus

intersect with the synthetic route described in the previ-

ous section (eq 5). As will be discussed, we were less

certain this “chiral pool” based synthesis could be devel-

oped into an efficient, manufacturable process.

Two alternative strategies for introduction of the

primary amide functionality utilized Heck carbonylation

technology. Two potential precursors to amide 15b were

envisioned to be esters or secondary amides, which could

be prepared by the reaction of an aryl halide with carbon

monoxide in the presence of a palladium catalyst (eq 4).¹³

While the aryl bromide 14c proved to be an unreactive

substrate for carbonylation, the iodide 14d reacted with

carbon monoxide and Pd(PPh₃)₂Cl₂catalyst in methanol

and ethanol to give methyl and ethyl esters 15c,d in high

yields. Both esters were extremely resistant toward

amidation with ammonia. Under forcing conditions

(NaNH₂) only cleavage of the benzoyl group occurred. We

were pleased to find, however, that iodide 14d smoothly

underwent carboxamidation in the presence of carbon

monoxide/ammonia with 0.05 equiv of Pd(PPh₃)₂Cl₂to

give the desired primary amide 15b in 65-75% yield. A

variety of palladium sources gave acceptable results in

this reaction (Table 1). Fortunately a very convenient

catalyst precursor, palladium black with added triphen-

ylphosphine, gave the best yields (consistently greater

than 80%) of 15b. Toluene was found to be an optimal

solvent for this process as the residual palladium was

easily removed by filtering the warm reaction mixture,

and the amide product was then crystallized directly from

the cooled reaction mixture. We believe this reaction was

the first reported example of preparation of a primary

amide via this methodology.¹⁴

With access to the optically enriched indoline, our

attention was now focused on the indole oxidation. The

literature offers many suggestions for this thermody-

namically favored conversion.¹⁵Manganese dioxide in

CH₂Cl₂was initially quite effective but required a large

excess of reagent, often >25 molar equiv. Solvent

exchange to acetic acid resulted in a dramatic rate

acceleration even at 1.5-2 equiv of MnO₂. This oxidative

(13) (a) Schoenberg, A.; Bartoletti, I.; Heck, R. J . Org. Chem. 1974,

39, 3318. (b) Schoenberg, A.; Heck, R. J . Org. Chem. 1974, 39, 3327.

(14) Kress, T. J .; Wepsiec, J . P. U.S. Patent 5039820, 1991.

(15) Sundberg, R. J . In The Chemistry of Indoles; Academic Press:

New York, 1970.

8644 J . Org. Chem., Vol. 62, No. 25, 1997

Carr et al.

Sch em e 2. Syn th esis of LY228729 fr om Olefin 5

(“Kor n feld Keton e Rou te”)^a

Sch em e 3. Red u ction a n d P r otection of

L-Tr yp top h a n ^a

a

(a) CF₃CO₂Et, MeOH, NEt₃; (b) Et₃SiH, TFA; (c) PhCOCl,

NEt₃, THF; (d) CHCl₃crystallization; (e) 1.5% PtO₂, H₂, H₂O, TFA;

2. NaHCO₃, PhCOCl; 3. CHCl₃.

hydrogenation had processing advantages over the ionic

hydrogenation, namely that indolines 25a ,b were not

isolated. Upon complete hydrogenation, the PtO₂was

filtered and aqueous NaHCO₃was added to the reaction

mixture followed by PhC(O)Cl. This process was coupled

with the CHCl₃fractional crystallization to provide

indoline 26a in 18% overall yield from 24. Notably, when

the hydrogenation was performed in acetic acid or 1 N

HCl, extensive overreduction of the aromatic ring oc-

curred (eq 6). Even with the TFA/H₂O reaction medium,

the reaction had to be carefully monitored to prevent

overreduction. Unfortunately, regardless of which indole

reduction and benzoylation sequence was used, consistent

crystallization of acid 26a was difficult (even with seed-

ing) and CHCl₃proved to be the optimal crystallization

solvent.

a

(a) MMPP, n-BuOH; (b) (S)-phenethylamine, n-BuOH; (c)

crystallize from i-PrOH; (d) MsCl, NEt₃; (e) 10% Pd-C, H₂, THF/

H₃PO₄; (f) H₅IO₆, I₂; (g) n-PrI K₂CO₃; (h) CO, NH₃, Pd black, PPh₃;

(i) NaOH, EtOH; (j) 10% Pd-C, MeOH; (k) hippuric acid.

Red u ction a n d P r otection of L-Tr yp top h a n . We

planned to form the C-ring of LY228729 using an

intramolecular Friedel-Crafts acylation reaction, as

precedented by Rebek. The N-protected indoline²²we

required was prepared by trifluoroacetylation²³of L-

tryptophan, followed by ionic hydrogenation of the C-2

double bond (TFA/Et₃SiH) and benzoylation of the indo-

line nitrogen (Scheme 3). Using this scheme, a 45:55

mixture of diastereomeric indolines (26a :26b) was ob-

tained in 67% yield from 24. Fractional crystallization

of the mixture from CHCl₃provided what was eventually

shown to be indoline diastereomer having the R C-2a

hydrogen (26a ) with >96% de in 21% overall yield from

24.

The Et₃SiH reduction required TFA as solvent. Al-

though convenient for small scale work, more economical

catalytic hydrogenation methods were investigated. Op-

timal catalytic hydrogenation conditions for 24 were

found to be effected by 5% PtO₂catalyst in water with

10-20 equiv of TFA at 10 psi of hydrogen. The catalytic

F r ied el-Cr a fts Acyla tion Rea ction a n d Ster eo-

ch em istr y. Preserving the configuration of the C-4

stereocenter during the Friedel-Crafts acylation reaction

was essential for using this strategy to prepare LY228729.

While several enantiomerically pure R-amido 1-tetralones

have been prepared via Friedel-Crafts cyclizations,²⁴

Rebek’s work suggested the desired tricyclic ketone and/

or the cyclization precursor (e.g. acid chloride, activated

ester) would be more prone to epimerization at C-4. As

shown in Scheme 4,¹²the azalactones derived from indo-

(16) Ketcha, D. Tetrahedron Lett. 1988, 29, 2151.

(17) Hagen, T. J .; Narayanan, K.; Names, J .; Cook, J . M. J . Org.

Chem. 1989, 54, 2170.

(18) Keirs, D.; Overton, K. J . Chem. Soc., Chem. Commun. 1987,

1660.

(19) Kiguchi, T.; Kuninobu, N.; Takahashi, Y.; Yoshida, Y.; Naito,

T.; Ninomiya, I. Synthesis 1989, 778.

(20) Kress, T. J .; Varie, D. L. U.S. Patent 5397799, 1995.

(21) For

a preliminary account of ths work, see: Varie, D. L.

Tetrahedron Lett. 1990, 31, 7583.

(24) Examples of intramolecular acylations of R-amino acid

derivatives: (a) Melillo, D. G.; Larsen, R. D.; Mathre, D. J .; Shukis,

W. F.; Wood, A. W.; Colleluori, J . R. J . Org. Chem. 1987, 52, 5143-50.

(b) Buckley, T. F.; Rapoport, H. J . Org. Chem. 1983, 48, 4222-4232.

(c) McClure, D. E.; Lumma, P. K.; Arison, B. H.; J ones, J . H.; Baldwin,

J . J . J . Org. Chem. 1983, 48, 2675-2679. (d) McClure, D. E.; Arison,

B. H.; J ones, J . H.; Baldwin, J . J . J . Org. Chem. 1981, 46, 2431-2433.

(22) Friedel-Crafts cyclizations of indole propionic acid derivatives

typically occur on the 2-position rather than the 4-position. For leading

references and a report of Friedel-Crafts cyclization conditions of

indole propionic acids which occur on the 4-position, see: Teranishi,

K; Hayashi S.; Nakatsuka, S.; Goto, T. Synthesis 1995, 506.

(23) Curphey, T. J . J . Org. Chem. 1979, 44, 2805-2807.

Partial Ergot Alkaloid LY228729

J . Org. Chem., Vol. 62, No. 25, 1997 8645

Ta ble 2. Isom er iza tion of Mixtu r es of Keton es 30a /30b

Sch em e 4. F r ied el-Cr a fts Cycliza tion

P r eced en t^12,a

initial

ratio^a

final

ratio

30a :30b

entry 30a :30b

conditions^b

1

2

3

4

63:37

4 equiv of AlCl₃powder;^cCH₂Cl₂; 24 h

4 equiv of AlCl₃powder; CH₂Cl₂; 55 h

4 equiv of AlCl₃flakes;^dCH₂Cl₂; 55 h

4 equiv of AlCl₃powder; 0.05 equiv DMF; 85:15

CH₂Cl₂; 55 h

66:34

69:31

73:27

5

6

7

76:24

2 equiv of NEt₃; CH₃CN; 1 h

1 equiv NaHCO₃; CH₃CN/H₂O: 24 h

1 equiv 1 N HCl; CH₃CN; 24 h

98:2

77:23

a

b

Determined by HPLC (see Experimental Section). All reac-

c

d

tions performed at 22 °C. Aldrich 99.9%. Witco technical grade.

tions, we were pleased to find that a 42:58 ratio of ketones

(+)-30a :30b was formed. However, chiral HPLC assay

of the crude reaction product showed that the enantio-

meric purity of ketone (+)-30a in this crude mixture was

unacceptable (82% ee). When the reaction was repeated

at 0 °C/27 h both ketone diastereomers were again

isolated, with ketone (+)-30a having 91% ee. In this

experiment 50% of the carboxylic acid starting materials

26a , 26b was recovered as well. This reaction essentially

stopped after 4 h at which time the reaction became

heterogeneous. Additionally, the majority of the loss in

enantiomeric purity of ketone (+)-30a (from 98.5% ee to

93% ee) occurred over the first 4 h (while the reaction

mixture was homogeneous). The amorphous residue that

precipitated from the reaction mixture appeared to be

aluminum complexes of the acid chloride and the ketone

product. Although performing the reaction at 0 °C slowed

the rate of epimerization at C-4, these reaction conditions

were not synthetically useful.

a

(a) Ac₂O, 100 °C; (b) 4 equiv of AlCl₃, EtCl₂, 80 °C.

Sch em e 5^a

Subsequent experiments showed that ketone diaste-

reomer (+)-30a is the themodynamically favored dias-

tereomer and that epimerization of 30b to (-)-30a during

the reaction is the likely cause for the decrease in ee of

(+)-30a (Scheme 6).²⁵When mixtures of ketones 30a :

30b were treated with 4 equiv of AlCl₃and 0.05 equiv of

DMF in CH₂Cl₂at 22 °C, ketone 30b isomerized to 30a

with a half-life of approximately 33 h (see Table 2). The

fact that a catalytic amount of DMF increased the rate

of epimerization (compare entries 2 and 4) indicated that

a variety of reaction parameters, such as the amount of

HCl present and AlCl₃solubility, may be significant for

isomerization catalysis. (Some differences in the rates

of both the cyclization and epimerization reactions were

also observed when different grades of AlCl₃were used.)

Ketone 30b was more rapidly (2 h) isomerized with 1

equiv of triethylamine in CH₂Cl₂and was stable in the

a

(a) 2 equiv of (COCl)₂, cat. DMF, -5 °C, CH₂Cl₂; (b) 4 equiv of

AlCl₃, 22 °C; (c) n-BuOH cryst.

line diastereomers 28a and 28b cyclized in the presence

of AlCl₃at 80 °C to give enantiomers of a single ketone

diastereomer 29, having the C-2a and C-4 hydrogens on

the same face of the molecule. As expected, cyclization

of 50:50 mixture of acid diastereomers (28a , 28b) gave

racemic ketone 29. In short, the configuration at C-2a

(which was not controlled) dictated the configuration at

C-4.

Considering this precedent, we desired to determine

if the Friedel-Crafts cyclization reaction of indoline 26

could be performed under (milder) conditions that would

enable us to use both diastereomers and avoid epimer-

ization at C-4 in the product ketones. If so, the covalent

resolution of acid 26 could be omitted from the synthesis,

providing a much more efficient large scale process.

Indeed we found that the single diastereomer of the

carboxylic acid 26a could be converted to the tricyclic

ketone 30a under relatively mild conditions. The acid

was treated with 2 equiv of oxalyl chloride and catalytic

DMF at -5 °C, and the resulting acid chloride (26c, not

isolated) was treated with 4 equiv of AlCl₃at 22 °C for

20-30 h. The crude ketone 30 was obtained as a 98:2

mixture of diastereomers and was recrystallized from

n-BuOH to provide (+)-30a in 70% overall yield with

>99% ee (Scheme 5). Other Lewis acids were examined

in this reaction. TiCl₄and FeCl₃/MeNO₂gave no reaction

at ambient temperature. Reactions with AlBr₃gave the

desired ketone product but offered no advantage over

AlCl₃.

(25) The data do not exclude the possibilty of azalactone formation

and epimerization prior to cyclization or epimerization of the acid

chloride. While small amounts (<5% yield) of azalactone i were isolated

from some of the cyclization reactions, it is not clear that any of the

azalactone cyclized. Treatment of 26a or a mixture of 26a and 26b with

DCC gave an identical (by NMR and HPLC) 1:1 mixture of diastere-

omeric azalactones i. Cyclization of i with AlCl₃at 22 °C in CH₂Cl₂was

much slower than the cyclization of the acid chloride. Less than 5%

ketone 30 was formed after 7 h, whereas 80% of the acid chloride 26c

was converted to 30 in the same time.

When a 45:55 mixture of indolines 26a :26b was

subjected to the above Friedel-Crafts cyclization condi-

8646 J . Org. Chem., Vol. 62, No. 25, 1997

Carr et al.

Sch em e 6. In tr a m olecu la r F r ied el-Cr a fts

Sch em e 8. Com p letion of th e Syn th esis of Am in e

13 fr om L-Tr yp top h a n ^a

Acyla tion Rea ction : Ster eoch em istr y

a

(a) NaBH₄, MeOH, THF, 0 °C; (b) TFAA, THF, H₂, 10% Pd-

C; (c) NaOH, THF, H₂O.

at C-5 in 90% yield. Alcohol 31 was then trifluoroacetyl-

ated in situ with TFAA under hydrogenolysis conditions

(10% Pd-C, 50 psi of H₂, THF) to give amide 32 in 86%

yield. Simple base hydrolysis of the trifluoroacetamide

gave amine 13 identical with that prepared from the

Kornfeld ketone.

Sch em e 7. P ossible Con for m a tion s of Keton e

Dia ster eom er s

In summary, we demonstrated that a key intermediate

in the LY228729 synthesis (amine 13) could be prepared

in seven steps and 8% overall yield from ^L-tryptophan.

However, due to the stereochemical consequences of the

Friedel-Crafts acylation reaction, this strategy still

required a difficult covalent resolution which detracted

from its large scale viability.

Com p a r ison of Syn th etic Rou tes. Our route selec-

tion was guided by our laboratory and pilot plant experi-

ences on all three synthetic approaches, including the

tetralone approach. Strategically, all three routes pro-

duced enantiomerically pure product via a covalent

resolution. A comparison of the syntheses of the key

intermediate, amine 13, from commercially available

materials is shown in Scheme 9. In summary, the

synthesis of LY228729 was accomplished in nine steps

in 9% overall yield from commercially available olefin 5.

While ^L-tryptophan cost is approximately one-tenth that

of olefin 5, the synthesis of LY228729 from ^L-tryptophan

required 13 steps and proceeded in 3% overall yield. In

addition to requiring fewer steps, the conversion of olefin

5 to amine 13 was operationally much simpler and more

reliable than the conversion of ^L-tryptophan to 13.

Notably the fractional crystallization of amino alcohol 8

was much more robust than the fractional crystallization

of the tryptophan-derived acid 26a . These factors all

pointed to the Kornfeld ketone route as the most viable

long-term route of choice.

presence of aqueous acid (e.g. acetonitrile containing 1

equiv of 1 N HCl/24 h). We found that any mixture of

the ketone diastereomers could be isomerized to an

equilibrium 98:2 ratio of diastereomers, 30a :30b, respec-

tively. Extensive isomerization of 30b also occurred on

silica gel flash chromatography columns, making it

impossible to isolate pure samples in this fashion.

Ketone 30b was eventually isolated in 97% purity by

rapid preparative HPLC on a silica gel column (see

Experimental Section). The thermodynamic preference

for diastereomer 30a versus 30b can be rationalized as

the difference between equatorial and axial C-4 amide

substituents, respectively (Scheme 7). With the above

data in hand we concluded that the Friedel-Crafts

cyclization of the mixture of diastereomeric acids 26

would not be a suitable process for the preparation of

enantiomerically pure LY228729.

In ter section w ith th e Kor n feld Keton e Rou te.

Focusing on ketone (+)-30a , all that remained to inter-

sect the Kornfeld ketone route was deoxygenation of C-5

and deprotection of the primary amide. Deoxygenation

of the ketone proved to be nontrivial. Catalytic reduction

of (+)-30a with 10% Pd/C gave the fully reduced com-

pound (32) in 47% yield but required TFA as solvent. The

most efficient deoxygenation method was a two-step

protocol (Scheme 8). Ketone (+)-30a was reduced with

NaBH₄to give alcohol 31 as an 8:1 mixture of epimers

Exp er im en ta l Section

Melting points are uncorrected. Unless otherwise

noted, reagents and solvents were used as received from

commercial suppliers. TLC was performed on Kieselgel

60 F254 plates (Merck) using reagent grade solvents.

Flash chromatography was performed using Merck silica

gel 60 (230-400 mesh). ¹H NMR spectra were recorded

at 500 or 300 MHz or ¹³C NMR spectra were recorded at

75 or 125 MHz, respectively, in CDCl₃unless otherwise

Partial Ergot Alkaloid LY228729

J . Org. Chem., Vol. 62, No. 25, 1997 8647

Sch em e 9. Com p a r ison of Syn th etic Rou tes

noted. Chemical shifts are in ppm downfield from

internal tetramethylsilane. Concentrations were made

using a rotary evaporator at a bath temperature <40 °C.

Mass spectral analysis and combustion analysis were

performed by the Eli Lilly and Co. Physical Chemistry

Department.

mixture was filtered through a small plug of SiO₂and

rinsed first with toluene and then with toluene:acetone

(1:1) until the filtrates were colorless. Filtration through

a plug of silica gel removes some of the color and a base

line impurity as well as the resin. The filtrates were

concentrated to dryness and then dried in vacuo, causing

crystallization of the olefin (317.8 g, 78.7%). The olefin

was used directly in the subsequent epoxidation, al-

though an analytical sample was prepared by recrystal-

lization from EtOAc/hexanes: mp 97-98.5 °C; IR (KBr)

3030 (m), 1639 (s), 1623 (s), 1607 (m), 1460 (s), 1456 (s),

1395 (s), 1233 (s) cm^-1; ¹H NMR (CDCl₃) δ 2.14 (m, 1H);

2.54 (m, 1H), 3.50 (m, 1H), 3.78 (t, 1H, J ) 10.2 Hz), 4.32

(br s, 1H), 6.01 (m, 1H), 6.54 (dd, 1H, J ) 3, 7.2 Hz),

1-Ben zoyl-1,2,2a ,3,4,5-h exa h yd r oben z[cd ]in d ol-5-

ol (A). The ketone³(947.9 g, 3.42 mol) was suspended

in absolute EtOH (3 L) in a 5-L, 3-neck round-bottom

flask equipped with a condenser topped with a drying

tube, a mechanical stirrer, and a thermocouple. NaBH₄

(126.6 g, 3.35 mol) was then added portionwise at rt over

a 15 min period. The internal temperature rose from 23

to 66 °C and began to cool slowly, whereupon the reaction

mixture became nearly homogeneous and then formed a

precipitate. When the exotherm subsided, the reaction

mixture was heated to reflux for 30 min to consume the

last 5-10% of starting ketone and then cooled to rt. The

resulting thick slurry was added to water (5 L) and the

pH adjusted to 7 with 6 N HCl. The precipitate was

filtered, washed well with 4:1 water:EtOH,and dried to

vacuo to afford 832.2 g (87.1%) as an off-white solid. The

crude reaction mixture was analyzed by reverse phase

HPLC, acetonitrile/water 1/1, and the ratio of epimers

determined to be 6:1. The NMR data given represent

the chemical shifts for the major isomer, although the

minor isomer is also apparent by NMR: mp 190.5-191.5

°C; IR (KBr) 3437 (br), 1637 (s), 1468 (s), 1466 (s), 1395

(s), 1249 (w) cm^-1; ¹H NMR (CDCl₃) δ 1.55 (m, 1H), 1.74

(m, 1H), 2.17 (m, 1H); 2.42 (m, 1H), 2.52 (br s, 1H), 3.31

(m, 1H), 3.58 (t, 1H, J ) 10.8 Hz), 4.24 (br s, 1H), 4.86

(m, 1H), 6.98-7.64 (m, 8H), ¹³C NMR (CDCl₃) δ 26.9,

33.9, 37.5, 59.5, 68.4, 127.4, 128.3, 128.6, 130.6, 141.3,

6.75 (br s, 1H), 7.36-7.66 (m, 7H); ¹³C NMR (CDCl ) δ

3

28.1, 34.5, 59.7, 119.6, 126.3, 127.3, 128.3, 128.4, 128.6,

130.5, 140.7, 168.8; EI-MS m/z ) 261; UV (EtOH) λ_max

)

265 (ꢀ ) 21 400), 235 (ꢀ )15 800); TLC R_f) 0.75 (SiO₂,

4:1 toluene:acetone). Anal. Calcd for C₁₈H₁₅NO: C,

82.73; H, 5.79; N, 5.36. Found: C, 82.49; H, 5.73; N, 5.24.

1-Ben zoyl-4,5-ep oxy-1,2,2a ,3,4,5-h exa h yd r oben z-

[cd ]in d ole (6). The olefin 5 (455.0 g, 1.74 mol) was

dissolved in CH₂Cl₂(2.5L) and cooled to -2 °C with

stirring. m-Chloroperbenzoic acid (300.6 g, 1.74 mol) was

then added, and a strong exotherm was noted after a 5

min induction period. After 2 h, another portion of

m-chloroperbenzoic acid (150.3 g, 0.87 mol) was added

with continued stirring for another 2 h. The precipitated

m-chlorobenzoic acid was filtered and washed with CH₂-

Cl₂. The filtrate was then successively washed with cold

2 N NaOH (1 L), water (1 L), 2 N NaOH (1 L), water (1

L), and brine (1 L). The organic phase was dried over

Na₂SO₄, filtered through HyFlo, and concentrated to

dryness. The residue was azeotroped several times from

toluene and used directly in the next step without further

purification. An analytical sample was prepared by

recrystallization from EtOAc/hexanes: mp 109-110 °C;

IR (KBr) 3035 (m), 1640 (s), 1619 (m), 1475 (s), 1406 (s),

1392 (s) cm^-1; ¹H NMR (CDCl₃) δ 1.28 (dd, 1H, J ) 12.0,

13.8 Hz), 2.68 (m, 1H), 3.36 (m, 1H), 3.70 (m, 2H), 3.86

(d, 1H, J ) 3.9 Hz), 4.38 (br s, 1H), 7.10 (br s, 1H), 7.20-

7.42 (m, 7H); ¹³C NMR (CDCl₃) δ 26.2, 32.0, 50.2, 55.1,

57.8, 123.0, 127.4, 128.0, 128.6, 130.7, 141.6, 168.7; EI-

MS m/z ) 277; UV (EtOH) λ_max) 295 (ꢀ ) 7100), 265 (ꢀ

) 12 000), 216 (ꢀ ) 30 000); TLC R_f) 0.44 (SiO₂, 1:1

hexanes:EtOAc). Anal. Calcd for C₁₈H₁₅NO₂: C, 77.96;

H, 5.45; N, 5.05. Found: C, 77.79; H, 5.20; N, 4.98.

1-Ben zoyl-5-(1-(S)-p h en et h yla m in o)-1,2,2a ,3,4,5-

h exa h yd r oben z[cd ]in d ol-4-ol (8). The epoxide 7 (482.5

168; EI-MS m/z ) 279, 105 (parent); UV (EtOH) λ_max

)

293 (ꢀ ) 8600), 266 (ꢀ ) 11 000); TLC R_f) 0.44 (two

spots, SiO₂, 4:1 toluene:acetone). Anal. Calcd for

C₁₈H₁₇NO₂: C, 77.40; H, 6.13; N, 5.01. Found: C, 77.10;

H, 6.17; N, 5.01.

1-Ben zoyl-1,2,2a ,3-tetr a h yd r oben z[cd ]in d ole (5).

Into a 5-L, 3-neck round-bottom flask were placed Am-

berlyst 15 ion-exchange resin (228 g) and toluene (2.5 L).

The flask was equipped with a mechanical stirrer, a

Dean-Stark water separator topped with a condenser,

a N₂inlet, and a thermocouple. The mixture was stirred

at reflux for 15-20 min and cooled to 95 °C (internal

temperature). The alcohol A (432.5 g, 1.55 mol) was then

carefully added and the mixture brought to reflux under

N₂. Water collected in the trap at a steady rate, and the

dehydration was complete after 3 h. The cooled reaction

8648 J . Org. Chem., Vol. 62, No. 25, 1997

Carr et al.

g, 1.74 mol) was dissolved in n-BuOH (4400 mL) in a 12-L

3-neck flask equipped with a mechanical stirrer, a

thermocouple, and a condenser topped with a N₂inlet.

The (S)-(-)-R-methylbenzylamine (900 mL, 14.0 mol) was

added, and the solution was stirred at 90 °C overnight.

After 24 h, the reaction mixture was allowed to cool to

rt, whereupon the desired amino alcohol crystallized

directly from the reaction mixture. The crystalline

material was filtered, washed with Et₂O (2 L), and dried.

The first crop weighed 168.26 g (24.3%) and was used

directly in the subsequent reaction. A second crop could

be obtained by concentrating the filtrates to dryness and

dissolution in toluene (200 mL) followed by the addition

of hexanes (100 mL) and Et₂O (100 mL). The resulting

solution was allowed to stand in the refrigerator over-

night to provide an additional 39.2 g (5.7%) after filtra-

tion. Recrystallization from i-PrOH affords the amino

alcohol as colorless fibrous needles: mp 192-193 °C; IR

(KBr) 3480 (br), 1638 (s), 1610 (w), 1470 (s), 1457 (s), 1394

(s) cm^-1; ¹H NMR (CDCl₃) δ 7.02-7.56 (m, 13H), 4.21 (q,

1H, J ) 6.6 Hz), 4.25 (br s, 1H), 3.63 (m, 2H), 3.42 (m,

2H), 2.72 (br s, 1H, exhanges with D₂O), 1.99 (m, 1H),

1.80 (m, 1H), 1.47 (d, 3H, J ) 6.6 Hz); ¹³C NMR (CDCl₃)

δ 168.8, 145.4, 141.1, 136.5, 130.6, 128.6, 128.2, 127.3,

126.9, 120.3, 71.1, 59.2, 57.5, 30.7, 24.9; EI-MS m/z )

398, 355, 249, 145, 105; UV (EtOH) λ_max) 292 (ꢀ ) 8900),

265 (ꢀ ) 11 400); TLC R_f) 0.68 (SiO₂, 42:42:16 EtOAc:

hexane:triethylamine) ) desired diastereomer, R_f) 0.62

undesired diastereomer. Anal. Calcd for C₂₆H₂₆N₂O₂: C,

78.37; H, 6.58; N, 7.03. Found: C, 78.14; H, 6.67; N, 6.77.

(-)-4-Ben zoyl-5,5a ,6,6a ,7,7a -h exa h yd r o-7-(1-(S)-

ph en eth yl)-4H-azir in o[f]ben z[cd]isoin dole (12). Mit-

sunobu Protocol. The amino alcohol 8 (278.0 g, 0.698

mol) was placed into a 5-L 3-neck round-bottom flask

equipped with a mechanical stirrer, a thermocouple, and

an addition funnel topped with a N₂inlet. Anhydrous

THF (3.0 L) was added followed by triphenylphosphine

(228 g, 0.873 mol, 1.25 molar equiv), and the resulting

solution was stirred at rt under N₂. Diethyl azodicar-

boxylate (DEAD, 150.1 g, 0.873 mol) dissolved in THF

(100 mL) was added dropwise over a 5 h period. A slight

exotherm was also noted during the course of the addi-

tion: 21-30.4 °C. The reaction mixture was stirred

vigorously at rt overnight and monitored by TLC showing

steady progress but requiring in excess of 16 h for

complete reaction. The reaction was then filtered through

Celite (HyFlo) and silica gel (ca. 60 g), and the solution

was washed with THF (500 mL). The solvent was

removed in vacuo, and the resulting semisolid (715 g) was

triturated with MeOH (100 mL) and Et₂O (600 mL) and

refrigerated overnight. The solid was filtered, washed

well with Et₂O (3 × 100 mL to remove DEAD-H₂), and

dried. The fairly crystalline sold (279.4 g, >100%) was

triturated again with MeOH (100 mL) and Et₂O (500

mL), filtered, washed, and dried to yield 248.0 g that

appeared to be >90% pure by NMR. The calculated

chemical yield was 84.5%. The thus obtained material

could be recrystallized from i-PrOH (approximately 20

volumes): mp 184-186 °C; IR (KBr) 2978 (m), 1638 (s),

1468 (s), 1455 (s), 1385 (s) cm^-1; ¹H NMR (CDCl₃) δ 7.18-

7.56 (m, 13H), 4.17 (br m, 1H), 3.55 (t, 1H, J ) 10.7 Hz),

3.41 (m, 1H), 2.75 (q, 1H, J ) 6.5 Hz), 2.56 (br m, 1H),

2.50 (d, 1H, J ) 6.3 Hz), 2.08 (m, 1H), 1.59 (m, 1H), 1.53

(d, 3H, J ) 6.5 Hz); ¹³C NMR (CDCl₃) δ 168.6, 144.4,

141.6, 136.5, 132.8, 130.5, 128.6, 128.3, 127.3, 127.0,

126.7, 124.1, 69.9, 59.0, 38.6, 37.7, 34.5, 31.6, 23.6; MS

m/z ) 380, 275, 261, 105, 77; UV (EtOH) λ_max) 294 (ꢀ )

8400), 265 (ꢀ ) 11 200); TLC R_f) 0.72 (SiO₂, hexane:

ethyl acetate 1:1) ) desired diastereomer, R_f) 0.60

undesired diastereomer. Anal. Calcd for C₂₆H₂₄N₂O: C,

82.07; H, 6.37; N, 7.36. Found: C, 81.79; H, 6.34; N, 7.28.

[R]_D) -41.1 (589 nm); [R]_D) -249.4 (365 nm).

Mesyla tion P r otocol. Methanesulfonyl chloride (438

g, 3.8 mol) was added over 55 min to a -10 °C solution

of amino alcohol 8 (1015 g, 2.55 mol) and triethylamine

(800 g, 7.9 mol) in 8.1 L of CH₂Cl₂. The reaction mixture

was allowed to warm to 25 °C and then stirred for 2 h.

The reaction mixture was then extracted sequentially

with 8.1 L each of water, a 5% NaHCO₃solution, and

brine. The organic layer was dried with Na₂SO₄and

added to 4 L of acetonitrile. The solution was concen-

trated in vacuo, and an additional 4 L of acetonitrile was

added. The solution was concentrated to a total volume

of 2 L, cooled to 0-5 °C for 18 h, and filtered. The solid

was washed with 2 L of cold acetonitrile and vacuum-

dried to give 855 g (88% yield) of aziridine 12 as a white

solid.

Physical data (for undesired aziridine diastereomer

derived from 9): mp 196-199 °C; IR (KBr) 3010 (m),

1

1636 (s), 1612 (s), 1596 (s), 1459 (s) 1398 (s) cm^-1; H

NMR (CDCl₃) δ 7.00-7.58 (m, 13H), 4.21 (br m, 1H), 3.59

(t, 1H, J ) 10.7 Hz), 3.47 (m, 1H), 2.77 (br m, 2H), 2.37

(d, 1H, J ) 6.3 Hz), 2.24 (m, 1H), 1.47 (d, 3H, J ) 6.5

Hz); ¹³C NMR (CDCl₃) δ 168.6, 144.3, 141.2, 130.4, 128.5,

128.4, 128.3, 128.0, 127.4, 127.3, 127.0, 126.9, 70.0, 59.0,

39.3, 36.9, 31.6, 23.2; EI-MS m/z ) 381, 277, 262, 105,

77; UV (EtOH) λ_max) 294 (ꢀ ) 8780), 264 (ꢀ ) 1600).

Anal. Calcd for CHNO: C, 82.07; H, 6.37; N, 7.36.

Found: C, 82.32; H, 6.54; N, 7.28. [R]_D) -34.4 (c 1.0,

THF).

(+)-1-Ben zoyl-1,2,2a ,3,4,5-h exa h yd r ob en z[cd ]in -

d ole-4-a m in e (13). To a mixture of 12 g of 10% Pd-C

and 20 mL of water under nitrogen was added 600 mL

of a 3:1 THF:85% phosphoric acid solution. The mixture

was cooled to 5 °C, and 60 g (0.15 mol) of aziridine 12

was added. The reaction vessel was evacuated and back-

filled with H₂three times. The reaction mixture was

stirred under 1 atm of H₂for 6 h, at which time HPLC

analysis showed complete aziridine hydrogenolysis. The

pH of the reaction mixture was then increased to 3 by

the addition of 127 mL of a 45% KOH solution. The

reaction vessel was again evacuated and back-filled with

H₂(3×). The reaction mixture was heated to 55 °C and

stirred under 1 atm of H₂for 17 h. The reaction mixture

was then cooled to ambient temperature, vented, and

diluted with 450 mL of CH₂Cl₂and 50 mL of water. The

pH was then adjusted to 10 with 45% KOH. This

mixture was stirred for 90 min and filtered through a

pressure filter. The catalyst was washed with 2 × 150

mL of CH₂Cl₂and 2 × 50 mL of water. The (upper)

CH₂Cl₂layer was separated, and the water layer was

extracted with 200 mL of CH₂Cl₂. The product was

extracted from the combined CH₂Cl₂layers by the addi-

tion of 350 mL of 0.5 M HCl. The layers were separated,

and the pH of the aqueous layer was adjusted to 11 with

45% KOH solution. The milky solution was extratcted

with 300 mL of CH₂Cl₂. The organic layer was separated

and concentrated by atmospheric distillation. To the

residual warm oil were added 98 mL of i-PrOH and 196

mL of water. After being cooled to 0 °C for 2 h the

resulting slurry was filtered. The solid was washed with

cold 1:2 i-PrOH:water (100 mL) and vacuum-dried to

provide 31.1 g (71% yield) of amine 13 as short needles:

mp 147-150 °C; IR (KBr) 1225 (w), 1396 (s), 1457 (s),

Partial Ergot Alkaloid LY228729

J . Org. Chem., Vol. 62, No. 25, 1997 8649

1488 (m), 1597 (m), 1612 (s), 1637 (s), 3009 (m) cm^-1; ¹H

NMR (CDCl₃) δ 7.38-7.57 (m, 5H), 6.99 (m, 1H), 6.78

(m, 2H), 4.25 (br m, 1H), 3.62 (t, 1H, J ) 11.5 Hz), 3.29

(m, 2H), 3.12 (dd, 1H, J ) 6.1, 16.7 Hz), 2.39 (dd, 1H, J

) 10.3, 16.7 Hz), 2.17 (m, 1H), 1.49 (br s, 2H), 1.31 (q,

1H, J ) 11.5 Hz); ¹³C NMR (CDCl₃) δ 168.5, 141.4, 136.6,

133.3, 132.6, 130.7, 130.1, 128.8, 128.1, 127.7, 127.6,

127.1, 123.1, 122.6 58.2, 48.6, 37.3, 37.2, 36.9; EI-MS m/z

) 278, 261, 235, 130, 105, 77; UV (EtOH) λ_max) 291 (ꢀ

) 8150), 266 (ꢀ ) 10 600); TLC R_f) 0.19 (SiO₂, CH₂Cl₂:

MeOH 4:1), R_f) 0.86 ) 2° amine. Anal. Calcd for

C₁₈H₁₈N₂O: C, 77.67; H, 6.52; N, 10.06. Found: C, 77.76;

H, 6.55; N, 9.61. [R]_D) +57.4 (EtOH). [R]₃₆₅) +341.6.

(+)-1-Ben zoyl-6-br om o-1,2,2a ,3,4,5-h exa h yd r oben -

z[cd ]in d ole-4-a m in e (14a ). The 1° amine 13 (29.4.g,

0.106 mol) was placed into a 500-mL, 3-neck round-

bottom flask equipped with a mechanical stirrer, a N₂

inlet, and a constant addition funnel. The substrate was

dissolved in glacial HOAc (250 mL), and NaOAc (34.7 g,

0.423 mol, 4 molar equiv) was then added. A solution of

bromine (5.45 mL, 0.106 mol) in HOAc (20 mL) was then

added dropwise over a 1 h period with vigorous stirring

and then stirred at rt overnight. The resulting thick

slurry was diluted with Et₂O, filtered, and washed well

with Et₂O. The material thus obtained was slurried in

H₂O (400 mL) and the pH adjusted to 11-12 with 5 N

NaOH. The solid was filtered, washed well with H₂O (2

× 200 mL), and dried in vacuo to provide 33.6 g (88.8%)

of the bromo 1° amine. An analytical sample could be

prepared by recrystallization from i-PrOH or 50% aque-

ous EtOH: mp 169-173 °C; IR (KBr) 3010, 2934, 1640,

128.9, 127.7, 118.6, 57.8, 53.1, 30.6, 29.2, 22.9, 12.1; EI-

MS m/z ) 440/442; UV (EtOH) λ_max) 272 (ꢀ ) 15 600);

TLC R_f) 0.46 (SiO₂, Hex:EtOAc:TEA 60:35:5); HPLC

Supelco C-18 25 cm column, 2 mL/min, 35.65 acetonitrile:

chloroacetic acid buffer t_R) 2.9 min (product), t_R) 2.0

min (starting material). Anal. Calcd for C₂₄H₂₉N₂OBr:

C, 60.52; H, 4.80; N, 7.82. Found: C, 60.33; H, 4.89; N,

7.72. [R]_D) +20.73 (EtOH).

(+)-1-B e n zo y l-4-(d ip r o p y la m in o )-1,2,2a ,3,4,5-

h exa h yd r o[cd ]in d ole-6-ca r bon itr ile (15a ). The bro-

mide 14c (154.48 g, 0.35 mol) was dissolved in N-meth-

ylpyrrolidinone (NMP, 850 mL) to which CuCN (37.6 g,

0.42 mol, 1.2 molar equiv) was added. The flask was

equipped with a condenser topped with a Firestone valve,

a thermocouple, and a mechanical stirrer. The mixture

was degassed five times (vacuum/N₂purge) and slowly

brought to 200 °C (internal temperature). After 1 h, TLC

indicated that the reaction was nearly complete. After

a total of 2.5 h, TLC showed that no starting material

present. The resulting dark reaction mixture had pre-

cipitated Cu on the flask walls and was then cooled to

rt. The mixture was diluted with CH₂Cl₂(1 L) and

washed with 15% NH₄OH. The combined organic layers

were washed with H₂O (4 x l L), and brine (1 L) and dried

over Na₂SO₄. The desiccant was removed by filtration

and the filtrate concentrated to dryness. The crude

residue was chromatographed in several small portions

over silica gel with a hexane/EtOAc gradient to provide

10.27 g of the nitrile (75.7%). This material was used in

the subsequent deprotection step without crystallization

but could be recrystallized from 50% aqueous EtOH or

EtOH: mp 109-111 °C; IR (KBr) 2959, 2213, 1661, 1616,

1

1580, 1468, 1454, 1384 cm^-1; H NMR (CDCl₃) δ 7.42-

1

7.58 (m, 7H), 4.27 (br s, 1H), 3.68 (t, 1H, J ) 11.1 Hz),

3.33 (m, 2H), 3.16 (dd, 1H, J ) 6.3, 17.3 Hz), 2.28 (dd,

1H, J ) 9.6, 17.3 Hz), 2.17 (m, 1H), 1.44 (br s, 2H), 1.32

(q, 1H, J ) 11.6 Hz); ¹³C NMR (CD₃OD) δ 170.6, 141.9,

137.3, 136.4, 134.1, 132.1, 132.0, 129.8, 128.1, 118.8,

116.2, 59.5, 49.3, 37.8, 37.1, 35.4; EI-MS m/z ) 356, 358,

339, 341, 105, 77; UV (EtOH) λ_max) 272 (ꢀ ) 14 400);

TLC R_f) 0.28 (SiO₂, MeOH); HPLC Supelco C-18 25 cm

column, 2 mL/min, 35.65 acetonitrile: chloroacetic acid

buffer t_R) 6.3 min (product); t_R) 4.2 min (intermediate

2° amine); t_R) 4.6 min (desbromo). Anal. Calcd for

C₁₈H₁₇N₂OBr: C, 65.31; H, 6.62; N, 6.35; Br, 18.10.

1470, 1453, 1368, 1355 cm^-1; H NMR (CDCl₃) δ 7.34-

7.58 (m, 7H), 4.35 (m, 1H), 3.72 (t, 1H, J ) 11.2 Hz), 3.30

(m, 2H), 3.13 (m, 1H), 2.72 (m, 1H), 2.45 (m, 4H), 2.27

(m, 1H), 1.46 (m, 5H), 0.90 (t, 6H, J ) 7.3 Hz); ¹³C NMR

(CDCl₃) δ 169.0, 145.0, 138.2, 135.8, 134.1, 133.2, 131.0,

128.6, 127.3, 117.5, 113.9, 106.3, 58.4, 56.9, 52.7, 37.7,

29.3, 27.9, 22.5, 11.7; EI-MS m/z ) 387; UV (EtOH) λ_max

) 304 (ꢀ ) 19 600), 287 (ꢀ ) 19 800), 225 (ꢀ ) 23 000);

TLC R_f) 0.38 (SiO₂, Hex:EtOAc:TEA 65:35:5). Anal.

Calcd for C₂₅H₂₉N₃O: C, 77.47; H, 7.55; N, 10.85.

Found: C, 77.09; H, 7.65; N, 10.74. [R]_D) +1.59 (EtOH).

(-)-4-(D ip r o p y la m in o )-1,2,2a ,4,5-h e x a h y d r o -

ben z[cd ]in d ole-6-ca r bon itr ile (16a ). The benzamide

15a (41.02 g, 0.106 mol) was dissolved in freshly distilled

THF (375 mL) and cooled to -78 °C under N₂. n-BuLi

(59.3 mL, 0.148 mol, 1.4 molar equiv, 2.5 M) was then

added dropwise at a rate to maintain the temperature

below -65 °C. When TLC analysis indicated complete

reaction, glacial HOAc (10 mL) was added carefully and

the reaction mixture was warmed to rt. Et₂O (250 mL)

and 1 N HCl (250 mL) were added and the layers

separated. The organic phase was extracted with ad-

ditional 1 N HCl (2 × 100 mL) and the combined aqueous

phase washed with Et₂O (2 × 250 mL); 5 N NaOH (100

mL) was added dropwise with stirring followed by

extraction with CH₂Cl₂(250 + 2 × 150 mL). The

combined organic phase was washed with brine, dried

over Na₂SO₄, and concentrated to dryness. The resulting

light tan, highly crystalline material was dried in vacuo

to a constant weight (28.4 g, 94.5%). This material was

recrystallized from hot aqueous EtOH (EtOH:H₂O 75:25),

cooled, filtered, and washed with ice cold solvent: mp

114.5-115.5 °C; IR (KBr) 3336, 2934, 2210, 1625, 1586,

Found: C, 65.15; H, 6.70; N, 6.36; Br, 18.31. [R]_D

+11.64 (EtOH).

)

(+)-1-Ben zoyl-6-br om o-N,N-d ip r op yl-1,2,2a ,3,4,5-

h exa h yd r oben z[cd ]in d ole-4-a m in e (14c). Amine 14a

(9.82 g, 0.0275 mol) was placed in a 500-mL, round-

bottom flask equipped with a mechanical stirrer, a

condenser topped with a N₂inlet, and a thermocouple.

Acetonitrile (175 mL) and K₂CO₃(0.275 mol) were added,

followed by the addition of iodopropane (13.2 mL, 0.137

mol) with vigorous stirring. The reaction mixture was

stirred at 75 °C under N₂overnight. After being cooled

to rt, the reaction mixture was diluted with CH₂Cl₂(200

mL), washed successively with H₂O, NaHCO₃solution,

H₂O, and brine, and dried over Na₂SO₄. After filtration,

the volatiles were removed in vacuo to provide 11.5 g

(94%) of crude product. This material was then recrys-

tallized from 95% EtOH to provide the desired product

as colorless needles, 9.7 g (80.0%): mp 93 °C; IR (KBr)

1

2958, 1655, 1464, 1453, 1381 cm^-1; H NMR (CDCl₃) δ

7.41-7.58 (m, 7H), 4.27 (m, 1H), 3.34 (m, 1H), 3.19 (m,

1H), 2.92 (dd, 1H, J ) 5.6, 18.1 Hz), 2.48 (m, 5H), 2.16

(m, 1H), 1.47 (m, 4H), 1.40 (m, 1H), 0.90 (t, 6H, J ) 7.3

Hz); ¹³C NMR (CDCl₃) δ 168.9, 140.9, 134.7, 131.3, 130.0,

1

805 cm^-1; H NMR (CDCl₃) δ 7.27 (1H, d, J ) 9.0 Hz),

6.39 (1H, d, J ) 9.0 Hz), 4.12 (1H, br s), 3.75 (1H, m),

8650 J . Org. Chem., Vol. 62, No. 25, 1997

Carr et al.

3.20 (1H, m), 3.03 (1H, dd, J ) 18, 6.0 Hz), 2.63 (1H,

ddd, J ) 18, 12, 2.0 Hz), 2.45 (4H, t, J ) 9.0 Hz), 2.19

(1H, dt, J ) 6.0, 3.0 Hz), 1.45 (5H, m), 0.89 (6H, t, J )

9.0 Hz); ¹³C NMR (CDCl₃) δ 154.0, 137.4, 134.0, 130.7,

119.2, 105.7, 99.6, 57.4, 55.7, 52.8, 38.9, 29.7, 27.6, 22.6,

11.8; EI-MS m/z ) 283, 254, 240, 183, 156, 128, 98, 72;

UV (EtOH) λ_max) 296 (ꢀ ) 16 500), 231 (ꢀ ) 14 100),

205 (ꢀ ) 16 300) in EtOH; TLC R_f) 0.30 (SiO₂, Hex:

EtOAc:TEA 65:35:5). Anal. Calcd for C₁₈H₂₅N₃: C,

76.28; H, 8.89; N, 14.83. Found: C, 76.56; H, 8.85; N,

14.71. [R]_D) -34.0 (THF, c ) 1). [R]₃₆₅) -217.7.

(-)-4-(Dip r op yla m in o)-1,2,2a ,3,4,5-h e xa h yd r o-

[cd ]in d ole-6-ca r boxa m id e (16b). Polyphosphoric acid

(PPA, 300 mL) was placed into a 500-mL, 3-neck flask

equipped with a mechanical stirrer, a stopper, and a

condenser topped with a N₂inlet. The reaction vessel

was degassed by vacuum/purge cycles (5×) and then

heated to 85-90 °C (internal temperature) while the

nitrile 16a (22.65 g, 0.080 mol) was added portionwise.

The reaction mixture became homogeneous as the hy-

drolysis occurred. After all of the nitrile had been added,

the mixture was stirred at this temperature for an

additional 2.0 h to ensure complete hydrolysis. The

reaction mixture was then carefully poured onto crushed

ice and stirred vigorously. After the ice had melted, the

pH was adjusted with 5 N NaOH to 11/12, and the

solution was extracted with several portions of CH₂Cl₂.

The organic phase was dried over Na₂SO₄, filtered, and

concentrated to afford 23.53 g of the amide as a foam:

mp 161-164 °C; IR (KBr) 3381 (s), 3377 (s), 2956 (m),

136.2, 134.5, 130.8, 128.6, 127.4, 115.9, 92.4, 58.1, 49.2,

42.8, 38.1, 36.8. IR (KBr) 1640 cm^-1; FD mass spectrum

m/z 404 (M⁺). Anal. Calcd for C₁₈H₁₇IN₂O: C, 53.48; H,

4.24; N, 6.93. Found: C, 53.20, H, 4.11; N, 6.64.

(2a S,4S)-1-Ben zoyl-4-(d i-n -p r op yla m in o)-6-iod o-

1,2,2a ,3,4,5-h exa h yd r oben z[cd ]in d ole (14d ). Primary

amine 14b (100 g, 0.25 mol), potassium carbonate (137

g, 1.0 mol), and 1-iodopropane (97 mL, 1.0 mol) were

slurried together in 1 L of CH₃CN. The heterogeneous

reaction mixture was warmed to approximately 75 °C and

stirred for 17 h. The mixture was cooled to room

temperature, and water (500 mL) and tert-butyl methyl

ether (500 mL) were added. The organic layer was

separated and washed with water. CH₃CN (500 mL) was

added to the organic layer, and the volume was reduced

in half by evaporation. A small amount of precipitated

black solids was removed by hot gravity filtrataion. The

filtrate was cooled to room temperature, and orange

crystals formed. The mixture was cooled in an bath and

filtered. The solid was washed with cold CH₃CN (2 ×

75 mL) and vacuum-dried to give 100 g (83% yield) of

tertiary amine 14d : mp 109-110 °C; [R]_D) 17.0 (c ) 1,

1

THF); H NMR (300 MHz, CDCl₃) δ 7.50 (m, 7H), 4.28

(bs, 1H), 3.66 (t, J ) 11 Hz, 1H), 3.34 (m, 1H), 3.20 (m,

1H), 3.05 (dd, J ) 18, 6 Hz, 1H), 2.47 (m, 5H), 2.16 (m,

1H), 1.38 (m, 5H), 0.91 (t, J ) 7 Hz, 6H); ¹³C NMR (75

MHz, CDCl₃) δ 168.6, 141.7, 137.9, 137.6, 137.6, 136.3,

130.7, 127.4, 106.8, 93.3, 58.0, 52.9, 38.5, 35.1, 29.1, 22.6,

11.9; IR (KBr) 1638 cm^-1; FD mass spectrum m/z 488

(M⁺). Anal. Calcd for C₂₄H₂₉IN₂O: C, 59.02; H, 5.99; N,

5.74. Found: C, 59.03, H, 5.87; N, 5.64.

1

2932 (m), 1645 (s), 1616 (s), 1585 (m), 1379 (s) cm^-1; H

NMR (CDCl₃) δ 7.30 (d, 1H, J ) 8.0 Hz), 6.40 (d, 1H, J

) 8.0 Hz), 5.67 (br s, 2H), 3.92 (br s, 1H), 3.70 (m, 1H),

3.24 (m, 1H), 3.18 (m, 3H), 2.82 (dd, 1H, J ) 10.1, 17.0

Hz), 2.46 (m, 4H), 2.17 (m, 1H), 1.46 (m, 5H), 0.87 (t, 6H,

J ) 7.3 Hz); ¹³C NMR (CDCl₃) δ 171.2, 152.4, 133.9,

130.6, 128.4, 122.8, 104.8, 57.5, 55.5, 52.5, 39.2, 29.0,

27.8, 22.3, 11.5; EI-MS m/z ) 301 (fd); UV (EtOH) λ_max

) 273 (ꢀ ) 15 400), 214 (ꢀ ) 22 300); TLC R_f) 0.44 (SiO₂,

CH₂Cl₂:MeOH, 1:1). Anal. Calcd for C₁₈H₂₇N₃O: C,

71.72; H, 9.02; N, 13.94. Found: C, 71.44; H, 8.88; N,

13.75. [R]_D) -70.5 (c ) 1.02, MeOH).

(2a S ,4S )-1-Be n zoyl-4-a m in o-6-iod o-1,2,2a ,3,4,5-

h exa h yd r oben z[cd ]in d ole (14b). H₂SO₄(4.4 g, 45

mmol) was added dropwise to a solution of amine 13 (10.0

g, 36 mmol) in 1:1 acetic acid/water (50 mL). Periodic

acid (2.2 g, 9.6 mmol) and I₂(4.8 g, 18.9 mmol) were

added in rapid succession. The solution was heated to

60 °C for 1 h. With vigorous agitation, a 20% NaHSO₃

solution (10 mL) was added to the warm mixture. After

the solution was cooled with an ice bath, CH₂Cl₂(50 mL)

was added and the pH adjusted to 12 with a 10 N NaOH

solution (60 mL). The two-phase mixture was warmed

to 25 °C, and 50 mL of CH₂Cl₂was added. The organic

layer was separated, and the aqueous layer was extracted

with CH₂Cl₂. The combined CH₂Cl₂layers were dried

with Na₂SO₄‚CH₃CN (70 mL) was added, and a solvent

exchange to CH₃CN was affected. A mixture with a

yellow precipitate resulted, which was cooled to -15 °C

overnight. The cold mixture was filtered, and the solid

was washed with cold CH₃CN (2 × 30 mL). After

vacuum-drying, 13.1 g (90% yield) of 14b was obtained

as a white solid; mp 178-80 °C. [R]_D) 5.0 (c ) 1, THF).

¹H NMR (300 MHz, CDCl₃) δ 7.55 (m, 7H), 4.27 (bs, 1H),

3.70 (t, J ) 15 Hz, 1H), 3.30 (m, 2H), 3.05 (dd, J ) 17, 6

Hz, 1H), 2.18 (m, 2H), 1.30 (q, J ) 12 Hz, 1H). ¹³C NMR

(75 MHz, CDCl₃) δ 168.5, 141.8, 137.7, 136.8, 136.2,

(2a S,4S)-1-Ben zoyl-4-(d i-n -p r op yla m in o)-6-(a m i-

n oca r b on yl)-1,2,2a ,3,4,5-h exa h yd r ob en z[cd ]in d ole

(15b). To a 500-mL stainless stell autoclave were added

10.0 g (20.5 mmol) of iodoindoline 14d , 0.12 g (1.1 mmol)

of Pd black, 1.07 mmol of triphenylphosphine, and 150

mL of toluene. The autoclave was sealed, purged with

anhydrous ammonia (3 × 50 psi), and then pressured to

50 psi with ammonia. Carbon monoxide was added to

bring the final pressure to 85 psi. The stirred mixture

was heated to 110 °C for 14 h and then cooled to 35 °C

and vented. The warm reaction mixture was filtered, and

the autoclave was rinsed with an additional 50 mL of

toluene. The filtrate was concentrated to 50 mL, and

then 10 mL of hexane was added. The resulting slurry

was cooled to -15 °C for 24 h and filtered. The filter

cake was rinsed with cold toluene (25 mL) and hexanes

(2 × 30 mL) and then vacuum-dried at 50 °C to give 7.39

g (89% yield) of the product as a white crystalline solid:

mp 120-122 °C; [R]_D) -14.5 (c ) 1, THF); ¹H NMR (500

MHz, CDCl₃) δ 7.59 (d, J ) 7.0 Hz, 2H), 7.51 (m, 1H),

7.47 (m, 3H), 7.31 (bs, 1H), 5.81 (bs, 2H), 4.30 (bs, 1H),

3.69 (t, 1H, J ) 7.3 Hz), 3.34 (m, 1H), 3.27 (dd, J ) 17.5,

6.0 Hz), 3.19 (m, 1H), 2.89 (dd, 1H, J ) 17.5, 10.0 Hz),

2.47 (m, 4H), 2.19 (m, 1H), 1.48 (m, 4H), 0.90 (t, J ) 5.9

Hz, 6H); ¹³C NMR (75 MHz, CDCl₃) δ 170.9, 168.9, 143.3,

138.2, 134.6, 134.2, 130.8, 129.2, 128.6, 127.7, 127.4, 57.4,

52.9, 38.4, 29.2, 27.9, 22.6, 11.8; IR (KBr) 3347, 3177,

2958, 2932, 2871, 1676, 1639, 1579, 1465, 1450, 1368

cm^-1; FD mass spectrum m/z 405 (M⁺). Anal. Calcd for

C₂₅H₃₁N₃O₂: C, 74.03; H, 7.70; N, 10.36. Found: C,

74.33, H, 7.90; N, 10.48.

(2aS,4S)-4-(Di-n -p r op yla m in o)-6-(a m in oca r bon yl)-

1,2,2a ,3,4,5-h exa h yd r oben z[cd ]in d ole (16b). To a

solution of 71.95 g (177 mmol) of indoline 15b in 720 mL

of EtOH were added 650 mL of deionized water and 47

mL of 50% NaOH. The solution was heated to 80 °C for

Partial Ergot Alkaloid LY228729

J . Org. Chem., Vol. 62, No. 25, 1997 8651

4 h at which time HPLC analysis showed the reaction to

be complete. EtOH was removed by distillation under

vacuum, and 720 mL of CH₂Cl₂was added to the

remaining aqueous mixture. The organic layer was

isolated and washed with water. Ethyl acetate (360 mL)

was added, and the organic layer was concentrated to

approximately 400 mL total volume by atmospheric

distillation. The solution was cooled to room temperature

and stirred for 2 h to initiate crystallization. The slurry

was then cooled to 5 °C, stirred for 2 h, and filtered. The

solid was washed with 200 mL of cold ethyl acetate and

vacuum-dried at 50 °C to obtain 43.69 g (82% yield) of

white solid: mp 163-64 °C. [R]_D) -48.5 (c ) 1, THF);

¹H NMR (500 MHz, DMSO-d₆) δ 7.29 (bs, 1H), 7.21 (d, J

) 8.0 Hz, 1H), 6.76 (bs, 1H), 6.27 (d, J ) 8.0 Hz, 1H),

5.69 (s, 1H), 3.55 (m, 1H), 3.01 (m, 4H), 2.67 (dd, J ) 17,

10.3 Hz, 1H), 2.42 (m, 4H), 2.06 (d, J ) 6.8 Hz, 1H), 1.40

(m, 4H), 1.30 (q, J ) 11.0 Hz, 1H), 0.86 (t, J ) 7.2 Hz,

6H); ¹³C NMR (75 MHz, DMSO-d₆) δ 171.35, 153.7, 134.1,

131.1, 129.4, 123.8, 104.9, 58.2, 55.8, 53.0, 29.4, 23.1,

12.6.7; IR (KBr) 3392, 3180, 2957, 2934, 2870, 2810, 1654,

1584, 1457, 1380 cm^-1; FD mass spectrum m/z 301 (M⁺).

Anal. Calcd for C₁₈H₂₇N₃O: C, 71.72; H, 9.03; N, 13.94.

Found: C, 71.48, H, 9.15; N, 13.75.

P r epar ation of LY228729. (4R)-4-(Di-n -pr opylam i-

n o)-6-(am in ocar bon yl)-1,3,4,5-tetr ah ydr oben z[cd]in -

d ole (3). P d /C Oxid a tion . To a slurry of 1.25 g of 10%

Pd/C in 100 mL of deionized water was added 4.1 g (41.8

mmol) of concentrated H₂SO₄. To this mixture was aded

25.0 g (83.1 mmol) of indoline 16b in 125 mL of MeOH.

The mixture was heated to reflux (approximately 80 °C)

for 7 h. The reaction mixture was cooled to 25 °C and

filtered to remove the catalyst. The catalyst was washed

with water (2 × 60 mL) and ethyl acetate (125 mL). An

additional 125 mL of ethyl acetate was added to the

filtrate, and the layers were separated. The aqueous

phase was diluted with 100 mL of H₂O and then

extracted with 2 × 200 mL of ethyl acetate. The aqueous

phase was combined with 500 mL of ethyl acetate, and

then 75 mL of 5 M NaOH solution was added to the

mixture. The organic layer was separated, washed with

250 mL of H₂O, dried with Na₂SO₄, and concentrated in

vacuo to approximately 100 mL total volume. The

resulting slurry was cooled (with stirring) to 5 °C for 2 h

and filtered. The filter cake was washed with cold ethyl

acetate and vacuum-dried at 50 °C to give 17.13 g (69%

yield) of LY228729 as a white solid: mp 178-80 °C (500

MHz, DMSO-d₆) δ 10.74 (s, 1H), 7.41 (s, 1H), 7.28 (d, J

) 8.4 Hz, 1H), 7.09 (d, J ) 8.4 Hz, 1H), 7.0 (s, 1H), 3.38

(d, J ) 14.1 Hz, 1H), 3.32 (s, 1H), 2.99 (m, 1H), 2.87 (dd,

J ) 14.2, 3.2 Hz), 2.6.8 (d, J ) 13.7 Hz, 1H), 2.52 (m,

4H), 1.40 (m, 4H), 0.88 (t, J ) 7.3 Hz, 6H); ¹³C NMR (125

MHz, DMSO-d₆) δ 171.6, 135.1, 132.3, 127.4, 123.9, 123.0,

120.7, 113.9, 108.6, 59.21, 53.2, 29.9, 24.7, 24.7, 23.1,

12.6; IR (KBr) 3320, 2960, 2880, 1645, 1602, 1591, 1467,

1447 cm^-1; EI-MS m/z ) 300 (M + 1), 199, 156, 129; UV

(EtOH) λ_max) 243 (ꢀ ) 34 700), 284 (ꢀ ) 5600); TLC R_f

) 0.11 (SiO₂, 42:42:16 EtOAc:Hex:TEA). Anal. Calcd for

C₁₈H₂₅N₃O: C, 72.33; H, 8.16; N, 13.83. Found: C, 72.21;

H, 8.42; N, 14.03. [R]_D) -104.03 (589 nm), 29.76 mg/

3.0 mL THF; 5.0 cm path. [R]_D) -480.04 (365 nm). pK_a

) 8.6, DMF/H₂O.

filtered through diatomaceous earth filter aid, and the

filter pad was washed with CH₂Cl₂(6 × 50 mL). The

filtrate was placed in a flask with an overhead stirrer,

and 100 mL of water was added. The mixture was cooled

to 0 °C, and 50% NaOH (75 mL) was added dropwise over

20 min. The two-phase mixture was filtered through a

filter aid to remove precipiated manganese salts. The

organic layer was isolated from the filtrate, and the

aqueous layer was extratcted with 200 mL of CH₂Cl₂. The

combined CH₂Cl₂layers were washed with 200 mL of

brine, dried (Na₂SO₄), and concentrated in vacuo to give

18.0 g of crude LY228729 as an amber foam. The foam

was flash chromatographed on silica gel (eluent 1:1

EtOAc:hexane) to give 13.06 g (62% yield) of purified

product as a yellow foam.

P r ep a r a tion of LY228729 Hip p u r a te (3). To a

slurry of 5.0 g of 10% Pd/C in 100 mL of deionized water

was added 9.9 g (33 mmol) of indoline 16b in 50 mL of

MeOH. An additional 50 mL of MeOH was added, and

the mixture was heated to reflux for 2 h. The reaction

mixture was cooled to 25 °C and filtered to remove the

catalyst. The catalyst was washed with ethyl acetate (3

× 50 mL). The filtrate was concentrated in vacuo to

remove most of the MeOH and ethyl acetate. The

remaining residue was partitioned between 200 mL of

ethyl acetate and 100 mL of 1 N HCl. The lower acid

layer was isolated, and the pH was adjusted to app-

proimately 11 by adding 55 mL of 2 N NaOH. The basic

aqueous phase was extracted with CH₂Cl₂(2 × 100 mL).

The combined CH₂Cl₂layers were dried (Na₂SO₄) and

concentrated in vacuo to give 7.0 g of LY228729 as a tan

foam. The foam was dissolved in 30 mL of i-PrOH and

added to a 70 °C solution of hippuric acid (4.19 g, 23.4

mmol) in 30 mL of i-PrOH maintaining the temperature

at 65-70 °C. The solution was allowed to cool to 25 °C

and stirred for 2 h. The resulting slurry was cooled to 5

°C for 2 h and filtered. The solid was washed with cold

i-PrOH (3 × 25 mL) and vacuum-dried at 40 °C to give

9.63 g (61% overall yield) of the hippurate salt as a white

1

solid: mp 187-88 °C; [R]_D) -39.4 (c ) 1, MeOH); H

NMR (500 MHz, DMSO-d₆) δ 10.80 (s, 1H), 8.73 (t, J )

5.6 Hz, 1H), 7.88 (d, J ) 7.1 Hz, 2H), 7.54 (m, 1H), 7.48,

(m, 3H), 7.31 (d, J ) 8.5 Hz, 1H), 7.11 (d, J ) 8.5 Hz,

1H), 7.06 (s, 1H), 7.02 (1H, s), 3.92 (d, J ) 5.8 Hz, 2H),

3.46 (dd, J ) 16.0, 2.4 Hz, 1H), 3.13 (m, 1H), 3.02 (m,

1H), 2.92 (dd, J ) 14.5, 3.3 Hz, 1H), 2.76 (t, J ) 13 Hz,

1H), 2.61 (m, 4H), 1.46 (m, 4H), 0.89 (t, J ) 7.3 Hz, 6H);

¹³C NMR (125 MHz, DMSO-d₆) δ 172.3, 171.6, 167.2,

135.2, 134.9, 132.7, 131.8, 129.2, 128.1, 127.3, 123.9,

123.0, 120.3, 113.4, 108.7, 62.9, 59.5, 53.2, 42.4, 29.7,

26.3, 24.4, 22.5, 12.5. IR (KBr) 3458, 3134, 2975, 1655,

1604, 1545, 1457, 1384 cm^-1; FAB mass spectrum m/z

300 (M + 1 for amine). Anal. Calcd for C₂₇H₃₄N₄O₄: C,

67.76; H, 7.16; N, 11.71. Found: C, 67.96, H, 7.18; N,

11.61.

N^r-(Tr iflu or oacetyl)-N¹-ben zoyl-2,3-dih ydr o-L-tr yp-

top h a n (26a ). Ca ta lytic Hyd r ogen a tion P r oced u r e.

N^R-(Trifluoroacetyl)-L-tryptophan (24)²³(20 g, 66.7 mmol),

platinum oxide (1.0 g, 4.4 mmol), and a mixture of 100

mL of water and 100 mL of TFA were added sequentially

to a Parr pressure bottle. The mixture was shaken under

10 psi of hydrogen at 25 °C for 2 h, at which time HPLC

analysis showed that all starting material had been

consumed. The catalyst was removed by filtration through

diatomaceous earth. The filtrate was cooled to 0 °C, and

the pH was adjusted to 8 by the addition of 150 mL of

25% NaOH solution. The resulting solution was diluted

P r ep a r a tion of LY228729 (3). Ma n ga n ese Dioxid e

Oxid a tion . Acetic acid (88 mL) was added to a slurry

of indoline 16b (21.0 g, 70 mmol) in 88 mL of CH₂Cl₂at

-10 °C. Solid MnO₂(9.1 g, 104 mmol) was added, and

the mixture was stirred for 5.5 h. The cold mixture was

8652 J . Org. Chem., Vol. 62, No. 25, 1997

Carr et al.

with 120 mL of of 1 M NaHCO₃solution, and benzoyl

chloride (7.3 mL, 63 mmol) was added dropwise. After

30 min the solution was cooled to 0 °C and 25 mL of 6 N

HCl was added to adjust the pH to 3. The precipitate

was filtered, washed with water, and vacuum-dried to

give 17.25 g of a mixture of diastereomers as a light

brown solid. The solid was dissolved in 85 mL of CHCl₃

at 25 °C and then cooled to -20 °C. After 7 days, the

precipitate was filtered, washed with cold CHCl₃, and

vacuum-dried to give 4.34 g (18% yield) of a 98:2 mixture

of amides. An analytical sample of the desired amide

was obtained by recrystallization from 90:10 dichloro-

ethane:EtOAc (8 mL/g): mp 175-78 °C; [R]_D) -36.1 (c

) 0.1, MeOH). ¹H NMR (300 MHz, DMSO-d₆) δ 13.15

(s, 1H), 9.85 (d, J ) 6 Hz, 1H), 7.5 (m, 6H), 7.32 (m, 2H),

7.05 (m, 1H), 4.40 (m, 1H), 4.15 (t, J ) 10 Hz, 1H), 3.77

(m, 1H), 3.40 (m, 1H), 3.27 (m, 1H), 3.05 (m, 1H); ¹³C

NMR (75 MHz, DMSO-d₆) δ 171.5, 167.9, 156.7, 156.2,

142.2, 136.8, 135.1, 130.1, 128.4, 127.6, 126.9, 124.3,

123.8, 117.6, 113.8, 56.2, 51.3, 45.0, 37.3, 34.8; IR (KBr)

3250, 3070, 2990, 2460, 1707, 1700, 1612, 1560, 1485,

1436 cm^-1; FD mass spectrum m/z 406 (M⁺), 105. Anal.

Calcd for C₂₀H₁₇N₂O₄F₃: C, 59.12; H, 4.22; N, 6.89.

Found: C, 58.98, H, 4.30; N, 6.91.

N^r-(Tr iflu or oacetyl)-N¹-ben zoyl-2,3-dih ydr o-L-tr yp-

top h a n (26a ). Red u ction w ith Tr ieth ylsila n e. N^R-

(Trifluoroacetyl-^L-tryptophan (24)²³(125 g, 0.42 mol) was

added to a stirred mixture of 755 mL of TFA and

triethylsilane (133 mL, 0.84 mol). The mixture was

stirred for 40 min at 25 °C and concentrated in vacuo.

The resulting oily residue was partitioned between 250

mL of 2 N HCl and 200 mL of EtOAc and 100 mL of

EtOAc. The layers were separated, and the organic layer

was extracted with 2 N HCl (5 × 100 mL). The pH of

the aqueous layer was adjusted to 3.0 by the dropwise

addition of a 50% NaOH solution. The cloudy solution

was then extracted with EtOAc (6 × 200 mL). The

combined EtOAc extracts were dried (Na₂SO₄) and

concentrated in vacuo to give 125.9 g of a mixture of

indoline diasteroemers as a tan foam.

To a solution of crude indoline (as prepared above,

100.0 g, 0.33 mol) and triethylamine (49 mL, 0.35 mol)

in 700 mL of dry THF was added, dropwise, a solution

of benzoyl chloride (36.5 mL, 0.31 mol) in 230 mL of THF.

The reaction mixture was then stirred for 1 h, diluted

with 1 L of Et₂O, and then added to 1 L of 2 N HCl. The

resulting mixture was stirred for 5 min, and the layers

were separated. The aqueous layer was extracted with

500 mL of Et₂O. The combined organic layers were

washed with brine, dried (Na₂SO₄), and concentrated in

vacuo to give an oily brown foam. The foam was

dissolved in 300 mL of hot CHCl₃and cooled to 0 °C for

24 h. The resulting solid was filtered, washed with cold

CHCl₃, and vacuum-dried to give 27.24 g (21% yield from

tryptophan trifluoracetamide) of the carboxylic acid as a

tan crystal. HPLC analysis showed this to be a 97:3

mixture of diastereomers. The mother liquor was con-

centrated to an an oil and redissolved in 150 mL of hot

CHCl₃. Cooling to 0 °C for 24 h provided a second crop

of 48.58 g. HPLC analalysis showed this to be a 44:56

ratio of diastereomers. Diastereomer ratios were deter-

mined by dissolving the indolines in HOAc and treating

the solution with excess 1 M Br₂in HOAc. The solution

was stirred for 5 min, dissolved in pH 3.5 chloroacetic

acid buffer, and treated with a 10% Na₂SO₃solution to

decompose the excess bromine. The resulting solution

of the 7-bromoindolines was assayed by HPLC: Zorbax

C-18 25 cm column; 35:65 CH₃CN:pH 3.5 chloroacetic

acid buffer eluent; 2 mL/min; UV detection at 275 nm.

(2a R,4S)-1-Ben zoyl-4-t r iflu or oa cet a m id o-5-oxo-

1,2,2a ,3,4,5-h exa h yd r oben z[cd ]in d ole ((+)-30a ). To

a mixture of indolinecarboxylic acid (26a ) (24.00 g, 59

mmol) and DMF (0.93 mL, 12 mmol) in 240 mL of CH₂-

Cl₂at -10 °C was added a solution of oxalyl chloride (10.3

mL, 118 mmol) in 10 mL of CH₂Cl₂over 8 min. The

mixture was warmed to -5 °C and stirred 1 h, at which

time all of the carboxylic acid had dissolved. The reaction

mixture was cooled to -10 °C, and solid AlCl₃(39.23 g,

295 mmol) was added. The mixture was stirred for 1 h

at -10 °C, and the cooling bath was removed. After 22

h, the reaction mixture was added in a stream to a stirred

mixture of 200 g of ice, 50 mL of concentrated HCl, and

100 mL of CH₂Cl₂. The reaction flask was rinsed with

100 mL of CH₂Cl₂which was added to the quench

mixture. The layers were separated, and the aqueous

layer was extracted with 100 mL of CH₂Cl₂. The combine

organic layers were washed with brine (250 mL), satu-

rated NaHCO₃solution (250 mL), and then brine (250

mL). After the solution was dried with Na₂SO₄, the

organic layer was concentrated in vacuo to give 20.6 g of

a light yellow foam. The foam was crystallized from 100

mL of boiling n-BuOH to give 15.70 g (69% yield) of the

ketone as a white solid: mp 210-212 °C; [R]_D) 183 (c )

1

0.1, CHCl₃); H NMR (500 MHz, CDCl₃) δ 8.0 (bs, 1H),

7.58 (m, 5H), 7.49 (m, 2H), 7.28 (m, 1H), 4.69 (m, 1H),

4.43 (m, 1H), 3.87 (m, 1H), 3.77 (t, J ) 10.6 Hz, 1H), 3.02

(m, 1H), 1.95 (q, J ) 12.0 Hz, 1H); ¹³C NMR (125 MHz,

CDCl₃) δ 192.3, 169.3, 157.8 (q for CF₃), 142.7, 141.6,

135.9, 131.6, 129.8, 129.2, 128.5, 127.7, 121.4, 119.4,

117.2, 114.9, 57.1, 35.5. IR (KBr) 3290, 3070, 1730, 1721,

1691, 1643, 1634, 1595, 1470, 1393 cm^-1

; FD mass

spectrum m/z 388 (M⁺). Anal. Calcd for C₂₀H₁₅N₂O₃F₃:

C, 61.86; H, 3.89; N, 7.21. Found: C, 61.78, H, 3.88; N,

7.14. Chiral HPLC assay: SS Whelk 01 S/N 100022

column; 60:40 ethanol:hexane eluent; 1 mL/min; UV

detection at 275 nm. Retention times: ketone diastere-

omer 30b, 4.9 min; ketone diastereomer 30a , (-)-enan-

tiomer, 5.8 min; ketone diastereomer 30a , (+)-enanti-

omer, 6.6 min.

(2a S,4S)-1-Ben zoyl-4-t r iflu or oa cet a m id o-5-oxo-

1,2,2a,3,4,5-h exah ydr oben z[cd]in dole (30b). Friedel-

Crafts reaction of 2.44 g of a 45:55 mixture of carboxylic

acids, as described above, gave 1.82 g of crude ketone as

a mixture of diastereomers. The mixture was flash

chromatographed on silica gel (eluent 1:1:1 hexane:Et₂O:

CH₂Cl₂) to give 1.05 g of ketone diastereomer (+)-30a

and 0.34 g of a 65:35 mixture (by HPLC, conditions

desribed above) of ketones diastereomers 30a :30b. Ke-

tone 30b was isolated as a white foam from this mixture

by preparative HPLC on silica gel (eluent 2.5% i-PrOH

in hexane): ¹H NMR (500 MHz, CDCl₃) δ 8.4-8.0 (bs,

1H), 7.55 (m, 3H), 7.49 (m, 3H), 7.28 (m, 2H), 4.62 (t, J

) 5.2 Hz, 1H), 4.42 (m, 1H), 3.78 (m, 2H), 2.67 (d, J )

13.2 Hz, 1H), 2.22 (m, 1H); ¹³C NMR (125 MHz, CDCl₃)

δ 191.6, 169.4, 157.9 (q for CF₃), 142.3, 141.5, 135.9,

131.6, 129.7, 129.2, 129.1, 127.6, 121.6, 119.4, 117.1,

114.9, 112.6, 53.8, 33.7; IR (KBr) 3290, 3070, 1730, 1721,

1691, 1643, 1634, 1595, 1470, 1393 cm^-1; FD mass

spectrum m/z 388 (M⁺).

(4S)-1-Ben zoyl-4-t r iflu or oa cet a m id o-5-h yd r oxy-

1,2,2a ,3,4,5-h exa h yd r ob en z[cd ]in d ole (31). Solid

NaBH₄(0.50 g, 13.2 mmol) was added to a -10 °C

suspension of ketone (+)-30a (13.90 g, 35.8 mmol) in 20

mL of THF. Methanol (20 mL) was added dropwise over

Partial Ergot Alkaloid LY228729

J . Org. Chem., Vol. 62, No. 25, 1997 8653

8 min, keeping the temperature below 2 °C. The reaction

mixture was stirred for 1 h at 0 °C and then quenched

by the addition of 70 mL of 1 N HCl over 20 min (the

temperature was maintained below 5 °C). Water (50 mL)

was added to preceipitate the product, and the mixture

was stirred for an additional 1 h at 0 °C. After filtration

the filter cake was washed with water (4 × 30 mL) and

dried to give 12.59 g (90% yield) of alcohol as an 89:11

mixture of epimers at C-5. An analytical sample of the

trans alcohol was obtained by recrystallization from

n-BuOH: mp 250-253 °C; [R]_D) 93.7 (c ) 0.1, THF);

¹H NMR (500 MHz, DMSO-d₆) δ 9.50 (d, J ) 8.5 Hz, 1H),

7.60 (m, 2H), 7.50 (m, 3H), 7.13 (m, 2H), 5.71 (d, J ) 7.3

Hz, 1H), 4.73 (t, J ) 7.6 Hz, 1H), 4.10 (m, 1H), 3.76 (m,

1H), 3.50 (m, 1H), 2.09 (m, 1H), 1.76 (q, J ) 12.1 Hz,

1H); ¹³C NMR (125 MHz, DMSO-d₆) δ 168.5, 156.8 (q),

141.7, 137.4, 133.6, 131.3, 129.3, 128.8, 128.0, 123.1,

120.4, 118.1, 115.9, 70.5, 55.4, 33.7; IR (KBr) 3279, 1698,

1627, 1593, 1471, 1461, 1232, 1214 cm^-1; FD mass

spectrum m/z 390 (M⁺). Anal. Calcd for C₂₀H₁₇N₂O₃F₃:

C, 61.54; H, 4.39; N, 7.18. Found: C, 61.80, H, 61.80; N,

7.24.

(4S )-1-Be n zoyl-4-t r iflu or oa ce t a m id o-1,2,2a ,3,4,

5-h exa h yd r oben z[cd ]in d ole (32). A solution of alcohol

31 (750 mg, 1.9 mmol) and trifluoroacetic anhydride (0.75

mL, 5.3 mmol) in 12 mL of THF was added to 375 mg of

10% Pd-C under N₂. The mixture was hydrogenated

under 45 psi of hydrogen for 22 h at 25 °C. The catalyst

was removed by filtration, and the filter cake was washed

with 60 mL of EtOAc. The filtrate was concentrated to

give the crude product as a white solid. The solid was

chromatographed on silica gel (1:1:1 hexane:CH₂Cl₂:Et₂O

eluent) to give 624 mg (86% yield) of white solid, mp

185-6 °C: [R]_D) 65 (c ) 0.1, THF); ¹H NMR (500 MHz,

CDCl₃) δ 7.52 (m, 6H), 6.87 (m, 3H), 4.38 (m, 2H), 3.61

(t, J ) 11 Hz, 1H), 3.40 (m, 1H), 3.28 (dd, J ) 16.8, 6.5

Hz), 2.60 (dd, J ) 16.8, 11 Hz), 2.36 (m, 1H), 1.47 (q, J )

11 Hz, 1H); ¹³C NMR (125 MHz, CDCl₃) δ 169.3, 157.3

(q), 141.8, 136.6, 132.0, 131.3, 129.1, 128.7, 127.7, 123.3,

119.7, 117.4, 115.1, 112.8, 58.0, 47.5, 37.0, 32.9, 32.7; IR

(KBr) 3290, 1693, 1652, 1555, 1458, 1390, 1205, 1164

cm^-1; FD mass spectrum m/z 374 (M⁺). Anal. Calcd for

C₂₀H₁₇N₂O₂F₃: C, 64.17; H, 4.58; N, 7.48. Found: C,

64.38, H, 4.50; N, 7.52.

(4S)-1-Ben zoyl-4-am in o-1,2,2a,3,4,5-h exah ydr oben -

z[cd ]in d ole (13). Amide 32 (480 mg, 1.28 mmol) was

dissolved in a mixture of 7 mL of THF and 2 mL of 1 N

NaOH. The reaction mixture was heated to reflux (70

°C) for 3 h, at which time HPLC indicated the reaction

to be complete. To the cooled reaction mixture were

added 10 mL of EtOAc and 10 mL of water. The layers

were separated, and the aqueous layer was extracted

with 10 mL of EtOAc. The combined organic layers were

extracted with 10 mL of a saturated NaHCO₃solution,

dried (Na₂SO₄), and concentrated in vacuo to obtain 327

mg (91% yield) of the amine as a white solid. This

compound was identical by HPLC and NMR to amine

13 prepared from Kornfeld’s ketone. An analytical

sample was obtained by recrystallization of the solid from

EtOAc: mp 146-48 °C; [R]_D) 61.6 (c ) 0.1, THF). Anal.

Calcd for C₁₈H₁₈N₂O: C, 77.67; H, 6.52; N, 10.06.

Found: C, 77.44, H, 6.52; N, 10.00. FD mass spectrum

m/z 278 (M⁺).

Ack n ow led gm en t. We are grateful to Professors

Marvin Miller, Leo Paquette, Bill Roush, Ted Taylor,

and Paul Wender for helpful discussions during the

course of this work. We are also indebted to Professor

Eric J acobsen for his help and advice with the asym-

metric epoxidation of olefin 5, as well as other aspects.

We thank Drs. Gerald Thompson and Charles Paget for

their encouragement and support.

J O971256Z

Article Doi

Synthetic studies toward the partial ergot alkaloid LY228729, a potent 5HT(1A) receptor agonist

DOI: 10.1021/jo971256z

Source and publish data:

Authors:

Article abstract of DOI:10.1021/jo971256z

Full text of DOI:10.1021/jo971256z

Products guided by the article

R&D Labs maybe for 132619-31-9

Relevant to this article

Hot Product