Total synthesis of (-)-7-epicylindrospermopsin, a toxic metabolite of the freshwater cyanobacterium Aphanizomenon ovalisporum, and assignment of its absolute configuration

VOLUME 70, NUMBER 6

MARCH 18, 2005

Total Synthesis of (-)-7-Epicylindrospermopsin, a Toxic

Metabolite of the Freshwater Cyanobacterium Aphanizomenon

ovalisporum, and Assignment of Its Absolute Configuration

James D. White* and Joshua D. Hansen

Department of Chemistry, Oregon State University, Corvallis, Oregon 97331-4003

james.white@oregonstate.edu

Received August 5, 2004

The Z and E nitrones 38 and 39 from condensation of aldehyde 20 with hydroxylamine 36 underwent

intramolecular dipolar cycloaddition to give the substituted 1-aza-7-oxobicyclo[2.2.1] heptanes 40

and 41 in a ratio of 2:1, respectively. Reductive N-O bond cleavage of 40 followed by carbonylation

gave cyclic urea 47 in which inversion of the secondary alcohol was effected via an oxidation-

reduction sequence. After conversion of the p-bromobenzyl ether 50 to azide 54, activation of the

cyclic urea as its O-methylisourea and reduction of the azide led to spontaneous cyclization to afford

the tricyclic nucleus 59 of cylindrospermopsin. Global deprotection, including hydrolysis of the 2,4-

dimethyoxypyrimidine appendage to a uracil, and then monosulfation of the resultant diol 60

afforded a substance identical with natural (-)-7-epicylindrospermopsin (1). The asymmetric

synthesis of (-)-7-epicylindrospermopsin defines its absolute configuration as 7S,8R,10S,12S,13R,-

14S.

Introduction

and consequently, C. raciborskii must now be considered

an environmental hazard of some importance.³

An outbreak of hepatoenteritis in 1979 among the

human population of Palm Island, situated off the

Queensland coast of Australia, was later traced to the

presence of a cyanobacterium Cylindrospermopsis raci-

borskii in the drinking water supply.¹C. raciborskii,

originally discovered in Indonesia, is now quite common

in phytoplankton of temperate waters and is often the

dominant species in algal blooms where incidence of

gastrointestinal disease is found.²For example, blooms

of the organism have been noted in eutrophicated lakes

and rivers in Hungary, Israel, Brazil, Japan, and Florida,

Aqueous extraction of C. raciborskii by Moore and co-

workers furnished a novel alkaloid which they named

cylindrospermopsin and to which they assigned structure

1.⁴The same alkaloid was isolated from the alga Umeza-

kia natans by Harada⁵and from Aphanizomenon oval-

isporum by Banker.⁶A. ovalisporum was also found to

produce a minor toxic metabolite which was shown to be

a stereoisomer of 1 and which was assigned structure

2.⁷A substance designated deoxycylindrospermopsin has

also been isolated from C. raciborskii,⁸but the structure

of this metabolite remains uncertain.

Cylindrospermopsin was found to be cytotoxic to rat

hepatocytes with an in vivo LD₅₀of 2.1 mg/kg for a 24 h

assay, decreasing to 0.2 mg/kg when a 5-day kill rate was

(1) Byth, S. Med. J. Aust. 1980, 2, 40.

(2) Hawkins, P. R.; Runnegar, M. T.; Jackson, A. R.; Falconer, I. R.

Appl. Environ. Microbiol. 1985, 1292.

Published on Web 11/02/2004

J. Org. Chem. 2005, 70, 1963-1977

1963

White and Hansen

Thus, cylindrospermopsin is correctly represented by 2,

and 1 is 7-epicylindrospermopsin. The absolute configu-

ration of the two metabolites was unknown, however, and

biosynthetic evidence¹⁵provided no reason to favor enan-

tiomers 1 and 2 over their antipodes. A major purpose of

our synthesis, therefore, was to establish the absolute

configuration of 1 and/or 2 while exploring an asymmetric

intramolecular dipolar cycloaddition strategy as a key

step in constructing a portion of the cylindrospermopsin

framework.¹⁶

measured.⁹Unlike other hepatotoxic metabolites of cy-

anobacteria such as the microcystins,¹⁰cylindrospermop-

sin does not exert its toxic effect through specific inhi-

bition of protein phosphatases but through interruption

of the biosynthetic steps leading to cell-reduced glu-

tathione.¹¹This inhibition is complete at 2.5 µM but the

precise mechanism of inhibition is unknown. Renal

toxicity as well as neurotoxicity have also been associated

with cylindrospermopsin.¹²Interestingly, although 7-epi-

cylindrospermopsin possesses toxicity similar to its ster-

eoisomer,⁷deoxycylindrospermopsin is reported to be

devoid of biological activity.⁸

The structure assignment made to cylindrospermopsin

by Moore⁴rested on the precarious assumption that the

uracil moiety existed as an anomalous enol tautomer

which permitted intramolecular hydrogen bonding with

a guanidine nitrogen and fixed a conformation in which

H-7 would occupy a gauche relationship to H-8. The same

assumption when applied to 7-epicylindrospermopsin

would place H-7 and H-8 in an antiperiplanar orientation.

Although the NMR evidence reported by Moore was not

entirely consistent with these conclusions, a synthesis

of (()-cylindrospermopsin by Snider appeared to confirm

Moore’s structural attribution to this alkaloid.¹³It was

apparent, however, that the Snider synthesis does not

distinguish between cylindrospermopsin stereoisomers

which differ in configuration at C-7. On the other hand,

the synthesis of 7-epicylindrospermopsin by Weinreb does

define the relative configuration at this center and proves

that the assignments previously made to cylindrosper-

mopsin and 7-epicylindrospermopsin must be reversed.¹⁴

Our goal at the outset was an asymmetric synthesis

of the presumed structure of cylindrospermopsin, i.e., 1,

but as events unfolded it became clear that the relative

configuration of cylindrospermopsin was incorrectly rep-

resented by this structure. When it was shown by

Weinreb that 1 was the C-7 epimer of cylindrospermop-

sin,¹⁴our target was redefined as the natural enantiomer

of 7-epicylindrospermopsin. It was hoped, nevertheless,

that by determining the absolute configuration of the

7-epi compound we could establish the absolute stereo-

chemistry of cylindrospermopsin as well even though no

structural correlation existed between the two compounds

at that time. Our synthesis plan for 1 is shown in Scheme

1 and centers upon construction of a 1-aza-7-oxabicyclo-

[2.2.1]heptane 3 as the prelude to elaboration of the

tetrasubstituted piperidine (ring A) of 4. The logical

precursor to 3 is (Z)-nitrone 5, which we anticipated

would undergo a stereoselective intramolecular dipolar

cycloaddition to the vinyl substituent in a boat conforma-

tion which has minimal steric interactions between

intervening substituents.¹⁸Nitrones are typically as-

sembled from condensation of an aldehyde with a hy-

droxylamine, and hence, the requisite building blocks for

the synthesis of 5 are projected as 6 (Western Fragment)

and 7 (Eastern Fragment). This plan partitions the four

stereogenic centers of 5 equally between its two precur-

sors and also positions members of each pair of stereo-

centers in a vicinal relationship. This enables a single

center in each subunit to exercise stereocontrol in the

asymmetric construction of that subunit. The uracil

substituent of 1 is concealed as a 2,4-dimethoxypyrimi-

dine through most of the synthesis, to be unmasked at a

late stage by hydrolysis, as was done in the Snider¹³and

Weinreb¹⁴routes.

(3) (a) Nagy, L.; Hiripi, L.; Kovacs, A.; Voros, L. Hidrol. Kozl. 1998,

78, 298. (b) Presing, M.; Herodek, S.; Voros, L.; Kobor, I. Arch.

Hydrobiol. 1996, 136, 553. (c) Sukenik, A.; Rosin, C.; Porat, R.; Telsch,

B.; Banker, R.; Carmeli, S. Isr. J. Plant Sci. 1998, 46, 109. (d) de Souza,

R. C. R.; Carvalho, M. C.; Truzzi, A. Environ. Toxicol. Water Qual.

1998, 13, 73. (e) Chapman, A. D.; Schelske, C. L. J. Phycol. 1997, 33,

191.

Results and Discussion

Synthesis of Eastern Fragment 7. Construction of

the dimethoxypyrimidine unit needed for this fragment

began with improvement to a procedure reported by

Langely¹⁹for the synthesis of 2,4,6-tribromopyrimidine

from barbituric acid (8).²⁰Treatment of 8 with phospho-

rus oxybromide and triethylamine in hot toluene was

found to give the volatile tribromopyrimidine in quanti-

tative yield if toluene was removed azeotropically from

(4) Moore, R. E.; Ohtani, I.; Runegar, M. T. C. J. Am. Chem. Soc.

1992, 114, 7941.

(5) Harada, K.; Ohtani, I.; Iwamoto, K.; Suzuki, M.; Watanabe, M.

F.; Watanabe, M.; Terao, K. Toxicon 1994, 32, 73.

(6) Banker, R.; Teltsch, B.; Sukenik, A.; Carmeli, S.; Hadas, O.;

Porat. R. J. Phycol. 1997, 33, 613.

(7) Banker, R.; Teltsch, B.; Sukenik, A.; Carmeli, S. J. Nat. Prod.

2000, 63, 387.

(8) Norris, R. L.; Eaglesham, G. K.; Pierens, G.; Shaw, G. R.; Smith,

M. J.; Chiswell, R. K.; Seawright, A. A.; Moore, M. R. Environ. Toxicol.

1999, 14, 163.

(9) Runnegar, M. T.; Kong, S.-M.; Zhong, Y.-Z.; Ge, J.-L.; Lu, S. C.

Biochem. Biophys. Res. Commun. 1994, 201, 235.

(10) Carmichael, W. Sci. Am. 1994, 270, 78.

(15) Burgoyne, D. L.; Hemscheidt, T. K.; Moore, R. E.; Runnegar,

M. T. C. J. Org. Chem. 2000, 65, 152.

(16) White, J. D.; Hansen, J. D. J. Am. Chem. Soc. 2002, 124, 4950.

(17) For a review of the biological properties and synthetic ap-

proaches to cylindropsermopsin, see: Murphy, P. J.; Thomas, C. W.

Chem. Soc. Rev. 2001, 30, 303.

(18) Oppolzer, W.; Siles, S.; Snowden, R. L.; Bakker, B. H.; Petrzilka,

M. Tetrahedron 1985, 41, 3497.

(19) Langley, B. W. J. Am. Chem. Soc. 1956, 78, 2136.

(20) Dimethylaniline facilitates the preparation of 2,4,6-trichloro-

pyrimidine from barbituric acid and phosphorus oxychloride, see:

Baddiley, J.; Topham, A. J. Chem. Soc. 1944, 678.

(11) Runnegar, M.; Kong, S.-M.; Zhong, Y.-Z.; Lu, S. Biochem.

Pharmacol. 1995, 49, 219.

(12) Falconer, I. R.; Hardy, S. J.; Humpage, A. R.; Froscio, S. M.;

Tozer, G. J.; Hawkins, P. R. Environ. Toxicol. 1999, 14, 143.

(13) Xie, C.; Runnegar, M. T. C.; Snider, B. B. J. Am. Chem. Soc.

2000, 122, 5017.

(14) Heintzelman, G. R.; Fang, W.-K.; Keen, S. P.; Wallace, G. A.;

Weinreb, S. M. J. Am. Chem. Soc. 2002, 124, 3939 and references

therein.

1964 J. Org. Chem., Vol. 70, No. 6, 2005

Total Synthesis of (-)-7-Epicylindrospermopsin

SCHEME 1

SCHEME 2

reaction time for the formation of 10 was 5-6 h; shorter

reactions resulted in significant quantities of the monoben-

zylamino lactone, whereas longer reaction periods de-

graded the enantiomeric purity of 10.

As a prelude to coupling the pyrimidine unit with 10,

halogen-metal exchange of bromopyrimidine 9 was

carried out with n-butyllithium. Exchange occurred

rapidly at -78 °C, but treatment of the lithio pyrimidine

alone with lactone 10 was accompanied by ca. 25%

racemization as measured by a deuterium incorporation

experiment. The addition of cerium trichloride to the

reaction mixture completely suppressed racemization of

10,²²and the adduct 11 was produced in high yield as a

mixture of two diastereomeric hemiketals. Not surpris-

ingly, there was no evidence of an equilibrium between

11 and its ring-opened isomer, but exposure of 11 to trityl

chloride in the presence of triethylamine containing a

catalytic quantity of 4-(dimethylamino)pyridine led

smoothly to the protected primary alcohol 12. The reduc-

tion of chiral R-aminoketones has been examined by

Reetz,²³who showed that tertiary amines do not direct

hydride reagents and that reduction follows the Felkin-

Anh model in yielding predominantly a syn amino

alcohol. Accordingly, the reaction of 12 with sodium

borohydride in methanol was found to yield 13 and its

anti epimer in the ratio 8.5:1, respectively. This ratio was

improved to 12:1 in favor of 13 when L-Selectride was

used as the reducing agent at -78 °C. Proof of configu-

ration of the major syn diastereomer 13 was obtained by

its hydrolysis in formic acid to give diol 14, which upon

oxidation with catalytic tetra-n-propylammonium per-

ruthenate and excess N-methylmorpholine N-oxide²⁴

produced γ-lactone 15. The relationship between H-3 and

the mixture (Scheme 2). The tribromopyrimidine was

used without purification in a reaction with sodium

methoxide which displaced two of the three bromo

substituents selectively from the pyrimidine nucleus.

Optimal conditions for this reaction required dropwise

addition of two equivalents of sodium methoxide in

methanol to the pyrimidine and afforded 4-bromo-2,6-

dimethoxypyrimidine (9) in excellent yield.

The second component required for the Eastern Frag-

ment was (R)-2-(dibenzylamino)butyrolactone (10), the

enantiomer of which was previously prepared from (S)-

(+)-methionine by Reetz.²¹We adapted Reetz’ method to

(R)-(-)-methionine and found that reaction of the amino

acid with benzyl bromide and potassium carbonate in

aqueous tetrahydrofuran gave γ-lactone 10 with an

acceptable enantiomeric purity (90% ee) which could be

improved to >98% ee by recrystallization. The optimal

(22) Imamoto, T.; Takiyama, N.; Nakamura, K.; Hatajima, T.;

Kamiya, Y. J. Am. Chem. Soc. 1989, 111, 4392.

(23) Reetz, M. T. Angew. Chem., Int. Ed. Engl. 1991, 30, 1531.

(24) Ley, S.; Norman, J.; Griffith, W. P.; Marsden, S. P. Synthesis

1994, 639.

(21) Reetz, M. T.; Schmitz, A.; Holdgroen, X. Tetrahedron Lett. 1989,

30, 5421.

J. Org. Chem, Vol. 70, No. 6, 2005 1965

White and Hansen

SCHEME 3

SCHEME 4

H-4 in this lactone was shown conclusively to be cis by

means of an NOE experiment in which H-4 was irradi-

ated. A signal enhancement of 18% was observed for H-3

confirming that the amino alcohol 13 from which 15

originated must have possessed syn configuration. There

was also an opportunity with 13 to determine its enan-

tiomeric purity and thus be reassured that no racemiza-

tion of 12 had occurred during its reduction. For this

purpose, racemic 13 was prepared and the enantiomers

were shown to be separable by HPLC on a chiral (OJ)

column. By this criterion, the enantiomeric ratio of 13

was found to be >97:3.

Completion of the Eastern Fragment from 13 required

only protection of the secondary alcohol, release of the pri-

mary alcohol from its trityl ether, and oxidation of the

latter to an aldehyde. To this end, 13 was converted to its

tert-butyldimethylsilyl ether 16 and the trityl group was

removed with formic acid to afford 17 in high yield

(Scheme 3). Unfortunately, no oxidant could be found

which would transform primary alcohol 17 to aldehyde 18.

In every case, mixtures were produced which were pre-

dominantly the result of a retro-Mannich reaction of the

initially generated â-aminoaldehyde 18 or of N-oxide for-

mation from this amine. It was clear from this outcome

that the reactivity of the tertiary amine of 17 should be

suppressed in order to accomplish oxidation of the pri-

mary alcohol, and when the two benzyl groups were

removed by hydrogenolysis in the presence of palladium

hydroxide and the resulting primary amine was protected

as carbamate 19, oxidation of the alcohol took place to

give aldehyde 20 in high yield. The sequence to Eastern

Fragment 20 requires nine steps from (R)-(-)-methionine

(10 from barbituric acid) and proceeds in an overall 20%

yield.

crotylation of this material would afford the (2S,3S)

configuration required for 6 (Scheme 4). Unfortunately,

treatment of the oxime of 21 with the (E)-crotylborane

prepared from trans-2-butene and (-)-diisopinylcam-

pheylmethoxyborane gave an impractically low yield of

the desired hydroxylamine.²⁵The same reagent with

O-protected versions of the oxime were also unsuccessful.

This failure prompted an alternative approach to the

Western Fragment from aldehyde 21. Syn homoallylic

alcohol 22²⁶was obtained with high enantio- and dias-

tereoselectivity from the reaction of 21 with the (Z)-

crotylborane prepared from (+)-diisopinylcampheyl-

methoxyborane and was converted to its mesylate 23 in

anticipation that displacement with O-benzylhydroxy-

lamine would lead to a protected version of 6. Instead,

the sole product of this reaction was oxazolidinone 24

resulting from intramolecular displacement of the me-

sylate by the Boc group.

In an attempt to find a route to 6 which would avoid

neighboring group participation in the displacement seen

with 23, the Boc protecting group of 21 was replaced by

Synthesis of the Western Fragment. Our initial

plan for constructing homoallylic hydroxylamine 6, the

designated reaction partner for 20, envisioned aldehyde

21 as the starting point. The latter was prepared in

straightforward fashion from ethanolamine and was

converted to its oxime in the hope that asymmetric

(25) For a similar result, see: Itsuno, S.; Watanabe, K.; Ito, K.; El-

Shehawy, A. A.; Sarhan, A. A. Angew. Chem., Int. Ed. Engl. 1997, 36,

109.

(26) Brown, H. C.; Bhat, K. S. J. Am. Chem. Soc. 1986, 108, 5919.

1966 J. Org. Chem., Vol. 70, No. 6, 2005

Total Synthesis of (-)-7-Epicylindrospermopsin

SCHEME 5

SCHEME 6

a pair of benzyl substituents. N,N-(Dibenzylamino)-

acetaldehyde (25) was prepared from ethanolamine and

was subjected to the same asymmetric crotylation²⁶used

with 21 (Scheme 4). The resulting homoallylic alcohol 26

was treated with triflic anhydride and then with the tert-

butyldimethylsilyl ether of hydroxylamine in the expec-

tation that displacement would give 6. The first indica-

tion that events had taken a different turn came when

the silyl group was removed from the product under basic

conditions and N-hydroxypyrrolidine 27 was obtained.

This result suggested that the product from 26 was the

rearranged hydroxylamine derivative 28, implying the

intervention of azidirinium ion 29 in the displacement

of triflate from 26.²⁷A firm assignment of structure to

28 became possible when it was reacted with p-nitroben-

zaldehyde in the presence of ammonium fluoride to yield

crystalline nitrone 30 (Scheme 5). X-ray crystallographic

analysis of 30 confirmed its relative stereostructure and

proved that an inversion consistent with 1,2-migration

of the dibenzylamino group had taken place in 28.

The foregoing difficulties associated with carrying a

protected amine into the sequence designed to afford 6

persuaded us to postpone introduction of the amino

function to a later stage when it would be needed for the

eventual cyclization of 4 to the amidine moiety of 1. This

revision to the synthesis plan required versions of 3 and

5 (Scheme 1) in which the X group is a protected alcohol,

and our starting point would now be ethylene glycol

rather than ethanolamine. The mono-p-bromobenzyl

ether of ethylene glycol was oxidized to aldehyde 31

which underwent asymmetric crotylation²⁶to give syn

homoallylic alcohol 32 with excellent stereoselectivity

(Scheme 6). Attempts to effect displacement of derivatives

of 32, such as its triflate, with O-tert-butyldimethylsilyl-

hydroxylamine led primarily to a conjugated diene by

elimination, but reaction of the mesylate of 32 with

sodium azide in dimethylformamide was more successful.

The inverted (anti) azide was obtained in good yield

accompanied by ca. 10% of elimination product, and the

azide was reduced to amine 33 with triphenylphos-

phine.²⁸This left us with the task of oxidizing the amine

to a hydroxylamine without causing oxidation of the vinyl

group. Fortunately, an excellent protocol has been pub-

lished by Holmes for exactly this transformation,²⁹and

we exploited this method by first condensing 33 with

p-anisaldehyde and then oxidizing the resultant imine

34 with m-chloroperbenzoic acid. Rapid oxaziridine for-

mation took place at 0 °C, uncomplicated by epoxidation

of the vinyl group, and gave 35 as a 1:1 mixture of

diastereomers at the oxaziridine carbon. This mixture,

upon exposure to hydroxylamine hydrochloride, smoothly

furnished hydroxylamine 36 after removal of the byprod-

uct, p-anisaldoxime (37). The sequence from ethylene

glycol to 36, although somewhat lengthier than was

originally projected, was amenable to scale-up and al-

lowed us to prepare multigram quantities of the hydroxy-

lamine 36 for condensation with aldehyde 20.

Union of Fragments and Completion of the Syn-

thesis. The coupling of Eastern and Western Fragments

20 and 36 proceeded efficiently in refluxing methanol

when 3 Å molecular sieves were present to remove the

water of condensation (Scheme 7). Under these condi-

tions, (Z)-nitrone 38 was obtained as a single isomer.

With the stage now set for an intramolecular 1,3-dipolar

cycloaddition leading to ring A of 1, we invested sub-

stantial effort in optimizing the conditions for this key

step. We quickly found that the preferred solvent was

toluene and that there was a narrow temperature

window within which cycloaddition was practical. Above

110 °C, the nitrone decomposed rapidly, whereas below

95 °C there was little reaction. In addition, we found that

within the 95-110 °C range there was slow isomerization

of the (Z)-nitrone 38 to its (E)-isomer 39 and that the

(E) isomer also underwent intramolecular cycloaddition.

The (Z)- to (E)-nitrone isomerization was accelerated by

Lewis acid catalysts such as 5 M lithium perchlorate in

ether,³⁰with no benefit to the cycloaddition process. The

optimized conditions for intramolecular cycloaddition of

(27) For a similar rearrangement, see: Gmeiner P.; Junge, D.;

Kaertner, A. J. Org. Chem. 1994, 59, 6766.

(28) Gololobov, Y. G.; Kasukhin, L. F. Tetrahedron 1992, 48, 1353.

(29) (a) Smith, A. L.; Williams, S. F.; Holmes, A. B.; Hughes, L. R.;

Swithenbank, C.; Lidert, Z. J. Am. Chem. Soc. 1988, 110, 8696. (b)

Holmes, A. B.; Smith, A. L.; Williams, S. F.; Hughes, L. R.; Swithen-

bank, C.; Lidert, Z. J. Org. Chem. 1991, 56, 1393.

J. Org. Chem, Vol. 70, No. 6, 2005 1967

White and Hansen

SCHEME 7

SCHEME 8

difficulty (Scheme 9). Although both samarium diiodide

and Raney nickel were effective reagents for the reduc-

tion of 40, the most convenient method involved the use

of zinc dust and ammonium chloride at elevated temper-

ature. An important attribute of this relatively mild

reduction was preservation of the silyl ether, and after

removal of the Boc protecting group in acidic methanol

the piperidine 46 was obtained in an overall 68% yield

from nitrone 38.

With 46, a substantial portion of the cylindrospermop-

sin functionality and five of the six stereocenters of 1

were in hand. However, the aberrant configuration at

C-12 remained a significant challenge since it appeared

unlikely for steric reasons that a Mitsunobu-type inver-

sion would be effective at this site. On the other hand,

reduction of a C-12 ketone by a bulky hydride reagent

should deliver an axial alcohol with the required (S)

configuration. Our concern with this tactic was that the

keto piperidine from oxidation of 45 or 46 could undergo

retro-Mannich fragmentation which would scramble the

stereochemistry of substituents attached to the piperi-

dine. Indeed, when oxidation of alcohol 45 was attempted,

it became clear that this fear was well grounded. Removal

of basicity from the piperidine nitrogen prior to oxidation

at C-12 was an obvious remedy for this situation, and

the proximal amino group at C-8 afforded a handy

partner for this purpose. Exposure of 46 to carbonyldimi-

dazole conveniently bridged the amine and piperidine

nitrogens to produce a cyclic urea 47 in which acylation

of the C-12 alcohol had accompanied urea formation. The

38 led to a 2:1 mixture of isomeric oxazabicyclo[2.2.1]-

heptanes 40 and 41 which could only be separated

chromatographically with substantial loss of material.

High dilution proved unnecessary for the reaction since

no products of intermolecular cycloaddition were ever

observed. The mode of cycloaddition of 38 conforms to

results published by Oppolzer who examined the influ-

ence of chain length between the nitrone and olefin on

the regiochemistry of the reaction¹⁸and is also consistent

with findings of Hoffman and Endesfelder,³¹who showed

that (Z)-nitrone 42 in xylene at reflux gave a 7:3 mixture

of exo,exo and exo,endo cycloadducts 43 and 44, respec-

tively (Scheme 8).

Since both 40 and 41 were unstable to chromatography

on silica and appreciable loss of material occurred when

their purification was attempted by this means, the

mixture of cycloadducts was subjected to immediate

reductive cleavage of the N-O bond. The resulting

piperidine 45 was stable and could be purified without

(30) Grieco, P. A.; Nunes, J. J.; Gaul, M. D. J. Am. Chem. Soc. 1990,

112, 4595.

(31) Hoffman, R. W.; Endesfelder, A. Justus Liebig Ann. Chem. 1986,

1823.

1968 J. Org. Chem., Vol. 70, No. 6, 2005

Total Synthesis of (-)-7-Epicylindrospermopsin

SCHEME 9

SCHEME 10

imidazoyl carbamate was readily cleaved from 47 with

methanolic potassium carbonate to give alcohol 48, and

oxidation of this secondary alcohol now proceeded un-

eventfully. As expected, reduction of ketone 49 with

L-Selectride³²yielded a single alcohol 50 which was

clearly epimeric with 46 at C-12.

relationship between hydrogens at C-7 and C-8 is gauche,

consistsent with the coupling constant measured in the

¹H NMR spectrum of 51, and the equatorial hydroxy-

methyl substituent at C-14 is positioned relative to the

urea carbonyl in a manner which suggests that ring

Acquisition of 50 gave us a structure with all six of

the stereocenters of 1 now set correctly, and our next task

was elaboration of the guanidine unit embedded in the

cylindrospermopsin core. This required replacement of

the p-bromobenzyl ether with a primary amine which, it

was hoped, could be induced to undergo intramolecular

condensation with the urea carbonyl. Cleavage of ether

50 by hydrogenolysis over Pearlman’s catalyst furnished

diol 51 which proved to be a crystalline substance

(Scheme 10). X-ray crystallographic analysis of 51 (Figure

1) fully confirmed its structure and demonstrated that

rings A and B occupied chair conformations with the C-12

hydroxyl group in an axial orientation. The spatial

FIGURE 1. ORTEP representation of the X-ray crystal

structure of 51. Ellipsoids are at the 20% probability level.

(32) Brown, H. C.; Krishnamurthy, S. J. Am. Chem. Soc. 1972, 94,

7159.

J. Org. Chem, Vol. 70, No. 6, 2005 1969

White and Hansen

SCHEME 11

closure to a cyclic imidate should be facile. Indeed, when

51 was treated with triphosgene a substance presumed

to be 52 was detected by mass spectrometry, and we

conjecture that this imidate is formed by intramolecular

displacement of chloride from 53. In any event, cyclic

imidate 52 was reopened by sodium azide in hot dim-

ethylformamide to give azide 54 which was reduced to

primary amine 55 under an atmosphere of hydrogen.

Unfortunately, many attempts to force cyclization of

amino urea 55 all met with failure, neither thermal

conditions (1,2-dichlorobenzene at 170 °C) nor Lewis

acids such as diethylaluminum chloride affording any

evidence for the formation of a guanidine moiety. Most

of these reactions returned the urea unchanged, indicat-

ing that modification of this functional group would be

required in order to assemble a guanidine in the manner

we envisioned.

Our inability to close amino urea 55 forced us to return

to azide 54 so that a suitable modification protocol could

be applied to the urea without interference from the

amino group. First, however, it was necessary to protect

the hydroxyl function of 54, and reaction of this alcohol

with triethylsilyl triflate gave ether 56 in excellent yield.

Interestingly, no silylation of the urea moiety took place

with this reagent, although when trimethylsilyl triflate

was used for protection of 54, some O-silylation at the

urea did occur. With the urea of 56 open for derivatiza-

tion, its exposure to potassium hexamethyldisilazide and

then to trimethyloxonium fluoroborate furnished the

unstable methylisourea 57 along with two minor products

which arose from monomethylation at each of the pyri-

midine nitrogens. Since chromatographic purification of

57 resulted in substantial loss of material, the crude

azide was hydrogenated over palladium-on-carbon with

the result that the transient primary amine 58 cyclized

spontaneously to yield 59. Although the facility with

which 58 underwent ring closure to guanidine 59 seemed

surprising, a similar observation was reported by Wein-

reb in his synthesis of 7-epicylindrospermopsin when an

azido urea similar to 56 was reacted with methyl triflate

and then was hydrogenated.

Conversion of 56 to 1 proved to be straightforward

since we could now take advantage of known chemistry¹³

for the global deprotection of this substance. Thus,

exposure of 59 to hot concentrated hydrochloric acid

accomplished hydrolysis of the dimethoxypyrimine to the

uracil residue as well as removal of the two silyl groups

to yield hydrochloride 60 after purification by reversed-

1

phase HPLC. The H and ¹³C NMR spectra of diol 60

matched those of the racemic hydrochloride synthesized

by Weinreb,¹⁴confirming that our synthesis had led to

7-epicylindrospermopsin rather than cylindrospermopsin

itself. Finally, treatment of diol 60 with sulfur trioxide-

pyridine complex as described by Snider¹³gave 7-epicy-

lindrospermopsin (1) accompanied by bissulfate 61 in the

approximate ratio 2.5:1. These two substances were

readily separated by reversed-phase HPLC and pure 1

was shown to have spectral properties which agreed with

those reported for natural 7-epicylindrospermospin.⁷The

optical rotation of our synthetic compound ([R]_D²⁵-17.2)

closely matched the value reported for natural material

([R]_D-20.5)⁷and established that the absolute config-

25

uration of natural 7-epicylindrospermopsin is correctly

represented by 1, i.e., 7S,8R,10S,12S,13R,14S.³³Since

7-epicylindrospermopsin has been correlated chemically

with cylindrospermopsin by Weinreb¹⁴and the two

compounds are known to differ in configuration only at

(33) A recent asymmetric synthesis of 7-epicylindrospermopsin by

Williams confirms this assignment (Looper, R. E.; Williams, R. M.

Angew. Chem., Int. Ed. 2004, 43, 2930).

1970 J. Org. Chem., Vol. 70, No. 6, 2005

Total Synthesis of (-)-7-Epicylindrospermopsin

2H), 3.70 (d, J ) 13.7 Hz, 2H), 3.78 (t, J ) 9.9 Hz,1H), 3.95

(d, J ) 13.7 Hz, 2H), 4.10-4.01 (q, J ) 8.8 Hz, 1H), 4.35-

4.28 (m, 1H), 7.49-7.26 (m, 10H); ¹³C NMR (75 MHz, CDCl₃)

δ 24.5, 54.5, 57.6, 65.1, 127.1, 128.2, 128.4, 138.7, 175.8; MS-

(EI) m/z 282 (M + 1)⁺, 254, 196, 190, 181, 146, 91; HRMS (El)

m/z 282.1494 (calcd for C₁₈H₁₉NO₂281.1416).

C-7, cylindrospermopsin is represented in absolute con-

figuration by 2, i.e., 7R,8R,10S,12S,13R,14S.

The asymmetric synthesis of (-)-7-epicylindrosper-

mopsin described above corroborates the structural cor-

rection made to this natural product by Weinreb in the

course of his synthesis of the racemic substance. The

stereochemical error in Moore’s assignment of relative

configuration to cylindrospermopsin,⁴and which passed

unnoticed in Snider’s synthesis of this compound,¹³must

now be attributed to an incorrect assumption regarding

the conformation of the molecule about the C-7-C-8 bond.

This led to a misinterpretation of the 4.0 Hz coupling be-

tween H-7 and H-8 as that of a syn-vicinal relationship. An

analogous conformation for 7-epicylindrospermopsin would

place H-7 and H-8 in an antiperiplanar orientation, a

stereochemical relationship difficult to reconcile with the

6.8 Hz coupling between these two protons. Although the

solution conformations of 1 and 2 are presently unknown,

the structural revisions made to cylindrospermopsin and

its epimer remove the awkward postulate of a previously

unknown tautomer of uracil as the agency which, through

hydrogen bonding, was thought to stabilize the confor-

mations previously proposed for these metabolites.

(2R,3R)- and (2S,3R)-3-Dibenzylamino-2-hydroxy(2,6-

dimethoxy-4-pyrimidinyl)tetrahydrofuran (11). A flask

containing anhydrous CeCl₃(dried at 120 °C and 0.5 Torr for

2 h, then cooled to room temperature) was filled with argon,

and 15 mL of dry THF was added. The suspension was stirred

for 5 min before addition of 10 (423 mg, 1.50 mmol), and the

mixture was stirred for 30 min at room temperature and then

cooled to -78 °C. In a separate flask, a solution of 9 (657 mg,

3.0 mmol) in THF-Et₂O (1:1, 40 mL) was stirred at -78 °C

for 5 min, and ice-cold n-butyllithium (1.6 M, 1.82 mL, 2.92

mmol) was added. The resulting solution was immediately

transferred via cannula to the solution containing 10, the

mixture was stirred at -78 °C for 45 min and then was allowed

to warm to room temperature during 5 h. Aqueous saturated

NH₄Cl solution (6 mL) was added, followed by saturated

aqueous NaHCO₃(20 mL) and water (5 mL), and the aqueous

layer was separated and extracted with Et₂O (30 mL). The

extract was dried over MgSO₄, filtered, and concentrated under

reduced pressure. Flash chromatography (petroleum ether-

EtOAc, 9:1 to 8:2 + 5% MeOH) of the residue yielded 617 mg

(97%) of 11 as a pale yellow oil: [R] - 43.3 (c 0.28, CHCl₃); IR

(neat) 3600-3100, 2955, 1720, 1567, 1448, 1352 cm^-1; ¹H NMR

(300 MHz, CDCl₃) δ 2.38-2.15 (m, 2H), 3.51 (d, J ) 14.0 Hz,

2H), 3.69 (t, J ) 8.1 Hz, 1H), 4.01 (d, J ) 22.8 Hz, 2H), 4.08-

4.00 (m, 1H), 4.20 (ddd, J ) 8.8, 8.8, 4.4 Hz, 1H), 6.36 (s, 1H),

7.31-6.93 (m, 10H); ¹³C NMR (75 MHz, CDCl₃) δ 24.7, 53.9,

54.7, 55.5, 66.9, 97.9, 102.0, 126.9, 128.0, 128.1, 128.4, 128.5,

128.7, 139.1, 164.5, 170.9, 172.4; MS (CI) m/z 422 (M + 1),

340, 323, 279, 209, 185, 91; HRMS (CI) m/z 422.2078 (calcd

for C₂₄H₂₇N₃O₄421.2002).

Experimental Section

2,4,6-Tribromopyrimidine. To a flame-dried, argon-filled

flask were added phosphorus(V) oxytribromide (17.88 g, 62.4

mmol), dry toluene (30 mL), barbituric acid (8, 2.00 g, 15.6

mmol), and N,N-dimethylaniline (3.60 mL, 28.8 mol). The

mixture turned red and was heated to reflux for 3 h and then

cooled to room temperature, and ice (ca 50 g) was added. The

aqueous layer was separated and extracted with benzene (10

mL), and the extract was dried over MgSO₄, filtered, and

diluted with MeOH (60 mL). The solution was concentrated

under reduced pressure to give 4.89 g (99%) of 2,4,6-tribro-

mopyrimidine as a yellow solid: ¹H NMR (300 MHz, CDCl₃) δ

7.71 (s, 1H); ¹³C NMR (75 MHz, CDCl₃) δ 128.1, 150.7, 153.0.

This material was used without purification for the next

reaction.

4-Bromo-2,6-dimethoxypyrimidine (9). To a solution of

2,4,6-tribromopyrimidine (4.89 g, 15.4 mmol) in anhydrous

MeOH (40 mL) and dry Et₂O (20 mL) was added dropwise

sodium methoxide (0.50 M, 61.7 mL, 30.9 mmol) during 1 h,

and the mixture was stirred for 20 h. After concentration of

the mixture under reduced pressure, the residue was diluted

with water (10 mL), and the mixture was extracted with Et₂O

(100 mL). The extract was concentrated under reduced pres-

sure to give 2.83 g (84%) of pure 9 as a colorless solid: IR (neat)

1588, 1568, 1466, 1360, 1344 cm^-1; ¹H NMR (300 MHz, CDCl₃)

δ 3.93 (s, 3H), 3.88 (s, 3H), 6.51 (s, 1H); ¹³C NMR (75 MHz,

CDCl₃) δ 54.2, 55.2, 104.8, 151.8, 164.3, 171.5; MS (FAB) m/z

219 (M + 1) 149, 136; HRMS (FAB) m/z 218.9767 (calcd for

C₆H₈N₂O₂⁷⁹Br 218.9769).

(R)-2-Dibenzylamino-4-hydroxybutyric Acid Lactone

(10). (R)-Methionine (5.00 g, 34.0 mmol) was added to a

solution of K₂CO₃(13.9 g, 101 mmol) in water (100 mL), and

the mixture was heated to reflux. Benzyl bromide (24.0 mL,

0.20 mol) was added dropwise, and the mixture was heated

at reflux for 3 h and then cooled to room temperature. The

aqueous phase was separated and extracted with Et₂O (150

mL), and the extract was washed with brine and dried over

MgSO₄. The solvent was removed under reduced pressure, and

the excess benzyl bromide and benzyl alcohol were distilled

off (1 Torr, 60 °C). The residual oil was purified by chroma-

tography (hexanes-ethyl acetate, 7:3) to yield a solid which

was crystallized from hexanes-Et₂O to yield 5.91 g (62%) of

10 as colorless prisms: [R] +28.3 (c 1.0, CHCl₃); IR (neat) 1757,

1643, 1602 cm^-1; ¹H NMR (300 MHz, CDCl₃) δ 2.33-2.24 (m,

(R)-2,6-Dimethoxy-4-(2′-dibenzylamino-4′-trityloxybu-

tanoyl)pyrimidine (12). A solution of 11 (617 mg, 1.46

mmol), Et₃N (0.365 mL, 2.62 mmol), triphenylmethyl chloride

(569 mg, 2.00 mmol), and 4-(dimethylamino)pyridine (7.2 mg,

0.06 mmol) in 10 mL of CH₂Cl₂was stirred at room temper-

ature for 48 h and then at reflux for 5 h. The solution was

cooled to room temperature and diluted with saturated aque-

ous NH₄Cl (2 mL) and saturated aqueous NaHCO₃(5 mL).

The aqueous layer was separated and extracted with Et₂O (30

mL), and the extract was dried over MgSO₄, filtered, and

concentrated under reduced pressure. Flash chromatography

(petroleum ether-Et₂O, 9:1, to petroleum ether-EtOAc, 8:2)

afforded 900 mg (93%) of 12 as a pale yellow oil: [R] +36.5 (c

1

1.0, CHCl₃); IR (neat) 1711, 1595, 1568, 1359 cm^-1; H NMR

(300 MHz, CDCl₃) δ 2.11-2.02 (m, 1H), 2.28-2.19 (m, 1H),

3.18 (q, J ) 5.5 Hz, 2H), 3.46 (s, 3H), 3.73 (ABq, J ) 14.3 Hz,

4H), 4.03 (s, 3H), 5.19 (t, J ) 7.5 Hz, 1H), 6.85 (s, 1H), 7.36-

7.13 (m, 25H, m); ¹³C NMR (75 MHz, CDCl₃) δ 27.0, 54.2, 54.3,

54.8, 57.2, 60.6, 86.6, 99.9, 126.8, 127.2, 127.6, 127.8, 128.1,

128.6, 139.5, 144.0 162.4, 165.2, 172.8, 200.8; MS (FAB) m/z

664 (M + 1), 496, 391, 243, 136, 91; HRMS (FAB) m/z 664.3099

(calcd for C₄₃H₄₂N₃O₄664.3097).

(1′S,2′R)-2,6-Dimethoxy-4-(2′-dibenzylamino-1′-hydroxy-

4′-trityloxybutyl)pyrimidine (13). To a solution of 12 (291

mg, 0.438 mmol) in THF (4.5 mL) at -78 °C was added

dropwise L-Selectride (1.0 M, 0.526 mL, 0.526 mmol). After

1.5 h, the mixture was warmed to 25 °C, and EtOH (2.5 mL)

and water (1.5 mL) were added. The mixture was stirred for

30 min, NaOH (1 M, 2.5 mL) and H₂O₂(30%, 2.0 mL) were

added, and the mixture was stirred for a further 45 min and

was extracted with CH₂Cl₂(3 × 5 mL). The extract was dried

over MgSO₄and concentrated, and the residual oil was purified

by flash chromatography (9:1-8:2, hexanes-EtOAc) to give

275 mg (84%) of 13: [R] -25.0 (c 1.0, CHCl₃); IR (neat) 3600-

1

3150, 1590, 1580, 1470, 1346; H NMR (300 MHz, CDCl₃) δ

2.24-1.95 (m, 2H), 3.08-2.98 (m, 3H), 3.43 (d, J ) 13.1 Hz,

J. Org. Chem, Vol. 70, No. 6, 2005 1971

White and Hansen

2H), 3.88 (m, 1H), 3.93 (s, 3H), 3.96 (s, 3H), 4.35 (d, J ) 6.6

Hz, 1H), 4.72, (br, 1H), 6.01 (s, 1H), 7.44-7.09 (m, 25H); ¹³C

NMR (75 MHz, CDCl₃) δ 14.4, 20.9, 25.8, 53.6, 54.4, 54.5, 54.6,

59.6, 60.3, 61.8, 73.5, 86.6, 98.6, 126.9, 127.0, 127.7, 128.2,

128.5, 129.0, 139.0, 143.9, 164.3, 171.7, 172.2; MS (CI) m/z 666

(M + 1), 496, 243, 267, 91; HRMS (CI) m/z 666.3351 (calcd for

C₄₃H₄₄N₃O₄666.3332).

(1′S,2′R)-2,6-Dimethoxy-4-[2′(tert-butoxycarbonyl)ami-

no-1′-(tert-butyldimethylsilanyloxy)-4′-hydroxybutyl]py-

rimidine (19). To a solution of 17 (201 mg, 0.372 mmol) in

EtOH (5 mL) was added Pd(OH)₂/C (20%, 100 mg), and the

mixture was stirred vigorously under an atmosphere of H₂for

7 h. The suspension was filtered through Celite, and the

filtrate was concentrated under reduced pressure to give 108

mg (81%) of an amino alcohol: IR (neat) 3600-3400, 2926,

(1′,2′R)-2,6-Dimethoxy-4-(2′dibenzylamino-1′,4′-dihy-

droxybutyl)pyrimidine (14). To 13 (330 mg, 0.496 mmol)

were added diethyl ether (13 mL) and formic acid (20 mL).

The mixture was stirred at room temperature for 12 h and

then was diluted with Et₂O (30 mL) and washed with brine (3

× 15 mL) and saturated aqueous NaHCO₃(3 × 20 mL). The

separated organic phase was dried over MgSO₄, filtered, and

concentrated. Purification of the residual oil by flash chroma-

tography (1:1, hexanes-EtOAc) gave 146 mg (70%) of 14: [R]

- 20.0 (c 1.0, CHCl₃); IR (neat) 3200-3650, 1594, 1569, 1450,

1355 cm^-1; ¹H NMR (CDCl₃) δ 1.70-1.85 (m, 1 H), 1.95-2.10

(m, 1 H), 2.25-2.50 (bs, 1H), 3.04 (q, J ) 13.2, 6.6, 6.6 Hz, 1

H), 3.40-3.52 (m, 2 H), 3.78 (dd, J ) 17.3, 13.2 Hz, 4 H), 3.91

(s, 3H), 3.93 (s, 3H), 4.49 (d, J ) 8.1 Hz, 1H), 4.47-4.62 (bs, 1

H), 6.18 (s, 1 H), 7.18-7.32 (m, 10 H); ¹³NMR (CDCl₃) δ 29.0,

53.9, 54.5, 54.7, 59.8, 60.7, 74.2, 98.8, 127.2, 128.4, 129.2, 139.0,

164.6, 172.0, 172.7; MS (FAB) m/z 424.2 (M + 1)⁺, 285.2, 154.1,

91.1; HRMS (FAB) m/z 424.2149 (calcd for C₂₄H₃₀N₃O₄

424.2158).

(1′S,2′R)-2,6-Dimethoxy-4-[2′dibenzylamino-1′-(tert-bu-

tyldimethylsilanyloxy)-4′-trityloxybutyl]pyrimidine (16).

To a solution of 13 (195 mg, 0.294 mmol) in CH₂Cl₂(0.7 mL)

were added 2,6-lutidine (0.066 mL, 0.572 mmol) and tert-

butyldimethylsilyl triflate (0.429 mmol, 0.098 mL), and the

mixture was stirred for 1.25 h. Aqueous NH₄Cl (1 mL) was

added, and the mixture was diluted with Et₂O. The aqueous

phase was separated and extracted with Et₂O (3 × 5 mL), and

the combined extract was dried over MgSO₄, filtered, and

concentrated. The residual oil was purified by flash chroma-

tography (9:1 to 8:2 petroleum ether-Et₂O) to give 199 mg

(87%) of 16: [R] -43.3 (c 0.28, CHCl₃); IR (neat) 2947, 2900,

1602, 1570, 1355 cm^-1; ¹H NMR (300 MHz, CDCl₃) δ -0.20 (s,

3H), -0.18 (s, 3H), 0.88 (s, 9H), 1.90-2.08 (m, 1H), 2.2-2.38

(m, 1H), 3.20-3.32 (m, 1H), 3.30-3.40 (m, 1H), 3.36 (s, 4H),

3.42 (s, 1H), 3.46-3.55 (m, 1H), 4.05 (s, 4H), 3.95-4.09 (bs,

1H), 4.55 (d, J ) 2.7 Hz, 1H), 6.72 (s, 1H), 7.05-7.57 (m, 25

H); ¹³C NMR (75 MHz, CDCl₃) δ -5.2, -4.7, 17.9, 25.0, 25.8,

53.8, 54.3, 55.4, 57.6, 60.9, 77.4, 86.5, 99.6, 126.3, 126.7, 127.6,

127.8, 128.5, 128.7, 140.5, 144.2, 164.4, 171.6, 174.8; MS (FAB)

m/z 780 (M + 1), 652, 496, 368, 243, 91; HRMS (FAB) m/z

780.4186 (calcd for C₄₉H₅₈N₃O₄Si 780.4197).

1

1573, 1360 cm^-1; H NMR (300 MHz, CDCl₃) δ 0.06 (s, 3H),

0.22 (s, 3H), 0.93 (s, 9H), 1.23 (m, 1H), 1.46 (t, J ) 7 Hz, 1H),

1.56-1.87 (m, 2H), 3.66-3.88 (m, 2H), 3.98 (s, 6H), 4.94 (d, J

) 3 Hz, 1H), 6.51 (s, 1H); ¹³C NMR (75 MHz, CDCl₃) δ -5.0,

-4.7, 14.1, 17.9, 25.8, 29.7, 54.0, 55.1, 55.5, 59.4, 72.1, 99.3,

164.5, 170.5, 172.4. This material was used immediately in

the next reaction.

To a solution of the amino alcohol obtained above (125 mg,

0.350 mmol) in CH₂Cl₂(6 mL) were added Boc₂O (84.4 mg,

0.386 mmol) and Et₃N (35.3 mg, 0.350 mmol), and the resulting

solution was stirred for 24 h at room temperature. The solvent

was removed, and the residual oil was purified by chromatog-

raphy on silica gel, with hexane/EtOAc (7:3, then 1:1) as

eluent, to give 131 mg (68%) of 19: [R] -9.4 (c 0.5, CHCl₃); IR

(neat) 2953, 2926, 1689, 1595, 1570, 1370 cm ^-1;¹H NMR (300

MHz, CDCl₃) δ 0.00 (s, 3H), 0.11 (s, 3H), 0.93 (s, 9H), 1.37 (s,

9H), 1.97-1.79 (m, 1H), 3.72-3.41 (m, 3H), 3.94 (s, 3H), 3.96

(s, 3H), 4.63 (d, J ) 2.2 Hz, 1H), 5.58 (d, J ) 9.5 Hz, 1H), 6.49

(s, 1H);¹³C NMR (75 MHz, CDCl₃), δ -5.0, -4.9, 14.1, 18.1,

20.9, 25.8, 28.2, 34.9, 51.8, 53.8, 54.7, 58.7, 75.5, 79.6, 98.8,

156.9, 164.7, 172.1, 172.3; MS (FAB) m/z 446 (M + 1), 358,

285, 185, 91; HRMS (FAB) m/z 446.2528 (calcd for C₂₀H₄₀N₃O₆-

Si 446.2530).

(3R,4S)-3-(tert-Butoxycarbonyl)amino-4-(tert-butyldim-

ethylsilanyloxy)-4-(2,6-dimethoxy-4-pyrimidinyl)bu-

tanol (20). To a solution of 19 (111 mg, 0.240 mmol) in CH₂Cl₂

(2.5 mL) under argon were added powdered molecular sieves

(4 Å, 50 mg), N-methylmorpholine N-oxide (117 mg, 0.360

mmol), and tetra-n-propylammonium perruthenate (8.6 mg,

0.02 mmol), and the resulting green mixture was stirred at

room temperature for 45 min. The mixture was filtered

through a pad of silica which was washed with EtOAc. The

filtrate was concentrated under reduced pressure to leave 102

mg (91%) of 20 as a colorless oil: [R] -34.1 (c 0.66, CHCl₃); IR

(neat) 3482-3205, 2959, 2928, 2858, 1720, 1598, 1568, 1477,

1358, 1096, 841, 777 cm ^-1; ¹H NMR (300 MHz, CDCl₃) δ 0.01

(s, 3H), 0.13 (s, 3H), 0.95 (s, 9H), 1.37 (s, 9H), 2.34 (ddd, J )

2.6, 8.2, 15.0 Hz, 1H), 2.53 (ddd, J ) 2.6, 5.1, 15.0 Hz, 1H),

3.96 (s, 3H), 3.99 (s, 3H), 4.43-4.55 (m, 1H), 4.69 (d, J ) 3

Hz, 1H), 5.66 (d, J ) 9.5 Hz, 1H), 6.49 (s, 1H), 9.73 (s, 1H);

¹³C NMR (75 MHz, CDCl₃), δ -4.8, -4.6, 18.4, 26.1, 28.5, 46.1,

51.2, 54.1, 55.0, 75.0, 79.8, 99.5, 155.4, 165.1, 171.2, 172.6,

200.5; MS (FAB) m/z 456 (M + 1), 312, 285, 227, 154, 91;

HRMS (FAB) m/z 456.2523 (calcd for C₂₁H₃₈N₃O₆Si 456.2530).

(1′S,2′R)-2,6-Dimethoxy-4-[2′dibenzylamino-1′-(tert-bu-

tyldimethylsilanyloxy)-4′-hydroxybutyl]pyrimidine (17).

A solution of 16 (235 mg, 0.301 mmol) in Et₂O (10 mL) and

formic acid (15 mL) was stirred for 12 h. The mixture was

diluted with Et₂O (50 mL), and the solution was washed with

brine (3 × 50 mL) and with saturated aqueous NaHCO₃(3 ×

40 mL). The organic phase, which contained a mixture of 17

and its formate ester, was concentrated to 20 mL, and MeOH

(30 mL) and K₂CO₃(663 mg, 4.80 mmol) were added. The

mixture was stirred for 1 h and extracted with Et₂O (3 × 15

mL), and the combined extract was dried over MgSO₄, filtered,

and concentrated. The residual oil was purified by flash

chromatography (9:1 to 8:2 hexane/EtOAc) to give 156 mg

(100%) of 17: [R] -43.3 (c 0.28, CHCl₃); IR (neat) 2940, 1595,

2-(tert-Butoxycarbonyl)aminoacetaldehyde (21). To

2-(tert-butoxycarbonyl)aminoethanol (40.0 mg, 0.248 mmol) in

CH₂Cl₂(2 mL) was added Dess-Martin periodinane (130 mg,

0.306 mmol) in one portion. A milky suspension formed, and

the mixture was stirred for 2 h. To the mixture was added 2

mL of saturated aqueous NaHCO₃-Na₂S₂O_4.Within 20 min

the solution became clear and biphasic. The mixture was

extracted with CH₂Cl₂, and the extract was washed with brine

and dried over MgSO₄. Filtration through a plug of Celite and

concentration gave 38.6 mg (98%) of 21 which was used

without further purification: IR (neat) 3400, 2990, 1702 cm^-1

;

1

1560, 1365 cm^-1; H NMR (300 MHz, CDCl₃) δ -0.07 (s, 3 H),

¹H NMR (300 MHz, CDCl₃) δ 1.45 (s, 9 H), 4.07 (d, J ) 4.6

Hz, 2 H), 5.18 (bs, 1 H), 9.65 (s, 1 H); ¹³C NMR (300 MHz,

CDCl₃), δ 27.8, 51.24, 80.0, 155.7, 197.4; HRMS (FAB) m/z

160.088 (calcd for C₇H₁₄NO₃160.0895).

1.00 (s, 9 H), 0.17 (s, 3 H), 1.42-1.58 (m, 1H), 1.98-2.09 (m,

1H), 2.60-2.90 (bs, 1H), 3.28 (q, J ) 6.6, 11.0 Hz, 1H), 3.48-

3.60 (m, 1H), 3.63 (d, J ) 13.7 Hz, 2H), 3.82 (s, 3H), 3.91 (d,

J ) 13.7 Hz, 2H), 4.06 (s, 3H), 4.71 (d, J ) 4.4 Hz, 1H), 6.73

(s, 1H), 7.23-7.33 (m, 10H); ¹³C NMR (75 MHz, CDCl₃) δ -4.9,

-4.4, 18.0, 25.9, 29.4, 29.6, 53.9, 54.5, 54.8, 54.9, 58.8, 58.9,

60.9, 61.2, 78.6, 78.8, 99.4, 126.8, 128.1, 129.1, 140.0, 164.5,

171.8, 174.5; MS (CI) m/z 538 (M + 1), 522, 492, 480; HRMS

(CI) m/z 538.3101 (calcd for C₃₀H₄₄N₃O₄Si 538.3101).

(2R,3S)-1-(tert-Butyoxycarbonyl)amino-2-hydroxy-3-

methyl-4-pentene (22). To a stirred mixture of potassium

tert-butoxide (113 mg, 1.0 mmol) in THF (1.33 mL) and cis-

2-butene (2 mL, 22 mmol) was added n-butyllithium (1.0 mmol,

0.667 mL, 1.6 M in hexane) at -78 °C. The mixture was stirred

1972 J. Org. Chem., Vol. 70, No. 6, 2005

Total Synthesis of (-)-7-Epicylindrospermopsin

at -45 °C for 10 min, and the resulting orange solution was

recooled to -78 °C. To this solution was added dropwise a

solution of (+)-methoxydiisopinocampheylborane (387 mg, 1.23

mmol) in Et₂O (2.2 mL). The mixture was stirred at -78 °C

for 30 min, BF₃‚Et₂O (0.137 mmol, 0.14 mL) was added

dropwise, and the mixture was stirred for 10 min. A solution

of 21 (159 mg, 1 mmol) in THF (1.2 mL) was added dropwise,

and the mixture was stirred at -78 °C for 5 h. The mixture

was allowed to slowly warm to room temperature and then

treated with NaOH (3 M, 3 mL) and H₂O₂(30%, 1 mL), and

the resulting mixture was refluxed for 1 h. The organic phase

was separated, washed with water and brine, and dried

(MgSO₄). After removal of solvent under reduced pressure, the

residual oil was chromatographed on silica gel (8:2, hexanes-

EtOAc, + 5% CH₂Cl₂) to give 85 mg (40%) of 22: [R] -28.4 (c

(300 MHz, CDCl₃) δ 57.1, 59.5, 63.5, 122.2, 126.9, 127.5, 128.2,

128.8, 129.1, 138.3, 138.8, 202.1.

(2R,3S)-1-Dibenzylamino-2-hydroxy-3-methyl-4-pen-

tene (26). To a stirred mixture of potassium tert-butoxide (339

mg, 3.03 mmol) in THF (4 mL) and cis-2-butene (2.0 mL, 22

mmol) at -78 °C was added n-butyllithium (3.2 mmol, 2.0 mL,

1.6 M in hexane), and the mixture was stirred at -45 °C for

10 min. The resulting orange solution was recooled to -78 °C,

and a solution of (+)-methoxydiisopinocampheylborane (1.16

g, 3.69 mmol) in Et₂O (3.7 mL) was added dropwise. The

mixture was stirred at -78 °C for 30 min, BF₃‚Et₂O (0.41

mmol, 0.45 mL) was added dropwise, and the mixture was

stirred for 10 min. A solution of 25 (980 mg, 4.10 mmol) in

THF (4 mL) was added dropwise, and the mixture was stirred

at -78 °C for 5 h and was then allowed to warm slowly to

room temperature. The mixture was treated with NaOH (3

M, 9.0 mL) and H₂O₂(30%, 3.0 mL) and was refluxed for 1 h.

The organic phase was separated, washed with water and

brine, and dried (MgSO₄). Isopinocampheol was removed by

distillation (75 °C, 1 mm), and the residual oil was chromato-

graphed on silica gel (9:1, then 8:2, hexanes-EtOAc) to give

590 mg (48%) of 26: [R] - 83.6 (c 1.0, CHCl₃); IR (neat) 3423,

1.0, CHCl₃); IR (neat) 3550-3300, 2973, 1690, 1516, 1171 cm^-1

;

¹H NMR (300 MHz, CDCl₃) δ 1.03 (d, J ) 6.6 Hz, 3H), 1.40 (s,

9H), 2.29-2.14 (m, 1H), 3.00-2.73 (m, 1H), 3.10-3.00 (bs, 1H),

3.38-3.25 (m, 1H), 3.52-3.40 (m, 1H), 4.68-4.58 (bs, 1H),

5.15-4.95 (m, 2H), 5.77-5.65 (ddd, J ) 18.4, 10.3, 8.1 Hz, 1H);

¹³C NMR (300 MHz, CDCl₃) δ 15.4, 28.1, 28.3, 42.3, 44.6, 74.6,

79.5, 115.5, 140.1, 156.8; MS (FAB) m/z 216.1(M + 1)⁺,160.0,

142.1, 104, 75; HRMS (FAB) m/z 216.1596 (calcd for C₁₁H₂₂O₃N

216.1599).

1

cm^-1; H NMR (300 MHz, CDCl₃) δ 1.00 (d, J ) 6.6 Hz, 3H),

2.10 (dd, J ) 12.1, 6.6 Hz, 1H), 3.39 (d, J ) 14.0 Hz, 2H), 3.41-

3.58 (m, 2H), 3.85 (d, J ) 14.0 Hz, 2H), 4.95-5.29 (m, 2H),

5.61-5.73 (ddd, J ) 17.0, 10.3, 8.1 Hz, 1H), 7.35-7.22 (m,

10H); ¹³C NMR (300 MHz, CDCl₃) δ 15.9, 42.5, 57.5, 58.4, 70.0,

114.8, 127.3, 128.5, 129.1, 138.5, 140.5; MS (FAB) m/z 296.1

(M + 1)⁺, 240.1, 210.1, 91.1; HRMS (FAB) m/z 296.20113 (calcd

for C₂₀H₂₆NO 296.20144).

(2R,3S)-1-(tert-Butoxycarbonyl)amino-2-methanesulfo-

nyloxy-3-methyl-4-pentene (23). To a solution of 22 (59 mg,

0.274 mmol) in THF (1 mL) were added pyridine (252 mg, 3.2

mmol), DMAP (cat), and methanesulfonyl choride (37.5 mg,

0.329 mmol). The mixture was stirred at room temperature

for 0.5 h and then diluted with Et₂O (6 mL) and aqueous NH₄-

Cl. The aqueous layer was extracted with Et₂O (2 × 5 mL),

and the combined Et₂O extract was washed with brine (5 mL)

and dried (MgSO4). Evaporation of the solvent under reduced

pressure yielded 67 mg (84%) of 23 which was used without

(2S,3R,4R)-4-Dibenzylamino-1-hydroxy-2,3-dimeth-

ylpyrrolidine (27). To a solution of 28 (5.81 mg, 0.014 mmol)

in anhydrous THF (1 mL) was added cesium fluoride (3.10 mg,

0.021 mmol), and the mixture was stirred at reflux for 9 h.

Anhydrous DMF (2 mL) was added to the solution, which was

stirred at 65 °C for 6 h. After the solvent was removed under

reduced pressure, the residual material was chromatographed

on silica gel (7:3, hexanes-EtOAc) to give 2.31 mg (55%) of

27: IR (neat) 2921, 1667, 1579, 1452 cm^-1; ¹H NMR (300 MHz,

CDCl₃) δ 1.25-1.20 (m, 6 H), 1.92-1.86 (m, 1H), 2.72-2.66

(m, 1H), 3.03-2.95 (m, 1H), 3.46-3.36 (m, 2H), 3.58 (d, 14.7

Hz, 2H), 3.68 (d, 14.7 Hz, 2H), 7.41-7.18 (m, 10H); ¹³C NMR

(300 MHz, CDCl₃) δ 13.1, 16.7, 56.8, 57.1, 57.8, 71.2, 126.9,

127.3, 128.3, 128.5, 144.0; MS (FAB) m/z 311 (M + 1)⁺, 309,

293, 91; HRMS (FAB) m/z 311.2127 (calcd for C₂₀H₂₇N₂O

311.21234).

further purification: [R]²⁴- 30.9 (c 1.0, CHCl₃); IR (neat)

D

3450-3350, 1705, 1173 cm^-1

;

¹H NMR (300 MHz, CDCl₃) δ

1.13 (d, J ) 6.6 Hz, 3H)1.43 (s, 9H), 2.61-2.54 (q, J ) 6.6 Hz,

1H), 3.05 (s, 3H), 3.25-3.18 (m, 1H), 3.53-3.45 (m, 1H), 4.66-

4.60 (m, 1H), 5.00-4.85 (bs, 1H), 5.18-4.93 (m, 2H), 5.83-

5.72 (ddd, J ) 18.4, 10.3, 8.1 Hz, 1H); ¹³C NMR (300 MHz,

CDCl₃) δ 15.4, 27.9, 28.3, 29.2, 38.2, 40.2, 42.2, 79.7, 84.7,

116.9, 137.8, 155.8; MS (FAB) m/z 294.1 (M + 1)⁺, 238.1, 194.1,

142.1; HRMS (FAB) m/z 294.13771 (calcd for C₁₂H₂₄NO₅S

294.13752).

2-(Dibenzylamino)ethanol. To a solution of ethanolamine

(1.00 g, 16.3 mmol) in CH₂Cl₂-H₂O (33 mL-16 mL) were added

Na₂CO₃(6.90 g, 65.4 mmol) and benzyl bromide (5.60 g, 32.7

mmol), and the solution was stirred at reflux for 3 h. The

solution was diluted with H₂O, and the aqueous phase was

extracted with CH₂Cl₂(2 × 15 mL). The combined organic

extract was evaporated under reduced pressure, and the excess

benzyl bromide was removed by distillation (55 °C, 1 mm) to

afford 1.42 g (37%) of the title compound which was used

(2S,3S)-2-Dibenzylamino-1-(tert-butyldimethylsilany-

loxy)amino-3-methyl-4-pentene (28). To a stirred solution

of 26 (100 mg, 0.34 mmol) in CH₂Cl₂/hexane (1:1, 0.8 mL) at

-78 °C were added triflic anhydride (68.7 µL, 0.41 mmol) and

lutidine (51.2 µL, 0.44 mmol) during 15 min, followed by

dropwise addition of O-tert-butyldimethylsilanylhydroxylamine

(99.9 mg, 0.68 mmol) in CH₂Cl₂/hexane (1:1, 0.8 mL). The

mixture was stirred at -78 °C for 3.5 h and warmed to room

terperature during 3 h. After removal of the solvent under

reduced pressure, the residue was chromatographed on silica

gel (100% hexane, then 98:2, then 95:5, and then hexanes-

without further purification: IR (neat) 3650-3200, 1450 cm^-1

;

¹H NMR (300 MHz, CDCl₃) δ 2.55-2.45 (bs, 1H), 2.67 (t, J )

5.5 Hz, 2H), 3.59 (m, 2H), 3.63 (s, 4H), 7.34-7.26 (m, 10H);

¹³C NMR (300 MHz, CDCl₃) δ 54.8, 58.3, 58.7, 127.2, 128.5,

129.0, 138.8; MS (FAB) m/z 242.2 (M + 1)⁺, 210.1, 91.1; HRMS

(FAB) m/z 242.1547 (calcd for C₁₆H₂₀NO 242.1545).

Et₂O) to afford 81 mg (56%) of 28: [R]²⁴+1.8 (c 1.0, CHCl₃);

D

IR (neat) 2900, 1450, 1249, 836, 699 cm^-1; ¹H NMR (300 MHz,

CDCl₃) δ 0.12 (s, 3 H), 0.13 (s, 3 H), 0.94 (s, 9 H),1.02 (d, J )

6.6 Hz, 3H), 2.58 (dd, J ) 14.7, 7.4 Hz, 1H), 2.80 (m, 1H), 2.92

(dd, J ) 12.5, 4.4 Hz, 1H), 3.14 (m, 1 H), 3.72 (d, J ) 13.5 Hz,

2-(Dibenzylamino)acetaldehyde (25). To a solution of

crude 2-(dibenzylamino)ethanol (50.0 mg, 0.207 mmol) and

triethylamine (125 mg, 1.24 mmol) in DMSO (1 mL) was added

dropwise SO₃‚Py (154 mg, 0.622 mmol) in DMSO (1 mL). The

resulting solution was stirred for 1 h at room temperature and

diluted with Et₂O (5 mL) and aqueous NH₄Cl. The mixture

was further diluted with H₂O, and the aqueous phase was

extracted with Et₂O (2 × 5 mL). The combined Et₂O extract

was washed with brine (5 mL) and dried (MgSO₄), and the

solvent was removed under reduced pressure to give 40.1 mg

(81%) of 25 which was unstable and was used immediately

without purification: ¹H NMR (300 MHz, CDCl₃) δ 3.18 (s,

2H), 3.69 (s, 4H), 7.40-7.19 (m, 10H), 9.50 (s, 1H); ¹³C NMR

2H), 3.77 (d, J ) 13.5 Hz, 2H), 5.30 (m, 2 H), 5.4-5.2 (bs, 1H),

13

5.88 (ddd, J ) 16.9, 10.3, 8.1 Hz, 1H), 7.29 (m, 10 H);

C

NMR (300 MHz, CDCl₃) δ -5.3, 18.0, 26.4, 39.0, 52.7, 54.6,

59.0, 113.6, 126.8, 128.2, 129.0, 140.2, 143.3; MS (FAB) m/z

425 (M + 1), 369, 264, 210, 91; HRMS (FAB) m/z 425.2994

(calcd for C₂₆H₄₁N₂OSi 425.2988).

Nitrone 30. To a solution of 28 (124 mg, 0.29 mmol) in

anhydrous MeOH (3.0 mL) at room temperature was added

p-nitrobenzaldehyde (66.3 mg, 0.44 mmol), and the solution

was stirred at room temperature for 5 min. To the solution

J. Org. Chem, Vol. 70, No. 6, 2005 1973

White and Hansen

were added ammonium fluoride (26.1 mg, 0.58 mmol) and 3 Å

molecular sieves. The resulting mixture was warmed to 60 °C

and stirred for 2 h. After being cooled to room temperature,

the mixture was diluted with saturated aqueous NaHCO₃(0.5

mL) and extracted with CH₂Cl₂. The extract was washed with

saturated aqueous NaCl, dried over Na₂SO₄, and concentrated

under reduced pressure. The resultant orange solid was

crystallized from Et₂O-hexane to give 109 mg (85%) of 30:

(45%) of 32: [R] -11.8 (c 2.0, CHCl₃); IR (neat) 3621-3162,

2964, 2909, 2854, 1646, 1587, 1482, 1098, 1006, 913, 795 cm^-1

;

¹H NMR (300 MHz, CDCl₃) δ 1.08 (d, J ) 6.8 Hz, 3H), 2.26 (d,

J ) 3.8 Hz, 1H), 2.22-2.39 (m, 1H), 3.38 (dd, J ) 9.6, 7.9 Hz,

1H), 3.54 (dd, J ) 9.6, 3.0 Hz, 1H), 3.61-3.66 (m, 1H), 4.49 (s,

2H), 5.01-5.09 (m, 2H), 5.73 (ddd, J ) 17.5, 10.4, 7.9 Hz, 1H),

7.19 (d, J ) 8.2 Hz, 2H), 7.48, (d, J ) 8.2 Hz, 2H); ¹³C NMR

(75 MHz, CDCl₃) δ 16.1, 41.4, 72.9, 73.4, 73.8, 115.5, 122.0,

129.7, 131.9, 137.5, 140.8; MS (FAB) m/z 285 (M + 1), 282,

171, 154, 136, 107, 89; HRMS (FAB) m/z 284.0406 (calcd for

C₁₃H₁₇O₂⁷⁹Br 284.0412).

[R]²⁴-10.0 (c 0.64, CHCl₃); IR (neat) 1595, 1516, 1340, 746,

D

1

695 cm^-1; H NMR (300 MHz, CDCl₃) δ 1.05 (d, J ) 6.8 Hz,

3H), 2.62 (ddd, J ) 14.4, 14.5, 7.4 Hz, 1H), 3.52-3.59 (m, 1H),

3.78 (ABq, J ) 27.1, 13.7 Hz, 4H), 3.99 (dd, J ) 12.3, 5.2 Hz,

1H), 4.11-4.21 (m, 1H), 5.08, 5.14 (m, 2H), 5.9 (ddd, J ) 18.6,

10.4, 8.2 Hz, 1H), 7.22-7.31 (m, 10 H), 7.35 (s, 1H), 8.32 (ABq,

J ) 20.0, 9.0 Hz, 4H); ¹³C NMR (300 MHz, CDCl₃), δ 18.6,

39.7, 55.5, 61.2, 67.3, 115.4, 124.2, 127.6, 128.7, 129.0, 129.4,

133.7, 136.6, 139.9, 142.3, 148.1; MS (FAB) m/z 444.2 (M +

1)⁺, 388, 264, 154, 91; HRMS (FAB) m/z 444.2291 (calcd for

C₂₇H₃₀N₃O₃444.2287).

(3S,4R)-5-(4-Bromobenzyloxy)-4-methanesulfonyloxy-

3-methylpent-1-ene. To a solution of 32 (100 mg, 0.35 mmol)

in anhydrous CH₂Cl₂(1 mL) were added pyridine (0.28 mL,

3.5 mmol), a catalytic quantity of DMAP (0.05 mol %), and

methanesulfonic anhydride (79 mg, 0.45 mmol). After being

stirred at room temperature for 16 h, the mixture was

concentrated under reduced pressure, and the residue was

diluted with Et₂O (4 mL). The solution was washed sequen-

tially with saturated aqueous NH₄Cl, H₂O, and saturated

aqueous NaCl and dried over Na₂SO₄. Removal of the solvent

under reduced pressure gave 127 mg of the title compound

which was used without further purification: [R] -10.9 (c 4.0,

2-(4-Bromobenzyloxy)ethanol. To a slurry of NaH (1.18

g, 0.49 mol) in anhydrous THF (90 mL) at room temperature

was added ethylene glycol (8.0 mL, 0.14 mol) during 1 h. The

mixture was warmed to 50 °C, after which a solution of

p-bromobenzyl bromide (11.7 g, 0.047 mol) in THF (30 mL)

was added during 2 h. The mixture was stirred under reflux

for 12 h, cooled to room temperature, and diluted with Et₂O

(100 mL) and water (100 mL). The mixture was extracted with

Et₂O, and the ethereal extract was washed with saturated

aqueous NaCl, dried over Na₂SO₄, and concentrated under

reduced pressure. Chromatography of the residue on silica

(hexanes-EtOAc, 8:2) gave 7.30 g (65%) of the title com-

CHCl₃); IR (neat) 2968, 2934, 2871, 1482, 1343, 1170, 909 cm^-1

;

¹H NMR (300 MHz, CDCl₃) δ 1.12 (d, J ) 6.8 Hz, 3H), 2.55-

2.62, (m, 1H), 3.01 (s, 3H), 3.58-3.68 (m, 2H), 4.48 (q, J )

18.6, 11.8 Hz, 2H), 4.69 (ddd, J ) 6.8, 6.8, 3.5 Hz, 1H), 5.08-

5.16 (m 2H), 5.74 (ddd, J ) 17.0, 10.1, 7.6 Hz, 1H), 7.19 (d, J

) 8.5 Hz, 2H), 7.47, (d, J ) 8.5 Hz, 2H); ¹³C NMR (75 MHz,

CDCl₃) δ 16.1, 39.2, 40.4, 70.7, 73.0, 85.3, 117.1, 122.2, 129.8,

132.0, 136.9, 138.6; MS (FAB) m/z 364 (M + 1), 185, 169, 91;

HRMS (FAB) m/z 362.0190 (calcd for C₁₄H₁₉O₄⁷⁹BrS 362.0187).

pound: IR (neat) 3600-3111, 2922, 2863, 1482, 1069 cm^-1

;

¹H NMR (300 MHz, CDCl₃) δ 3.15-3.21 (bs, 1H), 3.45-3.50

(m, 2H), 3.61-3.68 (m, 2H), 4.41 (s, 2H), 7.13 (d, J ) 8.5 Hz,

2H), 7.39 (d, J ) 8.5 Hz, 2H); ¹³C NMR (75 MHz, CDCl₃) δ

62.0, 72.1, 72.8, 121.9, 129.8, 131.9, 137.5; MS (FAB) m/z 230

(M + 1), 185, 169, 151, 107, 88; HRMS (FAB) m/z 229.9941

(calcd for C₉H₁₁BrO₂229.9942).

(3S,4S)-4-Amino-5-(4-bromobenzyloxy)-3-methylpent-

1-ene (33). To a solution of (3S,4R)-5-(4-bromobenzyloxy)-4-

methanesulfonyloxy-3-methylpent-1-ene (12.1 mg, 0.04 mmol)

in anhydrous DMF (0.25 mL) was added sodium azide (9.5 mg,

0.14 mmol), and the mixture was stirred for 16 h in a

preheated oil bath at 85 °C. After being cooled to room

temperature, the mixture was diluted with Et₂O (2 mL) and

washed with H₂O (3 × 2 mL). The ethereal solution was

washed with saturated aqueous NaCl, dried over Na₂SO₄, and

concentrated under reduced pressure to give the crude azide.

This material was taken up into THF (0.5 mL), and triph-

enylphosphine (11.5 mg, 0.044 mmol) and water (0.05 mL)

were added. The mixture was stirred for 4 h at 55 °C, the

solvent was removed under reduced pressure, and the residue

was taken up into Et₂O-hexane. The resulting suspension was

filtered through Celite, and the filtrate was concentrated under

reduced pressure. Chromatography of the residue on silica

(hexanes-EtOAc, 7:3 to CH₂Cl₂-saturated methanolic am-

monia, 5:1) gave 7.0 mg (56%) of 33 as a pale yellow oil: [R] -

8.3 (c 0.7, CHCl₃); IR (neat) 3500-3200 2968, 2934, 2871, 1482,

1343, 1170, 909 cm^-1; ¹H NMR (300 MHz, CDCl₃) δ 1.02 (d, J

) 6.8 Hz, 3H), 2.47-2.54 (m, 1H), 3.41-3.50 (m, 2H), 3.74-

3.79 (m, 1H), 4.40 (s, 2H), 5.00-5.06 (m, 2H), 5.73 (ddd, J )

7.4, 9.6, 17.2 Hz, 1H), 7.14 (d, J ) 8.5 Hz, 2H), 7.44 (d, J )

8.5 Hz, 2H); ¹³C NMR (75 MHz, CDCl₃) δ 17.1, 39.3, 54.4, 71.8,

72.8, 116.2, 121.9, 129.7, 131.9, 137.5, 140.2; MS (FAB) m/z

284 (M + 1), 273, 107, 89; HRMS (FAB) m/z 284.0649 (calcd

for C₁₃H₁₉ON⁷⁹Br 284.0650).

4-Bromobenzyloxyacetaldehyde (31). To a solution of

2-(4-bromobenzyloxy)ethanol (23.2 mg, 0.10 mmol) in wet CH₂-

Cl₂(1 mL) was added in one portion Dess-Martin periodinane

(46.5 mg, 0.11 mmol). After being stirred at room temperature

for 2 h, the mixture was diluted with saturated aqueous

NaHCO₃-Na₂S₂O₃and extracted with CH₂Cl₂. The extract was

washed with saturated aqueous NaCl, dried over Na₂SO₄, and

concentrated under reduced pressure to give 20.6 mg (90%) of

31 which was used without further purification: IR (neat)

1

2922, 2867, 1730, 1477, 1059, 1014, 795 cm^-1; H NMR (300

MHz, CDCl₃) δ 4.09 (s, 2H), 4.55 (s, 2H), 7.22 (d, J ) 8.5 Hz,

2H), 7.47 (d, J ) 8.5 Hz, 2H), 9.70 (s, 1H); ¹³C NMR (75 MHz,

CDCl₃) δ 73.2, 75.8, 122.5, 130.0, 132.1, 136.4, 200.3; MS (FAB)

m/z 230 (M + 1), 185, 169, 90; HRMS (FAB) m/z 227.9787

(calcd for C₉H₉O₂⁷⁹Br 227.9786).

(3S,4R)-5-(4-Bromobenzyloxy)-4-hydroxy-3-methylpent-

1-ene (32). To a stirred mixture of potassium tert-butoxide

(113 mg, 1.00 mmol) in THF (1.3 mL) and cis-2-butene (2.0

mL, 22 mmol) was added n-butyllithium (1.6 M in hexane, 0.67

mL, 1.00 mmol) at -78 °C. The mixture was stirred at -45

°C for 10 min, the resulting orange solution was recooled to

-78 °C, and a solution of (+)-methoxydiisopinocampheylborane

(387 mg, 1.23 mmol) in Et₂O (2.2 mL) was added dropwise.

The mixture was stirred at -78 °C for 30 min, BF₃‚Et₂O (0.14

mL, 0.137 mmol) was added dropwise, and the mixture was

stirred for 10 min. A solution of 31 (159 mg, 1.00 mmol) in

THF (1.2 mL) was added dropwise, and the mixture was

stirred at -78 °C for 5 h and then allowed to slowly warm to

room temperature. The mixture was treated with NaOH (3

M, 3 mL) and H₂O₂(30%, 1 mL) and was refluxed for 1 h. The

organic phase was separated, washed with water and brine,

and dried (MgSO₄). After removal of the solvent under reduced

pressure, the residual oil was chromatographed on silica gel,

eluting with hexane/EtOAc (8:2) +5% CH₂Cl₂, to give 95 mg

(3S,4S)-5-(4-Bromobenzyloxy)-4-hydroxylamino-3-me-

thylpent-1-ene (36). To a solution of 33 (60.8 mg, 0.214 mmol)

in anhydrous MeOH (0.30 mL) were added p-methoxybenzal-

dehyde (26 µL, 0.214 mmol) and solid Na₂CO₃(34.6 mg, 0.321

mmol). The mixture was stirred for 16 h at room temperature

and for 50 min at 60 °C. After being cooled to room temper-

ature, the mixture was concentrated under reduced pressure,

and the residual oil was taken up in Et₂O (2 mL). The mixture

was filtered, and the filtrate was concentrated under reduced

pressure to yield 84 mg of crude imine 34.

A solution of the crude imine obtained above in CH₂Cl₂(0.4

mL) was cooled to -60 °C, and a solution of m-CPBA (0.5 M,

1974 J. Org. Chem., Vol. 70, No. 6, 2005

Total Synthesis of (-)-7-Epicylindrospermopsin

0.314 mL, 0.157 mmol) was added. The mixture was allowed

to warm to room temperature during 2 h and was diluted with

saturated aqueous NaHCO₃(0.3 mL). The mixture was

extracted with CH₂Cl₂, and the extract was washed with H₂O

(1 mL) and saturated aqueous NaCl (1 mL) and dried over

Na₂SO₄. Removal of the solvent under reduced pressure

afforded 55.2 mg of oxaziridine 35.

pressure to yield 301 mg (98%, 68% from 38) of 46 as a mixture

of diastereomers: IR (neat) 3545-3180, 2960, 2926, 2863,

1650, 1595, 1569, 1472, 1367, 1248, 1086, 729 cm^-1; MS (FAB)

m/z 639 (M + 1), 613, 391, 285, 219, 149, 136, 89; HRMS (FAB)

m/z 639.2603 (calcd for C₂₉H₄₈N₄O₅Si⁷⁹Br 639.2577).

Urea 48. To a solution of 46 (301 mg, 0.470 mmol) in CH₂-

Cl₂(9.4 mL) at 0 °C under argon was added carbonyldiimida-

zole (166 mg, 1.03 mmol), and the mixture was allowed to

warm to room temperature during 1 h. The solution was

stirred at 40 °C for 7 h and concentrated under reduced

pressure to yield a mixture of 47 and 48 which was taken up

in MeOH (5 mL). To the resulting solution was added K₂CO₃

(650 mg, 4.7 mmol), and the mixture was stirred at 60 °C for

1 h. The solvent was removed under reduced pressure to afford

306 mg (85%) of 48 (mixture of stereoisomers) as a tan oil: IR

(neat) 3545-3166, 2960, 2926, 2863, 1717, 1650, 1595, 1566,

1469, 1360, 1246, 1204, 1086, 905, 833, 728 cm^-1; MS (FAB)

m/z 665 (M + 1), 587, 465, 285, 227, 169; HRMS (FAB) m/z

665.2382 (calcd for C₃₀H₄₆N₄O₆Si⁷⁹Br 665.2397).

Ketone 49. To a solution of 48 (360 mg, 0.470 mmol) in

CH₂Cl₂(9.4 mL) was added Dess-Martin periodinane (237 mg,

0.564 mmol), and the mixture was stirred at room temperature

for 1.5 h. The mixture was diluted with saturated aqueous

NaHCO₃-Na₂S₂O₃and extracted with CH₂Cl₂. The extract was

washed with saturated aqueous NaCl, dried over Na₂SO₄, and

concentrated under reduced pressure to give 355 mg (98%) of

49. The ketone was used for the next reaction without further

purification: IR (neat) 2959, 2934, 2852, 1717, 1653, 1592,

1568, 1455, 1352, 1090, 844 cm^-1; HRMS (FAB) m/z 663.2202

(calcd for C₃₀H₄₄N₄O₆Si⁷⁹Br 663.2213).

Alcohol 50. To a solution of 49 (320 mg, 0.483 mmol) in

THF (8 mL) at -78 °C under argon was added L-Selectride

(0.57 mL, 0.67 mmol), and the mixture was stirred for 1 h at

-78 °C. The mixture was allowed to warm to room tempera-

ture during 1 h, after which time EtOH (0.2 mL), water (0.2

mL), H₂O₂(0.2 mL, 30%), and sodium hydroxide (1 M, 0.2 mL)

were added sequentially. The mixture was stirred for 4 h at

room temperature, diluted with saturated aqueous NaHCO₃

(5 mL), and extracted with CH₂Cl₂. The extract was washed

with saturated aqueous NaCl, dried over Na₂SO₄, and con-

centrated under reduced pressure to give 273 mg (85%) of 50

which was used for the next reaction without further purifica-

tion: IR (neat) 3570-3132, 2947, 2926, 2850, 1650, 1578, 1469,

1351, 1204, 1094, 905, 842 cm^-1; MS (FAB) m/z 665 (M + 1),

587, 465, 285, 227, 169; HRMS (FAB) m/z 665.2382 (calcd for

C₃₀H₄₆N₄O₆Si⁷⁹Br 665.2397).

Diol 51. To a solution of the alcohol 50 (262 mg, 0.394 mmol)

in EtOH (16 mL) was added Pd(OH)₂/C (20%, 400 mg), and

the mixture was stirred at ambient temperature under an

atmosphere of H₂for 16 h. The suspension was filtered through

Celite, the filtrate was concentrated under reduced pressure,

and the residue was chromatographed on silica (hexanes-

EtOAc, 1:1, then EtOAc, then EtOAc-methanolic NH₃1-6%)

to give 149 mg (76%, 66% from 48) of 51 which was crystallized

from EtOH: mp 55-57 °C; [R] -59.3 (c 2.0, CHCl₃); IR (neat)

3515-3110, 2962, 2922, 2849, 1635, 1595, 1568, 1480, 1459,

1361, 1090, 838 cm^-1; ¹H NMR (300 MHz, CDCl₃) δ -0.08 (s,

3H), 0.06 (s, 3H), 0.90 (s, 9H), 1.03 (d, J ) 6.8 Hz, 3H), 1.23 (t,

J ) 7.3 Hz, 1H), 1.47 (m, 2H), 1.94 (m, 2H), 2.27 (bs, 1H),

3.19 (dd, J ) 3.8, 10.1 Hz, 1H), 3.50-3.69 (m, 3H), 3.88 (m,

1H), 3.95 (s, 6H), 4.43 (d, J ) 5.5 Hz, 1H), 5.03 (bs, 1H), 5.60

(dd, J ) 3.0, 9.6 Hz, 1H), 6.46 (s, 1H); ¹³C NMR (75 MHz,

CDCl₃) δ -4.7, -4.3. 14.5, 18.5, 26.2, 33.7, 36.8, 41.8, 50.9,

54.1, 54.4, 55.3, 61.3, 62.2, 69.9, 76.1, 99.1, 157.9, 165.4, 171.6,

172.7; MS (FAB) m/z 497 (M + 1), 465, 285, 154, 136; HRMS

(FAB) m/z 497.2795 (calcd for C₂₃H₄₁N₄O₆Si 497.2795).

To a solution of 35 obtained above in MeOH (0.2 mL) was

added hydroxylamine hydrochloride (19.8 mg, 0.286 mmol),

and the mixture was stirred for 2 h at 0 °C. The mixture was

diluted with saturated aqueous NaHCO₃(1 mL) and extracted

with CH₂Cl₂. The extract was washed with saturated aqueous

NaCl and concentrated under reduced pressure. Chromatog-

raphy of the residue on silica (hexanes-EtOAc, 8:2-1:1) gave

25.6 mg (40% from 33) of 36 as a colorless oil: [R] -10.0 (c

1.0, CHCl₃); IR (neat) 3490-3111, 2972, 2909, 2867, 1600,

1

1482, 1237 cm^-1; H NMR (300 MHz, CDCl₃) δ 1.02 (d, J )

6.8 Hz, 3 H), 2.42-2.49 (m, 1H), 2.91 (ddd, J ) 3.8, 7.1, 7.1

Hz, 1H), 3.50 (dd, J ) 6.6, 9.6 Hz, 1H), 3.62 (dd, J ) 3.6, 9.6

Hz, 1H), 4.47 (ABq, J ) 12.3, 17.8 Hz, 2H), 5.03-5.10 (m, 2H),

5.78 (ddd, J ) 8.2, 10.4, 18.4 Hz, 1H), 7.20 (d, J ) 8.5 Hz, 2

H), 7.46 (d, J ) 8.5 Hz, 2H); ¹³C NMR (75 MHz, CDCl₃) δ 16.8,

37.6, 65.7, 68.3, 73.0, 121.9, 129.0, 129.7, 131.9, 137.6, 141.8;

MS (FAB) m/z 300 (M + 1), 246, 244, 171, 169, 113; HRMS

(FAB) m/z 300.0595 (calcd for C₁₃H₁₉O₂N⁷⁹Br 300.0599).

Nitrone 38. To a flask containing 36 (39.0 mg, 0.129 mmol)

and 3 Å molecular sieves (10 mg) was added a solution of 20

(50.8 mg, 0.111 mmol) in CH₂Cl₂(2 mL). The mixture was

immediately concentrated under reduced pressure, and the

residue was taken up in anhydrous MeOH (0.3 mL). The

solution was stirred under argon for 5 h at room temperature

and was concentrated under reduced pressure. Chromatogra-

phy of the residue on silica (hexanes-EtOAc 6:4-4:6) gave

49.3 mg (60%) of 38: [R] -16.9 (c 1.8, CHCl₃); IR (neat) 3469-

3162, 2964, 2922, 2854, 1717, 1600, 1570, 1477, 1351, 1166,

1090, 833 cm ^-1;¹H NMR (300 MHz, CDCl₃) δ 0.02 (s, 3H), 0.14

(s, 3H), 0.94 (s, 9H), 1.02, (d, J ) 6.8 Hz, 3H), 1.34 (s, 9H),

2.37 (ddd, J ) 5.2, 11.2, 18.1 Hz, 1H), 2.62-2.74 (m, 2H), 3.57-

3.67 (m, 2H), 3.93 (s, 3H), 3.96 (s, 3H), 4.14 (m, 1H), 4.44 (ABq,

J ) 12.3, 23.3 Hz, 2H), 4.66 (d, J ) 2.7 Hz, 1H), 4.97-5.00

(m, 2H), 5.66 (d, J ) 9.6 Hz, 1H), 5.74 (ddd, J ) 8.2, 10.1,

17.3 Hz, 1H), 6.51 (s, 1H), 6.85 (t, J ) 4.9 Hz, 1H), 7.13 (d, J

) 8.5 Hz, 2H), 7.44 (d, J ) 8.5 Hz, 2H);¹³C NMR (75 MHz,

CDCl₃) δ -4.6, -4.4, 16.9, 18.5, 26.2, 28.6, 29.9, 38.0, 53.2,

54.2, 55.1, 68.6, 72.8, 75.4, 79.4, 79.8, 99.5, 116.1, 121.8, 129.5,

131.9, 137.5, 138.1, 139.4, 156.0, 165.1, 172.2, 172.6; MS (FAB)

m/z 737 (M + 1), 285, 171, 136, 89; HRMS (FAB) m/z 737.2937

(calcd for C₃₄H₅₄N₄O₇Si⁷⁹Br 737.2945).

Piperidine 45. A suspension of 38 (510 mg, 0.691 mmol)

and 4 Å molecular sieves (60 mg) in toluene (14 mL) was

stirred under argon for 72 h at 96 °C. The mixture containing

40 and 41 (ca. 2:1) was filtered through a pad of Celite, and

the filtrate was concentrated under reduced pressure. The

residue was taken up into THF (9 mL), and H₂O (9 mL),

saturated aqueous NH₄Cl (9 mL), and Zn dust (2.20 g, 34.5

mmol) were added sequentially. The mixture was stirred for

8 h at 65 °C, cooled to room temperature, and diluted with

CH₂Cl₂(30 mL). Saturated aqueous NaHCO₃(30 mL) was

added, and the mixture was extracted with CH₂Cl₂. The extract

was washed with saturated aqueous NaCl and dried over Na₂-

SO_4.Removal of the solvent under reduced pressure afforded

400 mg of crude 45. The crude material was carried forward

to the next reaction without further purification.

Amine 46. A solution of crude 45 obtained above in

methanolic HCl (1 M, 2 mL) was concentrated under reduced

pressure at 45 °C. This process was repeated three times, and

the residue was taken up into methanolic HCl (1 M, 1 mL),

MeOH (2 mL), and water (2 mL). The mixture was washed

with Et₂O, and the aqueous phase was diluted with CH₂Cl₂(5

mL). The organic phase was made basic (pH 8-9) with K₂-

CO₃and was extracted with CH₂Cl₂. The extract was washed

with saturated aqueous NaCl and concentrated under reduced

Azide 54. To a solution of 51 (12.4 mg, 0.02 mmol) in THF

(1.2 mL) at room temperature under argon was added triph-

osgene (3.0 mg, 0.01 mmol), and the mixture was stirred for

45 min. To this solution was added MeOH (1 mL), and the

mixture was stirred for 5 min and concentrated under reduced

pressure. The residue was taken up in anhydrous DMF (1.0

J. Org. Chem, Vol. 70, No. 6, 2005 1975

White and Hansen

FIGURE 2. ¹H NMR spectra of natural and synthetic (-)-7-epicylindrospermopsin (1) in MeOH-d₄.

mL), sodium azide (16.2 mg, 0.25 mmol) was added, and the

mixture, which darkened to an orange-red color, was stirred

at 60 °C for 3 h and then diluted with CH₂Cl₂(5 mL). The

solution was washed with water (3 × 3 mL) and saturated

aqueous NaCl, dried over Na₂SO₄, and concentrated under

reduced pressure. Chromatography of the residue on silica

(EtOAc to EtOAc-methanolic NH₃, 2%) gave 6.4 mg (49%) of

54: [R] +0.5 (c 0.64, CHCl₃); IR (neat) 3534-3122, 2956, 2921,

2856, 2101, 1643, 1628, 1593, 1566, 1458, 1351, 1093, 842

cm^-1; ¹H NMR (300 MHz, CDCl₃) δ -0.09 (s, 3H), 0.10 (s, 3H),

0.91 (s, 9H), 1.03 (d, J ) 6.8 Hz, 3H), 1.53-1.76 (m, 4H), 1.87

(ddd, J ) 3.8, 7.6, 13.7 Hz, 1H), 1.95-2.06 (m, 1H), 3.30 (dd,

J ) 3.0, 12.6 Hz, 1H), 3.50-3.75 (m, 3H), 3.92-3.98 (m, 1H),

3.97 (s, 3H), 3.98 (s, 3H), 4.10 (dd, J ) 4.6, 12.6 Hz, 1H), 4.33

(d, J ) 7.1 Hz, 1H), 5.07 (bs, 1H), 6.46 (s, 1H); ¹³C NMR (75

MHz, CDCl₃) δ -4.6, -4.2, 14.5, 16.4, 18.5, 23.1, 26.2, 30.1,

32.3, 33.2, 36.0, 38.3, 47.6. 53.0, 54.3, 55.3, 55.6, 56.1, 67.3,

77.0, 78.4, 99.3, 158.1, 165.4, 171.8, 172.6; MS (FAB) m/z 522

(M + 1), 465, 285, 219; HRMS (FAB) m/z 522.2863 (calcd for

C₂₃H₄₀N₇O₅Si 522.2860).

(FAB) m/z 636 (M + 1) 579, 285, 154, 136, 87; HRMS (FAB)

m/z 636.3747 (calcd for C₂₉H₅₃N₇O₅Si₂636.3725).

Diol 60. To a solution of 56 (6.0 mg, 9.4 µmol) in CH₂Cl₂

(400 µL) at 0 °C under argon was added KHMDS (0.5 M in

toluene, 90.0 µL, 47.0 µmol), and the mixture was stirred for

10 min. Solid trimethyloxonium tetrafluoroborate (2.8 mg, 18.0

µmol) was added, and the mixture was stirred at 0 °C for 8 h

and then was diluted with aqueous NH₄Cl and extracted with

Et₂O. The extract was washed with brine, dried over Na₂SO₄,

and concentrated under reduced pressure to give crude 57.

This material was taken up in MeOH (1 mL), Pd/C (20%, 3

mg) was added, and the mixture was stirred under an

atmosphere of H₂at room temperature for 8 h. The suspension

was filtered through a pad of Celite, the filtrate was concen-

trated under reduced pressure, and the resulting crude 59 was

taken up in concd HCl. The solution was heated at reflux for

7 h and concentrated under reduced pressure, and the residue

was purified by HPLC on a Zorbax 300SB-C18 column (9.4 ×

250 mm), eluting with 4% MeOH/H₂O containing 0.1% TFA

at 4 mL/min (retention time 20.1 min), to provide 0.65 mg (21%

from 56) of pure 60 as a colorless amorphous solid: [R] -8.3

(c 0.06, H₂O); ¹H NMR (400 MHz, D₂O) δ 0.98 (d, J ) 6.9 Hz,

3H), 1.51-1.79 (m, 3H), 2.07-2.19 (m, 2H), 3.28 (dd, J ) 8.8,

10.6 Hz, 1H), 3.58-3.68 (m, 1H), 3.73-3.89 (m, 3 H), 4.03 (s,

1H), 4.50 (d, J ) 6.6 Hz, 1H), 5.86 (s, 1H); ¹³C NMR (150 MHz,

D₂O) δ 13.1, 30.2, 37.8, 39.7, 44.1, 47.6, 52.6, 56.9, 68.3, 71.2,

99.7; MS (FAB) m/z 336 (M⁺) 282, 185, 93; HRMS (FAB) m/z

336.1673 (calcd for C₁₅H₂₂N₅O₄: 336.1672). The ¹H and ¹³C

NMR spectra of this compound were identical to those of the

corresponding racemic compound prepared by Weinreb.¹⁴

Triethylsilyl Ether 56. To a solution of 54 (6.0 mg, 12.0

µmol) in CH₂Cl₂(100 µL) at 0 °C under argon were added

triethylsilyl triflate (0.8 µL, 5.0 µmol) and Et₃N (0.5 µL, 5.0

µmol), and the mixture was stirred for 20 min. The mixture

was warmed to room temperature, diluted with aqueous NH₄-

Cl, and extracted with Et₂O. The extract was washed with

brine, dried over Na₂SO₄, and concentrated under reduced

pressure to afford 6.6 mg (99%) of pure 56: [R] -0.7 (c 0.6,

CHCl₃); IR (neat) 2962, 2922, 2874, 2852, 2095, 1659, 1592,

1

1568, 1462, 1358, 1115, 1090, 838 cm^-1; H NMR (300 MHz,

CDCl₃) δ -0.10 (s, 3H), 0.11 (s, 3H), 0.56 (q, J ) 7.6 Hz, 6 H),

0.88-0.99 (m, 21 H), 1.50-1.64 (m, 3H), 1.73 (ddd, J ) 3.8,

7.6, 13.7 Hz, 1H), 1.85-1.93 (m, 1H), 3.2 (dd, J ) 2.9, 12.4

Hz, 1H), 3.48 (q, J ) 6.9 Hz, 1H), 3.52-3.69 (m, 2H), 3.83-

3.90 (m, 1H), 3.97 (s, 3H), 3.98 (s, 3H), 4.13 (dd, J ) 4.3, 12.4

Hz, 1H), 4.32 (d, J ) 6.9 Hz, 1 H), 5.05 (bs, 1H), 6.47 (s, 1H);

¹³C NMR (75 MHz, CDCl₃) δ -4.6, -4.2, 5.3, 7.3, 15.7, 16.8,

18.5, 26.2, 30.1, 33.2, 36.7, 39.7, 39.4, 47.9, 52.9, 54.3, 55.3,

55.7, 56.3, 66.3, 67.7, 78.6, 99.3, 158.2, 165.4, 171.9, 172.6; MS

7-Epicylindrospermopsin (1). To a solution of 60 (0.65

mg, 1.90 µmol) in dry pyridine (0.1 mL) containing Na₂SO₄

(10 mg) was added a solution of SO₃‚DMF in DMF (0.1 M, 216

µL, 21.6 µmol, 11 equiv), and the mixture was stirred at

ambient temperature for 30 min. The solvent was removed in

vacuo, and the residue was taken up in MeOH. The mixture

was filtered, and the filtrate was concentrated. Purification

of the residual oil by HPLC on a Zorbax 300SB-C18 column

(9.4 × 250 mm), eluting with 1.5% MeOH in water containing

1976 J. Org. Chem., Vol. 70, No. 6, 2005

Total Synthesis of (-)-7-Epicylindrospermopsin

0.1% TFA at 4 mL/min, gave 0.51 mg (63%) of 1: [R] -17.2 (c

0.03, H₂O) (lit.⁷[R] -20.5 (c 0.06, H₂O)); HRMS (FAB) m/z

416.1238 (calcd for C₁₅H₂₂N₅O₇S 416.1240). The ¹H NMR

spectrum corresponded closely with that of natural 7-epicy-

lindrospermopsin (Figure 2).

support was provided by the National Science Founda-

tion (0076103-CHE) and by the National Institute of

General Medical Sciences (GM58889). We are greatful

to Dr. Ellen Laird, Array Biopharma, Boulder, CO, for

assistance with the preparation of the cover graphic.

Acknowledgment. We are indebted to Professor

Alexandre F. T. Yokochi, Department of Chemical

Engineering, Oregon State University, for the X-ray

crystal structures of 30 and 51, and to Professor Steven

Weinreb, Pennsylvania State University, for the NMR

spectra of 60 and for a preprint of ref 14. Financial

Supporting Information Available: X-ray crystallo-

graphic data for 30 and 51 and copies of NMR spectra of new

compounds. This material is available free of charge via the

Internet at http://pubs.acs.org.

JO0486387

J. Org. Chem, Vol. 70, No. 6, 2005 1977

Article Doi

Total synthesis of (-)-7-epicylindrospermopsin, a toxic metabolite of the freshwater cyanobacterium Aphanizomenon ovalisporum, and assignment of its absolute configuration

DOI: 10.1021/jo0486387

Source and publish data:

Authors:

Article abstract of DOI:10.1021/jo0486387

Full text of DOI:10.1021/jo0486387

Products guided by the article

R&D Labs maybe for 812646-16-5

Relevant to this article

Hot Product