Synthesis of the angiotensin-converting enzyme inhibitors (-)-A58365A and (-)-A58365B from a common intermediate

Article abstract of DOI:10.1021/jo9822941

(-)-A58365A (1) and (-)-A58365B (2), which are inhibitors of angiotensin-converting enzyme, were synthesized from the subunits 9 and 10. These were coupled, and the resulting individual amides 17a,b were converted by ozonolysis into aldehydes 18a,b, which underwent cyclodehydration to the enamides 19a,b. Treatment with a stannane served to generate the vinyl stannanes 20a,b, from which ketones 22a,b were produced by protodestannylation and ozonolysis. Base treatment and hydrogenolysis then afforded (-)-A58365A (1). The intermediates 17a,b were also converted into aldehydes 26a,b by hydroboration and oxidation, and a similar sequence to that used for (-)-A58365A was then applied in order to complete the first enantiospecific synthesis of the congener, (-)-A58365B (2).

Full text of DOI:10.1021/jo9822941

J . Org. Chem. 1999, 64, 1447-1454
1447
Syn th esis of th e An gioten sin -Con ver tin g En zym e In h ibitor s
(-)-A58365A a n d (-)-A58365B fr om a Com m on In ter m ed ia te
Derrick L. J . Clive,* Don M. Coltart, and Yuanxi Zhou
Chemistry Department, University of Alberta, Edmonton, Alberta, Canada T6G 2G2
Received November 19, 1998
(-)-A58365A (1) and (-)-A58365B (2), which are inhibitors of angiotensin-converting enzyme, were
synthesized from the subunits 9 and 10. These were coupled, and the resulting individual amides
17a ,b were converted by ozonolysis into aldehydes 18a ,b, which underwent cyclodehydration to
the enamides 19a ,b. Treatment with a stannane served to generate the vinyl stannanes 20a ,b,
from which ketones 22a ,b were produced by protodestannylation and ozonolysis. Base treatment
and hydrogenolysis then afforded (-)-A58365A (1). The intermediates 17a ,b were also converted
into aldehydes 26a ,b by hydroboration and oxidation, and a similar sequence to that used for (-)-
A58365A was then applied in order to complete the first enantiospecific synthesis of the congener,
(-)-A58365B (2).
We report the synthesis1 of (-)-A58365A (1) and (-)-
A58365B (2), two natural products which were isolated2a,b
in the Eli Lilly Laboratories at Indianapolis from the
fermentation broth of a soil bacterium. The compounds
are powerful inhibitors2c of angiotensin-converting en-
zyme, and this property should make their structures
relevant to the design of blood-pressure-lowering drugs.3
When we began our studies, a route to (-)-A58365A
had already been published,5 and during the course of
our work the synthesis of (()-A58365B was also re-
ported.6
for the corresponding saturated materials (cf. 3), with the
intention of then using one of the standard methods for
introducing a double bond R to a ketone carbonyl.
However, this approach is not straightforward and leads
to poor yields.6-8 Consequently, we decided to introduce
the troublesome C(2)-C(3)9 double bond in the disguised
form of an additional ring, as shown in 4 (eq 1, R )
protecting group). Such compounds (4) should be convert-
ible by the action of base into 5, which represent
protected versions of the target natural products. In the
event, this approach was very successful, and we applied
it first to (()-A58365B.1a Extensive exploratory work was
required, however, to devise an efficient synthesis of the
intermediate spirolactone 4 (n ) 2, R ) Me), and we
eventually developed a route based on the radical cy-
clization10 summarized in Scheme 1. With 4 (n ) 2, R )
Me) in hand, conversion into 5 (n ) 2, R ) Me) was easy,
As the compounds are formally enols of R,â-unsatur-
ated ketones, a potential synthetic approach is to aim
(7) de Lima, D. P., Ph.D. Thesis, 1994, Universidade Federal Minas
Gerais (Brazil). The studies on (()-A58365B were done at the
University of Alberta.
(8) For example, treatment of 3 (n ) 1, R ) Me) with DDQ leads to
i (see ref 6). Desaturation of 3 (n ) 1 or 2, R ) Me) had to be done in
an indirect way (see ref 6).
(1) Preliminary communications: (a) Clive, D. L. J .; Zhou, Y.; de
Lima, D. P. J . Chem. Soc., Chem. Commun. 1996, 1463-1464. (b) Clive,
D. L. J .; Coltart, D. M. Tetrahedron Lett. 1998, 39, 2519-2522.
(2) (a) Isolation: Mynderse, J . S.; Samlaska, S. K.; Fukuda, D. S.;
Du Bus, R. H.; Baker, P. J . J . Antibiot. 1985, 38, 1003-1007. (b)
Structure: Hunt, A. H.; Mynderse, J . S.; Samlaska, S. K.; Fukuda, D.
S.; Maciak, G. M.; Kirst, H. A.; Occolowitz, J . L.; Swartzendruber, J .
K.; J ones, N. D. J . Antibiot. 1988, 41, 771-779. (c) Biological activity:
O’Connor, S.; Somers, P. J . Antibiot. 1985, 38, 993-996.
(3) (a) Douglas, W. W. In The Pharmacological Basis of Therapeutics,
7th ed.; Gilman, A. G., Goodman, L. S., Rall, T. W., Murad, F., Eds.;
Macmillan: New York, 1985; pp 639-659. (b) For studies on analogues
of 1, see ref 4.
(4) (a) Mynderse, J . S.; Fukuda, D. S.; Hunt, A. H. J . Antibiot. 1995,
48, 425-427. (b) Mynderse, J . S.; Fukuda, D. S. EP 133038 (Chem.
Abstr. 1985, 103, 160306).
(5) Fang, F. G.; Danishefsky, S. J . Tetrahedron Lett. 1989, 30, 3621-
3624.
(6) For a synthesis of (()-A58365B and formal syntheses of (()- and
(-)-A58365A, see: Wong, P. L.; Moeller, K. D. J . Am. Chem. Soc. 1993,
115, 11434-11445.
(9) Nonsystematic numbering.
(10) For intermolecular addition of radicals (carrying electron-
withdrawing groups) to the distal terminus of an enamine double bond,
see: Renaud, P.; Schubert, S. Synlett 1990, 624-626. For intramo-
lecular cyclization of a radical onto the proximal terminus of a enamide
double bond, see: Yuasa, Y.; Kano, S.; Shibuya, S. Heterocycles 1991,
32, 2311-2314. Beckwith, A. L. J .; J oseph, S. P.; Mayadunne, R. T. A.
10.1021/jo9822941 CCC: $18.00 © 1999 American Chemical Society
Published on Web 02/18/1999

1448 J . Org. Chem., Vol. 64, No. 5, 1999
Clive et al.
Sch em e 1
Sch em e 3a
a
(a) Zn, BrCH2CH2Br, Me3SiCl, THF; (b) CuCN, LiCl, THF,
-15 °C to 0 °C; allyl chloride, -30 °C to 0 °C; 84% from 14; (c)
CF3CO2H, CH2Cl2; 97%.
Sch em e 4a
Sch em e 2a
a
(a) MeOH, TsOH‚H2O, CHCl3, azeotropic distillation; 94%; (b)
propargyl bromide, Al, HgCl2, THF, -78 °C; 95%; (c) LiOH, THF,
H2O; Amberlite IR-120; evaporate eluant at 60 °C; 95%.
and the last stepshydrolysis of the methyl esterscould
be accomplished1a by using Bu3SnOSnBu3.11
With these experiments as background, we were able
to improve our approach by using a benzyl instead of a
methyl ester, to facilitate the final deprotection. At the
same time we introduced some modifications so that both
1 and 2 could be made optically pure from a common
advanced intermediate.
a
(a) 1-(3-(Dimethylamino)propyl)-3-ethylcarbodiimide hydro-
chloride, 1-hydroxybenzotriazole, CH2Cl2, DMF; 95% of a 1:1
mixture of 17a (chromatographically less polar diastereoisomer)
and 17b; (b) O3, CH2Cl2, -78 °C; Ph3P, -78 °C to room temper-
ature; 97% for 18a , and 96% for 18b; (c) BaO, CH2Cl2, sonication;
P2O5, sonication; 71% or 93% (corrected for recovered 18a ) for 19a ;
87% for 19b.
at 60 °C. We assume that the intermediate hydroxy
diacid cyclizes during evaporation of the solvent.
The final version of our synthesis is based on the
racemic acid 9 and the optically pure amino acid benzyl
ester 10.
The other component, optically active amino acid
benzyl ester 10, was a known compound,14 which was
made as shown in Scheme 3. We actually first prepared
it in racemic form and found that considerable experi-
mentation was needed in order to develop a reliable and
efficient method for making the intermediate organozinc
(cf. 15). This was eventually accomplished by a judicious
blend of two literature procedures.15,16
Compounds 9 and 10 were coupled under standard
conditions [1-(3-(dimethylamino)propyl)-3-ethylcarbodi-
imide hydrochloride, 1-hydroxybenzotriazole; 95%] to
obtain a mixture of diastereoisomers 17a ,b, which could
be separated chromatographically (Scheme 4). Each of
these was then ozonized to the corresponding aldehyde
(18a ,b) in very high yield (g96%). The next step, cycliza-
tion to enamides 19a ,b, was first studied using racemic
compounds. Application of the reagent system (CF3CO2H,
4 Å molecular sieves) used in our earlier synthesis1a of
(()-A58365B was unsuccessful,17 but we eventually found
Syn th esis of (-)-A58365A. Acid 9 was prepared from
R-ketoglutaric acid (11) (Scheme 2) by methylation (11
f 12) and treatment with the reagent12 generated from
propargyl bromide and amalgamated aluminum. In this
second step (12 f 13) no allene was detected,13 and the
required tertiary alcohol could be isolated in high (95%)
yield. Conversion into acid 9 was achieved (95%) by base
hydrolysis (LiOH), followed by ion-exchange chromatog-
raphy (Amberlite IR-120), and evaporation of the eluant
J . Org. Chem. 1993, 58, 4198-4199. Beckwith, A. L. J .; Westwood, S.
W. Tetrahedron 1989, 45, 5269-5282. For cyclization onto the distal
terminus of an enamide double bond, see: Schultz, A. G.; Guzzo, P.
R.; Nowak, D. M. J . Org. Chem. 1995, 60, 8044-8050.
(14) Dunn, M. J .; J ackson, R. F. W. J . Chem. Soc., Chem. Commun.
1992, 319-320.
(15) Yeh, M. C. P.; Knochel, P. Tetrahedron Lett. 1989, 30, 4799-
(11) Mata, E. G.; Mascaretti, O. A. Tetrahedron Lett. 1988, 29,
6893-6896.
4802.
(16) Dunn, M. J .; J ackson, R. F. W.; Pietruszka, J .; Wishart, N.; Ellis,
D.; Wythes, M. J . Synlett 1993, 499-500.
(12) La¨uger, P.; Prost, M.; Charlier, R. Helv. Chim. Acta 1959, 42,
2379-2393. Schneider, D. F.; Weedon, B. C. L. J . Chem. Soc. (C) 1967,
1686-1689. J anardhanam, S.; Shanmugam, P.; Rajagopalan, K. J . Org.
Chem. 1993, 58, 7782-7788. J anardhanam, S.; Balakumar, A.; Raja-
gopalan, K. J . Chem. Soc., Perkin Trans. 1 1994, 551-556.
(13) Cf. Klein, J . In The Chemistry of the Carbon-Carbon Triple
Bond, Patai, S., Ed.; Wiley: Chichester, 1978; Part 1, pp 343-379.
(17) Cyclization to six-membered enamides appears to be easier than
for the five-membered series: Cf. Robl, J . A. Tetrahedron Lett. 1994,
35, 393-396. Ojima, I.; Tzamarioudaki, M.; Eguchi, M. J . Org. Chem.
1995, 60, 7078-7079.
(18) Cf. Paterson, I.; Cowden, C.; Watson, C. Synlett 1996, 209-
211.

Synthesis of Angiotensin-Converting Enzyme Inhibitors
J . Org. Chem., Vol. 64, No. 5, 1999 1449
Sch em e 5a
Sch em e 6a
a
(a) Bu3SnH, AIBN, PhMe, reflux; (b) CF3CO2H, THF; 93%
overall for 21a ; 94% overall for 21b; (c) O3, CH2Cl2, -78 °C; Ph3P,
-78 °C to room temperature; (d) Et3N, THF, 60 °C; 95% overall
from 21a ; 96% overall from 21b; (e) H2, Pd-C, MeOH; 93%.
a
(a) 9-BBN, THF, 0 °C to room temperature; PCC, CH2Cl2, 4 Å
molecular sieves, reflux; 83% for 26a ; 74% for 26b; (b) BaO,
CH2Cl2, sonication; P2O5, sonication; 83% for 27a , 80% for 27b;
(c) Bu3SnH, AIBN, PhMe, reflux; CF3CO2H, THF; 74% overall from
27a ; 92% overall from 27b; (d) O3, CH2Cl2, -78 °C; Ph3P, -78 °C
to room temperature; Et3N, THF, 60 °C; 95% overall from 28a ;
96% overall from 28b; (e) H2, Pd-C, MeOH; 96%.
that sonication of the amido aldehydes with BaO, fol-
18
lowed by addition of P2O5 (with continued sonication),
led to smooth ring closure and dehydration, and the same
conditions were used in the optically active series (Scheme
4, 18a f 19a , 71% or 93% based on recovered starting
material; 18b f 19b, 87%). Both 19a and 19b exist as a
mixture of rotamers, as judged by variable temperature
1H NMR measurements.
(-)-1 corresponded to an optical purity of 98.5%.19 This
value was confirmed by the following experiments. Ben-
zyl ester 23 was dimethylated (CH2N2) and the product
(24, n ) 1) examined by HPLC, using an optically active
Rapid addition (over a few seconds) of a PhMe solution
(best added in two equal portions) of Bu3SnH and AIBN
to a refluxing solution of 19a in the same solvent gave
20a (77%) after a reflux period of ca. 5 h (Scheme 5).
Similarly, 19b gave 20b (85%). However, as described
below, the efficiency of the process is much greater than
these yields would imply. Each vinyl stannane was a
single compound of undetermined double bond geometry.
Although the vinyl stannanes could be purified by flash
chromatography and then subjected to protodestannyla-
tion with CF3CO2H, it was far more efficient to use the
crude material directly for protodestannylation. In this
manner, 19a was converted into 21a (93% from 19a ), and
19b into 21b (94%). Ozonolytic cleavage of the exocyclic
double bond of each stereoisomer proceeded efficiently
(21a f 22a ; 21b f 22b), but the ketones were difficult
to separate from Ph3PO, formed during reductive workup
of the ozonolysis mixture. Therefore, the crude ketones
were treated directly with Et3N at 60 °C, and this
operation served to open the lactone, introduce the C(2)-
C(3) double bond, and release the propionate side chain
(21a f 23, 95%; 21b f 23, 96%). Finally, hydrogenolysis
of the benzyl group gave (-)-A58365A (1) in 93% yield
as a white foam.
stationary phase that gave baseline resolution with (()-
24 (n ) 1). Our optically active material was found to
have an ee of 99.5%. Synthetic 1 was converted into its
trimethyl derivative (25, n ) 1), and this, as well as the
corresponding racemic material, was examined by HPLC.
An ee of 96.2% was determined, but in this case clear
baseline resolution of the enantiomers was not obtained.
We conclude that, within the limits of experimental error,
the present route involves no loss of optical integrity.
Syn th esis of (-)-A58365B. Intermediates 17a,b served
also for the synthesis of 2. In the synthesis of 1, described
above, the olefinic carbon chain of 17a ,b was shortened
by oxidative cleavage of the double bond. In the present
case, to make 2, oxidation of the double bond is again
required, but with retention of the complete carbon chain,
so that a six-membered ring can eventually be generated
(see ring B in 2). This oxidation was effected (Scheme 6)
by taking advantage of the selectivity of 9-BBN for an
alkene over an alkyne.20 While the use of 9-BBN did
indeed serve to functionalize the terminal double bond,
we found it difficult to isolate the derived alcohols21
because of hydrolysis of the benzyl ester during the
borane oxidation step. We could, however, obtain the
corresponding aldehydes directly, by oxidizing the inter-
mediate boranes in situ with PCC22 (17a f 26a , 83%;
17b f 26b, 74%),23 and so this temporary problem was
actually an advantage, as we would otherwise have had
to oxidize the alcohols in a separate step.
The whole of the above sequence was also done using
the racemic amino acid benzyl ester corresponding to 10,
so as to afford (()-A58365A.
Our key amino acid 10 was made from serine of 97%
ee (Aldrich), and the specific rotation of our synthetic
(19) The reported [R]25 ) -199.5° (c 1.0, H2O) (see reference 2b);
D
our material had [R]25 ) -196.55° (c 0.87, H2O).
D
(20) Brown, C. A.; Coleman, R. A. J . Org. Chem. 1979, 44, 2328-
2329.
(21) These exploratory studies were done on the more polar (TLC,
silica gel) diastereoisomer [(+)-17b].
(22) Cf. Brown, H. C.; Kulkarni, S. U.; Rao, C. G. Synthesis, 1980,
151-153.
From 26a ,b, cyclization with the BaO/P2O5 combina-
tion again worked well and took the route as far as
(23) Oxidation of the boranes to the aldehydes could also be done
with TPAP, but in lower yield.

1450 J . Org. Chem., Vol. 64, No. 5, 1999
Clive et al.
ous NaCl (50 mL), dried (MgSO4), and evaporated. Flash
chromatography of the residue over silica gel (2 × 16 cm), using
30:70 EtOAc-hexanes, gave pure (1H NMR) 13 (185 mg, 95%)
as a colorless oil: FTIR (KBr) 3502, 3286, 1739 cm-1; 1H NMR
(CDCl3, 300 MHz) δ 2.02-2.15 [m, 3 H including a t at δ 2.05
(J ) 2.7 Hz)], 2.16-2.26 (m, 1 H), 2.41-2.52 (m, 1 H), 2.56 (A
of an ABX system, apparent dd, J ) 16.9, 2.7 Hz, 1 H), 2.66
(B of an ABX system, apparent dd, J ) 16.9, 2.7 Hz, 1 H),
3.52 (s, 1 H), 3.65 (s, 3 H), 3.78 (s, 3 H); 13C NMR (CDCl3, 75.5
MHz) δ 28.65 (t′), 30.26 (t′), 32.89 (t′), 51.81 (q′), 53.26 (q′),
71.66 (s′), 75.83 (d′), 78.42 (s′), 173.36 (s′), 174.79 (s′); exact
mass m/z calcd for C9H11O4 (M - OCH3) 183.0657, found
183.0657.
Tet r a h yd r o-5-oxo-2-[2-(p r op yn yl)]-2-fu r a n ca r b oxylic
Acid (9). A solution of LiOH (509 mg, 10.4 mmol) in water (2
mL) was added to a solution of diester 13 (1.00 g, 4.67 mmol)
in THF (20 mL), and the mixture was stirred overnight at room
temperature. Evaporation of the solvent gave what we assume
to be the dilithium salt, as a solid.
enamides 27a ,b (83% for 27a ; 80% for 27b). In our earlier
synthesis of (()-A58365B1a cyclization of the correspond-
ing methyl esters had been effected with CF3CO2H/4 Å
molecular sieves, but in lower yield for one of the
diastereoisomers.
With the individual diastereoisomers 27a and 27b in
hand, treatment with Bu3SnH/AIBN, this time added in
one portion, followed directly by protodestannylation,
gave 28a and 28b, respectively. One of the diastereoiso-
mers of 27a ,b exists as a mixture of rotamers (cf. 19a ,b),
1
each of which is observable by H NMR measurements,
and this isomer cyclizes less efficiently than the other
(74% overall yield of the protodestannylated product
versus 92%). The same phenomenon had been met1a in
our earlier synthesis of (()-A58365B, and in that case
use of Ph3SnH led to some improvement in the yield. In
the present instance we did not vary the nature of the
stannane, but did try the experiment in boiling xylene
and found no improvement.
The exocyclic olefins 28a ,b were individually ozonized
and treated with Et3N, so as to generate benzyl ester 29
(95% for 28a ; 96% for 28b), and final deprotection was
again done by hydrogenolysis (29 f 2; 96%), thus
completing the first synthesis of (-)-A58365B. The entire
sequence was repeated in the racemic series. The optical
purity of the synthetic natural product was established
in the same manner as for its congener: 99% ee on the
basis of optical rotation measurements,24 98.3% by HPLC
analysis of the derived trimethyl derivative (25, n ) 2),
and 99.6% by HPLC analysis of the dimethyl benzyl
derivative (24, n ) 2). Again, we conclude that the route
involves no loss of optical integrity.
A column packed with Amberlite IR-120 ion-exchange resin
(20-50 Å mesh, 2.5 × 16 cm) was washed with water until
the eluant was colorless. The column was then washed
successively with 2 M aqueous NaOH (4 bed volumes), water
(until the eluant was neutral to pH paper), 2 M HCl (4 bed
volumes), and finally with water (until the eluant was neutral
to pH paper). The above dilithium salt was dissolved in water
(2-3 mL), and the solution was passed down the column, using
water. The eluant was monitored with pH paper or by TLC
(silica gel, 10:90 MeOH-CHCl3). Evaporation of the combined
acidic fractions (water pump, rotary evaporator, bath temper-
ature 65-70 °C) gave 9 (746 mg, 95%) as a pure (1H NMR),
light-yellow, powder: FTIR (KBr) 3500-2500 (br), 1770, 1720
1
cm-1; H NMR (D2O, 300 MHz) δ 2.45-2.70 (m, 3 H), 2.75-
2.85 (m, 2 H), 2.92 (A of an ABX system, apparent dd, J )
16.9, 2.9 Hz, 1 H), 3.06 (B of an ABX system, apparent dd, J
) 16.9, 2.9 Hz, 1 H); 13C NMR (D2O, 75.5 MHz) δ 27.78 (t′),
29.03 (t′), 30.74 (t′), 73.35 (d′), 79.10 (s′), 86.90 (s′), 174.86 (s′),
180.54 (s′); exact mass m/z calcd for C5H5O4 (M - CH2CtCH)
129.0188, found 129.0189; FABMS m/z calcd for C8H8O4
168.15, found 168.9.
In the above work the use of spirolactones constitutes
an effective method for introduction of the C(2)-C(3)
double bond, and it is likely that our route can be
modified to make a variety of analogues, in particular,
those differing in the length of the side chain or by the
presence of substituents on the amino acid component,
especially R to the carboxyl group.
P h en ylm eth yl (S)-2-[[(1,1-Dim eth yleth oxy)ca r bon yl]-
a m in o]-5-h exen oa te [(-)-16]. Zn dust (0.7452 g, 11.40 mmol)
was added to a round-bottomed flask containing a magnetic
stirring bar. The flask was sealed with a rubber septum,
evacuated with an oil-pump, and refilled with Ar. The evacu-
ation-filling process was repeated twice more, and then the
system was placed under a static pressure of Ar. Dry THF (4
mL) and 1,2-dibromoethane (0.05 mL, 0.58 mmol) were added.
The resulting mixture was heated to boiling with the aid of a
heat gun and then allowed to cool for 1 min. The heating-
cooling process was repeated three more times, and then the
flask was allowed to cool for an additional 5 min. Freshly
distilled, dry Me3SiCl (0.07 mL, 0.55 mmol) was added, and
the resulting mixture was stirred for 10 min and then placed
in an oil bath set at 38 °C. Stirring was continued for 10 min.
A solution of (-)-1426 (1.1526 g, 2.84 mmol) in dry THF (5 mL)
was added by cannula over ca. 1 min, and additional THF (3
× 1 mL) was used as a rinse. The resulting mixture was stirred
at 38 °C (Ar atmosphere) until all the starting material had
been consumed (ca. 6 h, TLC control, silica gel, 10:90 EtOAc-
hexanes).
The above solution of alkylzinc iodide (15) was transferred
by cannula over ca. 1 min to a stirred and precooled (bath
temperature -15 °C) solution of CuCN (0.2811 g, 3.14 mmol)
and LiCl (0.2661 g, 6.28 mmol) in dry THF (7 mL), and further
portions of THF (2 × 1 mL) were used as a rinse to complete
the transfer. The reaction flask was removed from the cold
bath and placed in another at 0 °C. Stirring was continued
for 15 min, and the reaction flask was then transferred to a
cold bath at -30 °C. Stirring was continued for 10 min, allyl
chloride (0.46 mL, 5.64 mmol) was added, and, finally, the
Exp er im en ta l Section
Gen er a l P r oced u r es. The same general procedures as
used previously25 were followed. The symbols s′, d′, t′, and q′
used for 13C NMR signals indicate zero, one, two, or three
attached hydrogens, respectively.
Dim et h yl 2-H yd r oxy-2-[2-(p r op yn yl)]p en t a n ed ioa t e
(13). A mixture of Al powder12 (84.0 mg, 3.11 mmol) and HgCl2
(5 mg, 0.02 mmol) in dry THF (3 mL) was stirred vigorously
for 1 h (Ar atmosphere) in a three-necked flask fitted with a
reflux condenser and closed by septa. Most of the solvent was
then withdrawn by syringe from the resulting shiny Al, and
fresh THF (3 mL) was injected. The mixture was warmed in
an oil bath set at 40 °C, and propargyl bromide (363 mg, 3.05
mmol) in THF (1 mL) was then added slowly with vigorous
stirring and at such a rate that the THF did not boil. The
addition took ca. 10 min. After the addition, stirring at 40 °C
was continued until a dark gray solution was obtained (ca. 1
h). This was cooled to room temperature and added by cannula
at a fast dropwise rate to a stirred and cooled (-78 °C) solution
of 12 (158 mg, 0.90 mmol) in THF (5 mL). Stirring at -78 °C
was continued for 4 h, and the mixture was then poured into
ice-water (100 mL) and extracted with Et2O (4 × 30 mL). The
combined organic extracts were washed with saturated aque-
(24) The reported [R]25 ) -141.2° (c 0.16, H2O) (see reference 2b);
D
our material had [R]25 ) -139.82° (c 1.67, H2O).
D
(25) Clive, D. L. J .; Wickens, P. L.; Sgarbi, P. W. M. J . Org. Chem.
1996, 61, 7426-7437.
(26) J ackson, R. F. W.; Wishart, N.; Wood, A.; J ames, K.; Wythes,
M. J . J . Org. Chem. 1992, 57, 3397-3404.

Synthesis of Angiotensin-Converting Enzyme Inhibitors
J . Org. Chem., Vol. 64, No. 5, 1999 1451
reaction flask was transferred to a cold bath at 0 °C. Stirring
was continued for 14 h. The resulting solution was acidified
with hydrochloric acid (2 M) and shaken with water (25 mL)
and EtOAc (25 mL). The resulting white precipitate was
filtered off, using a Whatman no. 2 filter paper, and the
aqueous filtrate was extracted with EtOAc (2 × 15 mL), and
filtered, as before, after each extraction to remove a small
amount of precipitate. The combined organic extracts were
washed with water (30 mL), filtered (as above), washed with
saturated aqueous NaCl (30 mL), dried (MgSO4), and evapo-
rated. Flash chromatography of the residue over silica gel (2
× 30 cm), using 15:85 EtOAc-hexanes, gave (-)-16 (0.7631
overlap), 29.46 (t′), 29.59 (t′), 30.93 (t′), 52.05 (d′), 67.28 (t′),
72.43 (d′), 77.19 (s′), 85.31 (s′), 116.16 (t′), 128.37 (d′), 128.54
(d′), 128.61 (d′), 135.14 (s′), 136.50 (d′), 170.83 (s′), 171.02 (s′),
175.03 (s′); exact mass m/z calcd for C21H23NO5 369.1576, found
369.1565.
Compound (+)-17b had: [R]25D ) 0.63° (c 3.98, EtOH); FTIR
(CHCl3 cast) 3350, 3297, 1790, 1740, 1678 cm-1 1H NMR
;
(CDCl3, 400 MHz) δ 1.76-1.87 (m, 1 H), 1.95-2.14 [m, 4 H,
including a t at δ 2.08 (J ) 2.6 Hz)], 2.30-2.52 (m, 3 H), 2.58-
2.71 (m, 1 H), 2.76-2.89 (m, 2 H), 4.61 (dt, J ) 8.4, 4.7 Hz, 1
H), 4.97-5.05 (m, 2 H), 5.10 and 5.19 (AB q, ∆νAB ) 35.9 Hz,
J ) 12.2 Hz, 2 H), 5.74 (ddt, J ) 16.6, 9.7, 6.6 Hz, 1 H), 6.93
(d, J ) 8.4 Hz, 1 H), 7.27-7.40 (m, 5 H); 13C NMR (CDCl3,
75.5 MHz) δ 28.09 (t′), 28.43 (t′), 29.52 (t′), 29.96 (t′), 30.77
(t′), 51.96 (d′), 67.28 (t′), 72.24 (d′), 77.40 (s′), 85.40 (s′), 116.27
(t′), 128.35 (d′), 128.54 (d′), 128.63 (d′), 135.13 (s′), 136.44 (d′),
170.66 (s′), 171.13 (s′), 175.05 (s′); exact mass m/z calcd for
g, 84%) as a pure (1H NMR), colorless oil: [R]25 ) -18.90° (c
D
3.10, EtOH); FTIR (CHCl3 cast) 3366, 1741, 1716 cm-1
;
1H
NMR (CDCl3, 400 MHz) δ 1.44 (s, 9 H), 1.67-1.78 (m, 1 H),
1.86-1.98 (m, 1 H), 2.00-2.17 (m, 2 H), 4.32-4.42 (m, 1 H),
4.95-5.05 (m, 2 H), 5.09 (d, J ) 7.7 Hz, 1 H), 5.14 and 5.20
(AB q, ∆νAB ) 26.0 Hz, J ) 12.4 Hz, 2 H), 5.76 (ddt, J ) 17.1,
10.3, 6.6 Hz, 1 H), 7.29-7.40 (m, 5 H); 13C NMR (CDCl3, 75.5
MHz) δ 28.31 (q′), 29.43 (t′), 31.93 (t′), 53.12 (d′), 66.98 (t′),
79.85 (s′), 115.68 (t′), 128.26 (d′), 128.40 (d′), 128.58 (d′), 135.43
(s′), 136.94 (d′), 155.30 (s′), 172.61 (s′); exact mass (HR
electrospray) m/z calcd for C18H25NNaO4 (M + Na) 342.1681,
found 342.1681.
C
21H23NO5 369.1576, found 369.1574.
5-Oxo-N-[[tetr a h yd r o-5-oxo-2-(2-p r op yn yl)-2-fu r a n yl]-
ca r b on yl]-L-n or va lin e p h en ylm et h yl est er [(-)-18a ].
Freshly distilled CH2Cl2 (15 mL) was added to (-)-17a (0.5333
g, 1.44 mmol) contained in a three-necked flask closed by a
stopper and fitted with a condenser (not attached to a water
supply) closed by a drying tube packed with Drierite, and an
ozone-oxygen inlet. The resulting solution was stirred and
cooled (-78 °C), and ozone was then bubbled through the
solution until all of the starting material had been consumed
(ca. 8 min, TLC control, silica gel, 50:50 EtOAc-hexanes). The
solution was purged with oxygen for 10 min, and then Ph3P
(0.7672 g, 2.93 mmol) was added. The cooling bath was
removed and stirring was continued for 4 h, by which time
the mixture had warmed to room temperature. Evaporation
of the solvent and flash chromatography of the residue over
silica gel (3 × 20 cm), using 90:10 EtOAc-hexanes, gave (-)-
18a (0.5199 g, 97%) as an oil, containing trace amounts of what
we take to be the corresponding E and Z enol tautomers (1H
P h en ylm eth yl (S)-2-Am in o-5-h exen oa te [(+)-10]. CF3-
CO2H (5 mL) was added over ca. 5 min to a stirred and cooled
(ice-water bath) solution of (-)-16 (0.2871 g, 0.90 mmol) in
freshly distilled CH2Cl2 (5 mL). The ice-water bath was
removed and stirring was continued until all the starting
material had been consumed (ca. 30 min, TLC control, silica
gel, 15:85 EtOAc-hexanes), by which time the mixture had
warmed to room temperature. The solution was evaporated,
and the residue was dissolved in EtOAc (10 mL), washed with
saturated aqueous NaHCO3 (2 × 10 mL), water (2 × 10 mL),
and saturated aqueous NaCl (10 mL), dried (MgSO4), and
evaporated (<0.1 mmHg). Flash chromatography of the residue
over silica gel (1.5 × 20 cm), using 90:10 EtOAc-hexanes, gave
NMR): [R]25 ) -12.32° (c 3.79, CHCl3); FTIR (CHCl3 cast)
D
(+)-10 (0.1912 g, 97%) as a pure (1H NMR), colorless oil: [R]25
3356, 3288, 2834, 2731, 1789, 1740, 1677 cm-1
;
1H NMR
D
) 7.94° (c 5.19, EtOH); FTIR (CHCl3 cast) 3383, 3318, 1734
(CDCl3, 400 MHz) δ 1.96-2.16 [m, 2 H, including a t at δ 2.03
(J ) 2.7 Hz)], 2.24 (ddt, J ) 13.9, 6.8, 4.9 Hz, 1 H), 2.39-2.62
(m, 5 H), 2.67-2.91 [m, 3 H, including an apparent dd at δ
2.74 (A of an ABX system, J ) 17.2, 2.7 Hz), and an apparent
dd at δ 2.78 (B of an ABX system, J ) 17.2, 2.7 Hz)], 4.55 (dt,
J ) 8.0, 5.2 Hz, 1 H), 5.13 and 5.18 (AB q, ∆νAB ) 18.4 Hz, J
) 12.1 Hz, 2 H), 7.20 (d, J ) 8.0 Hz, 1 H), 7.28-7.40 (m, 5 H),
9.68 (s, 1 H); 13C NMR (CDCl3, 75.5 MHz) δ 24.02 (t′), 28.29
(t′), 28.44 (t′), 29.43 (t′), 39.84 (t′), 51.91 (d′), 67.49 (t′), 72.42
(d′), 77.21 (s′), 85.23 (s′), 128.50 (d′), 128.56 (d′), 128.64 (d′),
134.99 (s′), 170.50 (s′), 171.09 (s′), 175.16 (s′), 200.49 (d′); exact
mass m/z calcd for C20H21NO6 371.1369, found 371.1371.
P h en ylm eth yl (2S)-2,3-Dih yd r o-1-[[tetr a h yd r o-5-oxo-
2-(2-p r op yn yl)-2-fu r a n yl]ca r bon yl]-1H-p yr r ole-2-ca r box-
yla te [(-)-19a ]. BaO (0.3412 g, 2.23 mmol) was tipped into a
solution of (-)-18a (0.3865 g, 1.04 mmol) in dry CH2Cl2 (10
cm-1 1H NMR (CDCl3, 400 MHz) δ 1.58-1.72 [m, 3 H,
;
including a s (NH2) at δ 1.68], 1.78-1.89 (m, 1 H), 2.05-2.20
(m, 2 H), 3.48 (dd, J ) 7.7, 5.3 Hz, 1 H), 4.93-5.06 [m, 2 H,
including a dd at δ 5.01 (J ) 17.2, 1.6 Hz)], 5.12 and 5.14 (AB
q, ∆νAB ) 7.9 Hz, J ) 12.3 Hz, 2 H), 5.76 (ddt, J ) 17.2, 10.3,
6.6 Hz, 1 H), 7.26-7.40 (m, 5 H); 13C NMR (CDCl3, 75.5 MHz)
δ 29.77 (t′), 34.03 (t′), 53.84 (d′), 66.58 (t′), 115.33 (t′), 128.23
(d′), 128.31 (d′), 128.55 (d′), 135.72 (s′), 137.47 (d′), 175.95 (s′);
exact mass m/z calcd for C13H17NO2 219.1259, found 219.1262.
P h en ylm eth yl (2S)-2-[[[Tetr a h yd r o-5-oxo-2-(2-p r op y-
n yl)-2-fu r an yl]car bon yl]am in o]-5-h exen oate [(-)-17a, (+)-
17b]. Lactone acid 9 (2.8235 g, 16.79 mmol), 1-hydroxyben-
zotriazole(2.2733g,16.82mmol),and 1-(3-(dimethylamino)propyl)-
3-ethylcarbodiimide hydrochloride (3.2349 g, 16.87 mmol) were
added in that order to a stirred solution of (+)-10 (3.3419 g,
15.24 mmol) in a mixture of dry CH2Cl2 (112 mL) and freshly
distilled, dry (stored over 4 Å molecular sieves) DMF (29 mL)
(Ar atmosphere). Stirring at room temperature was continued
for 12 h, and the mixture was washed with water (2 × 60 mL)
and saturated aqueous NaCl (60 mL), dried (MgSO4), and
evaporated, to give a dark purple oil. Chromatography of the
oil over silica gel (4.5 × 30 cm), repeated three times, using
3.5:3.5:4 CH2Cl2-Et2O-hexanes, gave the faster-eluting di-
astereomer [(-)-17a ] (2.7374 g, 49%) as a pure (1H NMR),
colorless oil, and the slower-eluting diastereomer [(+)-17b]
(2.6112 g, 46%), also as a pure (1H NMR), colorless oil.
mL), contained in
a round-bottomed flask fused onto a
condenser (Ar atmosphere), and the suspension was sonicated
(Branson, model B-12, 80 W; Ar atmosphere). Sonication was
stopped after 1 h, and P2O5 (0.3922 g, 2.76 mmol) was tipped
into the flask. The system was resealed with a septum and
flushed with Ar, and the mixture was sonicated until no more
aldehyde was being consumed (ca. 2.5 h, TLC control, silica
gel, 60:40 EtOAc-hexanes). The suspension was then centri-
fuged. Evaporation of the supernatant liquid and flash chro-
matography of the orange residue over silica gel (2 × 25 cm),
using 60:40 EtOAc-hexanes gave (-)-19a (0.2612 g, 71%, 93%
after correction for recovered starting material) as a pure (1H
NMR), colorless oil, which was a mixture of rotamers. The
material should be used within 24 h. Further elution with 90:
10 EtOAc-hexanes gave starting material (0.0878 g, 0.24
mmol) (1H NMR) as a light-yellow oil.
Compound (-)-17a had: [R]25 ) -21.01° (c 2.68, EtOH);
D
FTIR (CHCl3 cast) 3350, 3295, 1789, 1741, 1677 cm-1; 1H NMR
(CDCl3, 400 MHz) δ 1.78-1.88 (m, 1 H), 1.92-2.10 (m, 4 H),
2.43-2.59 (m, 3 H), 2.69-2.84 [m, 3 H, including an apparent
dd at δ 2.76 (A of an ABX system, J ) 17.3, 2.6 Hz), and an
apparent dd at δ 2.80 (B of an ABX system, J ) 17.3, 2.6 Hz)],
4.57 (dt, J ) 8.1, 5.0 Hz, 1 H), 4.96-5.03 (m, 2 H), 5.12 and
5.19 (AB q, ∆νAB ) 24.8 Hz, J ) 12.2 Hz, 2 H), 5.72 (ddt, J )
17.8, 9.7, 6.4 Hz, 1 H), 6.98 (d, J ) 8.1 Hz, 1 H), 7.28-7.40
(m, 5 H); 13C NMR (CDCl3, 75.5 MHz) δ 28.38 (t′, two signals
Compound (-)-19a had: [R]25 ) -83.09° (c 5.56, CHCl3);
D
FTIR (CHCl3 cast) 3283, 1792, 1745, 1642 cm-1
;
1H NMR
(CDCl3, 400 MHz) δ 2.07-2.12 [m, 1 H, including a t at δ 2.10
(J ) 2.6 Hz)], 2.36-2.78 [m, 5 H, including a dd at δ 2.64 (J
) 17.4, 2.7 Hz), and a ddd at δ 2.72 (J ) 17.4, 10.2, 6.4 Hz)],

1452 J . Org. Chem., Vol. 64, No. 5, 1999
Clive et al.
2.79-3.24 (m, 3 H, including an apparent dd at δ 2.87 (A of
an ABX system, J ) 12.6, 2.6 Hz), and an apparent dd at δ
2.94 (B of an ABX system, J ) 12.6, 2.6 Hz), and a ddt at δ
3.19 (J ) 17.1, 11.2, 2.5 Hz)], 4.86-5.42 [m, 4 H, including a
dd at δ 4.89 (J ) 11.5, 4.2 Hz), an AB q at δ 5.07 and 5.20
(∆νAB ) 52.7 Hz, J ) 12.0 Hz), overlapping an AB q at δ 5.11
and 5.23 (∆νAB ) 47.4 Hz, J ) 12.2 Hz), overlapping a dd at δ
5.25 (J ) 7.0, 2.7 Hz), a dd at δ 5.28 (J ) 4.6, 2.6 Hz), and a
dd at δ 5.39 (J ) 11.1, 2.7 Hz)], 7.05-7.23 [m, 1 H, including
a ddd at δ 7.07 (J ) 4.4, 2.8, 1.6 Hz), and a dt at δ 7.21 (J )
4.4, 2.2 Hz)], 7.31-7.42 (m, 5 H); 13C NMR (CDCl3, 75.5 MHz)
δ 27.72 (t′), 28.07 (t′), 28.59 (t′), 29.34 (t′), 30.67 (t′), 31.21 (t′),
32.17 (t′), 36.39 (t′), 59.32 (d′), 59.71 (d′), 67.17 (t′), 67.86 (t′),
72.18 (d′), 72.79 (d′), 86.13 (s′), 86.28 (s′), 109.94 (d′), 111.33
(d′), 128.34 (d′), 128.50 (d′), 128.54 (d′), 128.63 (d′), 128.68 (d′),
129.44 (d′), 130.97 (d′), 135.16 (s′), 135.32 (s′), 165.94 (s′),
167.83 (s′), 170.22 (s′), 172.39 (s′), 175.14 (s′), 175.35 (s′), one
pair of d′ in the aromatic region overlap; the internal acetylene
carbon signals overlap the solvent signals; exact mass m/z
calcd for C20H19NO5 353.1263, found 353.1261.
If the aldehyde solution is not pretreated with BaO for ca.
1 h before addition of P2O5, the overall yield is <50%. In the
absence of BaO, the P2O5 turns into a reddish-brown gummy
material after sonication in CH2Cl2 for 1 h or more; when BaO
is present, the P2O5 stays as a fine powdered suspension. If
additional P2O5 is added to the system during the course of
the reaction in an attempt to improve the conversion, an
unidentified byproduct is formed, and the yield of (-)-19a is
reduced.
(ca. 10 min, TLC control, silica gel, 60:40 EtOAc-hexanes).
The solution was purged with oxygen for 10 min, and then
Ph3P (0.8892 g, 3.39 mmol) was added. The cooling bath was
removed and stirring was continued for 1.5 h, by which time
the mixture had warmed to room temperature. Evaporation
(<0.1 mmHg) of the solvent gave a light-yellow solid; the
ketonic product (22a ) could not be separated chromatographi-
cally from Ph3PO, and so the crude mixture was used directly.
Dry Et3N (1.0 mL, 7.17 mmol) was added to a stirred
solution of the above crude ozonolysis product in dry THF (10
mL) (Ar atmosphere). Stirring was continued at 60 °C (oil bath)
for 1.5 h, and the mixture was then cooled and evaporated.
Flash chromatography of the light-yellow oily residue over
silica gel (2 × 20 cm), using 80:20:5 EtOAc-hexanes-AcOH,
gave (-)-23 (0.5746 g, 95%) as pure (1H NMR), light-yellow
crystals: mp 159-161 °C; [R]25 ) -158.64° (c 2.28, MeOH);
D
FTIR (CHCl3 cast) 3450-2400, 1744, 1543 cm-1; 1H NMR (CD3-
OD, 400 MHz) δ 2.26 (ddt, J ) 13.3, 8.7, 3.3 Hz, 1 H), 2.48-
2.63 (m, 3 H), 2.68-2.83 (eight-line m, 2 H), 2.95-3.16 [m, 2
H, including a tt at δ 3.03 (J ) 17.2, 8.7 Hz), overlapping a
ddd at δ 3.11 (J ) 17.2, 9.5, 3.4 Hz)], 5.13-5.24 [m, 3H,
including a dd at δ 5.15 (J ) 9.7, 3.1 Hz), overlapping an AB
q at δ 5.17 and 5.21 (∆νAB ) 13.6 Hz, J ) 12.3 Hz)], 7.24 (s, 1
H), 7.27-7.38 (m, 5 H); 13C NMR (CD3OD, 50.3 MHz) δ 26.90
(t′), 27.48 (t′), 28.11 (t′), 33.75 (t′), 63.93 (d′), 68.34 (t′), 129.21
(d′), 129.40 (d′), 129.62 (d′), 129.80 (s′), 133.99 (s′), 134.92 (d′),
137.00 (s′), 137.12 (s′), 160.37 (s′), 171.47 (s′), 176.74 (s′); exact
mass m/z calcd for C19H19NO6 357.1213, found 357.1208.
(S)-3-Ca r boxy-1,2,3,5-tetr a h yd r o-8-h yd r oxy-5-oxo-6-in -
d olizin ep r op a n oic Acid [(-)-1]. 10% Pd-C (ca. 25 mg) was
added to a stirred solution of (-)-23 (0.1614 g, 0.45 mmol) in
MeOH (20 mL). The reaction flask was flushed with hydrogen,
and the mixture was stirred under hydrogen (balloon) until
all of the starting material had been consumed (ca. 15 min,
TLC control, silica gel, 95:5 EtOAc-AcOH). The mixture was
filtered through a sintered glass frit (grade D) and evaporated.
Flash chromatography of the residue over reverse phase C-18
silica gel (Toronto Research Chemicals Inc., 10% capped with
TMS) (2 × 20 cm), using 90:10 water-MeCN, gave (-)-1
-196.55° (c 0.87, H2O), lit.2b [R]25D ) -199.5° (c 1.0, H2O); FTIR
(microscope) 3650-2300, 1721 cm-1; 1H NMR (D2O, 400 MHz)
δ 2.30 (ddt, J ) 13.5, 8.4, 3.8 Hz, 1 H), 2.48-2.73 (m, 5 H),
2.95-3.13 [m, 2 H, including a tt at δ 3.03 (J ) 17.3, 8.8 Hz),
overlapping a ddd at δ 3.08 (J ) 17.3, 9.5, 3.9 Hz)], 5.06 (dd,
J ) 9.9, 3.0 Hz, 1 H), 7.25 (s, 1 H); 13C NMR (D2O, 75.5 MHz)
δ 25.56 (t′), 26.63 (t′), 27.52 (t′), 33.06 (t′), 63.72 (d′), 128.26
(s′), 134.73 (d′), 135.28 (s′), 135.88 (s′), 159.74 (s′), 174.63 (s′),
178.03 (s′); exact mass m/z calcd for C12H13NO6 267.0743, found
267.0735.
P h en ylm eth yl (3′S)-Octa h yd r o-8′-m eth ylen e-5,5′-d iox-
ospir o[fu r an -2(3H),6′(5′H)-in dolizin e]-3′-car boxylate [(+)-
21a ]. A solution of AIBN (0.0281 g, 0.17 mmol, 7.79 mM) and
Bu3SnH (0.63 mL, 2.34 mmol, 0.11 M) in dry PhMe (22 mL)
was injected by syringe over ca. 2 min into a stirred and
refluxing solution (0.05 M with respect to the acetylene) of (-)-
19a (0.3911 g, 1.11 mmol) in PhMe (22 mL) (Ar atmosphere).
Stirring at reflux was continued for 3 h. A solution of AIBN
(0.0277 g, 0.17 mmol, 15.3 mM) and Bu3SnH (0.63 mL, 2.34
mmol, 0.21 M) in dry PhMe (11 mL) was then injected by
syringe over ca. 30 s. Stirring under reflux was continued until
no more starting material remained (ca. 1.5 h, TLC control,
silica gel, 40:60 EtOAc-hexanes), and the mixture was allowed
to cool to room temperature. Evaporation (<0.1 mmHg) of the
solvent gave the crude vinyl stannane, which was treated as
follows.
(0.1128 g, 93%) as a pure (1H NMR), white foam: [R]25
)
D
Dry CF3CO2H (1.5 mL) was injected rapidly into a stirred
solution of the above crude vinyl stannane in THF (15 mL)
(Ar atmosphere). After ca. 45 min no more vinyl stannane
could be detected (TLC control, silica gel, 40:60 EtOAc-
hexanes). Evaporation of the solvent and flash chromatography
of the residue over silica gel (2.5 × 25 cm), using 60:40 EtOAc-
hexanes, gave (+)-21a (0.3658 g, 93%) as a pure (1H NMR)
Meth yl (S)-1,2,3,5-Tetr ah ydr o-8-m eth oxy-5-oxo-3-[(ph en -
ylm eth oxy)ca r bon yl]-6-in d olizin ep r op a n oa te [(-)-24, n
) 1]. An excess of ethereal CH2N2 was added to a stirred and
cooled (ice-water bath) solution of (-)-23 (0.0342 g, 0.096
mmol) in MeOH (10 mL). The cooling bath was removed, and
the solution was stirred for 4 h, by which time all of the
starting material had been consumed (TLC control, silica gel,
80:20:5 EtOAc-hexanes-AcOH). Evaporation of the solvent
and flash chromatography of the residue over silica gel (1.5 ×
20 cm), using 80:20 EtOAc-hexanes, gave (-)-24 (n ) 1)
white solid: mp 135-138 °C; [R]25 ) 21.02° (c 2.84, CHCl3);
D
FTIR (CHCl3 cast) 1779, 1739, 1664 cm-1; H NMR (CDCl3,
1
400 MHz) δ 1.82-2.01 [m, 2 H, including a dt at δ 1.88 (J )
12.8, 9.9 Hz)], 2.03-2.12 (m, 1 H), 2.16-2.32 (m, 2 H), 2.48
(ddd, J ) 17.4, 9.6, 2.9 Hz, 1 H), 2.62-2.74 (m, 2 H), 2.79 (dt,
J ) 17.4, 9.8 Hz, 1 H), 3.01 (dd, J ) 14.8, 1.3 Hz, 1 H), 4.31-
4.38 (m, 1 H), 4.53 (d, J ) 8.8 Hz, 1 H), 5.00-5.04 (m, 1 H),
5.05-5.24 [m, 3 H, including an AB q at δ 5.07 and 5.21 (∆νAB
) 55.1 Hz, J ) 12.3 Hz)], 7.29-7.40 (m, 5 H); 13C NMR (CDCl3,
75.5 MHz) δ 28.36 (t′), 28.54 (t′), 29.04 (t′), 31.03 (t′), 42.19
(t′), 58.14 (d′), 61.80 (d′), 67.11 (t′), 83.04 (s′), 112.03 (t′), 128.19
(d′), 128.42 (d′), 128.60 (d′), 135.42 (s′), 139.07 (s′), 167.30 (s′),
171.01 (s′), 176.39 (s′); exact mass m/z calcd for C20H21NO5
355.1420, found 355.1417.
(S )-1,2,3,5-Te t r a h yd r o-8-h yd r oxy-5-oxo-3-[(p h e n yl-
m eth oxy)ca r bon yl]-6-in d olizin ep r op a n oic Acid [(-)-23]
fr om (+)-21a . Freshly distilled CH2Cl2 (10 mL) was added to
(+)-21a (0.6011 g, 1.69 mmol) contained in a three-necked flask
closed by a stopper and fitted with a condenser (not attached
to a water supply) closed by a drying tube packed with Drierite,
and an ozone-oxygen inlet. The resulting solution was stirred
and cooled (-78 °C), and ozone was then bubbled through the
solution until all of the starting material had been consumed
(0.0277 g, 75%) as a pure (1H NMR), light-yellow oil: [R]25
)
D
-165.45° (c 1.32, CHCl3); FTIR (CDCl3 cast) 1739, 1591 cm-1
;
1H NMR (CDCl3, 360 MHz) δ 2.20-2.31 (m, 1 H), 2.47 (ddd, J
) 18.8, 13.4, 9.4 Hz, 1 H), 2.26-2.70 (m, 2 H), 2.75-2.93 (m,
2 H), 3.03-3.12 (m, 2 H), 3.64 (s, 3 H), 3.72 (s, 3 H), 5.11-
5.30 [m, 3 H, including a dd at δ 5.15 (J ) 9.5, 3.3 Hz),
overlapping an AB q at δ 5.17 and 5.26 (∆νAB ) 34.6 Hz, J )
12.4 Hz)], 7.26 (s, 1 H), 7.27-7.40 (m, 5 H); 13C NMR (CDCl3,
50.3 MHz) δ 26.37 (t′), 26.54 (t′), 27.40 (t′), 32.47 (t′), 51.52
(q′), 58.96 (q′), 61.96 (d′), 67.38 (t′), 128.17 (d′), 128.41 (d′),
128.61 (d′), 129.31 (s′), 131.42 (d′), 135.30 (s′), 135.86 (s′),
137.51 (s′), 159.15 (s′), 170.00 (s′), 173.64 (s′); exact mass (HR
electrospray) m/z calcd for C21H23NNaO6 (M + Na) 408.1423,
found 408.1432.

Synthesis of Angiotensin-Converting Enzyme Inhibitors
J . Org. Chem., Vol. 64, No. 5, 1999 1453
HPLC analysis [Chiralcel OD-H (0.46 × 25 cm), 15% EtOH
in hexane] of the above material and comparison with the
corresponding racemic compound indicated an ee of 99.5%.
Meth yl (S)-1,2,3,5-Tetr a h yd r o-8-m eth oxy-3-(m eth oxy-
ca r bon yl)-5-oxo-6-in d olizin ep r op a n oa te [(-)-25 (n ) 1)].
An excess of ethereal CH2N2 was added to a stirred and cooled
(ice-water bath) solution of (-)-1 (0.0835 g, 0.31 mmol) in
MeOH (10 mL). The cooling bath was removed, and the
solution was stirred for 4 h, by which time all of the starting
material had been consumed (TLC control, silica gel, 95:5
EtOAc-MeOH). Evaporation of the solvent and flash chro-
matography of the residue over silica gel (2 × 20 cm), using
95:5 EtOAc-MeOH, gave (-)-25 (n ) 1) (0.0705 g, 74%) as a
EtOAc-hexanes, gave (-)-27a (0.0704 g, 83%) as a pure (1H
NMR), colorless oil, which was a mixture of rotamers: [R]25
D
) -57.77° (c 1.21, CHCl3); FTIR (CDCl3 cast) 3278, 1792, 1742,
1
1642 cm-1; H NMR (CDCl3, 400 MHz) δ 1.80-2.16 (m, 4 H,
including a t at δ 2.09 (J ) 2.4 Hz), and a t at δ 2.14 (J ) 2.6
Hz)], 2.24-2.79 (m, 5 H), 2.88-3.10 (m, 2 H), 4.98-5.54 (m, 4
H, including a dd at δ 5.02 (J ) 8.1, 5.5 Hz), and a br s at δ
5.52)], 7.11-7.22 [m, 1 H, including a d at δ 7.15 (J ) 8.6 Hz),
and a d at δ 7.20 (J ) 8.6 Hz)], 7.27-7.42 (m, 5 H); 13C NMR
(CD2Cl2, 75.5 MHz) δ 19.06 (t′), 19.19 (t′), 23.70 (t′), 24.35 (t′),
27.76 (t′), 28.21 (t′), 28.73 (t′), 30.41 (t′), 30.70 (t′), 33.06 (t′),
54.20 (d′), 56.59 (d′), 67.43 (t′), 67.86 (t′), 72.47 (d′), 72.87 (d′),
77.12 (s′), 77.73 (s′), 86.64 (s′), 87.80 (s′), 108.64 (d′), 110.81
(d′), 124.58 (d′), 125.24 (d′), 128.46 (d′), 128.71 (d′), 128.89 (d′),
136.12 (s′), 167.93 (s′), 168.10 (s′), 170.23 (s′), 171.02 (s′), 174.95
(s′), 175.42 (s′), not all of the signals from the minor rotamer
were observed; exact mass (HR electrospray) m/z calcd for
pure (1H NMR), light-yellow oil: [R]25 ) -158.37° (c 4.54,
D
CHCl3); FTIR (CDCl3 cast) 1739, 1591 cm-1; 1H NMR (CDCl3,
400 MHz) δ 2.21-2.31 (m, 1 H), 2.47 (ddd, J ) 18.6, 13.4, 9.4
Hz, 1 H), 2.60-2.67 (m, 2 H), 2.71-2.90 (m, 2 H), 3.04-3.12
(m, 2 H), 3.61 (s, 3 H), 3.71 (s, 3 H), 3.76 (s, 3 H), 5.07 (dd, J
) 9.5, 3.4 Hz, 1 H), 7.24 (s, 1 H); 13C NMR (CDCl3, 100.6 MHz)
δ 26.30 (t′), 26.56 (t′), 27.39 (t′), 32.38 (t′), 51.46 (q′), 52.67 (q′),
58.88 (q′), 61.86 (d′), 129.19 (s′), 131.35 (d′), 135.87 (s′), 137.56
(s′), 158.99 (s′), 170.61 (s′), 173.55 (s′); exact mass (HR
electrospray) m/z calcd for C15H19NNaO6 (M + Na) 332.1110,
found 332.1112.
C
21H21NNaO5 (M + Na) 390.1317, found 390.1325.
P h en ylm eth yl (6′S)-Octa h yd r o-1′-m eth ylen e-4′,5-d iox-
osp ir o[fu r a n -2(3H ),3′(4′H )-[2H ]q u in olizin e]-6′-ca r b oxy-
la te [(+)-28a ]. A solution of AIBN (0.0083 g, 0.05 mmol, 7.49
mM) and Bu3SnH (0.19 mL, 0.71 mmol, 0.10 M) in dry PhMe
(6.75 mL) was injected by syringe over ca. 1 min into a stirred
and refluxing solution (0.05 M with respect to the acetylene)
of (-)-27a (0.1242 g, 0.34 mmol) in PhMe (6.75 mL) (Ar
atmosphere). Stirring at reflux was continued for 2 h, by which
time all of the starting material had been consumed (TLC
control, silica gel, 30:70 EtOAc-hexanes), and the mixture was
allowed to cool to room temperature. Evaporation (<0.1
mmHg) of the solvent gave the crude vinyl stannane, which
was treated as follows.
HPLC analysis [Chiralpak AS (0.46 × 25 cm), 10% EtOH
in hexane] of the above material and comparison with the
corresponding racemic compound (clear baseline resolution of
the enantiomers was not obtained) indicated an ee of 96.2%.
6-Oxo-N-[[tetr a h yd r o-5-oxo-2-(2-p r op yn yl)-2-fu r a n yl]-
ca r bon yl]-L-n or leu cin e P h en ylm eth yl Ester [(-)-26a ].
9-BBN (0.5 M in THF, 1.79 mL, 0.90 mmol) was added by
syringe pump over ca. 4 min to a stirred and cooled (ice-water
bath) solution of (-)-17a (0.2968 g, 0.80 mmol) in dry THF
(10 mL) (Ar atmosphere). The cooling bath was removed, and
stirring was continued for 1 h. The mixture was then trans-
ferred by cannula to a stirred and cooled (ice-water bath)
suspension of PCC (1.5620 g, 7.25 mmol) and crushed 4 Å
molecular sieves (ca. 500 mg) in dry CH2Cl2 (10 mL). Ad-
ditional dry CH2Cl2 (2 × 1 mL) was used as a rinse. The
resulting mixture was transferred to an oil bath and stirred
at reflux temperature for 1.5 h, allowed to cool to room
temperature, and filtered through a pad (2.5 × 5 cm) of silica
gel, using EtOAc as a rinse. Evaporation of the solvent and
flash chromatography of the residue over silica gel (2 × 20
cm), using 80:20 EtOAc-hexanes, gave (-)-26a (0.2539 g, 83%)
Dry CF3CO2H (0.5 mL) was injected rapidly into a stirred
solution of the above crude vinyl stannane in THF (5 mL) (Ar
atmosphere). After ca. 50 min no more vinyl stannane could
be detected (TLC control, silica gel, 30:70 EtOAc-hexanes).
Evaporation of the solvent and flash chromatography of the
residue over silica gel (1.5 × 25 cm), using 50:50 EtOAc-
hexanes, gave (+)-28a (0.0924 g, 74%) as a pure (1H NMR),
colorless oil: [R]25D ) 0.32° (c 1.58, CHCl3); FTIR (CHCl3 cast)
1
1785, 1743, 1653 cm-1; H NMR (CDCl3, 400 MHz) δ 1.60-
1.82 (m, 3 H), 1.85-2.14 (m, 4 H), 2.39-2.51 (m, 2 H), 2.69 (d,
J ) 13.4 Hz, 1 H), 2.80 (dt, J ) 17.8, 10.6 Hz, 1 H), 2.94 (d, J
) 13.4 Hz, 1 H), 4.09 (d, J ) 11.8 Hz, 1 H), 4.37 (dd, J ) 6.1,
5.0 Hz, 1 H), 5.05-5.25 [m, 4 H, including a d at δ 5.08 (J )
0.6 Hz), a d at δ 5.11 (J ) 1.2 Hz), and an AB q at δ 5.09 and
5.22 (∆νAB ) 54.6 Hz, J ) 12.2 Hz)], 7.28-7.39 (m, 5 H); 13C
NMR (CDCl3, 75.5 MHz) δ 18.23 (t′), 24.17 (t′), 27.69 (t′), 28.52
(t′), 32.38 (t′), 41.74 (t′), 56.31 (d′), 57.16 (d′), 67.02 (t′), 81.85
(s′), 113.83 (t′), 128.35 (d′, two signals overlap), 128.58 (d′),
135.70 (s′), 138.57 (s′), 169.51 (s′), 170.92 (s′), 176.39 (s′); exact
mass (HR electrospray) m/z calcd for C21H23NNaO5 (M + Na)
392.1474, found 392.1466.
as a pure (1H NMR), colorless oil: [R]25 ) -4.36° (c 1.56,
D
CHCl3); FTIR (CDCl3 cast) 3360, 3288, 2850, 2750, 1788, 1739,
1
1675 cm-1; H NMR (CDCl3, 400 MHz) δ 1.48-1.66 (m, 2 H),
1.67-1.78 (m, 1 H), 1.79-1.94 (m, 1 H), 2.03 (t, J ) 2.0 Hz, 1
H), 2.34-2.63 (m, 5 H), 2.65-2.83 [m, 3 H, including an
apparent dd at δ 2.74 (A of an ABX system, J ) 17.3, 2.6 Hz),
and an apparent dd at δ 2.79 (B of an ABX system, J ) 17.3,
2.6 Hz)], 4.53 (dt, J ) 8.1, 5.1 Hz, 1 H), 5.11 and 5.17 (AB q,
∆νAB ) 22.4 Hz, J ) 12.1 Hz, 2 H), 7.11 (d, J ) 8.1 Hz, 1 H),
7.26-7.39 (m, 5 H), 9.66 (s, 1 H); 13C NMR (CDCl3, 50.3 MHz)
δ 17.66 (t′), 28.29 (t′, two signals overlap), 29.51 (t′), 30.99 (t′),
42.75 (t′), 52.04 (d′), 67.27 (t′), 72.37 (d′), 77.22 (s′), 85.26 (s′),
128.36 (d′), 128.57 (d′, two signals overlap), 135.03 (s′), 170.80
(s′), 170.94 (s′), 175.21 (s′), 201.35 (d′); exact mass (HR
electrospray) m/z calcd for C21H23NNaO6 (M + Na) 408.1423,
found 408.1431.
(S )-1,3,4,6-Te t r a h yd r o-9-h yd r oxy-6-oxo-4-[(p h e n yl-
m eth oxy)car bon yl]-2H-qu in olizin e-7-pr opan oic Acid [(-)-
29] fr om (+)-28a . Freshly distilled CH2Cl2 (10 mL) was added
to (+)-28a (0.0809 g, 0.22 mmol) contained in a three-necked
flask closed by a stopper and fitted with a condenser (not
attached to a water supply) closed by a drying tube packed
with Drierite, and an ozone-oxygen inlet. The resulting solu-
tion was stirred and cooled (-78 °C), and ozone was then
bubbled through the solution until all of the starting material
had been consumed (ca. 7 min, TLC control, silica gel, 50:50
EtOAc-hexanes). The solution was purged with oxygen for
10 min, and then Ph3P (0.1198 g, 0.46 mmol) was added. The
cooling bath was removed, and stirring was continued for 1.5
h, by which time the mixture had warmed to room tempera-
ture. Evaporation (<0.1 mmHg) of the solvent gave a light-
yellow solid; the ketonic product could not be separated
chromatographically from Ph3PO, and so the crude mixture
was used directly.
P h en ylm eth yl (2S)-1,2,3,4-Tetr a h yd r o-1-[[tetr a h yd r o-
5-oxo-2-(2-p r op yn yl)-2-fu r a n yl]ca r bon yl]-2-p yr id in eca r -
boxyla te [(-)-27a ]. BaO (0.1742 g, 1.14 mmol) was tipped
into a solution of (-)-26a (0.0879 g, 0.23 mmol) in dry CH2Cl2
(5 mL), contained in a round-bottomed flask fused onto a
condenser (Ar atmosphere), and the suspension was sonicated
(Branson, model B-12, 80 W; Ar atmosphere). Sonication was
stopped after 1 h, and P2O5 (0.1633 g, 1.15 mmol) was tipped
into the flask. The system was resealed with a septum and
flushed with Ar, and the mixture was sonicated until no more
aldehyde remained (ca. 1 h, TLC control, silica gel, 50:50
EtOAc-hexanes). The suspension was then centrifuged. Evapo-
ration of the supernatant liquid and flash chromatography of
the orange residue over silica gel (1.5 × 20 cm), using 40:60
Dry Et3N (1.0 mL, 7.17 mmol) was added to a stirred
solution of the above crude ozonolysis product in dry THF (10
mL) (Ar atmosphere). Stirring was continued at 60 °C (oil bath)
for 1.5 h, and the mixture was then cooled and evaporated.
Flash chromatography of the light-yellow oily residue over

1454 J . Org. Chem., Vol. 64, No. 5, 1999
Clive et al.
silica gel (1.5 × 15 cm), using 80:20:5 EtOAc-hexanes-AcOH,
gave (-)-29 (0.0781 g, 95%) as a pure (1H NMR), light-yellow
175.85 (s′), 178.26 (s′); exact mass (HR electrospray) m/z calcd
for C13H16NO6 (M + H) 282.0978, found 282.0976.
oil: [R]25 ) -131.61° (c 1.18, MeOH); FTIR (CHCl3 cast)
D
3450-2400, 1743, 1538 cm-1; H NMR (CD3OD, 360 MHz) δ
1
Ack n ow led gm en t. We thank the Natural Sciences
and Engineering Research Council of Canada, and the
Merck Frosst Therapeutic Research Centre for financial
support. D.M.C. holds an Alberta Heritage Foundation
for Medical Research Fellowship. We thank Dr. Deˆnis
Pires de Lima (supported by CNPq, Brazil), A. Wu, and
I. Kong for assistance. A.W. and I.K. held AHFMR
Summer Undergraduate Studentships. We also thank
Dr. P. L. Beaulieu and D. Tung of Bio-Me´ga Boehringer
Ingelheim (Montreal) for measuring the optical purity
of our synthetic samples.
1.46-1.62 (m, 1 H), 1.69-1.81 (m, 1 H), 2.00-2.13 (m, 1 H),
2.21-2.33 (m, 1 H), 2.49-2.59 (m, 2 H), 2.60-2.82 (m, 3 H),
2.89 (dt, J ) 18.3, 4.5 Hz, 1 H), 5.09-5.24 [m, 3 H, including
an AB q at δ 5.13 and 5.20 (∆νAB ) 26.4 Hz, J ) 12.3 Hz)],
7.24 (s, 1 H), 7.25-7.37 (m, 5 H); 13C NMR (CD3OD, 75.5 MHz)
δ 16.72 (t′), 23.81 (t′), 26.58 (t′), 27.29 (t′), 33.52 (t′), 57.12 (d′),
68.17 (t′), 127.72 (s′), 129.18 (d′), 129.31 (d′), 129.54 (d′), 130.74
(s′), 132.99 (d′), 137.11 (s′), 138.25 (s′), 161.86 (s′), 172.36 (s′,
two signals overlap); exact mass (HR electrospray) m/z calcd
for C20H21KNO6 (M + K) 410.1006, found 410.1001.
(S)-4-Ca r boxy-1,3,4,6-tetr a h yd r o-9-h yd r oxy-6-oxo-2H-
qu in olizin e-7-p r op a n oic Acid [(-)-2]. Pd-C (10%) (ca. 25
mg) was added to a stirred solution of (-)-29 (0.0565 g, 0.152
mmol) in MeOH (5 mL). The reaction flask was flushed with
hydrogen, and the mixture was stirred under hydrogen (bal-
loon) until all of the starting material had been consumed (ca.
20 min, TLC control, silica gel, 80:20:5 EtOAc-hexanes-
AcOH). The mixture was filtered through a sintered glass frit
(grade D) and evaporated. Flash chromatography of the
residue over reverse phase C-18 silica gel (Toronto Research
Chemicals Inc., 10% capped with TMS) (1 × 20 cm), using 90:
10 water-MeCN, gave (-)-2 (0.0411 g, 96%) as a pure (1H
Su p p or tin g In for m a tion Ava ila ble: NMR spectra for all
compounds and experimental procedures for 12, (-)-18b, (-)-
19b, (-)-21b, (-)-23 [from (-)-21b], (+)-20a , (-)-20b, (+)-21a
(as part of two-step procedure), (-)-21b (as part of two-step
procedure), (-)-26b, (-)-27b, (-)-28b, (-)-29 [from (-)-28b],
(-)-24 (n ) 2), (-)-25 (n ) 2), (+)-28a (two-step procedure),
(-)-28b (two-step procedure), (()-14, (()-16, (()-10, (()-17a ,b,
(()-18a , (()-18b, (()-19a , (()-19b, (()-20a , (()-20b, (()-21a ,
(()-21b, (()-21a (one-step procedure), (()-21b (one-step pro-
cedure), (()-23 [from (()-21a ], (()-23 [from (()-21b], (()-1,
(()-24 (n ) 1), (()-25 (n ) 1), (()-26a , (()-26b, (()-27a , (()-
27b, (()-28a (two-step procedure), (()-28b (two-step proce-
dure), (()-28a (one-step procedure), (()-28b (one-step proce-
dure), (()-29 [from (()-28a ], (()-29 [from (()-28b], (()-2, (()-
24 (n ) 2), (()-25 (n ) 2). This material is available free of
charge via the Internet at http://pubs.acs.org.
NMR), white foam: [R]25D ) -139.82° (c 1.67, H2O), lit.2b [R]25
D
) -141.2° (c 0.16, H2O); FTIR (CHCl3-MeOH cast) 3525-
2375, 1723 cm-1; 1H NMR (D2O, 400 MHz) δ 1.58-1.72 (m, 1
H), 1.78-1.90 (m, 1 H), 2.08-2.20 (m, 1 H), 2.27-2.38 (m, 1
H), 2.59-2.82 (m, 5 H), 2.91 (dt, J ) 18.2, 4.9 Hz, 1 H), 5.11
(dd, J ) 6.3, 4.0 Hz, 1 H), 7.35 (s, 1 H); 13C NMR (D2O, 50.3
MHz) δ 15.83 (t′), 23.43 (t′), 25.79 (t′), 26.17 (t′), 33.07 (t′), 57.30
(d′), 126.53 (s′), 132.35 (s′), 133.12 (d′), 137.43 (s′), 161.50 (s′),
J O9822941