Anodic amide oxidation/olefin metathesis strategies: Developing a unified approach to the synthesis of bicyclic lactam peptidomimetics

Article abstract of DOI:10.1016/S0040-4020(00)00856-5

In connection with efforts to build constrained peptidomimetics for the endocrine hormone TRH, a general strategy for the construction of bicyclic lactam peptide building blocks has been developed. This strategy used an anodic amide oxidation to selective

Full text of DOI:10.1016/S0040-4020(00)00856-5

TETRAHEDRON
Pergamon
Tetrahedron 56 (2000) 10113±10125
Anodic Amide Oxidation/Ole®n Metathesis Strategies:
Developing a Uni®ed Approach to the Synthesis of Bicyclic
Lactam Peptidomimetics
*
Laura M. Beal, Bin Liu, Wenhua Chu and Kevin D. Moeller
Department of Chemistry, Washington University, St. Louis, MO 63130, USA
Received 7 December 1999; accepted 7 February 2000
AbstractÐIn connection with efforts to build constrained peptidomimetics for the endocrine hormone TRH, a general strategy for the
construction of bicyclic lactam peptide building blocks has been developed. This strategy used an anodic amide oxidation to selectively
functionalize proline and then an ole®n metathesis to build the desired lactam constraint. The route described provides a single approach for
synthesizing both fully functionalized TRH analogs having seven- and eight-membered ring lactam constraints, as well as six- and seven-
membered ring lactam analogs without the sidechain on the central amino acid. q 2000 Elsevier Science Ltd. All rights reserved.
Introduction
the result of the analog being too rigid, and analogs like 4
which were designed in order to explore whether the partial
agonist behavior of 2a was the result of the substituent on
the central amino acid not being properly positioned. We
report here that both families of analogs (3 and 4) can be
readily synthesized by taking advantage of an anodic amide
oxidation±ole®n metathesis strategy for constructing
bicyclic lactams.
As part of a general effort to develop new strategies for
constructing conformationally restricted peptidomi-
metics,1,2 we recently synthesized the fully constrained
thyroliberin (TRH) analogs 2a and 2b.3,4 These analogs
were synthesized in connection with efforts to determine
whether the conformation of TRH represented by 1 was a
viable proposal for the TRH conformation that binds the
endocrine receptor TRH-R.5±7 Interestingly, analogs 2a
and 2b were not full agonists of the TRH-R receptor.
While neither a bridge added to the pyroglutamate region
of the molecule (bridge 1)8 nor a bridge added to the central
region of the molecule (bridge 2)7 had stopped previous
analogs from being full agonists, the fully constrained
analog 2a having both bridges reached only 47% of maxi-
mal potency at high concentrations. Analog 2b reached only
25% of maximal potency. This observation was particularly
surprising for 2a which displayed an af®nity for TRH-R that
was a factor of 4 better than the unrestricted, and fully
potent, 2-cyclohexylAla-TRH.7 Typically, TRH analogs
have displayed an almost linear relationship between
af®nity for TRH-R and potency. Why was this not true for
2a? (Scheme 1)
From the start, it was clear that the synthetic methodology
developed for the synthesis of 2a would not allow for the
synthesis of analogs like 3 and 4. Efforts to functionalize
seven-membered ring lactams (7 and 9/XˆH) in analogy to
the conversion of 5 to 2a were not successful, and the anodic
oxidation of proline derivatives was not compatible with the
presence of either an oxygen or a nitrogen substituent on the
carbon alpha to the amide carbonyl (XˆOR or NR2).1a In
addition, intramolecular cyclization reactions of iminium
ions like 8 (XˆH) did not afford seven-membered ring
lactams whenever R1 was an electron donating group.
Instead, the cyclization reactions triggered subsequent
In order to address this question, a number of new analogs
were proposed. These suggestions included more ¯exible
analogs like 3 (Scheme 2) which were designed in order
to explore whether the partial agonist behavior of 2a was
Keywords: anodic amide oxidation; TRH-R; bicyclic lactum peptidomi-
metics.
* Corresponding author. Tel.: 1314-935-4270; fax: 1314-935-4481;
e-mail: moeller@wuchem.wust1.edu
Scheme 1.
0040±4020/00/$ - see front matter q 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020(00)00856-5

10114
L. M. Beal et al. / Tetrahedron 56 (2000) 10113±10125
Scheme 2.
rearrangement reactions that afforded six-membered ring
lactam products.1a,1d Even if these problems could be
resolved, the number of analogs that could be synthesized
was limited by the fact that the bridgehead stereochemistry
resulting from the intramolecular cyclization reaction was
set by the stereochemistry of the groups on the N-terminal
side chain.9 A sidechain with (S,S) stereochemistry would
give rise to a bicyclic lactam with an R-con®guration at the
bridgehead carbon. A cyclization substrate having a side-
chain with (R,R)-stereochemistry would give rise to a
bicyclic lactam having an S-con®guration at the bridgehead
carbon. This would be a problem since TRH analogs having
S-bridgehead stereochemistry are known not to bind or
activate TRH-R well.3,7
have been reported.9a,10 However, the application of this
chemistry to the TRH problem did not appear practical.
First, the existing routes required that separate strategies
be taken for making either the six- or the seven-membered
ring analogs. Second, a new route would be still required for
analogs like 3, because the existing routes to 9 did not
incorporate substituents on the carbon beta to the amide
carbonyl.
With this in mind, we began to investigate the route outlined
in Scheme 3. In this approach, both families of analogs (3
and 4) would be made from a common vinyl-substituted
proline derivative 13.11 This plan offered several important
advantages. First, the problems encountered with function-
alization of seven-membered ring lactams and oxidations of
substituted prolines would be avoided by functionalizing
the proline ring early. Second, the stereochemistry at the
Alternative routes to the six- and seven-membered ring
building blocks (9) needed for the synthesis of 4a and 4b
Scheme 3.

L. M. Beal et al. / Tetrahedron 56 (2000) 10113±10125
10115
geous to use a mixed Lewis acid catalyst (1:1 AlCl3/
SnCl4) for the synthesis of 16. Using the mixed catalyst
system improved the consistency of the yield obtained for
the addition, and avoided the occasional generation of
increased amounts of the trans-5-vinylproline product.
On a ®nal note, both the synthesis of 13 and the synthesis of
15 did use an anodic amide oxidation to generate the
Scheme 4.
Scheme 5. Reagents: (a) TBAF, THF, 2208C, 95%. (b) Lindlar cat., H2, EtOAc, 95%. (c) TMSI, CHCl3, 508C, 70%. (d) i. O3, MeOH, CH2Cl2, 2788C;
ii. NaBH4, 2788C to rt, 96%. (e) (n-Bu)3P, o-NO2(C6H4)SeCN, THF, 85%. (f) i. MCPBA, CH2Cl2, 2708C, 30 min; ii. Me2S, Et3N, 2708C to rt, 96%. (g) TFA,
CH2Cl2, 65%.
bridgehead carbon would be set during the synthesis of 13.
If a mixture of isomers were to arise, then the cis- and trans-
isomers of 13 could be separated and a sizable amount of the
desired isomer obtained with a minimum amount of impact
on the synthesis. Finally, the synthesis called for the use of
an ole®n metathesis reaction to complete the formation of
the lactam ring.12 Since the ole®n metathesis reaction was
known to be compatible with amino acid functionality, all of
the stereocenters in a desired analog would be set in the
starting materials (XˆNHBOC) before construction of
the bicyclic lactam ring skeleton. This would allow for the
synthesis of analogs having a variety of stereochemical
patterns without altering the bridgehead stereochemistry.
methoxylated amide 14.14 The use of the anodic amide
oxidation ensured that the pathway could take advantage a
variety of amino acid based starting materials,14c and hence
left open the possibility for using the same chemistry to
develop peptidomimetics with variations in the C-terminal
end of the building block.1c
Building more ¯exible analogsÐan initial approach
At the start, it appeared that the fastest way to constrain the
central region of TRH with a seven-membered ring lactam
would be to take advantage of the route suggested in
Scheme 6. In this route, the metathesis reaction would be
used to make 18, and then 18 converted into a TRH analog
using the chemistry developed for 2a. In this sequence, an
a-hydroxyamide was oxidized and the resulting ketone
converted into a enamine by treatment with ammonia in
methanol. The enamine was then coupled to a pyrogluta-
mate derivative and the C-terminal end of the building block
converted into the primary amide.
Synthesis of the vinylproline starting materials
When this work was started, syntheses for both the trans-
and the cis-5-vinylsubstituted prolines had been reported.
Several items concerning these syntheses warrant comment
here. The pure trans-5-vinylproline derivative 15 can be
conveniently made by the addition of a vinyl cuprates to
the N-acyliminium ion generated by BF3´Et2O treatment
of a methoxylated amide 14 (Scheme 4).13 While this pro-
cedure was extremely useful, it did not allow access to the
cis-5-vinylproline derivative 13 that was required for the
synthesis of TRH analogs. In this case, three options were
available.11 Of these options, the two outlined in Scheme 5
have proven to be the most effective for generating multi-
gram quantities of 13. Of these two, we prefer the route
developed by Manfree because it avoids the use of selenium.
While for the most part our use of the Manfree procedure
has followed the literature,11a we have found it advanta-
In practice, the synthesis of alcohol 18 proceeded nicely
Scheme 6.

10116
L. M. Beal et al. / Tetrahedron 56 (2000) 10113±10125
Scheme 7.
the oxidation of the alpha keto amide nor the conversion of
the resulting ketone into an enamine could be accomplished
in an acceptable yield.
Building more ¯exible analogsÐa second more versatile
approach
Figure 1.
The dif®culty associated with adding in the nitrogen
functionality to 18 and the compatibility of the ole®n
metathesis catalyst with peptide based substrates,12
prompted us to examine the utility of a synthetic route
that would directly generate the cyclic products with the
N-terminal amine group already in place (Scheme 8). This
route was extremely attractive because molecular modeling
suggested that a more ¯exible seven-membered ring lactam
analog would better mimic 2a if the double bond in 3c was
removed. For example, consider the overlapped structures
illustrated in Scheme 9.16 In each of the three drawings, a
potential new analog (3a±3c) has been overlayed with 2a in
order to illustrate how well the pharmacophoric groups in
the new analog overlap with the conformation of the same
groups in 2a. The distance between the pharmacophores for
the best possible ®t are indicated by the arrows. From
structure I, it was clear that the conformation of the
pharmacophoric groups in 3c would differ signi®cantly
from the conformation found in 2a. This was a problem
because analog 2a bound TRH-R well. The question being
asked was whether or not a more ¯exible analog having the
same conformation would be fully potent. An ideal analog
would add ¯exibility to the constraints without altering the
conformation of the pharmacophoric groups. To this end,
both analogs 3a and 3b appeared to be superior to 3c (struc-
tures II and III). Of these two, analog 3b looked to be best
(Scheme 7). The cis-vinylproline 13 was coupled to
2-benzyloxy-3-phenyl-4-pentenoic acid 19, and the ole®n
metathesis reaction accomplished with the use of the
ruthenium based catalyst developed by Grubbs and
coworkers.12 The metathesis reaction led to an 81% isolated
yield of the bicyclic lactam derivative 21. Hydrogenation of
the double bond in 21 afforded the desired seven membered
building block 18. The yield of the hydrogenation reaction
was not optimized due to subsequent problems with the
synthesis (see below).
In addition to 18, two other building blocks were
assembled (Fig. 1). The eight-membered ring analog 22
was made in an identical fashion to 18 except for the use
of 5-allylproline 17 in place of 13. The ole®n metathesis
reaction led to the 72% yield of the cyclized product. The
synthesis of 23 started from the trans-vinylproline
derivative 15 (R1ˆR2ˆH). Every other aspect of the
synthesis was identical.
While this synthetic strategy allowed for assembly of the
proposed alcohol building blocks, it was not capable of
generating the actual TRH analogs. In the case of 18, neither
Scheme 8.

L. M. Beal et al. / Tetrahedron 56 (2000) 10113±10125
10117
Scheme 9.
because it would allow for a close overlap between the
pharmacophoric groups while better maintaining the overall
conformation of the peptide backbone found in 2a. In either
case, it appeared that our ability to address the biological
questions of concern would bene®t greatly from the addition
of stereocenters into the lactam ring.
In both cases, the building blocks were readily converted
into TRH analogs (Scheme 12). As initial targets, the
partially constrained TRH analogs 30a and 30b were
selected so that the new constraints could be independently
evaluated for how they effected the binding and potency of
the analog. The synthesis involved cleavage of the t-Boc
protecting group, coupling of the resulting amine to pyro-
glutamic acid using standard EDCI conditions, and then
conversion of the methyl ester to the primary amide using
ammonia in methanol. The ®nal aminolysis step was
completed by dissolving the methyl ester in a saturated
solution of ammonia in methanol, placing the resulting
mixture into a sealed tube, and then heating the reaction
to 658C for two days. In this way, an 77% isolated yield
of 30a was obtained from the corresponding methyl ester.
An 85% yield of 30b was obtained.
The forward synthesis of the bicyclic building blocks
required for making 3a and 3b (as well as partially
constrained analogs) began with the construction of the
known acids 27a and 27b (Scheme 10).15 The synthesis of
building block 29a was then started by coupling the cis-5-
vinylproline methyl ester 13 with acid 27a (Scheme 11).
The resulting ole®n metathesis substrate was then cyclized
to afford a 65% yield of the seven-membered ring lactam
24a along with 24% of the recovered starting material.
Hydrogenation of the double bond in 24a afforded building
block 29a. The synthesis of 29b proceeded in an identical
fashion starting from acid 27b.
Interestingly, preliminary biological studies on 30a and 30b
con®rmed the suggestion made by molecular modeling.
Scheme 10.
Scheme 11.

10118
L. M. Beal et al. / Tetrahedron 56 (2000) 10113±10125
Scheme 12.
Analog 30b had an af®nity for TRH-R that was about 35
times greater than that measured for 30a.17 In fact, the
constraint used in 30b allowed for binding to TRH-R at a
level competitive with the constraint used in the central
region of 2a;7 an observation that suggested the use of the
constraint in 30b for making a fully constrained TRH analog
that is more ¯exible than 2a.
Initially, the coupling of vinylglycine to the vinyl substi-
tuted proline derivative was problematic. When the
coupling reaction was performed using HOBt conditions
with either EDCI or DCC as an activating group three
products were obtained in a 1:1:1 ratio (Scheme 14). Only
one of the products was the desired 12a. The other two
products resulted from either migration of the double
bond or epimerization of the vinylglycine a-carbon. The
competitive racemization and ole®n migration reactions
could not be avoided. Even when short reaction times
were used, side reactions dominated these procedures.
Fortunately, many of the problems with the coupling
reaction could be avoided with the use of a mixed anhydride
as the activated ester. When the anhydride derived from the
treatment of acid 32 with isobutylchloroformate and pyri-
dine was used to effect the coupling reaction, the desired
product was obtained in a 58% isolated yield (Scheme 15).
Alternatively, IIDQ (2-isobutoxy-1-isobutoxycarbonyl-1,2-
dihyroquinoline) could be used to make the mixed
anhydride. While the yield was not excellent using either
method, neither the product from migration nor the product
from epimerization (12b) was observed, and the pure ole®n
metathesis substrate could be obtained. Interestingly, the
yield of the coupling reaction was not dependent on the
steric size of the amine. Removing the vinyl group from
the proline ring did not improve the yield of the reaction.
The coupling of 34 to vinylglycine afforded a 52% yield of
product. The use of the more hindered (and normally more
dif®cult to couple) trans vinyl proline derivative 35 led to a
55% isolated yield of product. As earlier, no product from
Analogs without the sidechain on the central amino acid
As mentioned above, we were also interested in synthesiz-
ing TRH analogs that did not have the sidechain on the
central amino acid. These analogs were desired as probes
for determining if the presence of the sidechain was essen-
tial for the analog to be a full agonist for TRH-R. Because of
the ease with which analogs 30a and 30b were made, we
hoped that a nearly identical approach would be useful for
the synthesis of either analogs like 4 or partially constrained
analogs like 31 (Scheme 13). Two concerns immediately
arose. First, the route required the coupling of vinylglycine
to a secondary amine that was branched on both alpha
carbons.18 The double bond in vinylglycine was known to
undergo migration reactions,19 and it was not clear that the
coupling of vinylglycine to a sterically hindered amine
could be accomplished without such complications. Second,
the yield of ole®n metathesis reactions using vinylglycine
derived ole®ns can be dependent on the nature of the
surrounding substituents.20 Would the highly substituted
proline derivative needed for the synthesis of analogs like
4 or 31 interfere with the metathesis reaction?
Scheme 13.
Scheme 14.

L. M. Beal et al. / Tetrahedron 56 (2000) 10113±10125
10119
analog 31a was then obtained with the use of ammonia in
methanol.
A very similar approach was used to generate the
corresponding seven-membered ring analog 31b. In this
case, the synthesis began with the coupling of vinylglycine
to the 5-allylproline derivative 17 (Scheme 17).11b
Ole®n metathesis to generate the seven-membered
ring lactam, hydrogenation of the double bond, depro-
tection of the Boc group, coupling to the N-hydroxysuc-
cinimide ester of pyroglutamic acid, and conversion of
the methyl ester into a primary amide afforded TRH
analog 31b.
Scheme 15.
migration of the double bond or racemization was obtained
in either of these reactions.
For the synthesis of a seven-membered ring analog, an alter-
native strategy could be taken that avoided the use of vinyl-
glycine. For example, the synthesis of a TRH analog having
the opposite lactam stereochemistry (31c) was accom-
plished by starting the synthesis with a coupling reaction
between the vinyl substituted proline derivative 13 and the
commercially available vinylalanine derivative 41
(Scheme 18). Following this coupling reaction, the synthesis
of the TRH analog proceeded in a fashion directly
analogous to the earlier syntheses.
Once the ole®n metathesis substrate was in hand, the lactam
constraint was readily completed (Scheme 16). Substrate
12a was cyclized in order to afford an 85% isolated yield
of the bicyclic lactam 37. The double bond in the lactam
ring was immediately hydrogenated in order to avoid
migration reactions. The N-terminal Boc protecting group
on 37 was then removed using HCl in ethyl acetate, and the
resulting amine coupled to pyroglutamic acid by using the
N-hydroxysuccinimide activated ester. The desired TRH
Scheme 16.
Scheme 17. Reagents: (a) 3M HCl, EtOAc. (b) Et3N, pyroglutamate N-hydroxysuccinimide ester, CH2Cl2, RT, 70% (two steps). (c) NH3, MeOH, 08C±RT,
48 hr, 73%.

10120
L. M. Beal et al. / Tetrahedron 56 (2000) 10113±10125
Scheme 18. Reagents: (a) 3 M HCl, EtOAc. (b) Et3N, pyroglutamate N-hydroxysuccinimide ester, CH2Cl2, rt, 71% (two steps). (c) NH3, MeOH, 08C-rt, 48 h,
40%.
Preliminary biological testing of TRH analogs 31a±c
showed that the analogs did not have a high af®nity for
TRH-R relative to 2a.17 However, both 31a and 31b were
full agonists of the receptor, an observation that indicated
that the presence of the sidechain was not a requirement for
full agonist behavior.
and bis(trimethylsilyl)acetylene (5.00 g, 29.4 mmol) in
CH2Cl2 (38 mL) was added 1 M solution of SnCl4 in
CH2Cl2 (19 mL) at 2208C, then AlCl3 (2.7 g) was added.
The mixture was warmed to rt and stirred at room tempera-
ture for 36 h. Water (10 mL) was added, then neutralized
with saturated Na2CO3 to PHˆ10, the white solid was
®ltered out. The ®ltrate was extracted with ethyl acetate
(150 mL), and washed with saturated NaCl, dried over
Na2SO4. After evaporation of the solvent, the crude product
was puri®ed with hexane±ether (4:1) to afford 1.57 g (38%)
of the trans compound and 1.82 g (44%) of the cis
compound 16 which was converted into 13.
Conclusions
The ability of electrochemical amide oxidations to selec-
tively functionalize amino acid derivatives has made readily
available both vinyl- and allyl-substituted proline deriva-
tives. In this work, these building blocks were coupled to
vinylalanine and vinylglycine derivatives, and then an ole®n
metathesis reaction used to complete the synthesis of a
family of bicyclic lactam derivatives. In turn, the bicyclic
lactams were used to construct a series of previously
unavailable analogs for the tripeptidic hormone TRH.
For the purpose of providing a general synthetic protocols,
the synthesis of analog 30b is detailed below. For the
synthesis of analogs 30a and 31a±c characterization data
are provided along with alternative synthetic procedures
where noted. Details concerning the syntheses of building
blocks 18, 22, and 23 have not been included since they
proved ineffective for constructing the desired TRH
analogs. However, it should be noted that the synthetic
procedures followed for the construction of these building
blocks were identical to those reported for the synthesis of
30b.
It is clear that the use of a sequential amide oxidation/ole®n
metathesis strategy can provide a rapid approach to bicyclic
lactam peptidomimetics. Because of the availability of
chiral, vinyl-substituted amino acid derivatives, the route
not only offers rapid access to the desired ring skeletons,
but also access to derivatives that vary the stereochemistry
of the centers within the lactam constraint. As in the
synthesis of analogs 30a and 30b, the development of this
method allows for the construction of peptidomimetics
that ®x the pharmacophoric groups of a peptide segment
in the same orientation while altering the shape of the
scaffold that holds them in place. Such a change can
dramatically effect the biological activity of the analog,
and therefore the methodology developed here should
prove valuable for the synthesis of constrained peptidomi-
metics in the future.
N-[(2R)-t-Butoxycarbonylamino-(3R)-phenyl-4-pente-
noyl]-(5R)-vinylproline methyl ester (28b). A solution of
13 (274 mg, 0.94 mmol) and 27b (63 mg, 0.47 mmol) in
CH2Cl2 was added HOBt (152 mg, 1.13 mmol), then a
1 M solution of DCC in CH2Cl2 (1 mL) was added at
2208C. The reaction mixture was warmed to rt and stirred
at rt overnight. Ethyl acetate (50 mL) was added to the
reaction mixture, washed with 10% citric acid (15 mL£2),
5% NaHCO3 (15 mL£2) and saturated NaCl (20 mL),
dried over Na2SO4. After evaporation of the solvent, the
crude product was puri®ed with hexane±ether (1:1) to
give 28b 151 mg (75%) as a colorless oil. 1H NMR
(300 MHz/CDCl3) d: 7.35±7.20 (m, 5H), 6.36±6.24 (m,
1H), 5.82±5.71 (m, 1H), 5.44±5.10 (m, 4H), 4.77±4.72
(m, 1H), 4.26 (t, Jˆ8.1 Hz, 1H), 3.83±3.72 (m, 2H), 3.76
(s, 3H), 3.56 (m, 1H), 2.02±1.98 (m, 1H), 1.86±1.57 (m,
3H), 1.37 (s, 9H); 13C NMR (75 MHz/CDCl3) d 172.2,
169.8, 154.4, 139.8, 136.5, 135.5, 128.5, 128.4, 127.1,
Experimental
The synthesis of 13 was conducted as reported in the litera-
ture with one exception. The conversion of 14 to 16 was
modi®ed as follows: A solution of 14 (3.21 g, 14.8 mmol)

L. M. Beal et al. / Tetrahedron 56 (2000) 10113±10125
10121
118.2, 117.2, 79.3, 60.2, 59.8, 55.5, 53.7, 52.2, 31.6, 29.7,
28.2, 27.0; IR (neat/NaCl) 3430, 3310, 3060, 2979, 2932,
1750, 1710, 1643, 1498, 1437, 1366, 1265, 1174, 996, 921,
736, 702 cm21; LRMS (FAB) m/e (rel. int.) 429 (M11, 23),
373 (19), 329 (65), 211 (17), 156 (100); HRMS (FAB) m/e
calcd for C24H32N2O5 (M11) 249.2389, found 429.2395.
76 mg (75%) 6 as a white solid. 1H NMR (300 MHz/CDCl3)
d: 7.29±7.19 (m, 5H), 7.07 (d, Jˆ9.9 Hz, 1H), 6.61 (s, 1H),
5.11 (t, Jˆ9.9 Hz, 1H), 4.58 (dd, Jˆ8.1, 4.2 Hz, 1H), 4.12
(m, 1H), 3.83 (dd, Jˆ9.3, 3.9 Hz, 1H), 3.77 (s, 3H), 2.90 (m,
1H), 2.32 (m, 1H), 2.15±2.04 (m, 6H), 1.93±1.82 (m, 4H),
1.61±1.52 (m, 1H), 1.19±1.14 (m, 2H); 13C NMR (75 MHz/
CDCl3) d: 178.8, 172.3, 170.9, 170.5, 142.5, 128.1, 127.9,
126.7, 60.6, 58.8, 56.8, 52.4, 46.7, 37.1, 34.3, 32.9, 28.4,
27.7, 25.3; IR (neat/NaCl) 3282, 3054, 2946, 1742, 1682,
(3R,4R,7R,10S)-1-Aza-3-t-butoxycarbonylamino-4-phenyl-
10-carbomethoxybicyclo[5.3.0]dec-5-ene (24b). A solu-
tion of 28b (118 mg, 0.276 mmol) in CH2Cl2 (55 mL) was
added bis(tricyclohexylphosphine)-benzylidine ruthenium
(IV) dichloride (45 mg) and the mixture was heated to re¯ux
for 24 h. After evaporation of the CH2Cl2 in vacuo, the
crude product was puri®ed with hexane±ether (1:2) to afford
68 mg (62%) 24b as a white solid along with 36 mg (30%)
of recovered 28b. 1H NMR (300 MHz/CDCl3) d: 7.32±7.21
(m, 5H), 5.83±5.71 (m, 2H), 5.28 (d, Jˆ9.9 Hz, 1H), 5.00
(dd, Jˆ11.7, 9.0 Hz, 1H), 4.85 (m, 1H), 4.67 (dd, Jˆ7.05,
4.2 Hz, 1H), 3.76 (s, 3H), 3.56 (m, 1H), 2.41 (m, 1H), 2.14
(m, 2H), 1.96 (m, 1H), 1.17 (s, 9H); 13C NMR (75 MHz/
CDCl3) d: 171.8, 170.0, 154.6, 140.3, 132.6, 129.0, 128.6,
128.0, 126.8, 79.1, 59.8, 55.6, 54.5, 52.4, 48.1, 33.0, 28.1,
27.7; IR (neat/NaCl) 3420, 3057, 2983, 2929, 1745, 1708,
1654, 1507, 1436, 1362, 1265, 1168, 1064, 1023, 862,
735 cm21; LRMS (FAB) m/e (rel. int.) 401 (M11, 8), 345
(11), 301 (100), 282 (60); HRMS (FAB) m/e calcd for
C22H29N2O5 (M11) 401.2076, found 401.2072.
1619, 1538, 1431, 1364, 1267, 1204, 1174, 735, 702 cm21
;
LRMS (FAB) m/e (rel. int.) 420 (M1Li, 100), 303 (8), 160
(22); HRMS (FAB) m/e calcd for C22H27N3O5Li (M1Li)
420.2111, found 420.2104.
(3R,4R,7R,10S)-1-Aza-3-pyroglutamoylamino-4-phenyl-
10-carboxyamidebicyclo[5.3.0]decane (30b). In a tube, a
solution of the methyl ester made above (65 mg,
0.157 mmol) in methanol (5 mL) was saturated with NH3
gas at 08C. The tube was sealed and heated to 658C for 2
days. After evaporation of the methanol in vacuo, the crude
product was puri®ed with ether±methanol (4:1) to afford
53 mg (85%) 30b as a white solid along with 8 mg of the
1
recovered starting material. H NMR (300 MHz/CDCl3) d:
7.34 (s, 1H), 7.29±7.15 (m, 6H), 6.94 (d, Jˆ9.3 Hz, 1H),
5.94 (s, 1H), 5.10 (t, Jˆ9.6 Hz, 1H), 4.46 (dd, Jˆ7.2 Hz,
Jˆ2.7 Hz, 1H), 4.09 (m, 1H), 3.84 (dd, Jˆ9.4, 3.3 Hz, 1H),
2.91±2.89 (m, 1H), 2.35±2.23 (m, 3H), 2.19±2.09 (m, 3H),
1.99±1.83 (m, 4H), 1.51±1.25 (m, 2H); 13C NMR (75 MHz/
CDCl3) d: 183.4, 179.2, 173.7, 171.3, 171.1, 142.4, 128.3,
127.7, 126.9, 61.7, 59.3, 56.7, 54.8, 46.3, 37.2, 34.3, 33.3,
28.5, 27.2, 25.3; IR (neat/NaCl) 3405, 3291, 3050, 2983,
2930, 1689, 1635, 1445, 1421, 1264, 896, 739 cm21; LRMS
(FAB) m/e (rel. int.) 399 (M11, 100), 354 (65), 326 (17),
288 (22), 154 (52); HRMS (FAB) m/e calcd for C21H27N4O4
(M11) 399.2032, found 399.2025.
(3R,4R,7R,10S)-1-Aza-3-t-butoxycarbonylamino-4-phenyl-
10-carbomethoxybicyclo[5.3.0]decane (29b). A solution
of 24b (124 mg, 0.31 mmol) and 5% Pd/C (62 mg) in
methanol (5 mL) was hydrogenated under hydrogen balloon
overnight. After removal of the catalyst by ®ltration and
evaporation of the methanol, the residue was puri®ed with
hexane±ether (1:2) to afford 115 mg (93%) 5 as a white
1
solid. H NMR (300 MHz/CDCl3) d: 7.29±7.18 (m, 5H),
5.17 (d, Jˆ9.3 Hz, 1H), 4.71 (t, Jˆ9.9 Hz, 1H), 4.65 (dd,
Jˆ7.95, 4.2 Hz, 1H), 4.10 (m, 1H), 3.78 (s, 3H), 2.81 (m,
1H), 2.31 (m, 1H), 2.18±2.07 (m, 4H), 1.91 (m, 2H), 1.80
(m, 1H), 1.21 (s, 9H); 13C NMR (75 MHz/CDCl3) d: 172.5,
171.1, 155.0, 142.4, 128.1, 127.8, 126.3, 79.0, 60.3, 58.6,
56.5, 52.3, 47.1, 37.1, 34.5, 32.9, 28.1, 27.8; IR (neat/NaCl)
3424, 2976, 2936, 1746, 1710, 1650, 1500, 1433, 1369,
1249, 1172, 1066, 872, 762, 736 cm21; LRMS (FAB) m/e
(rel. int.) 409 (M1Li, 90), 359 (10), 309 (100); HRMS
(FAB) m/e calcd for C22H30N2O5Li (M1Li) 409.2315,
found 409.2315.
N-[(2S)-t-Butoxycarbonylamino-(3S)-phenyl-4-pentenoyl]-
(5R)-vinylproline methyl ester (28a). The procedure
followed was the same as that used for the synthesis of
28b above except for the substitution of EDCI as the
coupling reagent in place of DCC. An 83% yield of the
coupled product was obtained. 1H NMR (300 MHz/
CDCl3) 7.32±7.18 (m, 5H), 6.06±5.94 (m, 2H), 5.63 (d,
Jˆ16.8 Hz, 1H), 5.27 (d, Jˆ10.2 Hz, 1H),5.10 (d,
Jˆ10.2 Hz, 1H), 5.01±4.81 (m, 4H), 4.53 (t, Jˆ6.9 Hz,
1H), 3.80 (t, Jˆ8.1 Hz, 1H), 3.75 (s, 3H), 2.22±2.15 (m,
2H), 2.04±1.99 (m, 1H), 1.89 (m, 1H), 1.24 (s, 9H); 13C
NMR (75 MHz/CDCl3) 172, 171, 154, 139, 138.8, 137,
130, 128.8, 128.4, 128.2, 126, 117.3, 116.9, 79.5, 61, 60,
53.8, 53.0, 52.8, 52.1, 32, 28, 27; IR (neat/NaCl) 3422,
3300, 3055, 2973, 1742, 1698, 1634, 1494, 1427, 1359,
1260, 1162, 1033, 1009, 914, 731 cm21; LRMS (FAB) m/
e (rel. int.) 451 (M1Na, 100), 429 (M11, 20), 373 (30), 329
(96), 297 (12); HRMS (FAB) m/e calcd for C24H33N2O5
(M11) 429.2389, found 429.2382.
(3R,4R,7R,10S)-1-Aza-3-pyroglutamoylamino-4-phenyl-
10-carbomethoxybicyclo[5.3.0]decane. A solution of 29b
(100 mg, 0.25 mmol) in ethyl acetate±HCl (37%) (3:1) was
stirred for 2 h at rt and then the solvent removed in vacuo.
Methanol (10 mL) was added and then the solvent again
removed in vacuo. The residue was dried under vacuum
for 12 h to afford a white solid. To this residue, CH2Cl2
(5 mL), triethylamine (51 mg, 0.50 mmol) and HOBt
(68 mg, 0.50 mmol) were added at 08C, and then a 1 M
solution of DCC in CH2Cl2 (0.63 mL) was added. The
mixture was stirred at rt overnight, ethyl acetate (50 mL)
was added, washed with 10% citric acid (25 mL£2), 5%
Na2CO3 (25 mL£2) and saturated NaCl (30 mL), dried
over Na2SO4. After evaporation of the solvent, the crude
product was puri®ed with ether±methanol (10:1) to afford
(3S,4S,7R,10S)-1-Aza-3-t-butoxycarbonylamino-4-phen-
yl-10-carbomethoxybicyclo[5.3.0]dec-5-ene (24a). The
ole®n metathesis reaction was accomplished using the
conditions described above for the synthesis of 24b in
order to form 65% yield of 24a along with 24% of the
1
recovered starting material. H NMR (300 MHz/CDCl3) d
7.39 (m, 2H), 7.32±7.20 (m, 3H), 5.94 (d, Jˆ12.0 Hz, 1H),

10122
L. M. Beal et al. / Tetrahedron 56 (2000) 10113±10125
5.75 (b, 1H), 5.21 (d, Jˆ5.4 Hz, 1H), 4.71±4.45 (2 bm, 2H),
4.45±4.2 (bm, 1H), 3.87±3.74 (bm, 1H), 3.69 (s, 1H), 2.4±
2.29 (bm, 1H), 2.17 (m, 1H), 2.55±1.96 (m, 2H), 1.45 (s,
9H); 13C NMR (75 MHz/CDCl3) d 172.3, 167.7, 154.7,
137.4, 129.4, 129.0, 128.2, 127.3, 80.1, 60.8, 55.2, 51.9,
46.9, 33.3, 29.5, 28.1, 27.0; IR (neat/NaCl) 2916,
0.128 mmol) was dissolved in 4 mL of THF followed by
the addition of pyridine (103 mL/0.128 mmol) at 08C and
treated with 166 mL (0.128 mmol) of isobutylchloro-
formate. After 15 min, 198 mg (0.128 mmol) of 13 in
1 mL of THF was added, and another 1 mL of THF was
used to rinse the glassware. The mixture was stirred and
allowed to warm to room temperature over 3 h. The solvent
was removed in vacuo and the residue was redissolved in
25 mL of CH2Cl2. The organic phase was washed with 5%
NaHCO3, 5% HCl, and brine, and then dried over MgSO4.
After the solvent was removed, the residue was ¯ash chro-
matographed through silica gel using 20% EtOAc/hexane as
eluant to afford 250 mg (58%) of 12a as clear oil. 1H NMR
(300 MHz/CDCl3) d 5.94±5.70 (m, 2H), 5.45±5.13 (m,
5H), 4.96 (m, 1H), 4.73 (br s, 1H), 4.54 (m, 1H), 3.68 (s,
3H), 2.18±2.07 (m, 2H), 1.99±1.78 (m, 2H), 1.37 (s, 9H);
13C NMR(75 MHz/CDCl3) 172.1,170.7,155.1, 138.1, 133.4,
117.8,116.8, 79.8,61.2,59.8, 53.4, 52.1, 32.5, 28.2, 27.1; IR
(NaCl) 2974, 2931, 1750, 1701, 1651, 1504, 1434, 1356,
1180 cm21; LRMS (FAB) m/e (rel. int.) 339 (M11, 100%);
HRMS(FAB) m/e calcd for C17H27N2O5 (M11) 339.1920;
found 339.1915.
2822,2747, 1866, 1674, 1643, 1584, 1429, 1329 cm21
;
LRMS (FAB) m/e: 401 (M11, 28), 359 (20), 345 (35),
301 (M112CO2C3H9, 100), 282 (35); HRMS (FAB) m/e
calcd for C22H29N2O5 (M11) 401.2076, found 401.2080.
(3S,4S,7R,10S)-1-Aza-3-t-butoxycarbonylamino-4-phenyl-
10-carbomethoxybicyclo[5.3.0]decane (29a). The pro-
cedure utilized to make 29a (89%) was identical to the
1
procedure described above for the synthesis of 29b. H
NMR (300 MHz/CDCl3) d 7.32±7.18 (m, 5H), 5.05 (b,
1H), 4.58 (d, Jˆ8.4 Hz, with ®ne splitting, 1H), 4.29 (bm,
2H), 3.74 (s, 1H), 3.50±3.33 (bm, 1H), 2.25±1.76 (m, 9H),
1.32 (s, 9H); 13C NMR (75 MHz/CDCl3), d 172.8, 169.8,
157.7, 142.9, 128.2, 126.5, 79.5, 61.3, 60.7, 57.0, 52.1, 43.6,
32.4, 30.7, 29.9, 28.2, 27.2; IR (neat/NaCl) 2872, 2845,
1669, 1630, 1583 cm21; LRMS (FAB) m/e: 403 (M11,
14), 347 (M256, 9), 303 (M2CO2C3H9, 100),
287(M2115, 13); HRMS (FAB) m/e calcd for
C22H31N2O5 (M11) 403.2233, found 403.2217.
(3S,6R,9S)-1-Aza-3-t-butoxycarbonylamino-9-carbo-
methoxybicyclo[4.3.0]non-4-ene. Substrate 12a was
cyclized in a fashion identical to the cyclization of 28b to
afford a 90% yield of the bicyclic lactam product. 1H NMR
(300 MHz/CDCl3) 5.99 (m, 2H) 5.54 (d, Jˆ3.9 Hz, 1H),
4.66 (br s, 1H), 4.44 (d, Jˆ8.7 Hz, 1H), 4.21 (dt,
Jˆ11.7 Hz, 5.8 Hz, 1H), 3.71 (s, 3H), 2.32±2.06 (m, 3H),
1.99±1.89 (m, 1H), 1.45 (s, 9H); 13C NMR (75 MHz/
CDCl3) 173.4, 167.8, 157.6, 131.2, 127.8, 81.4, 60.2, 59.4,
53.9, 52.7, 30.8, 30.4, 29.8; IR (NaCl/neat) 2976, 2877,
1733, 1717, 1663, 1506, 1436, 1367, 1165 cm21; LRMS
(FAB) m/e 311(100), 309(76), 307(87), 312(25); HRMS
(FAB) m/e calcd for C15H23N2O5 (M11) 311.1607; found
311.1604.
(3S,4S,7R,10S)-1-Aza-3-pyroglutamoylamino-4-phenyl-
10-carbomethoxybicyclo[5.3.0]decane. The deprotection
and subsequent coupling to pyroglutamic acid of 29a was
done using the succinimide ester of pyroglutamic as
described below for the synthesis of 38. Using this pro-
cedure a 70% yield of the desired product was obtained
1
over the two steps. H NMR (300 MHz/CDCl3) d 7.50 (d,
Jˆ7.3 Hz, 1H), 7.29±7.14 (m, 5H), 6.98 (s, 1H), 4.55 (dd,
Jˆ8.0, 3.4 Hz, 1H), 4.41 (bt, Jˆ6.8 Hz, 1H), 4.29 (dd,
Jˆ10.6, 7.7 Hz, 1H), 3.82 (dd, J1ˆ9.4, 3.8 Hz, 1H), 3.71,
(s, 3H), 3.71±3.6 (m, 1H), 2.29±1.74 (m, 11H with H2O),
1.51±1.39 (m, 1H); 13C NMR (75 MHz/CDCl3) d 178.9,
172.9, 172.5, 169.7, 143.5, 128.2, 128.1, 126.7, 61.8, 59.1,
57.1, 56.8, 52.2, 42.9, 32.3, 31.6, 30.8, 28.9, 27.2, 25.3; IR
(neat/NaCl) 3282, 3063, 2950, 1743,1683, 1626, 1547,
1438, 1368, 1272, 1203, 1177, 737 cm1; LRMS (EI) m/e:
413 (M, 8), 354 (M259, 16), 329 (M284, 33), 287 (37), 269
(37), 257 (M2156, 100); HRMS (FAB) m/e calcd for
C22H28N3O5 (M11) 414.2029, found 414.2025.
(3S,6R,9S)-1-Aza-3-t-butoxycarbonylamino-9-carbo-
methoxybicyclo[4.3.0]nonane (37). Product 37 was
synthesized in a fashion identical to the synthesis of 29b
1
described above (100%). H NMR (300 MHz/CDCl3) 5.47
(d, Jˆ4.5 Hz, 1H), 4,46 (d, Jˆ7.2 Hz, 1H), 4.13±4.06
(m, 1H), 3.69±3.62 (s1m, 4H), 2,45±2.35 (m, 1H),
2.21±1.97 (m, 4H), 1.73±1.57 (m, 3H), 1.39 (s, 9H);
13C NMR (75 MHz/CDCl3) 172.1, 169.1, 155.7, 79.5,
58.1, 56.5, 52.3, 50.0, 32.1, 29.0, 28.3, 27.1, 26.8; IR
(NaCl/neat) 3403, 2977, 1739, 1700, 1653, 1506, 1169,
1073 cm21; LRMS (FAB) m/e 313.4 (M11); HRMS
(FAB) m/e calcd for C15H25N2O5 (M11) 313.1763; found
313.1754.
(3S,4S,7R,10S)-1-Aza-3-pyroglutamoylamino-4-phenyl-
10-carboxyamidebicyclo[5.3.0]decane (30a). Compound
30a was made (77%) in a fashion identical to that described
for the synthesis of 30b above. 1H NMR (300 MHz/CD3OD)
d 7.18±7.04 (m, 5H), 4.40 (d, Jˆ10.4 Hz, 1H), 4.35 (d,
Jˆ7.4 Hz, 1H), 4.21 (m, 1H), 3.81±3.78 (m, 1H), 3.36 (t,
Jˆ10.6 Hz, 1H), 2.87±1.70 (m, 11H); 13C NMR (75 MHz/
CDCl3) d 179.9, 175.7, 172.5, 170.6, 143.2, 128.5, 128.1,
126.6, 63.2, 58.5, 57.0, 56.7, 43.4, 32.9, 30.9, 30.2, 29.3,
27.9, 25.1; IR (neat/NaCl) 3280, 2952, 1743, 1688, 1632,
1437 cm21; LRMS (EI) m/e: 398(M, 4), 354 (M2HNCO,
79), 314 (M2C4NHO, 22), 269 (65), 226 (81), 84(100);
HRMS (FAB) calcd for C21H27N4O4 (M11) 399.2032,
found 399.2032.
(3S,6R,9S)-1-Aza-3-pyroglutamoylamine-9-carbomethoxy-
bicyclo[4.3.0]nonane (38). The N-hydroxysuccinimide
ester of pyroglutamic acid was prepared by adding
10 mmol of DCC to a slurry solution of 10 mmol pyro-
gulatmic acid and 10 mmol N-hydroxysuccinimide in
EtOAc at 08C. The mixture was stirred at 08C for 3 h and
then placed in a refrigerator overnight. The urea byproduct
was then removed by ®ltration and the solvent evaporated in
vacuo. The crude solid was recrystalized from CH2Cl2 to
afford a white solid which was used in the reaction below.
1H NMR (300 MHz/CD2Cl2) 6.19 (s, 1H), 4.60±4.56 (m,
N-[(2S)-t-Butoxycarbonylamino-3-butenoyl]-(5R)-vinyl-
proline methyl ester (12a). Compound 32 (257 mg/

L. M. Beal et al. / Tetrahedron 56 (2000) 10113±10125
10123
1H), 2.84 (s, 4H), 2.70±2.60 (m, 1H), 2.50±2.30 (m, 3H).
mp 133±1358C.
5.36±5.32 (m, 1H), 4.51 (t, Jˆ6.0 Hz, 1H), 4,42±4.35 (m,
1H), 3.71 (s, 3H), 2.48±2.36 (m, 1H), 2.32±2.17 (m, 3H),
2.15±1.98 (m, 1H), 1.79±1.62 (m, 1H), 1.40 (s, 9H); 13C
NMR (75 MHz/CDCl3) 172.6, 169.2, 155.7,128.0, 127.7,
79.6, 59.9, 55.5, 52.3, 51.7, 34.8, 31.5, 28.3, 27.9; IR
The protected amine 37 (80 mg) was dissolved in 2 mL of
3 M HCl in EtOAc (1:3). The solution was stirred at room
temperature for 30 min before the solvent was removed
under reduced pressure. The remaining residue was
dissolved in 4 mL of CH2Cl2, and then 89.5 mL
(0.64 mmol) of triethylamine was added followed by
145 mg (0.64 mmol) of the active ester made above. The
mixture was stirred at room temperature overnight. The
solvent was then removed in vacuo, and the residue chro-
matographed through silica gel to afford 80 mg (80%) of 38.
1H NMR (300 MHz/CDCl3) 7.31(d, Jˆ6.3 Hz, 1H), 6.59 (s,
1H), 4.49±4.39 (m,2H), 4.17±4.13 (m, 1H), 3.72 (m1s,
4H), 2.53±2.04 (m, 8H), 1.78±1.62 (m, 4H); 13C NMR
(75 MHz/CDCl3) 178.7, 172.3, 172.0, 169.0, 58.3, 57.0,
56.7, 52.5, 48.7, 32.1, 29.3, 29.0, 27.0, 26.4, 25.9; IR
(CHCl3) 3282, 2952, 1735, 1700, 1684, 1652, 1559, 1540,
1436, 1457 cm21; LRMS (FAB) 324(30) 338(50) 307(58)
279(100) 289(70); HRMS (FAB) m/e calcd for C15H22N3O5
(M11) 324.1559; found 324.1557.
(neat/NaCl) 3259, 1674, 1644, 1583, 1373, 1116 cm21
;
LRMS (FAB) 325(M11, 33), 307(100), 289(54), 225(44);
HRMS (FAB) m/e calcd for C16H25N2O5 (M1H) 325.1763;
found 325.1755.
(3S,7R,10S)-1-Aza-3-t-butoxycarbonylamino-10-carbo-
methoxybicyclo[5.3.0]decane (40). The hydrogenation
reaction was done in a fashion identical to that described
above for the synthesis of 29b to afford a 90% yield of 40.
1H NMR (300 MHz/CDCl3) 5.84 (br d, Jˆ6.3 Hz, 1H), 4.61
(dd, Jˆ7.2, 4.5 Hz, 1H), 4.20 (dd, Jˆ10.8, 6.3 Hz, 1H), 3.81
(q, Jˆ8.1 Hz, 1H), 3.72 (s, 3H), 2.26±2.17 (m, 1H), 2.06±
1.95 (m, 4H), 1.82±1.55 (m, 5H), 1.40 (s, 9H); 13C NMR
(75 MHz/CDCl3) 172.5, 171.3, 155.2, 79.3, 60.2, 59.1,
54.3, 52.3, 34.2, 32.8, 31.7, 28.3, 27.6, 27.5; IR (neat/
NaCl) 3132, 1625, 1674, 1576 cm21; LRMS (FAB) m/e:
327(M11, 45), 227(77), 154(100), 136(67); HRMS (FAB)
m/e calcd for C16H27N2O5 (MH1) 327.1920, found 327.1924.
(3S,6R,9S)-1-Aza-3-pyroglutamoylamine-9-carboxy-
amidebicyclo[4.3.0]nonane (31a). Methyl ester 38 (85 mg)
was treated with 8 mL of methanolic ammonia solution
(saturated) and stirred at room temperature for 4 days.
Evaporation of the MeOH and NH3 followed by silica gel
chromatography (2:1:1 ether/methanol/dichloromethane)
afforded 47 mg (57%) of the desired 31a. 1H NMR
(300 MHz/CDCl3) 7.63±7.24 (d1s, Jˆ12.6 Hz, 2H), 7.12
(s, 1H), 6.32 (s, 1H), 4.48 (d, Jˆ7.5 Hz, 1H), 4.40 (d,
Jˆ8.1 Hz, 1H), 4.20±4.17 (m, 1H), 3.64±3.61 (m, 1H),
2.50±1.69 (m, 12H); 13C NMR (75 MHz/CDCl3) 179.3,
174.1, 172.7, 169.5, 59.75, 58.3, 57.2, 48.4, 32.0, 29.6,
28.7, 27.0, 26.2, 25.6; IR (neat/NaCl) 3415(br), 2955,
2879, 1683, 1652, 1559, 1267, 1118 cm21; LRMS (FAB)
309(35), 154(100); HRMS (FAB) m/e calcd for C14H21N4O4
(M11) 309.1563; found 309.1550.
(3S,7R,10S)-1-Aza-3-pyroglutamoylamine-10-carbo-
methoxybicyclo[5.3.0]decane. The synthesis was done in a
fashion identical to the preparation of 38 in order to obtain a
70% yield of the desired pyroglutamte coupled product. 1H
NMR (300 MHz/CDCl3) 7.56 (d, Jˆ6.6 Hz, 1H), 7.24 (s,
1H), 4.54 (dd, Jˆ9.0, 4.2 Hz, 1H), 4.45 (dd, Jˆ10.2, 6.9 Hz,
1H), 4,13 (dd, Jˆ8.4, 5.7 Hz, 1H), 3.84 (q, Jˆ8.1 Hz, 1H),
3.72 (s, 3H), 2.56±1.54 (m, 14H); 13C NMR (75 MHz/
CDCl3) 179.1, 172.2, 171.1, 170.9, 60.4, 59.2, 57.2, 53.0,
52.4, 34.1, 32.9, 30.9, 29.5, 27.5, 25.8; IR (neat/NaCl) 3502,
1669, 1615, 1584, 1144 cm21; HRMS (FAB) m/e calcd for
C16H24N3O5 (M1H) 338.1716, found 338.1722.
(3S,7R,10S)-1-Aza-3-pyroglutamoylamine-10-carboxy-
amidebicyclo[5.3.0]decane (31b). Primary amide 31b was
made using the same procedure used to construct 31a
(73%). 1H NMR (300 MHz/CDCl3) 7.64 (d, Jˆ6.9 Hz,
1H), 7.51 (s,1H), 6.96 (br s, 1H), 6.00 (br s, 1H), 4.59±
4.56 (m, 1H), 4.53±4.50 (m, 1H), 4.17±4.13 (m, 1H),
3.87±3.79 (m, 1H), 2.50±1.52 (m, 14H); 13C NMR
(75 MHz/CDCl3) 179.4, 173.8, 172.3, 171.5, 61.6, 59.7,
57.1, 53.1, 34.3, 33.2, 30.8, 29.4, 27.5, 26.9, 25.5; IR
(neat/NaCl) 3191(br), 3086(br), 1613, 1584, 1566, 1393,
1200 cm21; HRMS (FAB) m/e calcd for C15H23N4O4
(M1H) 323.1719, found 323.1709.
N-[(2S)-t-Butoxycarbonylamino-3-butenoyl]-(5R)-vinyl-
proline methyl ester (39). Compound 39 was synthesized
using the IIDQ coupling method. As in the synthesis of 12a
this route involved the formation of a isobutoxy based
mixed anhydride. To this end, 443 mg (2.62 mmol) of the
5R-allylproline methyl ester (17), 527 mg (2.62 mmol) of
vinyl glycine 32, and 794 mg (2.62 mmol) of IIDQ were
dissolved in 10 mL of THF. The reaction mixture was stir-
red at 238C for 12 h, at which time the solvent was removed
and the crude product was puri®ed by ¯ash chromatography
to afford 778 mg of the desired product (84%) as clear oil.
The NMR spectra for the amide was complex due to the
presence of several rotamers. For this reason, the material
was carried on to the next step before complete character-
ization. HRMS (FAB) m/e calcd for C18H29N2O5 353.2076,
found 353.2072.
N-[(2R)-t-Butoxycarbonylamino-4-pentenoyl]-(5R)-vinyl-
proline methyl ester (42). Compound 42 was obtained
(62%) from the coupling of 5R-vinyl proline methyl
ester 13 with d-N-Boc-allylglycine (41) using EDCI and
HOBt as coupling reagents. The procedure followed was
the same as that used for the synthesis of 28b above
except for the substitution of EDCI as the coupling
reagent in place of DCC. Due to the number of rotomers
produced, compound 42 was converted into the cyclic
derivative before full characterization. LRMS (FAB) m/e
(rel. int.) 359 (M1Li, 100), 303 (20), 259 (54), 156 (18);
HRMS (FAB) m/e calcd for C18H28N2O5Li (M1Li)
359.2158, found 359.2165.
(3S,7R,10S)-1-Aza-3-t-butoxycarbonylamino-10-carbo-
methoxybicyclo[5.3.0]dec-5-ene. Substrate 39 was
cyclized in a fashion identical to that described for the
cyclization of 28b to afford a 66% yield of the desired
product. 1H NMR (300 MHz/CDCl3) 5.71 (br d,
Jˆ6.6 Hz, 1H), 5.62±5.56 (m, 1H), 5.40±5.38 (m, 1H),

10124
L. M. Beal et al. / Tetrahedron 56 (2000) 10113±10125
(3R,7R,10S)-1-Aza-3-t-butoxycarbonylamino-10-carbo-
methoxybicyclo[5.3.0]dec-5-ene. The ole®n metathesis
originating from 42 was accomplished using the same
addition, we thank the Washington University High
Resolution NMR Facility, partially supported by NIH grants
RR02004, RR05018, and RR07155, and the Washington
University Mass Spectrometry Resource Center, partially
supported by NIH RR00954, for their assistance.
1
procedure used for the cyclization of 28b (82%). H NMR
(300 MHz/CDCl3) 5.73±5.68 (m, 2H), 5.28 (br s, 1H),
4.60±4.50 (m, 2H), 4.27 (m, 1H), 3.68 (s, 3H), 2.79±2.74
(m, 1H), 2.43±2.37 (m, 1H), 2.29±2.23 (m, 1H), 1.99±1.90
(m, 3H), 1.42 (s, 9H); 13C NMR (75 MHz/CDCl3) 172.2,
170.2, 155.0, 129.8, 124.4, 79.9, 60.7, 60.2, 55.3, 52.0,
32.9, 28.6, 28.1, 26.6; IR (neat/NaCl) 3293, 2976, 1741,
1704, 1643, 1505, 1435, 1167 cm21; LRMS (FAB) m/e
(rel. int.) 331 (M1Li, 100), 275 (40), 231 (80); HRMS
(FAB) m/e calcd for C16H25N2O5 (M1H) 325.1764, found
325.1752.
References
1. For selected examples see: (a) Li, W.; Hanau, C. E.; d'Avignon,
A.; Moeller, K. D. J. Org. Chem. 1995, 60, 8155. (b) Fobian, Y. M.;
Moeller, K. D. Methods in Molecular Medicine. Peptidomimetics
Protocols; Kazmierski, W. M., Ed.; Humana: Totowa, NJ, 1998;
Vol. 23, p 259. (c) Tong, Y.; Fobian, Y. M.; Wu, M.; Boyd, N. D.;
Moeller, K. D. Bioorg. Med. Chem. Lett. 1998, 8, 1679. (d) Chu,
W.; Moeller, K. D. Tetrahedron Lett. 1999, 40, 7939.
(3R,7R,10S)-1-Aza-3-t-butoxycarbonylamino-10-carbo-
methoxybicyclo[5.3.0]decane (43). The hydrogenation
reaction was done in a fashion identical to that described
above for the synthesis of 29b to afford a 95% yield of 43.
1H NMR 5.32 (br s, 1H), 4.50 (d, Jˆ7.8 Hz, 1H), 4.24
(q, Jˆ4.8 Hz, 1H), 3.76 (m, 1H), 3.69 (s, 3H), 2.08±1.65
(m, 10H), 1.40 (s, 9H); 13C NMR (75 MHz/CDCl3) 172.6,
170.6, 155.2, 79.5, 61.0, 58.3, 54.7, 51.9, 32.8, 32.1,
28.1, 27.4, 27.1, 22.3; IR (neat/NaCl) 3302, 2967, 1708,
1741, 1636, 1514, 1436, 1169 cm21; LRMS (FAB) m/e
(rel. int.) 333 (M1Li, 100), 277 (26), 233 (62); HRMS
(FAB) m/e calcd for C16H27N2O5 (M1H) 327.1921, found
327.1929.
2. For a recent review concerning the use of lactam based pepti-
domimetics see: (a) Hanessian, S.; McNaughton-Smith, G.;
Lombart, H.-G.; Lubell, W. D. Tetrahedron 1997, 53, 12789. For
more recent references see: (b) Polyak, F.; Lubell, W. D. J. Org.
Chem. 1998, 63, 5937 and (c) Curran, T. P.; Marcaurell, L. A.;
O'Sullivan, K. M. Org. Lett. 1999, 1, 1998, as well as references
therein.
3. Chu, W.; Perlman, J. H.; Gershengorn, M. C.; Moeller, K. D.
Bioorg. Med. Chem. Lett. 1998, 8, 3093.
4. For the design of analogs 2a and 2b see (a) Marshall, G. R.;
Gorin, F. A.; Moore, M. L. In Annual Reports in Medicinal
Chemistry; Clarke, F. H., Ed.; Academic: New York, 1978; Vol.
13, p 227. (b) Font, J. PhD Thesis, Washington University, St.
Louis, MO, 1986. (c) For an alternative application of these models
to the construction of TRH mimetics see Olson, G. L.; Cheung,
H. C.; Chiang, E.; Madison, V. S.; Sepinwall, J.; Vincent, G. P.;
Winokur, A.; Gary, K. A. J. Med. Chem. 1995, 38, 2866.
5. For reviews on the biology of TRH see: Thyrotropin-Releasing
Hormone: Biomedical Signi®cance. Proceedings of the Biotech
USA 89 Meeting, San Francisco, October 1989; Metcalf Olson,
G. L., Cheung, H. C., Voss, M. E., Hill, D. E., Kahn, M., Madison,
V. S., Cook, C. M., Sepinwall, J., Eds.; 1989.
(3R,7R,10S)-1-Aza-3-pyroglutamoylamine-10-carbo-
methoxybicyclo[5.3.0]decane. The synthesis was done in a
fashion identical to the preparation of 38 in order to obtain a
71% yield of the desired pyroglutamte coupled product. 1H
NMR 7.37 (d, Jˆ5.1 Hz, 1H), 7.10 (s, 1H), 4.47±4.44 (m,
1H), 4.37±4.32 (m, 1H), 4.15 (dd, Jˆ8.1, 5.1 Hz, 1H),
3.81±3.69 (m, 1H), 3.68 (s, 3H), 2.45±2.24 (m, 3H),
2.18±1.70 (m, 11H); 13C NMR (75 MHz/CDCl3) 179.0,
172.4, 171.7, 170.1, 61.4, 58.9, 57.0, 53.6, 52.3, 32.9,
31.2, 29.6, 27.4, 27.2, 25.7, 22.6; IR (neat/NaCl) 3279,
2944, 1744, 1689, 1626, 1441, 1207, 1178 cm21; LRMS
(FAB) m/e (rel. int.) 344 (M1Li, 18), 313 (34); HRMS
(FAB) m/e calcd for C16H23N3O5Li (M1Li) 344.1798,
found 344.1794.
6. For alternative approaches to understanding TRH/TRH-R bind-
ing see: (a) Hinkle, P. M.; Woroch, E. L.; Tashjian, G. H. J. Biol.
Chem. 1974, 249, 3085. (b) Goren, H. J.; Bauce, L. G.; Vale, W.
Mol. Pharmacol. 1977, 13, 606. (c) Kamiya, K.; Takamoto, M.;
Wada, Y.; Fujino, M.; Nishikawa, M. J. Chem. Soc., Chem.
Commun. 1980, 438. (d) Sharif, N. A.; To, Z.; Whiting, R. L.
Biochem. Biophys. Res. Commun. 1989, 161, 1306. (e) Vonhof,
S.; Feuerstein, G. Z.; Cohen, L. A.; Labroo, V. M. Eur. J.
Pharmacol. 1990, 180, 1.
(3R,7R,10S)-1-Aza-3-pyroglutamoylamine-10-carboxy-
amidebicyclo[5.3.0]decane (31c). Primary amide 31c was
made using the same procedure used to construct 31a
(40%). 1H NMR (300 MHz/CDCl3) 8.56 (d, Jˆ6.9 Hz,
1H), 7.52 (br s, 1H), 7.08 (br s, 1H), 6.17 (br s, 1H), 4.53
(t, Jˆ8.1 Hz, 1H), 4.40 (d, Jˆ6.3 Hz, 1H), 4.19±4.18
(m, 1H), 3.80±3.78 (m, 1H), 2.39±2.27 (m, 3H), 2.15±
2.02 (m, 2H), 1.96±1.74 (m, 9H); 13C NMR (75 MHz/
CDCl3) 179.4, 175.5,172.9, 170.5, 62.6, 59.0, 56.9, 53.0,
33.1, 31.3, 30.0, 27.9, 27.7, 25.0, 22.6; IR (neat/NaCl)
7. Biological data for 2-cyclohexyAla-TRH can be found in (a) Li,
W.; Moeller, K. D. J. Am. Chem. Soc. 1996, 118, 10106.
(b) Laakkonen, L.; Li, W.; Perlman, J. H.; Guarnieri, F.; Osman,
R.; Moeller, K. D.; Gershengorn, M. C. Mol. Pharmacol. 1996, 49,
1092.
8. Rutledge, L. D.; Perlman, J. H.; Gershengorn, M. C.; Marshall,
G. R.; Moeller, K. D. J. Med. Chem. 1996, 39, 1571.
3293(br), 2927, 1678, 1622, 1540, 1439, 1260 cm21
;
HRMS (FAB) m/e calcd for C15H27N4O4Li (M1Li)
329.1801, found 329.1809.
9. (a) Robl, J. A. Tetrahedron Lett. 1994, 35, 393. (b) Flynn, G.;
Giroux, E. L.; Dage, R. C. J. Am. Chem. Soc. 1987, 109, 7914.
(c) Cornille, F.; Slomczynska, U.; Smythe, M. L.; Beusen, D. D.;
Moeller, K. D.; Marshall, G. R. J. Am. Chem. Soc. 1995, 117, 909±
917.
Acknowledgements
10. (a) Mueller, R.; Revesz, L. Tetrahedron Lett. 1994, 35, 4091.
(b) Lombart, H.-G.; Lubell, W. D. J. Org. Chem. 1996, 61, 9437.
(c) Colombo, L.; Di Giacomo, M.; Papeo, G.; Carugo, O.;
Scolastico, C.; Manzoni, L. Tetrhadron Lett. 1994, 35, 4031.
We thank the National Institutes of Health (RO1
GM5324001A1) for their generous ®nancial support. In

L. M. Beal et al. / Tetrahedron 56 (2000) 10113±10125
10125
(d) Colombo, L.; Di Giacomo., Scolastico, C.; Manzoni, L.;
Belvisi, L.; Molteni, B. Tetrahedron Lett. 1995, 36, 625.
11. (a) Manfre', F.; Kern, J.-M.; Biellmann, J.-F. J. Org. Chem.
1992, 57, 2060. (b) Beal, L. M.; Moeller, K. D. Tetrahedron Lett.
1998, 39, 4639. (c) Campbell, J. A.; Rapoport, H. J. Org. Chem.
1996, 61, 6313.
E., Ed.; Springer: Berlin Heidelberg New York, 1988; Vol. 148, p
131. (c) For applications to functionalizing amino acids see
Moeller, K. D. Topics in Current Chemistry 1997, 185, 49 and in
particular Ref. 13 therein.
15. Kazmaier, U.; Krebs, A. Angew. Chem., Int. Ed. Engl. 1995,
34, 2012.
12. For the initial paper describing the application of RCM to a
peptidomimetic see: (a) Miller, S. J.; Grubbs, R. H. J. Am. Chem.
Soc. 1995, 117, 5855. For reviews of RCM reactions see:
(b) Grubbs, R. H.; Miller, S. J.; Fu, G. C. Acc. Chem. Res. 1995,
28, 446 and (c) Blechert, S.; Schuster, M. Angew. Chem., Int. Ed.
Engl. 1997, 36, 2036. For a route to fused nitrogen heterocycles
16. These structures were generated by overlapping the key
pharmacophoric groups for the minimized structures (MM2).
Because the tetracyclic analogs are not ¯exible, the minimized
structure for each was assumed to be an average picture of the
conformational space available to the analog.
17. Unpublished results with Laura Matson, Wenhua Chu, and
Marvin C. Gershengorn. The full details of this work will be
published elsewhere.
È
see: (d) Martin, S. F.; Liao, Y.; Chen, H. J.; Patzel, M.; Ramser,
M. N. Tetrahedron Lett. 1994, 35, 6005.
13. (a) Wistrand, L.-G.; Skrinjar, M. Tetrahedron 1991, 47, 573.
(b) Collado, I.; Ezquerra, J.; Pedregal, C. J. Org. Chem. 1995, 60,
5011.
18. For the synthesis of vinylglycine see: Afzali-Ardakani, A.;
Rapoport, H. J. Org. Chem. 1980, 45, 4817±4820.
19. Gao, Y.; Lane-Bell, P.; Vederas, J. C. J. Org. Chem. 1998, 63,
2133.
14. For reviews of anodic amide oxidations see: (a) Shono, T.;
Matsumura, Y.; Tsubata, K. In Organic Synthesis; Saucy, G.,
Ed.; Organic Synthesis Inc., 1984; Vol. 63, p 206 and references
therein. (b) Shono, T. In Topics in Current Chemistry; Steckhan,
20. Campagne, J.-M.; Ghosez, L. Tetrahedron Lett. 1998, 39,
6175.