Conformationally restricted TRH analogs: The compatibility of a 6,5-bicylic lactam-based mimetic with binding to TRH-R

Article abstract of DOI:10.1021/ja961887v

A pair of conformationally restricted TRH analogs have been synthesized. Both analogs have the central histidine of TRH replaced by a cyclohexylalanine unit, and both analogs contain a 6,5-bicyclic lactam ring fusing the proline peptide bond into a trans conformation. The analogs differ at the bridgehead stereocenter. The synthesis of the analogs utilized an anodic amide oxidation reaction to selectively functionalize a proline derivative and a titanium tetrachloride-initiated cyclization-rearrangement sequence to assemble the desired bicyclic ring skeleton. The analogs were compared with the unrestricted cyclohexyalanine-containing TRH analog (cyclohexyl-Ala²-TRH) in order to determine how the added lactam ring affected the affinity and the potency of the analog for TRH-R. Both the affinity and potency of the restricted analogs were found to be critically dependent on the bridgehead stereochemistry of the bicyclic ring. The analog having R stereochemistry at the bridgehead was 478 times more potent than the S isomer. In addition, the R isomer was found to be approximately 4.7 times more potent than its unrestricted counterpart. Similarly, the R isomer was found to have an affinity for TRH-R that was approximately 3.4 times the affinity of the unrestricted cyclohexyl-Ala²-TRH.

Full text of DOI:10.1021/ja961887v

10106
J. Am. Chem. Soc. 1996, 118, 10106-10112
Conformationally Restricted TRH Analogs: The Compatibility
of a 6,5-Bicyclic Lactam-Based Mimetic with Binding to
TRH-R
Wenhao Li and Kevin D. Moeller*
Contribution from the Department of Chemistry, Washington UniVersity,
St. Louis, Missouri 63130
ReceiVed June 5, 1996X
Abstract: A pair of conformationally restricted TRH analogs have been synthesized. Both analogs have the central
histidine of TRH replaced by a cyclohexylalanine unit, and both analogs contain a 6,5-bicyclic lactam ring fusing
the proline peptide bond into a trans conformation. The analogs differ at the bridgehead stereocenter. The synthesis
of the analogs utilized an anodic amide oxidation reaction to selectively functionalize a proline derivative and a
titanium tetrachloride-initiated cyclization-rearrangement sequence to assemble the desired bicyclic ring skeleton.
The analogs were compared with the unrestricted cyclohexyalanine-containing TRH analog (cyclohexyl-Ala2-TRH)
in order to determine how the added lactam ring affected the affinity and the potency of the analog for TRH-R.
Both the affinity and potency of the restricted analogs were found to be critically dependent on the bridgehead
stereochemistry of the bicyclic ring. The analog having R stereochemistry at the bridgehead was 478 times more
potent than the S isomer. In addition, the R isomer was found to be approximately 4.7 times more potent than its
unrestricted counterpart. Similarly, the R isomer was found to have an affinity for TRH-R that was approximately
3.4 times the affinity of the unrestricted cyclohexyl-Ala2-TRH.
Introduction
Scheme 1
TRH is the hypothalamic tripeptide that controls the release
of thyroid stimulating hormone from the anterior pituitary
gland.1 In addition to this endocrine activity, TRH displays
effects in the brain, blood, spinal cord, and gut. The widespread
biological activity of TRH, the existence of extensive reviews
describing the structure-activity relationships of TRH analogs,2
the close correlation between measured binding affinity and
biological potency for TRH analogs (suggesting that binding
and activity have similar structural requirements), and the fact
that the conformation of TRH can be described to a first
approximation with the use of only six torsional angles3
combined to make TRH an ideal model system for developing
chemical probes that are capable of elucidating the relationship
between the predicted and actual biological activity of peptide
conformations. This possibility was recognized by Garland
Marshall and co-workers, and in the early 1980s they used the
available analog data to predict several potential conformations
for the binding of TRH to its endocrine receptor TRH-R. These
predictions were then used to design a series of TRH analogs
as conformational probes for examining the validity of the
predictions and for refining the three-dimensional picture of the
TRH receptor site.4,5 One of the possible conformations
predicted for the binding of TRH to TRH-R is illustrated in
Scheme 1 (1), along with a family of proposed conformationally
restricted analogs (2).6
ing conformation 1 and then replacing spacially close hydrogens
with carbon bridges. The net result was to embed the peptide
backbone into a polycyclic ring skeleton.7 The advantages of
such an approach included the ease with which new analogs
could be designed, the preservation of both the peptide backbone
and the amino acid side chains, and the increased proteolytic
stability of the lactam rings. The principal disadvantage of this
approach was the steric size of the bridges. If the bridges
interfered with binding, then no new information about the
conformational requirements of binding to TRH-R would be
gained. Second, molecular modeling experiments were not able
to differentiate the bridgehead isomers with respect to the
potential of the analogs to bind TRH-R. Hence, both isomers
at Ha were proposed for study. Finally, the initially designed
analogs avoided the use of histidine as the central amino acid
by replacing the histidine with either phenylalanine or cyclo-
hexylalanine. Although analogs based on phenylalanine and
cyclohexylalanine were not nearly as potent as TRH itself, they
Several items concerning the proposed restricted analogs
deserve comment. First, the analogs were designed by examin-
(5) For alternative approaches to understanding the conformation of TRH
binding, see: (a) Hinkle, P. M.; Woroch, E. L.; Tashjian, G. H. J. Biol.
Chem. 1974, 249, 3085-3090. (b) Goren, H. J.; Bauce, L. G.; Vale, W.
Mol. Pharmacol. 1977, 13, 606-614. (c) Kamiya, K.; Takamoto, M.; Wada,
Y.; Fujino, M.; Nishikawa, M. J. Chem. Soc., Chem. Commun. 1980, 438-
439. (d) Sharif, N. A.; To, Z.; Whiting, R. L. Biochem. Biophys. Res.
Commun. 1989, 161, 1306-1311. (e) Vonhof, S.; Feuerstein, G. Z.; Cohen,
L. A.; Labroo, V. M. Eur. J. Pharmacol. 1990, 180, 1-12.
(6) For an additional application of the models developed by Marshall
and Font 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.
X Abstract published in AdVance ACS Abstracts, October 1, 1996.
(1) For an outstanding series of reviews, see: Thyrotropin-Releasing
Hormone: Biomedical Significance; Metcalf, G., Jackson, I. M. D., Eds.;
Ann. N. Y. Acad. Sci. 1989, 553.
(2) For a summary, see: Marshall, G. R.; Gorin, F. A.; Moore, M. L. In
Annual Reports in Medicinal Chemistry; Clarke, F. H., Ed.; Academic
Press: New York, 1978; Vol. 13, p 227 and references therein.
(3) Reference 2, p 229.
(4) Font, J. Ph.D. Thesis, Washington University, St. Louis, MO, 1986,
and unpublished results with Lawrence D. Rutledge.
S0002-7863(96)01887-2 CCC: $12.00 © 1996 American Chemical Society

Conformationally Restricted TRH Analogs
J. Am. Chem. Soc., Vol. 118, No. 42, 1996 10107
Chart 1
Chart 2
Scheme 2
Scheme 3
were both known to completely displace TRH from TRH-R and
to lead to the same maximal extent of stimulation as TRH. By
mimicking one or both of these TRH derivatives, it was hoped
that the compatibility of the bridges used to restrict the analogs
with binding and activation of the receptor could be determined
before taking on the extra challenge of synthesizing restricted
analogs having an imidazole substituent.
The bridge added to restrict the pyroglutamate region of 2
has been shown to be compatible with activation of TRH-R.8
In this work, analog 3 (Chart 1) were synthesized and tested
for its ability to bind and activate TRH-R. The resulting data
were compared to similar data obtained for the unrestricted
[Phe2]TRH 4. Analog 3 was found to have an approximately
3-fold lower affinity for TRH-R and a 3-fold lower potency for
TRH-R than did the unrestricted analog 4.
Restricting the X-Pro region of TRH proved to be more
difficult. Our initial plan was to construct a seven-membered-
ring lactam-based analog from compound 6 which would, in
turn, be formed from the cyclization of 5. However, the
cyclization of 5 failed to form the desired seven-membered-
ring analog and instead led to the six-membered-ring lactam
compound 7 (Scheme 2).9
Bicyclic lactam 7 was converted into a TRH analog (8, Chart
2). Unfortunately, the extra benzylic methylene group interfered
badly with binding to the TRH-R receptor. Even the unre-
stricted analog 9 with the extra methylene group failed to bind
TRH-R. It was clear that compound 8 could not serve as a
probe for determing the effectiveness of the added conforma-
tional constraint. A TRH analog that did not possess the extra
methylene group was needed. We report herein the synthesis
of a pair of such TRH analogs. These analogs were used to
demonstrate that the six-membered-ring lactam constraint could
serve to enhance the affinity and potency of a TRH analog for
TRH-R and to identify the bridgehead stereochemistry required
for binding and activity.
Synthesis of TRH Analogs 10
In principle, the rearrangement reaction outlined in Scheme
2 could be used to synthesize TRH analogs without the extra
methylene group that interfered with the binding of both 8 and
9 to TRH-R. With this in mind, the construction of a pair of
six-membered-ring lactam-based TRH analogs (10a,b) was
designed as outlined in Scheme 3. As in earlier designs, these
TRH analogs avoided the use of an imidazole group in order to
circumvent potential synthetic problems (Vide infra). In this
case, a cyclohexyl group was chosen as the substituent in order
to simplify the conversion of the initially formed six-membered-
ring lactam derivative 11 into the eventual TRH analog. The
plan for10,11 synthesizing 11 called for the anodic methoxylation
of amide 12, followed by treatment of the resulting 5-methox-
yproline derivative with TiCl4 in order to initiate the desired
cyclization-rearrangement sequence.
The forward syntheses of 10a,b are outlined in Scheme 4.
The syntheses started from compound 13, which was utilized
as a substrate for an ortho ester Claisen rearrangement. The
product from the Claisen rearrangement was saponified to form
an acid, the acid coupled to prolinol using EDCI coupling
conditions, and the resulting amide alcohol benzylated to form
the electrolysis substrate 12. The anodic oxidation of 12 was
accomplished at a carbon anode in methanol solvent using a
constant current of 26.8 mA. A platinum cathode was used as
the auxiliary electrode. After 2.1 F/mol had been passed, the
reaction was stopped. A 74% isolated yield of the methoxylated
amide 14 was obtained along with a 16% yield of the recovered
starting material.
The cyclization-rearrangement sequence leading to the
formation of the six-membered ring-lactam proceeded smoothly.
In this experiment, the treatment of the methoxylated amide
with titanium tetrachloride led to compound 11 in a 94% isolated
yield. As illustrated in Scheme 5, the proposed mechanism for
this reaction involves an initial cyclization to form a seven-
membered ring, followed by migration of the C3-C4 bond to
generate the tertiary carbocation and six-membered ring-lactam.
The resulting cation was trapped by chloride to form 11.
Intermediate 11 was converted into peptide building blocks
15a,b as outlined in Scheme 4. In analogy to earlier syntheses
involving the functionalization of six-membered-ring lactams,
the oxidation of the carbon R to the amide with LDA and O2
(7) For selected examples of lactam-based peptide mimetics, see: (a)
Wolf, J. P.; Rapoport, H. J. Org. Chem 1989, 54, 3164-3173, as well as
refs 1-8 therein. (b) Freidinger, R. M.; Perlow, D. S.; Veber, D. F. J. Org.
Chem. 1982, 47, 104-109. (c) Freidinger, R. M. J. Org. Chem. 1985, 50,
3631-3633. (d) Zydowsky, T. M.; Dellaria, J. F. Jr.; Nellans, H. N. J.
Org. Chem. 1988, 53, 5607-5616. (e) Kempf, D. J.; Condon, S. L. J. Org.
Chem. 1990, 55, 1390-1394. (f) Kemp, D. S.; McNamara, P. E. J. Org.
Chem. 1985, 50, 5834-5838 and references therein. (g) Nagai, U.; Sato,
K. Tetrahedron Lett. 1985, 26, 647-650. (h) Hinds, M. G.; Richards, N.
G. J.; Robinson, J. A. J. Chem. Soc., Chem. Commun. 1988, 1447-1449.
(i) Paul, P. K. C.; Burney, P. A.; Campbell, M. M.; Osguthorpe, D. J. Bioorg.
Med. Chem. Lett. 1992, 2, 141-144. (j) Ward, P.; Ewan, G. B.; Jordon, C.
C.; Ireland, S. J.; Hagan, R. M.; Brown, J. R. J. Med. Chem. 1990, 33,
1848-1851. (k) Deal, M. J.; Hagan, R. M.; Ireland, S. J.; Jordan, C. C.;
McElroy, A. B.; Porter, B.; Ross, B. C.; Stephens-Smith, M.; Ward, P. J.
Med. Chem. 1992, 35, 4195-4204. (l) Flynn, G. A.; Giroux, E. L.; Dage,
R. C. J. Am. Chem. Soc. 1987, 109, 7914-7915. (m) Flynn, G. A.;
Burkholder, T. P.; Huber, E. W.; Bey, P. Bioorg. Med. Chem. Lett. 1991,
1, 309-312. (n) Burkholder, T. P.; Huber, E. W.; Flynn, G. A. Bioorg.
Med. Chem. Lett. 1993, 3, 231-234. (o) Robl, J. A. Tetrahedron Lett. 1994,
35, 393-396, p. Lombart, H. G.; Lubell, W. D. J. Org. Chem. 1994, 59,
6147-6149. (q) Colombo, L.; Giacomo, M. D.; Papeo, G.; Carugo, O.;
Scolastico, C.; Manzoni, L. Tetrahedron Lett. 1994, 35, 4031-4034.
(8) Rutledge, L. D.; Perlman, J. H.; Gershengorn, M. C.; Marshall, G.
R.; Moeller, K. D. J. Med. Chem. 1996, 39, 1571.
(9) (a) Li, W.; Hanau, C. E.; d’Avignon, A.; Moeller, K. D. J. Org. Chem.
1995, 60, 8155. (b) Kao, J.; Li, W.; Moeller, K. D. Submitted for publication.

10108 J. Am. Chem. Soc., Vol. 118, No. 42, 1996
Li and Moeller
Scheme 4a
were completed served to illustrate the generality of the overall
synthetic approach.
Biological Data
Once synthesized, the analogs were tested for their ability to
activate TRH-R and compared with the unrestricted cyclohexyl-
Ala2-TRH.12 The EC50s for second messenger inositol phos-
a Reagents: (a) CH3C(OEt)3, CH3CH2CO2H, reflux; (b) KOH,
MeOH, heat, 69% (over two steps); (c) L-prolinol, HOBt, EDCI,
CH2Cl2, 86%; (d) (i) NaH, THF, 0 °C to RT, (ii) BnBr, THF, 0 °C to
RT, 89%; (e) carbon rod anode, Pt cathode, 0.03 M Et4NOTs, MeOH,
26.8 mA, 2.1 F/mole 74% (16% recovered starting material); (f) TiCl4,
CH2Cl2, 94%; (g) H2, Pd on C, NaOMe, MeOH, 94%; (h) TBDMS,
DBU, CH2Cl2, 95%; (i) (i) LDA, THF, -78 °C, (ii) O2, 62%; (j) NH3,
MeOH, 91% (7% recovered starting material); (k) HOpGlu, HOBt,
EDCI, CH2Cl2, isomer a 35%, isomer b 61%; (l) H2SO4, THF, H2O,
isomer a 96%, isomer b 78%; (m) Jones oxidation, isomer a 63%,
isomer b 50%; (n) (i) i-BuOCOCl, CH2Cl2, N-methylmorpholine, (ii)
NH3, isomer a 90%, isomer b 84%.
phate (IP) formation, that is, potencies, and the Kis of binding,
that is, affinities, were obtained according to previously
published procedures.13 The maximal extents of stimulation of
IP formation were similar for TRH, cyclohexylAla2-TRH, 10a,
and 10b, and at high concentrations all three cyclohexyl-based
analogs displaced 100% of TRH-R-bound [3H]-[NτMe-His]TRH.
Comparison of the restricted analogs with the unrestricted
cyclohexyl-Ala2-TRH indicated that the added bridge in 10a
badly interfered with both the binding and the potency of the
analog for TRH-R. However, analog 10b proved to be
approximately 4.7 times more potent than its unrestricted
counterpart. Similarly, analog 10b had an affinity for TRH-R
that was approximately 3.4 times the affinity of the unrestricted
cyclohexyl-Ala2-TRH. The results of these studies are sum-
marized in Table 1.14
(step i in Scheme 4) led directly to the formation of a
1,2-dicarbonyl compound.9a The dicarbonyl was converted into
the desired enamine building blocks with the use of ammonia
in methanol. The bridgehead stereoisomers were then separated
at this point of the synthesis by gravity flow chromatography
through silica gel. Each isomer was then carried on independ-
ently. The stereochemistry of isomers 15a,b was assigned by
analogy to previously synthesized building blocks having the
same ring system.9 Specifically, enamine-based 6,5-bicyclic
lactam building blocks like 15a,b fall into two “stereochemical
families” that can be readily distinguished by proton NMR.
Initially the stereochemistry of the two families was identified
by the use of 2D-NOE data at the building block stage9a and
by a combination of 2D-NOE data and coupling constant
information once the building block was incorporated into a
peptide.9b Once the stereochemical families were characterized,
three proton NMR signals could be used to identify which
stereochemical family a new analog belonged to. An isomer
having S stereochemistry at the bridgehead could be character-
ized by a methine signal at 4.2 ppm, an AM of an AMX pattern
where A was centered around 4.0 ppm, M was centered around
3.8 ppm, and JAX was larger than JMX, and a methine signal at
1.6-1.5 ppm. An isomer having R stereochemistry at the
bridgehead could be characterized by a methine signal at 4.1
ppm (the appearance of which was significantly different than
that for the S isomer), an AM of an AMX pattern where A was
centered around 3.8 ppm, M was centered around 3.6 ppm, and
JMX was larger that JAX, and a methine buried under other signals
at approximately 1.8 ppm. For the cyclohexyl case examined
here, the methine signal around 1.6-1.5 ppm in 16a was
partially obscured by the cyclohexyl protons. However, the
other two characteristic signals could be easily seen for both
15a and 15b.
Clearly, the added lactam ring in analog 10b did not interfere
with binding but rather served to enhance both the potency and
the affinity of the analog for TRH-R. It would appear that the
(10) For pioneering work with anodic amide oxidations, see: (a) Ross,
S. D.; Finkelstein, M.; Peterson, C. J. Am. Chem. Soc. 1964, 86, 4139. (b)
Ross, S. D.; Finkelstein, M.; Peterson, C. J. Org. Chem. 1966, 31, 128. (c)
Ross, S. D.; Finkelstein, M.; Peterson, C. J. Am. Chem. Soc. 1966, 88,
4657. For reviews, see: (d) Shono, T. Tetrahedron 1984, 40, 811. (e) Shono,
T.; Matsumura, Y.; Tsubata, K. In Organic Synthesis; Saucy, G., Ed.;
Organic Synthesis Inc., 1984; Vol. 63, p 206 and references therein. (f)
Shono, T. In Topics in Current Chemistry; Steckhan, E. Ed.; Springer-
Verlag: Berlin-Heidelberg-New York, 1988; Vol. 148, p 131.
(11) For examples of amide oxidations being used to functionalize
peptides, see refs 8 and 10 as well as the following: (a) Shono, T.;
Matsumura, Y.; Inoue, K. J. Org. Chem. 1983, 48, 1388. (b) Thaning, M.;
Wistrand, L.-G. HelV. Chim. Acta 1986, 69, 1711. (c) Renaud, P.; Seebach,
D. Angew. Chem., Int. Ed. Engl. 1986, 25, 843. (d) Renaud, P.; Seebach,
D. HelV. Chim. Acta 1986, 69, 1704. (e) Seebach, D.; Charczuk, R.; Gerber,
C.; Renaud, P.; Berner, H.; Schneider, H. HelV. Chim. Acta 1989, 72, 401.
(f) Papadopoulos, A.; Heyer, J.; Ginzel K.-D.; Steckhan, E. Chem. Ber.
1989, 122, 2159. (g) Ginzel, K.-D.; Brungs, P.; Steckhan, E. Tetrahedron
1989, 45, 1691. (h) Papadopoulos, A.; Lewall, B.; Steckhan, E.; Ginzel,
K.-D.; Knoch, F.; Nieger, M. Tetrahedron 1991, 47, 563. (i) Barrett, A. G.
M.; Pilipauskas, D. J. Org. Chem. 1991, 56, 2787. (j) Moeller, K. D.;
Rutledge, L. D. J. Org. Chem. 1992, 57, 6360. (k) Cornille, F.; Fobian, Y.
M.; Slomczynska, U.; Beusen D. D.; Marshall, G. R.; Moeller, K. D.
Tetrahedron Lett. 1994, 35, 6989-6992. (l) Cornille, F.; Slomczynska, U.;
Smythe, M. L.; Beusen, D. D.; Moeller, K. D.; Marshall, G. R. J. Am.
Chem. Soc. 1995, 117, 909. (m) Slomczynska, U.; Chalmers, D. K.; Cornille,
F.; Smythe, M. L.; Beusen, D. D.; Moeller, K. D.; Marshall, G. R. J. Org.
Chem. 1996, 61, 1198-1204. (n) Fobian, Y. M.; d’Avignon, D. A.; Moeller,
K. D. Bioorg. Med. Chem. Lett. 1996, 6, 315-318.
(12) The unrestricted analog was synthesized from proline amide using
standard peptide coupling conditions.
(13) Perlman, J. H.; Thaw, C. N.; Laakkonen, L.; Bowers, C. Y.; Osman,
R.; Gershengorn, M. C. J. Biol. Chem. 1994, 269, 1610. Straub, R. E.;
Frech, G. C.; Joho, R. H.; Gershengorn, M. C. Proc. Natl. Acad. Sci. U.S.A.
1990, 87, 9514.
(14) For a detailed experimental discussion concerning biological data
and an interpetation of the data with respect to a second model for the
binding of TRH to TRH-R based on receptor mutants, see: Laakkonen,
L.; Li, W.; Perlman, J. H.; Guarnieri, F.; Osman, R.; Moeller, K. D. and
Gershengorn, M. C. Mol. Pharmacol. 1996, 49, 1092.
Once obtained, 15a,b were converted into TRH analogs using
standared peptide coupling techniques. The yield for the
coupling of isomer 15a with pyroglutamic acid was never
optimized due to the inability of 10a to serve as a potent ligand
for TRH-R (Vide infra). The ease with which these syntheses

Conformationally Restricted TRH Analogs
J. Am. Chem. Soc., Vol. 118, No. 42, 1996 10109
Table 1. Binding and Activation of TRH-R Receptors by TRH Analogs 10a and 10b
analog
Ki (nM)b
EC50 (nM)c
[NτMe-His]TRH
TRH
1.2 (0.92-1.5)d
nd
nd
1.1 (0.95-1.3)d
430 (380-480)d
44 000 (37 000-52 000)d
91 (76-110)d
cyclohexyl-Ala2-TRH
analog 10a
6500 (5400-7900)d
290 000 (220 000-370 000)d
1900 (1600-2200)d
analog 10b
a Experiments were performed with intact AtT-20 mouse pituitary tumor cells stably expressing TRH receptors. b For binding, cells were incubated
with 1 nM [3H]-[NτMe-His]TRH in the absence or presence of various doses of unlabelled TRH analogs for 1 h at 37 °C. The data are means of
duplicate determinations in two or three experiments. c For activation, cells prelabeled with [3H]myoinositol were incubated with various doses of
TRH analogs for 1 h at 37 °C, and inositol phosphate formation was measured. Maximal extents of stimulation were similar for all analogs. The
data are means of duplicate determinations in two or three experiments. d 95% confidence intervals. e nd, not determined.
was then raised to reflux and heated until about 42 mL of ethanol was
removed from the reaction. At this point, the reaction was cooled,
MeOH (100 mL) and KOH (5.8 g, 103.4 mmol) were added, and the
mixture was again raised to reflux for an additional 5 h. The MeOH
was then removed under reduced pressure, and the crude oil was
partitioned between a saturated NaHCO3 solution and Et2O. The
aqueous layers were combined, acidified to pH ) 1, and extracted with
CH2Cl2. The organic layers were combined, dried over MgSO4, and
concentrated in Vacuo. The crude oil was distilled to afford 6.000 g
(69%) of the desired acid product. The spectral data for the acid were
as follows: 1H NMR (300 MHz/CDCl3) δ 10.50 (br s, 1H), 5.80 (dd,
1H, J ) 17.7, 10.5 Hz), 5.15-5.01 (m, 2H), 2.35 (s, 2H), 1.71-1.64
(m, 2H), 1.55-1.45 (m, 8H); 13C NMR (75 MHz/CDCl3) δ 177.9,
144.5, 113.6, 45.7, 39.1, 35.5, 26.1, 22.0; IR (neat/NaCl) 3083, 2928,
2853, 1706, 1455, 1448, 1411, 1302, 1278, 1260, 914, 676 cm-1; LRMS
(PCI) m/e (relative intensity) 170 (M + 2, 5), 169 (M + 1, 22), 167
(10), 152 (12), 151 (56), 133 (17), 123 (17), 110 (19), 109 (100), 108
(10), 107 (16); HRMS (EI) m/e calcd for C10H16O2 168.1150, found
168.1148.
(2S)-2-(Hydroxymethyl)-1-(1′-vinylcyclohexylacetoyl)pyrroli-
dine. To a mixture of the acid generated above (6.000 g, 35.7 mmol)
in 70 mL of CH2Cl2 were added HOBt (6.755 g, 50.0 mmol) and
L-prolinol (4.335 g, 42.9 mmol). The mixture was stirred at 0 °C for
5 min, and then 1-[3-(dimethylamino)propyl]-3-ethylcarbodiimide
(EDCI) (8.876 g, 46.4 mmol) was added. The resulting mixture was
allowed to warm to room temperature and stirred at room temperature
overnight. The reaction was then poured into a separatory funnel and
washed with saturated NaHCO3, 10% citric acid, and an aqueous
saturated NaCl solution. The organic layers were combined, dried over
MgSO4, and concentrated in Vacuo. The crude product was chromato-
graphed through ca. 300 g of silica gel using a 5% MeOH in Et2O
solution as eluant to afford 7.703 g (86%) of the amide as a colorless
oil. The spectral data for the amide were as follows: 1H NMR (300
MHz/CDCl3) δ 5.80 (dd, 1H, J ) 17.4, 10.8 Hz), 5.24 (br s, 1H), 5.15-
5.02 (m, 2H), 4.26-4.18 (m, 1H), 3.65-3.50 (m, 3H), 3.47-3.39 (m,
1H), 2.32 and 2.31 (two s, 2H), 2.07-1.95 (m, 1H), 1.93-1.73 (m,
5H), 1.60-1.43 (m, 8H); 13C NMR (75 MHz/CDCl3) δ 172.6, 144.5,
113.5, 67.3, 60.7, 49.0, 46.0, 40.0, 36.1, 35.7, 28.2, 26.1, 24.3, 22.0;
IR (neat/NaCl) 3394 br, 2926, 2855, 1701, 1662, 1617, 1586, 1447,
1429, 1401, 1052, 914, 786, 657 cm-1; LRMS (PCI) m/e (relative
intensity) 280 (M + 29, 12), 253 (M + 2, 27), 252 (M + 1, 100), 251
(M, 9), 250 (17), 102 (27); HRMS (EI) m/e calcd for C15H25O2N
251.1885, found 251.1874.
Scheme 5
lactam ring used to restrict 10b is an excellent constraint for
beginning to experimentally “map” the TRH-R receptor. In
addition, the data indicated that the R configuration at the
bridgehead stereocenter was essential for both high potency and
high affinity for the receptor.
Conclusions
In summary, we have found that using electrochemistry to
systematically embed the X-Pro region of a TRH analog into a
1-aza-2-oxobicyclo[4.3.0]nonane ring skeleton can lead to a
new, conformationally restricted analog with enhanced potency
and affinity for TRH-R. The potency and affinity of the analog
for TRH-R were critically dependent on the bridgehead stereo-
chemistry of the bicyclic ring system. The R isomer at the
bridgehead was 478 times more potent than the S isomer. The
synthesis of future analogs will have to focus on controlling
the stereochemistry at this position. At this point, both bridges
proposed by Marshall and Font in the design of analog 2 appear
to be compatible with the receptor site. Efforts aimed at
synthesizing and testing the fully restricted analogs and at
synthesizing and testing TRH analogs having the imidazole ring
present are underway.
Experimental Section15
(1′-Vinylcyclohexyl)acetic Acid. A solution of triethyl phosphono-
acetate (22.419 g, 100 mmol) in 120 mL of DME was added dropwise
to a suspension of NaH (4.00 g of 60% mixture with mineral oil, 100
mmol) in 80 mL of DME at 0 °C. The mixture was allowed to warm
to room temperature, stirred for 1 h, cooled to 0 °C, and treated with
cyclohexanone (9.815 g, 100 mmol). The resulting mixture was stirred
at room temperature for an additional 30 min. The reaction mixture
was then poured into a separatory funnel and washed with water. The
organic layers were combined, dried over MgSO4, and concentrated in
Vacuo to afford 16.011 g (95%) of the crude product. A solution of
this crude product (12.979 g, 77.3 mmol) in 100 mL of dry ether was
added dropwise to a suspension of LiAlH4 (4.398 g, 115.9 mmol) in
50 mL of dry ether at 0 °C. The mixture was allowed to warm to
room temperature and stirred overnight. The reaction was quenched
by adding small chips of ice. The aqueous layer was extracted with
ether, and then the combined organic layers were dried over MgSO4
and concentrated in Vacuo to afford 8.710 g (89%) of the crude alcohol
13. The crude alcohol 13 (6.512 g, 51.7 mmol) was placed in a flame-
dried 250 mL round-bottom flask along with triethyl orthoacetate
(41.949 g, 258.4 mmol) and propionic acid (0.229 g, 3.1 mmol). To
the flask was attached a Dean-Stark trap and a condenser. The reaction
(2S)-2-(Benzyloxymethyl)-1-(1′-vinylcyclohexylacetonyl)pyrrol-
idine (12). A solution of the above amide (7.228 g, 28.8 mmol) in 40
mL of THF was added dropwise to a suspension of NaH (1.728 g of
a 60% mixture with mineral oil, 43.2 mmol) in 20 mL of THF at 0 °C.
The mixture was allowed to warm to room temperature, stirred for 30
min, cooled to 0 °C, and treated with benzyl bromide (5.418 g, 31.7
mmol). The reaction was allowed to warm to room temperature, stirred
overnight, and then quenched by adding an aqueous saturated NaCl
solution. The emulsion was transferred to a separatory funnel, diluted
with Et2O, and washed with water. The aqueous layers were combined
and extracted with Et2O. The organic layers were combined, dried
over MgSO4, and concentrated in Vacuo. The crude product was
chromatographed through ca. 250 g of silica gel using a 30% hexane
in Et2O solution as eluant to afford 8.723 g (89%) of the protected
amide as a colorless oil. The spectral data for the protected amide
were as follows: 1H NMR (300 MHz/CDCl3) δ 7.36-7.26 (m, 5H),
(15) For a general experimental section, see: Wong, P. L.; Moeller, K.
D. J. Am. Chem. Soc. 1993, 115, 11434.

10110 J. Am. Chem. Soc., Vol. 118, No. 42, 1996
Li and Moeller
5.86-5.73 (m, 1H), 5.08-4.94 (m, 2H), 4.56-4.45 (m, 2H), 4.34-
4.27 (m, 0.7H), 4.08-4.01 (m, 0.3H), 3.65 (dd, 1H, J ) 9.3, 3.0 Hz),
3.52-3.35 (m, 3H), 2.28 and 2.27 (two s, 2H), 2.04-1.71 (m, 6H),
1.60-1.41 (m, 8H); 13C NMR (75 MHz/CDCl3) δ 170.2, 170.0, 145.1,
138.5, 128.4, 128.2, 127.7, 127.5, 113.0, 73.2, 73.0, 71.0, 70.1, 57.2,
56.1, 48.2, 45.9, 45.2, 39.8, 36.8, 35.9, 35.7, 35.2, 28.6, 27.5, 26.2,
24.2, 22.2, 22.1, 22.0, 21.7; IR (neat/NaCl) 2928, 2856, 1636, 1452,
1412, 1363, 1197, 1102, 911, 736, 698 cm-1; LRMS (PCI) m/e (relative
intensity) 370 (M + 29, 13), 343 (M + 2, 55), 342 (M + 2, 100), 341
(M, 12), 340 (18); HRMS (EI) m/e calcd for C22H31O2N 341.2355,
found 341.2368.
diastereomers obtained were as follows: 1H NMR (300 MHz/CDCl3)
δ 4.27-4.12 (m, 1H), 3.73-3.38 (m, 3H), 2.52 (td, 1H, J ) 17.4, 6.0
Hz), 2.19-1.93 (m, 4H), 1.77-1.62 (m, 6H), 1.57-1.32 (m, 2H), 1.20-
0.91 (m, 8H); 13C NMR (75 MHz/CDCl3) δ 172.0, 68.1, 67.6, 61.3,
61.2, 61.0, 59.6, 58.2, 42.2, 41.8, 38.8, 38.2, 35.7, 35.4, 32.7, 32.5,
31.3, 29.8, 29.7, 27.3, 26.4, 26.2, 18.3; IR (neat/NaCl) 3276 br, 2931,
2854, 1694, 1546, 1455, 1419, 1321, 1265, 730 cm-1; LRMS (PCI)
m/e (relative intensity) 292 (M + 41, 10), 280 (M + 29, 24), 254 (M
+ 3, 2), 253 (M + 2, 44), 252 (M + 1, 100), 234 (3), 220 (3); HRMS
(EI) m/e calcd for C15H25O2N 251.1885, found 251.1887.
(9S)-1-Aza-9-[(tert-butyldimethylsiloxy)methyl]-4-cyclohexyl-2-
oxobicyclo[4.3.0]nonane. To a solution of the hydrogenated compound
(3.341 g, 13.3 mmol) in 30 mL of CH2Cl2 were added TBDMSCl (4.028
g, 26.6 mmol) and DBU (4.197 g, 28.0 mmol). The mixture was stirred
at room temperature for 1 h and then transferred to a separatory funnel.
The organic layer was separated and then washed with a 10% citric
acid solution followed by a saturated aqueous NaHCO3 solution. The
CH2Cl2 was removed under reduced pressure and the crude product
chromatographed through ca. 250 g of silica gel using 80% ether/hexane
as eluant to afford 4.251 g of the protected amide (95%). The spectral
data for the mixture of diastereoisomers were as follows: 1H NMR
(300 MHz/CDCl3) δ 4.18-4.11 (m, 0.5H), 4.04-3.98 (m, 0.5H), 3.95
(dd, 0.5H, J ) 10.2, 4.5 Hz), 3.86 (dd, 0.5H, J ) 9.6, 5.1 Hz), 3.73
(dd, 0.5H, J ) 9.9, 2.7 Hz), 3.67 (dd, 0.5H, J ) 10.2, 2.4 Hz), 3.48-
3.32 (m, 1H), 2.42 (td, 1H, J ) 18.6, 6.6 Hz), 2.10-1.80 (m, 5H),
1.78-1.55 (m, 6H), 1.39-0.90 (m, 8H), 0.88 and 0.87 (two s, 9H),
0.03, 0.02, 0.01 and 0 (four s, 6H); 13C NMR (75 MHz/CDCl3) δ 169.6,
169.1, 63.0, 62.1, 60.1, 59.9, 58.1, 57.8, 42.5, 42.0, 39.3, 38.7, 35.9,
33.1, 33.0, 32.4, 31.2, 29.9, 29.8, 29.7, 26.6, 26.5, 26.3, 25.9, 25.8,
25.7, 24.8, 18.1, -3.6, -5.4, -5.5; IR (neat/NaCl) 2927, 2854, 1620,
(2S)-2-(Benzyloxymethyl)-5-methoxy-1-(1′-vinylcyclohexylacetoyl)-
pyrrolidine (14). An oven-dried three-neck flask was charged with
the protected amide (13.321 g, 39.1 mmol), MeOH (78 mL), and
tetraethylammonium tosylate (0.705 g, 2.34 mmol). The flask was
equipped with a carbon anode, a platinum wire cathode, and a nitrogen
inlet. The reaction mixture was degassed by sonication while a slow
stream of nitrogen was passed through the solution for 5 min. The
mixture was then electrolyzed at a constant current of 26.8 mA until
2.1 F/mol had been passed. When complete, the MeOH was removed
under reduced pressure and the crude oil chromatographed through ca.
600 g of silica gel using a gradient elution from 60% ether/hexane to
80% ether/hexane. The column afforded 10.748 g (74%) of the desired
methoxylated amide as a mixture of diastereomers, along with 2.179 g
(16%) of the recovered starting material. The spectral data for the
mixture of diastereoisomers were as follows: 1H NMR (300 MHz/
CDCl3) δ 7.36-7.25 (m, 5H), 5.87-5.73 (m, 1H), 5.59 (d, 0.3H, J )
5.1 Hz), 5.11-4.93 (m, 2H), 4.90 (d, 0.7H, J ) 3.3 Hz), 4.58-4.33
(m, 2H), 4.30-4.09 (m, 0.7H), 3.91 (dd, 0.3H, J ) 9.0, 3.6 Hz), 3.59
(dd, 0.5H, J ) 9.6, 3.3 Hz), 3.54-3.32 (m, 1.5H), 3.28 and 3.23 (two
s, 3H), 2.41 and 2.35 (two s, 2H), 2.10-1.61 (m, 6H), 1.54-1.26 (m,
8H); 13C NMR (75 MHz/CDCl3) δ 171.6, 145.4, 145.2, 138.5, 128.4,
128.3, 127.7, 127.6, 127.5, 127.4, 113.3, 113.0, 112.9, 89.8, 87.4, 74.7,
73.3, 73.1, 72.1, 69.9, 57.1, 56.7, 56.3, 55.7, 53.9, 45.5, 45.3, 39.8,
37.0, 36.8, 36.6, 35.4, 35.2, 31.0, 30.5, 29.0, 27.0, 26.3, 26.2, 26.1,
24.7, 22.2, 22.1, 22.0; IR (neat/NaCl) 2929, 2855, 1701, 1689, 1647,
1463, 1413, 1361, 1327, 1253, 1088, 1049, 889, 834, 773, 665 cm-1
;
LRMS (PCI) m/e (relative intensity) 395 (M + 30, 6), 394 (M + 29,
18), 368 (M + 3, 8), 367 (M + 2, 29), 366 (M + 1, 100), 365 (M, 9),
364 (26), 351 (13), 350 (49), 309 (8), 308 (40); HRMS (EI) m/e calcd
for C21H39O2NSi 365.2750, found 365.2775.
(9S)-1-Aza-9-[(tert-butyldimethylsiloxy)methyl]-4-cyclohexyl-3-
hydroxy-2-oxobicyclo[4.3.0]non-3-ene. An oven-dried two-neck flask
was charged with 15 mL of THF and 4.071 g (40.2 mmol) of
diisopropylamine. To this mixture was added 15.7 mL (39.3 mmol)
of a 2.5 M n-butyllithium in hexanes solution at -78 °C. The mixture
was stirred at -78 °C for 0.5 h, and then a solution of the protected
amide from above (6.991 g, 19.2 mmol) in 20 mL of THF was
cannulated slowly into the flask. The flask originally containing the
protected amide was washed with an additional 5 mL of THF, and
these washings were added to the reaction. The resulting mixture was
stirred at -78 °C for another 45 min. Following this period, oxygen
was bubbled through the reaction for 30 min at -78 °C. The reaction
was quenched by the addition of a saturated aqueous NaHSO3 solution
to the flask. The reaction was then transferred to a separatory funnel,
the layers were separated, and the aqueous layer was extracted with
ether. The organic layers were combined, dried over MgSO4, and
concentrated in Vacuo. The crude product was chromatographed
through 400 g of silica gel using a gradient elution from 30% ether/
hexane to 70% ether/hexane as eluant to afford 4.524 g (62%) of the
product as a white solid. The spectral data for the mixture of
diastereoisomers were as follows: 1H NMR (300 MHz/CDCl3) δ 6.29
and 6.28 (two s, 1H), 4.19-4.06 (m, 2H), 3.85-3.56 (m, 3H), 2.80-
2.68 (m, 1H), 2.39 (dd, 1H, J ) 16.5, 6.3 Hz), 2.27-1.99 (m, 3H),
1.82-1.53 (m, 6H), 1.46-1.11 (m, 5H), 0.93 and 0.91 (two s, 9H),
0.09, 0.07, 0.06 and 0.05 (four s, 6H); 13C NMR (75 MHz/CDCl3) δ
162.5, 161.6, 137.1, 136.7, 123.2, 121.2, 62.9, 62.3, 58.5, 58.1, 57.7,
57.5, 37.4, 37.3, 32.2, 30.9, 30.8, 29.1, 28.9, 28.6, 26.9, 26.2, 26.1,
25.9, 25.8, 18.2, 18.1, -5.5; IR (neat/NaCl) 3393 br, 2928, 2854, 1637,
1455, 1450, 1330, 1257, 1111, 1056, 845, 837, 777, 667 cm-1; LRMS
(PCI) m/e (relative intensity) 408 (M + 29, 13), 382 (M + 3, 11), 381
(M + 2, 26), 380 (M + 1, 100), 379 (M, 7), 378 (20), 364 (29), 322
(28); HRMS (EI) m/e calcd for C21H37O3NSi 379.2543, found 379.2527.
1622, 1559, 1539, 1528, 1457, 1438, 1415, 1399, 1085, 757, 698 cm-1
;
LRMS (PCI) m/e (relative intensity) 372 (M + 1, 6), 341 (10), 340
(25), 218 (8), 191 (14), 190 (100), 91 (7); HRMS (EI) m/e calcd for
C23H33O3N 371.2460, found 371.2463.
(9S)-1-Aza-4-(1′-chlorocyclohexyl)-9-(hydroxymethyl)-2-oxobicyclo-
[4.3.0]nonane (11). To a -78 °C solution of the methoxylated amide
(10.745 g, 29.0 mmol) in 150 mL of CH2Cl2 was added 72.4 mL of a
1 M TiCl4 in CH2Cl2 solution. The mixture was allowed to warm to
room temperature and stirred overnight. The reaction was then
quenched by the addition of a 30% solution of Rochelle’s salt in water.
The reaction mixture was transferred to a separatory funnel, the layers
were separated, and the aqueous layer was extracted 10 times with
CH2Cl2. The organic layers were combined, dried over MgSO4, and
concentrated in Vacuo. The crude product was chromatographed
through 600 g of silica gel using a 10% solution of MeOH in ether as
the eluant to afford 7.770 g (94%) of the cyclized compound as a yellow
oil. The spectral data for the mixture of diastereoisomers obtained were
as follows: 1H NMR (300 MHz/CDCl3) δ 4.29-4.15 (m, 1H), 3.76-
3.42 (m, 3H), 2.69-2.42 (m, 2H), 2.30 (t, 1H, J ) 12.3 Hz), 2.18-
1.90 (m, 5H), 1.82-1.14 (m, 12H); 13C NMR (75 MHz/CDCl3) δ 171.1,
76.6, 68.2, 67.6, 61.4, 61.3, 60.5, 59.0, 45.1, 44.5, 37.7, 37.0, 36.9,
33.0, 32.7, 32.5, 31.4, 30.2, 30.1, 27.4, 26.3, 25.2, 22.0, 21.8; IR (neat/
NaCl) 3391 br, 2934, 2863, 1618, 1459, 1448, 1413, 1328, 1058, 1048,
873, 686 cm-1; LRMS (FAB) m/e (relative intensity) 286 (M + 1,
100), 250 (37), 154 (36), 136 (42), 117 (78), 115 (42), 91 (48), 89
(34); HRMS (FAB) m/e calcd for C15H25O2ClN 286.1573 (M + 1),
found 286.1584.
(9S)-1-Aza-4-cyclohexyl-9-(hydroxymethyl)-2-oxobicyclo[4.3.0]-
nonane. To a solution of the cyclized compound (7.720 g, 27.0 mmol)
in 54 mL of MeOH were added 1.93 g of 5% Pd on activated carbon
and 2.921 g (54.1 mmol) of sodium methoxide. The mixture was stirred
under a hydrogen balloon overnight. The catalyst was then removed
with the use of a fritted glass funnel and the filtrate concentrated in
vacuo. The crude product was chromatographed through 500 g of silica
gel using 10% MeOH in ether as the eluant to afford 6.381 g (94%) of
the hydrogenated product. The spectral data for the mixture of
(6S and 6R,9S)-1-Aza-3-amino-9-[(tert-butyldimethylsiloxy)-
methyl]-4-cyclohexyl-2-oxobicyclo[4.3.0]non-3-ene (15a,b). Am-
monia gas was passed through a solution of the above 1,2-dicarbonyl
compound (4.497 g, 11.9 mmol) in 500 mL of MeOH at 0 °C until the
solution was saturated. The reaction mixture was stirred at room

Conformationally Restricted TRH Analogs
J. Am. Chem. Soc., Vol. 118, No. 42, 1996 10111
temperature under a rubber septum for 3 days. The MeOH was
removed under reduced pressure, and the crude product was chromato-
graphed through 400 g of silica gel using 50% ether/hexane as the
eluant in order to afford 2.324 g (52%) of isomer a, 1.728 g (39%) of
isomer b, and 0.333 g (7%) of recovered starting material. The spectral
data for isomer a were as follows: 1H NMR (300 MHz/CDCl3) δ 4.21-
4.14 (m, 1H), 3.98 (A of AMX, 1H, JAM ) 9.6 Hz, JAX ) 4.8 Hz),
3.78 (M of AMX, 1H, JMA ) 9.6 Hz, JMX ) 2.7 Hz), 2.35-2.28 (m,
1H), 2.31 (dd, 1H, J ) 15.9, 5.1 Hz), 2.23-2.13 (m, 1H), 2.10-1.94
(m, 3H), 1.87-1.66 (m, 4H), 1.61-1.48 (m, 2H), 1.40-1.09 (m, 6H),
0.88 (s, 9H), 0.06 and 0.03 (two s, 6H); 13C NMR (75 MHz/CDCl3) δ
162.0, 129.3, 120.8, 62.6, 58.5, 56.8, 39.1, 32.1, 30.5, 29.4, 29.0, 26.4,
26.3, 26.1, 26.0, 25.8, -5.5; IR (neat/NaCl) 3450, 3354, 2928, 2856,
1663, 1631, 1583, 1450, 1361, 1323, 1255, 1189, 1155, 1116, 1035,
1007, 939, 837, 776, 736, 666 cm-1; LRMS (FAB) m/e (relative
intensity) 380 (M + 2, 57), 379 (M + 1, 100), 378 (M, 43), 322 (41),
321 (78), 233 (38), 154 (31); HRMS (FAB) m/e calcd for C21H39O2N2-
Si 379.2780 (M + 1), found 379.2788.
intensity) 491 (M + 2, 33), 490 (M + 1, 100), 432 (15), 307 (16);
HRMS (FAB) m/e calcd for C26H44O4N3Si (M + 1) 490.3101, found
490.3100.
(6S and 6R,9S,12S)-1-Aza-4-cyclohexyl-9-(hydroxymethyl)-2-oxo-
3-(pyroglutamylamino)bicyclo[4.3.0]non-3-ene. The tripeptide (iso-
mer a, 6S) (0.110 g, 0.23 mmol) was dissolved in a solution comprised
of 1 mL of 2 N H2SO4 and 3 mL of THF. The mixture was stirred at
room temperature overnight. The solution was transferred into a
separatory funnel, the layers were separated, and the aqueous layer was
extracted with CH2Cl2. The organic layers were combined, dried over
MgSO4, and concentrated in Vacuo. The crude product was then
chromatographed through silica gel with the use of 40% MeOH in ether
as eluant to afford 81.5 mg (96%) of the deprotected product (isomer
a) as a white solid. The spectral data for isomer a were as follows:
1H NMR (300 MHz/CDCl3) δ 8.20 (s, 1H), 7.33 (s, 1H), 4.31 (dd, 1H,
J ) 8.4, 4.8 Hz), 4.10-4.08 (m, 1H), 3.88-3.71 (m, 2H), 3.56-3.50
(m, 1H), 2.59-2.45 (m, 2H), 2.42-2.08 (m, 8H), 1.80-1.67 (m, 4H),
1.61-1.49 (m, 3H), 1.30-1.10 (m, 5H); 13C NMR (75 MHz/CDCl3) δ
179.2, 172.2, 162.9, 150.7, 121.6, 65.5, 62.0, 57.2, 56.3, 41.0, 32.0,
30.5, 30.2, 29.5, 28.6, 27.0, 26.2, 25.9; IR (neat/NaCl) 3387 br, 3267
br, 2931, 2854, 1685, 1645, 1616, 1522, 1449, 1319, 1265, 1098, 735
cm-1; LRMS (FAB) m/e (relative intensity) 386 (M + 11, 19), 376
(M + 1, 44), 308 (26), 307 (100), 289 (46), 235 (15), 219 (15); HRMS
(FAB) m/e calcd for C29H30O4N3 (M + 1) 376.2236, found 376.2235.
Isomer b (6R) was carried on in an identical fashion using pyridine
and HCl instead of H2SO4 with a yield of 78%. The spectral data for
isomer b were as follows: 1H NMR (300 MHz/CDCl3) δ 8.26 (s, 1H),
7.89 (s, 1H), 4.29 (dd, 1H, J ) 7.8, 3.0 Hz), 4.18-4.11 (m, 1H), 3.81-
3.69 (m, 2H), 3.52-3.44 (m, 1H), 2.59 (dd, 1H, J ) 16.8, 3.6 Hz),
2.52-2.28 (m, 5H), 2.18-2.07 (m, 3H), 2.01-1.93 (m, 2H), 1.83-
1.67 (m, 5H), 1.49-1.45 (m, 1H), 1.32-1.11 (m, 5H); 13C NMR (75
MHz/CDCl3) δ 179.1, 172.1, 163.7, 152.7, 122.1, 65.0, 59.5, 57.5, 57.3,
41.2, 31.0, 30.6, 30.2, 29.5, 29.0, 27.0, 26.2, 26.0, 25.9, 25.5; IR (neat/
NaCl) 3386 br, 3256, 2930, 2853, 1685, 1647, 1617, 1522, 1447, 1262,
1046, 736 cm-1; LRMS (FAB) m/e (relative intensity) 386 (M + 11,
29), 377 (M + 2, 24), 376 (M + 1, 100), 374 (14), 371 (15), 307 (10);
HRMS (FAB) m/e calcd for C29H30O4N3 (M + 1) 376.2236, found
376.2232.
(6S and 6R,9S,12S)-1-Aza-9-carboxy-4-cyclohexyl-2-oxo-3-(pyro-
glutamylamino)bicyclo[4.3.0]non-3-ene. To a solution of the depro-
tected alcohol (isomer a, 6S, 0.413 g, 1.10 mmol) in 3 mL of acetone
was added Jones reagent (about 1.1 mL, 1.43 mmol) dropwise at 0 °C.
The mixture was stirred at 0 °C for 0.5 h and then at room temperature
for 1.5 h. When complete, the reaction was diluted with MeOH, and
then the salt in the solution was removed by filtration through a plug
of glass wool. The filtrate was concentrated in Vacuo to afford a crude
oil, which was dissolved in water. Ammonium chloride was added in
order to saturate the solution, and then the mixture was extracted with
ethyl acetate. The organic layers were combined, dried over MgSO4,
filtered through a glass fritted funnel, and concentrated in Vacuo. The
crude acid was further purified by HPLC using 35% MeOH/H2O as
eluant to afford the acid (isomer a) as a white solid (0.263 g, 63%).
The spectral data for isomer a were as follows: 1H NMR (300 MHz/
CDCl3) δ 8.64 (s, 1H), 7.84 (s, 1H), 4.61-4.52 (m, 1H), 4.47-4.40
(m, 1H), 4.02-3.91 (m, 1H), 2.62-2.38 (m, 5H), 2.32-2.09 (m, 4H),
2.02-1.85 (m, 2H), 1.80-1.59 (m, 5H), 1.53-1.45 (m, 1H), 1.24-
1.05 (m, 5H); 13C NMR (75 MHz/CDCl3) δ 180.6, 174.0, 172.3, 162.1,
150.5, 121.6, 58.5, 57.5, 56.1, 41.2, 32.6, 30.3, 29.9, 29.7, 29.0, 28.5,
26.4, 26.3, 26.2, 26.0; IR (neat/NaCl) 3268 br, 2926, 2847, 1716, 1685,
1653, 1616, 1521, 1448, 1320, 1262, 728, 626 cm-1; LRMS (FAB)
m/e (relative intensity) 391 (M + 2, 7), 390 (M + 1, 27), 246 (8), 219
(7), 186 (9), 185 (100); HRMS (FAB) m/e calcd for C20H28O5N3 (M +
1) 390.2029, found 390.2042.
The spectral data for isomer b were as follows: 1H NMR (300 MHz/
CDCl3) δ 4.12-4.06 (m, 1H), 3.79 (A of AMX, 1H, JAM ) 10.2 Hz,
JAX ) 3.0 Hz), 3.64 (M of AMX, 1H, JMA ) 10.2 Hz, JMX ) 6.0 Hz),
3.59-3.53 (m, 2H), 2.37-2.30 (m, 1H), 2.33 (dd, 1H, J ) 16.8, 4.5
Hz), 2.11-1.92 (m, 3H), 1.89-1.66 (m, 6H), 1.57-1.53 (m, 1H), 1.41-
1.11 (m, 6H), 0.89 (s, 9H), 0.06 and 0.04 (two s, 6H); 13C NMR (75
MHz/CDCl3) δ 163.0, 129.4, 123.0, 62.8, 57.8, 57.7, 39.3, 31.2, 30.7,
29.2, 29.0, 26.5, 26.4, 26.3, 26.1, 25.9, 18.2, -5.4, -5.5; IR (neat/
NaCl) 33445, 3349, 2925, 2846, 1659, 1625, 1575, 1442, 1358, 1337,
1320, 1255, 1190, 1101, 1063, 1005, 936, 835, 775, 734, 663 cm-1
;
LRMS (FAB) m/e (relative intensity) 379 (M + 1, 12), 172 (14), 150
(100), 148 (61), 119 (64); HRMS (FAB) m/e calcd for C21H39O2N2Si
379.2780 (M + 1), found 379.2775.
(6S and 6R,9S,12S)-1-Aza-9-[(tert-butyldimethylsiloxy)methyl]-
4-cyclohexyl-2-oxo-3-(pyroglutamylamino)bicyclo[4.3.0]non-3-
ene. To a solution of 15a (2.304 g, 6.09 mmol) in 30 mL of CH2Cl2
were added HOBt (1.070 g, 8.53 mmol) and L-pyroglutamic acid (0.944
g, 7.31 mmol). The mixture was stirred at 0 °C for 5 min, and then
EDCI (1.398 g, 7.92 mmol) was added. The resulting mixture was
allowed to warm to room temperature and stirred overnight. When
complete, the solution was transferred to a separatory funnel, the layers
were separated, and the organic phase was washed with saturated
NaHCO3, 10% citric acid in water, and a saturated solution of aqueous
NaCl. The CH2Cl2 was removed under reduced pressure, and the crude
product was chromatographed through 150 g of silica gel using 7%
MeOH in ether to afford 1.253 g (35%) of the tripeptide (isomer a) as
a white solid along with 0.413 g (14%) of the silyl-deprotected amide.
The spectral data for the product from 15a were as follows: 1H NMR
(300 MHz/CDCl3) δ 7.97 (s, 1H), 7.02 (s, 1H), 4.30 (dd, 1H, J ) 8.7,
5.1 Hz), 4.15-4.08 (m, 1H), 3.90-3.72 (m, 1H), 3.87 (A of ABX,
1H, JAB ) 9.9 Hz, JAX ) 4.8 Hz), 3.74 (B of ABX, 1H, JBX ) 9.9 Hz,
JBX ) 3.0 Hz), 2.61-2.47 (m, 4H), 2.42-2.21 (m, 4H), 2.16-1.95
(m., 3H), 1.91-1.70 (m, 4H), 1.64-1.49 (m, 2H), 1.32-1.13 (m, 5H),
0.87 (s, 9H), 0.04 and 0.01 (two s, 6H); 13C NMR (75 MHz/CDCl3) δ
178.7, 171.2, 161.3, 148.3, 121.2, 62.6, 58.8, 57.3, 56.1, 41.3, 32.0,
30.6, 30.2, 29.4, 28.9, 26.3, 26.2, 26.1, 26.0, 25.8, 18.1, -5.4, -5.5;
IR (neat/NaCl) 32.56 br, 2929, 2855, 1700, 1653, 1617, 1522, 1448,
1254, 1102, 837, 778, 735 cm-1; LRMS (FAB) m/e (relative intensity)
492 (M + 3, 10), 491 (M + 2, 35), 490 (M + 1, 100), 488 (11), 432
(14), 386 (8); HRMS (FAB) m/e calcd for C26H44O4N3Si (M + 1)
490.3101, found 490.3094.
15b was carried on in an identical fashion with a yield of 61%.
The spectral data for the product from 15b were as follows: 1H NMR
(300 MHz/CDCl3) δ 8.33 (s, 1H), 7.54 (s, 1H), 4.25 (dd, 1H, J ) 9.3,
5.1 Hz), 4.08-4.01 (m, 1H), 3.78-3.67 (m, 1H), 3.76 (A of ABX,
1H, JAB ) 10.2 Hz, JAX ) 5.4 Hz), 3.70 (B of ABX, 1H, JBA ) 10.2
Hz, JBX ) 2.4 Hz), 2.64-2.46 (m, 3H), 2.43-2.22 (m, 3H), 2.15-
1.99 (m, 3H), 1.94-1.68 (m, 6H), 1.50-1.47 (m, 1H), 1.34-1.13 (m,
5H), 0.89 (s, 9H), 0.03 and 0.01 (two s, 6H); 13C NMR (75 MHz/
CDCl3) δ 179.2, 172.1, 162.3, 151.3, 121.8, 62.7, 58.0, 57.8, 56.7, 41.3,
31.3, 30.4, 30.2, 29.4, 29.2, 27.0, 26.3, 26.1, 25.9, 18.2, -5.5; IR (neat/
NaCl) 3263 br, 2930, 2855, 1700, 1653, 1618, 1522, 1447, 1362, 1321,
1258, 1096, 1055, 838, 778, 735 cm-1; LRMS (FAB) m/e (relative
Isomer b (6R) was carried on in an identical fashion with a yield of
50%. The spectral data for isomer b were as follows: 1H NMR (300
MHz/CDCl3) δ 8.42 (s, 1H), 8.17 (s, 1H), 4.51-4.45 (m, 1H), 4.38-
4.32 (m, 1H), 3.90-3.78 (m, 1H), 2.58-2.38 (m, 6H), 2.25-2.11 (m,
5H), 1.80-1.61 (m, 5H), 1.58-1.51 (m, 1H), 1.32-1.07 (m, 5H); 13
C
NMR (75 MHz/CDCl3) δ 179.9, 173.6, 172.9, 161.6, 152.5, 122.7, 57.9,
57.6, 56.1, 40.5, 31.6, 30.5, 29.9, 29.7, 28.7, 28.3, 26.1, 25.8, 25.7,
25.6; IR (neat/NaCl) 3260 br, 2927, 2854, 1680, 1653, 1517, 1449,
1265, 1215, 733 cm-1; LRMS (FAB) m/e (relative intensity) 390 (M

10112 J. Am. Chem. Soc., Vol. 118, No. 42, 1996
Li and Moeller
+ 1, 21), 384 (19), 307 (100), 289 (52), 280 (15); HRMS (FAB) m/e
calcd for C20H28O5N3 (M + 1) 390.2029, found 390.2033.
) 8.4 Hz), 4.29 (dd, 1H, J ) 8.4, 4.8 Hz), 3.81-3.70 (m, 1H), 2.59-
2.29 (m, 6H), 2.26-1.97 (m, 4H), 1.94-1.54 (m, 6H), 1.30-1.10 (m,
5H); 13C NMR (75 MHz/CDCl3) δ 179.3, 174.4, 173.1, 162.5, 153.8,
122.3, 58.7, 57.1, 56.7, 40.9, 31.7, 30.7, 30.0, 29.7, 28.8, 28.2, 26.1,
26.0, 25.9, 25.8; IR (neat/NaCl) 3388 br, 2930, 2851, 1674, 1636, 1617,
1539, 1521, 1506, 1457, 1448, 1436, 1420, 1294, 1262, 667; LRMS
(FAB) m/e (relative intensity) 389 (M + 1, 100), 306 (59), 288 (33);
HRMS (FAB) m/e calcd for C20H29N4O4 (M + 1) 389.2189, found
389.2201.
(6S and 6R,9S,12S)-1-Aza-9-carboxamoyl-4-cyclohexyl-2-oxo-3-
(pyroglutamylamino)bicyclo[4.3.0]non-3-ene (10a,b). Isobutyl chlo-
roformate (79 mg, 0.58 mmol) was added to a solution of the acid
(isomer a) (224 mg, 0.58 mmol) in 6 mL of CH2Cl2, followed by
N-methylmorpholine (64 mg, 0.64 mmol) at 0 °C. The mixture was
stirred at 0 °C for 30 min, and then ammonia gas was passed through
it for 1 h. The reaction mixture was filtered through Celite, and the
filtrate was concentrated in Vacuo. The crude product was further
purified by HPLC using 50% MeOH/H2O as eluant to afford 200 mg
(90%) of 10a. The spectral data for 10a were as follows: 1H NMR
(300 MHz/CDCl3) δ 8.69 (s, 1H), 7.50 (s, 1H), 7.37 (s, 1H), 6.64 (s,
1H), 4.38 (dd, 1H, J ) 8.1, 5.1 Hz), 4.29 (t, 1H, J ) 7.8 Hz), 2.68-
2.10 (m, 10H), 2.04-1.91 (m, 1H), 1.78-1.57 (m, 5H), 1.32-1.12
(m, 5H); 13C NMR (75 MHz/CDCl3) δ 179.2, 175.1, 173.3, 162.0,
152.7, 121.9, 59.9, 56.9, 56.2, 40.9, 32.8, 30.5, 30.1, 29.8, 28.9, 28.5,
26.2, 26.0, 25.8; IR (neat/NaCl) 3457 br, 2925, 2850, 1701, 1670, 1636,
1617, 1457, 1449, 1419, 730, 657; LRMS (FAB) m/e (relative intensity)
389 (M + 1, 64), 306 (100), 288 (50), 233 (20); HRMS (FAB) m/e
calcd for C20H29N4O4 (M + 1) 389.2189, found 389.2207.
Acknowledgment. We thank the National Institutes of
Health (P01 GM24483-12) and Washington University for their
generous financial support. We also gratefully acknowledge
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 NIHRR00954, for their
assistance.
Supporting Information Available: Proton and carbon
NMR data for all new compounds (36 pages). See any current
masthead page for ordering and Internet access instructions.
Analog 10b was made in an identical fashion with a yield of 84%.
The spectral data for 10b were as follows: 1H NMR (300 MHz/CDCl3)
δ 8.69 (s, 1H), 7.80 (s, 1H), 7.39 (s, 1H), 6.41 (s, 1H), 4.35 (d, 1H, J
JA961887V