Conformationally constrained substance P analogues: The total synthesis of a constrained peptidomimetic for the Phe7-Phe8 region

Article abstract of DOI:10.1021/jo991649t

A lactam-based peptidomimetic for the Phe⁷-Phe⁸ region of substance P has been synthesized. The synthesis used an anodic amide oxidation to selectively functionalize the C₅-position of a 3-pheylproline derivative. The resulting proline derivative was coupled to a Cbz-protected phenylalanine, and an intramolecular reductive amination strategy used to convert the coupled material into a bicyclic piperazinone ring skeleton. The net result was a dipeptide building block that imbedded one of two proposed receptor bound conformations for the Phe⁷-Phe⁸ region of substance P into a bicyclic ring skeleton. The building block was then converted into a constrained substance P analogue with the use of solid-phase peptide synthesis. A similar intramolecular reductive amination strategy was used to synthesize a substance P analogue having only Phe⁷ constrained, and the original 3-phenylproline was converted into a substance P analogue having only Phe⁸ constrained. All of the analogues were examined for their ability to displace substance P from its NK-1 receptor.

Full text of DOI:10.1021/jo991649t

2484
J . Org. Chem. 2000, 65, 2484-2493
Con for m a tion a lly Con str a in ed Su bsta n ce P An a logu es: Th e Tota l
Syn th esis of a Con str a in ed P ep tid om im etic for th e P h e7-P h e8
Region
Yunsong Tong,† Yvette M. Fobian,† Meiye Wu,‡ Norman D. Boyd,‡ and Kevin D. Moeller*,†
The Department of Chemistry, Washington University, St. Louis, Missouri 63130, and
The Department of Pharmacology and Experimental Therapeutics Neuropeptide Laboratory,
Boston University School of Medicine, 80 East Concord Street, Boston, Massachusetts 02118-2394
Received October 21, 1999
A lactam-based peptidomimetic for the Phe7-Phe8 region of substance P has been synthesized. The
synthesis used an anodic amide oxidation to selectively functionalize the C5-position of a
3-phenylproline derivative. The resulting proline derivative was coupled to a Cbz-protected
phenylalanine, and an intramolecular reductive amination strategy used to convert the coupled
material into a bicyclic piperazinone ring skeleton. The net result was a dipeptide building block
that imbedded one of two proposed receptor bound conformations for the Phe7-Phe8 region of
substance P into a bicyclic ring skeleton. The building block was then converted into a constrained
substance P analogue with the use of solid-phase peptide synthesis. A similar intramolecular
reductive amination strategy was used to synthesize a substance P analogue having only Phe7
constrained, and the original 3-phenylproline was converted into a substance P analogue having
only Phe8 constrained. All of the analogues were examined for their ability to displace substance
P from its NK-1 receptor.
In tr od u ction
nists for substance P/NK1 binding have shown promise
as antidepressant drugs in clinical trials.3d
Ideally, constrained peptidomimetics for probing the
biological relevance of a proposed peptide conformation
can be designed by simply repacing spatially close
hydrogens in the proposed conformation with an ap-
propriate bridge. In this way, the backbone is imbedded
into a polycyclic ring skeleton that fixes the conformation
in place.1 However, such a transformation on paper often
leads to the suggestion of a peptidomimetic that is either
difficult or impossible to synthesize in the laboratory. For
this reason, we have been working to develop the
methodology needed to simplify the syntheses of a variety
of constrained amino acid building blocks.2 As part of this
effort, we have been building constrained mimetics for
the Phe7-Phe8 region of substance P. Substance P is an
11 amino acid peptide having the structure Arg1-Pro2-
Lys3-Pro4-Gln5-Gln6-Phe7-Phe8-Gly9-Leu10-Met11-NH2. It
is a member of the mammalian tachykinin family of
peptides and has been implicated in a number of disease
states including arthritis, asthma, inflammatory bowel
disease, and depression.3 Recently, high-affinity antago-
Marshall and Nikiforovich have proposed two low
energy conformations of substance P as potential models
for the binding of substance P to the NK1 receptor.4 These
two models were derived by determining the low energy
conformations available to both substance P and a series
of substance P analogues having a high affinity for the
NK1 receptor. Interestingly, the two proposed conforma-
tions differ greatly with respect to the conformation of
the critical Phe7-Phe8 region of the peptide (Scheme 1, 1
and 2).5 This difference suggested that the Phe7-Phe8
region of substance P might provide a handle for begin-
ning to assess the relevance of the two proposed models.
For this reason, a pair of constrained peptide building
blocks (3 and 4) for fixing the Phe7-Phe8 region of
substance P into the desired conformations were de-
signed. But, how could these building blocks and the
corresponding substance P analogues be synthesized, and
would the bridges interfere with the binding affinity and
potency of the analogues constructed? We report here the
synthesis of the first of these substance P analogues (3)
and the initial evaluation of how each added constraint
affects the activity of the analogue.
† Washington University.
‡ Boston University School of Medicine.
Th e Con str u ction of a Bicyclic P ip er a zin on e
(1) For selected examples of this approach to other systems, see:
(a) Wolf, J . P.; Rapoport, H. J . Org. Chem. 1989, 54, 3164 as well as
refs 1-8 therein. (b) Freidinger, R. M.; Perlow, D. S.; Veber, D. F. J .
Org. Chem. 1982, 47, 104. (c) Freidinger, R. M. J . Org. Chem. 1985,
50, 3631. (d) Zydowsky, T. M.; Dellaria, J . F., J r.; Nellans, H. N. J .
Org. Chem. 1988, 53, 5607 (e) Kempf, D. J .; Condon, S. L. J . Org. Chem.
1990, 55, 1390. (f) Kemp, D. S.; McNamara, P. E. J . Org. Chem. 1985,
50, 5834 and references therein. (g) Nagai, U.; Sato, K. Tetrahedron
Lett. 1985, 26, 647, (h) Hinds, M. G.; Richards, N. G. J .; Robinson, J .
A. J . Chem. Soc., Chem. Commun. 1988, 1447, (i) Paul, P. K. C.;
Burney, P. A.; Campbell, M. M.; Osguthorpe, D. J . Bioorg. Med. Chem.
Lett. 1992, 2, 141, (j) Lombart, H. G.; Lubell, W. D. J . Org. Chem. 1996,
61, 9437.
Bu ild in g Block for Con st r a in in g Bot h P h e7 a n d
(3) (a) Payan, D. G. Annu. Rev. Med. 1989, 40, 341. (b) Ichinose,
M.; Nakajima, N.; Takahashi, T.; Yamauchi, H.; Inote, H.; Takishima,
T. Lancet 1992, 340, 1248. (c) Maggi, C. A.; Patacchini, R.; Roverro,
P.; Giachetti, A. J . Auton. Pharmacol. 1993, 13, 23. (d) Kramer, M. S.
et. al. Science 1998, 281, 1640.
(4) Private communication from Nikiforovich, G. V. and Marshall,
G. R. The methods for the calculation have been published: (a)
Nikiforovich, G. V. THEOCHEM 1986, 134, 325. (b) Nikiforovich, G.
V. Int. J . Peptide Protein Res. 1994, 44, 513.
(2) (a) Long, R. D.; Moeller, K. D. J . Am. Chem. Soc. 1997, 119,
12394, (b) Li, W.; Moeller, K. D. J . Am. Chem. Soc. 1996, 118, 10106.
(c) Rutledge, L. D.; Perlman, J . H.; Gershengorn, M. C.; Marshall, G.
R.; Moeller, K. D. J . Med. Chem. 1996, 39, 1571.
(5) (a) Fobian, Y. M.; d’Avignon, D. A.; Moeller, K. D. Bioorg. Med.
Chem. Lett. 1996, 6, 315. (b) Fobian, Y. M.; Moeller, K. D. In Methods
in Molecular Medicine: Peptidomimetic Protocols; Kazmierski, W. M.,
Ed.; Humana Press Inc.: Totowa, NJ , 1998; p 259.
10.1021/jo991649t CCC: $19.00 © 2000 American Chemical Society
Published on Web 03/23/2000

Conformationally Constrained Substance P Analogues
J . Org. Chem., Vol. 65, No. 8, 2000 2485
Sch em e 1
Sch em e 3
F igu r e 1.
Sch em e 2
reaction. Fortunately, the anodic oxidation reaction was
not a problem (Scheme 3). The substrate for the elec-
trolysis was obtained by protecting the nitrogen of the
3(R)-phenylproline derivative (9) with a t-Boc group. The
carbamate was then subjected to oxidation using a carbon
rod anode, a 0.03 M Et4NOTS in methanol electrolyte
solution, and a constant current of 26.8 mA (Scheme 3).
A total of 3.0 F/mol was passed. Using these conditions,9
the oxidation could be run on a multigram scale and led
to the formation of an 87% isolated yield of the meth-
oxylated product 10.
P h e8 of Su bsta n ce P . Work on the synthesis of con-
strained substance P analogues began with the develop-
ment of an approach for constructing the bicyclic piper-
azinone ring skeleton needed for analogue 3.5 This
approach (outlined in Scheme 2) utilized an electrochemi-
cal amide oxidation to selectively functionalize proline
derivative 5.6 The resulting methoxylated product 6 was
treated with BF3‚Et2O and the cuprate derived from
2-methylpropenyllithium to form a 5-vinyl-substituted
proline. The Boc group was then removed and the
resulting amine coupled to a Cbz protected phenylalanine
to form 7. The olefin in 7 was cleaved using an ozonolysis
reaction and the subsequent aldehyde treated with
reductive amination conditions to complete the desired
ring system. The synthesis could be varied to employ
either an allyl-substitued proline (derived from 6) or an
N-terminal â-amino acid in order to construct seven- and
eight-membered ring lactam derivatives of the originally
proposed ring system.
In principle, one of the strengths of this approach was
that it could potentially allow for the use of a wide variety
of existing amino acid starting materials. For example,
the synthesis of a building block for making substance P
analogues such as 3 would involve starting with the
known 3(R)-phenyl-substituted proline 9.7 However, the
addition of the phenyl ring was worrisome. The anodic
oxidation of 3-substituted proline derivatives had not
been studied, and previous amide oxidation reactions had
proven to be sensitive to the presence of phenyl substit-
uents.8 In addition, it was not known how the presence
of the phenyl ring would affect the cuprate addition
Having successfully methoxylated the amide, attention
was turned to the addition of a vinyl group to the
substituted proline.10 Initially, the N-acyliminium ion
generated from 10 was treated with the vinyl cuprate
derived from 2-methylpropenyllithium in analogy to
Scheme 2. However, this reaction was not successful, and
none of the desired 5-vinyl-substituted proline was
obtained. Clearly, the phenyl ring interfered with the
incoming nucleophile. This was surprising since earlier
cuprate additions to N-acyliminium ions derived from
proline had not been hindered by substituents, even when
a substituent was placed at C4 of the proline cis to the
incoming nucleophile at C5.10b One suggestion for ex-
plaining the result obtained from 10 is illustrated in
Figure 1, where a complex between a copper atom, the
methyl ester of the proline ring, and the double bond of
the iminium ion would force the nucleophile to approach
C5 from the face of the ring opposite that of the methyl
ester. This model for the reaction was initially forwarded
by Wistrand and co-workers10a and has been consistent
with the stereochemical observations made for a variety
of related reactions.10b The fact that in the current
example the addition was blocked by the phenyl ring
argued for a scenario where the nucleophile approached
the iminium ion from “over the ring”. In this way, the
allylic methyl group cis to the anion would encounter the
phenyl ring leading to a large steric interaction that
blocked the favored approach to the iminium ion. The
net result was an unsuccessful addition reaction.
We reasoned that if this suggestion was correct, then
the addition reaction would benefit from the use of the
(6) For pioneering work, 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.; Matsumura, Y.; Tsubata, K. In Organic
Synthesis; Saucy, G., Ed.; Organic Synthesis Inc.: New York, 1984;
Vol. 63, p 206 and references therein. (e) Shono, T. In Topics in Current
Chemistry; Steckhan, E. Ed.; Springer-Verlag: Berlin, Heidelberg, New
York, 1988; Vol. 148, p 131. (f) ref 1b and references therein.
(7) Chung, J . Y. L.; Wasicak, J . T.; Arnold, W. A.; May, C. S.;
Nadzan, A. M.; Holladay, M. W. J . Org. Chem. 1990, 55, 270.
(8) (a) Barrett, A. G. M.; Pilipauskas, D. J . Org. Chem. 1991, 56,
2787. (b) Li, W.; Hanau, C.; d’Avignon, A.; Moeller, K. D. J . Org. Chem.
1995, 60, 8155. (c) Fobian, Y. M. Ph.D. Thesis, Washington University,
1996. (d) Moeller, K. D.; Wang, P. W.; Tarazi, S.; Marzabadi, M. R.;
Wong, P. L. J . Org. Chem. 1991, 56, 1058.
(9) (a) Wong, P. L.; Moeller, K. D. J . Am. Chem. Soc. 1993, 115,
11434. (b) Alternatively, a 6 V lantern battery can be used for exploring
a
new electrolysis reaction: Frey, D. A.; Wu, N.; Moeller, K. D.
Tetrahedron Lett. 1996, 37, 8317.
(10) Wistrand, L.-G.; Skrinjar, M. Tetrahedron 1991, 47, 573. (b)
Collado, I.; Ezquerra, J .; Pedregal, C. J . Org. Chem. 1995, 60, 5011.

2486 J . Org. Chem., Vol. 65, No. 8, 2000
Tong et al.
Sch em e 4
Sch em e 6a
Sch em e 5
a
Reagents: (a) (i) O3, MeOH, CH2Cl2, (ii) NaBH4, 75%; (b)
o-NO2-Phe-SeCN, (n-Bu)3P, THF, 72%; (c) (i) m-CPBA, CH2Cl2,
-78 °C, (ii) Me2S, Et3N, -78 °C to rt, 92%.
have found methoxylated amides such as 15 with an
electron-withdrawing group on the carbon bearing the
methoxy group do not afford iminium ions upon treat-
ment with BF3‚Et2O.12 Stronger Lewis acids are required
in these cases. In the current example, the inability to
regenerate iminium ion 14 from 15 with BF3‚Et2O meant
that the equilibrium was “drained” toward the formation
of 15. Hence, longer reaction times increased the yield
of 15 relative to the desired addition product 13. Finally,
no racemization at C2 was observed in the desired vinyl-
substituted product even when large amounts of the
2-methoxy derivative 15 were isolated. This result indi-
cated that the addition of methanol to C2 was nonrevers-
ible and fast enough so that an equilibrium was not
established between the two iminium ions 12 and 14.
While this rearrangement reaction was less of a problem
with the vinyl anion used in Scheme 4, the reaction times
for the process were kept short in order to avoid the
problem entirely.
In contrast to cuprate additions where the presence of
a 3-phenyl substituent did not alter the stereochemistry
of the product, the 3-phenyl substituent did control the
stereochemistry of an allylsilane addition to the N-
acyliminium derived from 10. When 10 was treated with
BF3‚Et2O and allyltrimethylsilane, the nucleophile at-
tacked the N-acyliminium ion from the R-face of the
molecule. The product (16) having the allyl substituent
trans to the phenyl ring and cis to the methyl ester at
C2 was formed (Scheme 6). None of the product having
the opposite stereochemistry at C5 of the proline ring was
obtained. The stereochemistry of product 16 was assigned
as being R at C5 with a NOESY experiment that showed
a cross-peak between the benzylic methine at C3 and one
of the allylic methylene protons. No cross-peak was
observed for the methine at C3 and the methine at C5.
These two observations indicated that the allyl group was
on the same face of the molecule as the methine at C3,
and trans to the phenyl group.
anion generated from trans-1-bromopropene. In this case,
the nucleophile would have a proton cis to the anion and
the size of the steric interaction with the phenyl ring
would be greatly reduced. Since the synthesis called for
a subsequent cleavage of the olefin, the change in
substitution of the olefin product would not matter. In
practice, the use of a cuprate derived from trans-1-
bromopropene led to a 75% isolated yield of the desired
addition product (Scheme 4). In addition, a small amount
of a second isomer, assumed to be the cis product, was
obtained. At this point, the isomers could not be sepa-
rated. The stereochemistry of the major product was
assigned with the use of 2-D NOE data after 11 was
converted into the bicyclic building block (Scheme 7).11
This assignment is discussed below. The formation of the
5(S) isomer in this reaction was consistent with the
previous synthesis of 8 and the picture for the addition
illustrated in Figure 1. It was clear from this work that
the cuprate addition could not be made to afford a product
where the incoming vinyl group and the methyl ester
group wound up cis to each other. Depending on the size
of the steric interaction, either the reactions afforded a
high yield of the trans product, or the addition reaction
failed.
On a final note, the short reaction time for the cuprate
addition (15 min) was important. Longer reaction times
led to the generation of a 2-methoxyproline-derived side
product. The formation of this side product was consistent
with what was observed for the cuprate-based addition
of 2-methylpropene to the N-acyliminium ion derived
from 6 (Scheme 5). In this case, short reaction times led
to high yields of the vinyl-substituted product (13). When
the reaction was run for 15 min, an 89% isolated yield of
13 was obtained.5b However, after 20 min the yield of the
product decreased to 65-70%. This decrease in yield was
accompanied by an increase in the formation of the
2-methoxyproline byproduct 15. After 30 min, a 54% yield
of the vinyl substituted product 13 was obtained along
with 41% of 15. Apparently, the formation of the 5-vinyl-
substituted product 13 was reversible. Given enough
time, the iminium ion isomerized from 12 to 14 and then
trapped methoxide at C2 to generate 15. In the past, we
The formation of 16 suggested that the chemistry being
developed could be used to generate bicyclic piperazinone
building blocks with either S or R bridgehead stereo-
chemistry. For this reason, the 5-allylproline derivative
16 was converted into the (5R)-vinylproline derivative 17
by cleaving the double bond of the allyl group with an
(12) Unpublished results with Mr. Lawrence D. Rutledge. In this
example, the treatment of a 5-methoxypyroglutamate derivative with
BF3‚Et2O led to complete recovery of the starting material.
(11) See, for example: Kao, J .; Li, W.; Moeller, K. D. Magn. Reson.
Chem. 1997, 35, 267.

Conformationally Constrained Substance P Analogues
J . Org. Chem., Vol. 65, No. 8, 2000 2487
Sch em e 7a
F igu r e 2.
Sch em e 8a
a
Reagents: (a) BF3‚Et2O, THF; (b) Cbz-Phe-F, N-ethylmorpho-
line, CH2Cl2, yield 18a ) 60% over two steps, yield 18b ) 73%
over two steps; (c) (i) O3, MeOH, CH2Cl2, -78 °C, (ii) Me2S, -78
°C to rt, yield from 18a ) 84%, yield from 18b ) 99%; (d) H2, Pd
on BaSO4, MeOH, yield for 19a ) 75%, yield for 19b ) 66%; (e)
Boc2O, N-ethylmorpholine, 83%; (f) LiOH, THF/MeOH/H2O, 62%.
a
Reagents: (a) allyl bromide, Et3N, DMF, 60%; (b) t-Boc-Phe-
F, NEM, CH2Cl2, 75%; (c) (i) O3, MeOH, -78 °C to rt, (ii) Me2S,
88%; (d) Et3SiH, TFA (15 equiv), CH2Cl2, 89%; (e) Boc2O, NEM,
83%; (f) LiOH, THF/MeOH/H2O, 72%.
ozonolysis reaction, reducing the resulting aldehyde to
an alcohol, and then eliminating the alcohol to form the
vinyl substituent.
designed analogue,4 the bicyclic piperazinone was con-
verted into a building block for solid-phase peptide
synthesis (20). This transformation was accomplished by
reprotecting the proline nitrogen with a t-Boc group and
saponifying the methyl ester with lithium hydroxide
(Scheme 7).
Bu ild in g Block s for Con str a in in g Eith er P h e7 or
P h e8 of Su bsta n ce P . Along with the bicyclic building
block (20) needed for making a substance P analogue
with both Phe7 and Phe8 constrained, monocyclic building
blocks were desired for individually constraining either
the Phe7 or Phe8 region of substance P. These analogues
were needed so that the effects of each individual bridge
could be determined.
For constraining Phe8, the known t-Boc-protected pro-
line derivative 21 could be used (Figure 2).7 However,
the monocylcic piperazinone building block (22) required
for constraining the Phe7 group of substance P repre-
sented a larger synthetic challenge. Fortunately, the
intramolecular reductive amination route used for con-
structing bicyclic building block 20 also proved useful for
the synthesis of 22 (Scheme 8). This synthesis started
with the selective monoallylation of the phenylalanine‚
HCl salt. To our suprise, this reaction could be ac-
complished by directly treating phenylalanine with al-
lylbromide in the presence of triethylamine to afford 60%
of the monoallylated amine along with only 5% of the
doubly allylated byproduct. Attempts to monoallylate the
t-Boc-protected phenylalanine using a variety of reaction
conditions were not as successful. Once the allylated 24
was obtained, it was coupled to a second, t-Boc-protected
phenylalanine group in a 75% yield by activating the
carboxylic acid with cyanuric fluoride. We found that the
use of an acid fluoride as the activated acid derivative
was essential for this transformation.13 For example, the
use of EDC, HOBt conditions led to only recovered
starting material and no coupling product.
Both complimentary vinyl-substituted prolines (11 and
17) were converted into bicyclic piperazinone derivatives
by removing the t-Boc protecting group and coupling the
resulting amine to a Cbz-protected phenylalanine using
the acid fluoride derivative of the acid (Scheme 7).13 In
each case, the vinyl substituent was cleaved using an
ozonolysis reaction. Following workup with dimethyl
sulfide, both molecules spontaneously cyclized to form a
six-membered ring hemiaminal. These products were
reduced with hydrogen and Pd on barium sulfate in
methanol in order to complete the synthesis of the
bicyclopiperazine building blocks (19a and 19b). The Cbz
group on the N-terminus was removed during this step.
The in situ removal of the Cbz group proved important
for completion of the reductive amination reaction. When
an analogous synthesis was attempted using a t-Boc-
protected phenylalanine, the reductive amination reac-
tion led to recovery of the cyclic N-R-hydroxyalkyl amide
product from the ozonolysis without subsequent reduction
by hydrogen.
At this point, the stereochemistry of the bridgehead
carbon in 19a was determined with the use of a NOESY
experiment. While the stereochemistry could be estab-
lished by systematically “walking” around the five-
membered ring, two key long-range interactions were
especially helpful. First, an NOE cross-peak for the
interaction between the bridgehead methine at C6 and
the benzylic methine at C8 was present. Second, a strong
NOE cross-peak between the C5 methylene proton on the
â-face of the molecule and the C7 methylene proton on
the â-face of the molecule was present along with NOE
cross-peaks for the interactions between the bridgehead
methine at C6 and the C5 and C7 methylene protons on
the R-face of the molecule. This collection of NOE cross-
peaks was only consistent with 19a having the bridge-
head methine proton on the R-face of the molecule.
Having established that 19a possesed the stereochem-
istry needed to complete the synthesis of the initially
Ozonolysis of 25 generated an aldehyde that under-
went a spontaneous cyclization to form an N-R-hydroxy-
alkylamide in an 88% yield. The reductive amination was
then completed with the use of triethylsilane and 15
equiv of TFA in dichloromethane. These conditions also
(13) Vojkovsky, T. Peptide Res. 1995, 6, 316.

2488 J . Org. Chem., Vol. 65, No. 8, 2000
Tong et al.
Sch em e 9
Under these binding assay conditions, the unmodified
peptide Substance P exhibited an IC50 of 0.3 nM, indicat-
ing that the modification in the analogues reduced their
affinity by approximately 4 orders of magnitude. Because
of the low affinity of the analogues, the relative differ-
ences between their binding affinities may not be sig-
nificant.
Sch em e 10
The most surprising of these results was the poor
affinity of analogue 30 for the NK-1 receptor. This result
was not anticipated because the 3-phenylproline used to
constrain Phe8 in this analogue was proposed for both
model 1 and model 2 and because sterically bulky Phe8
replacements have been compatible with the binding of
an analogue to NK-1 in the past.16 At this point, it is not
clear whether the bridge sterically interfered with bind-
ing to the NK-1 receptor or whether the phenyl ring was
not located in a position compatible with binding. What
is clear is that until a suitable mimic for the Phe8 region
is found future work to examine the comformation of the
Phe7-Phe8 region of substance P responsible for binding
NK-1 will need to be confined to analogues having just
the Phe7 constraint. The result obtained for the bicyclic
analogue 29 was consistent with this observation, al-
though other factors such as the wrong bridgehead
stereochemistry could also contribute to the lack of
analogue 29/ NK-1 binding as well.
The poor affinity of constrained analogue 31 for the
NK-1 receptor was most likely due to one (or a combina-
tion) of three problems. Either model 1 does not represent
a viable backbone conformation for the bound Phe7
region, the Phe7 side chain in the analogue is not
positioned correctly for binding to the receptor, or the
added constraint in 31 sterically interferes with the
analogues ability to bind the receptor. Clearly, a Phe7
building block that constrains this region into a confor-
mation consistent with model 2 for substance P/ NK-1
binding is needed in order to begin resolving this issue.
deprotected the N-terminal end of the amino acid. A
reduction in the amount of TFA used did not lead to a
reductive amination product having the t-Boc protecting
group intact. For example, when 2 equiv of TFA was
used, the reaction failed to produce any of the reductive
amination product. Instead, the corresponding enamide
28 was formed (Scheme 9). Eventually, a route was
settled on that used a large excess of TFA and then
reprotected the N-terminal amino acid (Scheme 8). Fol-
lowing the reprotection, the methyl ester was saponified
to complete the synthesis of building block 22.
Su bsta n ce P An a logu e Syn th eses. Once the build-
ing blocks were constructed, attention was turned toward
synthesizing substance P analogues 29-31 (Scheme 10).
In each case, the full substance P analogue was con-
structed using solid-phase peptide synthesis techniques.
The solid support chosen was the MBHA (p-meth-
ylbenzhydrylamine‚HCl). The monomers were protected
on the N-terminus with a t-Boc group. Methionine,
arginine, glutamine, and lysine monomers were “perma-
nently” protected by oxide, 4-toluenesulfonyl (Tos), 9-xan-
thenyl (Xan), and 2-chlorobenzyloxycarbonyl (2-ClZ),
respectively. During the synthesis, the t-Boc groups were
removed with 50% TFA/dichloromethane. The coupling
reactions were accomplished using the standard TBTU,
HOBT, and DIEA conditions. Further details concerning
the solid-phase peptide synthesis are given in the Ex-
perimental Section. After completion of each substance
P analogue, thiophenol was used to remove the oxide
protecting group on the methionine moiety. The remain-
ing protecting groups were removed while cleaving the
peptide from the resin with HF. The three substance P
analogues were then purified using HPLC.
Con clu sion s
The synthetic methodology needed to rapidly construct
both monocyclic and bicyclic piperazinone building blocks
for use in the construction of constrained peptidomimetics
has been developed. The analogues have been used to
construct substance P analogues with conformational
constraints in the Phe7Phe8 region. In both cases, an
intramolecular reductive amination reaction was utilized
to complete the lactam ring used as the conformational
constraint. In the case of the bicyclic building blocks, the
intramolecular reductive amination reaction was used in
a sequential strategy with an anodic amide oxidation to
effect the net annulation of the piperazinone ring onto a
proline derivative. The building blocks synthesized proved
to be compatible with solid-phase peptide synthesis and
were used to construct three constrained substance P
analogues. None of the constrained analogues synthe-
sized proved to be high affinity analogues for the NK-1
receptor, although analogue 31 did inhibit the binding
of substance K to the receptor.
Equ ilibr iu m Disp la cem en t Bin d in g of th e An a -
logu es.14 All three substance P analogues were assayed
for their ability to compete with radioiodinated SP
binding to the NK-1 receptor expressing in CHO cells.15
Analogues 29-31 inhibited binding only at high concen-
trations with IC50 values of 32, 80, and 5 µM, repectively.
Finally, the biological data obtained identified a prob-
lem with the proposed constraint for the Phe8 moiety of
substance P and has triggered the search for a more
(14) The hard data are included in the Supporting Information.
(15) Kage, R.; Hershey, A. D.; Krause, J . E.; Boyd, N. D.; Leeman,
S. E. J . Neurochem. 1995, 64, 316.
(16) J osien, H.; Lavielle, S.; Brunissen, A.; Saffroy, M.; Torrens, Y.;
Beaujouan, J . C.; Glowinski, J .; Chassng, G. J . Med. Chem. 1994, 37,
1586.

Conformationally Constrained Substance P Analogues
J . Org. Chem., Vol. 65, No. 8, 2000 2489
52.0, 47.4, 46.6, 41.9, 40.6, 28.3, 28.1; FAB MS m/z (rel
intensity) 334 (M - 1, 11), 305 (20), 304 (M+ - OCH3, 100),
276 (15), 260 (10), 250 (12), 235 (11), 234 (62), 220 (20). Due
to stability issues, this product was carried on without further
characterization.
effective constraint for this region. In analogy to the work
described above, once a more effective constraint has been
identified, the anodic amide oxidation - intramolecular
reductive amination strategy should allow for the rapid
synthesis of the corresponding bicyclic analogues directly
from the constrained Phe8 building block.
(3R,5S)-N-ter t-Bu tyloxyca r bon yl-3-p h en yl-5-(tr a n s-1-
p r op en yl)-L-p r olin e Meth yl Ester (11). In a flame-dried 25
mL flask under argon 150 mg (7.5 mmol) of lithium as a 30%
dispersion in mineral oil was washed three times with hexane.
Anhydrous ether (6 mL) was added, and the mixture was
cooled to -20 °C. To this solution was added 359 mg (3.0 mmol)
of trans-1-bromo-1-propene. After 2 h at -20 °C, the solution
was cannulated into a 50 mL flask containing a heterogeneous
mixture of 613 mg (3.0 mmol) of copper(I) bromide-dimethyl
sulfide complex in 7.5 mL of anhydrous ether at -40 °C. The
resulting dark brown solution was stirred at -40 °C for 1 h,
cooled to -78 °C, and then treated with 424 mg (3.0 mmol) of
boron trifluoride etherate. After 5 min, 501 mg (1.5 mmol) of
10 was added and the -78 °C bath removed. After 15 min,
the reaction mixture was quenched with a 1:1 solution of
ammonium hydroxide and saturated ammonium chloride and
diluted with dichloromethane. The aqueous layer was ex-
tracted three times with dichloromethane. The organic layers
were combined, dried over magnesium sulfate, and concen-
trated in vacuo. The crude product was chromatographed
through 15 g of silica gel that was slurry packed with a 1:1
hexane in ether solution. Elution with the same solvent
afforded 385 mg (75%) of the desired product (11). A small
amount of a second stereoisomer (tentatively assigned as the
cis compound) was also observed. For the trans isomer, the
spectral data for the mixture of carbamate rotamers were as
Exp er im en ta l Section 17
(3R)-N-ter t-Bu tyloxycar bon yl-3-ph en yl-L-pr olin e Meth -
yl Ester . In a 1 L flask was placed 8.13 g (24.2 mmol) of 2-R-
methylbenzylamide-trans-3-phenyl-L-proline6 along with 61
mL of glacial acetic acid and 245 mL of 8 N hydrochloric acid.
The mixture was then heated at reflux for 20 h. When
complete, the reaction mixture was concentrated in vacuo and
the residue taken up in water and washed once with ethyl
acetate. The aqueous layer was concentrated in vacuo to afford
the trans-3-phenyl-L-proline hydrochloride. This crude product
was dissolved in 50 mL of anhydrous methanol and saturated
with hydrogen chloride gas. After 18 h, the reaction mixture
was concentrated in vacuo. The residue was taken up in 1 N
hydrochloric acid and washed twice with ethyl acetate. The
aqueous layer was adjusted to pH 10 with 1 N sodium
bicarbonate and then extracted three times with chloroform.
The combined chloroform layers were dried over magnesium
sulfate and concentrated in vacuo to afford the trans-3-phenyl-
L-proline methyl ester 9. The crude 9 was dissolved in 100 mL
of dichloromethane with 4.89 g (42.5 mmol) of N-ethylmor-
pholine and cooled to 0 °C. To this mixture was added 4.61 g
(21.1 mmol) of di-tert-butyl dicarbonate. After gas evolution
had ceased, the reaction mixture was warmed to room tem-
perature, stirred for 18 h, and then washed with 0.5 M citric
acid, 5% sodium bicarbonate, and water. The organic layer was
dried over magnesium sulfate and concentrated in vacuo. The
crude product was chromatographed through 225 g of silica
gel that was slurry packed with 10% hexane in ether. Elution
with the same solvent afforded 5.75 g of the desired product,
which was contaminated with some di-tert-butyl dicarbonate.
The yield corrected for the contamination was 78%. The
spectral data for the mixture of carbamate rotomers were as
follows: 1 H NMR (300 MHz, CDCl3) δ 7.48-7.20 (m, 5H), 4.38
(d, J ) 6.1 Hz, 0.33H), 4.25 (d, J ) 6.8 Hz, 0.66H), 3.75 (m,
0.5H), 3.70 (s, 3H), 3.61 (m, 1.5H), 3.44 (m, 1H), 2.38-2.24
(m, 1H), 2.10-1.96 (m, 1H), 1.49 (s, 3H), 1.42 (s, 6H);13C NMR
(75 MHz, CDCl3) δ 173.2, 153.6, 140.9, 140.5, 128.8, 127.2,
126.9, 80.2, 65.8, 65.1, 52.2, 52.0, 49.9, 48.7, 46.1, 46.0, 33.0,
32.3, 28.4, 28.2; FAB MS m/z (rel intensity) 306 (M + 1, 25),
304 (M - 1, 14), 289 (8), 279 (7), 251 (17), 250 (100), 246 (M+
- CO2CH3, 10); HRMS FAB calcd for C17H24NO4 (M + 1)
306.1705, found 306.1708.
1
follows: H NMR (300 MHz, CDCl3) δ 7.30 (m, 5H), 5.69-
5.23 (bm, 2H), 4.52-4.33 (m, 2H), 3.70 and 3.69 (two s, 3H),
3.38 (m, 1H), 2.49 (m, 1H), 1.94 (m, 1H), 1.65 (bs, 3H), 1.42
and 1.40 (two bs, 9H); FAB MS m/z (rel intensity) 346 (M + 1,
14), 291 (13), 290 (70), 247 (17), 246 (100), 245 (19), 244 (M+
- (CH3)3COCO, 46), 230 (31); HRMS FAB calcd for C20H28
-
NO4 (M + 1) 346.2018, found 346.2011. In this case, the rate
of equilibration between the carbamate rotomers was such that
it was difficult to determine the purity of the product by NMR.
For this reason, the product was deprotected and converted
to 18a prior to complete characterization.
(3R,5R)-N-ter t-Bu t yloxyca r b on yl-3-p h en yl-5-a llyl-L-
p r olin e Meth yl Ester (16). To a -40 °C solution of 3.915 g
(11.7 mmol) of 10 in 24 mL of ether was added 8.35 mL (53
mmol, 4.5 equiv) of allyltrimethylsilane and 1.48 mL (12 mmol,
1.03 equiv) of BF3‚Et2O. The cold bath was removed after the
addition, and the reaction was stirred for 30 min. The resulting
mixture was diluted with 20 mL of ether and extracted with
20 mL of NaHCO3 solution. The aqueous layer was washed
two more times with ether. The combined organic layers were
dried over Na2SO4 and concentrated. The crude product (3.9
g) was chromatographed through 180 g of silica gel using ether/
hexane (1:1) as the eluant to afford 3.093 g (77%) of pure
product as a colorless oil. The spectral data for the mixture of
carbamate rotamers were as follows: 1H NMR (300 MHz,
CDCl3) δ 7.34-7.21 (m, 5H), 5.89-5.79 (m, 1H), 5.16-5.05 (m,
2H), 4.37 (d, 0.34H, J ) 8.3 Hz), 4.22 (d, 0.66H, J ) 8.9 Hz),
4.17-4.10 (m, 0.66H), 4.04-4.01 (m, 0.34H), 3.68 (s, 3H), 3.56-
3.47 (m, 1H), 2.82-2.66 (m, 1H), 2.40-2.30 (m, 1H), 2.17-
2.13 (m, 2H), 1.49 (s, 3H), 1.41 (s, 7H); 13C NMR (125 MHz,
CDCl3) δ 173.3, 153.3, 139.6, 128.7, 127.3, 127.2, 117.2, 80.3,
66.8, 66.1, 57.9, 57.8, 52.1, 51.9, 47.9, 47.1, 39.2, 38.4, 38.1,
36.7, 28.4, 28.3; IR (neat/NaCl) 3072, 1750, 1701 cm-1; LRFAB
MS m/e (rel intensity) 346 (MH+, 24), 290 (MH+ - C4H8, 46),
246 (MH+-C5H8O2, 89), 204 (MH+-C8H12O2, 100); HRFAB
calcd for C20H28NO4 (M + 1) 346.2018, found 346.2017. Anal.
Calcd for C20H27N1O4: C, 69.54; H, 7.88. Found: C, 69.50; H,
7.99.
(3R)-N-ter t-Bu t yloxyca r b on yl-5-m et h oxy-3-p h en yl-L-
p r olin e Meth yl Ester (10). In a 100 mL three-neck round-
bottom flask equipped with a carbon rod anode and a platinum
wire cathode was dissolved 4.93 g (16.1 mmol) of (3R)-N-tert-
butyloxycarbonyl-3-phenyl-L-proline methyl ester was dis-
solved in 33 mL of a 0.03 M solution of tetraethylammonium
tosylate in methanol. The reaction mixture was degassed by
sonication under a slow stream of nitrogen for 10 min and then
electrolyzed at a constant current of 26.8 mA until 3.0 F/mol
had passed. The solution was concentrated in vacuo, and the
residue was chromatographed through 170 g of silica gel that
was slurry packed with 20% hexane in ether. Elution with the
same solvent afforded 4.95 g (87%) of the desired product (10).
The product was a mixture of stereoisomers with respect to
1
the methoxy group and carbamate rotomers: H NMR (300
MHz, CDCl3) δ 7.80-7.40 (m, 5H), 5.41 (d, J ) 4.5 Hz, 0.66H),
5.31 (d, J ) 4.4 Hz, 0.33H), 4.42 (d, J ) 9.3 Hz, 0.33H), 4.27
(d, J ) 9.5 Hz, 0.66H), 3.80-3.59 (m, 1H), 3.68 (s, 3H), 3.51
and 3.49 (two s, 3H), 2.32-2.21 (m, 1H), 2.18-2.04 (m, 1H),
1.52 and 1.43 (two s, 9H); 13C NMR (75 MHz, CDCl3) δ 172.6,
172.5, 154.0, 153.9, 139.6, 139.2, 128.7, 128.4, 128.1, 127.4,
127.0, 126.9, 88.2, 87.9, 81.0, 80.9, 66.4, 65.5, 55.6, 55.2, 52.1,
(3R,5R)-N-ter t-Bu tyloxyca r bon yl-3-p h en yl-5-(2′-h yd r o-
xyeth yl)-L-p r olin e Meth yl Ester . In a 100 mL flask, 3.043
g (8.8 mmol) of 16 was dissolved in 17 mL of anhydrous
methanol and 17 mL of dichloromethane. Ozone was bubbled
through the solution at -78 °C. The solution turned blue after
10 min, and the ozone bubbling was continued for another 10
(17) For a description of general experimental details, see ref 9a.

2490 J . Org. Chem., Vol. 65, No. 8, 2000
Tong et al.
min. The ozone was stopped and then oxygen bubbled through
the solution until the blue color disappeared. To this mixture
was added 0.400 g (10.6 mmol, 1.2 equiv) of NaBH4. The the
flask was then capped with a septum and connected to a
nitrogen bubbler. The dry ice bath was removed 10 min later,
and the reaction was stirred at room temperature for 1 h before
it was quenched with 50 mL of H2O at 0 °C. The mixture was
extracted two times (30 mL, 15 mL) with dichloromethane,
and then the combined organic layers were dried over mag-
nesium sulfate and concentrated in vacuo. The crude product
(3.4 g) was chromatographed through 140 g of silica gel using
ether/hexane (8:2) as the eluant to afford 2.314 g (75%) of pure
product as a colorless oil: 1H NMR (300 MHz, CDCl3) δ 7.38-
7.25 (m, 5H), 4.45-4.43 (m, 1H), 4.26 (d, 1H, J ) 9.5 Hz),
3.81-3.71 (m, 3H), 3.67 (s, 3H), 3.58-3.49 (m, 1H), 2.35 (dt,
1H, J ) 7.8, 12.4 Hz), 2.03 (dd, 1H, J ) 6.6, 12.4 Hz), 1.86-
1.79 (m, 1H), 1.42 (s, 9H); 13C NMR (125 MHz, CDCl3) δ 173.3,
155.2, 139.2, 128.8, 127.5, 127.3, 81.2, 66.8, 58.8, 54.5, 52.0,
(rel intensity) 332 (MH+, 16), 276 (MH+ - C4H8, 77), 232 (MH+
- C5H8O2, 100), 215 (58); HRFAB calcd for C19H26N1O4 (M +
1) 332.1862, found 332.1867. Anal. Calcd for C19H25N1O4: C,
68.86; H, 7.60; N, 4.23. Found: C, 68.86; H, 7.65; N, 4.21.
(3R,5S)-N-Ben zyloxyca r bon yl-L-p h en yla la n in e-3-p h en -
yl-5-(tr a n s-1-p r op en yl)-L-p r olin e Meth yl Ester (18a ). In
a 10 mL flask, 361 mg (1.04 mmol) of 11 was dissolved in 2.1
mL of anhydrous ether. To this mixture was added 156 mg
(1.10 mmol) of boron trifluoride etherate. The reaction was
then stirred at room temperature for 18 h. When complete,
the crude reaction mixture was washed with 5% sodium
bicarbonate and concentrated in vacuo. The crude product,
which contained some of the cis-substituted proline, was
carried on to the coupling step without purification. The
1
spectral data for the crude amine were as follows: H NMR
(300 MHz, CDCl3) δ 7.30 (m, 5H), 5.74-5.41 (m, 2H), 3.87 (m,
2H), 3.69 (s, 3H), 3.40 (m, 1H), 2.62 (bs, 1H), 2.31 (m, 1H),
1.76 (app q, J ≈ 11 Hz, 1H), 1.70 (d, J ) 6.4 Hz, 3H); 13C NMR
(75 MHz, CDCl3) δ 175.5, 142.7, 132.9, 132.5, 128.5, 127.3,
126.8, 126.7, 126.2, 67.6, 66.6, 61.2, 60.9, 55.3, 52.1, 49.9, 42.5,
42.4, 41.2, 17.7, 13.3; FAB MS m/z (rel intensity) 247 (M + 2,
10), 246 (M + 1, 100); HRMS FAB calcd for C15H20NO2 (M +
1) 246.1494, found 246.1491.
48.3, 39.4, 38.2, 28.2; IR (neat/NaCl) 3459 br, 1744, 1697 cm-1
LRFAB MS m/e (rel intensity) 350 (MH+, 68), 294 (MH+
;
-
C4H8, 63), 250 (MH+ - C5H8O2, 100); HRFAB calcd for
C
C
19H28N1O5 (M + 1) 350.1967, found 350.1969. Anal. Calcd for
19H27N1O5: C, 65.31; H, 7.79. Found: C, 65.15; H, 7.90.
(3R,5R)-N-ter t-Bu t yloxyca r b on yl-3-p h en yl-5-(2′-o-n i-
For the coupling step, a solution of 305 mg (1.03 mmol) of
N-carbobenzyloxy-L-phenylalanine and 83 mg (1.05 mmol) of
pyridine in 2.6 mL of dichloromethane were placed in a 25
mL flask and cooled to -20 °C. To this mixture was added
698 mg (5.16 mmol) of cyanuric fluoride. After 1.5 h, the
reaction was quenched by the addition of ice and extracted
twice with dichloromethane. The organic layer was dried over
magnesium sulfate and concentrated in vacuo. The resulting
residue was taken up in 2.0 mL of dichloromethane and added
to a solution of 202 mg (0.82 mmol) of the amine derived from
11 and 179 mg (1.55 mmol) of N-ethylmorpholine in 2.0 mL of
dichloromethane at room temperature. After 18 h, the reaction
mixture was washed with 0.5 M citric acid and 1 N sodium
bicarbonate, dried over magnesium sulfate, and concentrated
in vacuo. The crude product was chromatographed through
13 g of silica gel that was slurry packed with a 2:1 ether in
hexane solution. Elution with the same solvent afforded 218
mg (60% over two steps) of the desired product (18a ). The
spectral data for the major rotamers were as follows: 1 H NMR
(300 MHz, CDCl3) δ 7.40-7.10 (m, 15H), 5.84-5.30 (m, 2.5H),
5.20 (app q, J ≈ 8.5 Hz, 0.5H), 5.02 (m, 2.5H), 4.84 (m, 0.5H),
4.60-4.35 (m w/ dd at δ 4.52, J ) 8.7, 14.6 Hz, 1.5H), 3.91
(app q, J ≈ 7.2 Hz, 0.5H), 3.67 and 3.66 (two s, 3H), 3.40-
3.20 (m, 1H), 3.10 (m, 1H), 2.89 (app dd, J ≈ 13.8, 6.9 Hz,
1H), 2.52-2.30 (m, 1H), 2.16-1.76 (m, 1.5H), 1.71 (app d, J ≈
6.0 Hz, 1.5H), 1.67-1.44 (m, 1.5H); 13C NMR (75 MHz, CDCl3)
δ 171.1, 171.6, 155.3, 138.4, 138.3, 136.5, 136.4, 136.3, 131.0,
129.7, 129.5, 129.4, 129.3, 128.7, 128.6, 128.5, 128.3, 128.2,
128.1, 128.0, 127.9, 127.6, 127.4, 127.1, 127.0, 126.8, 126.7,
126.6, 66.7, 66.6, 66.4, 66.3, 65.9, 61.2, 55.8, 53.3, 53.1, 52.0,
46.6, 46.4, 41.9, 41.1, 39.3, 39.2, 31.5, 22.5, 17.5 14.0; IR (neat/
NaCl) 3295, 3063, 3030, 3004, 1721, 1641 cm-1; FAB MS m/z
(rel intensity) 659 (13), 529 (M + 3, 12), 528 (M + 2, 39), 483
(M+ - CH2CHCH2, 10), 336 (9), 246 (100), 244 (20); HRMS
FAB calcd for C32H35N2O5 (M + 1) 527.2546, found 527.2538.
(3S,6S,8R,9S)-N-Ben zyloxycar bon yl-1,4-diaza-3-ben zyl-
9-ca r bom eth oxy-5-h yd r oxy-8-p h en yl-2-oxobicyclo[4.3.0]-
n on a n e. To a 25 mL round-bottom flask was added 0.358 g
(0.68 mmol) of 18a along with 4 mL of anhydrous methanol
and 4 mL of dichloromethane. Ozone was bubbled through the
solution at -78 °C until a blue color persisted. After an
additional 10 min, the bubbling of ozone was stopped and
oxygen was bubbled through the solution until the blue color
disappeared. The reaction flask was then capped with a
septum and connected to a nitrogen bubbler, and 0.17 mL (2.4
mmol, 3.5 equiv) of methyl sulfide was added. The resulting
mixture was allowed to warm to room temperature and stirred
overnight. The reaction mixture was then concentrated in
vacuo, and the crude product was chromatographed through
25 g of silica gel using ether/hexane (8:2) as the eluant to afford
0.293 g (84%) of the product as a white solid. Two products
that were isomeric about the newly formed hydroxyl group
tr op h en ylselen o)eth yl-L-p r olin e Meth yl Ester . To a solu-
tion of 2.274 g (6.5 mmol) of (3R,5R)-N-tert-butyloxycarbonyl-
3-phenyl-5-(2′-hydroxyethyl)-L-proline methyl ester in 65 mL
of THF was slowly added 1.923 g (8.5 mmol, 1.3 equiv) of
o-nitrophenyl selenocyanate and 2.12 mL (8.5 mmol, 1.3 equiv)
of tributylphosphine. The reaction was stirred overnight, and
then the mixture was concentrated. The crude product was
chromatographed through 120 g of silica gel using gradient
hexane to hexane/ether (7:3) as the eluant. The product (3.4
g) was mixed with some impurities. The material collected
from the first column was chromatographed again through 150
g of silica gel using gradient hexane/ether (8:2 to 1:1) as the
eluant to afford 2.5 g (72%) of pure product as a yellow solid.
The spectral data for the mixture of carbamate rotamers were
as follows: 1H NMR (300 MHz, CDCl3) δ 8.31 (d, 1H, J ) 8.2
Hz), 7.70 (d, 1H, J ) 8.2 Hz), 7.59-7.53 (m, 1H), 7.37-7.24
(m, 6H), 4.45-4.22 (m, 2H), 3.69 (s, 3H), 3.58-3.47 (m, 1H),
3.16-3.03 (m, 2H), 2.40-2.25 (m, 2H), 2.14-2.03 (m, 2H), 1.48
(s, 3H), 1.43 (s, 6H); 13C NMR (150 MHz, CDCl3) δ 173.4, 153.9,
146.8, 139.7, 139.3, 133.7, 129.2, 128.8, 127.5, 127.1, 126.5,
125.3, 125.2, 80.6, 66.7, 66.1, 58.4, 58.2, 52.3, 52.0, 48.3, 47.4,
39.3, 38.6, 34.5, 34.0, 28.5, 28.2, 22.9, 22.6; IR (neat/NaCl)
1744, 1694 cm-1; LRFAB MS m/e (rel intensity) 535 (MH+,
19), 435 (MH+ - C5H8O2, 100), 276 (MH+ - C8H6NO4Se, 76),
185 (60); HRFAB calcd for C25H31N2O6Se (M + 1) 535.1347,
found 535.1331. Anal. Calcd for C25H30N2O6Se: C, 56.29; H,
5.67. Found: C, 56.62; H, 5.93.
(3R,5R)-N-ter t-Bu t yloxyca r b on yl-3-p h en yl-5-vin yl-L-
p r olin e Meth yl Ester (17). To a -78 °C solution of 0.115 g
(0.22 mmol) of the selenium compound made above in 24 mL
of dichloromethane was added 0.074 g of m-CPBA (57-86%).
After 90 min, methyl sulfide (0.79 mL, 50 equiv) and triethyl-
amine (0.78 mL, 26 equiv) were added to the mixture, and the
reaction was allowed to warm to room temperature. After 5
h, 10 mL of a saturated NaHCO3 solution was added to the
reaction. The layers were separated, and the aqueous layer
was extracted with dichloromethane. The combined organic
layers were dried over magnesium sulfate and concentrated
in vacuo. The crude product was chromatographed through
silica gel using gradient hexane/ether (8:2 to 1:1) as the eluant
to afford 0.066 g (92%) pure product as a yellow oil. The
spectral data for the mixture of carbamate rotamers were as
follows: 1H NMR (300 MHz, CDCl3) δ 7.36-7.22 (m, 5H),
6.03-5.93 (m, 1H), 5.57-5.39 (m, 1H), 5.30-5.17 (m, 1H),
4.61-4.59 (m, br, 0.6H), 4.47-4.46 (m, br, 0.4H), 4.40 (d, 0.4H,
J ) 7.8 Hz), 4.26 (d, 0.6H, J ) 8.7 Hz), 3.67 (s, 3H), 3.54-3.45
(m, 1H), 2.36-2.25 (m, 1H), 2.14-2.08 (m, 1H), 1.46 (s, 3.3H),
1.42 (s, 5.7H); 13C NMR (125 MHz, CDCl3) δ 173.1, 153.8,
139.1, 138.3, 137.7, 128.7, 127.4, 127.3, 127.2, 115.6, 115.3,
80.4, 66.6, 65.9, 60.1, 59.7, 52.1, 51.9, 47.8, 46.9, 39.8, 39.0,
28.3, 28.2; IR (neat/NaCl) 1757, 1701 cm-1; LRFAB MS m/e

Conformationally Constrained Substance P Analogues
J . Org. Chem., Vol. 65, No. 8, 2000 2491
were obtained in a roughly 2.2:1 ratio. Both isomers were
converted into 19a . The spectral data for the major isomer
were as follows: 1H NMR (300 MHz, CDCl3) δ 7.39-7.09 (m,
15H), 4.98-4.95 (m, 3H), 4.51 (d, 1H, J ) 9.4 Hz), 4.35 (x of
abx, 1H, J ax ) J bx ) 8.1 Hz), 4.08 (ddd, 1H, J ) 5.0, 8.6, 13.4
Hz), 3.71 (s, 1H), 3.41 (ddd, 1H, J ) 6.2, 9.2, 15.6 Hz), 3.28 (a
in 9 mL of brine and 18 mL of ethyl acetate. After separation,
the aqueous layer was washed twice with ethyl acetate. The
combined organic layers were dried over sodium sulfate and
concentrated. The crude product was chromatographed through
50 g of silica gel using ether/methanol (9:1) as the eluant to
afford 0.423 g (62%) of pure product as a white solid. The
spectral data for the mixture of carbamate rotamers were as
follows: 1H NMR (300 MHz, CDCl3) δ 7.41-7.13 (m, 10H),
4.96-4.92 (m, 0.2H), 4.83-4.80 (m, 0.8H), 4.65-4.62 (m,
0.25H), 4.57 (d, 0.75 H, J ) 8.9 Hz), 4.49-4.46 (m, 0.22H),
4.41 (dd, 0.78H, J ) 4.1, 13.6 Hz), 3.92-3.61 (m, 1.5H), 3.48-
3.38 (m, 0.5H), 3.23 (d, 2H, J ) 5.6 Hz), 2.37-2.27 (m, 1H),
2.18-2.10 (m, 1H), 1.94-1.85 (m, 0.5H), 1.73-1.63 (m, 0.5H),
1.46 (s, 2H), 1.32 (s, 7H); 13C NMR (125 MHz, CDCl3) δ 171.3,
169.2, 153.4, 139.1, 137.1, 130.2, 129.7, 128.9, 128.6, 127.6,
127.3, 127.0, 81.2, 66.2, 58.6, 57.7, 45.4, 42.7, 38.1, 37.3, 28.3,
28.0; IR (neat/NaCl) 3451 br, 1694, 1645 cm-1; LRFAB MS
m/e (rel intensity) 473 (MNa+, 28), 451 (MH+, 10), 395 (MH+
- C4H8, 100), 303 (13), 277 (54), 213 (16), 154 (33); HRFAB
calcd for C26H31N2O5 (M + 1) 451.2233, found 451.2214.
(3R,5R)-N-Ben zyloxycar bon yl-L-ph en ylalan in e-3-ph en -
yl-5-(cis-1-vin yl)-L-p r olin e Meth yl Ester (18b). Using the
same deprotection and coupling procedure outlined for the
synthesis of 18a , 0.586 g (1.77 mmol) of 17 was converted into
0.668 g (73%) of the coupled product 18b. The proton NMR
data for the intermediate amine were as follows: 1H NMR (300
MHz, CDCl3) δ 7.36-7.24 (m, 5H), 5.96 (ddd, 1H, J ) 7.1, 10.1,
17.1 Hz), 5.25 (d, 1H, J ) 17.0 Hz), 5.11 (d, 1H, J ) 10.4 Hz),
3.96-3.91 (m, 1H), 3.88 (d, 1H, J ) 7.0 Hz), 3.69 (s, 3H), 3.45
(dd, 1H, J ) 6.9, 15.5 Hz), 2.37 (s, 1H), 2.37-2.01 (m, 2H).
The spectral data for 18b were as follows: 1H NMR (300 MHz,
CDCl3) δ 7.35-7.16 (m, 15H), 6.05 (ddd, 1H, J ) 6.0, 10.2,
16.8 Hz), 5.60 (d, 1H, J ) 16.8 Hz), 5.30-5.27 (m, 2H), 5.11-
4.96 (m, 3H), 4.78-4.71 (m, 1H), 4.53 (d, 1H, J ) 9.3 Hz), 3.67
(s, 3H), 3.66-3.45 (m, 1H), 3.14 (dd, 1H, J ) 5.1, 13.8 Hz),
2.84 (dd, 1H, J ) 9.0, 13.8 Hz), 2.49-2.38 (m, 1H), 2.20-2.13
(m, 1H); 13C NMR (125 MHz, CDCl3) δ 171.9, 156.0, 138.6,
138.1, 136.4, 136.2, 129.5, 128.8, 128.5, 128.4, 128.1, 127.8,
127.5, 127.2, 126.8, 117.2, 66.9, 66.0, 60.6, 52.8, 52.2, 46.0, 40.7,
38.4; IR (neat/NaCl) 3282 br, 3064, 3029, 1743, 1707, 1645
cm-1; LRFAB MS m/e (rel intensity) 513 (MH+, 81), 332 (MH+
- C14H13, 40), 276 (94), 232 (MH+ - C17H15NO3, 100); HRFAB
calcd for C31H33N2O5 (M + 1) 513.2389, found 513.2398. Anal.
Calcd for C31H32N2O5: C, 72.64; H, 6.29. Found: C, 72.66; H,
6.42.
(3S,6R,8R,9S)-N-Ben zyloxycar bon yl-1,4-diaza-3-ben zyl-
9-ca r bom eth oxy-5-h yd r oxy-8-p h en yl-2-oxobicyclo[4.3.0]-
n on a n e. Using the same ozonolysis procedure described above
for converting 18a into (3S,6S,8R,9S)-N-benzyloxycarbonyl-
1,4-diaza-3-benzyl-9-carbomethoxy-5-hydroxy-8-phenyl-2-
oxobicyclo[4.3.0]nonane, 0.613 g (1.2 mmol) of 18b was con-
vertedinto0.6104g(99%)ofthe(3S,6R,8R,9S)-N-benzyloxycarbonyl-
1,4-diaza-3-benzyl-9-carbomethoxy-5-hydroxy-8-phenyl-2-
oxobicyclo[4.3.0]nonane. The molecule was obtained as a 2:1
mixture of stereoisomers about the N-R-hydroxy amide group.
Both isomers were readily converted into the reduced 19b
below. The spectral data for the major isomer were as
follows: 1H NMR (300 MHz, CDCl3) δ 7.43-7.09 (m, 15H), 5.35
(d, 1H, J ) 8.2 Hz), 4.99-4.94 (m, br, 1H), 4.82-4.80 (m, br,
1H), 4.77-4.73 (m, br, 1H), 4.66-4.63 (m, br, 1H), 3.78-3.71
(m, 4H), 3.61 (d, 1H, J ) 7.1 Hz), 3.15 (dd, 1H, J ) 4.8, 13.6
Hz), 3.03 (dd, 1H, J ) 9.2, 13.5 Hz), 2.46-2.36 (m, 2H); 13C
NMR (150 MHz, CDCl3) δ 171.1, 167.5, 156.0, 141.0, 136.1,
135.1, 129.9, 129.1, 128.4, 128.2, 127.4, 126.9, 126.2, 82.2, 67.9,
63.8, 60.3, 57.1, 52.8, 45.9, 38.6, 37.5; IR (neat/NaCl) 3444 br,
3036, 1743, 1673 cm-1; LRFAB MS m/e (rel intensity) 537
(MNa+, 4), 515 (MH+, 87), 497 (M+-OH, 100), 460 (28), 425
(MH+-C7H6, 23), 386 (36); HRFAB calcd for C30H31N2O6 (M
+ 1) 515.2182, found 515.2188. Anal. Calcd for C30H30N2O6:
C, 70.02; H, 5.88. Found: C, 69.52; H, 6.01.
of abx, 1H, J ax ) 4.3, J ab ) 13.2 Hz), 3.15 (b of abx, 1H, J bx
)
8.2, J ab ) 13.9 Hz), 2.58-2.50 (m, 1H), 1.73 (ddd, 1H, J ) 12.2
Hz); 13C NMR (125 MHz, CDCl3) δ 171.6, 166.0, 156.5, 138.5,
136.8, 134.9, 130.1, 129.6, 128.9, 128.8, 128.5, 128.4, 128.0,
127.7, 127.3, 127.1, 82.7, 67.7, 65.7, 63.0, 60.4, 52.5, 47.4, 38.4,
37.2; IR (neat/NaCl) 3396 br, 3030, 1750, 1708, 1665 cm-1
;
LRFAB MS m/e (rel intensity) 515 (MH+, 3), 167 (22), 149
(100); HRFAB calcd for C30H31N2O6 (M + 1) 515.2182, found
515.2166. Anal. Calcd for C30H30N2O6: C, 70.02; H, 5.88.
Found: C, 69.58; H, 6.11.
(3S,6S,8R,9S)-1,4-Dia za -3-b en zyl-9-ca r b om et h oxy-8-
p h en yl-2-oxobicyclo[4.3.0]n on a n e (19a ). To a solution of
1.557 g (3.03 mmol) of the R-hydroxyamide generated from 18a
in 27 mL of dry methanol was added 0.389 g (1/4 wt of the
starting material) of 5% palladium/BaSO4. The reaction was
run under a hydrogen balloon for 26 h, and then the reaction
mixture filtered through Celite and concentrated. The crude
product was chromatographed through 120 g of silica gel using
ether/methanol (9:1) as the eluant to afford 0.824 g (75%) of
pure product as a yellow oil: 1H NMR (300 MHz, CDCl3) δ
7.38-7.22 (m, 10H), 4.60 (d, 1H, J ) 8.9 Hz), 4.04-3.97 (m,
1H), 3.79-3.75 (m, 1H), 3.71 (s, 1H), 3.46-3.37 (m, 1H), 3.27
(dd, 1H, J ) 3.3, 13.7 Hz), 3.14 (dd, 1H, J ) 3.8, 12.6 Hz),
2.97 (dd, 1H, J ) 10.3, 13.6 Hz), 2.65 (dd, 1H, J ) 9.9, 12.3
Hz), 2.39-2.32 (m, 1H), 1.85 (s, br, 1H), 1.70 (dd, 1H, J ) 11.7,
23.2 Hz); 13C NMR (75 MHz, CDCl3) δ 172.1, 169.2, 139.4,
138.7, 129.3, 128.7, 128.5, 127.4, 127.1, 126.5, 64.8, 60.2, 58.1,
52.2, 47.5, 44.0, 39.4, 37.9; IR (neat/NaCl) 3334, 1746, 1633
cm-1; LRFAB MS m/e (rel intensity) 365 (MH+, 100), 307 (MH+
- C2H2O2, 12), 273 (M+ - C7H7, 79), 154 (92); HRFAB calcd
for C22H25N2O3 (M + 1) 365.1865, found 365.1863. Anal. Calcd
for C22H24N2O3: C, 72.51; H, 6.64; N, 7.69. Found: C, 72.84;
H, 6.19; N, 7.06.
(3S,6S,8R,9S)-N-ter t-Bu tyloxycar bon yl-1,4-diaza-3-ben -
zyl-9-ca r b om e t h oxy-8-p h e n yl-2-oxob icyclo[4.3.0]n o-
n a n e. To a 0 °C solution of 0.108 g (0.30 mmol) of 19a in 1
mL of dichloromethane was added 0.13 mL (1.0 mmol, 3.5
equiv) of 4-ethylmorpholine and 0.258 g (1.18 mmol, 4 equiv)
of di-tert-butyl dicarbonate dissolved in 1.5 mL of dichlo-
romethane. The reaction mixture was warmed to room tem-
perature and stirred overnight. Since TLC still showed the
starting material, the reaction mixture was heated to reflux
for another 9 h. The solution was concentrated in vacuo, and
the crude product was chromatographed through 30 g of silica
gel using ether as the eluant to afford 0.103 g (75%) of pure
product as a white crystal. The spectral data for the mixture
of carbamate rotamers were as follows: 1H NMR (300 MHz,
CDCl3) δ 7.38-7.17 (m, 10H), 4.91 (m, 0.2 H), 4.81-4.77 (m,
0.8H), 4.53 (d, 1H, J ) 9.2 Hz), 4.40 (dd, 1H, J ) 4.0, 13.6
Hz), 3.99-3.91 (m,1H), 3.72 (s, 3H), 3.44-3.35 (m, 1H), 3.23
(d, 2H, J ) 5.7 Hz), 2.35-2.27 (m, 1H), 2.14 (dd, 1H, J ) 10.8,
13.4 Hz), 1.91-1.78 (m, 0.5H), 1.58-1.50 (m, 0.5H), 1.47 (s,
3H), 1.32 (s, 6H); 13C NMR (125 MHz, CDCl3) δ 171.9, 166.8,
153.6, 138.9, 137.5, 129.8, 129.7, 128.8, 128.4, 128.2, 127.6,
127.0, 126.8, 80.8, 65.0, 58.2, 57.8, 52.4, 47.5, 42.9, 38.8, 37.5,
28.3, 28.0; IR (neat/NaCl) 1744, 1694, 1659 cm-1; LRFAB MS
m/e (rel intensity) 465 (MH+, 42), 409 (MH+ - C4H8, 100), 349
(22), 273 (MH+ - C5H9O2 - C7H7, 30), 154 (100); HRFAB calcd
for C27H33N2O5 (M + 1) 465.2389, found 465.2383. Anal. Calcd
for C27H32N2O5: C, 69.81; H, 6.94; N, 6.03. Found: C, 70.04;
H, 6.84; N, 5.97.
(3S,6S,8R,9S)-N-ter t-Bu tyloxycar bon yl-1,4-diaza-3-ben -
zyl-9-ca r b oxy-8-p h en yl-2-oxob icyclo[4.3.0]n on a n e (20).
To a 0 °C solution of 0.700 g (1.5 mmol) of the t-Boc-protected
derivative of 19a in 25.5 mL of THF/MeOH/H2O (12:4:1) was
added 0.164 g (3.9 mmol, 2.6 equiv) of lithium hydroxide
monohydrate. After 3 h at 0 °C, the reaction mixture was
acidified to pH ) 3 using saturated NaHSO4 solution. The
solution was then concentrated, and the residue was dissolved
(3S,6R,8R,9S)-1,4-Dia za -3-b en zyl-9-ca r b om et h oxy-8-
p h en yl-2-oxobicyclo[4.3.0]n on a n e (19b). Using the proce-
dure described above for the synthesis of 19a , 0.376 g (0.73
mmol) of the N-R-hydroxyamide derived from 18b was con-
verted into 0.175 g (66%) of 19b: 1H NMR (300 MHz, CDCl3)

2492 J . Org. Chem., Vol. 65, No. 8, 2000
Tong et al.
δ 7.36-7.05 (m, 10H), 4.71 (s, 1H), 3.80 (s, 3H), 3.79-3.76 (m,
1H), 3.61-3.52 (m, 1H), 3.48-3.44 (m, 1H), 3.32 (dd, 1H, J )
3.7, 14.6 Hz), 3.16 (dd, 1H, J ) 3.4, 12.2 Hz), 3.09 (dd, 1H, J
) 8.0, 13.6 Hz), 2.76 (dd, 1H, J ) 10.3, 12.2 Hz), 2.17 (ddd,
1H, J ) 7.3, 12.0, 12.0 Hz), 1.85 (dd, 1H, J ) 5.3, 12.2 Hz),
1.62 (s, br, 1H); 13C NMR (125 MHz, CDCl3) δ 172.1, 169.8,
142.0, 137.9, 129.5, 128.8, 128.6, 127.1, 126.8, 126.6, 64.3, 59.8,
57.9, 52.5, 48.1, 45.5, 38.2, 36.3; IR (neat/NaCl) 3641, 3331
br, 3072, 3030, 1743, 1651 cm-1; LRFAB MS m/e (rel intensity)
365 (MH+, 100), 307 (MH+ - C2H2O2, 62), 273 (M+ - C7H7,
92); HRFAB calcd for C22H25N2O3 (M + 1) 365.1865, found
365.1856. Anal. Calcd for C22H24N2O3: C, 72.51; H, 6.64; N,
7.69. Found: C, 71.93; H, 6.79; N, 7.54.
(3S)-3-Ben zyl-4-ter t-bu tyloxycar bon yl-5-h ydr oxy-1-(1′S-
ben zyl-1′-ca r bom eth oxy)m eth yl-1,4-dia za -2-oxocycloh ex-
a n e. In a 250 mL flask was dissolved 1.964 g (4.21 mmol) of
25 in 30 mL of anhydrous methanol. Ozone was bubbled
through the solution at -78 °C. The solution turned blue 20
min later. After an additional 15 min, the addition of ozone
was stopped, and oxygen was bubbled through the solution
until the blue color disappeared. The reaction flask was then
capped with a rubber septum and connected to a nitrogen
bubbler, and 1.1 mL (15 mmol, 3.5 equiv) of methyl sulfide
was added. The reaction was allowed to warm to room
temperature and stirred overnight. The reaction mixture was
then concentrated in vacuo. The crude product was chromato-
graphed through 160 g of silica gel using ether/hexane (9:1)
as the eluant to afford 1.736 g (88%) of the purified product
as a white solid. Two isomers were obtained with respect to
the N-R-hydroxyamide moiety. The isomers were combined and
then used in the next reaction as a mixture: 1H NMR (300
MHz, CDCl3, major isomer) δ 7.30-7.11 (m, 10H), 5.53-5.41
(m, 2H), 4.51 (dd, 1H, J ) 3.4, 11.0 Hz), 4.36 (s, 1H), 3.79 (s,
3H), 3.74-3.68 (m, 1H), 3.51-3.33 (m, 2H), 3.00 (dd, 1H, J )
13.1, 14.4 Hz), 2.48-2.29 (m, 2H), 1.04 (s, 9H); 13C NMR (75
MHz, CDCl3, major isomer) δ 170.8, 169.4, 154.8, 136.9, 135.8,
130.0, 129.8, 129.1, 128.9, 128.6, 128.5, 128.2, 127.2, 127.1,
126.5, 81.3, 75.0, 59.7, 56.3, 52.6, 45.6, 38.1, 35.5, 28.2, 27.5;
IR (neat/NaCl, major isomer) 3448 br, 1743, 1687 cm-1; LRFAB
MS (major isomer) m/e (rel intensity) 491(MNa+, 6), 469 (MH+,
22), 468 (M+, 5), 451 (M+ - OH, 100), 438 (7); HRFAB (major
isomer) calcd for C26H33N2O6 (M + 1) 469.2338, found 469.2328;
LRFAB MS (minor isomer) m/e (rel intensity) 491 (MNa+, 3),
469 (MH+, 8), 451 (M+ - OH, 100), 438 (3); HRFAB (minor
isomer) calcd for C26H33N2O6 (M + 1) 469.2338, found 469.2340.
(3S)-N-(1′S-Ben zyl-1′-ca r bom eth oxy)m eth yl-3-ben zyl-
1,4-d ia za -2-oxocycloh exa n e (26). To a room-temperature
solution of 1.321 g (2.82 mmol) of the N-R-hydroxyamide
derived from 25 in 20 mL of dichloromethane was added 0.90
mL (5.6 mmol, 2 equiv) of triethylsilane and 3.26 mL (42.3
mmol, 15 equiv) of trifluoroacetic acid. The reaction mixture
was stirred for 1 day and then concentrated. The residue was
dissolved in 10 mL of dichloromethane and 10 mL of triethyl-
amine at 0 °C and stirred for 1 h at room temperature, and
the solution was concentrated. The residue was dissolved in
20 mL of 20% sodium bicarbonate and 20 mL of dichlo-
romethane. After separation, the aqueous layer was extracted
two more times with dichloromethane. The crude product was
chromatographed through 70 g of silica gel (slurry prepared
in 95% ether/methanol with 30 drops of triethylamine) using
ether/methanol (9.5:0.5) as the eluant to afford 0.82 g (83%)
of pure product as a yellow oil: 1H NMR (300 MHz, CDCl3) δ
7.34-7.16 (m, 10H), 5.08 (dd, 1H, J ) 5.4, 10.9 Hz), 3.76 (s,
3H), 3.63 (dd, 1H, J ) 3.6, 9.7 Hz), 3.42-3.26 (m, 3H), 3.14
(dd, 1H, J ) 10.9, 14.5 Hz), 2.97-2.89 (m, 2H), 2.80-2.71 (m,
1H), 2.61 (dd, 1H, J ) 9.8, 13.4 Hz), 1.51 (s, br, 1H); 13C NMR
(75 MHz, CDCl3) δ 171.0, 169.8, 138.3, 137.0, 129.3, 128.9,
128.6, 128.5, 126.8, 126.6, 60.6, 58.7, 52.4, 46.3, 41.8, 38.2, 34.2;
N-Allyl-L-p h en yla la n in e Meth yl Ester (24). To a solution
of 1.224 g (5.7 mmol) of l-phenylalanine methyl ester hydro-
chloride 23 in 18 mL of N,N-dimethylformamide was added
2.77 mL (19.9 mmol, 3.5 equiv) of triethylamine at 0 °C. Allyl
bromide (1.23 mL, 14.2 mmol, 2.5 equiv) was added to the
mixture, and the reaction was warmed to room temperature.
After 47 h of stirring, the reaction was diluted with 15 mL of
water and extracted three times with 20 mL of ether. The
combined organic layers were washed with saturated sodium
chloride solution and then dried over sodium sulfate and
concentrated. The crude product (1.3 g) was chromatographed
through 60 g of silica gel (slurry prepared by ethyl acetate with
15 drops of triethylamine) using ethyl acetate as the eluant
to afford 0.750 g (60%) of the purified product as a pale yellow
oil: 1H NMR (300 MHz, CDCl3) δ 7.32-7.16 (m, 5H), 5.80 (ddt,
1H, J ) 10.3, 17.1, 6.0 Hz), 5.14-5.05 (m, 2H), 3.64 (s, 3H),
3.56 (x of abx, 1H, J ax ) J bx ) 6.8 Hz), 3.26 (a of abx, 1H, J ab
) 14.0, J ax ) 5.9 Hz), 3.15 (b of abx, 1H, J ab ) 14.0, J bx ) 6.2
Hz), 2.96 (d, 1H, J ) 6.9 Hz), 1.59 (s, br, 1H); 13C NMR (75
MHz, CDCl3) δ 175.0, 137.1, 136.1, 129.1, 128.4, 126.7, 116.4,
62.0, 51.6, 50.6, 39.7; IR (neat/NaCl) 3331 br, 3083, 3063, 1736,
1643 cm-1; LRFAB MS m/e (rel intensity) 220 (MH+, 100), 160
(M+- C2H3O2, 58), 128 (M+- C7H7, 42); HRFAB calcd for
C
13H18NO2 (M + 1) 220.1337, found 220.1336.
t-Boc-P h e-F . To a -15 °C (ethylene glycol/dry ice bath)
solution of 4.996 g (0.0188 mol) of t-Boc-Phe-OH in 45 mL of
dichloromethane was added 1.52 mL (0.0188 mol, 1 equiv) of
pyridine and 8.48 mL (0.0942 mol, 5 equiv) of cyanuric fluoride.
The reaction was stirred at -15 °C for 1 h and then diluted
with a large amount of ice-water. The aqueous layer was
extracted four times with dichloromethane. The combined
organic layers were dried over MgSO4 and concentrated to
afford 3.7 g (74%) of product as a white solid. The product was
used in the next step without further purification: 1H NMR
(300 MHz, CDCl3) δ 7.39-7.18 (m, 5H), 4.86-4.74 (m, br, 1H),
3.18-3.17 (m, 2H), 1.43 (s, 9H).
N-Allyl-ter t-bu tyloxycar bon yl-L-ph en ylalan in yl-L-ph en -
yla la n in e Meth yl Ester (25). To a room-temperature solu-
tion of 2.580 g (11.8 mmol) of 24 in 25 mL of dichloromethane
was added 1.72 mL (13.5 mol, 1.15 equiv) of N-ethylmorpho-
line. To this mixture was cannulated 3.775 g (14.1 mol, 1.2
equiv) of t-Boc-Phe-F dissolved in 20 mL of dichloromethane.
The reaction mixture was stirred for 2 days and then diluted
with 15 mL of dichloromethane followed by extraction with
20 mL of 5% citric acid and 40 mL of 5% sodium bicarbonate.
The crude product was chromatographed through 300 g of
silica gel using ethyl acetate/hexane (1:1) as the eluant to
afford 4.122 g (75%) of the purified product as a white
crystal: 1H NMR (300 MHz, CDCl3) δ 7.32-7.15 (m, 8H),
7.06-6.99 (m, 2H), 5.48-5.39 (m, 1H), 5.14-5.02 (m, 3H), 4.74
(dd, 1H, J ) 7.2, 16.5 Hz), 4.41 (dd, 1H, J ) 5.9, 9.1 Hz), 3.79
(dd, 1H, J ) 4.8, 16.8 Hz), 3.68 (s, 3H), 3.47-3.30 (m, 3H),
3.07-2.99 (m, 2H), 2.90-2.82 (m, 1H), 1.38 (s, 9H); 13C NMR
(75 MHz, CDCl3) δ 172.0, 170.6, 154.9, 137.7, 136.6, 133.0,
129.6, 129.2, 128.4, 126.8, 118.3, 79.7, 60.7, 52.2, 51.8, 51.0,
IR (neat/NaCl) 3495 br, 3323, br, 3028, 1745, 1644 cm-1
;
LRFAB MS m/e (rel intensity) 391 (MK+, 6), 353 (MH+, 100),
325 (25), 289 (17), 261 (M+ - C7H7, 75); HRFAB calcd for
C
C
21H25N2O3 (M + 1) 353.1865, found 353.1865. Anal. Calcd for
21H24N2O3: C, 71.57; H, 6.86; N, 7.95. Found: C, 70.76; H,
6.64; N, 7.98.
(3S)-3-Ben zyl-4-ter t-bu tyloxyca r bon yl-1-(1′S-ben zyl-1′-
ca r bom eth oxy)m eth yl-1,4-d ia za -2-oxocycloh exa n e. To a
0 °C solution of 0.105 g (0.30 mmol) of 26 in 1 mL of
dichloromethane was added 0.057 mL (0.45 mmol, 1.5 equiv)
of 4-ethylmorpholine and 0.162 g (0.742 mmol, 2.5 equiv) of
di-tert-butyl dicarbonate dissolved in 2.5 mL of dichlo-
romethane. The reaction was allowed to warm to room
temperature. After 16 h, starting material was still evident
by TLC. To the reaction mixture was added 0.034 mL of
4-ethylmorpholine and 0.097 g of di-tert-butyl dicarbonate. The
reaction was then heated to reflux. Six hours later, the reaction
was stopped, and the mixture was concentrated in vacuo. The
residue was chromatographed through 30 g of silica gel using
ether/hexane (9:1) as the eluant to afford 0.120 g (89%) of pure
39.6, 35.0, 28.3; IR (neat/NaCl) 3318 br, 3028, 1711, 1650 cm-1
;
LRFAB MS m/e (rel intensity) 467 (MH+, 30), 411 (MH+
-
C4H8, 9), 379 (16), 367 (MH+ - C5H8O2, 42), 246 (M+
-
C
13H18O2N, 4), 220 (C13H18O2N, 100), 160 (63), 136 (27), 120
(75); HRFAB calcd for C27H35N2O5 (M + 1) 467.2546, found
467.2547. Anal. Calcd for C27H34N2O5: C, 69.51; H, 7.34; N,
6.00. Found: C, 69.65; H, 7.29; N, 6.03.

Conformationally Constrained Substance P Analogues
J . Org. Chem., Vol. 65, No. 8, 2000 2493
product as a colorless oil. The spectral data for the mixture of
carbamate rotamers were as follows: 1H NMR (300 MHz,
CDCl3) δ 7.33-7.21 (m, 8H), 7.04-7.02 (m, 2H), 5.29-5.26 (m,
br, 1H), 4.66-4.65 (m, br, 1H), 3.96 (d, br, 0.7H), 3.77 (s, 3H),
3.3.66-3.53 (m, br, 0.3H), 3.41-3.35 (m, 1H), 3.19-2.94 (m,
4.2H), 2.72-2.64 (m, 1.8H), 1.36 (s, br, 2H), 1.19 (s, 7H); 13C
NMR (75 MHz, CDCl3) δ 170.5, 168.3, 153.4, 137.3, 136.4,
129.7, 128.8, 128.6, 128.3, 127.0, 126.5, 80.4, 58.8, 57.9, 52.4,
44.0, 37.4, 37.0, 34.1, 27.9; IR (neat/NaCl) 3023, 1744, 1694,
1652 cm-1; LRFAB MS m/e (rel intensity) 453 (MH+, 7), 397
(MH+ - C4H8, 100), 337 (36), 265 (27), 146 (15); HRFAB calcd
for C26H33N2O5 (M + 1) 453.2389, found 453.2392. Anal. Calcd
for C26H32N2O5: C, 69.01; H, 7.13; N, 6.19. Found: C, 68.91;
H, 7.10; N, 6.08.
demanding secondary amine. In these cases, the reactions were
allowed to stir for 15-18 h. Due to the secondary amine, the
Kaiser test was not useful for determining if the coupling
reactions had proceeded to completion. For this reason, the
reactions were monitored for completeness using the chloranil
(tetrachlorobenzoquinone)-acetaldehyde procedure.18
After the final amino acid, arginine, was coupled to the
peptide, the “permanent” protecting groups were removed. To
this end, the oxide protecting the methionine moiety was
removed by treatment of the solid-phase peptide with thiophen-
ol in DMF. The t-Boc group was removed from the N-terminal
Arg using the conditions described above. The peptide was
then cleaved from the resin and the “permanent” protecting
groups removed from the Arg, Lys, and Gln residues with the
use of anisole, methyl sulfide, and a dithioethanol in HF
solution. The reactions were performed at 0 °C using an HF
reaction apparatus (model I) made by Immuno Dynamics, Inc.
The analogues were purified by filtering away the resin and
then isolating the analogue by HPLC. The preparative HPLC
purification of the crude peptide was carried on a Rainin HPXL
solvent delivery system with a Rainin Dynamax UV-1 UV/
Visible Absorbance Detector. A Vydac C18 (5 µm, 10 × 250
mm) column was used with a 20% to 80% acetonitrile/water
linear gradient containing 0.1% (by volume) of TFA. The
elution rate was 15 mL/min and the UV detection wavelength
was set at 220 nm. The retention times for analogues 29, 30,
and 31 were 12.9 min (50.9% acetonitrile), 12.4 min (49.9%
acetonitrile), and 10.9 min (46.4% acetonitrile), respectively.
The only major byproduct for each analogue came off the
column about 2 min earlier. Using HRMS, the byproduct was
shown to be the peptide with an oxidized Met side chain.
Apparently, some of the Met side chain was oxidized during
the final t-Boc deprotection and/or HF cleavage process. The
purity of the peptide analogues was confirmed by analytical
HPLC using a Hewlett-Packard 1100 series system with a
Xpertek (C18, 5 µm, 4.6 × 250 mm) column. The gradient
elution solvent conditions were the same as for the preparative
purification, while the elution rate was changed to 1.4 mL/
min. The retention times for analogues 29, 30, and 31 were
14.2, 12.9, and 13.5 min respectively. The mass spectrum
evidence for analogue 29 is as follows: LRFAB MS (rel
intensity) 1385.6 (MH+, 100), 693.6 (35); HRFAB calcd for
(3S)-3-Ben zyl-4-ter t-bu tyloxyca r bon yl-1-(1′S-ben zyl-1′-
ca r boxy)m eth yl-1,4-d ia za -2-oxocycloh exa n e (22). To a 0
°C solution of 0.852 g (1.9 mmol) of (3S)-3-benzyl-4-tert-
butyloxycarbonyl-1-(1′S-benzyl-1′-carbomethoxy)methyl-1,4-
diaza-2-oxocyclohexane in 30.6 mL of THF/MeOH/H2O (12:4:
1) was added 0.197 g (4.7 mmol, 2.5 equiv) of lithium hydroxide
monohydrate. The reaction was kept at 0 °C and stirred for 2
h. When complete, the reaction mixture was acidified to pH )
3 using saturated sodium bisulfate solution (about 3 mL). The
solution was concentrated and the residue dissolved in 10 mL
of brine and 20 mL of ethyl acetate. After separation, the
aqueous layer was extracted two more times with ethyl
acetate. The combined organic layers were dried over sodium
sulfate and concentrated. The crude product was chromato-
graphed through 50 g of silica gel using gradient ether/
methanol (9:1 to 8:2) to afford 0.591 g (72%) of pure product
as a white solid. The spectral data for the mixture of carbamate
rotamers were as follows: 1H NMR (300 MHz, CDCl3) δ 10.32
(s, br, 1H), 7.32-7.20 (m, 8.5H), 7.05-7.02 (m, 1.5H), 5.13 (s,
br, 0.75H), 4.95 (s, br, 0.25H), 4.85 (s, br, 0.25H), 4.85, (s, br,
0.75H), 3.96 (d, 0.7H, J ) 13.0 Hz), 3.67-3.64 (m, 0.3H), 3.49-
3.36 (m, 1.5H), 3.28-3.11 (m, 2H), 3.08-2.93 (m, 1.5H), 2.80-
2.65 (m, 2H), 1.35 (s, br, 2H), 1.12 (s, 7H); 13C NMR (75 MHz,
CDCl3) δ 173.6, 169.0, 153.6, 137.2, 136.3, 129.7, 128.6, 128.7,
128.4, 127.1, 126.6, 80.7, 59.3, 58.8, 45.0, 37.4, 36.9, 33.9, 27.9;
IR (neat/NaCl) 2945 (br), 1735, 1701, 1665 cm-1; LRFAB MS
m/e (rel intensity) 439 (MH+, 17), 383 (MH+ - C4H8, 100), 337
(M+ - C5H9O2, 27), 265 (19), 154 (100); HRFAB calcd for
C
C
25H31N2O5 (M + 1) 439.2233, found 439.2223. Anal. Calcd for
25H30N2O5: C, 68.48; H, 6.90; N, 6.39. Found: C, 68.20; H,
C
66H101N18O13S1 (M + 1) 1385.7516, found 1385.7452. The
mass spectral data for analogue 30 is as follows: LRFAB MS
(rel intensity) 1373.3 (MH+, 100), 687.4 (25), 492.2 (25);
HRFAB calcd for C65H101N18O13S1 (M + 1) 1373.7516, found
1373.7495. The mass spectral data for analogue 31 is as
follows: LRFAB MS (rel intensity) 1373.7 (MH+, 100), 687.4
(25); HRFAB calcd for C65H101N18O13S1 (M + 1) 1373.7516,
found 1373.7491.
7.16; N, 6.26.
Gen er a l P r oced u r e for Solid -P h a se P ep tid e Syn th esis.
Substance P analogues 29-31 were made using standard
solid-phase peptide synthesis techniques. The amine group of
each monomer added was protected with a t-Boc group. Four
of the eleven amino acids were also protected by a “permanent”
protecting group. To this end, oxide, 4-toluenesulfonyl (Tos),
9-xanthenyl (Xan), and 2-chlorobenzyloxycarbonyl (2-LZ) pro-
tecting groups were used to protect the side chain functionality
in the methionine, arginine, glutamine, and lysine residues,
respectively.
In each case (29-31), the synthesis began with 0.571 g (0.4
mmol) of t-Boc-Met(O) that had already been linked to the
MBHA resin (p-methylbenzyhydrylamine-resin‚HCl, 0.7 mmol/
g). Prior to the coupling sequence, the polymer was swelled
by allowing it to stand in dichloromethane for 30 min. The
t-Boc group was removed with the use of a freshly prepared
50% TFA/dichloromethne solution containing 2-5% of anisole
as a Boc scavenger. Following deprotection, the resin was
neutralized by washing it with 5% DIEA (diisopropylethyl-
amine).
Ack n ow led gm en t. We thank the National Insti-
tutes of Health (RO1 GM5324001A1) for their generous
financial support. We thank Professor Garland R.
Marshall and his group for the use of their solid-phase
peptide synthesis facilities. Finally, 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.
Su p p or tin g In for m a tion Ava ila ble: Included are the
COSY and NOESY spectra for products 16 and 19a , the
biological data for analogues 29-31, and the hard NMR
spectral data (as evidence of purity) for all compounds lacking
CHN analysis. This material is available free of charge via
the Internet at http://pubs.acs.org.
The coupling steps were accomplished with the common
TbTu, HOBt, and DIEA conditions in DMF solvent. For
coupling reactions using a primary amine, the reactions were
allowed to stir for 30 min. The subsequent washing process
involved DMF (1 min), 2-propanol (1 min), and dichlo-
romethane three times (1 min each). The Kaiser test was used
to ensure that the coupling reactions proceeded to complete
conversion.
J O991649T
The coupling of the constrained building blocks proved to
be slightly more difficult because of the more sterically
(18) Carpino, L. A.; Mansour, E. M. E.; Sadat-Aalaee, D. J . Org.
Chem. 1991, 56, 2611.