Constrained peptidomimetics for TRH: cis-peptide bond analogs

Article abstract of DOI:10.1016/S0040-4020(00)00886-3

Two constrained analogs of the tripeptide hormone thyroliberin (TRH) have been synthesized. In both, the HisPro peptide bond of the hormone has been 'locked' into a cis-conformation. In the first analog, the constraint used was identical to the constraint used in an earlier trans-peptide bond analog that was a potent agonist for the TRH endocrine receptor TRH-R₁. The second analog was built in order to take advantage of a tetrazole ring as the cis-peptide bond constraint. Neither analog showed agonist or antagonist behavior toward TRH-R₁ suggesting that the cis-peptide bond conformation may not play a role in the affinity of TRH for TRH-R₁. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0040-4020(00)00886-3

TETRAHEDRON
Pergamon
Tetrahedron 56 (2000) 9791±9800
Constrained Peptidomimetics for TRH: cis-Peptide Bond Analogs
Yunsong Tong,a Jacek Olczak,b,d Janusz Zabrocki,b,d Marvin C. Gershengorn,c
Garland R. Marshallb and Kevin D. Moellera,
*
aDepartment of Chemistry, Washington University, St. Louis, MO 63130, USA
bDepartment of Molecular Biology and Pharmacology, and the Center for Molecular Design, Washington University,
St. Louis, MO 63130, USA
cDivision of Molecular Medicine, Department of Medicine, Cornell University Medical College and The New York Hospital,
New York, NY 10021, USA
dInstitute of Organic Chemistry, Technical University, Zeromiskiego, 116, 90-924 Lodz, Poland
Received 28 April 2000; accepted 14 August 2000
AbstractÐTwo constrained analogs of the tripeptide hormone thyroliberin (TRH) have been synthesized. In both, the HisPro peptide bond
of the hormone has been `locked' into a cis-conformation. In the ®rst analog, the constraint used was identical to the constraint used in an
earlier trans-peptide bond analog that was a potent agonist for the TRH endocrine receptor TRH-R1. The second analog was built in order to
take advantage of a tetrazole ring as the cis-peptide bond constraint. Neither analog showed agonist or antagonist behavior toward TRH-R1
suggesting that the cis-peptide bond conformation may not play a role in the af®nity of TRH for TRH-R1. q 2000 Elsevier Science Ltd. All
rights reserved.
Thyroliberin (TRH) is the hypothalamic tripeptide
(p-GluHisPro-NH2) that controls the release of thyroid
stimulating hormone from the pituitary gland.1 In addition,
it is active in the central nervous system.2 Because of its
biological relevance, TRH has been extensively studied, and
a number of proposals have been made concerning the
conformation responsible for the binding of TRH to its
endocrine receptor (TRH-R1).3 Marshall and coworkers
used the active analog approach to computer assisted drug
design to predict an active conformation for the binding of
TRH to TRH-R1.4,5 In this work, two main conformational
possibilities were proposed. Of these two possibilities, the
more frequently populated conformation is illustrated in
Scheme 1 (1). In order to probe whether this proposed
conformation represented an `active' conformation of the
hormone, a series of analogs having the overall structure
of 2 were examined. Analogs having only the pyroglutamate
region constrained (mˆ0),6 analogs having only the HisPro
region of the molecule constrained (nˆ0),7 and fully
constrained analogs8 have all been studied. The studies
have resulted in several constrained analogs that bind the
TRH-R1 receptor more tightly than their unconstrained
counterparts7,8 and at least one analog that also shows
improved potency relative to its unconstrained counterpart.7
The conformation of TRH illustrated by 1 was determined
by assuming that all of the analogs used in the original basis
set for the active analog approach had a trans-peptide bond
connecting the central amino acid to the proline moiety.
This assumption represented a very reasonable starting
point. However, because proline residues involve a tertiary
amide, the presence of a proline in a peptide is known to
destabilize trans-peptide bonds relative to their cis-peptide
bond counterparts. A signi®cant percentage (ca. 10%)9 of
the proline residues found in natural proteins have a cis-
conformation for the N-terminal peptide bond. In a number
of cases, the cis-conformation for this peptide bond has been
implicated as the conformation responsible for receptor
binding and the observed biological activity.10 Could a
cis-peptide bond conformation involving the proline
amide bond also play a role in the binding of TRH to
TRH-R1? While the observation that analogs constraining
the HisPro region of TRH into a trans conformation were
potent agonists for TRH-R1 certainly implicated that the
trans-conformation was important for potency, one could
Scheme 1.
Keywords: tripeptide hormone thyroliberin; tetrazole ring; TRH endocrine
receptor.
* Corresponding author. Tel.: 1314-935-4270; fax: 1314-935-4481;
e-mail: moeller@wuchem.wustl.edu
0040±4020/00/$ - see front matter q 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020(00)00886-3

9792
Y. Tong et al. / Tetrahedron 56 (2000) 9791±9800
the cis-peptide bond mimetic (4/Scheme 3).11 In this case,
the constraint incorporated the nitrogen of the peptide bond;
an alteration that forced a replacement of the proline amino
acid at the C-terminal end of the hormone. For TRH,
replacements of this type can be made without losing
activity. TRH analogs having a methylalanine (which has
the same steric parameters as the proline) group in the
C-terminal position retain both their af®nity and potency
for TRH-R1.12
Scheme 2.
Two approaches to the synthesis of 3 seemed reasonable
(Scheme 4). Both plans called for taking advantage of
what had been learned during the synthesis of 2a in order
to convert the a bicyclic lactam 5 into the desired TRH
analog. In turn, the bicyclic building block would be
constructed from the known allylated proline 8.13 The two
routes differed in how the lactam ring constraint was to be
built. For path a, the amine in 8 would be converted into an
amide and then the two sidechains coupled to form the
lactam ring. For path b, the sidechain in 8 would be devel-
oped into a longer chain carboxylic acid derivative and then
the lactam ring synthesized by an intramolecular amide
formation. On paper both routes had their advantages.
Path a appeared more divergent and compatible with the
rapid incorporation of substituents at the carbon beta to
the amide carbonyl in 5, while path b was directly analogous
to the route used by others to make unsubstituted (RˆH)
building blocks like 5.14,15
Scheme 3.
imagine a scenario where TRH initially bound TRH-R1 with
a cis-peptide bond conformation in the HisPro region and
then equilibrated to a trans-peptide bond conformation
during the reorganization of the ligand/receptor complex
that leads to signaling. In this case, the potency of the
constrained trans-peptide mimetics might have resulted
from the analogs either inducing a ®t with the receptor or
`draining off' an equilibrium between various receptor
conformations. If this were the case, then a constrained
analog ®xed into a cis-conformation might prove to be an
antagonist for the receptor. Such an analog would bind
TRH-R1, but then be unable to undergo the conformational
change needed for potency.
In practice, efforts along path a were not successful. Three
different approaches were attempted including an intra-
molecular aldol reaction (substrate 9), a ring closing ole®n
metathesis reaction (substrates 10 and 11),16 and an intra-
molecular Barbier reduction (substrate 12).17 (Scheme 5)
Neither the intramolecular aldol nor the ring closing meta-
thesis reaction led to the formation of any cyclized product.
A variety of conditions were tried for each, but in each case
only starting material was recovered. The lack of reactivity
for these substrates was rationalized by suggesting that the
steric size of the quaternary center forced the amide bond to
adopt a conformation that did not allow for the two ends of
the proposed cyclizations to approach each other. The
Barbier route to effecting the cyclization (SmI2) did lead
to low yields of the bicyclic product (23±34%), but the
reaction could not be optimized further.
Initially, two approaches were taken in order to investigate
this possibility. In one, a TRH analog having a cis-peptide
bond constraint (3) identical to the one used successfully for
the trans-peptide bond analog 2a was designed (Scheme 2).
The only difference between the two analogs was the posi-
tion of the primary amide group on the fused bicyclic ring
system used to lock the C-terminal region of the molecule in
place. In the second approach, a tetrazole ring was used as
Scheme 4.

Y. Tong et al. / Tetrahedron 56 (2000) 9791±9800
9793
CyclohexylAla2TRH. Both are a factor of 102±103 less
active than the natural hormone in terms of binding and
potency, both completely displace TRH from TRH-R1 at
high concentrations, and both are full agonists. In addition,
conclusions about the compatibility of the bridges are made
by comparing the constrained analog to its acyclic counter-
part not having the constraint. Conclusions about 2a were
made by directly comparing the data obtained with that
obtained using CyclohexylAla2TRH.7 Along these lines,
conclusions about 3 would be made by directly comparing
the data obtained for 3 with the data obtained for Phe2TRH.6
Second, the cuprate addition reaction on 14 needed to be
stirred well and kept below 2158C. Above 2158C, the reac-
tion led to substantial amounts of a byproduct tentatively
assigned as 16. Even when the reaction was optimized, a
10% yield (assuming 16) of this product was obtained along
with a 61% isolated yield of the desired Michael product.
Scheme 5.
Fortunately, path b proved to be much more useful. As in the
earlier syntheses of unsubstituted analogs,13,14 the 2-allyl-
proline starting material was treated under ozonolysis
conditions and the resulting aldehyde subjected to a Wittig
reaction to form an a,b-unsaturated ester (Scheme 6). At
this point, the substituent needed for the central amino acid
of the TRH derivative was added with the use of a higher
order cuprate.18 Two aspects of this reaction deserve
comment. First, the cyclohexyl substituent utilized in 2a
was replaced with a phenyl ring in 3, because the required
phenyl cuprate could be generated more easily and because
Phe2TRH showed roughly the same biological activity as
The conversion of 15 into desired bicyclic intermediate 5
was straightforward (Scheme 7). The Cbz protecting group
was removed using hydrogen in methanol and Pd/BaSO4 as
the catalyst, and the six-member ring lactam was formed
using triethylamine in toluene. An 85% yield of the bicyclic
product was obtained over the two steps. The required car-
bonyl functionality alpha to the amide was then added by
deprotonating the amide and quenching the resulting enolate
with oxygen. Addition of tri¯uoroacetic anhydride to the
workup ensured formation of the dicarbonyl equivalent.19
Scheme 6.
Scheme 7.

9794
Y. Tong et al. / Tetrahedron 56 (2000) 9791±9800
using a two step procedure. Attempts to oxidize the alcohol
directly to an acid using either a Jones oxidation or PDC in
DMF met with failure. For the two step procedure, a
methanol workup was used in order to directly form methyl
ester 23. The yield of this procedure was not optimized.
Finally, treatment of methyl ester 23 with ammonia in
methanol afforded analog 3.
Analog 3 was evaluated both as a potential agonist and a
potential antagonist for TRH-R1. For agonist behavior,
monkey kidney COS-1 cells transiently expressing TRH
receptors and pre-labelled with [3H]-inositol were incubated
with various doses (1.0±1000 mM) of 3.20 Inositol phos-
phate formation was measured. In no case did the presence
of 3 lead to an increase in the amount of inositol phosphate
generated, and it was concluded that 3 was not an agonist for
TRH-R1. For antagonist behavior, the cells were incubated
with both 3 and varying concentrations of TRH itself. In no
case did the presence of 3 alter the extent of inositol phos-
phate released. Based on this result, it was concluded that 3
was also not an antagonist for TRH-R1.
Scheme 8.
However, this initial pass at the synthesis failed when treat-
ment of the a-ketoamide with ammonia in methanol failed
to generate the desired product 18. From the crude spectral
data, it was clear that the presence of the methyl ester inter-
fered with the reaction.
A similar observation was made for the analog using a tetra-
zole ring to constrain the HisPro region of the hormone (4).
As illustrated in Scheme 10, the tetrazole ring was intro-
duced into the molecule by treating a Cbz-HisAla-OBz
dipeptide derivative with PCl5, quinoline, and HN3. The
N-terminal Cbz group was then removed from the
constrained building block with the use of HBr in acetic
acid and the resulting amine coupled to a Cbz protected
pyroglutamic acid using a mixed anhydride as the activating
group. Cleavage of the N-terminal Cbz group and the
C-terminal benzyl ester afforded tripeptide 26. The desired
TRH analog was completed by converting the C-terminal
acid to a primary amide with the use of freshly prepared
ammonium salt of HOBt and DCC and then deprotecting
the imidazole benzyl protecting group. As in the case of 3,
analog 4 did not show any af®nity for TRH-R1.
In order to avoid this problem, the methyl ester in 5 was
selectively reduced with lithium borohydride and the result-
ing alcohol protected as the t-butyldimethylsilyl ether. This
transformation both increased the yield obtained for func-
tionalization of the carbon alpha to the amide carbonyl and
allowed for conversion of the resulting dicarbonyl product
20 into enamine 21 (Scheme 8). As in earlier cases, the
enamine group was surprisingly stable, and compound 21
was puri®ed by chromatography through triethylamine
treated silica gel.
Conversion of enamine 21 into the cis-peptide bond TRH
analog 3 proceeded as outlined in Scheme 9. The enamine
functionality was coupled to the N-terminal pyroglutamate
group by means of the mixed anhydride intermediate using
isobutylchloroformate. It should be noted that other
coupling procedures using the standard HOBT/EDCI,
HOBT/TBTU, and BOP-Cl/Et3N conditions were not effec-
tive. Following the coupling reaction, the silyl protecting
group was removed and the resulting alcohol oxidized
The observation that neither 3 nor 4 showed any af®nity for
TRH-R1 suggested that the conformation of TRH having a
cis-peptide in the HisPro region does not play a role in the
binding of TRH to its endocrine receptor, TRH-R1. In this
case, the data for the two analogs worked together to make
the argument more compelling. Typically, a lack of activity
Scheme 9.

Y. Tong et al. / Tetrahedron 56 (2000) 9791±9800
9795
Scheme 10.
for a single constrained analog can not be used to draw
conclusions about the biological relevance of the conform-
ation being mimicked. The conformational constraints
added to the analog simply make other changes in the
molecule. For example, the methylene bridge used in 3 to
constrain the cis-peptide bond also added steric bulk to the
central portion of the analog and ®xed the orientation of the
side chain of residue two. In 4, the tetrazole ring used to
constrain the cis-peptide bond both added steric bulk to the
amide region of the molecule and caused a loss in the hydro-
gen-bonding capacity of the analog. However, when taken
together the lack of activity for multiple analogs can shed
light on the biological relevance of a proposed conform-
ation. In the current example, both analogs 3 and 4 may
have had multiple differences relative to the natural
hormone, but only the cis-peptide bond was common to
both. The fact that the analogs behaved similarly implied
that it was the presence of this cis-peptide bond that inter-
fered with TRH/TRH-R1 binding. The synthesis and testing
of a third analog ([AzPro3]-TRH) has recently added
further support to this conclusion (Zhang, Gershengorn
and Marshall, unpublished). [AzPro3]-TRH proved to be
35-fold less active than the natural hormone. Azaproline
(AzPro) has been proposed by Zouikri et al.21 to enhance
the stability of the cis-amide conformation and has been
shown using both theory and NMR (Berglund, Zhang and
Marshall, unpublished) to do so in the TRH analog.
Experimental
(2R)-N-(Benzyloxycarbonyl)-2-[(E)-30-(carbomethoxy)-
prop-20-enyl]proline methyl ester (14). In a 250 mL ¯ask,
10.2 g (0.0336 mol) of the Cbz protected 2-allylproline
derivative 13 was dissolved in 70 mL of dichloromethane.
Ozone was bubbled though the solution at 2788C. The solu-
tion turned blue 70 min later. After an additional 5 min, the
bubbling of ozone was stopped and nitrogen was bubbled
through the solution until the blue color disappeared. To the
2788C solution was added 25.8 g (0.0773 mol, 2.3 equiv.)
of methyl (triphenylphosphoranylidene)acetate and the
reaction mixture was warmed to room temperature. After
20 h of stirring, 100 mL of water and 150 mL of dichloro-
methane were added to the reaction mixture. The aqueous
layer was separated and washed twice with 100 mL of
dichloromethane. The combined organic layers were dried
over MgSO4 and concentrated. The crude product was
chromatographed through 300 g of silica gel using ether/
hexane (8:2) as the eluant to afford 9.69 g (80%) of pure
product as a pale yellow oil. 1H NMR (300 MHz, CDCl3) A
1:1 mixture of rotamers was observed. d 7.36±7.27 (m, 5H),
6.89 (ddd, 1H, Jˆ7.0, 15.2, 15.2 Hz), 5.92 (d, 0.5H,
Jˆ15.7 Hz), 5.83 (d, 0.5H, Jˆ15.7 Hz), 5.22 (a of ab,
0.5H, Jabˆ12.2 Hz), 5.13 (a0 of a0b0, 0.5H, Ja b ˆ12.5 Hz),
0
0
5.10 (b0 of a0b0, 0.5H, Ja b ˆ12.6 Hz), 5.01 (b of ab, 0.5H,
Jabˆ12.1 Hz), 3.80±3.65 (m, 6H, two s at 3.73 and 3.72
representing 5H), 3.56±3.46 (m, 2H, one s at 3.46 represent-
ing 1H), 3.24 (dd, 0.5H, Jˆ7.1, 14.5 Hz), 3.00 (a of abx,
0.5H, Jaxˆ7.0 Hz, Jabˆ14.5 Hz), 2.92±2.76 (m, 1H), 2.16±
1.78 (m, 4H); 13C NMR (125 MHz, CDCl3) d 173.9, 166.6,
154.4, 143.8, 143.4, 136.7, 128.4, 128.3, 127.9, 127.6,
124.6, 67.8, 67.3, 66.8, 52.5, 52.3, 51.5, 49.0, 48.2, 38.2,
37.4, 36.8, 35.9, 23.1, 22.6; IR (neat/NaCl) 2960, 2882,
1707, 1659, 1406, 1349, 1244, 1180, 1124, 1026, 998,
744, 703 cm21; LRFAB MS m/e (rel. intensity) 362
(MH1, 100), 302 (M12C2H3O2, 12), 258 (24), 226
0
0
The observations made using all three TRH analogs combined
with the fact that the closely related trans-peptide mimetic 2a
bound TRH-R1 more tightly and was more potent than its
unconstrained counterpart con®rmed that the initial modelling
efforts were correct to assume a trans-amide bond conform-
ation for the HisPro region of TRH. It appears clear that the
design of both agonists and potential antagonists for TRH-R1
in the future will need to incorporate a trans-peptide bond in
the HisPro region of the hormone.

9796
Y. Tong et al. / Tetrahedron 56 (2000) 9791±9800
(M12C8H7O2, 41), 154 (50); HRFAB calcd for C19H24N1O6
(M11) 362.1603, found 362.1583; Anal. calcd for
C19H23N1O6: C, 63.15, H, 6.41, found: C, 63.12, H, 6.57.
isomers was observed. d 7.37±7.17 (m, 5H), 3.55 (s, 2.2H),
3.51 (s, 2.2H), 3.44 (s, 1.6H), 3.27±3.12 (m, 1H), 2.95±2.76
(m, 1H), 2.68±2.37 (m, 3H), 2.28±2.11 (m, 2H), 1.98±1.63
(m, 4H); 13C NMR (125 MHz, CDCl3) d 174.0, 169.4,
142.1, 128.5, 127.2, 126.8, 67.5, 52.5, 44.8, 39.2, 38.6,
37.4, 36.5, 20.6. The crude product was carried on to the
next step without further characterization.
(2R)-N-(Benzyloxycarbonyl)-2-[20-(phenyl)-30-(carbo-
methoxy)-propyl]proline methyl ester (15). Copper (I)
cyanide was dried in an Abderhalden apparatus (acetone
re¯uxing at 568C) overnight. To a ¯amed dried 250 mL
¯ask was added 2.938 g (0.0328 mol, 2 equiv.) of copper
cyanide, which was gently ¯ame-dried again under high
vacuum. The ¯ask was then connected to a mechanical
stirrer and protected by an argon balloon. To a suspension
of copper cyanide in 20 mL of dry ether and 20 mL of THF
was added 36.4 mL (0.0656 mol, 4 equiv.) of 1.8 M (cyclo-
hexane/ether) phenyllithium solution at 2788C. The ¯ask
was transferred into a ice bath and cooled back to 2788C
10 min later. To the reaction mixture was added 4.03 mL
(0.0328 mol, 2 equiv.) of BF3´Et2O. After 10 min,
compound 14 (5.9 g/16.4 mmol) in 20 mL of ether was
added to the ¯ask. The reaction temperature was raised to
about 2458C (CH3CN/dry ice bath) 30 min later and further
warmed up to about 2208C (CCl4/dry ice bath) after an
additional 100 min. The temperature was kept at 2208C
for 40 min before the reaction was quenched by 150 mL
of 10% NH4OH/NH4Cl. The mixture was then stirred at
room temperature for 30 min and was diluted with
160 mL of ether afterwards. The organic layer was sepa-
rated, and the aqueous layer was extracted by 160 mL of
ether two more times. The combined organic layers were
washed with 10% NH4OH/NH4Cl and 5% NaHCO3 and
dried over MgSO4 before being concentrated. The crude
product was chromatographed through 550 g of silica gel
using ether/hexane (8:2) as the eluant to afford 4.40 g (61%)
of pure product as a yellow oil along with 670 mg of bypro-
duct tentatively assigned as 16 (10% if 16). 1H NMR
(300 MHz, CDCl3) A mixture of rotamers and stereochem-
ical isomers was observed. d 7.41±7.16 (m, 10H), 5.26±
5.09 (m, 2H), 3.77±3.63 (m, 3H, one s at 3.65 representing
1H), 3.49, 3.46, 3.40 (3 s, 5H), 3.22±3.15 (m, 1H), 2.78 (dd,
0.5H, Jˆ4.2, 14.7 Hz), 2.66±2.34 (m, 3.5H), 1.84±1.61
(m, 4H); 13C NMR (75 MHz, CDCl3) isomers observed d
174.9, 174.5, 172.2, 171.8, 154.7, 144.9, 144.4, 136.9,
136.3, 129.0, 128.6, 128.5, 128.4, 128.1, 128.0, 127.8,
127.6, 126.7, 126.5, 68.8, 67.9, 67.2, 66.7, 52.3, 52.2,
51.3, 49.4, 48.5, 43.2, 43.0, 39.8, 38.6, 38.0, 37.9, 37.0,
35.6, 22.6, 22.1; IR (neat/NaCl) 3065, 3030, 2953, 2882,
1736, 1701, 1602, 1497, 1447, 1406, 1342, 1244, 1167,
1124, 1019, 913, 857, 772, 695 cm21; LRFAB MS m/e
(rel. intensity) 440, (MH1, 100), 380 (M12C2H3O2, 49),
336 (66), 304 (M12C8H7O2, 54), 246 (27); HRFAB calcd
for C25H30N1O6 (M11) 440.2073, found 440.2076; Anal.
calcd for C25H29N1O6: C, 68.32, H, 6.65, N, 3.19, found:
C, 68.38, H, 6.78, N, 3.06.
(6R)-1-Aza-6-carbomethoxy-4-phenyl-2-oxobicyclo[4.3.0]-
nonane (5). To a solution of 3.2 g of crude deprotected
material made in the previous step in 135 mL of toluene
was added 8.7 mL of Et3N. The solution was heated to
re¯ux for 21 h. The reaction mixture was then concentrated.
The crude product (3.6 g) was chromatographed through
180 g of silica gel using ether/dichloromethane/methanol
(8:1:1) as the eluant to afford 2.4 g (85%, two steps
combined) of pure product as a yellow oil. 1H NMR
(300 MHz, CDCl3) d 7.33±7.18 (m, 5H), 3.86±3.79
(m, 1H), 3.55 (ddd, 1H, Jˆ2.7, 9.4, 9.4 Hz), 3.44 (s, 3H),
3.28±3.23 (m, 1H), 2.71±2.62 (m, 2H), 2.43±2.37 (m, 1H),
2.20 (a of abx, 1H, Jaxˆ6.1, Jabˆ14.0 Hz), 1.98±1.68
(m, 4H); 13C NMR (75 MHz, CDCl3) d 173.9, 169.5,
142.1, 128.4, 127.2, 126.8, 67.4, 52.5, 44.8, 39.1, 38.6,
37.3, 36.5, 20.5; IR (neat/NaCl) 3023, 2953, 2889, 1736,
1652, 1497, 1434, 1349, 1216, 1173, 1118, 1019, 759,
703 cm21; LRFAB MS m/e (rel. intensity) 274 (MH1,
100), 214 (M12C2H3O2, 23), 131 (10); HRFAB calcd for
C16H20N1O3 (M11) 274.1443, found 274.1431; Anal. calcd
for C16H19N1O3: C, 70.31, H, 7.01, found: C, 69.88, H, 7.27.
(6R)-1-Aza-6-hydroxymethyl-4-phenyl-2-oxobicyclo[4.3.0]-
nonane. To a 08C solution of 2.423 g (8.864 mmol) of 5 in
35 mL of THF was added 6.65 mL (0.0133 mol, 1.5 equiv.)
of 2 M LiBH4. The reaction was warmed to room tempera-
ture and stirred overnight. The reaction was then recooled to
08C, and 35 mL of 1N HCl, 10 mL of brine, and 35 mL of
ethyl acetate added to the reaction solution. After separa-
tion, the aqueous layer was extracted twice with 35 mL
of ethyl acetate. The combined organic layers were dried
over Na2SO4 and concentrated. The crude product was
chromatographed through 100 g of silica gel using ether/
dichloromethane/methanol (8:1:1) as the eluant to afford
1
2.024 g (93%) of pure product as a white solid. H NMR
(300 MHz, CDCl3) A mixture of two isomers was observed.
d 7.34±7.20 (m, 5H), 4.54±4.47 (m, 0.1H), 4.06±3.94
(m, 1.9H), 3.61(d, br,1H, Jˆ10.9 Hz), 3.40±3.25 (m, 2H),
2.89 (tt, 1H, Jˆ4.4, 12.5 Hz), 2.65±2.27 (m, 3H), 2.06 (ddd,
1H, Jˆ2.5, 5.0, 11.1 Hz), 1.97±1.76 (m, 4H); 13C NMR
(150 MHz, CDCl3) d 171.5, 143.2, 128.6, 126.8, 126.7,
66.2, 65.6, 43.1, 39.4, 37.0, 36.7, 34.8, 20.2; IR (neat/
NaCl) 3340 br, 3023, 2945, 1645, 1462, 1420, 1237,
1180, 1069, 1040, 998, 913,765, 695 cm21; LRFAB MS
m/e (rel. intensity) 246 (MH1, 49), 214 (M12CH3O, 18),
154 (100); HRFAB calcd for C15H20N1O2 (M11) 246.1494,
found 246.1487; Anal. calcd for C15H19N1O2: C, 73.44, H,
7.81, N, 5.71, found: C, 73.56, H, 8.04, N, 5.68.
(2R)-2-[20-(Phenyl)-30-(carbomethoxy)-propyl]proline
methyl ester. To a solution of 4.57 g (0.0104 mol) of 15 in
100 mL of methanol was added 1.14 g of 5% palladium in
barium sulfate. The reaction mixture was hydrogenated with
a hydrogen balloon. After being stirred overnight, the
mixture was ®ltered through celite. The ®ltrate was then
concentrated to give 3.2 g of crude product as a pale yellow
oil, which was used in the following step without further
puri®cation. 1H NMR (300 MHz, CDCl3) A mixture of two
(6R)-1-Aza-6-[(t-butyldimethylsiloxy)methyl]-4-phenyl-2-
oxobicyclo[4.3.0]nonane (19). To a 08C solution of 6.4 g
(0.026 mol) of the alcohol made above in 110 mL of
dichloromethane was added 7.2 mL (0.030 mol,
1.2 equiv.) of TBDMSOTf and 6.1 mL (0.052 mol,
2 equiv.) of 2,6-lutidine. The reaction was kept at 08C for

Y. Tong et al. / Tetrahedron 56 (2000) 9791±9800
9797
1 h and then diluted with 100 mL of dichloromethane. The
organic solution was washed two times with 100 mL of 5%
aqueous citric acid, 100 mL of 5% NaHCO3 and once with
brine. The organic layer was then dried and concentrated.
The crude product was chromatographed through 400 g of
silica gel using ether/dichloromethane/methanol (8:1:1) as
the eluant to afford 9.0 g (96%) of pure product as a color-
less oil. 1H NMR (300 MHz, CDCl3) d 7.36±7.19 (m, 5H),
3.98 (ddd, 1H, Jˆ8.4, 12.2, 12.2 Hz), 3.43 (a of ab, 1H,
Jabˆ9.9 Hz), 3.36 (b of ab, 1H, Jabˆ9.9 Hz), 3.36±3.29
(m, 1H), 2.95±2.88 (m, 1H), 2.53±2.50 (m, 2H), 2.23±
1.74 (m, 6H), 0.90 (s, 9H), 0.06 (s, 3H), 0.05 (s, 3H); 13C
NMR (150 MHz, CDCl3) d 170.8, 143.6, 128.7, 126.7, 67.2,
64.9, 43.4, 39.4, 37.8, 37.0, 34.9, 25.8, 20.2, 18.2, 25.5,
25.6; IR (neat/NaCl) 3459, 3030, 2945, 2882, 1652,
1455, 1426, 1335, 1258, 1208, 1103, 998, 843, 772, 703,
660 cm21; LRFAB MS m/e (rel. intensity) 360 (MH1, 100),
214 (M12C7H17OSi, 43)156 (8), 131 (23); HRFAB calcd
for C21H34N1O2Si (M11) 360.2359, found 360.2350; Anal.
calcd for C21H33N1O2Si: C, 70.15, H, 9.25, found: C, 70.21,
H, 9.53.
was stirred at 508C for 6 days. If during the 6 days, the
ammonia pressure inside the ¯ask decreased (determined
based on the appearance of the septum), ammonia gas was
bubbled through the solution again. The reaction solution
was then concentrated. The crude product was chromato-
graphed through 25 g of silica gel (slurry neutralized with 4
drops of Et3N) using ether/hexane as the eluant to afford
1
0.290 g (69%) of pure product as a yellow resin. H NMR
(300 MHz, CDCl3) d 7.50±7.27 (m, 5H), 4.14 (s, br, 2H),
3.79 (a of ab, 1H, Jabˆ9.6 Hz), 3.76±3.67 (m, 2H), 3.45 (b of
ab, 1H, Jabˆ9.3 hz), 2.97 (d, 1H, Jˆ16.1 Hz), 2.75 (d, 1H,
Jˆ16.0 Hz), 2.48 (ddd, 1H, Jˆ3.1, 6.4, 12.4 Hz), 2.10±2.02
(m, 2H), 1.79±1.72 (m, 1H), 0.90 (s, 9H), 0.08 (s, 2.3H),
0.06 (s, 3.7H); 13C NMR (150 MHz, CDCl3) d 161.7, 140.0,
131.0, 128.6, 127.2, 126.6, 110.5, 63.6, 62.4, 45.3, 35.8,
35.7, 25.7, 21.4, 18.1, 25.6, 25.7; IR (neat/NaCl) 3445
br, 3346, br, 3057, 2945, 2868, 1630, 1567, 1455, 1349,
1258, 1103, 1005, 843, 722, 703 cm21; LRFAB MS m/e
(rel. intensity) 373 (MH1, 67), 227 (M12C7H17OSi, 100),
156 (23); HRFAB calcd for C21H33N2O2Si (M11)
373.2311, found 373.2278.
(6R)-1-Aza-6-[(t-butyldimethylsiloxy)methyl]-3-hydroxy-
4-phenyl-2-oxobicyclo[4.3.0]non-3-ene (20). To a 2788C
solution of 0.331 g (0.921 mmol) of 19 in 18 mL of
dichloromethane was added 5.52 mL (2.76 mmol,
3 equiv.) of 0.5 M KHMDS in toluene. After 1 h, oxygen
was bubbled through the solution for 75 s and the reaction
mixture was then quenched with 0.21 mL (1.5 mmol) of
TFAA. Oxygen ¯ow was stopped 50 seconds later, and
the reaction was protected under argon. The reaction was
then warmed up to room temperature over 45 min. To the
mixture was added 20 mL of NaHSO3, and the solution was
then extracted three times with 20 mL of ethyl acetate. The
combined organic layers were dried over Na2SO4 and
concentrated. The crude product (0.67 g) was chromato-
graphed through 30 g of silica gel using ether/hexane (8:2)
as the eluant to afford 0.247 g (72%) of pure product as a
(6R,12S)-1-Aza-3-(pyroglutamylamino)-6-[(t-butyldimethyl-
siloxy)methyl]-4-phenyl-2-oxobicyclo[4.3.0]non-3-ene. To a
2208C (CCl4/dry ice bath) solution of 0.078 g
(0.60 mmol, 1.1 equiv.) of pyroglutamic acid in 9 mL of
THF was added 0.076 mL (0.60 mmol, 1.1 equiv.) of
NEM and 0.078 mL (0.60 mmol, 1.1 equiv.) of isobuytl-
chloroformate. After the reaction mixture was stirred at
2208C for 40 min, compound 21 (0.204 g, 0.548 mmol,
dissolved in 5 mL of THF) was added to the ¯ask. The
reaction was warmed to room temperature and 3 h later
the solution was concentrated. The residue was then
dissolved in 40 mL of chloroform and washed two times
with 25 mL of 5% citric acid and once with brine solution.
The organic layer was dried over Na2SO4 and concentrated.
The crude product (0.41 g) was chromatographed through
20 g of silica gel using ether/dichloromethane/methanol
(70:15:15) as the eluant to afford 0.196 g (74%) of pure
1
brown oil. H NMR (300 MHz, CDCl3) d 7.77 (d, 2H,
1
Jˆ7.1 Hz), 7.48±7.43 (m, 2H), 7.35±7.31 (m, 1H), 7.00
(s, 1H), 3.79±3.70 (m, 2H), 3.74 (a of ab, 1H,
Jabˆ8.4 Hz), 3.48 (b of ab, 1H, Jabˆ9.0 Hz), 3.20 (a of ab,
1H, Jabˆ16.2 Hz), 2.78 (b of ab, 1H, Jabˆ16.1 Hz), 2.52
(ddd, 1H, Jˆ3.5, 6.0, 12.4 Hz), 2.17±2.08 (m, 2H), 1.83
(dd, 1H, Jˆ8.8, 11.1 Hz), 0.92 (s, 9H), 0.08 (s, 3H), 0.06
(s, 3H); 13C NMR (150 MHz, CDCl3) A mixture of tauto-
mers was observed. d 179.7, 161.4, 157.9, 145.1, 140.0,
138.2, 136.8, 133.9, 128.7, 128.4, 128.3, 128.1, 127.6,
127.2, 111.3, 66.8, 65.6, 64.0, 62.8, 45.1, 44.3, 35.8, 33.7,
31.5, 25.7, 21.7, 21.0, 18.1, 25.6; IR (neat/NaCl) 3318 br,
3058, 2960, 2861, 1645, 1462, 1349, 1258, 1208, 1103, 998,
842, 772, 695 cm21; LRFAB MS m/e (rel. intensity) 374
(MH1, 100), 284 (12), 228 (M12C7H17OSi, 89), 136
(14); HRFAB calcd for C21H32N1O3Si (M11) 374.2151,
found 374.2149; Anal. calcd for C21H31N1O3Si: C, 67.52,
H, 8.36, N, 3.75, found: C, 67.55, H, 8.70, N, 3.93.
product as a white solid. H NMR (300 MHz, CDCl3) d
8.63 (s, 1H), 7.70 (s, 1H), 7.36±7.27 (m, 5H), 4.06±4.02
(m, 1H), 3.98 (a of ab, 1H, Jabˆ9.9 Hz), 3.62±3.57 (m, 2H),
3.42 (b of ab, 1H, Jabˆ9.9 Hz), 3.27 (d, 1H, Jˆ17.7 Hz),
2.59 (d, 1H, Jˆ17.6 Hz), 2.46±2.15 (m, 4H), 2.03±1.95
(m, 2H), 1.79±1.64 (m,2H), 0.92 (s, 9H), 0.10 (s, 3H),
0.08 (s, 3H); 13C NMR (150 MHz, CDCl3) d 179.6, 170.8,
161.8, 142.1, 138.2, 128.3, 128.2, 127.4, 123.1, 63.8, 62.0,
57.7, 45.3, 37.4, 35.5, 29.2, 25.8, 25.0, 21.6, 18.1, 25.5,
25.6; IR (neat/NaCl) 3241 br, 2960, 2882, 1701, 1645,
1511, 1447, 1244, 1103, 1004, 836 cm21; LRFAB MS m/e
(rel. intensity) 506 (MNa1, 18), 484 (MH1, 100), 373
(MH12C5H5NO2, 96), 338 (M12C7H17OSi, 39), 241
(64), 227 (85); HRFAB calcd for C26H38N3O4Si (M11)
484.2631, found 484.2644.
(6R,12S)-1-Aza-3-(pyroglutamylamino)-6-hydroxymethyl-
4-phenyl-2-oxobicyclo[4.3.0]non-3-ene (22). The coupled
product synthesized in the previous step (0.926 g,
1.92 mmol) was dissolved in 8.2 mL of 2N H2SO4 and
25 mL of THF. The mixture was stirred overnight. To the
mixture was then added saturated Na2SO4 solution until the
pH of the reaction mixture reached 9 to 10. The solution was
evaporated in vacuo and the residue was dissolved in
(6R)-1-Aza-3-amino-6-[(t-butyldimethylsiloxy)methyl]-4-
phenyl-2-oxobicyclo[4.3.0]non-3-ene (21). Ammonia gas
was bubbled through a solution of 0.423 g (1.13 mol) of
20 in 32 mL of methanol at 508C for 20 min and then at
room temperature for 10 min. The gas inlet and outlet
needles were removed from the septum, and the mixture

9798
Y. Tong et al. / Tetrahedron 56 (2000) 9791±9800
methanol. The solid material was ®ltered, and the ®ltrate
was dried over Na2SO4 and concentrated. The crude product
was chromatographed through 50 g of silica gel using ether/
dichloromethane/methanol (2:1:1) as the eluant to afford
CDCl3) A mixture of rotamers was observed. d 8.51
(s, 1H), 7.40±7.26 (m, 6H), 4.03±3.99 (m, 1H), 3.85±3.77
(m, 4H), 3.72±3.62 (m, 1H), 3.56 (d, 0.9H, Jˆ17.2 Hz),
3.45 (d, 0.1H, Jˆ16.9 Hz), 2.93 (d, 0.1H, Jˆ17.1 Hz),
2.81 (d, 0.9H, Jˆ15.8 Hz), 2.46±2.41 (m, 1H), 2.38±1.92
(m, 6H), 1.72±1.67 (m, 1H); 13C NMR (150 MHz, CDCl3) d
179.6, 174.2, 170.1, 162.4, 142.1, 137.5, 128.6, 128.3,
127.4, 126.9, 123.9, 67.7, 57.6, 53.3, 45.3, 39.4, 37.6,
29.1, 25.1, 22.1; IR (neat/NaCl) 3255 br, 2953, 1701,
1
0.610 g (86%) of pure product as a white solid. H NMR
(300 MHz, CDCl3) d 8.56 (s, 1H), 7.60 (s, 1H), 7.36±7.27
(m, 5H), 4.05±4.01 (m, 1H), 3.90 (a of ab, 1H,
Jabˆ12.5 Hz), 3.77±3.71 (m, 1H), 3.74 (b of ab, 1H,
Jabˆ11.6 Hz), 3.65±3.58 (m, 1H), 2.96 (a of ab, 1H,
Jabˆ17.7 Hz), 2.81 (b of ab, 1H, Jabˆ18.0 Hz), 2.37±1.64
(m, 8H); 13C NMR (150 MHz, CDCl3) d 179.1, 172.8,
161.6, 143.7, 137.7, 128.6, 128.3, 127.3, 123.3, 66.6, 64.1,
57.4, 45.8, 38.8, 36.7, 29.1, 25.2, 22.4; IR (neat/NaCl) 3276
br, 2953, 2882, 1694, 1659, 1518, 1462, 1335, 1258, 1075,
1047, 913 cm21; LRFAB MS m/e (rel. intensity) 370 (MH1,
100), 338 (M12CH3O, 10), 259 (MH12C5H5NO2, 26), 227
(M12C6H8NO3, 13), 185 (22), 154 (16), 93 (43); HRFAB
calcd for C20H24N3O4 (M11) 370.1767, found 370.1749.
1659, 1511, 1441, 1279, 1230, 1110, 921, 766, 731 cm21
;
LRFAB MS m/e (rel. intensity) 398 (MH1, 27), 246 (27),
185 (100); HRFAB calcd for C21H24N3O5 (M11) 398.1716,
found 398.1703.
(6R,12S)-1-Aza-3-(pyroglutamylamino)-6-carboxamoyl-
4-phenyl-2-oxobicyclo[4.3.0]non-3-ene (3). Ammonia gas
was bubbled through a solution of 0.054 g (0.14 mmol) of
23 in 6 mL of dry methanol for 20 min at room temperature.
The ammonia inlet and outlet needles were then removed
from the septum, and the reaction mixture was then stirred at
408C for 4 days. The solution was concentrated and the
crude material was chromatographed through 7 g of silica
gel using ether/dichloromethane/methanol (2:1:1) as the
eluant to afford 0.027 g (51%) of pure product as a white
(6R,12S)-1-Aza-3-(pyroglutamylamino)-6-formyl-4-phenyl-
2-oxobicyclo[4.3.0]non-3-ene. To a solution of 0.281 g
(0.761 mmol) of 22 in 2 mL of dichloromethane was
added 0.134 g (1.14 mmol, 1.5 equiv.) of NMO, 0.38 g of
Ê
ground and dried 4 A molecular sieves, and 0.0214 g
1
(0.0609 mmol, 8 mol%) of TPAP. The mixture was stirred
for 3 h and then diluted with 10 mL of dichloromethane.
The suspension was passed through a funnel packed with
glass wool and the ®ltrate was concentrated. The crude
product was chromatographed through 40 g of silica gel
using ether/dichloromethane/methanol (50:35:15) as the
eluant to afford 0.129 g (46%) of product and acetal bypro-
duct (formed in the column). The acetal compound could be
converted back to the desired aldehyde by treatment with
Amberlyst 15 resin in acetone and water. This oxidation
reaction also resulted in 0.0521 g (19%) of recovered start-
solid and 0.021 g (39%) of recovered starting material. H
NMR (300 MHz, CDCl3) d 8.56 (s, 1H), 7.75 (s, 1H), 7.56
(s, 1H), 7.40±7.25 (m, 5H), 5.89 (s, 1H), 4.14±4.09 (m, 1H),
3.88±3.83 (m, 1H), 3.71±3.64 (m, 1H), 3.59 (d, 1H,
Jˆ16.9 Hz), 2.81 (d, 1H, Jˆ17.0 Hz), 2.70±2.65 (m, 1H),
2.35±1.72 (m, 7H); 13C NMR (150 MHz, CDCl3) d 179.0,
175.6, 173.8, 161.6, 143.2, 136.6, 129.0, 128.5, 127.3,
127.1, 124.8, 67.6, 57.4, 46.1, 39.3, 38.8, 28.9, 25.7, 22.3;
IR (neat/NaCl) 3248 br, 2974, 1680, 1511, 1434, 1328,
1272, 1103, 1019, 766, 695 cm21; LRFAB MS m/e (rel.
intensity) 383 (MH1, 64), 339 (MH12CH2NO, 39), 246
(25), 185 (100); HRFAB calcd for C20H23N4O4 (M11)
383.1719, found 383.1726.
1
ing material. H NMR (300 MHz, CDCl3) A mixture of
rotamers was observed. d 9.84 (s, 1H), 8.42 (s, 1H), 7.37±
7.27 (m, 6H), 4.09±4.00 (m, 1H), 3.86±3.71 (m, 2H), 3.33
(d, 0.7H, Jˆ17.6 Hz), 3.24 (d, 0.3H, Jˆ17.1 Hz), 3.04 (d,
0.3H, Jˆ17.5 Hz), 2.88 (d, 0.7H, Jˆ17.6 Hz), 2.42±1.67
(m, 8H); 13C NMR (150 MHz, CDCl3) d 201.5, 179.2,
170.9, 161.7, 140.7, 137.7, 128.8, 128.4, 127.1, 126.7,
124.5, 70.1, 57.4, 45.8, 36.6, 34.9, 29.1, 25.2, 22.2; IR
(neat/NaCl) 3261 br, 3058, 2960, 2889, 1701, 1659, 1511,
1441, 1279, 1075, 921, 772, 723 cm21; LRFAB MS m/e
(rel. intensity) 368 (MH1, 4), 307 (21), 279 (9), 154
(100); HRFAB calcd for C20H22N3O4 (M11) 368.1610,
found 368.1604.
Z-His(Bzl)-c[CN4]-Ala-OBzl (24). To a stirred suspension
of PCl5 (458 mg, 2.2 mmol) in chloroform (8 mL), quin-
oline (0.496 mL, 4.4 mmol) was added at 10±158C. The
mixture was stirred for 20 min at room temperature before
the crystalline dipeptide Z-His(Bzl)-Ala-OBzl (1.081 g,
2 mmol) was added in portions with stirring. After 30 min
at room temperature a benzene solution of hydrazoic acid in
benzene (6 mL) was added. The reaction mixture was stirred
at room temperature for 2 h before being evaporated. The
crude residue was partitioned between ethyl acetate and
water (20 mL of each). The organic layer was washed
with 1N KHSO4 (2£20 mL), 1N NaHCO3 (2£20 mL),
H2O (2£20 mL), and saturated NaCl solution (20 mL).
When the dried (Na2SO4) ethyl acetate solution was evapo-
rated and the residue puri®ed by ¯ash chromatography (pure
ethyl acetate as an eluent), the tetrazole derivative (323 mg,
28.5%) was isolated: mp 125±1268C; [a]Dˆ2388 (cˆ0.5,
MeOH); TLC Rfˆ0.55 (ethyl acetate), Rfˆ0.3 (CH2Cl2: ace-
toneˆ10:1), Rfˆ0.73 (butanol: acetic acid: waterˆ4:1:1);
FABMS m/e 568 (MH1) calcd for C31H33N7O4 567.
(6R,12S)-1-Aza-3-(pyroglutamylamino)-6-carbomethoxy-
4-phenyl-2-oxobicyclo[4.3.0]non-3-ene (23). To a solution
of 0.010 g (0.027 mmol) of the aldehyde made above in
0.8 mL of DMF was added 0.021 g (0.054 mmol, 2 equiv.)
of pyridinium dichromate (PDC). The reaction mixture was
stirred for 21 h. Methanol (2 mL) was then added to the
reaction and the mixture stirred for 15 min. The solution
was concentrated at approximately 408C under high
vacuum. The residue was then dissolved in 2 mL of
dichloromethane and 2 mL of methanol, and the solution
was concentrated again. The crude product was chromato-
graphed through silica gel using ether/dichloromethane/
methanol (2:1:1) as the eluant to afford 4 mg (38%) of
pure product as a colorless oil. 1H NMR (300 MHz,
HBr£His(Bzl)-c[CN4]-Ala-OBzl. A solution of Z-His(Bzl)-
c[CN4]-Ala-OBzl (323 mg, 0.571 mmol) in 0.5 mL acetic
acid was treated with 3.9 mL of 36% HBr in acetic
acid while stirring. After 20 min at room temperature, the

Y. Tong et al. / Tetrahedron 56 (2000) 9791±9800
9799
solution was poured into 25 mL of the mixture ether:hex-
aneˆ 1:1 (precooled to 2108C) with vigorous stirring. The
oily hydrobromide precipitated, and the upper phase was
discarded. The oil was washed with hexane (3£20 mL)
and dried in high vacuo over KOH to give the dipeptide
HBr salt (256 mg, 87%) as a very hygroscopic glass. The
material was carried on without further characterization.
in 25 min); FABMS m/e 364 (MH1) calcd for C14H21N9O3
363.
Acknowledgements
We thank the National Institutes of Health (RO1
GM5324001 and R01 DK43036) for their generous ®nan-
cial support. In addition, we thank the Washington Univer-
sity 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.
Z-Pyr-His(Bzl)-c[CN4]-Ala-OBzl. A solution of 145 mg
(0.55 mmol) of Z-Pyr-OH in 8 mL of CH2Cl2 was cooled
to 2208C and treated with 60 mL (0.55 mmol) of N-Methyl-
morpholine, followed by 74 mL (0.55 mmol) of isobutyl-
chloroformate. After stirring for 10 min, solid HBr£
His(Bzl)-c[CN4]-Ala-OBzl (256 mg, 49.8 mmol) was intro-
duced, followed by the addition of N-methyl-morpholine
(55 mL, 0.5 mmol) at a rate such that the temperature stayed
below 2108C. After being stirred for 1 h at 2108C, the
mixture was warmed slowly to room temperature and left
with stirring overnight. The solvent was removed in vacuo.
The crude residue was taken up in ethyl acetate (30 mL) and
washed with 1N KHSO4 (3£15 mL), 1N NaHCO3 (3£
15 mL), water (2£15 mL), and saturated NaCl solution
(20 mL). The ethyl acetate solution, dried over anhydrous
Na2SO4, was evaporated to dryness. The crude crystalline
material was washed with ether and ®lter off serving 230 mg
(67.8%) of pure product: mp 96±978C; HPLC purity 97%,
tRˆ13.5 min. The HPLC puri®cation was run in 25 min
linear gradient mode (30±70% solvent B in A22); Rfˆ0.54
(ethyl acetate:MeOHˆ20:1), Rfˆ0.45 (CH2Cl2:acetoneˆ
15:1), Rfˆ0.56 (CHCl3:MeOHˆ10:1); FABMS m/e 679
(MH1) calcd for C36H38N8O6 678.
References
1. For reviews on the biology of TRH see: Thyrotropin-Releasing
Hormone: Biomedical Signi®cance; Metcalf Olson, G. L., Cheung,
H. C., Voss, M. E., Hill, D. E., Kahn, M., Madison, V. S., Cook,
C. M., Sepinwall, J., Eds.; Proceedings of the Biotech USA 89
Meeting, San Francisco, October 1989.
2. For recent examples see: (a) Olson, G. L.; Cheung, H.-C.;
Chaing, E.; Madison, V. S.; Sepinwall, J.; Vincent, G. P.; Winokur,
A.; Gary, K. A. J. Med. Chem. 1995, 38, 2866. (b) Coa, J.;
O'Donnell, D.; Vu, H.; Payza, K.; Pou, C.; Godbout, C.;
Jakob, A.; Pelletier, M.; Lembo, P.; Ahmad, S.; Walker, P. J. Bio.
Chem. 1998, 273, 32281, and references therein.
3. For approaches to understanding TRH/TRH-R1 binding see:
(a) Hinkle, P. M.; Woroch, E. L.; Tashjian, G. H. J. Biol. Chem.
1974, 249, 3085. (b) Goren, H. J.; Bauce, L. G.; Vale, W. Mol.
Pharmacol. 1977, 13, 606. (c) Kamiya, K.; Takamoto, M.; Wada,
Y.; Fujino, M.; Nishikawa, M. J. Chem. Soc., Chem. Commun.
1980, 438. (d) Sharif, N. A.; To, Z.; Whiting, R. L. Biochem.
Biophys. Res. Commun. 1989, 161, 1306. (E) Vonhof, S.;
Feuerstein, G. Z.; Cohen, L. A.; Labroo, V. M. Eur. J. Pharmacol.
1990, 180, 1. (f) Perlman, J. H.; Laakkonen, L. J.; Guarnieri, F.;
Osman, R.; Gershengorn, M. C. Biochemistry 1996, 35, 7643.
(g) Laakkonen, L. J.; Guarnieri, F.; Perlman, J. H.; Gershengorn,
M. C.; Osman, R. Biochemistry 1996, 35, 7651.
Pyr-His(Bzl)-c[CN4]-Ala-OH (26). Z-Pyr-His(Bzl)-c[CN4]-
Ala-OBzl (150 mg, 0.22 mmol) in methanol (15 mL) was
hydrogenated in the presence of 10% palladized charcoal
(40 mg) at a pressure of 4±5 kg/cm2 on a Parr apparatus for
3 h (monitored by TLC). After evaporation of the ®ltered
solution, the residual glassy solid was washed few times
with hexane. Yield 86 mg (85%); Rfˆ0.48 (butanol:acetic
acid:waterˆ4:1:1); HPLC purity 98%, tRˆ9.86 min (gradi-
ent 5±35% B in A22 in 25 min); FABMS m/e 455 (MH1)
calcd for C21H26N8O4 454.
4. For an overview of the approach to active site modeling utilized
see: (a) Beusen, D. D.; Shands, E. F. B.; Karasek, S. F.; Marshall,
G. R.; Dammkoehler, R. A. J. Mol. Struct. (Theochem) 1996, 370,
157. (b) Buesen, D. D.; Shands, E. F. B. Drug Discovery Today
1996, 1 (10), 429. (c) Dammkoehler, R. A.; Karasek, S. F.; Shands,
E. F. B.; Marshall, G. R. J. Comput.-Aided Mol. Des. 1995, 9, 491.
(d) Dammkoehler, R. A.; Karasek, S. F.; Shands, E. F. B.;
Marshall, G. R. J. Comput.-Aided Mol. Des. 1989, 3, 3.
Pyr-His(Bzl)-c[CN4]-Ala-NH2. To the solution of Pyr-
His(Bzl)-c[CN4]-Ala-OH (57 mg, 0.126 mmol) in 1 mL of
CH2Cl2 and 1.5 mL DMF, ammonium salt of HOBt was
added followed by DCC (28.5 mg, 0.138 mmol). The reac-
tion mixture was left with stirring overnight. Solvents were
removed in vacuo and the crude material has been puri®ed
on preparative HPLC (gradient 5±35%B) giving 36 mg
(63%) of the desired product. HPLC purity 97%,
tRˆ8.06 min (gradient 5±35% B in A22 in 25 min);
FABMS m/e 454 calcd for C21H27N9O3 453.
5. For applications of the active analog approach to TRH see:
(a) Marshall, G. R.; Gorin, F. A.; Moore, M. L. In Annual Reports
in Medicinal Chemistry; Clarke, F. H., Ed.; Academic: New York,
1978; Vol. 13, p 227 and references therein. (b) Font, J. Ph.D.
thesis, Washington University in St. Louis, 1986.
Pyr-His-c[CN4]-Ala-NH2 (4). Pyr-His(Bzl)-c[CN4]-Ala-
NH2 (21 mg, 0.046 mmol) in MeOH (3 mL) and acetic
acid (0.5 mL) was hydrogenated in the presence of 10%
palladized charcoal (10 mg) at a pressure of 4±5 kg/cm2
on a Parr apparatus for 48 h (monitored by TLC). After
evaporation of the ®ltered solution, the residual crude
product was puri®ed on preparative HPLC (gradient
5±35% B in A22) giving 15 mg of the desired analog.
HPLC purity 98%, tRˆ6.12 min (gradient 5±35% B in A
6. Rutledge, L. D.; Perlman, J. H.; Gershengorn, M. C.; Marshall,
G. R.; Moeller, K. D. J. Med. Chem. 1996, 39, 1571.
7. (a) Li, W.; Moeller, K. D. J. Am. Chem. Soc. 1996, 118, 10106.
(b) Laakkonen, L.; Li, W.; Perlman, J. H.; Guarnieri, F.; Osman, R.;
Moeller, K. D.; Gershengorn, M. C. Mol. Pharmacol. 1996, 49,
1092.
8. Chu, W.; Perlman, J. H.; Gershengorn, M. C.; Moeller, K. D.
Bioorg. Med. Chem. Lett. 1998, 8, 3093.
9. Richardson, J.; Richardson, D. C. Prediction of Protein

9800
Y. Tong et al. / Tetrahedron 56 (2000) 9791±9800
Structure and Principles of Protein Conformation; Fasman, G.,
Ed.; Plenum: New York, 1989.
Soc. 1995, 117, 5855. For reviews of RCM reactions see:
(b) Grubbs, R. H.; Miller, S. J.; Fu, G. C. Acc. Chem. Res. 1995,
28, 446 and (c) Blechert, S.; Schuster, M. Angew. Chem., Int. Ed.
1997, 36, 2036. For a route to fused nitrogen heterocycles see:
10. Liakopoulou-Kyriakides, M.; Galardy, R. E. Biochemistry
1979, 18, 1952.
11. (a) Marshall, G. R.; Humblet, C.,; Opdenbosch, N. V.;
Zabrocki, J. In Proceedings of the 7th American Peptide
Symposium, Rich, D. H., Gross, E., Eds.; Pierce Chemical
Company: Rockford IL; 1989; p 669. (b) Zabrocki, J.; Dunbar,
Jr., J. B.; Marshall, K. W.; Toth, M. V.; Marshall, G. R. J. Org.
Chem. 1992, 57, 502.
È
(d) Martin, S. F.; Liao, Y.; Chen, H. J.; Patzel, M.; Ramser,
M. N. Tetrahedron Lett. 1994, 35, 6005.
17. (a) Curran, D. P.; Gu, X.; Zhang, W.; Dowd, P. Tetrahedron
1997, 53, 9023. (b) Molander, G. A.; Harris, C. R. J. Am. Chem.
Soc. 1995, 117, 3705.
18. Lipshutz, B. H.; Sengupta, S. Org. React. (NY) 1992, 41, 135.
19. Li, W.; Hanau, C. E.; d'Avignon, A.; Moeller, K. D. J. Org.
Chem. 1995, 60, 8155.
12. For reviews of TRH analog activity data see: (a) Vale, W.;
Rivier, C. Handbook of Psychopharmacology; Iverson, L. I. et al.,
Eds.; Plenum: New York, 1975; Vol. 5, 195. (b) Goren, H. J.;
Bauce, L. G.; Vale, W. Mol. Pharmacol. 1977, 13, 606.
13. Seebach, D.; Boes, M.; Naef, R.; Schweizer, W. B. J. Am.
Chem. Soc. 1983, 105, 5390.
20. For published procedures see: (a) Perlman, J. H.; Thaw, C. N.;
Laakkonen, L.; Bowers, C. Y.; Osman, R.; Gershengorn, M. C.
J. Biol. Chem. 1994, 269, 1610. (b) Straub, R. E.; Frech, G. C.;
Joho, R. H.; Gershengorn, M. C. Proc. Natl. Acad. Sci. USA 1990,
87, 9514.
14. Gramberg, D.; Robinson, J. A. Tetrahedron Lett. 1994, 35,
861.
21. Zouikri, M.; Vicherat, A.; Aubry, A.; Marraud, M.; Boussard,
G. J. Peptide Res. 1998, 52, 19.
15. Dumas, J.-P.; Germanas, J. P. Tetrahedron Lett. 1994, 35,
1493.
22. HPLC has been performed using solvents system: A: 0.5%
TFA in water; B: 0.038% TFA, 90% acetonitrile, 10% water.
16. For the initial paper describing the application of RCM to a
peptidomimetic see: (a) Miller, S. J.; Grubbs, R. H. J. Am. Chem.