Methods to initiate synthetic re-structuring of peptides

Article abstract of DOI:10.1016/j.tet.2003.04.008

Sculpture during assembly is an appropriate description of procedures used to transform unprotected peptide i into complex macrocycle ii. These methods appear a general means to begin manipulating the form and characteristics of common polyamides. Herein we demonstrate initial phases of a type of synthesis that has the potential to produce large numbers of novel structural classes beginning with machine-made heteropolymers.

Full text of DOI:10.1016/j.tet.2003.04.008

Tetrahedron 59 (2003) 8947–8954
Methods to initiate synthetic re-structuring of peptides
†
Qi Wei, Susan Harran and Patrick G. Harran
*
Department of Biochemistry, University of Texas Southwestern Medical Center at Dallas, 5323 Harry Hines Blvd,
Dallas, TX 75390-9038, USA
Received 21 March 2003; accepted 2 April 2003
Abstract—Sculpture during assembly is an appropriate description of procedures used to transform unprotected peptide i into complex
macrocycle ii. These methods appear a general means to begin manipulating the form and characteristics of common polyamides. Herein we
demonstrate initial phases of a type of synthesis that has the potential to produce large numbers of novel structural classes beginning with
machine-made heteropolymers.
q 2003 Elsevier Ltd. All rights reserved.
1. Introduction
able to advance our processing sequence. In particular, the
allylic carbonate present in 4 can be decomposed by
catalytic amounts of soluble palladium salts.5 If there is a
nucleophile of appropriate pKa in proximity, a putative
intermediate p-allyl palladium complex is trapped intern-
ally and a large ring is formed. Among the set of
proteinogenic amino acid side-chain functional groups, we
find the phenol of tyrosine (pKa,10.1) uniquely competent
in such cyclizations. However, generality and efficiency in
the transformation were achieved only after extensive
experimentation. Numerous combinations of metal pre-
catalyst and stabilizing ligand were either poisoned by
substrate or led, slowly and inefficiently, to a mixture of
desired cyclic cinnamyl ether 6 and oligomeric material6—
the latter dominating in most instances. The situation
changes considerably with the use of a catalyst formed by
pre-treating [(h3-allyl)PdCl]2 with van Leeuwen’s Xant-
phos7 (2:1 5/Pd atom). In this instance, the conversion of 4
to 6 is complete within minutes at rt (DMF, 5 mM in 4) and
competing oligomerization is minimized. We are aware
neither of large rings being formed in this manner
previously nor of the [(h3-allyl)PdCl]2/5 combination
finding use in catalyzed allylic etherification—despite its
marked effectiveness here.
We are attempting to develop synthetic procedures able to
incrementally re-cast the form and properties of common
peptides. In particular, reaction sequences which begin by
incorporating a hydrophobic ‘reaction nucleus’ (N in
Scheme 1A) within peptide chains. Doing so, the aim is to
offset polarity and restrict conformational flexibility while
generating intermediates poised for additional manipu-
lation. The long-term goal is to systematically move beyond
peptidomimetics with a form of synthesis that molds new
structural types into existence as much as it builds them
from parts.1 Herein we describe our initial results.
2. Results and discussion
Compound 2 (Scheme 1B) represents a minimal prototype
of reaction nucleus N. Conventional peptide synthesis
occurs such that a free N-terminus is a feature available
on all sequences prepared and aldehyde 2 is designed to
exploit this commonality. For example, mixing 2 and
dipeptide 1 ligates the two via Schiff-base formation. This
joining is then made irreversible by treatment in situ with
a-(p-toluenesulfonyl)-4-fluorobenzylisonitrile (3).2 The
product of the resulting cycloaddition/sulfinate elimination3
(4 isolated simply by precipitation from a concentrated
EtOAc solution4) integrates the peptide amino terminus into
a heterocyclic ring that now displays tethered functionality
As shown in Table 1, the protocol of multi-component
condensation/metal-catalyzed cycloetherification appears a
general means to begin manipulating the structure of linear,
unprotected peptides.8 In fact, we have yet to identify a
sequence displaying a free N-terminus and a tyrosine
residue that will not participate.9 The catalyzed cycloetheri-
fication step tolerates free carbinols and carboxamides,
thioethers, and selected heteroaromatics. Sequences con-
taining multiple tyrosine residues ring-close with compar-
able efficiency although, in the case examined (entry 5),
regioselectivity is modest.10 Minor technical demands of the
Keywords: peptides; heterocycles; palladium catalysis; rearrangements;
solid-phase synthesis.
*
Corresponding author. Tel.: þ1-214-648-3612; fax: þ1-214-648-6455;
e-mail: pharra@biochem.swmed.edu
Present Address: Cumbre Inc., 1502 Viceroy Drive, Dallas, TX 75235-
2304, USA.
†
0040–4020/$ - see front matter q 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2003.04.008

8948
Q. Wei et al. / Tetrahedron 59 (2003) 8947–8954
Scheme 1. Reaction conditions: (a) 1 equiv. 2, K2CO3, DMF, rt, 5 h; then add 3 (1.05 equiv.), rt, 17 h (80%). (b) 10 mol% [(h3-allyl)PdCl]2, 40 mol% 5,
degassed DMF, rt, 1 h (63%).
cyclization include a need for oxygen-free media and
workup with aqueous NaCN11 to avoid partial destruction of
allylic ether products during isolation.
esis, we have adapted initial stages of our processing
sequence to a solid-phase format (Scheme 3). This will
permit use of automated parallel synthesis techniques
during library construction. To date, we have found an
amine-presenting form of Foley’s silane-functionalized
polystyrene resin (23)18 most well-suited for our needs.
Fmoc-based peptide synthesis19 is facile in this format and
subsequent isonitrile-based imidazole synthesis and cata-
lyzed cycloetherification have both been performed suc-
cessfully. While optimization continues,20 the solid-phase
synthesis of 26 suggests we now have superior means
to quickly widen our starting material base (to include
N-alkylated, D-configured, and unnatural residue-containing
peptides), to incorporate and evaluate second generation
reaction nuclei N with enhanced functionality, and to screen
for new dehydrogenative (ideally, biaryl-forming) oxidation
methods as part of more complete processing sequences-
those able to close the gap between abundant oligopeptides
and novel natural product-like substances.
In line with our original design criteria, the cinnamyl ether
unit formed via cycloetherification is a staging ground for
further chemoselective modifications. Scheme 2 outlines
two simple demonstrations of this potential. Following
desiccation,12 cyclization product 13 undergoes diastereo-
selective13 dihydroxylation using a modified Sharpless
protocol.14 This transforms its allylic ether substructure
into a monoaryl glycerol (namely 18, Scheme 2A)—a
moiety likely amenable to derivatization with, for example,
fatty acids and/or sugars. A second perturbation available in
the allylic ether series is Claisen rearrangement. For
example, thermolysis of 8 provides phenolic a-olefin 19—
a change accompanied by four-atom ring-contraction of the
macrocycle (Scheme 2B). As currently executed,15 the
sigmatropic rearrangement is not stereoselective. However,
this is viewed as an asset given the diversity-oriented goals
of the program. In addition, the functionality in 19 produced
as a result of rearrangement can be modified in direct and
useful ways. Mitsunobu etherification with allyl carbinol
followed by ruthenium catalyzed ring-closing olefin
metathesis16 annulates a hydrophobic, 7-membered ring
onto the existing 21-membered macrocycle-affording yet
another novel polycyclic structure (namely 21).
3. Experimental
3.1. Data for compounds 2
3.1.1. tert-Butyl-{1-[3-((4-oxo)butoxy)phenyl]-2-propenyl}
carbonate (2). Anhydrous Cs2CO3 (103.2 g, 317 mmol)
and 3-hydroxybenzaldehyde (32.2 g, 264 mmol) are sus-
pended in 600 mL DMF. 4-(t-Butyldimethylsiloxy)-1-
chlorobutane (21.6 g, 106 mmol) is added and the mixture
brought to 1158C with vigorous stirring. Beginning at 4 h,
two additional portions (21.6 g each) of alkyl chloride are
added at 1.5 h intervals. The mixture is cooled to rt, diluted
with 1.5 L EtOAc, washed with H2O (2£500 mL), dried
over MgSO4, and concentrated. Filtration through a column
of silica gel (5% EtOAc/hexanes) provides a light yellow oil
(48.6 g, 60%). A portion of this material (38.9 g, 126 mmol)
is dissolved in 500 mL dry THF and cooled to 2708C under
N2. A solution of vinyl magnesium bromide (Aldrich, 1.0 M
The elaborations above are a small sampling of pertur-
bations intended to impart increasing ‘alkaloid-like’17
character to initially formed macrocycles. As the method-
ology is refined and expanded, and our collection of
polycyclic compounds grows, the opportunity to discover
members having new biochemical functions likewise
increases. It is conceivable, given the complexity and
functional group content of the compounds in question, that
specificity in, for example, protein binding, will be observed
at a frequency higher than that typical of conventional
pharmaceutical collections. To facilitate testing this hypoth-

Table 1. Macrocyclic cinnamyl ethers prepared by sequential three-component condensation/palladium-catalyzed cycloetherification
Entry
1
Peptidea
GGY
Condensationb product (%)
(64)
Cycloetherificationc product (%)
Entry
6
Peptide
GSY
Condensation product (%)
(65)
Cycloetherification product (%)
2
3
4
GWY
GPY
ACY
(80)
(65)
(55)
7
8
9
AYS
ASSY
GQYS
(42)
(49)
(60)
(continued on next page)

Table 1 (continued)
Peptidea
Entry
Condensationb product (%)
(64)
Cycloetherificationc product (%)
Entry
10
Peptide
Condensation product (%)
(48)
Cycloetherification product (%)
5
GYY
AH(3-Me)Y
aPrepared in the form H2N-AA1-AAn-CONHBu-n; where amino acid residues (AA) are listed in one-letter codes. bSee Section 3 and Ref. 4. Yields quoted refer to .90% pure (1H NMR) precipitated powders. cSee
d
Section 3. Yields quoted are for isolated, chromatographically homogeneous material. Isolates contaminated with ,15% unidentified by-products.

Q. Wei et al. / Tetrahedron 59 (2003) 8947–8954
8951
Scheme 2. Reaction conditions: (a) [MeO(CH2)2]2NSF3, CH2Cl2, 2208C (62%). (b) K2Fe(CN)6, 1.5 mol% K2OsO2(OH)4, 15 mol% (DHQD)PHAL, K2CO3,
MeSO2NH2, t-BuOH/H2O, rt (61%). (c) 1808C (microwave), silica gel, CH3CN, 40 min (74%). (d) allyl alcohol, DIAD, PPh3, THF (81%). (e) 20 mol% 22,
CH2Cl2, 408C (65%).
Scheme 3. Reaction conditions: (a) Ref. 19. (b) 2 (4 equiv.), 1:1 THF/(MeO)3CH, rt, 4 h; THF wash (3£); 3 (5 equiv.), piperazine (5 equiv.), DMF, rt, 24 h. (c)
[(h3-allyl)PdCl]2/5 complex (,10 mol%), degassed DMF, rt, 2£1 h. (d) HF/pyridine, THF, rt, 3 h.

8952
Q. Wei et al. / Tetrahedron 59 (2003) 8947–8954
in THF, 150 mL) is added over 30 min via dropping funnel.
When complete, the cooling bath is removed and the
solution brought to rt over 45 min. Sat. aqueous NH4Cl
(100 mL) is added and the mixture partitioned between
EtOAc and H2O. The layers are separated, the aqueous layer
is extracted with EtOAc (1£), and the combined organics
are dried over Na2SO4 and concentrated. The residue is
dissolved in 350 mL CH2Cl2 and cooled in an ice bath under
N2. Di-t-butyldicarbonate (27.5 g, 126 mmol) is added
followed by p-dimethylaminopyridine (15.4 g, 126 mmol)
in 5 equal portions over 20 min. The solution is stirred at rt
for 3.5 h and quenched with sat. aqueous NH4Cl. The
organic layer is separated and washed with H2O and brine,
dried over Na2SO4, and concentrated. Flash chromato-
graphy on silica gel (5% EtOAc/hexanes) provides a
homogenous carbonate product (46.0 g, 84% yield-two
steps). This material is transferred to a polypropylene bottle
and dissolved in 600 mL THF. Anhydrous pyridine (75 mL)
and HF/pyridine complex (Aldrich, ,70% HF, 30 mL) are
added and the solution stirred at rt for 4 h. The reaction is
concentrated to ,1/3 its original volume, diluted with
500 mL EtOAc, and washed with sat. CuSO4 (3£150 mL),
H2O, sat. aqueous NaHCO3, and brine. Rotary evaporation
provides an oil that is chromatographed on silica gel (30%
EtOAc/hexanes) to afford an alcohol product (26.1 g, 77%).
A portion of this substance (21.43 g, 66.4 mmol) is
dissolved in CH2Cl2 (250 mL), cooled in an ice bath, and
treated successively with pyridine (8.1 mL, 0.10 mol) and
solid Dess–Martin periodinane (42.3 g, 0.10 mol). The
mixture is stirred for 1 h at 48C and 2 h at rt. A solution of
sat. aqueous NaHCO3 containing Na2S2O3 (80 mL) is added
and the two-phase system is mixed vigorously until the
organic layer clears (,20 min). The layers are separated and
the organics are washed with sat. aqueous NaHCO3 and
brine, dried over Na2SO4 and concentrated. Flash chromato-
graphy on silica gel (10!20% EtOAc/hexanes) gives
aldehyde 2 (15.73 g, 74%) as a cream colored, pasty solid.
2: Rf¼0.40 (20% EtOAc/hexanes). IR (film): 2981, 2936,
2827, 2726, 1742, 1602, 1587, 1370, 1254, 1161, 1087, 941,
854, 792 cm21. 1H NMR (300 MHz, CDCl3): d 9.86 (s, 1H),
7.27 (dd, J¼5.4, 4.8 Hz, 1H), 6.96 (d, J¼5.4 Hz, 1H), 6.89
(s, 1H), 6.83 (d, J¼4.8 Hz, 1H), 5.90–6.18 (m, 2H), 5.24–
5.35 (m, 2H), 4.01 (t, J¼4.5 Hz, 2H), 2.68 (t, J¼4.9 Hz,
2H), 2.10–2.16 (m, 2H), 1.49 (s, 9H). 13C NMR (75 MHz,
CDCl3): d 201.6, 158.8, 152.6, 140.3, 136.1, 129.6, 119.3,
117.1, 114.1, 112.9, 82.2, 78.9, 66.6, 40.5, 27.7, 21.9.
LRMS (positive electrospray) calcd for C18H24O5: [MþH]þ
321.17. Found: 321.14.
combined and the precipitation procedure repeated once
more to afford additional 4 (152 mg, combined yield¼80%)
as a light yellow powder. ES-MS: calcd for C41H49FN4O7
[MþþH]: 729.37. Found: 729.37.
Cinnamyl ether 6. [(h3-allyl)PdCl]2 (2.0 mg, 5.5 mmol) and
bis-phosphine 5 (13.0 mg, 22 mmol) are dissolved in 1 mL
degassed (N2 purge) THF. The solution is stirred for 20 min,
diluted with 8 mL degassed DMF and treated with 4 (40 mg,
55 mmol in 2 mL degassed DMF). Stirring is continued for
1 h and then 100 mL 1.0 M aq. NaCN is added. The solution
is diluted with EtOAc, washed with water and brine, dried
over Na2SO4, and concentrated. Purification by flash
chromatography (3!5% i-PrOH/CHCl3) affords cyclic
ether 6 (21 mg, 63%) as a white film. 6: Rf¼0.38 (10%
i-PrOH/CHCl3). [a]2D0¼þ65.78 (c¼1.17, MeOH). IR (film):
1
3284, 2931, 1639, 1558, 1509, 1221, 1157, 839 cm21. H
NMR (400 MHz, CD3OD): d 7.56 (d, J¼8.8 Hz, 1H), 7.55
(d, J¼8.8 Hz, 1H), 7.52 (s, 1H), 7.22 (d, J¼8.4 Hz, 2H),
7.17 (t, J¼7.8 Hz, 1H), 7.11 (app t, J¼8.8 Hz, 2H), 6.92 (m,
3H), 6.70 (dd, J¼8.0, 2.4 Hz, 1H), 6.60 (s, 1H), 6.48 (d,
J¼16.2 Hz, 1H), 6.16 (dt, J¼16.2, 5.6 Hz, 1H), 4.88 (d,
J¼5.6 Hz, 2H), 4.63 (d, J¼16.4 Hz, 1H), 4.53 (dd, J¼12.0,
3.6 Hz, 1H), 4.40 (d, J¼16.4 Hz, 1H), 3.90 (sym 7 line m,
1H), 3.78 (sym 7 line m, 1H), 3.21 (sym 8 line m, 2H), 3.13
(dd, J¼14.4, 3.2 Hz, 1H), 2.91 (app t, J¼7.2 Hz, 2H), 2.79
(dd, J¼14.4 Hz, 12.0, 1H), 1.93 (m, 1H), 1.77 (m, 1H), 1.48
(sym 7 line m, 2H), 1.34 (sym 6 line m, 2H), 0.92 (t,
J¼7.4 Hz, 3H). 13C NMR (75 MHz, CD3OD): d 174.0,
169.5, 160.0, 157.9, 139.9, 138.4, 137.6, 134.5, 131.1,
131.0, 130.7, 130.2, 130.1, 129.5, 126.9, 120.1, 118.3,
116.5, 116.3, 114.9, 113.6, 68.6, 66.8, 57.0, 47.6, 40.4, 38.3,
32.6, 29.7, 25.4, 21.2, 14.2. ES-MS: calcd for
C36H39FN4O4: [MþH]þ 611.31. Found: 611.39.
3.1.3. Oxazoline-containing glycol 18. Alcohol 13 (45 mg,
0.64 mmol) is dissolved in 1 mL dry CH2Cl2, cooled to
2208C, and treated with bis(2-methoxyethyl)aminosulfur
trifluoride (Deoxo-Fluor, 0.3 mL aliquot of CH2Cl2 stock
solution, 0.96 mmol). The reaction is stirred at 2208C for
3 h, quenched with saturated NaHCO3, and warmed to rt.
The aqueous layer is washed with CHCl3 (3£) and the
combined organics dried over Na2SO4. Concentration and
flash chromatography on silica gel (CHCl3/EtOAc/MeOH,
8/1.7/0.3) affords an oxazoline product (25 mg, 57%) as a
white solid. Rf¼0.59 (10% MeOH/CH2Cl2). [a]2D0¼þ40.28
(c¼0.33, MeOH). IR (film): 3280, 2960, 2927, 2873, 1657,
1513, 1293, 1223, 973, 840 cm21 1H NMR (300 MHz,
.
CD3OD): d 7.70 (s, 1H), 7.48–7.58 (m, 2H), 7.02–7.22 (m,
5H), 6.82–6.96 (m, 4H), 6.67 (dd, J¼9.0, 2.4 Hz, 1H), 6.56
(d, J¼16.2 Hz, 1H), 6.33 (dt, J¼15.9, 5.4 Hz, 1H), 4.80–
4.96 (m, 2H), 4.52–4.64 (m, 2H), 4.27 (dd, J¼10.5, 8.7 Hz,
1H), 4.04 (dd, J¼8.7, 8.1 Hz, 1H), 3.80–3.92 (m, 2H), 3.22
(t, J¼6.9 Hz, 2H), 3.13 (dd, J¼14.1, 3.3 Hz, 1H), 2.95 (t,
J¼7.2 Hz, 2H), 2.81 (dd, J¼14.1, 11.4 Hz, 1H), 1.80–2.00
(m, 2H), 1.44–1.60 (m, 2H), 1.30–1.44 (m, 2H), 0.94 (t,
J¼6.9 Hz, 3H). 13C NMR (75 MHz, CD3OD): d 173.8,
173.4, 167.9, 160.3, 158.4, 139.5, 139.3, 138.4, 133.9,
131.3, 130.8, 130.5, 130.4, 130.3, 128.7, 126.4, 120.8,
116.6, 116.5, 116.2, 114.9, 113.4, 72.9, 69.6, 68.7, 67.0,
56.3, 42.7, 40.4, 38.2, 32.7, 29.5, 21.2, 20.6, 14.2. LRMS
(positive electrospray) calcd for C39H42FN5O5: [MþH]þ
680.33. Found: 680.53.
3.1.2. General condensation and cyclization procedures
(1!4!6). A mixture of dipeptide 1 (413 mg, 1.4 mmol),
aldehyde 2 (451 mg, 1.4 mmol), powdered anhydrous
˚
K2CO3 (486 mg, 3.5 mmol), and 4 A molecular sieves
(2.0 g) is suspended in 4 mL DMF. The mixture is stirred
under N2 for 5 h (rt) and then treated with solid isonitrile 3
(428 mg, 1.478 mmol). Stirring is continued for 17 h. The
mixture is filtered through a plug of celite and concentrated.
The residue remaining is dissolved in EtOAc (200 mL) and
washed with H2O and brine. Concentration of the organic
layer to ,20 mL induces precipitation of 4 which is pelleted
by centrifugation (6 min at 3500 rpm). The supernatant is
decanted and the solid (670 mg) washed with 30% EtOAc/
hexanes (3£5 mL). The washings and supernatant are

Q. Wei et al. / Tetrahedron 59 (2003) 8947–8954
8953
A portion of the above oxazoline (6.0 mg, 8.8 mmol) is
dissolved in 1:1 t-BuOH/H2O (0.2 mL). Solid CH3SO2NH2
(2.5 mg, 26 mmol) and a pre-ground mixture of K3Fe(CN)6
(8.7 mg, 26 mmol), K2OsO2(OH)4 (32.4 mg, 0.09 mmol),
(DHQD)2PHAL (0.69 mg, 0.88 mmol), and K2CO3 (3.7 mg,
26 mmol) are added. The resultant mixture is stirred at rt for
20 h. Volatiles are removed in vacuo and the residue
remaining is diluted with 5 mL H2O, filtered, and concen-
trated. Purification by PTLC affords glycol 18 (4 mg, 63%)
as a white film. 18: Rf¼0.38 (10% MeOH/CH2Cl2).
[a]2D0¼þ46.58 (c¼0.33, MeOH). IR (film): 3333, 2953,
10 line m, 1H), 5.29 (dd, J¼17.2, 1.6 Hz, 1H), 5.20 (d,
J¼10.0 Hz, 2H), 5.02 (d, J¼6.4 Hz, 1H), 4.91 (d,
J¼17.2 Hz, 1H), 4.61–4.70 (m, 2H), 4.49 (d, J¼16.8 Hz,
1H), 4.41 (d, J¼16.8 Hz, 1H), 4.33–4.40 (m, 2H), 3.70–
3.78 (m, 1H), 3.30–3.38 (m, 1H), 3.13–3.30 (m, 3H), 3.01–
3.13 (m, 2H), 2.89 (dd, J¼14.4, 8.0 Hz, 1H), 2.54–2.68 (m,
2H), 1.60–1.74 (br m, 2H), 1.45–1.52 (m, 2H), 1.29–1.40
(m, 2H), 0.93 (t, J¼6.8 Hz, 3H). 13C NMR (75 MHz,
CDCl3): d 170.79, 170.75, 167.6, 158.3, 155.0, 144.8, 140.1,
137.2, 136.3, 133.4, 132.2, 130.7, 129.4, 128.7, 128.6,
128.2, 127.4, 127.1, 122.9, 122.8, 122.5, 112.0, 118.6,
117.3, 116.7, 115.8, 115.6, 115.5, 112.4, 111.6, 110.9,
109.5, 69.1, 66.2, 54.2, 54.0, 48.8, 47.8, 39.7, 37.3, 31.6,
28.8, 27.2, 20.3, 19.9, 14.0. LRMS (positive electrospray)
calcd for C50H53FN6O5: [MþH]þ 837.42. Found: 837.15.
20: more polar diastereomer: Rf¼0.43 (CHCl3/EtOAc/
MeOH¼6/3/1). [a]2D4¼þ9.98 (c¼1.2, CHCl3). IR (film):
3301, 3065, 2926, 2855, 1652, 1506, 1456, 1247, 1157, 840,
743 cm21. 1H NMR (400 MHz, CDCl3): d 8.49 (s, 1H), 7.46
(dd, J¼8.4, 6.0 Hz, 2H), 7.40 (d, J¼8.0 Hz, 1H), 7.17–7.25
(m, 3H), 7.07–7.17 (m, 2H), 6.94–7.07 (m, 4H), 6.74 (s,
1H), 6.75 (d, J¼8.4 Hz, 1H), 6.66 (d, J¼8.4 Hz, 1H), 6.54–
6.42 (m, 4H), 6.30 (sym 7 line m, 1H), 6.09 (s, 1H), 5.90
(sym 8 line m, 1H), 5.28 (d, J¼17.2 Hz, 1H), 5.19 (t,
J¼9.6 Hz, 2H), 5.13 (d, J¼7.2 Hz, 1H), 5.00 (d, J¼17.2 Hz,
1H), 4.66 (dd, J¼12.4, 6.8 Hz, 1H), 4.30–4.48 (m, 5H),
3.53–3.62 (m, 1H), 3.43–3.53 (m, 1H), 3.24–3.10 (m, 3H),
2.97–3.09 (m, 2H), 2.91 (dd, J¼14.4, 7.2 Hz, 1H), 2.57–
2.72 (m, 2H), 1.46–1.64 (m, 2H), 1.35–1.46 (m, 2H), 1.24–
1.34 (m, 2H), 0.89 (t, J¼7.2 Hz, 3H). 13C NMR (75 MHz,
CDCl3): d 170.9, 170.7, 167.2, 160.4, 158.3, 155.2, 145.6,
140.3, 137.9, 137.4, 136.5, 133.5, 132.5, 131.2, 130.1,
129.2, 128.9, 128.8, 128.0, 127.3 127.1, 123.3, 122.7, 122.6,
120.0, 118.9, 117.0, 116.7, 115.81, 115.75, 115.5, 112.3,
111.6, 110.4, 109.9, 69.1, 66.3, 53.9, 48.0, 47.8, 39.7, 37.3,
31.7, 28.2, 27.9, 20.3, 20.2, 14.0. LRMS (positive electro-
spray) calcd for C50H53FN6O5: [MþH]þ 837.42. Found:
837.25.
2933, 1656, 1507, 1247, 1037, 977 cm21 1H NMR
.
(300 MHz, CD3OD): d 7.75 (s, 1H), 7.52–7.62 (m, 2H),
7.27 (t, J¼7.8 Hz, 1H), 7.12–7.02 (m, 3H), 6.94 (d,
J¼8.7 Hz, 2H), 6.82–6.72 (m, 2H), 6.62 (d, J¼8.7 Hz,
2H), 4.96 (d, J¼16.8 Hz, 1H), 4.84 (d, J¼16.8 Hz, 1H), 4.71
(d, J¼8.1 Hz, 2H), 4.47–4.60 (m, 2H), 4.27 (t, J¼8.4 Hz,
1H), 3.78–3.92 (m, 3H), 3.73 (dd, J¼10.2, 2.4 Hz, 1H),
3.47 (dd, J¼10.2, 4.8 Hz, 1H). 3.12–3.24 (m, 2H), 3.05 (dd,
J¼14.4, 3.9 Hz, 1H), 3.00–2.80 (m, 3H), 1.82–2.04 (m,
2H), 1.42–1.54 (m, 2H), 1.26–1.40 (m, 2H), 0.92 (t,
J¼7.2 Hz, 3H). 13C NMR (75 MHz, CD3OD): d 173.12,
173.10, 167.4, 159.7, 159.3, 144.5, 139.1, 138.0, 132.5,
131.5, 130.8, 130.3, 130.2, 129.0, 120.3, 116.6, 116.3,
116.2, 115.9, 114.1, 76.5, 76.0, 72.4, 70.6, 69.9, 67.5, 55.3,
42.5, 40.4, 38.1, 32.6, 29.6, 21.2, 20.8, 14.2. LRMS
(positive electrospray) calcd for C39H44FN5O7: [MþH]þ
714.33. Found: 714.44.
3.1.4. Claisen rearrangement products 19. Cinnamyl
ether 8 (100 mg, 0.125 mmol) is placed in a 5 mL conical
Smith Process Vial. Chromatography grade silica gel
(500 mg) is added and the mixture is suspended in 4 mL
CH3CN. The tube is sealed under N2 and inserted into the
reaction chamber of a SmithCreator desktop microwave
instrument. After heating (2.45 GHz) at 1808C for 40 min,
the mixture is concentrated and the residue purified by
flash chromatography (4!7% MeOH/CH2Cl2) to afford 19
(,1:1 diastereoisomeric mixture) as a pale yellow solid
(74 mg, 74%). Rf¼0.38 (10% MeOH/CH2Cl2). HRFAB
calcd for C47H49FN6O5: [MþH]þ 797.3827. Found:
797.3834.
3.1.6. Dihydrobenzoxepins 21. To a dry reaction flask
containing 20 (less polar diastereomer, 10 mg, 11.9 mmol)
is added a degassed (N2 purge) CH2Cl2 solution (0.8 mL) of
ruthenium alkylidene 22 (Strem chemical, 1.0 mg,
10 mol%). The resultant solution is heated to reflux for a
total of 10 h. Additional aliquots of catalyst solution
(5 mol%) are added at 3 and 7 h. The mixture is cooled to
rt, concentrated, and the residue purified by PTLC to afford
21 (6.2 mg, 65%) as a white film.
3.1.5. Mitsunobu etherification products 20. To phenols
19 (110 mg, 0.138 mmol) is added a solution of allyl alcohol
(24 mg, 28 mL, 0.414 mmol) in anhydrous THF (0.37 mL)
followed by PPh3 (109 mg, 0.414 mmol in 92 mL PhH). The
solution is cooled in an ice-water bath and treated with
diisopropyl azodicarboxylate (84 mg, 82 mL, 0.414 mmol)
dropwise via syringe. Stirring is continued at 08C for 30 min
and at rt for 30 min. Concentration and purification by flash
chromatography (CHCl3/EtOAc/MeOH 8/1.6/0.4!8/1.3/
0.7) affords 20 as white solids (two separated diastereomers
(1:1), 116 mg total, 81%). 20: less polar diastereomer:
Rf¼0.53 (CHCl3/EtOAc/MeOH¼6/3/1). [a]2D4¼-20.98 (c¼
0.46, MeOH). IR (film): 3296, 3065, 2927, 2855, 1648,
1
Compound 21. H NMR (400 MHz, CD3OD): d 7.54 (d,
J¼7.6 Hz, 1H), 7.41–7.47 (m, 3H), 7.31 (d, J¼8.0 Hz, 1H),
7.20 (app t, J¼8.0 Hz, 1H), 7.14 (d, J¼8.0 Hz, 1H), 7.05–
7.16 (m, 2H), 6.96–7.05 (m, 5H), 6.84 (d, J¼7.6 Hz, 1H),
6.62 (dd, J¼8.0 Hz, 1.6, 1H), 6.54 (s, 1H), 6.07–6.14 (m,
1H), 5.66 (app d, J¼11.2 Hz, 1H), 4.58–4.66 (m, 4H),
4.52–4.58 (m, 2H), 4.35 (dd, J¼19.0, 1.2 Hz, 1H), 3.70 (t,
J¼5.2 Hz, 2H), 3.26 (dd, J¼14.8, 6.0 Hz, 1H), 3.07–3.15
(m, 3H), 2.98 (dd, J¼14.8, 4.4 Hz, 1H), 2.89 (dd, J¼14.4,
8.8 Hz, 1H), 2.62–2.74 (m, 2H), 1.60–1.71 (m, 1H), 1.47–
1.60 (m, 1H), 1.37–1.46 (m, 2H), 1.28–1.37 (m, 2H), 0.92
(t, J¼7.6 Hz, 3H). 13C NMR (75 MHz, CDCl3): d 170.7,
170.3, 167.1, 158.3, 156.9, 145.1, 139.0, 137.7, 137.5,
136.3, 132.3, 131.4, 130.4, 129.9, 129.5, 129.3, 128.6,
1
1509, 1375, 1245, 1156, 1104, 837, 741 cm21. H NMR
(400 MHz, CDCl3): d 7.86 (s, 1H), 7.47 (dd, J¼8.8, 5.6 Hz,
2H), 7.35 (s, 1H), 7.34 (d, J¼7.6 Hz, 1H), 7.21 (t, J¼8.0 Hz,
1H), 7.10–7.19 (m, 2H), 6.96–7.06 (m, 4H), 6.92 (d,
J¼2.0 Hz, 1H), 6.84 (dd, J¼8.4 Hz, 2.0, 1H), 6.68 (d,
J¼8.4 Hz, 1H), 6.47 (dd, J¼8.4, 2.0 Hz, 1H), 6.36–6.44 (m,
3H), 6.22–6.34 (m, 2H), 6.16 (t, J¼4.0 Hz, 1H), 5.92 (sym

8954
Q. Wei et al. / Tetrahedron 59 (2003) 8947–8954
and hydrophobicity of a spacer used for solid-phase synthesis
experiments (Scheme 3).
128.5, 128.2, 126.8, 123.6, 122.9, 122.6, 121.3, 120.4,
118.7, 116.5, 115.8, 115.5, 111.7, 110.4, 109.2, 71.1, 65.5,
53.9, 53.4, 48.9, 47.5, 39.6, 36.9, 31.6, 28.7, 27.1, 20.2,
19.5, 14.0. LRMS (positive electrospray) calcd for
C48H49FN6O5: [MþH]þ 809.38. Found: 809.30. HRFAB
calcd for C48H49FN6O5: [MþH]þ 809.3826. Found:
809.3827.
9. Synthetic peptides Gly-His, Gly-Thr, Gly-Pro-Ser, Gly-Pro-
His, and Gly-Pro-Trp (each prepared as a C-terminal n-
butyramide) readily undergo condensation with 2 followed by
cycloaddition with 3. The products (analogous to 4), however,
do not cyclize when treated with the [(h3-allyl)PdCl]2/5
complex. While initial studies have focused on peptides
containing N-terminal Gly or Ala residues, we see no
indication of a limited tolerance at this position.
Acknowledgements
10. The structures assigned to regioisomers 11 and 12 are tenta-
tive. The chemical shift (400 MHz, CD3OD) of the b-styryl
proton in 11 (C16H, d 6.37) is more similar to that observed in
25-membered macrocycles (7–10, 13, 16, 17—range: d 6.29–
6.45) than it is to the 22-membered ring series [namely 6 (d
6.16) and 14 (d 6.21)]. The minor isomer (assigned as 12) has
this resonance appearing at d 6.15.
Funding provided by the NIH (RO1-GM60591), the NSF
(CAREER 9984282), the Howard Hughes Medical Institute,
the Robert A. Welch Foundation, and an unrestricted
research award from AstraZeneca. We thank Dr. Jia Zhou
for experimental assistance. P.G.H. is a fellow of the Alfred
P. Sloan Foundation.
11. Isolated 6 decomposes (t1/2,5 h) when re-subjected to the
same conditions used for its synthesis. Aqueous cyanide
minimizes product loss by poisoning the etherification
catalyst.
References
12. Williams, A. J.; Uto, Y.; Wipf, P.; Reno, M. J.; Williams, D. R.
Org. Lett. 2000, 2, 1165–1168.
1. Outlines of this general concept are also evident in Houghten’s
“Libraries from libraries experiments”. See: Ostresh, J. M.;
13. Dihydroxylation using ‘Super AD-mix-b’ (Ref. 14) provides
glycol 18 as a mixture (94:6) of diastereomers. The
stereochemistry of the major isomer (shown) is assigned
using an established mnemonic. The use of Super AD-mix-a
provides the same two diols (confirmed by HPLC co-injection)
but in a 92:8 ratio in which 18 is minor.
¨
Husar, G. M.; Blondelle, S. E.; Dorner, B.; Weber, P. A.;
Houghten, R. A. Proc. Natl. Acad. Sci. USA 1994, 91,
11138–11142.
2. Sisko, J.; Mellinger, M.; Sheldrake, P. W.; Baine, N. H.
Tetrahedron Lett. 1996, 37, 8113–8116.
3. (a) Sisko, J.; Kassick, A. J.; Mellinger, M.; Filan, J. J.; Allen,
A.; Olsen, M. A. J. Org. Chem. 2000, 65, 1516–1524.
(b) van Leusen, A. M.; Wildeman, J.; Oldenziel, O. H. J. Org.
Chem. 1977, 42, 1153–1159.
14. Nicolaou, K. C.; Yue, E. W.; La Greca, S.; Nadin, A.; Yang,
Z.; Leresche, J. E.; Tsuri, T.; Naniwa, Y.; De Riccardis, F.
Chem. Eur. J. 1995, 1, 467–494.
15. Aryl Claisen rearrangement is best performed by microwave
heating (1808C, 40 min) solutions of 8 containing flash
chromatography-grade silica gel (230–400 mesh, 4 g/mmol
8). Added silica gel allows the reaction to proceed to full
conversion without substantial degradation. Soluble Lewis-
acidic additives have thus far failed in this regard.
4. Imidazole 4 is produced as an inconsequential mixture of
diastereomers. This material, and others like it examined in
this study (‘condensation’ products in Table 1) are highly
insoluble in most common organic solvents as well as water.
Preparative scale purification therefore is a challenge.
Precipitation is found most effective although, because
recovery using this method is not complete, isolated yields
(Table 1) underestimate the true efficiency of these reactions.
5. For palladium catalysis relevant to this work see: (a) Trost,
B. M.; Toste, F. D. Int. Pat. App. (PCT) Publication number
WO 00/14033, 2000. (b) Trost, B. M.; Toste, F. D. J. Am.
Chem. Soc. 1998, 120, 9074–9075. (c) Trost, B. M.; Toste,
F. D. J. Am. Chem. Soc. 1998, 120, 815–816. (d) Goux, C.;
Lhoste, P.; Sinou, D. Synlett 1992, 725–727.
16. Chatterjee, A. K.; Morgan, J. P.; Scholl, M.; Grubbs, R. H.
J. Am. Chem. Soc. 2000, 122, 3783–3784.
17. In this discussion, ‘alkaloid-like’ is meant primarily to contrast
‘peptide-like’ Our goal is the systematic preparation, from
peptides, of non-reactive molecules that passively diffuse into
the cytoplasm of living cells from culture media.
18. Professor Matthew Shair (Harvard University) generously
provided us with resin that fueled our initial experiments. See:
Tallarico, J. A.; Depew, K. M.; Pelish, H. E.; Westwood, N. J.;
Lindsley, C. W.; Shair, M. D.; Schreiber, S. L.; Foley, M. A.
J. Comb. Chem. 2001, 3, 312–318.
6. A more polar by-product repeatedly generated in this reaction
has been isolated and found to have 1H NMR and mass
spectral data consistent with a symmetrical dimer.
19. Fmoc Solid Phase Peptide Synthesis; Chan, W. C., White,
P. D., Eds.; Oxford University: Oxford, 2000; ISBN
0199637245.
20. At present, alcohol 26 (assignment supported by 1H NMR
and mass spectrometry) is isolated together with un-reacted
H2N-Gly-Tyr-NH(CH2)3OH. Achieving complete conversion
in the synthesis of 25 from 24 is the subject of ongoing
experiments.
7. (a) Yin, J.; Buchwald, S. L. J. Am. Chem. Soc. 2002, 124,
6043–6048. (b) Kamer, P. C. J.; van Leeuwen, P. W. N. M.;
Reek, J. N. H. Acc. Chem. Res. 2001, 34, 895–904.
(c) Kranenburg, M.; van der Burgt, Y. E. M.; Kamer, P. C.
J.; van Leeuwen, P. W. N. M. Organometallics 1995, 14,
3081–3089.
8. Peptides used in this study were prepared (Fmoc method-
ology) as C-terminal n-butyramides to approximate the size