Synthesis of structurally diverse bis-peptide oligomers

Article abstract of DOI:10.1021/jo0609125

We have developed second-generation monomers 1 and 2 and improved conditions for rapidly and simultaneously closing multiple diketopiperazines on solid support. These new conditions involve either the microwave heating of a suspension of solid-supported a

Full text of DOI:10.1021/jo0609125

Synthesis of Structurally Diverse Bis-peptide Oligomers
Sharad Gupta, Megan Macala, and Christian E. Schafmeister*
Department of Chemistry, UniVersity of Pittsburgh, Pittsburgh, PennsylVania 15260
meister@pitt.edu
ReceiVed May 2, 2006
We have developed second-generation monomers 1 and 2 and improved conditions for rapidly and
simultaneously closing multiple diketopiperazines on solid support. These new conditions involve either
the microwave heating of a suspension of solid-supported amino-tetrafluoropropyl esters in acetic acid/
triethylamine catalyst solution or continuous flow of catalyst solution through the resin, heated in a flow
cell apparatus. We demonstrate that the new monomers 1 and 2 can be combined with the new conditions
easily to synthesize previously inaccessible hetero and homo spiro ladder oligomers 3 and 4 and others.
The diketopiperazine (DKP) motif is found throughout
chemistry. Many natural products containing 2,5-piperazinedione
rings have been isolated, displaying a wide variety of biological
functions.1 Medicinal chemistry also makes considerable use
of the diketopiperazine scaffold,2 and DKPs have been found
to be a useful scaffold for molecular recognition.3 Our laboratory
has developed an approach to the synthesis of nanometer-scale
spiro ladder oligomers with controlled three-dimensional
structures.4-6 Our long-term goal is to develop these molecules
for biomimetic and nanotechnology applications. Our oligomers
are composed of cyclic monomers (Figure 1) joined by
diketopiperazine rings (e.g., compound 3). We assemble these
oligomers in two phases: assembly followed by rigidification.
FIGURE 1. Protected bis-amino acid monomers used to optimize DKP
formation on solid support. We name the monomers with m ) 1 the
“pro4” monomers and those with m ) 2 the “pip5” monomers.
(1) Isaka, M.; Palasarn, S.; Rachtawee, P.; Vimuttipong, S.; Kongsaeree,
P. Org. Lett. 2005, 7, 2257-2260.
(2) Zeng, Y. B.; Li, Q. S.; Hanzlik, R. P.; Aube, J. Bioorg. Med. Chem.
Lett. 2005, 15, 3034-3038.
In the assembly phase, we assemble stereochemically pure,
cyclic bis-amino acid monomers on solid support using the
techniques of solid-phase peptide synthesis to form flexible
oligomers. In the rigidification phase, we promote the simul-
taneous formation of a diketopiperazine between each adjacent
(3) Krattiger, P.; Wennemers, H. Synlet 2005, 706-708.
(4) Levins, C. G.; Schafmeister, C. E. J. Am. Chem. Soc. 2003, 125,
4702-4703.
(5) Habay, S. A.; Schafmeister, C. E. Org. Lett. 2004, 6, 3369-3371.
(6) Gupta, S.; Das, B. C.; Schafmeister, C. E. Org. Lett. 2005, 7, 2861-
2864.
10.1021/jo0609125 CCC: $33.50 © 2006 American Chemical Society
Published on Web 10/18/2006
J. Org. Chem. 2006, 71, 8691-8695
8691

Gupta et al.
SCHEME 1. Solid Supported Diketopiperazine Formation
Model Reactionsa
pair of monomers in the oligomer to form a rigid spiro ladder
oligomer. To form a spiro ladder oligomer containing N
monomers, we must simultaneously close (N-1) diketopipera-
zines. To achieve this with useful yields, it is very important
that the diketopiperazine formation reaction be fast and exhibit
no side reactions.
Diketopiperazine formation has been the subject of several
reviews.7,8 A particularly facile route and the one that we initially
used is an intramolecular aminolysis reaction between an amine
and an ester catalyzed by 20% piperidine in dimethylformamide
(DMF).4 We found these conditions to be quite effective for
pro4 monomers, but they are not ideal for several reasons.
Occasionally, precipitate would form in the reaction mixture
and the yield of fully closed product would be low. We also
discovered that for oligomers containing hin(2S4R7R9R)5 or
pip5(2S5S)6 the DKP formation between monomers under 20%
piperidine in DMF at room temperature was very slow or
undetectable. The rate of DKP formation is known to be
sensitive to the stereochemistry and nature of the substituents
on the forming DKP ring.9 We cannot use high temperatures
or long reaction times to drive the aminolysis reaction toward
completion because DKPs epimerize under basic conditions.10
DKP formation is catalyzed by carboxylic acids9 as well as
base.11 The mechanism and the rate-determining steps are
different under acidic conditions than basic conditions.9 We
hypothesized that we could use carboxylic acids to catalyze DKP
formation in difficult cases. We also hypothesized that using
acid-catalyzed conditions would allow us to use higher tem-
perature to accelerate the reaction without fear of epimerization.
Thus, we have undertaken a search for improved conditions for
closing diketopiperazines.
a These bis-amino acid dimers are assembled from one of 6a or 6b
followed by one of 7a or 7b.
TABLE 1. Half-Life of Diketopiperazine Formation for Different
Dimers in 80 mM AcOD, DMF, 100 °C
sequence
half-life, t1/2 min
pro4-pro4 (8a f 9a)
pro4-pip5 (8b f 9b)
pip5-pro4 (8c f 9c)
pip5-pip5 (8d f 9d)
6.5
12
16
78
TABLE 2. Effect of Solvent and Temperature on Diketopiperazine
Closure in the Reaction 8b f 9b Using 80 mM AcOD as the
Catalyst
To explore different solvents, catalyst concentrations, and
temperatures, we chose to close the diketopiperazines on solid
support. Accordingly, we substituted the benzyloxycarbonyl
(Cbz) protecting group of monomers 5a and 5b with the tert-
butoxycarbonyl (Boc) protecting group to create 7a and 7b and
substituted the methyl esters of monomers 5a and 5b with para-
nitrobenzyl esters to give monomers 6a and 6b. We carried out
oligomer syntheses on trifluoromethylsulfonic acid cleavable
linkers (MBHA and hydroxyl methyl) and then measured the
kinetics of DKP formation in a flow cell containing resin bound
dimers 8a-d (Scheme 1) by quantitating the amount of para-
nitrobenzyl alcohol in the effluent as different catalyst solutions
were flowed over the resin at varying temperatures.
We carried out a series of diketopiperazine closures examin-
ing how the nature of the monomers affects the rate of
intramolecular aminolysis using acetic acid-d as a bifunctional
catalyst (Table 1, Scheme 1). An 80 mM solution of acetic
acid-d in DMF at 100 °C is a good catalyst for the formation
of 9a, 9b, and 9c but four times slower for the formation of 9d.
Deutero acetic acid was used because epimerization of 8a-d
would give rise to an increase in molecular weight of at least 1
Da. When the dimers were cleaved from the resin, no epimer-
ization was observed using these conditions (data not shown).
To explore more optimal acetic acid-catalyzed DKP formation
conditions, we investigated the effect of solvent and temperature
half-life, t1/2 (min), for
temp (°C)
DMF
o-xylene
40
70
100
130
126
31
12
n/a
115
18
<4
<2
on AcOD-catalyzed formation of 9b (Scheme 1, Table 2).
Increasing the temperature accelerates the rate of DKP formation
in both solvents, and the temperature effect is stronger in
o-xylenes. Dimethylformamide decomposes above 100 °C and
is thus not a suitable solvent at high temperatures. These
observations led us to choose o-xylene as the standard solvent
for subsequent experiments. Gisin and Merrifield found that the
maximum rate of acetic acid-catalyzed DKP formation occurred
at an acetic acid (AcOH) concentration of 100 mM.9
The work of Capasso and Mazzarella suggests that triethyl-
ammonium acetate is at least 10 times faster at catalyzing DKP
formation than acetic acid by itself.12 In this study, it is likely
that the triethylamine served to neutralize the acidic trifluoro-
acetate salt of the dipeptides that underwent diketopiperazine
formation. In our system, we free-base our amines on solid
support by washing the deprotected resin with triethylamine in
DMF prior to diketopiperazine closure. Nonetheless, we exam-
ined the effect of adding triethylamine to the catalyst solution
and found that DKP formation was accelerated by a factor of 2
to 3 in the formation of 9a or 9d (Scheme 1, Table 3) relative
to when no triethylamine was added. The acid/base equilibrium
between Et3N:AcOH and Et3NH+:AcO- is likely to lie on the
(7) Fischer, P. M. J. Pept. Sci. 2003, 9, 9-35.
(8) Dinsmore, C. J.; Beshore, D. C. Tetrahedron 2002, 58, 3297-3312.
(9) Gisin, B. F.; Merrifield, R. B. J. Am. Chem. Soc. 1972, 94, 3102-
3106.
(10) Gund, P.; Veber, D. F. J. Am. Chem. Soc. 1979, 101, 1885-1887.
(11) Pedroso, E.; Grandas, A.; Delasheras, X.; Eritja, R.; Giralt, E.
Tetrahedron Lett. 1986, 27, 743-746.
(12) Capasso, S.; Mazzarella, L. Peptides 1998, 19, 389-391.
8692 J. Org. Chem., Vol. 71, No. 23, 2006

Structurally DiVerse Bis-peptide Oligomers
TABLE 3. Half-Life of Diketopiperazine Formation in the
Reactions 8a f 9a and 8d f 9d: Exploring Different
Concentrations of Acetic Acid and Triethylamine
SCHEME 3. Test of Our Second-Generation Monomers
and Diketopiperazine Formation Conditions Using a
Pentamer of 2
half-life
t1/2 (min)
half-life
1/2 (min)
t
conditions
(solvent: o-xylene)
at 130 °C
for 8d f 9d
at 80 °C
for 8a f 9a
50 mM AcOD
80 mM AcOD
100 mM AcOD
9.5
12.5
9.2
18.8
18
50 mM AcOD + 50 mM Et3N
100 mM AcOD + 50 mM Et3N
1 mM Et3N
5
4.3
266
4.6
3.3
SCHEME 2. Effect of pKa of the Leaving Group on
AcOH-Catalyzed Diketopiperazine Formation
of DKP; however, only 5 min are required for dimer containing
TFP ester to obtain the same amount of DKP product. This
experiment allowed us to quantify epimerization under these
conditions. In the case of the methyl ester, a separate peak was
observed for epimerized product. This epimer peak had a mass
spectrum that indicated 50% deuterium incorporation, suggested
by a large change in the ratios of M + 1, M + 2, and M + 3
peaks of the mass spectrum. This new peak represented only
3% epimerized product; thus, even after 6 h of heating at 80
°C in a flow cell in the presence of 80 mM AcOD/o-xylene,
there is very little epimerization. We did not detect any
epimerized product in the case of the TFP ester. These
observations suggest that the acid-catalyzed DKP formation
results in very little epimerization if the reaction time is kept
short. By exchanging the methyl ester for more reactive esters,
we can shorten the reaction time considerably.
To test our best catalysis conditions, we synthesized the
monomer 2 and assembled a pentamer on a 4-methylbenzhy-
drylamine (MBHA) resin and on solid support removed the Boc
protecting group on each monomer to form intermediate 13.
We suspended the resin bound 13 in 100 mM AcOD and 50
mM Et3N and exposed it to microwaves at 130 °C for 30 min
(Scheme 3). The resin cleavage product 14 was characterized
by C18 reverse-phase HPLC with mass spectrometry (Figure 2).
The major peak was the desired product 14. Two small peaks
representing intermediates that had failed to close one dike-
topiperazine were also observed, but there was no evidence of
epimerization of 14.
side of the neutral species in the apolar solvent o-xylene, and
thus it should not reduce the concentration of acetic acid catalyst.
Triethylamine could be increasing the rate of diketopiperazine
formation by deprotonating any secondary ammonium ions on
the resin in an acid/base pre-equilibrium.
After screening several esters, we developed the second-
generation monomers 1 and 2 as a replacement for the methyl
ester version of our first-generation monomers 7a and 7b. We
explored trifluorethyl esters, trichloroethyl esters, and acetoxy
esters and found that they were not suitable because they react
with piperidine and DBU during our solid-phase oligomer
assembly (data not shown). Ultimately, we found that 2,2,3,3-
tetrafluoropropyl (TFP) esters are a good compromise of stability
to piperidine and reactivity in diketopiperazine formation.
To examine the effect of the leaving group on the rate of
DKP formation, we set out to measure the rate of the intra-
molecular aminolysis reaction in pro4-pro4 dimers 10a and 10b
under the same solvent and temperature conditions (Scheme 2).
In 10a a secondary amine attacks a methyl ester, while in 10b
it attacks a tetrafluoropropyl ester. Unlike para-nitrobenzyl
alcohol, the release of methyl alcohol or tetrafluoropropyl
alcohol cannot be observed and quantified by HPLC. Hence, a
different method was used to monitor the progress of the
reaction. We carried out the reaction by flowing catalyst solution
over separate batches of resin-supported dimer (80 mM AcOD
in o-xylene at 80 °C) for definite time intervals: 1 h for 10a
and 5 min for 10b. After each reaction, we cleaved the oligomer
from the resin and quantified the amount of closed product and
open product. We found that the acceleration provided by
substitution of the methyl ester by a tetrafluoropropyl ester is
too large to be accurately quantified by this experiment. In the
case of the methyl ester, it takes 6 h to obtain 70% formation
We then set out to synthesize a small library of structurally
diverse bis-peptide oligomers containing pip5(2S5S) and pro4-
(2S4S) monomers and test the efficacy of our newly optimized
conditions for diketopiperazine formation. We designed four
hetero sequences 15-18 and assembled them on hydroxymethyl
polystyrene resin (Scheme 4). This was followed by on resin
diketopiperazine formation in a flow cell apparatus catalyzed
by 100 mM AcOH and 50 mM Et3N in o-xylene at 130 °C.
The resin cleavage products 3, 4, 19, and 20 were purified by
C18 reverse-phase preparative HPLC and characterized by high-
resolution mass spectrometry. The analysis of the crude products
before purification by C18 reverse-phase HPLC reveals that in
all four cases diketopiperazine formation was very clean. Only
trace amounts of any product with one diketopiperazine ring
unclosed were detected (Figure 3).The oligomers 3, 4, 19, and
20 gave rise to high-resolution mass spectra with m/z )
762.2927, 762.2844, 762.2880, and 762.2865, respectively
(calculated 762.2847).
J. Org. Chem, Vol. 71, No. 23, 2006 8693

Gupta et al.
FIGURE 2. Unpurified reverse-phase C18 HPLC chromatograms of
14 (0.1% HCO2H, 0-25% AcN over 30 min). The main peak is the
desired product (m/z ) 941).
FIGURE 3. Reverse-phase C18 HPLC chromatograms of 20 (0.1%
HCO2H, 0-25% AcN over 30 min) before purification. The main peak
is the desired product (m/z ) 762).
resulting coupling mixture was mixed well and immediately added
to the SPPS reactor carrying the resin. The resin and coupling
mixture were agitated for 2 h. The coupling solution was drained,
and the resin was washed two times each with DMF and CH2Cl2
alternately. Any residual free amine was acylated by treatment with
10% acetic anhydride/pyridine (250 µL) for 5 min. The resin was
then washed five times with DMF and CH2Cl2 alternately, and the
terminal Fmoc-protected amine was deblocked by treatment with
2% DBU/DMF for 5 min. UV-visible spectroscopic analysis of
the piperidine fluorenyl adduct, formed by diluting the deprotection
reaction solvent with 20% piperidine/DMF at 301 nm (ꢀ ) 7800
M-1 cm-1), indicated quantitative coupling based on resin loading.
The resin was washed five times with DMF and CH2Cl2 alternately.
Monomer 6b (19.4 mg, 28.5 µmol) was coupled to the resin
using HATU in a fashion similar to that described above. The
coupling reaction was repeated one additional time followed by
capping of any residual free amine and Fmoc deprotection using
the procedure given above. The overall sequential procedure for
In conclusion, we have developed the second-generation
monomers 1 and 2 and AcOD/Et3N-catalyzed, microwave-
assisted, and continuous flow conditions for simultaneously
closing multiple diketopiperazines on solid support with insig-
nificant amounts of epimerization. We are now using this new
methodology to synthesize larger and more complex spiro ladder
oligomers for biomimetic and nanotechnology applications.
Experimental Section
Compound 8c. MBHA resin (23 mg, 14.6 µmol free amine,
0.62 mmol loading) was transferred to a 1-mL polypropylene solid-
phase peptide synthesis (SPPS) reactor and allowed to swell for
30 min in DMF. In a 1.5-mL polypropylene microcentrifuge vial,
Fmoc-L-Tyr(t-Bu)-OH (13.1 mg, 28.5 µmol) was dissolved in 20%
CH2Cl2/DMF (143 µL). To this mixture was added DIPEA (9.9
µL, 57 µmol), followed by HATU (10.8 mg, 28.5 µmol). The
SCHEME 4. Assembly of Short Hetero Oligomers and on Resin Diketopiperazine Formation by Continous Flow Method Using
Optimized Catalytic Conditions
8694 J. Org. Chem., Vol. 71, No. 23, 2006

Structurally DiVerse Bis-peptide Oligomers
coupling an amino acid moiety was repeated one additional time
for monomer 5a. The coupling yields were estimated as 90-95%
on the basis of spectroscopic analysis described above.
product was precipitated a second time from ether. After centrifuga-
tion, ether was decanted and the pellet was dissolved in 200 µL of
1:1 acetonitrile/water. Quantities of 10 µL of this solution were
injected into HPLC-MS for analysis. HPLC-MS: column, Waters
XTerra MS C18 column 4.6 mm × 150 mm; mobile phase, MeCN
(0.05% formic acid)/water (0.1% formic acid), 0-25% MeCN over
30 min; flow rate 0.80 mL/min; UV detection at 274 nm; tR for
14, 16.03 min; ES-LRMS m/z (ion), calcd 941.4 (M + H+), obsd
941.2.
Compounds 15-18. Each of the hetero oligomers 15-18 was
assembled on hydroxymethyl polystyrene resin (0.98 mmol/g
loading) on a 20-mg scale. Monomer X1 was attached to the
hydroxymethylpolystyrene resin using the MSNT/MeIm method
(Novabiochem 2004/05 catalogue, Method 2-12). The resin was
allowed to swell in dry CH2Cl2 in a SPPS reactor for 10 min. In
the case of oligomer 18, 25 mg of 2 (40 µmol, 2 equiv) was
dissolved in 300 µL of dry CH2Cl2. To this solution were added
2.3 µL of MeIm (28 µmol, 1.4 equiv) followed by 12 mg of MSNT
(40 µmol). This coupling solution was added to the pre-swelled
resin and gently agitated for 30 min. The coupling solution was
drained, and the resin was washed many times with MeOH and
CH2Cl2 alternately. Any unreacted alcohol groups were capped by
treatment with 500 µL of 10% Ac2O in pyridine for 5 min. Finally,
the resin was drained and washed many times with MeOH and
CH2Cl2, followed by only CH2Cl2. Other monomers, X2-4, followed
by Fmoc-L-Tyr(t-Bu)OH were coupled sequentially using HATU
following a procedure similar to the one as described for oligomer
13. Subsequently, Fmoc and Boc group were removed, and the
resins were prepared for diketopiperazine closure using the previ-
ously described method for 13.
Compounds 3, 4, 19, and 20. The resins carrying desired
sequence (15-18) were transferred to the PTFE tubing of flow
cell apparatus, and the catalytic solution (100 mM AcOH and 50
mM Et3N in o-xylene) was pumped through the resin at 200 µL/
min for 2 h at 130 °C. The resins were recovered and then cleaved
using trifluoromethanesulfonic acid in the presence of scavengers
as described for 14. After a single precipitation in ether and
centrifugation, the resulting pellets were dissolved in 10% aceto-
nitrile in water and purified by preparative HPLC: mobile phase
MeCN (0.05% formic acid)/water (0.1% formic acid), 0-25%
MeCN over 30 min; flow rate 15 mL/min. The desired fractions
were pooled and lyophilized giving compounds 3, 4, 19, and 20;
HRESIQTOFMS calcd for C35H40N9O11 (M + H+) 762.2847, found
762.2927 for 3, found 762.2844 for 4, found 762.2880 for 19, found
762.2865 for 20.
After the coupling of the second monomer, the terminal amine
was acylated by treatment with 10% acetic anhydride/pyridine
(2 × 5 min). The resin was washed three times with DMF and
CH2Cl2 alternately and then three times with MeOH and CH2Cl2
alternately. Boc group removal was affected by treatment of the
resin with 1 mL of 1:1 TFA/DCM (2 × 15 min). The resin was
washed with CH2Cl2 and MeOH many times and neutralized by
washing a few times with 10% Et3N/CH2Cl2. Finally, the resin was
washed with CH2Cl2 and MeOH many times, followed by only
CH2Cl2 and then dried under reduced pressure overnight. Other
bis-amino acid sequences (8a, 8b, and 8d) were synthesized in the
same manner.
Measurement of Kinetics of Diketopiperazine Formation. A
tiny amount of resin (2 to 3 mg) carrying desired dimer (e.g., 8c)
was loaded in a homemade flow cell apparatus, and catalyst solution
was made to flow through the resin at 100 µL/min at the desired
temperature (details given in Supporting Information). The fractions
were collected at 15-min intervals and analyzed by HPLC. Areas
under the peaks were recorded for the reference and the p-
nitrobenzyl alcohol and used for further calculations as shown in
Supporting Information.
Compound 13. The homo oligomer 13 was synthesized on
MBHA resin (10-mg resin, 0.62 mmol/g loading) following a
procedure similar to the one described for 8c. The only difference
was that, instead of 2% DBU in DMF, 20% piperidine in DMF
solution for 20 min was used to effect Fmoc deprotection.
Accordingly, 5 units of monomer building block 2 were coupled
sequentially to the resin, followed by Fmoc-L-Tyr(t-Bu)-OH. This
was followed by removal of terminal Fmoc group to unmask the
free amine and removal of Boc groups to reveal secondary amino
groups. The resin was washed with CH2Cl2 and MeOH many times
and neutralized by washing a few times with 10% Et3N/CH2Cl2.
Finally, the resin was washed with DCM and MeOH many times,
followed by only CH2Cl2, and then dried under reduced pressure
overnight to yield 13.
Compound 14. A few beads of MBHA resin carrying homo
oligomer sequence 13 were placed in a 10-mL microwave reaction
vessel (10-mL capacity, glass) containing 5 mL of a solution of
100 mM AcOH and 50 mM Et3N in o-xylene. The vessel was
capped and placed in the microwave reactor (CEM Discover) and
irradiated (300 W, maximum power, 130 °C, 5 min ramp) with
continuous stirring. After 60 min, the tube was removed from the
microwave reactor, and the resin was transferred to a 1-mL
polypropylene SPPS reactor. The resin was washed with CH2Cl2
and MeOH many times, followed by only CH2Cl2, and then dried
under reduced pressure overnight.
Acknowledgment. This material is based on work supported
in part by the National Science Foundation under Grant 0348823
(monomer 1 and 2) and the NIH/NIGMS (GM067866) (mono-
mer 7b).
The reactor carrying the resin was kept in an ice bath, and 25
µL of thioanisole, 12.5 µL of 1,2-ethanedithiol (EDT), and 250 µL
of TFA were added. The cleavage mixture was stirred for 5 min,
and 25 µL of TFMSA was added. The stirring was continued for
1 h at 0 °C followed by 1 h at room temperature. The cleavage
mixture was dripped into 50 mL of ether in an Eppendorf tube that
caused desired product 14 to precipitate out. This precipitate was
pelleted by centrifugation for 30 min at 4 °C at 3200 rpm. Ether
was decanted, precipitate was dissolved in 200 µL of TFA, and
Supporting Information Available: Experimental procedures,
schematics of the flow cell apparatus, methodology for half-life
estimation, and optimization of reaction condition for diketopiper-
azine formation, relevant NMR spectra, and other characterization
data for new compounds. This material is available free of charge
via the Internet at http://pubs.acs.org.
JO0609125
J. Org. Chem, Vol. 71, No. 23, 2006 8695