Dynamic combinatorial libraries of hydrazone-linked pseudo-peptides: Dependence of diversity on building block structure and chirality

Article abstract of DOI:10.1039/b617217b

Expanding on our earlier building block architecture [(MeO) ₂CH-Linker-Pro-X-NHNH₂ where X = Phe, Cha], we have produced a series of new pseudo-dipeptides [(MeO)₂CH-Linker-Pro-X- NHNH₂ where X = Val, Leu, Ile, A

Full text of DOI:10.1039/b617217b

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Dynamic combinatorial libraries of hydrazone-linked pseudo-peptides:
dependence of diversity on building block structure and chirality
Jingyuan Liu, Kevin R. West, Chantelle R. Bondy and Jeremy K. M. Sanders*
Received 27th November 2006, Accepted 8th January 2007
First published as an Advance Article on the web 26th January 2007
DOI: 10.1039/b617217b
Expanding on our earlier building block architecture [(MeO)2CH–Linker–Pro–X–NHNH2 where X =
Phe, Cha], we have produced a series of new pseudo-dipeptides [(MeO)2CH–Linker–Pro–X–NHNH2
where X = Val, Leu, Ile, Ala] for use in hydrazone-based dynamic combinatorial libraries (DCLs);
reverse order analogues [Phe-Pro and Val-Pro] and two enantio-analogues [Pro-Phe and Pro-Val] were
also prepared. The behaviours of these building blocks in DCLs, as single components and in mixtures,
were studied systematically using HPLC and mass spectrometry in order to gain insight into the
relationship between building block structure and good library diversity. Subtle changes in building
block structure lead to significant changes in library distribution and in the ability to produce diverse
libraries in mixtures.
that all the components of the library and the template are stable
Introduction
under acidic conditions.
Dynamic Combinatorial Chemistry is a powerful approach for the
Few principles for the design of successful building blocks have
discovery of new functionally successful receptors and ligands.1 Its
emerged. Many factors need to be addressed when designing a
key feature is the dynamic combinatorial library (DCL), in which
bifunctional building block, such as solubility, rigidity/flexibility
each library member is assembled from building blocks that bond
and suitable molecular recognition sites. Excessive rigidity may
reversibly to one another under specific reaction conditions. As a
limit the diversity of the library members, favouring the formation
of a single product or giving self-sorting in mixed libraries.9–13
Conversely, building blocks with excessive flexibility may lead to
internal cyclisation. Previously, we reported the DCL behaviour
of a pseudo-dipeptide building block, pPF, which contains a para-
substituted aromatic spacer (p), L-Pro (P) and L-Phe (F). The
untemplated library produced from pPF has been successful at
generating diversity. This may result from the Pro enforcing a
degree of curvature, while the remainder of the structure gives
flexibility.6
It is not obvious what structural changes to a building block will
have a subtle or dramatic effect on the diversity of a DCL. One
can imagine several possible scenarios in a DCL containing a
mixture of building blocks. At one extreme they could form
mixed products present in a statistical distribution with all the
products having very similar geometries and properties. At the
other extreme, geometrical incompatibilities could lead to self-
sorting.10–13 We have therefore investigated the effects of varying
the amino acids, their chirality and the order in which they appear
in the building blocks to explore the diversity of the resulting
DCLs. Herein, we report how the composition of DCLs derived
from these various building blocks, as single components and in
mixtures, produces libraries of varying diversity. The composition
of the library members in these DCLs has been determined by
LC-MS, while MS-MS was used to elucidate the sequence of the
constitutional isomers from mixed libraries.14 This investigation
focuses purely on analysing the diversity of the library, so not
all library members were isolated for full characterisation; in
a practical library one is usually interested only in successfully
amplified species. Templating results from these new building
blocks will be reported elsewhere.
result of this reversibility, library members can interconvert to give
a distribution that is under thermodynamic control. Addition of
a guest or template molecule that can selectively recognise at least
one receptor in the library will increase the concentration of that
host at the expense of poorly binding members of the library.2–4
The successful host is then isolated, identified and studied. Such
systems demonstrate amplification of molecules with the desired
properties in a manner that is reminiscent of the mammalian im-
mune system. Perhaps the most remarkable example of the power
of dynamic combinatorial chemistry to uncover new receptors
is our unexpected discovery of a catenane that is amplified by
acetylcholine chloride from the simple hydrazone DCL generated
from a pseudo-dipeptide building block.5 If this approach is to
achieve its full potential, we need both a diverse library in order to
maximise the probability of finding successful structures and an
understanding of the relationship between building block structure
and DCL behaviour.
Hydrazone exchange is one of the most successful reactions used
so far in dynamic combinatorial chemistry.1,5–8 A hydrazone link-
age is formed from the acid-catalysed condensation of a hydrazide
and an aldehyde. Hydrazone bonds are readily broken and formed
under acidic conditions, making the process reversible, while
neutralisation yields stable, isolable products. These conditions
are compatible with molecular recognition events that are capable
of influencing the equilibrium distribution of products, provided
University Chemical Laboratory, Lensfield Road, Cambridge, CB2 1EW,
UK. E-mail: jkms@cam.ac.uk; Fax: +44 1223 336017; Tel: +44 1223
336411
778 | Org. Biomol. Chem., 2007, 5, 778–786
This journal is
The Royal Society of Chemistry 2007
©

View Article Online
Table 1 Summary of building blocks
Name
Amino acid 1 (AA1) Amino acid 2 (AA2) Overall yield
pPF
pPC
pPL
pPV
pPI
pPA
pFP
pVP
D-pPF
L-Pro
L-Pro
L-Pro
L-Pro
L-Pro
L-Pro
L-Phe
L-Val
D-Pro
L-Phe
L-Cha
L-Leu
L-Val
L-Ile
L-Ala
L-Pro
L-Pro
D-Phe
D-Val
43% (ref. 6)
68% (ref. 6)
62%
49%
45%
45%
68%
39%
43%
D-pPV D-Pro
49%
Results and discussion
Syntheses of building blocks and preparation of DCLs
The structural variations made to the building blocks were
restricted to the hydrophobic amino acids to ensure solubility
in a single solvent system. Using the pPF building block as our
model, the modifications explored consist of exchanging Phe for
Val, Leu, Ile and Ala, as well as varying the order of the Pro
amino acid between the first and second position. Enantiomeric
building blocks, D-pPF and D-pPV, and a mixed chirality building
block, LD-pPF, were also prepared; however, all DCLs prepared
using the LD-pPF building block suffered precipitation problems,
preventing further studies. Fig. 1 shows the general structure of
the building blocks, while Table 1 summarises the building blocks
explored and identifies the nomenclature used throughout.
Scheme 1 Syntheses of building blocks using pPL as an example, where
amino acid 1 is L-Pro and amino acid 2 is L-Leu. Reagents and conditions:
(i) EDC, DMAP and Et3N in dry CH2Cl2, inert atmosphere, 90–96%.
(ii) 5% Pd/C and H2 in MeOH, quantitative. (iii) EDC, DMAP and Et3N
in dry DCM, inert atmosphere, 54–88%. (iv) NH2NH2·H2O in MeOH,
63–87%.
TFA (100 mM) was used to initiate hydrazone exchange. The
solution was stirred at room temperature for 4 days, at which point
thermodynamic equilibrium was reached. Control experiments
were performed to confirm that DCLs were under thermody-
namic control. The library members were separated using HPLC
(UV = 290 nm) and identified using electrospray ionisation
mass spectrometry (ESI-MS) and MS-MS to distinguish between
constitutional isomers.
Fig. 1 General building block structure, where AA1 and AA2 are defined
in Table 1.
The building blocks were prepared as shown in Scheme 1.
The general procedure involved the EDC coupling of CBZ-
protected amino acid 1 with amino acid 2 methyl ester. This was
followed by the deprotection of amino acid 1 via Pd/C-catalysed
hydrogenation. The amine of the dipeptide was then coupled to 4-
carboxybenzaldehyde dimethoxyacetal (pDMA) using EDC, once
again. The final product was obtained by hydrazinolysis of the
dipeptide methyl ester, giving an overall yield that varied between
39 and 68%.
Product distributions in libraries of single building blocks
These DCLs gave relatively simple distributions dominated by
macrocyclic dimers, trimers and tetramers. The overall distribu-
tions fell into four structure-related families (Table 2). Building
blocks forming the first family contain a Pro residue as amino
acid 1 and a secondary b-carbon on the side chain of amino acid
2. The building blocks pPF, pPC and pPL make up this family,
producing similar distributions of dimers, trimers and tetramers,
as exemplified by the chromatogram of pPF in Fig. 2a. The second
family consists of two building blocks, pPV and pPI, which possess
The DCLs were prepared in a CHCl3–DMSO (97 : 3 v/v)
solution containing a total building block concentration of 2 mM;
Table 2 Building block families
Family
Building blocks
AA2 Side-chain properties
Library composition
1
2
3
4
pPF, pPC, pPL
pPV, pPI
pPA
2◦ b-Carbon
Dimer ≈ trimer ≈ tetramer
Tetramer
Trimer > dimer
Trimer > dimer
3◦ b-Carbon
1◦ b-Carbon
Reverse order amino acids
pFP, pVP
This journal is
The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 778–786 | 779
©

View Article Online
Fig. 3 HPLC traces of DCLs (a) pPF (F) + pPC (C) and (b) pPV (V) +
pPI (I).
pPC2, is 1 : 2 : 1, and likewise the distributions of trimers and
tetramers are found to be 1 : 3 : 3 : 1 and 1 : 4 : 6 : 4 : 1, respectively.
The overall distribution of dimers, trimers and tetramers in general
agrees with the distribution from the single building block libraries.
Comparable diversity was also observed for the mixed DCL of
pPV and pPI (Fig. 3b).
Mixing building blocks from different families
When building blocks of different families are mixed, the overall
product distributions of the DCLs are unpredictable. The mixed
DCL of pPF and pPV contains a broad distribution of oligomers
that is dominated by mixed tetramers (Fig. 4). Two of the mixed
tetramer constitutional isomers, pPF–pPF–pPV–pPV and pPF–
pPV–pPF–pPV, are well separated and distinguished by their MS-
MS fragmentation patterns (Fig. 5). The first-eluted tetramer (18.5
min) fragments into three distinct dimers, pPF–pPF, pPF–pPV
and pPV–pPV, in an approximate 1 : 2 : 1 ratio, strongly suggesting
the cyclic oligomer sequence of pPF–pPF–pPV–pPV (Fig. 5a).
The second tetramer (19 min) produces only pPF–pPV fragments,
consistent with the pPF–pPV–pPF–pPV sequence (Fig. 5b).
Fig. 2 HPLC traces of DCLs (a) pPF (F), (b) pPV (V), (c) pPA (A) and
(d) pFP (u). The subscript denotes the number of building blocks in each
oligomer.
tertiary b-carbons on the side chains of amino acid 2 and strongly
favour the formation of tetramers, although dimers and trimers
are also detected (Fig. 2b). The corresponding building block with
a primary b-carbon on amino acid 2, pPA, favours the formation
of the trimer, with a modest proportion of dimer also observed
(Fig. 2c); however, low solubility prevented studies of this building
block in mixed libraries. The DCLs from the reverse peptide
sequence building blocks, pFP and pVP, form trimers as the dom-
inant product, with small amounts of other oligomers (Fig. 2d).
Whether the composition of the library from the first and
second families resulted from stabilisation of the tetramer by
intramolecular interactions or from steric destabilisation of the
smaller oligomers is unclear. The much greater dominance of
tetramer in the second family actually reflects a rather small
energetic difference from the first family.
Fig. 4 HPLC trace of pPF (F) + pPV (V).
A mixture of the pPF and pFP building blocks results in a
DCL with a distribution of both homo- and hetero-oligomers.
Using LC-MS to compare retention times from the pure libraries,
the dimers plus the homo-trimers and -tetramer were assigned
(Fig. 6a). However, the composition of the hetero-oligomers
cannot be identified using MS alone since pPF and pFP have
identical molecular weights. To overcome this problem, pPC was
used as a pPF replacement and MS marker. The mixed library of
pPC and pFP shows a product distribution of mainly self-sorting
Mixing building blocks from the same family
As indicated above, pPF and pPC behave similarly in single
building block libraries. When mixed in a DCL, these two building
blocks combine in a statistical manner. As can be seen in Fig. 3a,
three dimers, four trimers, and five tetramers are found. The
statistical and observed ratio of possible dimers, pPF2 : pPF-pPC :
780 | Org. Biomol. Chem., 2007, 5, 778–786
This journal is
The Royal Society of Chemistry 2007
©

View Article Online
DCLs of mixed L- and D-amino acid based building blocks
The DCL of the D-pPF building block alone behaves identically
to pPF, as expected. The mixed DCL of these two building blocks
shows a large bias towards trimers, while the tetramers are present
minimally (Fig. 7a). The homo-dimers and the mixed (meso)
dimer are also seen in this library. Further characterisation of
the diastereomeric trimers was frustrated by lack of resolution of
these oligomers in the chromatogram combined with the identical
masses of the building blocks. In order to overcome this problem,
pPC was introduced as the pseudo-enantiomer to D-pPF.7 Since
these two building blocks are from the same family, pPC and pPF
may be interchangeable.
Fig. 5 MS and MS-MS of constitutional isomers of pPF2–pPV2 (F2V2)
tetramers: (a) MS from LC peak at 18.5 min (inset: MS-MS); (b) MS from
LC peak at 19 min (inset: MS-MS).
Fig. 7 HPLC traces of DCLs (a) pPF (F) + D-pPF (U) and (b) pPC (C) +
D-pPF (U).
The DCL obtained from the mixture of D-pPF and pPC shows
the resolution of the trimers and confirms the presence of a
very small abundance of tetramers (Fig. 7b). The four possible
trimers are present in a statistical proportion. This implies that
the amplification of trimers in the racemic pPF library occurs
because the mixed tetramers are in effect suppressed by their lack
of stability relative to the trimers.
Similar experiments were performed by mixing D-pPV and pPV
(Fig. 8a). The chromatogram reveals the presence of both homo-
and hetero-dimers, -trimers and -tetramers. The mixed trimer (13.5
min) is more than three times as abundant as the homo-trimer (15
min), which indicates the preference for the mixed trimers. Once
again, in order to elucidate the structure of the mixed oligomers
the pPI building block was used as the pseudo-enantiomer to D-
pPV, and again the mixed trimers are formed at the expense of the
homo-trimers (Fig. 8b).
Fig. 6 HPLC traces of DCLs (a) pPF (F) + pFP (u), (b) pPC (C) + pFP
(u) and (c) pPV (V) + pFP (u).
homo-dimers, -trimers and -tetramers, with only trimers showing
any mixing (Fig. 6b). This distribution differs from that of the
mixed library of pPF and pFP. Therefore, when the order of the
amino acids is changed Phe and Cha are not interchangeable, as
was the case for the mixed library of pPF and pPC. This obser-
vation presumably results from inter-building block interactions
that are only accessible in the mixed-order macrocycles. Similarly,
the HPLC trace from the mixed library of pPV and pFP reveals
the presence of homo-dimers, -trimers and -tetramers, as well as
hetero-trimers and -tetramers, while the hetero-dimer is absent
(Fig. 6c).
The final investigation into the effects of chirality on the
diversity of a DCL was explored by mixing a D-building block with
one L-building block from a different family. A total of six DCLs
were examined, where the first three contained D-pPF plus one of
pPV, pPI or pFP, and the final three contained D-pPV plus one
of pPF, pPC or pFP. Generally, self-sorting is more pronounced
in these DCLs compared to those of the same chirality. For
example, the resulting chromatogram from mixing D-pPF and
pPV shows the presence of a much lower abundance of mixed
oligomers (Fig. 9) compared with the DCL from the pPF and pPV
mixture (Fig. 4). The resulting DCLs from the enantiomeric and
This journal is
The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 778–786 | 781
©

View Article Online
This study illustrates how relatively simple changes in the
composition of an established pseudo-dipeptide building block
can be used to produce subtle or major changes in the product
distributions of hydrazone DCLs. The observed variation in
macrocyclic product distributions at equilibrium reflects subtle
and as-yet unidentified differences in thermodynamic stability. In
principle this might be explored by molecular modelling, but the
energy differences observed appear to be too small to allow reliable
conclusions to be drawn. From a practical perspective, in searching
for amplification of new receptors, it will clearly be necessary to
explore a range of building block mixtures in order to explore as
many potential receptors as possible.
Experimental
General
Analytical HPLC was carried out on either an Agilent 1100 or
1050 instrument coupled to a UV analyser, set to 290 nm with a
reference wavelength of 550 nm, at 45 ◦C. The data was processed
using HP Chemstation software. All separations were performed
using H2O–acetonitrile gradients, which were adjusted according
to the polarity of the libraries. A 5 lL library solution was
injected and pumped through a symmetry C18 column (25.0 cm ×
4.6 mm, 5 lm) at a flow rate of 1.00 mL min−1. Preparative scale
isolation was performed using a Nucleodur C18 preparativecolumn
Fig. 8 HPLC traces of DCLs (a) pPV (V) + D-pPV (Y) and (b) pPI (I) +
D-pPV (Y).
˚
(25.0 cm × 2.1 cm, 100 A, 5 lm) with a Nucleodur C18 guard
˚
column (5.0 cm × 2.1 cm, 100 A, 5 lm). A 2.5 mL solution of
concentrated library was injected and run through the column at
a flow rate of 20.00 mL min−1. The desired fractions were collected
and the solvent was removed.
Fig. 9 HPLC trace of DCL of pPV (V) + D-pPF (U).
The LC-MS was performed using an Agilent LC-MSD-Trap-
XCT system. The LC is an Agilent 1100 series HPLC equipped
with an online degasser, binary pump, autosampler, heated column
compartment and diode array detector. The MS was performed
using an Agilent XCT ion trap MSD mass spectrometer. Mass
spectra (positive mode) wer◦e acquired in ultra scan mode using a
drying temperature of 350 C, a nebuliser pressure of 60.00 psi,
a drying gas flow of 11.00 L min−1, a capillary voltage of 4000 V
and capillary current of 39 nA. HRMS-ESI was performed using
a Waters LCT Premier instrument. Mass spectra (positive mode)
were acquired using a capillary voltage of 3000.0 V and a sample
cone of 50.0 V. The desolvation temperature was 200.0 ◦C with
a source temperature of 100.0 ◦C, a desolvation gas flow of
350.0 L h−1 and an ion energy of 30.0 V.
pseudo-enantiomeric building blocks are examples of how DCLs
operate to optimise the free energy of the entire system.
Conclusions
Working from the starting point of an established building block,
pPF, systematic alterations to its structure were made to produce
DCLs with a view to producing greater diversity. A change to
pPF such as substituting the Phe residue with Cha (pPC) does
little to affect the behaviour of the building block. Indeed, in our
experiments, the building blocks pPF and pPC produce similar
product distributions with each other as they do with themselves.
More dramatic changes in the behaviour of the DCLs are achieved
by making changes that are more significant to the building block
structure. For example, substitution of the Phe for Val (pPV)
changes the distribution of library members significantly. A single
building block library of pPV strongly favours the formation of
tetramer as opposed to the relatively even distribution of dimer,
trimer and tetramer seen with pPF. A mixed DCL of pPF and
pPV favours the formation of mixed species, and the resulting
constitutional isomers were identified by MS-MS.
The NMR spectroscopy was performed on a Bruker DRX
500 MHz spectrometer at 300 K unless otherwise stated. Chemical
shifts are quoted in parts per million with reference to residual
protons from the deuterated solvent for the 1H spectra. The
coupling constants are reported in hertz (Hz).
Materials
All chemicals were purchased from Aldrich, Fluka, Sigma,
Lancaster or Avocado in reagent-grade quality or better and
used without further purification. All solvents where distilled
prior to use and dry solvents where freshly distilled from CaH2
under argon, with the exception of DMSO (Lancaster), which was
used without further purification. Thin layer chromatography was
The effects on library diversity when the chirality of the amino
acid residues is mixed were also investigated. Both L- and D-pPF
have a broadly even distribution of dimers, trimers and tetramers
when equilibrated on their own although, when mixed, the two
possible hetero-trimers are strongly favoured.
782 | Org. Biomol. Chem., 2007, 5, 778–786
This journal is
The Royal Society of Chemistry 2007
©

View Article Online
carried out on silica gel plates loaded with Merck silica gel 60
F254. Column chromatography was performed on Merck 60 silica
gel (230–400 mesh). HPLC-grade CHCl3 (Aldrich) and HPLC-
grade acetonitrile (Romil) were filtered with a 0.45 lm Millipore
filter and used without further purification. Ultrapure H2O was
obtained from a Millipore H2O purification system.
Synthesis of pPL
CBZ-L-Pro-L-Leu methyl ester (Reaction i). CBZ-L-proline
(0.25 g, 1.0 mmol) and L-leucine methyl ester hydrochlo-
ride (0.18 g, 1.0 mmol) were coupled through the stan-
dard amide coupling process. Silica gel column chromato-
graphy (EtOAc–Hex = 7 : 3, Rf = 0.55) yielded the product as
1
a colourless oil (0.34 g, 90%). H NMR (500 MHz, CDCl3) d
General DCL setup and screening. The single-component
libraries were prepared by making a 2 mM building block solution
in CHCl3–DMSO (97 : 3 v/v) containing 100 mM of TFA. Mixed
libraries were prepared by mixing two of the above solutions in
a 1 : 1 ratio. The resulting solutions were allowed to equilibrate
for at least 4 days. The library members were then separated and
identified using HPLC and LC-MS. Four gradient methods were
used depending on the polarity of the library members. In all four
methods, eluent A is acetonitrile and eluent B is H2O. Method 1
(eluent A : B 35 : 65 to 80 : 20 over 45 min and 80 : 20 to 90 : 10
over 10 min) was used for all libraries containing pPC. Method
2 (eluent A : B 30 : 70 to 80 : 20 over 45 min) was used for all
libraries containing pPF and pFP, except those containing pPC.
Method 3 (eluent A : B 20 : 80 to 40 : 60 over 30 min, 40 : 60
to 90 : 10 over 2 min and 90 : 10 for 6 min) was used for all the
libraries containing pPV, pVP, pPI, pPL, except those containing
pPC, pPF and pFP. Method 4 (eluent A : B 10 : 90 to 20 : 80 over
10 min, 20 : 80 to 25 : 75 over 15 min, 25 : 75 to 40 : 60 over 10 min,
40 : 60 to 90 : 10 over 2 min and 90 : 10 for 6 min) was used for the
library containing pPA.
7.38–7.24 (5H, br, Ar-H), 7.06 and 6.25 (1H, br, Leu-NH), 5.19–
5.09 (2H, m, OCH2Ph), 4.52 (1H, br, Leu-a), 4.34 (1H, br, Pro-a),
3.69–3.63 (3H, br, OCH3), 3.53–3.39 (2H, br, Pro-b), 2.39–1.83
(4H, br, Pro-c(2H) and Pro-d(2H)), 1.65–1.35 (3H, br, Leu-b(2H)
ꢀ
and Leu-c(1H)), 0.88–0.85 (6H, br, Leu-d(3H) and Leu-d(3H));
13C NMR (125 MHz, CDCl3) d 173.1, 171.3, 156.1, 136.4, 128.5,
128.0, 127.8, 67.3, 60.3, 52.1, 51.0, 46.9, 41.3, 27.9, 24.9, 24.6, 22.7,
21.9; HRMS (ESI) [M + H]+ C20H29N2O5 requires 377.2076, found
377.2089.
L-Pro-L-Leu methyl ester (Reaction ii). Hydrogenation was
carried out with CBZ-L-proline-L-leucine methyl ester (0.34 g,
0.90 mmol) following the general procedure above. Product was
1
yielded as a colourless oil (0.21 g, 96%). H NMR (500 MHz,
CDCl3) d 8.00 (1H, d, J = 8.5 Hz, Leu-NH), 4.46 (1H, m, Leu-a),
3.96 (1H, br, Pro-NH), 3.83 (1H, dd, Pro-a), 3.63 (3H, s, OCH3),
3.01 (1H, m, Pro-d), 2.90 (1H, m, Pro-d), 2.11 (1H, m, Pro-b),
1.84 (1H, m, Pro-b), 1.67 (2H, m, Pro-c), 1.54 (3H, m, Leu-b(2H)
and Leu-c(1H)), 0.84 (6H, t, J = 6.5 Hz, Leu-d(3H) and Leu-
ꢀ
d(3H)); 13C NMR (125 MHz, CDCl3) d 173.9, 173.2, 60.2, 52.0,
50.3, 47.0, 41.1, 30.7, 25.7, 24.8, 22.7, 21.7; HRMS (ESI) [M +
H]+ C12H23N2O3 requires 243.1709, found 243.1719.
General procedure for amide coupling (Reactions i and iii).
Amide coupling reactions were performed using EDC and
DMAP. The acid and amine were dissolved in 45 mL of dry
DCM, under nitrogen. Dry Et3N (2 equivalents) was added and
the resulting solution was cooled to 0 ◦C on an ice bath for 30 min,
after which EDC (1.2 equivalents) and DMAP (3 equivalents
N-(4-(Dimethoxymethyl)benzoyl)-L-Pro-L-Leu methyl ester (Re-
action iii). L-Proline-L-leucine methyl ester (0.21 g, 0.87 mmol)
and pDMA (0.17 g, 0.87 mmol) were coupled through the general
amide coupling procedure. Silica gel chromatography (EtOAc,
Rf = 0.45) yielded the product as colourless oil (0.30 g, 82%). 1H
NMR (500 MHz, CDCl3) d 7.46 (4H, br, Ar-H), 7.26 (1H, d, J =
7.5 Hz, Leu–NH), 5.37 (1H, s, CH(OCH3)2), 4.76 (1H, m, Pro-a),
4.53 (1H, m, Leu-a), 3.69 (3H, s, OCH3), 3.29 (6H, s, CH(OCH3)2),
3.48 (1H, m, Pro-d), 3.40 (1H, m, Pro-d), 2.41 (1H, m, Pro-b), 2.01
(2H, m, Pro-b), 1.79 (2H, m, Pro-c), 1.63–1.53 (3H, m, Leu-b(2H)
◦
with respect to acid) were added. The reaction was kept at 0 C
for 1 hour, allowed to warm to room temperature and stirred
(under nitrogen) overnight. The work-up involved the addition
of DCM (3-fold dilution) and subsequent washing of the organic
solution with two portions of H2O. The organic layer was dried
over MgSO4, filtered and the solvent was removed in vacuo. The
crude product was purified by silica gel column chromatography.
ꢀ
and Leu-c(1H)), 0.87 (6H, t, J = 6 Hz, Leu-d(3H) and Leu-d(3H));
13C NMR (125 MHz, CDCl3) d 173.1, 170.8, 170.7, 140.2, 136.3,
126.9, 126.8, 102.5, 59.6, 52.7, 52.1, 51.1, 50.3, 41.2, 27.0, 25.4,
24.9, 22.7, 21.8; HRMS (ESI) [M + Na]+ C22H32N2O6Na requires
443.2158, found 443.2176.
General procedure for hydrogenation (Reaction ii). Hydrogena-
tion was carried out to deprotect the CBZ group. CBZ-protected
compounds were dissolved in 30 mL of 4 : 1 EtOAc–MeOH, 5%
Pd/C was added and H2 was bubbled through the reaction until
completion (approximately four hours; monitored by TLC with
EtOAc as eluent). The resulting suspension was filtered through a
pad of Celite and the solvent was removed in vacuo to afford the
product, which was used without further purification.
pPL: N-(4-(Dimethoxymethyl)benzoyl)-L-Pro-L-Leu carboxylic
acid hydrazide (Reaction iv). The hydrazinolysis was carried out
with N-(4-(dimethoxymethyl)benzoyl)-L-proline-L-leucine methyl
ester (0.30 g, 0.71 mmol) following the general procedure above,
1
yielding the product as a white solid (0.26 g, 87%). H NMR
General procedure for hydrazinolysis (Reaction iv). The methyl
ester was dissolved in 15 mL of MeOH, and 1 mL of hydrazine
monohydrate was added. The reaction was allowed to stir at room
temperature until completion (monitored by TLC with EtOAc as
eluent). The solvent was removed at room temperature in vacuo,
100 mL of DCM was added to the residue, and the organic
phase was washed twice with H2O, dried over MgSO4, filtered
and evaporated. Washing the solid residue with Et2O yielded the
product as a white solid.
(500 MHz, CDCl3) d 7.77 (1H, br, NHNH2), 7.49 (4H, br, Ar-H),
7.17 (1H, d, J = 7.5 Hz, Leu-NH), 5.39 (1H, s, CH(OCH3)2), 4.72
(1H, m, Pro-a), 4.41–4.36 (1H, m, Leu-a), 3.86 (2H, br, NHNH2),
3.54 (1H, m, Pro-d), 3.47 (1H, m, Pro-d), 3.31 (6H, s, CH(OCH3)2)
2.34 (1H, m, Pro-b), 2.10 (1H, m, Pro-b), 2.02 (1H, m, Pro-c),
1.84 (1H, m, Pro-c), 1.73 (1H, m, Leu-b), 1.63–1.53 (2H, m, Leu-
b(1H) and Leu-c(1H)), 0.87 (6H, m, Leu-d(3H) and Leu-d(3H));
13C NMR (125 MHz, CDCl3) d 172.4, 171.4, 171.1, 140.6, 135.8,
127.0, 126.9, 102.4, 60.1, 52.8, 50.8, 50.5, 40.2, 27.5, 25.5, 24.9,
ꢀ
This journal is
The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 778–786 | 783
©

View Article Online
22.9, 21.8; HRMS (ESI) [M + H]+ C21H33N4O5 requires 421.2451,
found 421.2445.
25.5, 19.3, 17.7; HRMS (ESI) [M + Na]+ C20H30N4O5Na requires
429.2114, found 429.2117.
Synthesis of pPI
Synthesis of pPV or D-pPV (with L-amino acids as an example)
CBZ-L-Pro-L-Ile methyl ester (Reaction i). CBZ-L-proline
(0.25 g, 1.0 mmol) and L-isoleucine methyl ester hydrochloride
(0.18 g, 1.0 mmol) were coupled through the standard amide
coupling process. Silica gel column chromatography (EtOAc–
Hex = 7 : 3, Rf = 0.55) yielded the product as colourless oil
CBZ-L-Pro-L-Val methyl ester (Reaction i). CBZ-L-proline
(0.25 g, 1.0 mmol) and L-valine methyl ester hydrochloride (0.17 g,
1.0 mmol) were coupled through the standard amide coupling
process. Silica gel column chromatography (EtOAc–Hex = 7 : 3,
1
Rf = 0.7) yielded the product as colourless oil (0.33 g, 91%). H
1
(0.34 g, 90%). H NMR (500 MHz, CDCl3) d 7.36–7.25 (5H,
NMR (500 MHz, CDCl3) d 7.32–7.25 (5H, br, Ar-H), 7.18 and
6.38 (1H, br, Val-NH), 5.19–5.07 (2H, m, OCH2Ph), 4.44 (1H,
br, Val-a), 4.35 (1H, br, Pro-a), 3.69–3.63 (3H, br, OCH3), 3.57–
3.39 (2H, br, Pro-d), 2.39–1.83 (4H, br, Pro-b and Pro-c), 2.12 (1H,
br, Ar-H), 7.19 and 6.39 (1H, br, Ile-NH), 5.19–5.08 (2H, m,
OCH2Ph), 4.49 (1H, br, Ile-a), 4.36 (1H, br, Pro-a), 3.72–3.64
(3H, br, OCH3), 3.58–3.39 (2H, br, Pro-d), 2.36–1.80 (5H, br, Pro-
b(2H), Pro-c(2H) and Ile-b), 1.34 (1H, br, Ile-c), 1.09 (1H, br, Ile-c),
0.87–0.77 (6H, br, Ile-bꢀ(3H) and Ile-d(3H)); 13C NMR (125 MHz,
CDCl3) d 173.1, 171.3, 156.1, 136.4, 128.5, 128.0, 127.8, 67.3, 60.3,
52.1, 51.0, 46.9, 41.3, 27.9, 24.9, 24.6, 22.7, 21.9; HRMS (ESI) [M
+ H]+ C20H29N2O5 requires 377.2076, found 377.2086.
ꢀ
br, Val-b), 0.85–0.79 (6H, br, Val-c(3H) and Val-c(3H)); 13C NMR
(125 MHz, CDCl3) d 172.0, 171.4, 156.0, 136.4, 128.4, 127.9, 127.8,
67.3, 60.2, 57.3, 51.9, 46.9, 31.1, 27.9, 24.5, 18.9, 17.6; HRMS (ESI)
[M + H]+ C19H27N2O5 requires 363.1920, found 363.1910.
L-Pro-L-Val methyl ester (Reaction ii). Hydrogenation was
carried out with CBZ-L-proline-L-valine methyl ester (0.33 g,
0.91 mmol) following the general procedure above. Product was
L-Pro-L-Ile methyl ester (Reaction ii). Hydrogenation was
carried out with CBZ-L-proline-L-isoleucine methyl ester (0.34 g,
0.90 mmol) following the general procedure above, yielding the
1
yielded as colourless oil (0.21 g, 100%). H NMR (500 MHz,
1
product as a colourless oil (0.22 g, 100%). H NMR (500 MHz,
CDCl3) d 8.38 (1H, d, J = 8.0 Hz, Val-NH), 4.53 (1H, br, Pro-a),
4.28 (1H, m, Val-a), 3.63 (3H, s, OCH3), 3.30–3.23 (2H, m, Pro-d),
2.40 (1H, m, Pro-b), 2.15 (1H, m, Val-b), 1.97 (1H, m, Pro-b),
CDCl3) d 8.11 (1H, d, J = 9.0 Hz, Ile-NH), 4.46 (1H, m, Ile-a),
3.85 (1H, br, Pro-a), 3.72 (1H, br, Pro-NH), 3.65 (3H, s, OCH3),
3.04 (1H, m, Pro-d), 2.93 (1H, m, Pro-d), 2.15 (1H, m, Pro-b), 1.88
(2H, m, Pro-b(1H) and Ile-b), 1.70 (2H, m, Pro-c), 1.36 (1H, m,
Ile-c), 1.12 (1H, m, Ile-c), 0.90 (6H, m, Ile-bꢀ(3H) and Ile-d(3H));
13C NMR (125 MHz, CDCl3) d 174.1, 172.3, 60.3, 56.1, 51.8,
47.1, 37.6, 30.8, 25.8, 25.0, 15.5, 11.4; HRMS (ESI) [M + H]+
C12H23N2O3 requires 243.1709, found 243.1720.
ꢀ
1.95–1.80 (2H, m, Pro-c), 0.90 (6H, m, Val-c(3H) and Val-c(3H));
13C NMR (125 MHz, CDCl3) d 171.7, 170.7, 59.7, 58.1, 51.9, 46.8,
30.7, 30.3, 24.7, 19.0, 18.0; HRMS (ESI) [M + H]+ C11H21N2O3
requires 229.1552, found 229.1546.
N-(4-(Dimethoxymethyl)benzoyl)-L-Pro-L-Val methyl ester (Re-
action iii). L-Proline-L-valine methyl ester (0.21 g, 0.92 mmol)
and pDMA (0.18 g, 0.92 mmol) were coupled through the general
amide coupling procedure. Silica gel chromatography (EtOAc,
Rf = 0.40) yielded the product as colourless oil (0.27 g, 72%).
1H NMR (500 MHz, CDCl3) d 7.46 (4H, br, Ar-H), 7.37 (1H,
d, J = 8.5 Hz, Val-NH), 5.35 (1H, s, CH(OCH3)2), 4.77 (1H, m,
Pro-a), 4.46 (1H, m, Val-a), 3.69 (3H, s, OCH3), 3.49 (1H, m,
Pro-d), 3.39 (1H, m, Pro-d), 3.28 (6H, s, CH(OCH3)2), 2.41 (1H,
m, Pro-b), 2.14 (1H, m, Val-b), 2.08–2.00 (2H, m, Pro-b(1H) and
Pro-c(1H)), 1.79 (1H, m, Pro-c(1H)), 0.87 (6H, m, Val-c(3H) and
N-(4-(Dimethoxymethyl)benzoyl)-L-Pro-L-Ile methyl ester (Re-
action iii). L-Proline-L-isoleucine methyl ester (0.22 g,
0.91 mmol) and pDMA (0.18 g, 0.92 mmol) were coupled through
the general amide coupling procedure. Silica gel chromatography
(EtOAc, Rf = 0.50) yielded the product as colourless oil (0.26 g,
1
68%). H NMR (500 MHz, CDCl3) d 7.46 (4H, br, Ar-H), 7.37
(1H, d, J = 7.5 Hz, Val-NH), 5.36 (1H, s, CH(OCH3)2), 4.77 (1H,
m, Pro-a), 4.51 (1H, m, Ile-a), 3.69 (3H, s, OCH3), 3.49 (1H, m,
Pro-d), 3.39 (1H, m, Pro-d), 3.29 (6H, s, CH(OCH3)2), 2.40 (1H, m,
Pro-b), 2.02 (2H, m, Pro-b(1H) and Pro-c(1H)), 1.90 (1H, m, Ile-
b), 1.79 (1H, m, Pro-c(1H)), 1.39 (1H, m, Ile-c), 1.16 (1H, m, Ile-c),
0.88–0.82 (6H, m, Ile-bꢀ(3H) and Ile-d(3H)); 13C NMR (125 MHz,
CDCl3) d 172.1, 170.8, 170.7, 140.2, 136.3, 126.8, 102.5, 59.7, 56.8,
52.7, 51.9, 50.3, 37.6, 27.1, 25.3, 25.0, 15.5, 11.5; HRMS (ESI) [M
+ Na]+ C22H32N2O6Na requires 443.2158, found 443.2178.
ꢀ
Val-c(3H)); 13C NMR (125 MHz, CDCl3) d 172.1, 170.8, 170.7,
140.2, 136.3, 126.8, 126.7, 102.4, 59.7, 57.4, 52.7, 52.0, 50.3, 31.0,
27.0, 25.3, 19.0, 17.6; HRMS (ESI) [M + H]+ C21H31N2O6 requires
407.2182, found 407.2188.
pPV: N-(4-(Dimethoxymethyl)benzoyl)-L-Pro-L-Val carboxylic
acid hydrazide (Reaction iv). The hydrazinolysis was carried out
with N-(4-(dimethoxymethyl)benzoyl)-L-proline-L-valine methyl
ester (0.27 g, 0.67 mmol) following the general procedure above,
pPI: N-(4-(Dimethoxymethyl)benzoyl)-L-Pro-L-Ile carboxylic
acid hydrazide (Reaction iv). The hydrazinolysis was carried
out with N-(4-(dimethoxymethyl)benzoyl)-L-proline-L-isoleucine
methyl ester (0.26 g, 0.62 mmol) following the general procedure
above, yielding the product as a white solid (0.19 g, 73%). 1H NMR
(500 MHz, CDCl3) d 7.55 (1H, br, NHNH2), 7.49 (4H, m, Ar-H),
7.26 (1H, d, J = 8.0 Hz, Ile-NH), 5.39 (1H, s, CH(OCH3)2), 4.75
(1H, m, Pro-a), 4.27 (1H, m, Ile-a), 3.87 (2H, br, NHNH2), 3.54
(1H, m, Pro-d), 3.46 (1H, m, Pro-d), 3.32 (6H, s, CH(OCH3)2),
2.38 (1H, m, Pro-b), 2.12–1.96 (3H, m, Pro-b(1H), Pro-c(1H) and
Ile-b), 1.84 (1H, m, Pro-b), 1.44 (1H, m, Ile-c), 1.10 (1H, m, Ile-c),
0.89 (3H, d, J = 7.0 Hz, Ile-bꢀ), 0.84 (3H, t, J = 7.5 Hz, Ile-d);
1
yielding the product as a white solid (0.20 g, 74%). H NMR
(500 MHz, CDCl3) d 7.87 (1H, br, NHNH2), 7.49 (4H, br, Ar-H),
7.34 (1H, d, J = 8.5 Hz, Val-NH), 5.38 (1H, s, CH(OCH3)2), 4.75
(1H, m, Pro-a), 4.27 (1H, m, Val-a), 3.93–3.87 (2H, br, NHNH2),
3.56 (1H, m, Pro-d), 3.46 (1H, m, Pro-d), 3.31 (6H, s, CH(OCH3)2)
2.34 (1H, m, Pro-b), 2.24 (1H, m, Val-b), 2.12–1.96 (2H, m, Pro-
b(1H) and Pro-c(1H)), 1.83 (1H, m, Pro-c), 0.90 (6H, m, Val-c(3H)
ꢀ
and Val-c(3H)); 13C NMR (125 MHz, CDCl3) d 171.6, 171.4,
171.0, 140.5, 136.0, 127.0, 102.5, 60.1, 57.7, 52.8, 50.5, 30.1, 27.6,
784 | Org. Biomol. Chem., 2007, 5, 778–786
This journal is
The Royal Society of Chemistry 2007
©

View Article Online
13C NMR (125 MHz, CDCl3) d 171.6, 171.3, 171.2, 140.5, 135.9,
127.0, 126.9, 102.5, 60.2, 57.0, 52.8, 50.5, 36.1, 27.5, 25.5, 24.5,
15.7, 11.3; HRMS (ESI) [M + Na]+ C21H32N4O5Na requires
443.2270, found 443.2285.
Synthesis of pVP
CBZ-L-Val-L-Pro methyl ester (Reaction i). CBZ-L-valine
(0.25 g, 1.0 mmol) and L-proline methyl ester hydrochloride (0.17 g,
1.0 mmol) were coupled through the standard amide coupling
process. Silica gel column chromatography (EtOAc–Hex = 7 : 3,
Rf = 0.55) yielded the product as colourless oil (0.35 g, 97%). 1H
NMR (500 MHz, CDCl3) d 7.33–7.25 (5H, m, Ar-H), 5.47 (1H,
d, J = 9.0 Hz, Val-NH), 5.09–5.01 (2H, m, OCH2Ph), 4.49 (1H,
m, Pro-a), 4.30 (1H, m, Val-a), 3.76 (1H, m, Pro-d), 3.67 (3H, s,
OCH3), 3.62 (1H, m, Pro-d), 2.17 (1H, m, Pro-b), 2.02 (2H, m,
Pro-c(1H) and Val-b), 1.94 (2H, m, Pro-b(1H) and Pro-c(1H)),
1.01 (3H, d, J = 7.0 Hz, Val-c), 0.92 (3H, d, J = 7.0 Hz, Val-
Synthesis of pPA
CBZ-L-Pro-L-Ala methyl ester (Reaction i). CBZ-L-proline
(0.25 g, 1.0 mmol) and L-alanine methyl ester hydrochloride
(0.14 g, 1.0 mmol) were coupled through the standard amide
coupling process. Silica gel column chromatography (EtOAc–
Hex = 7 : 3, Rf = 0.5) yielded the product as colourless oil (0.31 g,
1
93%). H NMR (500 MHz, CDCl3) d 7.36–7.25 (5H, br, Ar-H),
ꢀ
c); 13C NMR (125 MHz, CDCl3) d 172.3, 170.7, 156.4, 136.4,
7.10 and 6.42 (1H, br, Ala-NH), 5.16–5.08 (2H, m, OCH2Ph), 4.47
(1H, br, Ala-a), 4.29 (1H, br, Pro-a), 3.72–3.60 (3H, br, OCH3),
3.56–3.38 (2H, br, Pro-d), 2.30–1.80 (4H, br, Pro-b(2H) and Pro-
c(2H)), 1.36–1.20 (3H, br, Ala-b); 13C NMR (125 MHz, CDCl3)
d 173.0, 171.3, 155.9, 136.4, 128.4, 127.9, 127.8, 67.2, 60.3, 52.2,
48.0, 46.9, 28.3, 24.4, 18.0; HRMS (ESI) [M + H]+ C17H23N2O5
requires 335.1607, found 335.1597.
128.4, 127.9, 127.8, 66.7, 58.7, 57.4, 52.0, 47.1, 31.3, 28.9, 24.9,
19.1, 17.4; HRMS (ESI) [M + H]+ C19H27N2O5 requires 363.1920,
found 363.1938.
L-Val-L-Pro methyl ester (Reaction ii). Hydrogenation was
carried out with CBZ-L-valine-L-proline methyl ester (0.35 g,
0.97 mmol) following the general procedure above, yielding the
1
L-Pro-L-Ala methyl ester (Reaction ii). Hydrogenation was
carried out with CBZ-L-proline-L-alanine methyl ester (0.31 g,
0.93 mmol) following the general the procedure above, yielding
the product as a colourless oil (0.18 g, 97%). 1H NMR (500 MHz,
CDCl3) d 8.60 (1H, d, J = 7.0 Hz, Ala-NH), 4.50 (1H, br, Pro-a),
4.42 (1H, m, Ala-a), 3.67 (3H, s, OCH3), 3.32 (2H, m, Pro-d), 2.43
(1H, m, Pro-b), 2.04 (1H, m, Pro-b), 1.93 (2H, m, Pro-c), 1.42
(3H, d, J = 7.0 Hz, Ala-b); 13C NMR (125 MHz, CDCl3) d 172.8,
170.2, 59.8, 52.3, 48.6, 46.8, 30.6, 24.7, 17.2; HRMS (ESI) [M +
H]+ C9H17N2O3 requires 201.1239, found 201.1240.
product as a colourless oil (0.22 g, 100%). H NMR (500 MHz,
CDCl3) d 4.59 (1H, br, Pro-a), 4.20 (1H, d, J = 6.0 Hz, Val-a),
3.95 (1H, m, Pro-d), 3.66 (3H, s, OCH3), 3.51 (1H, m, Pro-d), 2.32
(1H, m, Val-b), 2.25 (1H, m, Pro-b), 2.01 (1H, m, Pro-c), 1.94
(1H, m, Pro-c), 1.88 (1H, m, Pro-b), 1.11 (6H, m, Val-c(3H) and
ꢀ
Val-c(3H)); 13C NMR (125 MHz, CDCl3) d 172.3, 167.9, 59.3,
57.0, 52.0, 47.8, 30.1, 29.0, 25.0, 18.6, 17.8; HRMS (ESI) [M +
H]+ C11H21N2O3 requires 229.1552, found 229.1541.
N-(4-(Dimethoxymethyl)benzoyl)-L-Val-L-Pro methyl ester (Re-
action iii). L-Valine-L-proline methyl ester (0.22 g, 0.96 mmol)
and pDMA (0.19 g, 0.97 mmol) were coupled through the general
amide coupling procedure. Silica gel column chromatography
(EtOAc, Rf = 0.50) yielded the product as colourless oil (0.21 g,
N-(4-(Dimethoxymethyl)benzoyl)-L-Pro-L-Ala methyl ester (Re-
action iii). L-proline-L-alanine methyl ester (0.18 g, 0.91 mmol)
and pDMA (0.18 g, 0.91 mmol) were coupled through the general
amide coupling procedure. Silica gel chromatography (EtOAc,
1
54%). H NMR (500 MHz, CDCl3) d 7.73 (2H, d, J = 8.0 Hz,
1
Rf = 0.4) yielded the product as colourless oil (0.27 g, 79%). H
Ar-H), 7.44 (2H, d, J = 8.0 Hz, Ar-H), 7.03 (1H, d, J = 8.5 Hz,
Val-NH), 5.36 (1H, s, CH(OCH3)2), 4.79 (1H, m, Val-a), 4.43
(1H, m, Pro-a), 3.87 (1H, m, Pro-d), 3.67 (3H, s, OCH3), 3.66
(1H, m, Pro-d), 3.24 (6H, s, CH(OCH3)2), 2.19 (2H, m, Pro-b(1H)
and Val-b), 2.04–1.88 (3H, m, Pro-b(1H) and Pro-c(2H)), 1.04
NMR (500 MHz, CDCl3) d 7.46–7.38 (4H, m, Ar-H), 7.32 (1H,
d, J = 7.0 Hz, Ala-NH), 5.31 (1H, s, CH(OCH3)2), 4.65 (1H, m,
Pro-a), 4.42 (1H, m, Ala-a), 3.63 (3H, s, OCH3), 3.49 (1H, m,
Pro-d), 3.39 (1H, m, Pro-d), 3.22 (6H, s, CH(OCH3)2), 2.24 (1H,
m, Pro-b), 2.02 (1H, m, Pro-b), 1.94 (1H, m, Pro-c), 1.72 (1H, m,
Pro-c), 1.30 (3H, d, J = 7.0 Hz, Ala-b); 13C NMR (125 MHz,
CDCl3) d 173.0, 170.8, 170.2, 140.1, 136.1, 126.9, 126.5, 102.3,
59.7, 52.5, 52.1, 50.2, 48.1, 27.8, 25.1, 17.6; HRMS (ESI) [M +
Na]+ C19H26N2O6Na requires 401.1689, found 401.1705.
(3H, d, J = 6.5 Hz, Val-c), 0.96 (3H, d, J = 6.5 Hz, Val-c); 13C
ꢀ
NMR (125 MHz, CDCl3) d 172.2, 170.8, 166.9, 141.5, 134.0, 126.9,
126.8, 102.2, 58.8, 55.8, 52.4, 52.0, 47.2, 31.5, 28.9, 24.9, 19.2, 17.7;
HRMS (ESI) [M + Na]+ C21H30N2O6Na requires 429.2002, found
429.1996.
pPA: N-(4-(Dimethoxymethyl)benzoyl)-L-Pro-L-Ala carboxylic
acid hydrazide (Reaction iv). The hydrazinolysis was carried out
with N-(4-(dimethoxymethyl)benzoyl)-L-proline-L-alanine methyl
ester (0.27 g, 0.71 mmol) following the general procedure above,
pVP: N-(4-(Dimethoxymethyl)benzoyl)-L-Val-L-Pro carboxylic
acid hydrazide (Reaction iv). The hydrazinolysis was carried out
with N-(4-(dimethoxymethyl)benzoyl)-L-valine-L-proline methyl
ester (0.21 g, 0.52 mmol) following the general procedure above,
1
1
yielding the product as a white solid (0.17 g, 63%). H NMR
yielding the product as a white solid (0.16 g, 76%). H NMR
(500 MHz, CDCl3) d 7.86 (1H, br, NHNH2), 7.51 (4H, m, Ar-H),
7.16 (1H, d, J = 7.0 Hz, Ala-NH), 5.38 (1H, s, CH(OCH3)2), 4.65
(1H, m, Pro-a), 4.42 (1H, m, Ala-a), 3.86 (2H, br, NHNH2), 3.58
(1H, m, Pro-d), 3.49 (1H, m, Pro-d), 3.31 (6H, s, CH(OCH3)2), 2.28
(1H, m, Pro-b), 2.14 (1H, m, Pro-b), 2.01 (1H, m, Pro-c), 1.83 (1H,
m, Pro-c), 1.38 (3H, d, J = 7.0 Hz, Ala-b); 13C NMR (125 MHz,
CDCl3) d 172.5, 171.3, 171.0, 140.6, 135.8, 127.2, 126.9, 102.4,
60.4, 52.7, 50.6, 48.0, 28.0, 25.5, 17.5; HRMS (ESI) [M + H]+
C18H27N4O5 requires 379.1981, found 379.1980.
(500 MHz, CDCl3) d 7.88 (1H, br, NHNH2), 7.78 (2H, d, J =
8.0 Hz, Ar-H), 7.50 (2H, d, J = 8.0 Hz, Ar-H), 6.81 (1H, d,
J = 9.0 Hz, Val-NH), 5.42 (1H, s, CH(OCH3)2), 4.81 (1H, m,
Val-a), 4.48 (1H, m, Pro-a), 3.87 (1H, m, Pro-d), 3.66 (1H, m,
Pro-d), 3.30 (6H, s, CH(OCH3)2), 2.34 (1H, m, Pro-b), 2.15 (2H,
m, Val-b and Pro-c(1H)), 2.01(1H, m, Pro-c), 1.94 (1H, m, Pro-b),
ꢀ
1.03 (3H, d, J = 6.5 Hz, Val-c), 0.98 (3H, d, J = 6.5 Hz, Val-c);
13C NMR (125 MHz, CDCl3) d 172.1, 171.9, 167.0, 141.8, 134.0,
127.1, 127.0, 102.2, 58.4, 55.9, 52.6, 47.8, 31.8, 27.2, 25.2, 19.4,
This journal is
The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 778–786 | 785
©

View Article Online
17.9; HRMS (ESI) [M + Na]+ C20H30N4O5Na requires 429.2114,
found 429.2134.
pFP: N-(4-(Dimethoxymethyl)benzoyl)-L-Phe-L-Pro carboxylic
acid hydrazide (Reaction iv). The hydrazinolysis was carried out
with N-(4-(dimethoxymethyl)benzoyl)-L-phenylalanine-L-proline
methyl ester (0.36 g, 0.79 mmol) following the general procedure
above, yielding the product as a white solid (0.31 g, 86%). 1H NMR
(500 MHz, CDCl3) d 7.75 (2H, d, J = 8.0 Hz, Ar-H), 7.69 (1H, br,
NHNH2), 7.50 (2H, d, J = 8.0 Hz, Ar-H), 7.32–7.16 (5H, m, Phe-
Ar), 6.92 (1H, d, J = 8.0 Hz, Phe-NH), 5.41 (1H, s, CH(OCH3)2),
5.16 (1H, m, Phe-a), 4.48 (1H, m, Pro-a), 3.64 (1H, m, Pro-d), 3.30
(6H, s, CH(OCH3)2), 3.15 (2H, m, Phe-b), 3.06 (1H, m, Pro-d), 2.30
(1H, m, Pro-b), 1.88 (3H, m, Pro-b(1H) and Pro-c(2H)); 13C NMR
(125 MHz, CDCl3) d 171.5, 171.1, 166.4, 141.9, 135.6, 133.7, 129.5,
128.7, 127.4, 127.1, 127.0, 102.2, 58.6, 52.6, 52.3, 47.4, 39.1, 27.0,
25.0; HRMS (ESI) [M + Na]+ C24H30N4O5Na requires 477.2114,
found 477.2116.
Synthesis of pFP
CBZ-L-Phe-L-Pro methyl ester (Reaction i). CBZ-L-
phenylalanine (0.30 g, 1.0 mmol) and L-proline methyl ester
hydrochloride (0.17 g, 1.0 mmol) were coupled through
the standard amide coupling process. Silica gel column
chromatography (EtOAc–Hex = 7 : 3, Rf = 0.55) yielded the
product as colourless oil (0.38 g, 93%). 1H NMR (500 MHz,
CDCl3) d 7.33–7.16 (10H, m, Ar-H and Phe-Ar), 5.64 (1H, d,
J = 8.5 Hz, Phe-NH), 5.08–4.96 (2H, m, OCH2Ph), 4.69 (1H,
m, Phe-a), 4.47 (1H, m, Pro-a), 3.70 (3H, s, OCH3), 3.59 (1H, m,
Pro-d), 3.17 (1H, m, Pro-d), 3.09 (1H, m, Phe-b), 2.91 (1H, m,
Phe-b), 2.16–1.60 (4H, m, Pro-b(2H) and Pro-c(2H)); 13C NMR
(125 MHz, CDCl3) d 172.1, 171.7, 170.2, 136.3, 136.1, 129.6,
128.3, 128.2, 127.9, 127.8, 126.7, 66.6, 58.8, 53.7, 52.0, 46.7, 38.9,
28.9, 24.7; HRMS (ESI) [M + H]+ C23H27N2O5 requires 411.1920,
found 411.1933.
Acknowledgements
We thank the Engineering and Physical Sciences Research Council
for studentships to J. L. and K. R. W., the National Sciences
and Engineering Research Council of Canada for a post-doctoral
fellowship for C. R. B., Dr Ana Belenguer for managing the HPLC
facility and Dr Sijbren Otto for fruitful discussions.
L-Phe-L-Pro methyl ester (Reaction ii). Hydrogenation was
carried out with CBZ-L-phenylalanine-L-proline methyl ester
(0.38 g, 0.92 mmol) following the general procedure above, yielding
the product as a colourless oil (0.25 g, 98%). 1H NMR (500 MHz,
CDCl3) d 7.29–7.18 (5H, m, Phe-Ar), 4.50 (2H, br, Pro-a and
Phe-a), 3.65 (3H, s, OCH3), 3.60 (1H, m, Pro-d), 3.46 (1H, m,
Phe-b), 3.21 (1H, m, Phe-b), 2.54 (1H, m, Pro-d), 2.10 (1H, m,
Pro-b), 1.79 (2H, m, Pro-c(1H) and Pro-b(1H)), 1.66 (1H, m, Pro-
c); 13C NMR (125 MHz, CDCl3) d 172.0, 167.9, 134.4, 130.3, 128.5,
127.3, 59.4, 53.4, 52.1, 47.1, 37.3, 28.9, 24.8; HRMS (ESI) [M +
H]+ C15H21N2O3 requires 277.1552, found 277.1564.
References
1 P. T. Corbett, J. Leclaire, L. Vial, K. R. West, J.-L. Wietor,
J. K. M. Sanders and S. Otto, Chem. Rev., 2006, 106, 3652–
3711.
2 P. T. Corbett, S. Otto and J. K. M. Sanders, Chem. Eur. J., 2004, 10,
3139–3143.
3 K. Severin, Chem. Eur. J., 2004, 10, 2565–2580.
4 I. Saur and K. Severin, Chem. Commun., 2005, 1471–1473.
5 R. T. S. Lam, A. Belenguer, S. L. Roberts, C. Naumann, T. Jar-
rosson, S. Otto and J. K. M. Sanders, Science, 2005, 308, 667–
669.
6 S. L. Roberts, R. L. E. Furlan, S. Otto and J. K. M. Sanders, Org.
Biomol. Chem., 2003, 1, 1625–1633.
7 S. M. Voshell, S. J. Lee and M. R. Gagne´, J. Am. Chem. Soc., 2006, 128,
12422–12423.
8 T. Bunyapaiboonsri, O. Ramstrom, S. Lohmann, J. M. Lehn, L. Peng
and M. Goeldner, ChemBioChem, 2001, 2, 438–444.
9 M. G. Simpson, S. P. Watson, N. Feeder, J. E. Davies and J. K. M.
Sanders, Org. Lett., 2000, 2, 1435–1438.
N-(4-(Dimethoxymethyl)benzoyl)-L-Phe-L-Pro methyl ester (Re-
action iii). L-Phenylalanine-L-proline methyl ester (0.25 g,
0.91 mmol) and pDMA (0.18 g, 0.92 mmol) were coupled
through the general amide coupling procedure. Silica gel column
chromatography (EtOAc, Rf = 0.55) yielded the product as
1
colourless oil (0.36 g, 88%). H NMR (500 MHz, CDCl3) d 7.64
(2H, d, J = 8.5 Hz, Ar-H), 7.38 (2H, d, J = 8.5 Hz, Ar-H), 7.32–
7.16 (6H, m, Phe-Ar and Phe-NH), 5.34 (1H, s, CH(OCH3)2),
5.11 (1H, m, Phe-a), 4.48 (1H, m, Pro-a), 3.71 (1H, m, Pro-d),
3.68 (3H, s, OCH3), 3.25 (6H, s, CH(OCH3)2), 3.24–3.08 (3H, m,
Phe-b(2H) and Pro-d(1H)), 2.16 (1H, m, Pro-b), 1.92 (2H, m, Pro-
b(1H) and Pro-c(1H)), 1.65 (1H, m, Pro-c); 13C NMR (125 MHz,
CDCl3) d 172.1, 170.5, 166.3, 141.4, 136.2, 133.8, 129.7, 128.3,
126.9, 126.8, 126.7, 102.3, 59.0, 52.5, 52.5, 52.1, 46.9, 38.3, 28.9,
24.9; HRMS (ESI) [M + Na]+ C25H30N2O6Na requires 477.2002,
found 477.1996.
10 S. J. Rowan, D. G. Hamilton, P. A. Brady and J. K. M. Sanders, J. Am.
Chem. Soc., 1997, 119, 2578–2579.
11 S. J. Rowan, P. S. Lukeman, D. J. Reynolds and J. K. M. Sanders,
New J. Chem., 1998, 22, 1015–1018.
12 I. Saur, R. Scopelliti and K. Severin, Chem. Eur. J., 2006, 12, 1058–
1066.
13 S. Liu, C. Ruspic, P. Mukhopadhyay, S. Chakrabarti, P. Y. Zavalij and
L. Isaacs, J. Am. Chem. Soc., 2005, 127, 15959–15967.
14 S.-A. Poulsen, P. J. Gates, G. R. L. Cousins and J. K. M. Sanders, Rapid
Commun. Mass Spectrom., 2000, 14, 44–48.
786 | Org. Biomol. Chem., 2007, 5, 778–786
This journal is
The Royal Society of Chemistry 2007
©