The development and preparation of the 2,4-dimethoxybenzyl arylhydrazine (DMBAH) "latent" safety-catch linker: solid phase synthesis of ketopiperazines.

Article abstract of DOI:10.1039/b301701j

The development and preparation of the 2,4-dimethoxybenzyl arylhydrazine (DMBAH) linker 3, a new class of "latent" safety-catch linker which is stable under Mitsunobu alkylation conditions and in the presence of amines and hydrazine, is reported. The util

Full text of DOI:10.1039/b301701j

View Article Online / Journal Homepage / Table of Contents for this issue
The development and preparation of the 2,4-dimethoxybenzyl
arylhydrazine (DMBAH) “latent” safety-catch linker:
solid phase synthesis of ketopiperazines
Frédéric Berst,a,b Andrew B. Holmesb and Mark Ladlow*a
a GSK Cambridge Technology Centre, University Chemical Laboratory, Lensfield Road,
Cambridge, UK CB2 1EW. E-mail: Mark.2.Ladlow@gsk.com
b University Chemical Laboratory, Lensfield Road, Cambridge, UK CB2 1EW
Received 12th February 2003, Accepted 26th March 2003
First published as an Advance Article on the web 11th April 2003
The development and preparation of the 2,4-dimethoxybenzyl arylhydrazine (DMBAH) linker 3, a new class of
“latent” safety-catch linker which is stable under Mitsunobu alkylation conditions and in the presence of amines and
hydrazine, is reported. The utility of the new linker is exemplified by the synthesis of ketopiperazines (MKPs) 24
bearing up to four points of diversity using a cyclitive cleavage approach.
a strong nucleophile (hydrazinolysis of Dde protecting group),
Introduction
whilst being amenable to mild activation towards nucleophilic
Over the last thirty years, solid phase organic synthesis has
cleavage at the last step in a synthesis. A safety-catch linker,
evolved into a valuable chemical technology for the generation
inert during the synthetic process but chemically activated prior
of compound libraries.1 This has been facilitated by the avail-
to the cleavage step, would therefore be most suitable for our
ability of an increasing number of linkers which have permitted
purposes.
the transposition of many synthetic transformations to the
However, a survey of existing safety-catch linkers did not
solid phase.2
reveal a candidate that would be compatible with the proposed
Mono-ketopiperazines (MKPs) represent an attractive small
scaffold for library generation which are complementary to
chemistry.5 For example, the 4-alkylsulfamylphenol safety-
catch linker developed by Liener and Marshall6 has been
diketopiperazines and are able to display multiple points of
observed to undergo premature nucleophilic cleavage without
diversity. An attractive synthetic route to MKP libraries is
the need for oxidative activation.7 In addition, an acylsulfon-
embodied in a solid phase cyclitive cleavage approach, whereby
amide linker would also be unsuitable for our purpose since their
only cyclised compounds are released from the resin, leaving
any acyclic impurities trapped on the support.
As outlined in Scheme 1, it was envisaged that the required
resin-bound linear precursors, incorporating four points of
activation to nucleophilic attack is effected by N-alkylation.8
Similarly, in our hands, the recently reported arylhydrazine
“safety-catch” linker9,10 was found to be incompatible with
Mitsunobu N-alkylation conditions, giving rise to multiple
products.
Herein, we report in detail the development and preparation
of a new safety-catch linker which is compatible with both
diversity, could be prepared using Fukuyama’s Mitsunobu
N-alkylation methodology3 to couple an immobilised α-amino
acid N-2-nitrobenzenesulfonamide (first point of diversity) and
an N-protected 1,2-amino alcohol (second point of diversity).
In the resulting adduct replacement of the sulfonamide serves
to introduce a third point of diversity. A fourth point of diver-
Mitsunobu alkylation, amines and hydrazine. The utility of the
new linker was exemplified by the synthesis of MKPs bearing
up to 4 points of diversity.11 In particular, we disclose a detailed
sity may then be incorporated by reductive amination of the
study of the cleavage of the new linker under nucleophilic
terminal amino group prior to cyclitive cleavage.
cyclitive cleavage conditions in the presence of copper cations
as an oxidant.
Results and discussion
Preliminary investigations showed that the incompatibility of
the reported arylhydrazine linker9,10 with Mitsunobu reaction
conditions was attributable to the presence of the reactive
hydrazine moiety which was susceptible to N-alkylation. It
occurred to us that temporary protection of the anilino nitro-
gen would circumvent this possibility. Prior to cleavage, a suit-
able protecting group could be removed to reveal the hydrazine
moiety which could then be subjected to oxidative cleavage
(Scheme 2). In this sense, the proposed new linker would serve as
Scheme 1 Retrosynthetic analysis of mono-ketopiperazines.
a “latent” safety-catch linker.
Pivotal to the success of any solid phase synthesis is the
Synthesis of the 2,4-dimethoxybenzyl arylhydrazine (DMBAH)
choice of linker used. For the synthesis of MKPs, preliminary
linker
solution phase studies established that the Mitsunobu alkyl-
ation was most successful when performed using an amino
alcohol protected with an N-Dde (N-2-acetyldimedone) group.4
Therefore, as a prerequisite, we required a linker stable under
both Mitsunobu N-alkylation conditions and in the presence of
The choice of a suitable linker protecting group was pivotal to
the successful implementation of our strategy. We chose to use
the acid-labile 2,4-dimethoxybenzyl group (DMB) to block the
reactive nitrogen, since its removal could be effected under
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 3
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 7 1 1 – 1 7 1 9
1711

View Article Online
Scheme 2 Principle of the “latent” arylhydrazine safety-catch linker.
acidic conditions which are orthogonal to those required for the
synthesis of MKPs.
Thus, 4-aminobenzoic acid and 2,4-dimethoxybenzaldehyde
were condensed in refluxing toluene under Dean–Stark condi-
tions to form the intermediate imine, which was immediately
reduced to the corresponding aniline 1 with sodium triacetoxy-
borohydride. Nitrosation of 1 afforded the nitrosamine 2, which
in turn was carefully reduced to the hydrazine in the presence of
zinc.12 The linker was most conveniently isolated in multigram
quantities as the Dde derivative 3 (Dde–DMBAH linker,
Scheme 3).
Scheme 4 Reagents and conditions: i) ArgoGel–NH2, PyBOP, HOBt,
DIPEA, CH2Cl2, DMF, rt; ii) 20% (v/v) hydrazine hydrate–DMF, rt;
iii) Fmoc–Phe–OH, DIC, CH2Cl2, DMF, rt; iv) piperidine–CH2Cl2–
DMF (1 : 2 : 2), rt; v) 2-nitrobenzenesulfonyl chloride, DIPEA, CH2Cl2,
rt; vi) a) 10% (v/v) TFA–CH2Cl2, rt, b) Cu(OAc)2, pyridine, MeOH, rt.
nucleophile. Gratifyingly, in a small scale experiment, the
desired methyl ester 8a13 was obtained in greater than 95%
purity by LC-MS analysis and subsequent re-exposure of
the recovered resin to the two stage cleavage protocol gave no
additional 8a.
In progressing the intermediate resin 8, we were concerned
that the final ketopiperazine formation might be dependant
upon the relative configuration of the stereocentres present.
Therefore, we elected to work in two representative diastereo-
meric series. Thus, the sulfonamide resin 8 was alkylated under
Mitsunobu conditions with N-Dde--phenylalaninol 9 and
N-Dde--phenylalaninol 10 to yield resins 11 and 12 respect-
ively (Scheme 5).
In order to introduce another point of diversity onto the
dipeptoid template, the 2-nitrobenzenesulfonamide N-protect-
ing group was removed from the resins 11 and 12 in the
presence of sodium thiophenolate to afford the secondary
amino-resins 13 and 14. These, in turn, were acylated with
2-naphthoyl chloride to yield the corresponding resins 15
and 16. Resins 11–16 were N-deprotected in the presence of
hydrazine to yield the primary amino-resins 17, 18, 19, 20, 21
and 22. Notably, no evidence for the premature cyclitive
cleavage of these amino-resins was observed prior to removal
of the DMB N-protecting group and oxidative activation
of the DMBAH linker. To exemplify reductive amination,14
the primary amine resin 18 was treated with p-tolylaldehyde
in the presence of TMOF and NaBH(OAc)3 to give the
corresponding secondary amine resin 23.
To promote the cyclitive cleavage of the resin-bound pre-
cursors 17–23, the amino-resins 17–23 were first treated with a
solution of TFA in dichloromethane (10% v/v) for 10 min, and
then a solution of copper() acetate (2 equiv.) and pyridine
(10 equiv.) in acetonitrile (a non-nucleophilic solvent in which
the copper salt is reasonably soluble) for 2–3 h. In all cases,
LC-MS analysis of the resulting solutions showed the pres-
ence of the desired products and the crude materials were
readily purified by preparative HPLC to give the MKPs 24a–g
(Table 1). Notably, the relative configuration did not signifi-
cantly influence the yield and purity of the products obtained
(24c and 24d). However, we noted that the crude purities and
overall yields obtained for 24c, 24d and 24g were only moderate
to good—in contrast to the other examples. In particular, the
Scheme
3
Reagents and conditions: i) p-TsOH, toluene, reflux;
ii) NaBH(OAc)3, CH2Cl2, rt; iii) NaNO2, H2SO4, EtOH, rt; iv) Zn,
AcOH, aq. NaOH, 0 ЊC to 100 ЊC; v) 2-acetyldimedone, EtOH, reflux.
Solid phase synthesis of MKPs
In order to establish that the new DMBAH linker 3 was indeed
suitable for the solid phase synthesis of MKPs and compatible
with both Mitsunobu N-alkylation and Dde hydrazinolysis, we
targeted a series of representative MKPs incorporating up to
four points of diversity. The key starting resin 8 was prepared as
shown in Scheme 4. The protected DMBAH linker 3 was first
loaded onto ArgoGel–NH2 to afford resin 4 which was
in turn treated with hydrazine hydrate to afford the free
hydrazino-resin 5. Resin 5 was treated with N-Fmoc-phenyl-
alanine in the presence of diisopropylcarbodiimide (DIC) to
afford the acylated resin 6. Subsequent treatment with piper-
idine in a mixture of DMF and dichloromethane (1 : 1) effected
complete removal of the Fmoc group to yield the amino-resin
7, which was then sulfonylated with 2-nitrobenzenesulfonyl
chloride in the presence of diisopropylethylamine (DIPEA).
At this stage, a sample of the resulting resin 8 was cleaved for
analysis to confirm that the linker was compatible with the pre-
ceding chemistries. The cleavage of the DMBAH linker was
performed in two steps: acidic treatment with a solution of TFA
in dichloromethane (10% v/v) to remove the DMB N-protecting
group, followed by oxidation of the resulting acyl arylhydrazine
in the presence of copper() acetate in methanol as an external
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 7 1 1 – 1 7 1 9
1712

View Article Online
Table 1 Isolated yields and crude purities of MKPs 24a–g
Resin
MKP
R1
R2
R3
R4
Purity (%)a
Yield (%)b
17
18
19
20
21
22
23
24a
24b
24c
24d
24e
24f
24g
o-NO2PhSO2
o-NO2PhSO2
H
H
Benzyl
H
Benzyl
H
Benzyl
Benzyl
Benzyl
H
Benzyl
H
Benzyl
H
H
H
H
H
H
H
H
>99
90
74
63
96
>99
75
68
65
49
54
69
70
34
H
2-Naphthyl
2-Naphthyl
o-NO2PhSO2
p-MePhCH2
a HPLC product peak area in crude cleavage mixture (230 nm). b Overall purified isolated yield for the whole synthetic sequence starting from
ArgoGel–NH2 resin.
a secondary amine was low. Thus, in order to gain a better
understanding of the cyclitive cleavage process we initiated a
more thorough study with the objective of identifying robust
and general conditions that could be applied to the parallel
synthesis of compound arrays.
A study of the cyclitive cleavage process
To simplify product clean-up,15 we wanted to limit the amount
of copper salts used for the oxidation of the linker whilst main-
taining acceptable cleavage rates. We therefore studied the rate
of release of the target MKPs under a range of conditions
chosen so as to be compatible with any future library format.
In an initial study, we investigated the rate of cyclitive cleav-
age of the resins 17, 18, 21–23 when treated under identical
conditions with 1.0 equivalents of copper() acetate and 10
equivalents of pyridine in acetonitrile (Fig. 1a). The numbers of
equivalents used were calculated with respect to the maximum
theoretical loading of a given resin, assuming that all the steps
were quantitative and that no loss of material occurred
Scheme 5 Reagents and conditions: i) 9 or 10, di-tert-butyl azo-
dicarboxylate (TBAD), Ph3P, CH2Cl2, rt; ii) PhSNa in DMF, rt;
iii) 2-naphthoyl chloride, DIPEA, CH2Cl2, rt; iv) hydrazine hydrate,
DMF, rt; v) p-tolylaldehyde, CH2Cl2–trimethylorthoformate (3 : 1)
followed by NaBH(OAc)3, AcOH (1%) in CH2Cl2.
Fig. 1 (a) Cleavage profiles for resins 17, 18, 21–23 to give MKPs
24a, 24b, and 24e–g. Conditions: 1 equiv. Cu(OAc)2; 10 equiv. pyridine.
amine-containing MKPs 24c and 24d were contaminated with a
(b) Cleavage profile for resin 18 to give MKP 24b in the presence of
number of unidentified impurities not observed for the other
examples and the isolated yield of 24g derived by cyclisation of
0.25, 0.5 and 1.0 equiv. of Cu(OAc)2, 10 equiv. pyridine. Additional
Cu(OAc)2 added at t = 40 min and t = 70 min.
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 7 1 1 – 1 7 1 9
1713

View Article Online
throughout the solid phase synthesis. The rate of cleavage of
each MKP was monitored over time by HPLC analysis of the
crude supernatants. Interestingly, under identical experimental
conditions, the cyclitive cleavage rates of the five resins 17, 18,
21–23 to afford MKPs 24a, 24b, 24e, 24f and 24g were found to
be similar and complete cleavage of each compound was
observed in less than 40 min. The similarities in cyclitive cleav-
age rates observed suggest a low structural dependence on the
resin-bound precursors. Moreover, it suggests that the oxid-
ation of the linker to the acyldiazene dictates the rate of release,
and that subsequent intramolecular nucleophilic attack on the
carbonyl is fast. This is also true for the cyclitive cleavage
of the secondary amino-resin 24g. Therefore, the low yield
obtained in this case (34%) might be better attributed to a
reduced resin loading arising from premature cleavage, or
degradation of the linker during the prolonged reductive
amination step (72 h).
In principle, the oxidation of the hydrazine to the acyldiazene
is catalytic in copper(), as the catalyst can be reoxidised by the
presence of oxygen dissolved in the reaction solvent. Indeed,
typically 0.5 equivalents of a copper() salt were sufficient to
cleave the previously reported arylhydrazine linker in less than 2
h.9 Therefore, in a second investigation (Fig. 1b), it was decided
to carry out a series of kinetic experiments to study the effect of
copper stoichiometry and concentration on the cyclisation–
release process. Since the cyclisation profiles of compounds 24a,
24b, 24e, 24f and 24g appeared to be very similar, resin 18 was
selected as a representative example for this study.
Resin 18 was cleaved in three parallel experiments in the pres-
ence of 0.25 (expt. 1), 0.5 (expt. 2) and 1.0 (expt. 3) theoretical
equivalents of copper() acetate in the same volume of
acetonitrile–pyridine. The experiments were carried out in open
vessels under vigorous stirring. The curves obtained for the
release of MKP 24b clearly indicate that the rate of release is
strongly dependent on the stoichiometry and concentration of
copper() acetate.
After 40 min, additional copper() acetate was added to each
experiment as required to bring the total amount present in
each to 1.0 equivalent. This resulted in a rapid increase in the
rate of cleavage in expts. 1 and 2. After a further 30 min,
another equivalent of copper() acetate was added to each
experiment, but the amount of 24b in the supernatants
remained constant, indicating that all the substrate had been
released from the resin. Extrapolation of the curves obtained
from expts. 1 and 2 suggested that even a considerable increase
in reaction time would still give rise to a reduced cleavage yield
if less than 1.0 theoretical equivalent of copper() acetate was
used. By contrast, in expt. 3, with 1.0 theoretical equivalent of
copper, at a concentration of 1.0 mg cmϪ3, complete release of
the desired product occurred in less than 1 h.
Fig. 2 (a) Cleavage profile for resins 19 and 20 to give MKPs 24c and
24d in the presence of 1 equiv. Cu(OAc)2. Additional 1 equiv. Cu(OAc)2
added at t = 100 min. (b) Cleavage profile for resin 20 to give MKP 24d
in the presence of 2 equiv. Cu(OAc)2. Additional 1 equiv. of Cu(OAc)2
added at t = 120 min.
but that this progressively decreased with time. After 120 min,
another equivalent of copper() acetate was added to the cleav-
age mixture and a further increase in the rate of disappearance
of 24d was noted along with the appearance of a number of
impurities.16 This experiment indicates that the behaviour
of MKPs containing a free secondary amine differs under the
cleavage conditions from that of the corresponding acylated or
sulfonylated analogues (which were found to be stable for
several hours in the presence of excess Cu() cations – data not
shown). Although an increased amount of copper ions is
required for complete cleavage, shorter reaction times are neces-
sary to avoid subsequent degradation of the product. This
observation would be particularly relevant to the preparation
of compounds containing unprotected amino groups using the
DMBAH linker.
Conclusions
However, in a third study (Fig. 2a), the cleavage of resins 19
and 20 in the presence of 1.0 equivalent of Cu() to give MKPs
24c and 24d respectively showed a different behaviour. In the
course of the experiment, the supernatant gradually lost the
blue colour characteristic of copper() ions. Addition of a fur-
ther 1.0 equivalent of copper() acetate after 100 min restored
the colour and an increase in conversion was immediately
observed. The decolouration of the solution and dark colour-
ation of the resin beads indicated that reduction to Cu()
occurred, but that in this case subsequent aerobic oxidation to
Cu() did not occur. Notably, MKPs 24c and 24d differ from
the other examples prepared in that they contain a basic
secondary amine.
A further experiment was carried out to study the stability of
the MKP 24d in the presence of excess copper cations (Fig. 2b).
Thus, after removal of the DMB N-protecting group, resin 20
was exposed to 2.0 theoretical equivalents of copper() acetate
in acetonitrile–pyridine. The formation of MKP 24d was
closely monitored and it was observed that a maximum concen-
tration of the desired product was attained after only 10 min,
In summary, we have described the preparation and develop-
ment of a new “latent” arylhydrazine solid phase linker
(DMBAH, 3) and have shown that it is stable under Mitsunobu
conditions and in the presence of a good nucleophile such as
hydrazine. The utility of the new linker has been demonstrated
in a versatile solid phase synthesis of mono-ketopiperazines
(MKPs) 24 utilising a cyclitive cleavage strategy. Up to four
points of diversity were incorporated into the MKP scaffold in
this way. A study to identify robust cyclitive cleavage conditions
for the preparation of MKPs has determined that the treatment
of the DMBAH resin with a solution of TFA in dichloro-
methane (10% v/v) followed by exposure to 1.0 theoretical
equivalent of copper() acetate in acetonitrile containing 10
equivalents of pyridine reliably gives complete cleavage in
under 1 h – except where the product MKP contains a second-
ary amine in the ring. In this case, 2.0 equivalents of copper()
acetate and a shorter reaction time (20 min) is preferred to
avoid subsequent degradation of the resulting MKP. Further
studies to demonstrate the utility of the new linker 3 for
the preparation of libraries based upon related heterocyclic
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 7 1 1 – 1 7 1 9
1714

View Article Online
scaffolds by a cyclitive cleavage strategy using the protocols
described herein are currently in progress and will be the subject
of a future report.
solution of Cu(OAc)2 in pyridine–methanol (1 : 30; 50 µl).
The mixture was left to stand for 30 min and then filtered. The
filtrate was collected in an HPLC vial and submitted for HPLC
and LC-MS analysis. Nb. Under these cleavage conditions, the
methyl ester of the substrate cleaved from the resin was
obtained.
Experimental
Unless otherwise specified, all reactions involving resins were
carried out using Alltech tubes mounted on a IKA Vibrax
VXR laboratory shaker. All moisture sensitive reactions were
carried out under a nitrogen atmosphere in oven-dried glass-
ware. All solvents and reagents were used as supplied unless
otherwise stated. Analytical thin layer chromatography (TLC)
was carried out on Polygram SIL G/UV254 plates. Column
chromatography was performed over Merck Kieselgel 9385
under nitrogen pressure. Analytical high pressure liquid chrom-
atography (HPLC) was performed using a Hewlett Packard
Series 1050 instrument. Column: Supercosil ABZϩPLUS
3.3 cm × 4.6 mm, 3 µm. Eluent A: water, 0.1% TFA; B: aceto-
4-(2,4-Dimethoxybenzylamino)benzoic acid (1). p-TsOH (200
mg) was added to a suspension of 4-aminobenzoic acid (12.4 g,
90.2 mmol) and 2,4-dimethoxybenzaldehyde (15.0 g, 90.2
mmol) in toluene (500 cm3) and the resulting mixture was
heated under reflux with Dean–Stark water removal for 3 h.
The solvent was evaporated under reduced pressure and the
residual solid was suspended in CH2Cl2 (500 cm3) and treated
with acetic acid (15 cm3) and sodium triacetoxyborohydride
(28.0 g, 135 mmol). The resulting suspension was stirred at
room temperature for 72 h when the solvent was removed under
reduced pressure. The yellow oil obtained was triturated with
water (500 cm3) for 1 h and the resulting off-white powder
was collected on a filter. The powder was washed with water
(2 × 20 cm3) and diethyl ether (2 × 20 cm3) and then triturated
with diethyl ether to give the benzoic acid 1 as a white solid
(20.7 g, 80%). Mp: 185–187 ЊC (from EtOH); HPLC (254 nm):
tR = 4.45 min (100%) [Found: C, 67.06; H, 5.81; N, 4.74%;
(M ϩ H)ϩ 288.1244. C16H17NO4 requires C, 66.89; H, 5.96; N,
4.87%; MH, 288.1235]; νmax (KBr)/cmϪ1 3403, 1672, 1604, 1315,
1295 and 1176; δH (400 MHz, DMSO-d6) 11.92 (1 H, s), 7.61
(2 H, d, J 9), 7.08 (1 H, d, J 8), 6.71 (1 H, t, J 6), 6.55–6.50 (3 H,
m), 6.46 (1 H, dd, J 9), 4.15 (2 H, d, J 6), 3.78 (3 H, s) and 3.71
(3 H, s); δC (100 MHz, DMSO-d6) 167.6, 159.9, 158.1, 152.7,
131.2, 128.9, 118.8, 117.1, 111.1, 103.4, 98.5, 55.6, 55.3 and
40.7; m/z (ESI) 288 [M ϩ H]ϩ.
nitrile 95%, water 5%, TFA 0.05%. Flow rate: 1 cm3 minϪ1
.
Detection: UV (diode array: 215, 230, 254 nm). Method:
gradient 10–95% B in A over 7 min. Auto-preparative HPLC
purifications were carried out on a Gilson AutoPrep system.
Column: Supercosil ABZϩPLUS 10 cm × 2.12 cm, 5 µm.
Eluent: A, B. Flow rate: 6 cm3 minϪ1. Detection: 215 nm.
Method: (20%–95% B in A) over 20 min. Melting points were
measured on a Mettler FP5 automatic melting point apparatus
in open capillaries and are uncorrected. Optical rotations were
measured at the sodium D-line using an Optical Activity AA10
polarimeter at 20 ЊC (c is measured in g 100 cmϪ3, in the indi-
cated solvent, impling units of deg dm2 gϪ1). Infrared spectra
were collected on a Bio Rad FT-IR machine using DRIFTs
from a potassium bromide (KBr) surface. Liquid chromato-
graphy-mass spectra (LC-MS) were recorded on a Micromass
Platform I (8084) under electrospray positive and negative ion-
isation conditions. Column: Supercosil ABZϩPLUS 3.3 cm ×
4.6 mm, 3 µm. Eluent A: 10 mM solution of ammonium acetate
in water, 0.1% formic acid; B: acetonitrile 95%, water 5%,
formic acid 0.05%. Flow rate: 1 cm3 minϪ1. Detection: UV
(diode array: 215, 230, 254 nm). Method: gradient 0–100% B in
A over 3.5 min. Accurate mass spectra were recorded on a VG
Autospec mass spectrometer in positive electrospray mode.
NMR spectra were recorded in the indicated solvent. Chemical
shifts (δ) are reported in parts per million (ppm) relative to
tetramethylsilane, the following abbreviations are used for
multiplicities: s = singlet; d = doublet; t = triplet; m = multiplet;
dd = doublet of doublets; br = broad; and coupling constant
4-(2,4-Dimethoxybenzylnitrosoamino)benzoic acid (2). To an
ice-cold suspension of 1 (19.6 g, 68.2 mmol) in absolute ethanol
(500 cm3) was added sodium nitrite (4.94 g, 71.6 mmol) fol-
lowed, slowly, by conc. H2SO4 (4.65 cm3, 87.3 mmol). The
resulting mixture was stirred at room temperature for 18 h,
when additional NaNO2 (2.40 g, 34.1 mmol) and conc. H2SO4
(2.3 cm3 as a solution in 20 cm3 of ethanol) were added. After 2
h, ice-water (400 cm3) was added and the mixture stirred for 10
min before being concentrated under reduced pressure to a vol-
ume of approximately 500 cm3. Water (500 cm3) was added and
the resulting suspension was collected by filtration to give the
acid 2 as a brown powder (19.55 g, 90%). Mp: 165–167 ЊC (from
EtOH); HPLC (254 nm): tR = 4.76 min (100%) [Found: C,
60.50; H, 4.92; N, 8.36. C16H16N2O5 requires C, 60.76; H, 5.10;
N, 8.86%]; νmax (KBr)/cmϪ1 1680, 1606, 1294 and 901; δH (400
MHz, DMSO-d6) 12.92 (1 H, s), 7.99 (2 H, d, J 9), 7.67 (2 H, d,
J 9), 6.73 (1 H, d, J 9), 6.49 (1 H, d, J 3), 6.36 (1 H, dd, J 9, 3),
5.14 (2 H, s), 3.69 (3 H, s) and 3.67 (3 H, s); δC (100 MHz,
DMSO-d6) 189.6, 168.9, 162.5, 160.0, 146.8, 132.9, 131.2,
121.7, 115.9, 107.1, 100.8, 57.8, 57.5 and 43.8; m/z (ESI) 339
[(M ϩ Na)ϩ, 80%], 287 [(M Ϫ NO ϩ H)ϩ, 100%].
1
J-values are quoted in Hz. H NMR spectra were recorded on
either Bruker AM 400 or Varian Inova 750 spectrometers
at 400 MHz and 750 MHz respectively. Variable temperature
1H NMR were recorded at the specified temperature using
a Bruker DPX 250 at 250 MHz. 1H HRMAS NMR were
recorded on a Bruker AM 400 fitted with a Varian Nano
NMR probe and a CPMG pulse sequence was utilised to
minimise line broadening and aid interpretation. 13C NMR
were recorded on a Bruker AM 400 at 100 MHz with proton
decoupling.
Where possible, to characterise resins analytical cleavage of
small samples of beads was performed according to the follow-
ing procedure and the supernatant solution obtained analysed
by both HPLC and LC-MS:
4-{N-(2,4-Dimethoxybenzyl)-NЈ-[1-(4,4-dimethyl-2,6-dioxo-
cyclohexylidene)ethyl]hydrazino}benzoic acid (3). To a solution
of 2 (3.80 g, 12.0 mmol) in aqueous sodium hydroxide (1 M,
12 cm3) and water (100 cm3) was added zinc (16 g, 0.24 mol,
freshly activated with dilute HCl). The resulting suspension was
cooled to 0 ЊC and acetic acid (6 cm3) was added dropwise over
20 min with vigorous stirring, causing the suspension to thicken
to a paste. The mixture was heated slowly to 100 ЊC (bath tem-
perature) and maintained at this temperature for 6 hours. The
suspension was then filtered and the resulting grey solid tritur-
ated with aqueous NaOH (2 M, 2 × 10 cm3) which caused a
white solid to precipitate. The pH was adjusted to pH 4 using
concentrated HCl and the suspension was filtered to yield a
white solid, which was dried under reduced pressure. The solid
was suspended in ethanol (60 cm3) and treated with 2-acetyl-
Analytical cleavage of resin-bound substrates for HPLC–LC-
MS analysis
A small sample of resin (typically between fifty and a hundred
beads) was transferred to a RAININ RT-10GF pipette filter tip.
The beads were washed with DMF (1 cm3) and CH2Cl2 (1 cm3),
and then treated with a solution of TFA in CH2Cl2 (10% v/v;
1 cm3). After 10 min, the beads were filtered, washed (DMF:
1 × 1 cm3, CH2Cl2: 1 × 1 cm3), and treated with a saturated
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 7 1 1 – 1 7 1 9
1715

View Article Online
dimedone (Dde–OH: 1.13 g, 6.20 mmol) under reflux periodic-
ally adding additional Dde–OH until a slight excess persisted
according to TLC analysis. The solvents were removed under
reduced pressure to leave an oil which was purified by column
chromatography (EtOAc–hexane–acetic acid: 31 : 66 : 3) to give
the Dde-protected linker 3 as a white solid (3.12 g, 56%). Mp:
199–201 ЊC (from MeCN–water); HPLC (254 nm): tR = 4.81
min (100%) [Found: C, 67.10; H, 6.41; N, 5.71%; (M ϩ H)ϩ
467.2173. C26H30N2O6 requires C, 66.94; H, 6.48; N, 6.00%;
MH, 467.2182]; νmax (KBr)/cmϪ1 1709, 1684, 1604, 1588 and
1288; δH (400 MHz, DMSO-d6) 14.00 (1 H, s), 12.45 (1 H, s),
7.81 (2 H, d, J 9), 7.11 (1 H, d, J 8), 6.89 (2 H, d, J 9), 6.49 (1 H,
d, J 2), 6.41 (1 H, dd, J 8, 2), 4.67 (2 H, br s), 3.73 (3 H, s), 3.69
(3 H, s), 2.30 (4 H, br s), 2.22 (3 H, s) and 0.93 (6 H, s); δC (100
MHz, DMSO-d6) 175.1, 166.9, 160.7, 158.5, 151.2, 131.9, 130.9,
128.8, 128.1, 121.9, 114.4, 112.7, 106.1, 104.3, 98.1, 55.2, 55.1,
52.1, 29.6, 27.7 and 15.9; m/z (ESI) 467 [(M ϩ H)ϩ, 100%].
max. th. loading 0.35 mmol gϪ1) in CH2Cl2 (2 cm3) was added
diisopropylethylamine (0.36 cm3, 2.08 mmol) followed by a
solution of o-nitrobenzenesulfonyl chloride (0.46 g, 2.08 mmol)
and diisopropylethylamine (0.36 cm3, 2.08 mmol) in CH2Cl2
(10 cm3). After 1 h the suspension was filtered, washed (DMF:
6 × 5 cm3, CH2Cl2: 6 × 5 cm3) and dried under reduced pressure
to give the resin 8 (1.28 g). Analytical cleavage of a sample of 8
gave the following data: HPLC (254 nm): tR = 4.96 min (100%);
LC-MS: tR = 4.98 min, [M ϩ H]ϩ = 365, [M ϩ NH4]ϩ = 382
(data consistent with that expected for the corresponding
methyl ester 8a: M 364).
2-[1-((1,R)-Hydroxymethyl-2-phenylethylamino)ethylidene]-
5,5-dimethylcyclohexane-1,3-dione (9). To a solution of (R)-2-
amino-3-phenylpropanol (1.0 g, 6.6 mmol) in absolute ethanol
(5 cm3) was added, dropwise, a solution of 2-acetyldimedone
(1.2 g, 6.6 mmol) in absolute ethanol (5 cm3). The mixture was
stirred at room temperature for 24 h before the solvent was
evaporated under reduced pressure to give a yellow oil. The oil
was taken up in a minimum of EtOAc and hexane was added
until the solution became turbid. The solvents were removed
under reduced pressure and the resulting solid triturated with
hexane to give the alcohol 9 as a white powder (1.64 g, 79%).
Rf 0.43 (EtOAc); HPLC (230 nm): tR = 3.95 min (100%); [α]D
ϩ299 (c 1.00 in MeOH) [Found: C, 72.29; H, 8.07; N, 4.26%;
(M ϩ H)ϩ 316.1909. C19H25NO3 requires C, 72.35; H, 7.99; N,
4.44%; MH, 316.1912]; νmax (KBr)/cmϪ1 3362, 2955, 2878, 1631,
1573, 1465 and 1337; δH (400 MHz, CDCl3) 13.62 (1 H, br s),
7.28–7.14 (5 H, m), 4.05 (1 H, m), 3.77 (1 H, dd, J 4, 11.5), 3.71
(1 H, dd, J 5.5, 11.5), 3.17 (1 H, br s), 3.00 (1 H, dd, J 6, 14),
2.80 (1 H, dd, J 9, 14), 2.32 (4 H, s), 1.26 (3 H, s) and 1.00 (6 H,
s); δC (100 MHz, CDCl3) 198.4, 173.9, 137.2, 129.6, 129.1,
127.4, 108.3, 64.3, 57.7, 57.6, 53.1, 38.8, 30.4, 28.6 and 18.1;
m/z (ESI) 316 [(M ϩ H)ϩ, 100%].
4-{N-(2,4-Dimethoxybenzyl)-NЈ-[1-(4,4-dimethyl-2,6-dioxo-
cyclohexylidene)ethyl]hydrazino}benzoic acid, amide with Argo-
Gel–NH2 (4). To a slurry of ArgoGel–NH2 (1.50 g, theor-
etical (th.) loading 0.41 mmol gϪ1) in CH2Cl2 (10 cm3) was
added diisopropylethylamine (1 cm3) and the resulting mixture
was shaken for 5 min. The resin was thoroughly washed with
distilled CH2Cl2 (3 × 10 cm3) and then treated with a solution of
3 (0.57 g, 1.20 mmol), PyBOP (0.96 g, 1.80 mmol) and HOBt
(0.24 g, 1.80 mmol) in DMF–CH2Cl2 (1 : 1; 10 cm3). After
4 min, diisopropylethylamine (0.64 cm3, 3.60 mmol) was added
and the suspension was stirred at room temperature for 18 h.
The resin was washed (DMF: 6 × 5 cm3, CH2Cl2: 6 × 10 cm3)
and dried under reduced pressure to give the resin 4 (1.8 g);
Kaiser test: Ϫve;17 δH (400 MHz, CDCl3, MAS) 7.70 (d), 6.92
(d), 6.81 (d), 6.32 (s), 6.26 (d), 5.20 (s), 3.75–3.70 (2 × s), 2.29
(br s), 2.25 (s) and 0.95 (s).
4-[N-(2,4-Dimethoxybenzyl)hydrazino]benzoic acid, amide
with ArgoGel–NH2 (5). To a slurry of resin 4 (1.20 g, max. th.
loading 0.31 mmol gϪ1) in DMF (8 cm3) was added hydrazine
hydrate (55% N2H4 in water, 2 cm3). After 15 min the resin was
washed with DMF (3 × 5 cm3). This procedure was repeated
three times and the resin was finally filtered, washed (DMF:
6 × 5 cm3, CH2Cl2: 6 × 5 cm3) and dried under reduced pressure
to give the resin 5 (1.13 g). δH (400 MHz, CDCl3, MAS) 7.70
(d), 7.05 (m), 6.48 (s), 6.40 (d), 4.65 (s) and 3.75–3.70 (2 × s).
2-[1-((1,S )-Hydroxymethyl-2-phenylethylamino)ethylidene]-
5,5-dimethylcyclohexane-1,3-dione (10). To a solution of (S)-2-
amino-3-phenylpropanol (1.00 g, 6.61 mmol) in ethanol
(10 cm3) was added 2-acetyldimedone (1.26 g, 6.94 mmol).
After 2 h, additional (S)-2-amino-3-phenylpropanol (0.20 g,
1.32 mmol) was added to the mixture. After a further 1 h, the
yellow solution was passed through an SCX solid phase extrac-
tion cartridge (strongly acidic sulfonic acid) to scavenge excess
amine. The cartridge was washed with methanol (ca. 20 cm3)
and the combined filtrates were concentrated under reduced
pressure to leave a thick yellow oil. The oil was taken up in
CH2Cl2 (20 cm3) and evaporated to dryness. This process was
repeated 3× to remove any residual ethanol. The resulting clear
glass was then triturated in boiling hexane to give the alcohol 10
as a white solid (1.90 g, 91%). Rf 0.43 (EtOAc); HPLC (230
nm): tR = 3.95 min (100%); [α]D Ϫ299 (c. 1.00 in MeOH)
[Found: C, 72.29; H, 7.99; N, 4.34%; (M ϩ H)ϩ 316.1906.
C19H25NO3 requires C, 72.35; H, 7.99; N, 4.44%; MH,
316.1912]; νmax (KBr)/cmϪ1 3364, 2955, 2878, 1631, 1573, 1465
and 1337; δH (400 MHz, CDCl3) 13.62 (1 H, br s), 7.28–7.14
(5 H, m), 4.05 (1 H, m), 3.77 (1 H, dd, J 4, 11.5), 3.71 (1 H, dd,
J 5.5, 11.5), 3.17 (1 H, br s), 3.00 (1 H, dd, J 6, 14), 2.80 (1 H,
dd, J 9, 14), 2.32 (4 H, s), 1.26 (3 H, s) and 1.00 (6 H, s); δC (100
MHz, CDCl3) 198.4, 173.9, 137.2, 129.6, 129.1, 127.4, 108.3,
64.3, 57.7, 57.6, 53.1, 38.8, 30.4, 28.6 and 18.1; m/z (ESI) 316
[(M ϩ H)ϩ, 100%].
4-{N-(2,4-Dimethoxybenzyl)-NЈ-[(2,S )-(9H-fluoren-9-yl-
methoxycarbonylamino)-3-phenylpropionyl]hydrazino}benzoic
acid, amide with ArgoGel–NH2 (6). To a slurry of resin 5 (1.13
g, max. th. loading: 0.36 mmol gϪ1) in CH2Cl2–DMF (1 : 1,
2 cm3) was added a solution of Fmoc–Phe–OH (1.61 g, 4.16
mmol) and DIC (0.65 cm3, 4.16 mmol) in CH2Cl2–DMF (1 : 1,
10 cm3). The suspension was shaken for 18 h and then washed
(DMF: 6 × 5 cm3, CH2Cl2: 6 × 5 cm3) and dried under reduced
pressure to give the resin 6 (1.28 g). Fmoc UV quantification18
indicated a loading of 0.26 mmol gϪ1; Kaiser test: Ϫve.
4-[NЈ-((2,S )-Amino-3-phenylpropionyl)-N-(2,4-dimethoxy-
benzyl)hydrazino]benzoic acid, amide with ArgoGel–NH2 (7).
To a slurry of resin 6 (1.18 g, max. th. loading: 0.32 mmol gϪ1
)
in CH2Cl2 (1 cm3) was added a solution of piperidine in
CH2Cl2–DMF (1 : 2 : 2; 10 cm3). The resulting suspension was
stirred for 10 min, before the resin was washed (DMF: 3 ×
5 cm3). The above procedure was repeated three additional
times. Finally, the resin was thoroughly washed (DMF: 6 ×
5 cm3, CH2Cl2: 6 × 5 cm3) and dried under reduced pressure to
give the resin 7 (1.19 g); Kaiser test: ϩve.
General procedure for on-resin Mitsunobu alkylation
4-(N-(2,4-Dimethoxybenzyl)-NЈ-{(2,S )-[{(2,R)-[1-(4,4-di-
methyl-2,6-dioxocyclohexylidene)ethylamino]-3-phenylpropyl}-
(2-nitrobenzenesulfonyl)amino]-3-phenylpropionyl}hydrazino)-
benzoic acid, amide with ArgoGel–NH2 (11). To a slurry of
resin 8 (640 mg, max. th. loading: 0.33 mmol gϪ1) in CH2Cl2
(3 cm3), was added triphenylphosphine (569 mg, 2.17 mmol)
4-{N-(2,4-Dimethoxybenzyl)-NЈ-[(2,S )-(2-nitrobenzene-
sulfonylamino)-3-phenylpropionyl]hydrazino}benzoic acid, amide
with ArgoGel–NH2 (8). To a suspension of resin 7 (1.19 g,
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 7 1 1 – 1 7 1 9
1716

View Article Online
and N-Dde--phenylalaninol 9 (710 mg, 2.17 mmol) as a
solution in CH2Cl2 (3 cm3). A solution of di-tert-butyl azodi-
carboxylate (500 mg, 2.17 mmol) in CH2Cl2 (1.2 cm3) was
added in five portions and the resulting yellow suspension was
stirred for 19 h. The resin was filtered, washed (DMF: 6 ×
5 cm3, CH2Cl2: 6 × 5 cm3) and dried under reduced pressure
to give the resin 11 (780 mg). Analytical cleavage of a sample of
11 gave the following data: HPLC (254 nm): tR = 6.17 min
(100%); LC-MS: tR = 5.52 min, [M ϩ H]ϩ = 662 (data consist-
ent with that expected for the corresponding methyl ester:
M 661).
4-(N-(2,4-Dimethoxybenzyl)-NЈ-{(2,S )-[{(2,S )-[1-(4,4-di-
methyl-2,6-dioxocyclohexylidene)ethylamino]-3-phenylpropyl}-
(naphthalene-2-carbonyl)amino]-3-phenylpropionyl}hydrazino)-
benzoic acid, amide with ArgoGel–NH2 (16). According to
the general procedure for resin 15, the resin 14 (188 mg) was
treated with 2-naphthoyl chloride (228 mg, 1.2 mmol) to
give the resin 16 (201 mg). Analytical cleavage of a sample
of 16 gave the following data: HPLC (254 nm): tR = 6.40 min
(100%); LC-MS: tR = 5.63 min [M ϩ H]ϩ = 631 (data
consistent with that expected for the corresponding methyl
ester: M 630).
4-(N-(2,4-Dimethoxybenzyl)-NЈ-{(2,S )-[{(2,S )-[1-(4,4-di-
methyl-2,6-dioxocyclohexylidene)ethylamino]-3-phenylpropyl}-
(2-nitrobenzenesulfonyl)amino]-3-phenylpropionyl}hydrazino)-
benzoic acid, amide with ArgoGel–NH2 (12). According to the
general procedure for resin 11, resin 8 (640 mg, max. th. load-
ing: 0.33 mmol gϪ1) was treated with triphenylphosphine (569
mg, 2.17 mmol), N-Dde-phenylalaninol 10 (710 mg, 2.17 mmol)
and di-tert-butyl azodicarboxylate (500 mg, 2.17 mmol) in
CH2Cl2 for 19 h to give the resin 12 (766 mg). Analytical cleav-
age of a sample of 12 gave the following data: HPLC (254 nm):
tR = 6.11 min (100%); LC-MS: tR = 5.45, min, [M ϩ H]ϩ = 662
(data consistent with that expected for the corresponding
methyl ester: M 661).
General procedure for the on-resin removal of the Dde
N-protecting group
4-[NЈ-{(2,S )-[((2,R)-Amino-3-phenylpropyl)-(2-nitrobenzene-
sulfonyl)amino]-3-phenylpropionyl}-N-(2,4-dimethoxybenzyl)-
hydrazino]benzoic acid, amide with ArgoGel–NH2 (17). To a
slurry of resin 11 (200 mg, max. th. loading 0.30 mmol gϪ1) in
DMF (1.6 cm3) was added hydrazine hydrate (55% N2H4 in
water, 0.4 cm3) and the resulting mixture stirred for 15 min
before the resin was washed with DMF (3 × 2 cm3). The pro-
cedure was repeated. Finally, the resin was filtered, washed
(DMF: 6 × 2 cm3, CH2Cl2: 6 × 2 cm3) and dried under reduced
pressure to give the resin 17 (189 mg). Analytical cleavage of a
sample of 17 gave the following data: HPLC (230 nm): tR
=
5.39 min (88%); LC-MS: tR = 5.09 min, [M ϩ H]ϩ = 466 (data
consistent with that expected for the MKP 24a: M 465).
General procedure for the on-resin removal of the o-nitrobenzene-
sulfonyl group
4-[N-(2,4-Dimethoxybenzyl)-NЈ-((2,S )-{(2,R)-[1-(4,4-di-
methyl-2,6-dioxocyclohexylidene)ethylamino]-3-phenylpropyl-
amino}-3-phenylpropionyl)hydrazino]benzoic acid, amide with
ArgoGel–NH2 (13). To a slurry of resin 11 (200 mg, max. th.
loading 0.30 mmol gϪ1) in DMF (2 cm3) was added a solution
of sodium thiophenolate (158 mg, 1.2 mmol, 20 equiv.) in DMF
(2 cm3). The resin was stirred for 15 min and then the dark blue
solution was filtered. The resin was thoroughly washed (DMF
3 × 5 cm3, allowed to stand for 10 min, 3× [MeOH: 3 cm3,
2 minutes then CH2Cl2: 3 cm3, 2 minutes]) and dried under
reduced pressure to give the resin 13 (188 mg). Analytical cleav-
age of a sample of 13 gave the following data: HPLC (254 nm):
tR = 4.14 min (100%); LC-MS: tR = 5.17 min, [M ϩ H]ϩ = 477
(data consistent with that expected for the corresponding
methyl ester: M 476).
4-[NЈ-{(2,S )-[((2,S )-Amino-3-phenyl-propyl)-(2-nitrobenzene-
sulfonyl)-amino]-3-phenyl-propionyl}-N-(2,4-dimethoxybenzyl)-
hydrazino]benzoic acid, amide with ArgoGel-NH2 (18).
According to the general procedure for resin 17, the resin 12
(200 mg, max. th. loading 0.30 mmol gϪ1) was treated with
hydrazine hydrate (55% N2H4 in water, 0.4 cm3) to give the resin
18 (190 mg). Analytical cleavage of a sample of 18 gave the
following data: HPLC (230 nm): tR = 5.42 min (89%); LC-MS:
tR = 5.12 min, [M ϩ H]ϩ = 466 (data consistent with that
expected for the MKP 24b: M 465).
4-[NЈ-[(2,S )-((2,R)-Amino-3-phenylpropylamino)-3-phenyl-
propionyl]-N-(2,4-dimethoxybenzyl)hydrazino]benzoic acid,
amide with ArgoGel–NH2 (19). According to the general pro-
cedure for resin 17, the resin 13 (100 mg, max. th. loading 0.31
mmol gϪ1) was treated with hydrazine hydrate (55% N2H4 in
water, 0.4 cm3) to give the resin 19 (95 mg). Analytical cleavage
of a sample of 19 gave the following data: HPLC (254 nm): tR =
2.81 min (63%); LC-MS: tR = 4.02 min, [M ϩ H]ϩ = 281 (data
consistent with that expected for the MKP 24c: M 280).
4-[N-(2,4-Dimethoxybenzyl)-NЈ-((2,S )-{(2,S )-[1-(4,4-di-
methyl-2,6-dioxocyclohexylidene)ethylamino]-3-phenylpropyl-
amino}-3-phenylpropionyl)hydrazino]benzoic acid, amide with
ArgoGel–NH2 (14). According to the general procedure for
resin 13, the resin 12 (200 mg, max. th. loading 0.30 mmol gϪ1
)
was treated with sodium thiophenolate (158 mg, 1.2 mmol,
20 equiv.) to give the resin 14 (190 mg). Analytical cleavage of
4-[NЈ-[(2,S )-((2,S )-Amino-3-phenylpropylamino)-3-phenyl-
propionyl]-N-(2,4-dimethoxybenzyl)hydrazino]benzoic acid,
amide with ArgoGel–NH2 (20). According to the general pro-
cedure for resin 17, the resin 14 (100 mg, max. th. loading 0.31
mmol gϪ1) was treated with hydrazine hydrate (55% N2H4 in
water, 0.4 cm3) to give the resin 20 (95 mg). Analytical cleavage
of a sample of 20 gave the following data: HPLC (254 nm): tR =
2.46 min (74%); LC-MS: tR = 3.82 min, [M ϩ H]ϩ = 281 (data
consistent with that expected for the MKP 24d: M 280).
a sample of 14 gave the following data: HPLC (254 nm): tR
=
4.10 min (100%); LC-MS: tR = 5.17 min, [M ϩ H]ϩ = 477 (data
consistent with that expected for the corresponding methyl
ester: M 476).
General procedure for on-resin acylation with acid chlorides
4-(N-(2,4-Dimethoxybenzyl)-NЈ-{(2,S )-[{(2,R)-[1-(4,4-di-
methyl-2,6-dioxocyclohexylidene)ethylamino]-3-phenylpropyl}-
(naphthalene-2-carbonyl)amino]-3-phenylpropionyl}hydrazino)-
benzoic acid, amide with ArgoGel–NH2 (15). To a slurry of
resin 13 (189 mg) in CH2Cl2 (2 cm3) was added a solution of 2-
naphthoyl chloride (228 mg, 1.2 mmol) in CH2Cl2 (2 cm3). The
resulting suspension was stirred for 5 h and then the resin was
filtered, washed (DMF: 6 × 2 cm3, CH2Cl2: 6 × 2 cm3) and dried
to give the resin 15 (200 mg). Analytical cleavage of a sample of
15 gave the following data: HPLC (254 nm): tR = 6.40 min
(100%); LC-MS: tR = 5.64 min, [M ϩ H]ϩ = 631 (data consistent
with that expected for the corresponding methyl ester: M 630).
4-[NЈ-{(2,S )-[((2,R)-Amino-3-phenylpropyl)(naphthalene-2-
carbonyl)amino]-3-phenylpropionyl}-N-(2,4-dimethoxybenzyl)-
hydrazino]benzoic acid, amide with ArgoGel–NH2 (21).
According to the general procedure for resin 17, the resin
15 (200 mg, max. th. loading 0.30 mmol gϪ1) was treated with
hydrazine hydrate (55% N2H4 in water, 0.4 cm3) to give the resin
21 (191 mg). Analytical cleavage of a sample of 21 gave the
following data: HPLC (254 nm): tR = 5.51 min (100%); LC-MS:
tR = 5.79 min, [M ϩ H]ϩ = 435 (data consistent with that
expected for the MKP 24e: M 434).
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 7 1 1 – 1 7 1 9
1717

View Article Online
4-[NЈ-{(2,S )-[((2,S )-Amino-3-phenylpropyl)(naphthalene-2-
carbonyl)amino]-3-phenylpropionyl}-N-(2,4-dimethoxybenzyl)-
hydrazino]benzoic acid, amide with ArgoGel–NH2 (22).
According to the general procedure for resin 17, the resin 16
(200 mg, max. th. loading 0.30 mmol gϪ1) was treated with
hydrazine hydrate (55% N2H4 in water, 0.4 cm3) to give the title
resin 22 (185 mg). Analytical cleavage of a sample of 22 gave
the following data: HPLC (254 nm): tR = 5.40 min (92%);
LC-MS: tR = 5.16 min, [M ϩ H]ϩ = 435 (data consistent with
that expected for the MKP 24f: M 434).
134.0, 133.8, 132.6, 131.4, 130.0, 129.6, 129.5, 128.8, 128.0,
127.6, 124.9, 60.3, 53.8, 44.9, 39.9 and 38.4; m/z (ESI) [Found:
(M ϩ H)ϩ, 466.1436. C24H24N3O5S requires MH, 466.1436],
m/z (ESI) 466 ([M ϩ H]ϩ, 100%).
(3S,6R)-3,6-Dibenzylpiperazin-2-one (24c). According to the
general procedure for 24a, the resin 19 (94 mg, max. th. loading
0.33 mmol gϪ1) was treated with a solution of TFA in CH2Cl2
(10% v/v), then copper() acetate (10.8 mg, 0.06 mmol) and
pyridine (24 µl, 0.30 mmol) in acetonitrile (5 cm3) for 3 h to give,
after purification by HPLC and lyophilisation of the combined
fractions, the trifluoroacetate salt of the MKP 24c as a white
powder (18 mg, 65%). HPLC (230 nm): tR = 2.47 min (100%);
[α]D Ϫ68 (c. 0.31 in MeOH); νmax (KBr)/cmϪ1 3205, 2941, 1677
and 1202; δH (750 MHz, DMSO-d6) 9.19 (1 H, br), 8.43 (1 H,
br), 7.36–7.20 (10 H, m), 4.12 (1 H, br m), 3.82 (1 H, br m), 3.40
(1 H, dd, J 15, 4), 3.05 (1 H, br), 3.01 (1 H, dd, J 14, 4), 2.93
(1 H, dd, J 15, 9), 2.84 (1 H, br), and 2.71 (1 H, dd, J 14, 8.5);
δC (100 MHz, CDCl3) 165.8, 134.5, 134.1, 129.9, 129.7, 129.4,
129.3, 128.4, 128.2, 70.0, 50.7, 45.8, 40.1 and 35.8; m/z (ESI)
[Found: (M ϩ H)ϩ, 281.1658. C18H21N2O requires MH,
281.1653], (ESI) 281 ([M ϩ H]ϩ, 100%).
4-(N-(2,4-Dimethoxybenzyl)-NЈ-{2-[[2-(4-methylbenzyl-
amino)-3-phenylpropyl]-(2-nitrobenzenesulfonyl)amino]-3-
phenylpropionyl}hydrazino)benzoic acid, amide with ArgoGel–
NH2 (23). To a slurry of resin 18 (140 mg, max. th. loading: 0.31
mmol gϪ1) in CH2Cl2–trimethylorthoformate (3 : 1; 2 cm3) was
added p-tolylaldehyde (50 µl, 0.42 mmol). The resulting suspen-
sion was shaken for 3.5 h then filtered and washed with
anhydrous CH2Cl2 (6 × 2 cm3). The resin was then immediately
suspended in CH2Cl2 (3 cm3) and treated with sodium tri-
acetoxyborohydride (89 mg, 0.42 mmol) and acetic acid (30 µl).
The resulting slurry was stirred for 72 h at room temperature.
Finally, the resin was filtered and washed (DMF: 3 × 1 cm3,
H2O: 3 × 1 cm3, DMF: 3 × 1 cm3, CH2Cl2: 6 × 2 cm3) to give the
resin 23. Analytical cleavage of a sample of 23 gave the follow-
(3S,6S )-3,6-Dibenzylpiperazin-2-one (24d). According to the
general procedure for 24a, the resin 20 (95 mg, max. th. loading
0.33 mmol gϪ1) was treated with a solution of TFA in CH2Cl2
(10% v/v), then copper() acetate (10.8 mg, 0.06 mmol) and
pyridine (24 µl, 0.30 mmol) in acetonitrile (5 cm3) for 3 h to give,
after purification by HPLC and lyophilisation of the combined
fractions, the trifluoroacetate salt of the MKP 24d as a white
powder (6.4 mg, 54%). HPLC (230 nm): tR = 2.83 min (100%);
[α]D Ϫ7.5 (c. 4 in MeOH); νmax (KBr)/cmϪ1 3376, 2924, 1677,
1202, 1136 and 1032; δH (750 MHz, DMSO-d6) 9.22 (1 H, br),
8.43 (1 H, br), 7.35–7.20 (10 H, m), 4.12 (1 H, br m), 3.82 (1 H,
br m), 3.21 (1 H, dd, J 15, 5.5), 3.06 (1 H, br d, J 13.5), 3.03
(1 H, dd, J 15, 8), 2.92 (1 H, dd, J 13.5, 4.5), 2.79 (1 H, dd,
J 13.5, 7.5) and 2.73 (1 H, dd, J 13.5, 9); δC (100 MHz, CDCl3)
164.5, 133.1, 132.8, 128.3, 127.9, 127.8, 127.7, 126.6, 126.4,
68.3, 49.0, 41.2, 38.6, and 34.2; m/z (ESI) [Found: (M ϩ H)ϩ,
281.1662. C18H21N2O requires MH, 281.1653], (ESI) 281
([M ϩ H]ϩ, 100%).
ing data: HPLC (215 nm): tR = 6.61 min (85%); LC-MS: tR
=
5.64 min, [M ϩ H]ϩ = 570 (data consistent with that expected
for MKP 24g: M 569).
General procedure for the cyclitive cleavage of resin-bound
intermediates (unoptimised)
(3S,R)-3,6-Dibenzyl-4-(2-nitrobenzenesulfonyl)piperazin-2-
one (24a). The resin 17 (190 mg, max. th. loading 0.31 mmol
g
Ϫ1) was treated with a solution of TFA in CH2Cl2 (10% v/v;
2 cm3) for 10 min and then washed thoroughly (DMF: 6 ×
1 cm3; CH2Cl2: 6 × 1 cm3). The resin was then treated with
a solution of copper() acetate (10.8 mg, 0.06 mmol) and
pyridine (48 µl, 0.60 mmol) in acetonitrile (10 cm3). After stir-
ring for 2 h at room temperature, the resin was filtered, washed
(CH2Cl2: 3 × 2 cm3) and the resulting combined organic phases
evaporated to dryness under reduced pressure to give a thick
blue oil. The oil was purified by autopreparative HPLC and the
combined fractions lyophilised to give the MKP 24a as a white
powder (19 mg, 68%). HPLC (230 nm): tR = 5.41 min (100%);
[α]D ϩ295 (c. 0.41 in CHCl3); νmax (KBr)/cmϪ1 3360, 3209, 3090,
1675, 1543, 1368, 1165, 751 and 701; δH (400 MHz, CDCl3) 7.90
(1 H, d, J 7.5), 7.66 (2 H, m), 7.57 (1 H, m), 7.32–7.23 (3 H, m),
7.05 (7 H, m), 6.57 (1 H, br s), 4.82 (1 H, ddd, J 6.5, 5, 1), 3.77 (1
H, dd, J 15, 3), 3.34 (2 H, m), 3.22 (1 H, dd, J 6.5, 4), 3.12 (1 H,
dd, J 5, 4), 2.76 (1 H, dd, J 9.5, 7) and 2.67 (1 H, dd, J 9.5, 14);
δC (100 MHz, CDCl3) 167.5, 145.7, 134.1, 134.0, 132.1, 131.7,
130.4, 129.6, 127.9, 127.4, 127.2, 126.6, 125.4, 122.9, 117.7,
57.9, 51.3, 41.8, 39.0 and 36.3; m/z (ESI) [Found: (M ϩ H)ϩ,
466.1449. C24H24N3O5S requires MH, 466.1436], (ESI) 466
([M ϩ H]ϩ, 100%).
(3S,6R)-3,6-Dibenzyl-4-(naphthalene-2-carbonyl)piperazin-2-
one (24e). According to the general procedure for 24a, the resin
21 (95 mg, max. th. loading 0.33 mmol gϪ1) was treated with a
solution of TFA in CH2Cl2 (10% v/v), then copper() acetate
(10.8 mg, 0.06 mmol) and pyridine (48 µl, 0.60 mmol) in
acetonitrile (10 cm3) for 3 h to give, after purification by HPLC
and lyophilisation of the combined fractions, the MKP 24e as a
yellow powder (17 mg, 69%). HPLC (230 nm): tR = 5.37 min
(100%); [α]D ϩ143 (c. 0.30 in CHCl3); νmax (KBr)/cmϪ1 3242,
3062, 3029, 2932, 1673, 1640, 1421, 755 and 702; the NMR
spectrum at room temperature was very broad and showed the
presence of several stable rotamers on the NMR time scale. At
high temperature, some distinct signals could be identified
although the spectrum remained badly resolved: δH (250 MHz,
DMSO-d6, 120 ЊC) 7.95–7.83 (3 H, m), 7.62–7.54 (3 H, m),
7.32–7.15 (6 H, m), 7.16–6.93 (5 H, m), 4.88 (1 H, br s), 3.85
(1 H, br d, J 15), 3.50 (1 H, br m), 3.27 (1 H, dd, J 14, 7.5), 3.21
(1 H, d, J 15), 3.16 (1 H, dd, J 14, 6.5), 2.81 (1 H, dd, J 13.5, 6)
and 2.63 (1 H, dd, J 13.5, 8.5); m/z (ESI) [Found: (M ϩ H)ϩ,
435.2077. C29H27N2O2 requires MH, 435.2072], (ESI) 435
([M ϩ H]ϩ, 100%).
(3S,6S )-3,6-Dibenzyl-4-(2-nitrobenzenesulfonyl)piperazin-2-
one (24b). According to the general procedure for 24a, the resin
18 (185 mg, max. th. loading 0.31 mmol gϪ1) was treated with a
solution of TFA in CH2Cl2 (10% v/v), then copper() acetate
(10.8 mg, 0.06 mmol) and pyridine (48 µl, 0.60 mmol) to give,
after purification by HPLC and lyophilisation of the combined
fractions, the MKP 24b as a white powder (18 mg, 65%). HPLC
(230 nm): tR = 5.36 min (100%); [α]D ϩ148 (c 0.33 in CHCl3);
νmax (KBr)/cmϪ1 3373, 3202, 3030, 1673, 1545, 2368, 1165 and
755; δH (400 MHz, CDCl3) 7.66–7.58 (3 H, m), 7.47 (1 H, m),
7.35–7.25 (3 H, m), 7.07 (7 H, m), 6.14 (1 H, br s), 4.65 (1 H, m),
4.02 (1 H, dd, J 15, 4.5), 3.74 (1 H, m), 3.19 (1 H, dd, J 4.5,
14), 3.13 (1 H, dd, J 8, 14), 2.79 (2 H, m) and 2.36 (1 H, dd,
J 9, 13.5); δC (100 MHz, CDCl3) 169.4, 147.9, 136.6, 134.8,
(3S,6S )-3,6-Dibenzyl-4-(naphthalene-2-carbonyl)piperazin-2-
one (24f). According to the general procedure for 24a, the resin
22 (191 mg, max. th. loading 0.30 mmol gϪ1) was treated with a
solution of TFA in CH2Cl2 (10% v/v), then copper() acetate
(10.8 mg, 0.06 mmol) and pyridine (48 µl, 0.60 mmol) in
acetonitrile (10 cm3) for 3 h to give after purification by HPLC
and lyophilisation of the combined fractions, the MKP 24f as a
yellow powder (18 mg, 70%). HPLC (230 nm): tR = 5.48 min
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 7 1 1 – 1 7 1 9
1718

View Article Online
(100%); [α]D ϩ91 (c. 0.35 in CHCl3); νmax (KBr)/cmϪ1 3221,
3062, 3029, 2931, 1671, 1426, 756 and 702; the NMR spectrum
at room temperature was very broad and showed the presence
of several stable rotamers on the NMR time scale. At high
temperature, some distinct signals could be identified although
the spectrum remained badly resolved: δH (250 MHz, DMSO-
d6, 120 ЊC) 7.92–7.73 (3 H, m), 7.60–7.45 (3 H, m), 7.31–7.23
(3 H, m), 7.20–7.03 (8 H, m), 4.83 (1 H, m), 3.81 (1 H, m), 3.72
(1 H, br d, J 14.5), 3.13 (2 H, m), 2.89 (1 H, dd, J 15, 5), 2.74
(1 H, dd, J 14, 10.5) and 2.54 (1 H, dd, J 14, 8.5); m/z (ESI)
[Found: (M ϩ H)ϩ, 435.2077. C29H27N2O2 requires MH,
435.2072], (CI) 435 ([M ϩ H]ϩ, 100%).
References
1 R. E. Dolle, J. Comb. Chem., 2001, 3, 477.
2 F. Guillier, D. Orain and M. Bradley, Chem. Rev., 2000, 100,
2091.
3 T. Fukuyama, C.-K. Jow and M. Cheung, Tetrahedron Lett., 1995,
36, 6373.
4 P. J. Murray, M. Kranz, M. Ladlow, S. Taylor, F. Berst, A. B.
Holmes, K. N. Keavey, A. Jaxa-Chamiec, P. W. Seale, P. Stead,
R. J. Upton, S. L. Croft, W. Clegg and M. R. J. Elsegood,
Bioorg. Med. Chem. Lett., 2001, 11, 773.
5 M. Patek and M. Lebl, Biopolymers, 1999, 47, 353.
6 D. L. Marshall and I. E. Liener, J. Org. Chem., 1970, 35, 867.
7 P. P. Fantauzzi and K. M. Yager, Tetrahedron Lett., 1998, 39,
1291.
(3S,6S )-3,6-Dibenzyl-1-(4-methylbenzyl)-4-(2-nitrobenzene-
sulfonyl)piperazin-2-one (24g). (Chelex work-up). According
to the general procedure for 24a, the resin 23 (151 mg, max. th.
loading 0.30 mmol gϪ1) was treated with a solution of TFA in
CH2Cl2 (10% v/v), then copper() acetate (7.6 mg, 42 µmol) and
pyridine (34 µl, 0.42 mmol) in acetonitrile (10 cm3) for 70 min.
Chelex ion exchange resin (ca. 1 g, prewashed) was added and
then, after 3 h, the resins were filtered and washed (CH2Cl2:
3 × 5 cm3, MeOH: 3 × 5 cm3) to give after purification by
HPLC, the MKP 24g as a yellow powder (8.2 mg, 34%). HPLC
(230 nm): tR = 6.71 min (100%); [α]D ϩ65 (c. 0.29 in MeOH);
νmax (KBr)/cmϪ1 3292, 3029, 2922, 2853, 1649, 1545, 1454, 1370,
1167 and 737; δH (400 MHz, CDCl3) 7.59–7.55 (2 H, m), 7.52
(1 H, m), 7.42 (1 H, m), 7.25 (3H, m), 7.14–1.05 (9H, m), 6.90–
6.87 (2H, m), 5.48 (1 H, d, J 15), 4.63 (1 H, m), 4.11 (1 H, d,
J 15), 3.47 (2 H, m), 3.30 (1 H, dd, J 13.5, 4.5), 3.15 (1 H, dd,
J 14, 8.5), 2.96 (1 H, dd, J 13.5, 8), 2.77 (1 H, dd, J 14, 3.5), 2.32
(3 H, s) and 2.25 (1 H, dd, J 14, 8.5); δC (100 MHz, CDCl3)
166.0, 147.8, 135.3, 134.3, 133.5, 131.5, 130.5, 129.8, 129.7,
129.1, 127.8, 127.4, 127.1, 126.8, 126.2, 126.0, 125.2, 124.9,
122.4, 58.9, 52.5, 44.4, 41.6, 35.5, 34.4 and 19.1; m/z (ESI)
570 ([M ϩ H]ϩ, 100%), (ESI) [Found: (M ϩ H)ϩ, 570.2049.
C32H32N3O5S requires MH, 570.2062].
8 G. W. Kenner, J. R. McDermott and R. C. Sheppard, J. Chem. Soc.,
Chem. Commun., 1971, 636; B. J. Backes and J. A. Ellman, J. Am.
Chem. Soc., 1994, 116, 11171.
9 C. R. Millington, R. Quarrell and G. Lowe, Tetrahedron Lett., 1998,
39, 7201.
10 F. Stieber, U. Grether and H. Waldmann, Angew. Chem., Int. Ed.,
1999, 38, 1073; T. Wieland, J. Lewalter and C. Birr, Liebigs Ann.
Chem., 1970, 740, 31.
11 F. Berst, A. B. Holmes, M. Ladlow and P. J. Murray, Tetrahedron
Lett., 2000, 41, 6649.
12 M. Di Fonzo and C. Saracini, Farmaco, 1958, 13, 639.
13 J. F. Reichwein and R. M. J. Liskamp, Tetrahedron Lett., 1998, 39,
1243–1246.
14 A. F. Abdel-Magid, K. G. Carson, B. D. Harris, C. A. Maryanoff
and R. D. Shah, J. Org. Chem., 1996, 61, 3849.
15 Removal of copper salts during work-up could be readily achieved
either by acidic work-up, by the addition of Chelex ion-exchange
resin and/or preparative reverse phase HPLC.
16 Complexation of related diketopiperazines by Cu() has been
reported and this may result in an increased propensity towards
hydrolysis: G. R. Brubaker and N. Y. Sakkab, Bioinorg. Chem., 1974,
3, 243; H. Zajworoniuk and F. Karczynski, Chem. Anal., 1983, 28,
657.
17 C. Kay, O. E. Lorthioir, N. J. Parr, M. Congreve, S. C. McKeown,
J. J. Scicinski and S. V. Ley, Biotechnol. Bioeng., 2001, 71,
110.
18 Fmoc quantification was performed according to the following
procedure: A sample of dried resin (ca 5–10 mg) was weighed into a
volumetric flask and sonicated with 20% v/v piperidine–DMF for
15 mins. The supernatant was sampled and the absorbance
Acknowledgements
measured at 302 nm. Resin loading (mmol gϪ1) = (vol (cm3) × Abs302
(7.8 × wtresin (mg)).
/
We are grateful to GlaxoSmithKline for a fully funded student-
ship (to F. B.).
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 7 1 1 – 1 7 1 9
1719