Synthesis and binding properties of a macrocyclic peptide receptor

Article abstract of DOI:10.1002/(SICI)1521-3765(19990104)5:1<79::AID-CHEM79>3.0.CO;2-W

Macrocyclic receptor 1 has been synthesised, as a racemate and as a single enantiomer, utilising a Stille coupling for the formation of the biphenyl portion and a macrolactamisation as the final step. The binding properties for the racemic and the homochi

Full text of DOI:10.1002/(SICI)1521-3765(19990104)5:1<79::AID-CHEM79>3.0.CO;2-W

FULL PAPER
Synthesis and Binding Properties of a Macrocyclic Peptide Receptor
James Dowden,[a] Peter D. Edwards,[b] Stephen S. Flack,[a] and Jeremy D. Kilburn*[a]
Abstract: Macrocyclic receptor 1 has
been synthesised, as a racemate and as a
single enantiomer, utilising a Stille cou-
pling for the formation of the biphenyl
portion and a macrolactamisation as the
final step. The binding properties for the
racemic and the homochiral macrocycle
with amino acid and dipeptide deriva-
tives, in CDCl3 solution, have been
investigated. In the case of racemic 1,
addition of homochiral peptide sub-
strates led to two distinct diastereomeric
complexes, and the well separated sig-
for the two diastereoisomeric com-
plexes, and indicating that 1 is capable
of enantioselective recognition. Titra-
tion of homochiral 1 with the same
peptide substrates allowed the sense of
the enantioselectivity to be determined,
and experiments with a greater range of
substrates indicated that 1 is particularly
effective for the recognition of N-Cbz-b-
alanyl-l-amino acids, the strongest bind-
ing being observed with N-Cbz-b-alanyl-
tably the binding of N-Cbz-b-alanyl-l-
lactic acid was considerably weaker
1
(
DGass ˆ 13.1 kJmol ), presumably
due to replacement of an NH hydro-
gen-bond donor in the case of N-Cbz-b-
alanyl-l-alanine with an oxygen lone-
pair in the case of N-Cbz-b-alanyl-l-
lactic acid. Molecular modelling and 2D
NMR studies on the free macrocycle 1
and associated complexes did not pro-
vide conclusive evidence for the struc-
ture of the host ± guest complexes, but
did serve to emphasise the flexibility of
1, which despite this flexibility, shows
strong, selective binding of certain pep-
tide guests.
1
l-alanine ( DGass ˆ 19.9 kJmol ). No-
1
nals for several protons in the H NMR
Keywords: host ± guest chemistry ´
macrocycles ´ peptides ´ receptors
´ synthetic methods
spectrum could be conveniently fol-
lowed in titration experiments, allowing
determination of both binding constants
for example, to new biosensors, therapeutics and catalysts for
peptide hydrolysis. In our own efforts to develop novel
receptors for amino acids and peptides we have targeted
peptidic guests with a carboxylic acid terminus by preparing a
range of macrocyclic structures which feature a specific
binding site for the carboxylic acid and additional hydrogen
bonding functionality to bind to the backbone of the peptidic
guest.[3] Macrocycle 1 is an example of such a receptor, which
incorporates a diamidopyridine unit to serve as a carboxylic
acid binding site,[4] in a cavity lined on one side with an amide
functionality, and on the other with a rigid biaryl fragment,
linked by a benzyl ether to a further aromatic ring, to hold
open the binding cavity and, in principle, to provide a site for
p-stacking interactions or hydrophobic interactions in water-
soluble variants of the basic structure.
Introduction
Over the last thirty years the rapid development of host ±
guest chemistry has led to the development of synthetic
receptors for a whole range of substrates, including many of
biological importance, such as nucleotides, carbohydrates and
amino acids.[1] The study of such synthetic model systems aids
the understanding of fundamental aspects of non-covalent
interactions and in many cases inspires the design of new
functional molecules and supramolecular systems. The devel-
opment of synthetic receptors for peptides and amino acid
derivatives[2] is of particular interest because the intermolec-
ular interactions involved in complexes between small
molecules and peptides are of direct relevance to many
biological peptide ± protein interactions, and may also lead,
[a] Dr. J. D. Kilburn, J. Dowden, S. S. Flack
Department of Chemistry
University of Southampton
Southampton SO17 1BJ (UK)
Fax: (‡44)1703-593781
E-mail: jdk1@soton. ac.uk
[b] Dr. P. D. Edwards
Department of Medicinal Chemistry
SmithKline Beecham Pharmaceuticals
New Frontiers Science Park (North)
Third Avenue
Harlow CM19 5AW (UK)
Chem. Eur. J. 1999, 5, No. 1
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0501-0079 $ 17.50+.50/0
79

J. D. Kilburn et al.
FULL PAPER
Diacid 11 was assembled by an initial Mitsunobu coupling[6]
of commercially available alcohols 2 and 3 (see Scheme 1),
and subsequent reduction of nitrile 4 by using borane ± di-
methyl sulfide complex[7] to give amine 5. Alternative
methods for reduction of the nitrile (LiAlH4, H2/Raney
nickel, NaBH4/AlCl3) were less successful, since cleavage of
the benzyl ether occurs as a competing side reaction. Acid 6
was simply prepared by a base-mediated opening of succinic
anhydride with methyl phenylalaninate, and was coupled to
amine 5 using dicyclohexyl carbodiimide (DCC) to give aryl
bromide 7 in 85% yield. Formation of the biaryl unit then
required coupling of aryl bromide 7 with a suitable organo-
metallic reagent derived from bromide 8. Thus we prepared
stannane 9 by treatment of bromide 8 with bis(tributyltin) and
a palladium catalyst.[8] The best yields for this reaction (72%)
were ultimately achieved by using an excess of bis(tributyltin),
(Ph3P)4Pd as catalyst, and carrying out the reaction under
highly concentrated conditions that minimised unwanted
biaryl formation. With the required stannane 9 in hand, the
Stille coupling[9] with bromide 7 using (Ph3P)4Pd and a
stoichiometric amount of silver(i) oxide[10] gave the diester
10 in reasonable yields (50% by NMR analysis of the crude
material). Purification of the diester proved to be problem-
atic, and it was instead hydrolysed directly to give the
corresponding diacid 11 in 24% overall yield from bromide 7.
Numerous attempts were made to couple diacid 11 with
diaminopyridine. Formation of the bis(acid chloride) was
attempted by using oxalyl chloride, thionyl chloride or
cyanuric chloride[11] but failed to yield any of the desired
product in the subsequent coupling with diaminopyridine.
Similarly all efforts with in situ activating agents, such as
dicyclohexylcarbodiimide (DCC; with or without auxiliary
reagents), carbonyldiimidazole and 2-(1H-benzotriazole-1-
yl)-1,1,3,3-tetramethyluronium tetrafluoroborate (TBTU)[12]
were in vain. The use of the bis(acylhydrazide) or bis(acti-
vated esters),[13] such as the bis(pentafluorophenol ester), also
failed to give the desired macrocycle.
We envisioned that variation of the amide-containing side
wall of the macrocycle would provide a means of tuning the
binding selectivity for specific peptide guests, but initially we
wished to prepare a relatively simple example of these
structures and establish that it contained appropriate basic
features for peptide recognition. Thus we chose to prepare
macrocycle 1, in which the amide side wall is made up of a
succinic acid moiety linked to a phenylalanine unit; the latter
unit provides a single chiral centre for the structure and thus
the possibility of some stereoselective binding properties. In
this paper we describe our synthetic efforts to prepare
macrocycle 1 in both racemic and chiral forms, and discuss
the binding properties of the macrocycle with a range of
peptidic guests.[5]
Results and Discussion
Synthesis: Our first attempts to prepare macrocycle 1
involved the synthesis of diacid 11 (Scheme 1) which we
hoped to activate and couple in one step to diaminopyridine.
This route would be short and allow easy variation of the
amide side wall in later variants.
O
OH
OH
DEAD, Ph3P
THF (78%)
R
+
Br
Br
CN
2
3
BH3.Me2S
THF, reflux
(94%)
4 R = CN
5 R = CH2NH2
O
O
O
Cl- H3N+
CO2Me
H
Et3N,
HO2C
N
CO2Me
Ph
+
CH2Cl2
(89%)
Ph
O
6
The failure of this approach led us to consider an alternative
synthesis which would culminate in the macrolactamisation of
amino acid derivative 21 (see Scheme 3), which in turn could
be assembled, with appropriate protecting groups, by sequen-
tial coupling of a suitable peptidic fragment and a biaryl acid
fragment with diaminopyridine.
The peptidic fragment was prepared, as the allyl ester 14
(Scheme 2), by reaction of succinic anhydride with allyl
alcohol, followed by conversion of monoacid 12[14] to the acid
chloride 13, which, in turn, was used to acylate phenylalanine.
5, DCC,
HOBT, DMF
(85%)
Bu3Sn
Br
Bu3SnSnBu3,
10 mol %
Pd(PPh3)4,
toluene (76%)
O
H
N
MeO2C
CO2Me
O
H
8
9
O
Br
10 mol %
O
N
Pd(PPh3)4,
Ag2O, DMF,
50˚C
N
A
PyBOP- (PyBOP ˆ benzotriazol-1-yloxy-tripyrrolidino-
H
phosphonium hexafluorophosphate[15] or DCC-mediated cou-
pling of acid 14 to 2,6-diaminopyridine gave the desired
monoacylated product 15, but with complete racemisation of
the phenylalanine derived chiral centre, presumably via an
oxazolone intermediate.[16] In retrospect the formation of 15
as a racemate is not surprising given that 14 is an acylated
amino acid (and therefore prone to racemisation on activa-
tion) and diaminopyridine is a poor nucleophile.
Ph
MeO2C
O
H
7
O
N
Ph
RO2C
CO2R
LiOH,
H2O,
dioxan
10 R = Me
11 R = H
(24% overall
from 7)
Racemisation was avoided in the subsequent synthesis of
the enantiomerically pure macrocycle by first coupling 2,6-
diaminopyridine to N-tert-butoxycarbonyl-l-phenylalanine to
Scheme 1. Synthesis of diacid 11.
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
80
0947-6539/99/0501-0080 $ 17.50+.50/0
Chem. Eur. J. 1999, 5, No. 1

Macrocyclic Peptide Receptor
79±89
O
H2N
N
NH2
O
O
O
ROC
NHBoc
NHBoc
COCl)2,
cat. DMF,
CH2Cl2
12 R = OH
13 R = Cl
9, 13 mol %
(Boc)2O,
Boc-Phe
5
DCC, HOBT
CH2Cl2/DMF
(10:1, v/v)
(76%)
Pd(PPh3)2Cl2,
toluene, 80˚C
(63%)
Et3N,
CH2Cl2
(81%)
Br
phenylalanine
1M Na2CO3
(63%)
17
O
NHBoc
H
N
H
N
COR
HO2C
Ph
H2N
N
O
18 R = OMe
19 R = OH
20 R = F
LiOH, H2O,
dioxan
(92%)
cyanuric fluoride
pyridine, CH3CN
(96%)
O
Ph
O
16
14
15, N-methyl-
1. TFA/CH2Cl2
(1:1, v/v)
2. 13, DIPEA
CH2Cl2
diaminopyridine
PyBOP, DIPEA,
DMF (40%)
morpholine
THF, reflux
(62%)
O
O
O
O
O
NHR1
O
OR2
O
slow addition of 24 to
DMAP, DIPEA,
CH3CN, reflux
H N
H
H N
H
O
H2N
N
N
H2N
N
N
1
(25-35% overall from
21)
Ph
O
Ph
O
O
H N
O
H
N
H
N
15 (racemic)
(S)-15 (72% from 16)
N
Scheme 2. Synthesis of monoacylated product 15.
Ph
O
21 R1 = Boc, R2 = allyl
22 R1 = Boc, R2 = H
10 mol % Pd(PPh3)4,
pyrrolidine, CH2Cl2
C6F5OH, DCC,
cat. DMAP,
CH2Cl2
give 16, followed by removal of the amine protecting group,
and selective acylation of the more reactive amino function-
ality with acid chloride 13 (Scheme 2).
23 R1 = Boc, R2 = C6F5
4M HCl,
dioxan
24 R1 = H2+ Cl-, R2 = C6F5
Biaryl ester 18 was prepared by Stille coupling of stannane
9 with the protected amino bromide 17 (Scheme 3). Initial
attempts at this reaction under Pd(PPh3)4 or Pd(PPh3)2Cl2
catalysis in N-methylpyrollidinone gave poor yields of the
desired biaryl ester 18 (30 ± 40%). The use of more polar
solvents,[9] copper iodide co-catalysis[17] and softer ligands
such as triphenylarsane[18] gave no improvement. Variations of
the Pd2(dba)3 catalytic system developed by Farina and
Krishman[19] gave none of the desired product with apparent
rapid decomposition of the catalyst. However, when the
reaction was performed in a less polar solvent such as toluene
with Pd(PPh3)2Cl2 as catalyst, the yield of the biaryl ester
improved to an acceptable and reproducible level of 60 ±
65%. The ester 18 was readily hydrolysed to give the required
acid 19 in good yield.
Coupling of the relatively unreactive monoacylated diami-
nopyridine 15 (as the racemate or as a single enantiomer) with
biaryl acid 19 could be achieved in moderate yield (30 ± 40%)
by using either DCC or 2-ethoxy-N-ethoxycarbonyl-1,2-dihy-
droquinoline (EEDQ) in refluxing THF. Better yields how-
ever, were achieved by converting acid 19 to the acid fluoride
20[20] and refluxing this with 15 and N-methylmorpholine in
THF to give the protected macrocyclisation precursor 21 in
62% yield (Scheme 3). The allyl ester was deprotected to give
acid 22 by using (Ph3P)4Pd in dioxan, with water or pyrrolidine
as the allyl scavenger.[21] Subsequent DCC coupling with
pentafluorophenol and removal of the amine protecting
group by treatment with 4m HCl in dioxan, followed by
trituration with diethyl ether, provided the hydrochloride salt
24. The crude hydrochloride salt, dissolved in DMF, was
Scheme 3. Synthesis of 1.
added by syringe pump to a refluxing solution of diisopropyl-
ethylamine (DIPEA) in acetonitrile to give the desired
macrocycle (as a racemate or as a single enantiomer) in 25 ±
35% yield from the allyl ester 21.
Macrocycle 1 was isolated as an amorphous solid which
resisted all our attempts to obtain single crystals for X-ray
analysis. It is sparingly soluble in relatively apolar solvents
such as chloroform (ꢀ2 mg per mL), and the analysis of 1 by
1H NMR spectroscopy (in CDCl3) gave a well resolved
spectrum which could be fully assigned with the help of 2D
NMR experiments. A 2D ROESY[22] spectrum revealed no
NOEs between protons on opposite sides of the macrocyclic
ring, suggesting that 1 exists in an open conformation in
CDCl3 solution, as desired.
Binding studies and molecular modelling
In order to determine the extent to which the receptor was
able to bind amino acids and peptides with a free carboxylic
acid terminus, as it was designed to do, a series of binding
studies with a range of carboxylic acid substrates was carried
out. All binding studies with macrocycle 1 were carried out in
deuteriochloroform (in which hydrogen bonds can be ex-
pected to provide the dominant binding interactions), using a
standard NMR titration experiment, monitoring the shift of
the NH signals, and analysing the resultant binding curves
using the Hostest programme.[23] In each titration experiment
significant downfield shifts of NHb were observed with no
Chem. Eur. J. 1999, 5, No. 1
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0501-0081 $ 17.50+.50/0
81

J. D. Kilburn et al.
FULL PAPER
1
apparent shift of NHa, consistent with a strong association
between the carboxylic acid and the amidopyridine moiety,
presumably involving NHb and not NHa. (Addition of
corresponding methyl esters to a solution of 1 led to no
significant changes in the NMR spectrum for 1 further
confirming the importance of the interaction between the
carboxylic acid and the amidopyridine in the observed
binding). In each binding experiment a 1:1 binding stoichi-
ometry has been assumed which was generally supported by
the good fit of the measured data to the theoretical model, on
analysis. (A Job plot[24] was attempted to confirm the 1:1
binding stoichiometry, but the low solubility of 1 meant that
acceptable NMR spectra could not be obtained across a
suitable concentration range). Dilution experiments were also
carried out on both the receptor 1 and on the guest molecules
to provide estimates of dimerisation constants for each
compound. No appreciable dimerisation or aggregation was
detected for 1 and although dimerisation constants were
measurable for the dipeptide substrates these were sufficient-
ly small (0 ± 30m 1) that inclusion of these numbers did not
affect the measured binding constants with receptor 1 within
the margin of error for the measurements. As would be
glycine ( DGass ˆ 14.4 kJmol ) showed a more pronounced
increase. When the racemic receptor was titrated with
homochiral peptide substrates two distinct diastereomeric
1
complexes were immediately evident in the H NMR spec-
trum, and the well separated signals for NHb of the two
complexes (and to a lesser extent for NHc and NHd) could be
conveniently monitored throughout the titration experiment.
Analysis of these data using the Hostest software[23] gave
estimates of the two binding energies for the diastereomeric
complexes. Thus, titration of racemic 1 with N-Cbz-l-alanine
1
gave DGass ˆ 15.8 and 15.2 kJmol for the two diastereo-
meric complexes (which was essentially mirrored, as expect-
ed, by titration with N-Cbz-d-alanine) and titration with N-
1
Cbz-l-phenylalanine gave DGass ˆ 14.2 and 13.7 kJmol
for the two diastereomeric complexes.
Titration of racemic 1 with the dipeptide N-Cbz-l-alanyl-l-
alanine showed no significant increase in binding over simple
amino acid derivatives for either of the two diastereomeric
complexes formed, and N-Cbz-b-alanylglycine similarly
1
showed only weak binding ( DGass ˆ 15.0 kJmol ) (al-
though the measured binding constant of the latter must be
treated with some caution as the poor solubility of the
substrate in CDCl3 meant that only 49% saturation was
achieved in the titration experiment). However, titration of 1
with N-Cbz-b-alanyl-l-alanine gave two significantly different
binding energies for the two diastereomeric complexes
anticipated,[25] for the more strongly bound substrates
1
(
DGass > 17 kJmol ), the titration data showed a poorer
fit to the theoretical model and were therefore repeated
several times to give an average value. Repeat experiments
did, however, give consistent results and the comparison
between data obtained for the homochiral macrocycle and
that obtained for the racemate was also good (vide infra).
Furthermore, titration of receptor 1 with strongly binding
substrates in CDCl3 contaminated with '2% MeOH, gave, as
expected, reduced binding energies, but now the data again
gave an excellent fit for the assumed 1:1 binding.
1
1
(
DGass ˆ 19.4 and 16.2 kJmol , DDG ˆ 3.2 kJmol ), rep-
resenting a binding enantioselectivity of ꢀ80:20, although the
sense of the enantioselectivity could not be determined from
these experiments).
1
The H NMR of a 1:1 complex between racemic 1 and N-
Cbz-b-alanyl-l-alanine (7.5mm in CDCl3) also showed the
most significant differences between the diastereomeric
complexes so formed (as compared to 1:1 complexes with
any other substrate) (Figure 1).
Thus NHb shifted downfield (relative to uncomplexed 1)
from d ˆ 7.97 to d ˆ 8.79 and 8.28, respectively, for the two
Macrocycle 1 was at first obtained as a racemate, but was
none-the-less used in initial binding studies (Table 1).
Titration of racemic
1
with phenylacetic acid gave
1
DGass ˆ 11.5 kJmol which places an upper limit on the
strength of the amidopyridine ± carboxylic acid interaction.[26]
1
N-Cbz-b-alanine ( DGass ˆ 12.9 kJmol ) showed a modest
increase in binding over phenylacetic acid, while N-Cbz-
Table 1. Binding constants (Kass) and free energies of complexation[a]
(
DGass) for the 1:1 complexes formed between racemic macrocycle 1
and various acid substrates in CDCl3 at 208C.
1
[b]
1
Substrate
Kass [m
]
DGass
[kJmol
]
phenylacetic acid
Cbz-Gly-OH
113
376
11.5
14.4
Cbz-l-Ala-OH
Cbz-d-Ala-OH
Cbz-l-Phe-OH
tBoc-l-Ala-OH
Cbz-b-Ala-OH
665/505
608/492
344/279
335/275
202
15.8/15.2
15.6/15.2
14.2/13.7
14.2/13.7
12.9
Cbz-l-Ala-l-Ala-OH
Cbz-b-Ala-l-Ala-OH
Cbz-b-Ala-Gly-OH
723/405
2920/781
313
16.0/14.6
19.4/16.2
15.0
[a] Where two figures are reported these refer to the binding constants or
binding energies for the two diastereomeric complexes formed (see text).
[b] Errors for binding energies were estimated as 0.5 ± 0.9 kJmol for the
Figure 1. 1H NMR spectra of the aromatic/NH region, in CDCl3 of:
a) Macrocycle 1; b) Racemic macrocycle 1 ‡ N-Cbz-b-alanyl-l-alanine;
c) homochiral macrocycle 1 ‡ N-Cbz-b-alanyl-l-alanine
1
substrates listed.
82
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0501-0082 $ 17.50+.50/0
Chem. Eur. J. 1999, 5, No. 1

Macrocyclic Peptide Receptor
79±89
diastereomeric complexes. NHc shifted downfield from d ˆ
6.25 to d ˆ 6.84 and 6.52, respectively, and NHd shifted
downfield from d ˆ 5.95 to d ˆ 6.21 and 6.09, respectively. The
position of the signal for NHa, however, again appeared to be
unaffected by the addition of the substrate. Essentially all
side chain incorporated into the dipeptide, the binding for N-
Cbz-b-alanyl-l-valine was of a similar order of magnitude
1
(
DGass ˆ 16.8 kJmol ) to N-Cbz-b-alanyl-l-phenylalanine,
but the binding energy for the enantiomer N-Cbz-b-alanyl-d-
1
valine was much lower ( DGass ˆ 12.7 kJmol ), in fact
1
other signals in the H NMR spectrum were separated and
reduced to a value only slightly larger than that for phenyl-
1
could be assigned to the two diastereomeric complexes. (Thus
macrocycle 1 may serve as a chiral shift reagent for the
1H NMR spectrum of such peptidic substrates, although
perhaps not a very practical one!).
With homochiral receptor 1 in hand we were able to extend
the binding studies and determine the sense of the enantio-
selective binding initially observed with the racemic receptor.
Titration experiments were carried out with homochiral 1 and
a range of dipeptide substrates (Table 2).
acetic acid. The enantioselectivity (DDG ˆ 4.1 kJmol , ee ꢀ
70%) for the N-Cbz-b-alanylvaline dipeptides is quite high
and although it does not compare with the high enantiose-
lectivities observed for other recently described peptide
receptors,[2b] it is notable for such a flexible receptor that
bears only one chiral centre.
From these results it would appear that the presence of a
methyl group attached to the chiral centre adjacent to the
carboxylic acid receives more stabilisation from complexation
than other substituents, reflected by the higher levels of
binding for all alanyl containing peptides compared to other
substrates, and in particular the strong binding of N-Cbz-b-
alanyl-l-alanine in comparison to the binding of N-Cbz-b-
Table 2. Binding constants (Kass) and free energies of complexation
(
DGass) for the 1:1 complexes formed between homochiral macrocycle
1 and dipeptide substrates in CDCl3 at 208C.
1
alanylglycine (DDG ˆꢀ 5 kJmol ). In the 1H NMR spectrum
1
Substrate
Kass [m
]
DGass
1
[kJmol
]
of the 1:1 complex between macrocycle 1 and N-Cbz-b-alanyl-
l-alanine the alanine methyl signal was shifted upfield from
its unbound position at d ˆ 1.44 to d ˆ 1.28 in the complex,
which could be due to a weak interaction between the methyl
group and the biphenyl unit.
Cbz-l-Ala-l-Ala-OH
Cbz-d-Ala-d-Ala-OH
Cbz-b-Ala-l-Ala-OH
Cbz-b-Ala-d-Ala-OH
Cbz-b-Ala-l-Phe-OH
Cbz-b-Ala-d-Phe-OH
Cbz-b-Ala-l-Val-OH
Cbz-b-Ala-d-Val-OH
Cbz-b-Ala-l-Lac-OH
Cbz-b-Ala-d-Lac-OH
1940 Æ 485
869 Æ 124
3598 Æ 900
925 Æ 230
321 Æ 82
18.4 Æ 0.7
16.5 Æ 0.3
19.9 Æ 0.7
16.6 Æ 0.7
16.2 Æ 0.5
14.0 Æ 0.5
16.8 Æ 0.6
12.7 Æ 0.2
13.1 Æ 0.3
11.2 Æ 0.2
We also measured the binding of the ester-linked N-Cbz-b-
alanyllactic acids as substrates for receptor 1, effectively
replacing a hydrogen bond donor substituent (NH) with a
hydrogen bond acceptor substituent (lone pair of electrons on
oxygen) in the guest structure. The binding energies for both
enantiomers of N-Cbz-b-alanyllactic acid ( DGass ˆ 13.1,
828 Æ 143
995 Æ 364
186 Æ 15
221 Æ 24
100 Æ 10
1
11.2 kJmol ) are comparable to the binding energy measured
Whereas titration of such guests with the racemic macro-
cycle had led to the formation of two diastereomeric
1
for phenyl acetic acid ( DGass ˆ 11.5 kJmol ) indicating that
1
these substrates apparently gain little stabilisation other than
from the carboxylic acid interaction with the amidopyridine
binding site. The substantially lower binding energies for
these esters compared to the corresponding N-Cbz-b-alanyl-
alanine dipeptides, strongly suggests that a hydrogen bond
from the guest NH to the receptor side wall is a key
interaction in the binding of the latter.[27] This result is also
of interest in view of recent observations that certain bacteria
have mutated their cell wall precursor structure from a
terminal d-Ala-d-Ala-CO2H sequence to a d-Ala-d-Lac-
CO2H sequence, thus providing immunity to conventional
antibacterials such as the natural d-Ala-d-Ala-CO2H receptor
vancomycin.[28, 29]
complexes, clearly distinguishable in the H NMR spectra,
we now observed formation of a single diastereomeric
complex (see Figure 1), allowing independent determination
of binding energies for each enantiomer of the various guests,
and incidentally confirming that no racemisation had occur-
red in the modified synthesis of 1. As before, in each titration
experiment significant downfield shifts of NHb were observed
with no apparent shift of NHa.
Initially, identical dipeptide substrates to those explored
with racemic 1 were examined. The results compared well
with those obtained for racemic 1 with the best agreement
observed for the N-Cbz-b-alanylalanine enantiomeric pair,
which were bound with very similar levels of enantioselectiv-
ity (ee ꢀ 60%) to those observed in the racemic experiment.
The overall levels of binding were also in good agreement, N-
Cbz-b-alanyl-l-alanine bound with a free energy of binding of
The 1:1 complex of N-Cbz-b-alanyl-l-alanine with both
racemic and homochiral 1, in chloroform, was studied by 2D
ROESY and NOESY NMR spectroscopy in an attempt to
identify which intermolecular interactions were contributing
to complexation. No useful intermolecular NOEs, however,
were observed in any of these experiments. One possible
explanation for this is that, due to the flexibility of the
macrocycle and guest peptides a large number of conforma-
tions are occupied that are suitable for binding and it is
therefore, impossible to observe any intermolecular interac-
tions. If this is correct then the level of selectivity observed for
some of the substrates is surprisingly good (in general higher
selectivity is anticipated with less flexible receptors).[2]
1
DGass ˆ 19.9 kJmol , which compared favourably with the
strongest diastereomeric complex in the racemic experiment
1
(
DGass ˆ 19.4 kJmol ).
In order to probe the relatively high enantioselective
binding of the N-Cbz-b-alanylalanine dipeptides the binding
of related substrates was studied. Thus with the N-Cbz-b-
alanylphenylalanine dipeptides as guests, binding energies
1
were somewhat lower ( DGass ˆ 16.2, 14.0 kJmol ), with
enantioselectivity also being slightly reduced (DDG ˆ
1
2.2 kJmol , ee ꢀ 44%). With the more sterically bulky valine
Chem. Eur. J. 1999, 5, No. 1
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0501-0083 $ 17.50+.50/0
83

J. D. Kilburn et al.
FULL PAPER
Molecular modelling was also carried out in an attempt to
visualise the structure of the complex. Thus the geometry of
the free macrocycle 1 and its complexes with dipeptide
substrates were examined by using a combination of simu-
lated annealing calculations and molecular dynamics using the
MacroModel[30] program. The AMBER*[31] force field was
used[32] as implemented in MacroModel V5.0 and the effect of
solvent was included through the use of GB/SA chloroform
model.[33] Initially, the free macrocycle was energy minimised
by using a conjugate gradient (PRCG) and five simulated
annealing calculations of 1 ns, involving slow cooling from 600
to ꢀ0 K, were performed, followed by 5 Â 1 ns calculations of
molecular dynamics at 300 K to examine the behaviour at
room temperature.
The results of the simulation were not conclusive as a large
number of conformations were generated in the course of the
simulation, reflecting the flexibility of the macrocycle, but it
was possible to draw some general conclusions. None of the
structures generated showed any intramolecular hydrogen
bonds and most of the conformations featured both of the
pyridyl-amide NH groups of the diamidopyridine binding site
pointing into the macrocyclic cavity. Furthermore, in all of the
structures the aromatic units provided an open, box shaped
binding cavity, as they were designed to do, and they could be
seen to effectively stretch out the amide side wall, holding it
apart from the carboxylic acid binding site. The most flexible
portion of the macrocycle was the amide side wall which
adopted many conformations during the calculation and as a
consequence it was not possible to identify a preferred
orientation for the benzyl residue of the phenylalanine
moiety. The five structures produced at the end of the
successive 1 ns molecular dynamics runs at 300 K are shown in
Figure 2, and reflect the open binding cavity, the consistent
orientation of the amidopyridine unit, and the flexibility of
the amide portion.
Dipeptide guests were docked by eye into the binding
cavity of the final structures from the molecular dynamic
calculations, such that the carboxylic acid carbonyl formed a
hydrogen bond with the pyridyl-amide NHb which was
justified by the earlier NMR experiments that showed that
only this pyridyl-amide NH group is involved in binding. This
single hydrogen-bonding interaction was constrained at a
reasonable hydrogen-bonding distance (2.0 Š) throughout all
future calculations, thus tethering the guest to the binding site,
although this was the only bias placed on the system. The
same simulated annealing and molecular dynamics protocol
was used as before to examine the conformation of the
complex between the macrocycle 1 and dipeptide substrates.
Now an even greater diversity of structures was produced
for the conformation of the complexes, in the course of the
simulation. Both the macrocycle and the dipeptide guests
were very flexible and did not show any clear preference for a
single binding geometry. However, complexation of the
carboxylic acid to the diamidopyridine unit through two
hydrogen bonds in an eight-membered ring arrangement was
commonly observed, and this generally preferred orientation
for carboxylic acid binding placed the rest of the dipeptide
guest in a position close to the amide side wall of the
macrocycle. Further hydrogen bonding interactions between
the backbone of the guest and the amide side wall of the
receptor frequently featured in the structures generated by
the molecular modelling, but no specific hydrogen bonds,
other than the constrained one, could be consistently identi-
fied. The structure for the complex of macrocycle 1 with N-
Cbz-b-alanyl-l-valine shown in Figure 3 is an illustrative
example.
From the modelling it was not possible to provide a
rationalisation for the measured enantioselective binding
properties of macrocycle 1. Similarly it was not possible to
Figure 3. Complex between macrocycle
1 and Cbz-b-alanyl-l-valine
Figure 2. Structures of macrocycle 1 generated at the end of each of five
successive molecular dynamics calculations (1 ns at 300 K).
illustrating possible H-bonding interactions that may be formed (all H
atoms, except those in NH groups, have been omitted for clarity).
84
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0501-0084 $ 17.50+.50/0
Chem. Eur. J. 1999, 5, No. 1

Macrocyclic Peptide Receptor
79±89
53.3 mmol), 4-cyanophenol (3) (6.32 g, 53.3 mmol) and triphenylphos-
phane (16.79 g, 64.0 mmol) in dry, freshly distilled THF (80 mL) at 08C
under nitrogen. The reaction mixture was gradually allowed to warm to
room temperature and stirred overnight. The reaction was concentrated in
vacuo to give an orange gum and trituration with MeOH gave a white
powder which was recrystallised from MeOH to give nitrile 4 (11.68 g,
78%), m.p. 130 ± 1318C (MeOH); IR (CHCl3): nÄ ˆ 2225, 1605, 1574, 1507,
elucidate how the selectivity for alanyl amides over lactate
esters may arise. The results do, however, support the
hypothesis that the macrocycle is very flexible, and indicate
that in addition to a strong interaction between the carboxylic
acid and the diamidopyridine unit, additional hydrogen
bonding interactions between the backbone of the guest and
the amide side wall of the receptor are perfectly feasible,
consistent with the observed stronger binding of N-Cbz-b-
alanyl-l-amino acid guests compared to simple amino acid
derivatives. The modelling of such a flexible system would
undoubtedly benefit from structural experimental data, since
it would allow known interactions to be confidently and
accurately constrained, but such data was not forthcoming
from 2D NMR experiments (vide supra).
1
1488, 1465, 1406, 1377, 1298, 1254 cm 1; H NMR (300 MHz, CDCl3): d ˆ
7.60 (d, J ˆ 8.0 Hz, 2H, ArH), 7.55 (d, J ˆ 8.0 Hz, 2H, ArH), 7.3 (d, J ˆ
8.0 Hz, 2H, ArH), 6.95 (d, J ˆ 8.0 Hz, ArH), 5.10 (s, 2H, ArCH2O);
13C NMR (68 MHz, CDCl3): d ˆ 161.9, 134.9, 134.2, 132.1, 129.3, 122.5,
119.3, 115.8, 104.6, 69.7; MS (EI): m/z (%): 289 (7), 287 (7), 171 (97), 169
(100), 90 (40); elemental analysis calcd for C14H10ONBr: C 58.36, H 3.5, N
4.86; found: C 58.09, H 3.38, N 4.96.
11-Bromo-5-amino-3-oxa-1(1,4), 4(1,4)-dibenzapentaphane (5): Borane ±
dimethyl sulfide reagent (22 mL of a 2m solution in THF, 44.0 mmol) was
added to a refluxing solution of nitrile 4 (10.02 g, 34.7 mmol) in dry THF
(200 mL) under nitrogen. The reaction was refluxed for two days, allowed
to cool, and 1m HCl (105 mL) was added. The resulting mixture was
refluxed for a further two hours, cooled to room temperature and sodium
hydroxide pellets (8.42 g, 209 mmol) were added. The mixture was stirred
at room temperature for thirty minutes. The organics were extracted into
Et2O (3 Â 100 mL) and the combined organic layer washed with H2O (3 Â
300 mL) and brine (2 Â 100 mL). The organic phase was dried (MgSO4),
filtered and concentrated under reduced pressure. Recrystallisation from
EtOH yielded amine 5 as a white solid (9.44 g, 94%), m.p. 155 ± 1578C
(EtOH/H2O); IR (Nujol): nÄ ˆ 2918, 2856, 1605, 1595, 1503, 1457, 1401, 1375,
1303, 1237, 1171 cm 1; 1H NMR (300 MHz, CDCl3): d ˆ 7.51 (d, J ˆ 8.5 Hz,
2H, ArH), 7.31 (d, J ˆ 8.5 Hz, 2H, ArH), 7.23 (d, J ˆ 8.5 Hz, 2H, ArH), 6.92
(d, J ˆ 8.5 Hz, 2H, ArH), 5.01 (s, 2H, ArCH2O), 3.81 (s, 2H, ArCH2NH2),
1.66 (bs, 2H, NH2); 13C NMR (68 MHz, CDCl3): d ˆ 157.5, 136.3, 136.1,
131.8, 129.2, 128.5, 122.0, 115.0, 69.4, 46.0; MS (EI): m/z (%): 293 (30), 291
(30), 171 (100), 169 (100), 90 (25); elemental analysis calcd for
C14H14ONBr: C 57.55, H 4.83, N 4.79; found: C 57.35, H 4.92, N 4.83.
Conclusion
In conclusion, we have been able to show that macrocycle 1 is
an effective receptor for dipeptides with a carboxylic acid
terminus, and shows some selectivity, particularly for N-Cbz-
b-alanyl-l-amino acids, and thus the basic receptor design was
shown to be appropriate for peptide recognition. Although
the observed binding enantioselectivity is not as great as has
been observed in some other peptide binding systems it is
surprisingly good given the flexibility of the receptor and the
minimal chirality which it incorporates. Evidence from the
binding studies suggests that the binding of N-Cbz-b-alanyl-l-
amino acid substrates involves interaction of the carboxylic
acid of the peptide guest with the amidopyridine unit (with a
strong hydrogen bond to NHb) and involves a hydrogen-bond
interaction with the amide NH group of the substrate.
Variation of the peptide side wall of the receptor, using the
synthetic route we have developed, should be straightforward,
allowing the preparation of less flexible variants, with
increased binding functionality, and potentially providing
structures which can be solubilised in water.
N-Succinyl-l-phenylalanine methyl ester (6): Succinic anhydride (1.0 g,
10 mmol), phenyl alanine methyl ester hydrochloride (2.16 g, 10 mmol) and
triethylamine (2.02 g, 2.8 mL, 20 mmol) were suspended in dry CH2Cl2
(50 mL) and the mixture was stirrred at room temperature for 12 h. 1m HCl
was added (50 mL) until the aqueous layer was at pH 1. The layers were
separated and the aqueous layer further extracted with CH2Cl2 (50 mL).
The combined organic layers were dried (Na2SO4) and concentrated to give
a clear oil which was crystallised overnight in a freezer from EtOAc/
toluene to give acid 6 as a white amorphous solid (2.48 g, 89%); m.p. 70 ±
1
728C; IR (CHCl3): nÄ ˆ 3425, 2926, 1736, 1676, 1514, 1438 cm
;
1H NMR
(270 MHz, CDCl3): d ˆ 7.08 ± 7.31 (m, 5H, ArH), 6.36 (d, J ˆ 8 Hz, 1H,
NHCO), 4.88 (q, J ˆ 7 Hz, 1H, CHNH), 3.72 (s, 3H, COOMe), 3.15 (dd,
J ˆ 6 Hz, 14 Hz, 1H, PhCHAHB), 3.07 (dd, J ˆ 6 Hz, 14 Hz, 1H,
PhCHAHB), 2.66 (t, J ˆ 6 Hz, 2H, CH2COOH), 2.50 (t, J ˆ 6 Hz, 2H,
CH2CONH); 13C NMR (68 MHz, CDCl3): d ˆ 177.1, 172.2, 171.6, 135.8,
129.4, 128.7, 127.3 (1) 53.5, 52.6, 37.9, 30.6, 29.4 (2); MS (CI, NH3): m/z (%):
Experimental Section
‡
‡
297 ([M‡NH4] , 30), 280 ([M‡H] , 100), 180 (45).
General methods: Whenever possible all solvents and reagents were
purified according to literature procedures.[34] Thin layer chromatography
(TLC) was performed on aluminium backed sheets (CamLab) coated with
either silica gel (SiO2; 0.25 mm) or neutral alumina, containing fluorescent
indicator UV254. Unless otherwise indicated column chromatography was
performed on Sorbsil C60, 40 ± 60 mesh silica. All melting points were
determined in open capillary tubes using Gallenkamp Electrothermal
Melting Point Apparatus and are uncorrected. Infrared spectra were
recorded on a Perkin-Elmer 1600 Fourier Transform Spectrophotometer.
1H NMR spectra at 270 MHz were obtained on a JEOL GX 270, at
300 MHz on a Bruker AC 300 and at 360 MHz on a Bruker AM 360
spectrometer. 13C NMR spectra were recorded at 68 MHz on the JEOL GX
270, and at 75 MHz on the Bruker AC 300 and at 90 MHz on the Bruker
AM 360. Microanalytical data were obtained from SmithKline Beecham
Pharmaceuticals, Brockham Park, Surrey, UK. Mass spectra were recorded
Methyl (1S)-1-benzyl-12-bromo-3,6-dioxo-10-oxa-2,7-diaza-9(1,4),12(1,4)-
dibenzenadodecaphane-1-carboxylate (7): Dicyclohexyl carbodiimide
(1.78 g, 8.63 mmol) was added to a stirred solution of amine 5 (2.09 g,
7.16 mmol), acid 6 (2 g, 7.16 mmol) and 1-hydroxybenzotriazole monohy-
drate (0.97 g, 7.16 mmol) in dry distilled DMF (65 mL) at 08C, under N2,
and stirred at 08C for 30 min then at room temperature for six hours. The
solvent was removed by distillation at reduced pressure and the crude solid
was purified by flash column chromatography (CH2Cl2/MeOH, 99:1 v/v) to
give ester 7 as a white amorphous solid (3.36g, 85%); m.p. 175 ± 1778C; IR
(CHCl3): nÄ ˆ 3281, 2924, 1740, 1633, 1538, 1516, 1456, 1427, 1375, 1254,
1
1178 cm
;
1H NMR (270 MHz, (CD3)2SO): d ˆ 8.49 (d, J ˆ 8 Hz, 1H,
MeOOCCHNH), 8.37 (t, J ˆ 6 Hz, 1H, ArCHNHCO), 7.69 (d, J ˆ 8 Hz,
2H, ArH), 7.51 (d, J ˆ 8 Hz, 2H, ArH), 7.25 ± 7.40 (m, 7H, ArH), 7.04 (d,
J ˆ 8 Hz, 2H, ArH), 5.18 (s, 2H, ArCH2O), 4.56 (q, J ˆ 7 Hz, 1H,
CHCOOMe), 4.28 (d, J ˆ 6 Hz, 2H, ArCH2NH), 3.70 (s, 3H, COOMe),
3.13 (dd, J ˆ 6 Hz,14 Hz, 1H, CHCHAHBPh), 3.05 (dd, J ˆ 6 Hz,14 Hz, 1H,
CHCHAHBPh), 2.41 ± 2.46 (m, 4H, OCCH2CH2CO); 13C NMR (68 MHz,
(CD3)2SO): d ˆ 172.0, 171.4, 170.9, 156.8, 137.1, 136.6, 131.8, 131.2, 129.6,
128.9, 128.4, 128.1, 126.4, 120.7, 114.5, 68.2, 53.5, 51.7, 41.4, 36.7, 30.4, 30.3;
MS (FAB): m/z (%): 553 (45), 262 (37), 225 (60), 169 (65); elemental
either on
a Micromass Platform quadrupole mass analyser with an
electrospray ion source, or on a VG Analytical 70-250-SE normal geometry
double focusing mass spectrometer at Southampton University.
11-Bromo-4-cyano-3-oxa-1(1,4),4(1,4)-dibenzatetraphane (4): Diethyl di-
azodicarboxylate (11.10 g, 64.0 mmol) was added dropwise over a period of
one hour to a stirred solution of 4-bromobenzyl alcohol (2) (10.00 g,
Chem. Eur. J. 1999, 5, No. 1
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0501-0085 $ 17.50+.50/0
85

J. D. Kilburn et al.
FULL PAPER
analysis calcd for C28H29BrN2O5: C 60.77, H 5.28, N 5.06; found: C 60.93, H
4.97, N 5.21.
drops of DMF were added and the mixture stirred for two hours under
nitrogen. The solvent was removed under reduced pressure to yield acid
chloride 13 as a pale yellow oil (6.0 g), which was used without further
Methyl 4-tri-n-butylstannylphenylacetate (9): Tetrakis(triphenylphospha-
ne)palladium(0) (250 mg, 2.2 mmol) was added to a thoroughly degassed
1
purification; IR (CH2Cl2): nÄ ˆ 1794, 1737 cm
;
1H NMR (300 MHz,
ˆ
CDCl3): d ˆ 5.85 (ddt, J ˆ 17.3 Hz, 10.3 Hz, 5.5 Hz, 1H, CH2 CH), 5.26
solution of methyl 4-bromophenylacetate
7 (5.00 g, 22.0 mmol) and
ˆ
(ddt, J ˆ 17.3 Hz, 3.3 Hz, 1.5 Hz, 1H, CH2CH CHtrans), 5.18 (ddt, J ˆ
bis(tributylditin) (22.0 mL, 25.0 g, 44 mmol) in toluene (5 mL) under
nitrogen. The reaction mixture was heated to reflux and stirred for five
hours. The mixture was cooled to room temperature and concentrated
under reduced pressure. Dry flash silica chromatography, using petroleum
ether as eluent, removed unwanted bis(tributylditin) and tributyltin
bromide and the product was then obtained by elution with petroleum
ether/EtOAc (99.5:0.5, v/v) to give stannane 9 as a clear, colourless oil
(7.05 g, 76%), IR (neat): nÄ ˆ 2957, 2925, 1736, 1458, 1440, 1371, 1340, 1249,
1154; 1H NMR (300 MHz, CDCl3): d ˆ 7.48 (dd, J ˆ 7.7 Hz, JH,Sn 37.9 Hz,
2H, ArH), 7.13 (d, J ˆ 7.7 Hz, 2H, ArH), 3.73 (s, 3H, COOMe), 3.65 (s, 2H,
CH2COO), 1.67 ± 1.61 (m, 6H, Sn(CH2)2CH2CH3), 1.51 ± 1.36 (m, 6H,
SnCH2CH2CH2CH3), 1.11 (t, J ˆ 8.1 Hz, 6H, SnCH2(CH2)2CH3), 0.95 (t,
J ˆ 7.4 Hz, 9H, Sn(CH2)3CH3); 13C NMR (75 MHz, CDCl3): d ˆ 172.1,
140.4, 136.7 (JC,Sn ˆ 32 Hz), 133.6, 128.8 (JC,Sn ˆ 40 Hz), 52.0, 41.2, 29.1
(JC,Sn ˆ 20 Hz), 27.4 (JC,Sn117/119 ˆ 55/57 Hz), 13.7, 9.6 (JC,Sn117/119 ˆ 326/
ˆ
10.3 Hz, 3.3 Hz, 1.5 Hz, 1H, CH2CH CHcis), 4.55 (dt, J ˆ 5.5 Hz, 1.5 Hz,
2H, CH2O), 3.22 (t, J ˆ 6.6 Hz, 2H, CH2CH2), 2.71 (t, J ˆ 6.6 Hz, 2H,
CH2CH2); 13C NMR (75.5 MHz, CDCl3): d ˆ 173.2, 170.0, 131.8, 118.9, 66.0,
41.9, 29.4.
4(S) 4-tert-butoxycarbonyl-4-benzyl-3-oxo-11,4-diamino-12,2-diaza-1(1,3)-
benzatetraphane (16): Dicyclohexyl carbodiimide (12.90 g, 56.5 mmol) was
added to a stirred solution of 2,6-diaminopyridine (6.17 g, 56.5 mmol),
HOBt (5.09 g, 37.7 mmol) and N-tert-butoxycarbonylphenylalanine
(10.00 g, 37.7 mmol), in CH2Cl2/DMF (100 mL, 10:1 v/v) . After stirring
overnight the reaction mixture was filtered through celite and concentrated
under reduced pressure. The resulting gum was diluted with EtOAc
(100 mL), washed with saturated aqueous NaHCO3 (2 Â 100 mL) and brine
(2 Â 100 mL) and the organic phase dried (MgSO4), filtered and concen-
trated under reduced pressure to give a green gum, which was purified by
flash column chromatography (CH2Cl2/MeOH, CH2Cl2/MeOH, 99.5:0.5 to
98:2, v/v) followed by recrystallization from aqueous EtOH to give 16 as a
‡
‡
340 Hz); MS (CI, NH3): m/z (%): 458 ([M‡NH4] , 55), 441 ([M‡H] ,
12), 400 (25), 308 (100), 168 (57), 35 (85).
white crystalline solid (10.21 g, 76%), m.p. 198 ± 2008C (EtOH); [a]D
ˆ
(14S)-14-Benzyl-9,12-dioxo-5-oxa-8,13-diaza-2(1,4),3(1,4),6(1,4)-tribenze-
natetradecaphane-1,14-dicarboxylic acid (11): Tetrakis(triphenylphospha-
ne)palladium (20 mg) and silver(i) oxide (83 mg, 0.36 mmol) were added to
a thoroughly degassed solution of bromide 7 (200 mg, 0.36 mmol) in dry
DMF (7 mL) and the mixture was heated with stirring to 508C under
nitrogen. A degassed solution of stannane 9 (192 mg, 0.44 mmol, 1.2 equiv)
in dry DMF (2 mL) was added and the mixture stirred for 12 hours at 508C.
The mixture was allowed to cool and was filtered through celite. The filtrate
was diluted with EtOAc (100 mL) and washed with H2O (100 Â 2 mL) and
brine (100 mL), dried (Na2SO4) and concentrated, to give a crude solid
which was purified by flash column chromatography (petroleum ether/
EtOAc 90:10 v/v, then MeOH/CH2Cl2/Et2O 1:50:50 v/v) to give the
intermediate diester 10 (152 mg) contaminated with the starting bromide 7
in 4:1 ratio (yield of 10 ꢀ54% by 1H NMR spectroscopy), which was
subjected to hydrolysis without further purification.
‡3.9 (c ˆ 1.0 in acetone); IR (CH2Cl2): nÄ ˆ 3467, 3347, 3257, 1695, 1685,
1
1671, 1617, 1555, 1463, 1388, 1365, 1159 cm
;
1H NMR (300 MHz,
(CD3)2SO): d ˆ 9.00 (s, 1H, pyrNH), 7.50 ± 7.10 (m, 7H, p-pyrH, PhH,
CHNH), 6.26 (d, J ˆ 7.3 Hz, 1H, m-pyrH), 6.25 (d, J ˆ 7.1 Hz, 1H, m-pyrH),
5.30 (bs, 2H, NH2), 4.56 (br. m, 1H, CHCO), 3.26 (dd, J ˆ 13.8 Hz, 4.8 Hz,
1H, PhCHAHB), 2.95 (dd, J ˆ 13.8 Hz, 9.6 Hz, 1H, PhCHAHB), 1.30 (9H, s,
(CH3)3); 13C NMR (68 MHz, (CD3)2SO): d ˆ 171.3, 158.6, 155.5, 150.2,
139.0, 138.2, 129.4, 128.9, 126.3, 103.6, 100.9, 78.2, 56.5, 37.2, 28.2; MS (CI):
m/z (%):357 (70), 229 (100), 246 (85); elemental analysis calcd for
C19H24N4O3; C 64.03, H 6.79, N 15.72; found: C 63.81, H 6.65, N 15.33.
4(S) Prop-2-enyl-4-benzyl-3,6-dioxo-11-amino-12,2,5-diaza-1(1,3)-benza-
nonaphane-9-carboxylate (15): Amine 16 (2.05 g, 5.64 mmol) was dissolved
with TFA/CH2Cl2 (25 mL, 1:1 v/v) and stirred for 30 minutes. The reaction
mixture was concentrated under reduced pressure and excess TFA was
removed by azeotropic distillation with toluene. The resulting gum was
triturated with Et2O to give the amine as the trifluoroacetate salt as a white
solid (2.50 g) which was used without further purification, 1H NMR
(300 MHz, CD3OD): d ˆ 7.86 (dd, J ˆ 8.8 Hz, 7.7 Hz, 1H, p-pyrH), 7.38 (m,
5H, ArH), 6.82 (d, J ˆ 7.7 Hz, 1H, m-pyrH), 6.68 (d, J ˆ 8.8 Hz, 1H, m-
pyrH), 4.41 (dd, J ˆ 8.4 Hz, 6.4 Hz, 1H, CHNH), 3.39 (dd, J ˆ 14.0 Hz,
6.4 Hz, 1H, PhCHAHB), 3.20 (dd, J ˆ 14 Hz, 8.4 Hz, 1H, PhCHAHB);
13C NMR (75.5 MHz, CD3OD): d ˆ 170.3, 155.4, 146.0, 142.0, 134.9, 130.5,
130.1, 129.0, 108.2, 101.4, 56.4, 38.3.
Diester 10 (contaminated with bromide 7) (152 mg) was dissolved in 1,4-
dioxane (20 mL) with gentle warming. Lithium hydroxide monohydrate
(70 mg, 1.6 mmol) in H2O (5 mL) was added and the mixture stirred at
room temperature overnight. The solution was adjusted to pH 1 (1m HCl)
and the solvents were removed in vacuo to give a crude solid, which was
purified by flash column chromatography (CH2Cl2/MeOH, 98:2 to 95:5 v/v)
to give the diacid 11 as a yellowish solid (52 mg, 24% from 7); m.p. 185 ±
1
1878C; H NMR (270 MHz, (CD3)2SO): d ˆ 8.39 (t, J ˆ 6 Hz, 1H, ArCH2-
NHCO), 8.33 (d, J ˆ 8 Hz, 1H, MeOOCCHNH ), 7.79 (d, J ˆ 8 Hz, 2H,
ArH), 7.74 (d, J ˆ 8 Hz, 2H, ArH), 7.63 (d, J ˆ 8 Hz, 2H, ArH), 7.47 (d, J ˆ
8 Hz, 2H, ArH) 7.27 ± 7.42 (m, 7H, ArH), 7.08 (d, J ˆ 8 Hz, 2H, ArH), 5.24
(s, 2H, ArCH2O), 4.53 (m, 1H, CHCOOMe), 4.30 (d, J ˆ 6 Hz, 2H,
ArCH2NH), 3.74 (s, 2H, ArCH2COOH), 3.19 (dd, J ˆ 5 Hz, 14 Hz, 1H,
CHCHAHBPh), 2.97 (dd, J ˆ 9 Hz, 14 Hz, 1H, CHCHAHBPh), 2.41 ± 2.46
(m, 4H, OCCH2CH2CO); 13C NMR (68 MHz, (CD3)2SO): d ˆ 173.1, 172.6,
171.3, 171.0, 157.1, 139.3, 138.1, 137.6, 136.2, 134.3, 131.7, 129.9, 129.0, 128.4,
128.1, 126.5, 126.4, 126.3, 114.5, 68.7, 53.5, 41.4, 40.2, 36.7, 30.6, 30.5; IR
(CHCl3): nÄ ˆ 1698, 1634, 1231, 1004 cm 1; MS (FAB): m/z (%): 595 (5), 225
(20), 169 (52), 85 (100); elemental analysis calcd for C35H34N2O7: C 70.69, H
5.76, N 4.71; found C 70.39, H 5.61, N 4.63.
Acid chloride 13 (1.00 g, 5.57 mmol) in THF (5 mL) was added dropwise to
a stirred solution of the trifluoroacetate salt described above (1.44 g,
5.64 mmol) and a large excess of DIPEA (0.14 mL, 29 mmol) in dry freshly
distilled THF (5 mL). The reaction was stirred for two hours before being
concentrated under reduced pressure and dissolved in EtOAc (100 mL),
washed with saturated aqueous sodium bicarbonate (2 Â 50 mL), dried
(MgSO4), filtered and concentrated under reduced pressure to give a
yellow oil. Purification by flash column chromatography (CH2Cl2/MeOH,
99.5:0.5 to 98:2, v/v) gave (S)-15 as a colourless foam (1.58 g, 72%), [a]D
ˆ
24.3 (c, 1.0 in CH2Cl2); IR (KBr): nÄ ˆ 3407, 1729, 1662, 1458, 1380, 1298,
1247, 1162 cm 1; 1H NMR (300 MHz, CDCl3): d ˆ 8.46 (bs, 1H, pyrNHCO),
7.33 (d, J ˆ 7.6 Hz, 1H, m-pyrH), 7.29 (dd, J ˆ 7.8 Hz, 7.6 Hz, 1H, p-pyrH),
7.25 ± 7.15 (m, 7H, PhH, CHNH), 6.36 (d, J ˆ 7.8 Hz, 1H, m-pyrH), 6.10 (d,
J ˆ 8.1 Hz, 1H, CHNH), 5.87 (ddt, J ˆ 17.3 Hz, 10.3 Hz, 5.9 Hz, 1H,
4-Hydrogen-1-prop-2-enyl-butane dicarboxylate (12): A mixture of suc-
cinic anhydride (5 g, 50 mmol) and allyl alcohol (3 mL, 50 mmol) were
refluxed for 3 h. The mixture was then distilled at reduced pressure to
afford monoester 12 as a colorless liquid (5 g, 72%), bp 1288C (4 mm of
ˆ
ˆ
CH2 CH), 5.33 (ddt, J ˆ 17.3 Hz, 3.3 Hz, 1.5 Hz, 1H, CH2CH CHtrans),
ˆ
5.27 (ddt, J ˆ 10.3 Hz, 3.3 Hz, 1.5 Hz, 1H, CH2CH CHcis), 4.92 (dd, J ˆ
Hg) [ref.[14] 1488C (8 mm of Hg)]; IR (CH2Cl2): nÄ ˆ 2930, 1737, 1717, 1649,
6.6 Hz, 6.3 Hz, 1H, CHNH), 4.62 (dt, J ˆ 5.9 Hz, 1.5 Hz, 2H, CH2 ˆ
CHCH2), 4.32 (br, 2H, NH2), 3.25 (dd, J ˆ 14.0 Hz, 6.3 Hz, 1H, PhCHAHB),
3.14 (dd, J ˆ 14.0 Hz, 6.6 Hz, 1H, PhCHAHB), 2.80 ± 2.50 (m, 4H, CH2CH2);
13C NMR (75 MHz, CDCl3): d ˆ 172.8, 171.9, 170.0, 157.3, 149.4, 140.2,
136.3, 132.1, 129.5, 128.8, 127.2, 118.7, 104.8, 103.6, 65.7, 55.2, 38.3, 31.0, 29.5;
1
1420, 1170 cm
;
1H NMR (300 MHz, CDCl3): d ˆ 10.90 (br. s, 1H,
ˆ
COOH), 5.85 (ddt, J ˆ 17.3 Hz, 10.3 Hz, 5.5 Hz, 1H, CH2 CH), 5.26 (ddt,
ˆ
J ˆ 17.3 Hz, 3.3 Hz, 1.5 Hz, 1H, CH2CH CHtrans), 5.18 (ddt, J ˆ 10.3 Hz,
ˆ
3.3 Hz, 1.5 Hz, 1H, CH2CH CHcis), 4.55 (dt, J ˆ 5.5 Hz, 1.5 Hz, 2H,
CH2O), 2.62 (m, 4H, CH2CH2); 13C NMR (75.5 MHz, CDCl3): d ˆ 178.5,
‡
‡
LRMS (ES‡): m/z (%): 793 ([2MH] , 10%), 397 ([MH] , 100). HRMS
172.1, 132.0, 118.4, 65.5, 29.1, 28.9.
(ES‡): m/z (%): calcd for C21H25N4O4 [M‡H]: 397.1876, found 397.1894.
Acid chloride (13): Oxalyl chloride (4.2 mL, 47.5 mmol) was added to a
stirred solution of monoester 12 (5 g, 31.7 mmol) in CH2Cl2 (20 mL). Two
Synthesis of racemic prop-2-enyl-4-benzyl-3,6-dioxo-11-amino-12,2,5-dia-
za-1(1,3)-benzanonaphane-9-carboxylate (15): Acid chloride 13 was added
86
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0501-0086 $ 17.50+.50/0
Chem. Eur. J. 1999, 5, No. 1

Macrocyclic Peptide Receptor
79±89
dropwise to l-phenylalanine (7.84 g, 47.5 mmol) in 1m aqueous sodium
carbonate solution (100 mL) and the mixture stirred overnight. The
reaction mixture was extracted with CH2Cl2 (100 mL), and the aqueous
phase acidified with 1m HCl. The resulting oily suspension was extracted
with CH2Cl2 (100 mL), the organic phase dried (Na2SO4) and the solvent
removed at reduced pressure to give crude 14 as a yellow oil (9 g) which was
used directly without further purification. Dicyclohexyl carbodiimide
(7.84 g, 38 mmol) was added to 2,6-diaminopyridine (3.44 g, 31.7 mmol),
4-dimethylaminopyridine (3.84 g, 31.7 mmol) and the crude 14 (9 g) in
CH2Cl2 (100 mL). After stirring overnight the mixture was filtered, silica
gel (20 g) added and concentrated to dryness. Purification by flash column
chromatography (CH2Cl2/MeOH, 99.5:0.5 to 98:2, v/v)) gave racemic 15 as
a yellow foam (9 g, 72% overall), identical to (S)-15 except [a]D ˆ 0.0 (c ˆ
1.0 in CH2Cl2);
138.1, 136.2, 134.3, 132.3, 129.9, 128.2, 128.1, 126.6, 126.5, 114.4, 77.6, 68.7,
‡
‡
42.7, 40.2, 28.2; MS (ES ): m/z (%): 465 ([MNH4] , 100); elemental
analysis calcd for C27H29NO5: C 72.48, H 6.48, N 3.13; found C 72.08, H
6.50, N 3.19.
1-tert-butoxycarbonylamino-3-oxa-2(1,4), 5(1,4), 6(1,4)-tribenzahepta-
phane-7-carboxylic acid fluoride (20): Cyanuric fluoride (59 mg,
0.44 mmol) was added to a solution of carboxylic acid 19 (100 mg,
0.22 mmol) and pyridine (0.018 mL, 0.22 mmol) in dry acetonitrile
(2 mL) and stirred overnight at room temperature. H2O was added and
the precipitated cyanuric acid removed by filtration through a celite pad.
CH2Cl2 (10 mL) and H2O (10 mL) were added and the organic layer
separated, dried (MgSO4), filtered and concentrated under reduced
pressure to give acid fluoride 20 as a pale yellow solid (96 mg, 96%),
m.p. > 2308C; IR (CH2Cl2): nÄ ˆ 3054, 2986, 2305, 1842, 1710, 1509,
1
1421 cm
;
1H NMR (300 MHz, CDCl3): d ˆ 7.60 (d, J ˆ 8.5 Hz, 4H,
1-Bromo-8-tert-butoxycarbonylamino-7-oxo-3-oxa-6-aza-1(1,4), 4(1,4)-di-
benzaoctaphane (17): Di-tert-butyldicarbonate (4.51 g, 20.6 mmol) was
added to a stirred suspension of amine 5 (5.00 g, 17.3 mmol) and triethyl-
amine (3.6 mL, 25.7 mmol) in CH2Cl2 (50 mL). After one hour the reaction
mixture was poured into aqueous 1m citric acid solution (50 mL). The
organic phase was washed with brine (3 Â 50 mL), dried (MgSO4), filtered
and concentrated under reduced pressure to afford a crude pale yellow
solid. Precipitation from EtOAc/hexane gave the protected amine 17 as a
ArH), 7.54 (d, J ˆ 8.5 Hz, 2H, ArH), 7.37 (d, J ˆ 8.5 Hz, 2H, ArH), 6.96
(d, J ˆ 8.5 Hz, 2H, ArH), 5.11 (s, 2H, CH2O), 4.82 (br, 1H, NH), 4.26 (d,
J ˆ 5.1 Hz, 2H, CH2NH), 3.87 (s, 2H, CH2COF), 1.47 (s, 9H, (Me)3).
13(S) Prop-2-enyl 1-tert-butoxycarbonylamino-13-benzyl-8,12,15-trioxo-3-
oxa-9,102,11,14-tetraaza-2(1,4), 5(1,4), 6(1,4), 10(1,3)-tetrabenzenahepta-
decaphane-17-carboxylate
(21):
N-Methylmorpholine
(0.027 mL,
0.24 mmol) was added to a solution of acid fluoride 20 (0.05 g, 0.11 mmol)
and amine 15 (40 mg, 0.11 mmol) in THF (2 mL). The mixture was heated
at reflux under nitrogen for two hours, cooled, diluted with EtOAc (50 mL)
washed with saturated sodium bicarbonate (2 Â 50 mL) and brine (2 Â
50 mL). The organic phase was dried (MgSO4), filtered and concentrated
under reduced pressure to give an orange oil. Flash column chromatog-
raphy on basic alumina Brockman grade II (Et2O/MeOH (99.5:0.5 to 97:3,
v/v)) followed by recrystallisation from EtOAc/Et2O/hexane gave 21 as a
pale yellow solid (56 mg, 62%), m.p. 58 ± 608C (EtOAc/Et2O/hexane); [a]D
ˆ ‡7.1 (c ˆ 1.0 in MeOH); IR (CH2Cl2): n ˆ 1846, 1492, 1410, 1086, 1014,
842, 804; 1H NMR (300 MHz, CDCl3): d ˆ 9.11 (s, 1H, pyrNHCO), 8.48 (s,
1H, pyrNHCO), 7.86 (d, J ˆ 8.1 Hz, 1H, m-pyrH), 7.73 (d, J ˆ 7.7 Hz, 1H,
m-pyrH), 7.58 (d, J ˆ 8.1 Hz, 2H, ArH), 7.56 (d, J ˆ 8.1 Hz, 2H, ArH), 7.52
(1H, obscured, p-pyrH), 7.48 (d, J ˆ 8.1 Hz, 2H, ArH), 7.39 (d, J ˆ 8.1 Hz,
2H, ArH), 7.25-7.15 (m, 8H, PhH, ArH, ArCH2NH), 6.95 (d, J ˆ 8.5 Hz,
2H, ArH), 6.65 (d, J ˆ 8.1 Hz, 1H, COCHNH), 5.85 (ddt, J ˆ
white solid (5.22 g, 81%), m.p. 98 ± 1008C (EtOAc/hexane); IR (CH2Cl2):
1
n ˆ 3445, 1698, 1613, 1510, 1241, 1167 cm
;
1H NMR (300 MHz, CDCl3):
d ˆ 7.50 (d, J ˆ 8.5 Hz, 2H, ArH), 7.39 (d, J ˆ 8.5 Hz, 2H, ArH), 7.21 (d, J ˆ
8.5 Hz, 2H, ArH), 6.91 (d, J ˆ 8.5 Hz, 2H, ArH), 4.99 (s, 2H, CH2O), 4.90
(br. s, 1H, NH), 4.24 (d, 5.9 Hz, 2H, CH2NH), 1.47 (s, 9H, (CH3)3);
13C NMR (68 MHz, CDCl3): d ˆ 157.6, 156.0, 136.1, 131.8, 131.7, 129.1, 129.0,
121.9, 115.0, 79.5, 69.3, 44.2, 28.5; IR (CHCl3): nÄ ˆ 3448, 3312, 2981, 2925,
2873, 1682, 1610, 1585, 1531, 1510, 1434, 1392, 1366 cm 1; MS (CI, NH3):
‡
‡
m/z (%): 411 ([M‡NH4] , 30), 409 ([M‡NH4] , 30), 355 (100), 353 (100),
277 (30), 275 (37); elemental analysis calcd for C19H22O3NBr: C 58.17, H
5.65, N 3.57, found: C 57.86, H 5.57, N 3.55.
Methyl-1-tert-butoxycarbonylamino-3-oxa-2(1,4), 5(1,4), 6(1,4)-tribenza-
heptaphane-7-carboxylate (18): Bis(triphenylphosphane) palladium(II)
dichloride (94 mg, 0.13 mmol) was added to a thoroughly degassed solution
of bromide 17 (390 mg, 1.00 mmol) and stannane 9 (450 mg, 1.00 mmol) in
toluene (5 mL) under argon. The reaction was refluxed for 12 h, cooled to
room temperature, diluted with EtOAc (50 mL), washed with brine
(3x50 mL), dried (MgSO4), filtered and concentrated under reduced
pressure to give the crude product as a brown solid. Purification by flash
column chromatogaphy (petroleum ether (b.p. 40 ± 608C) to petroleum
ether (b.p. 40 ± 608C)/EtOAc (7:3, v/v)) followed by recrystallisation from
EtOAc/hexane gave ester 18 as a white solid (0.29 g, 63%), m.p. 116 ±
1188C (EtOAc/hexane); IR (CH2Cl2): nÄ ˆ 3450, 2979, 2870, 1732, 1707,
ˆ
17.3 Hz,10.3 Hz, 5.5 Hz, 1H, CH2CH CH2), 5.26 (ddt, 17.3 Hz, 3.3 Hz,
ˆ
1.5 Hz, 1H, CH2CH CHtrans), 5.20 (ddt, J ˆ 10.3 Hz, 3.3 Hz, 1.5 Hz, 1H,
ˆ
CH2CH CHcis), 5.08 (s, 2H, CH2O), 4.90 (bm, 1H, COCHNH), 4.50 (dt,
ˆ
J ˆ 5.9 Hz, 1.5 Hz, 2H, CH2CH CH2), 4.25 (d, J ˆ 5.2 Hz, 2H, CH2NH),
3.75 (s, 2H, ArCH2CONH), 3.18 (dd, J ˆ 13.6 Hz, 6.6 Hz, 1H, PhCHAHB),
3.10 (dd, J ˆ 13.6 Hz, 7.0 Hz, 1H, PhCHAHB), 2.68 (t, J ˆ 7.0 Hz, 2H,
CH2CH2), 2.49 (t, J ˆ 7.0 Hz, 2H, CH2CH2), 1.46 (s, 9H, (Me)3); 13C NMR
(75 MHz, CDCl3): d ˆ 172.9, 172.1, 170.4, 169.9, 158.2, 156.1, 149.9, 149.3,
140.5, 140.4, 140.0, 136. 3, 136.2, 133.8, 131.9, 131.6, 130.1, 129.5, 129.0, 128.7,
128.2, 127.7, 127.4, 127.2, 118.8, 115.1, 110.1, 109.9, 79.7, 69.9, 65.8, 55.2, 44.3,
1
1509, 1456, 1437, 1367, 1240, 1168 cm 1; H NMR (300 MHz, CDCl3): d ˆ
7.61 (d, J ˆ 8.1 Hz, 2H, ArH), 7.56 (d, J ˆ 8.1 Hz, 2H, ArH), 7.49 (d, J ˆ
8.1 Hz, 2H, ArH), 7.37 (d, J ˆ 8.1 Hz, 2H, ArH), 7.22 (d, J ˆ 8.5 Hz, 2H,
ArH), 6.96 (d, J ˆ 8.5 Hz, 2H, ArH), 5.07 (s, 2H, CH2O), 5.00 (1H, br. s,
NH), 4.26 (d, J ˆ 5.6 Hz, 2H, CH2NH), 3.73 (s, 2H, CH2COOMe), 3.69 (s,
3H, COOMe), 1.39 (s, 9H, (Me)3); 13C NMR (68 MHz, CDCl3): d ˆ 172.2,
158.2, 156.0, 140.7, 139.8, 136.1, 133.3, 131.5, 129.9, 129.0, 128.1, 127.5, 127.4,
‡
‡
44.2, 39.5, 30.8, 29.5, 28.8; LRMS (ES ): m/z (%): 826 (100); HRMS (ES ):
m/z calcd for C48H52N5O8 (M‡H): 826.3816; found m/z 826.3821; elemental
analysis calcd for C48H51N5O8; C 69.80, H 6.22, N 8.48; found: C 69.45, H
6.28, N 8.41.
‡
‡
115.1, 79.5, 69.9, 52.3, 44.3, 41.0, 28.6; MS (ES ): m/z (%): 945 ([2MNa] ,
12(S) 12-Benzyl-2-oxa-8, 92, 10, 13, 18-pentaaza-1(1,4), 4(1,4), 5(1,4),
9(1,3)-tetrabenzenacycloicosaphane-7,11,14,17-tetraone (1): Tetrakis(tri-
phenylphosphane)palladium(0) (17 mg, 10 mol%, 0.01 mmol) was added
to 21 (0.12 g, 0.15 mmol) and pyrrolidine (0.12 mL, 1.5 mmol) in CH2Cl2
(5 mL) at room temperature. The reaction was stirred under argon for half
an hour, before being diluted with CHCl3 (25 mL) and washed with 1m HCl
(2 Â 25 mL). The organic layer was dried (MgSO4), filtered, concentrated
under reduced pressure and dried under high vacuum to give the crude acid
22 as a pale yellow foam (0.12 g 0.15 mmol) which was dissolved in CH2Cl2
(5 mL) with DMAP (2 mg, 0.01 mmol) and pentafluorophenol (80 mg,
0.44 mmol). Dicyclohexyl carbodiimide (90 mg, 0.44 mmol) in CH2Cl2
(2 mL) was added and the mixture stirred for an hour. The mixture was
concentrated under reduced pressure to give the crude pentafluorophenyl
ester 22 as an orange gum (0.33 g), which was dissolved in dioxane (2 mL)
and HCl (5 mL of a 4m solution in dioxane) was added. The reaction
mixture was stirred for half an hour, concentrated under reduced pressure
and the resulting gum was triturated with Et2O to give the crude
hydrochloride salt 22 as a pale yellow solid (0.25 g), which was dried for
half an hour under high vacuum.
‡
‡
‡
12), 940 ([2MNH4] , 5), 500 ([MK] , 15), 484 ([MNa] , 100), 479
‡
([MNH4] , 15).
1-tert-butoxycarbonylamino-3-oxa-2(1,4), 5(1,4), 6(1,4)-tribenzahepta-
phane-7-carboxylic acid (19): Aqueous lithium hydroxide (1.3 mL of a 1m
solution, 1.3 mmol) was added to ester 18 (0.52 g, 1.1 mmol) in 1,4 dioxane
(4 mL) and the reaction mixture was stirred for two hours. The mixture was
concentrated under reduced pressure and diluted with EtOAc (50 mL). 1m
HCl (25 mL) was added and the mixture shaken until all of the solid
dissolved. The organic phase was separated, dried (MgSO4), filtered and
concentrated under reduced pressure. Recrystallisation from acetone/
hexane gave carboxylic acid 19 as a white solid (0.45 g, 92%), m.p. >2308C
(acetone/hexane); IR (KBr): nÄ ˆ 3303, 2726, 1719, 1686, 1517, 1248, 1161,
1
1109, 1055 cm
;
1H NMR (300 MHz, (CD3)2SO): d ˆ 12.35 (br, 1H,
COOH), 7.68 (d, J ˆ 8.1 Hz, 2H, ArH), 7.63 (d, J ˆ 8.1 Hz, 2H, ArH), 7.53
(d, J ˆ 8.1 Hz, 2H, ArH), 7.36 (d, J ˆ 8.1 Hz, 2H, ArH), 7.31 (br, 1H, NH),
7.17 (d, J ˆ 8.5 Hz, 2H, ArH), 6.98 (d, J ˆ 8.5 Hz, 2H, ArH), 5.14 (s, 2H,
CH2O), 4.06 (d, J ˆ 4.5 Hz, 2H, CH2NH), 3.63 (s, 2H, CH2COOH), 1.48 (s,
9H, (Me)3); 13C NMR (75.5 MHz, (CD3)2SO): d ˆ 172.6, 157.1, 156.0, 139.3,
Chem. Eur. J. 1999, 5, No. 1
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0501-0087 $ 17.50+.50/0
87

J. D. Kilburn et al.
FULL PAPER
The solid 22 was dissolved in dry, freshly distilled DMF (10 mL) and added
by syringe pump (2 mLh 1) to a refluxing solution of DIPEA (0.075 mL,
0.44 mmol) in dry, freshly distilled MeCN (100 mL). After the addition was
complete the reaction mixture was refluxed for a further 12 h after which it
was concentrated under reduced pressure, diluted with EtOAc (100 mL)
and washed with saturated aqueous sodium bicarbonate solution (2 Â
100 mL). The organic layer was dried (MgSO4), filtered and concentrated
under reduced pressure. Flash column chromatography (on silica pre-
washed with Et2O/ammonia saturated MeOH (99:1, v/v)) and gradient
elution (Et2O/ammonia saturated MeOH (99:1 to 97:3, v/v)) gave macro-
cycle 1 as a white flaky solid (31 mg, 33%). Further purification was
achieved by dissolving the product in acetone or CHCl3 and precipitating
Y. Cheng, Y. Shao, R. M. J. Liskamp, Tetrahedron Lett. 1996, 37, 8253;
d) M. Asakawa, C. L. Brown, D. Pasini, J. F. Stoddart, P. G. Wyatt, J.
Org. Chem. 1996, 61, 7234; e) C. Gennari, H. P. Nestler, B. Salom,
W. C. Still, Angew. Chem. 1995, 107, 1894; Angew. Chem. Int. Ed. Engl.
1995, 34, 1765; f) Y. Kuroda, Y. Kato, T. Higashioji, J. Hasegawa, S.
Kawanami, M. Takahashi, N. Shiraishi, K. Tanabe, H. Ogoshi, J. Am.
Chem. Soc. 1995, 117, 10950; g) S. R. LaBrenz, J. W. Kelly, J. Am.
Chem. Soc. 1995, 117, 1655; h) H. J. Schneider, Angew. Chem. 1993,
105, 890; Angew. Chem. Int. Ed. Engl. 1993, 32, 848. See also
references [3, 5].
[3] a) G. J. Pernia, J. D. Kilburn, J. W. Essex, R. J. Mortishire-Smith, M.
Rowley, J. Am. Chem. Soc. 1996, 118, 10220; b) G. J. Pernia, J. D.
Kilburn, M. Rowley, J. Chem. Soc., Chem. Commun. 1995, 305; c)
C. P. Waymark, J. D. Kilburn, I. Gillies, Tetrahedron Lett. 1995, 36,
3051; d) H. K. Patel, J. D. Kilburn, G. J. Langley, P. D. Edwards, T.
Mitchell, R. Southgate, Tetrahedron Lett. 1994, 35, 481.
[4] a) F. Garcia-Tellado, S. Goswami, S. K. Chang, S. J. Geib, A. D.
Hamilton, J. Am. Chem. Soc. 1990, 112, 7393; b) F. Garcia-Tellado, J.
Albert, A. D. Hamilton, J.Chem. Soc. Chem. Commun. 1991, 1761.
[5] For preliminary communications of parts of this work, see: a) J.
Dowden, P. D. Edwards, J. D. Kilburn, Tetrahedron Lett. 1997, 38,
1095; b) S. S. Flack, J. D. Kilburn, Tetrahedron Lett. 1995, 36, 3409.
[6] O. Mitsunobu, J. Kimura, K. Iizumu, N. Yanagida, Bull. Chem. Soc.
Jpn. 1976, 49, 510.
[7] a) H. C. Brown, Y. M. Choi, S. Narasimhan, Synthesis 1981, 605;
b) H. C. Brown, Y. M. Choi, S. Narasimhan, J. Org. Chem. 1982, 47,
3153.
[8] H. Azizian, C. Eaborn, A. Pidcock, J. Organomet. Chem. 1981, 215, 49.
[9] J. K. Stille, Angew. Chem. 1986, 98, 504; Angew. Chem. Int. Ed. Engl.
1986, 25, 508.
[10] J. Malm, P. BjoÈrk, S. Gronowitz and A. B. HoÈrnfeldt, Tetrahedron Lett.
1992, 33, 2199.
[11] K. Venkataraman, D. R. Wagle, Tetrahedron Lett. 1979, 20, 3037.
[12] R. Knorr, A. Trzeciak, W. Bannwarth, D. Gillessen, Tetrahedron Lett.
1989, 30, 1927.
[13] M. Bodanszky, A. Bodanszky, The Practice of Peptide Synthesis,
Springer, Berlin, 1984.
[14] C. D. Hurd, R. Roach, C. W. Huffman, J. Am. Chem. Soc. 1956, 78,
104.
with hexane. [a]D
nÄ ˆ 3392, 2919, 1635, 1584, 1509, 1449, 1289, 1241 cm
ˆ
16.3 (c ˆ 0.3 in CHCl3/MeOH, 1:1 v/v); IR (CH2Cl2):
1
;
1H NMR
(360 MHz, CDCl3): d ˆ 7.97 (s, 1H, NHCOCHCH2Ph), 7.91 (d, 1H, J ˆ
8 Hz, m-pyrH), 7.78 (d, 1H, J ˆ 8 Hz, m-pyrH), 7.71 (d, 2H, J ˆ 8 Hz, ArH),
7.65 (t, 1H, J ˆ 8 Hz, p-pyrH), 7.63 (d, 2H, J ˆ 8 Hz, ArH), 7.42 (d, 2H, J ˆ
8 Hz, ArH), 7.37 (d, 2H, J ˆ 8 Hz, ArH), 7.28 (s, 1H, NHCOCH2Ar), 7.26 ±
7.12 (m, 5H, Ph), 7.05 (d, 2H, J ˆ 9 Hz, ArH), 6.78 (d, 2H, J ˆ 9 Hz, ArH),
6.25 (d, 1H, J ˆ 8 Hz, PhCH2CHNH), 5.95 (t, 1H, J ˆ 5 Hz, CH2NH), 5.30
(s, 2H, CH2O), 4.62 (m, 1H, CHCH2Ph), 4.41 (dd, 1H, J ˆ 6 Hz, 14 Hz,
CHAHBNH), 4.06 (dd, 1H, J ˆ 5 Hz, 14 Hz, CHAHBNH), 3.80 (s, 2H,
NHCOCH2Ar), 3.19 (dd, 1H, J ˆ 7 Hz, 14 Hz, CHAHBPh), 2.96 (dd, 1H,
J ˆ 8 Hz, 12 Hz, CHAHBPh), 2.52 ± 2.13 (4H, m, CH2CH2); 13C NMR
(75.5 MHz, 5% CD3OD in CDCl3): d ˆ 173.2, 172.8, 169.5, 169.1, 156.9,
149.3, 148.9, 140.8, 139.8, 138.9, 137.0, 136.2, 133.1, 130.5, 130.4, 129.4, 129.3,
128.9, 128.1, 128.0, 127.3, 127.1, 116.6, 110.0, 109.8, 69.3, 55.8, 45.0, 43.5, 37.5,
‡
31.3, 31.2; LRMS (ES ) 668 (100); HRMS (FAB) calcd for C40H37N5O5:
668.2874; found 668.2863.
General procedure for NMR titration experiments
Macrocycle 1 was first dissolved in CHCl3, washed with saturated aqueous
NaHCO3, and brine, then dried (MgSO4), filtered and the solvent removed
under reduced pressure. All deuterochloroform used was passed through
alumina and collected over activated 4Š molecular sieves. A standard
solution of 1 was prepared immediately prior to use by dissolving a known
amount (typically ꢀ4 mgs) in CDCl3 (2 mL). A known aliquot of this
solution (either 500 or 600 mL) was transferred to a dry NMR tube. A
standard solution of the guest solution at 10 ± 20 times the concentration of
the host solution was similarly prepared. After recording the spectra of the
free host solution, aliqouts of the guest solution were added and the NMR
spectra recorded after each addition. The chemical shift of the most
downfield pyr-NHCO relative to residual chloroform (d ˆ 7.27) was
recorded. The signals corresponding to the other NH protons were
observed to move during the titration but were either too broad or became
obscured as they merged with the aromatic signals and were consequently
not monitored throughout the titration. The volume of the added aliquots
of guest solution varied from addition to addition and from experiment to
experiment so as to obtain the best range of data points for analysis. The
experimental data so obtained was analysed by curve fitting to obtain the
association constant (Ka) using HOSTEST 5.0.[23] The quality of the curve
fit was assessed by stochastic analysis of the curve, also using HOSTEST
5.0, and is indicated by the associated errors (see Table 2).
[15] J. Coste, D. Le-Nguyen, B. Castro, Tetrahedron Lett. 1990, 31, 205.
[16] J. Jones, The Chemical Synthesis of Peptides, Clarendon Press, Oxford,
1991.
[17] V. Farina, S. Kapadia, B. Krishnan, C. Wang, L. S. Liebeskind, J. Org.
Chem. 1994, 59, 5905.
[18] V. Farina, B. Krishnan, D. R. Marshall, G. P. Roth, J. Org. Chem. 1993,
58, 5434.
[19] V. Farina, B. Krishnan, J. Am. Chem. Soc. 1991, 113, 9585.
[20] a) L. A. Carpino, D. Sadat-Aalaee, H. G. Chao, R. H. DeSelms, J.
Am. Chem. Soc. 1990, 112, 9651; b) J. N. Bertho, A. Loffet, C. Pinel, F.
Reuther, G. Sennyey, Tetrahedron Lett. 1991, 32, 1303.
[21] R. Deziel, Tetrahedron Lett. 1987, 28, 4371.
[22] A. Bax, D. G. Davis, J. Magn. Reson. 1985, 63, 207.
[23] Hostest 5 software: C. S. Wilcox, N. M. Glagovich, University of
Pittsburgh, USA, C. S. Wilcox and the University of Pittsburgh, 1993.
See also: a) C. S. Wilcox Design, Synthesis, and Evaluation of an
Efficacious Functional Group Dyad. Methods and Limitations in the
Use of NMR for Measuring Host ± Guest Interactions, in Frontiers in
Supramolecular Organic Chemistry and Photochemistry, (Ed.: H. J.
Schneider, H. Dürr) VCH, Weinheim, 1991; b) C. S. Wilcox, J. C.
Adrian, Jr, T. H. Webb, F. J. Zawacki, J. Am. Chem. Soc., 1992, 114,
10189.
Acknowledgements
We thank the EPSRC and SmithKline Beecham for a CASE award (J. D.),
Dr Jonathan Essex for assistance with the molecular modelling studies and
Peter Henley for assistance in preparing this manuscript.
[24] a) K. A. Connors, Binding Constants, The Measurement of Molecular
Complex Stability, Wiley, 1987, 24 ± 28; b) A. Job, Annales de Chimie
1928, 9, 113.
[25] G. Weber, Molecular Biophysics; (Ed.: B. Pullman, M. Weissbluth)
Academic Press, New York, 1965, 369 ± 397.
[1] For reviews see a) J. J. B. Perry, J. D. Kilburn, Contemp. Org. Synth.
1997, 4, 61; b) Tetrahedron Symposium-In-Print on Molecular Recog-
nition, (Ed.: A. D. Hamilton) Tetrahedron 1995, 51, 343; c) T. H.
Webb, C. S. Wilcox, Chem. Soc. Rev. 1993, 22, 383.
[2] For recent examples of synthetic receptors for amino acid and peptide
derivatives, see for example: a) R. J. Pieters, J. Cuntze, M. Bonnet, F.
Diederich, J. Chem. Soc., Perkin Trans. 2 1997, 1891; b) W. C. Still,
Acc. Chem. Res. 1996, 29, 155; c) D. W. P. M. LoÈwik, S. J. E. Mulders,
[26] Comparison of binding constants of structurally different carboxylic
acids should be treated with some caution since the strength of the
interaction with the amidopyridine unit will be influenced by the
acidity of the carboxylic acid moiety, although differences in the series
of substrates presented here seem unlikely to be larger than the
experimental errors in the measured binidng constants.
88
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0501-0088 $ 17.50+.50/0
Chem. Eur. J. 1999, 5, No. 1

Macrocyclic Peptide Receptor
79±89
[27] The substantially stonger binding of N-Cbz-b-alanyl-l-alanine com-
pared with N-Cbz-b-alanyl-l-lactic acid might alternatively result
from the replacement of the relatively good hydrogen bond acceptor
provided by the alanyl amide carbonyl with the less effective lactate
ester carbonyl, and as a consequence of the ester being inherently
more flexible than the amide.
[28] C. T. Walsh, S. L. Fisher, I. S. Park, M. Prahalad, Z. Wu, Chemistry and
Biology 1996, 3, 21.
[29] R. J. Dancer, A. C. Try, G. J. Sharman, D. H. Williams, J. Chem. Soc.
Chem. Commun., 1996, 1445.
[31] D. Q. McDonald, W. C. Still, Tetrahedron Lett. 1992, 33, 7742.
[32] a) W. D. Cornell, P. Ceiplak, C. I. Baoyly, I. R. Gould, K. M. Merz,
Jr., D. M. Ferguson, D. C. Spellmeyer, T. Fox, J. W. Caldwell, P. A.
Kollman, J. Am. Chem. Soc. 1995, 117, 5179; b) S. J. Weiner, P. A.
Kollman, D. A. Case, U. C. Singh, C. Chio, G. Alagona, S. Profeta, Jr.,
P. Weiner, J. Am. Chem. Soc. 1984, 106, 765; c) S. J. Weiner, P. A.
Kollman, D. A. Case, J. Comput. Chem. 1986, 7, 230.
[33] W. C. Still, A. Tempczyk, R. C. Hawley, J. Am. Chem. Soc. 1990, 112,
6127.
[34] D. D. Perrin, W. L. F. Armarego, Purification of Laboratory Chem-
icals, Pergamon Press, Oxford, 3rd ed, 1989.
[30] F. Mohamadi, N. G. J. Richards, W. C. Guida, R. Liskamp, M. Lipton,
C. Caufield, G. Chang, T. Hendrickson, W. C. Still, J. Comput. Chem.
1990, 11, 440.
Received: April 28, 1998 [F 1123]
Chem. Eur. J. 1999, 5, No. 1
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0501-0089 $ 17.50+.50/0
89