Artificial oligopeptide scaffolds for stoichiometric metal binding

Article abstract of DOI:10.1021/ja0510818

Two artificial peptides with pendant pyridine or bipyridine ligands have been synthesized and incorporated into oligomeric strands that are analogous to peptide nucleic acid. Spectrophotometric titrations with Cu²⁺ and Fe²⁺ show that

Full text of DOI:10.1021/ja0510818

Published on Web 06/14/2005
Artificial Oligopeptide Scaffolds for Stoichiometric Metal
Binding
Brian P. Gilmartin, Kristi Ohr, Rebekah L. McLaughlin, Richard Koerner, and
Mary Elizabeth Williams*
Contribution from the Department of Chemistry, The PennsylVania State UniVersity,
104 Chemistry Building, UniVersity Park, PennsylVania 16802
Received February 19, 2005; E-mail: mbw@chem.psu.edu
Abstract: Two artificial peptides with pendant pyridine or bipyridine ligands have been synthesized and
incorporated into oligomeric strands that are analogous to peptide nucleic acid. Spectrophotometric titrations
with Cu2+ and Fe2+ show that the oligomers bind stoichiometric quantities of transition metals based on
the number of pendant ligands. The identities of the titration products are confirmed by high resolution
mass spectrometry. In the case of the bipyridine tripeptides, the titration stoichiometry and mass spectra
indicate that the metal ions form interstrand cross-links between two oligopeptides, creating duplex structures
linked exclusively by metal ions. Calculated molecular structures of the metalated oligopeptides and duplexes
indicate that the peptide backbone acts as a scaffold for the directed assembly of metal ions. Electron
paramagnetic resonance spectroscopy of the Cu-containing molecules have varying degrees of electronic
interaction based on their charge and supramolecular structure. Cyclic voltammetry of the Fe2+- and Cu2+
-
linked bpy oligopeptide duplexes shows that they possess unique electrochemical signatures based on
the redox reactivity of the metal complex.
Introduction
The molecular structure of deoxyribonucleic acid (DNA)
contains the instructions for the production of genes and provides
a uniquely elegant template for self-replication in biological
systems. The double helical structure of DNA was first described
by Watson and Crick in the middle part of the previous century,
and it is well-known that the molecular recognition between
individual strands making up this duplex results from hydrogen
bonding between complementary base pairs that decorate the
sugar phosphate backbone. Using DNA as a model, we
synthesize artificial oligomeric peptides that are designed to
exhibit the conformational structure of DNA but which self-
assemble upon transition metal chelation. The metal binding
base pairs should enhance the stability and binding affinity of
duplex structures. In comparison with DNA, the wider range
of transition metal complexes (metals and ligands) available for
use in these structures greatly increases the number of possible
complementary components.1
While nucleic acids (and in some cases the phosphate
backbone) in DNA bind metal ions, their chelation affinities
are relatively weak, leading to stoichiometries and geometries
that are not well-defined or controllable. Construction of
synthetic analogues allows the incorporation of distinct, high-
affinity metal binding sites that circumvent these problems. To
additionally avoid metal interaction with the sugar phosphates,
Figure 1. Structure of the polyamide backbone showing positions of
attachments of x number of pendant ligrands. Below are the two ligands
used in this study, where p indicates the point of attachment to the peptide
backbone.
we instead have chosen to utilize a peptidic backbone. The
oligomer identified for our initial studies of the inorganic
helicates is identical to that used for peptide nucleic acids
(PNAs).2 This peptide was chosen based on its (1) stability
toward the enzymatic and chemical processes that can degrade
the sugar-phosphate backbone of DNA and (2) easy adaptation
toward large-scale automated synthesis. We utilize nitrogen-
containing heterocyclic ligands with predictable binding affini-
ties for transition metal ions to covalently link to the PNA
peptide backbone in place of nucleic acids, as shown in Figure
1. These artificial peptide oligomers therefore serve as scaffolds
(1) (a) Henry, A. A.; Olsen, A. G.; Matsuda, S.; Yu, C.; Ge´ıerstanger, B. H.;
Romesberg, F. E. J. Am. Chem. Soc. 2004, 126, 6923-6931. (b) Ogawa,
A. K.; Wu, Y.; McMinn, D. L.; Liu, J.; Schultz, P. G.; Romesberg, F. E.
J. Am. Chem. Soc. 2000, 122, 3274-3287. (c) Neuner, P.; Monaci, P.
Bioconjugate Chem. 2002, 13, 676-678.
(2) Nielsen, P. E.; Egholm, M.; Berg, R. H.; Buchardt, O. Science 1991, 254,
1497-1500.
9
9546
J. AM. CHEM. SOC. 2005, 127, 9546-9555
10.1021/ja0510818 CCC: $30.25 © 2005 American Chemical Society

Artificial Oligopeptide Scaffolds for Metal Binding
A R T I C L E S
quantities of transition metal ions to form multimetallic single
and double stranded structures, and examines their electron
paramagnetic resonance spectra and electrochemical properties.
Experimental Section
Chemicals. All materials were purchased from Aldrich and used as
received unless otherwise noted. N,N-Diisopropylethylamine (DIPEA,
Avocado) was distilled over CaH2, and CH2Cl2 (VWR) was dried on
an activated alumina column. For all experiments, ultrapure water was
used (Labconco Water Pro PS system, 18.2 MΩ). Fmoc-aeg-OtBu‚
HCl,6 aquo 2,2′:6′,2′′-terpyridine copper(II) diperchlorate7 (i.e.,
[Cu(tpy)(H2O)](ClO4)2), and bis(aquo) pyridine-2,6-dicarboxyl cop-
per(II)8 (i.e., [Cu(pda)(H2O)2]) were prepared according to previously
published procedures.
Oligomers were synthesized on an Fmoc-PAL-PEG-PS resin (Ap-
plied Biosystems). A 4-molar excess of the monomers and amino acids
was used during coupling reactions; the targeted oligomer was
synthesized on a 0.1 mmol scale, as determined by the maximum
loading level of the resin.
Instrumentation and Analysis. High performance liquid chroma-
tography (HPLC) was performed on a Varian system equipped with
two quaternary pumps (model 210), an autosampler (model 410), UV-
visible detector (model 320), and fraction collector (model 701).
Preparatory scale separations were performed with a 100 × 20 mm2
C18 column (S-5 µm, 12 nm, YMC, Co.) and a 2 mL injection loop.
Elution of the product was detected using the pyridine absorbance at
255 nm.
Figure 2. Side and top views of a calculated molecular model of a synthetic
oligomeric helicate, formed by interstrand cross-linking of peptide strands
by three metal ions.
Positive ion electrospray ionization (ESI+) mass spectrometry was
performed on a Mariner mass spectrometer (PerSeptive Biosystems).
Theoretical mass spectrometry peaks and relative intensities were
calculated using software available at http://www2.sisweb.com/mstools/
isotope.htm.
Electrochemical data were collected with a CH Instruments model
660A potentiostat with a picoamp booster. Solutions were prepared
from doubly distilled acetonitrile and thoroughly deoxygenated. Vol-
tammetry was obtained using a 12.5 micron radius Pt working electrode,
2 mm diameter Pt wire counter electrode, and 22 gauge Ag wire quasi-
reference electrode. The obtained voltammograms were manually
corrected for uncompensated resistance.
for the stoichiometric binding of metals in a geometrically
defined arrangement.
We predict that when metal ions are used to link these
artificial peptides, a structure similar to those calculated and
shown in Figure 2 will result. In these synthetic model systems,
molecular recognition and interchain linking between the new
“base pairs” is based entirely on a coordination chemistry motif
rather than hydrogen bonding.3,4 In addition to mimicking the
helical structure of DNA, these multimetallic analogues mimic
the information storage function of DNA, in that they are
expected to exhibit unique optical and electronic properties that
are strongly dependent on the peptide sequence and coordinated
metals. The metal-decoroated peptides therefore provide the
basis for new motifs of functional self-assembled nanostructures
for possible use as inorganic “bar codes” and may provide
opportunities for the development of biocompatible transition
metal pharmacological agents.5
Circular dichroism (CD) spectroscopy was performed using a Jasco
J-810 spectropolarimeter with a quartz cell with an optical path length
of 1 mm. Experiments were performed at room temperature with a
bandwidth of 1 nm and wavelength increment of 1 nm.
All NMR spectra were obtained with either a Bruker 300 or 400
MHz spectrometer at room temperature. X-band EPR spectra were
obtained using a 9.5 GHz Bruker eleXsys 500 spectrometer equipped
with a liquid helium cryostat. All experiments were performed at 16
K, with a modulation frequency of 100 kHz and modulation amplitude
of 5 G. For [Cu(2)2]3(PF6)6 and [Cu(pda)(1)]6, the microwave power
was 8.23 mW; for [Cu(tpy)(1)]6(ClO4)12, it was 0.15 mW.
Molecular structures were calculated using Hyperchem 6.0 using
molecular mechanics (MM+) with atomic charge based electrostatic
repulsions and a Polack-Ribiere conjugate gradient to a minimum
energy gradient of 0.01 kcal/mol.
This paper presents the synthesis and characterization of
oligopeptides bearing pyridine (py) and 2,2′-bipyridine (bpy)
ligands, demonstrates the ability of these to bind stoichiometric
(3) (a) Zelder, F. H.; Mokhir, A. A.; Kra¨mer, R. Inorg. Chem. 2003, 42, 8618-
8620. (b) Mokhir, A.; Stiebing, R.; Kraemer, R. Bioorg. Med. Chem. 2003,
13, 1399-1401. (c) Verheijen, J. C.; van der Marel, G. A.; van Boom, J.
H.; Metzler-Nolte, N. Bioconjugate Chem. 2000, 11, 741-743.
(4) (a) Popescu, D.-L.; Parolin, T. J.; Achim, C. J. Am. Chem. Soc. 2003, 125,
6354-6355. (b) Atwell, S.; Meggers, E.; Spraggon, G.; Schultz, P. J. Am.
Chem. Soc. 2001, 123, 12364-12367. (c) Meggers, E.; Holland, P. L.;
Tolman, W. B.; Romesberg, F. E.; Schultz, P. J. Am. Chem. Soc. 2000,
122, 10714-10715. (d) Tanaka, K.; Tengiji, A.; Kato, T.; Toyama, N.;
Shionoya, M. Science 2003, 299, 1212-1213. (e) Tanaka, K.; Shionoya,
M. J. Org. Chem. 1999, 64, 5002-5003. (f) Weizman, H.; Tor, Y. J. Am.
Chem. Soc. 2001, 123, 3375-3376. (g) Weizman, H.; Tor, Y. Chem.
Commun. 2001, 5, 453-454.
(5) (a) Blower, P. J.; Dilworth, J. R.; Mullen, R. I.; Reynolds, C. A.; Zheng,
Y. J. Inorg. Biochem. 2001, 85, 15-22. (b) Kurosaki, H.; Sharma, R. K.;
Aoki, S.; Inoue, T.; Okamoto, Y.; Sugiura, Y.; Doi, M.; Ishida, T.; Otsuka,
M.; Goto, M. J. Chem. Soc., Dalton Trans. 2001, 441-447. (c) Lebon, F.;
Boggetto, N.; Ledecq, M.; Durant, F.; Benatallah, Z.; Sicsic, S.; Lapouyade,
R.; Kahn, O.; Mouithys-Mickalad, A.; Deby-Dupont, G.; Reboud-Ravaux,
M. Biochem. Pharm. 2002, 63, 1863-1873.
Synthesis. A. Fmoc-aeg(py)-OH‚HCl (1): A mixture of 4.057 g
(0.010 23 mol) of Fmoc-aeg-OtBu‚HCl, 2.374 g of 1-[(3-dimethylamino)-
propyl]-3-ethylcarbodiimide (EDC, 0.012 38 mol), 2.200 g (0.012 67
mol) of 4-pyridylacetic acid hydrochloride, 6.50 mL (0.0373 mol) of
(6) Thomson, S. A.; Josey, J. A.; Cadilla, R.; Gaul, M. D.; Hassman, C. F.;
Luzzio, M. J.; Pipe, A. J.; Reed, K. L.; Ricca, D. J.; Wiethe, R. W.; Noble,
S. A. Tetrahedron 1995, 51, 6179-6194.
(7) Ohr, K. The Pennsylvania State University, unpublished results.
(8) (a) Murtha, D. P.; Walton, R. A. Inorg. Chim. Acta 1974, 8, 279-284. (b)
Ullah, M. R.; Bhattacharya, P. K. Indian J. Chem., Sect. A 1991, 976-
978.
9
J. AM. CHEM. SOC. VOL. 127, NO. 26, 2005 9547

A R T I C L E S
Gilmartin et al.
DIPEA, and 300 mL of CH2Cl2 was stirred for 1 h under N2. The yellow
solution was extracted with H2O (5 × 100 mL). The combined aqueous
washings were extracted with CH2Cl2 (50 mL). The combined organic
extracts were dried over anhydrous Na2SO4 and flash evaporated to
give a yellow oil. The oil was stirred in an aqueous solution of 3 M
HCl (150 mL). The acid was removed under reduced pressure to yield
a yellow oil, which was dried for 16 h under vacuum. The pure product
was obtained by recrystallization from CH3CN, followed by vacuum-
drying for 16 h. Yield ) 2.235 g (48.1%) (1H NMR, 300 MHz, d6-
DMSO): 3.08 (m, 1 H); 3.18 (m, 2 H); 3.29 (m, 2 H); 3.40 (m, 2 H);
3.91 (d, 2 H); 4.05 (s, 1 H); 4.13 (m, 1 H); 4.22 (m, 3 H); 7.23 (t, 2
H); 7.31 (t, 2 H); 7.41 (t, 1 H); 7.59 (d, 2 H); 7.72 (t, 2 H); 8.80 (d, 2
H); 8.71 (d, 2 H).
COSY, HMQC, and HMBC NMR; see Supporting Information. (ESI+)
Calculated: (m + H)+ ) 1493.7, (m + Na)+ ) 1515.6. Found: (m +
H)+ ) 1493.8, (m + Na)+ ) 1515.7.
D. Bz-(2)3G-NH2. A 25 mL fritted polypropylene reservoir (Alltech)
was used in the synthesis of the oligomer. The reservoir was heated to
∼50 °C with electric heating tape (Barnstead/Thermolyne), and N2 was
bubbled through the DMF solution. The synthesis uses 20% piperidine
in DMF for Fmoc-deprotection (15 min), and solutions of 0.5 M HBTU
and 0.5 M DIPEA for monomer and amino acid coupling (30 min),
and 0.5 M benzoic anhydride and 0.5 M DIPEA in DMF for capping
(5 min). Each coupling step was monitored for completion using a
Kaiser test.10 After the desired number of cycles, the resin was washed
3 × 10 mL with DMF, followed by 3 × 10 mL CH2Cl2. The oligomer
was cleaved from the resin by stirring the resin with 10 mL of 2.5%
H2O, 2.5% TIS in TFA for 2 h. The mixture was filtered in 4 equal
volumes through a glass frit into 4 × 40 mL of cold Et2O. The oligomer
was collected as pale yellow pellets by centrifugation, and the pellets
washed with 3 × 30 mL with Et2O and dried under vacuum for 1 h.
The desired oligomer was separated from deletion sequences using
an HPLC gradient elution of 10% (0.1% TFA in CH3CN):90% (0.1%
TFA in H2O) ramped to 28% (0.1% TFA in CH3CN):72% (0.1% TFA
in H2O) over 15 min at 20 mL/min. After fraction collection, the
oligomer was lyophilized to give a pale pink solid. Overall Bz-(2)3G-
NH2 yield ) 1.2 mg (1.19%). (1H NMR, 400 MHz, MeOH-d4): 2.22-
2.41 (m, 9 H); 3.38-4.32 (m, 26 H); 7.00-8.59 (m, 23 H).
Identification of these peaks and confirmation of purity were ac-
complished with HMQC NMR; see Supporting Information for detailed
analysis. (ESI+) Calculated: (m + H)+ ) 1109.5. Found: (m + H)+
) 1109.4, (m + Na)+ ) 1131.5, (m + K)+ ) 1147.5.
Spectrophotometric Titrations. A. Bz-(1)6G-NH2 + [Cu(pda)-
(H2O)2] or [Cu(tpy)(H2O)2](ClO4)2. Titrations were performed using
a Varian Cary 500 spectrophotometer. A solution of the Bz-(1)6G-NH2
was prepared in methanol (MeOH). The concentration of the oligomer
(2.222 mM) was determined using its molar extinction coefficient (ꢀ
) 31 860 M-1 cm-1) at 256 nm. This solution was titrated in 5, 10,
and 20 µL increments into 2.5 mL of a 0.6608 mM [Cu(pda)(H2O)2]
methanolic solution at 4 min intervals between each addition. For each
addition, the reference solution was diluted with an equal volume of
MeOH to account for concentration effects. UV-visible absorption
spectra were acquired for each iterative addition of oligomer. The
titration product was precipitated by slow addition of ether, filtered,
rinsed with ether, and reprecipitated from methanol. Titrations with
[Cu(tpy)(H2O)](ClO4)2 were performed analogously using a 2.804 mM
methanolic solution containing a small amount (3.361 µM) H2O. The
titration product was isolated as for the [Cu(pda)] adduct.
B. Fe(ClO4)2 + Bz-(2)3G-NH2. Titrations were performed on a
Varian Cary 50 spectrophotometer. A solution of Bz-(2)3G-NH2 was
prepared, and the concentration was determined to be 70.8 µM using
its molar extinction coefficient (ꢀ ) 6322.6 M-1 cm-1) at 284 nm. A
solution of 784 µM [Fe(ClO4)2] in MeOH was added in 2 µL increments
to the oligomer solution in 5 min intervals, and UV-visible absorption
spectra were obtained for each addition. The titration products were
precipitated with NH4PF6, the supernatant was decanted, and the solid
was rinsed with copious amounts of water.
C. Cu(OAc)2 + Bz-(2)3G-NH2. This titration was performed
similarly to that from the Fe(ClO4)2 experiment, except that a 734.5
µM Bz-(2)3G-NH2 solution in MeOH was titrated with a solution of
9.29 mM Cu(OAc)2 in MeOH in 20 µL increments at 15 min intervals.
Spectra were corrected for the background absorbance of unbound Cu
acetate. The titration product was isolated and purified as above.
Elemental Analysis Calculated: 47.99%, C; 4.64%, H; 11.87%, N.
Found: 47.72%, C; 4.96%, H; 12.04%, N.11
B. Fmoc-aeg(bpy)-OH‚HCl (2): Fmoc-aeg-OtBu‚HCl (1.333 g,
0.003 08 mol) was dissolved in CHCl3 (100 mL) and washed with an
aqueous solution of saturated NaHCO3 (3 × 100 mL). The washings
were extracted with CHCl3 (50 mL). The organics were then dried over
Na2SO4, and the solvent was removed under vacuum leaving a clear
oil. A solution of 4′-methyl-2,2′-bipyridine-4-acetic acid9 (1.004 g,
0.004 40 mol), 2-(1H-benzotriazole-1-yl)-1,1,3,3-tetramethyluronium
hexafluorophosphate (HBTU, Novabiochem, 2.236 g, 0.005 90 mol),
and DIPEA (1.015 mL, 0.005 84 mol) in N,N-dimethylformamide
(DMF, 100 mL) was prepared, added to the oil, and was stirred under
N2 for 2 h. The tert-butyl-protected product was then precipitated by
addition of the solution to ice water (1.3 L). The precipitate was
collected by filtration and dried in vacuo overnight, leaving a yellow
solid. The solid was dissolved in a solution of 2.5% water in
trifluoroacetic acid (TFA, 10 mL). The solution was stirred for 2 h
and then precipitated from cold ethyl ether (Et2O, 200 mL), giving the
product as an off-white powder. Yield ) 1.014 g (49.6%) (1H NMR,
300 MHz, d6-DMSO): 2.50 (s, 3 H); 3.17 (t, 1 H); 3.28 (t, 1 H); 3.38
(t, 1 H); 3.49 (t, 1 H); 3.84 (s, 1 H); 4.00 (d, 2 H); 4.25 (m, 4 H); 7.29
(t, 2 H); 7.40 (t, 2 H); 7.53 (m, 2 H); 7.65 (t, 2 H); 7.87 (d, 2 H); 8.37
(s, 2 H); 8.65 (m, 2 H). (ESI+) Calculated: (m + H)+ ) 551.2.
Found: (m + H)+ ) 551.3.
C. R1-(1)6R2-NH2 (R1 ) Acetyl (Ac) or Benzyl (Bz); R2 ) Lysine
(K) or Glycine (G)). The oligomer was synthesized at room temperature
on a Pioneer peptide synthesis system (Applied Biosystems) with DMF
as the solvent. A solution of 20% piperidine in DMF (Applied Bio-
systems) was used for Fmoc-deprotection (5 min), and 0.5 M diiso-
propylcarbodiimide (DIPCDI, Applied Biosystems) and 0.5 M DIPEA
were used for the couplings (30 min). A capping cycle was performed
after every coupling with 0.5 M benzoic anhydride (Avocado) or 0.5
M acetic anhydride and 0.5 M DIPEA (5 min).
Following synthesis, the resin was washed 3 × 10 mL with DMF,
followed by 3 × 10 mL alternating 2-propanol and CH2Cl2. The
oligomer was cleaved from the resin by stirring the resin with 10 mL
of 2.5% H2O, 2.5% triisopropylsilane (TIS) in TFA for 2 h. The mixture
was filtered in 4 equal volumes through a glass frit into 4 × 40 mL of
cold Et2O. The solutions were mixed and kept at 0 °C for 1 h. The
oligomer was collected by centrifugation as pale yellow pellets and
washed 3 × 10 mL with Et2O and dried under vacuum for 1 h.
Purification of the oligomer (the major product) was accomplished
by preparatory scale HPLC using a gradient elution of 5% (0.1% TFA
in CH3CN):95% (0.1% TFA in H2O) to 15% (0.1% TFA in CH3CN):
85% (0.1% TFA in H2O) ramped over 13 min with a total flow of 20
mL/min and a fraction collection program. TFA and CH3CN were
removed from the collected fractions under reduced pressure, and H2O
was removed by lyophilization. Overall [Bz-(1)6G-NH2] yield ) 0.072
g (33%) (1H NMR, 300 MHz, d6-DMSO): 3.15-3.75 (m, 26 H); 3.75-
4.30 (m, 24 H); 7.08 (s, 1 H); 7.21 (s, 1 H); 7.50 (m, 4 H); 7.71 (m,
14 H); 8.01 (m, 1 H); 8.30 (m, 3 H); 8.50 (m, 2 H); 8.70 (m, 12 H).
Further confirmation of purity and identity were accomplished with
(10) Kaiser, E.; Colescott R. L.; Bossinger, C. D.; Cook, P. I. Anal. Biochem.
1970, 34, 595-598
(11) Insufficient quantities of material were available to quantitatively determine
the percentage of metal ion in the sample.
(9) Della Ciana, L.; Hamachi, I.; Meyer, T. J. J. Org. Chem. 1989, 54, 1731-
1735.
9
9548 J. AM. CHEM. SOC. VOL. 127, NO. 26, 2005

Artificial Oligopeptide Scaffolds for Metal Binding
A R T I C L E S
Scheme 1. Synthesis of Ligand-Peptide Monomers and Structures of the Oligomers; Ac ) Acetyl; Bz ) Benzoyl; G ) Glycine; K ) Lysine
Results and Discussion
indicated in Scheme 1. These slight modifications in the peptide
backbone have no substantial effects on the physical properties
of the py oligomers. For the sake of clarity, we detail the
synthesis and metal binding of the two oligomer sequences
shown in Scheme 1, a py hexamer, Bz-(1)6G-NH2, and bpy
trimer, Bz-(2)3G-NH2.
Synthesis. Py and bpy were chosen as the initial target ligands
for incorporation onto the peptide backbone because of their
wide use in coordination chemistry for a variety of transition
metals. To prepare peptides incorporating these pendant ligands,
we use standard solid-phase synthesis methods. The peptide/
ligand monomers were prepared according to the method shown
in Scheme 1 from their acetic acid derivatives,12,13 by adaptation
of a literature procedure.6 Briefly, the ligands were reacted with
Fmoc-protected N-[2-aminoethyl] glycine tert-butyl ester (Fmoc-
aeg-OtBu‚HCl) using N,N-diisopropylethylamine (DIPEA) and
the coupling reagent shown in Scheme 1. Amide coupling to
the py ligand was accomplished with DIPEA and EDC at 25
°C for 1 h. The bpy analogue required the use of DIPEA and
HBTU. For both ligands, the terminal acid was deprotected by
acid cleavage of the tert-butyl to yield monomers 1 and 2 with
overall yields 48% and 50%, respectively.
The oligomers were cleaved from the resin using an excess
quantity of 2.5% water and 2.5% triisopropylsilane (TIS) in
trifluoroacetic acid (TFA), giving the oligomers shown in
Scheme 1. The products were separated from deletion sequences
and reaction byproducts by reverse-phase high-performance
liquid chromatography (HPLC), and the desired fraction was
collected and lyophilized to yield the purified oligomer. For
Bz-(1)6G-NH2, the final yield of the hexamer was 33% (based
on the loading of the resin) and purity was confirmed by NMR
and mass spectrometry. The typical yield of the Bz-(2)3G-NH2
was 1.19%, with purity again confirmed by NMR and mass
spectrometry. Detailed analysis of the two-dimensional NMR
spectra (COSY and HMQC), which confirms the purity of the
oligopeptides, is found in the Supporting Information. The
significantly lower yield of the bpy oligomer is thought to arise
from both steric hindrance and aggregation induced by π-stack-
ing of the ligands, and our ongoing efforts seek to improve the
overall yield for this and other multidentate ligand oligomers.14
Oligomers were prepared from monomers 1 and 2 by solid-
phase synthetic methods6 on a resin support (Fmoc-PAL-PEG-
PS, Applied Biosystems). Oligomers were typically prepared
using an automated peptide synthesizer (Applied Biosystems
Pioneer System), which enables the facile preparation of
sequences of varying length and composition.7 In some cases,
resin-supported synthesis was performed by hand. For the py
hexamer, we prepared both the lysine- and glycine-linked
oligomers and the benzoyl- and acetyl-terminated chains, as
Metal Coordination to Pyridine Hexamer. We next inves-
tigated the ability of the artificial peptide oligomers to bind
transition metal ions. Since it is well-known that metal
(12) Sasse, W. H. F.; Whittle, C. P. J. Chem. Soc. 1961, 83, 1347-1350.
(13) Kim, B. H.; Lee, D. N.; Park, H. J.; Min, J. H.; Jun, Y. M.; Park, S. J.;
Lee, W.-Y. Talanta 2004, 62, 595-602.
(14) Gilmartin, B. P.; Pantzar, L. The Pennsylvania State University, unpublished
results.
9
J. AM. CHEM. SOC. VOL. 127, NO. 26, 2005 9549

A R T I C L E S
Gilmartin et al.
[Cu(pda)(py)] absorption peak begins to level at a maximum
value of ∼6 molar equivalents of [Cu(pda)] per mole of Bz-
(1)6G-NH2.15 Metal complexation of the acetyl capped, lysine
analogue (i.e., Ac-(1)6K-NH2) has the same metal binding
stoichiometry (not shown; see mass spectrometry data in the
Supporting Information). These spectrophotometric data strongly
suggest the formation of an oligomeric strand of Bz-(1)6G-NH2
with six bound [Cu(pda)] centers, (Bz-[Cu(pda)(1)]6G-NH2).
High-resolution ESI+ mass spectrometry was employed to
analyze the purified titration product; the observed molecular
ion peak at 1460.7 m/z (z ) +2, (m + 2Na+)) agrees well with
the calculated mass of 2869.4 amu (i.e., an expected (m + 2Na+)
z ) +2 peak at 1459.7 m/z), confirming the py hexamer with
a stoichiometric quantity of bound [Cu(pda)] is the titration
product (Bz-[Cu(pda)(1)]6G-NH2).
To test the effect of ionic charge on the addition of complexes
to the pyridine oligomers, Bz-(1)6G-NH2 was also titrated with
[Cu(tpy)](ClO4)2. Figure 3B shows the difference spectra
collected during the titration of Bz-(1)6G-NH2 with [Cu(tpy)]2+
,
together with the titration curve (Figure 3B, Inset) that
conclusively demonstrates the formation of the complex (Bz-
[Cu(tpy)(1)]6-G-NH2)(ClO4)12. The Cu complexes in (Bz-
[Cu(tpy)(1)]6-G-NH2) each have a +2 charge, so that the overall
charge of the metal-coordinated oligopeptide is +12, versus the
charge-neutral (Bz-[Cu(pda)(1)]6-G-NH2) species. Importantly,
stoichiometric complexation is observed despite the large
positive charge, indicating that electrostatic repulsion between
metal centers is unimportant for binding. The isolated solid
products of these multimetallic structures are finely divided
powders, and our attempts to grow crystallographic quality
crystals have been fruitless to date. However, we have used
molecular modeling to predict and understand these molecules,
discussed further below.
Metal Coordination by Bpy Oligmer. The bpy ligand in
Bz-(2)3G-NH2 is bidentate toward transition metals and therefore
provides a distinctly different binding motif than the (mono-
dentate) py oligopeptide. Because of the presence of tridentate
ligands coordinated to the tetracoordinate metal center, the
[Cu(pda)] and [Cu(tpy)]2+ complexes used in the titrations of
the py oligomer do not bind to the bidentate bpy ligands and
were therefore not used to study binding to the bpy tripeptide.
However, it is known that free Fe2+ ions form hexacoordinate
complexes with bpy (i.e., [Fe(bpy)3]2+), whereas Cu2+ ions form
the tetracoordinate complex [Cu(bpy)2]2+. In the absence of
other ligands, metal ions added to solutions containing Bz-(2)3G-
NH2 should coordinate to multiple bpy ligands to form inter-
and/or intrachain linkages. Metal complexation could also induce
the formation of undesired coordination polymers, which we
prevent by use of solutions with low oligomer and metal
concentration.16 Spectrophotometric titrations of dilute solutions
of the Bz-(2)3G-NH2 oligomer were performed with Fe2+
perchlorate (Fe(ClO4)2) and Cu2+ acetate (i.e., Cu(OAc)2).
Figure 4A shows the change in visible absorbance during the
addition of Fe(ClO4)2 to a methanolic solution of Bz-(2)3G-
NH2; the strong peak at 540 nm is the well-known metal-to-
Figure 3. UV-visible absorption spectroscopy difference spectra for the
titration of Bz-(1)6G-NH2 with (A) [Cu(pda)] and (B) [Cu(tpy)]2+. The insets
contain plots of the peak absorbance as a function of relative molar ratio
of metal ion to oligomer.
coordination is typically accompanied by the appearance of a
peak in the UV-visible absorption spectrum, we used spectro-
photometric titrations to monitor the complexation of Cu2+ and
Fe2+ ions to the oligomers. We chose to strictly enforce one-
to-one complexation to the pyridine hexamers (i.e., one metal
to each pendant ligand) to confirm the availability of each of
the ligands for chelation. Therefore, tetracoordinate metal ions
bearing tridentate ligands (i.e., [Cu(pda)] and [Cu(tpy)]2+) were
used to preferentially bind to the pendant pyridine oligomers
to fill their coordination shell.8 Figure 3A shows the difference
spectra for the change in visible absorbance resulting from the
titration of a methanolic solution of Bz-(1)6G-NH2 oligomer with
[Cu(pda)(H2O)2].8 The peak at 660 nm, associated with a
[Cu(pda)(py)] absorption, increases during early additions of
oligomer and reaches a maximum value. A negative change in
absorbance at 840 nm is simultaneously observed and is
attributed to the consumption of free [Cu(pda)] in solution. The
titration curve for this experiment, plotted as the change in
absorbance as a function of added oligomer, is shown in the
Figure 3A inset. This curve indicates that the growth of the
(15) The slight rise in absorbance beyond the equivalence point is likely a result
of either background absorbance of free [Cu(pda)] in solution, whose broad
absorbance peak overlaps the [Cu(pda)(py)] transition, or the slow chelation
kinetics.
(16) No polymerization products have been observed under dilute solution
conditions; however these are expected to be large and insoluble, and the
reaction solutions are filtered prior to precipitation of the desired material.
9
9550 J. AM. CHEM. SOC. VOL. 127, NO. 26, 2005

Artificial Oligopeptide Scaffolds for Metal Binding
A R T I C L E S
Figure 4. UV-visible absorption spectroscopy difference spectra for the titration of Bz-(2)3G-NH2 with (A) Fe(ClO4)2 and (C) Cu(OAc)2. (B and D) Plots
of the peak absorbance as a function of relative molar ratio of metal ion to oligomer for the Fe2+ and Cu2+ ions, respectively.
ligand charge transfer (MLCT) transition for [Fe(bpy)3]2+ 17
.
The
would have a different stoichiometric equivalence point based
on the tetracoordinate nature of Cu2+. Visible absorption spectra
were again collected during the titration, Figure 4C. These
spectra contain a peak at 610 nm that grew and reached a
titration curve in Figure 4B shows that the intensity of the
MLCT peak increases upon addition of Fe2+ and levels off at
a stoichiometric equivalence point of 1 mol of Fe2+:1 mol of
Bz-(2)3G-NH2.
maximum value in the titration curve in Figure 4D at a Cu2+
/
oligomer equivalence point of ∼1.6. This molar ratio is
consistent with either a single bpy tripeptide binding 1.5 Cu2+
ions to form [Bz-[Cu(2)2]1.5G-NH2]2 (i.e., [Cu(2)2]1.5, calculated
mass 613.0 amu) or two bpy tripeptides linked by three Cu2+
ions to make [Bz-[Cu(2)2]3G-NH2]2 (i.e., [Cu(2)3]2, calculated
mass 1225.9 amu). Elemental analysis of the titration product,
isolated in larger quantity (∼3 mg) than the Fe analogue (<1
mg), confirmed its purity but could not differentiate between
the two possible structures. However, analysis of the product
using high-resolution mass spectrometry revealed a z ) +2
molecular ion peak (based on the isotopic splitting) at 1227.7
m/z. This molecular ion peak is consistent with the calculated
molecular weight of the [Cu(2)3]2 duplex structure, and taken
together, these data strongly suggest the formation of an
oligopeptide duplex linked by three Cu2+ ions.
Two possible structures would be consistent with this
stoichiometric ratio: either one oligopeptide coordinated to a
single Fe2+ ion, [Bz-[Fe(2)3]G-NH2)2] (i.e., [Fe(2)3], calculated
mass 1164.4 amu), or two oligopeptides cross-linked by two
Fe2+ ions to form [Bz-[Fe(2)3]G-NH2)]2 (i.e., [Fe(2)3]2, calcu-
lated mass 2328.8 amu). High-resolution mass spectrometry was
used to identify which of these was the titration product, and a
molecular ion was observed at 583.7 m/z. Both the z ) +2 ion
of [Fe(2)3] and the z ) +4 ion of [Fe(2)3]2 are expected to have
peaks at this m/z ratio. However, analysis of the observed
isotopic splitting of the high-resolution molecular ion peak
evidences sharp peaks with 0.25 m/z separation, consistent with
an ion of z ) +4 charge. Therefore, the high-resolution mass
spectrometry data leads to the conclusion that the titration
product is an oligopeptide duplex linked by two bound Fe2+
ions.18
In both the Fe and Cu bpy oligopeptide duplex cases, the
isolated metal-containing molecules form finely divided pow-
ders, which likely results from the large number of possible
structural isomers. Figure 5 shows just two examples (for each
To unambiguously confirm that metal coordination can be
utilized to induce duplex formation, we also titrated the Bz-
(2)3G-NH2 oligomer with [Cu(OAc)2],19 which we reasoned
(17) Lever, A. B. P. Inorganic Electronic Spectroscopy; Elsevier Publishing
Co.: New York, 1968.
(18) Because of the very small quantities of this duplex that were produced,
elemental analysis of this titration product was not possible.
(19) Cu2+ ion complexation is kinetically slower than Fe2+ binding, requiring
longer equilibration times between iterative additions. The difference spectra
show free Cu acetate ions (λ > 650 nm) because of a lower equilibrium
binding constant.
9
J. AM. CHEM. SOC. VOL. 127, NO. 26, 2005 9551

A R T I C L E S
Gilmartin et al.
Figure 5. Schematic diagrams of bipyridine tripeptides linked by (A) three
tetracoordinate (e.g., Cu2+) ions and (B) two octahedral (e.g., Fe2+) ions.
Circles on chain termini indicate benzoyl or acetyl caps.
of the metals) of the possible isomers that result from antiparallel
vs parallel alignment of the oligopeptides, an asymmetry that
arises from the sequence of the chain and the different termini.
Many more structures are possible based on changes in the
metal-ligand connectivity and the asymmetry of the bipyridine,
which is attached only in the 4-position.20 Coupled together,
the structural isomers and the small quantity of prepared
metalated material have made crystallization of the metal-linked
duplexes impossible to date. Our continuing efforts are focused
on the preparation of large enough quantities of the oligopeptides
to enable the isolation of a single isomer.
Figure 6. Energy-minimized structures of (A) [Cu(pda)(1)]6 and (B)
+12
[Cu(tpy)(1)]6
calculated using MM+ (Hyperchem 6.0) from (left) side
view and (right) top. Cu-Cu distances are (A) d ) 4.4 Å and (B) average
∼20 Å. Hydrogen atoms and oligomer termini are omitted for clarity.
differences between the calculated structures in Figure 6A and
B based on the charge of the Cu complexes, where electrostatic
repulsions between the [Cu(tpy)(1)]2+ complexes prevents π
stacking of the aromatic ligands. The structures in Figure 6
predict that, in the solid form, the degree of electronic interaction
Molecular Modeling. To better understand the molecular
structures of the metal-coordinated pyridine hexapeptides and
the metal-linked bpy tripeptide duplexes, we have turned to
computational molecular modeling. In the cases of the former
oligopeptides, because the metal complexes decorate a single
peptide chain, we expected a large amount of disorder in the
calculated structures. Figure 6A contains the energy minimized
structure of [Cu(pda)(1)]6, shown without the terminal benzoyl
and glycine caps for clarity, from the side and top. In agreement
with the crystal structures that we have obtained for the small
molecule analogue [Cu(pda)(py)], the bound Cu complexes each
have a square planar geometry. The most striking feature of
this calculated structure is the twist of the oligopeptide backbone
around the column of Cu metal complexes, most likely a result
of π-π interactions of the pda ligands, with a helical pitch of
16-17 monomers. The oligopeptide strand constrains the Cu
centers to a columnar arrangements with a regular spacing of
∼4 Å.
between Cu centers is much larger in the charge-neutral
12+
[Cu(pda)(1)]6 structure than in the [Cu(tpy)(1)]6
complex.
These differences are not expected to exist in solutions of these
multimetallic molecules.
We also performed calculations on the metal-linked bpy
tripeptide duplexes to visualize the most stable possible isomers.
Figure 7 contains the energy minimized, calculated structures
(shown from the side and top) for the [Cu(2)2]3 and [Fe(2)3]2
duplexes. In each of these calculated structures, the oligopeptide
acts as a scaffold to hold the metal complexes in close
proximity: for the Cu2+ and Fe2+ duplexes, the metal centers
are separated by 4.8 and 11 Å, respectively. In comparison with
the molecular structures in Figure 6, the spacing between the
Cu atoms in the duplex shown in Figure 7 is similar to that
predicted for [Cu(pda)(1)]6. It is evident from the models that
differing metal coordination geometries dramatically affect the
orientation of the peptide chains: whereas the octahedral Fe2+
complexes twist the peptide into a knotlike structure, the
(distorted) square planar geometry of the Cu2+ complexes allows
+12
In contrast, molecular modeling of the [Cu(tpy)(1)]6
molecule is shown in Figure 6B. The metal complexes are
splayed apart with an average distance of 20 Å. We justify the
(20) Ignoring possible rotamers of the bpy ligand around the ligand-peptide
bond, which gives hundreds of geometric isomers, there are 12 possible
tri-copper and 14 possible di-iron duplexes.
9
9552 J. AM. CHEM. SOC. VOL. 127, NO. 26, 2005

Artificial Oligopeptide Scaffolds for Metal Binding
A R T I C L E S
Figure 7. Energy-minimized structures of (A) [Cu(2)2]3 and (B) [Fe(2)3]2 duplexes, calculated using MM+ (Hyperchem 6.0), from the side and top. Calculated
Cu-Cu distances are d ) 4.8 Å; Fe-Fe distance is d ) 11 Å. Hydrogen atoms and oligomer termini are omitted for clarity.
the peptide strands to adopt a helical confirmation with a pitch
of ∼18 monomers.
The spectrum of the Cu-linked bpy trimer duplex (Figure 8C)
is similar to that of the [Cu(pda)(1)]6 hexapeptide, except that
it is even further broadened and lacks hyperfine coupling. The
Electron Paramagnetic Resonance Spectroscopy. Since
electron paramagnetic resonance (EPR) spectroscopy is sensitive
to the local environment and coupling between paramagnetic
species, we next turned to analysis of the metalated oligopeptides
to provide additional insight into their structures. Frozen EPR
spectra were obtained in methanolic solutions at 16 K for each
of the oligopeptides with bound Cu (ions or complexes) and
are shown in Figure 8. Values of the EPR spectroscopy
parameters and coupling constants are given in Table 1. The
spectrum for the pyridine hexapeptide with six-coordinated
[Cu(tpy)]2+ complexes (i.e., [Cu(tpy)(1)]612+) (Figure 8A) has
the classic line shape for a Cu complex with a hyperfine
coupling of A| )167 × 10-4 cm-1 21
. While this spectrum would
be expected for a single Cu complex, this observation for the
hexametallic peptide implies that each of the [Cu(tpy)(1)]2+
subunits is electronically identical and they must be separated
by a distance of at least 6 Å.22 This spectrum is contrasted by
that obtained for the charge neutral [Cu(pda)(1)]6 oligopeptide,
Figure 8B, in which the lines are broadened and the hyperfine
coupling is reduced to ∼130 × 10-4 cm-1. These are consistent
with weak interactions between Cu centers that give rise to some
delocalization of the unpaired electrons. The spectra of the
[Cu(pda)(1)]6 is further complicated by the presence of (at least)
a second set of peaks with lower intensity and similar A|, which
implies that several types of Cu complexes exist in this structure
with slightly different electronic environments. For both the
[Cu(tpy)(1)]612+ and [Cu(pda)(1)]6 hexapeptides, the values of
g| > 2.1 > g > 2.0 are typical for Cu complexes with a square
planar geometry.23
(21) Lippard, S. J.; Berg, J. M. Principles of Bioinorganic Chemistry; University
Science Books: California, 1994.
(22) Solomon, E. I.; Sundaram, U. M.; Machonkin, T. E. Chem. ReV. 1996, 96,
2563-2605.
(23) (a) Lucchesse, B.; Humphreys, K. J.; Lee, D.-H.; Incarvito, C. D.; Sommer,
R. D.; Rheingold, A. L.; Karlin, K. D. Inorg. Chem. 2004, 43, 5987-
5998. (b) Wei, N.; Murthy, N. N.; Karlin, K. D. Inorg. Chem. 1994, 33,
6093-6100.
Figure 8. Frozen EPR spectra of 0.5 mM solutions of [Cu(pda)(1)]6 (top),
[Cu(tpy)(1)]6(ClO4)12 (middle), and [Cu(2)2]3(PF6)6 (bottom).
9
J. AM. CHEM. SOC. VOL. 127, NO. 26, 2005 9553

A R T I C L E S
Gilmartin et al.
6+
Table 1. EPR Spectroscopy Parameters and Hyperfine Coupling
[Cu(2)2]3 bpy oligopeptide duplexes. Supporting electrolyte
was not added to these solutions to avoid contamination of the
very small quantities of available material with salt; therefore
slow potential scan rates (5 mV/s) and compensation for solution
resistance24 were used to produce the voltammograms in Figure
9. In the case of the trimetallic Cu duplex, a sigmoidal wave
attributed to the Cu(II/I) reduction was observed with a half
wave potential of 0.22 V vs Ag QRE. Using the microelectrode
limiting current at large overpotentials, and assuming that each
of the three Cu atoms are reduced during the reaction, the
diffusion coefficient of the Cu duplex is calculated to be 5 ×
10-7 cm2/s.24
Constants
4
A|
×
10-
1
complex
g|
g
(cm-
)
12+
[Cu(tpy)(1)]6
[Cu(pda)(1)]6
[Cu(2)2]3
2.28
2.35
2.06
2.09
2.09
167
130
6+
The dimetallic Fe duplex exhibited a voltammetric response
at more positive potentials, where the [Fe(bpy)3]3+/2+ oxidation
is expected to appear. However in contrast to the Cu duplex,
the Fe duplex voltammetry possesses an oxidative signature that
appears to have two sequential reactions: the first of these is
smaller and occurs at a half wave potential of 0.98 V and the
second larger wave at 1.2 V vs Ag QRE. Background cyclic
voltammograms of a nonmetalated peptide oligomer do not
possess any redox activity within this potential region. The
appearance of two waves is unusual and could arise from strong
electronic interactions between the two metal centers or from
the existence of two structural isomers of the [Fe(2)3]24+ duplex.
The former of these has been ruled out because similar coupling
in the Cu duplex (which is expected to be far more ordered)
voltammetry is not observed and our molecular modeling
predicts a substantial 1.1 nm separation between the Fe centers.
Additionally, it would be expected that strong electronic
coupling would result in two waves of equal magnitude in
current: the wave observed at lower potentials is about half
the size of the larger wave. We therefore believe that the two
peaks in the voltammetry are a result of two isomeric forms of
the Fe duplex that contain Fe complexes with different degrees
of conformational stabilization, which is the subject of ongoing
studies with larger quantities of duplex material.
Figure 9. Cyclic voltammograms of [Fe(2)3]2(ClO4)4 and [Cu(2)2]3(ClO4)6,
obtained using a 12.5 micron radius Pt working electrode at a scan rate of
5 mV/s, with backgrounds corrected for uncompensated resistance. Solutions
contained 1 mM peptide duplex in acetonitrile.
overall shape is again consistent with a structure containing Cu
atoms with several electronic environments. The lack of
hyperfine peaks may be due to very small coupling con-
stants but is most likely due to multiple Cu environments both
within one molecule and between the isomeric forms that are
present.
Taken together, the EPR spectra are consistent with the
structures from molecular modeling in Figures 6 and 7. Based
6+
on these, [Cu(pda)(1)]6 and [Cu(2)2]3 would be expected to
rotate polarized light because of their helical structures.
However, our measurements of the molecules’ circular dichro-
ism spectra (see Supporting Information) did not contain
evidence of optical activity for any of these multimetallic
complexes. We rationalize this result based on the expected
differences in the solution phase and frozen molecular structures
of the single-strand peptide and note that the observation for
the Cu-linked tripeptide duplex is again consistent with the
presence of several geometric isomers.
Using the currents obtained for the Fe duplex, the diffusion
coefficients for the waves at 0.98 and 1.2 V are determined24
to be 5 × 10-7 and 1 × 10-6 cm2/s, respectively. These values
compare well with that calculated for the Cu duplex. In both
cases, the solution phase diffusion coefficients are consistent
with mass transport rates that are expected for structures of this
size and charge.25 The voltammetry therefore demonstrates that
these relatively simple, homometallic oligopeptide duplexes have
distinguishable electrochemical signatures.
Electrochemistry. The oligopeptide complexes and duplexes
possess structural components (metal atoms and ligands), with
unique oxidation and reduction properties. Each of the above
oligopeptide sequences is expected to have a distinctive volta-
mmetric signature based on the type and number of metal
complex(es) linked to the peptide backbone. The voltammo-
grams (not shown) of the py hexapeptides containing [Cu(pda)]
and [Cu(tpy)]2+ complexes are distinctive only in the large
amount of adsorption to the electrode surface and extremely
slow heterogeneous kinetics for the Cu(II/I) reduction (for both
the tpy and pda complexes), which completely obfuscate further
quantitative analysis.
Conclusions
We have synthesized two new artificial oligopeptides with
pendant nitrogen-containing ligands and used these to prepare
peptidic scaffolds that bind transition metal ions in a stoichio-
metric manner. In the case of the bpy oligomers, the metal ions
cross-link the strands to self-assemble structures into double-
stranded oligopeptide duplexes with sequence-dependent elec-
(24) Bard, A. J.; Faulkner, L. R. Electrochemical Methods, Fundamentals and
Applications, 2nd ed.; John Wiley & Sons: New York, 2001.
(25) Compared to the diffusion coefficient of ferrocene (radius ≈ 2 Å) in
acetonitrile (∼10-5 cm2/s) and using the voltammetrically determined
diffusion coefficient (D), the Stokes-Einstein relationship (D ) kBT/6πηrH)
would predict a hydrodynamic radius (which would include stabilizing
counterions and solvent molecules) for the metal-linked tripeptide duplexes
of ∼4 nm.
Voltammetry of the Cu- and Fe-linked bpy oligopeptide
duplexes is however quantitatively informative; Figure 9
contains voltammograms acquired with a microdisk electrode
4+
in acetonitrile solutions containing either the [Fe(2)3]2 or
9
9554 J. AM. CHEM. SOC. VOL. 127, NO. 26, 2005

Artificial Oligopeptide Scaffolds for Metal Binding
A R T I C L E S
trochemical and spectroscopic properties. Our continuing efforts
seek to obtain crystallographic confirmation of both the Fe2+
and Cu2+ duplex structures, to expand the library of artificial
peptides, and to examine detailed optical and electronic proper-
ties of these supramolecular structures.
spectroscopy. This work was generously supported by a David
and Lucile Packard Foundation Fellowship for Science and
Engineering.
Supporting Information Available: High-resolution mass
spectrometry data, circular dichroism and two-dimensional NMR
spectra, with analysis. This material is available free of charge
via the Internet at http://pubs.acs.org.
Acknowledgment. This work is dedicated to the memory of
our friend and colleague R. Koerner. We gratefully acknowledge
the Penn State Department of Chemistry NMR facility and
especially Dr. Chris Falzone for assistance with 2-D NMR
JA0510818
9
J. AM. CHEM. SOC. VOL. 127, NO. 26, 2005 9555