Design and characterization of a synthetic electron-transfer protein

Article abstract of DOI:10.1021/ja0006954

A 30-residue polypeptide [H21(30-mer)] with the sequence Ac-K(IEALEGK)₂(IEALEHK)-(IEALEGK)G-NH₂ was synthesized. The circular dichroism (CD) spectrum of the peptide shows minima at 208 and 222 nm and θ₂₂₂/θ₂₀₈ = 1.06, which indicates the formation of a self-assembled coiled-coil when dissolved in aqueous solution. The concentration dependence of the CD data can be fit to an expression that describes a two-state monomer - dimer equilibrium for the apopeptide(K(d) = 1.5 ± 0.4 μM and θ(max) = -23 800 ± 130 deg cm² dmol^-1), showing that it has a maximum helicity of 69%. A [MTSL-C21(30-mer)] dimer was also prepared in which MTSL is the thiol-specific nitroxide spin label 1-oxyl-2,2,5,5-tetramethyl-Δ³-pyrroline-3-methyl-methanethiosulfonate attached to C21 of the 30-mer. Fourier deconvolution analysis of the dipolar line broadening of the electron paramagnetic resonance (EPR) spectrum yields a measure of the interchain Cα - Cα distance of 13.5 ± 0.9 A at position 21 of the coiled-coil, which is nearly identical to those distances observed for the isostructural family of bZip proteins. Two metallohomodimers, [Ru(trpy)(bpy)-H21(30-mer)]₂ and [Ru(NH₃)₅-H21(30-mer)]₂, in which the ruthenium complexes were coordinated with the H21 site of the 30-mer, were prepared. Sodium dodecyl sulfate - polyacrylamide gel electrophoresis (SDS - PAGE), chemical cross-linking studies, and analytical ultracentrifugation show that the peptides exist as a dimeric coiled-coil with a molecular weight of ~ 7.5 kDa. The electron transfer (ET) heterodimer, [Ru(trpy)(bpy)-H21 (30-mer)]/ [Ru(NH₃)₅-H21(30-mer)], was prepared, and molecular modeling shows that the two metal complexes are separated by a metal-to-metal distance of ~ 24 A across the noncovalent peptide interface. Pulse radiolysis was used to measure an ET rate constant of k(et) = 380 ± 80 s^-1 for the intracomplex electron transfer (ΔG°= -1.11 eV) from the Ru(II)(NH₃)₅-H21 donor to the Ru(III)(trpy)(bpy)-H21 acceptor. The value for k(et) falls within the range reported for modified proteins over comparable distances and supersedes the one reported in an earlier communication.

Full text of DOI:10.1021/ja0006954

J. Am. Chem. Soc. 2000, 122, 7999-8006
7999
Design and Characterization of A Synthetic Electron-Transfer Protein
Anna Y. Kornilova,‡ James F. Wishart,† Wenzhong Xiao,¶ Robin C. Lasey,‡
Anna Fedorova,‡ Yeon-Kyun Shin,¶ and Michael Y. Ogawa*,‡
Department of Chemistry and Center for Photochemical Sciences, Bowling Green State UniVersity,
Bowling Green, Ohio 43403, Chemistry Department, BrookhaVen National Laboratory, Upton, New York
11973, and Department of Chemistry and DiVision of Structural Biology, Lawrence Berkeley Laboratory,
UniVersity of California, Berkeley, California 94720
ReceiVed February 18, 2000
Abstract: A 30-residue polypeptide [H21(30-mer)] with the sequence Ac-K(IEALEGK)2(IEALEHK)-
(IEALEGK)G-NH2 was synthesized. The circular dichroism (CD) spectrum of the peptide shows minima at
208 and 222 nm and θ222/θ208 ) 1.06, which indicates the formation of a self-assembled coiled-coil when
dissolved in aqueous solution. The concentration dependence of the CD data can be fit to an expression that
describes a two-state monomer-dimer equilibrium for the apopeptide (Kd ) 1.5 ( 0.4 µM and θmax ) -23 800
( 130 deg cm2 dmol-1), showing that it has a maximum helicity of 69%. A [MTSL-C21(30-mer)] dimer was
also prepared in which MTSL is the thiol-specific nitroxide spin label 1-oxyl-2,2,5,5-tetramethyl-∆3-pyrroline-
3-methyl-methanethiosulfonate attached to C21 of the 30-mer. Fourier deconvolution analysis of the dipolar
line broadening of the electron paramagnetic resonance (EPR) spectrum yields a measure of the interchain
CRsCR distance of 13.5 ( 0.9 Å at position 21 of the coiled-coil, which is nearly identical to those distances
observed for the isostructural family of bZip proteins. Two metallohomodimers, [Ru(trpy)(bpy)-H21(30-mer)]2
and [Ru(NH3)5-H21(30-mer)]2, in which the ruthenium complexes were coordinated with the H21 site of the
30-mer, were prepared. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), chemical
cross-linking studies, and analytical ultracentrifugation show that the peptides exist as a dimeric coiled-coil
with a molecular weight of ∼7.5 kDa. The electron transfer (ET) heterodimer, [Ru(trpy)(bpy)-H21(30-mer)]/
[Ru(NH3)5-H21(30-mer)], was prepared, and molecular modeling shows that the two metal complexes are
separated by a metal-to-metal distance of ∼24 Å across the noncovalent peptide interface. Pulse radiolysis
was used to measure an ET rate constant of ket ) 380 ( 80 s-1 for the intracomplex electron transfer (∆G°
) -1.11 eV) from the RuII(NH3)5-H21 donor to the RuIII(trpy)(bpy)-H21 acceptor. The value for ket falls
within the range reported for modified proteins over comparable distances and supersedes the one reported in
an earlier communication.
Introduction
within appropriately derivatized R-helices,11-13 proline helices,14-17
artificial â-strands,18 â-sheets,19,20 and multihelical bundles.21-24
However, the design of these first-generation peptide-based ET
(6) (a) Wishart, J. F.; Sun, J.; Cho, M.; Su, C.; Isied, S. S. J. Phys. Chem.
B 1997, 101, 687. (b) Sun, J.; Wishart, J. F.; Gardineer, M. B.; Cho, M. P.;
Isied, S. S. Inorg. Chem. 1995, 34, 3301. (c) Isied, S. S.; Kuehn, C.;
Worosila, G. J. Am. Chem. Soc. 1984, 106, 1722.
Protein-based electron-transfer (ET) reactions have been the
focus of considerable interest because of their important role
in biological energy conversion.1-4 A central goal of this
research is to identify the role, if any, of the intervening protein
matrix in propagating long-range donor/acceptor interactions.
To this end, much work has been done to investigate the ET
properties of photosynthetic reaction centers,3b,5 surface-deriva-
tized proteins,4,6 protein-protein complexes,7-10 and peptide-
based model compounds.1 The last approach to this problem
includes the study of intramolecular ET reactions that occur
(7) Qiao, T. C.; Witkowski, R.; Henderson, R. McLendon G. JBIC 1996,
1, 432.
(8) (a) Zhou, J. S.; Hoffman, B. M. Science 1994, 265, 1693. (b) Stemp,
E. D. A.; Hoffman, B. M. Biochemistry 1993, 32, 10848.
(9) Mei, H.; Wang, K.; Peffer, N.; Weatherly, G.; Cohen, D. S.; Miller,
M.; Pielak, G.; Durham, B.; Millett, F. Biochemistry 1999, 38, 6846.
(10) Olesen, K.; Ejdeback, M.; Crnogorac, M. M.; Kostic, N. M.;
Hansson, O. Biochemistry 1999, 38, 16 695.
(11) (a) For a review, see Sisido, M. AdV. Photochem. 1997, 22, 197.
(b) Kuragaki, M.; Sisido, M. J. Phys. Chem. 1996, 100, 16 019.
(12) (a) Fox, M. A.; Galoppini, E. J. Am. Chem. Soc. 1997, 119, 5277.
(b) Galoppini, E.; Fox, M. A. J. Am. Chem. Soc. 1996, 118, 2299.
(13) (a) Basu, G.; Anglos, D.; Kuki, A. Biochemistry 1993, 32, 3067.
(b) Basu, G.; Kubasik, M.; Anglos, D.; Kuki, A. J. Phys. Chem. 1993, 97,
3956.
(14) (a) Isied, S. S.; Ogawa, M. Y.; Wishart, J. F. Chem. ReV. 1993, 92,
381. (b) Ogawa, M. Y.; Wishart, J. F.; Young, Z.; Miller, J. R.; Isied, S. S.
J. Phys. Chem. 1993, 97, 11 456. (c) Ogawa, M. Y.; Moreira, I.; Wishart,
J. F.; Isied, S. S. Chem. Phys. 1993, 176, 589.
(15) (a) Schanze, K. S.; Cabana, L. A. J. Phys. Chem. 1990, 94, 2740.
(b) Cabana, L.; Schanze, K. AdV. Chem. Ser. 1990, 226, 101.
(16) (a) DeFilippis, M. R.; Faraggi, M.; Klapper, M. H. J. Am. Chem.
Soc. 1990, 112, 5640. (b) Faraggi, M.; DeFilippis, M. R.; Klapper, M. H.
J. Am. Chem. Soc. 1989, 111, 5141.
‡ Bowling Green State University.
† Brookhaven National Laboratory.
¶ University of California.
(1) For a review, see Ogawa, M. Y. in Multimetallic and Macromolecular
Photochemistry; Ramamurthy, V., Schanze, K. S., Eds.; Marcel Dekker:
New York, 1999; pp 113-150.
(2) Photochemistry and Radiation Chemistry, Complementary Methods
for the Study of Electron-Transfer; Wishart, J. F.; Nocera, D. G., Eds.; ACS
Advances in Chemistry Series 254; American Chemical Society: Wash-
ington, DC, 1998.
(3) a) Moser, C. C.; Page, C. C.; Farid, R.; Dutton, P. L. J. Bioenerg.
Biomembr. 1995, 27, 263. (b) Moser, C. C.; Keske, J. M.; Warncke, K.;
Farid, R. S.; Dutton, P. L. Nature 1992, 355, 796.
(4) (a) Winkler, J. R.; Gray, H. B. J. Biol. Inorg. Chem. 1997, 2, 399.
(b) Gray, H. B.; Winkler, J. R. Annu. ReV. Biochem. 1996, 65, 537.
(5) The Photosynthetic Reaction Center, Deisenhofer, J.; Norris, J. R.,
Eds.; Academic: New York, 1993.
10.1021/ja0006954 CCC: $19.00 © 2000 American Chemical Society
Published on Web 08/08/2000

8000 J. Am. Chem. Soc., Vol. 122, No. 33, 2000
KorniloVa et al.
models suffers from the following limitation: whereas many
in vivo ET reactions occur across the noncovalent interface of
a protein-protein complex, the early model compounds all used
peptide bridges to form a covalent link between their donor and
acceptor sites. The structural features of these model systems
are thus inherently different from those often found in the native
biological environment. We describe here a new approach
toward the design of more nativelike ET models in which self-
organized peptide assemblies are used to separate metal-based
redox sites across a well-defined, yet noncovalent, peptide-
peptide interface. Future study of these synthetic redox proteins
should produce valuable insight into the mechanisms of biologi-
cal ETs.
The synthesis of the self-assembled peptide-peptide ET
complex is derived from recent advances in the rational design
of artificial peptide assemblies that retain nativelike structures.
A notable example of de novo designed proteins can be seen
within the family of synthetic, two-stranded R-helical coiled-
coils.25,26 This ubiquitous protein structure results from the
noncovalent self-assembly of two R-helices to form a left-handed
supercoil. Synthetic coiled-coils can be constructed from
polypeptides whose sequence is based on a seven-residue repeat,
(abcdefg)n, in which positions a and d are occupied by
hydrophobic amino acids, e and g are filled by oppositely
charged residues, and b, c, and f denote hydrophilic residues
(Figure 1a). This design creates a hydrophobic face of an R-helix
to serve as a noncovalent dimerization site for the coiled-coil
(Figure 1b). The current study utilizes this motif to prepare a
unique, coiled-coil protein designated H21(30-mer) on which
metal-based redox sites are positioned to create a synthetic
metalloprotein that can exhibit long-range ET. The physico-
chemical characterization of this protein is presented, along with
the results of electron-pulse radiolysis experiments that dem-
Figure 1. (a) Wheel representation showing the sequence of the H21-
(30-mer) protein. (b) Schematic representation of the third heptad repeat
of the ET metalloheterodimer.
obtained from Aldrich Chemical Company. The thiol-specific MTSL
spin-label (MTSL ) 1-oxyl-2,2,5,5-tetramethyl-∆3-pyrroline-3-methyl-
methanethiosulfonate) was purchased from Toronto Research Chemicals
Inc. All chemicals and solvents were used as received without further
purification.
General Methods. Reversed-phase high-performance liquid chro-
matography (HPLC) analyses were performed on a Whatman Magnum
9 (500 × 9.4 mm) semipreparative Partisil ODS-3 C18 column. A single-
pump system was used (Waters model 510) equipped with a binary
gradient controller (Autochrom. Inc.) and a Waters model 994 diode
array detector/spectrophotometer with a 1-cm cell path length. For
preparative separations, the monitoring wavelengths were set to the
tails of the absorption bands of the desired compounds. The purity of
the resulting samples was subsequently verified by analytical HPLC
runs from which the full ultraviolet-visibile (UV-vis) spectra could
be obtained. Gel filtration was carried out on columns packed with
Bio-Gel P-2 resin (Bio-Rad). The UV-vis spectra of the purified
compounds were recorded on a Hewlett-Packard model 8452A diode
array spectrophotometer. The molar extinction coefficient, ꢀ(478 nm)
) 8800 M-1 cm-1, was used to determine the concentration of the
Ru(trpy)(bpy)-H21(30-mer) peptide.27 Mass spectral (MS) data were
obtained by the Core Protein Facilities of the University of Michigan
(Ann Arbor, MI) and the Mass Spectrometry Center of the University
of Maryland (College Park, MD). Cyclic voltammetry (CV) was
conducted on a BAS 100W Electrochemical Analyzer using a small-
volume (200 µL) sample cell equipped with a platinum working
electrode, a platinum wire auxiliary electrode, and a Ag/AgNO3
reference electrode. The samples were dissolved in 50 mM phosphate
buffer (pH 7.0).
Synthesis of the H21(30-mer) Peptide. The 30-residue polypeptide,
with the sequence Ac-K(IEALEGK)2-(IEALEHK)-(IEALEGK)G-NH2,
was synthesized on a Perseptive Biosystems model 9050 Plus Pep-
Synthesizer (Framingham, MA) using the fluorenylmethoxycarbonyl
N-terminal protection strategy and the manufacturer’s PAL PEG-PS
resin. Activation was achieved using diisopropylcarbodiimide (DIPCDI)
and 1-hydroxybenzotriazole (HOBT) with a 60-min cycle time. This
polypeptide is designated H21(30-mer).
Prior to cleavage from the resin, the peptide was N-acetylated with
a capping solution containing 0.5 M acetic anhydride, 0.125 M N,N-
diisopropylethylamine (DIEA), and 0.015 M HOBT in dimethylfor-
mamide (DMF). Cleavage of the H21(30-mer) from the resin and
onstrate the occurrence of a unimolecular ET reaction (ket
)
380 ( 80 s-1) occurring over a metal-to-metal distance of ca.
24 Å across the noncovalent interface.
Experimental
Materials. The Fmoc-protected L-amino acids, diisopropylcarbodi-
imide, and 1-hydroxybenzotriazole were purchased from Peptides
International. N-N′-Dimethylformamide, dichloromethane, ethylene
glycol, and trifluoracetic acid were obtained from Fisher Scientific.
Piperidine, anisole, and [Ru(NH3)5Cl]Cl2 were purchased from Lancaster
Chemicals, Baker Chemicals, and Strem International, respectively. The
•
2,2′:6′,2′′-terpyridine (trpy), 2,2′-bipyridine (bpy), and RuCl3 H2O were
(17) (a) McCafferty, D. G.; Friesen, D. A.; Danielson, E.; Wall, C. G.;
Saderholm, M. J.; Erickson, B. W.; Meyer. T. J. Proc. Natl. Acad. Sci.,
U.S.A. 1996, 93, 8200. (b) Mecklenburg, S. L.; McCafferty, D. G.;
Schoonover, J. R.; Peek, B. M.; Erickson, B. W.; Meyer, T. J. Inorg. Chem.
1994, 33, 2974.
(18) Fernando, S. R. L.; Kozlov, G. V.; Ogawa, M. Y. Inorg. Chem.
1998, 37, 1900.
(19) Gretchikhine, A. B.; Ogawa, M. Y. J. Am. Chem. Soc. 1996, 118,
1543.
(20) Williamson, D. A.; Bowler, B. E. J. Am. Chem. Soc. 1998, 120,
10902.
(21) Kozlov, G. V.; Ogawa, M. Y. J. Am. Chem. Soc. 1997, 119, 8377.
(22) (a) Mutz, M. W.; Case, M. A.; Wishart, J. F.; Ghadiri, M. R.;
McLendon, G. L. J. Am. Chem. Soc. 1999, 121, 858. (b) Zhou, J; Case, M.
A.; Wishart, J. F.; McLendon G. L. J. Phys. Chem. B 1998, 102, 9975. (c)
Mutz, M. W.; McLendon, G. L.; Wishart, J. F.; Gaillard, E. R.; Corin, A.
F. Proc. Natl. Acad. Sci., U.S.A. 1996, 93, 9521.
(23) Rau, H. K.; DeJonge, N.; Haehnel, W. Proc. Natl. Acad. Sci., U.S.A.
1998, 95, 11526.
(24) Jones, G., II; Vullev, V.; Braswell, E. H.; Zhu, D. J. Am. Chem.
Soc. 2000, 122, 388.
(25) (a) For a review, see Hodges, R. S. Biochem. Cell Biol. 1996, 74,
133. (b) Zhou, N. E.; Zhu, B. Y.; Kay, C. M.; Hodges, R. S. Biopolymers
1992, 32, 419.
(26) Graddis, T. J.; Myszka, D. G.; Chaiken, I. M. Biochemistry 1993,
32, 12664.
(27) (a) Di Bilio, A. J.; Dennison, C.; Gray, H. B.; Ramirez, B. E.; Sykes,
A. G.; Winkler, J. R. J. Am. Chem. Soc. 1998, 120, 7551. (b) Di Bilio, A.
J.; Hill, M. G.; Bonander, N.; Karlsson, B. G.; Villahermosa, R. M.;
Malmstrom, B. G.; Winkler, J. R.; Gray, H. B. J. Am. Chem. Soc. 1997,
119, 9921.

Synthetic Electron-Transfer Protein Characterization
J. Am. Chem. Soc., Vol. 122, No. 33, 2000 8001
•
deprotection of the amino acid side chains were achieved by gently
stirring the resin in a mixture of 88% (v/v) trifluoroacetic acid, 5%
(v/v) phenol, 2% (v/v), triisopropylsilane and 5% (v/v) water for 2 h.
The crude peptide was then precipitated in cold diethyl ether, collected
by vacuum filtration, and dried under vacuum. Final purification was
achieved by semipreparative reversed-phase C18 HPLC with a linear
AB gradient of 1% (v/v) A/min, at a flow rate of 2 mL/min, in which
solvent A was 0.1% (v/v) trifluoroacetic acid in acetonitrile and solvent
B was 0.1% (v/v) trifluoroacetic acid in water. The collected peptide
was then lyophilized and analyzed by MALDI-TOF MS (m/z obsd:
3287.8, calcd for [M + H+]: 3287.8).
Synthesis of the Spin-Labeled Peptide. The 30-residue polypeptide,
C21(30-mer), with the sequence Ac-K(IEALEGK)2-(IEALECK)-(IEA-
LEGK)G-NH2, was synthesized and purified as just described. A 10-
mg (3-µmol) sample of the purified peptide was then dissolved in 2
mL of 50 mM phosphate buffer (pH 7.0) to which a 5-fold molar excess
of the thiol-specific MTSL spin-label was added (MTSL ) 1-oxyl-
2,2,5,5-tetramethyl-∆3-pyrroline-3-methyl-methanethiosulfonate). The
reaction mixture was stirred at room temperature for 3 h after which it
was purified by semipreparative reversed-phase C18 HPLC as already
described. The collected peptide was then lyophilized and analyzed by
MS (m/z obsd: 3398.9, calcd for [M + 4H+]: 3399.1).
Synthesis of Ru(trpy)(bpy)-H21(30-mer). To a 10-mg (3-µmol)
sample of H21(30-mer) dissolved in 2 mL of argon-purged 50 mM
phosphate buffer (pH 7.0) was added a concentrated (8-10 mM)
solution of [Ru(trpy)(bpy)Cl]+ 28 dissolved in either water or phosphate
buffer. The solution was then adjusted to pH 7.7-7.8 and stirred under
reduced light for 3 days at 30 °C. The progress of the reaction was
monitored by HPLC, which showed the growth of an absorption band
centered at 478 nm for the HPLC peak corresponding to the H21(30-
mer). After the reaction was complete, the solution was concentrated
to a small volume by rotary evaporation and passed through a size-
exclusion column equilibrated with water. The metallopeptide solution
was then concentrated to near dryness and lyophilized. HPLC analysis
showed the existence of a small amount (∼5%) of free metal complex
in the presence of the desired metallopeptide.
Synthesis of Ru(NH3)5-H21(30-mer). A 15-mg (34-µmol) sample
of [Ru(NH3)5Cl](TFA)2 was dissolved in 0.5 mL of water, and the
resulting solution was purged with argon for 20 min to remove oxygen.
A piece of zinc amalgam was then placed into the solution and the
solution was bubbled with argon for another 20 min. This solution was
then cannulated into a separate flask containing 10 mg (3 µmol) of
H21(30-mer) dissolved in 3 mL of argon-saturated 50 mM phosphate
buffer. The reaction mixture was stirred under an argon atmosphere
overnight, after which it was filtered and injected directly onto the
semipreparative HPLC column.
(G ) 6.13, ꢀ ) 7950 at 472 nm). The yield of N3 radicals (G value)
•
was assumed to be 6.0 per 100 eV absorbed. The concentration of N3
radicals in each shot was 5-15% of the ruthenated peptide concentra-
tion.
Solutions for pulse radiolysis were prepared by reducing a stock
solution of equal mole quantities of the [RuII(trpy)(bpy)-H21(30-mer)]
and [RuIII(NH3)5-H21(30-mer)] peptides in 50 mM phosphate buffer
(pH 7.0) over zinc amalgam under an argon atmosphere. Measured
quantities of the stock solution were then transferred via Hamilton
Gastight syringes and diluted to the desired peptide concentration by
addition to a N2O-saturated, 1 mM NaN3, 50 mM pH 7.0 phosphate
buffer solution in the radiolysis cell. In most cases, the peptide samples
were prepared with 1-2 equivalents of an exogenous reductant, Euaq
2+
,
per equivalent of [Ru(NH3)5-H21(30-mer)]. This reductant was added
to keep the sample in its reduced state prior to the application of the
radiolytic dose.29 However, under such conditions, the europous ion
can act to intercept the radiolytically generated azide radical. This
interception would result in decreasing the absolute yield of the RuIII-
(trpy)(bpy)-peptide, as evidenced by the observation of a smaller
absorption bleach than would be seen in the absence of the reductant.
The Eu2+ species can also react with the radiolytically generated RuIII-
(trpy)(bpy)-peptide to produce an observed recovery of the absorption
bleach that is faster than the true intramolecular ET rate. Thus, the
kinetic experiments were performed in such a manner as to eliminate
the Eu2+ species in situ before the desired intramolecular ET rate was
measured. Successive radiolytic doses were therefore applied to a given
peptide sample to oxidatively titrate the Eu2+ species until (1) the initial
bleach at 490 nm had grown to the value initially observed in the
absence of europous ion, and (2) the observed recovery rate had
decreased to a constant value over several successive shots.
Circular Dichroism Measurements. Circular dichroism (CD)
spectra were obtained with an Aviv and Associates model 62DS CD
spectrometer (Lakewood, NJ) equipped with a thermoelectric temper-
ature controller. A rectangular 1-mm path length cell was used (sample
volume e 400 µL). The spectrometer was routinely calibrated with an
aqueous solution of (1S)-(+)-10-camphorsulfonic acid. Mean residue
molar ellipticities were calculated according to the following equation:
[θ] ) [θ]obs/(10• lcn)
(4)
where [θ]obs is the observed ellipticity measured in degrees, l is the
path length of the cell in centimeters, c is the molar peptide
concentration, and n is the number of amino acid residues in the peptide.
The spectra were obtained as an average of 3-5 scans using a
wavelength step of 1 nm. The thermal denaturation results were obtained
by measuring the ellipticity at 222 nm as a function of temperature
using a six-second thermal equilibration time between data points. The
series of solutions used in the concentration studies were prepared by
serial dilution of a concentrated stock solution prepared from dissolving
a known mass of the lyophilized peptide into a known volume
phosphate buffer (pH 7.0). The values reported for Kd and θmax are
taken from an average of two separate experiments.
Gel Electrophoresis and Chemical Cross-Linking Experiments.
Discontinuous polyacrylamide gel electrophoresis (PAGE) was per-
formed on samples of the closely related peptide, Ru(bpy)2Im-H21-
(31-mer),21 dissolved in HEPES buffer in the absence of sodium dodecyl
sulfate (pH 7.0) at either room temperature or after heating to 80 °C
for 15 min prior to loading. Experiments were performed in the presence
of SDS (0.1% v/v) in the running buffer at 200 V for 45 min. The
molecular mass standards for aprotinin (6 kDa), plastocyanin (10 kDa),
lysozyme (14 kDa), and cytochrome c (21 kDa) were run in parallel
for comparison.
Pulse Radiolysis Measurements. Pulse radiolysis transient absorp-
tion experiments were performed on the 2 MeV Van de Graaff
accelerator at Brookhaven National Laboratory with a PC-controlled,
CAMAC-based control and data acquisition system. The optical
detection system consisted of either a pulsed or continuous wave (CW)
xenon arc lamp, or a tungsten lamp, appropriate long-pass optical filters,
quartz radiolysis cells with various path lengths (2.0 and 6.1 cm), a
monochromator, and photomultiplier tube. Digitizer traces were fit to
single or double first-order kinetics by nonlinear least-squares methods.
•
Azide radical, N3 , was produced by the reaction of radiolytically
generated hydroxyl radical (OH•) with the azide ion present in the N2O-
saturated water:
H2O 2 MeV electrons8 e-aq, OH•, H•, H2, H2O2
eaq- + N2O + H2O f N2 + OH• + OH-
OH• + N3- f N3• + OH-
(1)
(2)
(3)
Nonspecific intersubunit cross-linking was performed on samples
of the metalloheterodimer using glutaraldehyde as the linker.30 Glut-
araldehyde reacts with the free amine groups at the N-terminal of the
peptide and at the lysine residues. Mixtures containing equal molar
amounts of [Ru(bpy)2Im-H21(31-mer)] and [Ru(NH3)5-H21(31-mer)]
were dissolved in 50 mM phosphate buffer, heated to 80 °C for 30
Radiolytic dose levels were calculated for each shot based on the current
delivered and a calibration factor obtained from the (SCN)2- dosimeter
(28) (a) Sullivan, B. P.; Calvert, J. M.; Meyer, T. J. Inorg. Chem. 1980,
19, 1404. (b) Ware, D. C.; Lay, P. A.; Taube, H. Inorg. Synth. 1986, 24,
299. (c) Johnson, E. C.; Sullivan, B. P.; Salmon, D. J.; Adeyemi, S. A.;
Meyer, T. J. Inorg. Chem. 1978, 17, 2211.
(29) Sun, J.; Su, C.; Wishart, J. F. Inorg. Chem. 1996, 35, 5893.
(30) Habeeb, A. F.; Hiramoto, R. Arch. Biochem. Biophys. 1968, 126,
16.

8002 J. Am. Chem. Soc., Vol. 122, No. 33, 2000
KorniloVa et al.
min, and then allowed to cool back to room temperature before the
addition of glutaraldehyde (final concentrations ranged from 2 to 100
mM). The samples were incubated at 37 °C for 20 min before being
quenched by adding glycine to a final concentration of 100 mM. These
samples were then run on an 18% SDS-PAGE gel. Two different
polypeptide standards were run in parallel (3.4, 5.5, 8.7, and g15.3
kDa; 3.0, 5.0, 9, and g14 kDa) for comparison.
Analytical Ultracentrifugation. Sedimentation equilibria were
determined for samples of the [Ru(bpy)2Im-H21(31-mer)]/[Ru(NH3)5-
H21(31-mer)] heterodimer21 using a Beckman Optima XL-A ultracen-
trifuge run at 50 000 rpm for 24 h at 25 °C. The sample was prepared
as already described with an initial concentration of ∼50 µM. The
partial specific volume was estimated to be 0.74 cm3 g-1 from peptide
composition,31 with a correction made for the ruthenium complex. The
solvent density was taken to be 1.002 g cm-3 32
from four different runs and averaged.
.
Data were obtained
Electron Paramagnetic Resonance (EPR) Spectroscopy and
Spectral Analysis. EPR spectra were collected with a Bruker ESP 300E
spectrometer (Bruker, Germany) equipped with a loop-gap resonator
(Medical Advances, Milwaukee, WI) and a low-noise microwave
amplifier (Miteq, Hauppauge, NY). Spin concentrations were deter-
mined by comparing doubly integrated room-temperature spectra with
that of a TEMPO standard solution. Low-temperature spectra were
collected at 125 K at 8-microwatt microwave power and a modulation
amplitude of 1 G, and spin concentration was ∼300 µM. The EPR
spectra were independent of concentration up to 2 mM.
Electron dipolar interactions between two interacting nitroxide spin
labels lead to overall EPR spectral broadening. Quantitative analysis
of room-temperature spectra is nontrivial because of the motional line
broadening in addition to the dipolar line broadening. Low-temperature
spectra of motionally frozen samples permits extraction of the dipolar
broadening function directly from Fourier deconvolution analysis of
the interacting double-label spectra with respect to noninteracting
monoradical spectra.33 The motionally frozen state is readily achieved
by freezing the EPR samples near the liquid nitrogen temperature. This
method provides an accurate estimation for interspin distances in the
range 8-25 Å, and its applications to peptides and proteins have been
described previously.34-36
Because these experiments were designed to detect the relatively
weak dipolar interactions between two spin labels at the f positions of
the peptide dimer, which are >20 Å apart, special efforts were made
to minimize the experimental error. The average of five independently
obtained low-temperature spectra was used in the final analysis. The
noninteracting monoradical reference spectra were taken both from the
sample prepared by mixing the spin-labeled peptide with the unlabeled
peptide at a 1:10 mole ratio and from a sample of the spin-labeled
peptide sample denatured in 6 M GuHCl.
Molecular Modeling. Computational results were obtained using
software programs from Biosym/MSI (San Diego, CA). Energy
minimization calculations were performed with the DiscoVer® program
using the ESFF force field and the conjugate gradient method, setting
a final convergence criterion of 0.001 kcal/(mol Å) rms force. Molecular
dynamics simulations were performed by the Nose temperature control
method. All structures were solvated with a water box prior to
performing the calculations. Graphical displays were obtained from the
Insight®II molecular modeling program.
Figure 2. The circular dichroism spectrum of a 76 µM solution of
H21(30-mer) peptde in 50 mM phosphate buffer (pH 7, 298 K). Inset:
The concentration dependence of θ222 for H21(30-mer). The solid line
shows the fit to a two-state monomer-dimer equilibrium model (eq 6)
as described in the text.
H21(30-mer). The sequence (Ac-K(IEALEGK)2(IEALEHK)-
(IEALEGK)G-NH2) was based on a seven-residue repeat
developed by Hodges and co-workers to prepare artificial coiled-
coils.25 In the sequence, positions a, d, e, and g are occupied
by isoleucine, leucine, glutamic acid, and lysine, respectively,
as these substitutions have been shown to encourage the
formation of two-stranded coiled-coils in solution. In particular,
it has been found that dimeric coiled-coils are favored by the
incorporation of â-branched amino acids into position a of the
heptad. An additional feature of the sequence is that it places a
single histidine residue at position 21 to provide a metal-binding
site on the most highly solvent-exposed position of the third
heptad repeat.
The two metallopeptides, [Ru(trpy)(bpy)-H21(30-mer)] and
[Ru(NH3)5-H21(30-mer)], were prepared by appropriate modi-
fication of methods previously described for the attachment of
ruthenium complexes to the surface of cytochrome c.6c,27 In each
case, purification was achieved by a combination of semi-
preparative reversed-phase HPLC and size-exclusion chroma-
tography. The UV absorption spectrum of the [Ru(trpy)(bpy)-
H21(30-mer)] peptide exhibits maxima at λ ) 206 (log ꢀ )
5.0), 274 (log ꢀ ) 4.5), 288 (log ꢀ ) 4.5), 314 (log ꢀ ) 4.50),
and 476 nm (log ꢀ ) 3.9), and is quite similar to the spectrum
previously reported for the related model compound, [Ru(trpy)-
(bpy)Im]2+ 27b
When dissolved in water, the metallopeptide has
.
a short-lived emission at 650 nm (τ ) 23 ns) that is nearly the
same as that reported for the related ruthenated plastocyanin.27
Differential pulse polarography and CV show a reversible redox
couple for the metallopeptide at E1/2 ) +1.17 V versus NHE
that is similar to the value reported for the model compound,
[Ru3+/2+(trpy)(bpy)Im].27,37 Characterization of the [Ru(NH3)5-
H21(30-mer)] peptide was accomplished by CV, which showed
nearly identical behavior to that of Ru(NH3)5-modified cyto-
chrome c.6c A reversible one-electron reduction at E1/2 ) +0.065
V versus NHE was observed.
Results
Synthesis and Characterization of Peptides. Solid-phase
methods were used to prepare a 30-residue polypeptide called
Circular Dichroism Spectroscopy. Figure 2 shows the
circular dichroism (CD) spectrum of a 76 µM solution of H21-
(30-mer) in 50 mM phosphate buffer (pH 7, 298 K). The
spectrum consists of a positive signal located below 200 nm,
and a pair of negative signals at 208 and 222 nm, which
demonstrates the existence of an R-helical structure. At the
concentrations used for the ET measurements described later,
(31) McMeekin, T.L.; Groves, M. L.; Hipp, N. J. J. Am. Chem. Soc.
1949, 71, 3298.
(32) Weast, R. C. Handbook of Chemistry and Physics; CRC Press: Boca
Raton, FL, 1985.
(33) Rabenstein, M. D.; Shin, Y.-K. Proc. Natl. Acad. Sci., U.S.A. 1995,
92, 8239.
(34) Hall, J. A.; Thorgeirsson, T. E.; Liu, J.; Shin, Y.-K.; Nikaido, H. J.
Biol. Chem. 1997, 272, 17 610.
(35) Thorgeirsson, T. E.; Xiao, W.; Brown, L. S.; Needleman, R.; Lanyi,
J. K.; Shin, Y.-K. J. Mol. Biol. 1997, 273, 951.
(36) Ottemann, K. M.; Thorgeirsson, T. E.; Kolodziej, A. F.; Shin Y.-
K.; Koshland, D. E. Biochemistry 1998, 37, 7062.
(37) A second reversible wave seen at 0.87 V versus NHE is assigned
to the presence of [Ru3+/2+(trpy)(bpy)(H2O)] which is observed as a minor
component (<5%) by HPLC in the peptide sample.

Synthetic Electron-Transfer Protein Characterization
J. Am. Chem. Soc., Vol. 122, No. 33, 2000 8003
an ellipticity ratio of θ222/θ208 ) 1.06 was observed, which
indicates the existence of an R-helical coiled-coil.25 In contrast,
single-stranded R-helices are expected to display values of θ222
/
θ208 ≈ 0.86, as theoretical calculations have shown that the CD
band located at 208 nm is polarized parallel to the helical axis
and its magnitude is therefore sensitive to whether the helical
structure is comprised of a single peptide chain or interacting
R-helices.38 It is interesting to note that when H21(30-mer) is
dissolved in methanol, θ222/θ208 ) 0.9, indicating that the coiled-
coil structure has unfolded into noninteracting helical monomers.
Thermal denaturation studies performed in aqueous solution (not
shown) indicate that the self-assembled peptide structure is very
stable, producing a melting curve in which Tm ) 65 °C.
The ellipticity value at 222 nm can be used to estimate the
helical content of the peptide by comparison to the maximum
molar ellipticity (X∞H) calculated for a 30-residue polypeptide:
X∞H ) (-37 400 deg cm2 dmol-1)(1 - k/n) )
-34 283 deg cm2 dmol-1 (5)
Figure 3. Determination of the molecular weight of the Ru(bpy)2Im-
H21(30-mer)/Ru(NH3)5-H21(30-mer) peptide sample by sedimentation
equilibrium ultracentrifugation. The observed data can be fit to a dimeric
model (solid line).
where k is a wavelength-dependent constant equal to 2.5 at 222
nm, and n is the number of residues per helix.39 Significantly,
the observed helicity of the H21(30-mer) peptides increases with
increasing peptide concentration, which is consistent with the
formation of a self-assembled coiled-coil structure. Recent
studies40 of some coiled-coil peptides have analyzed the
concentration dependence of θ222 in terms of a two-state
monomer-dimer equilibrium for which the dissociation con-
stant, Kd, can be calculated from eq 6:
at positions corresponding to the monomer and dimer species
(MW < 8.7 kDa). Higher concentrations of glutaraldehyde
produced more intense dimer bands. However, no evidence for
higher cross-linked oligomers was observed, even under saturat-
ing conditions of glutaraldehyde. To directly determine the
molecular weight of the peptide in solution, the sedimentation
equilibrium of equimolar samples of the [Ru(bpy)2Im-H21(31-
mer)] and [Ru(NH3)5-H21(31-mer)] metallopeptides were stud-
ied by analytical ultracentrifugation. Figure 3 shows that the
results obtained within the concentration range 20-80 µM can
be accurately fit to a dimer model using the averaged molecular
weight of 7.5 kDa for the dimer. Metal derivatization of the
peptide does not seem to perturb its overall aggregation state
as was sometimes observed in the case of organic fluorophores
attached to coiled-coil peptides.41
Preparation of the ET Metallo-heterodimer. The results
just discussed show that the two noncovalent metallohomodimers,
[Ru(trpy)(bpy)-H21(30-mer)]2 and [Ru(NH3)5-H21(30-mer)]2,
can be readily prepared. In a recent study of a related peptide,
stopped-flow CD spectroscopy was used to examine the
unfolding kinetics of a dimeric coiled-coil peptide to yield a
half-life of several minutes.42 Thus, the desired ET heterodimer
(Figure 4) was generated in situ by mixing an approximately
equimolar solution of the two homodimers and allowing the
sample to stand under ambient conditions for at least 25 min.
This procedure yields a statistical mixture of the peptide
homodimers and heterodimers that cannot be separated from
one another. However, we note that within this statistical
distribution of metallopeptides, the [Ru(trpy)(bpy)-H21(30-
mer)]/[Ru(NH3)5-H21(30-mer)] heterodimer is the only species
that can display intracomplex ET.
Kd ) 2[M0] (1 - ∆θ/∆θmax)2/(∆θ/∆θmax
)
(6)
where [M0] is the total peptide concentration, ∆θ ) (θobs
-
θ0), ∆θmax ) (θmax - θ0), θmax is the ellipticity of the folded
dimer, and θ0 is the ellipticity of the unfolded monomer, which
is usually taken to be 2500 deg cm2 dmol-1. All ellipticities
are at 222 nm. The inset in Figure 2 shows the results of a
nonlinear least-squares fit of the concentration dependence of
θ222 to eq 6, which yields values of Kd ) 1.5 ( 0.4 µM and
θmax ) -23 800 ( 130 deg cm2 dmol-1, showing that the
peptide has a maximum helicity of 69%. These results are
consistent with those reported for studies25 of synthetic coiled-
coils with similar amino acid sequences to that of H21(30-mer)
in which the deviations from ideality are likely due to endgroup
fraying.
The CD properties of the [Ru(trpy)(bpy)-H21(30-mer)] and
[Ru(NH3)5-H21(30-mer)] metallopeptides are qualitatively simi-
lar to those of the apopeptide. Both show a positive CD band
at <195 nm and two minima at 208 and 222 nm. In addition,
both peptides display values of θ222/θ208 ) 1.05, indicating that
they adopt the coiled-coil structure when dissolved in aqueous
solution. The concentration dependence of θ222 for the two
metallopeptides can be fit to eq 6 to give Kd ) 0.27 ( 0.1 µM
and θmax ) -19 031 ( 76 deg cm2 dmol-1 for the [Ru(NH3)5-
H21(30-mer)] and Kd ) 0.24 ( 0.1 µM and θmax ) -26 176
( 114 deg cm2 dmol-1 for the [Ru(trpy)(bpy)-H21(30-mer)].
Gel Electrophoresis, Chemical Cross-Linking Experi-
ments, and Analytical Ultracentrifugation. SDS-PAGE ex-
periments (not shown) were used to determine the oligomer-
ization state of the H21(30-mer) and showed the presence of
monomers and dimers in solution. Subsequent intersubunit cross-
linking experiments using glutaraldehyde also produced bands
EPR Measurements. The nitroxide-nitroxide distance mea-
sured by EPR was used to estimate the distance between the
two R-carbons at positions 21 and 21′ within the peptide dimer.
These positions are the sites of ruthenium attachment in the
ET metalloprotein already described. The identity of the spin-
labeled variant was verified by ESI-MS and its dimeric
aggregation state was confirmed by analytical ultracentrifuga-
tion.
Figure 5a shows a comparison of the average of five
independently obtained spectra of the [MTSL-C21(30-mer)]2
(38) Cooper, T. M.; Woody, R. W. Biopolymers 1990, 30, 657.
(39) Chen, Y.-H.; Yang, J. T.; Chau, K. H. Biochemistry 1974, 13, 3350.
(40) Wendt, H.; Berger, C.; Baici, A.; Thomas, R. M.; Bosshard, H. R.
Biochemistry 1995, 34, 4097.
(41) Daugherty, D. L.; Gellman, S. H. J. Am. Chem. Soc. 1999, 121,
4325.
(42) Zitzewitz, J. A.; Bilsel, O.; Luo, J.; Jones, B. E.; Matthews, C. R.
Biochemistry 1995, 34, 12812.

8004 J. Am. Chem. Soc., Vol. 122, No. 33, 2000
KorniloVa et al.
Figure 5. EPR measurements of the interspin distance and modeling
of the helical geometry. (a) The low-temperature absorbance spectrum
for the [MTSL-C21(30-mer)]2 peptide dimer (thick line), overlapped
with the noninteracting spectrum from the wild-type titration experiment
(thin line) and from the GuHCl denaturation experiment (dashed line).
(b) The broadening function due to interspin dipolar interaction shown
in Fourier space (dots) and the fit to a two-Gaussian to eliminate random
noise (solid line). (c) Model of the peptide dimer with nitroxide spin
labels at the f positions. Helix wheels of the peptide dimer are shown
as two circles, with the center of the helix (X), the CR position (CR),
and the averaged nitroxide position (N-O).
Figure 4. Computer-generated model of the heterodimer electron-
transfer metalloprotein.
dimer, with the noninteracting spectra obtained from a magneti-
cally dilute sample of [MTSL-C21(30-mer)] dissolved in a
stoichiometric excess of unmodified [C21(30-mer)], and a
sample of [MTSL-C21(30-mer)] denatured in 6 M GuHCl. This
comparison reveals a small but distinct spectral line broadening
in the spin-labeled dimer due to electron-electron dipolar
interactions. Importantly, no evidence for the presence of
monomeric peptides was detected in these samples.
Fourier deconvolution analysis was performed to determine
the interspin distance from these spectra. Previous studies have
measured interspin distances in derivatized R-helices to show
that the experimentally determined values were identical to those
calculated from idealized models within a standard deviation
of (0.9 Å.33 The dipolar broadening function extracted from
the analysis in the Fourier space is shown in Figure 5b. From
this function, an average interspin distance of 22.5 Å was
obtained. These results were invariant to the choice of reference
sample used to compare the line shapes (i.e., the magnetically
diluted, or chemically denatured samples).
An estimate of the intersubunit CR(21)-CR(21′) distance can
be obtained from the spin-label distance using the following
geometrical argument (Figure 5c). In several previous studies,
it has been found that the average location of a nitroxide spin-
label attached to an R-helix lies 7 Å away from the center of
the helix on the line connecting the helical axis and the R-carbon
of the cysteine residue.33,43 The CR positions are located 2.5 Å
away from the helical axis along this radius. Thus, the two CR
sites at the f positions of the dimer lie nearly on the straight
line connecting the two helical axes. The distance between these
sites is therefore measured to be 13.5 ( 0.9 Å, which is within
experimental error to the average value of 14.2 Å seen in the
1.8 Å resolution crystal structure of GCN-4.44 The H21(30-
mer) is therefore believed to exist within the nativelike structure
of an R-helical coiled-coil. These results also indicate that the
peptide dimer consists of a parallel arrangement of R-helices
because an antiparallel arrangement would result in very large
interspin distances giving rise to narrow EPR signals.
Molecular Modeling. As described later, the pulse radiolysis
technique was used to determine the rates of a ground-state ET
occurring from the RuII(NH3)5 electron-donor (d6) to the RuIII-
(trpy)(bpy) acceptor (d5). Because this reaction involves metal-
based d-orbitals, an estimate of the metal-to-metal distance in
the ET heterodimer was obtained by energy-minimization
calculations based on the results previously obtained45 for the
related metalloheterodimer [Ru(bpy)2Im-H21(31-mer)]/[Ru-
(NH3)5-H21(31-mer)] whose synthesis has been described
earlier.21 These calculations were initiated by coordinating the
appropriate ruthenium complexes to the H21 positions of the
31-mer apopeptide using starting coordinates taken from the
crystal structures of the GCN-4 dimerization domain,44 and the
relevant model complexes, cis-[Ru(bpy)2Im2](BF4)246 and [(NH3)5-
Ru(His)]Cl3•H2O.47 Separate energy minimization calculations
were then performed for 640 initial structures in which the side-
(44) O’Shea, E. K.; Klemm, J. D.; Kim, P. S.; Alber, T. Science 1991,
254, 539.
(45) Unpublished results.
(46) Reddy, K. B.; Cho, M. P.; Wishart, J. F.; Emge, T. J.; Isied, S. S.
Inorg. Chem. 1996, 35, 7241.
(43) Macosko, J. C.; Kim, C. H.; Shin, Y.-K. J. Mol. Biol. 1997, 267,
1139.
(47) Krogh-Jespersen, K.; Westbrook, J. D.; Potenza, J. A.; Schugar, H.
J. J. Am. Chem. Soc. 1987, 109, 7025.

Synthetic Electron-Transfer Protein Characterization
J. Am. Chem. Soc., Vol. 122, No. 33, 2000 8005
Scheme 1
chain dihedral angles ø1 and ø2 for the H21 residue were
systematically varied from -180° to +180° and in which the
backbone structure of the peptide remained fixed. For the current
study, the two lowest energy structures were identified, and their
ruthenium polypyridyl sites were changed to [Ru(trpy)(bpy)].
Energy minimization calculations were subsequently performed
followed by molecular dynamics simulations at 298 K to
overcome local energy barriers. After a second round of energy
minimization was performed an average metal-to-metal distance
of 24.4 Å was found, which was within 0.2 Å of the value
obtained from the original starting coordinates. From these final
coordinates, the edge-to-edge distance is 18.4 Å as measured
from the Cγ positions of the histidine residues.
Electron-Transfer Studies. The monitoring wavelength for
studying the ET kinetics was determined by examining the
difference absorption spectra obtained after oxidizing the [RuII-
•
(trpy)(bpy)-H21(30-mer)] peptide with N3 radical. Correction
for the radiolytic yield permits calculation of the differential
extinction coefficient at each observed wavelength (data not
shown), and the largest absorption bleach was seen to occur at
490 nm (∆ꢀ ) -5980 M-1 cm-1).
Figure 6. (a) Transient absorption plot obtained at 490 nm for the
pulse radiolysis of a 5-µM solution of the [RuII(trpy)(bpy)-H21(30-
mer)] peptide in a N2O-saturated solution containing 1 mM NaN3 and
50 mM phosphate buffer at pH 7. A tungsten lamp was used as the
source of the analyzing beam. (b) Transient absorption plot obtained
at 490 nm for the pulse radiolysis of a 5.2-µM solution of the [Ru-
(trpy)(bpy)-H21(30-mer)]/RuII(NH3)5-H21(30-mer)] ET heterodimer
(total peptide concentration ) 10.4 µM) in a N2O-saturated solution
containing 1 mM NaN3 and 50 mM phosphate buffer at pH 7. A pulsed
xenon arc lamp was used as the source of the analyzing beam.
The kinetics of the intramolecular ET reaction occurring
within the [Ru(trpy)(bpy)-H21(30-mer)]/[Ru(NH3)5-H21(30-
mer)] heterodimer was measured by pulse-radiolysis, as shown
in Scheme 1. In this experiment, the radiolytically generated
azide radical was used to oxidize either of the two ruthenium
centers in the reduced heterodimer to create a nonequilibrium
distribution of the two singly oxidized species. Electron transfer
brings the system back to redox equilibrium in a process that
can be followed by transient optical absorption. The ET
measurements were carried out on solutions having concentra-
tions in the range of 1.2 to 20 µM for the heterodimer (total
peptide concentrations of 2.4 to 40 µM).48 Control experiments
were performed on similar solutions of the [Ru(trpy)(bpy)-H21-
(30-mer)] homodimer. The usable concentration range for these
experiments was limited on the low end by the monomer-dimer
equilibrium constant, and at the high end by the absorbance of
the sample.
preparation (total peptide concentration of 10.4 µM). A rapid
bleach is observed at 490 nm that indicates that the ruthenium
polypyridyl centers have been oxidized by the azide radical.
This process proceeds with an estimated second-order rate
constant of ∼kox ) 2 × 109 M-1 s-1, which is similar to that
already seen for the oxidation of the related homodimer.
However, the absorption change observed for the ET het-
erodimer recovers by a process that can be fit to a sum of two
exponentials within the 70-ms time scale of the experiment. The
rate of the faster major component is independent of concentra-
tion, and is thus assigned to the intracomplex ET reaction.
Measurements were performed on samples containing 1.2, 2.0,
4.0, 5.2, 10, and 20 µM solutions of the ET heterodimer to give
first-order rate constants of ket ) 290 ( 60 (12), 410 ( 50
(11), 400 ( 35 (9), 400 ( 50 (12), 480 ( 40 (11), and 310 (
50 s-1 (9), respectively, where the number of independent
observations made for each concentration are shown in paren-
theses. These data give an average value of ket ) 380 ( 80 s-1
for the intramolecular ET reaction occurring within the ET
heterodimer. This value supersedes the one previously reported
for a closely related peptide,21 in which the laser flash quench
method was improperly used because of the incorrect choice
of the exogeneous quencher that allowed for the pre-flash
oxidation of the donor site.49
Figure 6a shows a typical transient absorption trace from a
control experiment involving the pulse radiolysis of [Ru(trpy)-
(bpy)H21(30-mer)] (total peptide concentration ) 10 µM). A
rapid bleach at 490 nm is observed that indicates formation of
the oxidized [RuIII(trpy)(bpy)H21(30-mer)] species. The second-
order rate constant for this process is kox ≈ 1.5 × 109 M-1 s-1
,
based on the pseudo-first-order rate constants measured at
different peptide concentrations. As shown, the oxidized met-
allopeptide lives for >70 ms in the absence of a reductant.
Observations at longer times show evidence for a partial (∼20%)
recovery of the MLCT bleach occurring within ∼1000 ms.
Figure 6b shows the results of the radiolysis experiment
performed on a 5.2-µM solution of the ET heterodimer
(48) Samples of the ET heterodimer were obtained from a total of seven
independent preparations of peptide. Samples from six of these preparations
displayed consistent kinetic behavior that is described in the text. The
samples obtained from the remaining preparation displayed much faster
kinetics appearing on the submillisecond time scale. However, these kinetics
were observed for an initial MLCT bleach that was an order of magnitude
smaller than that expected from comparison with the behavior seen for the
[Ru(trpy)(bpy)-H21(30-mer)] homodimer (-∆ꢀ ) 3 to 6 × 102 M-1 cm-1),
and are therefore ascribed to the presence of an unidentified reductant in
solution.
Examination of Figure 6b also shows the existence of a
persistent bleach at 490 nm. This behavior is similar to that
seen in Figure 6a and can be attributed to the radiolytic oxidation
of the [RuII(trpy)(bpy)-H21(30-mer)] homodimers present in the
solution (these account for half of the [RuII(trpy)(bpy)] species

8006 J. Am. Chem. Soc., Vol. 122, No. 33, 2000
KorniloVa et al.
present), as well as to the accumulation of the ET-inert product,
[RuII(trpy)(bpy)-H21(30-mer)]/[RuIII(NH3)5-H21(30-mer)]. These
oxidized ruthenium polypyridyl species can only decay by a
bimolecular ET reaction with the a [RuII(NH3)5-H21(30-mer)]
species. However, only a small percentage can react in this way
because the concentration of the ruthenium(II) pentammine sites
has been lowered by repeated pulsing of the sample. The kinetic
fits to the data show that the slower, minor component to the
bleach recovery appears to occur with a concentration-
independent rate of kapparent ) 32 ( 9 s-1. However, the small
amplitude of this component and the 70-ms time domain of the
experiment prevents an unambiguous analysis of these data to
be made at this time.
within the range of those reported for ET reactions that occur
across comparable metal-to-metal distances in both native and
surface-modified metalloproteins. For example, Farver and
Pecht51 report a value of kmax ) 3 × 102 s-1 for the reaction
between the [Cys3-S-S-Cys26]•- disulfide radical and the
Cu2+ center of wild-type azurin that are located 24.6 Å apart.
In addition, Gray and co-workers4b report values of kmax ) 3.1
× 104 s-1 for His54-ruthenated cytochrome c (dM-M ) 22.5
Å), kmax ) 2.4 × 102 s-1 for His107-ruthenated azurin for (dM-M
) 25.7 Å), and kmax ) 1.3 × 102 s-1 for His126-ruthenated
azurin (dM-M ) 26 Å). The behavior of the H21(30-mer)
metalloprotein clearly falls within this range of data and
highlights the rich variability in the ET behavior of protein-
based systems.
Discussion
At this point, it should be acknowledged that relatively little
is presently known about the conformational dynamics of the
ET heterodimer. It is certainly possible that internal motions of
the metalloprotein may play an important role in regulating the
observed intracomplex ET rate. However, it appears likely that
the gross dynamics of the monomer/dimer equilibrium do not
gate the observed ET reaction because Zitzewitz et al.42 recently
used stopped-flow CD to determine that the monomer/dimer
equilibrium for the related GCN-4 coiled-coil peptide has an
unfolding rate constant of 3.3 × 10-3 s-1, which is significantly
smaller than the ET rate constant measured in the present
system. Nevertheless, future work will attempt to address this
issue by examining the temperature and/or viscosity dependence
of ket for this and related systems. In the absence of such
experiments, a pathway analysis of the H21(30-mer) ET
heterodimer identifies the primary coupling path (metal-to-metal)
in this system to consist of 22 covalent bonds and a critical
interhelix through-space jump of 3 Å between the Câ of Lys22
and Cγ of Ile23′ of the next heptad repeat.52 Interestingly, a
secondary pathway is also identified that has the same number
of covalent interactions and only a 20% lower coupling strength
than that of the primary path because of a slightly longer
through-space jump occurring between the CR of Glu20 and
the Cδ of Leu19′. Thus, ongoing work in our laboratory is
attempting to chemically modify the nature of the peptide-
peptide interface to probe the mechanisms of interfacial ET
reactions.
This work describes a new approach toward the preparation
of self-assembled peptide complexes that are capable of display-
ing long-range ET reactions. In recent years, several examples
have been reported in which the principles of de novo protein
design have been used to prepare a new generation of synthetic
ET proteins.22-24 However, a unique feature of the current
system is that instead of using polypeptide chains to provide
covalent bridges between the electron donor and acceptor sites,
the ET event is now made to occur across a well-defined, yet
noncovalent, peptide-peptide interface. Accordingly, this syn-
thetic protein provides the first peptide-based model system that
is amenable to mechanistic studies of interfacial ET reactions
as they occur within native protein-protein complexes.
The results of CD, EPR, analytical ultracentrifugation, and
molecular modeling studies show that both the apo and
metalated derivatives of the [H21(30-mer)] peptide form two-
stranded R-helical coiled coils that remain essentially isostruc-
tural to the family of native leucine zipper proteins. In particular,
the EPR spin-label experiments show that the interchain C21-
C21′ distance (CR-CR) of the peptide dimer is nearly identical
to that observed from the X-ray crystal structure of GCN-4.
Based on these results, computer modeling studies suggest that
the ET heterodimer, [Ru(trpy)(bpy)-H21(30-mer)]/[Ru(NH3)5-
Im-H21(30-mer)], has an interhelix, metal-to-metal distance of
∼24 Å.
Pulse radiolysis experiments were used to measure a con-
centration-independent value of ket ) 380 ( 80 s-1 for the
interhelix ET reaction occurring between the [RuII(NH3)5H21-
(30-mer)] electron-donor and the [RuIII(bpy)(trpy)H21(30-mer)]
acceptor.
Acknowledgment. The authors thank Professor George S.
Bullerjahn for many helpful discussions. This work was partially
supported by the Petroleum Research Fund administered by the
American Chemical Society (M.Y.O.) and by the National
Institutes of Health through grant nos. GM61171 (M.Y.O.) and
GM51290 (Y.K.S.). The pulse radiolysis experiments were
carried out at Brookhaven National Laboratory under contract
DE-AC02-98CH10886 with the U.S. Department of Energy and
supported by its Division of Chemical Sciences, Office of Basic
Energy Sciences. Support from the National Science Foundation
is acknowledged for the purchase of the circular dichroism
spectrometer at BGSU through grant no. BIR-9208356. A.Y.K.
and A.F. thank the Harold and Helen McMaster Foundation
for predoctoral fellowships. R.C.L. acknowledges support from
a George S. Hammond predoctoral fellowship from the Center
for Photochemical Sciences at BGSU.
The electrochemistry measurements performed on the two
metallo-homodimers show that the driving force for this reaction
is ∆G° ) -1.11 eV, which is slightly higher than the magnitude
of λ calculated for this system.50 The observed value of ket is
therefore believed to be close to that of kmax (i.e., the activa-
tionless ET rate constant) and can therefore be compared with
those reported for related systems. Interestingly, the value falls
(49) Our studies have also found that the Ru(bpy)2(im)-H21(31-mer)
homodimer used in the previous study is photolabile in the absence of
quencher. Photoexcitation of this species results in a transient bleach of
the ruthenium MLCT peak, which does not fully recover to the original
baseline on the millisecond time scale.
(50) An estimate of the reorganization energy for this system can be
made using the Marcus cross-relation: λ12 ) 1/2 [λ11 + λ22] where λ11 and
JA0006954
λ22 are the self-exchange reorganization energies for two metal complexes
involved in the electron-transfer. Substitution of the values (ref 4c) of λ11
) 0.6 eV for Ru(bpy)3, and λ22 ) 1.20 eV for Ru(NH3)5 (pyr) yields a
value of λ12 ) 0.9 eV.
(51) Farver, O.; Skov, L. K.; Gilardi, G.; van Pouderoyen, G.; Canters,
G. W.; Wherland, S.; Pecht, I. Chem. Phys. 1996, 204, 271.
(52) Kurnikov, I.; Beratan, D. N., private communication.