Computational de novo design and characterization of a four-helix bundle protein that selectively binds a nonbiological cofactor

Article abstract of DOI:10.1021/ja044129a

We report the complete de novo design of a four-helix bundle protein that selectively binds the nonbiological DPP-Fe(III) metalloporphyrin cofactor (DPP-Fe(III) = 5, 15-Di[(4-carboxymethyleneoxy)phenyl]porphinato iron(III)). A tetrameric, D₂-sy

Full text of DOI:10.1021/ja044129a

Published on Web 01/15/2005
Computational De Novo Design and Characterization of a Four-Helix Bundle
Protein that Selectively Binds a Nonbiological Cofactor
Frank V. Cochran,† Sophia P. Wu,‡ Wei Wang,‡ Vikas Nanda,† Jeffery G. Saven,*,‡
Michael J. Therien,*,‡ and William F. DeGrado*,†,‡
Department of Biochemistry and Molecular Biophysics, Johnson Foundation, School of Medicine,
UniVersity of PennsylVania, Philadelphia, PennsylVania 19104, and Department of Chemistry,
UniVersity of PennsylVania, Philadelphia, PennsylVania 19104
Received September 27, 2004; E-mail: wdegrado@mail.med.upenn.edu
De novo design tests current understanding of metalloproteins1,2
and offers the potential to construct novel biomaterials, such as
catalysts3,4 and bioelectrochemical devices.5 The use of nonbio-
logical cofactors in these efforts offers the possibility of creating
proteins with unusual properties. We have developed computational
design methodology6,7 to derive a unique protein framework for a
specified cofactor of interest as an alternative to re-engineering
natural protein scaffolds.8 Selective cofactor recognition is a
hallmark of achieving this goal and a significant challenge,
especially in the absence of covalent attachment.9 Previously
designed heme proteins bound various metalloporphyrins with
relatively low specificity, which is likely attributable to molten
globule character.2 In contrast, natural proteins generally bind
cofactors in well-structured environments with precisely positioned
amino acid side chains. Here, we report the complete de novo design
of a nativelike protein that selectively binds a nonbiological
cofactor.
The designed protein serves as a scaffold for future electron-
transfer studies by encapsulating a symmetrical dyad of DPP-Fe
Figure 1. (a) DPP-Fe cofactor. (b) Model of designed four-helix bundle
protein containing two DPP-Fe cofactors. (c) Axial ligand interactions.
units (Figure 1a) through bis(His) coordination. DPP-Fe(III)Cl was
used as the precursor complex for incorporating the DPP-Fe unit
in the protein matrix.10 Requirements for efficient cofactor binding
dictated the precise protein backbone geometry. A single low-energy
structure (Figure 1b) was computed with a Monte Carlo simulated
annealing protocol that considered the following constraints: (1) a
metal-metal distance between 17 and 19 Å, (2) optimal His N(ꢀ)
to Fe bonding interactions, (3) second-shell hydrogen bonds between
His N(δ) and Thr O(γ) (Figure 1c), (4) minimal steric clashes, (5)
maintenance of D2 symmetry. This process led to imidazole rings
in near-perpendicular alignment as a result of the second-shell
hydrogen bonds. The identities of the remaining positions were
determined by recursive calculations with the computational design
algorithm SCADS,6,7 which provided site-dependent side chain
probabilities. In particular, SCADS located sites appropriate for
complementary charge patterning to enforce an antiparallel bundle
topology11 (E7, E14, K21, K28). Ala side chains at critical sites
promoted efficient helix-helix and helix-cofactor packing (A4,
A8, A12, A15, A18, A25). A suitable position for tyrosine was
found to aid peptide concentration measurements (Y32). Finally,
S0 and G33 were added as N- and C-terminal capping residues12
(Figure 1d). The 34-residue peptide was prepared by Fmoc-based
solid-phase synthesis.13
(d) Peptide sequence.
UV-vis spectroscopy demonstrated that the peptide binds DPP-
Fe via bis(His) coordination to low-spin Fe(III) and with a peptide/
cofactor stoichiometry consistent with the design. An increase in
Figure 2. UV-vis spectra demonstrating cofactor selectivity.
Soret band absorbance at 408 nm and a Q-band blue shift from
580 to 530 nm was observed upon adding four equivalents of
peptide to two equivalents of DPP-Fe(III) (Figure 2). A titration
showed a linear increase in Soret band intensity with added peptide
† Department of Biochemistry and Molecular Biophysics, Johnson Foundation,
School of Medicine.
‡ Department of Chemistry.
9
1346
J. AM. CHEM. SOC. 2005, 127, 1346-1347
10.1021/ja044129a CCC: $30.25 © 2005 American Chemical Society

C O M M U N I C A T I O N S
a similar UV-vis spectrum to that of free heme (Figure 2). The
CD spectrum of this peptide/heme mixture was nearly identical to
that of the unfolded peptide. These findings indicate a lack of
defined assembly involving peptide and the heme cofactor under
these conditions. This dramatic display of cofactor discrimination
strongly suggests a successful design with a nativelike hydrophobic
core specific for the shape of the DPP-Fe unit.
In conclusion, we have computationally designed a four-helix
bundle protein that selectively binds a nonbiological metallopor-
phyrin cofactor. Successful incorporation of a nonbiological cofactor
indicates the design methodology is robust and may be extended
beyond purely natural systems. Detailed structural studies are
underway to uncover the basis of cofactor selectivity. Preliminary
studies indicate the assembled protein displays functional redox
properties previously unobtainable with natural cofactors. These
findings open a path for the selective incorporation of more
elaborate cofactors17,18 into designed scaffolds to construct mo-
lecularly well-defined nanoscale materials.
Figure 3. EPR spectrum of the assembled protein at 10 K. Signals at g )
6.08 and 4.33 are assigned to unbound DPP-Fe(III)Cl and high-spin Fe
impurities, respectively.
until reaching a 2:1 peptide/DPP-Fe(III) mole ratio, after which
there was no significant increase in absorbance.
CD spectroscopy, size-exclusion chromatography, and analytical
ultracentrifugation indicated the peptide undergoes a transition from
a predominantly random coil monomer to an R-helical tetramer
upon binding DPP-Fe(III). Peptide in the absence of DPP-Fe(III)
displayed a CD spectrum indicative of a mainly random conforma-
tion. A 4:2 DPP-Fe(III)/peptide mixture produced a CD spectrum
with minima at 208 and 222 nm, indicating a large degree of helical
content. Size-exclusion chromatography showed that the 4:2 peptide/
DPP-Fe(III) species exists in a single aggregation state with an
apparent molecular weight (MWapp ) 20,200) consistent with the
design (MWcalc ) 17,230), while the apopeptide displayed a longer
retention time that was consistent with a much smaller molecular
weight. Analytical ultracentrifugation confirmed the peptide alone
exists in a predominantly monomeric form below 200 µM and that
the assembled protein sediments as a single species with MW )
17,230 ( 69.
Acknowledgment. We thank Takahiro Yano for assistance with
EPR measurements and James Lear for help with acquiring and
interpreting AU data. We acknowledge support from NIH (GM61267,
GM071628, and GM54616) and NSF (DMR 00-79909). S.P.W. is
an NSF Access Science Fellow. J.G.S. is a Cottrell Scholar of
Research Corporation.
Supporting Information Available: MALDI-MS, HPLC, DPP-
Fe binding titration, NMR and CD spectroscopy, size-exclusion
chromatography, sedimentation equilibrium, redox potentiometry. This
material is available free of charge via the Internet at http://pubs.acs.org.
References
(1) DeGrado, W. F.; Summa, C. M.; Pavone, V.; Nastri, F.; Lombardi, A.
Annu. ReV. Biochem. 1999, 68, 779.
The EPR spectrum of the assembled protein at 10 K displayed
a signal at g ) 3.35 and can be characterized as a highly anisotropic
low-spin (HALS), or type-I, spectrum (Figure 3).14 This EPR signal
is diagnostic of low-spin, six-coordinate Fe(III) porphyrin centers
with coordinating imidazole rings in near perpendicular alignment.
This result is consistent with the design, which included putative
second-shell histidine N(δ) to threonine O(γ) interactions to enforce
this orientation. The T12A variant peptide bound DPP-Fe(III) yet
displayed a rhombic, type-II EPR spectrum (g ) 3.08, 2.21, 1.40),
similar to that seen in unconstrained “maquette” systems.2 These
results demonstrate the use of second-shell interactions to modulate
the structural and electronic properties of the Fe centers.
(2) Reedy, C. J.; Gibney, B. R. Chem. ReV. 2004, 104, 617.
(3) Dwyer, M. A.; Looger, L. L.; Hellinga, H. W. Science 2004, 304, 1967.
(4) Kaplan, J.; DeGrado, W. F. Proc. Natl. Acad. Sci. U.S.A 2004, 101, 11566.
(5) Topoglidis, E.; Discher, B. M.; Moser, C. C.; Dutton, P. L.; Durrant, J.
R. ChemBioChem 2003, 4, 1332.
(6) Kono, H.; Saven, J. G. J. Mol. Biol. 2001, 306, 607.
(7) Calhoun, J. R.; Kono, H.; Lahr, S.; Wang, W.; DeGrado, W. F.; Saven,
J. G. J. Mol. Biol. 2003, 334, 1101.
(8) Uneno, T.; Ohashi, M.; Kono, M.; Kondo, K.; Suzuki, A.; Yamane, T.;
Watanabe, Y. Inorg. Chem. 2004, 43, 2852.
(9) Carey, J. R.; Ma, S. K.; Pfister, T. D.; Garner, D. K.; Kim, H. K.; Abramite,
J. A.; Wang, Z.; Guo, Z.; Lu, Y. J. Am. Chem. Soc. 2004, 126, 10812.
(10) Acid-catalyzed condensation of dipyrrylmethane and an ester-substituted
benzaldehyde was followed by treatment with DDQ to give an ester-
protected form of the porphyrin core. Metalation of the porphyrin
macrocycle was achieved with FeCl2‚4H2O in refluxing THF and ethanol.
Saponification with NaOH provided DPP-Fe(III)Cl.
1
(11) Oakley, M. G.; Hollenbeck, J. J. Curr. Opin. Struct. Biol. 2001, 11, 450.
(12) Aurora, R.; Rose, G. D. Protein Sci. 1998, 7, 21.
The 1-D H NMR spectrum of the assembled protein has well-
dispersed peaks along with broad peaks outside the diamagnetic
window, resulting from proton nuclei in close proximity to the
paramagnetic Fe(III) centers. These results suggest a well-packed
interior suitable for further structural studies.
Potentiometric studies found an apparent E1/2(Fe2+/3+) ) 103 mV
vs NHE. This value is 180 mV more positive than a similar de
novo-designed four-helix bundle protein containing the natural
Fe(III) protoporphyrin IX cofactor15 and is consistent with estab-
lished redox properties of these Fe porphyrins in bis(imidazole)
environments.16 Since the midpoint potential is similar to that of
DPP-Fe(N-Me-imidazole)2 in CH2Cl2 (81 mV vs NHE), the
cofactor is most likely bound in the relatively hydrophobic
microenvironment defined by the design.
(13) Automated solid-phase peptide synthesis was performed on a Rink amide
AM resin using standard Fmoc-based methodology. Acetylation of the
N-terminal residue with acetic anhydride was followed by TFA-mediated
cleavage and deprotection. Peptide purification and assay of homogeneity
was achieved by RP-HPLC. The expected peptide molecular mass of 3975
Da was determined by MALDI-TOF mass spectrometry.
(14) Walker, F. A. Chem. ReV. 2004, 104, 589.
(15) Ghirlanda, G.; Osyczka, A.; Liu, W.; Antolovich, M.; Smith, K. M.;
Dutton, P. L.; Wand, A. J.; DeGrado, W. F. J. Am. Chem. Soc. 2004,
126, 8141.
(16) DPP-Fe(N-Me-imidazole)2: E1/2(Fe2+/3+) ) -160 mV vs SCE (81 mV
vs NHE); iron protoporphyrin IX dimethylester(N-Me-imidazole)2:
E
1/2(Fe2+/3+) ) -410 mV vs SCE (-169 mV vs NHE). Experi-
mental conditions: solvent ) CH2Cl2; [metalloporphyrin] ) 1-2 mM;
[TBAClO4] ) 0.1 M.
(17) Uyeda, H. T.; Zhao, Y.; Wostyn, K.; Asselberghs, I.; Clays, K.; Persoons,
A.; Therien, M. J. J. Am. Chem. Soc. 2002, 124, 13806.
(18) Redmore, N. P.; Rubtsov, I. V.; Therien, M. J. J. Am. Chem. Soc. 2003,
125, 8769.
Initial studies reveal remarkable cofactor selectivity. A 4:2
mixture of peptide and heme (Fe(III) protoporphyrin IX) displayed
JA044129A
9
J. AM. CHEM. SOC. VOL. 127, NO. 5, 2005 1347