De Novo Design and Molecular Assembly of a Transmembrane
Diporphyrin-Binding Protein Complex
Ivan V. Korendovych,†Alessandro Senes,†,⊥Yong Ho Kim,†James D. Lear,†H. Christopher Fry,|
Michael J. Therien,‡J. Kent Blasie,|F. Ann Walker,§and William F. DeGrado*,†,|
Department of Biochemistry and Biophysics, UniVersity of PennsylVania, Philadelphia, PennsylVania 19104,
Department of Chemistry, Duke UniVersity, Durham, North Carolina 27708, Department of Chemistry and
Biochemistry, UniVersity of Arizona, Tucson, Arizona 85721, and Department of Chemistry, UniVersity of
PennsylVania, Philadelphia, PennsylVania 19104
Received August 18, 2010; E-mail: email@example.com
Abstract: The de novo design of membrane proteins remains
difficult despite recent advances in understanding the factors that
drive membrane protein folding and association. We have
designed a membrane protein PRIME (PoRphyrins In MEmbrane)
that positions two non-natural iron diphenylporphyrins (FeIIIDPP’s)
sufficiently close to provide a multicentered pathway for trans-
membrane electron transfer. Computational methods previously
used for the design of multiporphyrin water-soluble helical proteins
were extended to this membrane target. Four helices were
arranged in a D2-symmetrical bundle to bind two Fe(II/III) diphe-
nylporphyrins in a bis-His geometry further stabilized by second-
shell hydrogen bonds. UV-vis absorbance, CD spectroscopy,
analytical ultracentrifugation, redox potentiometry, and EPR
demonstrate that PRIME binds the cofactor with high affinity and
specificity in the expected geometry.
Significant progress has been achieved in the computational
design of functional water-soluble proteins.1-6However, the de
novo design of membrane proteins remains difficult despite recent
advances in understanding the factors that drive membrane protein
folding and association.7Here, we present the de novo design of a
membrane protein PRIME (PoRphyrins In MEmbrane) that utilizes
a non-natural iron diphenylporphyrin (FeIIIDPP) with the ultimate
goal of facilitating electron transfer across a bilayer. Transmembrane
(TM) electron transfer lies at the heart of photosynthesis and ATP
production in a variety of organisms and thus is of great
fundamental interest and, potentially, of practical importance.
Considerable progress has been made in the design of water-
soluble multiheme proteins8-11and amphiphilic maquettes12-15
that position a single heme in the membrane phase. Also, in an
elegant study, Cordova et al.16designed a TM peptide that binds a
single heme between two helices whose geometry is defined by a
GXXXG motif. To be generally useful for transmembrane electron
transfer, it is important to also design systems that position multiple
redox-active cofactors sufficiently close to provide a multicentered
pathway for electrons to rapidly pass across the bilayer.
Here we extend previous computational methods used for the design
of water-soluble multiheme proteins to membrane targets. The design of
PRIME is based on the backbone of a water-soluble multiporphyrin-
binding peptide.17,18Its fold appeared to be particularly well suited for a
membrane environment, because it has a tight interhelical “Ala-coil”
Four helices were arranged in a D2-symmetrical bundle to bind two
Fe(II/III) diphenylporphyrins in a bis-His geometry. In the design, the
His ligands are stabilized by a bifurcated second-shell hydrogen bond
with a main chain carbonyl and a Thr (T18) hydroxyl from a
neighboring helix (Figure 1).20
Following the optimization of the coordination sphere of the iron,
the backbone was repacked to produce a final sequence. Side chains
were selected from an extended, energy-based conformer library
(Supporting Information) using Dead End Elimination, followed
†Department of Biochemistry and Biophysics, University of Pennsylvania.
|Department of Chemistry, University of Pennsylvania.
§University of Arizona.
⊥Present address: Department of Biochemistry, University of Wisconsins
Figure 1. (a) Evolution of the four-helix porphyrin binding bundles: from
water-soluble proteins through amphiphilic maquettes to PRIME. Hydro-
philic residues are shown in blue, hydrophobic residues are shown in green,
and metal cofactor is shown in brown. (b) The design of the iron
coordination site in PRIME. (c) General approach to the design of PRIME.
The final repacked model of PRIME in a bilayer (yellow) viewed along
the directions parallel (d) and normal (e) to the membrane. The sequence
of PRIME is Ac-AIYGILAHSL ASILALLTGF LTIW-CONH2.
Published on Web 10/14/2010
10.1021/ja107487b 2010 American Chemical Society
15516 9 J. AM. CHEM. SOC. 2010, 132, 15516–15518
by Monte Carlo/Self Consistent Mean Field. Pairwise energies were
calculated with the CHARMM22 force field and Lazaridis implicit
membrane solvation (IMM1). Models were ranked by oligomer-
ization energy, i.e. the difference between the energy of the complex
and that of the monomeric state (a membrane solvated helical state,
with relaxed side chain conformations), and the lowest energy model
was chosen for experimental characterization. The two iron(III)
diphenylporphyrin molecules in the PRIME model form a path for
electron transfer across the membrane. The cofactor in the model
is partially accessible near the ends of the bundle to allow water-
soluble reagents to access the cofactors (Figure 1).
PRIME is insoluble in water but could be solubilized in detergent
micelles and phospholipid bilayers. UV-vis spectroscopy and analyti-
cal ultracentrifugation demonstrated that the peptide assembled with
the iron(III) diphenylporphyrin cofactor in the expected stoichiometry
when solubilized in dodecyl phosphatidylcholine (DPC) micelles.
Analytical ultracentrifugation showed that PRIME is predominantly
monomeric in DPC micelles in the absence of the cofactor. However
in the presence of the cofactor the assembly occurs in a fully
cooperative process, and the holoprotein is greater than 90% formed
Supporting Information). The binding of the cofactor (12 µM) was
also monitored by the shift in the maximum and intensity of the Soret
band upon addition of the peptide at various concentrations. The
titration curve (Figure 2) is typical of a tight-binding isotherm with a
clean break at a stoichiometry of two peptides per porphyrin as
expected for a diporphyrin four-helix bundle. The stoichiometry was
additionally confirmed by a Job’s plot (Figure S6, Supporting Informa-
more weakly with the closely related hemin and octaethylporphyrin
(Figures 2 and S7, Supporting Information). Additionally, mutation
of key residues A15 and G19 to isoleucine, which introduces steric
clashes around the porphyrin binding cleft, effectively eliminates the
binding of iron(III) diphenylporphyrin (Figure 2). Mutation of the
second-shell ligand T18 to alanine also has a detrimental effect on
binding (Figure 2).
PRIME also binds the cofactor in palmitoyl oleoyl phosphati-
dylcholine bilayers as evidenced by UV-vis spectra. The position
of the Soret band maximum at 410 nm corresponds well to the
value observed in micelles (409 nm) (Figure S8, Supporting
CD and EPR spectroscopies showed that the peptide bound the
cofactors specifically as intended in the design. CD spectroscopy
confirmed the R-helical secondary structure both in the presence
and in the absence of the cofactor in DPC micelles (Figure 3).
Additionally, binding of the achiral porphyrin to PRIME leads to
an induced CD signal in the Soret region of the absorbance
spectrum. Negative and positive peak extrema are observed at 415
and 402 nm, respectively, with a crossover point (408 nm) near
the maximum in the absorption spectrum (409 nm) (Figure 3). These
spectral features are consistent with the exciton coupling expected
from the design in which adjacent porphyrins are oriented with a
left-handed propeller twist between their porphyrin planes.21
Additional insight into the structure of the designed assembly
can be obtained from EPR data. The designed model predicts the
angle of 82° between the planes of the His imidazole rings bound
to the iron. Such a coordination sphere should result in a “highly
anisotropic low spin” (HALS) or Type I iron(III) EPR signal.22
The observed X-band EPR spectrum agrees well with the prediction:
the HALS iron(III) signal (g ) 3.64) is indeed observed (Figure 4)
and is consistent with the spectra of related water-soluble designs
2PA and 4PA.17,18The shoulder at g ) 3.37 is characteristic of a
coordination geometry with a somewhat smaller angle between the
His planes, yet greater than 60°.23Previously studied Fe(III)
porphyrin binding maquettes showed multiple peaks between g )
2.0 and 3.0, indicating more relaxed orientations of the His
Reversible chemical reduction of the cofactor is accompanied
by the shift in the Soret band maximum to 419 nm and the
resolution of the broad Q-band into several components in the
510-560 nm region (Figure S11, Supporting Information), indica-
tive of bis-His ligated low-spin Fe(II) porphyrins.22Potentiometric
titration of the PRIME-FeIIIDPP assembly shows two redox
waves (Figure S12, Supporting Information) with apparent
E1/2(FeIIIFeIII/FeIIFeIII) and E1/2(FeIIIFeII/FeIIFeII) of -97 ( 3 mV
and -168 ( 3 mV vs NHE, respectively. The 71 mV difference is
typical for closely positioned hemes (70-100 mV observed for
maquettes with Fe-Fe separation of ∼12 Å),14consistent with the
design. Interestingly, a similar potential separation (80 mV) is also
observed for yeast cytochrome bc1hemes bHand bL.24
In conclusion, we have demonstrated that minimalistic principles
combined with computational design can be successfully applied
Figure 2. PRIME’s specificity. Titration of different cofactors (12 µM) with
method, 2 mM DPC; empty circles: 20 mM DPC); black triangles: hemin with
PRIME in DPC (20 mM); green squares: FeIIIDPP with PRIME T18A in DPC
concentration was 12 µM in all cases. Unless explicitly noted, all samples were
reconstituted using the fast equilibration method in 10 mM phosphate buffer (pH
7.4) as described in the Supporting Information.
Figure 3. CD spectra of PRIME in the absence (blue) and in the presence
(red) of FeIIIDPP. Experimental details are given in the Supporting
Information. In the case of PRIME-FeIIIDPP, MRE values in the 360-460
nm region are normalized to the number of cofactor molecules.
J. AM. CHEM. SOC. 9 VOL. 132, NO. 44, 2010
to create metalloproteins that assemble into well-defined structures
exclusively in a membrane environment. In the absence of the
hydrophobic driving force we were able to design a membrane
peptide PRIME that binds an unnatural cofactor tightly and
selectively assembling into a tetrameric bundle. This represents an
advance in our understanding of the factors that drive membrane
metalloprotein assembly, as well as a major step toward the design
of artificial electro- and photosystems.
Acknowledgment. We are grateful to Dr. Andrei V. Astashkin
for running the EPR spectra and for estimating the molar ratio of
high-spin to low-spin Fe(III) present. We thank Prof. Jeffery G.
Saven for helpful discussions. This work was supported by the
MRSEC program of the NSF and the NIH, Grant GM54616.
Supporting Information Available: Details of the computational
design and experimental characterization of PRIME. This material is
available free of charge via the Internet at http://pubs.acs.org.
(1) Lu, Y.; Yeung, N.; Sieracki, N.; Marshall, N. M. Nature 2009, 460, 855–
(2) Nanda, V. Nat. Chem. Biol. 2008, 4, 273–275.
(3) Ghosh, D.; Pecoraro, V. L. Curr. Opin. Chem. Biol. 2005, 9, 97–103.
(4) Faiella, M.; Andreozzi, C.; de Rosales, R. T. M.; Pavone, V.; Maglio, O.;
Nastri, F.; DeGrado, W. F.; Lombardi, A. Nat. Chem. Biol. 2009, 5, 882–
(5) Röthlisberger, D.; Khersonsky, O.; Wollacott, A. M.; Jiang, L.; DeChancie,
J.; Betker, J.; Gallaher, J. L.; Althoff, E. A.; Zanghellini, A.; Dym, O.;
Albeck, S.; Houk, K. N.; Tawfik, D. S.; Baker, D. Nature 2008, 453, 190–
(6) Rosenblatt, M. M.; Wang, J.; Suslick, K. S. Proc. Natl. Acad. Sci. U.S.A.
2003, 100, 13140–13145.
(7) Ghirlanda, G. Curr. Opin. Chem. Biol. 2009, 13, 643–651.
(8) Reedy, C. J.; Gibney, B. R. Chem. ReV. 2004, 104, 617–650.
(9) Koder, R. L.; Anderson, J. L. R.; Solomon, L. A.; Reddy, K. S.; Moser,
C. C.; Dutton, P. L. Nature 2009, 458, 305–309.
(10) Bender, G. M.; Lehmann, A.; Zou, H.; Cheng, H.; Fry, H. C.; Engel, D.;
Therien, M. J.; Blasie, J. K.; Roder, H.; Saven, J. G.; DeGrado, W. F.
J. Am. Chem. Soc. 2007, 129, 10732–10740.
(11) Jiang, L.; Althoff, E. A.; Clemente, F. R.; Doyle, L.; Röthlisberger, D.;
Zanghellini, A.; Gallaher, J. L.; Betker, J. L.; Tanaka, F.; Barbas, C. F.;
Hilvert, D.; Houk, K. N.; Stoddard, B. L.; Baker, D. Science 2008, 319,
(12) Ye, S.; Discher, B. M.; Strzalka, J.; Noy, D.; Zheng, S.; Dutton, P. L.;
Blasie, J. K. Langmuir 2004, 20, 5897–5904.
(13) Noy, D.; Discher, B. M.; Rubtsov, I. V.; Hochstrasser, R. M.; Dutton, P. L.
Biochemistry 2005, 44, 12344–12354.
(14) Discher, B. M.; Noy, D.; Strzalka, J.; Ye, S.; Moser, C. C.; Lear, J. D.;
Blasie, J. K.; Dutton, P. L. Biochemistry 2005, 44, 12329–12343.
(15) Ye, S.; Discher, B. M.; Strzalka, J.; Xu, T.; Wu, S. P.; Noy, D.; Kuzmenko,
I.; Gog, T.; Therien, M. J.; Dutton, P. L.; Blasie, J. K. Nano Lett. 2005, 5,
(16) Cordova, J. M.; Noack, P. L.; Hilcove, S. A.; Lear, J. D.; Ghirlanda, G.
J. Am. Chem. Soc. 2007, 129, 512–518.
(17) McAllister, K. A.; Zou, H.; Cochran, F. V.; Bender, G. M.; Senes, A.;
Fry, H. C.; Nanda, V.; Keenan, P. A.; Lear, J. D.; Saven, J. G.; Therien,
M. J.; Blasie, J. K.; DeGrado, W. F. J. Am. Chem. Soc. 2008, 130, 11921–
(18) Cochran, F. V.; Wu, S. P.; Wang, W.; Nanda, V.; Saven, J. G.; Therien,
M. J.; DeGrado, W. F. J. Am. Chem. Soc. 2005, 127, 1346–1347.
(19) Walters, R. F. S.; DeGrado, W. F. Proc. Natl. Acad. Sci. U.S.A. 2006,
(20) Berry, E. A.; Walker, F. A. J. Biol. Inorg. Chem. 2008, 13, 481–498.
(21) Holmes, A. E.; Das, D.; Canary, J. W. J. Am. Chem. Soc. 2007, 129, 1506–
(22) Walker, F. A. Chem. ReV. 2004, 104, 589–616.
(23) Yatsunyk, L. A.; Dawson, A.; Carducci, M. D.; Nichol, G. S.; Walker,
F. A. Inorg. Chem. 2006, 45, 5417–5428.
(24) T’sai, A. L.; Palmer, G. Biochim. Biophys. Acta 1983, 722, 349–363.
Figure 4. EPR spectrum of FeIIIDPP (1.5 mM) bound to PRIME (3.1 mM)
in DPC micelles (185 mM) in frozen glass containing 30% glycerol. The
peak at g ) 6.0 represents high-spin iron(III) diphenyl-porphyrin not bound
to two histidine residues; the peak at g ) 4.3 represents a small amount of
high-spin nonheme Fe(III).
15518J. AM. CHEM. SOC. 9 VOL. 132, NO. 44, 2010