R ES E A RC H

PROTEIN DESIGN with Rosetta. To specify the orientation of the

designs (11) in the membrane when expressed
in cells, at the designed lipid-water boundary
Accurate computational design of on the extracellular/periplasmic side, we incor-
porated a ring of amphipathic aromatic resi-

multipass transmembrane proteins dues and, at the lipid-water boundary on the

cytoplasmic side, a ring of positively charged
residues (Figs. 1A and 2A). Between these two
Peilong Lu,1,2 Duyoung Min,3 Frank DiMaio,1,2 Kathy Y. Wei,1,2* Michael D. Vahey,4 rings, the surface residues are exposed to the
Scott E. Boyken,1,2 Zibo Chen,1,2 Jorge A. Fallas,1,2 George Ueda,1,2 William Sheffler,1,2 hydrophobic membrane environment; these po-
Vikram Khipple Mulligan,1,2 Wenqing Xu,5 James U. Bowie,3 David Baker1,2,6† sitions in Rosetta sequence design calculations
were restricted to hydrophobic amino acids
The computational design of transmembrane proteins with more than one membrane- (supplementary materials). Consistent with the
spanning region remains a major challenge. We report the design of transmembrane design, TMHMM predicts that the dimer de-
monomers, homodimers, trimers, and tetramers with 76 to 215 residue subunits containing signs contain 2 TMs and the single-chain design
two to four membrane-spanning regions and up to 860 total residues that adopt the (scTMHC2) contains 4 TMs (fig. S1). On average,
target oligomerization state in detergent solution. The designed proteins localize to the for each residue ~68% of the side-chain surface
plasma membrane in bacteria and in mammalian cells, and magnetic tweezer unfolding area is buried in the design models, which could
experiments in the membrane indicate that they are very stable. Crystal structures provide substantial van der Waals stabiliza-
of the designed dimer and tetramer—a rocket-shaped structure with a wide cytoplasmic tion (12).
base that funnels into eight transmembrane helices—are very close to the design Synthetic genes encoding the designs were
obtained and the proteins expressed in E. coli

Downloaded from [Link] on March 5, 2018

models. Our results pave the way for the design of multispan membrane proteins with
new functions. and mammalian cells. The dimer design with
the shortest hydrophobic span (15 residues;
TMHC2_S) was poorly behaved in both E. coli

I
n recent years, it has become possible to de A major challenge for membrane protein and mammalian cells, but the dimer designs
novo design, with high accuracy, soluble pro- design stems from the similarity of the mem- with longer spans—TMHC2, TMHC2_E, and
tein structures ranging from short con- brane environment to protein hydrophobic cores. TMHC2_L—localized to the cell membrane when
strained peptides to megadalton protein In the design of soluble proteins, the secondary expressed in human embryonic kidney (HEK)
cages (1). There have also been advances in structure and overall topology can be specified 293T cells (Fig. 1B) and in E. coli. The designed
membrane protein design, as illustrated by an by the pattern of hydrophobic and hydrophilic proteins were purified by extracting the E. coli
elegant zinc-transporting transmembrane pep- residues, with the former inside the protein and membrane fraction with detergent, followed
tide tetramer named Rocker (2) and an engi- the latter outside, facing solvent. This core de- by nickel–nitrilotriacetic acid (NTA) chroma-
neered ion-conducting oligomer based on the sign principle cannot be used for membrane tography and size exclusion chromatography
C-terminal transmembrane segment (TMs) of proteins because the apolar environment of the (SEC) with a yield of ~2 mg/L (fig. S2, A and B).
the Escherichia coli polysaccharide transporter hydrocarbon core of the lipid bilayer requires The designed proteins TMHC2, TMHC2_E, and
Wza (3). Both are single membrane–spanning that outward-facing residues in the membrane TMHC2_L eluted as single peaks in SEC, and
synthesized peptides with fewer than 36 resi- also be nonpolar. Buried hydrogen bonds be- in analytical ultracentrifugation (AUC) experi-
dues. It has also been possible to design and tween polar side chains have been demonstra- ments in detergent solution, the proteins sedi-
confirm the transmembrane topology of multi- ted to play an important role in the association mented as dimers, which is consistent with the
pass membrane proteins by using simple se- of helical peptides within the membrane, over- design models (Fig. 1C and fig. S3). For the
quence hydrophobicity and charge-based models coming the degeneracy in the nonpolar inter- single-chain scTMHC2, the major species in SEC
(4), but the extent to which the transmembrane actions (5–7). was the monomer, with a small side peak that
helices pack with each other is not clear. Design We reasoned that a recently developed meth- was readily removed by purification (fig. S2B).
of structurally defined multipass membrane pro- od for designing buried hydrogen bond net- Circular dichroism (CD) measurements showed
teins has remained a major challenge because works (8) could allow specification of the packing that the designs were a-helical and highly ther-
of the difficulty in specifying structure within interactions of transmembrane helices in multi- mal stable; the CD spectra at 95°C were similar
the membrane and in experimentally determin- pass transmembrane proteins. We first explored to those at 25°C (Figs. 1D and 2B). TOXCAT-
ing membrane protein structures generally; crys- the design of helical transmembrane proteins b−lactamase (TbL) assays (13), which couple
tal structures of the full designed oligomeric with four TMs—dimers of 76- to 104-residue E. coli survival to oligomerization and proper
states of Rocker- and the Wza-derived channel hairpins or a single chain design of 156 residues— orientation of fused antibiotic resistance mark-
have not yet been determined, and to date, there with hydrophobic spanning regions ranging ers on the N and C termini, suggest that the N
are no crystal structures of de novo–designed multi- from 21 to 35 Å (Figs. 1A and 2A), repurposing and C termini of TMHC2 are in the cytoplasm,
pass membrane proteins. the Ser- and Gln-containing hydrogen bond as in the design models (fig. S4).
networks in a designed soluble four-helix dimer We more quantitatively characterized the fold-
1
with C2 symmetry [2L4HC2_23; Protein Data ing stability of scTMHC2 using single-molecule
Department of Biochemistry, University of Washington,
Seattle, WA 98195, USA. 2Institute for Protein Design,
Bank (PDB) ID 5J0K] (8) to provide structural forced unfolding experiments (Fig. 2) (14, 15).
University of Washington, Seattle, WA 98195, USA. specificity. Four-helix bundles of different lengths The designed protein reconstituted in a bicelle
3
Department of Chemistry and Biochemistry, UCLA-DOE with backbone geometries capable of host- was covalently attached to a magnetic bead and
Institute, Molecular Biology Institute, University of California, ing these networks were produced by using a glass surface through its N and C termini
Los Angeles (UCLA), CA 90095, USA. 4Department of
Bioengineering and Biophysics Group, University of California,
parametric generating equations (9), residues (Fig. 2A and fig. S5). The distance between the
Berkeley, Berkeley, CA 94720, USA. 5Department of Biological comprising the hydrogen bond networks and bead and the surface was determined as a
Structure, University of Washington, Seattle, WA 98195, neighboring packing residues were introduced, function of the applied mechanical tension. In
USA. 6Howard Hughes Medical Institute, University of and the remainder of the sequence was opti- unfolding experiments with the force slowly
Washington, Seattle, WA 98195, USA.
*Present address: Department of Bioengineering, University of
mized by using Rosetta Monte Carlo (10) design increasing (~0.5 pN/s), unfolding transitions
California, Berkeley, CA 94720, USA. calculations to obtain low-energy sequences. were observed at ~18 pN and, upon force de-
†Corresponding author. Email: dabaker@[Link] Connecting loops between the helices were built ramping, refolding transitions were observed

Lu et al., Science 359, 1042–1046 (2018) 2 March 2018 1 of 5

R ES E A RC H | R E PO R T

at ~9 pN (80.1% of the recorded unfolding traces and fig. S5), the two refolding step sizes were (I), and an unfolded state (U ) (fig. S9). During
had one-step unfolding transitions, and 84.6% very similar (fig. S8). This unfolding and refold- unfolding at high force, only the barrier be-
of the refolding transitions had two steps) (Fig. ing asymmetry is consistent with a three-state tween the native and intermediate states is ob-
2C and figs. S6 and S7). Consistent with the in- free-energy landscape: the native state (N), an served, whereas at the lower forces at which
ternal symmetry of the single-chain design (Fig. 2A intermediate state containing only one hairpin refolding occurs, both energy barriers become

Downloaded from [Link] on March 5, 2018

Fig. 1. Design and characterization of proteins with four transmembrane protein. Line scans (yellow lines) across the membranes show sub-
helices. (A and B) From left to right, designs and data for TMHC2 stantial increase in fluorescence across the plasma membranes for TMHC2,
(transmembrane hairpin C2), TMHC2_E (elongated), TMHC2_L (long span), TMHC2_E, and TMHC2_L, but less substantial increase for TMHC2_S.
and TMHC2_S (short span). (A) Design models with intra- and extra- (C) Representative AUC sedimentation-equilibrium curves at three
membrane regions with different lengths. Horizontal lines demarcate the different rotor speeds. Each data set is globally well fit as a single ideal
hydrophobic membrane regions. Ribbon diagrams are at left, electrostatic species in solution corresponding to the dimer molecular weight. “MW
surfaces are at right, and the neutral transmembrane regions are in gray. (D)” and “MW (E)” indicate the molecular weight of the oligomer design
(B) Confocal microscopy images for HEK293T cells transfected with and that determined from experiment, respectively. (D) CD spectra
TMHC2 fused to mTagBFP, TMHC2_E fused to mTagBFP, TMHC2_L and (inset) temperature melt. No apparent unfolding transitions are
fused to mCherry, and TMHC2_S fused to enhanced green fluorescent observed up to 95°C.

Lu et al., Science 359, 1042–1046 (2018) 2 March 2018 2 of 5

R ES E A RC H | R E PO R T

Fig. 2. Folding stability of the 156-residue

single-chain TMHC2 (scTMHC2) design
with four transmembrane helices.
(A) Design model (left) and electrostatic
surface (right) of scTMHC2. N- and
C-terminal helical hairpins are colored
green and blue, respectively. Numbers
indicate the order of the four TMs in
the sequence. The linker connecting
the two hairpins is colored magenta.
Single-molecule forced unfolding experiments
were conducted by applying mechanical
tension to the N and C termini of a
single scTMHC2 (fig. S5). (B) CD spectra
of scTMHC2 at different temperatures.
No unfolding transition is observed up
to 95°C. (C) Single-molecule force-extension
traces of scTMHC2. The unfolding
and refolding transitions are denoted
with red and blue arrows. (D) Folding
energy landscape obtained from the
single-molecule experiments. N, I,

Downloaded from [Link] on March 5, 2018

and U indicate the native, intermediate,
and unfolded state, respectively.

Fig. 3. Crystal structure

of the designed trans-
membrane dimer
TMHC2_E. (A and B)
Crystal lattice packing.
(A) The extended soluble
region mediates a large
portion of the crystal
lattice packing. The four
helical hairpins in the
asymmetric unit are
colored green, gray,
yellow, and blue,
respectively. The TMs,
in magenta, forms layers
in the crystal separating the soluble regions. (B) The C2 axis of the design aligns with the crystallographic twofold. Two monomers (gray and yellow)
are paired in a dimer, whereas the other two (green and blue) form two C2 dimers with two crystallographic adjacent monomers. The space group diagram
(C121) is shown in the background. (C) Superposition of the TMHC2_E crystal structure and design model (RMSD = 0.7 Å over the core Ca atoms). (D) The
side-chain packing arrangements at layers [(C), colored squares] at different depths in the membrane are almost identical to the design model.

prominent (fig. S9). The transition rates be- cytoplasmic region, TMHC2_E, in n-nonyl-b-D- tions (RMSDs), 0.60 to 0.84 Å] (fig. S11). Both
tween the folded, intermediate, and unfolded glucopyranoside (NG). The crystals diffracted the overall structure and the core side-chain
states were determined by using the Bell mod- to 2.95-Å resolution, and we solved the struc- packing are almost identical in the crystal struc-
el (16), yielding the relative free energies of the ture by means of molecular replacement with ture and the design model, with a Ca RMSD
states and the associated barrier heights (Fig. the design model. As anticipated, the extended of 0.7 Å over the core residues (Fig. 3C). Two
2D and fig. S10) (14). The overall thermodynamic soluble region mediates the crystal lattice pack- of the three buried hydrogen bonding resi-
stability of scTMHC2 is 7.8(±0.9) kcal/mol on a ing; there are large solvent channels around dues within the membrane have conforma-
per transmembrane helix basis, which is more sta- the designed TMs likely because of the sur- tions that almost exactly match the design
ble than the naturally occurring helical membrane rounding disordered detergent molecules (Fig. model (S13 and Q93), but Q17 adopts a different
proteins studied thus far [folding free energy per 3A). Each asymmetric unit contains four heli- rotamer, with the side-chain nitrogen donat-
helix for scTMHC2 is 2.0(±0.2) kcal/(mol
helix) cal hairpins: Two are paired in a dimer, whereas ing a hydrogen bond to the main-chain carbonyl
compared with 0.7 to 0.9 kcal/(mol
helix) for the other two form two C2 dimers through crys- oxygen (Fig. 3D).
GlpG (14, 17) and 1.6 to 1.8 kcal/(mol
helix) for tallographic symmetry with two monomers in We used a similar approach to design a trans-
bacteriorhodopsin (18); error estimates in parenthe- adjacent asymmetric units. The C2 axis in the membrane trimer with six membrane-spanning
ses are propagated from the standard errors of design is perfectly aligned with the crystallo- helices (TMHC3) based on the 5L6HC3_1 scaf-
the kinetics measurements]. graphic twofold (Fig. 3B). The conformations fold (PDB ID 5IZS) (8). Guided by the results
We carried out crystal screens in different of the dimers in the three biological units are with the C2 designs, we chose a hydrophobic
detergents for each of the designs and obtained nearly identical, with very small differences due span of ~30 Å (20 residues) (Fig. 4A). The de-
crystals of the design with the most extensive to crystal packing [Ca root-mean-square devia- sign was expressed in E. coli and purified to

Lu et al., Science 359, 1042–1046 (2018) 2 March 2018 3 of 5

R ES E A RC H | R E PO R T

Fig. 4. Stability and

structural characteriza-
tion of designs with
six and eight membrane-
spanning helices.
(A) Model of designed
transmembrane trimer
TMHC3 with six
transmembrane helices.
Stick representation
from periplasmic side
(left) and lateral surface
view (right) are shown.
(B) CD characterization
of TMHC3. The design is
stable up to 95°C.
(C) Representative AUC
sedimentation-equilibrium
curves at three different
rotor speeds for TMHC3.
The data fit to a single
ideal species in solution

Downloaded from [Link] on March 5, 2018

with molecular weight
close to that of the
designed trimer. (D) Model
of designed transmem-
brane tetramer TMHC4_R
with eight transmembrane
helices. The four proto-
mers are colored green,
yellow, magenta, and blue,
respectively. (E) AUC
sedimentation-equilibrium
curves at three different
rotor speeds for TMHC4_R
fit well to a single species,
with a measured molec-
ular weight of ~94 kDa.
(F) Crystal structure of
TMHC4_R. The overall tetramer structure is very similar to the design model, with a helical bundle body and helical repeat fins. The outer helices of
the transmembrane hairpins tilt off the axis by ~10°. (G) Cross section through the TMHC4_R crystal structure and electrostatic surface. The HRD
forms a bowl at the base of the overall structure with a depth of ~20 Å. The transmembrane region is indicated in lines. (H) Three views of the
backbone superposition of TMHC4_R crystal structure and design model.

homogeneity, eluting on a gel filtration column 5L8HC4_6 (8), and the bowl was derived from a NG in the P4 space group that diffracted to
as a single homogeneous species (fig. S2C). CD designed helical repeat protein homo-oligomer 3.9-Å resolution. We solved the crystal struc-
measurements showed that TMHC3 was high- (tpr1C4_2) (19). Helical linkers were built by ture by means of molecular replacement using
ly thermostable, with the a-helical structure using RosettaRemodel (20); a nine-residue junc- the design model (Rwork/Rfree = 0.29/0.32, with
preserved at 95°C (Fig. 4B). AUC experiments tion was found to yield the correct helical re- unambiguous electron density) (table S1 and
showed that TMHC3 is a trimer in detergent gister (fig. S13). After Rosetta sequence design fig. S14). The crystal lattice packing is primarily
solution, which is consistent with the design calculations, a gene encoding the lowest en- between the extended cytoplasmic domains;
model (Fig. 4C and fig. S12A). ergy design, TMHC4_R, was synthesized. The there may be minor detergent-mediated inter-
To explore our capability to design mem- protein was expressed in E. coli and purified by actions between the transmembrane and heli-
brane proteins with more complex topologies, using nickel affinity and gel filtration chroma- cal repeat (HR) domains as well (fig. S15).
we designed a C4 tetramer with a two-ring, tography; the final yield was ~3 mg/L, and the Although the resolution is insufficient for
helical membrane-spanning region composed purified protein chromatographed as a mono- evaluating the details of the side-chain pack-
of eight TMs and an extended bowl-shaped cyto- disperse peak in SEC (fig. S2C). CD experiments ing, it does allow backbone-level comparisons.
plasmic domain formed by repeating struc- showed that the design was a-helical and ther- There are four TMHC4_R monomers in one
tures emanating away from the symmetry axis mostable up to 95°C (fig. S12B). AUC measure- asymmetric unit, with nearly identical struc-
(Fig. 4D). The design has an overall rocket shape, ments showed that TMHC4_R is a tetramer in tures (Ca RMSDs between 0.2 and 0.6 Å) (fig.
with a height of ~100 Å, and can be divided detergent solution, which is consistent with S16A). The Ca RMSDs between the structure
into three regions: the helical bundle domain the design model (Fig. 4E and fig. S12C). After a and design model are 1.2 to 1.8 Å for the mono-
(HBD), the helical repeat domain (HRD), and systematic effort to screen detergents for crys- mer transmembrane helices, 0.3 to 0.4 Å for
the helical linker between the two. The cen- tallization, we obtained crystals in a combina- the linkers, 1.1 to 1.5 Å for the HR domains,
tral HBD was derived from the soluble design tion of n-decyl-b-D-maltopyranoside (DM) and and 3.3 to 3.6 Å for the overall structure (fig.

Lu et al., Science 359, 1042–1046 (2018) 2 March 2018 4 of 5

R ES E A RC H | R E PO R T

S16B). As in the case of the C2 design, the C4 and diverse oligomeric states—are substantial 23. X. Feng, P. Barth, Nat. Chem. Biol. 12, 167–173 (2016).
symmetry axis of the design coincides with steps toward the complexity of natural trans- 24. A. D. Meruelo, S. K. Han, S. Kim, J. U. Bowie, Protein Sci. 21,
1746–1753 (2012).
the crystallographic axes of the crystal lattice membrane proteins with multiple membrane-
25. Y. Shi, Cell 159, 995–1014 (2014).
(fig. S16C). The four tetramer structures on the spanning regions and extra-membrane domains 26. R. Fernandez-Leiro, S. H. W. Scheres, Nature 537, 339–346
crystal C4 axes have overall structures very that play important roles in ligand/substrate (2016).
similar to each other and to the design model recognition and structure stabilization, such
AC KNOWLED GME NTS
(Fig. 4, F and G, and fig. S16A); the tetrameric as in the adenosine 5′-triphosphate–binding cas-
transmembrane domain, HR domain, and over- sette transporters, ion channels, ryanodine re- We thank J. Sumida for AUC support; A. Kang for crystallization
support; D. Ma and Z. Wang for crystallography support; and
all tetramer structure have Ca RMSDs to the de- ceptor, and g-secretase (25, 26). The capability to the staff at the Advanced Light Source and P. Huang, Y. Hsia,
sign model of 1.3 to 1.5 Å, 3.3 to 3.8 Å, and 3.3 to accurately design complex multipass transmem- A. Ford, L. Stewart, C. Xu, and many other members of the Baker
3.8 Å, respectively (Fig. 4H and fig. S16D, left). brane proteins that can be expressed in cells opens laboratory for helpful discussions. This work was facilitated by
The deviation in the HR domain may result the door to the design of a new generation of multi- the Hyak supercomputer at the University of Washington. Funding:
This work was supported by the Howard Hughes Medical
from crystal packing interactions between the pass membrane protein structures and functions. Institute (D.B.) and the National Institutes of Health (grant
termini; the Ca RMSDs over the first 162 resi- R01GM063919 to J.U.B.). P.L. was supported by the Raymond
dues are 2.2 to 2.3 Å (fig. S16D, right). The main and Beverly Sackler fellowship. D.M. was supported by the Basic
RE FERENCES AND NOTES Science Research Program through the National Research
deviation from the design model is a tilting of
1. P.-S. Huang, S. E. Boyken, D. Baker, Nature 537, 320–327 Foundation of Korea funded by the Ministry of Education (grant
the outer helices of transmembrane hairpins from NRF-2016R1A6A3A03007871). Author contributions: P.L. and
(2016).
the axis by ~10° (Fig. 4, F and G). 2. N. H. Joh et al., Science 346, 1520–1524 (2014). D.B. designed the research, and P.L., D.M., F.D., K.Y.W., J.U.B., and
The agreement between the crystal struc- 3. K. R. Mahendran et al., Nat. Chem. 9, 411–419 (2017). D.B. wrote the manuscript. P.L. and D.B. carried out design
tures of TMHC2_E and TMHC4_R with the calculations and developed the membrane protein design method.
4. P. Whitley, I. Nilsson, G. von Heijne, Nat. Struct. Biol. 1,
P.L. purified and characterized the designed proteins. D.M. and
design models demonstrates that transmem- 858–862 (1994).
J.U.B. designed, performed, and analyzed single-molecule forced
5. C. Choma, H. Gratkowski, J. D. Lear, W. F. DeGrado,

Downloaded from [Link] on March 5, 2018

brane homo-oligomers containing multiple unfolding experiments. K.Y.W. and M.D.V. performed mammalian
Nat. Struct. Biol. 7, 161–166 (2000).
membrane-spanning regions and extensive ex- 6. F. X. Zhou, M. J. Cocco, W. P. Russ, A. T. Brunger,
cell localization experiment. P.L. crystallized the designed proteins.
tracellular domains can now be accurately de- P.L. collected and analyzed crystallographic data with help from
D. M. Engelman, Nat. Struct. Biol. 7, 154–160 (2000).
W.X.; F.D. solved structures with help from P.L.; and S.E.B., C.Z.,
signed. For future work, the general approach 7. C. D. Tatko, V. Nanda, J. D. Lear, W. F. Degrado, J. Am.
J.A.F., G.U., and W.S. contributed the soluble scaffolds. V.K.M. wrote
Chem. Soc. 128, 4170–4171 (2006).
of first designing and characterizing hydro- the amino acid composition–based energy term. All authors
8. S. E. Boyken et al., Science 352, 680–687 (2016).
gen bond network–containing soluble versions 9. P.-S. Huang et al., Science 346, 481–485 (2014).
discussed results and commented on the manuscript. Competing
of the desired transmembrane structures, and interests: D.B., P.L., S.E.B., C.Z., J.A.F., G.U., and W.S. are
10. C. A. Rohl, C. E. M. Strauss, K. M. S. Misura, D. Baker, in
inventors on a U.S. provisional patent application submitted by the
then converting to integral membrane proteins Methods in Enzymology (2004), pp. 66–93.
University of Washington that covers computational design of
11. G. von Heijne, Nat. Rev. Mol. Cell Biol. 7, 909–918 (2006).
by redesigning the membrane-exposed residues, multipass transmembrane proteins described in this paper. Data
12. A. Oberai, N. H. Joh, F. K. Pettit, J. U. Bowie, Proc. Natl. Acad.
could be quite robust. Single-molecule forced Sci. U.S.A. 106, 17747–17750 (2009).
and materials availability: Coordinates and structure files have
unfolding and thermal denaturation exper- been deposited to PDB with accession codes 6B87 (TMHC2_E)
13. A. Elazar et al., eLife 5, e12125 (2016).
and 6B85 (TMHC4_R). All other data needed to evaluate the
iments show that the designed proteins are 14. D. Min, R. E. Jefferson, J. U. Bowie, T.-Y. Yoon, Nat. Chem. Biol. conclusions in the paper are present in the paper or the
highly stable. Although the designs lack the 11, 981–987 (2015). supplementary materials.
15. D. Min, M. A. Arbing, R. E. Jefferson, J. U. Bowie, Protein Sci.
classic small residue packing in the core that
25, 1535–1544 (2016).
is thought to be an important driver of mem- 16. G. I. Bell, Science 200, 618–627 (1978). SUPPLEMENTARY MATERIALS
brane protein folding (21–24), like natural mem- 17. R. Guo et al., Nat. Chem. Biol. 12, 353–360 (2016). [Link]/content/359/6379/1042/suppl/DC1
brane proteins they bury more surface area 18. Y.-C. Chang, J. U. Bowie, Proc. Natl. Acad. Sci. U.S.A. 111, Materials and Methods
than do typical soluble proteins, maximizing 219–224 (2014). Figs. S1 to S16
19. J. A. Fallas et al., Nat. Chem. 9, 353–360 (2017). Table S1
van der Waals packing contributions (12). The References (27–39)
20. P.-S. Huang et al., PLOS ONE 6, e24109 (2011).
range of the design features—variable trans- 21. S.-Q. Zhang et al., Structure 23, 527–541 (2015).
membrane and extracellular helix lengths and 22. R. F. S. Walters, W. F. DeGrado, Proc. Natl. Acad. Sci. U.S.A. 10 October 2017; accepted 11 January 2018
superhelical twists, extensive soluble domains, 103, 13658–13663 (2006). 10.1126/science.aaq1739

Lu et al., Science 359, 1042–1046 (2018) 2 March 2018 5 of 5

Accurate computational design of multipass transmembrane proteins
Peilong Lu, Duyoung Min, Frank DiMaio, Kathy Y. Wei, Michael D. Vahey, Scott E. Boyken, Zibo Chen, Jorge A. Fallas,
George Ueda, William Sheffler, Vikram Khipple Mulligan, Wenqing Xu, James U. Bowie and David Baker

Science 359 (6379), 1042-1046.

DOI: 10.1126/science.aaq1739

Membrane protein oligomers by design

In recent years, soluble protein design has achieved successes such as artificial enzymes and large protein
cages. Membrane proteins present a considerable design challenge, but here too there have been advances, including
the design of a zinc-transporting tetramer. Lu et al. report the design of stable transmembrane monomers, homodimers,
trimers, and tetramers with up to eight membrane-spanning regions in an oligomer. The designed proteins adopted the

Downloaded from [Link] on March 5, 2018

target oligomerization state and localized to the predicted cellular membranes, and crystal structures of the designed
dimer and tetramer reflected the design models.
Science, this issue p. 1042

ARTICLE TOOLS [Link]

SUPPLEMENTARY [Link]
MATERIALS

REFERENCES This article cites 37 articles, 9 of which you can access for free
[Link]

PERMISSIONS [Link]

Use of this article is subject to the Terms of Service

Science (print ISSN 0036-8075; online ISSN 1095-9203) is published by the American Association for the Advancement of
Science, 1200 New York Avenue NW, Washington, DC 20005. 2017 © The Authors, some rights reserved; exclusive
licensee American Association for the Advancement of Science. No claim to original U.S. Government Works. The title
Science is a registered trademark of AAAS.