self-assembling supramolecular nanostructures constructed...

Self-Assembling Supramolecular Nanostructures Constructed fromde Novo Extender Protein Nanobuilding BlocksNaoya Kobayashi,†,§,⊥ Kouichi Inano,‡ Kenji Sasahara,† Takaaki Sato,‡,¶ Keisuke Miyazawa,#

Takeshi Fukuma,# Michael H Hecht,∇ Chihong Song,○ Kazuyoshi Murata,○

and Ryoichi Arai*,†,§,∥,●,◇

†Department of Applied Biology and ‡Department of Chemistry and Materials, Faculty of Textile Science and Technology,§Department of Bioscience and Textile Technology, Interdisciplinary Graduate School of Science and Technology, ShinshuUniversity, Ueda, Nagano 386-8567, Japan⊥Exploratory Research Center on Life and Living Systems, National Institutes of Natural Sciences, Okazaki, Aichi 444-8787, Japan¶Center for Energy and Environmental Science, Interdisciplinary Cluster for Cutting Edge Research, Shinshu University, Nagano,Nagano 380-8553, Japan#Division of Electrical Engineering and Computer Science, Kanazawa University, Kanazawa, Ishikawa 920-1192, Japan∇Department of Chemistry, Princeton University, Princeton, New Jersey 08544, United States○National Institute for Physiological Sciences, Okazaki, Aichi 444-8585, Japan∥Division of Structural and Synthetic Biology, RIKEN Center for Life Science Technologies, Tsurumi, Yokohama 230-0045, Japan●Institute for Biomedical Sciences, Interdisciplinary Cluster for Cutting Edge Research, Shinshu University, Matsumoto, Nagano390-8621, Japan◇Department of Supramolecular Complexes, Research Center for Fungal and Microbial Dynamism, Shinshu University,Minamiminowa, Nagano 399-4598, Japan

*S Supporting Information

ABSTRACT: The design of novel proteins that self-assembleinto supramolecular complexes is important for developmentin nanobiotechnology and synthetic biology. Recently, wedesigned and created a protein nanobuilding block (PN−Block), WA20-foldon, by fusing an intermolecularly foldeddimeric de novoWA20 protein and a trimeric foldon domain ofT4 phage fibritin (Kobayashi et al., J. Am. Chem. Soc. 2015,137, 11285). WA20-foldon formed several types of self-assembling nanoarchitectures in multiples of 6-mers, includinga barrel-like hexamer and a tetrahedron-like dodecamer. In thisstudy, to construct chain-like polymeric nanostructures, we designed de novo extender protein nanobuilding blocks (ePN−Blocks) by tandemly fusing two de novo binary-patterned WA20 proteins with various linkers. The ePN−Blocks with long helicallinkers or flexible linkers were expressed in soluble fractions of Escherichia coli, and the purified ePN−Blocks were analyzed bynative PAGE, size exclusion chromatography−multiangle light scattering (SEC−MALS), small-angle X-ray scattering (SAXS),and transmission electron microscopy. These results suggest formation of various structural homo-oligomers. Subsequently, wereconstructed hetero-oligomeric complexes from extender and stopper PN−Blocks by denaturation and refolding. The presentSEC−MALS and SAXS analyses show that extender and stopper PN−Block (esPN−Block) heterocomplexes formed differenttypes of extended chain-like conformations depending on their linker types. Moreover, atomic force microscopy imaging in liquidsuggests that the esPN−Block heterocomplexes with metal ions further self-assembled into supramolecular nanostructures onmica surfaces. Taken together, the present data demonstrate that the design and construction of self-assembling PN−Blocksusing de novo proteins is a useful strategy for building polymeric nanoarchitectures of supramolecular protein complexes.

KEYWORDS: de novo protein, nanostructure, protein engineering, protein nanobuilding block, protein-based supramolecular polymers,self-assembly

All organisms contain self-assembling biomolecules includ-ing proteins, nucleic acids, sugars, and lipids. The ability to

design and control such assemblies is a central goal ofbiomolecular engineering, nanobiotechnology, and syntheticbiology. The design and construction of artificial biomacromo-

Received: January 5, 2018Published: April 24, 2018

Research Article

pubs.acs.org/synthbioCite This: ACS Synth. Biol. 2018, 7, 1381−1394

© 2018 American Chemical Society 1381 DOI: 10.1021/acssynbio.8b00007ACS Synth. Biol. 2018, 7, 1381−1394

pubs.acs.org/synthbio

http://pubs.acs.org/action/showCitFormats?doi=10.1021/acssynbio.8b00007

http://dx.doi.org/10.1021/acssynbio.8b00007

lecules that self-assemble into supramolecular complexes areimportant steps toward achieving this goal. Proteins are themost versatile self-assembling biomacromolecules, whichperform complex and functional tasks in all organisms. Proteinfunctions are essentially determined by their three-dimensional(3D) structures, which are characterized into four hierarchicallevels. Complex and refined structures create versatilefunctionalities of proteins. The design of a novel protein andcomplex is in essence an exploration of untracked areas ofamino-acid sequence space. This exploration can be challeng-ing, both because sequence space is vast and because thecontribution of many cooperative and long-range interactionscauses a significant gap between the primary structures andtheir resulting tertiary and quaternary structures. Research intode novo protein designs has progressed toward the constructionof novel proteins,1,2 and has been achieved from (1) rationaland computational designs,3−6 (2) combinatorial methods,7,8

and (3) semirational approaches that include elements ofboth.8−10 As a semirational approach, the binary code strategywas developed to design patterned polypeptide libraries(primary structures) for constructing tertiary structures of denovo proteins. Using secondary structure motifs with binarypatterns of polar and nonpolar residues,9,10 de novo proteinswith α-helixes and/or β-sheets have been successfullycreated.9−12 From a third-generation library of de novo 4-helixbundle proteins with binary patterns,13,14 several de novoproteins with functions in vitro14−16 and in vivo17−22 have beenproduced. We solved the crystal structure of the de novo proteinWA20,23 which is a stable and functional de novo protein fromthe third-generation library.14 WA20 has an unusual dimericstructure with an intermolecularly folded (3D domain-swapped) 4-helix bundle.23 Each WA20 monomer (“nuncha-ku”-like structure) comprises two long α-helices that areintertwined with the helices of another monomer (Figure 1A).

The WA20 structure is stable (melting temperature, Tm, about70 °C) and forms a simple rod-like shape.23 Hence, the stable,simple, and unusual intermolecularly folded structure of the denovo protein WA20 can be applicable to basic framework toolsin nanotechnology and synthetic biology.In these years, several approaches have been developed to

design and construct self-assembling protein complexes using

artificial and fusion proteins as nanoscale building blocks.1,24

These include nanostructures constructed from fusion proteinsthat were designed for symmetric self-assembly,25−31 computa-tionally designed self-assembling protein nanocages with atomiclevel accuracy,6,32−35 and computationally designed β-propellerproteins.36,37

In addition, the design and controlled self-assembly ofproteins into polymeric architectures via supramolecularinteractions offers advantages in understanding the sponta-neously self-assembling process and fabrication of variousbioactive materials.38,39 These protein-based supramolecularpolymers include chemically controlled polymeric self-assemblyof protein nanorings,40 supramolecular hemeprotein polymersthrough heme−heme pocket interaction,41,42 3D domain-swapped oligomers,43−45 self-assembling nanostructures con-structed from designed coiled-coil peptide modules,46−50 thedesign of protein−protein interactions through modification ofresidues on intermolecular interfaces,51,52 and metal-directedself-assembling protein complexes.53−56 In the field of supra-molecular chemistry, the chain length of supramolecularpolymers can be controlled by the addition of a monofunctionalmonomer.57 This chain stopper or end-capper limits thepolymerization or the growth of self-assembling bifunctionalmonomers. These examples of protein-based supramolecularpolymers were also reported: supramolecular protein polymersvia helix−helix association;58 supramolecular nanotube ofchaperonin GroEL with length control using single-ring mutantas end-capper;59 and a series of supramolecular greenfluorescent protein oligomers that were assembled in polygonalgeometries (GFP nanopolygons).60

Recently, we designed and constructed the polyhedralprotein nanobuilding block (PN−Block) WA20-foldon61 byfusing the intermolecularly folded dimeric de novo proteinWA2023 with a trimeric foldon domain of T4 phage fibritin.62

The WA20-foldon formed several distinctive types of self-assembling nanoarchitectures from combinations of dimers andtrimers in multiples of 6-mers (6-, 12-, 18-, and 24-mer),including a barrel-like-shaped hexamer and a tetrahedron-like-shaped dodecamer. The basic objective of the “PN−Blockstrategy” is to create various self-assembling nanostructuresfrom a few types of simple and fundamental PN−Blocks. PN−Blocks comprising intermolecularly folded dimeric de novoproteins (e.g., WA20) as a key component for nano-architectures have the following advantages: (1) The simple,stable, and intertwined rod-like structure of the de novo PN−Block protein makes it easy to use for design and constructionof simple and stable frameworks for nanoarchitectures,61 and(2) PN−Blocks that are based on simple binary patterning of denovo proteins have great potential for redesigning proteindomains that function in vitro14−16 and in vivo.17−22 (“PN” inPN−Block has a different meaning from the polar and nonpolarabbreviations used in the binary code strategy for proteindesign.) Further investigations of new types of PN−Blocks andthe reassembly of various PN−Blocks are essential steps for thedevelopment of the PN−Block strategy.Herein, we designed and created de novo extender protein

nanobuilding blocks (ePN−Blocks) by tandemly fusing two denovo WA20 proteins with various linkers,63,64 and produced anew series of PN−Blocks that can be used to construct self-assembling extended or cyclized chain-like polymeric nano-structures (Figure 1).Moreover, to expand possibilities of the PN−Block strategy,

we reconstructed heterooligomeric complexes by denaturating

Figure 1. Schematics of extender protein nanobuilding blocks (ePN−Blocks). (A) Construction and homooligomeric self-assembly ofePN−Blocks. Ribbon representation and schematics of the inter-molecularly folded dimeric de novo protein WA20 (PDB code 3VJF)23

are shown in red and blue. The ePN−Blocks were constructed bytandemly fusing two de novo WA20 proteins using various linkers.63,64

Helical and flexible linkers are shown as yellow rods and black lines,respectively. (B) Schematics of ePN−Blocks with helical linkers (HL).(C) Schematics of ePN−Blocks with flexible linkers (FL).

ACS Synthetic Biology Research Article

DOI: 10.1021/acssynbio.8b00007ACS Synth. Biol. 2018, 7, 1381−1394

1382


and refolding extender and stopper PN−Blocks. In addition, wedemonstrate that the complexes can further self-assemble intosupramolecular nanostructures with metal ions.

■ RESULTS AND DISCUSSIONDesign of de Novo Extender PN−Blocks (ePN−Blocks)

with Various Linkers To Construct Self-AssemblingChain-like Nanostructures. To construct oligomeric chain-like extended nanostructures, we designed de novo ePN−Blocksby tandemly fusing two de novo WA20 proteins with linkers ofvarious type and length63,64 as shown in Figure 1 and Figure S1(Supporting Information). The two WA20 domains were fusedwith helical linkers (HL: (EAAAK)n; n = 2−5; Figure 1B) orflexible linkers (FL: (GGGGS)n; n = 3, 4; Figure 1C). Thehelical linkers were derived from stable helix-forming de novodesigned peptides described by Marqusee and Baldwin,65 andwe previously demonstrated by using fluorescence resonanceenergy transfer and small-angle X-ray scattering (SAXS)techniques that the helical linkers can separate two domainsof a fusion protein at controlled distances.63,64

Because WA20 forms a stable intermolecularly foldeddimeric structure,23 ePN−Blocks were expected to self-assemble into cyclized chain-like homooligomers (Figure 1A),thereby giving rise to various nanostructures that depend on thedifferent types and lengths of linkers. This notable series ofPN−Blocks are completely “de novo proteins,” which have nosequences derived from any natural proteins. The de novoePN−Block proteins were designed by tandemly linking thetwo de novo WA20 proteins, which were created from scratchusing the binary code strategy,9,10 with artificial peptide linkersequences.63

Self-Assembling Homooligomers of ePN−Blocks. Thepresent ePN−Block proteins with various linkers wereexpressed in Escherichia coli, and soluble and insoluble fractionsof cells were prepared by centrifugation after disruption bysonication. SDS-PAGE analysis of the resulting fractionsshowed that ePN−Block proteins with long helical linkers(HL4, HL5) and flexible linkers (FL3, FL4) were expressedmainly in soluble fractions from E. coli lysates (Figure 2A).However, ePN−Block proteins with short helical linkers (HL2,HL3) were expressed mainly in insoluble fractions. Afterpurification of soluble ePN−Block proteins using immobilizedmetal ion affinity chromatography (IMAC), SDS-PAGEshowed a predominating single band at every ePN−Blocklane (Figure 2B), whereas native PAGE showed migrationladders of ePN−Block proteins with the long helical linkers(HL4, HL5) and flexible linkers (FL3, FL4; Figure 2C). Incontrast, ePN−Blocks with short helical linkers (HL2, HL3)migrated as a few bands in native PAGE (Figure 2C). Theseresults suggest that ePN−Blocks with long helical linkers (HL4and HL5) and flexible linkers (FL3 and FL4) form severalstable homooligomeric states in soluble fractions from E. colilysates. However, PN−Blocks with short helical linkers (HL2and HL3) mainly precipitated in the insoluble fraction andformed only a few limited oligomeric states in the solublefraction possibly because characteristics of short and rigidhelical linkers (HL2 and HL3) may cause steric hindrances andlead to insoluble forms of ePN−Block homooligomers (FigureS2).Because insoluble samples are difficult to analyze using

standard techniques for protein solutions, we performed furtherexperiments on soluble samples of ePN−Blocks with HL4 andFL4, and compared these typical samples of helical and flexible

linkers at the same linker length. After purification by IMACfollowed by size exclusion chromatography (SEC) (FiguresS3−S6), the fractionated samples of ePN−Block (HL4) andePN−Block (FL4) homooligomers (Table S1) were analyzedby SEC-multiangle light scattering (SEC−MALS) and SAXS asdescribed in more detail in the Supporting Information. Inbrief, the results of SEC−MALS are summarized in Table 1,Tables S2 and S3, and Figures S7 and S8, and the results ofSAXS are summarized in Table S4 (Supporting Information),and Figure 3 and Figures S9 and S10, suggesting that ePN−Blocks form homoligomers from dimer to pentamer at least.The discussion is carried forward in the following part.Moreover, the SEC-fractioned samples of ePN−Block (HL4)

homooligomers were observed by transmission electronmicroscopy (TEM) with negatively staining (Figure 4 andFigures S11−S14). Oligomeric states of the samples werepreviously analyzed by SEC−MALS and native PAGE (FiguresS7 and S11, Supporting Information). TEM images (Figure4A−D) show that the relatively larger size samples (TEMsamples 61, 63, 65, and 67 including tetramer (e4), pentamer(e5), and higher oligomers) contained various sphericalparticles approximately 15−20 nm in diameter, roughlyconsistent with SAXS analyses. Numbers and sizes ofobservable particles gradually decreased with increasing samplenumbers (corresponding to elution volume of SEC) andparticles were rarely found in TEM sample 71 mainly includingtrimer (e3) (Figure 4E), suggesting that tetramer (e4) may bethe observable limit, and trimer (e3) and dimer (e2) cannot beobserved by TEM due to low molecular mass and small size. In

Figure 2. Polyacrylamide gel electrophoresis (PAGE) of ePN−Blockswith various linkers. (A) SDS-PAGE (17.5% polyacrylamide gel) ofePN−Blocks expressed in E. coli. M, protein molecular weight marker,broad (Takara Bio, Otsu, Japan); S, supernatant (soluble fraction); P,pellet (insoluble fraction); +, addition of 0.2 mM IPTG; −, noaddition of IPTG. (B) SDS-PAGE (17.5% polyacrylamide gel) ofePN−Blocks and (C) native PAGE (5.0% polyacrylamide gel) ofePN−Blocks after immobilized metal ion affinity chromatography(IMAC) purification. W, stopper PN−Block (WA20).



1383

http://pubs.acs.org/doi/suppl/10.1021/acssynbio.8b00007/suppl_file/sb8b00007_si_001.pdf















contrast, the TEM image of sample 71 with the addition ofnickel ions shows many large aggregations as nanoscale spottedpatterns (Figure 4F), suggesting further supramolecularassembly of ePN−Block (HL4) homooligomers induced bynickel ions. The addition of nickel ions can promoteintermolecular interactions between WA20 domains, assuggested by dynamic light scattering (DLS) experiments ofWA20 proteins with and without nickel ions (Figure S15). The

DLS data indicate that the size distribution significantly shifts tolarger size with the addition of nickel ions probably because ofthe intermolecular assembly induced by the nickel ioncoordination with several clusters of histidine residues on thesurface of the WA20 structure (PDB code 3VJF)23 (Figure S16,Supporting Information).Negatively stained ePN−Block (FL4) homooligomer

samples were also observed by TEM (Figures S17−S19,Supporting Information). TEM images show various shapenanoparticles and aggregations. Numbers and sizes ofobservable particles gradually decreased with increasing samplenumbers corresponding to elution volume of SEC, as describedin more detail in the Supporting Information.

Reconstruction of Heterooligomeric Extender andStopper PN−Block Complexes by Denaturation andRefolding. To expand the possibilities of the PN−Blockstrategy, we reconstructed multicomponent PN−Block com-plexes from extender PN−Block (ePN−Block) and stopperPN−Block (sPN−Block, i.e., WA20 protein) by denaturationand refolding. In analogy with supramolecular chemistry,57 thechain length of supramolecular polymers can be tuned by theaddition of a monofunctional monomer (sPN−Block). Thechain stopper can be expected to limit the growth of self-assembling bifunctional monomers (ePN−Blocks) and lead toaltering conformation and reduce polydispersity. Figure 5Ashows native PAGE analysis following reconstruction of ePN−Block (HL4) and sPN−Block (WA20) proteins. Beforedenaturation, band patterns of ePN−Block (HL4) did notchange in just mixed samples with sPN−Block. In contrast,after denaturation and refolding, new pattern bands (Figure 4A,black arrowheads) and diminished bands (Figure 5A, grayarrowheads) appeared with increasing sPN−Block contents,suggesting the formation of several heterooligomeric complexesof ePN−Blocks and sPN−Blocks (esPN−Block heterocom-plexes). After denaturation and refolding of ePN−Block (FL4)and sPN−Block proteins, a stronger band and several weakerbands (Figure 5B, black and gray arrowheads, respectively)were seen with increasing sPN−Block contents, suggestingformation of heterooligomeric esPN−Block complexes fromePN−Block (FL4) and sPN−Block proteins. These results

Table 1. Summary of SEC−MALS Analyses

PN−Block oligomers sample and peak molecular mass [kDa] ePN−Block: sPN−Block native PAGE band (Figure S6 or Figure 6)

ePN−Block (HL4) homooligomer Sample (E), Peak 2 48.8 2:0 (e2) Band 1ePN−Block (HL4) homooligomer Sample (D), Peak 1 48.6 2:0 (e2) Band 2ePN−Block (HL4) homooligomer Sample (C), Peak 2 73.5 3:0 (e3) Band 3ePN−Block (HL4) homooligomer Sample (B), Peak 1 98.8 4:0 (e4) Band 4ePN−Block (HL4) homooligomer Sample (A), Peak 1 129 5:0 (e5) Band 5ePN−Block (FL4) homooligomer Sample (E), Peak 2 48.3 2:0 (e2) Band 1ePN−Block (FL4) homooligomer Sample (D), Peak 1 47.6 2:0 (e2) Band 2ePN−Block (FL4) homooligomer Sample (C), Peak 1 69.6 3:0 (e3) Band 3ePN−Block (FL4) homooligomer Sample (B), Peak 1 104 4:0 (e4) Band 4ePN−Block (FL4) homooligomer Sample (A), Peak 1 135 5:0 (e5) Band 5esPN−Block (HL4) heterocomplex Sample (d), Peak 1 49.5 1:2 (e1s2) Band 1esPN−Block (HL4) heterocomplex Sample (b), Peak 1 83.4 2:2 (e2s2) Band 2esPN−Block (HL4) heterocomplex Sample (b), Peak 2 112 3:2 (e3s2) Band 3esPN−Block (HL4) heterocomplex Sample (b), Peak 3 149 4:2 (e4s2) Band 4esPN−Block (FL4) heterocomplex Sample (d), Peak 1 46.8 1:2 (e1s2) Band 1esPN−Block (FL4) heterocomplex Sample (c), Peak 1 75.8 2:2 (e2s2) Band 2esPN−Block (FL4) heterocomplex Sample (b), Peak 1 106 3:2 (e3s2) Band 3esPN−Block (FL4) heterocomplex Sample (b), Peak 2 132 4:2 (e4s2) Band 4sPN−Block (WA20) Peak 1 22.5 0:2 (s2)

Figure 3. Small-angle X-ray scattering (SAXS) analyses of the ePN−Block homooligomer samples. Concentration-normalized absolutescattering intensities I(q)/c of ePN−Block (HL4) [e(HL4)]homooligomer samples (A) and ePN−Block (FL4) [e(FL4)]homooligomer samples (B). Chicken egg lysozyme was used as amolecular mass reference standard. Their real-space information,concentration-normalized pair-distance distribution functions p(r)/c ofePN−Block (HL4) homooligomer samples (C) and ePN−Block(FL4) homooligomer samples (D), which were calculated usingindirect Fourier transformation (IFT).



1384







imply that several esPN−Block complexes of extended openchain-like heterooligomers were reconstructed after denatura-tion and refolding of multicomponent ePN−Blocks and sPN−Blocks (Figure 5C).Oligomeric State Analyses of esPN−Block Hetero-

complexes. To investigate the oligomeric states of esPN−Block heterocomplexes, we fractionated esPN−Blocks beforeand after reconstruction using SEC on a Superdex 200 increase10/300 GL column (Figures S20−S23). Before reconstruction(Figures S20 and S22), SEC chromatograms and SDS-PAGEanalyses indicated that ePN−Block (HL4 or FL4) and sPN−Block (WA20) proteins had not formed hetero-oligomericcomplexes because the sPN−Block was eluted in only lowmolecular size fractions that corresponded with those of theWA20 homodimer. In contrast, reconstruction by denaturationand refolding (Figures S21 and S23, Supporting Information)clearly led to the formation of hetero-oligomeric esPN−Blockcomplexes from ePN−Block (HL4 or FL4) and sPN−Block(WA20) because sPN−Block was eluted with ePN−Block

(HL4 or FL4) in a wide range of fractions across highermolecular size complexes.In subsequent SAXS analyses, samples were derived from

several SEC fractions of esPN−Block heterocomplexes (Figure6, Figure S24, and Table S5). Some samples were selected forSEC−MALS analysis and the results are summarized in Table 1and Figures S25 and S26 (Supporting Information). Molecularmasses of esPN−Block heterocomplexes reveal the presence ofseveral species of the esPN−Block heterocomplexes containingone ePN−Block and two sPN−Blocks (e1s2), two ePN−Blocks and two sPN−Blocks (e2s2), three ePN−Blocks andtwo sPN−Blocks (e3s2), and four ePN−Blocks and two sPN−Blocks (e4s2), as indicated in native PAGE analyses (Figure 6,bands 1, 2, 3, and 4, respectively).Figure 7A and B show SAXS intensities of esPN−Block

heterocomplex and sPN−Block (WA20) samples with chickenegg lysozyme as a molecular mass reference standard (Mw, 14.3kDa). Assuming that these proteins have practically identicalscattering length densities and specific volumes, and thestructure factor S(q) ≈ 1 for screened electrostatic repulsion

Figure 4. TEM imaging of SEC-fractioned and negatively stainedePN−Block (HL4) homooligomers. (A) TEM image of ePN−Block(HL4) sample 61 mainly including tetramer, pentamer (major), andhigher oligomers. (B) TEM image of ePN−Block (HL4) sample 63mainly including tetramer and pentamer. (C) TEM image of ePN−Block (HL4) sample 65 mainly including tetramer (major) andpentamer. (D) TEM image of ePN−Block (HL4) sample 67 mainlyincluding tetramer (major) and trimer. (E) TEM image of ePN−Block(HL4) sample 71 mainly including trimer. (F) TEM image of ePN−Block (HL4) sample 71 with the addition of 5 μM NiCl2. These full-size images are shown in Figure S12. Oligomeric states of thesesamples were analyzed by SEC−MALS and native PAGE (Figures S7and S11).

Figure 5. Reconstruction of heterooligomeric complexes frommulticomponent extender and stopper PN−Blocks by denaturationand refolding. Native PAGE (7.5% polyacrylamide gel) analysis of thereconstruction of extender PN−Block (ePN-block) and stopper PN−Block (sPN−Block): (A) ePN-block (HL4) and (B) ePN-block(FL4). In the left half, samples were just mixed (Mixture), and in theright half samples were denatured and refolded after mixing(Reconstruction). The sPN-block (WA20) was added at stepwiseincreases in the ratio of sPN−Block/ePN−Block of 1, 2, 4, and 8. (C)Schematics of the reconstruction process from ePN−Blocks (red andblue) and sPN−Blocks (light gray) to esPN−Block heterocomplexes.



1385











in dilute samples, the forward scattering intensity normalized byprotein concentration I(q → 0)/c is proportional to the weight-average molecular mass (Mw). Mw of these samples (Table 2) isroughly consistent with SEC−MALS analysis, when samplepurity of complex components (Figure 6 and Table 2) andexperimental errors are considered.Shape Analyses of esPN−Block Heterocomplexes. To

calculate intuitive real-space data from SAXS analyses, pair-distance distribution functions p(r) were determined usingindirect Fourier transformation,66−68 which were reflected byshapes and components of the esPN−Block heterocomplexsamples (Figure 7C,D), while the complex samples had somepolydispersity derived from equilibria of cyclized/extendedchain forms in the reconstruction process. The p(r) series ofthe esPN−Block (HL4) heterocomplex samples (Figure 7C)was characterized as an extended tail in the high-r regime,suggesting that the esPN−Block (HL4) heterocomplexes formextended shapes. In contrast, the p(r) series of the esPN−Block(FL4) heterocomplexes (Figure 7D) had shorter Dmax valuesthan those of the esPN−Block (HL4) heterocomplexes,implying that the esPN−Block (FL4) heterocomplexes formrelatively more compact shapes than the esPN−Block (HL4)heterocomplexes. In comparison of p(r) of ePN−Blockhomooligomers with those of esPN−Block heterocomplexes(Figure S27), the Dmax values of the esPN−Block (HL4)

heterocomplex samples tend to be longer than those of theePN−Block (HL4) homooligomer samples while the sampleshad polydispersity. In addition, the p(r) series of ePN−Block(FL4) homoligomers and esPN−Block (FL4) heterocomplexesshows relatively smaller differences in shapes possibly due toflexible characteristics of the flexible linker (FL4).Kratky plots and dimensionless Kratky plots of esPN−Block

heterocomplex samples are shown in Figure S28 (SupportingInformation) and Figure 7E,F. These plots are usefulrepresentations of the SAXS intensity to quickly assess theglobular nature of a polypeptide chain without any modeling.69

In dimensionless Kratky plots normalized by I(0) and Rg, the

Figure 6. Native PAGE analysis of the esPN−Block heterocomplexsamples (Table S5) separated by size exclusion chromatography(SEC). (A) Native PAGE (7.5% polyacrylamide gel) of esPN−Block(HL4) heterocomplexes; Q, mixed sample of ePN−Block and sPN−Block before reconstruction and SEC separation; R, esPN−Blockheterocomplex sample after reconstruction before SEC separation; W,sPN−Block (WA20). (B) Schematics of the esPN−Block (HL4)heterocomplexes. (C) native PAGE (7.5% polyacrylamide gel) ofesPN−Block (FL4) heterocomplexes. (D) schematics of esPN−Block(FL4) heterocomplexes. The same colors are used in Figures 1 and 4.Theoretical molecular masses of esPN−Block heterocomplexes arepresented in parentheses. SDS-PAGE analyses of these samples arealso shown in Figure S24.

Figure 7. SAXS analyses of the esPN−Block heterocomplex samples.Concentration-normalized absolute scattering intensities I(q)/c ofesPN−Block (HL4) heterocomplex [es(HL4)] samples (A) andesPN−Block (FL4) [es(FL4)] heterocomplex samples (B). WA20 wasused as a control and chicken egg lysozyme was used as a molecularmass reference standard. The real-space functions, concentration-normalized pair-distance distribution functions p(r)/c of the esPN−Block (HL4) heterocomplex samples (C) and the esPN−Block (FL4)heterocomplex samples (D). Dimensionless Kratky plots of the esPN−Block (HL4) heterocomplex samples (E) and the esPN−Block (FL4)heterocomplex samples (F) (0 < q < 3 nm−1).



1386






information about size of the protein is removed but theinformation about the shape and flexibility is kept.70 Thedimensionless Kratky plot of lysozyme (a reference sample)exhibits a bell-shape with a well-defined peak at a maximumvalue of 1.104 for qRg = √3, indicating a typical globularprotein shape. In the case of the dimensionless Kratky plots ofthe esPN−Block (HL4) heterocomplexes (Figure 7E), thepeaks of the bell-shapes tend to shift to the upper right in largercomplexes, suggesting that the larger complexes have the moreextended shapes as simulated in Figure S29A. In addition, thedimensionless Kratky plots of the esPN−Block (FL4)heterocomplex samples have a gentler hill-like broad peak

than those of esPN−Block (HL4) heterocomplex samples,implying that the esPN−Block (FL4) heterocomplexes hadelongated shapes and more flexible conformation of multi-domains, such as flexible and extended conformation proteinswith some domains tethered by linkers69 (e.g., calmodulin,71

Filamin C 23−24,72 p47phox73 and p67phox74) (Figure S29A andB), than the esPN−Block (HL4) heterocomplexes because ofdistinctive features of the different linkers. However, thepolydispersity of the samples may be not negligible andcomplicate the interpretation of these results.Moreover, the dimensionless Kratky plots of comparable

samples of the ePN−Block homooligomer sample (C) and

Table 2. Summary of SAXS Analyses of esPN−Block Heterocomplexes

esPN−Block heterocomplex samples I(q→0)/c [cm−1mg−1mL] Rg [nm] Dmax [nm] Mw [kDa] components of esPN−Block heterocomplexesa

es(HL4) sample (a) 0.133 11.0 52 219 e3s2, e4s2, higheres(HL4) sample (b) 0.0760 8.6 40 125 e2s2, e3s2, e4s2, higheres(HL4) sample (c) 0.0619 6.9 29 102 e2s2, e3s2es(HL4) sample (d) 0.0359 4.7 20 59.0 e1s2, e2s2es(HL4) sample (e) 0.0353 3.6 15 58.1 e1s2es(FL4) sample (a) 0.102 8.8 35 167 e4s2, higheres(FL4) sample (b) 0.0712 6.4 27 117 e3s2, e4s2, higheres(FL4) sample (c) 0.0401 4.8 20 65.9 e2s2, e3s2es(FL4) sample (d) 0.0310 3.6 15 51.0 e1s2, e2s2es(FL4) sample (e) 0.0248 3.3 12 40.8 e1s2sPN−Block (WA20) 0.0152 2.6 10 25.0 s2Lysozyme 0.00870 1.5 4.5 14.3

aMain components are indicated by boldface. These components were judged mainly by the SEC−MALS and native PAGE results.

Figure 8. Three-dimensional structural modeling of the esPN−Block (HL4) heterocomplex based on SAXS analysis. (A) A rigid-body modelstructure of one extender and two stoppers (e1s2) of the esPN−Block (HL4) heterocomplex [e1s2(HL4)]. The model is shown as a ribbonrepresentation in the translucent surface representation. The first WA20 domain (red), the helical linker (yellow), and the second WA20 domain(blue) of ePN−Block; sPN−Block (light gray). (B) The concentration-normalized scattering intensity of the esPN−Block (HL4) heterocomplexsample (e) obtained by the SAXS experiment (open circle) and fitting of I(q) simulated from the rigid-body model of e1s2 (HL4) (red line). (C)The pair-distance distribution function p(r) of the esPN−Block (HL4) heterocomplex sample (e) calculated from the SAXS data (black dash line)and p(r) simulated from the rigid-body model of e1s2 (HL4) (red line). (D) A dummy atom model of e1s2(HL4) reconstructed from the SAXSdata using ab initio modeling programs DAMMIF, DAMAVER, and DAMMIN. (E) The concentration-normalized scattering intensity of the esPN−Block (HL4) heterocomplex sample (e) (black open circle) and fitting of I(q) simulated from the DAMMIN model of e1s2(HL4) (red line). (F)Superimposition of the rigid-body model (magenta ribbon representation) and the dummy atom model (green) of e1s2 (HL4). Blue dots representan averaged model from 10 structural models calculated by DAMMIF and DAMAVER (Figure S31).



1387






esPN−Block heterocomplex sample (c) mainly containinghomooligomers (e3, e4) or heterocomplexes (e2s2, e3e4) areshown in Figure S30. The plots of the ePN−Block (FL4 andHL4) homooligomer (C) and esPN−Block (FL4) hetero-complex (c) have a gentle hill-like broad peak, suggesting theirflexible and dynamic conformation in contrast to a globularprotein (e.g., lysozyme) and the esPN−Block (HL4)heterocomplex samples. By contrast, the esPN−Block (HL4)heterocomplex samples have a distinctive feature of a big bell-shape peak, suggesting that they are well-folded proteins withlarge Rg (i.e., elongated conformations with some rigidity) assimulated in Figure S29A. Also noteworthy is that the plots ofthe ePN−Block (HL4) homooligomer (C) and the esPN−Block (HL4) heterocomplex sample (c) represent significantlydifferent features, implying that topology and shape weretransformed from cyclized chain-like closed forms of the ePN−Block homoligomers into extended chain-like open forms of theheterooligomeric esPN−Block complexes by the reconstructionprocess as shown in the schematics (Figure 5C).Since the esPN−Block (HL4) heterocomplex sample (e)

contained predominantly the e1s2 (HL4) complex in nativePAGE analysis (Figure 6A), for further structural analysis, arigid-body model structure of e1s2 (HL4) was constructedbased on the crystal structure of the WA20 dimer (PDB code3VJF)23 to explain the experimental p(r) with consideration ofthe helical linker rigidity. The model of e1s2 (HL4) shows an

extended “Z” shape (Figure 8A), and the simulated I(q) andp(r) from the rigid-body model closely resembles that from theSAXS experiment (Figure 8B,C).Moreover, the low-resolution dummy atom models of e1s2

(HL4) were constructed from the SAXS data using ab initioshape modeling programs DAMMIF,75 DAMAVER,76 andDAMMIN77 (Figure 8D,E and Figure S31). Ten timesindependent calculations of ab initio shape determination byDAMMIF reproducibly generated typical elongated shapemodels (Figure S31A). The refined DAMMIN model of e1s2(HL4) is superimposed on the rigid body model of e1s2 (HL4)as shown in Figure 8F. This Z-like shape conformation of e1s2(HL4) resembles the conformation of chimeric proteins ofgreen fluorescent protein variants diagonally linked by thehelical linkers previously analyzed by SAXS.64

In addition, we also tried to analyze shapes of the esPN−Block heterocomplex (HL4) sample (c) mainly containing e2s2and e3s2 complexes judged from the native PAGE and SEC−MALS data (Figure 6 and Table 1 and Figure S25). On thebasis of the p(r) profiles and Dmax values (Figure 7C), wespeculate that the possible structural models of e2s2 (HL4) ande3s2 (HL4) may be constructed by a chain-like extension ofe1s2 (HL4) model as a basic structural unit, as shown in theschematics in Figure 6B and Figure S29A. However, it isdifficult to construct reliable models of e2s2 and e3s2 from theSAXS data due to polydispersity of the samples.

Figure 9. Three-dimensional structural modeling of the esPN−Block (FL4) heterocomplex based on SAXS analysis. (A) Rigid-body modelstructures of one extender and two stoppers (e1s2) of the esPN−Block (FL4) heterocomplex [e1s2 (FL4)]. The models are shown as ribbonrepresentation in translucent surface representation. The first WA20 domain (red), the flexible linker (black), and the second WA20 domain (blue)of ePN−Block; sPN−Block (light gray). (B) The concentration-normalized scattering intensity of the esPN−Block (FL4) heterocomplex sample (e)obtained by the SAXS experiment (open circle) and fitting of I(q) simulated from the two rigid-body model structures of e1s2 (C form:V form =1:1) (red line). (C) The pair-distance distribution function p(r) of the esPN−Block (FL4) heterocomplex sample (e) calculated from the SAXS data(black dash line) and p(r) simulated from the two rigid-body model structures of e1s2 (C form:V form = 1:1) (red line). (D) A dummy atom modelof e1s2 (FL4) reconstructed from the SAXS data using ab initio modeling programs, DAMMIF, DAMAVER, and DAMMIN. (E) The concentration-normalized scattering intensity of the esPN−Block (FL4) heterocomplex sample (e) (black open circle) and fitting of I(q) simulated from theDAMMIN model of e1s2 (FL4) (red line). (F) Superimposition of the rigid-body models (ribbon representations of yellow C form and magenta Vform) and the dummy atom model (green) of e1s2 (FL4). Blue dots represents an averaged model from 10 structural models calculated byDAMMIF and DAMAVER (Figure S34).



1388









Rigid-body model structures of the esPN−Block hetero-complex e1s2 (FL4) were also constructed based on the WA20crystal structure. The “V” form of a rigid-body model for e1s2(FL4) was built in terms of the Dmax value (∼12 nm) (Figure9A). The p(r) value that was calculated from the V form modelonly poorly resembled that from the SAXS experiment (FigureS32, Supporting Information). Thus, an additional rigid-bodymodel of a compact form (C form) was constructed withconsideration of linker flexibility and domain interactions.Subsequently, the I(q) and p(r) of the esPN−Block (FL4)heterocomplex sample (e), which comprised mainly the e1s1(FL4) heterocomplex, was simulated well with the compositeI(q) and p(r) of an equal ratio of V and C form models (Figure9B,C). These observations imply that dynamic structures ofe1s2 (FL4) represent structural ensembles including the Vform, the C form, and many transient intermediates. Therefore,we tried multistate modeling with SAXS profiles usingMultiFoXS server78 as described in the Supporting Information.The summarized results shown in Figure S33 also suggest thatthe e1s2 (FL4) heterocomplex are conformationally heteroge-neous and dynamic. In addition, the low-resolution dummyatom models of e1s2 (FL4) were constructed from the SAXSdata using the ab initio shape modeling programs (Figure 9D,Eand Figure S34). Ten times calculations of ab initio shapedetermination by DAMMIF generated various structuralmodels (Figure S34A), also suggesting that the e1s2 (FL4)conformation is more heterogeneous and dynamic than e1s2(HL4). The rigid-body models (V and C forms) of e1s2 (FL4)are roughly superimposed on the DAMMIN model of e1s2(FL4) as shown in Figure 9F.Atomic Force Microscopy (AFM) Observations of Self-

Assembling Supramolecular Nanostructures of esPN−Block Complexes. To expand the potential of PN−Blockstrategy for nanotechnology, esPN−Block heterocomplexesfurther self-assembled into supramolecular nanostructures withmetal ions (Ni2+). As suggested by the TEM experiment(Figure 4E,F) and the DLS experiment (Figure S15), theaddition of nickel ions can promote intermolecular interactionsbetween the WA20 domains of PN−Blocks. The esPN−Blockheterocomplexes with nickel ions were observed in liquid usingfrequency modulation (FM)-atomic force microscopy (AFM)(Figure 10 and Figures S35−S37). The resulting image (Figure10A) shows self-assembling supramolecular nanostructures ofthe esPN−Block (HL4) heterocomplex fraction 20 (FigureS21, Supporting Information), comprising mainly e2s2 (HL4)and e3s2 (HL4) heterocomplexes. Moreover, many bundles ofrodlike structures of 10.9 ± 1.8 nm in length and 3.4 ± 1.8 nmin width were found in the AFM image (Table S6, SupportingInformation). The sizes of the rodlike structural domains areconsistent with the size of ∼10 nm in length and ∼3 nm inwidth estimated from the crystal structure (PDB code 3VJF)23

of the WA20 dimer (Figure S16, Supporting Information).Several domains of WA20 units seem to be aligned laterally inthe direction of the longitudinal axis (Figure S35A and B).In contrast, Figure 10B shows an AFM image of self-

assembling nanostructures of the esPN−Block (FL4) hetero-complex fraction 19 comprising e2s2 (FL4), e3s2 (FL4), ande4s2 (FL4) (Figure S23). The stripe features in Figure 10Bcould be AFM scan artifacts as they are mostly oriented in theslow scan direction. However, similar nanostructures werereproducibly observed in the images obtained with differentscan sizes, and more magnified images indicate that theirorientations are not necessarily the same (Figure S36). Thus,

these strip features are not due to the scan artifacts butrepresent the true surface structures. A number of bundles ofrodlike structures with lengths of 10.2 ± 2.0 nm and widths of3.2 ± 0.4 nm (Table S6) represented the WA20 structuraldomains, and these rodlike structural domains seem to line upand extend in the direction of the lateral axis (Figure S35C andD). These contrasting observations may reflect differentstructural properties of esPN−Block complexes, includingdifferences in rigidity and flexibility of helical (HL4) andflexible (FL4) linkers.

■ CONCLUSIONSWe designed and created fully de novo ePN−Block proteins bytandemly fusing the two de novo WA20 proteins with variouslinkers, as a new series of PN−Blocks. Analyses of ePN−Blocksby native PAGE, TEM, SEC−MALS, and SAXS indicate theformation of several homooligomeric states. Then, wereconstructed heterooligomeric esPN−Block complexes bydenaturating and refolding ePN−Blocks and sPN−Blocks.The series of comprehensive SEC−MALS and SAXS analysessuggested that the cyclized chain-like ePN−Block homoo-ligomers were transformed into different types of extendedchain-like structures of esPN−Block heterocomplexes depend-ing on their linker type, demonstrating great potential ofreconstructible PN−Blocks as artificial nanobuilding blockmolecules. Moreover, AFM observations revealed that theesPN−Block heterocomplexes further self-assembled intohigher-order supramolecular nanostructures with metal ions.These results demonstrate that the PN−Block strategy is auseful and systematic strategy for constructing novel nano-architectures of de novo protein-based supramolecular polymercomplexes for potential nanobiomaterials in biotechnology andsynthetic biology.

■ METHODSConstruction of Protein Expression Plasmids. A DNA

fragment encoding the de novo protein WA20 was preparedfrom the plasmid pET3a-WA2023 using polymerase chainreactions (PCR) with KOD-Plus-Neo DNA polymerase

Figure 10. Atomic force microscopy (AFM) imaging of self-assembling supramolecular nanostructures of the esPN−Blockheterocomplexes with metal ions (Ni2+) on mica surfaces in liquid.(A) AFM image of the esPN−Block (HL4) heterocomplex fraction 20(Figure S21, Supporting Information) comprising mainly e2s2 (HL4)and e3s2 (HL4). (B) AFM image of the esPN−Block (FL4)heterocomplex fraction 19 (Figure S23) comprising e2s2 (FL4),e3s2 (FL4), and e4s2 (FL4). The additional images obtained withdifferent scan sizes and more magnified images are shown in FigureS36.



1389























(Toyobo, Osaka, Japan) and primers for the T7 promoterprimer and WA20RV_HindIII (Table S7, Supporting Informa-tion). The amplified fragment was digested using NdeI andHindIII and was cloned into pET32/EBFP-HL5-EGFP63

between the NdeI and HindIII sites to construct the plasmidpET-WA20-HL5-GFP, in which the Trx tag and the EBFP genewere replaced with the WA20 gene. Another DNA fragmentencoding WA20 was prepared from the plasmid pET3a-WA20using PCR with primers for WA20FW_NotI and WA20RV_X-hoI (Table S7). The amplified fragment was then digestedusing NotI and XhoI and was cloned into pET-WA20-HL5-EGFP between the NotI and XhoI sites to give the expressionplasmid pET-WA20-HL5-WA20 for an extender PN−Block(ePN−Block). DNA fragments encoding the other linker genes(HL2, HL3, HL4, FL3, and FL4) were prepared by digestion ofthe plasmid pET32/EBFP-linker-EGFP63 with HindIII andNotI and cloning of fragments into pET-WA20-HL5-WA20between the restriction sites to give the plasmids pET-WA20-HL2-WA20, pET-WA20-HL3-WA20, pET-WA20-HL4-WA20,pET-WA20-FL3-WA20, and pET-WA20-FL4-WA20, respec-tively. The amino acid sequences of the ePN−Block proteinsare shown in Figure S1.Protein Expression and Purification. All ePN−Block

proteins were expressed in E. coli BL21 Star(DE3) (Invitrogen,Carlsbad, CA) harboring pET-WA20-Linker-WA20 in 2 L ofLB broth, Lennox (Nacalai Tesque, Kyoto, Japan) containing50 μg/mL ampicillin sodium salt at 37 °C. Protein expressionwas induced using 0.2 mM β-D-1-thiogalactopyranoside at anoptical density OD600 of about 0.8 at 600 nm, and cells werefurther cultured for 3−4 h at 37 °C. Proteins were extractedfrom harvested cells by sonication in lysis buffer containing 50mM sodium phosphate buffer (pH 7.0), 300 mM NaCl, and10% glycerol. Proteins were then purified using IMAC withTALON metal affinity resin (Clontech, Takara Bio, MountainView, CA) according to the manufacturer’s protocols. Theequilibration/wash buffer contained 50 mM sodium phosphatebuffer (pH 7.0) and 300 mM NaCl, and the elution buffercontained 50 mM sodium phosphate buffer (pH 7.0), 300 mMNaCl, 10% glycerol, and 250 mM imidazole. Since manyhistidine residues are exposed on the surface of the WA20structure (Figure S16),23 WA20 and ePN−Block proteins canbind TALON metal affinity resin even without a His-tag.Protein expression and IMAC purification of the stopper PN−Block (sPN−Block; WA20) were performed as describedpreviously.23

Sodium dodecyl sulfate polyacrylamide gel electrophoresis(SDS-PAGE) analyses were performed according to thestandard Laemmli procedure. The stacking gel contained4.5% polyacrylamide, 125 mM Tris−HCl (pH 6.8), and 0.1%SDS, and the separating gel comprised 17.5% polyacrylamide,125 mM Tris−HCl (pH 8.8). Proteins were electrophoresed inrunning buffer containing 25 mM Tris, 192 mM glycine, and0.1% SDS. For native PAGE analyses, stacking gels, separatinggels (5% or 7.5% polyacrylamide), and running buffer wereprepared as for SDS-PAGE, but with no SDS. Proteins in thegels were stained with Coomassie brilliant blue (CBB) R-250.Denaturation, Refolding, and Further Purification of

PN−Block Proteins. The ePN−Block (HL4 or FL4) proteinwas mixed with the sPN−Block protein (WA20), and they weredenatured in 6 M guanidine hydrochloride (GdnHCl) for 3 h at25 °C in 20 mM HEPES buffer (pH 7.5) containing 100 mMNaCl and 10% glycerol. For refolding, denatured proteins weredialyzed three times for about 4 h against 20 mM HEPES buffer

(pH 7.5) containing 100 mM NaCl, 10% glycerol, and 200 mML-arginine hydrochloride (ArgHCl) using a Bio-Tech oscillatorymicrodialysis system (BM Equipment, Tokyo, Japan). Con-centrated samples of refolded ePN−Block and sPN−Blockheterocomplexes (esPN−Blocks) were separated using sizeexclusion chromatography (SEC), and were eluted in 20 mMHEPES buffer (pH 7.5) containing 100 mM NaCl, 10%glycerol, and 200 mM ArgHCl from a Superdex 200 Increase10/300 GL column (GE healthcare, Little Chalfont Buck-inghamshire, UK).

Size Exclusion Chromatography−Multiangle LightScattering (SEC−MALS). SEC−MALS experiments wereperformed using a 1260 Infinity HPLC system (AgilentTechnologies, Santa Clara, CA) equipped with a Superdex200 Increase 10/300 GL column, which was connected in linewith a miniDAWN TREOS multiangle static light scatteringdetector (Wyatt Technology, Santa Barbara, CA). Data werecollected at 20 °C in phosphate buffered saline (PBS, pH 7.4)comprising 1 mM KH2PO4, 3 mM Na2HPO4, and 155 mMNaCl, and were analyzed using ASTRA 6 software (WyattTechnology). A dn/dc value of 0.185 mL/g was generally usedfor proteins, with extinction coefficients of 0.507 and 0.519 mLmg−1 cm−1 for ePN−Block/esPN−Block (HL4) and ePN−Block/esPN−Block (FL4), respectively, as calculated accordingto amino acid sequences.

Small-Angle X-ray Scattering (SAXS). SAXS measure-ments were performed on several fraction samples (Table S1,Supporting Information) of ePN−Block homooligomers andesPN−Block heterocomplexes after separation by SEC, WA20(sPN−Block), and chicken egg white lysozyme (Wako PureChemical Industries, Osaka, Japan) in 20 mM HEPES buffer(pH 7.5) containing 100 mM NaCl, 200 mM ArgHCl, and 10%glycerol at 20 °C using synchrotron radiation (λ = 0.1488 nm)with a PILATUS3 2 M detector (Dectris, Baden, Switzerland)at the Photon Factory BL-10C beamline (KEK, Tsukuba,Japan).79 Two-dimensional scattering images were integratedinto one-dimensional scattering intensities (I(q)) as a functionof the magnitude of the scattering vector q = (4π/λ) sin(θ/2)using the FIT2D program,80 where θ is the total scatteringangle.The scattering intensity for a colloidal dispersion is generally

given by the product of the form factor P(q) and structurefactor S(q). Hence, I(q) = nP(q) S(q), where n is the numberdensity of the particle. In the present experiments, the structurefactor was almost at unity (I(q) ≈ nP(q)), because interparticleinteractions such as the excluded volume effect and electrostaticinteractions can be neglected at low protein and high saltconcentrations. Thus, the form factor is given by the Fouriertransformation of the pair-distance distribution function p(r),which expresses the size and shape of the particle as follows:

∫= πP q p rqr

qrr( ) 4 ( )

sind

D

0

max

where Dmax is the maximum intraparticle distance. The indirectFourier transformation (IFT) technique was used to calculatep(r) for particles with a virtually model-free routine.66−68

Forward scattering intensity I(q→0) was extrapolated fromSAXS data, and the radius of gyration Rg, was estimated usingthe Guinier approximation.67

Modeling Analyses. Rigid-body models of esPN−Blockheterocomplexes were constructed using the program Coot81

and Foldit Standalone82 based on the crystal structure of the de



1390







novo protein WA20 dimer (protein data bank (PDB) code,3VJF)23 with consideration of N- and C-terminal directions.Rigid-body models were manually and iteratively refined tominimize differences between that calculated p(r) from modelsand those determined in SAXS experiments. The programCRYSOL83 in the ATSAS program suite84 was used forevaluating the solution scattering from the models and fitting itto experimental scattering curves with the χ2 value.71 The low-resolution dummy atom models were constructed from theSAXS data using ab initio shape modeling programs as follows(Figures S31 and S34). Calculations of rapid ab initio shapedetermination were performed 10 times by DAMMIF75

without a symmetry constraint, and the generated modelswere aligned and averaged by DAMAVER76 in the ATSASprogram suite84 for small-angle scattering data analysis frombiological macromolecules. The DAMAVER model wasmodified with fixed core by DAMSTART and furtherrefinement of the model was performed by DAMMIN.77

Superimposing model structures were performed by SUP-COMB.85 In addition, multistate structural modeling of theesPN−Block (FL4) heterocomplex e1s2 (FL4) was performedbased on SAXS profiles using MultiFoXS server78 as describedin more detail in the Supporting Information. The SAXS dataof the esPN−Block (HL4 and FL4) heterocomplex samples (e)and the rigid-body and dummy-atom models of e1s2 (HL4)and e1s2 (FL4) have been deposited into the Small AngleScattering Biological Data Bank86 (SASBDB accession codes:SASDD46 for e1s2 (HL4), SASDD56 for e1s2 (FL4)).Transmission Electron Microscopy (TEM) Imaging.

After IMAC purification, concentrated ePN−Block homoo-ligomer samples were fractionated in 20 mM HEPES buffer(pH 7.5) containing 100 mM NaCl, 10% glycerol, and 200 mMArgHCl using SEC on a HiLoad 16/600 Superdex 200 pg (GEhealthcare) column (Figures S11, S13, and S17). The samples(∼0.1 mg/mL protein concentration) were negatively stainedwith 2%(w/v) uranyl acetate on carbon-supported copper gridsglow-discharged beforehand, and then imaged by transmissionelectron microscopy (TEM) with a 200 kV field-emissionelectron source and a zero-loss imaging of omega-type energyfilter (JEM2200FS, JEOL, Tokyo, Japan).Atomic Force Microscopy (AFM) Imaging. Solutions of

5 mM NiCl2 were deposited onto freshly cleaved mica surfaces(ϕ12 mm). After 5 min, surfaces were gently rinsed 10 times insuperpure water and were then dried by blowing with nitrogengas. Samples of esPN−Block heterocomplexes were diluted in10 mM PBS buffer to a concentration of about 10 μg/mL. Priorto frequency modulation atomic force microscopy (FM-AFM)imaging, esPN−Block complex sample solutions (100 μL) weredeposited onto nickel ion-coated mica surfaces. Samples werethen incubated for 5 min at room temperature (25 °C). Self-assembled nanostructures of esPN−Block (HL4) heterocom-plex or esPN−Block (FL4) heterocomplex on mica surfaceswere then rinsed with PBS. AFM measurements wereperformed using an in-house-built, ultralow-noise FM atomicforce microscope87 combined with a commercially availableAFM controller (ARC2, Asylum Research, Santa Barbara, CA).All AFM experiments were performed at room temperature (25°C) in PBS buffer solution using a silicon cantilever (PPP-NCH, NanoWorld, Neuchatel, Switzerland) with a nominalspring constant of 42 N m−1 and a resonance frequency of 150kHz in liquid. A phase-locked loop circuit (Nanonis OC4,SPECS Zurich, Zurich, Switzerland) was used to detectfrequency shifts and to oscillate the cantilever with a constant

amplitude at its resonance frequency. Sizes of rod-like structuraldomains in AFM images were measured using the ImageJprogram.88

■ ASSOCIATED CONTENT

*S Supporting InformationThe Supporting Information is available free of charge on theACS Publications website at DOI: 10.1021/acssynbio.8b00007.

Supplementary text, Tables S1−S7, Figures S1−S37, andrelated descriptions (PDF)

■ AUTHOR INFORMATION

Corresponding Author*E-mail: [email protected].

ORCID

Takaaki Sato: 0000-0002-3455-4349Takeshi Fukuma: 0000-0001-8971-6002Ryoichi Arai: 0000-0001-5606-8483Author ContributionsN.K. and R.A. designed the research; M.H.H. created the denovo protein WA20; N.K., K.I. and K.S. performed proteinexpression and purification experiments; N.K., K.I., T.S., andR.A performed SAXS experiment and analysis; N.K., K.M., andT.F. performed AFM experiment and analysis; C.S., N.K. andK.M. performed TEM experiment and analysis; N.K. and R.A.wrote the manuscript; and all authors discussed results,commented on the manuscript, and revised it.

NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTS

We thank Prof. Nobuyasu Koga, Dr. Rie Koga, and Dr.Takahiro Kosugi at the Institute for Molecular Science (IMS)for assistance in SEC−MALS experiments. We are grateful toProf. Teruyuki Nagamune at University of Tokyo for kindlyproviding the plasmids pET32/EBFP-linker-EGFP. We alsothank Prof. Nobuaki Hayashida at Shinshu University and Dr.Shinya Honda at Advanced Industrial Science and Technologyfor helpful advice. This work was supported by Joint Researchof IMS and the Bio-AFM summer school at KanazawaUniversity. We thank Dr. Nobutaka Shimizu, Dr. NoriyukiIgarashi, and Photon Factory (PF) staff for their help insynchrotron SAXS experiments, which were performed at PF,KEK, under the approval of PF program advisory committee(Proposal Nos. 2014G111, 2016G153, and 2016G606). We areindebted to Divisions of Gene Research and InstrumentalAnalysis of Research Center for Supports to Advanced Science,Shinshu University, for providing facilities. This work wassupported by JSPS Research Fellowships (DC2) and JSPSKAKENHI Grant Nos. JP14J10185 and JP16H06837 to N.K.,and JSPS KAKENHI Grant Nos. JP22113508, JP24113707(Innovative Areas “Intrinsically Disordered Proteins”),JP24780097, JP26288101, JP16K05841, JP16H00761 (Innova-tive Areas “Dynamical Ordering & Integrated Functions”),JSPS Postdoctoral Fellowships for Research Abroad, andProgram for Dissemination of Tenure-Track System, to R.A.The work was also supported by NSF Grants MCB-1050510and MCB-1409402 to M.H.H.



1391



http://pubs.acs.org

http://pubs.acs.org/doi/abs/10.1021/acssynbio.8b00007


mailto:[email protected]

http://orcid.org/0000-0002-3455-4349

http://orcid.org/0000-0001-8971-6002

http://orcid.org/0000-0001-5606-8483


■ REFERENCES(1) Arai, R. (2018) Hierarchical design of artificial proteins andcomplexes toward synthetic structural biology. Biophys. Rev. 10, 391−410.(2) Hecht, M. H., Zarzhitsky, S., Karas, C., and Chari, S. (2018) Arenatural proteins special? Can. we do that? Curr. Opin. Struct. Biol. 48,124−132.(3) Dahiyat, B. I., and Mayo, S. L. (1997) De novo protein design:fully automated sequence selection. Science 278, 82−87.(4) Kuhlman, B., Dantas, G., Ireton, G. C., Varani, G., Stoddard, B.L., and Baker, D. (2003) Design of a novel globular protein fold withatomic-level accuracy. Science 302, 1364−1368.(5) Koga, N., Tatsumi-Koga, R., Liu, G., Xiao, R., Acton, T. B.,Montelione, G. T., and Baker, D. (2012) Principles for designing idealprotein structures. Nature 491, 222−227.(6) Huang, P. S., Boyken, S. E., and Baker, D. (2016) The coming ofage of de novo protein design,. Nature 537, 320−327.(7) Keefe, A. D., and Szostak, J. W. (2001) Functional proteins froma random-sequence library. Nature 410, 715−718.(8) Urvoas, A., Valerio-Lepiniec, M., and Minard, P. (2012) Artificialproteins from combinatorial approaches. Trends Biotechnol. 30, 512−520.(9) Kamtekar, S., Schiffer, J. M., Xiong, H., Babik, J. M., and Hecht,M. H. (1993) Protein design by binary patterning of polar andnonpolar amino acids. Science 262, 1680−1685.(10) Hecht, M. H., Das, A., Go, A., Bradley, L. H., and Wei, Y. (2004)De novo proteins from designed combinatorial libraries. Protein Sci. 13,1711−1723.(11) West, M. W., Wang, W., Patterson, J., Mancias, J. D., Beasley, J.R., and Hecht, M. H. (1999) De novo amyloid proteins from designedcombinatorial libraries. Proc. Natl. Acad. Sci. U. S. A. 96, 11211−11216.(12) Jumawid, M. T., Takahashi, T., Yamazaki, T., Ashigai, H., andMihara, H. (2009) Selection and structural analysis of de novoproteins from an α3β3 genetic library. Protein Sci. 18, 384−398.(13) Bradley, L. H., Kleiner, R. E., Wang, A. F., Hecht, M. H., andWood, D. W. (2005) An intein-based genetic selection allows theconstruction of a high-quality library of binary patterned de novoprotein sequences. Protein Eng., Des. Sel. 18, 201−207.(14) Patel, S. C., Bradley, L. H., Jinadasa, S. P., and Hecht, M. H.(2009) Cofactor binding and enzymatic activity in an unevolvedsuperfamily of de novo designed 4-helix bundle proteins. Protein Sci.18, 1388−1400.(15) Patel, S. C., and Hecht, M. H. (2012) Directed evolution of theperoxidase activity of a de novo-designed protein. Protein Eng., Des. Sel.25, 445−452.(16) Cherny, I., Korolev, M., Koehler, A. N., and Hecht, M. H.(2012) Proteins from an unevolved library of de novo designedsequences bind a range of small molecules. ACS Synth. Biol. 1, 130−138.(17) Fisher, M. A., McKinley, K. L., Bradley, L. H., Viola, S. R., andHecht, M. H. (2011) De novo designed proteins from a library ofartificial sequences function in Escherichia coli and enable cell growth.PLoS One 6, e15364.(18) Smith, B. A., Mularz, A. E., and Hecht, M. H. (2015) Divergentevolution of a bifunctional de novo protein. Protein Sci. 24, 246−252.(19) Digianantonio, K. M., and Hecht, M. H. (2016) A proteinconstructed de novo enables cell growth by altering gene regulation.Proc. Natl. Acad. Sci. U. S. A. 113, 2400−2405.(20) Hoegler, K. J., and Hecht, M. H. (2016) A de novo proteinconfers copper resistance in. Escherichia coli, Protein Sci. 25, 1249−1259.(21) Digianantonio, K. M., Korolev, M., and Hecht, M. H. (2017) Anon-natural protein rescues cells deleted for a key enzyme in centralmetabolism,. ACS Synth. Biol. 6, 694−700.(22) Donnelly, A., Murphy, G. S., Digianantonio, K. M., and Hecht,M. H. (2018) A de novo enzyme catalyzes a life-sustaining reaction in.Nat. Chem. Biol. 14, 253−255.(23) Arai, R., Kobayashi, N., Kimura, A., Sato, T., Matsuo, K., Wang,A. F., Platt, J. M., Bradley, L. H., and Hecht, M. H. (2012) Domain-

swapped dimeric structure of a stable and functional de novo four-helixbundle protein, WA20. J. Phys. Chem. B 116, 6789−6797.(24) Kobayashi, N., and Arai, R. (2017) Design and construction ofself-assembling supramolecular protein complexes using artificial andfusion proteins as nanoscale building blocks. Curr. Opin. Biotechnol. 46,57−65.(25) Padilla, J. E., Colovos, C., and Yeates, T. O. (2001) Nanohedra:using symmetry to design self assembling protein cages, layers, crystals,and filaments. Proc. Natl. Acad. Sci. U. S. A. 98, 2217−2221.(26) Lai, Y. T., Cascio, D., and Yeates, T. O. (2012) Structure of a16-nm cage designed by using protein oligomers. Science 336, 1129.(27) Lai, Y. T., Reading, E., Hura, G. L., Tsai, K. L., Laganowsky, A.,Asturias, F. J., Tainer, J. A., Robinson, C. V., and Yeates, T. O. (2014)Structure of a designed protein cage that self-assembles into a highlyporous cube. Nat. Chem. 6, 1065−1071.(28) Sinclair, J. C., Davies, K. M., Venien-Bryan, C., and Noble, M. E.(2011) Generation of protein lattices by fusing proteins with matchingrotational symmetry. Nat. Nanotechnol. 6, 558−562.(29) Sciore, A., Su, M., Koldewey, P., Eschweiler, J. D., Diffley, K. A.,Linhares, B. M., Ruotolo, B. T., Bardwell, J. C., Skiniotis, G., andMarsh, E. N. (2016) Flexible, symmetry-directed approach toassembling protein cages. Proc. Natl. Acad. Sci. U. S. A. 113, 8681−8686.(30) Yeates, T. O., Liu, Y., and Laniado, J. (2016) The design ofsymmetric protein nanomaterials comes of age in theory and practice,.Curr. Opin. Struct. Biol. 39, 134−143.(31) Lai, Y. T., Hura, G. L., Dyer, K. N., Tang, H. Y., Tainer, J. A.,and Yeates, T. O. (2016) Designing and defining dynamic protein cagenanoassemblies in solution. Sci. Adv. 2, e1501855.(32) King, N. P., Sheffler, W., Sawaya, M. R., Vollmar, B. S., Sumida,J. P., Andre, I., Gonen, T., Yeates, T. O., and Baker, D. (2012)Computational design of self-assembling protein nanomaterials withatomic level accuracy. Science 336, 1171−1174.(33) King, N. P., Bale, J. B., Sheffler, W., McNamara, D. E., Gonen, S.,Gonen, T., Yeates, T. O., and Baker, D. (2014) Accurate design of co-assembling multi-component protein nanomaterials. Nature 510, 103−108.(34) Hsia, Y., Bale, J. B., Gonen, S., Shi, D., Sheffler, W., Fong, K. K.,Nattermann, U., Xu, C., Huang, P. S., Ravichandran, R., Yi, S., Davis,T. N., Gonen, T., King, N. P., and Baker, D. (2016) Design of ahyperstable 60-subunit protein icosahedron. Nature 535, 136−139.(35) Bale, J. B., Gonen, S., Liu, Y., Sheffler, W., Ellis, D., Thomas, C.,Cascio, D., Yeates, T. O., Gonen, T., King, N. P., and Baker, D. (2016)Accurate design of megadalton-scale two-component icosahedralprotein complexes. Science 353, 389−394.(36) Voet, A. R., Noguchi, H., Addy, C., Simoncini, D., Terada, D.,Unzai, S., Park, S. Y., Zhang, K. Y., and Tame, J. R. (2014)Computational design of a self-assembling symmetrical beta-propellerprotein. Proc. Natl. Acad. Sci. U. S. A. 111, 15102−15107.(37) Voet, A. R., Noguchi, H., Addy, C., Zhang, K. Y., and Tame, J. R.(2015) Biomineralization of a cadmium chloride nanocrystal by adesigned symmetrical protein. Angew. Chem., Int. Ed. 54, 9857−9860.(38) Luo, Q., Dong, Z., Hou, C., and Liu, J. (2014) Protein-basedsupramolecular polymers: progress and prospect. Chem. Commun. 50,9997−10007.(39) Luo, Q., Hou, C., Bai, Y., Wang, R., and Liu, J. (2016) Proteinassembly: Versatile approaches to construct highly ordered nanostruc-tures. Chem. Rev. 116, 13571−13632.(40) Carlson, J. C., Jena, S. S., Flenniken, M., Chou, T. F., Siegel, R.A., and Wagner, C. R. (2006) Chemically controlled self-assembly ofprotein nanorings. J. Am. Chem. Soc. 128, 7630−7638.(41) Kitagishi, H., Oohora, K., Yamaguchi, H., Sato, H., Matsuo, T.,Harada, A., and Hayashi, T. (2007) Supramolecular hemoproteinlinear assembly by successive interprotein heme-heme pocketinteractions. J. Am. Chem. Soc. 129, 10326−10327.(42) Kitagishi, H., Kakikura, Y., Yamaguchi, H., Oohora, K., Harada,A., and Hayashi, T. (2009) Self-Assembly of One- and Two-Dimensional Hemoprotein Systems by Polymerization through



1392


Heme-Heme Pocket Interactions. Angew. Chem., Int. Ed. 48, 1271−1274.(43) Ogihara, N. L., Ghirlanda, G., Bryson, J. W., Gingery, M.,DeGrado, W. F., and Eisenberg, D. (2001) Design of three-dimensional domain-swapped dimers and fibrous oligomers. Proc.Natl. Acad. Sci. U. S. A. 98, 1404−1409.(44) Hirota, S., Hattori, Y., Nagao, S., Taketa, M., Komori, H.,Kamikubo, H., Wang, Z., Takahashi, I., Negi, S., Sugiura, Y., Kataoka,M., and Higuchi, Y. (2010) Cytochrome c polymerization bysuccessive domain swapping at the C-terminal helix. Proc. Natl.Acad. Sci. U. S. A. 107, 12854−12859.(45) Miyamoto, T., Kuribayashi, M., Nagao, S., Shomura, Y., Higuchi,Y., and Hirota, S. (2015) Domain-swapped cytochrome cb562 dimerand its nanocage encapsulating a Zn-SO4 cluster in the internal cavity.Chem. Sci. 6, 7336−7342.(46) Papapostolou, D., Smith, A. M., Atkins, E. D., Oliver, S. J.,Ryadnov, M. G., Serpell, L. C., and Woolfson, D. N. (2007)Engineering nanoscale order into a designed protein fiber. Proc.Natl. Acad. Sci. U. S. A. 104, 10853−10858.(47) Sharp, T. H., Bruning, M., Mantell, J., Sessions, R. B., Thomson,A. R., Zaccai, N. R., Brady, R. L., Verkade, P., and Woolfson, D. N.(2012) Cryo-transmission electron microscopy structure of a giga-dalton peptide fiber of de novo design. Proc. Natl. Acad. Sci. U. S. A.109, 13266−13271.(48) Boyle, A. L., Bromley, E. H., Bartlett, G. J., Sessions, R. B.,Sharp, T. H., Williams, C. L., Curmi, P. M., Forde, N. R., Linke, H.,and Woolfson, D. N. (2012) Squaring the circle in peptide assembly:from fibers to discrete nanostructures by de novo design. J. Am. Chem.Soc. 134, 15457−15467.(49) Fletcher, J. M., Harniman, R. L., Barnes, F. R., Boyle, A. L.,Collins, A., Mantell, J., Sharp, T. H., Antognozzi, M., Booth, P. J.,Linden, N., Miles, M. J., Sessions, R. B., Verkade, P., and Woolfson, D.N. (2013) Self-assembling cages from coiled-coil peptide modules.Science 340, 595−599.(50) Park, W. M., Bedewy, M., Berggren, K. K., and Keating, A. E.(2017) Modular assembly of a protein nanotriangle using orthogonallyinteracting coiled coils. Sci. Rep. 7, 10577.(51) Yagi, S., Akanuma, S., and Yamagishi, A. (2018) (2018)Creation of artificial protein-protein interactions using alpha-helices asinterfaces. Biophys. Rev. 10, 411−420.(52) Yagi, S., Akanuma, S., Yamagishi, M., Uchida, T., and Yamagishi,A. (2016) De novo design of protein-protein interactions throughmodification of inter-molecular helix-helix interface residues. Biochim.Biophys. Acta, Proteins Proteomics 1864, 479−487.(53) Salgado, E. N., Faraone-Mennella, J., and Tezcan, F. A. (2007)Controlling protein-protein interactions through metal coordination:assembly of a 16-helix bundle protein. J. Am. Chem. Soc. 129, 13374−13375.(54) Brodin, J. D., Ambroggio, X. I., Tang, C., Parent, K. N., Baker, T.S., and Tezcan, F. A. (2012) Metal-directed, chemically tunableassembly of one-, two- and three-dimensional crystalline proteinarrays. Nat. Chem. 4, 375−382.(55) Sontz, P. A., Bailey, J. B., Ahn, S., and Tezcan, F. A. (2015) Ametal organic framework with spherical protein nodes: Rationalchemical design of 3D protein crystals,. J. Am. Chem. Soc. 137, 11598−11601.(56) Bailey, J. B., Subramanian, R. H., Churchfield, L. A., and Tezcan,F. A. (2016) Metal-directed design of supramolecular proteinassemblies. Methods Enzymol. 580, 223−250.(57) Besenius, P. (2017) Controlling supramolecular polymerizationthrough multicomponent self-assembly. J. Polym. Sci., Part A: Polym.Chem. 55, 34−78.(58) Staples, J. K., Oshaben, K. M., and Horne, W. S. (2012) Amodular synthetic platform for the construction of protein-basedsupramolecular polymers via coiled-coil self-assembly,. Chem. Sci. 3,3387−3392.(59) Sim, S., Niwa, T., Taguchi, H., and Aida, T. (2016)Supramolecular Nanotube of Chaperonin GroEL: Length Control

for Cellular Uptake Using Single-Ring GroEL Mutant as End-Capper.J. Am. Chem. Soc. 138, 11152−11155.(60) Kim, Y. E., Kim, Y. N., Kim, J. A., Kim, H. M., and Jung, Y.(2015) Green fluorescent protein nanopolygons as monodispersesupramolecular assemblies of functional proteins with defined valency.Nat. Commun. 6, 7134.(61) Kobayashi, N., Yanase, K., Sato, T., Unzai, S., Hecht, M. H., andArai, R. (2015) Self-assembling nano-architectures created from aprotein nano-building block using an intermolecularly folded dimericde novo protein. J. Am. Chem. Soc. 137, 11285−11293.(62) Guthe, S., Kapinos, L., Moglich, A., Meier, S., Grzesiek, S., andKiefhaber, T. (2004) Very fast folding and association of atrimerization domain from bacteriophage T4 fibritin. J. Mol. Biol.337, 905−915.(63) Arai, R., Ueda, H., Kitayama, A., Kamiya, N., and Nagamune, T.(2001) Design of the linkers which effectively separate domains of abifunctional fusion protein. Protein Eng. 14, 529−532.(64) Arai, R., Wriggers, W., Nishikawa, Y., Nagamune, T., andFujisawa, T. (2004) Conformations of variably linked chimericproteins evaluated by synchrotron X-ray small-angle scattering.Proteins: Struct., Funct., Genet. 57, 829−838.(65) Marqusee, S., and Baldwin, R. L. (1987) Helix stabilization byGlu-···Lys+ salt bridges in short peptides of de novo design. Proc. Natl.Acad. Sci. U. S. A. 84, 8898−8902.(66) Glatter, O. (1980) Evaluation of small-angle scattering datafrom lamellar and cylindrical particles by the indirect Fouriertransformation method. J. Appl. Crystallogr. 13, 577−584.(67) Glatter, O., and Kratky, O. (1982) Small-Angle X-Ray Scattering,Academic Press, New York.(68) Brunner-Popela, J., and Glatter, O. (1997) Small-anglescattering of interacting particles 0.1. Basic principles of a globalevaluation technique. J. Appl. Crystallogr. 30, 431−442.(69) Receveur-Brechot, V., and Durand, D. (2012) How Random areIntrinsically Disordered Proteins? A Small Angle ScatteringPerspective,. Curr. Protein Pept. Sci. 13, 55−75.(70) Kikhney, A. G., and Svergun, D. I. (2015) A practical guide tosmall angle X-ray scattering (SAXS) of flexible and intrinsicallydisordered proteins,. FEBS Lett. 589, 2570−2577.(71) Trewhella, J., Duff, A. P., Durand, D., Gabel, F., Guss, J. M.,Hendrickson, W. A., Hura, G. L., Jacques, D. A., Kirby, N. M., Kwan,A. H., Perez, J., Pollack, L., Ryan, T. M., Sali, A., Schneidman-Duhovny, D., Schwede, T., Svergun, D. I., Sugiyama, M., Tainer, J. A.,Vachette, P., Westbrook, J., and Whitten, A. E. (2017) 2017publication guidelines for structural modelling of small-angle scatteringdata from biomolecules in solution: an update. Acta Crystallogr. D 73,710−728.(72) Sjekloca, L., Pudas, R., Sjoblom, B., Konarev, P., Carugo, O.,Rybin, V., Kiema, T. R., Svergun, D., Ylanne, J., and Djinovic Carugo,K. (2007) Crystal structure of human filamin C domain 23 and smallangle scattering model for filamin C 23−24 dimer. J. Mol. Biol. 368,1011−1023.(73) Durand, D., Cannella, D., Dubosclard, V., Pebay-Peyroula, E.,Vachette, P., and Fieschi, F. (2006) Small-angle X-ray scatteringreveals an extended organization for the autoinhibitory resting state ofthe p47phox modular protein. Biochemistry 45, 7185−7193.(74) Durand, D., Vives, C., Cannella, D., Perez, J., Pebay-Peyroula, E.,Vachette, P., and Fieschi, F. (2010) NADPH oxidase activator p67phox

behaves in solution as a multidomain protein with semi-flexible linkers.J. Struct. Biol. 169, 45−53.(75) Franke, D., and Svergun, D. I. (2009) DAMMIF, a program forrapid ab-initio shape determination in small-angle scattering. J. Appl.Crystallogr. 42, 342−346.(76) Volkov, V. V., and Svergun, D. I. (2003) Uniqueness of ab initioshape determination in small-angle scattering. J. Appl. Crystallogr. 36,860−864.(77) Svergun, D. I. (1999) Restoring low resolution structure ofbiological macromolecules from solution scattering using simulatedannealing. Biophys. J. 76, 2879−2886.



1393


(78) Schneidman-Duhovny, D., Hammel, M., Tainer, J. A., and Sali,A. (2016) FoXS, FoXSDock and MultiFoXS: Single-state and multi-state structural modeling of proteins and their complexes based onSAXS profiles. Nucleic Acids Res. 44, W424−429.(79) Igarashi, N., Watanabe, Y., Shinohara, Y., Inoko, Y., Matsuba, G.,Okuda, H., Mori, T., and Ito, K. (2011) Upgrade of the small angle X-ray scattering beamlines at the Photon Factory. J. Phys.: Conf. Ser. 272,012026.(80) Hammersley, A. P., Svensson, S. O., Hanfland, M., Fitch, A. N.,and Hausermann, D. (1996) Two-dimensional detector software: fromreal detector to idealised image or two-theta scan. High Pressure Res.14, 235−248.(81) Emsley, P., Lohkamp, B., Scott, W. G., and Cowtan, K. (2010)Features and development of Coot. Acta Crystallogr., Sect. D: Biol.Crystallogr. 66, 486−501.(82) Kleffner, R., Flatten, J., Leaver-Fay, A., Baker, D., Siegel, J. B.,Khatib, F., and Cooper, S. (2017) Foldit Standalone: a video game-derived protein structure manipulation interface using Rosetta.Bioinformatics 33, 2765−2767.(83) Svergun, D., Barberato, C., and Koch, M. H. J. (1995) CRYSOL- A program to evaluate x-ray solution scattering of biologicalmacromolecules from atomic coordinates. J. Appl. Crystallogr. 28, 768−773.(84) Franke, D., Petoukhov, M. V., Konarev, P. V., Panjkovich, A.,Tuukkanen, A., Mertens, H. D. T., Kikhney, A. G., Hajizadeh, N. R.,Franklin, J. M., Jeffries, C. M., and Svergun, D. I. (2017) ATSAS 2.8: acomprehensive data analysis suite for small-angle scattering frommacromolecular solutions. J. Appl. Crystallogr. 50, 1212−1225.(85) Kozin, M. B., and Svergun, D. I. (2000) Automated matching ofhigh- and low-resolution structural models. J. Appl. Crystallogr. 34, 33−41.(86) Valentini, E., Kikhney, A. G., Previtali, G., Jeffries, C. M., andSvergun, D. I. (2015) SASBDB, a repository for biological small-anglescattering data. Nucleic Acids Res. 43, D357−363.(87) Fukuma, T., Kimura, M., Kobayashi, K., Matsushige, K., andYamada, H. (2005) Development of low noise cantilever deflectionsensor for multienvironment frequency-modulation atomic forcemicroscopy. Rev. Sci. Instrum. 76, 053704.(88) Schneider, C. A., Rasband, W. S., and Eliceiri, K. W. (2012)NIH Image to ImageJ: 25 years of image analysis. Nat. Methods 9,671−675.



1394


self-assembling supramolecular nanostructures constructed...

Documents