Altering DNA-Programmable Colloidal Crystallization Paths by

Jul 21, 2017 - Colloidal crystal engineering with DNA can be used to realize precise .... yielding conjugates bearing a rigid, double-stranded DNA (ds...
0 downloads 0 Views 3MB Size
Download PDF
Letter pubs.acs.org/NanoLett

Altering DNA-Programmable Colloidal Crystallization Paths by Modulating Particle Repulsion Mary X. Wang,†,∥ Jeffrey D. Brodin,‡,∥ Jaime A. Millan,§,∥ Soyoung E. Seo,‡,∥ Martin Girard,§ Monica Olvera de la Cruz,*,‡,§ Byeongdu Lee,*,⊥ and Chad A. Mirkin*,†,‡,§,∥ †

Department of Chemical and Biological Engineering, ‡Department of Chemistry, §Department of Materials Science and Engineering, International Institute for Nanotechnology, Northwestern University, Evanston, Illinois 60208, United States ⊥ X-Ray Science Division, Argonne National Laboratory, 9700 S. Cass Avenue, Argonne, Illinois 60439, United States ∥

S Supporting Information *

ABSTRACT: Colloidal crystal engineering with DNA can be used to realize precise control over nanoparticle (NP) arrangement. Here, we investigate a case of DNA-based assembly where the properties of DNA as a polyelectrolyte brush are employed to alter a hybridization-driven NP crystallization pathway. Using the coassembly of DNAconjugated proteins and spherical gold nanoparticles (AuNPs) as a model system, we explore how steric repulsion between noncomplementary, neighboring NPs due to overlapping DNA shells can influence their ligand-directed behavior. Specifically, our experimental data coupled with coarse-grained molecular dynamics (MD) simulations reveal that, by changing factors related to NP repulsion, two structurally distinct outcomes can be achieved. When steric repulsion between DNA−AuNPs is significantly greater than that between DNA−proteins, a lower packing density crystal lattice is favored over the structure that is predicted by design rules based on DNA hybridization considerations alone. This is enabled by the large difference in DNA density on AuNPs versus proteins and can be tuned by modulating the flexibility, and thus conformational entropy, of the DNA on the constituent particles. At intermediate ligand flexibility, the crystallization pathways are energetically similar, and the structural outcome can be adjusted using the density of DNA duplexes on DNA−AuNPs and by screening the Coulomb potential between them. Such lattices are shown to undergo dynamic reorganization upon changing the salt concentration. These data help elucidate the structural considerations necessary for understanding repulsive forces in DNA-mediated assembly and lay the groundwork for using them to increase architectural diversity in engineering colloidal crystals. KEYWORDS: DNA, protein assembly, nanoparticle superlattice, nanomaterials, colloidal crystals, self-assembly

T

design rules can accurately predict their crystallization behavior based on the guiding principle of maximizing DNA hybridization between complementary particles.10 In many of these thermodynamically preferred arrangements, the DNA shells of neighboring noncomplementary PAEs overlap, resulting in energetically unfavorable interdigitation of polyanionic strands. Typically, these repulsive forces are considered negligible in comparison with the attractive force of DNA hybridization due to the highly curved NP surface and high local salt concentration.10,33,34 However, inspired by studies on electrostatically assembled particles, we hypothesized that the crystallization pathway could be altered when repulsive forces between particles are asymmetric, that is, when repulsion between one set of particles is significantly stronger than that between another.35,36 This would likely require the assembly of

he use of biomolecules with high information content as structure-directing units has led to powerful new methods for colloidal assembly.1−9 DNA, in particular, provides the ability to control both recognition and spacing with subnanometer precision. Virtually any particle, regardless of composition, can be transformed into a nanoscale building block or programmable atom equivalent (PAE) by modifying it with a dense layer of oligonucleotides.10−16 Proteins represent an important class of functional NPs that are highly monodisperse compared with their inorganic counterparts.17 Their structural diversity and well-defined surface chemistries make them ideal building blocks for synthesizing novel materials.18−30 They can be assembled into superlattices by covalently linking their surface lysine or cysteine residues with DNA.31,32 This allows the number of DNA strands on the surface of the protein to be easily tuned and makes them excellent candidates for providing new insight into DNAmediated NP crystallization behavior. Unlike molecular and atomic systems, PAE assembly occurs independently of the identity of the building block core, and © XXXX American Chemical Society

Received: June 13, 2017 Revised: July 13, 2017

A

DOI: 10.1021/acs.nanolett.7b02502 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

varying flexibility were used as a model system. AuNPs were densely functionalized with hexylthiolated DNA to produce conjugates modified with ∼275 strands. DNA−protein conjugates were synthesized by first reacting either BV or CG catalase with a heterobifunctional cross-linker with an aminereactive NHS ester and an azide group for conjugation to oligonucleotides. The azide-conjugated proteins were then reacted with 5′-dibenzocyclooctyne (DBCO) modified DNA via strain-promoted cycloaddition37,38 to produce proteins modified with ∼50 and 70 strands for CG and BV catalase, respectively (Tables S1 and S2).31 DNA “linkers” were hybridized to the surface-bound DNA of each PAE, yielding conjugates bearing a rigid, double-stranded DNA (dsDNA) region and an exposed, nonself-complementary 5′ overhang (“sticky end”) to dictate PAE assembly. Upon combining complementary protein and AuNP conjugates in buffer supplemented with 0.5 M NaCl, the collective hybridization of DNA sticky ends lead to the visible precipitation of aggregates. Thermal annealing of the aggregates slightly below their melting temperature (Tm, the temperature at which 50% of NPs are disassociated) results in their reorganization into crystals of the thermodynamically most stable arrangement, as determined by small-angle X-ray scattering (SAXS) of samples in solution and transmission electron microscopy (TEM) after the encapsulation of crystals in silica and resin. Both SAXS and TEM provide only information about AuNP position because the scattering cross sections of proteins relative to AuNPs are negligible due to the large difference in electron density. Interestingly, binary protein−AuNP superlattices annealed by heating at a constant temperature consistently produce single crystals (Figure S3). In contrast, AuNPs assembled in the same manner form polycrystalline structures10 and must be subjected to slow annealing to form the equivalent single crystals.39 This observation suggests that the uniformity of the protein building blocks helps to minimize defect formation and kinetically trapped states, enabling the formation of single crystalline domains at higher ionic strengths. To examine the consequence of changing oligonucleotide flexibility on the assembly process, we prepared DNA−CG catalase conjugates with 0, 2, and 5 subunits of hexaethylene glycol phosphate (SP18) spacers between the dsDNA region and the DBCO functional group of the ligand. These were assembled into superlattices with DNA−AuNPs bearing two such SP18 spacers (Figure 1). We hypothesized that increasing flexibility would increase the conformational freedom of the ligand and thus increase the energetic penalty for DNA shells on neighboring, noncomplementary particles to overlap. Superlattices formed from catalase-PAEs with five spacers exhibit a simple cubic scattering pattern, which is the expected product for hybridization-driven assembly. In comparison, when the spacer region is eliminated a different scattering pattern corresponding to the space group I4̅3d is observed, supporting our hypothesis that linker flexibility could alter crystallization. Proteins modified with oligonucleotides of intermediate flexibility (two spacers) produce a mixture of both lattice symmetries. The same trend is observed using DNA ligands of the same length but with varying number of spacers (Figure S6), confirming that this phase transition is the result of changing ligand flexibility rather than length. In contrast, when the flexibility of the ligand on the AuNP PAE is varied, the opposite trend is observed (Figures S7 and S8). The I4̅3d structure is preferred when the DNA on the AuNP is more flexible, while the CsCl structure is favored when the

two similarly sized particles with significantly different DNA densities. Here, we use a model system consisting of DNA−AuNPs coassembled with DNA−proteins [catalase from bovine liver (BV) or Corynebacterium glutamicum (CG)] to study the effect of asymmetric repulsion on PAE assembly (Scheme 1A). In this Scheme 1. Modulating Particle Repulsion in DNAAssembled Binary Protein−AuNP PAE Superlatticesa

a

For binary superlattices in which complementary DNA linkers hybridize to assemble DNA-functionalized AuNPs and catalase proteins, (A) a phase transition from CsCl to a lower packing density crystal structure occurs when steric repulsion (FR, magnitude of repulsion) between one set of noncomplementary particles is much larger than the other. (B) Steric repulsion is modulated by changing the flexibility of the DNA sequence through incorporation of hexaethylene glycol spacer units, the ratio of linker strands to AuNP-PAE anchor strands, and shielding of the negatively charged DNA phosphate backbone through salt concentration (the counterion is Cl−).

scheme, the AuNP is functionalized with 5 and 4 times more DNA than the protein. We systematically modulate the relative amount of repulsion between the two sets of particles and monitor the resulting changes in crystallization behavior (Scheme 1B). Specifically, we tune the interaction between PAEs by changing the number of flexible ethylene glycol spacer units in the oligonucleotide at each NP surface. As oligonucleotide ligand flexibility is decreased on the protein or increased on the AuNP, a phase transition with respect to crystal symmetry and protein orientation is observed to a lower packing density structure. Through MD simulations, we determine that this is a result of an increase in volume accessible to flexible DNA ligands, which effectively increases conformational entropy and repulsion between noncomplementary PAEs. Furthermore, we explore the sensitivity of this crystallization process to determine which other interparticle forces can be tuned to enhance repulsion. These studies provide a general framework for accessing, understanding, and exploring a new phase space of PAE assembly in which likeparticle interactions play a significant role. To test the effect of asymmetric repulsion on crystallization behavior, binary superlattices composed of the enzyme catalase and AuNPs (10 nm diameter) functionalized with DNA of B

DOI: 10.1021/acs.nanolett.7b02502 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

Figure 1. Phase transformation due to changing DNA ligand flexibility. (A) Radially averaged SAXS data (experimental and simulated) of crystals with increasing spacer number accompanied by 2D diffraction patterns. Crystals were assembled in aqueous 0.5 M NaCl with 400 linker equivalents added to the AuNP PAEs. (B) Corresponding TEM cross-sectional images of the Th3P4 (top) and CsCl lattices (bottom). Scale bars 100 nm. Figure 2. Structural differences between (A) CsCl and (B) Th3P4 lattices highlighting the difference in nearest-neighbor and next-nearest neighbor distances. MD simulations reveal that phase transitions result from changes in ligand distribution around the protein conjugate and steric repulsion between neighboring AuNPs. As the flexibility of the DNA on the protein conjugates increases, their accessible volume increases: (C) flexible DNA ligands arrange diffusely around the protein core in the CsCl lattice, whereas (D) rigid DNA ligands retain the shape of the protein core in the Th3P4 lattice. The protein is depicted in blue, with protein DNA sticky ends in light blue and AuNPs in red.

DNA on the AuNP contains no flexors. This suggests that the conformational entropy of the DNA on the AuNPs increases as a function of flexibility despite dense oligonucleotide packing, increasing repulsion between them and thus favoring the lower packing density structure. The same I43̅ d product is observed when the aggregates are annealed by increasing the temperature from room temperature (Figure S9) as in the case when they are slowly cooled from high temperature through the melting point39 (Figure S10). This supports the conclusion that the lattice is a thermodynamic structure and not the product of kinetic effects. To further understand the observed structures, we determined the stoichiometry of each lattice element using UV/vis spectroscopy to quantify the relative concentrations of AuNPs and proteins (Scheme S2, Table S7). This revealed that the lattice stoichiometry of AuNP to protein is 1:1 for the 5spacer sample, confirming the expected CsCl arrangement, and 3:4 for the 0-spacer sample, indicating the formation of crystals isostructural with Th3P4 (Figure S4). The Th3P4 phase has been previously observed in systems of spherical particles hybridized with trivalent DNA linkers.35 In a CsCl lattice, each protein and AuNP PAE has 8 nearest neighbors. In contrast, while each AuNP in the Th3P4 lattice has 8 nearest protein neighbors arranged in a distorted octahedron, each protein is surrounded by only 6 AuNPs, giving it a lower packing density than the CsCl structure (Figure 2A,B). Further analysis of the two lattices reveals that there are significant differences in the distances between noncomplementary neighboring particles. In the Th3P4 lattice, the distance between AuNPs is 14% larger, while the distance between proteins is 22% smaller than if the same set of conjugates were in a CsCl arrangement. On the other hand, the distance between complementary AuNPs and proteins is similar (2% shorter in the Th3P4 lattice), indicating that attractive forces are comparable. This suggests that asymmetric repulsion between PAEs plays a strong role in the formation of the Th3P4 lattice. To elucidate the origin of the CsCl to Th3P4 phase transition and to understand the nature of like-particle repulsion, the system was modeled using coarse-grained isobaric (NPT) MD simulations. These simulations build upon previous work that has accurately predicted PAE assembly. 40−42 In these simulations, AuNPs, proteins, and DNA chains were explicitly modeled, and the DNA chain rigidity was tuned by incorporating ssDNA units as an approximation for SP18

groups (Figure S13). Ideal Th3P4 and CsCl superlattices were constructed, each comprising 54 and 28 particles, respectively, and allowed to equilibrate at a temperature below Tm with protein and AuNP PAEs diffusing freely.43,44 By simulating both CsCl and Th3P4 binary superlattices for a specific DNA design, the thermodynamically preferred structure between these two competing phases can be determined based on relative energetic differences and the entropies of their constituent DNA chains (Table S15, Figure S14). Due to the complexity of DNA−NP lattices and the vast number of possible microstates, it is not feasible to calculate the DNA conformational entropy exactly. Instead, we calculated the volume accessible to the DNA chains (Vfree) as an estimation of conformational entropy (Table S17). Energetic contributions (Utotal) were determined from a sum of hybridization energy (Uhybridization), the bond and angle energies due to distortions in the DNA chain (Uchain), and the excluded volume energy (Uexcluded). The enthalpy (H) can be estimated using H = Utotal + PV, where P = 0 in our implicit solvent model.45 The free energy difference between the CsCl and Th3P4 phases can be estimated using ΔF = ΔUtotal − TΔS, where ΔS can be estimated as kBlogVfree. As observed experimentally, the CsCl structure is preferred for proteins with flexible ligands. As the number of spacers decreases and the DNA ligand rigidifies, the Th3P4 symmetry becomes more stable. Calculations show that this transition is due to a relative increase in Uchain and Vfree in flexible systems over rigid ones, indicating that, for systems with rigid sequences, the CsCl lattice is unfavorable due to arrangement of the DNA chains. Notably, flexible DNA strands explore 128% greater Vfree than rigid ones, supporting the idea that flexibility increases the conformational entropy of DNA ligands. This can be visualized in the diffuse distribution of sticky ends around the protein PAE when flexible DNA are used (Figure 2C,D). C

DOI: 10.1021/acs.nanolett.7b02502 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

that under certain conditions, protein shape, and orientation could be used to direct PAE assembly. Since the two phases are energetically similar when DNA of intermediate flexibility (two spacers) is used on both protein and AuNP PAEs, this system serves as a sensitive platform for investigating additional methods to modulate repulsion. To investigate the effect of changing the density of dsDNA on the AuNP PAE, different equivalents of linker strands were added (Figure S1, Table S3). The composition of crystalline mixtures was quantified by modeling the radially averaged SAXS diffraction patterns to determine the relative contribution from both lattices (Figure 4A, B).46,47 Higher ratios of linker loading on the AuNP shift the equilibrium toward the Th3P4 lattice (Figure 4C, E) and increase the interparticle distance (Figure 5) due to repulsion between neighboring DNA−AuNPs. As the distance between complementary PAEs increases, DNA hybridization decreases per the complementary contact model (CCM).10 By transforming to the Th3P4 lattice, the distance between complementary PAEs is reduced and hybridization increases, increasing the stability of the lattice while maintaining a larger distance between AuNP PAEs. The effect of changing the density of dsDNA on the protein was studied by comparing the assembly properties of BV and CG catalase. The two proteins exhibit the same phase transformation, although lattices assembled from BV catalase favor the CsCl symmetry. This can be attributed to the 40% higher DNA loading on the BV catalase (Table S2), which reduces its tolerance for packing closely. The effects of electrolyte concentration on repulsion and phase equilibrium were investigated by assembling superlattices at 0.25, 0.5, and 1 M NaCl. Due to its nature as a polyelectrolyte, the effective diameter of DNA (deff) decreases with electrolyte concentration, as predicted by polyelectrolyte theory.48,49 A larger deff should increase the energetic penalty of overlapping AuNP DNA shells. Indeed, at 0.25 M NaCl the equilibrium is shifted toward the Th3P4 structure (Figure 4D, F), while the CsCl lattice predominates in high ionic strength solution (1 M NaCl). Interestingly, regardless of salt concentration, proteins with 0 spacers form 100% Th3P4 lattices and proteins with 5 spacers form 100% CsCl, suggesting that the conformational entropy of DNA due to differences in flexibility is a dominant effect. Charge screening of the DNA not only decreases deff but also decreases DNA length (Table S6). Although much of the change in length arises from compression of the SP18 groups, the height per base pair also decreases with increased salt concentration. The sensitivity of the phase equilibrium to salt concentration suggests the possibility of dynamically reorganizing lattices. To evaluate this, annealed lattices were heated and cooled after changing the ionic strength of the environment (Figure 6). When a lattice containing a mixture of both symmetries is transferred to a solution of higher ionic strength and heated, peaks corresponding to Th 3 P 4 disappear, resulting in reorganization to a completely CsCl lattice (Figure 6A). On the other hand, CsCl samples assembled at higher salt concentration and diluted to lower ionic strength can only reorganize into the Th3P4 lattice after dissociation and cooling (Figure 6B,C), due to a reduction in the lattice Tm resulting from a lower electrolyte concentration. No distinct intermediate phases appear during the reorganization process. This work demonstrates that, for a highly asymmetric system, repulsive forces can change the energetics of crystallization in DNA-programmable NP assembly to favor a lower packing

The interactions between noncomplementary PAEs were investigated by calculating the distribution of their hydrodynamic radii and the distance between neighboring AuNPs and proteins in each phase (Figures S17 and S18). For shorter ligands, AuNPs in the CsCl structure are close enough that their nonhybridizing DNA shells overlap, which is sterically unfavorable due to interdigitation of DNA strands (Figure 3C).

Figure 3. Analysis of DNA shells of AuNPs and proteins (2 spacers) reveals that distances between next-nearest neighbor proteins (blue) are larger for (A) CsCl arrangement than in the (B) Th3P4. On the other hand, DNA shells of AuNPs (red, DNA−protein omitted for clarity) interdigitate in the (C) CsCl lattice, whereas there is no overlap in the (D) Th3P4 lattice.

In the Th3P4 arrangement, the AuNPs are sufficiently far apart that their DNA shells do not overlap (Figure 3D). When the protein is functionalized with rigid oligonucleotides, the reduced conformational freedom of the ligands allows the protein PAEs to pack closely to relieve steric repulsion between AuNP PAEs, resulting in the formation of the Th3P4 structure (Figure 3B). When the DNA on the protein is flexible, it is less favorable for DNA−proteins to pack closely (Figure S15B), and the CsCl lattice is favored (Figure 3A). Since the salt concentration is high enough that counterions effectively neutralize the electrostatic charge of the DNA ligands, interparticle repulsion is expected to be primarily steric in nature and, in effect, a measure of the tolerance of noncomplementary DNA shells for overlapping. To investigate the role of the anisotropic shape of the protein core in assembly, control simulations were run where the protein core was replaced by a spherical particle functionalized with the same number of DNA ligands (Table S16). The same phase transition can be observed from CsCl to Th3P4, showing that the assembly behavior results from differences in DNA density. However, catalase PAEs do exhibit preferential orientation in the Th3P4 lattice with respect to the surrounding AuNP PAEs. In the Th3P4 structure, 71% of the proteins have an axis of symmetry oriented between 0 and 24° of vectors projected to the tips of the octahedra (Figure S16). By comparison, the proteins in the CsCl structure exhibit a significantly more uniform distribution. This highly preferential alignment of the enzyme with the surrounding AuNPs suggests D

DOI: 10.1021/acs.nanolett.7b02502 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

Figure 4. Superlattice composition mediated by DNA flexibility, solution ionic strength, and AuNP linker loading. Composition of superlattices of DNA−AuNP assembled with (A) CG catalase and (B) BV catalase where the arrow indicates the direction of increasing Th3P4 content. Radially averaged SAXS patterns show that the Th3P4 lattice is favored by increasing AuNP linker loading (linker eq defined as linkers added per NP), (C) CG catalase, (E) BV catalase, and decreasing solution ionic strength, (D) CG catalase, (F) BV catalase.

Figure 5. Nearest neighbor distances (d′) between AuNPs and proteins. Center-to-center distances calculated for superlattices of varying DNA flexibility on the protein [0 spacers (red), 2 spacers (green), and 5 spacers (blue)], linker ratio, and salt concentration. The lattice symmetry is indicated by the shape of the icon; CsCl lattice (■) and Th3P4 (▲).

Figure 6. Superlattice transformations due to electrolyte concentration. Superlattices assembled and annealed at an initial salt concentration are transferred to a solution of a different ionic strength and allowed to equilibrate. These samples are slowly heated and cooled to study their reorganization: (A) lattices annealed at 0.25 M NaCl and transferred to 1 M NaCl, (B) annealed at 1 M NaCl and (C) 0.5 M NaCl, and transferred to 0.25 M NaCl.

E

DOI: 10.1021/acs.nanolett.7b02502 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

1-0280 and the Center for Bio-Inspired Energy Sciences (CBES), an Energy Frontiers Research Center (EFRC) funded by the US Department of Energy, award number DESC0000989. X-ray experiments were carried out at beamlines 12-ID-B and the Dupont-Northwestern-Dow Collaborative Access Team beamline at the Advanced Photon Source (APS), a U.S. DOE Office of Science User Facility 489 operated by Argonne National Laboratory under contract no. 490 DEAC02-06CH11357. Microtomy was performed by C. Wilkes at the NU Biological Imaging Facility supported by the NU Office for Research. M.X.W. was supported by an NSF Graduate Research Fellowship and a Ryan Fellowship. J.A.M. and S.E.S. were supported by the NU Center of Computation and Theory of Soft Materials. M.G. was supported by a Natural Sciences and Engineering Research Council of Canada (NSERC) graduate fellowship (grant PGS-D no. 6799-459278-2014).

density structure. Using this as a platform, we define a general strategy for tuning repulsion between DNA−NPs by harnessing the properties of DNA as a polyelectrolyte. Through structural and environmental variables, the effective diameter of the DNA ligands, their accessible volume, and surface density can be modulated to adjust steric repulsion. This provides new opportunities not only for making lower symmetry structures but also in designing dynamic materials. In principle, the effect of repulsion on DNA assembly could be modeled in a predictive fashion by calculating interparticle potentials. In this work, a protein serves as a highly uniform building block with a specific surface chemistry, directing the formation of lattices not readily accessible with traditionally functionalized inorganic NPs. Notably, the preferential orientation of the protein within the lattice suggests that the use of a protein with higher anisotropy could lead to perfectly oriented proteins or directional assembly of NPs. These results highlight the modular nature of DNA assembly using organic and inorganic building blocks and hint at the increased diversity of structures accessible by exploiting forces beyond Watson−Crick base pairing. We anticipate that these insights can be used to expand the structural and functional diversity of DNA-directed colloidal crystal assembly for the design of tunable materials.

■

■

ABBREVIATIONS DNA, deoxyribose nucleic acid; NP, nanoparticle; PAE, programmable atom equivalent; MD, molecular dynamics

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.nanolett.7b02502. 3D visualization of the Th3P4 unit cell (ZIP) Materials and methods, additional experiments, and coarse-grained MD simulation details (PDF)

■

REFERENCES

(1) Mirkin, C. A.; Letsinger, R. L.; Mucic, R. C.; Storhoff, J. J. Nature 1996, 382, 607−9. (2) Shevchenko, E. V.; Talapin, D. V.; Kotov, N. A.; O’Brien, S.; Murray, C. B. Nature 2006, 439, 55−59. (3) Min, Y.; Akbulut, M.; Kristiansen, K.; Golan, Y.; Israelachvili, J. Nat. Mater. 2008, 7, 527−538. (4) Bishop, K. J.; Wilmer, C. E.; Soh, S.; Grzybowski, B. A. Small 2009, 5, 1600−1630. (5) Jones, M. R.; Osberg, K. D.; Macfarlane, R. J.; Langille, M. R.; Mirkin, C. A. Chem. Rev. 2011, 111, 3736−3827. (6) Choi, C. L.; Alivisatos, A. P. Annu. Rev. Phys. Chem. 2010, 61, 369−389. (7) Boles, M. A.; Engel, M.; Talapin, D. V. Chem. Rev. 2016, 116, 11220−11289. (8) Seo, S. E.; Wang, M. X.; Shade, C. M.; Rouge, J. L.; Brown, K. A.; Mirkin, C. A. ACS Nano 2016, 10, 1771−1779. (9) Wang, M. X.; Seo, S. E.; Gabrys, P. A.; Fleischman, D.; Lee, B.; Kim, Y.; Atwater, H. A.; Macfarlane, R. J.; Mirkin, C. A. ACS Nano 2017, 11, 180−185. (10) Macfarlane, R. J.; Lee, B.; Jones, M. R.; Harris, N.; Schatz, G. C.; Mirkin, C. A. Science 2011, 334, 204−208. (11) Park, S. Y.; Lytton-Jean, A. K.; Lee, B.; Weigand, S.; Schatz, G. C.; Mirkin, C. A. Nature 2008, 451, 553−556. (12) Nykypanchuk, D.; Maye, M. M.; van der Lelie, D.; Gang, O. Nature 2008, 451, 549−552. (13) Jones, M. R.; Seeman, N. C.; Mirkin, C. A. Science 2015, 347, 1− 11. (14) Zhang, C.; Macfarlane, R. J.; Young, K. L.; Choi, C. H. J.; Hao, L.; Auyeung, E.; Liu, G.; Zhou, X.; Mirkin, C. A. Nat. Mater. 2013, 12, 741−746. (15) Zhang, Y.; Lu, F.; Yager, K. G.; van der Lelie, D.; Gang, O. Nat. Nanotechnol. 2013, 8, 865−872. (16) Macfarlane, R. J.; O’Brien, M. N.; Petrosko, S. H.; Mirkin, C. A. Angew. Chem., Int. Ed. 2013, 52, 5688−5698. (17) Mann, S. Angew. Chem., Int. Ed. 2008, 47, 5306−5320. (18) Dotan, N.; Arad, D.; Frolow, F.; Freeman, A. Angew. Chem., Int. Ed. 1999, 38, 2363−2366. (19) Padilla, J. E.; Colovos, C.; Yeates, T. O. Proc. Natl. Acad. Sci. U. S. A. 2001, 98, 2217−2221. (20) Ringler, P.; Schulz, G. E. Science 2003, 302, 106−109. (21) Ballister, E. R.; Lai, A. H.; Zuckermann, R. N.; Cheng, Y.; Mougous, J. D. Proc. Natl. Acad. Sci. U. S. A. 2008, 105, 3733−3738. (22) Cigler, P.; Lytton-Jean, A. K.; Anderson, D. G.; Finn, M.; Park, S. Y. Nat. Mater. 2010, 9, 918−922.

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected] (847-467-7302). *E-mail: [email protected] (630-252-0395). *E-mail: [email protected] (847-491-7801). ORCID

Byeongdu Lee: 0000-0003-2514-8805 Chad A. Mirkin: 0000-0002-6634-7627 Author Contributions

M.X.W. and J.D.B. contributed equally. M.X.W., J.D.B., C.A.M., and B.L. designed experiments. M.X.W. and J.D.B. collected experimental data. M.X.W., J.D.B., and B.L. performed data analysis and modeling. S.E.S. contributed to analysis. J.A.M., M.G., and M.O.d.l.C. designed and performed molecular dynamics simulations and analysis. All authors contributed to writing the manuscript and have given approval to the final version of the manuscript. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We thank J. Griffin for help with data visualization and analysis. The authors acknowledge support from the Vannevar Bush Faculty Fellowship program sponsored by the Basic Research Office of the Assistant Secretary of Defense for Research and Engineering and funded by the Office of Naval Research through grant N00014-15-1-0043. This material is also based upon work supported by the AFOSR under Award FA9550-12F

DOI: 10.1021/acs.nanolett.7b02502 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters (23) Sinclair, J. C.; Davies, K. M.; Vénien-Bryan, C.; Noble, M. E. Nat. Nanotechnol. 2011, 6, 558−562. (24) Brodin, J. D.; Ambroggio, X.; Tang, C.; Parent, K. N.; Baker, T. S.; Tezcan, F. A. Nat. Chem. 2012, 4, 375−382. (25) Lanci, C. J.; MacDermaid, C. M.; Kang, S.-g.; Acharya, R.; North, B.; Yang, X.; Qiu, X. J.; DeGrado, W. F.; Saven, J. G. Proc. Natl. Acad. Sci. U. S. A. 2012, 109, 7304−7309. (26) Gonen, S.; DiMaio, F.; Gonen, T.; Baker, D. Science 2015, 348, 1365−1368. (27) Kostiainen, M. A.; Hiekkataipale, P.; Laiho, A.; Lemieux, V.; Seitsonen, J.; Ruokolainen, J.; Ceci, P. Nat. Nanotechnol. 2012, 8, 52− 56. (28) Sakai, F.; Yang, G.; Weiss, M. S.; Liu, Y.; Chen, G.; Jiang, M. Nat. Commun. 2014, 5. (29) Brodin, J. D.; Smith, S. J.; Carr, J. R.; Tezcan, F. A. J. Am. Chem. Soc. 2015, 137, 10468−10471. (30) Suzuki, Y.; Cardone, G.; Restrepo, D.; Zavattieri, P. D.; Baker, T. S.; Tezcan, F. A. Nature 2016, 533, 369−373. (31) Brodin, J. D.; Auyeung, E.; Mirkin, C. A. Proc. Natl. Acad. Sci. U. S. A. 2015, 112, 4564−4569. (32) McMillan, J. R.; Brodin, J. D.; Millan, J. A.; Lee, B.; Olvera de la Cruz, M.; Mirkin, C. A. J. Am. Chem. Soc. 2017, 139, 1754−1757. (33) Macfarlane, R. J.; Jones, M. R.; Senesi, A. J.; Young, K. L.; Lee, B.; Wu, J.; Mirkin, C. A. Angew. Chem. 2010, 122, 4693−4696. (34) Tkachenko, A. V. Phys. Rev. Lett. 2002, 89, 148303. (35) Thaner, R. V.; Eryazici, I.; Macfarlane, R. J.; Brown, K. A.; Lee, B.; Nguyen, S. T.; Mirkin, C. A. J. Am. Chem. Soc. 2016, 138, 6119− 6122. (36) Kalsin, A. M.; Fialkowski, M.; Paszewski, M.; Smoukov, S. K.; Bishop, K. J.; Grzybowski, B. A. Science 2006, 312, 420−424. (37) Agard, N. J.; Prescher, J. A.; Bertozzi, C. R. J. Am. Chem. Soc. 2004, 126, 15046−15047. (38) Ning, X.; Guo, J.; Wolfert, M. A.; Boons, G. J. Angew. Chem., Int. Ed. 2008, 47, 2253−2255. (39) Auyeung, E.; Li, T. I.; Senesi, A. J.; Schmucker, A. L.; Pals, B. C.; de La Cruz, M. O.; Mirkin, C. A. Nature 2013, 505, 73−77. (40) Li, T. I.; Sknepnek, R.; Macfarlane, R. J.; Mirkin, C. A.; Olvera de la Cruz, M. Nano Lett. 2012, 12, 2509−2514. (41) Knorowski, C.; Burleigh, S.; Travesset, A. Phys. Rev. Lett. 2011, 106, 215501. (42) Thaner, R. V.; Kim, Y.; Li, T. I.; Macfarlane, R. J.; Nguyen, S. T.; Olvera de la Cruz, M.; Mirkin, C. A. Nano Lett. 2015, 15, 5545−5551. (43) Glaser, J.; Nguyen, T. D.; Anderson, J. A.; Lui, P.; Spiga, F.; Millan, J. A.; Morse, D. C.; Glotzer, S. C. Comput. Phys. Commun. 2015, 192, 97−107. (44) O’Brien, M. N.; Girard, M.; Lin, H.-X.; Millan, J. A.; de la Cruz, M. O.; Lee, B.; Mirkin, C. A. Proc. Natl. Acad. Sci. U. S. A. 2016, 113, 10485−10490. (45) Girard, M.; Millan, J. A.; Olvera de la Cruz, M. Annu. Rev. Mater. Res. 2017, 47, 33−49. (46) Senesi, A. J.; Lee, B. J. Appl. Crystallogr. 2015, 48, 1172−1182. (47) Li, T.; Senesi, A. J.; Lee, B. Chem. Rev. 2016, 116, 11128−11180. (48) Stigter, D. Biopolymers 1977, 16, 1435−1448. (49) Rybenkov, V. V.; Cozzarelli, N. R.; Vologodskii, A. V. Proc. Natl. Acad. Sci. U. S. A. 1993, 90, 5307−5311.

G

DOI: 10.1021/acs.nanolett.7b02502 Nano Lett. XXXX, XXX, XXX−XXX