Tunable Nanoscale Cages from Self-Assembling DNA and Protein

Mar 5, 2019 - Here, we report a cage constructed from both protein and DNA building blocks through the use of covalent protein–DNA conjugates...
0 downloads 0 Views 1MB Size
Subscriber access provided by Washington University | Libraries

Article

Tunable Nanoscale Cages from SelfAssembling DNA and Protein Building Blocks Yang Xu, Shuoxing Jiang, Chad R. Simmons, Raghu Pradeep Narayanan, Fei Zhang, Ann-Marie Aziz, Hao Yan, and Nicholas Stephanopoulos ACS Nano, Just Accepted Manuscript • Publication Date (Web): 05 Mar 2019 Downloaded from http://pubs.acs.org on March 5, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

Tunable Nanoscale Cages From Self-Assembling DNA and Protein Building Blocks Yang Xu1, Shuoxing Jiang1, Chad R. Simmons1, Raghu Pradeep Narayanan1,2, Fei Zhang1, AnnMarie Aziz1, Hao Yan1,2, Nicholas Stephanopoulos1,2* 1Biodesign

Center for Molecular Design and Biomimetics & 2School of Molecular Sciences, Arizona State University, Tempe, Arizona 85287, USA *email: [email protected]

ABSTRACT: Three-dimensional (3D) cages are one of the most important targets for nanotechnology. Both proteins and DNA have been used as building blocks to create tunable nanoscale cages for a wide range of applications, but each molecular type has its own limitations. Here, we report a cage constructed from both protein and DNA building blocks through the use of covalent protein-DNA conjugates. We modified a homotrimeric protein (KDPG aldolase) with three identical ssDNA handles by functionalizing a reactive cysteine residue introduced via sitedirected mutagenesis. This protein-DNA building block was co-assembled with a triangular DNA structure bearing three complementary arms to the handles, resulting in tetrahedral cages comprised of six DNA sides capped by the protein trimer. The dimensions of the cage could be tuned through the number of turns per DNA arm (3 turns ~ 10 nm, 4 turns ~ 14 nm), and the hybrid structures were purified and characterized to confirm the three-dimensional structure. Cages were

ACS Paragon Plus Environment

ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

also modified with DNA using click chemistry with using aldolase trimers bearing the noncanonical amino acid 4-azidophenylalanine, demonstrating the generality of the method. Our approach will allow for the construction of nanomaterials that possess the advantages of both protein and DNA nanotechnology, and find applications in fields such as targeted delivery, structural biology, biomedicine, and catalytic materials.

KEYWORDS: DNA nanotechnology, 3D nanocages, DNA-protein hybrids, DNA-protein bioconjuagation, aldolase trimer

Three-dimensional (3D) cages are one of the most important targets for nanotechnology, with applications including transporting cargo for imaging or drug delivery,1-4 confining catalysts in nanoscale reactors,5-7 positioning chemical elements to mimic photosynthesis,8, 9 and serving as programmable “molds” or amphiphiles for polymeric10-13 or inorganic14, 15 nanoparticle synthesis and encapsulation. Viruses like MS2 or CPMV represent one of nature’s solutions to biomolecular nano-cages, with multiple protein components self-assembling in a highly symmetrical fashion to form a closed-shell structure. Viral capsids have been employed extensively as nano-scale scaffolds for diverse applications16, 17 but have the disadvantage of being limited to naturallyoccurring structures and symmetries. Inspired by viral capsids, there has been extensive research in the re-design of non-assembling protein components to form self-assembled cages. For example, naturally oligomeric proteins have been engineered (often with the aid of software like Rosetta18) to self-assemble into highly symmetric structures such as tetrahedral,19 and megadalton scale cages.20 This elegant approach yields biologically relevant and often highly stable assemblies, but can require technical expertise in protein design that is challenging for the non-expert.

ACS Paragon Plus Environment

Page 2 of 33

Page 3 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

Furthermore, tuning the nano-cage parameters (such as the size and volume) usually requires reengineering the system with another set of protein building blocks, and it can be difficult to build highly anisotropic or asymmetrically-modified structures. As an alternative approach, DNA has been used as a building block to construct well-defined cages through the programmable assembly of individual strands by Watson-Crick pairing. A wealth of 3D structures have been reported, with designs using wireframe,21-25 tile-based,26-28 and densely packed arrangements of DNA helices.29-31 These nano-cages are tunable in both geometry and size, and can be functionalized by modifying the individually addressable component strands. Furthermore, software such as Cadnano32 and the low-cost availability of oligonucleotides from commercial suppliers results in easier access for the non-expert. This flexibility and accessibility comes at the expense of chemical homogeneity, where the final DNA nanostructures are restricted to the physical and chemical properties of the DNA duplex. As a result, DNA cages must be further elaborated with receptor-binding peptides or proteins to imbue them with bioactivity,33 and many structures require supra-physiological concentrations of magnesium for stability.34, 35 We envisioned that hybrid cages—merging self-assembling protein building blocks with addressable DNA scaffolds—could combine the bioactivity and chemical diversity of the former with the programmability of the latter. Towards this goal, we chose the synthesis of a simple polyhedron as a proof of principle: a three-dimensional tetrahedral cage containing a homotrimeric protein linked to a triangular “base” comprised entirely of DNA. In our design, the protein trimer is first functionalized with identical single-stranded DNA handles at a specific amino acid on each of the three monomers (Figure 1A). The triangular base bearing three complementary singlestranded DNA handles is self-assembled and purified separately via thermal annealing (Figure 1B). We reasoned that combining these two purified building blocks, and annealing at a lower

ACS Paragon Plus Environment

ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

temperature, would neither disrupt the protein-protein interface nor the DNA edges at the base of the structure, yielding a three-dimensional cage (Figure 1C). The size of this assembly could then be rationally tuned by changing the length of each DNA edge, whereas the protein would provide a scaffold for the attachment of small molecules, targeting peptides, or even fusion proteins. This hybrid structure is constructed through chemical conjugation of oligonucleotide handles on a protein building block, although several recent examples exist of one-dimensional nanofibers using covalent protein-DNA conjugates,36, 37 protein-DNA cages where the protein does not play a structural role,38 and non-covalent protein-DNA nanostructures.39, 40 Our approach will allow both a wide range of cages to be synthesized, or arrange other DNA-based materials such as tiles or other nanostructures (vide infra).

Results/Discussion For the trimeric protein building block, we selected 2-dehydro-3-deoxy-phosphogluconate (KDPG) aldolase (hereafter referred to as ald), a 25 kDa protein that forms a C3-symmetric homotrimer ald3 (Figure 1A).41-43 In addition to the high thermal stability of the trimer (~80 °C), ald is readily amenable to mutagenesis and recombinant expression in E. coli and has been used in protein nanocage engineering.44, 45 Furthermore, its size (6 nm diameter, 3 nm thickness, Figure 2A,B) is comparable with small oligonucleotide nanostructures like the four-strand DNA tetrahedron, first reported by Turberfield46 and employed extensively in DNA nanotechnology.3, 47, 48

Although ald is an enzyme, here we used it strictly as a structural building block, divorced

from its chemical functionality, similar to previous reports.44, 45 We selected Glu54, a solventexposed residue on the outer edge of the trimer, as a suitable site for modification with DNA. We mutated this residue to cysteine (E54C) in order to carry out thiol-selective chemistry with the

ACS Paragon Plus Environment

Page 4 of 33

Page 5 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

heterobifunctional crosslinker succinimidyl 3-(2-pyridyldithio)propionate (SPDP) and a 21-nt 5’amino modified oligonucleotide (Figure 2C).49 Wild-type ald possesses four cysteine residues, but these are involved in disulfide bonds in the protein interior (Supporting Information Figure S1A) so we reasoned that we could selectively target the mutagenically-introduced C54 on the surface (for details regarding the expression, purification, and characterization of the E54C mutant protein see Supporting Information Section S2). Following exposure to 6 equivalents of purified SPDP-DNA, a band with higher retention was observed by denaturing polyacrylamide gel electrophoresis (SDS-PAGE), corresponding to the ald monomer linked to the oligonucleotide, to approximately 50% yield (for reaction conditions and analysis see Supporting Information Section S3). Control experiments with wild-type protein did not show this upper band, suggesting that the native cysteines were unreactive under these conditions (Supporting Information Figure S3A). The mild coupling conditions were not expected to disrupt the protein-protein interactions holding the three monomers together, so we used anion exchange chromatography to isolate the ald trimer with three DNA handles (referred to hereafter as (ald-DNA)3) from trimers bearing two or fewer oligonucleotides (Supporting Figure S2). The complete modification of purified (ald-DNA)3 was verified by SDS-PAGE, which showed only a band corresponding to the DNA-modified ald monomer, with lower mobility than the unmodified ald (Figure 2D). We next probed whether the trimer-linked DNA handles could bind their complementary strand using native, non-denaturing PAGE (Figure 2E). The unmodified ald3 (lane 1) in Tris buffer has a net charge of -12.2 (pI = 6.9), so upon modification with three anionic DNA handles, it migrates farther towards the positive terminal of the gel (lane 2). Addition of ssDNA complementary to the handles results in a reduction of gel mobility due to the increased size of the complex (lane 3),

ACS Paragon Plus Environment

ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

whereas non-complementary DNA does not shift the (ald-DNA)3 band (lane 4). To verify that the increased mobility in lane 2 was not due to the trimer dissociating into monomers after DNA conjugation, we analyzed the samples by AFM (Figure 2F,G). Both unmodified ald3 and the (aldDNA)3 conjugate showed roughly circular structures with similar heights: ~3 nm, in good agreement from the predicted height measured from the crystal structure. The measured diameter of the ald3 (9.4 ± 0.25 nm (S.D., 30 particles counted); Supporting Information Figure S3) was slightly larger than expected from the crystal structure (due to AFM tip effects), but almost identical to the (ald-DNA)3 diameter (9.8 ± 0.26 nm), demonstrating that the trimer had not dissociated into monomers as a result of the DNA conjugation. Based on all these data, we were confident that we had produced pure samples of the ald trimer modified with three ssDNA handles, and that these handles were still functional for sequence-specific hybridization with a complementary oligonucleotide. Experiments with model DNA nanostructures. In order to test the ability of the trivalent (aldDNA)3 conjugate to organize DNA structures, we synthesized a model nanostructure that could be directly imaged by AFM. We designed a rigid tile consisting of two parallel DNA helices linked by four crossover points, with single-stranded regions for the (ald-DNA)3 handles to bind as a fifth crossover point (Figure 3A; for tile design and characterization see Supporting Information Section S4). This multi-crossover (“MX”) tile is 4 nm wide and 30 nm long, large enough to unambiguously image using AFM. We hypothesized that combining the tile with the (ald-DNA)3 conjugate would yield up to three tiles linked to the (ald-DNA)3 hub in a radial arrangement (Figure 3A). Indeed, annealing the preformed and purified MX tile with (ald-DNA)3 from 55 to 4 °C over 5 hours in 1 x TAE-Mg2+ buffer resulted in several distinct bands by native PAGE (Figure 3B), with a shift towards higher bands with increasing concentration of the tile relative to (ald-

ACS Paragon Plus Environment

Page 6 of 33

Page 7 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

DNA)3. We purified each of these bands from the gel and imaged them by AFM, and they corresponded to (in order of decreasing mobility): the unmodified MX tile (band i), and (aldDNA)3 bound to one, two, and three tiles, respectively (bands ii-iv) (Figure 3C). The length of the MX tiles in Figure 3C measured 31 ± 1.4 nm by AFM (Supporting Information Figure S4D, 30 particles counted), which is in good agreement with the design (30 nm). Because the DNA handles are chemically linked to ald at a single site via a hydrocarbon linker, the attachment is quite flexible, so the angles between the MX tiles and the protein vary greatly (Supporting Information Figure S4). However, the number of tiles could be readily visualized for each band, and the presence of a major band with three tiles at higher concentrations of the tile shows that the (aldDNA)3 does not dissociate even with several large DNA structures attached. We also highlight that the highly stable trimer protein was able to withstand annealing from 55 °C, making it attractive as a building block for these hybrid materials. Control experiments with mismatched handles resulted in no association of the protein with the tiles (Supporting Information Figure S4B), confirming that the binding was mediated by DNA hybridization. At lower temperatures, the MX tile association with (ald-DNA)3 was not efficient (Supporting Information Figure S4E), highlighting the importance of the homotrimer stability to higher temperatures. We also explored the binding of (ald-DNA)3 to a triangular DNA origami structure with a cavity bearing three complementary handles (Figure 3D and Supporting Information Section S5). Compared to the empty origami, annealing of (ald-DNA)3 to the scaffold resulted in a circular structure in the center, ~10 nm in diameter and 3 nm in height (Figure 3E,F and Supporting Information Figures S5B and S6), comparable to the dimensions of the (ald-DNA)3 (Figure 2G). Furthermore, in many images, the three DNA duplexes linking the trimer to the origami could be directly visualized (Figure 3E, zoom-in blue arrows), further confirming its ability to bind in a

ACS Paragon Plus Environment

ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

trivalent fashion. As with the MX tiles, control experiments with mismatched handles resulted in origami cavities lacking protein (Supporting Information Figure S5B). A trimeric protein like ald provides an attractive candidate for the scaffolding for the immobilization of other materials (including fusion proteins), so integrating it with DNA origami provides the potential for hierarchical organization of molecules. Potential applications for these multi-scale, hybrid biomolecular scaffolds including targeted delivery of medicinal species, light harvesting and artificial photosynthetic systems, or structural biology by using DNA origami as a fiducial marker to visualize proteins in techniques such as cryo-electron microscopy.50, 51 Self-assembly of hybrid tetrahedral cages. Confident that the (ald-DNA)3 could reliably bind three complementary oligonucleotide strands and organize large DNA assemblies, we next turned to making the hybrid tetrahedral protein-DNA cages outlined in Figure 1C. We designed a cage with three helical turns of DNA (31 bp) for each of its six edges, to yield a structure with predicted dimensions of 10 nm per side and 8 nm in height (Figure 4A). The base of the scaffold was constructed as previously shown in Figure 1B, with three turns of DNA per side and three arms consisting of one turn of dsDNA extended with three identical 21-nt ssDNA. Hybridization of these handles to the complementary handles on (ald-DNA)3 results in three helical turns for each of the three protein-linked edges in turn. We refer to this three-turn triangular base as 3t-Base (full design and sequences see Supporting Information Section S6). For comparison, we also designed an all-DNA tetrahedron with three helical turns edge, 3t-Tet (Supporting Information Figure S12A).33 Both 3t-Tet and 3t-Base were self-assembled via thermal annealing, purified by native PAGE (Figure 4B and Supporting Information Section S6), and imaged using AFM (Figure 4C and Supporting Information Figure S9). The images show well-formed, cage-like particles with a triangular base and six edges for 3t-Tet, with an average height of 4.7 ± 1.5 nm, and an average

ACS Paragon Plus Environment

Page 8 of 33

Page 9 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

edge length of 11.4 ± 1.2 nm (see Supporting Information Figure S12 for size distributions of all particles from AFM; for all samples, 300 particles were counted for height distributions and 30 particles for edge length distributions). The height measurement is slightly shorter than predicted from the structure design, indicating that the cages are flexible and deform upon adsorption to the mica surface, whereas the edge measurements correlated closely with the intended 10 nm design. AFM imaging of 3t-Base showed open triangular structures on the surface, with three short arms. These particles had a smaller height than 3t-Tet (3.9 ± 1.2 nm) but with a similar edge length of 12.0 ± 1.2 nm, as expected. Annealing of (ald-DNA)3 with the 3t-Base yielded a distinct, lower-mobility band by native PAGE (Figure 4B, and Supporting Information Figure S8), due to the larger molecular weight of the three-turn protein-DNA tetrahedron (3t-aldTet). AFM imaging (Figure 4C) showed compact structures with a white circular object at a vertex with straight edges and sharp corners, and in some images one or two of the arms connecting the top to the base can be directly visualized. The full triangular supporting DNA platform could not be seen due to the small size of the structures and partial occlusion by the protein. The average height for 3t-aldTet was 6.0 ± 1.9 nm, slightly larger than either 3t-Base or 3t-Tet, though still less than expected from the design, with an average edge length of 12.1 ± 1.3 nm as designed. We hypothesize that the 3t-aldTet particles are somewhat flexible and deform, or lie flat on the surface upon adsorption similar to 3t-Tet. The vast majority of particles in the purified 3t-aldTet sample were the desired cage structures, but a few particles were malformed, with only one or two arms of the triangular DNA anchor bound to the protein (Supporting Information Figure S9, S12). This could be due to co-migration of these structures with the fully assembled cage, or breakage of the cage during the processing steps involved in sample extraction from the gel.

ACS Paragon Plus Environment

ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

One of the key advantages of DNA, and a motivation for building hybrid materials, is the high degree of programmability that oligonucleotides afford. The dimensions of the DNA cage can be tuned independently of the protein component by extending the length of the constituent strands, whereas in an all-protein cage an entirely different building block would have to be used. To demonstrate the tunability of the DNA scaffold, we designed a larger tetrahedral cage with four helical turns of per side, which can assemble from a triangular base with arms containing two full turns of DNA followed by the 21-nt handles (Figure 4D and Supporting Information Section S6). This structure has predicted dimensions of 14 nm (width) and 12 nm (height), ~40% larger than the three-turn variant. The sequences of these handles were designed so that we could use the same (ald-DNA)3 conjugate as for the three-turn system. Once again, we designed a four-turn all-DNA tetrahedron (4t-Tet) to compare with the triangular base (4t-Base) and the protein-DNA cage (4taldTet). All samples were prepared as indicated previously (Figure 4E,F and Supporting Information Figure S10, S13). The 4t-Tet samples showed the expected tetrahedral structure with dimensions suggesting flexibility and deformation on the surface (average height 5.8 ± 1.7 nm, average edge length 15.2 ± 1.3 nm). The structure, with four individual triangular cavities comprising the sides, and all six distinct edges are more evident with these larger cages compared with the 3t-Tet. Similarly, the 4t-Base consisted of a larger triangular structure with three longer arms (average height 3.0 ± 1.3 nm, average edge length 15.3 ± 1.0 nm). Annealing (ald-DNA)3 with 4t-Base produced a lowermobility band when visualized by AFM yielded structures containing a triangular base with a circular structure that corresponded to the protein. The average height of these particles was noticeably larger than 4t-Tet (13.4 ± 2.7 nm), and closer to the expected value of 12 nm from the design. Unlike 3t-aldTet, which appears to collapse on the mica surface, the 4t-aldTet seems to

ACS Paragon Plus Environment

Page 10 of 33

Page 11 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

be more rigid and resists this deformation. This result could be because the protein in the larger cages is farther from the mica and thus does not interact with it as strongly, but more experiments are needed to determine the origin of the difference. The average edge length was 16.0 ± 1.0 nm, quite close to the 14 nm expected from the design. We also note that the base of the cage, with the protein at the apex of the structure, is more evident in 4t-aldTet due to the larger size of the DNA structure relative to the protein. Similar to the three-turn system, the majority of the particles in the 4t-aldTet structure were well-formed, with a minor fraction of broken or malformed structures. For both the three-turn and four-turn structures, control experiments with triangular bases bearing mismatched sequences to the (ald-DNA)3 handles showed no cage formation by PAGE (Supporting Information Figure S8). Although the images in Figure 4C,F were encouraging that we had formed the desired proteinDNA cages, their small size precluded unambiguous confirmation of the structure by AFM. Specifically, we wanted to confirm two things: (1) that the structures possessed three doublestranded arms connecting the base to the protein trimer, and (2) that the triangular base was still intact. To investigate the first point, we synthesized three-turn triangular bases bearing only one or two ssDNA handles complementary to the trimer, with the remaining arms consisting of a single, blunt-ended helical turn (Supporting Information Figure S14). When annealed with (aldDNA)3 these structures did not yield structures like 3t-aldTet as evidenced by either PAGE or AFM. Rather, two or three distinct triangular bases attached to a central protein hub were seen instead (Supporting Information Figure S14C,D), providing indirect evidence that all three arms of the base were bound to the protein in 3t-aldTet. To further confirm that all three arms were bound to the protein, we exposed 3t-aldTet to the reducing agent tris(2-carboxyethyl)phosphine (TCEP), Figure 5. The DNA handles on (ald-DNA)3 are linked to the protein by a disulfide bond,

ACS Paragon Plus Environment

ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

so reduction of this linkage should yield a structure with three full turns of double stranded DNA per arm (Figure 5A), as opposed to 3t-Base, which has only a single turn of dsDNA per arm. Indeed, treating 3t-aldTet with TCEP yielded a band by native PAGE (Figure 5B, lane 5), with a retention between the bands for 3t-aldTet (lanes 2 and 4) and 3t-Base (lane 4). This band had the same mobility as a sample of 3t-Base to which we added complementary oligonucleotides to the 21-nt handles in order to make them fully double-stranded (lane 3). Purification of the band in lane 5 from and imaging it by AFM clearly showed that it yielded the expected three-edge triangular platform (Figure 5C), and the arms with three full helical turns appear longer than the 3t-Base arms in Figure 4C. Taken together, these results suggest that the 3t-aldTet samples do indeed consist of a triangular base linked to the (ald)3 by three double-stranded DNA arms. We note that a faint lower band can be seen in both lanes 3 and 5 (marked by a red arrow), which is the 3t-Base with two double-stranded and one single-stranded arm. For lane 5, this band intensity corresponds to a minor fraction (~10%) of cages where only two of the arms are bound to the protein. Site-specific DNA conjugation using non-canonical amino acids. Although the thiol-selective chemistry we used for modification of ald is highly effective when site-specific bioconjugation with native amino acids is required, many proteins have endogenous reactive cysteines that preclude this approach. Furthermore, in our experiments we found that the E54C ald3 mutant was prone to aggregation due to spontaneous disulfide formation between Cys54 residues. We were able to circumvent this issue by reducing the expressed protein with dithiothreitol (DTT) prior to conjugation, but the reaction conditions had to be carefully tuned to avoid breakage of the two internal disulfide bonds, which in turn led to over-modification with DNA. For a full discussion of protein aggregation and optimization of reduction conditions, see Supporting Information Section S7. To avoid these issues, we asked whether we could modify ald3 with DNA using a

ACS Paragon Plus Environment

Page 12 of 33

Page 13 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

chemical reaction completely orthogonal to native amino acids. For this purpose, we selected the strain-promoted azide-alkyne cycloaddition (SPAAC), more commonly known as copper-free click chemistry.52 This reaction is selective, efficient, and has been used to conjugate DNA to peptides52-54 and proteins55 successfully. We introduced the non-canonical amino acid 4-azidophenylalanine (azF) into ald, creating the E54(azF) mutant via the Schultz amber codon suppression method56, 57 (Figure 6A; for details of protein expression and characterization see Supporting Information S7). The incorporation of azF did not affect the self-assembly of the protein into a homotrimer, and the purified protein was completely free from the aggregation seen in the E54C mutant (Supporting Information Figure S18A). We reacted the E54(azF) ald3 with a 21-nt DNA strand functionalized with a dibenzocyclooctyne (DBCO) moiety (Figure 6B), isolated the (ald-DNA)3 conjugate by anion exchange chromatography, and verified its complete conversion by SDS-PAGE (Figure 6C). For details regarding the DNA-DBCO synthesis and characterization, conjugation to E54(azF) ald3, and purification of the (ald-DNA)3 conjugate, see Supporting Information Section S7. Upon annealing with 3t-Base, the E54(azF) (ald-DNA)3 yielded a band with the same mobility as the original 3t-aldTet samples (Figure 6E), and analysis of the purified band by AFM showed cages similar to those in Figure 4C. The average height of these particles was 9.0 ± 2.7 nm, and the average edge length was 11.4 ± 1.1 nm (Supporting Information Figure S17). We note that with the exception of the chemical linkage and minor differences in the linkers between the protein and DNA, the E54C and E54(azF) (ald-DNA)3 conjugates are identical so it was not surprising that they would form highly similar three-dimensional hybrid cages. However, we were curious as to whether we could move the conjugation site of the DNA and still obtain well-formed cages. Modifying the attachment site of DNA should affect the angle of the dsDNA edges emanating

ACS Paragon Plus Environment

ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

from the protein, and might allow for the synthesis of cages with different sizes or symmetries as demonstrated using DNA hierarchical assembly.28, 58 We thus designed a mutant, E74(azF), that positioned the reactive residues closer together on one surface of the trimer as opposed to the outer edge (Figure 6A). We reasoned that this site would reduce the angle between the three edges of the cage and potentially lead to more efficient cage formation. The E74(azF) version of (ald-DNA)3 was synthesized and purified similar to the E54(azF) conjugate (Supporting Information Section S7). When annealed with 3t-Base, the E74(azF) mutant of (ald-DNA)3 gave a higher-retention band by native PAGE, similar to the E54(azF) mutant (Figure 6D). However, contrary to our expectation, the vast majority of isolated structures (~95%, Supporting Info Figure S17B) were not well-formed, closed cages. Rather, the E74(azF) (ald-DNA)3 primarily bound to a single arm of 3t-Base, resulting in open structures where both the trimer and the triangular base with two unbound arms are visible (Figure 6F). A number of structures where the (ald-DNA)3 has two arms attached to a single base, or one arm attached to each of two different bases, can also be seen. The upper bands in the gel are most likely protein with two 3t-Base structures bound, or oligomers of several proteins and DNA structures. The inefficiency in cage formation with the E74(azF) mutant is most likely due to the steric or electrostatic repulsion incurred by trying to confine the three dsDNA arms in a small area, though we cannot rule out entropic constraints imposed by the linkers. Overall, our results demonstrate that DNA-protein cage formation can be accomplished using a variety of chemical reactions, and that the conjugation site can be moved around on the protein surface—perhaps aided by computational simulation to determine the optimal placement—to influence the assembly of the final nanostructure.

ACS Paragon Plus Environment

Page 14 of 33

Page 15 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

Conclusions We have demonstrated the construction of hybrid nanoscale cages composed of both selfassembling protein and DNA components. This nanoscale cage integrates both proteins and DNA as distinct and integral structural elements through the use of well-defined protein-DNA covalent conjugates. The DNA scaffold allows for rational tuning of the cage size, and will in the future permit the attachment of additional functionality with nanoscale control by elaborating the constituent strands. The protein, by contrast, will provide an alternative molecular scaffold to oligonucleotides for functionalizing these cages. We are particularly excited by the prospect of fusing other proteins to the trimer scaffold, allowing for multivalent targeting and cargo delivery. In addition, it should be possible to extend our design to construct hybrid protein-DNA cages with protein oligomers at every vertex, similar to the cages by the Baker lab20 or naturally-occurring viral capsids.59 These larger, highly symmetric structures will be particularly attractive for structural biology applications, such as in cryo-electron microscopy (cryo-EM). The rigid, highly symmetric particles can display multiple orientations of the protein on a single particle, aiding in particle identification, averaging, and structural determination of the fused protein.60-62 Although this work focused on the KDPG aldolase trimer, a wealth of homo- and heterooligomeric protein building blocks exist naturally, or have been rationally designed.63 We also note that the current design required self-assembly at 55 °C; future investigations will tune the chemistry, linker length, and nanostructure design to allow structure formation at lower temperatures (e.g. 37 °C), and enable less stable protein assemblies to be used. Our approach will find broad applicability in the construction of more complex hybrid materials, especially in conjunction with larger DNA scaffolds (e.g. three-dimensional DNA origami).22, 29, 30 To achieve this, it will necessitate modification of the protein scaffold in more than one location with

ACS Paragon Plus Environment

ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

oligonucleotide handles in order to control the angular orientation and rigidity of the protein-DNA nanostructure interface. This goal will in turn require multiple site-specific bioconjugation reactions, and enable integrated protein-DNA nanomaterials with the advantages of each molecular class.

Methods/Experimental Protein expression. The gene for KDPG aldolase was inserted into the pet28b+ plasmid using standard molecular cloning techniques, and then used to transform E. coli BL21(DE3) competent cells (New England Biolab Inc. USA) for protein expression. All expression constructs described herein were verified by DNA sequencing subsequent to cloning. The E54C mutant was prepared using a primer with the mutated nucleotides in the center of the forward and reverse primers, and introduced using the Q5 Site-Directed Mutagenesis Kit following the kit protocol (New England Biolab Inc. USA). The mutated plasmid was then transformed into E. coli BL21(DE3) cells. For the generation of aldolase protein containing the non-canonical amino acid 4-azidophenylalanine (azF) (Bachem Americas Inc. USA), we introduced the amber (TAG) stop codon at either position 54 or 74 using a QuikChange Site-Directed mutagenesis kit (Stratagene, USA). The plasmids containing the amber mutation at position 54 or 74 were derived from the pEt28- kdgA parent plasmid. Using site directed mutagenesis, a single point mutation was made by replacing an existing amino acid in the aldolase with the TAG codon. PCR primers were designed to have 1021 bp of homology on either side of the TAG mutation. Each set of primers was used to amplify the aldolase amber mutation using pEt28-kdgA plasmid as the template. The amplified PCR product was digested with Dpn1 to yield the mutant plasmid. The BL21(DE3) cells that were previously transformed with both the native and E54C mutation were used for protein expression. Cells were grown in 1X Terrific Broth (TB) at 37°C

ACS Paragon Plus Environment

Page 16 of 33

Page 17 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

containing 100 µg/mL ampicillin to an OD600 = 0.8, then induced with 0.5 mM IPTG (isopropylβ-D-thiogalactopyranoside) (SigmaAldrich, USA) at 37°C for 4 hours, and then harvested by centrifugation at 6000 RPM for 20 minutes. For the non-canonical amino acid incorporation, two plasmids were required, one for expression of the aldolase gene containing the amber stop codon and a second containing the orthogonal tRNA and aminoacyl-tRNA synthetase pair. Both the pEt28- kdgA (E54azF) and pDule2 plasmids were co-transformed into BL21(DE3) cells and grown 1X Terrific Broth (TB) at 37°C containing supplemented with 100 µg/ml ampicillin and 50 µg/ml spectinomycin) (VWR, USA). Once the cells had reached OD600=0.8, the non-canonical amino acid (azF) was added to the growth media to a final concentration of 1 mM and protein expression was induced with 0.5 mM IPTG and 0.02% L-arabinose (VWR, USA). Cells were shaken overnight at 37°C, and then harvested by centrifugation. Protein purification and characterization. Following protein expression, the harvested cell pellets were resuspended in lysis buffer containing 20 mM Tris pH 8.0, 150 mM NaCl, 10 mM imidazole, 0.1 µM EDTA, and lysed using sonication. After sonication, the cell lysate was centrifuged at 13,000 rpm for 30 min to remove cell debris. The supernatant was transferred onto a Ni-NTA 5ml HisTrap HP column (GE Healthcare, USA) and rinsed with 20 mM Tris pH 8.0, 300 mM NaCl, 10 mM imidazole to remove non-specifically bound species. The His-tagged aldolase was isolated with a 20 column volume linear gradient of lysis buffer with elution buffer containing 20 mM Tris pH 8.0, 150 mM NaCl, 500 mM imidazole. Fractions were pooled and placed in 30 kDa molecular weight cutoff dialysis membranes and dialyzed overnight at 4 °C against 1L of PBS buffer (10 mM Na2HPO4, 1.8 mM KH2HPO4, 137 mM NaCl, 2.7 mM KCI pH = 7.6). The aldolase protein was subjected to size exclusion chromatography with a Superdex Increase 75 10/300 GL column (GE Healthcare, USA). The peak fractions were then analyzed via

ACS Paragon Plus Environment

ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

SDS-PAGE and fractions corresponding to the pure aldolase were combined and concentrated using an Amicon-30 kDa (Millipore Sigma, USA) cutoff filter out small-molecular weight impurities. Fractions were further characterized for purity and mass using MALDI-TOF on an ABI4800-TOF/TOF mass spectrometer. AFM imaging. Samples (2 µL) were deposited on freshly cleaved mica (Ted Pella, Inc. USA), followed by addition of 60 µl of 1 x TAE buffer containing 45 mM Tris pH 8.0, 12.5 mM Mg(OAc)2·6H2O, and 2 mM EDTA with 2 µl of 10 mM NiCl2 (hereafter referred to as “1 x TAEMg2+”). Samples were allowed to adsorb at room temperature for 5 min, and then scanned in a tapping mode on a Pico-Plus AFM (Molecular Imaging, Agilent Technologies) with NP-S tips (Veeco, Inc. USA). All images were collected under ambient condition in tapping mode using the AFM. Width and height profiles were determined from horizontal line scans. Synthesis of SPDP-DNA and DBCO-DNA conjugates. All single-stranded DNA oligonucleotides used for conjugation were purchased from Integrated DNA Technologies, Inc. (www.idtdna.com). SPDP was used to crosslink aldolase to an amine-modified oligo DNA using a modification of a previously described protocol.46 500 µL of 400 µM 5’-amine-modified DNA oligo (5’-AmMC6-TGAGTTCCGTCAGGTCTGCTC-3’) in 1xPBS (pH 7.6) was combined with 20 equivalents of 50 mM SPDP ((N-succinimidyl 3-(2-pyridyldithio) propionate (Thermo Scientific, USA) in DMSO. The mixture was shaken for one hour at room temperature and then purified by reverse phase HPLC to remove free DNA and excess SPDP. Following SPDP modification of DNA, the conjugate was separated on a Zorbax Eclipse 5 XDB-C18 column (150 x 4.6mm). 80 µL of the SPDP-DNA solution was purified via HPLC with an XDB-C18 column using an elution gradient from 10% to 60% methanol in 50mM TEAA (triethylammonium acetate)

ACS Paragon Plus Environment

Page 18 of 33

Page 19 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

(Thermo Fisher Scientific, USA). DNA was modified with DBCO using a modification of a previously described protocol.50 The amine-modified oligonucleotide was dissolved in 1 x PBS and combined with 20 equivalents of DBCO-sulfo-NHS (100 mM in DMSO), and the mixture was shaken overnight at room temperature. The DBCO-DNA conjugate was purified by HPLC to remove free DNA and excess DBCO. Protein modification and characterization. Prior to conjugation, the purified E54C aldolase protein was treated with 10 equivalents of 100 mM DTT (1,4-dithiothreitol) for 30 min at 4°C. A 30 kDa molecular weight cutoff (MWCO) Amicon filter was used to wash the protein 3 times with PBS to remove excess DTT. To the reduced protein solution was added 6 equivalents of the purified SPDP-DNA conjugate. The reaction mixture was shaken gently overnight at 4°C. The excess SPDP-DNA was removed by filtration using a 30 kDa MWCO filter and washed twice with 20 mM Tris buffer. Following SPDP-DNA conjugation to Cys54, the solution contained a mixture of trimer bearing between 0-3 DNA strands. To isolate the aldolase trimer containing 3 oligonucleotides ((ald-DNA)3), the reaction products were purified by anion exchange chromatography using a Mono Q 4.6/100 PE column (GE Healthcare, USA).

As shown in

Supporting Information Figure S2A, three peaks from the chromatogram were collected. The fractions corresponding to each peak were concentrated using 30 kDa MWCO filters and analyzed by SDS-PAGE (Supporting Information Figure S2B). For the click conjugation, freshly purified E54(azF) or E74(azF) aldolase samples were mixed with 6 equivalents of purified DBCO-DNA and shaken gently overnight at room temperature. Protein-DNA conjugates were purified using the same protocol as the thiol-mediated conjugation. MX tile assembly with (ald-DNA)3. All strands comprising the MX tile were mixed together in 1 x TAE-Mg2+ buffer. The tiles were formed by thermally annealing the mixture from 90 - 20

ACS Paragon Plus Environment

ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

°C over 2 hours. The annealed MX tiles were purified from a native PAGE gel, then mixed with 2 µl of 5 µM aldolase-DNA in in 1 x TAE-Mg2+ buffer. The MX tile and (ald-DNA)3 mixture was annealed from 55 - 4 °C over 5 hours, followed by purification by native PAGE electrophoresis using standard elution procedures in 1 x TAE-Mg2+. Triangular origami assembly with (ald-DNA)3. All strands for the triangle origami structure were combined in 1 x TAE-Mg2+ buffer and the mixture was annealed from 90 - 20 °C over 2 hours. Origami design and purification followed a previous report64. In order to attach the (aldDNA)3 samples, three staple strands were replaced with staples extended with complementary handles (see Supporting Information Figure S5, numbering same as in the reference).64 Protein-DNA cage assembly. All component strands of the tetrahedron were mixed together in an equimolar ratio in 1 x TAE-Mg2+ buffer. DNA samples (the Tet and Base samples) were formed by annealing the oligo mixtures from 90 °C to 20 °C for 2 hours, and then purified via size exclusion chromatography using a Superdex Increase 75 10/300 GL column. To form the aldTet samples, the corresponding Base samples were diluted 10-fold and combined with 2 µL of 5 µM aldolase-DNA in 1 x TAE-Mg2+ buffer, followed by annealing from 55 - 4 °C over 5 hours. The annealed aldTet samples were purified using the same procedure for the MX tiles described above. For the TCEP cleavage described in Figure 5, a 0.3 M TCEP solution was prepared and adjusted to pH = 7.0 with sodium hydroxide. The purified 3t-aldTet was incubated with 20 mM TCEP for 1 hour at room temperature in 1 x TAE-Mg2+ buffer. ASSOCIATED CONTENT Supporting Information Available: DNA sequences and designs; Protein expression, modification, and characterization; Design and characterization of MX tiles, origami, and protein-

ACS Paragon Plus Environment

Page 20 of 33

Page 21 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

DNA cages; Non-canonical amino acid protein design and characterization. This material is available free of charge via the Internet at http://pubs.acs.org. AUTHOR INFORMATION Corresponding Author Email: [email protected] Author Contributions The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. ACKNOWLEDGEMENT N.S. gratefully acknowledges startup funds from Arizona State University. This material is based upon work supported by the Air Force Office of Scientific Research under award number FA955017-1-0053. The authors thank Prof. Matthew B. Francis for providing the plasmids for noncanonical amino acid incorporation.

ACS Paragon Plus Environment

ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 33

i

Figure 1: Hybrid protein-DNA cage design. A) Modification of a homotrimer aldolase (ald3) protein building block at a site-specific reactive residue with ssDNA handles. B) Self-assembly of four different DNA strands to yield a triangular base with three complementary handles to the protein. C) Annealing of the protein-DNA conjugate and triangular base will result in a tetrahedral cage comprised of both the protein and the DNA structural units. The dimensions of the cage are tunable by changing the length of the DNA strands in the triangular base.

ACS Paragon Plus Environment

Page 23 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

Figure 2: Protein modification and characterization. A) Underside view of ald3 showing the three Cys54 residues in yellow. B) Side view of ald3. C) Strategy for modifying the protein with DNA using thiol-specific chemistry. D) Denaturing SDS-PAGE of protein and protein-DNA conjugate. Lane M: protein molecular weight ladder; 1: unmodified ald3; 2: purified (ald-DNA)3. E) Native PAGE of protein samples. Lane 1: unmodified ald3; 2: (ald-DNA)3; 3: (ald-DNA)3 + complementary DNA; 4: (ald-DNA)3 + mismatched DNA. F,G) AFM imaging and height profiles ald3 and (ald-DNA)3 in order to confirm that the trimers were intact and had not dissociated into monomers. Scale bar: 25 nm.

ACS Paragon Plus Environment

ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 3: Protein assembly with DNA nanostructures. A) Self-assembly of (ald-DNA)3 with multi-crossover (MX) DNA tiles bearing complementary handles. B) Native PAGE of (ald-DNA)3 and MX tile association. C) AFM images of indicated bands, isolated from the gel. Scale bar: 20 nm. D) Design of triangular DNA origami with three complementary handles to (ald-DNA)3. E, F) AFM images and height profiles along the white dotted lines of origami before and after protein association. Scale bar: 50 nm

ACS Paragon Plus Environment

Page 24 of 33

Page 25 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

Figure 4: Protein-DNA tetrahedral cages. A) Design of three-turn protein-DNA tetrahedron (3taldTet). B) Native PAGE of purified 3t-Tet (lane 1), 3t-Base (lane 2), 3t-aldTet (lane 3). C) AFM images of 3t-Tet, 3t-Base, and 3t-aldTet. D) Design of four-turn protein-DNA tetrahedron (4taldTet). E) Native PAGE of 4t-Tet (lane 1), 4t-Base (lane 2), 4t-aldTet (lane 3). F) AFM images of 4t-Tet, 4t-Base, and 4t-aldTet. Scale bars for AFM images: 15 nm (zoom-ins), 100 nm (zoomouts).

ACS Paragon Plus Environment

ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 5: Confirmation of 3D cage structure. A) Schematic of TCEP cleavage of 3t-aldTet. Reduction of the disulfide linkage between the protein and DNA should result in a base with double-stranded DNA arms if all three are bound to the protein. B) Native PAGE of 3t-Base (lane 1), 3t-aldTet (lane 2, 4), 3t-Base + complementary handles (lane 3), and 3t-aldTet following TCEP cleavage (lane 5). C) AFM imaging of major band isolated from lane 5 (scale bar: 10 nm)

ACS Paragon Plus Environment

Page 26 of 33

Page 27 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

Figure 6: Non-canonical amino acid cages. A) Location of azidophenylalanine residues in the two mutants used. B) SPAAC reaction between DNA-DBCO and ald containing (azF). C, D) Denaturing PAGE of the E54(azF) and E74(azF) mutants, respectively. In both gels lane 1 is the unmodified protein, and lane 2 is the (ald-DNA)3 conjugate. E, F) Native PAGE and AFM images of the structures formed by the E54(azF) and E74(azF) (ald-DNA)3 conjugates, respectively, with 3t-Base. The AFM images correspond to the bands in the red box. Scale bars: 25 nm.

ACS Paragon Plus Environment

ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Table of contents graphic.

ACS Paragon Plus Environment

Page 28 of 33

Page 29 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

REFERENCES 1. Liu, X. T.; Wu, L. N.; Wang, L.; Jiang, W., A Dual-Targeting DNA Tetrahedron Nanocarrier for Breast Cancer Cell Imaging and Drug Delivery. Talanta 2018, 179, 356-363. 2. Walsh, A. S.; Yin, H. F.; Erben, C. M.; Wood, M. J. A.; Turberfield, A. J., DNA Cage Delivery to Mammalian Cells. ACS Nano 2011, 5, 5427-5432. 3. He, L.; Lu, D. Q.; Liang, H.; Xie, S. T.; Luo, C.; Hu, M. M.; Xu, L. J.; Zhang, X. B.; Tan, W. H., Fluorescence Resonance Energy Transfer-Based DNA Tetrahedron Nanotweezer for Highly Reliable Detection of Tumor-Related mRNA in Living Cells. ACS Nano 2017, 11, 40604066. 4. Wang, S.; Xia, M. C.; Liu, J.; Zhang, S. C.; Zhang, X. R., Simultaneous Imaging of Three Tumor-Related mRNAs in Living Cells with a DNA Tetrahedron-Based Multicolor Nanoprobe. ACS Sens. 2017, 2, 735-739. 5. Zhao, Z.; Fu, J.; Dhakal, S.; Johnson-Buck, A.; Liu, M.; Zhang, T.; Woodbury, N. W.; Liu, Y.; Walter, N. G.; Yan, H., Nanocaged Enzymes With Enhanced Catalytic Activity and Increased Stability Against Protease Digestion. Nat. Commun. 2016, 7, 10619. 6. Lu, N.; Pei, H.; Ge, Z. L.; Simmons, C. R.; Yan, H.; Fan, C. H., Charge Transport within a Three-Dimensional DNA Nanostructure Framework. J. Am. Chem. Soc. 2012, 134, 13148-13151. 7. Simmons, C. R.; Schmitt, D.; Wei, X. X.; Han, D. R.; Volosin, A. M.; Ladd, D. M.; Seo, D. K.; Liu, Y.; Yan, H., Size-Selective Incorporation of DNA Nanocages into Nanoporous Antimony-Doped Tin Oxide Materials. ACS Nano 2011, 5, 6060-6068. 8. Stephanopoulos, N.; Carrico, Z. M.; Francis, M. B., Nanoscale Integration of Sensitizing Chromophores and Porphyrins with Bacteriophage MS2. Angew. Chem. Int. Ed. 2009, 48, 94989502. 9. Miller, R. A.; Presley, A. D.; Francis, M. B., Self-Assembling Light-Harvesting Systems from Synthetically Modified Tobacco Mosaic Virus Coat Proteins. J. Am. Chem. Soc. 2007, 129, 31043109. 10 Trinh, T.; Liao, C. Y.; Toader, V.; Barlog, M.; Bazzi, H. S.; Li, J. N.; Sleiman, H. F., DNAImprinted Polymer Nanoparticles With Monodispersity and Prescribed DNA-Strand Patterns. Nat. Chem. 2018, 10, 184-192. 11. Wilks, T. R.; Bath, J.; de Vries, J. W.; Raymond, J. E.; Herrmann, A.; Turberfield, A. J.; O'Reilly, R. K., "Giant Surfactants" Created by the Fast and Efficient Functionalization of a DNA Tetrahedron with a Temperature-Responsive Polymer. ACS Nano 2013, 7, 8561-8572. 12. Edwardson, T. G. W.; Carneiro, K. M. M.; McLaughlin, C. K.; Serpell, C. J.; Sleiman, H. F., Site-Specific Positioning of Dendritic Alkyl Chains On DNA Cages Enables Their GeometryDependent Self-Assembly. Nat. Chem. 2013, 5, 868-875. 13. Chidchob, P.; Edwardson, T. G. W.; Serpell, C. J.; Sleiman, H. F., Synergy of Two Assembly Languages in DNA Nanostructures: Self-Assembly of Sequence-Defined Polymers on DNA Cages. J. Am. Chem. Soc. 2016, 138, 4416-4425. 14. Sun, W.; Boulais, E.; Hakobyan, Y.; Wang, W. L.; Guan, A.; Bathe, M.; Yin, P., Casting Inorganic Structures with DNA Molds. Science 2014, 346, 1258361.

ACS Paragon Plus Environment

ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

15. Bhatia, D.; Arumugam, S.; Nasilowski, M.; Joshi, H.; Wunder, C.; Chambon, V.; Prakash, V.; Grazon, C.; Nadal, B.; Maiti, P. K.; Johannes, L.; Dubertret, B.; Krishnan, Y., Quantum Dot-Loaded Monofunctionalized DNA Icosahedra for Single-Particle Tracking of Endocytic Pathways. Nat. Nanotechnol. 2016, 11, 1112-1119. 16. Stephanopoulos, N.; Tong, G. J.; Hsiao, S. C.; Francis, M. B., Dual-Surface Modified Virus Capsids for Targeted Delivery of Photodynamic Agents to Cancer Cells. ACS Nano 2010, 4, 60146020. 17. Dedeo, M. T.; Finley, D. T.; Francis, M. B., Viral Capsids as Self-Assembling Templates for New Materials. Prog. Mol. Biol. Transl. Sci. 2011, 103, 353-392. 18. Das, R.; Baker, D., Macromolecular Modeling with Rosetta. Ann. Rev. Biochem. 2008, 77, 363-382. 19. King, N. P.; Bale, J. B.; Sheffler, W.; McNamara, D. E.; Gonen, S.; Gonen, T.; Yeates, T. O.; Baker, D., Accurate Design of Co-Assembling Multi-Component Protein Nanomaterials. Nature 2014, 510, 103-108. 20. Bale, J. B.; Gonen, S.; Liu, Y. X.; Sheffler, W.; Ellis, D.; Thomas, C.; Cascio, D.; Yeates, T. O.; Gonen, T.; King, N. P.; Baker, D., Accurate Design of Megadalton-Scale Two-Component Icosahedral Protein Complexes. Science 2016, 353, 389-394. 21. Chen, J. H.; Seeman, N. C., Synthesis From DNA of a Molecular With The Conncectivity of a Cube. Nature 1991, 350, 631-633. 22. Han, D.; Pal, S.; Nangreave, J.; Deng, Z.; Liu, Y.; Yan, H., DNA Origami with Complex Curvatures in Three-Dimensional Space. Science 2011, 332, 342-346. 23. Zhang, F.; Jiang, S.; Wu, S.; Li, Y.; Mao, C.; Liu, Y.; Yan, H., Complex Wireframe DNA Origami Nanostructures with Multi-Arm Junction Vertices. Nat. Nanotechnol. 2015, 10, 779-785. 24. Benson, E.; Mohammed, A.; Gardell, J.; Masich, S.; Czeizler, E.; Orponen, P.; Hogberg, B., DNA Rendering of Polyhedral Meshes at The Nanoscale. Nature 2015, 523, 441-444. 25. Aldaye, F. A.; Sleiman, H. F., Modular Access To Structurally Switchable 3D Discrete DNA Assemblies. J. Am. Chem. Soc. 2007, 129, 13376-13377. 26. Jiang, S. X.; Hong, F.; Hu, H. Y.; Yan, H.; Liu, Y., Understanding the Elementary Steps in DNA Tile-Based Self-Assembly. Acs Nano 2017, 11, 9370-9381. 27. Shi, X. L.; Lu, W.; Wang, Z. Y.; Pan, L. Q.; Cui, G. Z.; Xu, J.; LaBean, T. H., Programmable DNA Tile Self-Assembly Using a Hierarchical Sub-Tile Strategy. Nanotechnology 2014, 25, 075602. 28. He, Y.; Ye, T.; Su, M.; Zhang, C.; Ribbe, A. E.; Jiang, W.; Mao, C., Hierarchical SelfAssembly of DNA into Symmetric Supramolecular Polyhedra. Nature 2008, 452, 198-202. 29. Rothemund, P. W. K., Folding DNA to Create Nanoscale Shapes and Patterns. Nature 2006, 440, 297-302. 30. Douglas, S. M.; Dietz, H.; Liedl, T.; Hoegberg, B.; Graf, F.; Shih, W. M., Self-Assembly of DNA into Nanoscale Three-Dimensional Shapes. Nature 2009, 459, 414-418. 31. Ke, Y.; Ong, L. L.; Shih, W. M.; Yin, P., Three-Dimensional Structures Self-Assembled from DNA Bricks. Science 2012, 338, 1177-1183.

ACS Paragon Plus Environment

Page 30 of 33

Page 31 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

32. Douglas, S. M.; Marblestone, A. H.; Teerapittayanon, S.; Vazquez, A.; Church, G. M.; Shih, W. M., Rapid Prototyping of 3D DNA-Origami Shapes with caDNAno. Nucleic Acids Res. 2009, 37, 5001-5006. 33. Lee, H.; Lytton-Jean, A. K. R.; Chen, Y.; Love, K. T.; Park, A. I.; Karagiannis, E. D.; Sehgal, A.; Querbes, W.; Zurenko, C. S.; Jayaraman, M.; Peng, C.G.; Charisse, K. Borodovsky, A.; Manoharan, M.; Donahoe, J.S.; Truelove, J.; Nahrendorf, M.; Langer, R.; Anderson, D.G., Molecularly Self-Assembled Nucleic Acid Nanoparticles for Targeted In Vivo SiRNA Delivery. Nat. Nanotechnol. 2012, 7, 389-393. 34. Mei, Q. A.; Wei, X. X.; Su, F. Y.; Liu, Y.; Youngbull, C.; Johnson, R.; Lindsay, S.; Yan, H.; Meldrum, D., Stability of DNA Origami Nanoarrays in Cell Lysate. Nano Lett. 2011, 11, 14771482. 35. Hahn, J.; Wickham, S. F. J.; Shih, W. M.; Perrault, S. D., Addressing the Instability of DNA Nanostructures in Tissue Culture. ACS Nano 2014, 8, 8765-8775. 36. McMillan, J. R.; Mirkin, C. A., DNA-Functionalized, Bivalent Proteins. J. Am. Chem. Soc. 2018, 140, 6776-6779. 37. Kashiwagi, D.; Sim, S.; Niwa, T.; Taguchi, H.; Aida, T., Protein Nanotube Selectively Cleavable with DNA: Supramolecular Polymerization of "DNA-Appended Molecular Chaperones J. Am. Chem. Soc. 2018, 140, 26-29. 38. Erben, C. M.; Goodman, R. P.; Turberfield, A. J., Single-Molecule Protein Encapsulation In A Rigid DNA Cage. Angew. Chem. Int. Ed. 2006, 45, 7414-7417. 39. Praetorius, F.; Dietz, H., Self-Assembly Of Genetically Encoded DNA-Protein Hybrid Nanoscale Shapes. Science 2017, 355, eaam5488. 40. Jiang, T.; Meyer, T. A.; Modlin, C.; Zuo, X. B.; Conticello, V. P.; Ke, Y. G., Structurally Ordered Nanowire Formation from Co-Assembly of DNA Origami and Collagen-Mimetic Peptides. J. Am. Chem. Soc. 2017, 139, 14025-14028. 41. Mavridis, I. M.; Hatada, M. H.; Tulinsky, A.; Lebioda, L., Structure of 2-Keto-3-Deoxy-6Phosphogluconate Aldolase at 2.8 Å Resolution. J. Mol. Biol. 1982, 162, 419-444. 42. Fullerton, S. W. B.; Griffiths, J. S.; Merkel, A. B.; Cheriyan, M.; Wymer, N. J.; Hutchins, M. J.; Fierke, C. A.; Toone, E. J.; Naismith, J. H., Mechanism of the Class I KDPG aldolase. Bioorg. Med. Chem. 2006, 14, 3002-3010. 43. Bruun, T. U. J.; Andersson, A. M. C.; Draper, S. J.; Howarth, M., Engineering a Rugged Nanoscaffold To Enhance Plug-and-Display Vaccination. ACS Nano 2018, 12, 8855-8866. 44. 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.; Baker, D., Design of a Hyperstable 60-Subunit Protein Icosahedron. Nature 2016, 535, 136-139. 45. Patterson, D. P.; Desai, A. M.; Holl, M. M. B.; Marsh, E. N. G., Evaluation of a SymmetryBased Strategy for Assembling Protein Complexes. RSC Adv. 2011, 1, 1004-1012. 46. Goodman, R. P.; Schaap, I. A. T.; Tardin, C. F.; Erben, C. M.; Berry, R. M.; Schmidt, C. F.; Turberfield, A. J., Rapid Chiral Assembly of Rigid DNA Building Blocks for Molecular Nanofabrication. Science 2005, 310, 1661-1665.

ACS Paragon Plus Environment

ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

47. Li, Z.; Wei, B.; Nangreave, J.; Lin, C.; Liu, Y.; Mi, Y.; Yan, H., A Replicable Tetrahedral Nanostructure Self-Assembled from a Single DNA Strand. J. Am. Chem. Soc. 2009, 131, 1309313098. 48. Sadowski, J. P.; Calvert, C. R.; Zhang, D. Y.; Pierce, N. A.; Yin, P., Developmental SelfAssembly of a DNA Tetrahedron. ACS Nano 2014, 8, 3251-3259. 49. Dutta, P. K.; Levenberg, S.; Loskutov, A.; Jun, D.; Saer, R.; Beatty, J. T.; Lin, S.; Liu, Y.; Woodbury, N. W.; Yan, H., A DNA-Directed Light-Harvesting/Reaction Center System. J. Am. Chem. Soc. 2014, 136, 16618-16625. 50. Bai, X. C.; Martin, T. G.; Scheres, S. H. W.; Dietz, H., Cryo-EM Structure of a 3D DNAOrigami Object. Proc. Natl. Acad. Sci. U.S.A. 2012, 109, 20012-20017. 51. Martin, T. G.; Bharat, T. A. M.; Joerger, A. C.; Bai, X. C.; Praetorius, F.; Fersht, A. R.; Dietz, H.; Scheres, S. H. W., Design Of A Molecular Support for Cryo-EM Structure Determination. Proc. Natl. Acad. Sci. U.S.A. 2016, 113, E7456-E7463. 52. Baskin, J. M.; Prescher, J. A.; Laughlin, S. T.; Agard, N. J.; Chang, P. V.; Miller, I. A.; Lo, A.; Codelli, J. A.; Bertozzi, C. R., Copper-Free Click Chemistry For Dynamic In Vivo Imaging. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 16793-16797. 53. Stephanopoulos, N.; Freeman, R.; North, H. A.; Sur, S.; Jeong, S. J.; Tantakitti, F.; Kessler, J. A.; Stupp, S. I., Bioactive DNA-Peptide Nanotubes Enhance the Differentiation of Neural Stem Cells Into Neurons. Nano Lett. 2015, 15, 603-609. 54. Freeman, R.; Stephanopoulos, N.; Alvarez, Z.; Lewis, J. A.; Sur, S.; Serrano, C. M.; Boekhoven, J.; Lee, S. S.; Stupp, S. I., Instructing Cells With Programmable Peptide DNA Hybrids. Nat. Commun. 2017, 8, 15982. 55. Marth, G.; Hartley, A. M.; Reddington, S. C.; Sargisson, L. L.; Parcollet, M.; Dunn, K. E.; Jones, D. D.; Stulz, E., Precision Templated Bottom-Up Multiprotein Nanoassembly through Defined Click Chemistry Linkage to DNA. ACS Nano 2017, 11, 5003-5010. 56. Chin, J. W.; Santoro, S. W.; Martin, A. B.; King, D. S.; Wang, L.; Schultz, P. G., Addition Of p-azido-L-Phenylalanine to the Genetic Code of Escherichia coli. J. Am. Chem. Soc. 2002, 124, 9026-9027. 57. Liu, C. C.; Schultz, P. G., Adding New Chemistries to the Genetic Code. Ann. Rev. Biochem. 2010, 79, 413-444. 58. Wang, P. F.; Wu, S. Y.; Tian, C.; Yu, G. M.; Jiang, W.; Wang, G. S.; Mao, C. D., Retrosynthetic Analysis-Guided Breaking Tile Symmetry for the Assembly of Complex DNA Nanostructures. J. Am. Chem. Soc. 2016, 138, 13579-13585. 59. Zlotnick, A.; Aldrich, R.; Johnson, J. M.; Ceres, P.; Young, M. J., Mechanism of Capsid Assembly for an Icosahedral Plant Virus. Virology 2000, 277, 450-456. 60. Walls, A. C.; Tortorici, M. A.; Bosch, B. J.; Frenz, B.; Rottier, P. J. M.; DiMaio, F.; Rey, F. A.; Veesler, D., Cryo-Electron Microscopy Structure of a Coronavirus Spike Glycoprotein Trimer. Nature 2016, 531, 114-117. 61. Halldorsson, S.; Li, S.; Li, M. Q.; Harlos, K.; Bowden, T. A.; Huiskonen, J. T., Shielding and Activation of a Viral Membrane Fusion Protein. Nat. Commun. 2018, 9, 349.

ACS Paragon Plus Environment

Page 32 of 33

Page 33 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

62. Kirchdoerfer, R. N.; Cottrell, C. A.; Wang, N. S.; Pallesen, J.; Yassine, H. M.; Turner, H. L.; Corbett, K. S.; Graham, B. S.; McLellan, J. S.; Ward, A. B., Pre-Fusion Structure of a Human Coronavirus Spike Protein. Nature 2016, 531, 118-121. 63. Huang, P. S.; Boyken, S. E.; Baker, D., The Coming of Age of De Novo Protein Design. Nature 2016, 537, 320-327. 64. Ding, B.; Deng, Z.; Yan, H.; Cabrini, S.; Zuckermann, R. N.; Bokor, J., Gold Nanoparticle Self-Similar Chain Structure Organized by DNA Origami. J. Am. Chem. Soc. 2010, 132, 3248– 3249.

ACS Paragon Plus Environment