Using Modular Preformed DNA Origami Building Blocks to Fold Dynamic 3D Structures

Thesis

Presented in Partial Fulfillment of the Requirements for the Degree Master of Science in the Graduate School of The Ohio State University

Gunter Eickert, B.S.

Graduate Program in Mechanical Engineering

The Ohio State University

2014

Thesis Committee: Carlos Castro, Advisor Haijun Su Copyright by

Gunter Erick Eickert

2014 Abstract

DNA origami is a bottom-up approach that takes advantage of DNA’s structure and lock key sequencing to build nano scale machines and structures. By introducing specific single stranded DNA (ssDNA) “staples”, a ssDNA “scaffold” can be folded into a desired 2D or 3D structure. Using this method, a variety of shapes have been formed, including tetrahedrons and octahedrons. These structures have been explored as drug carriers, enzyme platforms, signal markers, and for other medical and research uses.

However, each of these structures must be specifically designed and often pertain to only one particular application.

A DNA origami structure that functions as a building block was designed to allow for the creation of several different 3D structures without the need for a lengthy design process. An annealing process called a thermal ramp was then used to fold the structures.

The building block was made of four equilateral triangles arranged into a parallelogram.

This is ideal since a parallelogram can be used to form three of the five platonic solids, in addition to many non-platonic shapes. The parallelogram was successfully folded into a tetrahedron and an octahedron by introducing staples that bound to overhangs on its edges. By utilizing strand displacement, the 3D structures were able to be unfolded back into building blocks and then refolded into new 3D shapes. All of these foldings and unfoldings were performed at room temperature, without the need for a thermal ramp.

ii Each structure was verified using transmission electron microscopy. The ability to switch between a tetrahedron and an octahedron with the same parallelograms suggests that the parallelograms could also be used to form other structures, such as an icosahedron.

The parallelogram building block is both modular and dynamic. These features could allow the structure to be used for carrying and releasing a drug and could make it possible to control the kinetics of enzymes attached to the DNA structure. Both of these applications can be explored since the parallelogram can take on multiple conformations and can be switched between them without the need for a complete redesign.

iii Acknowledgments

I would like to thank Dr. Castro for all of his advice and guidance. I would also like to thank all the students in the Nanoengineering and Biodesign Lab and the

OhioMOD team for their help in my research.

iv Vita

2007...... Gibsonburg High School

2012...... B.S. Biomedical Engineering, The Ohio State University

2014...... M.S. Mechanical Engineering, The Ohio State University

Fields of Study

Major Field: Mechanical Engineering

v Table of Contents

Abstract...... ii

Acknowledgments...... iv

Vita...... v

List of Figures...... vii

Chapter 1: An Introduction to DNA Origami...... 1

Chapter 2: Design of a Modular and Dynamic Structure...... 19

Chapter 3: Results...... 32

Citations...... 47

vi List of Figures

Figure 1. Structure of DNA...... 2

Figure 2. The Holliday Junction...... 3

Figure 3. DNA Origami Cube...... 4

Figure 4. 3D Tiled DNA Origami Canvases...... 5

Figure 5. 2D Tiled DNA Origami Images...... 7

Figure 6. 3D Tiled DNA Origami Canvases...... 8

Figure 7.3D Tiled DNA Origami Images...... 9

Figure 8. Scaffolded DNA Origami Explanation...... 10

Figure 9. Modular Scaffolded DNA Origami Images...... 10

Figure 10. Modular Scaffolded DNA Origami Using a Framework...... 11

Figure 11. 3D Scaffolded DNA Origami Explanation...... 12

Figure 12. Scaffolded DNA Polymer and Icosahedron...... 13

Figure 13. Strand Displacement...... 14

Figure 14. Dynamic Scaffolded DNA Frame...... 16

Figure 15. DNA Origami Enzyme Example Complexes...... 18

Figure 16. A Parallelogram and Platonic Solids it Can Form...... 21

Figure 17. Parallelogram Scaffold Routing...... 22

Figure 18. caDNAno User Interface...... 23

vii Figure 19. Complete Form 2 and Form 3 Parallelagram Design Schematics...... 24

Figure 20. Folding a Parrallelogram into a Tetrahedron...... 26

Figure 21. Folding Two Parallelograms into an Octahedron...... 27

Figure 22. Transforming a Form 2 Parallelogram into a Form 3 Parallelogram...... 28

Figure 23. Strand Displacement Kinetics...... 29

Figure 24. Experiment Design...... 31

Figure 25. Initial Design and Images of a Parallelogram...... 33

Figure 26. Folding Initial Design into Tetrahedrons and Unfolding with Heat...... 34

Figure 27. Agarose Gel of Final Structures...... 36

Figure 28. Images of Folding Sequence Starting with Octahedrons...... 37

Figure 29. Images of Folding Sequence Starting with Tetrahedrons...... 38

Figure 30. How a Parallelogram can Fold onto Itself...... 40

Figure 31. Ways to Interpret TEM Images of Tetrahedrons and Octahedrons...... 42

Figure 32. Outlined TEM Images...... 43

Figure 33. How to Fold an Icosahedron...... 45

Figure 34. An Eight Sided Non-platonic Solid...... 46

Figure 35. A DNA Origami Delivery System...... 47

viii Chapter 1: An Introduction to DNA Origami

The ability to design and build nanoscale structures and machines is an idea that

has the potential to impact every aspect of our lives. Nanostructures are being explored

to improve solar energy harvesting[1] and to carry drugs to more effectively combat tumors[2]. Nano- and microfabrications are created using either a top-down method (by

which a structure is made as the result of removing specific material from an existing

bulk material) or a bottom-up method (by which a structure is assembled from individual

smaller parts). An example of the top-down method in use today is photolithography, a

process used to fabricate the chips made of nanoscale structures that ushered our world

into the modern era of computers. Life, on the other hand, the most abundant user of

nanoscale structures and machines, uses a bottom-up approach. Both proteins and DNA,

which define a large portion of our biology, are synthesized using a bottom-up method -

proteins being made from amino acids and DNA being made of nucleic acids.

It should come as no surprise, then, that the bottom-up assembly found in

biology has been the inspiration for a wide range of nanotechnology. However, it may be

surprising that a relatively new and promising nanotechnology does not use the robust

and diverse protein, and instead employs DNA.

DNA is comprised of four distinct nucleotides (Adenine, Thymine, Guanine,

Cytosine) that are chained together to form a single-stranded polymer. Watson-Crick

1 Figure 1. Structure of DNA: DNA is made up of nucleotides that prefer to pair up, as shown on the right. Shown on the left are two single strands of DNA that coil around each other to form a double helix when their complimentary base pairs bond.[5]

base pairing specifies that Adenine (A) prefers to bind with Thymine (T) and Guanine (G)

prefers to bind with Cytosine (C). Two complementary single strands of DNA (ssDNA)

will assemble by base-pairing to form a double helix[3]. This double helix has well-defined dimensions and twist, being ~2.2 nm in diameter, ~.33 nm per base pair (BP), and making a complete rotation every 10.5 BP.[4] Figure 1, taken from Seeman, depicts the structure of

DNA.[5]

Seeman first proposed using this selective base pairing to form designed structures

in the 1980s.[6] While genomic DNA is made of several pairs of ssDNA that bind with

2 only their complement, ssDNA can also bind with several other ssDNA strands at the

same time, creating a junction or branch. Figure 2 is an example of a Holliday junction,

a type of DNA branching that Seeman proposed using to build structures out of DNA.

[6] The top left of Figure 2 from Eichman shows how four separate ssDNA connect to

each other to form a Holliday junction with 4 branches.[7] Each ssDNA is made up of domains that are not complementary to the same strand. This causes one of the domains

Figure 2. The Holliday Junction: At the top left, a Holliday junction is formed from four ssDNA, each of which have domains that bond to two other ssDNA. The double helical structure is maintained as shown in the bottom left of the figure. The four branches of the Holliday junctions can lay relatively parallel to each other, as shown at the right of the figure.[7]

3 of ssDNA A to pair to ssDNA B and another to pair to ssDNA D. The bottom right of

Figure 2 shows how the twist of the double helix is maintained in a Holliday junction. The right of Figure 2 shows how it is possible to have the same Holliday junction’s 4 branches lay relatively parallel to one another.[7,8,9] Junctions with different numbers of branches have also been created, such as 3-[10], 5-, and 6-branch junctions.[11] It was also shown that multiple Holliday junctions could be used in one structure.[12,13,14]

By utilizing multiple Holliday junctions to form several junctions, a diverse number of 2D and 3D structures can be formed or “folded”, using a method called DNA origami. Structures that have been formed using this method include cubes[15], DNA structure polymers[16 ,17], tetrahedrons[18-26], octahedrons[27], icosahedrons[28], and other 2D and 3D forms[29,30]. The particular method used above to form these structures is often referred to as “tiled DNA origami”, since many tile-like short ssDNA are used. Figure

3, altered from Seeman, shows an early example of a cube folded using 3D tiled DNA origami. The cube is made of many short strands of ssDNA that form the double helix edges of the cube. The corners of the cube are 3-way junctions. The cube was built up

Figure 3. DNA Origami Cube: The cube’s edges are double stranded DNA and each corner is a 3-way junction. The structure was created in a stepwise manner, adding a few ssDNA at a time. The final step was to ligate A and B to A’ and B’ respectively. [15]

4 piecewise, by only adding a few strands in each step. Figure 3 shows the last step in forming the cube.[15]

An annealing reaction known as a thermal ramp is performed in DNA origami to produce proper folding of the structures. The thermal ramp heats the solution of ssDNA and then gradually cools the solution. The solution is cooled in a stepwise manner, through which the solution is cooled slightly, reheated, and then cooled a little further.

This is repeated until the desired structure is believed to have been formed. Since longer ssDNA domains bind at a higher heat, the ssDNA is sequentially assembled as the solution is cooled in the cyclic manner described. [5,6] The use of ssDNA, Holliday junctions, and the thermal ramp developed into what is now called DNA origami.

In more recent years, the tiled DNA origami method has been successfully used

Figure 4. 2D Tiled DNA Origami Canvases: a: Each 42 base single stranded tile is comprised of 4 domains. b: A tile connects to 4 other tiles to form a pixel. The canvas is formed when all tiles are connected together . c: Gaps can be formed by removing tiles from the annealing reaction, resulting in complex pixeled shapes. Half strands line the edges to prevent any undesired connections. d: The half tiles lining opposite edges can be combined to form a full tile, resulting in the canvas being folded into a tube.[29] 5 to create both 2D motifs and 3D detailed structures.[29, 30] Figure 4, modified from Wei,

illustrates the use of tiled DNA origami to create 2D shapes. Wei et. al. used ssDNA

tiles that were each comprised of 4 domains, as shown in Figure 4a.[29] Each domain

on a tile connected four other tiles. Each tile connected to four other tiles was called a

pixel. Multiple pixels were chained together in a folding reaction to form a canvas, as

seen in Figure 4b. The tiles were designed so that each domain in the tile was either 10

or 11 nucleotides long. Since a double helix completes a full rotation every 10.5 bases,

each domain completed a full 360o rotation. This allowed all of the pixels of the canvas to

lay in the same plane. The tiles that made up the full canvas were predetermined, but by

removing tiles from the folding reaction, pixels could be removed from the canvas. As a

result the canvas appeared as a particular 2D shape, as shown in Figure 4c. This method

of removing pixels from a predesigned canvas could be considered a form of top-down

fabrication. The design process to form new shapes was greatly simplified by removing

pieces from a predesigned structure. Wei further simplified the design process by writing

a program which could scan an image and automatically determine which staples in the

canvas should not be included in the thermal ramp folding. Over 100 different shapes

were successfully designed and folded. Figure 5, taken from Wei, shows 100 atomic force

microscopy (AFM) images of these shapes.[29]

Ke et al. folded 3D shapes and objects using a similar procedure as above, which

is illustrated in Figure 6 taken from Ke.[30] However, instead of ssDNA tiles with 10 to 11 nucleotide long domains, ssDNA bricks were made of four 8 nucleotide long domains were used, as seen in Figure 6A. Like the tiles, each brick connected to four other breaks

6 Figure 5. 2D Tiled DNA Origami Images: Numbers, letters, emoticons, and symbols pro- duced using single stranded tiles. Images were obtained using AFM.[29]

via its four domains. Since each domain was 8 bases long, it only completed about 75% of a full double helix twist. This resulted in adjacent bricks oriented at 90-degree angles relative to one another, as shown in Figure 6B. This orientation between adjacent bricks allowed for 3D structures to be built. Each connection between 2 domains formed a voxel

7 Figure 6. A: 3D Tiled DNA Origami Canvases: Each 32 base single stranded brick is comprised of 4 domains. B: A tile connects to another brick to form a voxel. Each brick lays at 90-degree angles to adjacent bricks, allowing for 3D structure. B: All tiles connected together form the 3D canvas. D: The 3D canvas can also be represented with LEGO-like bricks. E: By selecting only a subset of the total number of bricks that make up a canvas to react together, different structures can be made. F: This top-down approach to designing a bottom-up system can allow for creating diverse structures with ease.[30]

(a 3D pixel). These voxels were then used to create a 3D canvas shown in Figure 6 C & D.

By selectively removing bricks from the folding reaction, a shape was created as shown in

Figure 6 E & F.[30]

A program written by Ke was created to automatically select the appropriate bricks to create a 3D shapes. Figure 7 from Ke shows 102 folded structures imaged by transmission electron microscopy (TEM).[30] This method presents a modular system that

8 Figure 7. 3D Tiled DNA Origami Images: Shown are 100 different shapes created from a 3D canvas made of 1000 voxels.[30]

shortened the design process between different shapes. This kind of modular system is

preferred, since it promotes consistency and reduces error between shapes. It produces a

static structure, however, and not a dynamic nanomachine.

In 2006, Rothemund introduced an alternative approach to fold DNA called

“scaffolded DNA origami”. In scaffolded DNA origami, a single long DNA sequence,

called a scaffold, is folded with short DNA strands, called staples. Each staple connects to

two or more sections of the scaffold, pulling them adjacent to each other.[31, 32, 33] Since the staple moves from one section of the scaffold to another section via a Holliday junction,

9 Figure 8. Scaffolded DNA Origami Explanation: On the left is a ribbon diagram showing the scaffold, in black and gray, being bound by several colored staples. On the right is the same shape in a line diagram, showing the scaffold in black being bound by the colored staples.[31]

Figure 9. Modular Scaffolded DNA Origami Images: Shown are TEM images of triangular structures with overhangs that allow them to bind to each other.[31]

10 the connection is called a crossover. Figure 8, modified from Rothemund, illustrates a scaffold bound by staples.[31]

When Rothemund developed this method in 2006, he used a 7,249 nucleotide long ssDNA sequence from the M13mp18 virus as the scaffold. This scaffold was favorable since it was easily duplicated and evolved to minimize secondary structures. By having the staples cross over every 16 bases (approximately 1.5 turns of a double helix) along the scaffold, flat 2D shapes were folded from the DNA. Several DNA origami shapes were also combined into larger structures as shown in Figure 9, modified from Rothemund.

Figure 10. Modular Scaffolded DNA Origami Using a Framework: DNA Origami tiles are folded using a scaffold and staples. A separate pre-formed scaffold frame is also folded. When these two components are added together, the tiles interact with the frame and form a larger structure.[34]

11 [31] These larger structures were built from triangles, seen in Figure 9 j-m. Each triangle had unbound overhang staples that extended from half of each of the triangle’s sides, as shown in Figure 9 n-o. The other half of each side were missing staples. This allowed the overhang staples of one side to bind with another triangle whose side was missing the overhang sequence. Figure 9 p-u shows several examples of 2D structures formed from the triangles.[31]

Zhao also formed larger structures from preformed individual DNA origami

tile structures.[34] Figure 10 shows how the larger structures are assembled. The tiles

interact with a preformed scaffold framework that pulls them into the desired form

instead of the interacting directly with each other.[34] This combination of structures offers

the opportunity to make larger, more complex structures, especially if the number of

Figure 11. 3D Scaffolded DNA Origami Explanation: The scaffold, in gray, is folded on top of itself using colored staples. This results in a helices of DNA that forms a honeycomb structure.[32]

12 Figure 12. Scaffolded DNA Polymer and Icosahedron: A scaffold DNA origami structure is polymerized by incorporating ssDNA overhangs that bind to unpaired scaffold sites on other structures (a). A structure made up of 6 helix bundle struts made from the same scaffold (b). Three distinct forms of the structure were designed. Each has staple overhangs on eight of its struts that bind to another structures’ overhangs (c). After the three individual sections have been folded, they are combined in another reaction to form an icosahedron.[32]

13 connections can be limited and controlled.

Scaffolded DNA origami has also been used to create 3D structures, such as

tetrahedrons[35-38],octahedrons[39], cubes[40, 41], icosahedrons[32], and others[38, 42]. Figure 11

from Douglas shows how this is achieved.[32] By designing staples to cross from one part

of a scaffold to another every 7 bases, each crossover will occur every 240o of the double helix. This causes the structure to have a honeycomb form as seen in figure 9 c.[32]

Figure 12 shows how 3D structures can also be formed by connecting several

preformed structures.[32] In Figure 12a, several structures were polymerized together by including overhang staples that can bind with overhang staples on other structures. This

Figure 13. Strand Displacement: A: Two ssDNA strands are connected and form the double helix b. One of the strands has a toehold overhang h*. B: Strand X can bind to h* and compete for binding with b. It will eventually win out, resulting in strand Y being separated from b and the new structure L being complete.[43]

14 caused the structures to chain together and form a polymer. Figure 12 b-d shows how a

icosahedron was formed from 3 pieces. Each piece had overhang staples on its ends that

would bind the ends together, forming the icosahedron.[32]

Several examples have been presented that show a building block made of DNA

used to build a more complex structure.[29-32] The early examples focused on short strands of ssDNA as the building blocks to make 2D and 3D structures. Scaffolded DNA origami was then shown to build base building block structures that could be combined to form a larger, more complex structure. The design of these structures was greatly simplified by building them from repeatable structures. However, the structures presented thus far have been static and non-functional.

One way to have a dynamic DNA structure is to induce a conformational change

in the structure by removing staples from the scaffold. This can be done with DNA

strand displacement as shown in Figure 13 taken from Winfree.[43] Strand displacement

is when one strand in a double helix is replaced by another. As seen in Figure 13 A, this

is achieved with the aid of a toehold (h*), which is a ssDNA sequence that extends from

the structure (b) and does not bind to any part of the structure. When a displacement

staple is introduced to its complimentary toehold, it first binds to the toehold as

illustrated in Figure 13 B. Due to normal thermodynamic fluctuations, the edge of

the staple’s connection to the scaffold will disconnect and reconnect, as shown by the

double arrows in Figure 13 B. However, if the displacement staple is attached to the toe

hold, the displacement staple and the sequence of b will compete for will compete for

the disconnected nucleotide on X. Since the displacement staple is firmly rooted via the

15 toehold, it will eventually win. As they compete back and forth, the displacement staple

will bind to more and more of X and eventually fully bind with it, disconnecting it from

b.[43-46]

Toehold strand displacement is a powerful tool that can be used to change the

conformation of DNA nanostructures and even reverse the effect by reintroducing the

previously displaced staples. Yan et al. illustrated this in Figure 14 with a hollow DNA

square structure.[46] Staples with toeholds were used so that portions of the square’s frame could be sequentially unbound and set into new forms using new staples. The frame

Figure 14. Dynamic Scaffolded DNA Frame: a: A frame structure was designed with toeholds to allow reconfiguration of the structure. b: Unzipping staples were added, then closure staples were added to lock it into a new form. c: Different unzipping staples and closure staples were added to deform the structure even further. This process was reversed to yield the original shape, as indicated by the red arrow.[46]

16 was transformed from a square frame shown in Figure 14 a so that two corners of the

frames inner border were bent inwards as shown in Figure 14 b. This structure was then

transformed into Figure 14 c. The process was reversed, returning the structure first to its

form in Figure 14 b and then to its original form in 14 a.[46]

The combination of reversible conformational changes as described above and

modular building blocks as discussed previously could be a powerful tool. I will outline

the design of such a system that uses a modular building block to build 3D structures that

are reconfigurable and reversible. Many DNA origami structures[31,32] require a lengthy

design process, so a preformed building block could allow the formation of new shapes

without the need of designing an entirely new thermal ramp reaction. Many dynamic

structures can only switch between a few predesigned forms[28,42,46], while the ones designed to be very modular are not geared towards a functional application.[29,30,34] The proposed building block design has the potential to be combined into an undetermined number of dynamic and reconfigurable 3D shapes, which could be applied to a variety of applications.

One application others are exploring for 3D DNA origami structures is as

containers for drug delivery. Nanoparticles have been trapped in tetrahedrons,[47]

octahedrons, [48] and icosahedrons.[28] DNA origami structures have also been triggered to open in a variety of ways. Tetrahedrons have been dissembled and reassembled based on

PH[25], tetrahedrons have been opened with photons,[18] cubes have been opened by strand displacement,[40] and DNA cage structures have been triggered to open with an analite.[42]

DNA origami structures have also been shown to improve the cancer drug daunorubicin’s

17 uptake into drug resistant cancer cells.[21, 49]

Another therapeutic application for DNA origami structures being explored is as

part of a biomolecule detection system.[20, 22, 23, 50] DNA origami structures have also been

used to study and control the association and kinetics between proteins and enzymes.

Figure 15 shows several examples where this has been done.[51] These applications are an

Figure 15. DNA Origami Enzyme Example Complexes: A: A linear NAD(P) H:FMN(NFOR) oxidoreductase and luciferase reaction assembled on a DNA scaffold. B: A structure to test the distance dependence of the kinetics between BMR reductase domain and the BMP porphyrin domain cascade. The angle of the structure could be set to different angles to control the distance between the two proteins. C: The dimensions of a hexagonal DNA latices are able to be changed so that the GOx/HRP kinetics change. The graph shows a clear difference when the structure is and is not assembled. A difference is also notable between trends I and II, which represent dimensional differences of the structure. D: An in vivo folding of RNA nanostructures to improve the [FeFe]- hydrogenase and ferredoxin enzyme reaction. E: The GOx/HRP cascade is organized on a DNA tile with a bridge protein to transfer the reactant. The kinetics improve inversely with distance as seen in the graph. [51]

18 exciting direction for DNA origami and could be potential applications for a modular and dynamic system as will be described.

19 Chapter 2: Design of a Modular and Dynamic Structure

The goal of the study was to design a scaffolded DNA origami 2D building block

that could be reconfigured, combined, and refolded an indefinite number of times into

various 3D structures. This would allow many different structures to be made from the

same set of staples and would not require a new thermal ramp between conformation

changes. Other modular structures that have been formed must be folded into their

various forms in a thermal ramp (if they have multiple forms) and cannot be changed

after the structure has formed.[17,29-34,38] The structures that were dynamic only have a

limited number of conformations (usually only 2) and do not have the potential number

of forms of the proposed design.[18,25,40,42,46] Since the proposed building blocks are reusable

and don’t require heating to form into larger structures or to change, it may also be

possible to use the building blocks in vivo.[21,49,52]

The target 3D structures were platonic solids. Many DNA origami platonic solids

have been folded in literature. These structures are usually formed by making the edges of

the shapes out of DNA.[15, 18-28] However, a platonic solid’s faces are all the same shape and dimensions, so they are ideal for building from a base tile that is the same shape as one of the sides. Tetrahedrons (4 sides), octahedrons (8 sides), and icosahedrons (20 sides) are

3 of the 5 platonic solids and all have sides that are equilateral triangles. A parallelogram made out of 4 connected equilateral triangles was chosen as the base building block

20 Tetrahedron Octahedron Icosahedron Figure 16. A Parallelogram and the Platonic Solids It Can Form: The base building block was 4 equilateral triangles connected to form a parallelogram. Parallelograms can be combined and folded to form a tetrahedron, octahedron, or a icosahedron. since each solid can be made from a multiple of 4 triangles. A building block of only two triangles was not used since a smaller building block would require more building blocks than if a 4 triangle building block was used to form a platonic solid. Also, by combining the triangles into the largest common denominator of the 3 platonic solids, the fewest alterations were necessary to switch between the 3D forms. Figure 16 shows the shape of the base building block and the 3 platonic shapes that can be formed from it.

Each parallelogram was made from one scaffold. The scaffold chosen was a 8064 nucleotide long ssDNA strand derived from the M13mp18 bacteriophage. M13mp18 is a ssDNA genome that has evolved to minimize secondary structures and is what is typically used in scaffolded DNA origami since Rothmund introduced the method.[31] The lack of

21 repetition in the sequence makes it ideal for selectively targeting only 1 small section of the scaffold with part of a staple. The 8064 base pair long sequence is the longest form of M13mp18 available and was used to fold the largest single helix thick parallelogram possible.

To form a parallelogram, the scaffold was routed as shown in Figure 17. The scaffold was folded into four triangles with staples that were called internal staples.

The structure was designed using a square lattice framework instead of a hexagonal framework, since the triangles are only 1 helix thick and each triangle needs to be flat.[53]

In order for a square framework to lay flat, staples must crossover every 16 base pairs since this is 1.5 turns of a double helix.

The scaffold was routed so that the scaffold connected the corners of adjacent triangles. The scaffold was unpaired for 5 nucleotides between each triangle to allow for a flexibly joint. To form the 4 triangles into a parallelogram, the adjacent edges of triangles were also connected with internal staples. Each of these internal staples included 7

Figure 17. Parallelogram Scaffold Routing: A single scaffold, derived from M13mp18, was routed into the form of four triangles which were connected to complete a parallelogram.

22 unpaired nucleotides between triangle edges to allow for a flexible hinge to enable folding of the triangular panels. Repeating Thymines were used for these hinges to prevent anything from binding to them.

To actually design the parallelograms and determine the staple sequences necessary to fold them, a program called caDNAno was used.[54] CaDNAno is an open sourced computer aided design software for designing scaffolded DNA origami structures.

The software simplifies the design process by helping the designer stay within the geometric constraints of DNA, mainly how the rotation of the double helix determines when a crossover is possible. With parallel ssDNA strands arranged in a honeycomb pattern, a staple can cross over from one ssDNA to another every 7 base pairs, since 7 base

Figure 18. caDNAno User Interface: caDNAno is a graphical interface to help in designing scaffolded DNA origami structures. The left window shows a cross section of the structure looking down the DNA helices. The right window shows the scaffold in blue and the colored staples that make up the structure.

23 Form 2 parallelogram caDNAno design 2 parallelogram schmatic Form caDNAno design 3 parallelogram schmatic Form caDNAno Figure 19. Complete Form 2 and Form 3 Parallelogram Design Schematics: Separate designs were made for form 2 and form form 2 and form for made were designs Separate Design Schematics: 3 Parallelogram Form 2 and Form 19. Complete Figure the design of be can seen as row in each 3 trapezoids of Each up triangle made was in the parallelogram caDNAno. 3 with schematics.

24 pairs is 66% of a double helix’s 10.5 base pair full twist. To arrange the strands in a square lattice, the staples must cross over every 16 base pairs, which is 150% of a double helix’s twist. caDNAno helps to not break these rules, even though it can be if desired.[54]

When designing with caDNAno, the cross section of parallel double stranded

DNA (dsDNA) is first determined. caDNAno automatically provides a template for each dsDNA strand in the cross section. One can determine the length of each strand and connect them end to end to form the complete scaffold. It is important to verify that the lengths of the ssDNA scaffold cross section add up to the length of the scaffold that will be used. In the case of the parallelogram, that was 8064 nucleotides.

Staples are added to the scaffold by first routing one long staple through the entire scaffold, crossing over whenever possible, which was every 16 bases for a square lattice.

This long strand is then broken up into 18 to 49 nucleotide long segments that cross over one or more times. Once the staples are in place, the scaffold sequence is inserted into the design from a text file of the scaffold’s sequence. CaDNAno then generates a list of all the staples necessary to fold the scaffold. At this point, the structure is fully designed. Figure

18 shows the caDNAno interface and the top schematic of figure 19 shows the entire design for two different parallelogram forms.

In order to connect one edge of a triangle in a parallelogram to another edge to form a platonic solid, the sides of the parallelogram had overhang staples added that extended out from the edge and did not bind to anything during the normal folding of the parallelogram. These staples were called polymerization staples. Figure 20 shows the polymerization staples extending from the parallelogram. Each edge of a triangle, that

25 Polymerization Connect sides with External Staples Staples External Staples 1 + 2 1 2 3 + 6 4 + 5 6 3

5 4

Figure 20. Folding a Parallelogram into a Tetrahedron: Polymerization staples extend from each open side of the triangles making up the parallelogram. Each polymerization staple has a unique sequence and connects to half of a specific external staple. Each external staple connects two polymerization staples tethering them together. Sides of the parallelogram can be brought together using this method to form a tetrahedron.

faced the outside of the parallelogram, had 8 polymerization staples, each of which had

a unique 10 base pair sequence extending from the edge. Since the unbound sequences

were not dependent on the scaffold, a matlab program was written to determine and

select sequences that would not interact with any part of the scaffold or the staples. The

program generated all possible 10 nucleotide long sequences and weeded out the ones that

interacted unintentionally with the structure or had repetitive nucleotides (e.g. AAA).

By adding in a connecting staple, called an external staple, two polymerization

staples could be tethered. Each external staple was 20 bp long. 10 bp of the external

staple were compliment to one polymerization staple and the other 10 bp of the external

staple were compliment to a polymerization staple on another edge. When connecting

2 polymerization staples, it was important to consider the twist of the DNA and whether

the free end of the staple was a 3’ or 5’ end, since a 3’ end would have to be connected

to a 5’ end. The staples that were designed also had to match the direction of both

polymerization staples in order to bind with them. These are all things that had to be 26 considered when designing both the polymerization staples and the external staples.

When all 8 polymerization staples were tethered to the polymerization staples of another edge, the edges were connected. Since each of the 8 polymerization staples of a edge were unique, controlling the orientation of the attachment was possible. Using this method, 24 staples could be introduced to fold a parallelogram into a tetrahedron. Figure

20 illustrates how the external staples connect sides in order to fold a parallelogram into a tetrahedron.

An octahedron has 8 sides, so two parallelograms are necessary to fold it. Since multiple parallelograms are being connected, it would be possible for more than 2. This is a common occurrence when combining a modular structure, resulting in a polymer.

[16, 17, 31, 32] Refer to figure 21. If two parallelograms (form 2 and form 3) connect via side

10 Form 3 11 9

12 Connect sides with 8 External Staples Form 2 7 1 + 8 2 + 7 1 2 3 + 12 4 + 11 6 3 5 + 10 6 + 9 5 4

Figure 21. Folding Two Parallelograms into an Octahedron: A form 2 and form 3 parallelogram can be folded into a octahedron in a single folding reaction.

27 2 of the first parallelogram and side 3 of the second, then there is nothing stopping a third parallelogram from connecting to the free side of the second parallelogram using the same staples that were used to connect the first two parallelograms. To minimize the possibility of continuous polymerization of parallelograms, two approaches can be taken.

The first approach is a two step process. Two parallelograms can be initially connected on one side. If the sides connected on each parallelogram are both side 1 on the respective parallelogram, then there would be no free side 1 on the two parallelogram complex to connect to a third parallelogram. After this reaction has been completed, the rest of the external staples could be added to fold the two parallelograms into an octahedron.

The second approach is to use two different forms of parallelograms, each of which have different polymerization staples presented on their outside edges, none of

Form 2 Form 3

Figure 22. Transforming a Form 2 Parallelogram into a Form 3 Parallelogram: Two forms of parallelograms were used, form 2 and form 3. Both forms used the same structure, but differed in what sides faced outward. To fold into form 3 instead of form 2, the triangles only needed to rotate around the connections between them. Form 1 did not have sides that were unique from either form 2 or form 3 and so was not used in the design.

28 which are in common. Having two different parallelograms that cannot polymerize with themselves but only when they interact with a different parallelogram form, reduces the chance of polymerization. This is assuming that two connected parallelograms have a faster fold rate than the rate that parallelograms meet and connect in solution. This means two connected parallelograms should normally fold into octahedrons before they can

Figure 23. Strand Displacement Kinetics: Double stranded DNA sequences with toeholds were displaced. The toeholds were categorized into 3 groups, strong toeholds with high G-C nucleotide content (Ss), weak toeholds with low G-C content, and normal toeholds with equal GC-AT content. The rates of displacement of different toehold lengths for these 3 toehold categories is shows. The rates appear to be depend on toehold length, increasing as the length of the toehold increases. However, the increase in rate plateaus around a toe hold lenth of 5 bp as indicated by the red bar.[44]

29 polymerize with more parallelograms.

In the design of the parallelograms, the second option was used since having

two different forms would be especially useful when folding five parallelograms into

an icosahedron. To insure that the two different forms could not have any common

edges, the triangles that make up the parallelogram were rotated about the corners that

connected each triangle to the other. This resulted in the edges that were facing outward

facing the inside of the triangle. The 2 forms were referred to as form 2 and form 3.

Figure 22 shows the transformation and Figure 19 shows the complete schematics for both

form 2 and form 3.

The 2 forms were joined into an octahedron as shown if figure 21. Since each

parallelogram form had 6 triangle edges facing outward, each parallelogram included 48

polymerization staples. To form an octahedron 48 external staples were needed to fold

two parallelograms into an octahedron.

To unfold the tetrahedrons and octahedrons, 5 nucleotide long toeholds were

added to each external staple. A 5 nucleotide long toehold was chosen, since the kinetics

of strand displacement dependent on toehold length and 5 nucleotides is when the speed

of the displacement reaction begins to plateau, as shown in Figure 23, modified from

Zhang.[44] A python program was written to generate toeholds and add them onto an existing external staple. The toeholds were generated from sequences that were already determined to be non interactive with the rest of the structure and staples but were not used in the creation of the external staples. The program would randomly choose 5 nucleotide long sequences from this list, add it on to an external staple, and then check

30 Form 2 Form 3

Add External Staples

Add Displacement Staples

Add Unzipping Staples

Add External Staples

Figure 24. Experiment Design: On the left, a form 2 parallelogram is folded into a tetrahedron by adding tetrahedron external staples. Tetrahedron unzipping staples are then added to unfold the tetrahedron into a form 2 parallelogram again. A form 3 is then added to the form 2 in addition to octahedron external staples to fold them into a octahedron. On the right, a form 2 and a from 3 parallelograms are folded into a octahedron by adding octahedron external staples. Octahedron unzipping staples are then added to unfold the octahedron into a form 2 and form 3 parallelograms again. Tetrahedron external staples are then added to both forms, but only form 2 folds into a tetrahedron. 31 to see if any part of the external staple + toehold sequence or its reverse compliment interacted with the structure. A list of toehold extended external staples were exported along with their compliment sequences. These reverse compliment sequences were what were used to displace the external staples and unfold the platonic solids into parallelograms.

The goal of the study was to design a scaffolded DNA origami 2D building block that could be reconfigured, combined, and refolded an indefinite number of times into various 3D structures. To prove the concept, two experiments were performed as outlined in Figure 24. Form 2 and form 3 parallelograms were folded into either tetrahedrons or octahedrons, unfold back into parallelograms, and then refold into the other platonic solid (octahedrons or tetrahedrons respectively). Only form 2 was ever folded into a tetrahedron. Form 3 was solely used with form 2 to fold octahedrons. These experiments were designed to be a proof of concept for a more robust application of the system in the future, where the parallelograms are formed into an innumerable number of shapes and fictionalized for a given application.

The form 2 and form 3 parallelograms are 2 different modular building blocks.

The polymerization staples and the external staples facilitate connections required to fold the structure into platonic solids. Complimentary unzipping staples are used to remove external staples and unfold the solid back into a parallelogram. New external staples then could be added to fold the parallelograms into the platonic solid that was not folded previously. Figure 24 outlines the entire system. With these capabilities, the parallelogram building block is a modular and dynamic structure that can take on

32 multiple conformations.

33 Chapter 3: Results

Several experiments were conducted to confirm that the parallelogram building

block system is both modular and dynamic. Two design iterations of the parallelogram

were studied.. Figure 25 shows a design schematic of the scaffold routing for the first

version of the modular parallelogram design and a TEM image of a folded parallelogram

structure. The early structure design had unintended flaws, the main being that the cross

overs between triangles were not lined up properly. This misalignment of the staples

meant that all the staple connections between triangles could not form due to physical

constraints. This was confirmed by many observations in TEM images where triangles

appeared well-folded, but disconnected.

These early parallelograms were used to fold tetrahedrons. Figure 26 shows TEM

100 nm 100 nm

Figure 25 Initial Design and Images of a Parallelogram: To the left is a TEM image of an early iteration of the parallelogram. the right is the same as the left but with a parallelogram overlay. 34 Tetrahedrons

Unfolded parallelograms at 40o C Refolded Tetrahedrons at 4o C

100 nm 100 nm Figure 26. Folding Initial Design into Tetrahedrons and Unfolding with Heat: Tetrahedrons were folded with external staples that have no toeholds. They were then unfolded into parallelograms by heating them to 40o C. When cooled to 4o C, they refolded into tetrahedrons.

35 images of the folded tetrahedrons. The staples closing the triangles into a tetrahedron

were designed to have a lower melting temperature so that they could be removed by

melting without disrupting the parallelogram structure. Since the external staples used to

fold the tetrahedrons did not have toeholds, heating the tetrahedrons to 40o C was used to dissociate the external staples from there polymerization staples. The structures were then refolded into tetrahedrons when they were cooled down to 4o C.

The form 2 and 3 parallelograms were both folded in a Magnesium Chloride

(MgCl2) salt concentration of 18 mM using a 2.5 day Thermal Ramp. MgCl is necessary

because it allows the Negatively charged DNA to overcome the charge repulsion of its

negative charge and interact more readily. Before parallelograms were folded into a

tetrahedrons or octahedrons, they were filtered by centrifuging them with polyethylene

glycol (PEG). This separated the larger structures from the remaining folding staples.

PEG filtration does not remove misfolded structures however.

To fold the Form 2 parallelograms into tetrahedrons. External staples were added

at a concentration that was a factor of 10 larger than the concentration of parallelograms.

This surplus of staples helps the structure fold quicker and guarantees there are enough

staples for all the parallelograms. Folding of the parallelograms was done at room

temperature for 2 hours at a MgCl2 concentration of 14 mM. The lower temperature and

salt concentration were used in all foldings and unfoldings of parallelograms to minimize

aggregation of the structures.

The current parallelograms, that were discussed in Chapter 2, were designed to fix

the first version design flaws and improve folding. Using the redesigned parallelograms of

36 A B

Structure Band

Dimer Band Ladder Form 2 Form Form 3 Form Refolded Refolded Refolded Refolded Unfolded Unfolded Form 2 + 3 Form Octahedron Octahedron Octahedron Tetrahedron Tetrahedron 8064 Scaffold Parallelogram Parallelogram Parallelogram Parallelogram Parallelograms Figure 27. Agarose Gel of Final Structures: Above is a agarose gel of two separate folding sequences which were performed at the same time. Sequence A folded a Octahedron out of form 2 and 3 parallelograms. These tetrahedrons were then unfold into parallelograms. The form 2 parallelograms were then refolded into tetrahedrons. Sequence B is the reverse of sequence A, going from a tetrahedron to an octahedron.

both Form 2 and Form 3, experiments to fold and unfold tetrahedrons and octahedrons were performed. Figure 27 shows an agarose gel, run at 70 volts for 3.5 hours, which shows the results of two separate folding sequences. Figure 27 A depicts a folding sequence that folded Form 2 and Form 3 parallelograms into octahedrons, unfolded them, and then refolded the form 2 parallelograms into tetrahedrons. TEM images of each structure band from figure 27 A are shown in figure 28. Figure 27 B is the reverse progression. Form 2 parallelograms were first folded into tetrahedrons, then unfolded back into parallelograms, and finally were folded with form 3 parallelograms to form octahedrons. TEM images of figure 27 B’s bands are shown in figure 29.

The same conditions were used for unfolding the structures. A complimentary

37 A Form 3 Parallelograms Form 2 Parallelograms

100 nm 100 nm 100 nm 100 nm B Octahedrons

100 nm

100 nm C Unfolded Parallelograms

100 nm D Refolded Tetrahedrons

100 nm Figure 28. Images of Folding Sequence Starting with Octahedrons: A: TEM images of form 2 and 3 parallelograms. B: Images of octahedrons folded from the parallelograms. C: Parallelograms which resulted from unfolding the octahedrons. D: Tetrahedrons which were refolded from parallelograms which had previously been octahedrons.

38 A Form 2 Parallelograms

100 nm 100 nm B Tetrahedrons

100 nm C Unfolded Parallelograms

100 nm D Refolded Octahedrons

100 nm

Figure 29. Images of Folding Sequence Starting with Tetrahedrons: A: TEM images of form 2 parallelograms. B: Images of tetrahedrons folded from the form 2 parallelograms. C: Form 2 parallelograms which resulted from unfolding the tetrahedrons. D: Octahedrons which were refolded from 2 parallelograms which had previously been tetrahedrons and added form 3 parallelograms. 39 staple concentration that was 10 times greater than the external staple concentration

was used however. This resulted in the complimentary staple concentration being 100

times greater than the structures concentration. This higher concentration was used

so that all the external staples, bound to structure and free in solution, would be bound

and “consumed” by the complimentary staples. This prevented the external staples from

interfering in a subsequent folding without the need to filter them out.

Folding an octahedron required from 2 and form 3 to be combined. Each of the

two structures is diluted to half its original concentration when combined. To maintain

a consistent staple excess ratio, the concentration of octahedron external staples and

unzipping staples were also added at half the concentration. The lower concentration of

parallelograms decreased the likely hood of misfolding polymerization and aggregation of

octahedrons.

The structure bands within sequences A and B in Figure 27 show differences in how far the bands ran, suggesting conformation changes did take place. It is especially helpful to look for differences in dimer bands when the differences in structure bands is not present. Tetrahedrons do appear to run a bit faster than form 2 parallelograms, which is expected, since a tetrahedron is more compact than a flexible parallelogram. A noticeable difference between octahedrons and parallelograms was not seen however.

For this reason, a form 2 and form 3 parallelograms that were connected at 1 side were also run in the gel for comparison. These do appear to have run slightly further than the octahedron and the parallelograms, suggesting the octahedron band is a fully folded structure, and not a structure that has loose unbound sections. The fact that the

40 connected form 2 and 3 parallelograms ran slightly further was surprising, since it is a larger structure than a single parallelogram. However, it is surmised that the connected parallelograms fold up to present a minimal front to the gel, causing it to provide not much more resistance than a single parallelogram. The larger structure also may result in

A 100 nm

100 nm

Figure 30. How a Parallelogram can Fold onto Itself: A: The triangles of a parallelogram can fold onto each other, giving the appearance of a trapezoid. This can either occur from a end triangle folding over or the entire structure folding across its middle. B: Parallelograms do not always form correctly. Often, one triangle is left dangling from the rest of the structure.

41 more pull from the electric current running through the gel. While differences in the gel are observable, they are not clearly significant differences to make claims on if the foldings and unfoldings were successful. However, TEM imaging confirmed that we obtained the desired structures.

Figure 28 Shows the TEM images of the resulting structures which began with the folding of octahedrons from From 2 and 3 parallelograms. In Figure 28 A, form 2 and 3 parallelograms were successfully folded. Figure 29 shows the TEM images for the structures which were first folded into tetrahedrons. Figure 29 A shows the form 2 parallelograms. While parallelograms did form, there were many occurrences when 4 triangles did not perfectly line up. Figure 30 illustrates some of these cases where form

2 and 3 parallelograms did not form properly. Figure 30 A shows how the parallelogram can fold up onto itself, giving the appearance of only 2 or 3 triangles. Figure 30 B shows examples of instances where not all of the connections between triangles were made, causing 1 or more triangles to dangle from the structure or not line up perfectly with the other triangles.

While the individual parallelograms showed large variation in their shape due to the flexible connections, the octahedrons shown in figure 28 B and the tetrahedrons from figure 29 B appear to have folded well, since parallelogram like structures were not observed. A TEM image is 2D, so it can be difficult to differentiate between 3D structures.

This is especially true when comparing octahedrons and tetrahedrons. Figure 31 shows the most common 2 dimensional views of these structures and what is looked for in order to differentiate them.

42 Top Tetrahedron Tetrahedron Side Tetrahedron Top Octahedron Side Octahedron

100 nm

Figure 31. Ways to Interpret TEM Images of Tetrahedrons and Octahedrons: Tetrahedrons have a very distinct triangular top view. This is easily distinguishable from the octahedrons square top view. The side views of both structure are very similar however. The only distinguishing feature of the side view is that the octahedron appears sightly longer, due to the larger angle between its sides.

The side views of both structures are very similar. They both appear as 2 connected triangles. An octahedrons sides have a wider angle between them however, causing the 2 triangles that make up this side view to be slightly longer. This is one way to differentiate between a tetrahedron and an octahedron. The more distinct differentiation

43 is between there top views. An octahedron will have a square shape, while a tetrahedron will have a triangular shape. The appearance of these top views in one sample and not the other is the clearest way to identify that the structures folded properly. One can also often see a difference in shading between two sides, or a line where two sides meet as shown in

Figure 32. These lines allow the 3 dimensional structure to be interpreted even further.

The point where multiple sides come together, or the slope difference between 2 sides can be distinguished in this manner.

Figure 28 C and figure 29 C show the unfolded parallelograms that were previously an octahedron or tetrahedron respectively. While the parallelogram TEM images again did not have consistent distinct parallelogram structure, the lack of tetrahedron and octahedron structures and the lack of parallelograms in the images

Octahedron Refolded Octahedron

100 nm 100 nm

Tetrahedron Refolded Tetrahedron

100 nm 100 nm Figure 32. Outlined TEM Images: The sides of the tetrahedrons and octaherdrons were often different shades as can be detected in the octahedron and tetrahedron. The edges of a octahedron and tetrahedron are also evident by dark lines as can be seen in the refolded octahedron and refolded tetrahedron 44 of the previous tetrahedrons and octahedrons, is a clear indication that the structures successfully unfolded.

Figure 28 D shows the unfolded parallelograms refolded into tetrahedrons. These structures, which at one point had been octahedrons, now showed the distinct triangular overhead view of a tetrahedron. There were a few parallelograms also evident, but this was expected, since the form 3 parallelograms left over from the octahedron unfolding were not removed or folded into tetrahedrons.

The tetrahedrons were also successfully folded into octahedrons as shown if figure 29 D. The square overhead view corresponding to the octahedron is evident and triangular overhead views that would indicate tetrahedrons were not observed, suggesting that all tetrahedrons had been unfolded and only octahedrons remained after the refolding. Parallelograms were not observed in these images as was the case in the samples of the images in figure 28 D.

The complete conformation changes that were observed provide evidence that a modular building block was successfully designed and folded. It also was dynamic, undergoing multiple conformation changes with one set of parallelograms. While tetrahedrons and octahedrons were the only structures folded, it should be possible to fold other structures with these parallelograms, as shown in Figure 33. An icosahedron is the remaining platonic solid to fold. This folding could be achieved through a 2 step process shown in Figure 33. First, in figure 33 A, a form 3 would need to be connected with two form 2s. Both form 2s would use the same edge to connect to different sides of the form 3 parallelogram. This would prevent further polymerization. Then, in figure

45 33 B, two more form 3 parallelograms could be connected to the structure along with all the other staple necessary to fold into a icosahedron, as shown in figure 33 C. It would be the hope that the icosahedron would fold before more than the two parallelograms could bind to complete the structure. An alternative non platonic 8 sided solid is also proposed as shown in Figure 34 in relation to a octahedron. This structure uses the same parallelograms and may be easier to distinguish from the octahedron since the shape is

A B

Figure 33. How to Fold an Icosahedron: An icosahedron can be formed in a two step process. A: Two form 2 parallelograms can first be connected to a form 3 parallelogram. This prevents the structure from polymerizing to much. B: Two more form 3 parallelograms can then be added to the ends of the structure along with the rest of the external staples necessary to fold the icosahedron. C: The icosahedron will then fold.

46 Figure 34. An Eight Sided Non-platonic Solid: Two parallelograms can also be used to form the solid on the left. This solid is not platonic, unlike the octahedron on the right, because one of its edges is concave and all the angles between its sides are not the same. These two structure could be switched between without fully disconnecting the same parallelograms that form them.

qualitatively different with a larger aspect ratio. Furthermore, an octahedron could be refolded into this shape without completely disconnecting the two parallelograms.

There are two possible application for the 3 dimensional, modular, and dynamic system presented. The first is as a carrier for a molecule held within the platonic shape.

This molecule could then be released when the solid is unfolded by a trigger molecule.

Other DNA origami carriers have also been devised.[20, 21, 25, 28, 40, 47, 48, 49, 55] Figure 35 from

Ansusuya shows a particle contained in an icosahedron which could be triggered to open and release its cargo by introducing a cyclic-di-GMP (cdGMP) analite. [55]

The second is as a catalyst or an enzyme that can be turned on or off. If two parts of a catalyst or 2 separate catalysts that are necessary for a particular reaction, are housed and tethered inside separate tetrahedron, then the catalyst parts would not be able to interact with each other since the tetrahedrons would be in the way and keep the parts distant from each other. These tetrahedron could then be unfolded and refolded

47 Figure 35. A DNA Origami Delivery System: Two parallelograms can also be used for form the solid on the left. This solid is not platonic, unlike the octahedron on the right, because one of its edges is concave and all the angles between its sides are not the same. These two structure could be switched between without fully disconnecting the same parallelograms that form them.[55]

into an octahedron bringing the 2 parts in proximity, allowing the reaction to occur. By folding the octahedron back into two tetrahedrons, the reaction would be kept from continuing. DNA origami has been used to facilitate enzyme function, but the DNA origami structures used are generally static platforms that only keep enzymes in proximity as discussed in Figure 14 in Chapter 1.[26]

The proposed system is the first step towards a more robust and functional process. More solids will have to be folded with the parallelogram building blocks and functionalization of these building blocks with other molecules will have to occur in order

48 to make the technology useful to other fields.

49 Citations

1. Kamat PV. “Meeting the Clean Energy Demand: Nanostructure Architectures for Solar Energy Conversion.” J. Phys. Chem. 111, 2834-2860 (2007).

2. Luo Z, King X, Hu Y, Wu S, Xiang Y, Zeng Y, Zhang B, Yan H, Zhang H, Zhu L, Liu J, Li J, Cai k, Zhao Y. “Engineering a Hollow Nanocontainer Platform with Multifunctional molecular Machines for Tumor-Targeted Therapy in Vitro and in Vivo.” ACS Nano. 7, 11, 10271-10284 (2013).

3. Watson JD, Crick FHC. “Molecular Structure of Nucleic Acids: A Structure for Deoxyribose Nucleic Acid.” Nature. 171, 737-738 (1953).

4. Mandelkern M, Elias JG, Eden D, Crothers DM. “The Dimensions of DNA in solution.” Journal of Molecular Biology, 152, 1, 152-161 (1981)

5. Seeman NC, and Philip PL. “Nucleic acid nanostructures; bottom-up contrrol of geometry on the nanoscale.” Reports on Progress in Physics. 68, 237-270 (2005).

6. Seeman NC. “nucleic Acid Junctions and Lattices.” Journal of Theoretical Biology. 99, 237-247 (1982).

7. Eichman BF, Vargason JM, Mooers BHM, Ho PS. “The Holliday junction in an inverted repeat DNA sequence: Sequence effects on the structure of four-way junctions.” PNAS. 97, 8, 3971-3976 (2000).

8. Holliday Robin. “A mechanism for gene conversion in fungi.” Genetical Research. 5, 2, 282-304 (1964).

9. Duckett DR, Murchie AIH, Diekmann S, Kitzing E, Kemper B, Lilley DMJ. “The Structure of the Holliday Junction and Its Resolution.” Cell. 55, 79-89 (1988).

10. Ma RI, Kallenbach NR, Sheardy RD, Petrillo ML and Seeman NC. “Three arm nucleic acid junctions are flexible.” Nucleic Acids Res. 14, 9745–53 (1986).

11. Wang Y, Mueller JE, Kemper B, and Seeman NC. “The assembly and characterization of 5-arm and 6-arm DNA junctions.” Biochem. 30, 5667–74 (1991).

50 12. LaBean T, Yan H, Kopatsch J, Liu F, Winfree E, Reif JH, and Seeman NC. “The construction, analysis, ligation and self-assembly of DNA triple crossover complexes/” J. Am. Chem. Soc. 122, 1848–60 (2000).

13. Li X, Yang X, Qi J, and Seeman NC. “Antiparallel DNA double crossover molecules as components for nanoconstruction.” J. Am. Chem. Soc. 118, 6131–40 (1996).

14. Fu TJ and Seeman NC. “DNA double crossover structures.” Biochemistry. 32, 3211–20 (1993).

15. Chen J, Seeman NC. “Synthesis from DNA of a molecule with connectivity of a cube.” Nature. 350, 631-633 (1991).

16. Xiaoping Y, Wenzler LA, Qi J, Li X, Seeman NC. “Ligation of DNA Triangles Containing Double Crossover Molecules.” J. Am. Chem. Soc. 120, 9779-9786 (1998).

17. Liu D, Wang M, Deng Z, Walulu R, and Mao C. “Tensegrity: Construcution of Rigid DNA TRiangles with Flexible Four-Arm DNA Junctions.” J. AM. Chem. Soc. 126, 2324- 2325 (2004).

18. Han Da, Huang J, Zhu Z, Yuan Q, You M, Chen Y, and Tane W. “Molecular engineering of photoresponsive three-dimensional DNA nanostructures.” Chem. Commun. 47, 4670-4672 (2011).

19. Dohno C, Atsumi H, and Nakatani K. “Ligand inducible assembly of a DNA tetrahedron.” Chem. Commun. 47, 3499-3501 (2011).

20. Wang Z, Xue Q, Tian W, Wang L, and Jiang W. “Quantitative detection of single DNA molecules on DNA tetrahedron decorated substrates.”

21. Kim KR, Kim DR, Lee T, Yhee JY, Kim BS, Kwon IC, and Ahn DR. “Drug delivery by a self-assembled DNA tetrahedron for overcoming drug resistance in breast cancer cells.” Chem. Commun. 49, 2010-2012 (2013).

22. Bu NN, Tang CX, He XW, and Yin XB. “Tetrahedron-structured DNA and functional oligonucleotide for construction of an electrochemical DNA-based biosensor.” Chem. Commun. 47, 7689-7691 (2011).

23. Pei H, Zuo X, Zhu D, Huang Q, and Fan C. “Functional DNA Nanostructures for Theranostic Applications.” Accounts of Chemical Research. 47, 2, 550-559 (2014).

24. Goodman RP, Berry RM, and Turberfield AJ. “The single-step synthesis of a DNA tetrahedron.” Chem. Commun. 12, 13722-1373 (2004).

51 25. Liu Z, Li Y, Tian C, and Mao C. “A Smart DNA Tetrahedron That Isothermally Assembles or Dissociates in Response to the Solution pH Value Changes.” American Chemical Society. 14, 1711-1714 (2013).

26. Ozhalici-Unal H and Armitage BA. “Fluorescent DNA Nanotags Based on a Self- Assembled DNA Tetrahedron.” ACS Nano. 3, 2, 425-433 (2009).

27. Zhang Y and Seeman NC. “Construction of a DNA-Truncated Octahedron.” J. Am. Chem. Soc. 116, 1661-1669 (1994).

28. Bhatia D. Mehtab S, Krishnan R, Indi SS, Basu a, krishnan Y. “Icosahedral DNA nanocapsules.” Angew. Chem. Int. Ed. 48, 4134-4137 (2009).

29. Wei B, Dai M, and Yin P. “Complex shapes self-assembled from single-stranded DNA tiles.” Nature. 485, 623-627 (2012).

30. Ke Y, Ong LL, Shih WM, and Yin P. “Three-Dimensional Structures Self-Assembled from DNA Bricks.” Science. 338, 1177-1183 (2012).

31. Rothemund PWK. “Folding DNA to create nanoscale shapes and patterns.” Nature. 440, 297-302 (2006).

32. Douglas SM, Dietz H, Liedl T, Hogberg B, Graf F, and Shih WM. “Self-assembly of DNA into nanoscale three-dimensional shapes.” Nature. 459, 114-118 (2009).

33. Castro CE, Kilchherr F, Kim D, Shiao EL, Wauer T, Wortmann P, Bathe M, and Dietz H. “A primer to scaffolded DNA origami.” Nature Methods. 8, 3, 221-229 (2011).

34. Zhao Z, Liu Y, Yan H. “Orginizing DNA Origami Tiles into Larger Structure Using Preformed Scaffold Frames.

35. Zhang C, Su M, He Y, Leng Y, Ribbe AE, Wang G, Jiang W, and Mao C. “Exterior modification of a DNA tetrahedron.” Chem. Commun. 46, 6792-6794 (2010).

36. Smith DM, Schuller V, Forthmann C, Schreiber R, Tinnefeld P, and Liedl T. “A Structurally Variable Hinged Tetrahedron Framework from DNA Origami.” Journal of Nucleic Acids. (2011)

37. Ke Y, Sharma J, Liu M, Jahn K, Liu Y, Yan H. “Scaffolded DNA Origami of a DNA Tetrahedron Molecular Container.” Nano Letters. 9, 6, 2445-2447 (2009).

38. Linuma R, Ke Y, Jungmann R, Schlichthaerle T, Woehrstein JB, and Yin P. “Polyhedra

52 Self-assembled from DNA Tripods and Characterized with 3D DNA-PAINT.” Science. 344, 65-69 (2014).

39. Shih WM, Quispe JD, Joyce GF. “A 1.7-kilobase single-stranded DNA that folds into a nanoscale octahedron.” Nature. 427, 618 - 621 (2004).

40. Andersen ES, Dong M, Nielsen MM, Jahn K, Subramani R, Mamdouh W, Golas MM, Sander B, Stark H, Olivera CLP, Pedersen JS, Birkedal V, Besenbacher F, Gothelf KV, Kjems J. “Self-assembly of a nanoscale DNA box with a controllable lid.” Nature. 459, 73- 76 (2009).

41. Kuzuya A, Komiyama M. “Design and construction of a box-shaped 3D-DNA origami.” Chem. Commun. 4182-4184 (2009).

42. Douglas SM. “A Logic-Gated nanorobot for Targeted Transport of Molecular Payloads.” Science 335, 831-834 (2012).

43. Srinivas N, Thomas OE, Sulc P, Schaeffer JM, Yurke B, Louis AA, Doye JPK, and Winfree E. “On the biophysics and kinetics of toehold-mediated DNA strand displacement.” Nucleic Acids Research. 41, 22, 10641-10658 (2013).

44. Zhang YD and Winfree E. “Control of DNA Strand Displacement Kinetics Using Toehold Exchange.” Journal of American Chemical Society. 131, 17303-17314, (2009).

45. Seelig G and Zhang DY. “Dynamic DNA nanotechnology using strand-displacement reactions.” Nature Chemistry. 957, 103-113 (2011).

46. Zhang F, Nangreave J, Liu Y, and Yan H. “Reconfigurable DNA Origami to Generate Quasifractal Patterns.” Nano Letters. 12, 3290-3295 (2012).

47. Ke Y, Sharma J, Liu M, Jahn K, Liu Y, Yan H. “Scaffolded DNA Origami of a DNA Tetrahedron Molecular Container.” Nano Letters. 9, 6, 2445-2447 (2009).

48. Zhang C, Li X, Tian C, Yu G, Li Y, Jiang W, Mao C. “DNA Nanocages Swallow Gold Nanoparticles (AuNPs) to Form AuNP@DNA Cage Core - Shell Structures.” ACS Nano. 8,2, 1130-1135 (2014).

49. Jiang Q, Song C, Nangreave J, Liu X, Lin L, Qiu D, Wang Z, Zou G, Liang X, Yan H, Ding B. “DNA Origami as a Carrier for Circumvention of Drug Resistance.” Journal of The American Chemical Society. 134, 13396-13403 (2012).

50. Niemeyer CM. “The developments of semisynthetic DNA-protein conjugates.” TRENDS in Biotechnology. 20, 9, 395 - 401 (2002).

53 51. Fu J, Liu M, Liu Y, Yan H. “Spatially-Interactive Biomolecular networks Organized by Nucleic Acid Nanostructures.” Accounts of Chemical Research. 45, 8, 1215-1226 (2012).

52. Li Z, Wei B, Nangreave J, Lin Chenxiang, Liu Y, Mi Y, and Yan H. “A Replicable Tetrahedral Nanostructrure Self-Assmbled from a Single DNA STrand.” J. AM. CHEM. SOC. 131,13093-13098 (2009).

53. Ke Y, Voigt NV, Gothelf KV, and Shih WM. “Multilayer DNA Origami Packed on Hexagonal and Hybrid Lattices.” J. AM. CHEM. SOC. 134, 1770-1774 (2012).

54. Douglas SM, Marblestone AH, Teerapittayanon S, Vazquez A, Church GM, and Shih WM. “Rapid prototyping of 3D DNA-roigami shapes with caDNAno.” Nucleic Acids REsearch. 37, 15, 5001-5006 (2009).

55. Banerjee A, Bhatia D, Saminathan A, Chakraborty S, Kar S, and Krishnan Y. “Controlled Release of Encapsulated Cargo from a DNA Icosahedron using a Chemical Trigger.” Angew. Chem. Int. Ed. 52, 6854-6857 (2013).