Self-Assembling DNA Self-Assembling DNA N. Jonoska Jonoska, N. C. - - PDF document

SLIDE 1

Self-Assembling DNA Self-Assembling DNA Graphs Graphs

Jibran Rashid Jibran Rashid CPSC 607 CPSC 607

References References

• “

“Self-Assembling DNA Graphs Self-Assembling DNA Graphs” ”, P. Sa- , P. Sa-Aradyen Aradyen, , N.

• N. Jonoska

Jonoska, N. C. , N. C. Seeman Seeman, DNA Computing, , DNA Computing, 8 8th

th International Workshop on DNA-Based

International Workshop on DNA-Based Computers, pp. 1-9, 2003 Computers, pp. 1-9, 2003

• “

“Computation by Self-Assembly of DNA Computation by Self-Assembly of DNA Graphs Graphs” ”, P. Sa- , P. Sa-Aradyen Aradyen, N. , N. Jonoska Jonoska, N. C. , N. C. Seeman Seeman, Genetic Programming and Evolvable , Genetic Programming and Evolvable Machines, Vol. 4, pp. 123-137, 2003 Machines, Vol. 4, pp. 123-137, 2003

Outline Outline

• Motivation

Motivation

• Proposed Methods

Proposed Methods

• General framework for problem solving via Self-

General framework for problem solving via Self- Assembly Assembly

• Experiment Design

Experiment Design

• Experiment Results

Experiment Results

• Conclusions

Conclusions

• Issues

Issues… …. .

Self-Assembly Self-Assembly

• Self-assembly is the ubiquitous process by which

Self-assembly is the ubiquitous process by which

• bjects autonomously assemble into complexes.
• bjects autonomously assemble into complexes.1

1

• It may be utilized as a tool to enable the

It may be utilized as a tool to enable the development of complex information processing development of complex information processing units e.g. DNA Computing where the inherent units e.g. DNA Computing where the inherent 3D structure of DNA can be used as a 3D structure of DNA can be used as a computational device computational device

[1] [1] http://www.usc.edu/dept/molecular-science/index.html http://www.usc.edu/dept/molecular-science/index.html

Different Proposals Different Proposals – – Tiling Tiling

• DNA tiles made of double and triple cross-over

DNA tiles made of double and triple cross-over molecules molecules – – for self-assembly of 2D arrays for self-assembly of 2D arrays

• Autonomous Molecular Computation

Autonomous Molecular Computation

• Based on Tiling Theory

Based on Tiling Theory – – Arrangement of basic shaped that Arrangement of basic shaped that cover infinite place. Infinite number of square tiles with 4 cover infinite place. Infinite number of square tiles with 4 colored slides can simulate Turing Machines. Use DNA to colored slides can simulate Turing Machines. Use DNA to simulate these tiles via self-assembly simulate these tiles via self-assembly1

1

• Sides of a tile correspond to the sticky ends

Sides of a tile correspond to the sticky ends

• Labeled by the sequences of the sticky ends

Labeled by the sequences of the sticky ends

• Tiles assemble with each other according to WC

Tiles assemble with each other according to WC complementarity complementarity

• XOR operation was experimentally confirmed

XOR operation was experimentally confirmed

[1] http://ai.stanford.edu/~serafim/CS374_2004/Presentations/CS374_2004_Lecture19a_I_DNAassembly.ppt [1] http://ai.stanford.edu/~serafim/CS374_2004/Presentations/CS374_2004_Lecture19a_I_DNAassembly.ppt

Different Proposals Different Proposals – – Branched Branched Junction Molecules Junction Molecules

• Branched junction molecules and graph-like DNA

Branched junction molecules and graph-like DNA structures structures

• Splicing of tree like structures

Splicing of tree like structures

• Simulate Horn Clause computation via self-assembly of three

Simulate Horn Clause computation via self-assembly of three junctions and hairpins junctions and hairpins

• Proposals for solutions to various problems e.g. 3SAT,

Proposals for solutions to various problems e.g. 3SAT, Hamiltonian Path & 3-Coloring a graph Hamiltonian Path & 3-Coloring a graph

• These solutions have not been confirmed experimentally

These solutions have not been confirmed experimentally

• Able to construct regular graphs (degrees of all vertices are

Able to construct regular graphs (degrees of all vertices are the same) the same)

SLIDE 2

Key Idea Key Idea

• Problem is encoded via a graph such that for a

Problem is encoded via a graph such that for a given problem instance the graph is self- given problem instance the graph is self- assembled using WC assembled using WC complementarity complementarity if and if and

• nly if a solution exists i.e. a satisfying
• nly if a solution exists i.e. a satisfying

assignment to the variables exists (3SAT) OR assignment to the variables exists (3SAT) OR the graph is 3-colorable the graph is 3-colorable

• Usually need to remove partially formed 3D

Usually need to remove partially formed 3D DNA structures or graphs that are larger than DNA structures or graphs that are larger than the original problem encoding the original problem encoding

Example Example – – 3SAT 3SAT

) ( ) ( ) ( z y x z y x z y x

• =
• The graph on the left corresponds to the above 3SAT instance. A possible

self-assembled graph structure for the formula is depicted on the right.

Examples Examples – – 3-Coloring 3-Coloring

• For a given graph G, a DNA graph structure corresponding to

For a given graph G, a DNA graph structure corresponding to G can be formed by vertex and edge building blocks if and G can be formed by vertex and edge building blocks if and

• nly if G is 3-colorable.
• nly if G is 3-colorable.
• The following

The following two laboratory two laboratory steps are required: steps are required:

1. 1.

Combine all building blocks and allow the complementary ends to Combine all building blocks and allow the complementary ends to hybridize and be hybridize and be ligated ligated (probably need to use PCR to amplify the (probably need to use PCR to amplify the number of solution graph structures) number of solution graph structures)

2. 2.

Determine whether the required graph structure has formed by: Determine whether the required graph structure has formed by:

a) a)

Remove partially formed 3D DNA structures Remove partially formed 3D DNA structures

b) b)

Use gel electrophoresis to remove graphs that are larger than the original Use gel electrophoresis to remove graphs that are larger than the original graph graph

c) c)

If graph structures remain in the test tube, then If graph structures remain in the test tube, then then then graphs is 3-colorable graphs is 3-colorable

Experiment Experiment

• Goal: To self-assemble the following non-

Goal: To self-assemble the following non- regular graph (5 vertices & 8 edges) in the regular graph (5 vertices & 8 edges) in the laboratory (Not solving any graph coloring laboratory (Not solving any graph coloring problem currently) problem currently)

Design of the Self-Assembly Design of the Self-Assembly

• The five vertices are designed to be junction molecules

The five vertices are designed to be junction molecules

• 3-armed junctions for vertices v

3-armed junctions for vertices v1

1, v

, v3

3, v

, v4

4 & v

& v5

5

• 4-armed junction for v

4-armed junction for v2

2

• Sticky ends (represented by shaded area) designed such that the final

Sticky ends (represented by shaded area) designed such that the final graph structure is one cyclic molecule graph structure is one cyclic molecule

• Six DNA strands required for each edge:

Six DNA strands required for each edge:

• Two strands each from the two junction molecules representing the

Two strands each from the two junction molecules representing the vertices vertices

• Two strands to form the duplex molecule of the edge

Two strands to form the duplex molecule of the edge

• Length of e

Length of e1

1 = e

= e2

2 =e

=e5

5 = e

= e8

8 = 4 helical turns = 42 base pairs

= 4 helical turns = 42 base pairs

• Length of e

Length of e3

3 = e

= e4

4 = e

= e7

7 = 6 helical turns = 63 base pairs

= 6 helical turns = 63 base pairs

• Length of e

Length of e6

6 = 8 helical turns = 84 base pairs

= 8 helical turns = 84 base pairs

• Sticky edges of length 6 and 8

Sticky edges of length 6 and 8 bp bp

Design of the Self-Assembly Design of the Self-Assembly

Strands labeled using Strands labeled using The labeling of a v The labeling of a v2

2 strand was not

strand was not successful and no successful and no ligation ligation was was visible visible

ATP P 32

SLIDE 3

Design of the Self-Assembly Design of the Self-Assembly

• Edges e

Edges e3

3, e

, e4

4, e

, e6

6 & e

& e7

7 contain an additional

contain an additional hairpin of one a half helical turn in the middle of hairpin of one a half helical turn in the middle of the molecule the molecule

• The three junctions of the hairpins contain

The three junctions of the hairpins contain bulges of T bulges of T’ ’s at each end (Expected to add s at each end (Expected to add flexibility) flexibility)

Design of the Self-Assembly Design of the Self-Assembly

DNA sequences for the vertices DNA sequences for the vertices

Bulges of T Bulges of T’ ’s s

Design of the Self-Assembly Design of the Self-Assembly

DNA sequences for the Edges DNA sequences for the Edges

Bulges of T Bulges of T’ ’s s

Experimental Procedure Experimental Procedure

• Molecules annealed by heating 90

Molecules annealed by heating 90_

_C for two

C for two minutes and then slowly cooled to room minutes and then slowly cooled to room temperature temperature

• Ligation

Ligation of the whole graph performed via:

• f the whole graph performed via:
• Annealing all junctions & edges separately

Annealing all junctions & edges separately

• Mixture treated with T4 DNA

Mixture treated with T4 DNA ligase ligase at room temp. at room temp.

• vernight
• vernight
• Final graph molecule length = 1084 base pairs

Final graph molecule length = 1084 base pairs

Experimental Results Experimental Results

Gel-electrophoresis results for junction molecules Gel-electrophoresis results for junction molecules representing vertices v representing vertices v2

2 and v

and v5

5

Experimental Results Experimental Results

Gel-electrophoresis results Gel-electrophoresis results for the for the ligation ligation product of product of the self -assembled graph the self -assembled graph First lane contains a linear First lane contains a linear marker, and lanes 2, 3, 4 & 5 marker, and lanes 2, 3, 4 & 5 contain contain ligation ligation products products with labeled strands of with labeled strands of junctions v junctions v1

1, v

, v3

3, v

, v4

4 & v

& v5

5

SLIDE 4

Final Product Results Final Product Results

• 32

32 ligation ligation sites in the self-assembly sites in the self-assembly

• With 90% efficiency of the

With 90% efficiency of the ligase ligase, will not have more , will not have more than 3% of the final product than 3% of the final product

• 2D gel-electrophoresis

2D gel-electrophoresis

• Electrophoresis in which a second perpendicular

Electrophoresis in which a second perpendicular electrophoretic electrophoretic transport is performed on the transport is performed on the separate components resulting for the first. separate components resulting for the first.1

1

• 3.5% & 5% denaturing gel

3.5% & 5% denaturing gel

• The behavior of the final product could not be

The behavior of the final product could not be predicted predicted

[1] http://crisp.cit.nih.gov/Thesaurus/00020523.htm [1] http://crisp.cit.nih.gov/Thesaurus/00020523.htm

Final Product Results Final Product Results

• Lanes B & E show a clear band of the expected

Lanes B & E show a clear band of the expected size ( approx 1100 size ( approx 1100 bp bp) )

• No analysis for the extra bands

No analysis for the extra bands

Conclusions Conclusions

• The required graph is formed in some cases:

The required graph is formed in some cases:

• Need to sequence the final product to confirm

Need to sequence the final product to confirm

• Verification performed via restriction analysis

Verification performed via restriction analysis

• Analysis not published

Analysis not published

• Yield of the required graph structure is very low

Yield of the required graph structure is very low

• Extracted product needs to be amplified via PCR

Extracted product needs to be amplified via PCR

Issues Issues

• Scalability

Scalability – – Not clear how this is handled and seems Not clear how this is handled and seems to be the main problem in general. Experimental to be the main problem in general. Experimental techniques for a 1000-junction molecule? techniques for a 1000-junction molecule?

• Explore better algorithms not just brute force

Explore better algorithms not just brute force

• Theoretical framework for the self-assembly procedure

Theoretical framework for the self-assembly procedure (need to know more about how things work when we (need to know more about how things work when we shake the test tube) shake the test tube)

• How easy is it to come up with an encoding of the

How easy is it to come up with an encoding of the problem that utilizes WC problem that utilizes WC

• Number of laboratory steps

Number of laboratory steps – – (good measure?) (good measure?)