DNA disPLAY & Conway's Game of Life
Self-Assembly Lab + Little Devices Lab + Gehrke Lab + Autodesk Inc.
The Self-Assembly Lab is a cross-disciplinary research lab at MIT composed of designers, scientists and engineers inventing programmable material and self-assembly technologies aimed at industrial-scale application. Self-Assembly is a process by which disordered parts build an ordered structure through local interaction. The Self-Assembly Lab has demonstrated that this phenomenon is scale-independent and can be utilized for self-constructing and manufacturing systems at nearly every length-scale. Little Devices Lab and the Gehrke Lab have a background in interactive diagnostics with signatures of molecular scale virus-host interactions by mapping self assembled patterns into colormetric topologies. The Bio/Nano/Programmable Matter group at Autodesk, a world leader in design software, researches the intersection of design and matter programming across domains and scales such as synthetic biology, 3D bioprinting, 4D printing, and DNA nanotechnology. With this in mind, the bio/nano team is building Project Cyborg, a platform of platforms (or meta-platform) to simplify the interaction within and across emerging design domains and reduce the friction of software development. Through a multidisciplinary collaboration of these four groups, we created a time resolved diagnostic display, which we call DNA disPLAY. Traditionally, colorimetric diagnostics develop as single frame images. With DNA disPLAY we are exploring the potential of time resolved diagnostics through multisequence, multiplex assays through pixelated deposition. By reserving adjacent pixels for additional reactions that have affinity to different types of proteins DNA sequences and other targets we can train the system to recognize the time series being developed. In order to make multisequence, multiplex assays, we developed methodologies and tools that demonstrate the proof of concept.
1.Software: Project CyborgApplication built on Project Cyborg to plan and map the diagnostic by simulating each interaction and endpoint. This application is built for designing the components of the diagnostic display: the sensors found at each pixel, the targets recognized by those sensors, the overall arrangement of those sensors into an assay, and the combination of multiple targets into a wash over the entire assay.DNA targets to be recognized by the assay are the building block of the application. They are simply composed of a DNA sequence of varying length.
Each sensor in the application can be attuned to any number of incoming targets, using a programmable rule for how the sensor is built. For example, for a sensor to be attuned to multiple distinct DNA targets, we can define the program used for each sensor as one that assembles the sensor sequence out of the corresponding reverse complement sequences. For other types of targets, a different program could be used for the sensors that incorporates the design needed to sense that type of target.DNA targets can be grouped into a ‘wash’, which represents a set of molecules that will be delivered to the assay and may (or may not) be recognized by the sensors. For the sensor program used in the initial version, some DNA targets may cancel each other out (via complementary) and so a given combination of washes may lead to a very different set of reacting sensors. The standard workflow for the application is as follows: 1. Place a grid of fixed size down to represent the assay array, discretized into individual sensor areas. 2. Define a set of DNA targets, including the full sequences for each, and auxiliary color information for display purposes. 3. Attune the sensors (by position) to the desired set of DNA targets they should react with. 4. Define the washes, the non exclusive groupings of DNA targets that will be used in the experiment. 5. Test combinations of washes, seeing if the sensors produce the desired output: based on the building rules for the sensors and a simulation model for how arbitrary DNA targets interact with the sensor, we can tell which targets will end up bound to the sensor. 6. Repeat 2-5, re-designing sequences, sensor attunements, etc as needed to achieve the desired patterns.
2. CNC Printer
Bio-printing deposition system to import the biological gcode from CAD and deposit proteins to nitrocellulose membrane in specific orientation. Existing Bio-printing systems include the use of robotic liquid handlers, contact spotters, soft lithography stamping techniques and blot spotting. In all cases the machine comprise expensive scientific equipment not usually accessible or affordable outside a well equipped research facility. Moreover, existing systems do not lend themselves well to complex patterning capabilities found in modern digital fabrication equipment. We decoupled the motion stage and deposition mechanism to create a versatile system that is delicate to proteins, is accurate and does not disturb the substrate. HP Ink Jet and cartridge preparation: Consumer grading jet cartridges have been used for protein deposition in the past with positive results. We opted to use an HP CE 5506 cartridge for lack of "kill switch" electronics which made it a viable candidate for refilling with a third-party substance. Sharp blade in a rotary tool were used to remove the top lid. The ink cartridge was cleaned using a sonicator for 20 minutes, then sprayed with 75% alcohol to create a sterile environment. Direct Driving the Ink Jet Printing Head: Native control of the Ink Jet head was the preferred bioprinting approach since it offered direct nozzle control and calibration potential. An InkShield Open Source inkjet shield was purchased (Nerd Labs) and tested using an original HP c6602a ink jet cartridge. Motion control was achieved by pairing the InkShield Assembly to a custom XY stage. The stage was fitted with stepper motors and a belt drive with a XYZ step distance. Motors are simultaneously controlled by a Adafruit Motor/Stepper/Servo Shield for Arduino v2 Kit using a TB6612 MOSFET driver. Deposition: 1.Add 200ul Thrombin 2.Add 500ul 20% BSA 3.Add 1300 ul PBS 4.Add protein mixture to HP c6602a ink jet cartridge. 5. In same ink jet cartridge, add a polyurethane foam sponge cut to width x length of ink jet cartridge and 0.5” thick. We used CVS brand foam makeup sponges for this application. 6.Adjust settings for quantity of liquid deposited. 7.Align nitrocellulose below XY stage. 8.Run arduino code to print protein mixture in programmed pixelated pattern. 9.Nitrocellulose paper with deposited protein is ready for device assembly.
3. Functionalizing with chemistry.We used ssDNA thrombin binding aptamers (TBA) coupled to gold nanoparticles as a colorimetric label with a real world application. Nanoparticles were prepared by the protocol outlined below developed by the Gehrke Lab. Protocol for Synthesis of Gold Nanoparticles 8.5nm: The synthesis consists in the preparation of 2 solutions: A and B which concentrations and volumes are in the following table: 1.Each chemical must be weighted and prepared in a separate glass container. 2.Solution A can be directly poured in a 250ml Erlenmeyer. Mix all the components of solution B in a new glass container. 3.Place the Erlenmeyer with solution A on a plate. Heat the solution to 60-70 ºC (no more than 80ºC), with medium stirring. 4.When the solution A is close to 60 degrees, place solution B (with a lid) on the plate. 5.Once solution A reached 70ºC, pour solution B on solution A as fast as possible. 6.Maintain the stirring and heat for 15 min (make sure the temperature is at least 60 ºC but not higher than 80º). 7.After 15 min, stop the heating (change the plate) and keep stirring until room temperature. Washing step: 1.Wash the solution by centrifuging at 12000rpm for 45 min. 2.Release the supernatant carefully with a pipette. 3.Regenerate the pellet with the same volume of the beginning (with water or buffer) and centrifuge again at 12000rpm for 45min. 4.Discard the supernatant and regenerate the pellet with the same volume or less (this way the nanoparticles will be more concentrated). 5.Make sure the solution is red and not blue. Thrombin binding aptamers were prepared in a SELEX process previously described by Bock et. all.
The device is assembled in a later flow format using Whatman 0.45um nitrocellulose paper with the thrombin deposited in pixelated pattern and 1mm Whatman chromatography paper with 3mm overlap. The papers were secured using 3M Scotch tape on the back side and supported at a 30 degree incline. To reveal the encoded pixelated MIT, we deposited 800ul of binded nanoparticles to the nitrocellulose base assembled diagnostic. After 20 min, the remaining nanoparticles were washed with PBS solution.
We describe a workflow that enables democratized production of interactive paper diagnostics in a time resolvable domain. Through the use of computerized modeling we have the ability to design and simulate DNA display mechanism. By leveraging a more accessible bioprinting platform we can rapid prototype these digital simulations and using specific thrombin aptamer molecules we present a significant application of our work. 1. Design Framework -Defined scope of DNA design using principles of Conway's Game of Life. 2. Application on Project Cyborg -Built application using Project Cyborg platform that implements the design workflow described above (1). -Tested application with a basic sensor implementation that recognizes specific DNA sequences and allows DNA targets to be removed via fully complementary DNA targets. -Ran through sample workflows with up to 4 different targets recognized per sensor location, including testing wash combinations and identifying possible design errors in the DNA targets and wash combinations. 3. CNC Deposition -Fabrication of a custom milled frame with X and Y axis control. -Functionalize InkShield, Motor Control and Arduino software. -Customize arduino code to translate pixelated image to 0/1 to determine deposition, scale image size and adjust the quantity of liquid deposited per pixel. -Successful deposition of DNA on transparency film to generate using HP c6602a ink jet cartridge to generate image “MIT”. 4. Ink Jet Bioprinting -Demonstrated deposition of protein mixture in pixelated pattern “MIT". 5. Chemistry -Demonstrated binding of thrombin and gold nanoparticles
Next steps for DNA disPLAY are outlined below: CNC Deposition: 1. Implement UI ( User Interface) for easier translation of image to 0/1 coordinates 2. Implement multi-reagent deposition Ink Jet Deposition: 1. Modify print settings to control for droplet size Chemistry and Device Assembly: 1. Optimize dimensions of nitrocellulose and absorbent pad 2. Integrate sample pad and conjugate pad 3. Implement multisequence wash to demonstrate programmable pixels Project Cyborg Application: 1. Implement sensor designs used in the experimental work. 2. Implement “Order” action, which allows the automatic collation of sensor design information into a document of actual components that can then be ordered to implement the design. 3. Implement “Print” action, which creates the G-code necessary to run the bioprinting process that implements the design.