This Small Business Innovation Research (SBIR) Phase I project proposes to develop a technology that enables a rapid DNA assembly system for synthetic biology. Wedding recent advances in DNA assembly methods, and the software algorithms used to design those DNA assemblies, the company will develop a platform technology for facilitating an optimized combination of direct synthesis DNA assembly to make large combinatorial libraries. The proposed technology is an enabling technology that will allow scientists to direct their resources to conducting experiments that address primary issues.
The broader impact/commercial potential of this project will be to accelerate the pace of microbe development for companies and organizations that develop valuable proteins, advanced enzymes for industry, or therapeutic medicines. DNA cloning is an everyday practice in the course of both Industrial and University based research. Cloning technology has remained largely unchanged for the last 20 years. As a consequence, researchers waste an inordinate amount of time and money designing and constructing DNA, rather than on designing and conducting experiments. Over the past few years, standardized experimental DNA construction methods have been developed that lend themselves well to automation and rapid assembly of DNA. Process automation is progressing from luxury to necessity, as target applications demand the fabrication of large combinatorial DNA libraries in the search for better antibodies, faster enzymes, and more productive microbial strains. The proposed technology will allow rapid forward engineered biological libraries of recombinant DNA. The commercial availability of this technology will provide a low cost alternative to current methods.
PURPOSE The purpose of this study was twofold. First, we wished to test our novel software for the computer-aided design of DNA. Second, we did an experiment where we made DNA according to the automatically generated instructions our software produced. These instructions are a precise computer generated recipe for synthesizing very long pieces of DNA up to 100,000 base-pairs in length. The motivation behind this preliminary research is that if our software can give us a precise recipe for building DNA, we can then send that recipe directly to machines that can build the DNA for us. Currently, this is done in a hybrid fashion where some things are done by machine, and some by hand. We wish to completely automate this process. If we can fully automate the process of making very long DNA, we can replace the very slow methods used now by a much faster automated ones. This will help the primary users of this technology, biologists who a building therapeutics for diseases like cancer, vaccines against viruses like AIDS and chemists who are making sustainable alternatives to building biochemicals. The research outlined in this report was performed at TeselaGen Biotechnology Inc., San Francisco, and at the Joint Bioenergy Institute, Emeryville, California. OUTCOMES We performed research on how to best transform and extend academic software algorithms for DNA assembly to a system that can solve real world problems. We determined that the process should be broken down into a modular system of DNA synthesis followed by DNA assembly. The synthesis part includes traditional oligo (short single strand bits of DNA ~100bp long) manufacture coupled with polymerase chain assembly (PCA) in order to get double stranded DNA of about 1000bp in length. From that point modern DNA assembly technologies (so called "type-II endonuclease" and "flanking homology") take over to make very long constructs of 40kb or more. We determined that we should focus on the difficult later stages in this process where very long complex combinatorial DNA libraries are being produced. This is because making short DNA is fairly easy to do, it is the long pieces that are difficult to produce, and this is where we can contribute the most to the overall goal of being able to produce arbitrary amounts of DNA of any length and complexity. Using our software, we generated instructions for making a collection of 64 constructs from full factorial combinations of 8 constitutive promoters and 8 translation initiation elements of varying strengths. Then, we used those auto-generated instructions to successfully build the collection using conventional liquid handling robots. We also did an experiment to test if we could do that later stage DNA assembly step on a microfluidic chip. A microfluidic format would use a lot less of the very expensive enzymes used to catalyze DNA assembly reactions. We were able to successfully perform the experiment by re-purposing a microfluidic chip used at Stanford for a different experiment, saving time and money. We were able to successfully assembly the DNA on the "mixing chip" opening up the possibility of much cheaper large scale DNA assembly. CONCLUSIONS We were able to develop our software for automating the construction of very long DNA, as well as complex combinatorial libraries of DNA. We also performed a microfluidic mixing experiment, showing the way towards greatly reduced costs of this type of DNA manufacture. Our goal is to enable researchers to put down their pipettes and stop having to do most types of repetitive cloning as they seek to develop better medicines. Instead, they can design their molecules on the computer, let robots do the busy work needed to push their projects to completion much more quickly and at reduced cost.
