doi:10.1038/nindia.2011.53 Published online 21 April 2011
When Craig Venter's team carried out the first synthetic genome transplantation some time back, it made international news. Though the popular press equated it to creating life or playing God, the experiment was not exactly that. The team had manually replaced the operating system of the existing chassis of a microbe with a new DNA molecule, step-by-step. It was the first such experiment of genome transplantation with the computer playing the parent.
Designing user-defined genome, and by extension an organism, still remains one of the key unmet goals of synthetic biology.
The design of synthetic genome by Venter's team was based on existing information. They computer-edited the existing DNA data, divided it into blocks, ordered synthesis of each block separately, assembled the final version with suitable watermarks and transplanted the reconstructed genome into an existing cell. This was the first evidence that a cell could be made to reboot with a made-to-order DNA.
This process of DNA synthesis, assembly and transplantation might sound simple but in reality it is time consuming, prohibitively expensive and not scalable. For example, in Venter's case assembling 1.08 Mbp genome took more than a decade and tens of millions of dollars.
Ideally, constructing a brand new organism would be efficient if each part is defined quantitatively, an inventory of genome parts is built, computational tools are used to model their interactive behaviour, the compiled sequence is computationally 'tested' and a biotech company ordered to chemically synthesize it. However, science hasn't reached a stage where AutoCAD like environment can be used to design and synthesize genome with predefined properties. For the DIY biology to be retailed, a lot of fundamental biology at the structure and interaction level needs to be worked out.
Given that synthetic biology is taking rapid strides in the direction of artificial DNA synthesis, the fundamental question is – what are the key drivers to enable routine and customized synthesis of genomes in future? Perhaps an error-free and low-cost DNA synthesis technology is essential to enable the paradigm shift from traditional recombinant DNA methods to custom ordering DNA.
Interestingly, though the technology for long DNA synthesis looks modern, its seeds were sown in the 1970s when H G Khorana's lab completely synthesized tRNA structural genes, . Since then, the DNA synthesis technology has made enormous strides in terms of quality and turnover rates.
Among several methods to synthesize contiguous pieces of long DNA sequences, the polymerase cycling assembly method is probably the most common. In practice it is straightforward and scalable. The chemically synthesized DNA is biological indistinguishable from naturally evolved DNA. Further, the community is also sensitive towards targeted replacement of the existing genetic code with more efficient codons without affecting the final protein sequence. The hope is to optimize codons to produce much higher protein expression. This means that for producing the same amino acid, one codon works better, say in one bacterium, and another in yeast.
Even though the cost of synthesizing long DNA sequences has gone down five times in the last decade, the profit margins are still not so attractive. With the rapid expansion of the user base, there is a fair chance that the technology would become commercially viable soon. Thus, instead of spending months in designing recombinant vectors, one could email a long DNA sequence to a company for chemical synthesis. Further, as the computational tools develop, one could order a computer-designed genome to execute predetermined health and environmental applications, in-vivo.
In the late 1970s, hundreds of base pairs were the upper technological limit of chemical synthesis. Today, more than half a million base pairs of DNA (0.5 Mb) can be chemically synthesized error free. Likewise the cost per base of DNA synthesis has exponentially fallen in the last 10 years – from several dollars per base to less than half a dollar. This trend of falling synthesis cost is likely to continue.
To scale up this effort at the chromosomal level, the synthesis of contiguous sequence of millions of nucleotides requires enormous technological advances in the accuracy, scale and processing of longer DNA assembled from short fragments. A method called megacloning promises to overcome these barriers. The beauty of this technique is that it can reduce error rates by a factor of 500. In principle, genome synthesis can be scaled up to the megabase level.
Commercial rates of DNA synthesis for short oligos range in the cost of 0.10 – 0.50 dollar a base. Given the emergence of novel DNA synthesis technologies, it is expected that the speed will increase and cost will decrease due to the reduced reagent usage.
Once the DNA synthesis costs reach the cost of making oligos, parts of recombinant DNA technology may get obsolete. It would be more efficient then to order a vector instead of designing one!