doi:10.1038/nindia.2012.189 Published online 19 December 2012
A paradigm shift in molecular biology has already set in. It is now possible to email a DNA sequence to a company and receive a recombinant plasmid vector of choice. We are soon heading into an era where construction of recombinant DNA will be through chemical synthesis, instead of conventional methods of putting together DNA from various sources.
Recombinant DNA technology helps generate specific fragments of DNA sequences in various organisms and glue them on a common backbone for expression and analysis. DNA sequences sourced from various organisms perform specific roles in their native state. To integrate sequences on a shared platform and study them, one needs a vector — usually a circular piece of DNA called plasmid —that replicates along with the cell. The choice of vector depends upon the length of the inserted DNA sequence, the host where the new sequence needs to be expressed and so on.
Depending upon the available technology, it can take several weeks or months to complete the process of culturing the cell, isolating DNA, preparing DNA fragments, correctly placing them in a predetermined order on the vector, transferring them to a non-native host, identifying cells that have accepted the recombinant plasmid and grow these cells for future studies. One also needs to take care that these sequences stay unmodified throughout the relocation process.
Given that recombinant DNA technology is central to the molecular understanding of biology, the question is: can something be done to speed up the design and production of recombinant vectors? In other words, can we accelerate genetic engineering?
In June 2004, synthetic biology was announced as a formal discipline in a meeting at the Massachusetts Institute of Technology (MIT). The meeting realised that synthesis of long DNA sequences could be a game changer in future. A desktop DNA printer was envisioned as standard equipment in a molecular biology lab of the future. Thus, instead of designing a recombinant vector, one would synthesize a custom-made vector, leading to enormous savings in cost and time.
The question is: what kind of preparation is required to enable such a phase shift?
The practice of artificially synthesizing DNA can be traced back to the 1960s when Har Gobind Khorana invented novel chemical methods to synthesize nucleotides1. It took his group five years of rigorous work to synthesize 17 oligonucleotides leading to standardization and synthesis of more than 200 base pair gene sequences several years later2. The 1990s saw Polymerase Chain Reaction (PCR)-enabled biochemical synthesis of longer DNA sequences3.
The decade of 2000 saw a huge jump from gene synthesis to whole genome synthesis. In 2002, an infectious polio virus was chemically synthesized — its genome was approximately 7.5 kilo base long4. Six years later, Hamilton Smith's lab designed the Mycoplasma genitalium genome from scratch, using chemical and biochemical methods5. Though claims vary, currently the longest documented chemically synthesized DNA seems to be of more than 500 kilo base.
This covers the size range of viruses and a number of bacteria and is close to the size of plasmid vector.
Given the rapid pace of technological development, one would expect custom ordering of recombinant DNA vector as a norm in future. However, that hasn't happened so far. It turns out that the key rate limiting step is the high cost of error-free long DNA synthesis.
In early 2000, the cost of DNA synthesis per base was several dollars. However, with regular improvement in synthesis methods, the cost of DNA synthesis has been going down along with corresponding increase in the ability to synthesize longer DNA sequences.
Though the overall cost of DNA synthesis varies with the length and complexity of a given sequence, the cost of synthesizing a base has now touched 30 cents a base. This is still three times higher than short oligo synthesis (10 cents a base).
With a further improvement in the DNA synthesis methods, it is hoped that the synthesis and sequencing cost lines will meet in the near future.
Currently, full length genes can be delivered within two to four weeks. Researchers are now trying to make the method cheaper and accurate.
To achieve significant improvement in the accuracy of DNA synthesis, a DNA mismatch binding protein called MutS has been used to locate DNA synthesis errors, bind to defective DNA sequences, remove them and enrich the fraction of correctly synthesized DNA. MutS seems to have reduced the error rate by 15 times over conventional methods6. Recently, George Church's group at Harvard reported a one-step error free DNA synthesis method to assemble one kilo base of DNA from 48 oligos using microarray based approach7.
The traditional methods of cutting and pasting DNA seem to be rapidly heading in the direction of "making recombinant DNA by synthesis".
However, for the synthesis route to be adopted by the scientific community, it is important that the mutation rates are minimized, synthesis steps made parallel, technology miniaturised and oligo waste eliminated.
This calls for developing radical new methods of gene synthesis. It appears that in future, bio-CAD (biological computer aided design) tools will be employed to design DNA sequence models and the text file will be emailed to companies for rapid chemical synthesis of recombinant plasmids and microbial genomes.