To many (notably Craig Venter), synthetic biology means simply synthesizing large amounts of DNA. This type of synthetic biology actually began in the 1980's, when Caruthers introduced a phosphoramidite-based solid phase DNA synthesis architecture. This allowed ICI and the Benner group to synthesize complete genes encoding biomedically useful proteins and enzymes for the first time. The second example showed how biotechnological goals could better be met if the synthetic sequences were different from the sequences presented by Nature, through codon optimization and watermarking, inter alia. Subsequent developments now allow semi-routine synthesis of genes; these are used in biotechnology, gene therapy, RNA therapy, and elsewhere. Extrapolation suggests that routine synthesis of whole genomes will soon be possible, hopefully for less than the ca. $30 million spent to synthesize one in the Venter laboratory. Unfortunately, DNA has a rich biophysical chemistry that defeats any architecture that relies on autonomous assembly to make large DNA (L-DNA) constructs by simply mixing synthetic fragments, even within recombinogenic organisms. However, two innovations from the FfAME provide a new approach to creating L-DNA constructs. The first is an artificially expanded genetic information system (AEGIS). AEGIS is a DNA-like molecule that adds eight additional nucleobases that form four additional pairs (the Z:P, S:B, K:X, V:J pairs) to the four natural nucleotides (which form G:C and A:T pairs). By increasing information density of DNA and avoiding non-canonical interactions, AEGIS allows autonomous self-assembly of dozens of fragments to generate L-DNA. The second innovation is transliteration technology. Transliteration allows rule-based replacement of AEGIS nucleotides by standard nucleotides after a L-DNA assembly is complete. By converting S, B, Z, and P to T, A, C, and G (respectively), the AEGIS components can be replaced after they have served their role to guide autonomous self-assembly, converting GACTSBZP L-DNA to entirely standard GACTTACG DNA. To persuade commercial partners to engage with this technology, FfAME scientists demonstrated this strategy using the simpler GACTSB six-nucleotide AEGIS DNA to give, in one assembly step, an active, full-length, and sequence-correct gene encoding kanamycin resistance. Parallel attempts with standard 4-base DNA failed. Phase 1 project will transfer these innovations to Firebird, which will deliver L-DNA and whole plasmids by custom synthesis using an eight-letter GACTSBZP alphabet. This will require (a) adapting OligArch software to support design with this strategy, (b) creating a pipeline to synthesize DNA 60-80 nucleotide fragments using this alphabet, and (c) providing demonstration products, plasmids coding multiple enzymes for complete metabolic pathways to natural products, assembled in a single step. In Phase 2, this technology will be merged with Firebird's E. coli SEGUE strain (Second Example of Genetics Undergoing Evolution) that manages expanded DNA alphabets in cloning vehicles, with one week order-to-clone times for plasmids and viruses.
Whole gene synthesis today has a major commercial market in biomedical research and clinical practice, with (for example) genes that include an RNA polymerase promoter being used to generate whole messenger RNA transcripts to be used in RNA-based gene therapy. The expensive part of whole gene construction is not the synthesis of the primary DNA oligonucleotide gene fragments, which have now become quite inexpensive. Rather, the cost is the assembly of those fragments to give the full gene, a process that requires considerable human involvement and risk of failure. Recently, scientists at the Foundation for Applied Molecular Evolution reported a technology that allows autonomous self-assembly of dozens of the gene fragments, largely without human attention [Merritt et al. 2014]. Once transferred to Firebird, this technology will support large DNA construction with applications ranging from biomanufacturing to diagnostics to therapy.
