Genetics just got SEXY: Sequences encoding XY.

A semi-synthetic organism with an extended genetic alphabet heralds a new era in synthetic biology.

In the May 15th 2014 issue of Nature, Floyd E Romesberg and colleagues unveiled the first semi-synthetic organism with an expanded genetic alphabet; heralding a new era in synthetic biology.

Since the dawn of life on earth, all genetic information has been coded by just 5 natural bases (Adenine [A], Guanine (G], Cytosine [C], and Thymine [T], which is replaced with Uracil [U] in RNA). However, it wasn’t until the 1960s, a decade after Watson and Crick revealed that the structure of DNA is mediated by the specific pairing of A to T and G to C, that scientists began to consider the possibilities of incorporating synthetic analogs of the natural bases as additional functional pairs. Almost 30 years later, Steven Benner, a pioneer of synthetic biology, described the incorporation of non-standard amino acids into polypeptides by ribosome-based translation. Benner’s team described the expansion of the genetic code through the creation of a 65th codon-anticodon pair from unnatural base pairs (UBPs) exhibiting non-standard hydrogen-bonding patterns. This non-canonical codon-anticodon pair supported translation in vitro to yield peptides containing a non-standard amino acid. Schweitzer and Kool extended these findings, showing that hydrogen bonds are not necessarily required for the formation of a stable duplex DNA-like structure, suggesting that hydrogen bonds are more important for specificity of pairing than for affinity in DNA. The ability to substitute hydrogen bonding with hydrophobic or indeed steric interactions enabled the design of increasingly more versatile nucleotide analogs.- An added advantage of the differing chemistries (i.e., hydrophobic interactions vs. hydrogen bonding) is that it prevents the natural and synthetic nucleotides from mispairing with one another. This repartitioning could potentially be used to design hybrid systems: semi-synthetic cells operating two separate genetic codes. The native code would run normal cellular processes, while the parallel synthetic code would allow the cell to act as a micro-factory producing novel proteins.

In 2012, Romesberg and colleagues, having synthesized and tested more than 300 artificial nucleotides, developed a class of UBPs, exemplified by d5SICS-dNaM (abbreviated as X and Y), formed between nucleotides bearing hydrophobic nucleobases. In vitro testing, involving PCR and PCR-based applications, found d5SICS-dNaM to be functionally equivalent to natural base pairs. However, while replication of UBPs in vitro is one thing, in vivo replication was a completely different proposition. Until now, the major stumbling block in moving from in vitro synthesis to in vivo replication was in getting the synthetic nucleotide analogs inside the bacterial cell in the first place. While DNA integrated nucleotides contain a single phosphate group, they require two additional phosphates before being incorporated into replicating DNA. Removal of the extra phosphates is achieved via a dephosphorylation reaction, providing the energy to power the replication process. Passive diffusion of free nucleosides into the cell, followed by conversion to the corresponding triphosphate via the nucleoside pathway, proved inefficient. Given that in vivo analog assembly from precursors was unlikely to work, at least in the short-term, Romesberg’s group circumvented this step by engineering the host cell to accumulate the pre-formed nucleoside triphosphates. This simple fix elegantly incorporates a failsafe bio-containment measure; because the E. coli host cannot synthesize the analogs, they need to be added exogenously. In the absence of unnatural triphosphates in the growth medium, the UBP is eventually lost. Furthermore, this loss was found to be the result of replication-mediated mispairing as opposed to the activity of DNA repair pathways.

The researchers exploited the fact that certain intracellular bacteria and algal plastids fail to synthesize their own nucleotides, instead accumulating them from the external environment via nucleotide triphosphate transporters (NTTs)., Of 8 NTTs tested, the Phaeodactylum tricornutum NTT (PtNTT2) proved most effective when heterologously expressed against an Escherichia coli host, resulting in 90 µM d5SICSTP and 30 µM dNaMTP in the cell cytoplasm 30 min after addition to the growth media. In addition to the PtNTT2 encoding plasmid, named pACS (for accessory plasmid), the E. coli host also harbors pINF (the information plasmid). Based on the pUC19 backbone, the in vitro synthesized pINF harbors a dTPT3-dNaM pair in place of dA-dT at position 505, 362 bp downstream of the ColE1 origin of replication and within the TK-1 Okazaki processing site, where DNA replication is mediated by Pol I (previously shown to efficiently replicate DNA containing d5SICS-dNaM in vitro). The use of dTPT3, an analog of d5SICS, in the in vitro synthesized pINF provided an efficient means of confirming in vivo replication; replacement of dTPT3 with d5SICS, supplied exogenously in the medium, only occurs if the plasmid is replicated in vivo. Indeed, Romesberg’s team found that DNA containing the UBP is replicated in vivo with at least 99.4% fidelity, corresponding to an error rate of ~10–3, equivalent to that seen in natural DNA.

Having proved that UBPs can be readily incorporated and replicated in DNA, the next step is to demonstrate that they can be transcribed into RNA in vivo. Such modified RNAs might well lead to improved functional RNA elements such as riboswitches, ribozymes, and ribonucleoproteins—an entire suite of new synthetic biology tools. However, perhaps the most exciting aspect of this research is the ability to artificially extend our genetic alphabet. While the natural genetic code builds proteins with just 20 amino acids building blocks, Romesberg’s expanded code incorporates up to 172 amino acids, circumventing the need to recode the translational functions of existing codons as proffered by Lajoie et al.

The obvious commercial potential of the research is not lost on Romesberg who is co-founder of Synthorx (http://www.synthorx.com), a San Diego based biotech company, the launch of which coincided with the publication of the Malyshev et al., paper. A stated aim of Synthorx is to exploit the potential of UBPs to improve the discovery and development of new drugs, diagnostics, and vaccines as well as creating innovative products, including research reagents, aptamers, and nanomaterials.

A mere 53 years after Nirenberg and Matthaei’s pioneering work to crack the genetic code, we stand poised to write a new one. In Romesberg’s own words; the increased information offered by this newly expanded code will allow us “to write more interesting words, bigger words, more complicated words, more nuanced words, better stories.”