by Matthew Cobb
Perhaps the most exciting recent development in DNA information storage has been the announcement by Fahim Farzadfard and Timothy Lu of MIT that they were able to engineer a population of bacterial cells to record real-time data as the cells were treated with a particular chemical inducer – or, as Farzadfard and Lu put it, they created ‘genomic “tape recorders” for the analog and distributed recording of long-term event histories’.14 This information was stored in a special form of DNA that only has one strand; it could be written and rewritten, and could be recovered after several days. This breakthrough brings the development of organic sensors, for both environmental and medical uses, much closer.
*
Although we generally describe the structure of DNA as a double helix, it is in fact a bit more complicated than this. Unlike a screw thread, which has a constant pitch or interval between each turn, the double helix has two different intervals, which alternate as the molecule spirals round. These are known as the major and minor grooves – the major groove, which is larger, tends to be the point at which DNA-binding proteins affect gene activity, as the sequence of bases is physically more accessible there. Above all, the DNA double helix spirals in only one direction – anticlockwise as seen from the top, or right-handed, like a normal screw. It is easy to get confused about which way the double helix should spiral, and in many representations of DNA the molecule spirals the wrong way. In 1996 Tom Schneider began posting images of leftward spiralling double helices on his web site, but he was soon overwhelmed by the number of examples.15 I will not cast the first stone over mistaken representations of DNA – I once managed to put a wrong double helix on the side of a building: it was not even left-handed, it was geometrically impossible.
The DNA double helix comes in several shapes. The two forms studied by Wilkins and Franklin – A-DNA and B-DNA – both have right-handed helices; B-DNA is the iconic version that exists in your cells, and A-DNA occurs under conditions of low humidity and can be found in organisms although its biological function (if any) is unclear. In 1961, a group including Wilkins observed a third right-handed form of DNA, known as C-DNA, which appears in the presence of particular salts and has a slightly different structure again.16 Left-handed DNA, known as Z-DNA, can be found in our cells. In one of the ironies of history, the first structure of a DNA molecule to be determined precisely was of the Z form, in 1979.17 For complex chemical reasons, the Z form lacks the minor groove, and it turns in a looser helix than the B form – it has twelve base pairs per twist, whereas the B form has ten base pairs.18 It is not simply an elegant left-handed version of the B form, but a kind of twisted Bizarro-DNA, with bases turned upside down relative to the B form, so the phosphate backbone of the molecule forms a zigzag rather than a smooth spiral.19 The function of Z-DNA is still being explored – early hopes that it would turn out to be of importance in gene regulation or a useful tool in biotechnology have yet to be fulfilled.20 DNA can also form other structures, including the four-stranded G-quadruplex and a cruciform shape; although it is assumed that these non-helical structures have some functional role, probably in regulating transcription, the evidence is still unclear.21
DNA is not the only molecule that can form a double helix. In 1961, Watson and Crick, together with Alexander Rich and David Davies, suggested that in certain circumstances, RNA, which is normally single-stranded, could double up.22 Over half a century later, researchers were able to crystallise double stranded RNA and to describe its structure. It, too, has a right-handed spiral.23 There is no evidence that the RNA double helix has any biological function in normal cells, but it may be possible that biotechnology will be able to employ this novel molecular structure. Although RNA is generally presented on diagrams as a single strand, in fact RNA molecules often bend around on themselves, forming complex double-stranded stems with a single-stranded loop at the top, like a hairpin. Such secondary structures may be important in the various functions of RNA – they give tRNA its distinctive shape, for example.
Both DNA and RNA are made of a common ribose-phosphate backbone, onto which are linked the bases (A, T, C and G) that carry the genetic information. In 2012 a collaborative project led by Philipp Holliger at Cambridge described the creation of six new kinds of informational molecule that did not use ribose, dramatically called xeno-nucleic acids or XNA. In place of ribose, these weird molecules each have different forms of sugar in their backbones onto which the usual bases can be attached. As well as creating these six forms of XNA, the group also engineered the enzymes necessary to enable DNA to be copied into XNA, and for XNA to be copied into DNA – this was a truly remarkable feat.24 In a series of experiments the group demonstrated that genetic information (that is, the sequence of bases) could be successfully copied from DNA into XNA and back again. They were even able to subject one of the XNAs to selection and to show that its sequence evolved as a result. In principle, DNA and RNA are not the only potential informational molecules. Alien life-forms – if there are any – may well use non-DNA or non-RNA information.
The biotechnological potential for XNA is immense. As Holliger’s group concluded:
‘synthetic genetics’ – that is, the exploration of the informational, structural, and catalytic potential of synthetic genetic polymers – should advance our understanding of the parameters of chemical information encoding and provide a source of ligands, catalysts, and nanostructures with tailormade chemistries for applications in biotechnology and medicine.
Commenting on the creation of XNA, the veteran biochemist Gerald Joyce recognised that it opened the road to what he called an alternative biology, but he also sounded a warning.25 Use of DNA-based and RNA-based synthetic molecules carries a fail-safe mechanism in that they are susceptible to degradation by enzymes that have evolved over billions of years – indeed, this is one of the obstacles that restricts their widespread use. Furthermore, all DNA-based life-forms are susceptible to attack by other organisms and the enzymes they contain. This would not necessarily apply to XNAs. As Joyce put it:
XNAs are unnatural and would pass through the biosphere unscathed. The benefits of their unusual chemical properties must be weighed against their greater cost, both literally and with regard to operating in the uncharted waters of XNA biochemistry. … Synthetic biologists are beginning to frolic on the worlds of alternative genetics but must not tread into areas that have the potential to harm our biology.
For the moment, no one has been able to create an informational molecule that does not contain phosphate. In December 2010 a twelve-person team including the NASA astrobiologist Felisa Wolfe-Simon published an article online in Science, suggesting that bacteria found in Mono Lake in California, which has high levels of arsenic, naturally replace the phosphorus in their DNA by arsenic.26 This claim, which was announced at a high-profile NASA news briefing to the excitement of the world’s press, was immediately contested on social media, such as Dr Rosie Redfield’s blog RRResearch. What became known on Twitter as #arseniclife took on the proportions of a major scientific row and also showed the power of social media to act as a form of peer review. In 2011, when the original paper was finally published in Science, it was accompanied by an unprecedented seven short articles that were critical of the paper’s claims. Redfield and others eventually published papers in Science that showed that, in this case, there was no integration of arsenic into DNA.27 The original Wolfe-Simon paper has not been retracted, and it remains possible that, perhaps on another world or eventually in an Earth laboratory, other forms of informational molecule without phosphate may exist.
The #arseniclife debacle was partly driven by an idea put forward in 2005 by Carol Cleland and Shelley Copley of the University of Colorado, who published a speculative paper exploring the possibility that our planet hosts microorganisms that do not use DNA or RNA or our set of amino acids, and which are therefore undetectable by traditional methods such as the polymerase chain reaction.28 This hypothesis, which has been given the dramatic names ‘the shadow biosp
here’ or ‘weird life’, has attracted some interest from those prone to theoretical speculation, but has not been treated with any degree of seriousness by the scientific community. As Cleland and Copley put it, ‘the fact that we have not discovered any alternative life forms cannot be taken as evidence that they do not exist’. That is logically correct, but it is hardly an enticing starting point for a research programme and would not be taken seriously by any funding agency.
Physicists have realised that 95 per cent of the Universe is made of stuff we cannot directly detect – dark matter and dark energy – because calculations of the amount of matter in the Universe based on gravitational effects revealed a substantial discrepancy between observed and expected values. For the shadow biosphere to be more than day-dreaming, a similarly overwhelming signature of its existence would be necessary. Those intrigued by the remote possibility of weird life have come up with some potential indicators of its existence, such as the supposedly anomalous varnish that is found on desert rocks.29 However, something as impressive as the indirect evidence of the existence of dark matter and dark energy would be required for this hypothesis to be taken seriously.
*
The potential for synthetic biology is enormous. Scientists are already able to integrate unnatural amino acids into proteins, for example by manipulating enzymatic machinery associated with an ‘amber’ stop codon (UAG) that has been introduced into bacteria, yeast, the nematode worm Caenorhabditis elegans and even mammalian cells.30 A synthetic form of transfer RNA is used to allow the UAG codon to code for an unnatural amino acid, thereby producing a novel protein. Some of these engineered proteins produce pulses of light when they are subjected to particular forms of biochemical activity in the cell, acting as an exquisitely sensitive marker. Some produce modifications in the three-dimensional structure of the protein, allowing greater understanding of the precise organisation of the molecule; others enable light to be used to activate molecules in the cell, giving an insight into the roles of key components of the cellular machinery. For the moment, many of these novel proteins are aimed at increasing our understanding of basic processes. But it is only a matter of time before they will be used to develop new forms of biotechnology with potentially massive implications for our future.
The ability to manipulate DNA has recently been extended to altering the genetic code itself. Although only two base pairs occur in nature (A binds with T and C binds with G), unnatural base pairs can be used to make novel forms of DNA in test tube reactions. More than twenty-five years ago, Steven Benner’s research group extended the alphabet of the genetic code by introducing two new base pairs into DNA and RNA molecules – one pair is known as κ and Π, the other as iso-G and iso-C.31 In 2011, Benner’s group was able to amplify and sequence DNA containing the four usual bases and an unnatural base pair (Z and P), which they called GACTZP DNA.32 For decades, the manipulation of base pairs has been used in some forms of everyday antiviral medicine, such as those regularly used by cold sore sufferers. The cold sore virus is persuaded to replace the G bases in its sequence by one of several proprietary molecules that cannot be copied by the infected cell’s machinery, thereby blocking reproduction of the virus.33
In 2014, a group of researchers led by Floyd Romesberg of the Scripps Research Institute in California took a giant step towards the creation of a truly synthetic organism when they were able to make Escherichia coli bacteria copy a piece of DNA containing a sequence that involved an unnatural base pair (the two new bases go by the unfriendly names of d5SICS and dNaM), which had previously been created in a test tube.34 Where the E. coli DNA replication machinery found a d5SICS in the artificial molecule, it inserted a dNaM on the new complementary strand, and vice versa. The bacteria seemed quite happy with this new alphabet, showing no serious problems, and the classic DNA repair pathways in the cell, which normally snip out and repair errors, did not make any move against the intruding pairs of bases.
For the moment, this remains at the level of a technical breakthrough – as the title of the paper put it, the researchers had created ‘A semi-synthetic organism with an expanded genetic alphabet’. The two extra letters that were introduced into the DNA are currently mute; the expanded genetic alphabet does not yet form new words. But the authors made clear that their aim is to create a system in which unnatural base pairs will code for unnatural amino acids. The possibilities for synthetic biology – using living cells to produce new molecules – are almost endless.
It is inevitable that these astonishing developments in our ability to manipulate the essential elements of life will soon come together. Someone will eventually create a system in which XNA carries unnatural base pairs that code for unnatural amino acids that are assembled into bizarre proteins, and an utterly novel form of life, created entirely through human ingenuity, will be only a step away. It does not seem too outlandish to imagine that by the end of the century, entirely synthetic life-forms will exist, able to produce drugs, foods and novel compounds at the service of humanity, perhaps in the most inhospitable of environments. Synthetic organisms able to survive in low levels of oxygen and cold temperatures might be one way in which we could terraform Mars, should it prove to be barren, and should we consider it ethical to destroy such a pristine environment. In this respect, as others, science and technology pose questions; they do not necessarily provide the answers.
*
Despite the optimism that surrounds synthetic biology and genetic engineering, ever since the first appearance of these techniques in the 1970s scientists have consistently expressed concern about the potential dangers. With the development of restriction enzymes (proteins that will snip a piece of DNA in two at a defined sequence) it became possible to create what is known as recombinant DNA – DNA from more than one organism, generally from two different species. This led scientists, including those involved in pioneering the approach, to be concerned that the introduction of new genes into organisms could have unforeseen consequences. They were particularly worried about what would happen if the organisms escaped and transferred their genes into the wild or if the new genes were inherently dangerous to humans or the ecosystem. This was not just an abstract concern: the era of genetic engineering was heralded by the introduction of SV40, a viral gene that can cause cancer in rodents, into the DNA of a bacteriophage virus, which was then used to transform E. coli.35 Faced with the novelty of this technique, it was legitimate to worry that the transformed E. coli might end up inducing cancer in humans. The experiment, by Paul Berg, David Jackson and Robert Symons, was published in 1972. Eight years later, Berg won the Nobel Prize in Chemistry for this feat, together with Wally Gilbert and Fred Sanger, who were recognised for their work on DNA sequencing. Within a year of publication, Berg, along with other scientists, was arguing for a partial moratorium on recombinant DNA research because of the potential dangers.36
In February 1975, a conference took place at Asilomar, on the edge of Monterey Bay in California, to discuss the risks associated with the new technique and above all how to minimise the dangers. The conference, which included journalists and lawyers among the attendees, adopted a set of laboratory procedures, including strict containment facilities and biosecurity measures, which would enable research to continue safely. Many of these are still in force, but others have been abandoned as it has been realised that the dangers are far less than was originally feared.37 In my own field, the study of behaviour in Drosophila, the introduction of DNA from other species into the fly’s genome has become widespread in order to mark and manipulate tissues, enabling us to turn genes on and off, simply by allowing two flies to mate. The technique is perceived as entirely risk-free, and recombinant fly stocks, which may contain genes from yeast, jellyfish or bacteria, are routinely sent around the world by ordinary post and are used in ordinary laboratories, with no restrictive containment procedures.
However, genetic engineering can pose very real dangers. In June 2014, a group of US and Japanese scientists, led by
Yoshihiro Kawaoka of the School of Veterinary Medicine at University of Wisconsin-Madison, attempted to recreate the Spanish Flu virus, which killed millions of people after the First World War.38 As is well known, we are in danger of another global flu pandemic, with avian flu being the most likely source because it seems also to have been the source of the Spanish Flu. Kawaoka and his colleagues took bits of avian flu virus that were similar to the Spanish Flu infectious agent and put them together in a new DNA sequence, which proved to be highly infectious, just as the Spanish Flu virus was.
They justified their study by arguing that it would help identify which are the most dangerous parts of the viral genome and would therefore increase our preparation to meet any future outbreak. Although the US National Institute of Allergy and Infectious Diseases, which funded the study, defended the research both in terms of the information it provided about the potential dangers of newly emerging flu strains and the stringent biosecurity measures that were applied, there are clear dangers. The newly created virus could escape, or it could conceivably be used by a hypothetical group of bioterrorists, although this would require them to breach the stringent security procedures around such facilities and to be highly trained microbiologists.