What is synthetic biology?

‘Synthetic biology’ is a term that has attracted an enormous amount of attention in the last few years, not to mention funding. But what is synthetic biology? The short answer: many things.

The term ‘synthetic’ is itself worth examining, for its multiple meanings tell us something about the ground to be explored—especially of the traditional divides that set the stage for this efflorescence, and of the various ways efforts in synthetic biology propose to challenge these divides. So I start here, with the meanings of ‘synthetic’.

In one quite familiar sense of the term, synthetic is understood as the opposite of analytic, but as such, its meaning can still point in a variety of directions. In logic, it refers to propositions: as Kant employed the distinction between analytic and synthetic, analytic refers to propositions that are true by virtue of the meaning of their terms, and synthetic to propositions that are not. But in fact this classical meaning has little to do with the ways in which the term is used today in synthetic biology.

More commonly, the analytic/synthetic distinction is intended as a methodological one: analytic implies analysis, a taking apart, and synthesis, a putting back together. But here too, even in this common usage, the distinction is ambiguous, referring to two quite different contexts: conceptual and physical. We might theoretically analyze a problem into either its first principles or its basic elements—sometimes referred to as the reductionist approach—or we might, again theoretically, address the challenge of putting them back together, of deducing the consequences from first principles or from the interactions of the basic elements that have been identified. Indeed, an image frequently evoked in the early celebrations of the new systems biology was that of putting Humpty Dumpty back together again.

Alternatively, however, we might think of this distinction in material terms, as in first disassembling a machine into its components (analysis) and second, reassembling it, piece by piece (synthesis). This shift from conceptual to material reconstruction simultaneously marks the rebirth of synthetic biology and its distinctiveness from systems biology. Interestingly, analysis always comes first whenever analysis and synthesis are paired together terminologically (and philosophically), and often (though not always) when the terms are paired methodologically as well. First we take things apart, and, or so goes the usual presumption, then we try to put them back together, using the understanding analysis had given us. The practice of analysis and synthesis in conceptual terms is generally referred to as science; but when it is performed in terms of concrete physical components and involves literal reconstruction of the system in question, it is generally called engineering. People often ask about the relation between systems and synthetic biology, and I suggest the answer lies just here: systems biology never promised to put a real-world Humpty Dumpty together again (although some systems biologists are hard at work at virtual, or in silica, reconstructions of the biological cell); by contrast, real-world reconstruction of the system is precisely what is promised by almost all the different strands of synthetic biology.

But there is something curious that has happened to the term ‘synthetic’ in this division of labor between systems and synthetic biology: for the latter, the meaning of synthetic has become narrowed to that of physical reconstruction, leaving conceptual synthesis somewhere else, either to systems biology or off the board altogether. At first glance, this seems to do nothing more than place its own activities squarely in the realm of engineering. But in fact I think a more radical move is at work, for in much of synthetic biology today, the very link between analytic and synthetic is loosened here, and to the extent that that link is still present, it is far from clear that analytic comes first, either terminologically or methodologically. Instead, there is at least the suggestion of an inversion, of what Paul Foreman (2007) refers to as the new primacy of technology over science. Or perhaps not simply an inversion, but a replacement, carrying the implication that real-world synthesis is science. What is of particular interest in many of these discussions is the repeated assertion that making is knowing, that the building of a machine is not only of obvious practical utility, but that that process is also, in itself, the royal route to an understanding of the machine. Making is knowing, and knowing is making. As Anthony Forster and George Church put it: ‘Until we can assemble a form of life in vitro from defined, functionally understood macromolecules and small-molecule substrates, how can we say that we understand the secret of life?’ (2007: 5).

Knowledge as making might be said to be the first credo of synthetic biology. A closely related credo is the at least tacit denial of any meaningful distinction between science and engineering. In this sense, synthetic biology might indeed be said to be the poster child of what we in science studies call technoscience, and I will talk about this at greater length in a bit. Bur first I need to point out that the term ‘synthetic’ also has another meaning that is absolutely crucial in this business, but this other meaning works along a rather different axis. Here, the opposite of synthetic is not analytic but real, authentic, natural. By contrast, the reference of synthetic is something not-real, inauthentic, artificial, false. As the title of a recent article in the New Scientist tells us, the efforts now under way under the name of synthetic biology may lead to a ‘second Genesis’, giving us life, but not life as we know it. Synthetic life. Similarly, the field referred to is not biology, but synthetic biology.

Technoscience or artificial biology?

These two meanings of synthetic biology—first as the epitome of technoscience, and second as an alternative, artificial biology—pervade most if not all the various efforts that currently go under that name, and they are deeply entangled. I will try to illustrate some of the ways in which this entanglement plays out in three quite different examples of synthetic biology, one current and two historical. But first I want to begin with a brief aside—some parenthetical remarks about the meaning, and innuendoes, of the term ‘technoscience’. This term was originally coined in the late 1970s by the Belgian philosopher Gilbert Hottois, and subsequently popularized in the science studies community by Bruno Latour, Donna Haraway and many others. The term came to be employed to underscore the illusory character of the classical divides between representing and intervening, looking and touching, pure and applied science, and to do away with these divides. It was a manifestly polemical intervention, aimed at debunking the myth of pure science, and recommending that henceforth we will not speak of science and engineering (or of science and technology), rather, we will speak only of the real world of technoscience.Footnote ¹

I myself participated in this demystification, arguing that there was no representing without intervening, no looking without touching and, furthermore, that the evocation of such a divide was itself a form of intervening, if not in the physical world then certainly in the social world. My favorite example of the complicated cultural and political work that divide performed was the classic image of Albert Einstein, the symbol of pure science, juxtaposed with his creation, E = mc², an equally classic icon, but hardly of pure science.

Furthermore, I argued that the double-speak of the relation between pure and applied science was a cornerstone of the funding policy for basic science throughout the post-war period: we needed to fund basic (or pure) science because of the applications it would bring us, most notably, for the military prowess it would underpin. But for me, this argument did not come easily. I am of a generation that was born and bred on the dream of a pure science. Indeed, I was lured into physics by the fantasy of creating a world out of pure thought, by the belief that the physical world at least was comprehensible, if only you thought hard enough. My sympathy with Rosalind Franklin ran deep when she was quoted in the film The race to the double helix as saying: ‘I just want to look, I don’t want to touch.’ For me, therefore, recognition of this dream as a phantasm was wrenching.

These days, however, we do not use the term ‘technoscience’ as a political/intellectual intervention, or as a statement of disillusionment, but simply as a characterization of the way things are. Furthermore, the naturalness, the ease, with which we speak of technoscience today is not simply a testament to the acuity of scholars in science studies, but rather to the real-world changes that have taken place over the last three decades in the conduct of scientific and technical activities in our academic institutions. However apt the term may have been as a descriptor of such activities in the past, there is no question that technoscience has become an increasingly apt descriptor of our current state of affairs. Our institutions are reorganizing into new configurations, traditional disciplinary boundaries are blurring, and the classical divide between science and technology is less and less meaningful to new generations of practicing scientists.

I recall as a telling example the debate around Frances Arnold’s tenure in the Division of Chemistry and Chemical Engineering at Cal Tech some years ago. Arnold is known for her ground-breaking work on directed evolution, which she describes as follows:

Her tenure dispute revolved around the question of whether this was science or not, a question raised by (at least some of) the senior members of the committee. The junior members, who apparently prevailed, failed to see any basis for such a question. Over the years since, the larger scientific community seems to have similarly rejected the question. Today, Arnold is a member of the National Academy of Science, and the recipient of the ‘Excellence in Science Award’ awarded by the Federation of American Societies of Experimental Biology (FASEB) in 2007. In other words, what for me came as a rude awakening, seems to be for many today a banal observation.

But perhaps the pendulum has swung too far. While there may be—arguably, even can be—no looking without touching, it does not follow that looking and touching are the same. It might not be possible to make any clear distinctions between science and engineering—certainly it is impossible to disconnect the workings of hand and mind, but I suggest that the radical dissolution of all such distinctions erases more than disciplinary boundaries. It erases distinctions between a variety of conceptual aims that may not sort by old boundaries but may still sort in ways that need marking. Perhaps nowhere does my unease surface more clearly than in thinking about the new interdiscipline of synthetic biology—as I have already suggested, a set of activities that serve as an illustration par excellence of the extent to which older divisions between science and technology have evaporated. It is probably fair to say that no discipline is unitary, but the term ‘synthetic biology’, as it is used today, refers to an especially wide range of endeavors, embodying an equally wide range of aims, and having correspondingly various relations to the activities generally included in the discipline of biology. I suggest that, in order to address the question of what synthetic biology has to do with biology, it would be useful to try to tease these various aims apart.

Synthetic biology as biological engineering

My first example is the brand of synthetic biology that has come to dominate the attention of the media, and that it is closest in spirit to what we used to call engineering. Ironically, it is Richard Feynman who, however unwittingly, seems to have provided the slogan for this brand of the new interdiscipline: ‘What I cannot create, I do not understand.’

This comment was left on Feynman’s blackboard when he died, and it is quoted in an astonishing number of articles in synthetic biology, indeed, in many of these articles, it appears as a leading epigram, a motto that has been claimed by virtually the entire community. Notice this is very like the quote from Forster and Church, but it is Feynman’s version that is evoked, not Forster and Church’s, and the reason is clearly the extra weight lent to the claim by the prestige of a Nobel laureate theoretical physicist. I do not begin to know what Feynman might have meant by this comment, but it must be said that its appropriation by synthetic biology is at the very least ironic. For a number of generations of physics students, Feynman too served as an icon, along with Einstein, of the ultimately pure scientist, of the inquirer who seeks to build the world not by construction but by deduction, reasoning always from first principles. Indeed, it was Feynman’s vision of the power of scientific thought—as well as of ‘the pleasure of finding things out’—that most directly inspired my own fantasy of what it meant to be a theoretical physicist.

It is not the power of thought, however, or even the Eros of understanding, for which he is here invoked. The primary context in which Feynman’s quip tends to be employed is the particular vision of synthetic biology most actively promulgated by Drew Endy, formerly at MIT, and now at Stanford—a vision about as far removed from the aims of theoretical physics as I can imagine, and that may embody the credo of making as knowing (or knowing as making) in its most radical form, i.e. in a form that does away with the epistemological aspects of the claim: here, knowing is making, and nothing more. Drew Endy deserves a lot of credit for what he has done, and not least for the public relations efforts that have launched his vision of synthetic biology into such a prominent (or even dominant) position in current discussions of the subject.

Endy is by training not a biologist, but an engineer, and unabashedly so. In an interview with the online magazine Edge recently, he explained his vision as growing out of the question: ‘How can I make biology easy to engineer?’ Frustrated by what he saw as the failure to realize the initial promise of biotechnology, he bemoaned the fact that: ‘Thirty years into biotechnology, despite all of the successes and attention and hype, we still are inept when it comes to engineering the living world.’

And he vowed to redress this failure. He doesn’t want to talk about it, he wants to do it. He doesn’t want it to be a research project, he wants an engineering project.

Endy’s vision has two parts: his first idea was that the way to make biology easy to engineer was to make it easier for everybody to engineer. And second, that the best way to make biology easy to engineer is to draw on the inspiration of William Sellers of the Franklin Institute in Philadelphia who, in 1864, presented a system for a more efficient production of nuts and bolts, based on the Sellers Screw Thread Standard—a standardized screw that is easy to manufacture, and led to the retooling of American machine shops to produce screws, nuts and bolts in accordance with the Sellers Standard.

The consequence for today, as Endy observes, is:

Endy credits Tom Knight of MIT with pointing out to him: ‘that it would be nice if we had standard biological parts that could snap together and then behave as expected when we snapped them together’ (in Edge, 2008). Finito experiment and theory, enter standardized design and production. The rest, as they say, is history. Endy’s two major contributions—his role, first, in the organization of the phenomenally successful international Genetically Engineered Machine competition (iGEM), and second, in the development of a community devoted to the production of standardized modules (Bio Bricks)—is well known.

But the question remains: what does Endy’s version of synthetic biology have to do with biology? Two obvious connections come to mind:

• training hordes of students in techniques for certain kinds of biological manipulations, and

• developing nucleotide modules that may be useful in some kinds of biological experiments.

But this is the standard role that technological development plays in any science. As to synthetic biology’s role in understanding the organization of biological systems as we know them, by Endy’s own account, it is non-existent. Synthetic biology’s role is not in understanding organisms as they have evolved, but possibly, he adds, in understanding how to remake these organisms to better and more efficiently serve our ends as human users. Here, the aim is to make use of the understanding of biological parts that has emerged from biological science as we have known it to launch an alternative—a synthetic—biology. But it is not clear what, if anything our expertise in synthetic biology contributes back to an understanding of traditional biology. Its object is rather to expand our known universe to include living entities not as they have evolved but as we design them. In other words, engineering as we used to understand that term before it got folded into technoscience: what is novel is its application to whole organisms. Of course Endy is not the first to envision either engineering biology, or biological engineering, but I don’t think I could have invented a purer exemplar, and advocate, of the engineering ideal.

Synthetic biology as creating life

Shifting to a more historical vein, we might remind ourselves of Jacques Loeb’s contributions to what was actually at the time referred to as ‘artificial life’, and the stir that work created. Philip Pauly has described Loeb as: ‘the major public advocate of what can be termed “the engineering standpoint” in biology’ (1987a: 5) for the period between 1890 and 1915. Yet, between Loeb and Endy, a rather considerable difference can be seen, suggesting that, in itself, the engineering standpoint houses multiple aims: two in particular stand out here: creation and efficiency. Loeb’s aims seem to focus on the former rather than the latter; to my knowledge, there is no concern with efficiency in Loeb’s ambitions, only with the creative power of man. As he wrote: ‘The idea is now hovering before me that man himself can act as a creator, even in living nature, forming it eventually according to his will.’

Loeb’s subject was not the origin of life, but reproduction, and by today’s standards his efforts at what he referred to as ‘creating life’ were modest indeed. Perhaps his focus on fertilization, and on the possibility of initiating the start of life without sex, together with the prospects it suggested for new modes of reproduction, must have made his subject particularly titillating. In any case, both the press and Loeb himself interpreted his work as a major step forward towards an engineering biology. As E.B. Wilson wrote, Loeb’s work: ‘opened up a variety of manipulative possibilities—of controlling sex and “creating wholly new organic forms by varying slightly the conditions of development”.’ And, in Loeb’s own words:

Indeed, he provided a motto of his own very like Feynman’s when he wrote: ‘[W]e must either succeed in producing living matter artificially, or we must find the reasons why this is impossible.’Footnote ²

But, however committed Loeb may have been to the vision of engineering biology, he was not first and foremost an engineer, indeed he was not an engineer at all. He was trained as a biologist, and although he may have waxed enthusiastically at the prospect of manipulating life or reproduction, his primary focus was on understanding that process in much the same sense of understanding as modern molecular biologists use the term. Indeed, Loeb might even be taken as an exemplar of the contemporary biologist for whom explanation has become virtually synonymous with the analysis of what the philosopher R.G. Collingwood, following Aristotle, called ‘practical cause’: as Collingwood wrote: ‘The “condition” which I call the cause … of an event in which I take a practical interest is the condition I am able to produce or prevent at will’ (Collingwood, Reference Collingwood1940: 302). Or as Philip Sharp of MIT once said to me when I proposed a seminar on the meanings of ‘explanation’ in developmental biology: ‘What’s the question? I’ll understand biological development when I can make it in the lab.’ These days, as I’ve already said, the term ‘synthetic biology’ covers a great deal of territory, but I think the aims in at least some of that territory correspond quite closely to those of biologists like Loeb, or like Sharp.

Synthetic biology as understanding of life

My last example tells a different story. Stephane Leduc was a contemporary of Loeb’s—a biophysisict in Nantes from the early part of the twentieth century about whom I wrote in my book Making sense of life (Fox Keller, Reference Fox Keller2002). By my reading, Leduc was interested neither in efficiency, nor in bending life to his will. In fact, he seems to have manifested little if any of the engineering impulse. Rather, he sought an understanding of the origin of life incrementally, by the simulation of ever more ‘life-like’ constructions from simple chemical precipitates. I invoked Leduc in Making sense of life to illustrate a notion of explanation that would seem manifestly absurd to contemporary readers. Today however I want to return to him to illustrate an epistemological aim that we might recognize as theoretically legitimate, however rarely it may surface in current research in either engineering or biology.

Leduc may or may not have coined the term, but his 1912 publication of La biologie synthetique pre-dated current fashion by almost a century. The demonstrations illustrated in this book (and in his other publications) were little short of astonishing and, at the time, they attracted a great deal of attention. For example, in The mechanism of life (1911) he published an account of artificial cell division, produced with nothing more than inorganic chemicals.Footnote ³ As he explained:

And he concluded: ‘The resemblance of these successive phenomena to those of natural karyokines is of the closest. The experiment shows that diffusion is quite sufficient to produce organic karyokinesis, and that the only physical force required is that of osmotic pressure’ (1911: 94).

In similar ways, employing similar techniques, he also succeeded in producing artificial organisms: ‘When a soluble substance in concentrated solution is immersed in a liquid which forms with it a colloidal precipitate, its surface becomes encased in a thin layer of precipitate which gradually forms an osmotic membrane round it’ (1911: 94).Footnote ⁴ Increase in osmotic pressure then gives rise to ‘osmotic growth’. Leduc describes the effect as follows:

Indeed, by employing a variety of metallic salts and alkaline silicates (e.g. ferrocyanide of copper, potash and sodium phosphate), and adjusting their proportions and the stage of ‘growth’ at which they were added, Leduc was able to produce many truly spectacular effects—inorganic structures exhibiting a quite dramatic similitude to the growth and form of ordinary vegetable and marine life.

By ‘appropriate means’, it proved possible to produce ‘terminal organs resembling flowers and seed-capsules’, ‘coral-like forms’, shell forms and:

Some of Leduc’s osmotic ‘organisms’ also exhibited an apparent capacity for both ‘free-swimming’ and reproduction:

So what was, or is, the point in these exercises? Leduc’s aim was to demonstrate the continuity between living and non-living matter, and, in his own words, to do so incrementally. Obviously, these were not living organisms. Yet they did bridge the gap with living systems in one dimension: they looked like organisms; that is, Leduc had demonstrated that structures that resembled the forms of living organisms—at least in their outward appearance—could arise spontaneously, from brute matter, without help from either a designing deity or a vital force. This is not nothing. If the outward morphology of living organisms is seen as a property that has been impossible to explain on the basis of what we know about physico-chemical systems, then here is an explanation of how it could have come to be. Not an account of how biological form actually does emerge, but a demonstration that such forms can emerge, in principle, without the help of any sort of vital force.

There is a sense in which this is not all that different from Turing’s (1952) reaction-diffusion account of morphogenesis—he too provided a demonstration of how differentiation and morphogenesis could come about. Again, not of how it does happen, but at least establishing that it could happen. The only difference is that Turing’s demonstration was mathematical rather than experimental.

It is also not very different from contemporary efforts in supramolecular chemistry to synthesize complex structures by self-assembly. Indeed, I suspect that current interest in the synthetic possibilities of self-assembly of large molecules helped to spur the efforts of a group of researchers in Paris and Geneva to reproduce Leduc’s results.

In the discussion of Leduc’s work in my book, I clearly implied that his particular brand of synthetic biology has nothing at all to do with biology as it is practiced today, but I am no longer quite so certain. To be sure, it is hard to imagine any place for such activities in contemporary efforts to understand morphogenesis, but there are other biological phenomena in which I would suggest that efforts very like those of Leduc’s might have a legitimate, and perhaps even important, role today. Most notably I think of problems in the origin of life. The emergence of form may no longer mystify us, but other, perhaps more quintessentially biological properties, certainly do remain without plausible explanation. How, for example, can we account for the spontaneous evolution of function, purpose, or of agency? For such questions, I want to suggest that the kinds of manipulations we saw in Leduc, and now see in work in supramolecular chemistry, can serve a theoretical role—i.e. here we have an illustration of the ways in which manual, experimental manipulations can stir, provoke, the theoretical imagination. This leads us to an understanding not of how life actually did evolve, but of how it could have evolved: an in principle solution, exhibited either in the lab or, as is increasingly common these days, on the computer, but of a kind that could play much the same kind of role as mathematical (in principle) models have played in the history of theoretical physics, i.e. as a guide to thought. In other words, I am suggesting a conceptual function served by certain kinds of material construction that is closer to that of traditional theoretical work than it is to traditional engineering aims. Such efforts may not have played much of a role in biological research in the past, but they may well do so in the future, especially as questions about the origin of life once again become a legitimate part of biological science. Or, to put it yet another way, Leduc’s brand of synthetic biology may bear a relation to what we call systems biology that cuts along a different axis from the one I earlier proposed: not so much conceptual v. material, or science v. engineering, but rather actual v. possible.

Understanding no, changing yes

But just because I argue for a conceptual role for such efforts does not mean that here too practical consequences are not close at hand. Work on the development of artificial cells, and on the dynamics of self-assembly, clearly can and does also serve other aims, aims that have nothing to do with understanding life as we know it, and that would be difficult if not impossible to distinguish from more traditional engineering goals. To quote from a recent article in the New Scientist entitled ‘Second genesis: Life, but not as we know it’: ‘A synthetic, made-to-order living system might even serve as a self-maintaining, self-improving, adaptable assembly line for producing everything from pharmaceuticals to petrochemicals.’ Or, we might add, biological weaponry. The race is on, and rewards abound. Or perhaps I should say the races are on. For the goals are multiple. Not only do researchers in self-assembly and minimal cells have their eyes on practical uses, but so too, do researchers on minimal genomes (like Craig Venter) or modular construction (like Drew Endy). As Jay Keasling observes, researchers seek: ‘the development of well characterized biological components that can be easily assembled into larger functioning devices and systems to accomplish many particular goals’.Footnote ⁵ For the creation of artificial life, David Deamer predicts 5–10 years. Other, perhaps less ambitious goals are at hand today. Success in these efforts may not further our understanding of biology, but it does, and will continue to, change our world. When such changes help us to cure cancer (as some of Adam Arkin’s work promises to do), we embrace them; when they increase the risk of bioterrorism, we deplore them. Synthetic biology may not be a single discipline or a single field, but many of the efforts included under this term raise issues that all of us, biologists or not, need to be concerned about.

Yet here too, the issues are not new. As we know from past experience, similar concerns frequently arise around the development of new technologies. Clearly, we need to develop a ‘code of ethics and standards’—as George Church describes it—to regulate the promises and dangers that face us here, a code that will be effective in meeting the particular challenges of biological engineering. It is true that our power to change the world has grown exponentially, but even so we might still be able to learn something from history.