TECHNICIAN Jim McIver peers round from the back of a large black rig, his face half obscured by a cascade of wires. On the bench top in front of him, a network of tubes supplies a series of small reaction flasks with a mysterious bubbling fluid. McIver lifts his hands to make an adjustment, revealing fingertips singed by hours of soldering.
It is a scene that fits the cinematic cliché of the mad scientist's laboratory, but this is no crazed loner's den. We are in the chemistry department of the University of Glasgow, UK, and Lee Cronin, McIver's energetic boss, is far from a sinister Victor Frankenstein figure. That said, he has a superficially similar goal. With a string of high-profile results already under his belt, he is about to flick the switch on his most ambitious experiment yet: to evolve the chemical complexity required for life in just 5000 hours of lab time.
If the experiment succeeds, it could go a long way to answering some perennially fascinating questions. What exactly is life? What does it take for inanimate matter to start doing the sort of things life does: to replicate, compete for resources and evolve? Is such behaviour possible only with the carbon-based organic chemistry that powers us and the living world we inhabit? In less than seven months, we might have some new answers.
Cronin is by no means the first chemist to be fascinated by the idea of re-enacting the processes that got life started. Perhaps most famously, in 1952 Stanley Miller and Harold Urey of the University of Chicago sparked electrical discharges through a brew of water vapour, hydrogen, methane and ammonia to mimic the turbulent atmospheric conditions of the early Earth. Within days, they were finding that organic molecules appeared in the mix, including some small amino acids, the building blocks of the proteins essential for life (Science, vol 117, p 528).
Though influential, that work is now generally seen to be a little simplistic. "They made some amino acids, it was exciting," says Mike Russell, who investigates the underpinnings of life at NASA's Jet Propulsion Laboratory in Pasadena, California. But he points out that life quite possibly first poked its head above the parapet somewhere where conditions were quite different, for instance in hydrothermal vents deep under the sea. What's more, neither Miller and Urey nor more recent experimenters managed to make anything beyond a few simple and distinctly lifeless amino acids. Attempts to meld these amino acids into sophisticated proteins, or to create long, complex molecules such as DNA and RNA - the sort of components that underlie even the simplest single-celled life forms - just result in a tarry mess.
Bootstrapping evolution
That, says Cronin, points to a fundamental problem: "The complexity even of simple life cannot spring out of nothing." So he reasons that a series of simpler chemical levels capable of autonomous evolution must have existed before. But what?
There have plenty of suggestions as to the nature of these missing links on Earth, notably RNA-based organisms or mineral-based life that fed on hot rocks. But given the lack of hard evidence, Cronin thinks such proposals are the wrong tack. "I can't do an experiment to tell me definitely where life on Earth came from," he says. "There I'm as stuck as theologians and philosophers." Rather than fixating on recreating life as it popped up on Earth, he wants to coax lifeless inorganic matter to do lifelike things. If he succeeds, that will begin to tell us something about the minimum chemical infrastructure needed for a system capable of autonomous evolution to emerge - on Earth or anywhere else.
"Demonstrating that simple inorganic material is capable of evolving in order to survive in changing environments would be a major discovery," says Craig Richmond, the postdoc in charge of the 5000 hours experiment. "It could be classed as a new inorganic life form," he claims - though he acknowledges that this depends on exactly what it means to describe something as being "alive" (see "Testing life").
To guide their quest, Cronin and his team have identified three basic characteristics that a chemical system capable of autonomous evolution is likely to need. First, there must be a "library" of related molecules whose structures represent encoded information that can be transferred between generations. Second, these molecules must support a metabolism, a set of chemical reactions that produces useful energy. Third, the molecules must be able to form enclosed spaces where this metabolism can proceed undisturbed.
Life as we know it has all those things. But to prove their point, Cronin and co have plumped for something as far away from anything in nature's bio-construction kit as you could imagine: metal-oxide assemblies centred on atoms of metals such as tungsten, molybdenum or vanadium.
They may seem an odd choice, yet these metal oxide building blocks tick all the boxes for evolutionary potential. When fed into reaction flasks in which their concentration and the acidity of the environment can be varied, they assemble themselves into a family of complex metal-oxide structures known as polyoxometalates. These can in turn act as templates for the formation of other structures. It doesn't take too big a leap of imagination to see the parallel with the way DNA and RNA act as templates for amino acids, which in turn build into proteins.
Last year, Cronin and his colleagues exploited this self-assembling ability to fashion one of the largest inorganic molecules yet synthesised: a huge metal-oxide wheel containing 150 molybdenum atoms that measured 3.5 nanometres across - as big as a protein molecule (Science, vol 327, p 72). And the resemblance to the molecules of life doesn't stop with mere size. Electrons skittering through the polyoxometalate structures can promote a wide range of reduction and oxidation reactions, just as enzyme proteins catalyse similar reactions in biological contexts. In the right environment, the polyoxometalates even produce closed structures in the shape of bubbles, tubes and bubbles within bubbles reminiscent of the membranes that enclose biological cells.
So it looks a bit like life - but is that enough to make it act like life? That is what the 5000 hours experiment will find out. The reaction rig will randomly flow the self-templating, shape-shifting, membrane-making metal-oxide blocks from flask to flask, together with any larger molecules they make. As they flow, acidities, concentrations and other parameters will be constantly varied. The idea is that the competition for limited resources will see certain forms of cluster - those that are chemically most adaptable or best suited to their environment - winning out over others in the fight for survival. Eventually, only these "fittest" clusters will self-replicate, transferring their structural information to further generations of similar clusters.
Cronin and Richmond will be looking for the formation of increasingly sophisticated structures at each stage of the experiment. If that is what they find, it would go some way to showing that much of what we see as defining life - competition for resources, natural selection and survival of the fittest - are not unique to the carbon-based life we know, but the result of far more general chemical principles. "It is not just about the selfish gene - it is about selfish matter," says Cronin.
Stripped down to its barest essentials, life can be seen as a struggle against thermodynamics and its drive towards ever more featureless uniformity, Cronin observes. There seems to be no reason why the fight against that uniformity should be confined to carbon-based chemistry. "I don't think for a second that on planet Earth molybdenum oxide was ever alive," Cronin says. Elsewhere, though, given the right chemical environment - who knows?
So how much can we expect from Cronin's experiment? In particular, will it display the kind of rise in complexity that would indicate a living system? "I predict it'll get a little more complex but will finally be limited by the relative simplicity of the system," says David Deamer, a chemist investigating the origins of life at the University of California, Santa Cruz.
Not even Cronin thinks we will see anything like the level of complexity found in evolving biological systems. But any sign that could be interpreted as autonomous evolution would be an exciting starting point for further investigation, he says. What might happen, for example, if you threw some simple organic molecules into the mix? Would the self-replicating metal-oxide clusters shepherd the organic molecules towards increasing complexity?
Jim Cleaves, a geochemist at the Carnegie Institution in Washington DC, welcomes blue-sky ideas like Cronin's, but in the end he thinks that new chemistries are unlikely to lead us to new biologies. There is more than mere chauvinism to most scientists' fixation on the organic chemistry that underlies life on Earth, Cleaves argues; quite simply, it works like no other kind of chemistry can. The main elements that go into making things like proteins and DNA - carbon, nitrogen, oxygen and phosphorus - can react in an unparalleled variety of ways. Even carbon's close cousin silicon, often fingered as an alternative basis for life, forms far fewer compounds than carbon, and on a cosmic scale is much less prevalent.
Similarly water, organic life's solvent, is far easier to come by than alternatives such as ammonia or ethane. "There just aren't any other cosmically abundant solvents that have the same properties," says Cleaves. That may mean that life as we know it on Earth is the universe's one big fluke. "It could be like trying to get a royal flush - it's just a hand that comes up every 10 million hands," he says. "It's really unlikely that scientists will ever do it in the laboratory."
Will Cronin, Richmond and McIver prove such scepticism wrong in 5000 hours? The beauty is, we'll soon find out.
Testing life
In his lab at the University of Oxford, chemist Ben Davis has created an artificial cell that bears a striking resemblance to its biological counterparts. Within walls made of artificial versions of phospholipids, the main constituent of biological cell membranes, a varying pH triggers self-catalysing metabolic reactions that turn one small organic molecule, formaldehyde, into simple sugars, releasing energy on the way (Nature Chemistry, vol 1, p 379).
That sounds rather lifelike, even though no life form we know of on Earth has that sort of metabolism. So is it life? Davis prefers not to use the word, not least because many of the possible definitions of what constitutes life are problematical to say the least. Several such definitions have been put forward: the ability to reproduce; the production of waste; or a metabolism, as in Davis's artificial cell. Yet each definition throws up complications.
To break the impasse, in 2006 the UK's Engineering and Physical Sciences Research Council detailed a group of scientists to thrash out a definitive protocol for identifying life. The group included both Davis and Lee Cronin of the University of Glasgow (see main story).
What they came up with contained elements of the old: that an artificial chemical cell should be self-replicating and possess some sort of contained metabolism. But it also contained a radically new element, inspired by the similarly vexed question of whether machines can think. In 1950, the mathematician Alan Turing developed a test for this: if a machine truly were intelligent, he said, then an intelligent interrogator, asking questions of both a machine and a human, would be unable to distinguish the two sets of answers. Similarly, Davis, Cronin and their colleagues argued, to be worthy of the title life an artificial chemical cell should be able to live with biological cells undetected as an impostor, and participate in the same processes as the natural cells (Nature Biotechnology, vol 24, p 1203).
Davis's artificial cells pass this aspect of the test. He mixed them with a sample of a marine bacterium that luminesces when levels of certain chemicals secreted by the colony as a whole exceed a threshold. Products of the formaldehyde metabolism of the cells counted towards this "quorum sensing" system, causing the bacteria to light up.
At the moment, the cells do not have the ability to replicate on their own, so do not fulfil all the requirements to qualify as life. It remains to be seen whether whatever ultimately emerges from Cronin's experiments will fare any better.