91av

Building artificial cells will tell us much about the origins of life – and may explain how Darwinian evolution began, says Nobel laureate Jack Szostak

IMAGINE Earth 4 billion years ago. It is a world of oceans, peppered with volcanic land masses resembling Hawaii and Iceland. The volcanoes spew poisonous gases and the atmosphere is rent by the violent impacts of asteroids and comets. Temperatures range from the incandescent heat of flowing lava to the frozen ice fields of the high polar regions. Shallow ponds on the volcanic islands dry out, then fill with rain, incubating the fragile chemistry that ultimately leads to the emergence of life.

How extraordinary that cellular life should have arisen in such a harsh environment. The exact nature of that first cell, the basic unit of all life today, is still unknown. It is an exciting puzzle that goes to the heart of the origins of life as we know it. But it is hard to solve: after all, reconstructing such ancient events seems an impossible task. Fortunately, there is much to be learned from a slightly more modest goal: building basic artificial cells in the lab.

“The exact nature of the first cell, the basic unit of all life today, is still unknown”

I should emphasise that there is a world of difference between this kind of artificial life and the recent feats of Craig Venter and his team at the J. Craig Venter Institute in Rockville, Maryland, and San Diego, California. That team, which includes Nobel laureate Hamilton Smith, achieved a technical milestone by replacing the genome of an existing bacterium with a synthetic genome stitched together from stretches of chemically synthesised DNA. Creating synthetic genomes may ultimately let us redesign existing life forms in ways we can scarcely imagine.

Now let’s return to the origin of life. In the late 1990s there were opposing views about the key steps in the process. Many of us argued for the emergence of RNA replication as the most important step because the inheritance of genetic information is central to evolution. Others, however, thought about life from the perspective of structures that would define a boundary between the living system and its environment. They championed the idea that the most important aspect was the emergence of the cell membrane, in the form of self-assembling and self-replicating vesicles.

For over a year in the genetics department of the Harvard Medical School, my then student David Bartel and I argued about the origin of life with , then at the Swiss Federal Institute of Technology in Zurich. Bartel and I were firmly in the “genetics first” camp, since our research was focused on nucleic acids. Luisi, a pioneer of vesicle research, thought only about membranes and compartmentalisation.

As we sparred, a remarkable thing happened: our views started to converge. Finally, we agreed that for life to emerge, both membranes and genetic material needed to come together, compatibly, to endow the first cells with the capacity to evolve. We presented this synthesis in a in Nature in 2001. It was a real turning point in my origins research. Up to that point, my work on RNA had been inspired by Tom Cech at the University of Colorado, Boulder, who shared the Nobel prize in chemistry in 1989 for showing how RNA could carry out catalysis as well as transmit information. For years, my team and I wrestled with the problem of how RNA could catalyse its own replication. But after our debates with Luisi, I came to appreciate that cell membrane growth and division was just as important and that both problems would have to be solved.

I was a bit reluctant to jump into working with membranes because they are so messy and squishy and the techniques for working with them so crude compared with the refined methods available for studying RNA. But since we had put forward our ideas of what a simple cell might look like, and conjectured it could be made in the lab, I thought that I ought to get to work building just that kind of simple cell.

When we started, we had no idea how to make a cell membrane grow and divide in a realistic way. Our work began with a simple proof-of-principle demonstration of protocell membrane growth and division, by colleagues Martin Hanczyc and Shelly Fujikawa. We found we could make fatty-acid vesicles grow by “feeding” them with more fatty acids – an approach first explored by Luisi – and then make the resulting big vesicles divide by squeezing them through small pores. Knowing that simple physical forces could in a sense replace the fancy cellular machinery of modern life, we continued to look for other ways of driving growth and division.

Working on experiments with my student Irene Chen, together with Richard Roberts at the California Institute of Technology in Pasadena, we showed that the mere presence of RNA could drive vesicle growth. Growth occurs by “stealing” membrane molecules from neighbouring vesicles that contain less or no RNA. The reason this works is that RNA inside vesicles exerts an osmotic pressure on the inside of these sacs. This internal pressure places tension on the membrane, which grows by absorbing molecules from surrounding vesicles that are less swollen as a consequence of having a smaller cargo of genetic material.

Thus protocells with RNA that replicated better, and those that ended up with more RNA inside them, would grow faster. This was a really exciting discovery because it showed us how there could be competition between cells based on purely physical phenomena. There is a problem with this approach, though: a cell growing due to osmotic pressure can only expand as a sphere. Making a sphere divide is hard and takes a lot of energy. Plus, you have to squeeze out some of the contents, which is not very satisfying. For a few years, we were stuck – we had a new way to make cells grow, and the competition would drive evolution, but we didn’t know how to make the cells divide after growing. As so often happens, the solution came from an unexpected direction. One of my students, Ting Zhu, figured out how to make fatty acid vesicles big enough – 4 micrometres across – to be viewed under the microscope. He could feed these spherical vesicles by adding more fatty acids, and watching what happened.

We thought the spheres might get a little elongated as they grew, yet what happened shocked us. They started to sprout what looked like hairs and filaments. Even now, we do not understand all the details, but it seems that tiny tubules grow out of the surface because the membrane grows faster than the volume can keep up with. The cool thing is that, after half an hour, the spherical vesicle has transformed into a long, filamentous structure so fragile that it divides into daughter vesicles just by gentle shaking. So division, that part of the early cell cycle that looked so hard, turns out to be easy. Even wind-blown waves on a pond could turn the filaments into next-generation protocells.

While finding a simple, effective solution to the problem of how protocell membranes might grow and divide is satisfying, it is only half of the problem. The basis of biology is Darwinian evolution; for that to work, genetic material must encode useful information and transmit it to the next generation.

So what might that first genetic material, trapped inside a vesicle, have done to give its own membranous home an advantage over surrounding vesicles? An exciting possibility we are exploring is that a primitive RNA might have catalysed the synthesis of phospholipids, a class of lipids that are a major component of modern membranes. This would have been a critical step in starting the gradual transition from primitive membranes based on fatty acids to modern membranes based on phospholipids. But what advantage could phospholipids have conferred on primordial cells? The answer may well explain the onset of Darwinian evolution and the subsequent avalanche of events leading to modern biology. Stay tuned to these pages for updates on the ancient past!