Abstract
Over the next few days and weeks, a hundred kilograms of black stones were collected by townspeople and a few early-bird scientists. One of the Murchison stones was soon delivered to the NASA Ames Research Center for analysis, and in 1971 Keith Kvenvolden and his colleagues published the first convincing report that extraterrestrial amino acids and other organic compounds were present in the meteorite. This represented a dramatic confirmation of Stanley Miller's claim that it was possible for the most fundamental chemicals of life to be synthesized by a nonbiological process.
At the time, I was just beginning an academic career at the Davis campus of the University of California. I recall reading news accounts of Murchison fall and its implications for research on the origin of life, but it was not until 1975 that I became actively interested in such questions during a sabbatical leave. I was visiting Alec Bangham's lab at the Babraham Institute a few miles south of Cambridge, England, in order to learn more about liposomes. Alec was a pioneer in this field, and 10 years earlier had discovered that phospholipid could self-assemble into vesicles composed of lipid bilayers. In one of our conversations, we came up with an intriguing question: What lipidlike compounds were available that could be used by the first forms of cellular life to form membranous boundary structures? No one knew, but it seemed like an interesting problem for further study.
Back in Davis, Will Hargreaves was beginning his Ph.D. research in my lab. He expressed interest in the problem, and we agreed that a good first step was to find the simplest kind of lipid that could assemble into liposomes. To our surprise, this turned out to be decanoic acid, with 10 carbons, which readily formed vesicles at a pH where the ionized and neutral form of the fatty acid were present in equal quantities. We also discovered that two Australian scientists, Gebicki and Hicks, had published an earlier paper in Nature in which they described what they called ufasomes, or liposomes composed of oleic acid, an unsaturated fatty acid. Will found that other single-chained amphiphiles such as dodecyl sulfate also produced very stable vesicles if mixed with dodecanol. Was it possible that amphiphilic lipidlike substances might have been available in the prebiotic inventory of organic compounds?
Will completed his degree, but now I was hooked. In the early 1980s I began to interact with researchers at NASA Ames in Mountain View, California, where Sherwood Chang was working on meteorite analyses. I suggested that it might be interesting to see if lipidlike compounds were present in carbonaceous meteorites. Sherwood gave me a very clean sample of the Murchison, which I took home to Davis. In previous work, I had often used a mixture of chloroform, methanol, and water to extract lipids from rat livers, spinach, and egg yolks, so it seemed reasonable to use the same procedure on Sherwood's Murchison powder. I ended up with about a milliliter of chloroform. It had a slightly yellow color, so something in the meteorite had dissolved. I evaporated a drop of the extract on a microscope slide. As it dried, a strange aroma wafted up, and I realized that I was smelling something that was over 4.5 billion years old, the odor of outer space, perhaps even an odor that would have permeated the atmosphere of early Earth before life began.
I added a coverslip with a drop of dilute buffer at a slightly alkaline pH range, the same pH at which fatty acids formed membranes. Under the microscope, it was easy to watch the dried material as it began to interact with the water. Over a period of several minutes, it swelled, and large numbers of spherical structures appeared in the matrix. I switched to an oil immersion phase objective that gave 1000× magnification, and was astonished to see faint membranous strings and vesicles appearing at the interface between the meteoritic extract and the buffer. There was definitely something in the meteorite that could self-assemble into membranes, which we later determined was a mixture of fatty acids, including decanoic acid. I excitedly snapped off a few pictures, developed the film (no digital cameras in 1984), and ran down the hall to show some colleagues who were having lunch. I'm afraid they were not impressed. The fact that meteorites could form membranes was little more than a curiosity.
I published the observation in Nature in 1985, and again, no one else seemed very interested. I can't recall getting a single reprint request, but a few people must have read it, including Harold Morowitz. Thus began a collegial friendship that continues today. We later published a paper with the title “The Chemical Logic of a Minimum Protocell,” which appeared in 1988, coauthored with Bettina Heinz. Harold went on to write a wonderful book called Beginnings of Cellular Life, the first to incorporate the idea that membranes are essential components of all life today, so the first forms of cellular life must have made membranes from amphiphilic molecules available in their environment. It is interesting to note that living organisms today still must take certain lipids from their surroundings rather than synthesizing them de novo. Examples include essential fatty acids in the human diet, as well as vitamin A and vitamin E. I also recall that in 1960 Harold studied a mycoplasma bacterial species that could not grow unless it was supplied with fatty acids to be incorporated into its membrane phospholipids.
Over the next two decades, other colleagues began to take an interest in primitive membranes composed of fatty acids. In 1993, Luigi Luisi invited me to attend a meeting he had organized in Maritea, Italy, sponsored by NATO. I was surprised to learn that Luisi and his research group had read all our papers and were diving deep into the problem of fatty acid membranes. Over the next few years, they published a series of discoveries in which they demonstrated that fatty acid vesicles could grow and even reproduce, if the definition of reproduction was not too strict. Luisi's group had a friendly competition with us to be the first to encapsulate a polymerase enzyme called polynucleotide phosphorylase, which synthesizes RNA from nucleoside diphosphates. It was a tie, and both groups reported positive results in 1994. Perhaps most important was that Luigi got together with Jack Szostak and David Bartel, and the three of them published a landmark paper in Nature in 2001 that made a strong case for a cellular origin of life.
That brings us to the present day. My expectation is that someone will soon discover a way to fabricate an encapsulated system of molecules exhibiting the basic properties of a living cell: growth by directed and catalyzed polymerization, replication of the molecules themselves, perhaps even a primitive version of division. An example of this progress was reported by Szostak's lab in 2008 in which a DNA that was half double helix, half single strand was encapsulated in a fatty acid membrane. If the vesicles were “fed” with activated nucleotides, the single stranded portion acted as a template, using the nucleotides to fill in the rest of the double helix. This was possible only because the fatty acid membranes were sufficiently leaky to allow the nucleotides into the vesicles. Another important advance was reported in 2011 by Phil Holliger's group, who were able to evolve a ribozyme molecule that could copy over 90 nucleotides of its own base sequence. This was done in a test tube, but an obvious next step is to encapsulate the ribozyme.
My own research is following up an idea I have been playing with since 1985, that lipids functioned in the origin of life not just to provide a compartment but also to guide polymerization processes. In 2008, Rajamani et al. showed that simple wet-dry cycles of lipid vesicles in the presence of ordinary mononucleotides resulted in the synthesis of small amounts of RNA-like polymers ranging from 20 to 100 nucleotides in length. Olasagasti et al. (2011) reported that, if a DNA template was present under these conditions, there was an apparent transfer of sequence information from the template strand to product strands. These results have produced some friendly skepticism from my colleagues, and more work needs to be done to make them convincing. But it's a start, and ideas, right or wrong, generate more interesting ideas and experiments to test them. That's why I wake up every morning thinking about how life could have begun on Earth, perhaps even on Mars.