How The Primordial Soup Became Haute Cuisine

(This article was originally published in Ingeniøren, 25 March 2011. Online link (in Danish). Translated by the author.)

In the beginning, a primordial soup covered the Earth. In fact, the primordial soup was Earth – a pile of cosmic waste, erupting in spasms caused by meteor crashes and spouting flames of molten sulfur and iron, slowly condensing into a solid planet. All life derives from this primordial soup, science says. But nobody knows exactly how.

We know that the first cyanobacteria appeared in the oceans some 700 million years after the formation of the Earth, which makes it approximately 3.8 billion years ago. But how could a bacterium, an organism so immensely complex, so sophisticated and well adapted, emerge all by itself?

This is one of the big questions, and no one knows the answer. Or rather, perhaps we know in theory, we just haven’t conducted the right experiments to confirm the theory yet. At least some veterans in the field of theoretical biology think so, and they have recently triggered a new round of speculations about the origin of life by publishing a paper on the subject in the Open-Access journal Biology Direct.

Autocatalytic Networks
Stuart Kauffman, an american theoretical biologist and complex systems researcher from the Santa Fe Institute in New Mexico, was one of the authors of the paper. In 1993, Kauffman proposed that the transition from chemistry to biology may have taken place through a network of molecules which mutually catalyzed the production of each other. In this way, every single molecule could be formed by at least one chemical reaction in the network, and each of these reactions would be catalyzed by at least one other molecule in the network. At any given time, the whole network would thus be able to show a primitive form of metabolism, and slowly learn to copy itself.

Physicists Doyne Farmer from Santa Fe Institute, New Mexico, and Steen Rasmussen from the University of Southern Denmark have previously been able to partially create such `autocatalytic networks ́ in the lab. However, it has been quite difficult to demonstrate that the networks are able to learn in a Darwinian sense, that is, to proliferate, create variation and adapt to their surroundings.

In 2010, another co-author of the paper, the Hungarian biologist Erös Szathmáry, published a paper in collaboration with Mauros Santos from the Autonomous University of Barcelona in which they claimed that the larger the autocatalytic network, the more inaccurate the actual copying will become, thus preventing natural selection from working properly. This was a huge disappointment for those who thought that life on Earth could begin with a primitive autocatalytic network with some sort of metabolism, but without a membrane or autonomous reproduction mechanism.

The Chicken or the Egg
The scientific thinking behind the origin of life is nowadays split into two camps. One is the `metabolism first ́ camp, the other `genes first ́, and their interrelation is hardly distinguishable from the chicken or the egg dilemma.

Those who think that metabolism came first, do not believe that RNA-like molecules (which we know are able to copy themselves) could arise out of nowhere, that is, without natural selection having paved the way through a long evolutionary process. Indeed, experiments have shown that it is impossible to synthesize RNA, PNA, TNA or similar replication templates without highly advanced enzymes that facilitate the process.

On the other hand, those who think that genes came first, find it impossible to imagine that a metabolism process could be sufficiently stable over the course of generations without a central genetic control to produce the proper molecules.

Stuart Kauffman's autocatalytic network theory united these opposing positions, because the center of operations in an autocatalytic network is the network itself. But without actual experiments to show that a basic autocatalytic network can be stable long enough for RNA-like templates to develop, it has been difficult to convince biologists that the mystery of the origin of life in the primordial soup has been solved.

According to Kauffman, Santos and Szathmáry, there is yet hope for their model. Prebiotic evolution in a scrap heap of simple molecules is possible. It just isn’t sufficient with one single autocatalytic network; it requires multiple autocatalytic networks, separated from each other by independent vessels or cavities (for instance, clay crystals, liposomes, micelles, nanotubes etc.), which interact with one another in a certain way.

Genotype and Phenotype
More specifically, the multiple chemical networks have to consist of an autocatalytic nucleus and a non-autocatalytic periphery made up by products that are continuously produced by the nucleus. In a setup like this, the nucleus is analogous to a genotype (hereditary information), and the periphery is analogous to a phenotype (the appearance of the organism). By the use of simulations, the authors have managed to demonstrate that a system like this is able to show not only true Darwinian selection, but also neutral drift by means of mutations, variations and selection on the basis of interactions with the surroundings.

One important new insight brought by this approach is that there must exist an entire zoological garden of different species of autocatalysts. In its simplest form, a chemical reaction is autocatalytic if the reaction product catalyzes its own formation. But there are many other options. Two reaction products can catalyze a third product, or they can cross catalyze each other. Some chemical products can inhibit others, while other products promote another product, but only if an additional product is formed, and so on. The multiple autocatalytic networks thus create a diversity of reaction times and products that make it possible for the networks to compete for food.

With their new theory, Kauffman, Santos and Szathmáry argue that the `genes first ́ camp is wrong: The autocatalytic network is responsible for the copying, which subsequently leads to a question of how stable such a system has to be for heredity to work. But, as the researchers replied to one of their peer- reviewers, “An understanding of the kinds of chemical organization that could sustain heredity is logically anterior to the problem of the stability of such organizations”. In other words: Only through a deep understanding of the first steps of life can we find a clue about the later stages, such as replication through a genetic code.

Mud and Prebiotic Pizzas
Literature on the subject has brought many theories about how protocells, lacking membrane and  replication mechanisms, could have evolved. Graham Cairns-Smith from the University of Glasgow imagines that early life developed in simple lumps of clay. Since clay crystals grow and break up continuously, they can carry information (initially only as impurities) through time and space.

Proto-life hidden within the clay crystal cavities would thus be able to survive, and eventually reproduce. According to this theory, the genetic machinery would only be able to take over the functions of the clay crystals, and liberate itself from life in the mud at a far later stage.

Another proposal, put forth by chemist Günther Wächtershäuser, is to place the hotbed of the origin  of life in an aquatic environment on the surface of pyrite crystals. Wächtershäuser suggests that a combination of iron, carbon dioxide and hydrogen sulfide, bound to the surface of pyrite (Fe2S) deep in the ocean, could eventually develop organic molecules such as sugar. On the surface of this so called `pre-biotic pizza ́, where the pyrite releases useful energy, independent semi-cellular organisms could possibly develop their own chemoautotrophic metabolism and enzymes, due to an abundant flow of minerals and other compounds from the hot springs that formed the pyrite in the first place.

Only much later, when the mechanisms of the processes have gradually grown more complex - by a primitive citric acid cycle, a protective, but selective, lipid membrane and possibly also some sort of cellular division - the organisms would slowly be able to liberate themselves from the pyrite, and take off and colonize the oceans.

In the Elevator
An elevator pitch version of the origin of life would thus sound like this: In the beginning, random chemical compounds bump into each other and form new chemical compounds with new properties. Some of the new compounds are able to catalyze their own production, which creates an exponential growth that will stop only when the food source is depleted. In places where food sources are abundant, in nutrient rich primordial soups or near hydrothermal vents, these autocatalytic reactions persist, and slowly create compounds with other autocatalytic reactions, and after some hundreds of thousands of years create a population of autocatalytic networks, competing for food sources by having different characteristics and special talents.

Some of the autocatalytic networks learn to avoid the constant collapses by linking up with inhibiting reactions and hiding the autocatalytic nucleus in clay crystals or some other form of container in close contact with the food sources. As the eons pass by, these protocells specialize to more specific tasks, for instance by using sunlight to metabolize sulfur, or by producing lipids to create a semipermeable membrane. 

Eventually, after many millions of years and billions of random attempts at improvement, these protocells developed a molecular template able to copy itself and its container in a much more reliable way than previously possible. This invention marks a crucial transition in the history of evolution, i.e. the transition to an RNA-based, and later a DNA-based, life form. The invention turns out to be so efficient that all other catalytic networks are eradicated and replaced by genetically programmed - and genetically programming - organisms.

Evolution has also led to other things, for instance cell nuclei, organelles, chemotaxis, photosynthesis, sexual reproduction, multicellular organisms etc. Scientific theses about each one of these inventions would require many more biologists than ever lived. But the very beginning, where chemistry turns into biology by means of autocatalytic networks, can, in principle, be considered demystified. The most important thing missing now are experiments to confirm the theory.

  • Vera Vasas, Chrisantha Fernando, Mauro Santos, Stuart Kauffman, Eors Szathmary (2012), Evolution before genes, Biology Direct 2012, 7:1,
  • Vasas V, Szathmáry E, Santos M: Lack of evolvability in self-sustaining autocatalytic networks constraints metabolism-first scenarios for the origin of life, Proceedings of the National Academy of Sciences 2010, 107:1470,
  • Stuart A. Kauffman, The Origins of Order: Self-Organization and Selection in Evolution, Oxford University Press, 1993.
  • Bagley RJ, Farmer JD: Spontaneous Emergence of a Metabolism, Artificial Life I, Santa Fe Institute Studies in the Sciences of Complexity - Proceedings Volumes 1992, 10:93-140.
  • Steen Rasmussen, Liaohai Chen, Bärbel M. R. Stadler, Peter F. Stadler, Proto-Organism Kinetics: Evolutionary Dynamics of Lipid Aggregates with Genes and Metabolism, Origins of life and evolution of the biosphere, February 2004, Volume 34, Issue 1-2, pp. 171-180.