The Origin of Life on Earth and the Design of Alternative Life Forms

By Jack W. Szostak, Nobel Laureate in Physiology or Medicine, 2009
Molecular Frontiers Journal
| Vol. 01, No. 02, pp. 121-131 (2017)

Yellowstone Lake frozen in the winter. (Credit: J. Schmidt / nps.gov / CC)

This article was transcribed from a presentation delivered by Professor Szostak at the Molecular Frontiers Symposium on Tailored Biology at the Royal Swedish Academy of Sciences, May 2017. It is published here and in MFJ with permission from Professor Szostak.
This is an Open Access article published by World Scientific Publishing Company. It is distributed under the terms of the Creative Commons Attribution 4.0 (CC-BY) License. Further distribution of this work is permitted, provided the original work is properly cited.

To understand the origin of life on Earth, and to evaluate the potential for life on exoplanets, we must understand the pathways that lead from chemistry to biology. Recent experiments suggest that a chemically rich environment that provides the building blocks of membranes, nucleic acids and peptides, along with sources of chemical energy, could result in the emergence of replicating, evolving cells. The broad scope of synthetic chemistry suggests that it may be possible to design and construct artificial life forms based upon a very different biochemistry than that of existing biology.

One of the things that I particularly love about the subject of how life got started is that there are so many interesting questions, and that they cover such a wide range of fields, from astronomy to geology to chemistry and biology. Just getting to talk to people from so many different fields and trying to figure out how everything came together to create the conditions for the emergence of life is really fascinating. The approach to understanding the world by always asking new and interesting questions is one of the great things about science, and the origins field is particularly special in the way that it brings people together to cooperate and collaborate and try to get answers to some of the mysteries of nature. When we see so many divisive and tragic things happening in the world today it’s really wonderful that we’re privileged to be able to work together.

Let me begin with a brief account of how I came to be interested in the subject of the origin of life. In the 1990s my lab spent a lot of time learning how to evolve populations of molecules[1]. It was a fascinating endeavor, and we along with others including Gerald Joyce and Larry Gold, learned how to evolve new molecules that could do interesting things such as bind to targets and catalyze reactions. Some of these experiments even led to molecules that are now in the clinic to treat diseases. But after evolving molecules in the lab for roughly a decade I became more and more fascinated by the question of how evolution started spontaneously on the early Earth[2]. It’s one thing to direct evolution in the lab when you have enzymes and expensive instruments and brilliant students. But somehow the process of Darwinian evolution got started all by itself on the early Earth. Since my original background is in biology, to me the origins of evolution and the origins of biology are really the same thing. As a result I started to focus more and more on the issues surrounding the origin of life, and that is the subject of this article.

What might a really simple primitive cell (or protocell) look like? One model is that protocells consisted of a primitive membrane enclosing some very short pieces of genetic material (figure 1). We’re not totally sure of the nature of that first genetic material, but for many reasons we think it is probably closely related to RNA[3]. What we’re trying to understand is how structures like this came to exist on the early Earth, and how they were able to grow and divide and start to evolve.

Figure 1. Illustration of a model protocell, consisting of a lipid membrane encapsulating small pieces of an RNA-like genetic material. The protocell is shown in a rich chemical environment filled with nucleotides and other building blocks of life. Image courtesy of Janet Iwasa.

People have been asking how we got here, i.e. how life got started, for a long time, but one of the things that has generated an explosion of interest in this question is that we now know that there are millions, probably hundreds of millions of planets that are reasonably Earth like, in our galaxy alone. We are pretty sure that many of them could support life, but we don’t know if they do. An example of a particularly interesting nearby solar system is the TRAPPIST-1 system that has been worked on a great deal by Didier Queloz and his colleagues[4] in Cambridge in the U.K. (figure 2). This solar system has at least seven planets and three of them are in the zone (green in the figure) where there could be liquid water on the surface. It’s quite likely that planets like this could support life, but it is going to be really hard to figure out if they are actually inhabited. The entire field of astronomy has been revolutionized by this kind of discovery and scientists are trying as hard as they can to work out the technology to detect signs of life on these exoplanets.

Figure 2. The TRAPPIST-1 planetary system is one of the many recently discovered planetary systems orbiting other stars in our galaxy. Three of the seven planets of this system lie within the habitable zone, where liquid water might exist on the planetary surface. It is likely that such planets could support life, at least in local environments. Detecting signs of life on exoplanets is one of the great questions energizing current astronomical research.

In parallel with that ongoing effort, we can ask simple questions that can be addressed here in our laboratories on earth, to try to understand how life might get started on different kinds of exoplanets, and how it did start on our planet. If we understood the whole process we could make a better estimate of how easy or hard it is for life to begin, and therefore how likely or unlikely it is that there is life out there. This involves a lot of questions on the whole pathway from planet formation to the early evolution of life, but here I will focus on how life could actually get started from the chemistry on a young planet.

We first need to understand how the building blocks of biology might be synthesized on a young planet. The study of this question is the field of prebiotic chemistry, and it is being pursued intensively in a number of laboratories. Then, once we have the right chemicals, we need to understand how a set of molecules can get together and start acting like a living cell. That is the topic that we have been focusing on in my laboratory. Recently, work in these two areas has advanced to the point that we can start to deduce something about the necessary environments for the origin of life[5]. This is bringing us into contact with planetary scientists, geologists, and atmospheric chemists to try to be as rigorous as we can in thinking about early planetary environments.

Let us begin by considering some of the hard problems with prebiotic chemistry. Stanley Miller’s famous apparatus, built in 1952, used a spark discharge as a source of energy in what was then thought to be a realistic primitive atmosphere[6]. In this simple experiment, Miller made an astonishing number of different compounds, but most surprisingly amino acids, one of the quintessential types of building blocks of biology. This was a revolutionary advance at the time, and Miller’s experiment generated huge optimism that the chemistry leading to life would soon be figured out. There was in fact a lot of progress in the ensuing 10 or 20 years. And then progress stalled and some of the problems with this approach started to become more evident[7]. In particular if you start to look at what is in this mixture of compounds, the closer you look the more you see. There are thousands, probably tens of thousands of different chemicals made in this crude approach of blasting a reducing atmosphere with energy. A lot of the materials we want are in there but at very low levels, while some of the materials we know we need are not there.

How can we think in a more realistic way about generating the building blocks that we need to put cells together? In my opinion a series of new approaches to this problem have transformed the field. I will try to give an impression of how the field has changed in the last 10 to 20 years as a result of these new approaches. The idea that one simple process could generate a primordial soup that contained everything needed for life is clearly unrealistic; what we need to figure out is how to make complex chemical pathways possible. Since it takes multiple steps to get to something as complicated as the nucleotides needed to make RNA, it is not realistic to expect this to happen by just mixing simple starting materials together. But you can break a pathway into smaller chunks if you can build up reservoirs of intermediates as crystalline deposits — in effect, organic minerals. Another approach is systems chemistry[8], which goes beyond the classical chemical approach of taking two compounds and making them react to produce a product that you want, by thinking about what else could or must be present and that could make a difficult step work in a better way.

Here I will give a few examples of potential organic minerals, and of systems chemistry approaches to the problem of nucleotide synthesis. One of the ideas that has been around for a long time is that life began from cyanide. Cyanide is not good for us now but it is a fantastic starting material for the chemical synthesis of all the building blocks of life[9]. But how would cyanide be made on the young earth? It is known that many different sources of energy (UV, lightning, impacts) acting on a primitive atmosphere can generate cyanide. The trouble is, how do you actually make use of cyanide that forms in the atmosphere? If it rains out into the ocean it will just hydrolyze and be lost. However John Sutherland has devised a scenario[10] for how to capture, store and accumulate cyanide, and then later turn it into useful products (figure 3). The idea is that cyanide rains out onto the surface and collects in lakes or ponds in an area where there is geothermal circulation of the water through fractured rocks. Water circulating through rocks and heated by the magma deep below comes up through vents and brings up ferrous iron and other ions. Ferrous iron reacts very rapidly with cyanide to make ferrocyanide. Certain ferrocyanide salts are insoluble and will precipitate. One can imagine building up layer after layer of insoluble ferrocyanide salts over thousands of years. As a result, very dilute cyanide could potentially be converted into a huge reservoir of material that could subsequently be processed by heat and water to make useful starting materials.

Figure 3. Geochemical scenario for accumulation of a ferrocyanide reservoir. Cyanide generated in the atmosphere rains out into lakes where it react with Fe2+ ions brought to the surface by hydrothermal circulation of water through fractured rocks. Certain ferrocyanide salts are insoluble and precipitate, building up large reservoirs of cyanide complexes over time.

Another example, also from the Sutherland lab, illustrates the principles of systems chemistry. The compound 2-aminooxazole (2AO) doesn’t look particularly biological but it’s very important because it is a potential prebiotic intermediate on the way to building nucleotides and RNA[11]. It turns out that 2AO can be made from two very simple starting materials: the simplest sugar, glycolaldehyde, and the cyanide derivative cyanamide (figure 4). If you just mix them together, not much interesting happens because you get a huge range of products and polymeric materials, and just a little bit of the compound that we’d like to have. What the Sutherland lab found is that if you put phosphate in the reaction mixture, it works much better. Now phosphate is not one of the starting materials, and it’s not in the product, but we know that phosphate had to be available in the environment because phosphate is part of nucleotides and it is part of RNA. Phosphate is both a good pH buffer and a good acid-base catalyst, and it turns out that in the presence of phosphate our two starting materials react rapidly to make much more of the key intermediate 2AO. This is an excellent example of systems chemistry: using molecules that aren’t directly part of the reaction, but that should be present and that can make the chemistry work better. Remarkably, it turns out that 2AO is volatile, so it sublimes at moderate temperatures and then crystallizes on nearby cold surfaces. As a result it could self-purify and accumulate as a crystalline reservoir.

Figure 4. Systems chemistry and the synthesis of 2-aminooxazole. Glycolaldehyde (red) reacts with cyanamide (blue) to generate insoluble polymeric tar and trace quantities of 2-aminooxazole (2AO). However, in the presence of phosphate (Pi), acting as both a pH buffer and as an acid-base catalyst, 2AO is obtained in high yield.

Another simple physical way of purifying an intermediate is by crystallization from the reaction mixture. In the very next step on the pathway to nucleotide synthesis, 2AO reacts with the sugar glyceraldehyde to form the key intermediate ribose aminooxazoline (RAO). Remarkably, RAO crystallizes out of solution (figure 5), leaving behind left over starting materials as well as all of the other products are made[12],[13]. You can imagine this occurring on a geological scale, and building up a large reservoir of this crystalline intermediate. This is a perfectly natural and plausible way of building up a reservoir of an intermediate so that if conditions change in the right way, the next step in the pathway can happen.

Figure 5. Crystallization of ribose aminooxazoline (RAO). 2AO reacts with glyceraldehyde to form RAO, a precursor of the pyrimidine nucleotides, but the arabinose, lyxose and xylose aminooxazolines are also formed. Only RAO crystallizes from the reaction mixture. Adapted from Hein, Tse and Blackmond, Nat. Chem., 2011.

The examples given above highlight some of the new ways of looking at prebiotic chemistry that make it easier to understand how complicated compounds like nucleotides could actually be formed on the early Earth. But what happens then? Suppose we have all the amino acids, peptides, nucleotides and lipids we need: we still have to understand how these molecules get together to make a living cell. We think a very simple stripped down version of a cell would have looked like this: essentially a primitive cell membrane enclosing little fragments of something like RNA (figure 6)[14]. Here the nature of the questions that we’re asking is different. How could such a cell assemble, and then manage growth and division? Since we are talking about the first cells, there can’t have been any evolved biological machinery to coordinate these processes. Therefore there had to be very simple chemical and physical processes that could drive the primordial cell cycle. The key questions are: how did the primitive cell membrane grow and divide, and how did the genetic material replicate without enzymes? These are the questions that we have been working on for the last 10–20 years; here I will discuss just a few highlights from our efforts. The membranes of primitive cells were certainly not as complicated as the modern cell membranes surrounding our cells; primitive cell membranes had to have been made of simpler materials, and must have had very different properties. One possibility that we and others[15] have explored is that primitive membranes might have been made from fatty acids, which easily self-assemble into membranes. For example, if you start with oleic acid, the fatty acid derived from olive oil and commonly found in soap, and shake it up in water, the fatty acid molecules will spontaneously make bilayer membranes that close up into beautiful spherical vesicles (figure 7).

Figure 6. Schematic model of a protocell. A simple cell might be based on a replicating vesicle for compartmentalization, and a replicating genome to encode heritable information. A complex environment provides nucleotides, lipids and various sources of energy. Mechanical energy (for division), chemical energy (for nucleotide activation), phase transfer and osmotic gradient energy (for growth) may be used by the system. From Mansy et al., Nature, 2008.
Figure 7. Model protocell membranes. Fatty acids such as oleic, myristoleic and capric acids (right) spontaneously assemble into bilayer membranes which form closed vesicles (left). Such vesicles are used to study the growth, division, permeability and other properties of model protocell membranes.

The spontaneous assembly of fatty acid vesicles is reasonably well understood, but what is truly remarkable but far less well understood is that fatty acid vesicles can grow and divide as a result of simple physical processes. What you can see in some of our videos are vesicles with a fatty acid membrane and a fluorescent dye on the inside. We can watch in the microscope and see what happens when we add more food, i.e. more fatty acids[16]. We expected the initial vesicle, which we think of as a model of a protocell membrane, to gradually get bigger, perhaps somewhat elongated if the surface grew faster than the volume. What we didn’t expect was what actually happened. If you watch the movie (http://molbio.mgh.harvard.edu/ szostakweb/videos/Zhu_2009_JACS_01.qt) you will see a thin filament growing out of the originally spherical vesicle, and this filament grows, getting longer and thicker while the contents of the original vesicle gradually spread throughout the filament. This process looks very biological because there is a lot of movement and the shape of the filament changes as it grows. But amazingly this is just soap and water, nothing else. One of the beautiful things about this mode of growth is that the filaments are very fragile, which makes division very easy. At the beginning of our work on modeling protocell division, we had no idea how to make vesicles divide in a way that could easily happen on the early earth. However, once we saw how vesicles actually grow, it turned out that division could easily be driven by gentle environmental forces. Once a vesicle has already grown into a long filamentous shape, a puff of air on the microscope slide is enough to cause the filament to snap in two. The two halves will gradually round up and form two spherical daughter vesicles over time. That kind of experiment taught us that very gentle shear forces are sufficient to cause one vesicle to divide into two daughter vesicles after growth. It turns out that there is at least one other way that protocell division could work[17]. This involves a little bit more chemistry: in an environment with a lot of sulfur, the filamentous vesicle undergoes a strong pearling instability, until it looks like beads on a string, and then the little beads separate from each other over time and float away from each other, as you can see in this movie: (http://molbio.mgh.harvard.edu/ szostakweb/videos/Zhu_2012_PNAS_02.mov). In this remarkable process many small daughter cells are generated from one initial cell. Satisfyingly, by combining these different approaches to growth and division, it is possible to make a vesicle grow into a filament and then divide into multiple daughters, then grow and divide again, and so on, indefinitely. In terms of a primitive cell membrane, we therefore have an environmentally controlled cycle of growth and division. In fact we now have many ways of driving growth in addition to the two different ways of driving division. All of this makes the cell-cycle like processes of growth and division appear to be quite simple in that they can be driven by very gentle and common physical forces.

Next, we need to consider the nature of the primordial genetic material. In order for primitive cells to be able to evolve and thus become better adapted to their environment, they had to have some way of having new functions arise, and those new functions had to be coded in a genetic material so that they could be passed on from generation to generation. The idea that RNA molecules played this early role is quite popular, in part because RNA does so many things in modern biology. One of the most important of those modern functions is protein synthesis: it is actually the RNA components of the ribosome that are responsible for synthesizing all of the proteins in our bodies, and indeed in all living cells on the modern earth[18]. This is a strong rationale for believing that ribozymes existed before protein enzymes. As a result, we think that protocells used RNA to catalyze important reactions such as RNA replication. But before there were ribozymes that could make RNA replicate more efficiently, there must have been chemistry that could have driven the copying of RNA sequences. How might that have worked? In order to illustrate the process, Janet Iwasa prepared a very nice animation to illustrate the nonenzymatic copying of an RNA template, which you can watch here: (http://exploringo-rigins.org/downloads/nonEnzymatic Replication_narr.mov). The idea is that there were short strands of RNA floating around in a rich chemical environment filled with activated building blocks. These activated nucleotides find their partners by base-pairing, G with C, and A with U, and then click together, building up a complementary strand. This process of chemical copying is easy to imagine and to illustrate, but demonstrating this experimentally is still an unsolved problem. People have been thinking about and working on this problem for a long time.

Much of the most important early work on chemical copying of RNA was done by the late Leslie Orgel and his students and colleagues including Jerry Joyce[7]. I think Orgel’s greatest single contribution was to recognize that the nucleoside triphosphates that are used for DNA replication and for making RNA in all modern life are great for cells that have sophisticated enzymes to catalyze polymerization, but without such enzymes, our modern substrates can’t do much except sit there and slowly hydrolyze. So, at the beginning of life, nucleoside triphosphates are not appropriate building blocks. Leslie therefore searched for and found different kinds of activation chemistry that work much better and allow for template copying without enzymes (figure 8). In the best of those molecules that were first made in the Orgel lab[19], two of the phosphates of the triphosphate moiety of modern NTPs have been replaced with a 2-methylimidazole group. As a result, these nucleotides are much more reactive. They don’t need an enzyme to polymerize, so they can copy simple RNA sequences without an enzyme.

Figure 8. Modern vs. prebiotic means of nucleotide activation. Modern cells use nucleoside triphosphates (NTPs, top) for the enzymatic synthesis of RNA and DNA. Primitive cells must have used more reactive and less polar substrates, such as the nucleoside phosphorimidazolides (bottom), for RNA replication.

The reactive nucleotides described above can be used to copy short stretches of RNA. In a typical experiment we begin with a primer, for example a little piece of RNA six nucleotides long, that is base paired to a template, which is a somewhat longer piece of RNA. The goal of the experiment is to copy a string of Cs in the template strand by extending the 6-mer primer using activated G building blocks. The Gs should pair with Cs, and then they will be in position to react with the primer, as a result of which the primer should grow longer. In a time course we start off with just the primer, the six nucleotide RNA. After a minute, the first G nucleotide has been added. A few minutes later the second one, then a few minutes after that, the third G has been added to the primer. This is amazing chemistry: there is no enzyme but the RNA template is being copied! Well, that’s the good news about this chemistry. The bad news is that is the only reaction that works really well. If we try to copy a sequence that has all four nucleotides in it, it basically doesn’t work. We can put in all four activated building blocks and let them sit with the primer and template and nothing happens.

Why is copying mixed sequence templates so hard? Are we thinking about this in the wrong way, and is there some missing chemistry? Due to lack of space I cannot review all of the interesting advances that have been made in tackling this problem[20], so I will skip ahead to discuss one advance from last year in my lab. Remarkably, after 30 years of thinking about and working on this problem, we discovered that if we change the methyl group on the imidazole to an amino group, and make the new leaving group 2-aminoimidazole (2AI) (figure 9), the copying chemistry works much better[21]. As you can see, the pace of progress in this field is sometimes a little slow, but nonetheless, progress is being made. These new 2AI activated monomers are nice because they speed up the copying chemistry; now when we try to extend an RNA primer by copying a template, we can completely copy up to seven nucleotides and sometimes a bit more. We still need to make this chemistry work even better, but we have many ideas about how to copy longer sequences. If we could copy templates that were 20 to 30 nucleotides long, that might be good enough to bootstrap evolutionary processes within synthetic protocells, because we know that we can make good catalytic RNAs by assembling complexes from chunks of RNA that are 20 to 30 nucleotides long. That is our current goal.

Figure 9. A new leaving group for activated nucleotides: 2-aminoimidazole (2AO). Nucleotides activated with 2-aminoimidazole lead to faster and more efficient primer extension than nucleotides activated with 2-methylimidazole.

Let’s return to our new activating group, 2-aminoimidazole (2AI). One question that arises is whether this is just something that works in the lab, or does it have a plausible connection with the origin of life? Would molecules like that have been made in the early chemistry of our planet? This time we think that it might really be relevant. The reason we think so is shown in figure 10. On the left is 2-aminooxazole, the intermediate that goes on to build nucleotides, as worked out in the Sutherland lab[11]. On the right is 2-aminoimidazole (2AI), which we use to activate nucleotides so that they will polymerize. These two molecules differ only in the identity of one atom — where there is an oxygen in 2AO there is a nitrogen in 2AI. It turns that 2 AI can be made in the same reaction mixture as 2AO just by adding ammonia[22]. Even more amazingly, there is another very similar molecule, 2-aminothiazole (2AT), in which the oxygen of 2AO is replaced with sulfur. This compound can react with and store, in a stable crystalline form, the simple 2 and 3 carbon sugars that are needed to make nucleotides and amino acids[23]. I find it amazing that these three very closely related molecules, all made by simple chemistry from simple starting materials in the same or very similar environments, can go on to play different roles in building RNA molecules — this seems like a strong clue that this chemistry is truly relevant to the prebiotic chemistry of the early earth.

Figure 10. Common origins and complementary functions of three closely related heterocycles. 2AO (left) is a precursor of nucleotides, 2AT (center) can store simple sugars as crystalline complexes, and 2AI (left) activates nucleotides for polymerization. All three compounds can be made in the same or very similar conditions.

What did the primitive genetic material look like? Despite the issues discussed above, I still think that the first genetic material was very similar to modern RNA. Many polymers similar to but distinct from RNA have been made in the laboratory and studied, but so far nothing that looks like a better candidate than RNA has emerged. Similarly, recent advances in prebiotic chemistry begin to suggest how RNA might have arisen, but so far don’t point to alternatives. As a result we think the primordial genetic material was more or less the same as RNA, but probably a messier version with some heterogeneity (figure 11). For example, the nucleotides in primitive RNA were probably not always linked up in the ‘correct’ way. We know that different backbone linkages (i.e. 2′-5′ instead of the biologically universal 3′-5′) can form during chemical copying, but we have recently found that this variation is not as harmful as we used to think[24]. It is also possible that there could have been small chemical changes in some of the nucleotides. For example, replacing O2 of uracil with sulfur makes the RNA copying chemistry work much better, and this change is actually still seen in biology in part of the protein synthesis apparatus[25]. There could also be other small changes, yet to be discovered. But overall, the structure of primitive RNA is likely to have been close enough to modern RNA that it could easily evolve step by step into what we see now in biology.

Figure 11. What was the primordial genetic polymer? The illustrated structure is RNA, but it contains some 2′-5′ linkages as well as some non-standard nucleobases such as 2-thio-uracil.

What is still missing from all this chemistry? There has been a huge amount of progress in prebiotic chemistry, but there are still major gaps in our understanding. From the Sutherland lab we think we know how to make the pyrimidine nucleotides U and C in an efficient and plausible manner[26], but we still don’t know how to make the purines. Recent work from the Powner lab describes a beautiful chemical pathway in which a common intermediate is processed in one way to make U and C, and in another way to make 8-oxopurines, but not the natural modern purines. Another problem is that in this chemistry, the phosphate ends up in the wrong position — on the 3′-hydroxyl of the ribose sugar instead of the 5′-hydroxyl as is the case universally in biology, so there’s clearly something missing here. As a result phosphorylation chemistry is a very active area of study. Finally, what are the correct sources of chemical energy? We can drive all these reactions in the lab in convenient but artificial ways, but we don’t understand how this could have happened in an early earth environment. Interestingly we have many good ways of activating amino acids to make peptides. For example, the volcanic gas carbonyl sulfide is a great activating agent. If you bubble carbonyl sulfide through a solution of amino acids, peptides will form and lots of other interesting chemistry will happen as well[27],[28]. So, the search is on for an analogous way to activate nucleotides so they can engage in template copying chemistry and polymerize without enzymes.

What environments on the early earth could have supported the chemistry that gave rise to biology? And what environments could have nurtured the first very primitive cells, and driven their growth, division and evolution? We think that life began in geological settings where organic compounds could accumulate to high concentrations. If useful molecules are made, but fall into the ocean, they will be diluted and lost. This makes surface lakes or ponds seem like more interesting possibilities to consider. Figure 12 is a picture of Yellowstone Lake in the western US. Perhaps in a similar environment on the early earth, organic materials could have built up over long periods of time, generating a richly concentrated chemical environment. We also think that the environment where life began was probably cool most of the time, because RNA is a delicate molecule that will degrade if it is heated for too long. On the other hand, short periods of high temperature might have been essential so that once an RNA template had been copied, making a double-helical product, the complementary strands could be separated and copied again. Geothermally active areas can provide such a thermally fluctuating environment, making volcanically active areas or impact crater lakes seem ideal. For example, in Yellowstone Lake there are hydrothermal vents that emit plumes of hot water (figure 13), and one can imagine primitive cells being caught up in these plumes, heated quickly and then quickly cooled back down as they mix with the surrounding cold lake water. Alternatively, the early earth was bombarded by many asteroids and comets that would create large crater lakes. Such lakes are also hydrothermally active for thousands of years after the impact, and therefore provide another environment that could have played a role in nurturing the beginnings of life.

Figure 12. Yellowstone Lake in winter. Organic compounds could accumulate in similar lakes and ponds on the early earth, and proximity to the atmosphere along with exposure to UV light could have supplied sources of chemical energy. Cool water is good for a delicate molecule such as RNA, but hydrothermal events in the lake emit streams of hot water, which could provide transient exposure to the high temperatures needed for strand separation following template copying.
Figure 13. Hydrothermal vents in Yellowstone Lake. The vents emit streams of water that has been heated by circulation through hot fractured rock, that has in turn been heated by the magma reservoir deep below the surface. Images from Morgan et al. J. Volc. Geotherm.

Because the first cells lacked internal biochemical machinery, it is likely that fluctuations in the environment controlled the processes of protocell growth, division, and replication (figure 14)[29]. Imagine protocells living in a cold water environment that is chemically rich and contains a concentrated mix of all the necessary building blocks of biology (nucleotides, peptides, lipids and chemical sources of energy). Under such conditions, RNA copying could proceed along with growth of the protocell membrane. Every now and then these primordial cells would be pulled into a plume of hot water emanating from a hydrothermal vent, and the brief pulse of high temperature would serve to separate the strands of the RNA duplexes, while at the same time allowing for a sudden influx of nutrients to the cell interior. Before the high temperature could do any damage, the protocells would be swept back out into the cold lake waters, ready to go through another round of genome replication and membrane growth. In this scenario, a complex and fluctuating environment would drive the primitive cell cycle, before the evolution of internal machinery that could direct the cell cycle from inside.

Figure 14. An environmentally driven protocell life cycle. Image from Ricardo and Szostak, Scientific American, 2009.

I have presented an overview of where we are in our thinking about how life got started on the early earth. But, thinking about that kind of problem also raises other interesting questions. In everything I’ve discussed so far we have been trying to address the chemical and physical processes that would eventually give rise to life as we know it, in the form of modern biology. But could there be other possibilities? Would life emerge in a different form in a slightly, or in a very, different chemical environment? Could we design new kinds of life where the biochemistry is different? I think that these are really interesting questions and a huge challenge for the field of chemistry. Below I will point out a few examples both from my lab and from other labs of how this challenge might be approached.

As long as we are considering cellular life, the questions around how different such life could be reduce to the questions of whether it is possible to make membranes and genetic materials from non-biological building blocks, through non-biological chemistry? A good example of artificial membranes that are chemically distinct from biological membranes but still biophysically similar comes from the Devaraj lab at UCSD[30]. They have taken advantage of the very widely used copper click chemistry in which an azide and an alkyne can be joined together with a copper catalyst. This can be used to make lipids that look similar to biological phospholipids, but are chemically a little bit different and are made in a very different way. Remarkably, they have developed a catalyst for this lipid coupling reaction that can also catalyze its own synthesis. As a result, these membrane systems can grow indefinitely as long as they are fed with the correct synthetic building blocks. This is a potential route to making a non-biological protocell membrane system that could grow and divide indefinitely given the necessary resources.

What about a non-biological genetic material? One of the molecules related to RNA and DNA that we’ve been working on in my lab for quite a while is called 3′-NP-DNA or phosphoramidate DNA. The chemical structure of this polymer is the same as that of DNA except that the oxygen atom at the 3′-position of every sugar has been replaced by a nitrogen atom; in the corresponding nucleotide building blocks the 3′-hydroxyl group has been changed to a 3′-amine. The advantage of this change is that in the nucleotides are much more reactive, because an amine is a much stronger nucleophile than a hydroxyl. As a result these molecules are actually easier to copy in a purely chemical system without enzymes[31]. We are continuing to work on NP-DNA because, even though we are not yet able to fully replicate NP-DNA oligonucleotides, we are close enough that we can start to imagine the possibility of building living cells that use a genetic material that is different from RNA and DNA.

The examples of non-biological membranes and genetic polymers discussed so far remain quite close to biology as we know it, which raises the question of whether we can go further afield? Can we imagine making new kinds of living systems where the chemistry is really completely different? Speculation about such highly divergent forms of life was greatly stimulated when the Cassini-Huygens mission discovered the lakes and seas on Saturn’s moon Titan (figure 15). These are not lakes of water, but of liquid methane and ethane. Can we imagine anything living in an environment like that? Liquid methane is hard to work with in the lab, but we can easily work with nonpolar organic solvents such as decane. The challenge of making membrane vesicles in decane was actually addressed by the Kunieda lab in Japan, more than 20 years ago[32]. figure 16 shows vesicles composed of inside-out membranes with the polar parts of the lipids in the membrane interior, and the hydrophobic parts sticking into the hydrophobic solvent. They look like completely normal vesicles but they’re inside out and the molecular components are totally different. Given that it is possible to make membranes in such a solvent, what kind of genetic material can we imagine in such a solvent? This is probably a more difficult challenge, but one that is almost irresistible. Our work on this is at a very early stage, but it is a fascinating project. One of the wonderful things about working on such different genetic polymers is that it makes us appreciate RNA and DNA even more, since we can see how hard it is to design something new that has the same properties.

Figure 15. Lakes of methane and ethane on the surface of Saturn’s moon, Titan. Could there be life in such environments?
Figure 16. Reverse vesicles in decane. The vesicle membranes are assembled from molecules that contain hydrophobic regions that point out to the decane phase, and hydrophilic regions that cluster together in the middle of the membrane. Image from H Kunieda, K Nakamura, and DF Evans. Formation of Reversed Vesicles. JACS 1991 113: 1051–1052.

To summarize, I have tried to present an overview of the new questions and new ideas that are changing how we think about the chemistry that led to life on the early Earth. I have also tried to point out some of the tantalizing beginnings of work that may lead to the ability to construct living systems that are chemically completely different from life on Earth. These are both fantastically interesting challenges, so stay tuned for future progress!

Acknowledgements: I would like to thank the many fantastic students, post-docs and collaborators that I have had the pleasure of working with over the years. Work in my lab on the origin of life has been supported by grants from the Simons Foundation (290363), the NSF (CHE-1607034) and NASA (NNX15AL18G). I am an Investigator of the Howard Hughes Medical Institute.

References

[1] Lorsch, J.R. & Szostak, J.W. Chance and necessity in the selection of nucleic acid catalysts. Accounts Chem. Res. 29, 103–110 (1996).

[2] Szostak, J.W., Bartel, D.P. & Luisi, P.L. Synthesizing life. Nature 409, 387–390 (2001).

[3] Szostak, J.W. The eightfold path to non-enzymatic RNA replication. J. System Chem. 3, 2 (2012).

[4] Gillon, M., Triaud, A.H., Demory, B.O., Jehin, E., Agol, E., Deck, K.M., Lederer, S.M., de Wit, J., Burdanov, A., Ingalls, J.G., Bolmont, E., Leconte, J., Raymond, S.N., Selsis, F., Turbet, M., Barkaoui, K., Burgasser, A., Burleigh, M.R., Carey, S.J., Chaushev, A., Copperwheat, C.M., Delrez, L., Fernandes, C.S., Holdsworth, D.L., Kotze, E.J., Van Grootel, V., Almleaky, Y., Benkhaldoun, Z., Magain, P. & Queloz, D. Seven temperate terrestrial planets around the nearby ultracool dwarf star TRAPPIST-1. Nature 542(7642), 456–460 (2017).

[5] Sutherland, J.D. The origin of life — Out of the blue. Angew. Chem. Int. Ed. Engl. 55(1), 104–121 (2016).

[6] Miller, S.L. A production of amino acids under possible primitive earth conditions. Science 117(3046), 528–529 (1953).

[7] Orgel, L.E. Prebiotic chemistry and the origin of the RNA world. Crit. Rev. Biochem. Mol. Biol. 39(2), 99–123 (2004).

[8] Islam, S. & Powner, M.W. Prebiotic systems chemistry: Complexity overcoming clutter. Chem. 2(4), 470–501 (2017).

[9] Patel, B.H., Percivalle, C., Ritson, D.J., Duffy, C.D. & Sutherland, J.D. Common origins of RNA, protein and lipid precursors in a cyanosulfi dic protometabolism. Nat. Chem. 7(4), 301–307 (2015).

[10] Ritson, D.J. & Sutherland, J.D. Synthesis of aldehydic ribonucleotide and amino acid precursors by photoredox chemistry. Angew. Chem. Int. Ed. Engl. 52(22), 5845–5847 (2013).

[11] Powner, M.W., Gerland, B. & Sutherland, J.D. Synthesis of activated pyrimidine ribonucleotides in prebiotically plausible conditions. Nature 459(7244), 239–242 (2009).

[12] Springsteen, G. & Joyce, G.F. Selective derivatization and sequestration of ribose from a prebiotic mix. J. Am. Chem. Soc. 126(31), 9578–9583 (2004).

[13] Hein, J.E., Tse, E. & Blackmond, D.G. A route to enantiopure RNA precursors from nearly racemic starting materials. Nat. Chem. 3(9), 704–706 (2011).

[14] Mansy, S.S., Schrum, J.P., Krishnamurthy, M., Tobé, S., Treco, D.A. & Szostak, J.W. Template-directed synthesis of a genetic polymer in a model protocell. Nature 454(7200), 122–125 (2008).

[15] Budin, I. & Szostak, J.W. Expanding roles for diverse physical phenomena during the origin of life. Annu. Rev. Biophys. 39, 245–263 (2010).

[16] Zhu, T.F. & Szostak, J.W. Coupled growth and division of model protocell membranes. J. Am. Chem. Soc. 131(15), 5705–5713 (2009).

[17] Zhu, T.F., Adamala, K., Zhang, N. & Szostak, J.W. Photochemically driven redox chemistry induces protocell membrane pearling and division. Proc. Natl. Acad. Sci. U.S.A. 109(25), 9828–9832 (2012).

[18] Nissen, P., Hansen, J., Ban, N., Moore, P.B. & Steitz, T.A. The structural basis of ribosome activity in peptide bond synthesis. Science 289(5481), 920–930 (2000).

[19] Inoue, T. & Orgel, L.E. Oligomerization of (guanosine 5′-phosphor)-2-methylimidazolide on poly(C). An RNA polymerase model. J. Mol. Biol. 162(1), 201–217 (1982).

[20] Szostak, J.W. The narrow road to the deep past: In search of the chemistry of the origin of life. Angew. Chem. Int. Ed. Engl. 56(37), 11037–11043 (2017).

[21] Li, L., Prywes, N., Tam, C.P., O’Flaherty, D.K., Lelyveld, V.S., Izgu, E.C., Pal, A. & Szostak, J.W. Enhanced nonenzymatic RNA copying with 2-aminoimidazole activated nucleotides. J. Am. Chem. Soc. 139(5), 1810–1813 (2017).

[22] Fahrenbach, A.C., Giurgiu, C., Tam, C.P., Li, L., Hongo, Y., Aono, M. & Szostak, J.W. Common and potentially prebiotic origin for precursors of nucleotide synthesis and activation. J. Am. Chem. Soc. 139(26), 8780–8783 (2017).

[23] Islam, S., Bučar, D.-K. & Powner, M.W. Prebiotic selection and assembly of proteinogenic amino acids and natural nucleotides from complex mixtures. Nat. Chem. 9, 584–589 (2017).

[24] Engelhart, A.E., Powner, M.W. & Szostak, J.W. Functional RNAs exhibit tolerance for non-heritable 2′-5′ versus 3′-5′ backbone heterogeneity. Nat. Chem. 5(5), 390–394 (2013).

[25] Heuberger, B.D., Pal, A., Del Frate, F., Topkar, V.V. & Szostak, J.W. Replacing uridine with 2-thiouridine enhances the rate and fidelity of nonenzymatic RNA primer extension. J. Am. Chem. Soc. 137(7), 2769–2775 (2015).

[26] Xu, J., Tsanakopoulou, M., Magnani, C.J., Szabla, R., Šponer, J.E., Šponer, J., Góra, R.W. & Sutherland, J.D. A prebiotically plausible synthesis of pyrimidine β -ribonucleosides and their phosphate derivatives involving photoanomerization. Nat. Chem. 9(4), 303–309 (2017).

[27] Leman, L., Orgel, L. & Ghadiri, M.R. Carbonyl sulfidemediated prebiotic formation of peptides. Science 306(5694), 283–286 (2004).

[28] Izgu, E.C., Björkbom, A., Kamat, N.P., Lelyveld, V.S., Zhang, W., Jia, T.Z. & Szostak, J.W. N-carboxyanhydride-mediated fatty acylation of amino acids and peptides for functionalization of protocell membranes. J. Am. Chem. Soc. 138(51), 16669–16676 (2016).

[29] Ricardo, A. & Szostak, J.W. Origin of life on earth. Sci. Am. 301(3), 54–61 (2009).

[30] Hardy, M.D., Yang, J., Selimkhanov, J., Cole, C.M., Tsimring, L.S. & Devaraj, N.K. Self-reproducing catalyst drives repeated phospholipid synthesis and membrane growth. Proc. Natl. Acad. Sci. U.S.A. 112(27), 8187–8192 (2015).

[31] Zhang, S., Blain, J.C., Zielinska, D., Gryaznov, S.M. & Szostak, J.W. Fast and accurate nonenzymatic copying of an RNA-like synthetic genetic polymer. Proc. Natl. Acad. Sci. U.S.A. 110(44), 17732–17737 (2013).

[32] Kunieda, H., Nakamura, K. & Evans, D.F. Formation of reversed vesicles. J. Am. Chem. Soc. 113, 1051–1052 (1991).

Republished with permission from the author.

Jack W. Szostak is a Canadian American biologist of Polish British descent, Nobel Prize laureate, Professor of Genetics at Harvard Medical School, and Alexander Rich Distinguished Investigator at Massachusetts General Hospital, Boston. Prof. Szostak has made significant contributions to the field of genetics. His achievement helped scientists to map the location of genes in mammals and to develop techniques for manipulating genes. His research findings in this area are also instrumental to the Human Genome Project. He was awarded the 2009 Nobel Prize for Physiology or Medicine, along with Elizabeth Blackburn and Carol W. Greider, for the discovery of how chromosomes are protected by telomeres. You can read Prof. Szostak’s Nobel Lecture here.

Jack Szostak: Origin of life on earth and design of alternatives

The handedness of life – with Jack Szostak

Dr. Jack Szostak and fellow researchers Carol Greider and Elizabeth Blackburn shared the Nobel Prize

10-on-10: The Chronicles of Evolution – Jack W. Szostak

Be sure to ‘like’ us on Facebook

1 COMMENT

  1. March 4, 2019

    Dear Professor Szostak,

    Thanks to the prescience of our editor, I have immensely enjoyed reading your illuminating article The Origin of Life on Earth and the Design of Alternative Life Forms http://churchandstate.org.uk/2019/02/the-origin-o

    I am one of a kind of reader which this site seeks to attract, namely people without all the huge scientific credentials that Nobel winners have, but rather those who simply want to enlarge their horizons from narratives written by the giants who receive these prizes in sufficiently simple language which us non-scientists can enjoy because a piece like yours tells such a riveting story. I have forwarded this to others who I think will find it equally interesting.

    Thank you. I shall be alert for your further offerings as your work progresses.

    Donald A. Collins, Sr.

LEAVE A REPLY

Please enter your comment!
Please enter your name here