Chemists are now tantalisingly close to re-inventing in the laboratory the initial spark that led to the evolution of life on earth. Michael Gross reports
Living organisms are incredibly complex. But we can trace back their diverse forms and functions to simpler ancestors – all the way back to one, single-celled species, known as the Last Universal Cellular Ancestor (LUCA).
On the molecular level, LUCA was already a very complex network of catalysed reactions, featuring DNA, RNA and hundreds of proteins. At its heart, there was already the highly complex protein synthesis machine – the ribosome – which reads a messenger RNA to produce specific proteins.
Detailed studies of the structure and evolution of the ribosome have confirmed it evolved from a precursor containing only RNA. The first protein ever to be produced was made by an RNA molecular machine. RNA is also able to take over the function of DNA for the storage of genetic information – as it still does in viruses including HIV and coronaviruses. The ‘RNA world’ hypothesis of an early form of life relying on RNA as the only information-carrying macromolecule is fairly well established as the most likely scenario for the first phase of genetically coded and evolving life forms.
The evolution of the ribosome tells us quite clearly how the RNA world ended – with the RNA ribosome making more and more proteins that took over more and more functions. But how did the RNA world begin? How did RNA molecules start to take part in a scheme to produce more of their own kind? Chemists are now close to finding out.
Remnants of the RNA world are scattered all around modern-day biochemistry. Beyond the use of ribosomal and transfer RNAs in protein biosynthesis, RNA building blocks have retained important roles in our metabolism, such as ATP as energy carrier and GTP in signal conversion.
Intriguingly, however, one area where evolution has wiped out all traces of previous versions is RNA synthesis. Unlike the ribosome and certain RNA processing enzymes, RNA polymerases do not contain RNA remnants and give us no indication how the RNA world conducted its most basic chore, making RNA.
Therefore, researchers aiming to understand the early evolution of life on our planet only have one option, to re-invent RNA-replicating ribozymes – enzymes made of RNA that preceded modern protein enzymes. Fortunately, they can use a methodology that in a way is very similar to what likely happened in the very beginning: producing a large number of random RNA sequences and letting in vitro evolution optimise the ones that might perform the desired task.
However, therein lies a dilemma. For a realistically sized small ribozyme with 70 building blocks, there are 1042 possible sequences. The combined weight of just a single copy of each of these molecules would match Saturn’s moon Mimas (4x1019kg). On the other hand, a laboratory sample of random sequences may contain some 1014 variations. This gap of around 28 orders of magnitude cannot be closed with the timescales and resources accessible to humans.
Faced with this problem, the group of David Bartel at the Massachusetts Institute of Technology, US, used a known ribozyme of a related activity as a starting point for the artificial evolution experiments. They started from a ligase, which Bartel had discovered in his graduate work with Jack Szostak, and which couples strands of RNA end-to-end. After many cycles of artificial mutation and selection, they obtained the first ever polymerase ribozyme, reported in 2001.1
For a realistically sized small ribozyme with 70 building blocks, there are 1042 possible sequences.
The faithful copying of nucleic acids – a necessity for life on earth to begin – may have been promoted by a heat-flux system in volcanic rocks on our new-born planet.
This molecule could not have started the evolution of life on Earth, as it had a length of 180 nucleotides but was only able to produce strands no longer than 14 nucleotides. Moreover, it had a complex folded structure required for its activity, which impeded its accessibility as a template for making further copies. As with our DNA, RNA is copied via a complementary strand, so making an identical copy of a given sequence involves producing a strand with the complementary sequence, and then using that strand as a template to synthesise the copy of the first one.
Modular design promises the solution for both these problems: If the required large and tightly folded ribozyme is assembled from several shorter RNA strands, these will be easier to access as templates and also more likely to be copied faithfully. The group of Gerald Joyce at the Salk Institute at La Jolla, US, has combined this modular design principle with new and improved selection strategies and is now approaching the goal that has vexed researchers for the last 25 years.
This research, still using Bartel’s ligase as a starting point, numbers the new ribozymes by the number of steps they are removed from the original ligase. Thus, one milestone on the way was the 24-3 polymerase reported by David Horning and Gerald Joyce in 2016, which showed a dramatically improved ability to copy complex templates and polymerised RNA a hundred times faster than the best rate previously achieved. It could synthesise a yeast transfer RNA with a complex structure consisting of 76 nucleotides.2
Katrina Tjhung and colleagues in Joyce’s lab have now added another 14 rounds of in vitro evolution, selecting on the ability to synthesise a natural ribozyme, the hammerhead ribozyme, variants of which occur in many species. Tjhung and colleagues came up with a new version, the 38-6 polymerase, which can successfully synthesise its own evolutionary ancestor, the ligase ribozyme from which Bartel started two decades earlier.3
Hammerhead ribozyme, molecular model. Ribozymes are RNA (ribonucleic acid) molecules that catalyse certain biochemical reactions | Credit: LAGUNA DESIGN/SCIENCE PHOTO LIBRARY
Implementing the modular design idea, the researchers dissected the ancestral ligase into three fragments that spontaneously assemble to form the fully functional ribozyme. They were able to show that the 38-6 polymerase can synthesise all these strands as well as their complementary strands. Thus, the goal appears within sight. If they can design an analogous assembly pattern for the 38-6 polymerase or something derived from it, the researchers will have a self-reproducing RNA molecule that will be alive and could just be bred in the lab to evolve further improvements.
One remaining worry is the poor fidelity of copying in 38-6 polymerase. Even though evolution depends on the occasional mutation to produce the variety needed for selection to work on, the current version of the polymerase is making several errors in every RNA strand it produces. Considering that duplicating a given strand requires two rounds of transcription, to the complementary sequence and back, this low level of fidelity risks producing chaos rather than a re-run of the RNA world.
Another issue of concern is the necessity to form a double helix for structural stability and enzymatic function, but to dissolve it when a strand needs to be copied. In a recent paper, Joyce and Dieter Braun at the Ludwig Maximilian University (LMU) in Munich, Germany, have proposed that a heat-flux system in volcanic rocks on our newborn planet may have enabled the copying of nucleic acids and separation of strands without risking the chemical integrity of the polymer.4 Such volcanic ‘thermocyclers’ may have anticipated the modern use of PCR equipment by a few billion years.
How might genetic information in an RNA World have been converted into functional RNA in response to a changing environment? This is a critical aspect of modern life but has largely been ignored by the RNA World community.
Peter Unrau Professor in the Department of Molecular Biology and Biochemistry, Simon Fraser University, Canada
Towards gene expression
If and when polymerase scientists come up with a ribozyme that can copy its own kind, this will be an invaluable tool to investigate the possible origins of life, as it can be left to evolve in an experimental setting. However, there are further hurdles that the nascent RNA world had to overcome to turn into something that we might recognise as life.
Peter Unrau and Razvan Cojocaru from Simon Fraser University at Burnaby, Canada, have identified two such hurdles, namely first the issue of gene expression: ‘How might genetic information in an RNA world have been converted into functional RNA in response to a changing environment?’ Unrau wonders. ‘This is a critical aspect of modern life but has largely been ignored by the RNA world community.’
Unrau’s second worry is parasitism. Once a working polymerase is operating, sequences that are very good at being copied, but no use as polymerases might come to dominate the pool and lead to a dead end.
Unrau, whose lab set a record for the most effective polymerase ribozyme in 2007, now thinks a common solution for all these problems exists. ‘A beautiful feature of all nucleic acid strands is they can recognise each other easily,’ Unrau says. Taking advantage of this fact, the team developed a strategy based on the polymerase recognising promoter sequence as a criterion for the in vitro evolution of new more biologically nuanced polymerases, published in Science in March 2021.5 In today’s cellular life forms, promoter sequences are present to identify genes that are to be read by the cellular machinery for making messenger RNA and then proteins.
‘We have found a polymerase ribozyme that does something highly analogous to this, ie modern gene expression. It finds RNA promoters and then clamps onto the appropriate RNA template. As a result of this it becomes much more processive,’ Unrau explains. Ideal processivity in the synthesis of biological macromolecules means that the molecular machinery stays on the template track and keeps going until it hits a stop signal or the end of the track.
The important difference to previous polymerases is there is now a two-way dialogue reminiscent of modern gene expression. Recognition of the promoter sequence switches the polymerase into an activated state. This state is not only enzymatically active but also clamped tight around the template track, so it cannot fall off before it has completed the synthesis.
A complicating factor is that the promoter recognition and activation of the ribozyme requires a third party, a primer sequence, whose function is analogous to the sigma factors in bacteria. While bacterial sigma factors are proteins, the RNA world experiments naturally use RNA primers. In another step leading closer to the recreation of a workable RNA world, Unrau’s group could show that the new polymerase can synthesise at least part of this primer, so it does not rely on separate, unlikely events coming together. The polymerase uses itself as a template to synthesise part of its primer, allowing only the specific template to be copied, based on the sequence of the polymerase. This addresses the issue of replicative parasites, or sequences good at being copied, giving the polymerase control over which sequences it copies.
‘We got all of this adding one simplifying idea to the basic RNA world polymerase story,’ Unrau concludes, ‘namely that polymerase conformation states have likely to have been of central importance since the dawn of life.’
Detailed studies of the structure and evolution of the ribosome have confirmed it evolved from a precursor containing only RNA.
A RNA ribozyme that can copy its own kind will be an invaluable tool to investigate the possible origins of life.
Finding the spark
When Stanley Miller first pondered the primordial soup experiments that were to make him famous, his advisor, Harold Urey, tried to dissuade him, worrying that recreating processes that may have taken millions of years on primeval Earth would exceed the usual length of time accorded to a PhD thesis. He need not have worried, as Miller obtained exciting results within days, demonstrating the formation of important building blocks of life including amino acids.
Similarly, the next step towards the evolution of living things, the emergence of the first molecules to replicate and evolve, may have taken millions of years and may have faced astronomical odds. However, after finding important shortcuts towards molecules that are almost ready to spring to life, scientists are now tantalisingly close to making that improbable spark happen in the laboratory and turn chemistry into biology.
1 W. K. Johnston et al.; Science, 2001, 292, 1319
2 D. P. Horning and G. F. Joyce; Proc. Natl. Acad. Sci. USA, 2016, 113, 9786
3 K. Tjhung et al.; Proc. Natl. Acad. Sci. USA, 2020, 117, 2906
4 A. Salditt et al.; Phys. Rev. Lett. 2020, 125, 048104
5 R. Cojocaru and P. Unrau; Science, 2021, 371, 1225