On being the right size

The first replicating molecules were probably composed of RNA and undoubtedly small, limited in size by a self-destructing error rate. A new study shows that a relatively minor increase in replication fidelity may have had a large effect on the size, and hence complexity, of early replicators.

1953 was a landmark year for biological science. To many, it is best known as the beginning of a new age in molecular genetics, starting with the publication of Watson and Crick's structure of DNA¹. Yet 1953 also saw the dawn of modern studies on the origin of life. By simulating probable conditions of the earth's early atmosphere and using 'spark discharge' as an energy source, Stanley Miller achieved the prebiotic synthesis of a variety of organic compounds including amino acids, thereby showing that an endogenous origin of life on earth was feasible². In stark contrast to the developments made since Watson and Crick's work, retracing the steps in the origin of life remains an elusive goal, hampered by a multitude of interconnected problems. On page 1008 of this issue, Kun et al.³ shed new light on one of the key hurdles associated with the development of life on earth: how to evolve genetic complexity.

The error threshold

Understanding the origin of life requires us to solve some of the most fundamental problems in evolution: how to produce the first polymers, how to generate the first replicating molecule, how to make the first cell. Perhaps a more generic problem is how a simple replicating system can evolve increased genotypic, and hence phenotypic, complexity. Since the discovery of ribozymes, self-replicating RNA molecules with catalytic properties, it has been generally accepted that early life at one point existed as an 'RNA world', in which the only replicators were RNA molecules⁴, perhaps functioning together as 'bags of genes'⁵. Although this is an interesting theory, reliance on RNA replication comes at a cost; compared with DNA, the copying of RNA is highly error-prone. This, in turn, means that there is an upper limit on the size of primitive RNA replicators, as those larger than this intrinsic 'error threshold' will be unable to copy themselves with sufficient fidelity, and the system will degenerate. Hence, many of the ribozymes we know today are only a few hundred nucleotides in length. To create more genetic complexity, it is therefore necessary to encode more information in longer genes by using a replication system with greater fidelity. But there's a catch: to replicate with greater fidelity requires a more accurate and hence complex replication enzyme, but such an enzyme cannot be created because this will itself require a longer gene, and longer genes will breach the error threshold. This evolutionary chicken-and-egg dilemma has been dubbed Eigen's paradox (following Manfred Eigen's seminal work on the nature of early replicators⁶) and is one of the most intractable puzzles in the origin of life. Kun et al.³ now go a long way to providing an answer.

A distinctive feature of RNA molecules is that they generally form complex secondary structures, complete with loops, hairpins and bulges. This is central to understanding the evolution of early replicators, because such structures, and the complex fitness landscapes they enforce, means that many of the mutations that arise either will be neutral, and hence have no effect on fitness, or will interact epistatically. In sum, the effect of RNA secondary structure is to remove the one-to-one relationship between genotype and phenotype, producing what Kun et al. call a relaxed error threshold³, buffering early replicators against mutational meltdown. Crucially, if the fitness landscapes of present-day ribozymes can be calculated, as Kun et al. attempt to do³, then it should also be possible to estimate, albeit roughly, what sizes early RNA replicators would have been able to achieve under given mutation rates. As raw data, Kun et al. examined the fitness landscapes of two small ribozymes, the hairpin and Neurospora VS ribozymes³. The secondary structures of these ribozymes can be predicted, and extensive mutagenesis studies can be used to estimate fitness. Fitting data to model suggested that a decrease in the error rate from 0.1–0.01 to 0.001 mutations per nucleotide per replication would result in many replicators in the size range of 7–8 kb, equivalent to that of contemporary tRNAs and RNA viruses (Fig. 1).

Figure 1: Error thresholds and the origin of life.

For evolution to proceed in the RNA world, a reduction in the error rate (μ) to ∼0.001 mutations per nucleotide per replication must be achieved. For more complex genomes to evolve, with sizes >10⁴ nucleotides, a second threshold needs to be crossed in which μ ≪ 0.001. This necessitates the evolution of DNA replication, which provides a higher copying fidelity than RNA. Neutral and compensatory changes tend to dampen the effects of deleterious mutations, allowing a relaxed error threshold. In contrast, antiviral drugs that increase the error rate tend to push contemporary RNA viruses over the error threshold and into lethal mutagenesis.

From RNA to DNA

Despite the results of Kun et al.³, some key questions remain unanswered. In particular, although an increase in replicator size from several hundred nucleotides to >7 kb represents a large enhancement of genetic complexity and is undoubtedly sufficient for an RNA world, it is very much smaller than either the tiniest cellular life forms known today or the estimated size of the genome of the last universal common ancestor⁷. More fundamentally, the most abundant RNA life forms, RNA viruses, usually have genomes <12 kb in length. Hence, although most RNA viruses are larger and more complex than ribozymes and have greater copying fidelity, they are also at the mercy of an error threshold (Fig. 1). This recognition is the theoretical basis of a new and promising form of viral treatment that involves forcing RNA viruses over the error threshold through the application of antiviral agents like ribavirin that act as mutagens, thereby producing a “lethal mutagenesis”⁸. How, then, to move from the small genomes of RNA viruses to those of more complex life forms? Given that the largest replicating RNA molecules are the coronaviruses, at 30 kb, the answer must involve the evolution of DNA replication that comes equipped with proof-reading. That is another intricate evolutionary story, particularly given the growing evidence for RNA proof-reading⁹. Explaining the full history of the origin of life will continue to perplex evolutionary biologists for many generations to come.