Brevia

Extremely High Mutation Rate of a Hammerhead Viroid

See allHide authors and affiliations

Science  06 Mar 2009:
Vol. 323, Issue 5919, pp. 1308
DOI: 10.1126/science.1169202

Abstract

The mutation rates of viroids, plant pathogens with minimal non-protein-coding RNA genomes, are unknown. Their replication is mediated by host RNA polymerases and, in some cases, by hammerhead ribozymes, small self-cleaving motifs embedded in the viroid. By using the principle that the population frequency of nonviable genotypes equals the mutation rate, we screened for changes that inactivated the hammerheads of Chrysanthemum chlorotic mottle viroid. We obtained a mutation rate of 1/400 per site, the highest reported for any biological entity. Such error-prone replication can only be tolerated by extremely simple genomes such as those of viroids and, presumably, the primitive replicons of the RNA world. Our results suggest that the emergence of replication fidelity was critical for the evolution of complexity in the early history of life.

Mutation rates vary by orders of magnitude across species (1, 2), with the highest rates measured so far corresponding to RNA viruses (3), but little is known about other RNA replicons. Viroids are plant pathogens with minimal non-protein-coding RNA genomes replicated by host RNA polymerases (4). We estimated the mutation rate of Chrysanthemum chlorotic mottle viroid (CChMVd), a 399-nucleotide chloroplastic viroid with hammerhead ribozymes. Hammerheads are RNA motifs formed by three double-helix regions flanking a core of 15 highly conserved nucleotides critical for catalytic activity (5), which mediate self-cleavage of replicative intermediates and, hence, are essential for viroid replication. Hammerhead viroids show elevated genetic variability (6), but this variability results from the combined action of mutation and selection and therefore cannot be used to directly estimate mutation rates.

To achieve this goal, we inoculated plants with an in vitro transcript of CChMVd (7), and at the onset of symptoms we screened for mutations at the 15 core nucleotides plus the nucleotide preceding the self-cleavage site in each of the two hammerheads (32 sites). Considering that these mutations are lethal for the viroid, their population frequency must equal the mutation rate because, despite multiple replication rounds downstream from inoculation, they have necessarily been generated during the last one. In three independent experiments, we found three, seven, and five mutations in 63, 64, and 61 reverse transcription polymerase chain reaction (RT-PCR) clones, respectively (188 × 32 = 6016 total target sites), yielding a mutation rate of 0.0025 ± 0.0006 (SEM) per site and replication cycle, that is, one mutation per 400 nucleotides (fig. S1).

In a control experiment, we sequenced RT-PCR clones from the in vitro transcript used for inoculations and found a single substitution in 6525 sites. This result gives an error rate 17-fold lower than the estimated CChMVd mutation rate and discards any significant effect of RT-PCR artifacts. To confirm the lethality of the hammerhead mutations sampled in vivo, we recreated the mutations by site-directed mutagenesis and assayed for infectivity. Northern-blot hybridizations indicated that plants inoculated with these mutants had no detectable viroid RNA (fig. S2A). Further, self-cleavage analyses confirmed that all except one of the mutant hammerheads showed severely reduced or null catalytic activity (fig. S2B).

To determine the strength of selection against mutations elsewhere in the viroid genome, we competed 24 random-point mutants against the wild type. Sequencing of 138 RT-PCR clones revealed that 20/24 mutations had been purged by selection at the onset of symptoms. In contrast, 51 new polymorphisms appeared in this time interval, showing that genetic variability is rapidly regenerated because of highly error-prone replication (fig. S3). We also inferred that hammerheads are unlikely to constitute mutational hotspots because polymorphisms did not map more frequently in hammerheads than in the rest of the genome (Fisher exact test, P = 0.963) whereas the fraction of point mutations that were selected against was also similar for these two regions (7/8 and 13/16, respectively).

The CChMVd mutation rate is the highest reported for any biological entity (Fig. 1). Hammerhead viroids are replicated by a proofreading-deficient chloroplastic DNA-dependent RNA polymerase that is redirected to use RNA instead of its native DNA template (4). This, together with the presence of mutagenic free radicals or unbalanced nucleotide pools, would lead to extremely error-prone replication. Viroids can tolerate such elevated per-site mutation rates owing to their minimal genomes, whereas more complex genomes would accumulate an excessive mutational load (8). Given their genomic simplicity and autocatalytic activity, hammerhead viroids are reminiscent of the postulated RNA world replicons (9). These primitive replicons would also resemble hammerhead viroids in their extremely error-prone replication. Thus, our results support the notion that the emergence of replication fidelity mechanisms was central to the evolution of complexity in the early history of life.

Fig. 1.

Per-site mutation rate versus genome size for CChMVd and other biological entities [reviewed in (2) and updated with more recent data from (3)]. RNA viruses (left to right) are tobacco mosaic virus, human rhinovirus, poliovirus, vesicular stomatitis virus, bacteriophage Φ6, and measles virus. Single-stranded DNA viruses are bacteriophage ΦX174 and bacteriophage m13. Double-stranded DNA viruses are bacteriophage λ, herpes simplex virus, bacteriophage T2, and bacteriophage T4. Bacteria is Escherichia coli. Lower eukaryotes are Saccharomyces cerevisiae and Neurospora crassa. Higher eukaryotes are Caenorhabditis elegans, Drosophila melanogaster, Mus musculus, and Homo sapiens. When several estimations were available, the mean value is shown.

Supporting Online Material

www.sciencemag.org/cgi/content/full/323/5919/1308/DC1

Materials and Methods

Figs. S1 to S3

References

References

View Abstract

Navigate This Article