Unique cost dynamics elucidate the role of frameshifting errors in promoting translational robustness

Abstract

There is now considerable evidence supporting the view that codon usage is frequently under selection for translational accuracy. There are, however, multiple forms of inaccuracy (missense, premature termination, and frameshifting errors) and pinpointing a particular error process behind apparently adaptive mRNA anatomy is rarely straightforward. Understanding differences in the fitness costs associated with different types of translational error can help us devise critical tests that can implicate one error process to the exclusion of others. To this end, we present a model that captures distinct features of frameshifting cost and apply this to 641 prokaryotic genomes. We demonstrate that, although it is commonly assumed that the ribosome encounters an off-frame stop codon soon after the frameshift and costs of mis-elongation are therefore limited, genomes with high GC content typically incur much larger per-error costs. We go on to derive the prediction, unique to frameshifting errors, that differences in translational robustness between the 5' and 3' ends of genes should be less pronounced in genomes with higher GC content. This prediction we show to be correct. Surprisingly, this does not mean that GC-rich organisms necessarily carry a greater fitness burden as a consequence of accidental frameshifting. Indeed, increased per-error costs are often more than counterbalanced by lower predicted error rates owing to more diverse anticodon repertoires in GC-rich genomes. We therefore propose that selection on tRNA repertoires may operate to reduce frameshifting errors.

FIG. 1.—

Schematic representation of frameshifting cost. n_pre (n_post) is the number of codons translated before (after) the frameshift occurs, either in (A) the +1 or (B) the –1 direction. n_post is determined as the number of codons translated until an off-frame stop (highlighted in gray) is encountered, which minimally requires a T in the 2nd (3rd) position of the original frame for +1 (−1) events. In this example, p_{i + 1} (p_{i − 1}) is the probability of shifting from codon CCA onto CAG (TCC).

FIG. 2.—

The relationship between GC3 content and components of frameshifting cost. (A) n_post (median number across sites), (B) n_pre + n_post (median number across sites), (C) ∑ p_i (n_pre + n_post) (median cost across genes, i.e., for each gene, we sum cost across all codons in that gene to obtain a gene-specific cost and then plot the median by-gene cost across all genes). Genomic GC3 content is the proportion of G and C nucleotides across all 3rd sites across all coding sequences analyzed. Each data point represents one genome.

FIG. 3.—

Frameshifting probabilities decline with GC content for the majority of frameshifting contexts. We computed nonparametric correlations (Kendall’s tau) between genomic GC3 content and p_i for all possible frameshifting contexts (NNN|N or N|NNN for +1 and –1 shifts, respectively) across genomes. Each data point represents one such context. The majority of (A) +1 and (B) –1 contexts exhibit negative correlation coefficients, indicating that, for the respective context, the probability of frameshifting decreases with increasing GC3. In particular, there is typically a negative correlation for frameshifting contexts that can be especially prone to shifting (high maximum p_i).

FIG. 4.—

Anticodon repertoire (the number of different anticodons among all tRNA genes in the genome) increases with GC3 content.

FIG. 5.—

Anticodon sparing strategies as a function of GC content. The “absence” of a particular anticodon (rows) from a particular genome (columns, ordered by genomic GC3 content) is indicated to highlight that tRNAs with C or G in the first position of the anticodon are more frequently spared in genomes toward the lower end of the GC spectrum.

FIG. 6.—

Differences in frameshifting robustness become less pronounced with increasing GC3. Frameshifting robustness scores (FRSs) were computed as 1 − p_i for each of the 5′ and 3′ terminal 100 codons and averaged across codons. Averages were then subtracted (5′ − 3′; pairwise by gene) to determine differential robustness across gene ends. Both for (A) +1 and (B) –1 frameshifts, mean differential robustness (computed across all sites and all genes) approaches zero, consistent with decreasing differences in process cost in high-GC genomes (see main text).