RNA virus attenuation by codon pair deoptimisation is an artefact of increases in CpG/UpA dinucleotide frequencies.

Mutating RNA virus genomes to alter codon pair (CP) frequencies and reduce translation efficiency has been advocated as a method to generate safe, attenuated virus vaccines. However, selection for disfavoured CPs leads to unintended increases in CpG and UpA dinucleotide frequencies that also attenuate replication. We designed and phenotypically characterised mutants of the picornavirus, echovirus 7, in which these parameters were independently varied to determine which most influenced virus replication. CpG and UpA dinucleotide frequencies primarily influenced virus replication ability while no fitness differences were observed between mutants with different CP usage where dinucleotide frequencies were kept constant. Contrastingly, translation efficiency was unaffected by either CP usage or dinucleotide frequencies. This mechanistic insight is critical for future rational design of live virus vaccines and their safety evaluation; attenuation is mediated through enhanced innate immune responses to viruses with elevated CpG/UpA dinucleotide frequencies rather the viruses themselves being intrinsically defective.

Introduction

Protein encoding regions of all organisms, eukaryotic, bacterial and viral, are subject to a number of functional constraints in addition to coding capacity, many of which contribute to regulation of translation. These include the widely reported biases in the relative frequencies of codons encoding the same amino acid (Bennetzen and Hall, 1982; Sharp et al., 2005; Wu et al., 2010) which in some organisms represents optimisation of the coding sequence for specific tRNAs, elongation rates and translation accuracy (reviewed in Gingold and Pilpel, 2011). There are, in addition, consistent under- and over-representations of codon pairs (CPs) in all organisms (Yarus and Folley, 1985; Gutman and Hatfield, 1989; Boycheva et al., 2003; Moura et al., 2005; Tats et al., 2008) that have been proposed to influence gene expression through alterations in translation efficiency.

Because of its potential effect on gene expression, altering CP frequencies towards those that are disfavoured in their hosts has recently been advocated as a novel strategy to reduce RNA virus replication (Coleman et al., 2008; Wimmer et al., 2009; Mueller et al., 2010; Martrus et al., 2013; Yang et al., 2013; Le Nouen et al., 2014; Ni et al., 2014). This procedure potentially provides the means to produce a new generation of safer, non-reverting, live attenuated vaccines. Classically, virus genomes have been empirically attenuated by serial passage in tissue-culture leading to the accumulation of mutations. This lengthy, stochastic process produced attenuated virus vaccine strains which have produced major effects on human (eg., poliovirus—Sabin vaccines) and animal health (eg., eradication of Rinderpest using the Plowright vaccine). However, reversion to virulence by back-mutation of characteristically a small number of key, attenuating, mutations is a well-known problem. Novel strategies by which synonymous coding changes are introduced to modify codon usage (Mueller et al., 2010) (Martrus et al., 2013; Yang et al., 2013; Le Nouen et al., 2014; Ni et al., 2014) have the advantage that the resulting virus attenuation is dependent on a large number of mutations each of which only slightly reduce replicative fitness, but taken together produce significant attenuation with greatly enhanced genetic stability.

As one of the first examples, Coleman et al. (2008) generated synthetic poliovirus capsid gene sequences containing codon pairs that were specifically disfavoured in human coding sequences. These CP de-optimised sequences were inserted into an infectious cDNA clone of poliovirus. Virus generated from these mutants showed a remarkably attenuated replication phenotype attributed by the authors to impaired translation efficiency. Codon pair de-optimisation (CPD) has since been developed as a strategy for the production of a wide range of other live attenuated virus vaccines including influenza A virus (IAV), porcine reproductive and respiratory syndrome virus (PRRSV), human immunodeficiency virus type 1 (HIV-1) and respiratory syncytial virus (Mueller et al., 2010; Martrus et al., 2013; Yang et al., 2013; Le Nouen et al., 2014; Ni et al., 2014).

While altering translation efficiency through manipulation of codon or codon pair usage may attenuate virus replication, other virus compositional features may additionally contribute to replication phenotypes. One prominent compositional abnormality among RNA and small DNA viruses infecting mammals and plants is the marked suppression of the frequencies of CpG and UpA dinucleotides (Karlin et al., 1994; Rima and McFerran, 1997; Simmonds et al., 2013). The functional basis for this suppression was recently demonstrated by the marked attenuating effect of artificially increasing the numbers of CpG and UpA dinucleotides in the genome of echovirus 7 (E7; (Atkinson et al., 2014)). We and others (Burns et al., 2009) have speculated that effects of CpG/UpA frequencies on virus replication may indeed account for, at least in part, the attenuating effect of selecting disfavoured codon pairs in CPD mutant of poliovirus and other candidate attenuated vaccines. Supporting this conjecture, regression analysis of the effects of numerous compositional variables in a range of codon- and codon pair deoptimised mutants on poliovirus replication demonstrated the primary effect of CpG and UpA frequencies on replication ability rather than alterations in codon or codon pair usage (Burns et al., 2009).

In the current study, we have used a variety of bioinformatic analyses to investigate the relationship between dinucleotide frequencies and codon pair usage. We have subsequently designed and assessed the replication phenotypes and fitness of mutants of E7 constructed in such a way that allows effects of codon pair and dinucleotide frequency alterations to be separately altered. The findings demonstrate the primary influence of CpG and UpA frequencies on virus replication that was independent of codon pair usage and translation efficiency.

Results

Virus attenuation, CP and dinucleotide frequencies

Coding regions of poliovirus, IAV, PRRSV and HIV-1 have all been subjected to CP de-optimisation and effects on virus replication quantified (Coleman et al., 2008; Mueller et al., 2010; Martrus et al., 2013; Yang et al., 2013; Ni et al., 2014). Despite their diversity of replication and translation mechanisms, each showed a similar relationship between the extent of CPD and reduction in virus replication ability (Table 1). Typically, 10-fold or greater attenuation in cell culture required >12–15% replacement of WT genome with CPD sequences. It is notable, however, that for each virus, CPD invariably increased frequencies of CpG and UpA dinucleotides (Table 1), typically from 0.4–0.6 to 1.4–1.6 (CpG) and from 0.5–0.8 to 1.1–1.4 (UpA) in the mutated regions.

Table 1.

Relationship between codon pair de-optimisation, CpG and UpA frequencies and virus fitness reduction

DOI: http://dx.doi.org/10.7554/eLife.04531.003

			WT			CPD
Virus	Gene	Prop'n	CP bias	CpG	UpA	CP bias	CpG	UpA	Replication Reduction	Ref
Poliovirus
PV-X	Capsid	14.8%	−0.03	0.52	0.75	−0.46	1.34	1.25	×25	Coleman et al., 2008
PV-XY	Capsid	25.9%	−0.03	0.54	0.75	−0.46	1.31	1.27	×400
Influenza A virus*
HAMin	Segs.4	11.4%	0.02	0.43	0.64	−0.42	1.65	1.11	×3.5	Mueller et al., 2010
HA/NPMin	Segs.4,5	21.3%	0.02	0.44	0.55	−0.42	1.56	1.14	×14
PR83F	Segs.1,4,5	29.1%	0.01	0.43	0.53	−0.41	1.55	1.07	×35
HIV-1
A	gag	4.6%	0.03	0.47	1.04	−0.43	1.43	1.25	×7	Martrus et al., 2013
B	gag	4.7%	0.08	0	0.91	−0.37	1.22	1.15	×3
C	gag	4.8%	0.03	0.31	1.00	−0.38	1.50	1.09	× 8
D	gag	2.1%	−0.02	0	0.49	−0.42	1.47	0.99	×1.5
PRRSV
SAVE5	gp5	2.6%†	−0.06	0.63	0.73	−0.38	1.37	1.14	×4‡	Ni et al., 20142

Codon pair minimised sequences of IAV were not provided in (Coleman et al., 2008) and for the purposes of comparison these have been reconstructed in SSE. Note that the CP scores described in Table 1 of that paper (−0.386, −0.420 and −0.421 for PB1, HA and NP respectively) are not minimum scores; these are in fact −0.533, −0.585 and −0.602. Therefore, for the purposes of comparison, CP score minimisation in the current study was targeted to the former values. Although the sequences generated by SSE were not identical to those obtained previously, they would demonstrate a similar distortion of dinucleotide frequencies to those used in the previous study (Coleman et al., 2008).

Mutated region only (positions 147–542 in gp5).

Data from replication assay in PAM cells.

This linkage can be accounted for at least in part by the association between CP choice and the identity of the dinucleotide between the third and first (3–1) codon positions. We analysed these parameters in coding regions of a curated dataset of over 35,170 human mRNA sequences. The representation of each of the 3904 codon pairs found in coding sequences (ie., 61 × 64) was calculated, taking into account both the nucleotide composition of the sequences and the amino acid usage as previously described (Gutman and Hatfield, 1989). Relative under- and over-representation of each was indicated in a heat map, with values ranging from −0.222 to +0.271 (mean 0.072). Values were plotted on x- and y-axes using values that reflected base identities at each of six positions in the codon pair (Figure 1A). Most of the 256 CPs with CpG at the 3–1 position (sixth main column) were markedly under-represented. There was further influence of codon pair position 6 (CP score was more suppressed if A or U) but with generally minimal and inconsistent influences of nucleotide identities at other codon positions (Bennetzen and Hall, 1982; Buchan et al., 2006; Atkinson et al., 2014). In the overall distribution of human codon pair representations, codon pairs containing CpG across the codon boundary distributed towards the negative tail of the distribution of CP score (Figure 1B) and accounted for almost all those with scores below −1.25.

Figure 1.

(A) CP usage in human coding sequences arranged in a 64 x 64 grid.

CP frequencies relative to those expected from nucleotide and amino acid frequencies (CP bias) are colour coded in a heat map. The primary division on the x-axis is by identity of the 3–1 dinucleotide as annotated. Within these, further divisions within each of the 16 columns show the identity of the nucleotide at position 2 (A, C, G or U). The y-axis records nucleotides at positions 4 (4 main divisions on the y-axis), 1 (4 subdivisions of position 4) and 6 (4 subdivisions of position 1). Positions of unused codon pairs containing a 5′ stop codon (translated as *|x) are shaded in grey. CP usage heat maps for A. thaliana, C. elegans and E. coli coding sequences are shown in Figure 1—figure supplement 2A–2C. (B) Distribution of codon pair bias scores in human coding sequences; separate labelling of the 64 codon pairs with CpG (red) or UpA (blue) across the codon junction (3–1) demonstrates their consistent under-representation based on their component nucleotide and amino acid frequencies. The distribution of codon pair scores for A. thaliana, C. elegans and E. coli are shown in Figure 1—figure supplement 3A–3C. Correlations between codon pair scores between human coding sequences and those of A. thaliana, C. elegans and E. coli are shown in Figure 1—figure supplement 4.

DOI: http://dx.doi.org/10.7554/eLife.04531.004

Codons with CpG and UpA at the 1–2 or 2–3 codon position are shaded as indicated in the key (data derived from http://bioinformatics.weizmann.ac.il/databases/codon).

DOI: http://dx.doi.org/10.7554/eLife.04531.005

The primary division on the x-axis is by identity of the 3–1 dinucleotide (labelled on y-axis), divisions within each column show the identity of codon position 2. The y-axis records codon positions 5 (1 cycle), 1 (4 cycles) and 6 (16 cycles). Positions of codon pairs translated as *|x are shaded grey.

DOI: http://dx.doi.org/10.7554/eLife.04531.006

DOI: http://dx.doi.org/10.7554/eLife.04531.007

DOI: http://dx.doi.org/10.7554/eLife.04531.008

Figure 1—figure supplement 2.

CP scores of codon pairs of (A) A. thaliana, (B) C. elegans and (C) E.coli ORFeomes.

The primary division on the x-axis is by identity of the 3–1 dinucleotide (labelled on y-axis), divisions within each column show the identity of codon position 2. The y-axis records codon positions 5 (1 cycle), 1 (4 cycles) and 6 (16 cycles). Positions of codon pairs translated as *|x are shaded grey.

DOI: http://dx.doi.org/10.7554/eLife.04531.006

Figure 1—figure supplement 3.

Distribution of codon pair scores for other organisms-(A) A thaliana, (B) C. elegans and (C) E. coli, with separate representation of codon pairs with CpG and UpA across the codon junction.

DOI: http://dx.doi.org/10.7554/eLife.04531.007

Figure 1—figure supplement 4.

Correlation between representations of human codon pairs (x-axis) with those of other organisms-(A) A thaliana, (B) C. elegans and (C) E. coli (y-axis).

DOI: http://dx.doi.org/10.7554/eLife.04531.008

CPs with UpA at the 3–1 position were also specifically under-represented in human mRNA sequences (Figure 1A,B), consistent with global under-representation of this dinucleotide in coding sequences from eukaryotes (Beutler et al., 1989; Duan and Antezana, 2003). The dataset additionally demonstrated over-representation of CpA and UpG dinucleotides at the 3–1 position; these are typically created by the (methylation-associated) C->T transition upstream of G (fifth and 14th main columns in Figure 1A) and of CpC and CpU (Simmen, 2008). However, with a few exceptions, such as the prominent over-representation of GCG|GCG and CCG|CCG, other codon pairs showed infrequent or minor differences in representation. The avoidance of CpG and UpA in human mRNA sequences at the 3–1 position was further manifested at other three codon position (Bennetzen and Hall, 1982; Atkinson et al., 2014); among the 61 degenerate codons, those containing CpG or UpA at these positions showed lower relative synonymous codon usage than those containing other dinucleotides (Figure 1—figure supplement 1).

Figure 1—figure supplement 1.

Distribution of relative synonymous codon usage values for degenerate codons in the human genome (stop codons were excluded).

Codons with CpG and UpA at the 1–2 or 2–3 codon position are shaded as indicated in the key (data derived from http://bioinformatics.weizmann.ac.il/databases/codon).

DOI: http://dx.doi.org/10.7554/eLife.04531.005

Avoidance of codon pairs with CpG at the 3–1 position was also observed in the plant genome of A. thaliana that also possesses a methylation-dependent suppression of CpG dinucleotides (Figure 1—figure supplements 2A and 3). Codon pair usage of human and plant coding sequences was indeed significantly correlated (R2 = 0.146; Figure 1—figure supplement 4). In contrast to plant coding sequences, no equivalent avoidance of CpG-containing codon pairs was observed in organisms with non-methylated genomes (Caenorhabditis elegans and Escherichia coli; Figure 1—figure supplements 2B,C,3,4).

Separate assessment of effects of CP and dinucleotide frequencies on virus replication

The close association between CP usage and the identity of dinucleotides at codon boundaries immediately complicates any observational assessment of the potentially separate contributions of CP bias and CpG/UpA dinucleotide frequencies on virus replication. On the one hand, it could be hypothesised that the suppression of CpG and UpA at position 3–1 in mammalian codon pairs was a simple consequence of avoiding disfavoured codon pairs. Conversely, it could be conceptualised that codon pair choice is driven in part through avoidance of specific dinucleotides. To resolve this functionally, we compared replication dynamics and relative fitness of native E7 with a series of novel mutants of E7 in which dinucleotide frequencies and CP usage were independently manipulated (Figure 2; Table 2). To achieve this, a mutational program was developed (Sequence Mutate in the SSE package (Simmonds, 2012) that allowed synonymous changes to be introduced into a coding sequence to achieve a pre-specified CP score target while under constraints such as retaining CpG and UpA dinucleotide frequencies and mononucleotide composition.

Figure 2.

Codon pair scores and numbers of CpG and UpA dinucleotides in native (WT) and mutated regions of E7.

Mean CP scores for Regions 1 and 2 combined are shown on the x-axis; the total numbers of CpG and UpA dinucleotides in each sequence are shown on the y-axis. The histogram shows CP scores for the 35,170 human mRNA sequences >200 bases in length (mean 0.072; standard deviation ±0.031).

DOI: http://dx.doi.org/10.7554/eLife.04531.009

Table 2.

Composition and codon usage of E7 wt and mutant insert sequences

DOI: http://dx.doi.org/10.7554/eLife.04531.010

Region	Sequence (Symbol)	G+C content	CpG Total*	O/E ratio†,‡	UpA Total*	O/E ratio†,‡	Codon Usage
							CAI¶	ENc	CP Bias
1	Native (WT)	47.6%	51 (−)	0.730	62 (−)	0.742	0.685	56.5	−0.043
	Permuted (P)	47.6%	51 (0)	0.730	2 (0)	0.742	0.694	55.8	−0.025
	CpG/UpAL (cu)	47.5%	0 (−51)	0	19 (−43)	0.227	0.686	43.5	0.087
	Max-U	50.1%	47 (−4)	0.610	43 (−19)	0.573	0.708	49.6	0.328
	Min_E	47.5%	51 (0)	0.736	62 (0)	0.735	0.748	54.3	−0.131
	Min_U	47.5%	69 (+18)	0.992	76 (+14)	0.939	0.709	58.3	−0.134
	Min_H	49.8%	106 (+55)	1.400	79 (+17)	0.981	0.696	49.2	−0.130
2	Native (WT)	47.1%	18 (−)	0.320	48 (−)	0.695	0.743	53.2	0.015
	Permuted (P)	47.6%	18 (0)	0.320	48 (0)	0.695	0.739	49.0	0.013
	CpG/UpAL (cu)	48.5%	0 (−18)	0	48 (0)	0.214	0.739	47.2	0.118
	Max-U	46.3%	24 (+6)	0.440	43 (−3)	0.601	0.750	46.1	0.311
	Min-E	45.7%	18 (0)	0.343	48 (0)	0.657	0.785	53.3	−0.091
	Min-U	47.4%	37 (+19)	0.649	50 (+2)	0.738	0.767	57.6	−0.083
	Min-H	47.8%	68 (+50)	1.172	65 (+15)	0.970	0.715	49.7	−0.085

Total number of CpG and UpA dinucleotides in sequence. Changes in numbers between mutated and original WT sequence are indicated in parentheses.

Ratio of observed dinucleotide frequency (O) to that expected based on mononucleotide composition (E) that is, f(CpG)/f(C) × f(G).

Values deliberately changed are shown in red (maximised) and blue (minimised).

Calculated from http://genomes.urv.es/CAIcal/ (Puigbo et al., 2008).

The mutant, Min-E was constructed from two genome regions, together comprising 31% of the E7 genome, in which the coding sequence possessed the minimum possible CP score (−0.111) while retaining identical CpG and UpA frequencies as WT virus (CP score: −0.014; CpG: 0.525; UpA: 0.718; Figure 2, Table 2). Inserts with the same CP frequencies as Min-E but without dinucleotide frequency constraints (Min-U; CpG: 0.82; UpA: 0.95) or where CpG and UpA frequencies were maximised (Min-H; CpG: 1.3; UpA: 0.98) were generated similarly. The three mutants provided the opportunity to investigate effects of dinucleotide frequency differences of viral fitness without the compounding effect of CP bias. It was similarly possible to compare fitness of the mutant, Max-U, with a maximised CP score (0.320) but with similar CpG and UpA frequencies to WT with the previously described mutant, cu|cu, with minimised CpG and UpA frequencies (0 and 0.22 respectively) but a CP score marginally greater that WT (0.11; Figure 2). P|P was the permuted mutant control with randomised codon order but identical coding and dinucleotide frequencies to WT sequence.

If CP usage solely determined virus replication ability, the seven mutants would be expected to display the following fitness ranking:

Max-U > cu|cu > (WT = P|P) > (Min-H = Min-U = Min-E).

Conversely, if virus fitness was determined by CpG and UpA dinucleotide frequencies, the following ranking would be expected:

cu|cu > Max-U > (WT = P|P = Min-E) > Min-U > Min-H.

These predictions were determined by generation and infectivity measurement of virus stocks corresponding to the seven mutants and comparing their relative fitness in competition and replication assays (Figures 3,4).

Figure 4—figure supplement 1.

Competition assays between E7 mutants showing competing variants (lanes 1 and 2) andout at indicated passage number (lane 3) for each.

DOI: http://dx.doi.org/10.7554/eLife.04531.013

Replication phenotypes

Full length RNA transcribed from each E7 mutant cDNA constructs all generated infectious virus after transfection into RD cells. Stocks of virus were generated from WT and each mutant and infectivity quantified by quantal limiting dilution. To investigate replication kinetics, RD cells were infected with WT and each mutant at an MOI of 0.03 in triplicate and infectivity of supernatants measured at 8, 18 and 30 hr (Figure 3;). During the exponential period of replication (8 and 8 hr), Min-U and Min-H mutants showed 1 and >2 log reductions in virus replication respectively compared to WT E7. Contrastingly, the CpG/UpA-minimised mutant, cu|cu replicated to approximately 1 log higher levels that WT. Significantly for the analysis of effects of CP and dinucleotide frequencies on replication, virus titres obtained from mutant with identical (Min-E, CDLR) or similar (Max) CpG/UpA frequencies to WT were highly similar at both time points. At the last timepoint (30 hr), RD cells infected with WT, CDLR, Min-E, Max and cu|cu were entirely destroyed or almost entirely destroyed (Max-U) and showed similar residual infectivities, while those infected with Min-H showed an incomplete cytopthic effect.

Figure 3.

Replication of WT and mutants of E7 with altered CP and dinucleotide frequencies.

Bars are shaded diagrammatically based on their relative CpG/UpA composition. RD cells were infected with E7 WT, at an MOI of 0.03 and infectious titres quantified at 8, 18 and 30 hr time points post inoculation (p.i.) by TCID50 determination. Results are the mean of three biological replicates; error bars show standard errors of the mean.

DOI: http://dx.doi.org/10.7554/eLife.04531.011

Competition assays were used as a more stringent measure of fitness differences in mutants with different codon pair biases. Equal MOIs of WT and mutants were co-inoculated onto RD cells and serially passaged up to ten times. Population compositions were determined by amplification of sequences across modified regions and cleavage with restriction enzymes that differentiated WT mutant sequences from each other (Figure 4A,B; Table 3). In the examples of competition assays (Figure 4A), Max-U showed similar fitness to WT but a greater population representation at passage 10. cu|cu completely out-competed Max-U by passage 10 while in the final example, WT and Min-E showed equal fitness at passage 5 and at passage 10 (Figure 4B). A total of 12 pairwise comparisons were made and outcomes in terms of population representation recorded at passage 10 (Figure 4B; see Key). The results are internally consistent and with their replication kinetics (Figure 3) and indicate the following fitness ranking:

Figure 4.

RD cells were co-infected with pairs of WT (W|W) and E7 mutants at equal MOI and the supernatant serially passaged through cells after development of CPE. RNA was isolated and the composition of each virus determined through selective restriction digests using enzymes listed in Table 3.

(A) Examples of three competition assays showing cleavage patterns of individual viruses (lanes 1, 2), the starting inoculum (lane 3) and two biological replicates after 10 (panels 1, 2) or 5 (panel 3) passages in lanes 4 and 5. Results from the other competition assays are shown in Figure 4—figure supplement 1. (B) Summary of pairwise fitness comparisons of viruses with outcomes for the viruses listed in columns at passages 5 and 10 indicated by colour shading. For example, Min-E and WT showed equal fitness (yellow shading) and cu|cu outcompeted WT by passage 5 (red) and Max-U by passage 10.

DOI: http://dx.doi.org/10.7554/eLife.04531.012

DOI: http://dx.doi.org/10.7554/eLife.04531.013

Table 3.

Enzymes used in selective digests for competition ASSAYS

DOI: http://dx.doi.org/10.7554/eLife.04531.014

Virus 1	Virus 2	Region	Enzyme	Target
W|W	P|P	1	SpeI	Permuted
W|W	Max-U	1	SacI	Max
W|W	Min-E	1	NcoI	WT
W|W	Min-U	1	NcoI	WT
W|W	Min-H	1	EcoRV	WT
W|W	cu|cu	1	EcoRV	WT
P|P	cu|cu	1	SpeI	Permuted
Max-U	P|P	1	SpeI	Permuted
Max-U	cu|cu	1	SacI	Max
Min-E	Min-U	1	ClaI	Min-U
Min-E	Min-H	1	EcoRV	Min-E
Min-U	Min-H	1	ClaI	Min-U

Using the Spearman rank correlation test, fitness ranking was significantly associated with CpG and UpA frequencies in the insert region (p < 0.001) but showed no association with CP frequencies and other measures of codon usage that potentially influence translation rates, codon adaptation index (CAI) and effective number of codons (ENc) (Table 4). Consistently, these results demonstrate that when altered independently from CP bias, only dinucleotide frequencies were associated with replication fitness.

Table 4.

Correlation between fitness ranking and sequence composition

DOI: http://dx.doi.org/10.7554/eLife.04531.015

Variable	Spearman R	p value†
CpG/UpA*	1.0	<0.001
CP bias	−0.70	0.1 (n.s.‡)
CAI	−0.334	>0.5 (n.s.)
ENc	0.593	>0.5 (n.s.)
G + C content	0.075	>0.5 (n.s.)
Translation efficiency	−0.074	>0.5 (n.s.)

Number of CpG and UpA dinucleotides in insert region.

From values tabulated in (Ramsey, 1989).

n.s. : not significant.

Comparison of translation efficiencies

The maxim that any effects of CP frequencies on replication are mediated through its influence on translation efficiency was investigated for the mutants constructed in the study. Translation assays were evaluated in vitro to avoid effects mediated through stress response-related RNA recognition mechanisms that restrict E7 translation and subsequent replication immediately after entry (Atkinson et al., 2014). Viral RNA transcripts from E7 WT and mutant cDNA clones were used to program rabbit reticulocyte lysates in the presence of [35S]-methionine. Electrophoresis of reactions after 3 hr showed translation of several bands representing cleaved and partially cleaved E7 proteins (Figure 5; Figure 5—figure supplement 1). Translation efficiencies of each of the mutant E7 transcripts were comparable to WT RNA; what variability there was between mutants (Figure 5—figure supplement 1) did not correlate with replication fitness (R = −0.075; p > 0.5; Table 4). This indicates that, at least in a whole genome context, alteration of either CP or dinucleotide frequencies had no significant effect on viral polyprotein translation and therefore cannot be attributed to the marked differences in replication phenotypes observed.

Figure 5.

Translation of RNA templates generated from WT and mutant E7 cDNAs in a rabbit reticulocyte cell free assay.

Assignments of bands to E7 proteins were based on molecular weights on SDS-PAGE. A comparison of densitometry values for viral proteins is shown in Figure 5—figure supplement 1.

DOI: http://dx.doi.org/10.7554/eLife.04531.016

Translation efficiencies of mutant E7 cDNAs were quantified relative to expression from the WT template. Bars show mean values for seven viral proteins; error bars show standard errors of the mean.

DOI: http://dx.doi.org/10.7554/eLife.04531.017

Figure 5—figure supplement 1.

Translation efficiencies estimated by densitometry of band intensities of viral proteins translated in a rabbit reticulocyte cell free assay.

Translation efficiencies of mutant E7 cDNAs were quantified relative to expression from the WT template. Bars show mean values for seven viral proteins; error bars show standard errors of the mean.

DOI: http://dx.doi.org/10.7554/eLife.04531.017

Discussion

This study sought to disentangle effects of codon pair usage and nucleotide frequencies in a re-examination of their effects on the replication of an RNA virus, E7. In the literature, studies have documented effects of CP de-optimisation on virus attenuation without reference to effects of this procedure on dinucleotide frequencies (Coleman et al., 2008; Mueller et al., 2010; Martrus et al., 2013; Yang et al., 2013; Le Nouen et al., 2014; Ni et al., 2014). While frequencies of both of these dinucleotides are suppressed in most classes of mammalian RNA viruses (Rima and McFerran, 1997), all sequences modified to select disfavoured CPs (Coleman et al., 2008; Mueller et al., 2010; Martrus et al., 2013; Yang et al., 2013; Ni et al., 2014) consistently elevated frequencies of CpG and UpA dinucleotides to levels to 2.5–threefold higher levels than the original native sequences (Table 1). As documented in other studies (Burns et al., 2009; Atkinson et al., 2014), these dinucleotide frequencies may contribute additionally to the observed attenuation of virus replication.

Through construction of mutants of E7 in which CP frequencies was altered while keeping dinucleotide frequencies constant (WT/Min-E) and conversely, generating viruses with the same or similar CP biases but different dinucleotide frequencies (eg., Min-E/Min-U/Min-U and cu|cu/WT/Max-U), we were able to separate potential influences of these compositional variables on replication phenotype and fitness. The fitness ranking derived from competition assays (cu|cu > Max > (WT = P|P = Min-E) > Min-U > Min-H) demonstrated that it was dinucleotide frequencies that significantly influenced fitness while differences in CP usage showed no detectable phenotypic effect. Moreover, if CPD were to influence virus replication then its effect would be manifested through changes in translation rate; however, no measurable differences in translation efficiency were detected between WT and CP-optimised (Max) or de-optimised (Min) template RNA sequences. These findings are broadly consistent with results of previous translation assays of poliovirus mutants differing in CP choice in which relatively small differences in translation of PV-Min, WT and PV-Max mutants were clearly incompatible with the marked differences in their replicative ability (Coleman et al., 2008).

In a broader context, the finding that alterations in codon pair frequencies has no independent effect on virus replication are consistent with current understanding of the nature and driving forces behind codon pair usage in other organisms. Most importantly, there is no evidence that disfavoured CPs in eukaryotes, archaea or prokaryotes are those that specifically retard translation rates. Indeed, where specifically investigated, the opposite was observed. mRNA templates containing disfavoured codon pairs in E. coli were translated faster than those containing over-represented CPs (Irwin et al., 1995). The 16 codon pairs identified as most retarding translation of the E. coli his operon leader peptide gene has codon pair scores ranging from −0.94 to +0.54 and distributed around the centre of the distribution of codon pair scores (Figure 1—Figure Supplement 3) (Chevance et al., 2014). The current consensus view is that CP usage in prokaryotes is governed functionally as a means to regulate gene expression rather to maximise translation (Folley and Yarus, 1989; Irwin et al., 1995; Boycheva et al., 2003; Buchan et al., 2006).

In eukaryotic genomes, other factors underlie codon pair representation since coding regions in mRNA sequences and in non-transcribed genomic DNA showed similar biases in codon pair frequencies (Moura et al., 2007). CP frequencies must therefore be substantially determined by mutational events operating on DNA such as methylation and specific context-dependent errors during genome replication instead of any kind of optimisation or regulation of translation. The consistent under-representation of codon pairs with CpG in the 3–1 position (Figure 1; Buchan et al., 2006) in mammalian genomes indeed likely originates from DNA methylation-induced mutations in the nucleus. Our data showing similar rates of translation of Min-H and Max-U that show major differences in frequencies of CpG-containing CPs (Figure 5) are consistent with this interpretation.

Finally, there is no theoretical basis for the assumption that CPs are disfavoured because of their negative effects on translation efficiency and this concept runs counter to our growing understanding of the intricate mechanisms that govern gene expression. In all organisms, coding sequences differ in codon usage, match to tRNA abundances, mRNA stability and initiation sites to regulate rates and fidelity of protein expression (reviewed in Gingold and Pilpel, 2011). Some of the variability in CP usage observed in the three domains of life (Moura et al., 2005; Sharp et al., 2005; Tats et al., 2008; Wang and Li, 2009) likely represents aspects of that control, rather than as a means to simply maximise translation.

Mechanism of attenuation

Understanding what limits the replication of viruses with altered CP and dinucleotide frequencies is critical in the evaluation of their broader safety as attenuated virus vaccines. The proposed mechanism in which alterations in CP bias alter translation efficiency and it is this that inhibits virus replication introduces a conceptual model in which it is the virus that is intrinsically defective. With the large number of mutations required for reversion, such viruses should be stably attenuated in whatever context they are used. However, as we have now shown, the replication defect of CPD viruses is actually mediated through alterations in dinucleotide frequencies in the genome that influence their recognition by the cell. In this alternative paradigm, viruses with elevated frequencies of CpG and UpA are not intrinsically defective but they are more readily recognised by the cell and prevented from initiating replication. Their attenuation is therefore dependent on the efficacy of the host innate immune response.

The cellular mechanisms responsible for differential recognition and response to RNA sequences with different dinucleotide composition are currently unknown. In our previous study, we obtained evidence that replication inhibition of high CpG/UpA mutants of E7 occurred shortly after cell entry and was not mediated though conventional pattern recognition receptors (Atkinson et al., 2014). In that study, we additionally demonstrated that it was additionally not the result of differences between high and low CpG/UpA viruses in their sensitivity to the cellular interferon response. We did observe, however, that the attenuated phenotype of high mutants could be entirely reversed by the kinase inhibitor, C16, a finding that suggests that recognition may occur through an as yet uncharacterised PKR-related component of the stress response pathway in the cell.

Both the adaptive and innate arms of the human immune system are highly polymorphic with remarkable variability in function and expression of many key components of recognition or effector proteins mediating antiviral responses (Thomas et al., 2009; Everitt et al., 2012; Hambleton et al., 2013; Pothlichet and Quintana-Murci, 2013). Although uncharacterised mechanistically, there is clearly a potential danger that pathways that restrict the replication of high CpG/UpA RNA viruses may be similarly variable in the efficacy in humans and in veterinary species with different genetic backgrounds. The attenuation of live vaccines and safety margins established for their large scale use may be similarly variable; investigation of population differences in innate cellular responses to viruses of different dinucleotide compositions is essential in the evaluation of the safety of this new generation of high CpG/UpA live attenuated vaccines.

Materials and methods

Cell culture and cell lines

RNA transcripts of the pT7:E7 infectious cDNA clone of the isolate Wallace (accession number AF465516) were used to generate E7 viral stocks. E7 was propagated in rhabdomyosarcoma (RD) cells using Dulbecco modified Eagle medium (DMEM) with 10% foetal calf serum (FCS), penicillin (100 U/ml) and streptomycin (100 µg/ml). All cells were maintained at 37°C with 5% CO2.

In silico design of CpG and UpA modified viruses

The two regions in the E7 genome used previously to investigate effects of dinucleotide frequencies on virus replication (Atkinson et al., 2014) were used in the current study (Region 1: 1878–3119 and 5403–6462). Previously characterised mutants comprised the CpG/UpA-low mutant cu|cu with all CpG dinucleotides and as many UpA dinucleotides possible eliminated and the permuted mutant P|P in which codon order was permuted while retaining protein coding and native dinucleotide frequencies. Further mutants (Max-U, Min-E, Min-U, Min-H) are described in the main text; sequences listed in Supplementary file 1.

Bioinformatics analysis

Manipulation of dinucleotide frequencies and codon pair scores in coding sequences was performed using the program Sequence Mutate in version 1.2 of the SSE package (Simmonds, 2012). Reference datasets of human, A. thaliana, C. elegans and E. coli messenger RNA sequences were obtained from the Refseq database (http://www.ncbi.nlm.nih.gov/nuccore). Codon pair usage tables were generated from coding regions of each mRNA sequence datasets from the four organisms using the program Composition Scan in the SSE package (Simmonds, 2012). Codon pair tables generated by SSE were used to calculate CP frequencies for all mRNA sequences with coding regions >200 bases in length from each organism. These comprised 35,770 human mRNA sequences, 32,768 from A. thaliana, 24,093 from C. elegans and 4316 from E. coli.

The codon pair table generated by SSE from our dataset of human mRNA sequences was used in preference to that previously described (Table S2 in reference Coleman et al., 2008) because of the larger number of human mRNA sequences now available. The previously published dataset as presented additionally unaccountably omitted a large number of codon pairs (3586 were listed instead of the 3904 expected—61 × 64). There was however a good correlation between CP frequencies between the two datasets (Figure 6).

Figure 6.

Comparison of codon pair scores generated by SSE using a dataset of 35,770 human mRNA sequences (y-axis) with those used in a previous analysis (Coleman et al., 2008).

DOI: http://dx.doi.org/10.7554/eLife.04531.018

The codon adaptation index for human codon usage was calculated through the website http://genomes.urv.es/CAIcal/ (Puigbo et al., 2008). The effective number of codons (Enc; Wright, 1990) and CP usage (Buchan et al., 2006; Coleman et al., 2008) were calculated using the program Composition Scan in SSE.

Clone construction and recovery of mutant viruses

Mutant E7 constructs with altered CP frequencies were generated from custom synthesised DNA sequences (Eurofins Genomics, Ebersberg, Germany). Mutant clones were constructed as previously described (Atkinson et al., 2014). All clones were sequenced over the insert regions prior to further applications. Infectious virus from each cDNA clone was recovered by transfection of RNA transcripts produced from plasmids linearised using NotI using a Riboprobe System-T7 in vitro transcription kit (Promega Ltd. Southampton, UK). 100 ng of RNA was transfected into RD cells using Lipofectamine 2000 (Invitrogen, Life Technologies Ltd., Paisley, UK) according to the manufacturer's instructions. The resulting cell lysates were used to generate passage 1 stocks by re-infecting RD cells. Viral titres were determined by TCID50 titration in RD cells.

Replication phenotype

Multi-step growth curves for each virus were generated by infecting RD cells in triplicate in 24-well plates at an MOI of 0.03 as previously described (Atkinson et al., 2014). Supernatant collected at time points (8, 18 and 30 hr post-infection) were assayed for infectivity by quantal dilution. Competition assays were performed as previously described. Briefly, equal titres of virus pairs (combined MOI = 0.01) were applied simultaneously to RD cells in 25 cm2 bottles. Following the development of CPE, supernatant was collected and 300 µl applied to fresh RD cells. This was continued for up to 10 passages. The results of the competition assays were determined by restriction enzyme digestion of the amplicon amplified from Region 1 by combined reverse transcription—PCR (Atkinson et al., 2014). Restriction enzymes used to differentiate each mutant pair are listed in Table 3.

In vitro transcription and translation

RNAs were produced by in vitro T7 transcription (Riboprobe System T7, Promega) of the various cDNA plasmids, each linearised with NotI (Promega). Transcript RNAs were used to program nuclease-treated rabbit reticulocyte lysates (Promega) supplemented with HeLa cell S10 cytoplasmic extracts (Dundee Cell Products, Dundee, UK). Reactions were set-up as follows; 7 μl rabbit reticulocyte lysate, transcript RNA (0.25–2 μg), 0.5 μl 1 mM amino acid mix (minus methionine), 0.5 μl [35S]-methionine (1200 Ci/mmol), 10 U RNasin Ribonuclease Inhibitor and 2.25 μl HeLa cell extract in a total volume of 12.5 μl. Reactions were incubated at 30°C for 3 hr and analysed by SDS-PAGE (4–20% Tris-Glycine, Expedeon Ltd. Cambridge, UK). Gels were exposed to film (Thermo Scientific, Basingstoke, UK) for 1–4 days at −70°C. To determine the relative density of the protein bands, densitometry was carried out on the scanned gel image using ImageJ 1.48 software (http://imagej.nih.gov/ij).

eLife posts the editorial decision letter and author response on a selection of the published articles (subject to the approval of the authors). An edited version of the letter sent to the authors after peer review is shown, indicating the substantive concerns or comments; minor concerns are not usually shown. Reviewers have the opportunity to discuss the decision before the letter is sent (see review process). Similarly, the author response typically shows only responses to the major concerns raised by the reviewers.

Thank you for sending your work entitled “RNA virus attenuation by codon pair deoptimisation is an artefact of increases in CpG/UpA dinucleotide frequencies” for consideration at eLife. Your article has been favorably evaluated by Richard Losick (Senior editor), a Reviewing editor, and 3 reviewers, one of whom, Raul Andino, has agreed to reveal his identity.

The Reviewing editor and the reviewers discussed their comments before we reached this decision, and the Reviewing editor has assembled the following comments to help you prepare a revised submission. The comments in full from the three reviewers are attached below.

Our reviewers were in close agreement that the paper makes important points and is topical and appropriate for eLife. There was also a consensus that there were some failings in reviewing the history of this topic, and I agree that an improved review is needed (apparently key citations are missing). This should be easily fixed. There were some good suggestions to improve clarity. There were also requests for more experimentation from one (the third) reviewer, which I think sound reasonable, but I cannot evaluate how much work is involved. I would encourage the authors to consider doing what is easy. It seems to this reviewer that the quantitation might be improved without huge effort. We are willing to leave it to the authors to decide how much more wet work is desired.

All told, we are supportive of publication.

Reviewer #1

This manuscript presents a careful and interesting work, and even more thoroughly documents on the attenuating effect of dinucleotides CpG and UpA in the genomes of RNA viruses than previous papers have. They argue, convincingly, that attenuating effects previously ascribed to 'codon pair bias' in attenuating viruses can be completely explained by the resulting increases in CpG and UpA dinucleotides. They do this via a systematic analysis of the attenuating effects of altering 'codon-pair bias' in previous papers and by carefully design experiments of their own, which include documentation that the attenuation does not occur by reducing translational efficiency, but by reducing the infectivity of the resulting virions.

With respect to the possible attenuation mechanism of CpG and UpA dinucleotides, the authors discuss the possibility that it is a mechanism of the innate immune response that is not PKR- or pattern receptor-dependent. Data relevant to this are published separately (Atkinson et al., 2014).

The curious thing about this field, and this manuscript, is that the 'straw man' of codon-pair bias was already destroyed by Burns et al., in 2009. These authors were pursuing their own hypothesis, that the substitution of rare codons would attenuate viruses. They found that this was the case, but careful mutagenesis revealed that it was, in fact, the increase in CpG and UpA nucleotides in the RNA that accounted for the attenuation. They showed that this effect was not due to translational efficiency and was manifested as the increased particle-to-PFU ratio of the resulting viruses. They reflected that bias against CpG and UpA codons might be the reason, in fact, that codons that contain them, and codon pairs that create them, are rare to begin with.

They noted, quite politely, that this also accounted for the effect of the highly publicized “codon-pair bias” published by Coleman et al., 2008. They stated: “A prominent feature of the most disfavored codon pairs is the presence of CpG or UpA across codons. Thus, the observed CPB in poliovirus and in humans and higher eukaryotes may be driven primarily by CpG and UpA dinucleotide suppression. In this context, it is notable that in cassette C of the construct with the lowest fitness, ABc12, within-codon CpG and UpA frequencies were maximized but the CPB score was similar to those of higher-fitness constructs, including ABC.” Thus, like the current manuscript, they re-analyzed the data of Coleman et al., and found it to be completely explicable in terms of CpG and UpA dinucleotides.

This conclusion, and the published data, were clearly insufficient to prevent the design and interpretation of several more manuscripts using 'codon pair bias' to design attenuated viruses for vaccine and basic science purposes (Mueller et al., 2010; Martrus, 2013; Yang, 2013; Ni, 2014).

In short, this is a well-executed, important study that extends the very clear conclusions of the also excellent, but completely ignored, 2009 paper of Burns et al. Therefore its publication would bring an important discussion into the limelight. Here are my suggestions, mostly having to do with writing and scholarship:

1) That CpG and UpA dinucleotide frequency correlates with attenuation of viruses, rather than 'codon pair bias' or 'codon de-optimization', is a feature of these approaches rather than an artifact, in this reviewer's opinion. The combative nature of the title is not necessary to bring this fact to light.

2) It does not detract from the quality of the present work to acknowledge more explicitly the contributions of Burns et al. It may, however, detract from its novelty.

3) The readers understanding of the data is compromised by the poor description of the experiments and analysis in the figure legends. These should be rewritten in such a way that the figures can be independently understood. For example, nowhere in the manuscript can the metric for “CPB score” be discovered.

4) Similarly, the quantitative findings in Figure 4 are difficult to dissect. With so many changes in the genomes being analyzed, qRT-PCR of viral competitions would be helpful. As it is, the actual data from the competition experiments should be presented, rather than a cartoon depicting which viruses 'won' and 'lost'.

5) The last paragraph of the Discussion uses 'data not shown' to make its point. These data should be shown or the sentences deleted.

Reviewer #2

There has been interesting discussions on why the nucleotide composition of many human viruses present distinct nucleotide, di-nucleotide and codon preferences. In the last years synthetic viruses have been generated with alternative codon distributions that show attenuated replication. However these constructs change other variables (nucleotide or dinucleotide composition) in addition to the codon distribution and it is unclear what of those changes affects replication. The authors address this issue by generating mutants of the echovirus 7 in which CpG and UpA dinucleotides were varied independent of the codon distribution. The results clearly showed that the main factor dominating replication was dinucleotide content, and that translation efficiency was unaffected by the two variables. The authors then argue that attenuation is mediated through innate immune response to viruses with high CpG/UpA content.

The paper is very interesting and results are compelling. The main critiques are:

1) The topic of dinucleotide biases in RNA viruses has been explored extensively and most references are absent.

2) The discussion on the mechanisms of attenuation and the association to the innate immune response is not well elaborated. They mention the lack of association to interferon response and the effect of C16. However, the results are not presented in detail. I suggest the authors to strengthen their conclusions.

Reviewer #3

This study examines the role of nucleotide composition in RNA virus genomes. The authors introduced a number of synonymous mutations into the EV71 genome to modify the frequency of CpG and UpA dinucleotide (DN) or codon pair (CP). Their results support the idea that DN frequencies determine virus replication, while neither DN nor CP affected translation efficiency. They concluded that DN affect virus replication because the nucleotide composition of the virus genome influences the host-cell innate response to the virus.

The most interesting contributions of this study are: 1) they are able to identify mutations that will affect CP without affecting CpG/UpA dinucleotide composition and vice versa; (Bennetzen and Hall, 1982) based on this bioinformatics information they constructed mutants that change CP or DN composition and experimentally evaluated the effect of synonymous mutation on virus replication. Their conclusions are supported by the available results, and I think this study represents an important contribution to the field because it seems to address a molecular mechanism for virus genome nucleotide composition bias. However, I believe that a more quantitative analysis of the competition experiment may be required to determine the degree of correlation between DN composition and virus replication fitness.

The authors choose competition assays to precisely analyze fitness between different virus and with respect to WT. This is the correct experiment in my opinion, however I believe the analysis of the experiment is somewhat casual and not very quantitative, and therefore limits the value of this data. Figure 4 presents the results using a differential restriction enzyme pattern to distinguish between the two competing viruses, but they can only determine when one virus is lost (no longer detected by the assay). I think that it will be a lot more powerful to use digital PCR to precisely quantify the ratio between virus genomes in the given competition assay. This will provide parameters that can then be fit in a simple mathematical model to determine with more accuracy the correlation between DN or CP and fitness, which at this point seems a bit circumstantial.

Similarly, it would be desirable to improve the quality of the in vitro translation assay and quantify protein production to determine that, in this case, there is little correlation between translation efficiency and fitness, as this is one of the central claims of the study.

Reviewer #1

1) That CpG and UpA dinucleotide frequency correlates with attenuation of viruses, rather than 'codon pair bias' or 'codon de-optimization', is a feature of these approaches rather than an artifact, in this reviewer's opinion. The combative nature of the title is not necessary to bring this fact to light.

As co-authors, we did spend some time discussing the title of the paper. The final choice was in fact motivated for reasons alluded to by the reviewer: that those involved in the codon pair programme have simply ignored the published evidence, by Burns et al. and more recently from our lab (Atkinson et al.), that the attenuating effect was mediated through inadvertently increasing CpG ad UpA dinucleotide frequencies rather any effect on translation. A clear, declarative title of the paper seems required to counter mistaken views on this. If we changed the title as the reviewer suggests, it would imply that codon pair de-optimisation has been used deliberately as a way to increase CpG and UpA dinucleotide frequencies. This is absolutely not the case.

That said, we would have liked to qualify that statement by stating why we make this assertion, but the low character limit imposed on eLife papers titles prevents us from doing this.

2) It does not detract from the quality of the present work to acknowledge more explicitly the contributions of Burns et al. It may, however, detract from its novelty.

We have cited that study in the original manuscript. In the Introduction of the revised manuscript, we have described the observation made by Burns and colleagues in more detail as requested by the reviewer.

3) The readers understanding of the data is compromised by the poor description of the experiments and analysis in the figure legends. These should be rewritten in such a way that the figures can be independently understood. For example, nowhere in the manuscript can the metric for “CPB score” be discovered.

We apologise for these omissions and have endeavoured to make the figure legends clearer (more explanatory) and describe the various abbreviations more fully. We have dropped the abbreviations CPB as it can be expressed more clearly by other wordings (e.g. biased codon pair usage).

4) Similarly, the quantitative findings in are difficult to dissect. With so many changes in the genomes being analyzed, qRT-PCR of viral competitions would be helpful. As it is, the actual data from the competition experiments should be presented, rather than a cartoon depicting which viruses 'won' and 'lost'.

We fully agree that the results presentation in Figure 4B was unduly diagrammatic and lacked primary data (with the exception of the two example gel images in Figure 4A). This point was also made by Reviewer #3. To address this, we have made use of the possibility to include figure supplements by now including gel images of the other competition assays so that relative fitness can be directly evaluated. We have, however, retained the original Figure 4B as a summary of the experimental data.

The use of restriction enzymes to differentiate E7 mutants is a widely used method and its results can be made fully quantitative using appropriately calibrated controls. Had the phenotypes been more subtle we agree that quantitative qRT-PCR would likely have been necessary to discriminate between the fitness of viruses with modified dinucleotide ratios. However, in the current study, outcomes were either elimination of one or other of the competing viruses or a draw. These results are readily visualised as presented, and we believe the investigation does not need more precise quantitation as suggested by the reviewer.

5) The last paragraph of the discussion uses 'data not shown' to make its point. These data should be shown or the sentences deleted.

We agree, and that part of the discussion has entirely been removed. We have also removed reference to unpublished investigations of Theiler’s virus and influenza A virus from the Introduction.

Reviewer #2

1) The topic of dinucleotide biases in RNA viruses has been explored extensively and most references are absent.

We have now cited, in the Introduction, the Rima and Karlin studies that originally noted the suppression of CpG and UpA dinucleotide frequencies in RNA viruses.

2) The discussion on the mechanisms of attenuation and the association to the innate immune response is not well elaborated. They mention the lack of association to interferon response and the effect of C16. However, the results are not presented in detail. I suggest the authors to strengthen their conclusions.

That part of the Discussion was based on the results presented in the Atkinson et al. paper from earlier in the year. We have modified that paragraph to clarify that we were referring to this previous study.

Reviewer #3

This study examines the role of nucleotide composition in RNA virus genomes. The authors introduced a number of synonymous mutations into the EV71 genome to modify the frequency of CpG and UpA dinucleotide (DN) or codon pair (CP). Their results support the idea that DN frequencies determine virus replication, while neither DN nor CP affected translation efficiency. They concluded that DN affect virus replication because the nucleotide composition of the virus genome influences the host-cell innate response to the virus.

The most interesting contributions of this study are: 1) they are able to identify mutations that will affect CP without affecting CpG/UpA dinucleotide composition and vice versa; () based on this bioinformatics information they constructed mutants that change CP or DN composition and experimentally evaluated the effect of synonymous mutation on virus replication. Their conclusions are supported by the available results, and I think this study represents an important contribution to the field because it seems to address a molecular mechanism for virus genome nucleotide composition bias. However, I believe that a more quantitative analysis of the competition experiment may be required to determine the degree of correlation between DN composition and virus replication fitness.

The authors choose competition assays to precisely analyze fitness between different virus and with respect to WT. This is the correct experiment in my opinion, however I believe the analysis of the experiment is somewhat casual and not very quantitative, and therefore limits the value of this data. presents the results using a differential restriction enzyme pattern to distinguish between the two competing viruses, but they can only determine when one virus is lost (no longer detected by the assay). I think that it will be a lot more powerful to use digital PCR to precisely quantify the ratio between virus genomes in the given competition assay. This will provide parameters that can then be fit in a simple mathematical model to determine with more accuracy the correlation between DN or CP and fitness, which at this point seems a bit circumstantial.

This comment is related to that of Reviewer #1 and has been addressed above (the limited number of observed outcomes of competition assays can be effectively demonstrated through restriction enzyme analysis).

To address the issue of relative fitness further, we have repeated the experiment to determine the replication kinetics of WT and mutants of E7 that was depicted in Figure 3. The re-formatting of data as histograms with error bars now allows replication rates from all mutants at different time points to be shown and these entirely back up the competition assay results.

Similarly, it would be desirable to improve the quality of the in vitro translation assay and quantify protein production to determine that, in this case, there is little correlation between translation efficiency and fitness, as this is one of the central claims of the study.

We did indeed quantify translation of the individual proteins on the blot by densitometry and presented the results in the form of a histogram in Figure 5–figure supplement 1. We additionally used a translation efficiency metric based on this quantitation to analyse potential associations with virus replication rate (Table 4) along with other variables (dinucleotide composition, CB bias, CAI, ENc and G + C content).