Genetic inactivation of poliovirus infectivity by increasing the frequencies of CpG and UpA dinucleotides within and across synonymous capsid region codons

Abstract

Replicative fitness of poliovirus can be modulated systematically by replacement of preferred capsid region codons with synonymous unpreferred codons. To determine the key genetic contributors to fitness reduction, we introduced different sets of synonymous codons into the capsid coding region of an infectious clone derived from the type 2 prototype strain MEF-1. Replicative fitness in HeLa cells, measured by plaque areas and virus yields in single-step growth experiments, decreased sharply with increased frequencies of the dinucleotides CpG (suppressed in higher eukaryotes and most RNA viruses) and UpA (suppressed nearly universally). Replacement of MEF-1 capsid codons with the corresponding codons from another type 2 prototype strain (Lansing), a randomization of MEF-1 synonymous codons, increased the %G+C without increasing CpG, and reductions in the effective number of codons used had much smaller individual effects on fitness. Poliovirus fitness was reduced to the threshold of viability when CpG and UpA dinucleotides were saturated within and across synonymous codons of a capsid region interval representing only approximately 9% of the total genome. Codon replacements were associated with moderate decreases in total virion production but large decreases in the specific infectivities of intact poliovirions and viral RNAs. Replication of codon replacement viruses, but not MEF-1, was temperature sensitive at 39.5 degrees C. Synthesis and processing of viral intracellular proteins were largely unaltered in most codon replacement constructs. Replacement of natural codons with synonymous codons with increased frequencies of CpG and UpA dinucleotides may offer a general approach to the development of attenuated vaccines with well-defined antigenicities and very high genetic stabilities.

FIG. 1.

(A) MEF-1 codon replacement constructs aligned with a schematic of the poliovirus genome, with the single open reading frame represented by an open rectangle, flanked by the 5′- and 3′-untranslated regions, represented as lines. Bounding restriction endonuclease sites and locations of codon replacement cassettes (A, EcoRV-AgeI fragment [nt 748 to 1773; 1,026 nt]; B, AgeI-AfeI fragment [nt 1774 to 2610; 837 nt]; and C, AfeI-XhoI fragment [nt 2611 to 3300; 690 nt]) are indicated above the capsid region. Unmodified cassettes are indicated by italicized capital letters. ABC nucleotide sequences completely matched the MEF-1 sequence; ABC had two synonymous substitutions introduced to generate an AgeI site. (B) Plaque morphologies on HeLa cell monolayers (37°C, 60 h, 10-cm plates). Relative amounts of infected cell culture lysates yielding the plaques shown in each dish were as follows: ABC, 1; ABC, 0.1; aBC, 1; AbC, 1; ABc₀, 1; abC, 10; and abc₀, 4,000. (C) Mean plaque areas as a function of the number of nucleotide substitutions in the capsid region. (D) Growth properties of different virus constructs in single-step growth experiments in HeLa S3 cells (37°C). Single-step growth experiments were performed as described in Materials and Methods, and virus yields were determined by plaque assay. One milliliter of culture contained 2 × 10⁶ HeLa S3 cells. (E) Virus yields (averages for 10-h and 12-h time points) of the single-step growth experiments as a function of the total number of nucleotide substitutions in the capsid region.

FIG. 2.

Distribution of synonymous codon usage in cassette C of different virus constructs, described in Table 1 (also see Table S1 in the supplemental material). Encoded amino acids are identified by one-letter codes (top). Bars in graphs alternate between black and gray to facilitate visualization among amino acid codon sets.

FIG. 3.

(A) Plaque morphologies of constructs ABC, ABc₀ to ABc₁₂, and abc₀ on HeLa cell monolayers. Relative amounts of infected cell culture lysates yielding the plaques shown in each dish were as follows: ABC, 1; ABc₀, 4; ABc₁, 1; ABc₂, 1; ABc₃, 4; ABc₄, 1; ABc₅, 1; ABc₆, 1; ABc₇, 1; ABc₈, 1; ABc₉, 1; ABc₁₀, 1,600; ABc₁₁, 16,000; ABc₁₂, 160,000; and abc₀, 16,000. (B) Plaque yields of constructs ABC, ABc₀ to ABc₁₀, and abc₀ in single-step growth experiments in HeLa S3 cells.

FIG. 4.

Mean plaque areas (see Fig. 3A) as a function of different parameters for synonymous codon substitution in cassette C, including total nucleotide substitutions (A), total codon changes (B), %G+C (C), effective codon usage (N_C) (D), CAI (E), CPB (F), CpG dinucleotides (G), UpA dinucleotides (H), CpG-plus-UpA dinucleotides (I), CpA dinucleotides (J), UpG dinucleotides (K), and CpA-plus-UpG dinucleotides (L). All virus constructs had unmodified A and B cassettes. Numbers at data points correspond to cassette c subscripts (Table 1).

FIG. 5.

Virus infectivity yields (averages for 10-h and 12-h time points) in single-step growth experiments (Fig. 3B) as a function of the different parameters for synonymous codon substitution in cassette C described in the legend to Fig. 4. Virus constructs ABc₁₁ and ABc₁₂ were excluded from the single-step growth experiments because their titers were too low for infection at an MOI of 10.

FIG. 6.

Virus particle yields (solid squares) and infectivity yields (PFU/cell; solid circles) of virus constructs ABC, ABc₀, abC, and abc₀ in RD cells as a function of the total number of CpG-plus-UpA dinucleotides in the capsid region.

FIG. 7.

SDS-PAGE analysis of poliovirus-specific proteins produced by ABC, ABc₀, abc₀, ABc₁₀, ABc₁₁, and ABc₁₂ viruses in vivo. Lysates were obtained from HeLa cells infected with purified virus (MOI = 25 PFU/cell) labeled with [³⁵S]methionine at 4 to 7 h postinfection. Noncapsid proteins were identified by their electrophoretic mobilities and band intensities; capsid proteins were identified by their comigration with proteins from purified virions.

FIG. 8.

Temperature sensitivities of codon replacement virus constructs. Plaque morphologies (A), mean plaque areas (B), and plaque titers (C) are shown for constructs ABC, ABc₀, ABc₁₀, ABc₁₁, ABc₁₂, and abc₀ incubated on HeLa cell monolayers at 34.5°C, 37.0°C, and 39.5°C. Relative amounts of infected cell culture lysates yielding the plaques shown in each dish at 34.5°C, 37.0°C, and 39.5°C, respectively, were as follows: ABC, 1, 1, 1; ABc₀, 10, 10, 10; ABc₁₀, 10⁴, 10⁴, 10⁶; ABc₁₁, 10⁵, 10⁵, 10⁶; ABc₁₂, 10⁶, 10⁶, 10⁸; and abc₀, 10⁴, 10⁴, 10⁶.

FIG. 9.

Thermal inactivation kinetics of ABC, ABc₀, abc₀, ABc₁₀, ABc₁₁, and ABc₁₂ virus constructs at 46°C. Input titers (PFU/ml) were 4.0 × 10⁹ (ABC), 6.4 × 10⁸ (ABc₀), 2.6× 10⁷ (abc₀), 6.9 × 10⁶ (ABc₁₀), 7.1 × 10⁵ (ABc₁₁), and 1.5 × 10⁴ (ABc₁₂). The infectivity of ABc₁₂ was below detectable levels at 5 min of incubation at 46°C.