Fig 1.
Experimental model of recombination in the 5’ untranslated region of enteroviruses.
(A) Schematic representation of the genome of MAD4 cVDPV. The poly-adenylated single positive-strand RNA genome is covalently linked to the viral protein VPg (also named 3B) at the 5’ terminus. The unique large open-reading frame is flanked by two untranslated regions (5’ and 3’ UTRs). The 5’ UTR (nt 1 to 747) is magnified to indicate the seven stem-loop structures (I to VII) forming two functional units, the cloverleaf (CL: I) and the internal ribosome entry site (IRES: II-VI). Genomic regions encoding viral proteins VP4 to 3Dpol are indicated. The MAD4 genome is a PV/EV-C recombinant between mutated Sabin 2 sequences (light shading) and non-vaccine sequences derived from coxsackieviruses A (CV-A) (dark shading). (B) Substitutions in the CL leading to a noninfectious MAD4 genome. The native CL structure can be subdivided into four domains: stem a and stem-loops b to d. Mutations (lowercase) introduced into the CL of MAD4 disrupt stem a and create new predicted base pairing interactions. Secondary structure predictions were generated with mfold, version 3.6 [63]. (C) Rescue of the defective MAD4 genomic RNA (3’ partner) by co-transfection with 5’ UTR sequences from EVs (5’ partners). 5’ partners included the complete 5’ UTR followed by sequences encoding VP4, VP2 and part of VP3 from eight different EV strains belonging to the four human EV species (see Fig 2). Human HEp-2c and murine L20B cells were co-transfected with each RNA partner pair and then incubated in semisolid medium until plaques appeared, indicating the generation of viable recombinants.
Fig 2.
Phylogenetic relationships between the 5’ UTR sequences of the eight 5’ partners.
This neighbor-joining tree was constructed with MEGA, version 6.06 [65] using aligned 5’ UTR nt sequences. The reliability of tree topology was estimated using 1000 bootstrap replicates. The nt sequence of rhinovirus A16 (RV-A16) was used as an outgroup. The name of each isolate includes the type of the isolate followed by the laboratory number or name for the prototype strains. For each 5’ partner, the percentage similarity with MAD4 is shown in brackets. The enterovirus species (EV-A to -D) and the 5’ UTR group I and II are indicated.
Table 1.
Recombination efficiency following the co-transfection of L20B and HEp-2c cells with the MAD4 3’ RNA partner and each 5’ RNA partner.
Fig 3.
Location of recombination sites in group I/I recombinants.
(A) Genomes of recombinant viruses resulting from the co-transfection of the 3’ partner MAD4 with 5’ partners from group I 5’ UTRs (S2 File). Each recombinant is plotted on the graph according to the position of its recombination site in the 3’ partner sequence (on the x-axis) and in the 5’ partner sequence (on the y-axis). To facilitate the comparison between the different 5’ partners, the location of 5’ UTR domains and crossover sites are indicated, following alignment of all considered 5’ UTR sequences, numbered according to the MAD4 sequence. Homologous recombinants are located on the diagonal (identity line), whereas recombinants displaying insertions or deletions are below or above this line, respectively. Three examples numbered from 1 to 3 are shown (arrows). The corresponding 5’ UTR structures are drawn on the right side. Recombinants 1, 2 and 3 are examples of a homologous and two nonhomologous recombinants with an insertion and a deletion, respectively. The three recombinant classes identified on the graph, A, B and C, are circled, and the hotspot (HS) within each class, HSA, HSB and HSC, are framed. The size of the dot is proportional to the number of recombinants having the same recombination site coordinates (from 1 to 9). The color code indicates the respective recombinant type (see table below). (B) The table shows the percentage and number of recombinants within each class. The number of recombinants with recombination sites located in hotspots are indicated in brackets.
Fig 4.
Location of recombination sites in group II/I recombinants.
Genomes of recombinant viruses resulting from the co-transfection of the 3’ partner MAD4 with 5’ partners from group II 5’ UTRs (S3 File). For details, see the legend of Fig 3.
Table 2.
Stability of selected CV-A13.67900/MAD4 and CV-B4/MAD4 recombinants upon serial passaging.
Fig 5.
Competition assays comparing the relative fitness of homologous and nonhomologous recombinants.
Each of the four nonhomologous selected CV-B4/MAD4 recombinants presented in Table 2 were competed against the homologous recombinant CV-B4/MAD4 B.38, the recombination site of which is located in the spacer between IRES domains dIII and dIV (CV-B4 nt 240 / MAD4 nt 234) (A to D), and against the parental cVDPV MAD4 (E to H). Viruses were mixed at a 1:1 ratio and HEp-2c cells were inoculated in triplicate at an MOI of 0.01 TCID50/cell for three passages. At each passage, the progeny RNA was amplified by RT-PCR and the relative amount of each competitor was evaluated by measuring the intensity of the bands on agarose gel electrophoresis (ImageJ 1.47 software, NIH) corresponding either to the nonhomologous recombinant or to its homologous competitor. The ratio of band intensity for the nonhomologous recombinant versus the homologous competitor is shown for each well for the three passages (S6 Fig). An increase in this ratio indicates that the nonhomologous recombinant is more fit than the homologous one. Error bars indicate the standard deviation (Student’s t-test, n = 3; *P<0.05, **P<0.01, ***P<0.001).
Table 3.
Location of recombination site in selected recombinants.
Fig 6.
Growth and plaque size of selected recombinant viruses.
HEp-2c (A), Caco2 (B) and IMR5 (C) cells were infected with MAD4 and the 12 selected recombinants at an MOI of 10 TCID50/cell for 5 hours. Each point shows the log10 of the mean total virus titer from three independent experiments. Error bars indicate the standard deviation. *P<0.05, **P<0.01, ***P<0.001 in a Student’s t-test comparing each selected recombinant virus with MAD4 (n = 3). D. Plaque assays were performed on HEp-2c cell monolayers (two or three experiments). The relative area of plaques was calculated, with the average value of MAD4 plaque area set at 100%. Box plots show median values (thick bars) as well as maximum and minimum values. Averages are indicated (◊). (Student’s t-test, n = 39–578; *P<0.05, **P<0.01, ***P<0.001).
Fig 7.
Neurovirulence of MAD4 and selected recombinant viruses in PVR-Tg mice.
PVR-Tg21 mice expressing the human PV receptor were inoculated intracerebrally with 105 TCID50 of virus (six mice per virus). Animals were checked daily for 21 days post-inoculation for paralysis or death. The number of healthy mice following inoculation with group I/I recombinants (A) or group II/I recombinants (B) relative to that following inoculation with the parental cVDPV MAD4 is shown. Survival curves are numbered to indicate overlap. No additional mice suffered paralysis or died after day 15 post-inoculation. *P<0.05, **P<0.01, ***P<0.001 in Log Rank tests comparing each selected recombinant virus with MAD4 (S6 File).
Fig 8.
Competition assays comparing the fitness of MAD4 and selected CV-A17/MAD4 or EV-70/MAD4 homologous recombinants.
HEp-2c cells were infected in triplicate with a 1:1 mix of a recombinant virus and the parental strain MAD4 at an MOI of 0.01 TCID50/cell for three passages. The proportion of each virus at each passage was determined by real time RT-PCR and expressed as the ratio of the CT values of MAD4 competitor versus recombinant. The mean ± the standard deviation is shown (Student’s t-test, n = 3; *P<0.05, **P<0.01, ***P<0.001). A CT value ratio above 1 indicates that the recombinant is fitter than MAD4.