Development and characterization of a reverse genetic system for studying dengue virus serotype 3 strain variation and neutralization

Abstract

Dengue viruses (DENV) are enveloped single-stranded positive-sense RNA viruses transmitted by Aedes spp. mosquitoes. There are four genetically distinct serotypes designated DENV-1 through DENV-4, each further subdivided into distinct genotypes. The dengue scientific community has long contended that infection with one serotype confers lifelong protection against subsequent infection with the same serotype, irrespective of virus genotype. However this hypothesis is under increased scrutiny and the role of DENV genotypic variation in protection from repeated infection is less certain. As dengue vaccine trials move increasingly into field-testing, there is an urgent need to develop tools to better define the role of genotypic variation in DENV infection and immunity. To better understand genotypic variation in DENV-3 neutralization and protection, we designed and constructed a panel of isogenic, recombinant DENV-3 infectious clones, each expressing an envelope glycoprotein from a different DENV-3 genotype; Philippines 1982 (genotype I), Thailand 1995 (genotype II), Sri Lanka 1989 and Cuba 2002 (genotype III) and Puerto Rico 1977 (genotype IV). We used the panel to explore how natural envelope variation influences DENV-polyclonal serum interactions. When the recombinant viruses were tested in neutralization assays using immune sera from primary DENV infections, neutralization titers varied by as much as ∼19-fold, depending on the expressed envelope glycoprotein. The observed variability in neutralization titers suggests that relatively few residue changes in the E glycoprotein may have significant effects on DENV specific humoral immunity and influence antibody mediated protection or disease enhancement in the setting of both natural infection and vaccination. These genotypic differences are also likely to be important in temporal and spatial microevolution of DENV-3 in the background of heterotypic neutralization. The recombinant and synthetic tools described here are valuable for testing hypotheses on genetic determinants of DENV-3 immunopathogenesis.

Figure 1. Design of DENV3 infectious clone and chimeras.

Panel A provides a schematic representation of the DENV genome, divided into structural and nonstructural genes. Arrows indicate primer name and approximate primer locations and orientation on the genome. These primers were used to amplify the different cDNA fragments as well as adding appropriate, terminal restriction enzyme recognition sequences. The T7 promoter is located at the 5′end of primer DEN#1. Primer pairs that generate complete fragments are aligned opposite one another. The clone was propagated in in three circular and one linear plasmid. DENV fragments within each plasmid are highlighted in blue. The final fragments assembled to generate the clone are represented at the bottom of the figure as blue lines. Lengths and restriction site locations are approximate and supported by exact primer sequences, positions and PCR fragment sizes as noted in Text S1. Panel B illustrates the strategy used for E gene construct insertion. The top figures represent the DEN A and B fragments, which encode for the E protein. Arrows indicate approximate locations of primers EGENE+ and EGENE− used to silently introduce a BsaI recognition sequence into the 5′ end of fragment DEN B. The Bio Basic E construct in black represents the synthesized E gene, and the schematic below it illustrates approximate arrangement of E gene domains and restriction sites. Roman numerals indicate E domains; TM = transmembrane region. Arrows indicate primer names, approximate location and pair orientation used to introduce restriction enzyme recognition sequences Lengths and positions shown are approximate. The E construct is amplified for insertion into the DEN A fragment with primers pUC57 and with DEN2kb− and for the B fragment with primers DEN2kb+ and EGENE−. After sequence confirmation, constructs and DEN A and B fragments are digested with indicated enzymes, desired fragments gel purified and subjected to ligation to generate new DEN A and DEN B fragments containing E constructs. *Type IIS restriction endonuclease.

Figure 2. Phylogenetic relationship of DENV-3 viruses.

The phylogenetic tree illustrates genetic relatedness of DENV-3 virus genotypes, including those viruses from which representative E genes were synthesized. This tree is meant to display DENV-3 diversity but does not include all 175 sequences used to evaluate the genetic variability of the DENV-3 E gene. Instead, representative sequences from each genotype were selected for inclusion in the tree. The tree was constructed using Maximum Likelihood method based on the Tamura-Nei model . The tree is drawn to scale, with branch lengths measured in the number of substitutions per site. The analysis involved 37 nucleotide sequences. There were a total of 1479 positions in the final dataset. Evolutionary analyses were conducted in MEGA4 . *Parent virus for E gene variants.

Figure 3. Recombinant dengue virus growth kinetics in tissue culture.

Vero (figure 3A) and C6/36 (figure 3B) cells were inoculated at a multiplicity of infection (m.o.i) of 0.01 FFU. Cell culture supernatants were harvested at indicated times and the virus released from the infected cells was quantitated by immunofocus assay. Points show geometric mean titer (GMT) calculated from triplicate titrations. Error bars indicate 95% confidence intervals for each GMT. For Vero cells no focus forming units were observed at 0 hrs and for C6/36 cells no focus forming units were observed at 0 hrs and 24 hrs.

Figure 4. Mean FRNT₅₀ values for homotypic primary and secondary sera.

Homotypic primary (anti DENV-3) sera (Figures 4A–4H) or secondary serum (Figure 4I) FRNT₅₀ titers against each of the E variant isogenic clones. Each serum sample is identified above the graph. Serum histories are summarized in Table 2 and Text S1. Fold-dilution of serum is on the Y-axis and each clone is identified on the X-axis. Columns show GMT FRNT₅₀ values calculated from FRNT done in triplicate. Error bars indicate 95% confidence intervals. Columns connected by horizontal lines indicate groups of clones that did not have statistically significantly different FRNT₅₀ values (P<0.05). Connecting bars indicate individual or groups of clones that differed statistically (P<0.05) by Tukey's HSD. Sera 003 (Figure 4A), 005 (Figure 4B) and 103 (Figure 4E) did not have significantly different titers against the recombinant viruses.

Figure 5. Mean FRNT₅₀ values for heterotypic primary sera.

Each serum sample is identified above the graph. Serum 001(Figure 5A) is from a primary DENV-2 infection in; Serum 006 (Figure 5B) is from a primary DENV-1 infection; Serum 102 (Figure 5C) is from a primary DENV-4 infection. Serum histories are summarized in Table 2 and Text S1. Fold-dilution of serum is on the Y-axis and each clone is identified on the X-axis. Columns show GMT FRNT₅₀ values calculated from FRNT done in triplicate.