Two complex, adenovirus-based vaccines that together induce immune responses to all four dengue virus serotypes

Abstract

Dengue virus infections can cause hemorrhagic fever, shock, encephalitis, and even death. Worldwide, approximately 2.5 billion people live in dengue-infested regions with about 100 million new cases each year, although many of these infections are believed to be silent. There are four antigenically distinct serotypes of dengue virus; thus, immunity from one serotype will not cross-protect from infection with the other three. The difficulties that hamper vaccine development include requirements of the natural conformation of the envelope glycoprotein to induce neutralizing immune responses and the necessity of presenting antigens of all four serotypes. Currently, the only way to meet these requirements is to use a mixture of four serotypes of live attenuated dengue viruses, but safety remains a major problem. In this study, we have developed the basis for a tetravalent dengue vaccine using a novel complex adenovirus platform that is capable of expressing multiple antigens de novo. This dengue vaccine is constructed as a pair of vectors that each expresses the premembrane and envelope genes of two different dengue virus serotypes. Upon vaccination, the vaccine expressed high levels of the dengue virus antigens in cells to mimic a natural infection and induced both humoral and cellular immune responses against multiple serotypes of dengue virus in an animal model. Further analyses show the humoral responses were indeed neutralizing against all four serotypes. Our studies demonstrate the concept of mimicking infections to induce immune responses by synthesizing dengue virus membrane antigens de novo and the feasibility of developing an effective tetravalent dengue vaccine by vector-mediated expression of glycoproteins of the four serotypes.

FIG. 1.

Dengue virus and cAdVax vector genomes. (A) Schematic of the dengue virus positive-stranded RNA genome. (B) Schematic of the cAdVaxD(1-2) vaccine construct. cAdVaxD(1-2) contains DEN1 and DEN2 prM-E sequences in the left and right arms of the vector, respectively. (C) Schematic of the cAdVaxD(3-4) vaccine construct. cAdVaxD(3-4) contains DEN3 and DEN4 prM-E sequences in the left and right arms of the vector, respectively. TR, terminal repeat; CMV, cytomegalovirus.

FIG. 2.

cAdVaxD vectors express dengue virus proteins in infected cells in vitro. (A) Lysates of HEK293 cells infected with cAdVaxD(1-2) were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, and Western blotting was performed to analyze DEN1 and DEN2 prM-E expression. The results are a composite of multiple Western blots probed with serotype- and protein-specific antibodies (shown in parentheses): lane 1, DEN2 E (clone 2H3); lane 2, DEN2 E and DEN2 prM (DEN2 hyperimmune ascitic fluid); lane 3, DEN1 prM (clone 8C2); lane 4, DEN1 E (clone 8B9). The positions of molecular mass markers (in kilodaltons) are shown. (B) Lysates of HEK293 cells infected with cAdVaxD(3-4) were analyzed as described above for panel A for DEN3 and DEN4 prM-E expression. Results are a composite of multiple Western blots: lane 1, DEN3 E (clone 8D2-1); lane 3, DEN4 E (clone 1H10-6-7). Lanes 2 and 4 were probed with MAb clones 14A4-8 and 1A2-6, respectively. While these two clones are specific for dengue virus, the serotype specificity is undetermined.

FIG. 3.

Indirect immunofluorescent staining of DEN proteins expressed from cAdVax vectors. (A) Vero cells were mock infected (control) or infected with cAdVaxD(1-2) at a MOI of 20 or with wild-type DEN1 or DEN2 viruses at a MOI of 2 for 48 h. Following infection, the cells were fixed in formaldehyde and immunostained with serotype-specific MAb 13E7-9-10 (DEN1/DEN3 E) or 3H5-1 (DEN2 E). (B) Vero cells were mock infected (control) or infected with cAdVaxD(3-4) at a MOI of 20 or with wild-type DEN3 or DEN4 viruses at a MOI of 2 for 48 h. Cells were fixed and immunostained with serotype-specific MAb 5D4-11 (DEN3 E) or 1H10-6 (DEN4 E). Positive cells were detected using a phycoerythrin-conjugated secondary antibody and visualized with a fluorescence microscope.

FIG. 4.

Vaccination of mice with cAdVaxD vectors induces anti-DEN2 and anti-DEN4 antibody responses. CD-1 mice were vaccinated with 1 × 10⁸ PFU of the indicated vaccines on weeks 0 and 8. After vaccination, serum samples were collected biweekly for antibody analyses using ELISA. The wells in 96-well plates were coated with culture supernatants from Vero cells infected with wild-type DEN2 (A and C) or DEN4 (B and D) and probed with serial dilutions of serum obtained from vaccinated animals at the indicated time points. Dilution curves were used to calculate antibody titers from individual animals. Each data point represents the mean ± standard deviation (error bar) for at least five individual vaccinated mice. (A and B) Sera from mice vaccinated with cAdVaxD(1-2) or S3G were tested against DEN2 (A) or DEN4 (B) serotype. (C and D) Sera from mice vaccinated with cAdVaxD(3-4) or CHB3 were tested against DEN2 (C) or DEN4 (D) serotype.

FIG. 5.

Vaccination of mice with cAdVaxD vectors induces neutralizing antibody responses against all four DEN serotypes. Sera from mice vaccinated with cAdVaxD(1-2), cAdVaxD(3-4), S3G, or CHB3 were used to perform DEN PRNTs. Serial dilutions of sera that were collected 18 weeks after primary immunizations were incubated with each DEN serotype prior to infection of Vero cells. Seven days later, DEN plaques were visualized by immunostaining and then counted. Raw data were converted to percent inhibition of DEN plaque formation, and each data point represents the mean ± standard deviation (error bar) for four individual animals. (A) DEN1 neutralizing activity; (B) DEN2 neutralizing activity; (C) DEN3 neutralizing activity; (D) DEN4 neutralizing activity.

FIG. 6.

Vaccination of mice with cAdVaxD vectors induces anti-dengue virus cellular immune responses. C57BL/6 mice were vaccinated with 1 × 10⁸ PFU of cAdVaxD(1-2), cAdVaxD(3-4), or CHB3. Animals were sacrificed 4 and 10 weeks after primary immunizations, and isolated splenocytes from each animal were incubated with a pool of overlapping peptides derived from the DEN2 E protein (amino acids 1 to 110). After incubation, anti-mouse IFN-γ ELISPOT assays were performed to detect DEN-specific CTL activities. Results are presented as the number of spots formed per 10⁶ cells, and each data point represents the mean plus standard deviation (error bar) for four individual animals.