Both antigen optimization and lysosomal targeting are required for enhanced anti-tumour protective immunity in a human papillomavirus E7-expressing animal tumour model

Abstract

DNA immunization is a new approach for cancer immune therapy. In this study, we constructed human papillomavirus (HPV) 16 E7 expression vector cassettes and then compared the abilities of these constructs to induce antitumour protection. Lysosome-targeted E7 antigens, and to a lesser degree signal sequence-conjugated and transmembrane region sequence-conjugated E7 antigens in a DNA form, displayed tumour protection significantly higher than wild-type E7 antigens. This enhanced tumour protection was mediated by CD8+ cytotoxic T lymphocytes (CTL), as determined by in vivo T-cell depletion and in vitro interferon-gamma (IFN-gamma) production. Subsequent co-injection with interleukin-12-expressing cDNA showed insignificantly enhanced antitumour protection. However, E7 codon optimization plus lysosomal targeting resulted in a dramatic enhancement in antitumour protection both prophylactically and therapeutically through augmentation of the E7-specific CTL population, compared to either one of them alone. However, wild-type or codonoptimized E7 antigens without intracellular targeting displayed no protection against tumour challenge. Thus, these data suggest that antigen codon optimization plus lysosomal targeting strategy could be important in crafting more efficacious E7 DNA vaccines for tumour protection.

Figure 1

Construction of E7 DNA vaccine cassettes and their expression in vitro. (a) HPV 16 E7 genes were conjugated to the signal sequence of HSV-2 gB genes. E7 genes were further conjugated to the transmembrane region (TMR)-targeting sequence of HSV-2 gD genes. Finally, E7 genes were conjugated to the lysosomal targeting sequence (LAMP). (b) Renal carcinoma cells were transfected with E7 DNA cassettes. After 2 days incubation, 40 μg of cell lysates were run on 12% sodium dodecyl sulphate–polyacrylamide gel electrophoresis, followed by Western blot assay. Expression of size-corresponding E7 fusion proteins was confirmed.

Figure 2

CD8⁺ T-cell (CTL)-mediated antitumour protective immunity by lysosomal targeting of E7 antigens. (a) Two animal groups (n = 10, Sig/E7 and Sig/E7/LAMP) showing complete tumour protection for 50 days after tumour challenge (Table 1) were re-challenged s.c. with 20 times more tumour cells (2 × 10⁵ cells/mouse). Tumour size was measured using a caliper twice a week. The mean tumour size [(length + width)/2] in cm was recorded. Values and bars represent mean tumour size and SD, respectively. (b) An animal group (n = 10, Sig/E7/LAMP) showing complete tumour protection for 30 days after tumour re-challenge as shown in (a) was divided into two groups and then one group of animals (n = 5) was depleted in vivo of CD8⁺ T cells. Animals with or without CD8⁺ T-cell depletion were re-challenged s.c. with tumour cells (4 × 10⁵ cells/mouse). Values and bars represent mean tumour size and SD, respectively. (c) pcDNA3-Sig/E7/LAMP-immunized mice showing no tumour in (b) and age-matched control mice were killed for collection of splenocytes. Immune cells were stimulated in vitro with 1 μg E7 proteins or CTL peptides per ml for 3 days. Cell supernatants were obtained to measure IFN-γ production levels. Values and bars represent mean IFN-γ levels and SD, respectively. (d) Comparison of primary antitumour protection induced by E7 DNA vaccines (Sig/E7 and Sig/E7/LAMP). Each group of animals (n = 5) was immunized i.m. with 50 μg of pcDNA-Sig/E7 and pcDNA-Sig/E7/LAMP at 0 and 2 weeks. Two weeks after the last DNA immunization, animals were challenged s.c. with TC-1 tumour cells (1 × 10⁴ cells/mouse). Values and bars represent mean tumour size and SD, respectively. This was repeated with similar results. *Statistically significant at P < 0·05 using the paired Student's t-test compared to pcDNA3-Sig/E7 or control.

Figure 3

Anti-tumour prophylactic efficacy of pcDNA3-Sig/E7/LAMP and pcDNA3-Sig/sE7/LAMP at a higher tumour challenge dose (a) and evaluation of E7 protein expression levels in vitro (b). (a) Animals were immunized i.m. with 50 μg E7 DNA vaccines (pcDNA3-Sig/E7/LAMP and pcDNA3-Sig/sE7/LAMP) at 0 and 2 weeks. At 4 weeks, animals were challenged s.c. with 2 × 10⁵ TC-1 cells/mouse. Animals were checked for tumour formation twice a week. (b) Renal carcinoma cells were transfected with pcDNA3-LacZ, pcDNA3-Sig/E7/LAMP and pcDNA3-Sig/sE7/LAMP as shown in the Materials and method sections; 20 μg of cell lysates were run on 12% sodium dodecyl sulphate–polyacrylamide gel electrophoresis for Western blot analysis. The filter was re-hybridized with anti-actin antibodies to show that equal amounts of proteins were loaded. Arrow and arrowhead show E7 fusion proteins and actin proteins, respectively. A representative blot is shown.

Figure 4

Enhancement of therapeutic efficacy of E7 DNA vaccines by E7 codon optimization. Each group of animals (n = 5) was inoculated s.c. with 1 × 10⁴ TC-1 tumour cells per mouse. Next day, animals were immunized i.m. with 50 μg of control DNA vector (a), pcDNA3-Sig/E7/LAMP (b) and pcDNA3-Sig/sE7/LAMP (c), followed by two more injections at 1-week intervals. Tumour size was measured using a caliper twice a week. The mean tumour size [(length + width)/2)] in cm was recorded. This was repeated with similar results. (d) Tumour protection rates (%) of mice treated with E7 DNA vaccines. Numbers in (/) represent the number of animals showing complete tumour regression/the total number of animals treated with E7 DNA vaccines.

Figure 5

Requirement of lysosomal targeting of codon-optimized E7 genes for enhanced tumour protection. Each group of animals (n = 5) was immunized i.m. with 50 μg of control DNA vector, pcDNA3-Sig/sE7/LAMP, and pIn2-eE7 at 0 and 2 weeks. After 4 weeks, animals were challenged s.c. with 2 × 10⁴ (a) and 2 × 10⁵ (b) TC-1 tumour cells per mouse. Tumour size was measured using a caliper twice a week. The values and bars represent mean tumour size and SD, respectively.

Figure 6

Induction of E7-specific antibody and cellular immune responses by codon optimized E7 DNA vaccines (a, b, c), and effects of T-cell subsets on tumour growth (d). Each group of animals (n = 10) was immunized i.m. with 50 μg of pcDNA3-Sig/E7/LAMP and pcDNA3-Sig/sE7/LAMP at 0, 2 and 6 weeks. Eight weeks after the first DNA injection, animals were bled and sera were diluted to 1 : 50 for ELISA (a) The values and bars represent optical density (OD) values of each serum and the mean, respectively. (b) Equally pooled sera per group were diluted to 1 : 50 for determination of IgG isotype patterns. The values and bars represent means of OD values of equally pooled sera and the SD, respectively. (c) Eight weeks following the first DNA injection, animals were killed and spleen cells were pooled. Splenocytes were stimulated in vitro for 3 days with 1 μg HPV 16 E7 proteins or E7 CTL peptides per ml. Cell supernatants were tested for IFN-γ. The values and bars represent means of released IFN-γ concentrations and the SD, respectively. This was repeated with similar results. (d) Animals were immunized i.m. with 50 μg of pcDNA3-Sig/sE7/LAMP at 0, 2 and 6 weeks. After 2 weeks following the third injection, each group of animals (n = 5) was depleted in vivo of CD4⁺ or CD8⁺ T cells as described in the Materials and methods. Animals were challenged s.c. with 5 × 10⁴ TC-1 cells per mouse and then observed for tumour formation at 20 days following tumour challenge. The values represent the percentage of the number of animals showing no tumour/the total number of animals challenged with TC-1 tumour cells. *Statistically significant at P < 0·05 using the paired Student's t-test compared to negative control. **Statistically significant at P < 0·05 using the paired Student's t-test compared to pcDNA3-Sig/E7/LAMP.