Human papillomavirus (HPV) type 16 E7 protein bodies cause tumour regression in mice

Abstract

Background: Human papillomaviruses (HPV) are the causative agents of cervical cancer in women, which results in over 250 000 deaths per year. Presently there are two prophylactic vaccines on the market, protecting against the two most common high-risk HPV types 16 and 18. These vaccines remain very expensive and are not generally affordable in developing countries where they are needed most. Additionally, there remains a need to treat women that are already infected with HPV, and who have high-grade lesions or cervical cancer.

Methods: In this paper, we characterize the immunogenicity of a therapeutic vaccine that targets the E7 protein of the most prevalent high-risk HPV - type 16 - the gene which has previously been shown to be effective in DNA vaccine trials in mice. The synthetic shuffled HPV-16 E7 (16E7SH) has lost its transforming properties but retains all naturally-occurring CTL epitopes. This was genetically fused to Zera®, a self-assembly domain of the maize γ-zein able to induce the accumulation of recombinant proteins into protein bodies (PBs), within the endoplasmic reticulum in a number of expression systems.

Results: High-level expression of the HPV 16E7SH protein fused to Zera® in plants was achieved, and the protein bodies could be easily and cost-effectively purified. Immune responses comparable to the 16E7SH DNA vaccine were demonstrated in the murine model, with the protein vaccine successfully inducing a specific humoral as well as cell mediated immune response, and mediating tumour regression.

Conclusions: The fusion of 16E7SH to the Zera® peptide was found to enhance the immune responses, presumably by means of a more efficient antigen presentation via the protein bodies. Interestingly, simply mixing the free PBs and 16E7SH also enhanced immune responses, indicating an adjuvant activity for the Zera® PBs.

Figure 1

Constructs used for making DNA vaccines. (A) The HPV-16 E7 gene was cleaved at the positions corresponding to the pRB binding site and in between the two Cys-XX-Cys motifs. The cleavage points between amino-acid numbers are shown above the gene. The resulting fragments were rearranged (“shuffled”) forming the core-element, and an appendix containing the junctions where the cleavage took place was added to avoid loss of putative CTL epitopes to form 16E7SH. (B) Depiction of the 7 constructs utilised in this study. The core genes and fusions of genes were cloned into the respective pTH or pTRAc expression vectors.

Figure 2

Western blots of crude plant extracts and of purified ZERA-16E7SH and ZERA-eGFP protein. (A) Samples from plants expressing ZERA-16E7, ZERA-16E7SH and 16E7SH were harvested at 3, 5, 7 and 10 dpi, separated on a polyacrylamide gel and blotted onto nitrocellulose membrane. Proteins were detected with HPV-16 E7 antibody. 0.6 μg of E. coli- produced purified His-E7 protein was used as a positive (+ve) and comparative control. SH indicates the use of the shuffled 16E7 gene. + or - indicates the fusion of Zera®. Black arrows indicate the E7 positive control protein (18 kDa) and the Zera®-fused E7 proteins. (B) Leaves from vacuum-infiltrated N. benthamiana co-infiltrated with pBIN-NSS and either pTRAc-ZERA-16E7SH or pTRAc-ZERA-eGFP were extracted 5 dpi, ground in liquid nitrogen, homogenized, filtered and separated on sucrose density gradients. Interphase fractions (IF) were aspirated, pellets (P) were resuspended and run on acrylamide gels, blotted on nitrocellulose membranes and probed with HPV-16 E7 monoclonal antibody or anti-GFP monoclonal antibody. Black arrows indicate protein bands of expected sizes.

Figure 3

Tumour regression in mice inoculated with different vaccines. Mice were inoculated with 0.5 × 10⁶ C3 tumour cells to induce tumours and subsequently injected with vaccine (5 μg protein or 100 μg DNA in 100 μl) in two sites per animal when the tumours were clearly palpable. Surface tumour size was measured over time. Because some tumours became bloody in some animals of the control group (empty vectors), the experiment was terminated at day 14. Data gives the mean ± SEM of the indicated group (n = 10). A) Animal groups injected with 100 μg pTH DNA, or 100 μg pTH-16E7SH DNA, or 50 μg plant-produced ZERA-16E7SH protein or ZERA-16E7SH protein plus adjuvant (IFA). B) Animal groups injected with control DNA (pTH) or plant-produced ZERA-eGFP protein.

Figure 4

IFN-γ response, CTL activity and Granzyme B ELISPOT assays on mouse splenocytes. Four mice per group were injected either with 5 μg protein or with 100 μg DNA per animal, respectively. (A)Ex vivo IFN-γ ELISPOT responses. Given are the means of IFN-γ secreting cells/10⁴ splenocytes ± SEM. ZERA-16E7SH compared to ZERA-eGFP (p < 0.001) and to pTH (p < 0.001). (B) Splenocytes were tested by ⁵¹Cr-release assays after one round of in vitro re-stimulation for lysis of E7-wildtype expressing 2 F11 target cells. Data is given as mean ± SEM. ZERA-16E7SH compared to ZERA-eGFP (p < 0.001) and to pTH (p < 0.001). (C)Ex vivo Granzyme B ELISPOT responses. Four mice per group were immunized with 5 μg protein injected per animal. Given are the means of Granzyme B-secreting cells/10⁴ splenocytes ± SEM. ZERA-16E7SH compared to ZERA-eGFP (p < 0.001) and to OVA (p < 0.001).

Figure 5

Tumour regression in mice inoculated with Zera® separately. Mice received 0.5 × 10⁶ C3 tumour cells s.c. and when the tumours were clearly palpable they were immunized with 100 μg DNA vaccine or 2.5 μg 16E7SH +/- 2.5 μg Zera® PBs +/- 100 μl IFA per animal and surface tumour size was measured. Experiments were repeated (Exp. I and Exp. II) and both results are shown. The experiment was terminated at day 39 due to the size of tumours in the control groups (empty vectors). Data gives the mean ± SEM. of the indicated group at day 39 but in the second exp: on day 43 (n = 10). pTH-ZERA-16E7SH DNA compared to pTH-16E7SH (Exp. I - p < 0.05, Exp. II p = 0.2). 16E7SH protein compared to 16E7SH protein plus Zera® PBs (Exp. I - p < 0.005, Exp. II p < 0.0005).

Figure 6

Ex vivo IFN-γ, CTL activity and Granzyme B ELISPOT responses in mice inoculated with Zera® separately. Two groups of four mice were immunized with 100 μg DNA vaccine, or 2.5 μg 16E7SH +/- 2.5 μg Zera® PBs +/- 100 μl IFA per animal for each assay. (A)Ex vivo IFN-γ ELISPOT responses. Given are the means of IFN-γ secreting cells/10⁴ splenocytes ± SEM. (B)Ex vivo Granzyme B ELISPOT responses. Four mice per group were immunized with 5 μg protein injected per animal. Given are the means of Granzyme B-secreting cells / 10⁴ splenocytes ± SEM. 16E7SH protein compared to 16E7SH protein plus Zera® PBs (Exp. I - p = 0.0001, Exp. II p < 0.05). (C) The splenocytes were tested by ⁵¹Cr-release assays after one round of in vitro restimulation for lysis of syngeneic E7-wildtype expressing 2 F11 target cells. Data gives the mean ± SEM of the indicated group (n = 4). One representative of the two experiments is shown.

Figure 7

Humoral response of mice inoculated with Zera® separately. Four mice per group were immunized with 100 μg DNA vaccine or 50 μg protein and blood was taken post-immunization (day of splenectomy). Direct ELISA was performed against both Zera® and 16E7 protein using anti-Zera and anti-16E7 IgGs. Data gives the mean ± SEM of duplicates of the indicated group (n = 4). One representative of the two experiments is shown. pTH-ZERA-16E7SH DNA compared to pTH-16E7SH (p < 0.005), 16E7SH protein compared to 16E7SH protein plus Zera® PBs (p < 0.005).