Preventive And Therapeutic Roles Of Lentiviral Vector Prime And Vaccinia Virus Vector Boost In Mouse Melanoma

Based on the great success of vaccines in prevention from many infectious diseases, scientists are trying to generate cancer vaccines to treat or prevent tumors. As early as1902, Leyden and Blumentha attempted to treat tumors with tumor cell vaccines. But, their work did not succeed. In1959, Graham treated232women bearing gynecological cancer with autologous cancer vaccines, and found some therapeutic efficiency without any side effects. However this treatment could not change the course of tumor growth. After human tumor specific antigens (TSA) and human tumor associated antigens (TAA) identified and their genes being cloned, active immunotherapy based cancer vaccines has been revitalized and intensively pursued. Cancer vaccines can treat cancers by inducing specific cellular and humoral immune response to kill tumor cells, which differs from the classical vaccine used in prevention of infectious diseases. As therapeutic vaccine, the efficiency of tumor vaccine is restrained by the low magnitude of immune responses due to the weak immunogenicity of tumor self-antigen and by the multiple immunosuppressive factors in tumor microenvironment. One approach for resolving above problems is to develop efficient antigen delivery system to elicit potent immune responses against weak antigen in immunosuppressive microenvironments. Here, we would use recombinant viral vectors as TSA/TAA delivery system to elicit the specific anti-tumor immune response.Recombinant vaccinia virus vector (w) is one of the vectors being earliest used for tumor gene therapy. The several unique features of w include1) high immunogenicity;2) high capacity of accommodating large Ag,3) pan-trpoism of infecting different types of human and mouse cells;4) capability of lyzing tumor cells;5) well established safety;6) easy preparation of large amount of virus. Therefore, w has been widely used in developing vaccines for infectious diseases and cancers. For example, w carrying gp160was immunized to induce anti-HIV immune response in human, and w encoded HPV can elicit HPV specific immune response. Moreover, a series of studies proved that w expressing cytokines (e.g.GM-CSF) or TAA (e.g.CEA) can be used to treat tumors. However the dominant immune responses against viral vector will limit the desired immune responses against target tumor Ags. On the other hand, lentiviral vector (lv) is a very promising vector for delivering tumor antigens because lv often only expresses desired antigen and induces dominant specific antigen immune response. Secondly, lv can efficiently transfect the non-dividing dendritic cells (DCs), resulting in the effective processing and presentation of lv encoded Ags to effector T cells to initiate the immune responses. Additionally, lv can induce persistent Ag expression and elicit long lasting immune response. For example, lv expressing melanoma associated Ag like NY-ESO or TRP1can induce effective specific anti-melanoma immune response and prevent tumor growth in vivo.As the application of single virus vector cannot induce ideal therapeutic effect even though both w and lv are good virus vectors for tumor vaccine, our consideration is to use lv and w with a prime-boost strategy which is expected to induce more potent immune response and enhance the tumor prevention and therapeutic effect. In the past decade, heterologous prime-boost immunizations, which was performed by giving different vaccine delivery system carrying the same antigen was found to be more effective than the ’homologous’ prime-boost approach. In current work, we constructed lv and w which expresses human melanoma antigen gp100respectively, and the corresponding virus particles were prepared for immunization in mice. The preventive and therapeutic effects of lv and vv carrying tumor antigen were observed in vaccinated mouse melanoma model. There are three parts in our work, which are briefly described as follows.Part Ⅰ. Construction and immunization of recombinant lv and vv-expressing human melanoma antigen gp100We constructed recombinant lv-expressing melanoma Ag human gp100because it was previously demonstrated that xenogenic hgp100DNA could better protect mice from B16tumor cell challenge than mouse gp100(mgp100). The gene fragment containing the N-terminal340aa of hgp100was obtained using high-fidelity PCR and cloned into the lv plasmid. The recombinant lv was designated as hgp100-lv. The hgp100-lv DNA was transfected into human293T cells with calcium phosphate transfection method, and the virus particles could be detected in supernatant. We successfully obtained the highly concentrated virus particles (1-5×109Tu/ml). We also constructed hgp100-pG10vector, which was co-cultured with WR vaccinia virus in CV-1cells. The virus particle of hgp100-w with a high titer of1-5×109Tu/ml was obtained by homologous recombination and several rounds of purification. Compare to hgp100-w, hgp100-lv induced more profound in vivo proliferation of hgp10025-33-specific CD8T cells, pmel-1. In addition, using intracellular staining, we found that the activation of endogenous CD8T cells by lv was substantially higher than of recombinant w. Notably, the lv-expressing xenogenic human gp100could induce potent CD8responses, which showed a cross-react with mouse gp100.In summary, we obtained hgp100-lv and hgp100-vv virus particles which can induce specific immune response. The hgp100-lv is more potent than hgp100-w in inducing specific T cell immune responses.Part Ⅱ The preventive effect of vaccination with hgp100-lv-hgp100-w prime-boost in melanoma model.Thus far, most of cancer vaccines are used to treat tumors, which, not surprisingly, did not generate significant antitumor effect because vaccines are intended to prevent diseases. In this work, we first studied if lv and w can prevent melanoma. The mice were immunized with hgp100-lv by footpad injection, and5to10%of CD8+T cells can secrete IFNy after ex vivo stimulated with hgp100peptide.40%of CD8+T cells are hgp10025-33tetramer positive and～20%of tetramer-positive CD8+T cells with CD127+were observed in the peak of the response,10to14days after immunization, which suggests that lv-activated T cells contain a high level of precursors of effector memory CD8+T cells in the peripheral at early stage of immune responses. The CD8immune response decreases to a plateau, when only2%of CD8+T cells can specifically secrete IFNy in30to40days after lv immunization. Boosting immunization with w followed by primary immunization with lv was performed, and the results showed that vv boosting can induce very strong secondary response and substantially enhance the immune response. About15%of CD8+T cells in PBMC can secrete IFNy after ex vivo specific stimulation at5to10days after vv boost, while vv prime-lv boost only induced the same magnitude of immune response as that of the lv-elicited primary responses. And the lv prime-lv boost group also obtained similar level of response. More importantly, the CD8responses induced by lv prime-vv boost were long lasting. More than2months (70d) after boost, the ratio of IFNγ+CD8+T cells in the PBMC was still at the level of5%, while this number decreased to1%in lv or vv-lv immunized groups. The immunization with lv-vv markedly increased the number of hgp100specific CD8+T cells and memory CD8+T cells. Surprisingly, memory CD8+T cells in the lv-vv prime-boosted mice could immediately sense the B16tumor cell challenge, be expanded significantly, and exert the cytotoxic function to prevent tumor occurrence.In order to study the role of CD8+and CD4+effector T cells in cancer immune prevention, CD8, CD4, or both were depleted at2weeks after lv immunization. We found that both CD4+and CD8+effector cells were required for immune prevention while CD8+T cells played dominant role. On the other hand, both primary and secondary immune response were greatly impaired by CD4+T cells depletion, either prior to lv prime or prior to vv boost immunization, indicating that CD4help is required for both primary and secondary CD8immune response in this system. More importantly, the lv-vv prime-boost immunization strategy can prevent tumor growth in autochthonous tumor model, the Grm1Tg mice, suggesting the applicability of this strategy in clinical cancer therapy.In this part, we demonstrated that lv-vv prime-boost immunization can induce stronger (secondary) immune responses. This strategy can potently promote CD8+T cell proliferation and generates a large number of memory CD8+T cells, which can rapidly sense and eliminate the tumor cells and results in effective tumor prevention.Part Ⅲ. The affect of local acute inflammation in TIL function and therapeutic effect of hgp100-lv immunization in melanoma modelThe recent research including our previous work reveals that tumor vaccines could not provoke the efficient antitumor activity on solid tumor, which has been attributed to the tumor microenvironment inducing failure of tumor infiltration lymphocytes (TILs) in tumor tissue. Our previous work also observed the same phenomena. Based on the antigen specific lymphocytes in acute infection tissue can quickly eliminate pathogenic microbial, we hypothesized that induction of acute inflammation in tumor lesions may be capable of rescuing and restoring the effector function of TILs to achieve enhanced antitumor effect. The oncloytic vaccinia vector as an excellent candidate can be used for such purpose. In the current study, the effect of vaccinia vector on rescuing the effector function of TILs in tumor milieu were observed.In this study, a mouse advanced melanoma model was established by which0.3×106B16cells were transplanted to grow in the mice for5days. The mice bearing tumor were immunized with hgp100-lv by footpad injection and boosted in effector phase (7days after lv prime when the immune response can be detected) with hgp100-vv intratumorally (i.t.) or i.p.. The tumor growth curve was recorded and the immunological changes of tumor lesions were analyzed in6to10days after boost. The results showed that hgp100-lv prime and hgp100-vv boost by i.p. at effector phase can enhance the systemic immune response, and also increase the number of tumor infiltrating CD8+and CD4+cells. However, this strategy did not improve the anti-tumor effect significantly by comparing the tumor weight. Interestingly, the boosting with hgp100-w (i.t.) can rescue the effector function of tumor infiltrating CD8T cells which were primed by hgp100-lv, and decrease the Treg percentage in CD4cells and greatly improve the anti-tumor effect, although this i.t. boosting cannot enhance systemic immune response and neither changes the number of TILs cells.Furthermore, we replaced w with TLR3and9ligands (PolyI:C and CpG) that can induce acute inflammation and antitumor effect. The results showed that intratumoral (i.t) injection of TLR3/9ligands (PolyI:C/CpG) during the effector phase effectively rescued the function of lentivector (lv)-activated CD8TILs, decreased the Treg ratio in tumor lesions, and markedly improved the antitumor efficacy of lv immunization. It was proved that the rescue of TILs’ effector function was dependent on type-1IFN that was induced in situ in the tumor lesion by hgp100-w(i.t.) or PolyI:C/CpG (i.t.). But i.t injection of PolyI:C/CpG failed to restore the effector function of lv primed tumor-infiltrating Pmel-1cells in the IFNαPR-/-mice and resulted in similar antitumor effect as lv immunized group. In addition, PolyI:C/CpG can mature and activate tumor infiltrating dendritic cells (TIDCs). For example, the TIDCs expressed higher level of mature markers such as CD40, and CD86, and they were more potent to induce T cell proliferation. Moreover, the activated TIDCs restored the effector function by re-activating hgp100-lv primed CD8T cells in situ and increased the antitumor efficacy.In this part, we proved that i.t injection of vv or PolyI:C/CpG during the effector phase could rescue the function of lv-activated CD8TILs, decreased the Treg ratio in the tumor lesion, and markedly improved the antitumor efficacy of lv immunization.Taken together, lv-vv prime-boost immunization strategy can provoke stronger immune responses, by inducing more potent CD8+T cell proliferation and larger numbers of memory CD8T cells, which can rapidly sense and eliminate the tumor cells to result in effective tumor prevention. In the therapeutic setting, i.t injection of w or PolyI:C/CpG during the effector phase can effectively rescue the function of lv-activated CD8TILs, decrease the Treg ratio in the tumor lesion, and markedly improve the anti-solid tumor efficacy resulted from lv immunization. The whole study has provided new strategy and insight for cancer immunotherapy based tumor vaccine. It is the first time to prove that the memory CD8T cells induced by lv-vv prime-boost can rapidly sense and eliminate the tumor cells as well as effectively prevent tumors growth. The preventive protection can protect the mice for at least one year, which is much more persistent than that induced by other methods. We also found that CD4T cell help is required for CD8secondary response during hgp100-lv immunization, which is different from the prevailing view. It is the first report to prove that the acute inflammation in tumor lesions can rescue the CD8T cell effector function and improve the antitumor efficacy.