Delivery of Peptides for Anti-Tumor Immunotherapy

Discussion

Nanotechnology-based delivery of antigens has been developed for immunotherapy due to its dose-sparing and prolonged antigen presentation features.^{[27–31,34–43]} Although we previously demonstrated the feasibility of PLGA-NPs carrying TRP2 peptide and MPLA for cancer vaccination in the B16 melanoma model,^[33] more sophisticated approaches still remain to be developed to enhance anti-tumor effects. Since tumor cells have been reported to often escape from the immunological pressure by losing tumor antigens,^{[10,11,40–42]} simultaneous immunization with multiple antigens could prevent tumor escape from anti-tumor immunity. Indeed, vaccine formulations with multiple tumor antigens have recently been tested in clinical trials in humans, with promising results.^{[14–17,43,44]} There have been several reports on novel vaccine formulations with multiple antigens to enhance anti-tumor immunity in murine systems. For example, McCormick et al. demonstrated that chemical conjugation of multiple CTL peptides, including p15E and TRP2, to the tobacco mosaic virus improved vaccine efficacy in the B16 melanoma model.^[45] Mansour et al. also found that immunization of mice bearing B16 tumors with both TRP2- and P53-derived peptides formulated in a novel liposome-based vaccine delivery platform, VacciMax® (Nova Scotia, Canada), significantly inhibited tumor growth.^[46] However, no information is currently available regarding the NP formulation employing multiple antigen peptides. In the current study, we demonstrated the feasibility of combinational delivery of multiple NPs carrying different antigen epitopes for cancer immunotherapy.

The loading efficiency of peptides in PLGA-NPs is affected by the hydrophilicity/hydrophobicity of peptides. Since hgp100 and p15E are quite hydrophilic, the loading efficiency of these peptides in PVA-emulsified PLGA-NP was too low to induce substantial antigen-specific immune responses through vaccination (data not shown). Even if additives, such as glucose and sodium dodecyl sulfate, were used, the loading efficiencies of hgp100 and p15E were <1 and <2%, respectively. In the current study, therefore, lipid-coated NPs were formulated for encapsulation of these hydrophilic peptides. Since the lipid hydrophilic layer covers the surface of NPs with a core–shell structure in the lipid-coated NPs, hydrophilic peptides may be dispersed in the lipid layer or close to the surface of particles and efficiently loaded to the NPs.^[38] In fact, compared with PVA-emulsified PLGA-NP without lipids, the loading efficiencies of TRP2, p15E and hgp100 were increased from 24.1 ± 3.6, 1.05 ± 0.45 and 0.40 ± 0.35% to 30.1 ± 5.1, 12.1 ± 2.4 and 2.30 ± 0.70%, respectively. It seems that the lipid-coated structure increased the loading efficiencies of hydrophilic peptides of p15E and hgp100 by up to 11.5- and 5.8-fold in NPs. Notably, TRP2, p15E and hgp100 peptides could not be loaded together in the same NP by this procedure, since hydrophilic peptides affected the recovery of other hydrophobic peptides (data not shown). Therefore, the peptides were fabricated and loaded in NPs separately, and then mixed together before immunizations.

The immunized dosages of TRP, hgp100 and p15E in T-NP, G-NP and P-NP were selected as 100 µg, 20 µg and 50 µg, respectively. This was decided due to the T-cell response and loading efficiencies of peptides, as well as literature reports. In our previously published data,^[33] the dose of TRP2 was chosen as 100 µg, since low doses (10 µg or 50 µg) of TRP2 peptide significantly reduced antigen-specific T-cell responses in preliminary experiments. The dose of hgp100 was selected as 20 µg, as previously we used the dosage of 10 µg in adoptive transfer experiment to determine the stimulation of pmel-1-Tg T cells by vaccination with PLGA-NP-containing hgp100.^[33] The dose of p15E was determined as 50 µg. The doses could not be increased because the encapsulation efficiencies of hgp100 and p15E were as low as 2.3 and 12.1%, respectively. The final dispersible concentrations of G-NP and P-NP in PBS were around 0.334 and 1.95 mg/ml, respectively. The volumes of vaccines were a limiting factor for immunization, here we selected as 120 µl for all the immunized samples and thus higher doses of the peptides could not be employed. Considering the TPG-NP group, each mouse was injected with 120 µl PBS solution mixed with 35 µl T-NP (2.86 mg/ml in PBS, 100 µg), 60 µl G-NP (20 µg) and 25 µl P-NP (50 µg). Moreover, the spot number of antigen-specific T-cell response of p15E and hgp100 exhibited a slight decrease while the concentration of peptide decreased from 40 µg (120 µl injection) to 20 µg for G-NP and from 100 µg to 50 µg for P-NP (data not shown).

Combinational delivery of multiple lipid-coated NPs carrying antigenic peptides induced antigen-specific immune responses. However, compared with the NPs carrying each single peptide, combinations of multiple peptide-loaded NPs showed decreased antigen-specific T-cell responses assessed by IFN-γ production. This might possibly be explained by the competition of peptides in binding to MHC molecules and/or other antigen-presenting machineries. For example, the numbers of MHC molecules on cell surfaces of APCs are limited and the capability of each peptide for binding to MHC molecules has been reported to be dependent on the affinity between them; peptides encapsulated in the NPs might compete with each other through combinational delivery.^[47,48] Alternatively, competition of antigen-specific T cells might be related to the decreased antigen-specific T-cell responses in combinational delivery of different peptide-loaded NPs, since, based on the current paradigm, the size and composition of the adaptive immune system might be invariable and individual immune cells might be constantly competing for limited space.^[49]

Although single peptide-loaded NPs can induce a strong T-cell response, no significant suppression of tumor growth was demonstrated in mice that were prophylactically immunized with lipid-coated NPs containing a single antigenic peptide. It may be attributed to the downregulation of MHC I expression on tumor cells as an escape mechanism from antigen-specific T cells.^[33] Tumor cells may escape from peptide-specific immune responses by losing the antigen employed for vaccination. Combinations of multiple different peptide-loaded NPs showed decreased T-cell responses to each of vaccinated antigens, assessed by IFN-γ production, but demonstrated significantly higher anti-tumor effects than the NPs carrying each single peptide. This discrepancy between antigen-specific T-cell responses and tumor suppression might possibly be explained by the following reasons: the T-cell responses to each of vaccinated antigens, which were induced by combinations of different peptide-loaded NPs, were weaker, but might be strong enough to control tumor growth. Furthermore, employment of multiple different epitopes for vaccination could reduce the risk of tumor escape through induction of antigen-loss variants escaping peptide-specific immune responses, since it would be relatively rare that tumor cells escape from peptide-specific immune responses by simultaneously losing all multiple antigens employed for vaccination. Although this study did not examine whether vaccination with NPs carrying a single peptide actually induced antigen-loss variants, selection of antigen-loss variants through immunotherapies has already been reported in the B16-F10 melanoma model.^[50]

In the current study, combinations of multiple NPs carrying antigen peptides still did not completely prevent tumor growth, possibly because tumors possess several different mechanisms to inhibit anti-tumor T-cell responses.^[7–9] For example, tumors have been demonstrated to increase inhibitory immune cell subsets, such as regulatory T cells and myeloid-derived suppressor cells, which inhibit antigen-specific T-cell responses through several different mechanisms.^[51–53] Indeed, in the current study, the numbers of antigen-specific T cells induced by NP-mediated vaccinations were significantly reduced in tumor-bearing mice. To further enhance anti-tumor effects, the precise mechanisms, by which T-cell responses were inhibited, remain to be clarified. In addition, more issues, including the vaccination protocol, such as dose and timing of administration and selection/combination of antigens and adjuvants, also need empirical development to more effectively prevent tumor escape mechanisms for future clinical application.