Transgenic 4-1BBL-engineered vaccine stimulates potent Gag-specific therapeutic and long-term immunity via increased priming of CD44⁺CD62L^high IL-7R⁺ CTLs with up- and downregulation of anti- and pro-apoptosis genes

Abstract

Human immunodeficiency virus type-1 (HIV-1)-specific dendritic cell (DC) vaccines have been used in clinical trials. However, they have been found to only induce some degree of immune responses in these studies. We previously demonstrated that the HIV-1 Gag-specific Gag-Texo vaccine stimulated Gag-specific effector CD8⁺ cytotoxic T lymphocyte (CTL) responses, leading to completely protective, but very limited, therapeutic immunity. In this study, we constructed a recombinant adenoviral vector, adenovirus (AdV)_4-1BBL, which expressed mouse 4-1BB ligand (4-1BBL), and generated transgenic 4-1BBL-engineered OVA-Texo_/4-1BBL and Gag-Texo_/4-1BBL vaccines by transfecting ovalbumin (OVA)-Texo and Gag-Texo cells with AdV_4-1BBL, respectively. We demonstrate that the OVA-specific OVA-Texo_/4-1BBL vaccine stimulates more efficient OVA-specific CTL responses (3.26%) compared to OVA-Texo-activated responses (1.98%) in wild-type C57BL/6 mice and the control OVA-Texo_/Null vaccine without transgenic 4-1BBL expression, leading to enhanced therapeutic immunity against 6-day established OVA-expressing B16 melanoma BL6-10_OVA cells. OVA-Texo_/4-1BBL-stimulated CTLs, which have a CD44⁺CD62L^high IL-7R⁺ phenotype, are likely memory CTL precursors, demonstrating prolonged survival and enhanced differentiation into memory CTLs with functional recall responses and long-term immunity against BL6-10_OVA melanoma. In addition, we demonstrate that OVA-Texo_/4-1BBL-stimulated CTLs up- and downregulate the expression of anti-apoptosis (Bcl2l10, Naip1, Nol3, Pak7 and Tnfrsf11b) and pro-apoptosis (Casp12, Trp63 and Trp73) genes, respectively, by RT² Profiler PCR array analysis. Importantly, the Gag-specific Gag-Texo_/4-1BBL vaccine also stimulates more efficient Gag-specific therapeutic and long-term immunity against HLA-A2/Gag-expressing B16 melanoma BL6-10_Gag/A2 cells than the control Gag-Texo_/Null vaccine in transgenic HLA-A2 mice. Taken together, our novel Gag-Texo_/4-1BBL vaccine, which is capable of stimulating potent Gag-specific therapeutic and long-term immunity, may represent a new immunotherapeutic vaccine for controlling HIV-1 infection.

Introduction

CD8⁺ cytotoxic T lymphocytes (CTLs) are considered an important immune component for effective immunity against human immunodeficiency virus type 1 (HIV-1), and the induction of such responses using vaccines has become a major objective in the strategy to halt the pandemic.^1,2 Dendritic cells (DCs), the most potent antigen-presenting cells, which express the HIV-1 structure proteins gp120 and Gag, have been used as vaccines and have been shown to stimulate HIV-1-specific CTL responses in animal models.^3,4,5,6 However, HIV-1-specific DC vaccines have been found to only induce some degree of immune responses in clinical trials,⁷ warranting the search for other more efficient vaccine strategies.

We previously developed a novel T cell-based vaccine (ovalbumin (OVA)-Texo) using ConA-stimulated CD4⁺ T cells with the uptake of OVA-specific DC-released exosomes (EXOs).^8,9,10 We demonstrated that the OVA-Texo vaccine induces more efficient immunity than the DC vaccine and is capable of stimulating potent CD4⁺ T cell-independent CTL responses and long-term antitumor immunity via IL-2/CD80 and CD40L signaling, counteracting regulatory T cell-mediated immune suppression.^8,9,10 In addition, we also developed HIV-1 gp120- and Gag-specific T cell-based (gp120-Texo and Gag-Texo) vaccines using ConA-stimulated mouse CD8⁺ T cells for the uptake of gp120- and Gag-specific DC-released EXOs.^11,12,13 We demonstrated that gp120-Texo- and Gag-Texo-stimulated gp120- or Gag-specific CTL responses developed in transgenic HLA-A2 mice.^11,12,13 However, the therapeutic efficacy of these vaccines was still very limited. For example, the gp120-Texo vaccine only cured 2/8 transgenic HLA-A2 mice bearing 6-day-established HLA-A2/gp120-expressing BL6-10_Gp120/A2 B16 melanoma, although the vaccine cured 8/8 HLA-A2 mice bearing a 3-day-established tumor.^11,12,13 The Gag-Texo vaccine could only induce some degree of therapeutic immunity, resulting in a decreased number and size of 6-day-established HLA-A2/Gag-expressing BL6-10_Gag/A2 B16 melanomas in transgenic HLA-A2 mice.¹³

4-1BB ligand (4-1BBL) is a member of the tumor necrosis factor family.¹⁴ It is inducible on activated antigen-presenting cells and can provide a CD28-independent signal leading to cell division, the induction of effector function and the enhancement of CD8⁺ T-cell survival and memory development.^{15,16,17,18,19} 4-1BBL synergizes with CD80 and PD-1 blockade to re-activate anergic T cells²⁰ and to augment CTL responses during chronic viral infection.²¹ In addition, 4-1BBL signaling is also a critical component in the costimulation-dependent rescue of exhausted HIV-specific CTLs,²² and the combination of 4-1BBL and CD40L signaling enhances the stimulation of HIV-1-specific CTLs.²³ Therefore, 4-1BBL becomes an attractive candidate signaling molecule to improve the efficacy of immunotherapy.¹⁹ We assume that the incorporation of 4-1BBL into our Gag-Texo vaccine may enhance its therapeutic effect against Gag-expressing tumor cell challenge.

In this study, we generated transgenic 4-1BBL-engineered OVA-Texo_/4-1BBL and Gag-Texo_/4-1BBL vaccines and assessed the OVA- and Gag-specific CTL responses and therapeutic and long-term immunity against OVA- and Gag-expressing B16 melanoma cells in vaccinated wild-type C57BL/6 and -transgenic HLA-A2 mice, respectively.

Materials and methods

Reagents, cell lines and animals

Biotin-labeled or fluorescein isothiocyanate (FITC)-labeled antibodies (Abs) were obtained from BD Biosciences (Mississauga, Ontario, Canada). The Gag/HLA-A2-expressing BL6-10_Gag/A2 tumor cell line was generated by transfecting BL6-10 tumor cells with two expression vectors, pcDNA_NeoGag and pcDNA_HygroHLA-A2, that express Gag and HLA-A2, respectively.¹³ Female C57BL/6 and transgenic (Tg) HLA-A2 mice (#003475) carrying the transgene Tg(HLA-A2.1)1Enge were obtained from the Jackson Laboratory (Bar Harbor, MA, USA). All mice were treated according to the animal care committee guidelines of the University of Saskatchewan.

Recombinant adenovirus construction

The construction of recombinant adenovirus (AdV) that expresses 4-1BBL (AdV_4-1BBL) was performed by inserting the mouse 4-1BBL gene cloned from bone-marrow-derived DCs into a pShuttle vector (Stratagene Inc., La Jolla, CA, USA), using the cloned 4-1BBL cDNA to form pLpA_4-1BBL that expresses the 4-1BBL gene. The PmeI-digested shuttle vector was then co-transformed into BJ5183 Escherichia. coli cells already containing the backbone vector for homologous recombination to form the recombinant vector AdV_4-1BBL, as described previously (Figure 1a).¹³ The adenoviral vector AdV_Gag, which expresses HIV-1 Gag, and the control AdV_Null, without any transgenic expression, were previously constructed in our laboratory (Figure 1a).¹³

Figure 1.

Phenotypic analysis of OVA-Texo/4-1BBL. (a) Schematic representation of the AdV vectors AdVNull, AdVGag and AdV4-1BBL. The E1/E3-depleted, replication-deficient AdV is under the regulation of the CMV early/immediate promoter/enhancer. (b) The engineered OVA-Texo/4-1BBL and the control OVA-Texo/Null and OVA-Texo cells were stained with a panel of Abs (solid lines) or isotype-matched irrelevant Abs (dotted lines) and then analyzed by flow cytometry. One representative experiment of two is shown. Ab, antibody; AdV, adenoviral; 4-1BBL, 4-1BB ligand; CMV, cytomegalovirus; ITR, inverted terminal repeat; OVA, ovalbumin.

Preparation of engineered OVA-Texo_/4-1BBL and Gag-Texo_/4-1BBL vaccines

OVA- and Gag-specific OVA-Texo and Gag-Texo vaccines were generated by incubating ConA-activated CD8⁺ T (ConA-T) cells with OVA protein-pulsed bone marrow-derived dendritic cells (DC_OVA)- and AdV_Gag-transfected DC (DC_Gag)-released exosomes (EXO_OVA and EXO_Gag), as previously described.¹³ CD8^± OVA-Texo_/4-1BBL and Gag-Texo_/4-1BBL or the control CD8^± OVA-Texo_/Null and Gag-Texo_/Null vaccines were then generated by transfecting CD8^± OVA-Texo and Gag-Texo cells with AdV_4-1BBL or the control AdV_Null, as previously described.¹³

Flow cytometry analysis

To assess CTL responses, blood samples from C57BL/6 mice i.v. immunized with OVA-Texo_/4-1BBL or OVA-Texo_/Null cells (2×10⁶ cells/mouse) were taken 6 or 10 days after the immunization and were doubly stained with FITC-conjugated anti-CD8 Ab (FITC-CD8) and PE-conjugated H-2K^b/OVA_257–264 tetramer (PE-Tetramer) or triply stained with FITC-CD8 Ab, PE-tetramer and PE-Cy5-conjugated anti-CD44, CD62L and IL-7R Abs, respectively, and then analyzed by flow cytometry.^8,9,10 To assess the intracellular expression of granzyme B and IFN-γ, splenocytes from immunized mice were stimulated with 2 µM OVAI (SIINFEKL) peptide in vitro in the presence of 2 µM monensin (GolgiStop, BD Biosciences, San Diego, CA) for 4 h and then stained with a FITC-anti-CD8 antibody. The cells were then fixed, and the cell membranes were permeabilized in Cytofix/Cytoperm solution (BD Biosciences, San Diego, CA) and stained with a PE-anti-granzyme B antibody or a PE-anti-IFN-γ antibody for flow cytometry analysis.²⁴ To assess recall responses, immunized mice were i.v. boosted with DC_OVA (1×10⁶ cells/mouse) 30 days after immunization. CTL responses were analyzed by flow cytometry. The absolute numbers of OVA-specific tetramer-positive CD8^± T cells in each spleen of the immunized mice during the primary and recall responses were calculated as previously described.²⁴

Cytotoxicity assay

The in vivo cytotoxicity assay was performed in immunized mice with the transfer of both OVA-specific CFSE^high-labeled (H) and the control irrelevant CFSE^low-labeled (L) target splenocytes at a ratio of 1∶1 (each 1×10⁶ cells), as previously described.¹³ Sixteen hours after cell transfer, the residual CFSE^high and CFSE^low target cells remaining in the recipients' spleens were analyzed by flow cytometry.

RT² profiler PCR array analysis

T cells were enriched from splenocytes from C57BL/6 mice immunized with OVA-Texo_/4-1BBL and OVA-Texo_/Null vaccines using a nylon wool-column.^8,9,10 OVA-specific CTLs were then purified from the enriched T-cell population using PE-tetramer staining followed by anti-PE microbeads (Miltenyi Biotec).²⁵ A highly purified population of OVA-specific CTLs was obtained by passing the anti-PE microbeads-bound cell sample through a MACS column. The expression of a pathway-focused panel of 84 genes related to apoptosis in the above two groups of CTLs was examined using RT² Profiler PCR Array Mouse Apoptosis (SuperArray Bioscience, Betheoda, MD).²⁵

Animal studies

To examine the therapeutic antitumor immunity conferred by the Gag-Texo_/4-1BBL vaccine, the wild-type C57BL/6 and transgenic HLA-A2 mice (n=8) were first injected i.v. with 0.5×10⁶ BL6-10_OVA and BL6-10_Gag/A2 cells. Six days after tumor cell inoculation, C57BL/6 and HLA-A2 mice were then injected i.v. with OVA-Texo_/4-1BBL and Gag-Texo_/4-1BBL cells (2×10⁶ cells/mouse), respectively. OVA-Texo_/Null and Gag-Texo_/Null cells (2×10⁶ cells/mouse) were used as vaccine controls. To assess the long-term antitumor immunity, C57BL/6 mice immunized with the OVA-Texo_/4-1BBL vaccine and HLA-A2 mice immunized with the Gag-Texo_/4-1BBL vaccine were i.v. challenged with a large dose of BL6-10_OVA cells (1×10⁶ cells/mouse) and a regular dose of BL6-10_Gag/A2 cells (0.5×10⁶ cells/mouse), respectively, 30 days after immunization. The mice were killed 3 weeks after tumor cell injection, and the lung metastatic tumor colonies were counted in a blind fashion. Metastases on freshly isolated lungs appeared as discrete black-pigmented foci that were easily distinguishable from normal lung tissues and confirmed by histological examination. Metastatic foci too numerous to count were assigned an arbitrary value of >300.⁹

Statistical analyses

Statistical analyses were performed using Student's t-test or the Mann–Whitney U test for the comparison of variables from different groups. A value of P<0.05 was considered to be statistically significant.¹⁰

Results

Generation of AdV_4-1BBL-infected OVA-Texo (OVA-Texo_/4-1BBL) vaccine expressing transgenic 4-1BBL

We first constructed a recombinant 4-1BBL-expressing adenoviral vector, AdV_4-1BBL, by recombinant DNA technology (Figure 1a). We generated the OVA-Texo vaccine using ConA-stimulated CD8⁺ T cells to take up OVA-pulsed DC (DC_OVA)-released EXOs (EXO_OVA).¹¹ We then generated the transgenic 4-1BBL-engineered and control OVA-Texo vaccines (OVA-Texo_/4-1BBL and OVA-Texo_/Null) by transfecting OVA-Texo cells with AdV_4-1BBL and the control AdV_Null, respectively, followed by a phenotypic assessment of OVA-Texo_/4-1BBL and OVA-Texo_/Null by flow cytometry. We demonstrated that OVA-Texo_/4-1BBL cells expressed a comparable amount of T-cell CD8, CD25, CD28, CD44, CD40L, CD62L, IL-7R, inhibitory PD-1 and exosomal costimulatory CD80, but an enhanced amount of costimulatory 4-1BBL compared to the control OVA-Texo_/Null and OVA-Texo cells (Figure 1b), indicating that the OVA-Texo_/4-1BBL cells express the transgene-encoded cell-surface 4-1BBL.

OVA-Texo_/4-1BBL stimulates enhanced OVA-specific CD8⁺ effector CTL responses in wild-type C57BL/6 mice

To assess whether the OVA-Texo_/4-1BBL vaccine stimulates enhanced CTL responses, we immunized wild-type C57BL/6 mice with OVA-Texo_/4-1BBL and then assessed the OVA-specific CTL responses using FITC-CD8 and PE-tetramer antibody staining by flow cytometry. We demonstrated that the OVA-Texo_/4-1BBL vaccine was able to more efficiently stimulate OVA-specific CTL responses (3.26%, equivalent to 0.33×10⁶ cells per spleen²⁴) than the OVA-Texo vaccine (1.98%, equivalent to 0.20×10⁶ cells per spleen) (P<0.05) (Figure 2a). To assess their phenotypes, the splenocytes of immunized mice were stimulated in vitro with OVAI peptide, and the CD8^± T cells were examined for intracellular IFN-γ and Granzyme B expression using flow cytometry. We demonstrated that 2.69% and 1.98% of IFN-γ- and Granzyme B-expressing CD8^± T cells, respectively, were found in the OVA-Texo_/4-1BBL-immunized mice (Figure 2b), which are more than those in either the OVA-Texo- or the OVA-Texo_/Null-immunized mice. Next, we assessed the ability of the OVA-Texo_/4-1BBL vaccine to induce the differentiation of stimulated CD8⁺ T cells into effector CTLs. We adoptively transferred OVA_257–264 peptide-pulsed splenocytes that had been strongly labeled with CSFE (CFSE^high), as well as the control peptide Mut1-pulsed splenocytes that had been weakly labeled with CFSE (CFSE^low), into recipient mice that had been vaccinated with OVA-Texo_/4-1BBL or OVA-Texo_/Null. Thus, the loss of CFSE^high target cells represents the killing activity of CTLs in immunized mice. As expected, there was substantial loss (90%) of the CFSE^high (OVA peptide-pulsed) cells in the OVA-Texo_/4-1BBL-immunized mice, whereas significantly less cytotoxicity (77%) was induced in mice immunized with the control OVA-Texo_/Null vaccine (P<0.05) (Figure 2c), indicating that the OVA-Texo_/4-1BBL vaccine efficiently stimulates OVA-specific CD8⁺ T-cell differentiation into functional effector CTLs with OVA-specific cytotoxicity.

Figure 2.

OVA-Texo_/4-1BBL stimulates potent OVA-specific effector CTL responses and therapeutic immunity. (a) C57BL/6 mice (3–4 per group) were immunized with OVA-Texo_/4-1BBL, OVA-Texo_/Null and OVA-Texo. Six days after immunization, tail blood samples from the immunized mice were stained with FITC-anti-CD8 Ab and PE-tetramer and then analyzed by flow cytometry. The value in each panel represents the percentage of OVA-specific CD8⁺ T cells in the total CD8⁺ T-cell population. The value in parentheses represents the s.d. *P<0.05 versus cohorts of the control OVA-Texo_/Null and OVA-Texo groups (Student's t-test). (b) Intracellular staining. To assess the intracellular expression of granzyme B and IFN-γ, splenocytes from immunized mice were stimulated with OVAI peptide in vitro in the presence of monensin, followed by FITC-anti-CD8 antibody staining. After permeabilization, the cells were further stained with a PE-anti-granzyme B antibody or a PE-anti-IFN-γ antibody. CD8-positive T cells were gated for the assessment of granzyme B and IFN-γ expression by flow cytometry. (c) In vivo cytotoxicity assay. Six days after immunization, the immunized mice (3–4 per group) were i.v. injected with a mixture of CFSE^high- and CFSE^low-labeled splenocytes (at a 1∶1 ratio) that had been pulsed with OVA_257–264 (SIINFEKL) peptide and the control Mut1 (FEQNTAQP) peptide of an irrelevant 3LL lung carcinoma antigen, respectively. After 16 h, the spleens of immunized mice were removed, and the percentages of the residual CFSE^high (H) and CFSE^low (L) target cells remaining in the recipients' spleens were analyzed by flow cytometry. The value in each panel represents the percentage of CFSE^high vs. CFSE^low target cells remaining in the spleen. The value in parentheses represents the s.d. *P<0.05 versus cohorts of the control OVA-Texo_/Null and OVA-Texo groups (Student's t-test). (c) C57BL/6 mice (eight per group) were first i.v. injected with BL6-10_OVA tumor cells (0.5×10⁶ cells/mouse) followed by immunization of the mice with OVA-Texo_/Null and OVA-Texo_/4-1BBL (2×10⁶ cells/mouse) for the assessment of therapeutic immunity against 6-day-established tumors. The mice were killed 3 weeks after tumor cell challenge. The numbers of black lung metastatic tumor colonies were counted. *P<0.05 versus cohorts of OVA-Texo_/4-1BBL group (Mann–Whitney U test). (d) H&E staining of the lung tissues. The lung tissues of immunized mice were fixed in 10% neutral-buffered formalin and then embedded in paraffin. Tissue sections were stained with H&E and examined by microscopy. Tumor colonies are marked with arrows. Magnification, ×150. One representative experiment of two is shown. Ab, antibody; CTL, cytotoxic T lymphocyte; H&E, hematoxylin and eosin; OVA, ovalbumin; s.d., standard deviation.

The OVA-Texo_/4-1BBL vaccine stimulates enhanced therapeutic immunity against OVA-expressing BL6-10_OVA melanoma in wild-type C57BL/6 mice

To assess the potential therapeutic effect of OVA-Texo_/4-1BBL, we first challenged wild-type C57BL/6 mice with BL6-10_OVA tumor cells. Six days after tumor cell challenge, mice bearing 6-day-established tumors were then immunized with OVA-Texo_/4-1BBL. We demonstrated that both the OVA-Texo_/Null and the OVA-Texo vaccines were able to significantly reduce the average number of lung tumor metastatic colonies (P<0.05) compared to the control mice with PBS injection, although they failed at curing any mice bearing 6-day-established tumors (Figure 2d). However, the OVA-Texo_/4-1BBL vaccine was able to cure 8/8 mice (Figure 2d), indicating that the OVA-Texo_/4-1BBL vaccine induces an excellent therapeutic immunity against 6-day-established lung tumor metastasis in C57BL/6 mice. The presence and absence of lung tumor metastatic colonies in OVA-Texo_/Null- and OVA-Texo_/4-1BBL-immunized mouse lungs were confirmed by histopathologic analysis (Figure 2e).

OVA-Texo_/4-1BBL-stimulated CD8⁺ T cells are likely CD62L^highIL-7R⁺ memory CTL precursors

To assess the phenotypic characteristics of OVA-Texo_/4-1BBL-stimulated CD8⁺ T cells, we first gated OVA-Texo_/4-1BBL-, OVA-Texo_/Null- or OVA-Texo-stimulated CD8⁺ T cells for further analysis of the expression of CD44, CD62L and IL-7R 10 days after immunization. We demonstrated that all vaccine-stimulated CD8⁺ T cells expressed these molecules 10 days after immunization (Figure 3a). However, OVA-Texo_/4-1BBL-stimulated CD8⁺ T cells had a higher amount of CD62L and IL-7R expression than OVA-Texo_/Null- or OVA-Texo-stimulated CD8⁺ T cells, indicating that they are likely CD62L^highIL-7R⁺ memory CTL precursors.²⁶

Figure 3.

Characterization of OVA-Texo_/4-1BBL-stimulated CTLs. (a) The tail blood samples were harvested from the immunized C57BL/6 mice 10 days after immunization and triply stained with FITC-anti-CD8 and PE-tetramer and PE-Cy5-anti-CD44, CD62L and IL-7RAb for flow cytometric analysis. The value in each panel represents the percentage of OVA-specific CD8⁺ T cells in the total CD8⁺ T-cell population. The value in parentheses represents thes.d. *P<0.05 versus cohorts of the control OVA-Texo_/Null and OVA-Texo groups (Student's t-test). OVA-specific CD8⁺tetramer⁺ CTLs were sorted for the assessment of CD44, CD62L and IL-7R expression by flow cytometry (solid lines). Irrelevant isotype-matched Abs were used as controls (dotted lines). (b) An apoptosis RT-PCR array was performed to compare apoptosis-related gene expression differences between OVA-Texo_/4-1BBL-stimulated and OVA-Texo_/Null-stimulated CTLs. Total RNA was isolated from purified OVA-specific CTLs using an RNeasy extraction kit (Qiagen, Missisauga, Ontario, Canada) and reversely transcribed using an RT² First Strand kit (SuperArray Bioscience). The mRNA expression of each gene in the array system was performed using a StepOnePlus thermocycler (Applied Biosystem, BioRad, Missisauga, Ontario, Canada) and analyzed using Hprt1, Gapdh and β-actin as internal controls in web-based software per the manufacturer's instructions. Only genes with mRNA changes greater than threefold are shown. (c) Recall responses. The tail blood samples were harvested from the immunized C57BL/6 mice 30 days after immunization or 4 days after the boost and then doubly stained with FITC-anti-CD8 and PE-tetramer for flow cytometric analysis. The value in each panel represents the percentage of OVA-specific CD8⁺ T cells in the total CD8⁺ T-cell population. The value in parentheses represents the s.d. *P<0.05 versus cohorts of the control OVA-Texo_/Null and OVA-Texo groups on day 30 after immunization or day 4 after the boost (Student's t-test). One representative experiment of two is shown. Ab, antibody; CTL, cytotoxic T lymphocyte; OVA, ovalbumin; RT-PCR, real-time PCR; s.d., standard deviation.

OVA-Texo_/4-1BBL-stimulated CD8⁺ T cells display up- and downregulation of anti- and pro-apoptosis genes, respectively

To assess the expression of anti- and pro-apoptosis genes, we extracted RNA from mouse OVA-specific CTLs purified from the immunized mouse splenic T-cell population, using an RNeasy extraction kit, 10 days after immunization. Apoptosis real-time PCR arrays were then performed to compare apoptosis-related gene expression between OVA-Texo_/4-1BBL-stimulated and OVA-Texo_/Null-stimulated CTLs using their RNA samples and the RT² Profiler PCR Array Mouse Apoptosis kit. We demonstrated that anti-apoptosis-related genes, such as Bcl2l10 (3.3-fold), Naip1 (14.6-fold), Nol3 (3.0-fold), Pak7 (7.1-fold) and Tnfrsf11b (7.2-fold), were found to be upregulated in OVA-Texo_/4-1BBL-stimulated CTLs compared to OVA-Texo_/Null-stimulated CTLs (Figure 3b). In contrast, pro-apoptosis-related genes, such as Casp12 (−4.5-fold), Trp63 (−4.5-fold) and Trp73 (−14.1-fold), were found to be down-regulated in OVA-Texo_/4-1BBL-stimulated CTLs (Figure 3b). Our results indicate that compared to OVA-Texo_/Null-stimulated CTLs, OVA-Texo_/4-1BBL-stimulated CTLs up- and downregulate the expression of anti- and pro-apoptosis related genes, respectively.

The OVA-Texo_/4-1BBL vaccine stimulates enhanced long-term immunity against OVA-expressing BL6-10_OVA melanoma in wild-type C57BL/6 mice

To assess whether OVA-Texo_/4-1BBL-stimulated CD8⁺ T cells differentiate into memory CD8⁺ T cells, we harvested blood samples from immunized C57BL/6 mice 30 days after the primary immunization, when OVA-Texo_/4-1BBL-stimulated effector CD8⁺ T cells should become memory CD8⁺ T cells. We then stained the blood samples with a FITC-conjugated anti-CD8 Ab and PE-conjugated H-2K^b/OVA_257–264 tetramer and then analyzed the cells by flow cytometry. We demonstrated that CD8⁺ T cells (0.27%) in the OVA-Texo_/4-1BBL-vaccinated group became memory CD8⁺ T cells, which are significantly more than those in the OVA-Texo_/Null-vaccinated group (0.15%) (P<0.05) (Figure 3c). To assess whether these memory CD8⁺ T cells are functional, we then i.v. boosted the immunized mice with DC_OVA cells 30 days after the primary immunization. Four days after the boost, we assessed the OVA-specific CTL responses by flow cytometry. We demonstrated that more recall responses (2.98%, equivalent to 0.30×10⁶ cells per spleen) were observed in the OVA-Texo_/4-1BBL-vaccinated group than in the OVA-Texo_/Null-immunized group (0.73%, equivalent to 0.07×10⁶ cells per spleen) (P<0.05) (Figure 3c), indicating that OVA-specific CTLs in the OVA-Texo_/4-1BBL-vaccinated group increase by fourfold compared with those in the OVA-Texo_/Null-vaccinated group after the boost stimulation. To assess the potential long-term protective immunity, we also i.v. challenged the immunized mice with BL6-10_OVA tumor cells 30 days after the primary immunization, when vaccine-stimulated CTLs should have become memory CTLs. We demonstrated that OVA-Texo_/4-1BBL-immunized, but not OVA-Texo_/Null- or OVA-Texo-immunized, C57BL/6 mice obtained long-term immunity efficiently against challenge with a large dose of BL6-10_OVA tumor cells (1×10⁶/mouse) in 8/8 mice (Figure 4), indicating that the OVA-Texo_/4-1BBL vaccine also induces enhanced long-term immunity against OVA-expressing BL6-10_OVA tumors in C57BL/6 mice.

Figure 4.

The OVA-Texo_/4-1BBL vaccine stimulates efficient long-term immunity in C57BL/6 mice. To assess long-term immunity, C57BL/6 mice (eight per group) were first i.v. immunized with OVA-Texo, OVA-Texo_/Null and OVA-Texo_/4-1BBL (2×10⁶ cells/mouse). Thirty days after vaccination, mice were then i.v. injected with a large dose of OVA-expressing BL6-10_OVA tumor cells (1×10⁶ cells/mouse). The mice were killed 3 weeks after tumor cell challenge. The numbers of black lung metastatic tumor colonies were counted. *P<0.05 versus cohorts of the control PBS group (Mann–Whitney U test). One representative experiment of two is shown.

The Gag-Texo_/4-1BBL vaccine induces enhanced preventive, therapeutic and long-term immunity against Gag/HLA-A2-expressing BL6-10_Gag/A2 melanoma in transgenic HLA-A2 mice

To assess preventive immunity, we first immunized transgenic HLA-A2 mice with Gag-Texo_/4-1BBL or Gag-Texo_/Null followed by challenging the immunized mice with HLA-A2/Gag-expressing BL6-10_Gag/A2 tumor cells 6 days after immunization. We demonstrated that both the Gag-Texo_/4-1BBL and the Gag-Texo_/Null vaccines were able to protect all 8/8 mice from tumor challenge (Figure 5a), indicating that the Gag-Texo_/4-1BBL vaccine stimulates potent preventive immunity. To assess the potential therapeutic effect of Gag-Texo_/4-1BBL, we first challenged transgenic HLA-A2 mice with BL6-10_Gag/A2 tumor cells. Six days after tumor cell challenge, the mice were then immunized with Gag-Texo_/4-1BBL or Gag-Texo_/Null. We demonstrated that the Gag-Texo_/Null vaccine was able to significantly reduce the average number of lung tumor metastatic colonies (P<0.05) compared to the control mice with PBS injection, but it failed at curing any mice bearing 6-day-established tumors (Figure 5b). However, the Gag-Texo_/4-1BBL vaccine was still able to cure 8/8 mice (Figure 5b), indicating that the Gag-Texo_/4-1BBL vaccine stimulates potent therapeutic immunity. Next, we assessed long-term immunity by challenging immunized mice with a regular dose of BL6-10_Gag/A2 tumor cells (0.5×10⁶/mouse) 30 days after immunization. We demonstrated that the Gag-Texo_/4-1BBL vaccine but not the Gag-Texo_/Null vaccine was able to protect all 8/8 mice from tumor challenge 30 days after immunization, although the Gag-Texo_/Null vaccine significantly reduced the average number of lung tumor metastatic colonies (P<0.05) compared to the control mice with PBS injection (Figure 5c), indicating that the Gag-Texo_/4-1BBL vaccine also stimulates efficient long-term immunity in HLA-A2 mice.

Figure 5.

The Gag-Texo_/4-1BBL vaccine stimulates potent immunity in transgenic HLA-A2 mice. (a) Transgenic HLA-A2 mice (eight per group) were first i.v. immunized with Gag-Texo_/Null and Gag-Texo_/4-1BBL (2×10⁶ cells/mouse) followed by the i.v. injection of HLA-A2/Gag-expressing BL6-10_Gag/A2 tumor cells (0.5×10⁶ cells/mouse) 6 days after immunization for the assessment of preventive immunity. (b) Transgenic HLA-A2 mice (eight per group) were first i.v. injected with BL6-10_Gag/A2 tumor cells (0.5×10⁶ cells/mouse) followed by the immunization of the mice with Gag-Texo_/Null and Gag-Texo_/4-1BBL (2×10⁶ cells/mouse) for the assessment of therapeutic immunity against 6-day-established tumors. (c) HLA-A2 mice (eight per group) were first i.v. immunized with Gag-Texo_/Null and Gag-Texo_/4-1BBL (2×10⁶ cells/mouse) followed by the i.v. injection of BL6-10_Gag/A2 tumor cells (0.5×10⁶ cells/mouse) 30 days after immunization for the assessment of long-term immunity. The mice were killed 3 weeks after tumor cell challenge. The numbers of black lung metastatic tumor colonies were counted. *P<0.05 versus cohorts of the control PBS group (Mann–Whitney U test). One representative experiment of two is shown.

Discussion

We previously developed a novel OVA-specific T cell-based vaccine (OVA-Texo) capable of stimulating efficient CTL responses and protective immunity by counteracting regulatory T cell-mediated immune suppression and by inducing long-term antitumor immunity via IL-2/CD80 and CD40L signaling.^8,9,10 To assess its therapeutic effect, C57BL/6 mice bearing 6-day-established OVA-expressing B16 melanoma BL6-10_OVA tumors were immunized with OVA-Texo in this study. We demonstrate that the OVA-Texo vaccine is unable to cure any mice bearing 6-day tumors, although OVA-Texo vaccination significantly reduces the average number of lung tumor colonies compared to the PBS control. To improve its therapeutic efficacy, we generated an engineered OVA-Texo_/4-1BBL vaccine that expresses transgenic 4-1BBL. We demonstrate that the OVA-Texo_/4-1BBL vaccine stimulates more efficient OVA-specific effector CTL responses, leading to potent therapeutic immunity and curing 8/8 C57BL/6 mice bearing 6-day-established BL6-10_OVA melanoma. In addition, we also demonstrate that OVA-Texo_/4-1BBL-stimulated CTLs are more likely to be memory CTL precursors, with preferential differentiation into CD44^±CD62L^highIL-7R^± central memory CTLs.²⁶ Furthermore, OVA-Texo_/4-1BBL-stimulated mice demonstrated enhanced recall responses and more efficient long-term immunity against a large dose challenge of OVA-expressing B16 melanoma BL6-10_OVA cells.

Several apoptosis-related molecules that play key roles in regulating apoptosis in T cells have been identified,^27,28 including the Bcl-2 family, with either anti-apoptotic or pro-apoptotic effects. It has been reported that 4-1BBL signaling promotes the survival of CD8⁺ T cells via 4-1BB-mediated NF-κB activation, leading to the upregulation of Bcl-_XL and Bfl-1,¹⁷ and inhibits T-cell apoptosis via the phosphatidylinositol 3-kinase–AKT/protein kinase B-mediated upregulation of Bcl-_XL and c-FLIP_short.²⁹ Recently, it has been shown that 41BB signaling enhances T-cell survival via the cooperation of tumor-necrosis factor (TNF) receptor-1 (TNFR1) and leukocyte-specific protein-1, leading to the activation of ERK and Bcl-2-interacting mediator of cell-death.³⁰ In this study, we demonstrated that 4-1BBL signaling up- and downregulated the expression of anti-apoptosis (Bcl2l10, Naip1, Nol3, Pak7 and Tnfrsf11b) and pro-apoptosis (Casp12, Trp63 and Trp73) genes, respectively, in CD8⁺ CTLs with prolonged survival. It has also been reported that 4-1BBL can interact with Toll-like receptors (TLRs) to sustain TNF production.³¹ Active CD8⁺ T cells express TLRs, and TLR signaling has been shown to play a role in T-cell survival independently of antigen-presenting cells.^32,33,34 TNF receptor-2 (TNFR2) is also present on active CD8⁺ T cells, and the TNF/TNFR2 interaction is important for sustaining their survival.^35,36 Therefore, it is possible that the transgenic 4-1BBL expression on OVA-Texo_/4-1BBL cells interacts with the TLRs of CD8⁺ T cells, leading to the enhanced TNF production and prolonged survival of activated CD8⁺ T cells via TLR signaling. Enhanced TNF production may also induce the prolonged survival of activated CD8⁺ T cells via its autocrine TNF/TNFR2 loop.^35,36

The Gag vaccine has been shown to stimulate persistent and broader HIV-1-specific CTL responses against conserved Gag epitopes in animal models.^37,38,39,40 HLA-B57 HIV-1-infected individuals have been found to have autologous CTL responses against four conserved Gag epitopes, leading to reduced virus replication and viral control.⁴¹ In addition, clinically, effective CTL responses against Gag, but not other viral antigens, have been found to correlate with a significant suppression of HIV-1 replication in patients.^42,43,44,45 We have recently generated a Gag-specific, exosome-targeted Gag-Texo vaccine and demonstrated that Gag-Texo could stimulate Gag-specific CTL responses, but only induced some degree of therapeutic immunity against Gag/HLA-A2-expressing B16 melanoma BL6-10_Gag/A2 cells in transgenic HLA-A2 mice.¹³ To improve its vaccination efficacy, we incorporated 4-1BBL into the Gag-Texo vaccine to generate an engineered Gag-Texo_/4-1BBL vaccine that expresses transgene 4-1BBL. We demonstrate that the Gag-Texo_/4-1BBL vaccine stimulates potent preventive immunity, leading to the protection of 8/8 transgenic HLA-A2 mice from challenge with BL6-10_Gag/A2, and induces potent therapeutic immunity, leading to curing 8/8 transgenic HLA-A2 mice bearing 6-day-established BL6-10_Gag/A2 tumors. In addition, the Gag-Texo_/4-1BBL vaccine also induces an enhanced long-term immunity against BL6-10_Gag/A2 challenge in transgenic HLA-A2 mice. Because HIV-1-specific CTLs are essential for effective immunity against HIV-1,^1,2 Gag-Texo-stimulated Gag-specific CTLs, which have cytolytic effects against Gag-expressing tumor cells in HLA-A2 mice in this study, should also play important roles in controlling HIV-1 infection by their cytolytic effects against virus-infected T cells or DCs in HIV-1 patients. The assessment of Gag-Texo_/4-1BBL vaccination efficacy at controlling Gag-specific viral infection in wild-type C57BL/6 mice infected by a recombinant replication-competent poxvirus rTTV-luc-gag expressing Gag ⁴⁶ is underway in our laboratory.

In this study, we also demonstrated that the Gag-Texo_/4-1BBL-vaccinated cells express PD-1, which was originally reported to be a T-cell activation antigen ⁴⁷ and was later found to be an inhibitory regulator, promoting immune exhaustion⁴⁸ and suppressing T-cell responses.^49,50 PD-1 blockade with its antibody could restore T-cell function.⁵¹ It has been reported that the HIV-1 early regulatory protein Tat can broaden T-cell responses to HIV-1 envelope proteins.^52,53 To enhance the therapeutic efficacy of the Gag-Texo_/4-1BBL vaccine, the construction and the assessment of a Gag-Tat-Texo_/4-1BBL vaccine expressing both late HIV-1 Gag and early HIV-1 Tat proteins and the examination of Gag-Texo_/4-1BBL vaccinated cells with PD-1 blocked by anti-PD-1 antibody in our transgenic HLA-A2 mouse model are also underway in our laboratory.

Taken together, our novel engineered Gag-Texo_/4-1BBL vaccine, which is capable of stimulating potent therapeutic and long-term immunity, may represent a new immunotherapeutic vaccine for controlling HIV-1 infection.