miRNA-mediated silencing in hepatocytes can increase adaptive immune responses to adenovirus vector-delivered transgenic antigens

Abstract

Adenovirus vectors based on human serotype 5 can induce potent CD8 T cell responses to vector-encoded transgenic antigens. However, the individual contribution of different cell types expressing antigen upon adenovirus vector injection to the generation of antigen-directed adaptive immune responses is poorly understood so far. We investigated the role of hepatocytes, skeletal muscle, and hematopoietic cells for the induction of cellular and humoral immune responses by miRNA-mediated tissue-specific silencing of antigen expression. Using hepatitis B small surface antigen (HBsAg) as the vector-encoded transgene we show that adenovirus vector dissemination from an intramuscular (i.m.) injection site into the liver followed by HBsAg expression in hepatocytes can limit early priming of CD8 T cells and the generation of anti-HBsAg antibody responses. However, hepatocyte-specific miRNA122a-mediated silencing of HBsAg expression overcame these limitations. Early clonal expansion of K(b)/S(190-197)-specific CD8 T cells was significantly enhanced and improved polyfunctionality of CD8 T cells was found. Furthermore, miRNA122a-mediated antigen silencing induced significantly higher anti-HBsAg antibody titers allowing an up to 100-fold vector dose reduction. These results indicate that miRNA-mediated regulation of antigen expression in the context of adenovirus vectors can significantly improve transgene product-directed immune responses. This finding could be of interest for future adenovirus vaccine vector development.

Figure 1

In vitro and in vivo expression profiles of miRNA-regulated adenovirus (Ad) vectors. (a) Schematic representation of miRNA-regulated expression cassettes inserted into E1-deleted adenoviral vectors. (b) miRNA-target sequences specifically downregulate Ad vector-delivered transgenes in vitro. A549, Huh7.5, and K562 cells were transduced with indicated pMOI of EGFP-expressing miRNA-regulated Ad vectors or control vector without mirT (AdEGFP) in a 24-well plate format. EGFP expression was quantified after 24 hours by flow cytometry as (GMean_transduced × %GFP⁺ cells_transduced) − (GMean_untransduced × % GFP⁺ cells_transduced). The mean values (± SD) of one out of three independent experiments are shown. (c) Differentiated C2C12 myotubes were transduced with 2,000 pMOIs of EGFP-expressing miRNA-regulated Ad vectors or control vector (AdEGFP). 48 hours post transduction cell lysates were prepared and EGFP expression was analyzed by western transfer (20 µg protein per lane). (d,e) BALB/c mice (three mice per group) were injected i.v. with reporter gene expressing AdEGFPmirTs. Injected mice were sacrificed 48 hours postinjection and tissues were analyzed for EGFP expression. (d) EGFP expression of the different reporter gene expression vectors in livers was analyzed by western transfer of lysates of isolated hepatocytes 48 hours after i.v. delivery of 2 × 10¹⁰ Ad VP. EGFP expression was strongly reduced in hepatocytes of mice injected with AdEGFPmir122aT. Lysates of two out of three injected mice per vector are shown. (e) EGFP expression in CD11c⁺ cells of hematopoietic lineage is downregulated by mir-142-3p. Isolated blood cells and splenocytes of mice injected i.v. with indicated doses of AdEGFPmirTs were stained for CD11c and analyzed for EGFP expression by flow cytometry. Representative dot plots of blood cells are shown (upper panels). The mean percentages (± SD) of GFP⁺/CD11c⁺ blood cells in gate 2 for two different vector doses are presented in the lower left panel. The mean EGFP expression (± SD) of splenocytes (lower right panel) was calculated as (events GFP⁺CD11c⁺_injected × Gmean of GFP⁺CD11c⁺_injected) − (events GFP⁺CD11c⁺_injected × Gmean of GFP⁺CD11c⁺_uninjected). Significance levels of data were calculated by unpaired Student's t-test. *P < 0,05; **P < 0,01. CMV, cytomegalovirus; contr, control; EGFP, enhanced green fluorescence protein; i.v., intravenously; mirT, microRNA target sequence; pMOI, particle multiplicity of infection; SV, simian virus; UTR, untranslated region; VP, vector particles.

Figure 2

Hepatocytes are the main source of transgenic HBsAg particles after i.v. delivery of adenovirus (Ad) vectors. BALB/c mice were injected intravenously with 2 × 10¹⁰ VP of AdS or AdSmirTs. Blood samples were taken from the tail vein of Ad vector-injected mice at the indicated time-points and S-particle concentrations in sera were measured. The mean serum concentrations (± SD) of three mice per group and time-point are shown. conc., concentration; HBsAg (or: S), hepatits B small surface antigen; rel., relative; VP, vector particles.

Figure 3

miRNA-mediated antigen silencing in hepatocytes can increase early priming of HBsAg-specific CD8 T cells. (a) Kinetics of L^d/S_28–39-specific CD8 T cell frequencies. H-2^d-restricted BALB/c mice were immunized intramuscularly with 10⁸ VP of AdSmirTs. At the indicated time points (8d n = 8, 21d n = 7, 35d n = 12) frequencies of specific IFNγ⁺ CD8 T cells were determined after ex vivo stimulation of splenocytes with L^d/S_28-39 peptide in the presence of brefeldin A. Combined data of two independent experiments at time-point day 8/35 days are shown. Boxplots show the median percentages of IFNγ⁺ CD8 T cells. (b) Kinetics of K^b/S_190–197-specific CD8 T cell frequencies. H-2^b-restricted C57BL/6 mice were immunized intramuscularly with 10⁹ VP of AdSmirTs. At the indicated time points (8d n = 8, 21d/35d n = 7) K^b/S_190–197-specific CD8 T cell numbers in spleen of vaccinated C57BL/6 mice were determined by tetramer staining. Combined data of two out of three independent experiments at time point day 8 are shown. Boxplots show the median percentage of tetramer⁺ CD8 T cells. (c) K^b/S_190–197-specific CD8 T cell frequencies in liver. C57BL/6 mice were immunized intramuscularly with 10⁹ vector particles of AdS or AdSmir122aT. At the indicated time points nonparenchymal liver cells were analyzed for specific CD8 T cells by tetramer staining (seven mice per vector and time point). (d) Liver transduction after intramuscular vector injection. BALB/c mice were injected intramuscularly with 10⁹, 10¹⁰, 2 × 10¹⁰, or 5 × 10¹⁰ VP of AdEGFP in a total volume of 100 µl with 50 µl per limb. 48–72 hours postinjection liver cryosections were analyzed by fluorescence microscopy. Annotated numbers indicate vector genome copies per cell (vg/cell) determined by quantitative PCR. P values are based on two-sided Wilcoxon test. *P < 0.05; **P < 0.01; ***P < 0.001. HBsAg (or: S), hepatitis B small surface antigen; IFN, interferon; n.s., not significant; VP, vector particles.

Figure 4

Immunization with mir-122a-regulated adenovirus (Ad) vector results in a different cytokine phenotype of HBsAg-specific CD8 T cells. (a) Groups of BALB/c mice (n = 7) were vaccinated intramuscularly with 10⁸ VP of AdSmirTs. Twenty- one days postvaccination splenocytes were restimulated ex vivo overnight with L^d/S_28–39 peptide in the presence of brefeldin A with subsequent intracellular cytokine staining of CD8 T cells for IFNγ, TNFα, and IL-2. (b) Groups of C57BL/6 mice (n = 8, two independent experiments) were vaccinated intramuscularly with 10⁹ VP of AdSmirTs. Eight days postvaccination splenocytes were stimulated ex vivo overnight with K^b/S_190–197 peptide in the presence of brefeldin A with subsequent intracellular cytokine staining of CD8 T cells for IFNγ, TNFα, and IL-2. Pie charts show the mean percentages ± SD of multicytokine-secreting CD8 T cells of total HBsAg-specific CD8 T cells (see Figure 3a,b). Significance levels of data were calculated by unpaired Student's t-test. *P < 0.05; **P < 0.01. IFN, interferon; IL, interleukin; TNF, tumor necrosis factor.

Figure 5

Suppression of antigen expression in hepatocytes enhances anti-HBsAg antibody responses of adenovirus (Ad) vector-based genetic vaccines. Female C57BL/6 (a) or BALB/c (b) mice were immunized intramuscularly with 10⁹ (a) or 10⁸ (b) AdSmirT VP (seven mice per group). Blood samples were taken at days 21, 28, and 35 after vaccination. Serum was prepared and analyzed for anti-HBsAg antibody titers. Boxplots show the median of HBsAg-specific antibody titers. P values are based on two-sided Wilcoxon test. For day 35 one out of two (b) or three (a) independent experiments are shown. HBsAg (or: S), hepatitis B small surface antigen; VP, vector particles.

Figure 6

Coinjection of ubiquitously expressing adenovirus (Ad) vector partly counteracts the effects of hepatocyte-silenced Ad vector on antibody titers. Groups of C57BL/6 mice were vaccinated intramuscularly with 10⁹ VP of AdSmir122aT in the right limb and coinjected with 10⁸ VP (n = 5) or 10⁹ VP (n = 5) of AdS in the left limb. Control groups received a single injection of 10⁹ VP of AdSmir122aT (n = 11) or AdS (n = 12). Five weeks postvaccination serum was analyzed for anti-HBsAg antibody titers by enzyme linked immunosorbent assay. Boxplots show the median with 25% and 75% percentiles. P values are based on one-sided Wilcoxon test. HBsAg (or: S), hepatitis B small surface antigen; VP, vector particles.

Figure 7

mir-122a-regulated adenovirus (Ad) vector-induced humoral immune responses at low doses are superior to those of nonregulated Ad vector at higher doses. (a) anti-HBsAg antibody responses to increasing doses of Ad vector. BALB/c mice were immunized intramuscularly with 10⁸ (n = 10), 10⁹ (n = 7–8), or 10¹⁰ (n = 3) VP of AdS or AdSmir122aT, respectively. Thirty- five days postinjection mice were killed and sera were analyzed for anti-HBsAg antibody titers. Combined data of two independent experiments are shown for vector doses of 10⁸ and 10⁹. Boxplots show the median with 25% and 75% percentiles. (b) A subset of each group (n = 3) of the sera in a was analyzed for anti-adenovirus IgG by enzyme- linked immunosorbent assay. Boxplots show the median antibody titers calculated by the reciprocal dilution where a fixed threshold of absorbance is reached. P values are based on two-sided Wilcoxon test. HBsAg (or: S), hepatitis B small surface antigen; n.s., not significant; VP, vector particles.