Enhancing vaccine antibody responses by targeting Clec9A on dendritic cells

Abstract

Targeting model antigens (Ags) to Clec9A on DC has been shown to induce, not only cytotoxic T cells, but also high levels of Ab. In fact, Ab responses against immunogenic Ag were effectively generated even in the absence of DC-activating adjuvants. Here we tested if targeting weakly immunogenic putative subunit vaccine Ags to Clec9A could enhance Ab responses to a level likely to be protective. The proposed "universal" influenza Ag, M2e and the enterovirus 71 Ag, SP70 were linked to anti-Clec9A Abs and injected into mice. Targeting these Ags to Clec9A greatly increased Ab titres. For optimal responses, a DC-activating adjuvant was required. For optimal responses, a boost injection was also needed, but the high Ab titres against the targeting construct blocked Clec9A-targeted boosting. Heterologous prime-boost strategies avoiding cross-reactivity between the priming and boosting targeting constructs overcame this limitation. In addition, targeting small amounts of Ag to Clec9A served as an efficient priming for a conventional boost with higher levels of untargeted Ag. Using this Clec9A-targeted priming, conventional boosting strategy, M2e immunisation protected mice from infection with lethal doses of influenza H1N1 virus.

Fig. 1

Targeting vaccine Ags to Clec9A generates long-lasting enhanced Ab responses. a Schematic representation of rat anti-Clec9A-Ag 10B4 constructs b Schematic representation of the vaccination strategy. C57BL/6 mice were injected i.v. with 2 μg anti-Clec9A-M2e mAb construct (clone 10B4), 2 μg isotype control-M2e mAb construct (clone GL117), 2 μg anti-Clec9A-SP70 mAb construct, or 2 μg isotype control-SP70 mAb construct in the absence or presence of adjuvant, 5 nmol CpG-1668. A group of mice were injected with PBS as unprimed control. At day 42 after priming, all groups of mice were boosted with 2 μg anti-Clec9A-M2e mAb construct or with 2 μg anti-Clec9A-SP70 mAb construct. Serum samples were collected at the indicated times post immunisation and Ag-specific IgG responses measured by ELISA. c Targeted versus (vs.) untargeted anti-M2e responses. d Targeted vs. untargeted anti-SP70 responses. e The anti-rat Ig response associated with c. f The anti-rat Ig response associated with d. Each point represents the geometric mean end point titre with 95% confidence interval (CI) from a group of five mice. The dotted line in c and d represents a “maximal” response obtained in separate experiments by injection of C57BL/6 mice with 50 μg KLH-Ag in Freund’s complete adjuvant followed by boosting with 50 μg KLH-Ag in Freund’s incomplete adjuvant and sampling the blood 30 days later. This full kinetic experiment was carried out once (n = 5 mice per group) but the results were confirmed at restricted time points in two other experiments

Fig. 2

Responses to a single injection of Clec9A-targeted vaccine Ag: mouse strain variation, effect of adjuvant and resistance to infection. a, b Ab responses. Groups of C57BL/6 (n = 5) and BALB/c (n = 5) mice were immunised as in Fig. 1b. Serum samples were collected 2 weeks post immunisation and a the anti-M2e IgG response or b the anti-SP70 IgG response was measured by ELISA. Each point represents one individual mouse and the end point titre is shown as geometric mean with 95% CI. This side by side comparison was performed once but the individual strain responses were confirmed in multiple experiments. Data were analysed by an unpaired two-tailed Student's t-test, and significance was indicated as *p < 0.05, **p < 0.01. ns represents no significant differences. c Body weight following influenza infection. BALB/c mice (n = 10 mice per group) were immunised with 2 μg anti-Clec9A-TpD-M2e construct adjuvanted with 5 nmol of CpG (i.v.). An unimmunised control group was injected with 100 μl PBS. One month (30 days) later they were challenged by intratracheal administration of a lethal dose (400 PFU) of H1N1/PR8 virus. Body weight was recorded daily and monitored over a period of 14 days post infection. Mice were euthanised when they lost 30% or more of body weight. All mice in the unimmunised group were euthanised by day 8. One mouse in the immunised group was euthanised at day 8. Results are means with SEM. Statistical analysis was performed using two-way ANOVA analysis, followed by multiple comparisons with significance indicated as *** p < 0.001, **** p < 0.0001

Fig. 3

Effects of pre-injection with constructs on the persistence of the anti-Clec9A-Ag constructs in the serum and on Ab responses. a, b Serum persistence. C57BL/6 mice were injected i.v. with a 2 μg rat anti-Clec9A-M2e mAb construct or b 2 μg rat anti-Clec9A-SP70 mAb construct, both constructs being made using the rat mAb clone 10B4. At day 28 after priming, all groups of mice were given a second injection of 1 μg anti-Clec9A-M2e or 1 μg anti-Clec9A-SP70, respectively. Serum samples were collected at the indicated times post immunisation and the concentration of the rat Ig mAb in the serum was measured by ELISA. Open circles represent the values from the primary response, closed circles the values from the boost response. Each point represents pooled mean with SEM from five mice. c Ab responses following pre-injection with anti-Clec9A mAb only. BALB/c mice (n = 5 mice per group) were injected i.v. with 100 μl PBS for the primary response group or with 2 μg anti-Clec9A mAb (clone 10B4, not linked to M2e) for the Ab pre-injected group. Two weeks later, both groups of mice were given an injection of 2 μg anti-Clec9A-M2e mAb construct. A further 2 weeks later, serum samples were collected and the anti-M2e IgG response was measured by ELISA. Each point represents one individual mouse and the end point titre is shown as geometric mean with 95% CI. Results represent a single experiment. Data were analysed by an unpaired two-tailed Student's t-test, and significance was indicated as *p < 0.05

Fig. 4

Priming and heterologous boosting of BALB/c mice using human–mouse chimeric anti-Clec9A Ab. a Schematic representation of the two new chimeric human–mouse anti-Clec9A-TpD-M2e constructs prepared from phage display Abs, compared with the rat anti-Clec9A-TpD-M2e construct. b Schematic representation of the experimental setup. BALB/c mice (n = 5 mice per group) were injected i.v. with 2 μg human–mouse chimeric anti-Clec9A-TpD-M2e (clones #2 or #19) or 2 μg rat anti-Clec9A-TpD-M2e (clone 10B4) in the presence of 5 nmol CpG. An unprimed control group was injected with 100 μl PBS. At day 28 after priming, each group of mice primed with the human–mouse chimeric Ab constructs were boosted with 2 μg rat anti-Clec9A-TpD-M2e (clone 10B4) and the groups of mice primed with the rat anti-Clec9A construct boosted with 2 μg human–mouse chimeric anti-Clec9A-TpD-M2e (clones #2 or 19) in the presence of 5 nmol CpG. Unprimed control groups were likewise injected at day 28 to give a primary response for direct comparison. c Serum samples were collected at the indicated time points post immunisation and the anti-M2e IgG response was measured by ELISA. Each point represents one individual mouse and the end point titre is shown as geometric mean with 95% CI. This experiment was performed twice and the results pooled (n = 10 mice per group). The boost responses were analysed by unpaired two-tailed Student's t-test and significance was indicated as ** p < 0.01, ***p < 0.001, ****p < 0.0001. The differences of Ab response between day 28 and at day 42 were indicated in Fig. 4c and the differences of Ab response between primed-boosted groups and unprimed-boosted groups at day 42 are followed; red open circle vs. black open triangle (**p = 0.0012), blue open circle vs. black open triangle (****p < 0.0001), red closed circle vs. red open triangle (***p = 0.0008), and blue closed circle vs. blue open triangle (**p = 0.0024)

Fig. 5

Ab responses generated by targeting M2e to Clec9A can be boosted using conventional, high-dose untargeted M2e complexes. a Schematic representation of the rat anti-Clec9A-TpD-M2e (clone 10B4) construct used for priming. b Schematic representation of the experimental setup. BALB/c mice (n = 5 mice per group) were injected i.v. with 0.2 μg or 2 μg anti-Clec9A-TpD-M2e in the presence of 5 nmol CpG. An unprimed control group was injected with 100 μl PBS. At day 28 after priming, each group of mice was boosted with either 10 μg of KLH-M2e or TpD-M2e, in the presence of 5 nmol CpG. c Serum samples were collected at the indicated times post immunisation and the anti-M2e IgG response was measured by ELISA. Each point represents one individual mouse and the end point titre is shown as geometric mean with 95% CI. Results represent a single experiment. The data with KLH-M2e boost were confirmed in a second experiment. Data were analysed by an unpaired two-tailed Student's t-test, and significance was indicated as **p < 0.01, ****p < 0.0001

Fig. 6

Immunisation with M2e via Clec9A-targeted priming then conventional boosting protects mice against lethal influenza challenge. a Schematic representation of the experimental setup. BALB/c mice (n = 8–10 mice per group) were primed with 2 μg anti-Clec9A-M2e construct or the non-targeting isotype control-M2e construct, adjuvanted with 5 nmol of CpG (i.v.), then boosted 42 days later with 20 μg M2e-KLH adjuvanted with 2.5 nmol of CpG accordingly to the schedule shown. An unimmunised control group was injected with 100 μl PBS. One month (30–31 days) later they were challenged by intratracheal administration of a lethal dose (400 PFU) of H1N1/PR8 virus. b Body weight following influenza infection. Body weight was recorded daily and monitored over a period of 14 days post infection. Mice that lost 30% or more of their body weight were euthanised; by day 8 this amounted to 60% of the unimmunised group, 20% of the isotype control primed group, but none of the Clec9A-targeted primed group. Results are means with SEM of the surviving mice. Statistical analysis was performed using two-way ANOVA. (anti-Clec9A-M2e + CpG vs. Isotype control-M2e + CpG), and significance was indicated as **p < 0.01, ***p < 0.001. This result represents a single experiment, but two additional experiments gave equivalent protection results. c Lung viral titres. Lungs were harvested on days 3 and 5 post challenge with influenza virus. Influenza virus titres were determined by the plaque assay. Results represent a single experiment, initiated with n = 5 mice per group. However, one mouse of the unimmunised control group and one mouse of the isotype control group died by day 3; all targeted mice survived. Error bars represent mean with SD of surviving mice. Data were analysed by Mann–Whitney test and, significance was indicated as *p < 0.05, **p < 0.01