Alphavirus replicon-based enhancement of mucosal and systemic immunity is linked to the innate response generated by primary immunization

Abstract

Venezuelan equine encephalitis virus replicon particles (VRP) function as an effective systemic, cellular and mucosal adjuvant when codelivered with antigen, and show promise for use as a component in new and existing human vaccine formulations. We show here that VRP are effective at low dose and by intramuscular delivery, two useful features for implementation of VRP as a vaccine adjuvant. In mice receiving a prime and boost with antigen, we found that VRP are required in prime only to produce a full adjuvant effect. This outcome indicates that the events triggered during prime with VRP are sufficient to establish the nature and magnitude of the immune response to a second exposure to antigen. Events induced by VRP in the draining lymph node after prime include robust secretion of many inflammatory cytokines, upregulation of CD69 on leukocytes, and increased cellularity, with a disproportionate increase of a cell population expressing CD11c, CD11b, and F4/80. We show that antigen delivered 24h after administration of VRP does not benefit from an adjuvant effect, indicating that the events which are critical to VRP-mediated adjuvant activity occur within the first 24h. Further studies of the events induced by VRP will help elucidate the mechanism of VRP adjuvant activity and will advance the safe implementation of this adjuvant in human vaccines.

Fig. 1

VRP adjuvant activity is intact when using intramuscular delivery of VRP which lack the 26S promoter. (A) The VEE parental genome contains sequence for four nsPs and a 26S promoter followed by a polyprotein which is processed to produce the capsid and two envelope glycoproteins (C, E1, E2). The VRP genome lacks the structural protein sequences. VRP16M contains 16 nucleotides and a multiple cloning site (MCS) after the 26S promoter, while VRP(−5) lacks the sequence between the nsP4 gene and the 3’UTR, including the 26S start site. (B & C) Balb/c mice were primed and boosted in the footpad with OVA (10 µg) alone or in a coinjection with VRP16M or VRP(−5) (10³ IU, ≈10⁶ GE). (D & E) For analysis of the route of immunization, Balb/c mice were primed and boosted either by footpad or intramuscular injection with OVA (10 µg) in the absence or presence of VRP(−5) (10⁵ IU, ≈10⁸ GE). Levels of anti-OVA IgG in the serum (B & D) and anti-OVA IgA in fecal extracts (C & E) were determined by ELISA. Six mice per group were injected, and values represent the geometric mean +/− SEM. Statistical significance (p<0.05) was determined by t test relative to immunoglobulin levels detected for mice injected with OVA only, unless otherwise indicated. Similar results were achieved in two separate experiments.

Fig. 2

VRP are an effective adjuvant at a wide range of doses. Balb/c mice were primed and boosted i.m. with OVA (10 µg) in the absence or presence of the indicated dose (infectious units) of VRP (1 IU ≈ 10³ GE). After boost, anti-OVA serum IgG (A) and anti-OVA fecal IgA (B) were measured by ELISA. Six mice per group were injected, and values represent the geometric mean +/− SEM. Statistical significance (p<0.05) was determined by t test relative to OVA only.

Fig 3

C57Bl/6 mice were primed and boosted i.m. with OVA (10 µg) alone or in coinjections with the indicated VRP dose (1 IU ≈ 10³ GE). 8 days after boost, spleens were harvested and 4×10⁶ spleen cells incubated with OVA peptide (2 µg/ml) or an irrelevant peptide. Spleens from uninjected mice were assayed as well. After 5 hours, surface CD8 and intracellular IFN-γ were stained and measured by flow cytometry. There was no response to irrelevant peptide. For OVA stimulated cells, IFN-γ-positive CD8 T cells are shown. Six mice per group were injected, and statistical significance (p<0.05) was determined by t test relative to OVA only.

Fig. 4

VRP are required only during prime to induce a mucosal adjuvant effect. Balb/c mice were primed in the footpad with OVA (10 µg) alone or in a coinjection with VRP (10³ IU, ≈10⁶ GE). Mice were boosted in the same way either in the same or contralateral footpad. Other mice were primed and boosted with OVA, but received VRP only in prime, or in boost.After boost, anti-OVA serum IgG (A) and anti-OVA fecal IgA (B) were measured by ELISA. Six mice per group were injected, and values represent the geometric mean +/− SEM. Statistical significance (p<0.05) was determined by t test relative to OVA only, or as indicated. Similar results were achieved in two separate experiments.

Fig. 5

VRP trigger rapid secretion of inflammatory cytokines in the draining lymph node. Balb/c mice were injected in the footpad with OVA (10µg) alone or in the presence of VRP at the indicated dose (1 IU ≈ 10³ GE). After 6 hours the draining popliteal lymph node was harvested, homogenized, and cytokine levels in the homogenate analyzed by multiplex analysis. Values represent the geometric mean +/− SEM. For cytokines which were undetectable, a value of one half of the lower limit of detection was assigned. The maximum limit of detection was 10,000 pg/ml. Cytokines detected at levels above 1000 pg/ml (A) are shown separately from those detected at lower levels (B). Four mice per group were injected, and values represent the geometric mean +/− SEM.

Fig. 6

VRP-induced inflammation in the draining lymph node is dose responsive. Balb/c mice were injected with diluent or VRP at the indicated dose (1 IU ≈ 10³ GE). Draining lymph nodes were harvested after 6 or 24 hours post injection (hpi), homogenized, and cells counted (A). Expression of CD11b, CD11c, F4/80 and CD69 was identified by fluorescent antibody staining and analysis by flow cytometry (B–D). Results were gated on live cells as determined by forward and side scatter, and receptor expression is shown as a percent of total live cells. Four mice per group were injected, and values represent the geometric mean +/− SEM. Statistical significance (p<0.05) was determined by t test relative to diluent. Similar results were achieved in at least two separate experiments.

Fig. 7

VRP-induced inflammation does not affect the mucosal response to antigens delivered 24 hours after VRP. Balb/c mice were primed and boosted with OVA alone or in the presence of VRP (10⁵ IU, ≈10⁸ GE). VRP was delivered at time 0 and OVA was delivered either at time 0, −24 hours, or +24 hours, for both prime and boost. After boost, anti-OVA serum IgG (A) and anti-OVA fecal IgA (B) were measured by ELISA. In (C), mice were injected in the footpad with OVA (0.5 ug) conjugated to Alexa-Fluor-488. In some mice, VRP (10⁵ IU, ≈10⁸ GE) was injected with OVA, and some mice were injected with VRP (10⁵ IU, ≈10⁸ GE) 24 hours before OVA injection. After 6 hours the draining popliteal lymph node was harvested, homogenized, and analyzed by flow cytometry for the presence of OVA-positive cells, based on detection of cells containing Alexa Fluor-488. Six mice per group were injected, and values represent the geometric mean +/− SEM. Statistical significance (p<0.05) was determined by t test relative to OVA only.