Route and method of delivery of DNA vaccine influence immune responses in mice and non-human primates

Background: In spite of the large number of studies that have evaluated DNA -based immunization, few have directly compared the immune responses genera ted by different routes of immunization, particularly in non-human primates . Here we examine the ability of a hepatitis B surface antigen (HBsAg)-enco ding plasmid to induce immune responses in mice and non-human primates (rhe sus monkeys: Macaca mulatta) after delivery by a number of routes. Materials and Methods: Eight different injected [intraperitoneal (IP), intr adermal (ID), intravenous (IV), intramuscular (IM), intraperineal (IPER), s ubcutaneous (SC), sublingual (SL), vaginal wall (VW)I and six noninjected [ intranasal inhalation (WH), intranasal instillation (INS), intrarectal (IR) , intravaginal (IVAG), ocular (Oc), oral feeding (oral)] routes and the gen e gun (GG) were used to deliver HBsAg-expressing plasmid DNA to BALB/c mice . Sera were assessed for HBsAg-specific antibodies (anti-HBs, IgG, IgG1, Ig G2a) and cytotoxic T lymphocyte (CTL) activity measured. Three of the most commonly used routes (LM, ID, GG) were compared in rhesus monkeys, also usi ng HBsAg-expressing vectors. Monkeys were immunized with short (0-, 4- and 8-week) or long (0-, 12- and 24-week) intervals between boosts, and in the case of GG, also with different doses, and their sera were assessed for ant i-HBs. Results: In one study, anti-HBs were detected in plasma of mice treated by five of eight of the injected and none of the six noninjected routes. The h ighest levels of anti-HBs were induced by IM and IV injections, although si gnificant titers were also obtained with SL and ID. Each of these routes al so induced CTL, as did IPER and VW and one noninjected route (INH) that fai led to induce antibodies. In a second study, GG (1.6 mu g) was compared to ID and TM (100 mu g) delivery. Significant titers were obtained by all rout es after only one boost, with the highest levels detected by IM. Delivery t o the skin by GG induced exclusively IgG1 antibodies (Th2-like) at 4 weeks and only very low IgG2a levels at later times; ID-immunized mice bad predom inantly IgG1 at 4 weeks and this changed to mixed IgG1/IgG2a over time. Res ponses with IM injection (in the leg or tongue) were predominantly IgG2a (T h1-like) at all times. IV injection gave mixed IgG1/IgG2a responses. In mon keys, in the first experiment, 1 mg DNA IM or ID at 0, 4, and 8 weeks gave equivalent anti-MB titers and 0.4 mu g at the same times by GG induced lowe r titers. In the second experiment, 1 mg DNA IM or ID, or 3.2 mu g by GG, a t 0, 12, and 24 weeks, gave anti-MR values in the hierarchy of GG > IM > ID . Furthermore, high titers were retained after a single immunization in mic e but fell off overtime in the monkeys, even after boost. Conclusions: Route of administration of plasmid DNA vaccines influences the strength and nature of immune responses in mice and non-human primates. Ho wever, the results in mice were not always predictive of those in monkeys a nd this is likely true for humans as well. Optimal dose and immunization sc hedule will most likely vary between species. It is not clear whether resul ts in non-human primates will be predictive of results in humans, thus addi tional studies are required.