Immunogenicity of HIV Virus-Like Particles in Rhesus Macaques by Intranasal Administration

Abstract

Female rhesus macaques were immunized with HIV virus-like particles (HIV-VLPs) or HIV DNA administered as sequential combinations of mucosal (intranasal) and systemic (intramuscular) routes, according to homologous or heterologous prime-boost schedules. The results show that in rhesus macaques only the sequential intranasal and intramuscular administration of HIV-VLPs, and not the intranasal alone, is able to elicit humoral immune response at the systemic as well as the vaginal level.

TEXT

The vast majority of new HIV infections worldwide are acquired via the genital mucosa, and women account for close to 50% of them (39). The development of vaccination strategies able to elicit protective systemic and mucosal immune response represents a major goal in the HIV vaccine field, possibly providing a crucial method for halting the spread of HIV/AIDS.

Mucosal secretory immunoglobulin A (sIgA) specific for HIV-1 envelope glycoproteins is consistently detected in seropositive subjects (1, 14) and has been strongly associated with protection from HIV-1 infection in uninfected individuals having unprotected sexual intercourse with HIV-1-seropositive partners (13, 23, 25, 28). Furthermore, intravenous or intravaginal passive administration of the gp120-specific human neutralizing monoclonal antibody (Ab) b12 has been shown to be highly effective in protecting monkeys from a vaginal challenge (12, 30, 40).

Considering this epidemiological and experimental evidence, specific mucosal immunity is extremely relevant for controlling the primary HIV-1 infection. Intranasal (i.n.) immunization has been shown to be effective for protection against infectious respiratory diseases such as influenza (2, 22, 33, 35, 38, 43). However, the effectiveness of mucosal immunization often relies upon coadministration of appropriate adjuvants that can initiate and support the transition from innate to adaptive immunity (recently reviewed in reference 20). In addition to adjuvants, particulate antigens (e.g., virus-like particles [VLPs]) have been shown to be advantageous for intranasal immunization, given that they efficiently target antigen-presenting cells (APCs) and facilitate the induction of potent immune responses (7, 9, 10, 11, 16, 18, 31, 34, 41, 43, 44). However, several vaccine concepts have been evaluated in nonhuman primates (NHPs) by intranasal administration with inconsistent immunogenicity results, probably related to the different vaccination strategy (3, 17, 24, 26, 27, 29).

HIV-VLPs developed in our laboratory, and used in the present study, are based on HIV Gag protein and express the whole HIV gp120/140 envelope protein derived from an Ugandan clade A field isolate (4, 5, 6, 36, 37, 42). Elicitation of immune response at systemic as well as mucosal (vaginal and intestinal) levels has been previously evaluated in mice by intraperitoneal as well as intranasal administration (7, 8, 11). In particular, the mucosal immunogenicity of such HIV-VLPs has been evaluated by comparing a homologous (VLP + VLP) and a heterologous (DNA + VLP) prime-boost strategy by intranasal administration, in an adjuvant formulation (7).

In the present study, the immunogenicity of HIV-VLPs was evaluated in rhesus macaques immunized with HIV-VLPs administered via a sequential combination of mucosal (intranasal) and systemic (intramuscular [i.m.]) routes, according to homologous (VLP prime + VLP boost) or heterologous (DNA prime + VLP boost) prime-boost schedules.

A total of 24 female rhesus macaques were equally divided into four experimental arms and immunized by the intranasal route as described in Fig. 1. Groups 2 and 3 were immunized using the homologous prime-boost protocol in the absence (group 2) or in the presence (group 3) of the Eurocine L3 nasal lipid adjuvant. Group 4 was immunized using the heterologous prime-boost protocol in the presence of Eurocine L3 and N3 adjuvants. Additionally, group 3 received two further boosting doses of VLPs by the intramuscular (i.m.) route, 22 weeks after the last intranasal (i.n.) administration. Group 1 was the control group administered adjuvants. VLPs were administered at 100 μg per immunization dose; DNA plasmids were administered at 200 μg per immunization dose. Antigens as well as adjuvants used in the study have all been previously described (5, 7, 8, 11, 15, 19, 21, 32).

Fig 1.

Immunization scheme in NHPs. Six animals per group were immunized as described at indicated weeks.

Sera were collected from 10 ml of whole blood 1 week before and 1 week after each antigen administration, and enzyme-linked immunosorbent assays (ELISAs) were performed in microwell plates coated with recombinant HIV gp120 or p24 of subtype B. The data show that intranasal administration of HIV-VLPs, in either the homologous or heterologous prime-boost protocol, does not elicit measurable serum anti-Env or anti-Gag Ab titers (Fig. 2A). Moreover, the i.n. administration protocol does not appear to efficiently prime the systemic immune system, since two subsequent i.m. injections were needed to observe significant serum Ab titers (>1:1,000) (Fig. 2B). In particular, evaluating the individual animals in such group, it is possible to identify the best responders for both Env and Gag (no. 4642 > 4635 > 4636) (Fig. 2C and D).

Fig 2.

Systemic immune response. Specific anti-env and anti-gag immune responses in serum of immunized animals were evaluated by ELISA. (A and B) Average Ab titer in each group. (C and D) Ab titer for each animal in group 3 after the two i.m. immunizations. Results are expressed as the reciprocal last dilution with a 3-fold optical density at 492 nm of the preimmune sera.

Vaginal washes were collected on the same days as the serum, and ELISAs were run in parallel. The data show that intranasal administration of HIV-VLPs, in either a homologous or a heterologous prime-boost protocol, does not elicit measurable mucosal titers (data not shown); however, it seems to prime the mucosal immune system which, 6 months after the last i.n. boost, is able to respond after the i.m. immunizations. The effect is evident in 2/6 animals in group 3 and appears to be selective for Env (Fig. 3). Furthermore, antibody titers elicited by the two i.m. administrations of VLPs do not show HIV neutralization or antibody-dependent cell-mediated cytotoxicity (ADCC) activity (data not shown).

Fig 3.

Vaginal immune response. Specific anti-env and anti-gag immune responses in vaginal washes of immunized animals were evaluated by ELISA. The Ab titer for each animal in group 3 is shown. Results are expressed as the reciprocal last dilution with a 3-fold optical density at 492 nm of the preimmune sera.

The results obtained in NHPs in the present study by i.n. administration of VLPs are in contrast to those obtained in mice (7); however, this could be due to either the administered dose (i.e., too low in NHPs) or lower permeability to antigens of the nasal epithelium in macaques. Indeed, our data are in agreement with results from others who have previously shown the limited or absent immune response in NHP by i.n. administration. Such results have been obtained using different vaccine delivery systems, which suggests that they are not vaccine related (3, 24, 26, 27).

In conclusion, the described NHP preclinical trial shows the elicitation of specific immune response by HIV-VLPs when administered by the i.m. route. On the other hand, at least in our experimental model, the i.n. administration is possibly only priming the humoral mucosal immunity for subsequent i.m. boosting doses in a few animals. However, such an observation needs further investigation and must be taken into consideration for future preclinical vaccine evaluations.