The studies presented above were designed to further assess practical aspects of the recombinant gas vesicle display/delivery system through in vitro and in vivo studies. As a self-adjuvanting, highly stable component , r-GV have significant potential as an alternative delivery system and thus the studies reported were designed to provide additional characterizations. The SIV fragments tested were selected because they are key immunogens examined as relevant gene products in other published research [14, 15] Those studies had shown toxoids of Tat or Tat and Rev, or delivery of aggregated multiantigens, provided some disease attenuation [5, 6] whereas use of Tat alone for example, provided only temporary infection containment. This was interpreted to mean a single antigen might not be adequate to provide full and/or long term protection [14–19]. Our recombinant system inherently supports the expression and delivery of various epitopes [8, 9, 13] and could be readily adapted if important, new peptides were identified. Thus combinations of r-GV particles could be easily created for use in vivo and this would provide exposure to multiple antigen fragments and thereby elicit a variety of peptide specific antibodies.
Different isotypes bind to cell surface expressed Fc receptors in vivo and the resulting receptor derived cell stimulation then will initiate a cascade of effects that can impact the local milieu, and ultimately cell functions. Utilizing the optimum antigen specific dilution of each antibody sample as determined by pre-titration using the specific Tat, Rev, and Nef1 antigens, the present studies identified the isotype(s) elicited by r-GV immunizations. These analyses showed IgG (primarily IgG1 with some instances of minimal IgG2a) was the predominant immunoglobulin species (Figure 1). Although for the R3 and N3 groups, IgM was predominant at 10 days, at 43 weeks, in the absence of any subsequent booster, IgG1 was the dominant isotype in all samples.
This can be relevant in vivo since antibody isotypes impact downstream events through antibody binding to cell displayed Fc receptors . Additionally, the continued presence of antigen specific mouse antibodies 301 days (43 weeks) after the final re-immunization, is potentially an important feature for an immunizing agent and likely to be relevant in terms of practical utility. The isotyping data also inherently demonstrated that the peptides encoded by the expressed small DNA inserts (tat -150 bp and. rev -243 bp), are immunogenic and as shown in Figure 3, relative to the expressed nef1 encoded insert (642 bp) appear to be degraded more slowly than their actual size would predict. This finding may reflect the central location of these encoded small peptide inserts, 50 and 81 amino acids respectively, within the GvpC protein. At that site they may be less readily accessible to the intracellular degradation processes. Nevertheless they are immune visible and evoke specific antibody responses.
Gas vesicles are particulate moieties and when phagocytosed can impact various cell functions, including cytokine production. In turn, both individually and in combination, cytokine levels significantly affect the nature and level of immune responses and response polarization in terms of antibody (Th2), and cell mediated immunity (Th1) [21, 22]. In vivo, cytokines impact immune related processes and certainly specific roles for different cytokines have been identified and extensively investigated. Studies of HIV for example, have demonstrated cytokine correlations with protection from infection, prediction of disease progression and impact on immune response polarization [23–26]. Based on the potential for an insert specific effect, the present in vitro studies assessed the r-GV immunogens for their possible impact on selected cytokines, quantified by testing the media of monolayers treated with r-GVTat.Rev and Nef1, vs. wt-GV, or left untreated. Although the SIV specific peptide inserts constitute a very small segment in the recombinant GvpC protein, nevertheless in vitro studies showed there were detectible, insert specific differences in the cytokine patterns and the temporal profiles of their changing concentrations (Table 2). Given the known effects of cytokines IL-18, IL-10 and IL-12, and data in the literature relating to Tat, Ref and Nef1, the present findings show that despite the preponderance of GvpC specific protein, even very small segments of these pathogen proteins could play specific and significant roles in vivo by influencing the local milieu [27–29]. Although in vitro effects can not fully reflect in vivo process or responses, they nevertheless can indicate the potential to exert an effect. Analogous consequences in vivo could alter the local milieu and thereby alter the Th1/Th2 responses equation. In turn, this might affect vaccine efficacy and certainly adds to a characterization of this immunization vehicle. Thus similar examinations would be relevant in the broader context of vaccine development, and of characterizing pathogen impact on tissues or systems. Known examples are inhibitory effects of IL-10 on HIV-1 production in monocytes/macrophages resulting from IL-10 stimulated inhibition of cytokine syntheses e.g. tumor necrosis factor alpha (TNF-α) and IL-6. In the context of HIV-1, this alteration can up-regulate pathogen expression in the infected cells . Additionally, studies by others also have described IL-10 inhibitory effects on cytokine synthesis [31–33], and on pathogen (HIV) replication in monocytes [30, 34].
The inherent physical stability of the gas vesicle structure, along with its resistance to chemical dissociation or proteolytic cleavage [12, 35, 36] and its particulate structure, likely play a role in the natural slow release of the incorporated epitope containing fusion proteins and could enhance stimulation of the immune system. Ultimately however, the full degradation of the particulate r-GV immunogen is important. Their long term presence potentially could elicit chronic, local inflammatory responses that can characterize degradation resistant particles. The in vitro immunostaining and the Western blot findings presented here were aimed at assessing intracellular GV breakdown. The results indicate that degradation of these naturally stable immunogenic particles does occur and demonstrates the sizes of degradation products. In this context, it is noteworthy that even long term, oral ingestion of gas vesicles on a regular basis  has provided no evidence that gas vesicles exhibit any detectible short or long term toxicity. Thus per-oral usage of r-GV also could be an option.
Micrographs from the immunofluorescence staining allow the tracking of r-GV intracellular fate after the application to J774A.1 cells monolayers in vitro. These studies demonstrated that as expected, r-GV are phagocytosed. Microscope focusing indicated r-GV associated with and/or were within J774A.1 macrophages at 12 h post application and clearly internalized at later times (see Figures 2 and 3). Results from the subsequent immunostain monitoring of both the gas vesicle protein GvpC, and the SIV specific inserts confirmed their slow, progressive disappearance over time. An intracellular accumulation/coalescence of individually internalized r-GV particles clearly takes place and is evident in the micrographs of 12 h post application monolayers. Additionally, samples from various times (12-120 h) following removal and rinsing off of the applied r-GVTat, Rev or Nef1 were immunostained using either anti- SHIV or anti-GvpC antibody (Figure 3 and 4) and examined. These clearly show a decrease in detectible immunostained material during this time course. The results from this group of studies therefore provides critical information in terms of potential r-GV utility as a delivery vehicle since it indicates that in vivo there should be an ongoing but distinctly finite host cell perturbation following cellular uptake. This likely would impact immune stimulation, potentially by processes including the stimulation of cytokine production detected and quantified above. Importantly the present studies tracking degradation indicated that over time in vitro, the recombinant particles did undergo degradation in the individual cells. This was evidenced by the immunomicroscopy data for samples (Figures 3 and 4) and was verified in more detail by the Western blot assessments which tested lysates of r-GV treated monolayer at 96 and 120 h post r-GV applications (Figure 5).
Since recombinant inserts are expressed in gas vesicle protein 'C', studies assessed breakdown of this protein species and also of the specific expressed SIVsm inserts. In SDS-PAGE gels isolated, intact GvpC would appear as a single, ~60 kDa protein species . The Western blot analysis using anti-GvpC showed clearly that wt-GV and the r-GVTat, Rev and Nef1 recombinants all exhibited multiple molecular weight species at both 96 and the 120 h times after application to monolayer cells. This is the predicted result when proteins that become internalized are broken down and is consistent with the immunostaining detected in Figures 3 and 4. It also would be expected that the band patterns in SDS-PAGE gels, and the density of Western blot immunostained bands would change as a function of time and this is the case (Figure 5, left hand panels). Thus, by 120 h, there are crisp GvpC positive bands of lower molecular weight species (ex. somewhat greater than 18 kDa and less than 14 kDa), as well as higher molecular weight species which are now more lightly stained by the anti-GvpC antibody. All of these bands are reactive with the anti-GvpC antibody, and since in SDS-PAGE gels, GvpC that is un-degraded behaves as a 60 kDa moiety ; these clearly are degradation products from the GvpC protein. They exhibit discrete, decreasing molecular weights consistent with degradation that is ongoing over time. Similarly the companion Western blots probed with anti-SHIV specific antibody demonstrated breakdown of the SIVsm derived inserts and showed the disappearance of bands over time (Figure 5, right hand upper and lower panels). In addition to the detailed findings, the results overall support the conclusion that intracellular r-GV are susceptible to degradation and therefore can be cleared from cells that may phagocytose them. Furthermore, unlike the effects on cytokine concentrations, inserts in r-GVTat, Rev or Nef1 do not appear to significantly alter the r-GV time course of degradation in an insert dependant way.