The TLR4 Agonist Vaccine Adjuvant, GLA-SE, Requires Canonical and Atypical Mechanisms of Action for TH1 Induction

The Toll-like receptor 4 agonist glucopyranosyl lipid adjuvant formulated in a stable emulsion (GLA-SE) promotes strong TH1 and balanced IgG1/IgG2 responses to protein vaccine antigens. This enhanced immunity is sufficient to provide protection against many diseases including tuberculosis and leishmaniasis. To better characterize the adjuvant action it is important to understand how the different cytokines and transcription factors contribute to the initiation of immunity. In the present study using T-bet-/- and IL-12-/- mice and a blocking anti-IFNαR1 monoclonal antibody, we define mechanisms of adjuvant activity of GLA-SE. In accordance with previous studies of TLR4 agonist based adjuvants, we found that TH1 induction via GLA-SE was completely dependent upon T-bet, a key transcription factor for IFNγ production and TH1 differentiation. Consistent with this, deficiency of IL-12, a cytokine canonical to TH1 induction, ablated TH1 induction via GLA-SE. Finally we demonstrate that the innate immune response to GLA-SE, including rapid IFNγ production by memory CD8+ T cells and NK cells, was contingent on type I interferon, a cytokine group whose association with TH1 induction is contextual, and that they contributed to the adjuvant activity of GLA-SE.


Introduction
Subunit vaccines, in combination with adjuvants to enhance the immune response to the target antigens, represent a significant advance in the development of better defined, easier to produce and safer vaccines. Importantly an optimal adjuvant should induce a targeted innate response to tailor the desired adaptive response needed for vaccine efficacy. To that end we have developed a number of Toll-like receptor (TLR) agonists containing adjuvants that promote T H 1 T cell responses to vaccine antigens. The most notable of these is the GLA-SE adjuvant which contains a synthetic TLR4 agonist, Glucopyranosyl Lipid Adjuvant (GLA), formulated in a stable nano-emulsion of squalene oil-in-water (SE) [1,2]. GLA-SE drives strong T H 1 responses to a variety of antigens that are protective against intracellular infections [2][3][4][5][6][7]. In combination with the tuberculosis vaccine fusion protein antigen ID93, GLA-SE induces a poly-functional T H 1 response characterized by CD4 T cells producing CD154, IFNγ, TNF, GM-CSF, and IL-2, as well as a humoral response skewed towards IgG2c class-switching [8][9][10]. In order to better understand the mechanism of action of adjuvants it is important to define the role of different cytokines and transcription factors in initiating the immune response from the naive polyclonal repertoire.
The differentiation of CD4 T cells into T H 1 effectors is orchestrated by the transcription factor T-bet [11]. This differentiation occurs in two steps: first, during the initial polarization phase, simultaneous signaling via the T cell antigen receptor TCR and the IFNγ receptor of the naïve T cell results in T-bet expression which enables IL-12 receptor expression, and subsequently a second wave of T-bet expression is induced by IL-12 signaling in the absence of TCR stimulation [12][13][14][15]. T-bet induction and IL-12 production are therefore likely necessary for the potent T H 1 response induction to vaccination with GLA-SE.
Type I interferons (IFNα and IFNβ) induce an antiviral state in most nucleated cells, providing protection against infection [16,17]. Furthermore, type I IFN can shape the adaptive responses to infection (reviewed in [18]). These cytokines signal via the heterodimeric IFNαR1/2 receptor and act on both antigen presenting cells (APC) and lymphocytes to enhance maturation, proliferation and survival to a variety of stimuli [19]. In the present study, using T-bet -/and IL-12 -/mice and IFNαR1 antibody blockade we demonstrate that T-bet induction, IL-12 production and IFNαR1 signaling are necessary for the adjuvant activity of GLA-SE and that IFNαR1 signaling is also crucial for the early innate response initiation to this adjuvant.

GLA-SE adjuvant activity is dependent on T-bet expression and IL-12 production
GLA-SE, a synthetic TLR4 agonist formulated in a stable nano-emulsion of squalene oil induces a strong T H 1 response to vaccines antigens that otherwise elicit minimal cellular immune responses with a T H 2 bias [1,9,10,20]. IL-12 is important for T H 1 induction with LPS, another TLR4 agonist, and monophosphorylated lipid a (MPLA), a detoxified derivative of LPS [21,22]. Mouse and human dendritic cells stimulated with GLA produce IL-12 in a MyD88 and TRIF dependent manner [2,9]. To determine whether IL-12 production and/or Tbet expression are important for GLA-SE driven antibody and CD4 T cell responses, we immunized wild type (B6) or IL-12 or T-bet deficient mice with GLA-SE and the recombinant antigen ID93. Both T-bet and IL-12 were essential for T H 1 induction as indicated by CD4+ T cells production of CD154, IFN-γ, or TNF upon stimulation with ID93 ( Fig 1A). Following immunization with ID93+GLA-SE induction of a strong IgG2c skewed response to ID93 was completely dependent on T-bet expression, but, surprisingly, not on IL-12 ( Fig 1B). These data suggest discordance between an IL-12 dependent induction of IFNγ-producing CD4 T cells and an IL-12 independent class-switching to IgG2c, with T-bet being essential for both. adjuvant activity [9]; suggesting a role for type I IFN in the T H 1 differentiation induction by GLA-SE. To determine whether GLA-SE induced type I IFN in vivo, we measured IFNα levels in the draining lymph node (LN) and sera shortly after immunization with GLA-SE. IFNα levels in the draining LN peaked between 6 and 24 hours and returned to almost undetectable levels at 48 hours (Fig 2). No IFNα was detected in the serum indicating that the response was highly localized to draining LNs (Fig 2).
Type I IFN receptor signaling is critical during the innate response to GLA-SE Type I IFN drive expression of the cell surface activation marker CD69 which in turn facilitates trapping of naïve lymphocytes in the draining LN with antigen-loaded APCs [26]. Immunization with GLA-SE induces CD69 on lymphocytes, independent of antigen-specificity and this induction is partially dependent on cell intrinsic expression of type I IFN receptor as well as innate phase production of IFNγ. Similar to our previous findings [23] immunization with GLA-SE drove a transient expression of CD69 on the total lymphocyte population in the draining LN, which was dependent on type I IFN signaling (Fig 3A). GLA-SE immunization also produced a strong burst of IFNγ in the draining LN 6hr after immunization which returned to baseline by 24 hours (Fig 3B). This was also dependent on type I IFN as blocking the IFNαR1 ablated most of the IFNγ production ( Fig 3B).
The occupancy of IFN-γ receptor, and therefore the amount of IFNγ present on the surface of cells, can be determined by staining with the empty-IFNγ receptor-specific antibody clone GR-20 [14]. Correlating with our IFNγ production results, 12hrs after immunization the empty-IFNγ receptor staining on lymphocytes from draining LNs was substantially reduced in mice treated with GLA-SE compared with contralateral LNs. IFNαR1 blockade preserved empty-IFNγ receptor staining confirming that the early IFNγ burst depends on IFNαR1 signaling ( Fig 3C). This early burst of IFNγ at 6 hr was produced by both memory CD8 T cells and NK cells (Fig 3D and 3E). This is likely driven by IL-18 expression which is induced by GLA-SE and necessary for the adjuvanticity of GLA-SE [23]. IFNγ production by both cell types was strongly reduced in mice that were previously treated with anti-IFNαR1 (Fig 3D and  3E). These results indicate that type I IFN, in sequential manner, was necessary for the early innate production of IFNγ by CD8+ T cells and NK cells in response to GLA-SE.
Concurrent with the early IFNγ production there is an increase in IL-12 production in the draining LN following GLA-SE immunization as early as 6 hours post immunization which is sustained for at least 2 days and is critical for the adjuvant activity of GLA-SE (Figs 1 and 3F). Surprisingly IFNαR1 blockade augmented the levels of IL-12 in the draining LN despite IFNγ levels being lower in the treated animals at the same time-points ( Fig 3F). These results suggest that type I IFN production inhibits early IL-12 production, which is supported by prior reports [27][28][29][30]. Overall the innate activity of GLA-SE is strongly dependent on type I IFN which is necessary for lymphocyte activation, controls IL-12p40 production and promotes early IFNγ production, a feature associated with T-bet expression in T cells, and subsequently T H 1 induction [31].  To further investigate the role of type I IFN in the adaptive response elicited by immunization with GLA-SE and ID93, we blocked IFNαR1 and examined the cellular and humoral responses one week after immunization. IFNαR1 blockade reduced the frequency of antigen-specific CD4 T cells as measured by CD154 expression upon stimulation ( Fig 4A). Further, T H 1 responses were reduced and T H 2 responses enhanced, as measured by antigen induced production of IFNγ, TNF and IL-5 ( Fig 4A). In parallel, samples were stained for ID93-specific CD4 T cells via peptide-MCHII tetramer (Tet) staining. Consistent with CD154 expression upon stimulation there was a significant reduction in the frequency of Tet+ CD4 T cells from the IFNαR1 blockade group (Fig 4B). Congruent with the reduction in IFNγ observed after antigen stimulation, the remaining Tet+ T cells from the IFNαR1 blockade group had lower levels of T-bet staining ( Fig 4C). The reduction in T H 1 skewing correlated with modestly decreased isotype switching to IgG2c in ID93-specific antibody titers ( Fig 4D). Taken together these results suggest that GLA-SE induced type I IFN promote T H 1 responses by enhancing CD69 expression and increasing innate IFNγ production.

Discussion
GLA-SE is a clinical stage vaccine adjuvant that robustly augments T H 1 immunity yet its mode of action is still incompletely understood. In this study we assessed the contribution of T-bet, IL-12 and IFNαR1 signaling to the adjuvant activity of GLA-SE. We show that T-bet is required for IgG2c isotype skewing as well as T H 1 induction by GLA-SE. Thereupon we sought to investigate the role of the pro-inflammatory cytokine IL-12 and, surprisingly, found that IgG2c class-switching was not impaired even though production of IFN-γ secreting CD4 T H 1 cells was ablated. Finally, our results suggest that type I IFN, which are induced early after immunization with GLA-SE, are necessary for an optimal innate response including rapid lymphocyte activation and IFNγ expression by NK cells and memory CD8 T cells and, consequently, the induction of T H 1 immunity.
IL-12 is a key factor that drives T H 1 responses and IFNγ production [23,[32][33][34][35]. However, the relevance of IL-12 in vivo depends on the nature of the infection or immunization. In IL-12-deficient mice, T H 1-type responses were generated upon immunization with inactivated MHV, demonstrating the existence of an IL-12 independent T H 1 induction that has been since then described in several models of viral infection and immunization [36][37][38][39][40][41]. Conversely, IL-12 has been shown to be important for T H 1 induction by the TLR4 agonists LPS and MPL, although antibody production was not evaluated in those papers [21,22]. Using IL-12 -/and Tbet -/mice, we demonstrated that IL-12 is needed for the CD4 T H 1 induction by GLA-SE but not for IgG2c skewing, whereas T-bet is necessary for both. Since B-cell intrinsic T-bet has been shown to mediate early IgG2c production [42] further experiments will be needed to determine if T-bet expression upon ID93-GLA-SE vaccination is required in the CD4 T cells to promote IgG2c skewing. Furthermore we have previously shown that MyD88 and TRIF are both necessary for GLA-SE T H 1 skewing capacity as well as IL-12 production by BMDC, DC and macrophages [9]. This suggests that IL-12 production might be a responses by ID93+GLA-SE immunization. Moreover IL-18 has been shown to synergize with IL-12 to induce IFNγ in T cells [34,35]. IL-18 is produced very early upon immunization with GLA-SE and IL-18R -/mice have a reduced antigen-specific T H 1 response [23]. We hypothesize that this IL-18/IL-12 synergy is required for GLA-SE-induced T H 1 responses.
Type I IFN links the innate and adaptive immune responses. They cause APC maturation in vitro while inhibiting secretion of T H 2 cytokines leading to IFNγ producing T cells [43][44][45]. Type I IFN inhibit the death of CD4 activated T cells and their key role in priming adaptive T cells has been demonstrated in a wide variety of immunization schemes [46,47]. Recombinant type I IFN given in vivo has direct adjuvant activity, enhancing both antibody and T cell responses [48,49]. TLR agonist and non-TLR agonist require type I IFN activity to exert their adjuvant activities [18,48,50]. In some cases this may be due to IFNαR signaling altering the expression of some TLRs [43]. Furthermore Longhi et al observed an IL-12 independent impairment of T H 1 responses to poly (I:C) when blocking IFNαR1 and in IFNαR -/mice [39]. Thus it has been hypothesized that robust type I IFN induction may be necessary for effective adjuvant activity [51]. However other adjuvants including MF59 and Pam3CSK (TLR2) do not elicit type I IFN and do not require IFNαR signaling to enhance antibody responses [50].
Our results show that that type I IFN is critical for the adjuvanticity of GLA-SE. Type I IFN was required for maximal response to this adjuvant illustrated by defects in early innate responses as indicated by reduced CD69 expression and defects in innate IFNγ production by CD8+ T cells and NK cells. CD69 expression inhibits lymphocytes egress from the LN which is a mechanism to retain useful clonal specificities in the activated LN [26,52]. Early IFNγ produced by innate cells upon immunization with GLA-SE and other adjuvants is critical for CD4 T cells T H 1 polarization in the LN [13,15,23]. IFNαR1 blockade caused multiple defects in T H 1 adaptive responses including: (1) reduction in the frequency of antigen-specific cells determined by CD154 production or tetramer staining, (2) reduction in T H 1 skewing measured by IFNγ or TNF production and lower amounts of the T H 1 controlling transcription factor Tbet in the remaining tetramer positive cells, and (3) enhanced frequencies of CD4 T cells that produced IL-5, a hallmark of T H 2 differentiation.
In some settings type I IFN is necessary for IL-12 production by murine and human dendritic cells [53,54] but they also have been shown to decrease IL-12 production in other studies [27][28][29][30]. IFNαR1 blockade increases IL-12 production early after immunization with GLA-SE but diminished IFNγ production at the same time indicating a complex regulation of IFNγ upon vaccination. We hypothesize that type I IFN driven IFNγ is necessary to increase expression of the IL-12R on CD4 T cells, thus making them responsive to IL-12. Alternatively production of the IL-12p35 subunit to form active IL-12p70 may be dependent on IFNαR1 signaling. In addition to their role in stimulating APCs to enhance CD4 responses, type I IFN can also have a direct effect on T cells. CD8 T cells require direct type I IFN signals for their expansion in response to some infections [55]. Likewise direct type I IFN action on CD4 T cells is important for clonal expansion in vivo following LCMV, but inhibitory during Listeria infection [56] showing that the priming milieu determines the extent to which CD4 T cells are dependent on direct signal mediated by type I IFN. Conversely CD4 T cell responses are dependent on type I IFN in a cell extrinsic manner in response to polyIC:CD40 vaccination [57]. Interestingly deletion of IFNαR2 on T cells significantly diminished CD69 expression induced by GLA-SE at 18h after immunization suggesting a dependence on T cell intrinsic expression of the type I IFN receptor for GLA-SE innate response induction [23]. Based on our results we suggest a model where GLA-SE induces IFNα to trap cells in the LN by CD69 expression and to augment the early innate IFNγ production by CD8+ T cells and NK cells. In parallel GLA-SE induces production of IL-12 in vivo which synergizes with IFNα effects to promote Tbet expression and T H 1 commitment. Understanding the mechanism by which adjuvants engage the immune responses is critically important for development of vaccines. Our results suggest that early type I and II IFNs are signatures of the optimal innate response to the GLA-SE adjuvant, predicting subsequent T H 1 responses. Further research is warranted to determine if this early type I IFN induction can be used as an early gating strategy in developing new adjuvants or as a signature of adjuvanticity in human clinical trials.

Ethics statement
The study was conducted under protocols approved by the Infectious Disease Research Institute Institutional Animal Care and Use Committee. Animals 6-8 week old female C57BL/6 (B6), IL-12 -/and T-bet -/mice were purchased from Jackson Lab and maintained in Specific Pathogen Free conditions. All animal study protocols were approved by the IDRI Institutional Animal Care and Use Committee (IACUC) and were performed according to IACUC regulations and guidelines.

IFNα ELISA
At indicated times following immunization peripheral blood and inguinal LNs were collected and LNs were homogenized in 500μL PBS. Total production of IFNα was assessed using the Mouse IFNα ELISA kits (R&D Systems) according to the manufacturer's instructions. Concentrations below the limit of detection were reported as 0.

Antibody responses
Mouse sera were prepared by collection of retro-orbital blood into microtainer serum collection tubes (VWR International, West Chester, PA), followed by centrifugation at 10,000 rpm for 5 minutes. Each serum sample was then analyzed by antibody capture ELISA. Briefly, ELISA plates (Nunc, Rochester, NY) were coated with 2 μg/ml recombinant antigen ID93 in 0.1 M bicarbonate buffer and blocked with 1% BSA-PBS. Then, in consecutive order and following washes in PBS/Tween20, serially diluted serum samples, anti-mouse IgG1 or IgG2c-HRP (all Southern Biotech, Birmingham, AL) and ABTS-H 2 O 2 (Kirkegaard and Perry Laboratories, Gaithersburg, MD) were added to the plates. Plates were analyzed at 405nm (EL X 808, Bio-Tek Instruments Inc, Winooski, VT).

Splenocytes recalls and intracellular cytokine staining
Splenocytes were isolated from three to five animals per treatment regimen. Red blood cells were lysed using Red Blood Cell Lysis Buffer (eBioscience) and resuspended in RPMI 1640, 10% FBS. Total viable cells were enumerated using ViaCount assay with a PCA system (Guava Technologies), plated at 2x10 6 cells/well in 96-well plates and stimulated for 2 hours with media or ID93 (10 μg/mL) at 37°C. GolgiPlug (BD Biosciences) was added and the cells were incubated for an additional 8 hours at 37°C. Cells were washed and surface stained with fluorochrome labeled antibodies to CD4 (clone GK1.5) and CD8 (clone 53-6. 7) (BioLegend and eBioscience) in the presence of 20% normal mouse serum for 20 minutes at 4°C. Cells were washed and permeabilized with Cytofix/Cytoperm (BD Biosciences) for 20 minutes at room temperature. Cells were washed twice with Perm/Wash (BD Biosciences) and stained intracellularly with fluorochrome labeled antibodies to CD154 (clone MR1) IFN-γ (clone XMG-1.2), TNF (MP6-XT22) and IL-5 (TRFK5) (BioLegend and eBioscience) for 20 minutes at room temperature. Cells were washed and resuspended in PBS. Up to 10 6 events were collected on a four laser LSRFortessa flow cytometer (BD Biosciences). Cells were gated as singlets > lymphocytes > CD4+ CD8->cytokine positive. ID93-specific response frequencies were determined by subtracting the frequency of response positives of unstimulated cells from ID93 stimulated cells in matched samples.

Statistical methods and figures
Statistical analyses and figures were performed using Prism software (GraphPad Software, Inc., La Jolla, CA). T-test were used when comparing one group against another and one-way or two ways ANOVA analyses were used when more than two groups were compared over. Nonnormal data sets were log-transformed prior to analysis. Statistical significance was considered when the p-values were <0.05 and noted Ã or # when the group was statistically different to all the other groups.