The Ultra-Potent and Selective TLR8 Agonist VTX-294 Activates Human Newborn and Adult Leukocytes

Background Newborns display distinct immune responses that contribute to susceptibility to infection and reduced vaccine responses. Toll-like receptor (TLR) agonists may serve as vaccine adjuvants, when given individually or in combination, but responses of neonatal leukocytes to many TLR agonists are diminished. TLR8 agonists are more effective than other TLR agonists in activating human neonatal leukocytes in vitro, but little is known about whether different TLR8 agonists may distinctly activate neonatal leukocytes. We characterized the in vitro immuno-stimulatory activities of a novel benzazepine TLR8 agonist, VTX-294, in comparison to imidazoquinolines that activate TLR8 (R-848; (TLR7/8) CL075; (TLR8/7)), with respect to activation of human newborn and adult leukocytes. Effects of VTX-294 and R-848 in combination with monophosphoryl lipid A (MPLA; TLR4) were also assessed. Methods TLR agonist specificity was assessed using TLR-transfected HEK293 cells expressing a NF-κB reporter gene. TLR agonist-induced cytokine production was measured in human newborn cord and adult peripheral blood using ELISA and multiplex assays. Newborn and adult monocytes were differentiated into monocyte-derived dendritic cells (MoDCs) and TLR agonist-induced activation assessed by cytokine production (ELISA) and co-stimulatory molecule expression (flow cytometry). Results VTX-294 was ∼100x more active on TLR8- than TLR7-transfected HEK cells (EC50, ∼50 nM vs. ∼5700 nM). VTX-294-induced TNF and IL-1β production were comparable in newborn cord and adult peripheral blood, while VTX-294 was ∼ 1 log more potent in inducing TNF and IL-1β production than MPLA, R848 or CL075. Combination of VTX-294 and MPLA induced greater blood TNF and IL-1β responses than combination of R-848 and MPLA. VTX-294 also potently induced expression of cytokines and co-stimulatory molecules HLA-DR and CD86 in human newborn MoDCs. Conclusions VTX-294 is a novel ultra-potent TLR8 agonist that activates newborn and adult leukocytes and is a candidate vaccine adjuvant in both early life and adulthood.


Introduction
Newborns display distinct immune responses that leave them vulnerable to higher rates of infections as compared to older children and adults [1], with over 2,000,000 deaths per year worldwide due to infection in those less than 6 months of age [2]. In particular, neonatal deficiencies of innate cellular immune responses include a decreased production of interferons, IL-12p70, TNF and other proinflammatory/Th1-polarizing cytokines [3]. Divergent neonatal immune responses also pose a challenge to efforts to enhance early life immunization, which is a key global health strategy [4]. Lack of effective immunization strategies in the neonatal period (first 28 days of life) leads to a window of susceptibility to infections in newborns and infants lasting several months. For the past decade, there has been an ongoing focus on understanding the mechanisms by which the immune system of very young is distinct from adults, particularly in response to immunization [5]. New insights into the ontogeny of innate immunity are informing development of age-specific adjuvanted vaccine formulations [6,7,8], with safety and efficacy remaining central priorities [9]. The coming decade is likely to focus on developing technologies that overcome or circumvent these immunological obstacles to the development of more effective early life vaccines [2].
A promising approach to enhancing immunity of newborns and infants is to use TLR agonists as novel stand-alone immunomodulators and/or as adjuvants in vaccine design [10]. Indeed, TLR agonists have been unknowingly used for decades in vaccines such as Haemophilus influenzae type b (HIb) vaccine containing Neisseriaderived outer membrane proteins (TLR2), Bacille Calmette-Guérin (Mycobacterium bovis; TLR 2, 4 and 8) [11], and certain hepatitis B vaccines that contain lipid A (TLR4) [12]. First generation vaccines, including those consisting of inactivated or attenuated virus, contained inherent TLR7 and/or TLR8 adjuvant activity [13]. Both the inactivated polio and Japanese Encephalitis vaccines, which were introduced in 1955 and 1968, respectively, contained single stranded RNA (ssRNA) TLR7/8 agonists. Additionally, the highly effective live attenuated yellow fever vaccine 17D (YF-17D) activates multiple dendritic cell (DC) subsets via TLR2, TLR7, TLR8, and TLR9 [14] and exhibits activity in children [15].
VentiRx Pharmaceuticals (Seattle, WA) has produced a series of novel low molecular weight (400 -475 Da) synthetic benzazepine derivatives that selectively activate cells via TLR8. Due to its ability to re-orient immune responses towards Th1 immunity, the TLR8 agonist VTX-1463 is in development for the treatment of allergic rhinitis and has shown significant symptom reduction in atopic individuals [29]. The TLR8 agonist VTX-2337 activates monocytes, DCs and NK cells, induces IFN-c production and increased cytolytic activity, and is being developed as an immunotherapy in multiple oncology indications [30]. In the present study, we assessed novel VentiRx benzazepine compounds (VTXs), characterized their specificity using TLR-transfected human embryonic kidney (HEK) cells, and have evaluated their bioactivity towards human newborn and adult primary leukocytes. We find that VTX-294 is an ultra-potent TLR8-selective agonist with greater potency and activity than IMQs in both neonatal and adult leukocytes and MoDCs, suggesting that it may be a particularly attractive candidate as an adjuvant towards neonatal leukocytes.

Ethics Statement
Non-identifiable cord blood samples were taken with approval from the Ethics Committee of The Brigham & Women's Hospital, Boston, MA (protocol number 2000-P-000117). All blood samples from adult patients included in the experiments provided written informed consent with approval from the Ethics Committee of Boston Children's Hospital, Boston, MA (protocol number X07-05-0223). All experiments were performed in accordance with relevant institutional and national guidelines, regulations and approvals.

Human TLR Expressing HEK293 Cell Assays
Human embryonic kidney cells (HEK293) expressing TLR2, TLR3, TLR4, TLR5, TLR7, TLR8, or TLR9 were purchased from InvivoGen (San Diego, CA) and cultured in DMEM (Cambrex, Walkersville, MD) containing 4.5 g/L L-glucose (Sigma-Aldrich, St. Louis, MO) and 10% fetal bovine serum (FBS) (Gibco). The activity of specific TLR agonists was assessed using the secretory embryonic alkaline phosphatase (SEAP) reporter gene that is linked to NF-kB activation in response to TLR stimulation. Measurement of SEAP activity was determined using the Quanti-blue TM substrate (InvivoGen) after TLR agonist treatment for 18-24 hours using a Victor 3V 1420 Multilabel Counter (Perkin Elmer) to measure OD at 650 nM.

Human Blood
Peripheral blood was collected from healthy adult volunteers, while human newborn cord blood was collected immediately after Cesarean section delivery of the placenta. Births to HIV-positive mothers were excluded. Human experimentation guidelines of the U.S. Department of Health and Human Services, The Brigham & Women's Hospital, and Boston Children's Hospital were observed, following protocols approved by the local institutional review boards. Human blood was anti-coagulated with 15-20 units/ml pyrogen-free sodium heparin (American Pharmaceutical Partners, Inc., Schaumberg, IL). All blood products were kept at room temperature and processed within 4 hrs of collection.

Blood Sample Processing and in vitro Stimulation
For assessment of TLR agonist activity in whole blood, we used an adaptation of the method of Kollmann et al. [32,33]. Neonatal cord blood or adult whole blood (WB) was mixed 1:1 with sterile pre-warmed (37uC) RPMI 1640 medium (Invitrogen, Carlsbad, CA) and 180 mL of the 1:1 suspension was added to each well of a 96 well U-bottom plate (Becton Dickinson, Franklin Lakes, NJ, USA) containing 20 ml freshly prepared specific TLR agonists at 10x the final concentration. Suspensions containing 200 ml/well were gently mixed by pipetting and incubated for 2-24 h at 37uC in a humidified incubator at 5% CO 2 . After culture, plates were centrifuged at 5006g and 110-150 ml of supernatant was carefully removed by pipetting without disturbing the cell pellet. Superna-tants derived from human leukocyte stimulations were assayed by ELISA for TNF (BD Biosciences, San Jose, CA, USA) and IL-1b (eBiosciences, San Diego, CA). Additionally, WB supernatants were analysed by Multi-Analyte Profiling (MAP). Briefly, supernatant samples were frozen at 280uC and submitted to Myriad RBM (Austin, TX, USA) for screening with a modified Human InflammationMAPH v. 1.6 antigen panel, which analyzed the samples for the presence of 98 analytes associated with inflammation. The minimum threshold for each analyte was set at the minimum detectable dose for that particular assay (defined as three standard deviations above the mean background).

Monocyte-derived Dendritic Cells (MoDCs)
Heparinized human newborn cord blood or adult peripheral blood was layered onto Ficoll-Hypaque gradients (Ficoll-Paque PREMIUM, GE Healthcare, Waukesha, WI) to collect CBMC or peripheral blood mononuclear cell layers (PBMC), respectively. Monocytes were isolated from mononuclear cell fractions by positive selection by magnetic microbeads according to the manufacturer's instructions (Miltenyi Biotec, Auburn, CA) using CD14 or CD33 as a pan-marker for monocytes [34]. Monocyte preparations were .95% pure as assessed by flow cytometry for CD14 as previously described [25]. Isolated monocytes were cultured in tissue culture dishes at 10 6 cells/ml in RPMI 1640 media containing fresh 10% autologous plasma, supplemented with rhIL-4 (50 ng/ml) and rhGM-CSF (100 ng/ml) (R&D Systems, Minneapolis, MN) with one additional supplement of fresh media and cytokines at day 3 of culture. After 5-6 days, immature MoDCs (HLA-DR + , CD14 2 , DC-SIGN + as measured by flow cytometry) of .90% purity were harvested by gently pipetting only the loosely adherent fraction and re-plated (10 5 cells/well) in 96-well U-bottom plates in the presence or absence of TLR agonists.

Flow Cytometry
Following TLR agonist stimulation, MoDCs were resuspended in staining buffer (16PBS, 0.5% human serum albumin (HSA)) and stained for 30 min at 4uC in the dark (1.5610 5 /per staining) with fluorophore-labeled antibodies (FITC/HLA-DR and PE/ CD86) (BD Biosciences). Cells were then centrifuged, washed, fixed (1% paraformaldehyde (PFA)) and filtered prior to flow cytometry analysis using a MoFlo Legacy cytometer (DakoCytomation, Fort Collins, CO) and analyzed using FlowJo software (Tree Star). MFI fold change was plotted after subtraction of the MFI of equivalent samples stained with an isotype control antibody.

Statistical Analyses and Graphics
Data were analyzed using Prism for MacIntosh v. 5.0b (GraphPad Software Inc., San Diego, CA). Data in figures represent means 6 SEM. p values ,0.05 were considered significant. Normalcy was tested using the Shapiro-Wilk normality test. For normal sample sets, two-tailed t-test or One-way ANOVA with Bonferroni post-test were applied as appropriate. Nonnormal sample sets were analyzed by Kruskal-Wallis test with Dunn's multiple comparison post-test, by Wilcoxon signed-rank test or by Mann-Whitney test, as appropriate.

The Novel TLR Agonist VTX-294 Potently Activates Human Newborn and Adult Leukocytes
We tested the ability of VTX agonists to induce dose dependent cytokine production in human neonatal and adult blood ( Figure 2). In line with the HEK293 data in Figure 1 (C-D), the negative control compound VTX-217 did not demonstrate activation of either human neonatal cord or adult peripheral blood (data not shown). As VTX-294 was the most potent of all three VTX compounds, we tested the dose-dependent cytokine induction of VTX-294 compared with MPLA, a TLR4 agonist whose congeners are adjuvant components of vaccines licensed by both the U.S. Food and Drug Administration (FDA) and European Medicines Agency (EMA), and two imidazoquinoline TLR agonist adjuvants ( Figure 2). VTX-294 activated neonatal and adult blood in a dose-dependent manner, significantly inducing production of TNF and IL-1b over baseline (p,0.001). In both newborn cord and adult peripheral blood, MPLA (TLR4), R848 (TLR7/8) and CL075 (TLR8) significantly induced production of TNF and IL-1b, in a dose dependent manner ( Figure 2). When compared to R848 at 0.1 mM, VTX-294 demonstrated greater efficacy, stimulating greater TNF and IL-1b production in neonatal blood (p,0.05). VTX-294 was also more potent than R848 and CL075 in inducing IL-1b production in adult WB (p,0.05). Greater potency of VTX-294 as compared to R848 and CL075 was also noted for adult TNF production, but did not reach significance (p = 0.054). Overall, VTX-294 demonstrated the greatest potency (i.e. lowest EC 50 values) of all agonists tested for both TNF and IL-1b production in newborn cord and adult peripheral blood (Table  S1). With respect to efficacy of inducing cytokine production, the maximum cytokine levels produced at higher agonist concentrations of VTX-294 (e.g., 100 mM) was similar to those induced by R848 and CL075 (Table S2).

Comparison of TLR Agonist-induced TNF and IL-1b Production in Newborn and Adult Blood
Our results demonstrate an impaired newborn WB TNF response to MPLA by between group comparison (newborn vs. adult) (Figure 3 A, p,0.01). In line with previous findings on the unique efficacy of TLR8 agonists in newborn blood [18], we demonstrated a preservation of responses to R848 and CL075, both inducing TNF production in newborn cord blood at concentrations equal to or greater than that in adult blood (p,0.01 and p,0.001 respectively) by between group comparison (newborn vs. adult). VTX-294 not only induced greater concentrations of TNF in newborn WB compared to adults (p,0.001), but also did so at individual doses (newborn vs. adult) as low as 1 mM (p,0.05). In contrast, R848 was not more effective at inducing TNF in newborn relative to adult WB, while CL075 demonstrated greater efficacy at inducing TNF at 100 mM than VTX-294. IL-1b production followed a similar pattern, with VTX-294 inducing a greater concentration of IL-1b in newborn cord blood compared to adult peripheral blood at concentrations of .1 mM (p,0.01). CL075-and R848-induced IL-1b production was significantly increased in newborn compared to adult WB at 10 mM (p,0.05) (Figure 3 B).

Multi-Analyte Comparison of Agonist Combinationinduced Cytokine Production in Newborn and Adult Blood
We next employed a multiplexed bead array-based approach to provide a broader view of differences between newborn and adult TLR agonist-induced inflammatory biomarkers after 6 h of stimulation ( Figure 4 and Figure S2). This approach confirmed that compared to adult blood, newborn blood contained: 1) lower basal CRP concentrations (Figure 4 A), 2) greater human growth hormone [35] (Figure S2 A), 3) increased basal CXCL-8 (IL-8) concentrations [36], and 4) greater TLR4-mediated production of IL-10 (Figure 4 B) [32]. We also confirmed that TLR7/8 agonists induce similar concentrations of IL-12p40 in both newborn and adult WB [32], although IL-12p70 production in neonates was reduced [37] (Figure 4 C-F).
Newly identified differences in newborn as compared to adult blood, included a consistent increase in both EN-RAGE (extracellular newly identified RAGE-binding protein) and myeloperoxidase (MPO) under all stimulated conditions ( Figure S2). The increase in MPO may reflect the greater number of neutrophils in newborn cord blood compared to adult [3]. TLRmediated production of oncostatin M (OSM), an IL-6 family cytokine, was greater in newborn, as compared to adult blood for MPLA (100 ng/ml), R848 (10 mM) and VTX-294 (10 mM; Figure  S2). Conversely, under all conditions tested, newborn cord blood

VTX-294 Induces Greater Cytokine Responses than R848 in Human Neonatal Blood
At the lowest tested concentration (0.1 mM), VTX-294 induced greater concentrations of nearly all cytokines tested than R848 (Figure 5 A). Specifically, when expressed as fold change over R848 induction, VTX-294 induced .10 fold more TNF, IL-1b and IL-10 and .5 fold more IL-6 and IL-12p40 (Figure 5 A). To further characterize the immunostimulatory potential of VTX-294, we assessed dose-dependent stimulation of newborn WB measuring 50 analytes, at concentrations of 0.1-10 mM (Figure 6 A-B). When graphed as fold change over unstimulated, a dose-dependent induction was clearly evident for multiple cytokines including TNF, IL-1b, IL-1a, IL-1Ra, IL-6, CXCL-8 (IL-8), G-CFS, IL-12p40 and IL-10 (Figure 6 A). The low dose of VTX-294 (0.1 mM) induced a similar cytokine profile to MPLAmediated responses ( Figure A4 A-B). In newborn whole blood, a

VTX-294 Potently Activates Human Neonatal MoDCs
As antigen-presenting cells are key targets for candidate adjuvants, we next compared dose-dependent responses of human neonatal MoDCs to MPLA (1-100 ng/ml), R848 and VTX-294 (both 0.1-10 mM) with respect to production of TNF, as well as expression of HLA-DR and the co-stimulatory molecule CD86 (Figure 7). MPLA induced dose-dependent TNF production at both 10 and 100 ng/ml (p,0.05) compared to unstimulated DCs. Whereas VTX-294 induced MoDC TNF production at concentrations as low as 0.3 mM, R848 induced significant levels of this cytokine only at 10 mM. The greater potency of VTX-294 compared to R848 in neonatal MoDCs was also observed with respect to induction of surface HLA-DR and the significantly increased CD86 expression (Figure 7 B, C).

Discussion
Multiple licensed vaccines induce effective activation of innate immune responses by incorporating live attenuated pathogens or microbial products now known to contain TLR agonists [38]. However, with the continued development and use of inactivated subunit based vaccines, adjuvants have become increasingly important in order to overcome underdeveloped and distinct immune responses in immunocompromised populations such as newborns and infants [4]. Recent studies highlight that immunologic signatures can predict vaccine efficacy [14], suggesting that characterizing the immunostimulatory properties of candidate adjuvants may provide insight into their translational potential. We report here a broad characterization of secreted cytokine/ chemokine and inflammatory proteins in a neonatal whole blood system in response to the ultra-potent and selective TLR8 Agonist  VTX-294. We confirm prior studies that revealed a neonatal bias towards TLR-mediated IL-6 and IL-10 responses [6,39] and have also identified a similar neonatal bias towards TLR-mediated CXCL-8 (IL-8) production. Characterizing the ontogeny of innate immune responses may eventually inform development of adjuvant vaccine formulations that are tailored to induce distinct immune polarization for targeted neonatal and infant vaccines. Additional distinct characteristics of neonatal innate cells, including their reduced poly-functionality -the capability of an individual cell to produce multiple cytokines simultaneously, likely contribute to age-specific responses to TLR agonists [32].
TLR agonists are by now an established class of molecules with potential vaccine adjuvant properties [40]. Indeed, the detoxified LPS MPLA is a component of several licensed vaccines including Cervarix that is approved for school-aged girls [41]. Although both LPS and MPLA are agonists of TLR4, LPS triggers toxic inflammation whereas MPLA is potently immunostimulatory, but exhibits only 0.1 to 1% of the toxicity of its parent molecule and is employed as an adjuvant in [42]. Newborns demonstrate an impaired TNF response to LPS [32,43,44], but to our knowledge, ours is the first study to compare MPLA-induced cytokine production in cord and adult WB, revealing an impaired ability of human neonatal blood leukocytes and MoDCs to mount inflammatory responses to this TLR agonist.
We demonstrate for the first time that VTX-294, a small benzazepine molecule that specifically and potently targets TLR8, presents a unique opportunity for augmenting innate responses in human newborns. TLR8 (including dual TLR7/8) are refractory to the inhibitory neonatal adenosine system [27] and can therefore overcome suboptimal newborn innate immune responses to induce adult-level concentrations of several key cytokines [45,46]. We demonstrate here that VTX-294 induced key cytokines to concentrations comparable or greater than those observed in adult blood. We believe this is the first reported TLR agonist with such activity. Monocytes account for the majority of TLR-, including TLR8-, mediated TNF production in whole blood [18,25]. Additionally, as we compared TLR agonists in blood derived form the same newborns, our data suggest that robust responses induced by VTX-294 in newborn cord blood are not due to differences in cellular composition but rather to the inherent potency and efficacy of VTX-294. VTX-294 was not only at least as effective as R848 in inducing cytokine production, but also demonstrated ,10-fold greater potency in both WB and MoDC systems. When comparing to MPLA in newborn and adult blood, VTX-294 achieved higher maximum cytokine production for TNF, IL-10 and IL-12p40. When added to human neonatal MoDCs cultured in autologous plasma, VTX-294 was significantly more potent than R848, induced greater maximum TNF responses than MPLA. Even at the lowest concentrations tested VTX-294 also tended to enhance surface expression of HLA-DR and the co-stimulatory surface molecule CD86.
The potency of VTX-294 compared to R848 was also evident in comparing combinations of these agonists with MPLA. Combinations of TLR agonists may synergistically generate optimal cytokine production important for adjuvant activity [47]. For example, the live attenuated yellow fever vaccine YF-17D is one of the most effective vaccines available and activates multiple DC subsets via TLR2, 7, 8, and 9 [48]. Effective adjuvant systems already exist that combine MPLA and alum (AS04, developed by GlaxoSmithKline) [49] and in vitro studies with defined combinations of TLR agonists re-enforce the potential of such combined-adjuvant design [50,51]. More potent adjuvants (or adjuvant combinations) may allow for antigen dose sparing or may allow for single immunization strategies. Furthermore, addition of TLR8 agonists, either in conjugated form or as a freely soluble mixture, with vaccinal antigens enhances induction of adaptive immune responses to a greater extent than do TLR4 or TLR9 agonists [52]. The impact of impaired MPLA responses in early life may be seen in a recent phase 3 trial of the RTS,S/ AS01, an MPLA-adjuvanted malaria vaccine, that demonstrated reduced efficacy in African infants aged 6 to 12 weeks [53], as compared to those who were 5 to 17 months of age [54].
Overall, TLR8 agonists possess greater immunostimulatory properties than agonists of other TLRs, not only at birth (cord blood) but also in peripheral blood obtained during infancy, an important phase of susceptibility and immunization schedules [6]. Future in vitro studies should focus on characterizing the mechanisms underlying the potency and efficacy of VTX-294. TLR8 is preferentially localized to the early endosome and the endoplasmic reticulum but not to the late endosome or lysosome [55]. Subsequent studies should determine if the VTX compounds more effectively co-localize to these compartments. The phagosome is an efficient platform for the recognition of diverse ligands, sometimes involving co-operative interactions between TLR8 and other TLRs [56], or even other classes of PRRs [57]. Additionally, our lab has previously shown that CL075 can also activate DCs via other classes of PRRs, such as the inflammasome [17], which is supported by reports that imidazoquinolines can trigger induction of inflammatory cytokines in the absence of TLR 7 or 8 agonist activity [58]. Lastly, using molecular modeling studies, Govindaraj et al. demonstrated that the binding affinity of the TLR8 agonist R848 plays a key role in species-specificity [59]. Therefore, future studies to determine the binding affinity of VTX-294 in comparison with IMQs would be of interest. Future translational studies should evaluate the safety and efficacy of vaccine formulations containing TLR8 agonists in newborn animals, focusing on non-human primates that express TLR8 with structural and functional similarities to humans [60,61].
As safety and reactogenicity will be key considerations in developing novel adjuvanted vaccines, both the immunological and toxicological properties of novel vaccine formulations should be benchmarked against currently used adjuvants such as aluminum salts, oil emulsions, MPLA and approved pediatric vaccines that contain endogenous TLR agonists, such as Bacille Calmette-Guérin [62]. Finally, novel means of localizing the action of adjuvants, as has been done with the TLR7/8 agonist 3M-052, that contains a lipid tail modification of the imidazoquinoline that serves to localize its adjuvant action in tissues and prevent systemic cytokine induction, may represent additional approaches for enhancing safety while maintaining efficacy [63].
The data presented here support the potential adjuvant activity of VTX-294 in vitro. Furthermore, VTX-294 activates human newborn and adult leukocytes and MoDCs with potency and efficacy that exceed several other TLR agonists, including MPLA. VTX compounds can be produced to GMP quality in quantities that allow clinical trials, and have been recently tested in a Phase I clinical trials in patients with advanced solid tumors [30]. Our results suggest that VTX compounds may have additional applications as stand-alone immunomodulators or as vaccine adjuvants that can enhance neonatal DC function. Figure S1 VTX agonists do not activate TLRs-2, -3, -4, or -9. HEK-293 cells transfected with various human TLR and an NF-kB-driven reporter SEAP gene were stimulated for 18-24 h with TLR agonists. The y-axis shows the level of SEAP activity in the Quanti-blue TM assay by OD. The x-axis shows the concentration of each compound in mM with the exception of poly I:C, LPS, and flagellin which are expressed in mg/ml, ng/ml and mg/ml, respectively. Each data point represents the mean 6 SD of OD at 650 nm of triplicate culture wells. The positive control compound for each TLR is indicated in black.