Chemical and biological characterization of vaccine adjuvant QS-21 produced via plant cell culture

Abstract

Many vaccines, including those using recombinant antigen subunits, rely on adjuvant(s) to enhance the efficacy of the host immune responses. Among the few adjuvants clinically approved, QS-21, a saponin-based immunomodulatory molecule isolated from the tree bark of Quillaja saponaria (QS) is used in complex formulations in approved effective vaccines. High demand of the QS raw material as well as manufacturing scalability limitation has been barriers here. We report for the first-time successful plant cell culture production of QS-21 having structural, chemical, and biologic, properties similar to the bark extracted product. These data ensure QS-21 and related saponins are broadly available and accessible to drug developers.

Keywords: Bioactive plant product; Phytochemistry; cell; medical biochemistry; natural product synthesis.

Graphical abstract

Figure 1

Summary scheme demonstrating key steps of a sustainable manufacturing process to generate high quantities of QS saponin derived adjuvant using QS plant cell culture (A) Plant cells were initially expanded on solid base from stems isolated from different Quillaja saponaria trees in Chile. Then, in vitro plant cell culture was developed from solid callus to liquid suspension. Expanding cell cultures have been selected based on their profile of production of the QS-21 saponin fraction and went through an upstream process development variation in induction protocols to maximize biomass and saponin production. (B) The QS-21 isolated from plant cell culture process has been characterized to demonstrate its analytical, biochemical and biological comparability with QS-21 isolated from bark extract. (C) Cell culture of QS-21 is now being translated into a scalable and sustainable Good Manufacturing Process. Selected and validated cell lines are now being used in a scalable plant cell fermentation process to produce biomass. To isolate ccQS-21, similar methods of downstream processing used to produce clinical grade GMP beQS-21 composing the QS-21 Stimulon formulation are being implemented to extract and purify high yields of ccQS-21.

Figure 2

Analytical characterization of a representative development batch of ccQS-21 versus a reference batch of beQS-21 (A) UPLC-UV chromatogram detection demonstrating similar profile for cc and be crude extracts. (B) HPLC-UV chromatogram detection demonstrating similar retention time and purity of ccQS-21 and beQS-21. (C and D) The mass spectrometric (MS) fragmentation pattern of ccQS-21 is the same as beQS-21. The QS-21 1017.5 [M+2Na]²⁺ m/z parent ion (exact mass calculated for C₉₂H₁₄₈O₄₆Na⁺₂ as [M+2Na]²⁺: 1017.4519; observed: 1017.4512 [be], 1017.4511 [cc]) was subjected to MS/MS analysis and the same pattern of characteristic fragments was observed corresponding to the loss of the 3-O-linked trisaccharide producing a 1541.8 m/z ion (exact mass calculated. for C₇₅H₁₂₂O₃₁Na⁺ as [M+Na]⁺: 1541.7868; observed: 1541.7861 [be], 1541.7884 [cc]), and the elimination of quillaic acid from the remaining glycoside to form an apparent alkoxy enol ether resulting in a 1055.5 m/z ion (exact mass calculated for C₉₂H₁₄₈O₄₆Na⁺ as [M+Na]⁺: 1055.4523; observed 1055.4487 [be], 1055.4532 [cc]).

Figure 3

In vitro biological characterization; ccQS-21 and beQS-21 equivalently induce hemolysis, trigger the innate immune cell stress response, and enhance antigen processing and presentation to enhance the potency of the vaccinal immune synapse (A) Upon incubation with human red blood cells both beQS-21 and ccQS-21 retain equivalent saponin-specific hemolytic properties at pH-7.4 with matching half maximal effective concentration (EC50). (B) Upon pre-conditioning of THP1 human myeloid cells with a TLR4 agonist (LPS), both beQS-21 and ccQS-21 induce inflammasome-dependent secretion of IL-1β and the production of alarmins associated with pyroptosis such as HMGB1. Both IL-1β secretion and pyroptosis can be regulated by the canonical inflammasome pathway and can be suppressed by the disruption of the genes encoding for NLRP3, ASC and Caspase-1 or by the use of Caspase-1 inhibitor, ZVAD. (C) Both beQS-21 and ccQS-21 can comparably induce a broader inflammatory response in vitro. The breadth of the inflammatory response induced by the stimulation of human granulocyte macrophage colony stimulating factor (GMCSF) and IL-4 differentiated Monocyte derived Dendritic Cells (Mo-DCs) using both QS-21 was assessed with multiplex cytokine quantification. Both beQS-21 (BE) and ccQS-21 (CC) induce the same inflammatory signature which is further enhanced by formulating with TLR4 agonist. (D) Pseudo cross-presentation assay using cell viability to demonstrate that beQS-21 and ccQS-21 equivalently induce cytotoxicity through the regulation of the cytosolic translocation from the phagolysosome (e.g., as of the indicated toxin: saporin), a critical process for antigen cross presentation. (E) Flow cytometry analysis demonstrating that both QS-21 equivalently enhance large antigen processing and cross-presentation on MHC-I using a T cell receptor (TCR)-like antibody against the H-2K^b-SIINFEKL complex. (F) In vitro antigen cross-presentation assay demonstrating that ccQS-21 and beQS-21 equivalently enhance antigen processing and minimal epitope cross-presentation to activate transgenic T cells with matched OT1 TCR to express NFAT luciferase reporter. (G) OT1 T cell proliferation shows ccQS-21 and beQS-21 equivalently enhance antigen cross-presentation to induce the proliferation of primary T cells with matched OT1 TCR. OVA: Ovalbumin, OT1 long peptide: 29mer peptide including SIINFEKL, BFM: Bafilomycin A1, inhibiting the vacuolar type H⁺-ATPase (v-ATPase) to transfer protons into the lysosome.

Figure 4

In vivo biological characterization; both beQS-21 and ccQS-21 can comparably induce an inflammatory response in vivo and enhance the draining lymph nodes (DLN) priming (A) Mice were conditioned using the air pouch experimental model to characterize the site of injection and DLNs innate immune response 12 h post-stimulation. (B and C) Both QS-21 induces the same spectrum of inflammatory cytokine response at the site of injection (B-C) and systemically (Figure S4A). (D and E) Immune phenotyping by multiplex fluorescent Immunohistochemistry (IHC) of DLNs after overnight (O/N) exposure to adjuvants revealed that both beQS-21 and ccQS-21 comparably enhance the trafficking of antigen (OVA) to the DLNs and induce specific DLN reorganization by targeting CD169+ subcapsular macrophages and inducing the infiltration of myeloid cells, and the activation of B cells in the germinal centers (Figure S4C). (F) Immune phenotyping by flow cytometry of DLNs and non-draining lymph nodes (NDLNs) after O/N exposure to adjuvants confirmed that both beQS-21 and ccQS-21 can equivalently induce innate immune myeloid and lymphoid cell recruitment and activation in the vaccinal DLNs. OVA, Ovalbumin.

Figure 5

In vivo biological characterization; ccQS-21 and beQS-21 have equivalent vaccine adjuvant potency in young and elderly mice (A) Prime-boost (day 0/day 14) approach used in young versus elderly C57BL6/J mice with sub-unit ovalbumin vaccine model. (B and C) Both ccQS-21 and beQS-21 can equivalently induce a broad anti-ovalbumin T_H1 humoral response with higher ratios of anti-OVA Ig2c versus IgG1 as compared to reference adjuvants in young (B) and elderly (C) mice. (D‒G) the ex vivo anti-ovalbumin T cell response demonstrated that both ccQS-21 and beQS-21 enhance anti-ovalbumin immune response priming. (F and G) The ex vivo T cell restimulation cytokine signature further demonstrated an equivalent ability to trigger a T_H1 biased response. These properties can further be enhanced by combining the beQS-21 and ccQS-21 with MPLA to potentiate the T_H1 vaccinal response in young animals (Figure S5A).

Figure 6

In vivo biological characterization; both ccQS-21 and beQS-21 formulated with MPLA in liposomes (AS01 mimetic formulations) efficiently adjuvant the R21 malaria vaccine and Shingrix vaccine (A) Using a prime-boost vaccination regimen with the Shingrix antigen (VZV gE) adjuvanted with the different test adjuvant formulations, we demonstrated that AS01 could be substituted by mimetic formulations using ccQS-21 and beQS-21. (B–E) All QS-21 formulations were equivalent at priming T_H1 humoral (B-C) and a broad T_H1-2-17 cellular (D-E) responses against the Shingrix antigen. (F) Using an experimental prime-boost-boost anti-malaria vaccination model with the R21 VLP vaccine followed by rechallenge with a surrogate malaria parasite, we further demonstrated the equivalent adjuvant potency of ccQS-21 and beQS-21. (G and H) Both ccQS-21 and beQS-21 prime an anti-CSP-NANP repeat humoral response leading to an efficient protection of 100% of the animal against parasite challenge, as measured by the daily parasitemia evaluation over a 10-day postexposure period. VLP, virus like particles; CSP, circumsporozoite protein; NANP, Immunogenic NANP repeats.