characterization of the innate stimulatory capacity …...characterization of the innate stimulatory...

Vaccine xxx (xxxx) xxx

Contents lists available at ScienceDirect

Vaccine

journal homepage: www.elsevier .com/locate /vacc ine

Characterization of the innate stimulatory capacity of plant-derivedvirus-like particles bearing influenza hemagglutinin

https://doi.org/10.1016/j.vaccine.2018.10.0990264-410X/� 2018 The Author(s). Published by Elsevier Ltd.This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).

⇑ Corresponding author.E-mail address: [email protected] (C.M. Krawczyk).

1 Authors contributed equally to this work.

Please cite this article as: S.-Y. Won, K. Hunt, H. Guak et al., Characterization of the innate stimulatory capacity of plant-derived virus-like particlesinfluenza hemagglutinin, Vaccine, https://doi.org/10.1016/j.vaccine.2018.10.099

So-Yoon Won a,1, Kristin Hunt a,1, Hannah Guak a, Benedeta Hasaj a, Nathalie Charland b,Nathalie Landry b, Brian J. Ward c, Connie M. Krawczyk a,⇑aDepartments of Physiology and Microbiology & Immunology, Goodman Cancer Research Center, 3655 Promenade Sir William Osler, Montréal, Québec H3G 1Y6, CanadabMedicago Inc., 1020 route de l’Église, suite 600, Québec, Québec G1V 3V9, CanadacResearch Institute of the McGill University Health Centre, 1001 Décarie Street, Montréal, Québec H4A 3J1, Canada

a r t i c l e i n f o

Article history:Received 19 March 2018Received in revised form 29 October 2018Accepted 31 October 2018Available online xxxx

Keywords:InfluenzaVirus-like particlePlant-derived vaccineDendritic cellInnate immunityType 1 response

a b s t r a c t

Cell-mediated immunity is an important component of immediate and long-term anti-viral protection.Dendritic cells (DCs) are essential for the induction of cell-mediated immunity by instructing the activa-tion and differentiation of antigen-specific T cell responses. Activated DCs that express co-stimulatorymolecules and pro-inflammatory cytokines are necessary to promote the development of type 1 immuneresponses required for viral control. Here we report that plant-derived virus-like particles (VLPs) bearinginfluenza hemagglutinins (HA) directly stimulate mouse and human DCs. DCs exposed to H1- and, to alesser extent, H5-VLPs in vitro rapidly express co-stimulatory molecules and produce pro-inflammatory cytokines including IL-12, IL-6 and TNFa. Furthermore, these VLPs support the activationand differentiation of antigen-specific T cell responses. Mechanistically, H1-VLPs stimulate the activationof kinases typically activated downstream of pattern recognition receptors including AKT, p38, andp42/44 ERK. In vivo, immunization with plant-derived VLPs induce the accumulation of both cDC1sand cDC2 in the draining lymph node and a corresponding increase in T and B cells. VLPs devoid of HAprotein activate DCs, suggesting they are intrinsically immunostimulatory. Together, the results demon-strate that these candidate plant-derived VLP vaccines have an inherent and direct stimulatory effect onDCs and can enhance the ability of DCs to promote Type 1 immune responses.

� 2018 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license(http://creativecommons.org/licenses/by/4.0/).

1. Introduction

Influenza viruses infect the human respiratory tract where theycause a spectrum of pathologies ranging from mild illness to life-threatening disease [1]. Because these viruses mutate constantlyin both humans and reservoir species, yearly vaccination remainsthe most effective method of prevention and containment [2]. Tra-ditionally, hemagglutination-inhibition (HI) antibody responsewas used as a surrogate marker correlating with vaccine efficacy[3]. However, the cellular response (CD4+ and CD8+ T cells) cancontrol infection in both animal models and humans [4–7] andmay be particularly important in elderly subjects who often mountweak or no antibody responses to standard, split virus vaccines[8,9]. CD4+ T cells contribute to the control of viral infections bysustaining CD8+ T cell responses, supporting B cell functions andsupporting long-term T and B cell memory [10,11]. Cross-reactive

CD4+ T cells that recognize conserved epitopes from the HA proteinmay therefore recognize drifted strains and participate in cross-protection [12]. Although the currently available vaccines haveproved to be safe and to provide some degree of protection (i.e.efficacy averaging 40–60% in young healthy adult populations ifvaccine matches circulating strain), it is increasingly clear thatnew vaccines capable of inducing broader and more durableimmune responses (i.e. humoral, cellular, cross-reactive) areneeded to improve vaccine efficacy [13,14].

Virus-like particle (VLP) vaccines for influenza are self-assembling protein structures that contain viral antigens embed-ded in a lipid bilayer but lack any genetic material [15,16]. Bothpre-clinical (e.g. mouse, ferrets) and clinical studies have demon-strated that plant-derived VLP vaccine candidates bearing influ-enza hemagglutinin (HA) proteins can generate long-lived, cross-reactive HA-specific humoral responses as well as poly-functionalCD4+ T cell responses to both seasonal and avian strains [17–20].The mechanisms by which these vaccines induce this broaderimmune response are not fully understood. Because dendritic cells(DCs) are regarded as the principal antigen-presenting cell to both

bearing

http://creativecommons.org/licenses/by/4.0/

https://doi.org/10.1016/j.vaccine.2018.10.099

http://creativecommons.org/licenses/by/4.0/

mailto:[email protected]


http://www.sciencedirect.com/science/journal/0264410X

http://www.elsevier.com/locate/vaccine


2 S.-Y. Won et al. / Vaccine xxx (xxxx) xxx

stimulate and direct T cell immunity [21] and the activation of DCsis a necessary step in eliciting effective cellular responses againstviruses, we examined whether these plant-derived VLPs hadintrinsic immunostimulatory potential. We tested two VLP formu-lations; H1-VLP (Influenza A/California/04/09) and H5-VLP (A/Indonesia/05/2005 strain) [16] with and without adjuvant.

We observed that unadjuvanted H1-VLPs, and to a lesser extentH5-VLPs, stimulate both murine and human DCs to express co-stimulatory molecules and secrete pro-inflammatory cytokinesincluding the T helper 1 (Th1) promoting cytokine IL-12. DCs stim-ulated with H1-VLP, and to a lesser extent H5-VLP, promoted theproliferation, differentiation and effector function of CD4+ T cells.Exposure of murine DCs to H1-VLP also promoted cross-presentation of a simultaneously-administered model antigenovalbumin (OVA) and CD8+ T cell effector differentiation. Mecha-nistically, we found that the plant-derived VLPs stimulated theactivation of kinase cascades that are typically associated with pat-tern recognition receptor (PRR) ligation and necessary to drive thetranscriptional program that supports pro-inflammatory DC func-tion. When injected in vivo, plant-derived VLPs stimulated theaccumulation of DCs, T cells and B cells in the lymph node. Theseresults demonstrate that plant-derived VLPs can directly stimulateDCs to acquire an activation phenotype that supports the develop-ment of Type 1 cellular immune responses.

2. Materials & methods

2.1. Mice

Female wild-type C57BL/6 or BALB/c mice were purchased fromCharles Rivers Laboratories at 6–8 weeks of age (Montreal, QCCanada). OTI and OTII transgenic mice were purchased from theJackson Laboratory (Bar Harbor, ME USA) and bred in house. Ani-mals were maintained in a specific pathogen-free environment.All experiments were conducted following the guidelines of theCanadian Council of Animal Care, as approved by the animal carecommittee of McGill University.

2.2. Production of influenza VLP vaccines

VLP vaccines were obtained from Medicago Inc. (Quebec City,QC) and produced as previously described [15,17]. H1-VLPs containHA from A/California/7/09 (H1N1) and H5-VLPs contain HA from A/Indonesia/5/05 (H5N1). A detailed characterization of the mor-phology and biochemical content is reported here [15,22]. In someexperiments, ‘empty’ VLPs (eVLP) were used as an additional con-trol. These nanoparticles are generated from plants transientlytransfected with ‘null’ A. tumifasciens. They are approximately thesame size as the HA-bearing VLPs and have very similar lipid com-position (data not shown). However, since the influenza HA trimersdrive budding of the VLPs, the down-stream processes used to iso-late and purify the eVLPs are not identical to those used to generatethe HA-VLPs.

2.3. Murine BMDC culture

Bone marrow was extracted from the femur and tibia ofC57BL/6 or BALB/c mice. Briefly, bone marrow cells were differen-tiated to DCs in the presence of GM-CSF (20 ng/ml) in filtered com-plete DC media (cDCM: RPMI containing 10% FCS, 100 U/mlpenicillin, 100 mg/ml of streptomycin, 50 mM b-Mercaptoethanol,2 mM L-Glutamine (Wisent) and 1% non-essential amino acids)(Life Technologies/Thermofisher) on uncoated TC plates (Falcon)as described [23]. On day 8–10 of culture, immature DCs were har-vested and stimulated in 96-well non-TC plates with either

Please cite this article as: S.-Y. Won, K. Hunt, H. Guak et al., Characterization of tinfluenza hemagglutinin, Vaccine, https://doi.org/10.1016/j.vaccine.2018.10.09

10 ng/mL Escherichia coli LPS (serotype 0111:B4, Sigma), 5 mg/mLof H1-VLP, 5 mg/mL of H5-VLP (Medicago Inc, Ste Foy, QC: seeabove), 5 mg/ml of Alhydrogel (alum) (Brenntag/Cedarlane), VLPswith alum at 1:1 ratio, or 8.58 EU/mL of Agrobacterium-LPS(Agro-LPS) (List Biologicals Laboratories USA), 5 mg/ml OK432 (gen-erously provided by Chugai Pharmaceuticals Co. INC., Japan) or asotherwise noted. After 18 h of incubation at 37 �C, DCs were col-lected and analyzed by flow cytometry (FACS CANTOII, FORTESSA,Cell Vision Flow Cytometry Facility, McGill University).

2.4. Human monocyte-derived DC culture

After written informed consent, whole blood was obtained,peripheral blood mononuclear cells (PBMCs) were collected fromthe venous blood of healthy young adults (18–30 years old) atthe MUHC Vaccine Study Center (Pierrefonds, QC). PBMCs wereseparated on a Ficoll density gradient and CD14+ monocytes werepositively selected (Stem Cell Technologies, B.C. Canada), resus-pended in cDCM and then plated in 6-well tissue culture plates(Falcon) at a concentration of 1 million cells/mL. Differentiationinto moDCs was completed using human GM-CSF (100 ng/mL)and human IL-4 (50 ng/mL) (Peprotech). Human moDCs were fedwith 1 ml of fresh media on day 3 and 6 of culture. On day 7,moDCs were harvested and activated with either 10 ng/ml LPS,5 mg/mL of H1-VLP, 5 mg/mL of H5-VLP, 5 mg/mL of Alhydrogel(alum), VLPs with alum at 1:1 ratio, or 8.58 EU/mL of Agro-LPS.After 30 min of incubation at 37 �C, moDCs were collected, proteinexpression was analyzed byWestern. Otherwise, after 18 h of incu-bation at 37 �C, moDCs were collected and analyzed by flowcytometry. These studies were approved by the Research EthicsBoard of the McGill University Health Centre.

2.5. ELISA

Cell culture supernatants from DCs stimulated with LPS, VLPs orVLPs with alum, were collected after 18 h and quantified for IL-12p70, IL-12p40, IL-6, IL-10, IL-1b and TNFa by enzyme-linkedimmunosorbent assay (ELISA) kits (eBioscience/Affymetrix)according to the manufacturer’s instructions.

2.6. DC: T cell co-culture assays

On day 10 of culture, BMDCs from C57BL/6 mice were harvestedand cultured for 6 h with whole ovalbumin (OVA) protein (inhouse) (0.001 mg, 0.01 mg or 0.1 mg per well) with either Agro-LPS (0.5 ng/mL), H1-VLP (3 mg/mL) or H5-VLP (3 mg/mL). T cellsfrom OTII or OTI transgenic mice were isolated from the spleenand lymph nodes using CD4+ and CD8+ positive selection kits (Stemcell), respectively. Sorted T cells were then co-cultured with theactivated DCs at a 1:10 ratio of DCs to T cells. Prior to co-culture,half of the isolated T cells were additionally stained with a violetproliferation dye (eBioscience/Affymetrix) to examine prolifera-tion. On day 3–5, T cells were collected and analyzed by flowcytometry.

2.7. Flow cytometric analysis

Surface staining: Following activation, live DCs or T cells werestained with fluorochrome-conjugated antibodies and acquiredon a FACSCANTO II or FACSFortessa flow cytometer (BD Pharmin-gen). Analysis was performed using FlowJo software (Tree Star,Oregon, USA). See figure legends for analysis gates. Intracellularstaining: T cells were stimulated with 3 mg/mL Brefeldin A (eBio-science/Affymetrix), 0.5 mg/ml ionomycin (Sigma) and 0.05 mg/mlPMA (Sigma) for 4 h prior to their harvest. T cells were stained withfluorochrome-conjugated antibodies against surface markers, for

he innate stimulatory capacity of plant-derived virus-like particles bearing9


S.-Y. Won et al. / Vaccine xxx (xxxx) xxx 3

60 min at 4 �C in the dark. T cells were then fixed and permeabi-lized with Foxp3 fix/perm buffer from eBioscience/Affymetrix for40 min. Intracellular proteins were stained overnight at 4 �C. Seefigure legends for analysis gates. All conjugated antibodies usedin flow cytometry analyses were purchased from eBioscience/Affy-metrix: Anti-human CD11c-Percp710 3.9; anti-human CD86-PeCy7 IT2.2; anti-human CD83-PE HB15e; anti-human CD14-FITC61D3; anti-mouse CD11c-e450 N418; anti-mouse CD11c-APC-780N418; anti-mouse CD86-FITC GL1; anti-mouse CD86-PeCy7GL1; anti-mouse CD40-APC 1C10; anti-mouse MHCII-AF700M5/114.15.2; anti-mouse MHCII-APC-780 M5/114.15.2; anti-mouse MHCII-PE M5/114.5.2; anti-mouse Flt3 (CD135)-PE A2F10;anti-mouse CD115-AlexaFluor488 AFS98; anti-mouse CD62L-APCMEL-14; anti-mouse CD4-APC-780 RM4-5; anti-mouse CD4-e450RM4-5; anti-mouse CD8-APC-780 53–6.7; anti-mouse CD8-e45053–6.7; anti-mouse CD44-e450 IM7 anti-mouse CD44-PercpCy5.5IM7; anti-mouse CD44-PeCy7 IM7; anti-mouse CD25-PE PC61.5;anti-mouse CD25-PeCy7 PC61.5; anti-mouse XCR1-BV650 ZET;anti-mouse CD3- FITC 17A2; anti-mouse CD3- Biotin 17A2; strep-tavidin V500 (BDHorizon 561419); anti-mouse B220-APC RA3-6B2; anti-mouse B220-FITC RA3-6B2; anti-mouse Ly6G- FITC1A8; anti-mouse NK1.1-FITC PK136; anti-mouse CD69 FITCH1.2F3; anti-mouse CD172a-PeCy7 P84; anti-mouse T-bet-PeCy7eBio4B10; anti-mouse IFNc-FITC XMG1.2; anti-mouse IFNc-APCXMG1.2; anti-mouse Granzyme B-FITC NGZB; anti-mouse IL-2-PEJES6-5H4; anti-mouse CD16/CD32 FcR Block 93; 7AAD ViabilityStaining Solution; Fixable Viability APC-780; Cell Proliferationdye efluor 450; 123count eBeads Counting Beads.

2.8. qRT-PCR

RNA from BMDCs was isolated using Trizol (Life Technologies,ThermoFisher), cDNA were synthesized using random primersand multiscribe reverse transcriptase (Applied Biosystems). Real-time qRT-PCR analysis was performed using SYBR green (Frogga-Bio). Relative expression of IL-12p40 and pro-IL-1b to HPRT wascalculated using the 2-DDCt method. Real-time primers for HPRT,IL-12p40 and pro-IL-1bwere designed by Primer3 and PrimerBlast.The primer sequences used are listed below: Hprt Fwd: CTC CGCCGG CTT CCT CCT CA Rev:ACC TGG TTC ATC ATC GCT AAT C IL-12p40 Fwd: CTG GAG CAC TCC CCA TTC CT Rev: CGC CTT TGCATT GGA CTT CG Il-1b Fwd: GCC ACC TTT TGA CAG TGA TGA GRev:AAG GTC CAC GGG AAA GAC AC.

2.9. In vivo footpad injections

5 lg of VLPs, 0.1 lg of LPS, or PBS were injected into each foot-pad of female wild-type C57BL/6 mice at 6–8 weeks of age. After24–36 h, the draining popliteal lymph nodes were digested andstained with antibodies for flow cytometry.

2.10. Statistics

Statistical analysis on murine data was performed using one-way ANOVA using GraphPad Prism. Statistical analysis on humandata was performed using the Friedman test (nonparametric test).Single stars above bars indicate statistically significant differencerelative to control. Lines indicate where direct comparisons weremade between groups. Differences were considered significant atp < 0.05.

3. Results

To examine whether plant-derived VLPs bearing influenza HAproteins could directly activate DCs, BM-derived DCs from both


C57BL/6 and BALB/c mouse strains were stimulated with H1-VLPs or H5-VLPs and examined for expression of co-stimulatorymolecules and cytokine production. LPS from E. coli was used asa control in all experiments. DCs exposed to H1-VLPs increasedthe expression of co-stimulatory molecules CD86 and CD40, mark-ers of proinflammatory DC activation which were further increasedby alum (Fig. 1A and B). H5-VLPs were less stimulatory than H1-VLPs although their activity was also enhanced by alum (Fig. 1Aand B). H1-VLPs induced the secretion of pro-inflammatory cytoki-nes IL-12p40, IL-6, and TNFa and levels were enhanced with theaddition of alum (Fig. 1C). In contrast, pro-inflammatory cytokinesecretion in response to H5-VLPs was generally weak and wasnot significantly enhanced by alum (Fig. 1C). IL-10 secretion, whichcan be induced in murine DCs by LPS, also increased in response toH1-VLP stimulation, but not H5-VLPs. Increasing the concentrationof H5-VLPs improved their immunostimulatory capacity, but cyto-kine levels were never comparable to those stimulated by H1-VLPs(Fig. S1). Together, these findings demonstrate that H1-containingVLPs, and to a lesser extent H5-VLPs, can directly stimulate murineDCs to acquire a pro-inflammatory phenotype.

Since these VLPs are produced in plants using a Gram-negativebacterial vector (Agrobacterium tumefaciens), small amounts ofAgro-LPS are found in VLP preparations (range �312–1610 EU/mL in Drug Substance). Thus, the possible effect of Agro-LPS onmurine DC activation was evaluated. DCs from young mice werestimulated for 18 h with 1.72 ng/ml (8.58 EU/ml) of Agro-LPS(amount found in 5 mg/ml of H1-VLP) and the induction of activa-tion marker expression and cytokine production was compared to1.72 ng/ml of E. coli LPS (serotype 0111:B4). DCs stimulated withAgro-LPS alone expressed CD86 and CD40 levels almost identicalto the unstimulated control (Fig. S2). The addition of alum withAgro-LPS did not increase CD86 or CD40 expression (data notshown).

We also examined whether the VLPs could similarly stimulatehuman DCs. Human monocyte-derived DCs (moDCs) were stimu-lated with either 5 mg/mL of H1- or H5-VLP for 18 h with or with-out alum. Following stimulation, the activation status wasevaluated using CD83 and CD86 expression as well as cytokinesecretion. We observed that both H1- and H5-VLPs significantlyinduced the expression of CD83 and CD86 compared to the nega-tive control and expression was not increased by alum (Fig. 1D).H1-VLP (with or without alum) significantly increased the secre-tion of IL-6, TNFa and IL-10 (Fig. 1E). Although there was a trendobserved in some experiments for increased cytokine secretionwith H5-VLP stimulation, significance was not reached as com-pared to the control cultures (Fig. 1E). Neither alum nor Agro-LPSon their own induced up-regulation of activation marker expres-sion (Fig. 1D) and did not stimulate any significant cytokine secre-tion (Fig. 1E). With respect to IL-6, IL-10, and TNFa production andCD83 expression, H1-VLPs were generally more stimulatory thanH5-VLPs. However, differences between H1- and H5-VLPs werenot observed for CD86 upregulation and IL-12 production.

The promotion of effective Type 1 immunity (including bothTh1 CD4+ and CD8+ cytotoxic T lymphocyte responses (CTL) is nec-essary for optimal protection from most viruses and is thus animportant criterion for evaluating novel influenza vaccine candi-dates. Because the plant-derived HA-VLPs stimulated the produc-tion of Type 1 immunity promoting cytokines by both murineand human DCs (IL-12), we examined whether the VLP-stimulated DCs could promote the development of Th1 responses.DCs from C57BL/6 mice were exposed to H1- and H5-VLPs for 6 hin the presence of whole ovalbumin (OVA) and subsequently co-cultured with OVA-specific TCR transgenic CD4+ T cells from OTIItransgenic mice. CD4+ T cell responses were evaluated by flowcytometry based on the extent of proliferation, activation markerexpression (CD44 and CD25), Th1 transcription factor (T-bet) up-



Fig. 1. DC activation by H1- and H5-VLPs. DCs from BALB/c (A) or C57BL/6 mice (B and C) were stimulated for 18 h with 5 mg/ml of H1 or H5-VLPs, with and without 5 mg/mlof alum, and compared to LPS (10 ng/ml), Agro-LPS (8.58 EU/ml) and the unstimulated control. Flow cytometry was used to evaluate the expression of CD86, CD40, MHCII andMHCI (A and B). (C) Secretion of IL-12p70, IL-12p40, IL-6, TNFa and IL-10 were measured by ELISA. Results are shown for one representative of at least three independentexperiments. Data are shown as mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ****0 < 0.0001. Single stars indicate significance relative to control. Line indicates comparisonbetween 2 groups. (D and E) Human moDCs from 14 young donors (18–30 years old) were stimulated with 5 mg/ml of H1 or H5-VLPs (with and without 5 mg/ml of alum), LPS(10 ng/ml) or Agro-LPS (8.58 EU/ml) compared to the unstimulated control. D) Flow cytometry was used to evaluate the expression of CD83 and CD86 on CD11c+ CD14- livecells. E) Production of IL-12p70, IL-6, TNFa and IL-10 from human moDCs was measured by ELISA. Each dot represents a different donor and data are shown as mean ± SEM.*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.


Please cite this article as: S.-Y. Won, K. Hunt, H. Guak et al., Characterization of the innate stimulatory capacity of plant-derived virus-like particles bearinginfluenza hemagglutinin, Vaccine, https://doi.org/10.1016/j.vaccine.2018.10.099


Fig. 2. H1- and H5-VLPs promote CD4+ T-cell proliferation, differentiation and effector function. DCs from C57BL/6 mice were activated for 6 h with LPS (0.5 ng/ml) or HA-VLPs (3 mg/ml) in the presence of 0.001 mg (A–C) or 0.1 mg (D) of whole OVA protein and co-cultured with CD4+ T-cells from OTII transgenic mice at a 1:10 ratio. On day 4,CD4+ T-cells were analyzed by flow cytometry and stained for (A) CD44 expression and proliferation, B) CD25 expression and (C) T-bet expression. B, (C) numbers to right ofhistograms is geometric MFI for sample shown. On day 5, T-cells were also stained for IFNc production (D). Histograms and dot plots are gated on CD4+ live cells. Data isshown from one representative of at least three independent experiments. Data are shown as mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. #statistical significancenot achieved in every experiment.


regulation and effector cytokine secretion (IFNc). We observed thatboth H1- and H5-VLP-stimulated DCs promoted the activation(measured by CD44 expression) and proliferation of CD4+ T cells(Fig. 2A and B). Both H1- and H5-VLP-stimulated DCs promoted asignificant increase in T-bet expression compared to OVA-stimulated DCs, highlighting the promotion of a Th1 program(Fig. 2C). Consistently, the percentage of activated CD4+ T cells pro-ducing IFNc (CD44+IFNc+) was also increased by H1- and H5-stimulated DCs, indicating that both VLPs can induce Th1 effectorfunction (Fig. 2D). Although the increase in Th1 differentiation byH5-stimulated DCs was routinely observed, statistical significancewas not achieved in every experiment. Overall, the CD4+ T cellresponses promoted by DCs exposed to H1-VLPs were strongerand more consistent than those observed following H5-VLP stimu-lation, similar to the differences observed in murine DC activation.

VLPs are extracellular particles therefore antigen must be cross-presented on MHCI in order to stimulate antigen-specific CD8+ Tcell responses. Thus, the ability of VLP-stimulated DCs to cross-present antigen and stimulate CD8+ T cells was also evaluated.DCs from C57BL/6 mice were stimulated for 6 h with H1- or H5-VLPs in the presence of OVA and subsequently co-cultured withOVA-specific TCR transgenic CD8+ T cells from OTI transgenic mice.Activation marker expression (CD44), T-bet upregulation andeffector molecule secretion (IL-2, IFNc and granzyme B) wereexamined. The percentage of activated CD8+ T cells producing IL-2 (CD44+IL-2+) was significantly increased with H1-, but not H5-stimulated DCs compared to OVA alone (Fig. 3A). A trend for H1-and H5-VLP stimulated DCs to activate CD8+ T cells to produce


IFNc (CD44+IFNc+) was also observed (Fig. 3A). The expression ofT-bet and granzyme B, as well as the percentage of activatedCD8+ T cells producing T-bet and granzyme B, were significantlyincreased with H1-VLP-stimulated DCs compared to negative con-trols (Fig. 3B). Again, the changes induced with H5-VLP-activatedDCs trended in the same direction but did not reach significancecompared to the OVA control (Fig. 3B). Together, these findingssuggest that H1-VLP, and to some extent H5-VLP, can stimulateDCs to cross-present antigen to CD8+ T cells and promoteantigen-specific CD8+ T cell responses.

To begin to understand the mechanisms by which the plant-derived VLPs mediate DC activation, we first determined whetherthey stimulate a priming signal (much like pattern recognitionreceptor ligation), or whether their particulate nature activatedmechanisms associated with inflammasome activation. In murine,but not human DCs, IL-1b secretion is dependent on two signals; apriming signal that leads to Il1b gene transcription and then a sec-ond signal mediated by the inflammasome that results in IL-1bprotein cleavage and secretion. We found that exposure of DCs tothe H1-VLP increased both Il1b and Il12b transcription. (Fig. 4A).Although H5-VLP stimulated significant transcription of pro-IL-1bcompared to the control, the levels were similar to Agro-LPS andwere less than H1-VLP (Fig. 4A). However, VLPs on their own couldnot induce the secretion of active IL-1b and required the additionof alum, which stimulates inflammasome activation, for significantsecretion of IL-1b (Fig. 4B). Together these results indicate that theplant-derived VLPs, in particular H1-VLP, acts as signal 1 to primethe activation of DCs and that, as expected, alum then acts as signal



Fig. 3. H1-VLPs promote CD8+ T-cell effector function and cross presentation of antigen. DCs from C57BL/6 mice were activated for 6 h with LPS (0.5 ng/ml) or HA-VLPs (3 mg/ml) in the presence of 0.01 mg of whole OVA protein and cultured with CD8+ T-cells from OTI transgenic mice at a 1:10 ratio. On day 3, CD8+ T-cells were analyzed by flowcytometry and stained for (A) CD44, IFNc and IL-2 expression and (B) T-bet and granzyme B expression. Histograms and dot plots are gated on CD8+ live cells. Results shownare from one experiment and representative of at least three independent experiments. Data are shown as mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.


2 to trigger inflammasome assembly, resulting in pro-IL-1b cleav-age. Human moDCs stimulated with H1-VLP or LPS were able toproduce significant amounts of IL-1b in the absence of alum(Fig. 4C). This result is consistent with previous reports that humanmonocytes have a path to IL-1b production distinct from mousecells that does not require a second signal [24–26].

Since H1-VLP was shown to act as a strong priming signal for DCactivation, we determined which intracellular signaling pathwaysH1-VLP might activate to induce the expression of inflammatorycytokines and co-stimulatory molecules. DCs were exposed toH1- or H5-VLPs and the activation of signaling pathways knownto be important for DC activation were examined. As expected,LPS strongly activated many kinases known to be downstream ofPRRs (Fig. 4D). In DCs, H1-VLP induced the phosphorylation and


thus activation of AKT, p42/44 and p38, but only minimally acti-vated JNK (Fig. 4D). H5-VLP induced activation of AKT, p38 andto a lesser extent p42/44 but all were less strongly induced thanwith H1-VLP exposure. Similar results were obtained from humanmonocyte-derived DCs (Fig. 4E). Together, these results demon-strate that H1-VLP, and to a lesser extent H5-VLPs, directly activatekinase cascades that lead to the activation of DCs with their pro-duction of pro-inflammatory cytokines and ability to stimulate Tcell responses.

To determine whether H1 and H5-VLPs stimulated DCs in vivo,the footpads of mice were injected with H1-VLP or H5-VLP. Injec-tion of the VLPs led to a trend for increased numbers of total DCs,conventional DC1s (cDC1s; XCR1+) and cDC2s (CD172a+) in thedraining lymph nodes 36 h following immunization (Fig. 5A). Con-



Fig. 4. H1-VLPs are a priming signal in murine DCs. A) The mRNA levels of Il-1b and Il-12p40 were measured by qRT-PCR after DCs were activated for 6 h with 5 mg/ml of H1-or H5-VLPs, with or without 5 mg/ml of alum, compared to Agro-LPS (8.58 EU/ml) and LPS (10 ng/ml). B) The secretion of active IL-1b by DCs in response to VLP and alumstimulation was measured by ELISA. Data from (A) and (B) are shown as mean ± SD. and are representative of three independent experiments. (C) IL-1b secretion by humanmoDCs from 14 adult donors (18–30 years old). Each dot represents one donor and data are shown as mean ± SEM. Mouse (D) and human (E) DCs were activated for theindicated times with 10 ng/ml of LPS, 5 mg/ml of H1-VLP or 5 mg/ml of H5-VLP. Western blot results show the phosphorylation (p) of: AKT, JNK, p42/44 (ERK1/2), and p38.Tubulin (D) or b-actin (E) was used as a loading control. Results are from one experiment and representative of three independent experiments. *p < 0.05, **p < 0.01,***p < 0.001, ****p < 0.0001.


sistent with DC activation and lymphocyte recruitment, we found asignificant increase in the total number of B cells and T cells(Fig. 5B), some of which expressed markers of activation (CD69+

and CD44+ respectively). These experiments demonstrate thatH1- and H5-containing VLPs have inherent immunostimulatoryactivity in vivo.

Influenza HA proteins are not thought to have any intrinsicimmunostimulatory properties on their own. Therefore, weassessed the ability of the VLPs devoid of HA proteins to stimulate


murine and human DCs. Biochemical analyses of the VLPs revealthe presence of low levels of residual host cell proteins and gluco-sylceramides associated with plant lipid rafts [15]. Empty VLPs(eVLP), the VLP without the HA proteins, stimulated co-stimulatory molecule expression and cytokine production by DCs(Figs. 6A, B and S3). LPS and OK432 bacterial vaccine were usedas controls. DCs stimulated with eVLPs also induced in vitro T cellresponses (Fig. 6C), demonstrating enhanced function of the stim-ulated DCs. As in Fig. 5, eVLPs were injected into the footpads of



Fig. 5. Immunization of H1- and H5-VLPs stimulates the accumulation of DCs, B cells, and T cells in lymph nodes in vivo. 5 lg of VLPs, 0.1 lg of LPS, or PBS were injected intoeach footpad of female wild-type C57BL/6 mice at 6–8 weeks of age. After (A) 24 h or (B) 36 h, the draining popliteal lymph nodes were digested and analyzed for lineage-specific and cell activation markers. Cell numbers were calculated using counting beads. Each dot represents one mouse and data are shown as mean ± SEM. *p < 0.05,**p < 0.01.


mice and compared to PBS controls. eVLPs induced substantial DCaccumulation in the draining LN indicating that the eVLPs werestimulatory in vivo. The proportion of cDC subsets were not signif-icantly different upon eVLP immunization, with an increase in bothcDC1s and cDC2s (Fig. 6D). To determine whether eVLPs also stim-ulate human DCs, moDCs were stimulated with eVLP. eVLPsinduced the activation of p38 and p44/42 but minimal activationof AKT (Fig. 6E).

4. Discussion

High serum antibody titers can protect against many viral infec-tions, including influenza. Although antibodies that mediatehemagglutination inhibition (HI) have long been used as surro-


gates to license influenza vaccines [3], there is increasing interestin formulations that can induce more effective and broader immu-nity. Plant-derived VLP vaccines bearing the influenza virus HAprotein have recently been shown to generate broad and cross-reactive cellular and humoral immune responses in both animalmodels [16] and clinical trials [17]. In animal challenge studieswith highly pathogenic avian strains, some of the VLP-vaccinatedanimals are protected despite very low or even undetectable anti-body titers [16,18]. To better understand the mechanism(s) thatunderlie the broader pattern of immunity induced by the plant-derived vaccines, we evaluated the innate stimulatory potentialof H1- and H5-containing VLPs and eVLPs devoid of any HA pro-tein, in DCs isolated from mice and humans. We found that theplant-derived VLPs can directly activate DCs and stimulate subse-quent CD4+ and CD8+ T cell responses. In vitro, H1-VLPs were gen-



Fig. 6. eVLPs stimulate murine and human DCs. DCs from C57BL/6 were stimulated for 18 h with eVLPs at 1.69 mg/ml of lipid-content (approximate lipid content in 5 mg/mlH1-VLP), or 3.37 mg/ml of lipid-content (eVLP*2, approximately double the lipid content) and compared to LPS (10 ng/ml), OK432 (5 mg/ml) and the unstimulated control. (A)Flow cytometry evaluation of CD86, CD40, and MHCII expression and geometric MFIs of the CD11c+Gr-1� live cells. (B) Secretion of IL-6, IL-12p40 and IL-12p70 as measuredby ELISA. Results are representative of at least three independent experiments. (C) CD8+ T cell cocultures were performed as in Fig. 2. (D) 0.7 mg/ml (lipid-content) of eVLPs orPBS were injected into each footpad of male and female wild-type C57BL/6. After 18 h, the draining popliteal lymph nodes were digested and analyzed for cDC subsets. Cellnumbers and percentage of accumulated DCs, XCR1+ cDC1s and CD172a+ cDC2s are shown. (E) HumanmoDCs were stimulated as in Fig. 4. Data are shown as mean ± SD (A-C)or SEM (D). *p < 0.05; **p < 0.01; ***p < 0.001.


erally superior to H5-VLPs in stimulating these responses. How-ever in vivo, H1 and H5-VLPs showed similar ability to stimulateDC migration to the draining LN and lymphocyte recruitment.

Activated, or mature, DCs play a pivotal role in the induction ofanti-viral immunity. Activated DCs express increased levels of co-


stimulatory molecules including CD86 and CD40, antigen present-ing machinery (MHC molecules), and cytokines such as Type 1interferons and IL-12. The latter two are necessary for the develop-ment and function of CD4+ Th1 cells, CD8+ cytolytic cells and NKcells. Our data demonstrate that plant-derived VLPs, with and




without HA protein, can intrinsically stimulate DCs. In vitro, plant-based VLPs alone were able to stimulate DCs to induce the activa-tion and proliferation of CD4+ T cells and promote Th1 differentia-tion and effector function. Even though the H5-VLPs were lesseffective than their H1 counterparts in activating murine DCs, theystill induced readily detectable and significant antigen-specific Th1CD4+ T cell responses. H1-VLP- and eVLP-stimulated DCs also pro-moted cross presentation of the model antigen OVA to stimulateantigen-specific CD8+ T cell activation, proliferation and effectorfunction. H5-VLP-stimulated DCs showed a similar trend forincreased CD8+ T cell responses but these changes did not reachstatistical significance.

Overall, these animal data are consistent with early studies of Tcell responses in humans with the same H1- and H5-VLPs, in whichstrong CD4+ and more limited CD8+ T cell responses to the H1 anti-gen were detectable up to 6 months after H1-VLP vaccinationwhile only CD4+ T cell responses were seen in the H5-VLP vaccinerecipients [17]. Furthermore, the H5-VLP vaccine candidate neededto be adjuvanted in order to see significant CD4+ T cell responses,while the H1-VLP vaccine did not [17]. Although our mouse datasuggest that H1- and H5-VLPs that differ in only their HA cargohave different capacity to stimulate DCs, the differences in appar-ent T cell immunogenicity of these two formulations in humansmay be attributable, at least in part, to prior experience of mosthuman study subjects with H1 but not H5 antigens. However, priorexposure would not affect the studies reported here.

Although our principal interest was in the intrinsic stimulatorycapacity of the plant-derived VLPs, alum was included in many ofour experiments because of its widespread use. Alum remains themost commonly used adjuvant in human vaccines, even thoughaluminum-containing adjuvants have long been thought of as pro-moting Th2 responses [27,28]. In pre-clinical studies (ferret) andearly clinical trials with the plant-derived VLP vaccines, alum hadshown unexpected promise in eliciting T cell responses and protec-tion from challenge [17,29]. In the current work, we focused ontesting the combined effects of alum and VLPs on DC activation.Similar to previous findings [30,31], we observed that alum byitself does not activate murine or human DCs to any significantdegree. However, when alum was combined with H1-VLP, therewas a marked increase in activation marker expression and cyto-kine secretion from murine DCs compared to H1-VLP alone. Alumwas less effective at promoting enhanced responses to H5-VLPs. Incombination with H1-VLPs, alum drove increased secretion of Th1-promoting cytokines such as IL-12p70 and IL-12p40. This findingsuggests an unexpected feature of combined VLP plus alum stimu-lation and demonstrates that alum can play a role in promotingTh1 responses in certain contexts. Because particles are known toactivate the inflammasome, a known target of alum [32], we exam-ined whether the VLPs had any impact on inflammasome activa-tion. Although IL-1b secretion by DCs was only detected whenH1-VLP and alum were used together, Il1b transcription occurredfollowing exposure of DCs to H1-VLP alone. These findings suggestthat the plant-derived VLPs serve as a priming rather than an acti-vating signal for the inflammasome. When primed by the plant-derived VLP, alum can then support greater Th1 responses. Thisobservation raises the possibility that aluminum-containing adju-vants, in some circumstances, may stimulate immune responsesbeyond promoting antibody production.

In in vitro assays used to evaluate the innate immune-stimulatory potential of H1- and H5-VLPs on murine DCs, we foundthat H1-VLPs were more stimulatory than H5-VLPs. For humancells, even though there were fewer differences in the ability ofH1-VLPs and H5-VLPs to activate DCs, H1-VLPs still stimulatedgreater cytokine production. Interestingly, in vivo, the ability ofH1- and H5-VLPs to stimulate in vivo the recruitment of DCs, Bcells, and T cells to the draining lymph node was equivalent. The


observations that the H5-VLP effectively elicits an immuneresponse in vivo are in accordance with the findings obtained dur-ing clinical trials [17]. Considering that both these VLPs are pro-duced using the same expression platform, it was unexpected toobserve such a striking difference between their innate stimulatorycapacities. Detailed morphological analyses of H1- and H5- con-taining VLPs by cryostatic electron microscopy did not reveal dif-ferences between the two types of VLPs [22]. Even when muchhigher doses of H5-VLPs were used to stimulate murine DCs, theexpression of activation markers never reached the same level aswith H1-VLP in vitro. These differences were also evident in theactivation of signaling pathways that are essential for the develop-ment of a pro-inflammatory phenotype capable of promoting Type1 immune responses. Interestingly DCs activated by eVLPs resem-ble those activated by H1-VLPs. Further work is needed to deter-mine the mechanism behind the differential activation of DCs byempty VLPs, H1 and H5-containing VLPs. Such intrinsic stimulatorycapacity could be the result of the non-HA components of the VLPs(ie: plant-origin lipids/glycolipids) or to the physical structure ofthe VLPs themselves. We also cannot definitively rule out a rolefor residual Agrobacterium- or plant-derived proteins, particularlyin response to eVLP. Although >98.5% of the protein in the VLP vac-cine is viral HA, the eVLPs cannot be produced in the same way andhave slightly higher residual protein content. None of the proteinsidentified to date in either the VLP vaccine or the eVLPs is known tohave immunostimulatory activity. Recent work suggests that theinitial interactions of plant-derived VLPs with antigen presentingcells recapitulate those of intact virions [33]. However, the differ-ent binding specificities of the hemagglutinin trimers themselves(a2,6 and a2,3 sialic acid respectively) may result in distinct VLPinteractions with immune cells [34].

Injection in vivo of eVLPs, H1- and H5-VLPs resulted in anincrease in cDC1 and cDC2 populations in the draining lymph node.Both cDC1s and cDC2s are capable of stimulating antigen-specificCD4+ T cell responses. While XCR1+ cDC1s are more widely knownto be able to cross-present antigen and stimulate CD8+ T cells, cDC2scan also stimulate cross-presentation in some contexts, in particularfor human DCs [35]. These findings suggest that plant-derived VLPsmay stimulate the activation of multiple DC types to induce protec-tive immunity. Future studies will assess the contribution of cDC1sand cDC2s to the induction of CD4+ and CD8+ T cell immuneresponses in vivo and the generation of protective immunity.

In this study, we demonstrate the intrinsic stimulatory poten-tial of plant-derived VLPs. These VLP vaccine candidates effectivelyactivate murine and human DCs in vitro. The combination of alumand plant-derived VLPs had the unexpected effect of enhancing theproduction of Th1-promoting cytokines by DCs. The cooperationbetween alum and plant-derived VLPs raises the possibility thatthis well-established adjuvant could be re-purposed to drivebroader immune responses in certain circumstances. Overall, theunusual innate stimulatory activities of the plant-derived VLPsmay prove to be useful in generating more effective vaccines forinfluenza as well as other targets for which both humoral and cel-lular responses are important.

Author contributions

K.H., C.K., N.L., N.C. and B.W. conceived and designed the study.K.H., C.K. and SY.W. prepared the manuscript. K.H., SY.W. H.G. andB.H. performed experiments.

Acknowledgments

We thank G. Boukhaled, B. Cordeiro, M. Corrado and the membersof the ‘’Flu Crew’’ for stimulating discussions. This work was sup-




ported by a grant to C.M.K and B.J.W. from Medicago Inc. and theQuébec Ministry for the Economy, Science and Innovation (MESI)with oversight by Genome Canada. We thank J. Leconte, V. Mottaand the Cell Vision Core Facility (supported by the Canadian Foun-dation for Innovation) for their assistance with flow cytometry.

Conflict of interest

C.M.K. holds peer-reviewed support for collaborative, basicscience work with Medicago Inc. B.J.W. has been a principal inves-tigator of vaccine trials for several manufacturers, including Med-icago Inc., for which his institution obtained research contracts.Since 2010, B.J.W. has served as Medical Officer for Medicago Inc.In addition, B.J.W. has held and continues to hold peer-reviewedsupport from CIHR and other sources for collaborative, basicscience work with Medicago Inc. B.J.W. has received honorariafrom several vaccine manufacturers for participation on ScientificAdvisory Boards (including Medicago Inc.). N.L. and N.C. are cur-rent employees of Medicago Inc.

Appendix A. Supplementary material

Supplementary data to this article can be found online athttps://doi.org/10.1016/j.vaccine.2018.10.099.

References

[1] Taubenberger JK, Morens DM. The pathology of influenza virus infections.Annu Rev Pathol 2008;3:499–522.

[2] Yin JK, Salkeld G, Heron L, Khandaker G, Rashid H, Booy R. The threat of humaninfluenza: the viruses, disease impacts, and vaccine solutions. Infect DisordDrug Targets 2014;14(3):150–4.

[3] Ohmit SE, Petrie JG, Cross RT, Johnson E, Monto AS. Influenzahemagglutination-inhibition antibody titer as a correlate of vaccine-inducedprotection. J Infect Dis 2011;204(12):1879–85.

[4] Laurie KL, Carolan LA, Middleton D, Lowther S, Kelso A, Barr IG. Multipleinfections with seasonal influenza A virus induce cross-protective immunityagainst A(H1N1) pandemic influenza virus in a ferret model. J Infect Dis2010;202(7):1011–20.

[5] Weinfurter JT, Brunner K, Capuano 3rd SV, Li C, Broman KW, Kawaoka Y, et al.Cross-reactive T cells are involved in rapid clearance of 2009 pandemic H1N1influenza virus in nonhuman primates. PLoS Pathog 2011;7(11):e1002381.

[6] Hillaire ML, van Trierum SE, Kreijtz JH, Bodewes R, Geelhoed-Mieras MM,Nieuwkoop NJ, et al. Cross-protective immunity against influenza pH1N1 2009viruses induced by seasonal influenza A (H3N2) virus is mediated by virus-specific T-cells. J Gen Virol. 2011;92(Pt 10):2339–49.

[7] Subbramanian RA, Basha S, Brady RC, Hazenfeld S, Shata MT, Bernstein DI. Age-related changes in magnitude and diversity of cross-reactive CD4+ T-cellresponses to the novel pandemic H1N1 influenza hemagglutinin. HumImmunol 2010;71(10):957–63.

[8] McElhaney JE, Hooton JW, Hooton N, Bleackley RC. Comparison of single versusbooster dose of influenza vaccination on humoral and cellular immuneresponses in older adults. Vaccine 2005;23(25):3294–300.

[9] McElhaney JE, Xie D, Hager WD, Barry MB, Wang Y, Kleppinger A, et al. T cellresponses are better correlates of vaccine protection in the elderly. J Immunol2006;176(10):6333–9.

[10] Matloubian M, Concepcion RJ, Ahmed R. CD4+ T cells are required to sustainCD8+ cytotoxic T-cell responses during chronic viral infection. J Virol 1994;68(12):8056–63.

[11] Makedonas G, Betts MR. Polyfunctional analysis of human t cell responses:importance in vaccine immunogenicity and natural infection. Springer SeminImmunopathol 2006;28(3):209–19.

[12] Wilkinson TM, Li CK, Chui CS, Huang AK, Perkins M, Liebner JC, et al.Preexisting influenza-specific CD4+ T cells correlate with disease protectionagainst influenza challenge in humans. Nat Med 2012;18(2):274–80.

[13] Doherty PC, Turner SJ, Webby RG, Thomas PG. Influenza and the challenge forimmunology. Nat Immunol 2006;7(5):449–55.


[14] Swain SL, Agrewala JN, Brown DM, Jelley-Gibbs DM, Golech S, Huston G, JonesSC, Kamperschroer C, Lee WH, McKinstry KK, et al. CD4+ T-cell memory:generation and multi-faceted roles for CD4+ T cells in protective immunity toinfluenza. Immunol Rev 2006;211:8–22.

[15] Le Mauff F, Mercier G, Chan P, Burel C, Vaudry D, Bardor M, et al. Biochemicalcomposition of haemagglutinin-based influenza virus-like particle vaccineproduced by transient expression in tobacco plants. Plant Biotechnol J 2015;13(5):717–25.

[16] D’Aoust M-A, Couture MMJ, Charland N, Trépanier S, Landry N, Ors F, et al. Theproduction of hemagglutinin-based virus-like particles in plants: a rapid,efficient and safe response to pandemic influenza. Plant Biotechnol J 2010;8(5):607–19.

[17] Landry N, Pillet S, Favre D, Poulin J-F, Trépanier S, Yassine-Diab B, et al.Influenza virus-like particle vaccines made in Nicotiana benthamiana elicitdurable, poly-functional and cross-reactive T cell responses to influenza HAantigens. Clin Immunol 2014;154(2):164–77.

[18] Pillet S, Aubin E, Trepanier S, Bussiere D, Dargis M, Poulin JF, Yassine-Diab B,Ward BJ, Landry N. A plant-derived quadrivalent virus like particle influenzavaccine induces cross-reactive antibody and T cell response in healthy adults.Clin Immunol (Orlando, Fla) 2016;168:72–87.

[19] Pillet S, Racine T, Nfon C, Di Lenardo TZ, Babiuk S, Ward BJ, et al. Plant-derivedH7 VLP vaccine elicits protective immune response against H7N9 influenzavirus in mice and ferrets. Vaccine 2015;33(46):6282–9.

[20] Hodgins B, Yam KK, Winter K, Pillet S, Landry N, Ward BJ. A Singleintramuscular dose of a plant-made virus-like particle vaccine elicits abalanced humoral and cellular response and protects young and aged micefrom influenza H1N1 virus challenge despite a modest/absent humoralresponse. Clin Vac Immunol: CVI. 2017;24(12).

[21] Carreno BM, Becker-Hapak M, Huang A, Chan M, Alyasiry A, Lie WR, et al. IL-12p70-producing patient DC vaccine elicits Tc1-polarized immunity. J ClinInvest 2013;123(8):3383–94.

[22] Lindsay BJ, Bonar MM, Costas-Cancelas IN, Hunt K, Makarkov AI, Chierzi S,et al. Morphological characterization of a plant-made virus-like particlevaccine bearing influenza virus hemagglutinins by electron microscopy.Vaccine 2018;36(16):2147–54.

[23] Krawczyk CM, Holowka T, Sun J, Blagih J, Amiel E, DeBerardinis RJ, et al. Toll-like receptor-induced changes in glycolytic metabolism regulate dendritic cellactivation. Blood 2010;115(23):4742–9.

[24] van de Veerdonk FL, Netea MG, Dinarello CA, Joosten LAB. Inflammasomeactivation and IL-1b and IL-18 processing during infection. Trends Immunol2011;32(3):110–6.

[25] Elliott Eric I, Sutterwala Fayyaz S. Monocytes take their own path to IL-1b.Immunity 2016;44(4):713–5.

[26] Viganò E, Diamond CE, Spreafico R, Balachander A, Sobota RM, Mortellaro A.Human caspase-4 and caspase-5 regulate the one-step non-canonicalinflammasome activation in monocytes. Nat Commun 2015;6:8761.

[27] Kool M, Fierens K, Lambrecht BN. Alum adjuvant: some of the tricks of theoldest adjuvant. J Med Microbiol 2012;61(7):927–34.

[28] Marrack P, McKee AS, Munks MW. Towards an understanding of the adjuvantaction of aluminium. Nat Rev Immunol 2009;9(4):287–93.

[29] Landry N, Ward BJ, Trépanier S, Montomoli E, Dargis M, Lapini G, Vézina L-P.Preclinical and clinical development of plant-made virus-like particle vaccineagainst avian H5N1 influenza. PLoS One 2010;5(12).

[30] Sokolovska A, Hem SL, HogenEsch H. Activation of dendritic cells andinduction of CD4+ T cell differentiation by aluminum-containing adjuvants.Vaccine 2007;25(23):4575–85.

[31] Sun H, Pollock KGJ, Brewer JM. Analysis of the role of vaccine adjuvants inmodulating dendritic cell activation and antigen presentation in vitro. Vaccine2003;21(9–10):849–55.

[32] Eisenbarth SC, Colegio OR, O’Connor W, Sutterwala FS, Flavell RA. Crucial rolefor the Nalp3 inflammasome in the immunostimulatory properties ofaluminium adjuvants. Nature 2008;453(7198):1122–6.

[33] Makarkov AI, Chierzi S, Pillet S, Murai KK, Landry N, Ward BJ. Plant-madevirus-like particles bearing influenza hemagglutinin (HA) recapitulate earlyinteractions of native influenza virions with human monocytes/macrophages.Vaccine 2017;35(35 Pt B):4629–36.

[34] Hendin HE, Pillet S, Lara AN, Wu CY, Charland N, Landry N, et al. Plant-madevirus-like particle vaccines bearing the hemagglutinin of either seasonal (H1)or avian (H5) influenza have distinct patterns of interaction with humanimmune cells in vitro. Vaccine 2017;35(19):2592–9.

[35] Gutierrez-Martinez E, Planes R, Anselmi G, Reynolds M, Menezes S, Adiko AC,Saveanu L, Guermonprez P. Cross-presentation of cell-associated antigens byMHC class I in dendritic cell subsets. Front Immunol 2015;6:363.



http://refhub.elsevier.com/S0264-410X(18)31496-8/h0005


















































































































characterization of the innate stimulatory capacity …...characterization of the innate stimulatory...

Documents