Yeast-surface expressed BVDV E2 protein induces a Th1/Th2 response in naïve T cells

Yeast species such as Saccharomyces cerevisiae are known to be potent activators of the immune system. S. cerevisiae activates the innate immune system by engaging pattern recognition receptors such as toll like receptor 2 (TLR2) and dectin-1. In the current project, we express the immunogenic envelope protein E2 of bovine viral diarrhoea virus (BVDV) on the surface of S. cerevisiae. After successful expression, components of the innate and adaptive immune response induced by the recombinant S. cerevisiae in vitro were analysed to determine if expression in yeast enhances the immunogenicity of the viral protein. Recombinant S. cerevisiae stimulated production of the chemokine CXCL-8 in primary bovine macrophages, but did no stimulate production of reactive oxygen species (ROS) in the same cells. Additionally, bovine macrophages primed with S. cerevisiae expressing viral envelope proteins had a greater capacity for stimulating proliferation of CD4+ T-cells from BVDV-free animals compared to macrophages primed with envelope protein alone or S. cerevisiae without envelope protein expression. Heat inactivation of recombinant S. cerevisiae increased ROS production and capacity to stimulate CD4+ T-cells in macrophages but did not alter CXCL-8 release compared to the live counter-part. Additionally, heat-inactivation of recombinant S. cerevisiae induced less INFc and IL-4 but equal amounts of IL-10 compared to live yeast T-cell cultures. Our studies demonstrate a use for S. cerevisiae as a vehicle for transporting BVDV vaccine antigen to antigen-presenting cell in order to elicit cellmediated immunity even in naïve animals. 2011 Elsevier Ltd. Open access under CC BY license.


Saccharomyces cerevisiae
Dectin-1 TLR2 BVD E2 Cattle a b s t r a c t Yeast species such as Saccharomyces cerevisiae are known to be potent activators of the immune system. S. cerevisiae activates the innate immune system by engaging pattern recognition receptors such as toll like receptor 2 (TLR2) and dectin-1. In the current project, we express the immunogenic envelope protein E2 of bovine viral diarrhoea virus (BVDV) on the surface of S. cerevisiae. After successful expression, components of the innate and adaptive immune response induced by the recombinant S. cerevisiae in vitro were analysed to determine if expression in yeast enhances the immunogenicity of the viral protein. Recombinant S. cerevisiae stimulated production of the chemokine CXCL-8 in primary bovine macrophages, but did no stimulate production of reactive oxygen species (ROS) in the same cells. Additionally, bovine macrophages primed with S. cerevisiae expressing viral envelope proteins had a greater capacity for stimulating proliferation of CD4+ T-cells from BVDV-free animals compared to macrophages primed with envelope protein alone or S. cerevisiae without envelope protein expression. Heat inactivation of recombinant S. cerevisiae increased ROS production and capacity to stimulate CD4+ T-cells in macrophages but did not alter CXCL-8 release compared to the live counter-part. Additionally, heat-inactivation of recombinant S. cerevisiae induced less INFc and IL-4 but equal amounts of IL-10 compared to live yeast T-cell cultures. Our studies demonstrate a use for S. cerevisiae as a vehicle for transporting BVDV vaccine antigen to antigen-presenting cell in order to elicit cellmediated immunity even in naïve animals.

Introduction
Recognition of invading antigen by the immune system is key for tailoring an appropriate immune response in order to quell infection. This recognition is facilitated through pattern recognition receptors (PRRs), which recognise conserved structures on invading microbes termed pathogen associated molecular patterns (PAMPs). Ligation of these PAMPs to their corresponding PRRs induces a very specific immune response. Of the many PRRs now known, the toll like receptors (TLRs) are the best characterised and most extensively studied.
TLRs 1-10 have been identified in humans and in cattle, and bovine TLRs have been shown to share high sequence homology with human TLRs (ranging between 77% and 86% sequence similarity at the nucleotide level) (Werling et al., 2006). Of these TLRs, TLR2 seems to be the one involved in the recognising antigens from Gram-positive bacteria and fungi, and ligation of TLR2 induces cellular signalling via NF-kB.
Another group of PRRs are the c-type lectin receptors (CTLRs), of which dectin-1 is the best characterised. This receptor has been shown to bind to b-glucan, a component of yeast cell wall. Engagement of dectin-1 facilitates internalisation of yeast particles by phagocytosis (Dennehy and Brown, 2007;Herre et al., 2004). Dectin-1 and TLR2 have previously been shown to co-operate in mounting an immune response to yeast antigen. Receptors colocalise on the surface of RAW-D1 macrophages (MØ) at the point of zymosan contact, and also in nascent zymosan-induced phagosomes (Brown et al., 2003). In addition to these observations, previous work in our lab has demonstrated that stimulating HEK293 cells transfected with bovine dectin-1 and TLR2 receptors with zymosan induces co-localisation of receptors at points of zymosan contact (Willcocks and Werling, unpublished results Although dectin-1 and TLR2 both recognise and bind yeast antigen and co-operate in initiating the ensuing immune response, they appear to play very different roles. Experiments by others also suggest that TLR2 is mainly responsible for the inflammatory response to yeast antigen (Gantner et al., 2003;Gersuk et al., 2006;Underhill, 2007). Cells expressing a mutated murine TLR2 receptor fail to show NF-kB activity or produce TNF-a in response to zymosan compared to cells expressing the wild-type receptor. However, cells expressing the mutated receptor still engulfed zymosan particles and produced reactive oxygen species (ROS) with equal efficiency to those expressing the wildtype receptor indicating that TLR2 is not obligatory for phagocytosis or production of ROS (Gantner et al., 2003;Underhill et al., 1999a,b).
Bovine viral diarrhoea virus (BVDV) is a single stranded positive RNA pestivirus from the family Flaviviridae (Schweizer et al., 2006). Infection of cattle with BVDV presents a wide spectrum of disease, ranging from mild acute infection to fatal mucosal disease, and causes significant loss to the live-stock industry. The USDA licence over 180 commercially available vaccines for BVDV with varying levels of success and duration of protection (Kelling, 2004). In general, the use of modified live vaccine induces high titres of virus neutralising antibodies and gives the animal longer protection from clinical disease than inactivated vaccines. However, there is a risk that modified-live vaccines can revert to a virulent form and cause disease and indeed calves vaccinated with modified live vaccines often develop transient viraemia and shed vaccine virus (Fulton et al., 2003;Kelling, 2004). Neither modified-live nor inactivated vaccines allow for discrimination by clinical test as to which animals have been vaccinated and which animals have contracted BVDV by natural infection. This failure to discriminate could be rectified by development of an effective subunit vaccine for BVDV and the bovine immune response to individual BVDV antigens has been the subject of much research recently (Thomas et al., 2009). The 53 kDa E2 protein of BVDV is part of the viral envelope and is widely accepted as being the most immunogenic protein of this virus with high titres of neutralising antibodies against E2 being found in the host after infection (Baxi et al., 2000;Pande et al., 2005;Thomas et al., 2009). For any vaccine to be successful, the host immune system must recognise it as antigenic.
In the current study we chose to assess the value of using recombinant Saccharomyces (S.) cerevisiae yeast expressing the BVDV envelope protein 2 (E2) on its surface as a vaccine candidate for BVDV, by targeting dectin-1/TLR2 expressed on the surface of bovine macrophages. Due to the immunogenicity of the E2 protein and ability of yeast species to activate both inflammatory and phagocytic receptors on antigen presenting cells (APCs) our recombinant yeast represents an ideal expression system to determine if the efficiency of protein sub-unit vaccine may be improved by expression in yeast, and prevents the negative effects seen when expressed in baculovirus (Pande et al., 2005).

Animals
Whole blood was collected from healthy Friesian-Holstein bullocks into 10% acid citrate dextrose buffer as anti-coagulant in accordance with Home Office regulations. All animals were tested for the absence of BVDV antibodies and viral RNA by the Veterinary Laboratories Agency (Weybridge, UK) and tested negative for both. Thus, these animals were naïve with regards to recognising BVDV derived peptides.

Cloning of BVDV E2 sequence
The partial coding sequence of the BVDV Ky1203ncp E2 gene in pSecTag (Invitrogen) was kindly provided by Dr. C. Thomas (Royal Veterinary College) already. To obtain the full coding sequence, this sequence was amplified by PCR adding start and stop codon on the 5 0 and 3 0 end respectively using PCR and forward (5 0 -ATGCAC-GACTGCAAACC-3 0 ) and reverse (5 0 -TGGCCCTAATGAAGCGGTCA-3 0 ) primers. The following PCR conditions were used: 95°C for 2 min followed by 35 cycles of 95°C for 1 min, 59°C for 1 min, 72°C for 2 min. After successful cloning of the full E2 sequence was confirmed by sequencing, the pYD1 vector was linearised by a double restriction digest using BamHI and NotI (New England Bio-labs) and the E2 gene was ligated into the linearised pYD1 vector using T4 DNA ligase (Promega). Presence of the E2 insert in the correct orientation and in frame with the pYD1 vector was confirmed by PCR and sequencing (Gene service, Cambridge, UK).

Transformation of competent EBY100 S. cerevisiae with pYD1-E2
Competent EBY100 cells pre-prepared using the S. cerevisiae EasyComp kit (Invitrogen) according to the manufacturer's instructions. Stored competent cells were thawed and 50 ll was pipetted into a sterile Eppendorf tube for each transformation reaction. 1 lg of pYD1 or pYD1-E2 was added to the cells. An Eppendorf tube containing competent cells but no plasmid DNA was set up as a negative control. 500 ll of solution 3 of the S. cerevisiae EasyComp kit (Invitrogen) was added to each transformation tube and tubes were vortexed for 5 s each. Tubes were incubated at 30°C for 60 min and during this incubation were vortexed every 15 min for 5 s per tube. After incubation, 100 ll of transformation reaction from each tube was spread on a separate minimal dextrose plates pre-spread with 100 ll of 10 mg ml À1 L-leucine (Sigma-Aldrich).
Plates were incubated at 30°C for 2-4 days and checked for growth of transformed colonies.

Surface protein induction and staining in EBY100 S. cerevisiae
Wild-type EBY100 and EBY100 transformed with the pYD1 plasmid (Invitrogen) containing the E2 gene insert of BVDV or the empty pYD1 plasmid were grown overnight in 10 ml of YNB-CAA media containing 2% glucose at 30°C and 250 rpm. The absorbance of the overnight cultures was read at OD600 and the cells were pelleted by centrifugation at 4000g for 10 min at room temperature. The cell pellet was re-suspended to an OD600 of 0.5 in YNB-CAA medium containing 2% galactose and a volume of cells equivalent to 2 U OD600 was removed and stored on ice as a zero time point. Cell cultures were incubated for 48 h at 20°C and 250 rpm with a volume of cells equivalent to 2 U OD600 being removed every 12 h and stored on ice until all time-points were collected. Samples were pelleted and washed with PBS (Sigma) at 4000g at 4°C for 10 min. Cells were suspended in 250 ll of PBS with 1 mg ml À1 BSA (PAA) and 1 lg of X-press antibody (Invitrogen) or WB162 antibody (provided by VLA, Weybridge) against the BVDV E2 protein. Cells were incubated on ice for 30 min with occasional mixing. Antibody was removed with a centrifugation step and cells were again washed in PBS. After washing, cell pellet was re-suspended in 250 ll PBS with 1 mg ml À1 BSA and 1 lg donkey anti-mouse-Alexafluor 488 antibody (Invitrogen). Samples were incubated on ice and away from light for 30 min with occasional mixing. Antibody was removed with a centrifugation step and cells were washed twice in PBS. After washing, cells were resuspended in 100 ll of PBS and 10 ll of this was spotted on a microscope slide. A cover slip was applied and cells were viewed by fluorescent microscopy using 530 nm band-pass filter and 20Â magnification lens.
Heat-killed yeast was termed DEBY100-pYD1 and DEBY100-pYD1-E2. The viability of heat-killed cells was determined by plating on YPD plates.
2.6. Culture of bovine macrophages (MØ) MØ were prepared as described (Werling et al., 2004). Briefly, blood was collected from animal in a vacuum bottle containing 10% final volume of ACD anticoagulant buffer, the buffy coat harvested, cells washed with citrate-buffer, and red blood cells lysed. Subsequently, PBMC were generated by density-gradient centrifugation over Histopaque (d = 1.083 g dl À1 ; Sigma) for 45 min, 1200g at room temperature. The interphase was carefully removed, washed, cells counted in trypan blue and adjusted to a concentration of 5 Â 10 6 cells per ml, placed into Teflon bags and incubated at 37°C and 5% CO 2 for 7 days.

ELISA for CXCL8
ELISA for CXCL-8 was performed according manufacturer's recommendation (R&D Systems). Optical density of each well was determined using a Spectra Max M2 plate reader (Molecular Devices) set to 450 nm with wavelength correction of 540 nm. Results were analysed in Excel (Microsoft).

Measurement of ROS production
1 Â 10 7 MØ in a volume of 1 ml of HBSS (Gibco) were placed in sterile 1.5 ml Eppendorf tubes. Cells were stimulated with 1 mg ml À1 zymosan, 390 lg ml À1 isolated E2 protein, or the equivalent amount of E2 expressed by yeast, resulting in 5ppc EBY100-pYD1, 5ppc EBY100-pYD1-E2, 5ppc DEBY100-pYD1, 5ppc DEBY100-pYD1-E2, or HBSS as a negative control. ROS production was analysed as recently described (Conejeros et al., 2011), and plates were read at 485 nm excitation and 530 nm emission in a Spectra Max M2 plate reader (Molecular Devices) and results recorded over a 135 min period with readings being taken every 15 min. Results were analysed in Excel (Microsoft).

T-cell presentation assay
Bovine MØ were seeded at a density of 1 Â 10 4 per well in 200 ll RPMI (Gibco) supplemented with 10% FCS and 200 U ml À1 penicillin and streptomycin in 96-well round bottom plates (Cellstar). Plates were incubated at 37°C and 5% CO 2 for 2 h after which 100 ll of media was removed from each well. Cells were primed with either 5ppc EBY100-pYD1, 5ppc DEBY100-pYD1, 5ppc EBY100-pYD1-E2, 5ppc DEBY100-pYD1-E2 or 390 lg ml À1 E2 in a volume of 100 ll media per well. Plates were incubated at 37°C and 5% CO 2 for 2 h after which 100 ll was removed from each well. CD4+ autologues T cells were prepared by MidiMacs separation using mAb specific for bovine CD4 (CC8), as described (Werling et al., 2002). The purity of the cells was evaluated by flow cytometry and shown to be >98%. Subsequently, 1 Â 10 5 autologous CD4+ T-cells were added to each well. Plates were incubated at 37°C and 5% CO 2 for up to 5 days, pulsed with 0.04 MBq of tritium labelled thymidine (3-H-TdR) (Perkin-Elmer) in 25 ll media and plates were incubated for another 16 h before being stored at À20°C. Subsequently, cells were harvested (Tomtec 96, Mach II) onto glass-fibre mats (Perkin-Elmer) and analysed by liquid-scitillation counting using a a1450MicroBeta counter (Perkin-Elmer).

MSD cytokine measurement
To evaluate the presence of IFNc, IL-4 and IL-10 in culture supernatants, the Meso-Scale Discovery (MSD) technology (Gaithersburg, MD, USA) was used as previously described (Coad et al., 2010). Custom coated 7-plex 96-well plates were supplied by MSD. In brief, MSD plates were blocked with MSD's proprietary assay buffer prior to addition of sample or standards. Standards were serially diluted in MSD dilution buffer with top concentration of standard shown in parenthesis: IFNc (Endogen, 100 ng ml À1 ); IL-4 (bovine IL-4 calibrator (MSD), 2 ng ml À1 ) and IL-10 (IAH, 30 U ml À1 ). 25 ll of cell culture supernatant or standards were added to duplicate wells. Incubations were performed at room temperature. Following a 2 h incubation, plates were washed and then incubated for a further 2 h with a combined cocktail of secondary antibodies conjugated to SulfoTag TM(MSD) (IFNc, IL-4 and IL-10 all from MSD) in MSD dilution buffer. After a final wash, plates were coated with MSD Buffer-T and luminescence signal measured on an MSD-6000 reader. Results are shown as calibrated units extrapolated from standard curves.

Statistical analysis
Statistical analysis was performed using GraphPad Prism (Version 5) by means of 2-way analysis of variance (ANOVA) and post hoc analyses for multiple significance test comparisons were provided by Bonferroni correction. Graphical data are presented as the mean ± SEM. Significance values were set at p < 0.05 for 95% confidence intervals.

Results
3.1. Display of the mature E2 protein on the surface of EBY100 S. cerevisiae EBY100 cells transformed with the pYD1-E2 vector and untransformed EBY100 cells were grown overnight, stained and visualised using fluorescent microscopy. Fig. 1 shows expression of the X-press epitope on the surface of EBY100 cells transformed with pYD1-E2 (Fig. 1, right panels) can be detected in as little as 12 h after induction with galactose. This expression of the X-press epitope increased for up to 48 h and then decreased over a period of time not being detectable at 96 h (data not shown). No X-press epitope was detected on the surface of wild-type EBY100 cells (Fig. 1, left panels) for any time point analysed. All further experiments inducing protein expression on the surface of EBY100 cells were completed using 48 h as the time point at which surface expression was tested for.
As seen in Fig. 2, EBY100 cells transformed with the pYD1-E2 vector (EBY100-pYD1-E2) (Fig. 2, left hand panels) and grown in media containing galactose for 48 h stained positive for both the X-press epitope ( Fig. 2, lower panels) and the E2 protein (Fig. 2, upper panels) as tested using the WB162 antibody. EBY100 cells transformed with the pYD1 vector (EBY100-pYD1) (Fig. 2, middle panels) tested positive for the X-press epitope but negative for the E2 protein.
Wild-type EBY100 cells tested negative for both the X-press epitope and the E2 protein. The data suggest that working E2 protein is displayed on the surface of EBY100 cells transformed with the pYD1-E2 vector.

CXCL8 and ROS production by bovine MØ is dependent on the form of S. cerevisiae expressing BVDV E2 protein
CXCL8 is produced predominantly by MØ and acts as a potent chemo-attractant for neutrophils and other granulocytes inducing their migration to areas of infection in body tissue. Once at the site of infection neutrophils, MØ and other phagocytes engulf antigen and destroy it by pathways including the ROS pathway. We therefore assessed whether recombinant S. cerevisiae induces CXCL-8 and ROS production in bovine MØ. Bovine MØ stimulated with both, life or heat-inactivated EBY100-pYD1 and EBY100-pYD1-E2 produced significant amounts of CXCL-8 compared with those stimulated E2 protein alone (Fig. 3A). In contrast, neither EBY100-pYD1 nor EBY100-pYD1-E2 induced large amounts of ROS production by MØ. In contrast, DEBY100-pYD1 and DEBY100-pYD1-E2 induced significantly amounts of ROS.

Bovine MØ primed with EBY100-pYD-E2 induce stronger CD4+ Tcell proliferation compared to EBY100-pYD1
Antigen presenting cells such as MØ phagocytose and degrade antigen which is processed and presented at the cell surface in the context of MHC class II. Antigen specific CD4+ T cells recognise this presented antigen through their T-cell receptor and proliferate and release inflammatory cytokines. MØ primed with both EBY100-pYD1 and EBY100-pYD1-E2 induce significantly higher CD4+ T-cell proliferation than those primed with the E2 protein alone (Fig. 4A). Additionally, MØ primed with EBY100-pYD1-E2 induced significantly more CD4+ T-cell proliferation than those primed with EBY100-pYD1 indicating that a population of CD4+ cells specific for E2 antigenic peptides presented in the context of MHC class II had been expanded. Controls tested were media alone, CD4+ cells alone, MØ alone, CD4+ cells co-cultured with unstimulated MØ and MØ primed with either EBY100-pYD1, EBY100-pYD1-E2 or E2 alone. None of these controls displayed bcounts above 1500 cpm for any day tested (data not shown).
3.4. Bovine MØ primed with DEBY100-pYD1-E2 or EBY100-pYD1 induces stronger CD4+ T cell proliferation than those primed with live yeast Although live S. cerevisiae expressing hepatitis C virus proteins has been used in phase II clinical trials in humans (Habersetzer Fig. 1. Fluorescent microscope pictures showing staining of X-press epitope on the surface of Wild-type EBY100 and EBY100 transformed with pYD1-E2. Wild-type EBY100 S. cerevisiae and EBY100 S. cerevisiae transformed with pYD1-E2 were grown as described, and stained using an anti-X-press antibody followed by a secondary anti-mouse-Alexafluor488 antibody. Cells were viewed by fluorescent microscopy using a 20Â objective. et al., 2009), inactivation of yeast would be an attractive and safe option for use in vaccines if it were to be as immunogenic as its live counter-part. For this reason we tested the capacity of MØ primed with DEBY100-pYD1-E2 to stimulate proliferation of CD4+ T-cells. Priming MØ with the DEBY100-pYD1-E2 increases their capacity to stimulate proliferation of CD4+ T-cells compared to live EBY100-pYD1-E2 at 5 days of incubation (Fig. 4B). No significant difference between the proliferative capacities of MØ primed with live or DEBY100-pYD1-E2 was observed for days 2-4 of incubation. Controls tested were media alone, MØ alone and CD4+ alone. None of these controls displayed b-counts above 1500 cpm for any day tested (data not shown).

Live EBY100-pYD1-E2 stimulates increased production of INFc and IL-4
Next, we wanted to determine the cytokine production by CD4+ T cells in response to MØ primed with live and DEBY100-pYD1-E2. Cell-free supernatants were collected from the same plates as shown in Fig. 4B, and assayed for INFc, IL-4 and IL-10. Both, EBY100-pYD1-E2 and DEBY100-pYD1-E2 induced a higher production of all three cytokines tested compared to E2 protein alone. Surprisingly, live EBY100-pYD1-E2 induced more INFc and IL-4 compared to DEBY100-pYD1-E2. This was statistically significant at days 2, 4 and 5 for INFc and at days 2 and 4 for IL-4 ( Fig. 5A and C). Very little difference is observed between the amount of IL-10 released in response to live or DEBY100-pYD1-E2 (Fig. 5B). Controls tested were media alone, MØ alone and CD4+ alone. The controls did not exceed 1 pg ml À1 for INFc, 160 pg ml À1 for IL-10 and 0 pg ml À1 for IL-4 for any of the days tested (data not shown).

Discussion
Several previous studies have highlighted the importance of dectin-1 as key innate microbial receptor. Dectin-1 collaborates with TLR2, as well as acting independently, to initiate a wide range of microbial responses, especially against fungal pathogens. Yeast such as S. cerevisiae are well documented as being potent activators of the immune system and therefore do not require the addition of an additional adjuvant (Romani et al., 2002). In the current study, we successfully expressed the E2 protein of BVDV on the surface of the S. cerevisiae EBY100 strain, and started to assess the potential of this recombinant yeast as a vaccine antigen for BVDV by measuring the bovine immune response in vitro. We used bovine MØ stimulated with the S. cerevisiae displaying the E2 protein at the surface and also S. cerevisiae with no expression of the E2 surface protein and measured their capacity to produce ROS, CXCL8 and also their capacity to induce proliferation of CD4+ T-cells.
Both yeast types induced significantly more CXCL8 than cells stimulated with equal amounts of isolated E2 fusion protein. These data indicate that in our system, the CXCL8 produced by MØ is due to stimulation of the cells with yeast components. This may be mainly as a result of TLR stimulation and subsequent NF-kB activation. However, the induction of CXCL8 may not be completely dependent on dectin-1, as shown recently in human neutrophils using depleted zymosan (van Bruggen et al., 2009). When neutrophils were stimulated with normal zymosan which contains both TLR2 and dectin-1 ligands, CXCL8 production were observed (van Bruggen et al., 2009). The small reduction in CXCL8 production observed in response to heat-killed yeast could be accounted for by a reduction in the quantity of mannan at the cell surface and therefore possibly less TLR activation (Gantner et al., 2005).
Interestingly, our data demonstrates that MØ produce very little ROS in response to stimulus with live version of EBY100-pYD1 and EBY100-pYD1-E2. However, when cells are stimulated with heatkilled versions of the same recombinant yeast, a significant rise in ROS production is observed. This increase in ROS production would indicate an increase in dectin-1 ligation as it is well documented that dectin-1 activation is responsible for ROS production in phagocytic cells (Gantner et al., 2003;Goodridge and Underhill, Fig. 2. Fluorescent microscope pictures showing staining of E2 protein of BVDV and X-press epitope on the surface of transformed EBY100. Wild-type EBY100 S. cerevisiae and EBY100 S. cerevisiae transformed with pYD1 or pYD1-E2 were grown overnight in YNB-CAA media containing 2% glucose. The following day, cells were pelleted and resuspended to an OD600 of 0.5-1.0 in YNB-CAA containing 2% galactose. Cultures were grown for 48 h and the E2 protein and X-press epitope were stained on the surface of the cells using WB162 and X-press antibody respectively followed by the use of a secondary anti-mouse-Alexafluor488 antibody. Cell were viewed by fluorescent microscopy using a 20Â objective. 2008). Nerren et al. have demonstrated that pre-treating chicken heterophils with the dectin-1 specific antagonist laminarin reduces their ability to produce ROS in response to stimulus with curdlan (a b-1,3-glucan) (Nerren and Kogut, 2009). Additionally, experiments from our lab have demonstrated a significant drop in ROS production by MØ in response to zymosan when pre-treated with laminarin (data not shown). The fact that heat-killed recombinant yeast is capable to stimulate ROS production can be explained by the fact that heat-inactivation exposes more of the dectin-1 ligand b-glucan at the yeast surface, as shown recently (Gantner et al., 2005). Furthermore, recent work by Wellington et al. has suggested that live Candida albicans not only inhibits ROS production by masking b-glucan at its surface but also actively suppresses ROS production. However, when using S. cerevisiae, no active suppression was observed. The authors concluded that this is due to the fact that S. cerevisiae is non-pathogenic (Wellington et al., 2009). In our system, this increased recognition of recombinant yeast by dectin-1 may lead to increased uptake and processing of both EBY100 and E2 antigen which would be advantageous from the perspective of using the recombinant yeast as a vaccine. The mature E2 protein expressed on the yeast surface did not appear to be affected by the heat-inactivation process as determined by fluorescent staining (data not shown).
EBY100-pYD1-E2 induced significant CD4+ proliferation compared with those primed with the E2 protein alone. We are fully aware of the fact that the stimulation of naïve CD4+ T cells should not occur when using MØ as antigen presenting cells. We can only assume at the moment that either the MØ prepared using the method described possess functions associated normally with dendritic cells, or that the use of yeast particles induces such a strong stimulatory signal for MØ that these are able to subsequently drive T cell proliferation. Our data indicate that either yeast antigen readily enters endocytic processing pathways and is expressed on the cell surface in the context of MHC II, or stimulates such a strong innate immune response with the secretion of pro-inflammatory cytokines that it drives T cell proliferation. Additionally, priming MØ with EBY100-pYD1-E2 induced a significantly stronger CD4+ proliferative response at day 4 and 5 than priming with EBY100-pYD1. Thus, in addition to the proliferation of potential yeast antigen specific CD4+ cell, an E2 specific CD4+ T cell population may expand. Our data are in line with recent data using murine dendritic cells primed with S. cerevisiae expressing recombinant ovalbumine (Stubbs et al., 2001). Here, dendritic cells primed with as little as 5 recombinant yeast particles induced CD4+ proliferation to the same degree as those primed with saturating amounts of ovalbumin peptide. The authors also found that stimulation of dendritic cells with yeast induced up-regulation of CD80, CD86, CD40, CD54 and MHC II as well as an increase in IL-12 production. These increases were comparable to stimulation with LPS. These data clearly demonstrate the adjuvant-like properties of S. cerevisiae yeast and offer some explanation of why conjugating protein antigen to whole yeast appears to increase the cellular immune response to this antigen. Given the differences seen between live and heat-inactivated yeast, and given the fact that a heat-inactivated delivery system still showing immunogenicity, we also assessed the ability of heat-killed EBY100-pYD1-E2 primed MØ to stimulate CD4+ proliferation, and compared the INFc, IL-4 and IL-10 production induced in the responding T cells.
In correlation with the ROS data showing increased ROS production in response to heat-killed EBY100-pYD1-E2, a stronger CD4+ T-cell proliferative response was seen in MØ primed with heat-killed EBY100-pYD1-E2 after 5-days of incubation (Fig. 4B). Similar as for the ROS data, this may be a result of more b-glucan being exposed at the yeast surface (Gantner et al., 2005), resulting in increased dectin-1 ligation and therefore facilitate uptake of EBY100-pYD1-E2 by MØ. Indeed, work by Lu et al. and by Franzusoff et al. has demonstrated that immunisation with inactivated S. cerevisiae expressing tumour antigens elicits an equal anti-tumour response to immunisation with their live counter-parts (Ardiani et al., 2010;Franzusoff et al., 2005;Lu et al., 2004), but to our knowledge this is the first report that recombinant heat-inactivated S. cerevisiae may actually increase the adaptive immune. We were expecting to see an increased amount of INFc produced by T-cell assays co-cultured with MØ exposed to DEBY100-pYD1-E2 compared to those exposed to live yeast, reflecting the increased CD4+ proliferation (Fig. 4B). Surprisingly this was not the case. In fact, higher INFc and IL-4 concentrations were measured in supernatants of co-cultures where live EBY100-pYD1-E2 was used ( Fig. 5A and C), whereas amounts of IL-10 were similar (Fig. 5B). Whereas we can currently not fully explain this observation, it is possible that heat-inactivated, but not live yeast induces an anti-inflammatory state in MØ, without affecting the ability of these cells to stimulate T cells. Such effect has been recently described for Murine macrophages (RAW 264.7) using either heat-inactivated or live C. albicans on the proteomics level (Martinez-Solano et al., 2009).
In conclusion, both live and heat-killed S. cerevisiae are an attractive vaccine vehicle for delivering the weakly immunogenic proteins to APC in order to stimulate cellular immunity directed against this protein. Such delivery system could not only improve current BVDV vaccine strategies, but potentially any vaccine which relies on such protein antigen. The ease with which S. cerevisiae can be genetically manipulated means that potentially any antigenic protein could be expressed at the yeast surface, potentially expressing even more than one immunogenic protein, either from Fig. 5. Bovine MØ primed with live EBY100-pYD1-E2 induce a stronger INFc and IL-4, but similar IL-10 production compared with those primed with DEBY100-pYD1-E2. Bovine MØ were seeded in 96-well plates and primed with 5ppc EBY100-pYD1-E2, 5ppc DEBY100-pYD1-E2 or 390 lg ml À1 E2. CD4+ T-cells were added and plates were incubated for 2-5 days before cell-free supernatants were harvested and assayed for INFc, IL-4 and IL-10 content as described. Asterisks denote a significant difference between groups. ( ⁄ p < 0.05, ⁄⁄⁄ p < 0.001). Graphs are representative of 3 biological repeats.E2, 5ppc DEBY100-pYD1 or 5ppc DEBY100-pYD1-E2 and probed for ROS every 15 min for 2 h. Graphs are representative of 2 biological repeats. the same pathogen or derived from different pathogens. Furthermore, such delivery-systems could subsequently allow for the discrimination of vaccinated versus naturally infected animals.