JH is acting CEO, board member and shareholder of Recopharma AB. At the time of the study, GA, LS, TJ, AN, and NC were employees of Recopharma. MS has been partly funded by Recopharma through a research grant to LTU. The technology described in the manuscript has been patent-sought by Recopharma, “Production of Human Glycosylated Proteins in Yeast and Methods of Use Thereof” (United States application number 11/626156. Filing date 23-Jan-2007.) (WO application number PCT/IB2007/004164. Filing date 23-Jan-2007.) This does not alter the authors' adherence to all the PLOS ONE policies on sharing data and materials.
Conceived and designed the experiments: GA LS JH UR. Performed the experiments: GA LS TJ NC MS. Analyzed the data: GA LS TJ JH. Contributed reagents/materials/analysis tools: TJ AN NC MS. Wrote the paper: GA LS TJ MS JH.
Targeting antigens to antigen-presenting cells (APC) improve their immunogenicity and capacity to induce Th1 responses and cytotoxic T lymphocytes (CTL). We have generated a mucin-type immunoglobulin fusion protein (PSGL-1/mIgG2b), which upon expression in the yeast
OVA antibody class and subclass responses were determined by ELISA, the generation of anti-OVA CTLs was assessed in 51Cr release assays using
Immunizations with the OVA − mannosylated PSGL-1/mIgG2b conjugate, especially when combined with the AbISCO®-100 adjuvant, lead to faster, stronger and broader (with regard to IgG subclass) OVA IgG responses, a stronger OVA-specific CTL response and stronger Th1 and Th2 responses than if OVA was used alone or together with AbISCO®-100. Also non-covalent mixing of mannosylated PSGL-1/mIgG2b, OVA and AbISCO®-100 lead to relatively stronger humoral and cellular responses. The O-glycan oligomannoses were necessary because PSGL-1/mIgG2b with mono- and disialyl core 1 structures did not have this effect.
Mannosylated mucin-type fusion proteins can be used as versatile APC-targeting molecules for vaccines and as such enhance both humoral and cellular immune responses.
Targeting antigens to endocytic receptors on professional antigen-presenting cells represents an attractive strategy to enhance the efficacy of vaccines. Mannosylated antigens have been demonstrated to enhance MHC class I- and MHC class II-restricted antigen presentation, increase T-cell proliferation, and promote T cell effector responses
Several studies have shown that polymeric mannose (mannan) improve antigen presentation and that oxidized and reduced mannan can induce antigen-specific Th1/CTL and Th2/humoral responses, respectively
In contrast to protein-protein interactions, individual protein-carbohydrate interactions are characterized by a low affinity binding with Kd values many times in the mM range
P-selectin glycoprotein ligand-1/mouse IgG2b (PSGL-1/mIgG2b) is a mucin-like immunoglobulin fusion glycoprotein with 106 potential sites for O-linked glycosylation and six potential sites for N-linked glycosylation in its dimeric form
We hypothesize that a mucin-type fusion protein carrying multiple O-linked oligomannose structures has the potential of working as a universal antigen presenting cell (APC)-targeting molecule for a broad repertoire of protein antigens in different vaccine compositions. As such, it may amplify both humoral and cellular immune responses and may be used together with different antigens for which already established manufacturing bioprocesses can be maintained. Here, we present data on the OVA-specific immune responses in mice immunized with OVA, OVA-mannosylated PSGL-1/mIgG2b conjugates or mixtures, with or without an additional adjuvant in the form of Imject®Alum or AbISCO®-100.
Inbred C57Bl/6J (H-2b) mice were bred and housed at Karolinska Institutet, Division of Comparative Medicine, Clinical Research Center, Karolinska University Hospital, Huddinge. The animals were caged at five to ten mice per cage and fed a commercial diet with free access to food and water. All animals were six to eight weeks of age at the start of the experiment.
All mice were bred and maintained according to the regulations of the Ethical Committee for Animal Research at Karolinska Institutet. All animal experiments were approved by the regional committee (Stockholms södra djuretiska nämnd) on animal ethics, S-184-06 and S-132-09.
For ELISpot and proliferation assays different proteins and peptides at varying concentrations were used: the OVA-SIINFEKL (MHC Class I) CTL peptide (Innovagen, Lund, Sweden, OVA 257-264, SP-O257-5) was used at a concentration of 1 µg/mL – 0.0001 µg/mL, the FILKSINE control (MHC Class I) CTL peptide (Innovagen, SP-CS) at 1 µg/mL, the Th-OVA (MHC Class II) Th peptide (Innovagen, OVA 323-339, SP-O323A-5) at 10 µg/mL – 0.01 µg/mL, the OVA protein grade VII (Sigma Aldrich, St Louis, MO, USA, A7641) was used at a concentration of 625 µg/mL – 5 µg/mL, BSA (Sigma Aldrich, A8806) at 25 µg/mL and Concanavalin A (Sigma Aldrich, L7647) at 5 and 1 µg/mL.
The LC-SPDP linker (Pierce) and Imject Alum were from Thermo Fischer Scientific (Waltham, MA, USA) and AbISCO®-100 was from Isconova AB (Uppsala, Sweden).
Mannosylated PSGL-1/mIgG2b (PPM) was produced in
PSGL-1/mIgG2b with mono and disialylated core 1 structure (CP) was produced in a stable CHO (Chinese hamster ovary) cell line given the name, C-P55.2. The cells were cultured in serum-free ProCHO4 medium (Lonza, Basel, Switzerland) in repeated batch mode in a 20 L Wave bioreactor (Wave System 20/50 EH, GE Health Care, Uppsala, Sweden). The bioreactor was inoculated at 0.8×106 viable cells/mL in a volume of 5.2 L. At regular intervals, fresh cultivation medium with 2 mM glutamine was added as a bolus until the final volume in the bioreactor reached 10.3 L. The culture was harvested when the final cell density was 4.6×106 total cells/mL and the viability dropped to 88%. The glucose, glutamine and pH levels were monitored daily and adjusted to optimal levels. Total cultivation time was 11 days.
Cell culture supernatants were clarified using a 540 cm2 Millistak+ POD C0HC filter (Millipore, Billerica, MA, USA) connected to the Quattroflow pump on the Cogent M TFF system (Millipore). The clarified supernatant was subsequently concentrated 22× using a 0.11 m2 Pellicon 3 ultrafilter (Millipore) on the Cogent M TFF system, and then further diafiltered against six volumes of PBS. Finally, 1 mL/L of protease inhibitor cocktail (Sigma Aldrich, P8215) and 0.02% NaN3 (Sigma Aldrich, 71289) were added to the product solution, which was stored at 4°C until purification.
All chromatographic procedures were carried out on an ÄKTAExplorer 100 (GE Healthcare) controlled by the Unicorn software (v. 5.11). The clarified supernatants were sterile filtered with a 0.22 µm polyethersulfone (PES) filter (TPP, Trasadingen, Switzerland) before loading onto a MabSelect SuRe column (GE Healthcare) pre-equilibrated with PBS. The column was washed with 10 column volumes (CV) of PBS, and elution of recombinant fusion protein was achieved using 5 CV of 0.1 M sodium citrate, pH 3.0. After elution, selected fractions were pooled, neutralized with 250 µL per mL of 1 M Tris-HCl, pH 9.0 and then dialyzed extensively (12–14 kDa cut-off) against MilliQ water at 4°C. After dialysis, the samples were frozen, lyophilized and stored at −80°C before further purification.
Lyophilized samples were dissolved at a concentration of approximately 5 mg/mL in gel filtration buffer (0.1 M sodium phosphate pH 7.2, 0.5 M sodium chloride). Gel filtration of PSGL-1/mIgG2b was carried out on a pre-equilibrated HiPrep 26/60 Sephacryl S-300 HR column (GE Healthcare, 17-1196-01). Typically, 5 mL of sample was applied to the gel filtration column and eluted with a flow rate of 1 mL/minute. Eluted fractions were kept at 4°C until pooling was done on the basis of Western blot analysis. Pooled fractions were dialyzed as above, frozen, lyophilized and stored at −80°C.
A 5 mg/mL (113 µM) solution of OVA was prepared in conjugation buffer (0.1 M sodium phosphate pH 7.2, 0.15 M sodium chloride, 1 mM EDTA). A 20 mM solution of the LC-SPDP linker was prepared in MilliQ water immediately prior to activation of OVA. For activation of OVA, 300 µL of the linker solution was added to 2.1 mL of OVA solution at a 25× molar excess of linker. The mixture was allowed to react under rotation for 30 minutes at room temperature, yielding activated OVA (OVA*). The solution was applied to a PD-10 desalting column (GE Healthcare, 17-0851-01) and OVA* eluted in 3.5 mL conjugation buffer thereby removing free linker. The eluted OVA* had a concentration of 3 mg/mL (68 µM) as determined with the BCA method (Pierce, 23225) using BSA (Sigma Aldrich) as standard.
A 10 mg/mL (67 µM) solution of the PSGL-1/mIgG2b fusion protein was prepared in conjugation buffer. To obtain the fusion protein in a form suitable for conjugation with OVA*, the fusion protein solution was reduced with 100× molar excess of 0.5 M dithiothreitol (DTT, Fluka, 43815) solution. The reduction was carried out in a heat block at 37°C for 15 minutes. The DTT was removed from the fusion protein by passing the reaction solution twice through a PD MiniTrap G25 column (GE Healthcare).
To assess the amount of linker present on OVA*, a pyridine-2-thione assay was performed. 50 µL OVA* was mixed with 950 µL PBS in a plastic cuvette and the absorbance at 343 nm was measured. 2 µL 0.5 M DTT was added to the cuvette to cleave the chromophore from the LC-SPDP linker. After 15 minutes at room temperature the absorbance at 343 nm was measured again. The molecular substitution ratio (MSR) was calculated as follows:
To estimate the number of accessible thiol groups on the reduced fusion protein, Ellman's assay was performed. 2 µL reduced fusion protein, 998 µL Ellman's buffer (0.1 M sodium phosphate pH 8.0, 1 mM EDTA) and 50 µL 10 mM Ellman's reagent (5,5′-Dithio-
For conjugation, 2.816 mL OVA* and 1.0 mL reduced PPM (study A), 3.15 mL OVA* and 0.8 mL reduced PPM (study B), and 2.412 mL OVA* and 1 ml reduced CP, respectively, were mixed and split into two parallel reactions. The reaction was carried out at room temperature over night under rotation. See
Study A | Study B | ||
PPM | PPM | CP | |
Amount fusion protein reduced | 48 nmol | 67 nmol | 41 nmol |
Amount fusion protein in conjugation | 48 nmol | 54 nmol | 41 nmol |
Amount OVA activated | 237 nmol | 237 nmol | 212 nmol |
Amount LC-SPDP linker | 6 µmol | 6 µmol | 5.4 µmol |
Linker∶OVA ratio | 25∶1 | 25∶1 | 25∶1 |
Amount OVA* in conjugation | 192 nmol | 214 nmol | 164 nmol |
OVA*∶fusion protein molar ratio | 4∶1 | 4∶1 | 4∶1 |
Yields | |||
70% | 22% | 17% | |
OVA- |
Not analysed | 76% | 46% |
OVA∶fusion protein molar ratio | - | 1.1∶1 | 1.5∶1 |
After the conjugation reaction, the sample was centrifuged at 4,250× g and applied to a 26/60 HiPrep Sephacryl S300 gel filtration column (GE Healthcare) pre-equilibrated with 0.1 M sodium phosphate pH 7.2 with 0.5 M sodium chloride. Eluted fractions were kept at 4°C until pooling was done on the basis of Western blot analysis. Pooled fractions were then dialyzed as above, frozen, lyophilized and stored at −80°C.
Quantification of conjugated OVA used for immunizations in study A was done by anti-OVA Western blot analysis using OVA of known concentration as standard. The OVA standard was determined with the BCA method using BSA as standard. The concentration of OVA in the stock solution was 2.0 mg/mL. A dilution series of DTT-reduced samples was heat-inactivated for 10 minutes at 70°C prior to separation on a 4–12% Novex Bis-Tris gel (MES buffer, 200 V for 45 minutes, Invitrogen). Two identical SDS-PAGE gels were run, blotted and analyzed as described below. Blotting was performed in an Invitrogen iBlot device for 10 minutes using an iBlot Transfer Stack (Invitrogen, San Diego, CA, USA, SKU# IB1001EU) with a nitrocellulose membrane (Invitrogen, SKU# IB3010-01). After washing of the membrane in PBS-Tween (PBS-T; 2×5 minutes), it was incubated in blocking solution (3% BSA in PBS-T) for one hour at room temperature (RT). After additional washes (3×5 minutes), the membrane was incubated over night at +4°C with an anti-OVA antibody (Sigma Aldrich, A6075) diluted 1∶20,000 in 3% BSA/PBS-T. The membrane was after washing (3×5 minutes in PBS-T) incubated for one hour at RT with a horseradish peroxidase (HRP)-conjugated secondary goat anti-mIgG Fab antibody (Sigma Aldrich, A2304) diluted 1∶25,000 in 3% BSA/PBS-T. The membranes were washed three times before they were incubated with chemo-luminescent HRP substrate for 5 minutes (Immobilon Western WBKLS0050; Millipore). The membranes were then exposed to Amersham Hyperfilm ECL for 30 seconds to 5 minutes for visualization of anti-OVA staining. The films were scanned using a Fluor-S MAX MultiImager (BioRad, Hercules, CA, USA) and the concentration of OVA released from the conjugate was quantified using the Quantity One v. 4.6.5 software (BioRad). The anti-OVA staining of the known OVA samples and the dilutions of the reduced conjugate were integrated and a standard curve was created. By fitting the integrated volumes of the anti-OVA staining of the reduced conjugate samples to the standard curve, the concentration of OVA in the conjugate estimated from the two gels was 2.4 and 2.5 mg/mL, respectively. With a volume of 2.5 mL this amounts to 6.0 mg of OVA conjugated to the PSGL-1/mIgG2b fusion protein. The final amount of conjugated PSGL-1/mIgG2b fusion protein was assumed to be equal to the amount applied in the conjugation reaction. The amounts in the conjugation and coupling yield for OVA for study A are shown in
In study B, quantification of conjugated OVA, PPM and CP (in conjugates OVA−PPM and OVA−CP) used for immunizations was done by SDS-PAGE and SilverQuest analysis (Invitrogen, LC6070) using OVA, PPM and CP of known concentration as standard. The OVA standard was determined with the BCA method using BSA as standard. The concentration of OVA in the stock solution was 2.3 mg/mL. Concentrations of stock solutions of PPM and CP were determined by ELISA to 4 mg/mL and 0.33 mg/mL, respectively. A dilution series of DTT-reduced samples was heat inactivated for 10 minutes at 70°C prior to separation on a 4–12% Novex Bis-Tris gel (MES buffer, 200 V for 45 minutes). Two SDS-PAGE gels were run, stained and analysed as described below. SilverQuest staining was performed according to the manufacturer's instructions. Briefly, the gels were rinsed in ultrapure water and fixed in 40% ethanol and 10% acetic acid in Ultrapure water, for 20 minutes with gentle rotation. The gels were washed in 30% ethanol for 10 minutes. The ethanol was decanted and sensitizing solution added to the gels and incubated for 10 minutes. The sensitizing solution was decanted, the gels washed in 30% ethanol for 10 minutes, and further washed in ultrapure water for 10 minutes. The gels were incubated in staining solution for 25 minutes, after which it was decanted and the gels washed with ultrapure water for 30 seconds. The gels were incubated in developing solution for 8 minutes both for the fusion protein and the OVA gels. Stop solution was then immediately added to the gels that were agitated for 10 minutes. The stop/developing solution was decanted and the gels washed in ultrapure water for 10 minutes. The gels were scanned using the Fluor-S MAX MultiImager (BioRad) and the concentration of OVA, PPM and CP released from the conjugate was quantified by fitting the integrated volumes of the OVA/PPM/CP staining of the reduced conjugate samples to the standard curve. For OVA the concentrations were 3.8 mg/mL and 3.4 mg/mL for the OVA−PPM and OVA−CP conjugates used in study B, respectively. PPM in OVA−PPM was estimated to 11.3 mg/mL and CP in OVA−CP was estimated to 7.9 mg/mL. The amounts and coupling yields for the conjugations as well as the molar ratios of OVA to fusion protein are shown in
Two separate studies were performed here described as study A and B. However, each experimental group has been repeated one to three times. In study A, 8 mice/group were immunized subcutaneously (s.c.) in the base of the tail using a dose of 50 µg OVA, either free or conjugated 1∶1 with mannosylated PSGL-1/mIgG2b (PPM) produced in
The antigen and antigen conjugate were either given alone or in combination with AbISCO®-100 or Alum (Imject®). The mice were immunized three times with three-week intervals (
Schematic picture of the experimental design and the different read-outs used in study A and B.
In study B, 8–10 mice/group were immunized as described above using a dose of 35 µg OVA either free or conjugated 1∶1 with mannosylated PSGL-1/mIgG2b (PPM) produced in
Mouse sera analyzed with regard to antibody levels and isotypes were collected before the first immunization (w0) and then 2 weeks after each immunization (w.2, w.5 and w.8) by retro-orbital bleeding of isofluorane-anesthetized mice. To free the sera of cells, the samples were centrifuged twice at 6,000× g. OVA-specific mouse immunoglobulins were quantified by ELISA. ELISA plates (Corning, Lowell, MA, USA, 3590) were coated using a 10 µg/mL OVA (Sigma Aldrich, A7641) solution, which was incubated in the plates o/n at +4°C. After every incubation step the plates were washed 4 times with 400 µL of wash solution (9 g NaCl/L H2O+0.05% Tween). Plates were blocked with 1% BSA in PBS for 1 hour at 37°C. Serial dilutions of sera in 1% BSA/PBS were analyzed in duplicates or triplicates and incubated for 1 hour at 37°C. Anti-mouse IgG (Southern Biotech, Birmingham, AL, USA, 1030-05), IgG1 (Southern Biotech, 1070-05), IgG2a (Southern Biotech, 1080-05), IgG2b (Southern Biotech, 1090-05) or IgG3 (Southern Biotech, 1100-05)-HRP conjugates diluted 1∶2,000–8,000 and incubated for 1 hour at 37°C was used for detection of the different IgG subclasses. The TMB (tetramethylbenzidine, Sigma Aldrich, T3405) substrate (1 tablet) was dissolved in 10 mL phosphate citrate buffer, pH 5.0 containing 3 µL 30% H2O2 per 10 mL buffer and was used for detection of HRP conjugates. The reaction was stopped after 3–5 minutes with 2 M H2SO4. Optical density (OD) was measured in a TECAN Sunrise spectrophotometer (TECAN, Männedorf, Switzerland) at 450 nm within 2 hours after addition of H2SO4. An antibody titer was considered positive if the OD value was three times that of the animal serum collected prior to the first immunization. A hyperimmune serum of known titer was used as positive control. Pooled serum from non-immunized wt C57Bl/6J mice was used as negative control.
ELISA plates (Corning, Lowell, MA, USA, 3590) were coated using a 2 µg/mL PPM, CP or mIgG Fc fragment (Jackson ImmunoResearch, 015-000-008) o/n at +4°C. After every incubation step the plates were washed 4 times with 400 µL of wash solution (9 g NaCl/L H2O+0.05% Tween). Plates were blocked with 1% BSA in PBS for 1 hour at RT. Serial dilutions of pooled sera in 1% BSA/PBS were analyzed in duplicates and incubated for 2 hour at RT. An anti-mouse IgG Fab specific HRP antibody (Sigma, A2304) diluted 1∶5,000 and incubated for 2 hours at RT was used for detection. HRP conjugate was detected exactly as described in section 2.7. An antibody titer was considered positive if the OD value was three times that of the animal serum collected from the control group of mice only receiving PBS at each immunization.
Spleens from OVA immunized C57Bl/6J mice were collected two weeks after the final immunization and single cell suspensions were prepared in RPMI-1640 medium containing 100 U/mL penicillin and 100 µg/mL streptomycin (GIBCO, Invitrogen, 15140). Red blood cells were removed using Red blood cell lysing buffer (Sigma Aldrich, R7757). Immune spleen cells (25×106) were stimulated
Spleen cells from four to five individual mice in each group were pooled and immediately tested for the presence of OVA-specific T cells. Spleen cells from the other four to five individuals in each group were used to repeat the experiment with consistent results. The ability of OVA-specific Th and CTLs to produce IFN-γ, IL-2, IL-4 and IL-5 after exposure to different peptides SIINFEKL (CTL, OVA257–264; Innovagen), FILKSINE (CTL, irrelevant peptide; Innovagen), and ISQAVHAAHAEINEAGR (Th, OVA323–339; Innovagen), proteins (OVA grade VII and BSA, Sigma Aldrich), Concanavalin-A (Sigma Aldrich) and media was assessed. The production of the different cytokines was determined by a commercially available ELISpot assay. In brief, ELLIP plates (Millipore, cat. no MAIPSWU) with PVDF membranes were treated with 70% ethanol for 1 minute, washed in sterile water and coated o/n at +4°C with 10 µg/mL of monoclonal antibodies specific for IFN-γ (AN18), IL-2 (1A12), IL-4 (11B11) or IL-5 (TRFK5) (Mabtech AB, Nacka strand, Sweden) in PBS. After washing 5 times in PBS, the plates were blocked for 2 hours with complete RPMI-1640 medium. All stimulations (36 hours at 37°C, 5% CO2) were carried out using 250,000 immune cells/well. Various concentrations of the different antigens were added in triplicates to a total volume of 200 µL. After stimulation, the wells were washed and incubated for 2 hours at 37°C with the following biotinylated antibodies, respectively: anti-IFN-γ (R4-6A2-biotin), anti-IL-2 (5H4-biotin), anti-IL-4 (BVD6-24G2-biotin) and anti-IL-5 (TRFK-biotin) (Mabtech AB) at 2 µg/mL in 0.5% FBS/PBS (Sigma Aldrich, F6178). After washing, Strep-ALP (Mabtech AB, 3310-10) diluted 1∶1,000 in 0.5% FBS/PBS was added and incubated for 1 hour in RT. Sterile-filtered substrate, BCIP/NBT (Mabtech AB), was used to develop spots; IFN-γ and IL-2 for 10 minutes, IL-4 for 12 minutes and IL-5 for 14 minutes. The substrate reaction was stopped by rinsing extensively with dH2O, after which the plates were left to dry. The number of spots was counted using the AID ELISpot reader and software ver. 3.2.3 (AID, Strassberg, Germany). The number of spots (cytokine producing cells) was determined at each concentration of peptide or protein and the results given as the number of IFN-γ, IL-2, IL-4 or IL-5-producing cells per 106 cells. A mean number of cytokine-producing cells of <50 per 106 cells was considered as negative.
Proliferative responses to OVA and OVA peptides were determined by stimulation of splenocytes from groups of mice immunized with the different vaccine compositions. A total of 600,000 cells/well in complete RPMI-1640 medium were seeded in 96-well flat bottom plates with lid (Corning, 3595). Stimulation was carried out for 48 hours at 37°C in a 5% CO2 atmosphere using the same antigens and concentrations as in the ELISpot assay. At 48 hours, 0.1 Ci/mL 3H-thymidine (TRA120-5MC; GE Heathcare, Chalfont St. Giles, United Kingdom) was added and 16–20 hours later the cells were harvested onto filtermat A filters (Wallac,145–421) and the radioactivity counted in a TRILUX 1450 MicroBeta counter (Wallac). Proliferation was determined by dividing the radioactivity as counts per minute (cpm) of cells incubated with an Ag with the cpm of the cells incubated with medium alone (sample to negative (S/N) ratio). Groups were compared by the mean S/N ratios at each time point after subtraction of proliferative responses seen in the negative control group receiving PPM alone. All samples were run in triplicates.
The ANOVA with Tukey post-hoc test was used for statistical analyses using the JMP version 8.0.1 for PC software (SAS Institute Inc., Cary, North Carolina, USA). P-values <0.05 were considered statistically significant.
Total IgG was compared between all experimental groups at week 0 (data not shown), 2, 5 and 8 using mouse serum from individual mice. In study A (
Serum IgG antibody titers against OVA in study A (
In order to examine the importance of the oligomannose residues on the fusion protein for its adjuvant effect on the IgG antibody response, we included in study B a control fusion protein (CP) lacking oligomannose residues. At week two the group immunized with OVA−PPM+AbISCO®-100 had significantly higher antibody titers than the other groups including the group immunized with a fusion protein (CP) conjugate lacking oligomannose residues (
The IgG isotype distribution within each group was evaluated at week eight using sera from individual mice (
In study B (
Anti-mIgG responses were compared between the experimental groups at week 8 using pooled mouse sera from groups of mice included in study A and B (
Serum IgG antibody titers against PPM, CP or mIgG Fc in groups of mice from study A and B. Titres were determined from pooled sera of mice in each group (8–10 mice/group) at week 8, and were defined as the reciprocal endpoint dilution giving an optical density at 450 nm of 3 times the background value.
To study if the different immunizations were able to induce cellular immune responses with detectable OVA-specificity, lytic activity mediated by cytotoxic T lymphocytes were tested in 51Cr release assays using
Detection of OVA peptide-specific cytolytic activity of splenocytes isolated from the different groups of mice and restimulated
In Study B, there was significantly higher specific lysis at E∶T ratios of 80∶1 and 40∶1 in the groups immunized with OVA−PPM+AbISCO®-100 (mean 80.2±12.5; p<0.05 lysis and 73.3±14.8; p<0.05 lysis, respectively) and OVA mixed with PPM+AbISCO®-100 (mean 72.2±16.6; p<0.05 lysis and 62.9±17.0; p<0.05 lysis, respectively) compared to the other groups (
To further characterize and quantify the type of immune response elicited, we studied the
In study A, IFN-γ producing cells were mainly seen in groups of mice immunized with compositions containing AbISCO®-100 (
Quantification of OVA-specific, γ-IFN-producing spot forming cells (SFC) after a 36 hour
In study B, the IFN-γ response was shown to be highest in the group of mice immunized with OVA−PPM+AbISCO®-100, which was the only group that responded to the OVA-Th peptide (
In study A, the differences observed with regard to the number of IFN-γ secreting cells were also seen when the number of IL-2 producing cells was assessed (
Quantification of OVA-specific, IL-2-producing spot forming cells (SFC) after a 36 hour
In study B, the differences observed between the groups with regard to the number of IFN-γ secreting cells was not as clear when the number of IL-2 producing cells was assessed (
In relation to the Con-A control, the IL-4 ELISpot assay revealed fewer SFC as compared to both the IFN-γ and IL-2 ELISpot (
IL-5 producing cells were detected in low numbers and only when using OVA concentrations of 25–625 µg/mL (
In an attempt to compare the proliferative responses between the groups, a 3H-thymidine proliferation assay was performed. In this assay, the ratio between the cpm value of splenocytes stimulated with the different antigens (in triplicate) and the cpm value of splenocytes incubated in medium alone was first calculated. These values were then compared to the proliferation seen in the control group immunized with PPM alone. Each graph describes the proliferative responses from a pool of splenocytes isolated from four mice within each group. The experiment was repeated twice with consistent results. The analysis revealed that OVA-specific proliferation was detectable mainly in the groups of mice immunized with antigen compositions containing AbISCO®-100 (
Proliferative OVA-specific responses after a 68 hour stimulation of splenocytes (pools) with indicated antigens. The values in each group, presented as the S/N (sample to negative) ratio, indicates the difference between antigen-induced and spontaneous proliferation.
Numerous strategies have been suggested aiming at developing vaccine compositions targeting antigen presenting cells (APC) in order to improve antigen immunogenicity and elicit a Th1 response with the development of cytotoxic T lymphocytes (CTLs). The ability to generate CTL responses, and the killing of tumor cells and cells harboring intracellular pathogens, is maybe the most important feature of a therapeutic vaccine. One APC-directed strategy involves targeting mannose-binding receptors on macrophages and dendritic cells in order to improve vaccinogen uptake and MHC presentation (reviewed in
In combination with AbISCO®-100, the OVA − mannosylated PSGL-1/mIgG2b conjugate elicited a significantly faster and stronger antibody response compared to when OVA alone was used as antigen. Mannosylated PSGL-1/mIgG2b improved the anti-OVA IgG response also without AbISCO®-100, but the response was weaker. The anti-OVA response was broader with regard to the IgG subclasses being induced and only the OVA − mannosylated PSGL-1/mIgG2b conjugate with AbISCO®-100 induced an IgG2a response. IgG1 was the predominant IgG subclass detected suggesting an immune response skewed also towards a Th2 type of response. IgG2a and IgG2b antibody titers were, however, only detectable after inclusion of AbISCO®-100 and was stronger in the OVA − mannosylated PSGL-1/mIgG2b+AbISCO®-100 groups. These IgG subclasses would indicate a Th1 immune profile. The Th1 response is further evidenced by the generation of a strong OVA-specific CTL response and increased numbers of IFN-γ and IL-2 producing splenocytes in groups immunized with the OVA − fusion protein conjugate in the presence of AbISCO®-100. Increased CD4+ T-cell proliferation was also found, further confirming the ability of the OVA − mannosylated PSGL-1/mIgG2b+AbISCO®-100 combination to strongly activate the immune system. However the strong CTL response observed with the combination of OVA−PPM+AbISCO-100 was not seen when OVA−PPM was combined with Alum. Alum is the foremost vaccine adjuvant for clinical applications but it is a poor inducer of cellular immunity and the adjuvant mechanism of alum is not fully understood. Nevertheless in our study the combination with OVA−PPM+alum resulted in a somewhat broader humoral immune response compared to the OVA+alum group. It is thus evident that the combination of PPM and alum does not result in the strong synergy effect seen when PPM is combined with AbISCO-100. Further studies with PPM in combination with other adjuvant systems and/or antigens are necessary to evaluate PPMs adjuvant potency.
We found that groups of mice injected with the mucin-type fusion protein developed antibodies against the fusion protein irrespective of O-glycan substitution. Whether these antibodies were directed against the recombinant PSGL-1 part or the mouse IgG2b Fc part is at the moment not clear. The fact that antibodies binding to the fusion protein itself were induced could limit its clinical use, especially upon re-utilization. However, such antibodies against the fusion protein may not necessarily have limitations. In fact, antibodies against the fusion protein may act to further improve antigen uptake and presentation. More serious, however, was the observation that groups receiving AbISCO®-100 developed antibodies reactive with mouse IgG Fc fragments, which indicates that self tolerance was broken. Low titers of antibodies against the CP55.2 fusion protein and mouse IgG was observed in the OVA alone+AbISCO®-100 group, suggesting that a weak polyclonal B cell activation might have occurred. Clinical relevance of such responses will need to be assessed in the clinical development of any AbISCO®-100 mannosylated PSGL-1 mIg-based candidate. Whether this can lead to autoimmunity remains to be seen.
The fact that mannosylated PSGL-1/mIgG2b binds to MMR, DC-SIGN and MBL
Mannose receptor-mediated uptake of antigen has been shown to improve T-cell presentation a 100-fold compared to fluid phase uptake
A role for mannose-binding receptor targeting and enhanced antigen uptake is also suggested by the fact that O-glycan oligomannoses are required on PSGL-1/mIgG2b for an optimal immune-stimulating effect. When OVA was conjugated to a fusion protein expressed in CHO cells and carrying mono and disialylated core 1 structures, weaker humoral and cellular anti-OVA responses were detected. When comparing conjugated OVA with just mixing, conjugation of OVA to mannosylated PSGL-1/mIgG2b appear to give more rapid, stronger and broader antibody responses than when OVA is just mixed with mannosylated PSGL-1/mIgG2b.
Antigen-specific CTL activities are important for control of virus infected cells and tumors
In addition to the mentioned receptors, other lectins may also be involved. For example, Dectin-1 belongs to the C-type lectins like MBL, MR and DC-SIGN and has been shown to bind cell wall components and beta-glucans of fungal pathogens including
In conclusion, we have shown that the mannose structures in the fusion protein play a decisive role for inducing a broad immune response with a rapid and strong antibody response and a strong CTL response. When comparing conjugated OVA with just mixing, conjugation of OVA to mannosylated PSGL-1/mIgG2b appear to give a more rapid, stronger and broader antibody response than when OVA is mixed with mannosylated PSGL-1/mIgG2b. Hence this study demonstrates that PSGL-1/mIgG2b produced in
IL-4 producing cells determined by ELISpot assay. Quantification of OVA-specific, IL-4-producing spot forming cells (SFC) after a 36 hour
(TIFF)
IL-5 producing cells determined by ELISpot assay. Quantification of OVA-specific, IL-5-producing spot forming cells (SFC) after a 36 hour
(TIFF)