|
|
||||||||





* Institute of Virology, Prion Research Group, Technical University of Munich, Munich, Germany;
Institute of Medical Microbiology, Immunology and Hygiene, Technical University of Munich, Munich, Germany;
Institute of Pathology, GSF National Research Center for Environment and Health, Neuherberg, Germany;
Department of Chemistry, Institute of Biotechnology, Technical University of Munich, Munich, Germany; and
¶ Unité de Virologie et Immunologie Moléculaires, Biologie Physico-chimique des Prions, Jouy-en-Josas, France
| Abstract |
|---|
|
|
|---|
-helical cellular prion protein (PrP) into the abnormal β-sheet-rich infectious isoform (PrPSc). The immune system neither shows reaction against cellular PrP nor PrPSc, most likely due to profound self-tolerance. In previous studies, we were able to partly overcome self-tolerance using recombinantly expressed dimeric PrP (tandem PrP (tPrP)), in association with different adjuvants. Proof of principle for antiprion efficacy was obtained in vitro and in vivo. In this study, we demonstrate the induction of a specific Th1 T cell response in wild-type mice immunized with tPrP and CpG-oligonucleotide (ODN). Biochemical influences such as refolding conditions, ionic strength, pH, and interaction with CpG-ODN affected antigenic structure and thus improved immunogenicity. Furthermore, s.c. immunization with tPrP and CpG-ODN coencapsulated in biodegradable polylactide-coglycolide microspheres (PLGA-MS) enhanced CD4 T cell responses and, more prominent, the induction of CD8 T cells. In this vaccination protocol, PLGA-MS function as endosomal delivery device of Ag plus CpG-ODN to macrophages and dendritic cells. In contrast, PLGA-MS-based DNA vaccination approaches with a tPrP construct generated poor humoral and T cell responses. Our data show that prophylactic and therapeutic immunization approaches against prion infections might be feasible using tPrP Ag and CpG-ODN adjuvant without detectable side effects. | Introduction |
|---|
|
|
|---|
-helix-rich PrPC turns into highly insoluble, mainly β-sheeted PrPSc becoming partially resistant to proteolytic digestion (2, 5, 6). PrPC is a highly conserved protein of unknown function, which is posttranslationally modified in the secretory pathway and is attached to the outer leaflet of the plasma membrane via a glycolipid anchor. The subcellular locale of conversion to PrPSc is assumed to be the plasma membrane or a compartment in the early endocytic pathway, probably rafts (7, 8, 9). PrPC is widely expressed in the body, with notably high levels in neurons. PrPC is essential for the propagation of PrPSc, as shown by knockout mice (PrP0/0) mice, which are resistant to prion infection (10). It was shown some time ago that polyclonal anti-PrP Abs can reduce prion infectivity of hamster brain homogenates (11) and that anti-PrP Abs inhibit the formation of protease-resistant PrP in a cell-free conversion assay (12). Anti-PrP mAbs (13, 14, 15) and anti-rPrP-directed Fab (16) were shown to suppress prion replication in cell culture models. Passive immunization with mAbs was effective when mice were infected by the i.p. route, albeit only when very high amounts of mAbs were applied (17). Transgenic expression of an anti-PrP Ab showed the most promising proof of principle that protection against prion disease by immunization is feasible, interestingly, without inducing obvious side effects (18). Presumably, the major obstacle for obtaining specific anti-PrP immune responses represents the pronounced self-tolerance to host-encoded PrP. Of note, the sequence of prions generated de novo in an infected individual always reflects the species of the recipient host and is therefore not foreign in its primary structure.
Many groups attempted to overcome self-tolerance, and some were able to induce a humoral immune response either by using synthetic peptides with or without modifications (19, 20, 21, 22), full-length rPrP (23, 24, 25), or recombinant tandem PrP (tPrP) (24, 26). Other approaches dealt with DNA vaccines (27, 28), retroviral particles (29), or alternative application routes (30, 31), the latter ones being also effective against prion challenge in vivo. Some groups were reasonably successful in inducing a PrP-specific T cell response in PrP0/0 mice (32, 33), and to a much lower extent in wild-type (wt) animals, by either peptide immunization (19, 20, 34) or using recombinant proteins of a different species (35).
Immunostimulatory CpG-oligonucleotide (ODN) was shown to be a potent stimulator of innate immune cells, such as macrophages and dendritic cells (DCs) (36, 37, 38). In DCs and macrophages, CpG-DNA activates maturation by binding to TLR9, thereby facilitating cross-presentation of exogenous Ags (39, 40, 41). As TLR9 is expressed within phago-endosomes, concurrent endosomal translocation of Ag and CpG-ODN results in a significant enhancement of T cell stimulation (39, 40, 41). Endosomal translocation of Ag and CpG-ODN can also be achieved by coencapsulation of Ag plus CpG-ODN into biodegradable microparticles that act as endosomal delivery device (42).
In this study, we coencapsulated tPrP and CpG-ODN in biodegradable polylactide-coglycolide microspheres (PLGA-MS) in attempts to vaccinate for anti-PrP immune responses. We show that the mode of Ag preparation affects the magnitude of immune response, influenced by refolding conditions and interaction with adjuvant CpG-ODN. We show for the first time that tPrP is able to induce a significant CD4+ T cell response in spleen of immunized wt mice. Furthermore, coencapsulation of tPrP and CpG-DNA in PLGA-MS caused a significant enhancement of T cell responses, as indicated by increased CD4+ responses and also by significant PrP-specific CD8+ T cell responses. Despite these pronounced immune responses, no side effects related to autoimmune diseases were observed. Our results indicate that PLGA-MS-based anti-PrP vaccination strategies may provide a feasible and effective vaccination approach for prevention and therapy of prion-based diseases.
| Materials and Methods |
|---|
|
|
|---|
Female PrP wt NMRI mice were obtained from Janvier. NMRI mice are of prnp-A genotype, as are, for example, CD1 and C57BL/6 mice. All animal experiments were in accordance with German animal experimentation regulations.
Recombinant tPrP
Murine PrP gene subunit (prnp-A) for monomeric PrP (mPrP) was amplified by PCR, as described (26). tPrP consists of two murine PrP sequences (aa 23–231), lacking the C- and N-terminal signal peptides, and covalently linked by a 7-aa linker (AGAIGGA). Each PrP moiety contains a 3F4 epitope tag. By cloning into bacterial expression vector pQE30 (Qiagen), an N-terminal poly-histidine tag for purification was added (26). tPrP expression was done in Escherichia coli strain BL21-Gold(DE3) pLysS (Stratagene). Cell lysis with 6 M guanidinium hydrochloride and purification in 8 M urea buffer with ProBond Ni2+ columns (Sigma-Aldrich) was done, as described previously (26). Protein was refolded by dialysis against ddH2O or 10 mM NaAc buffer (pH 4.5) overnight at 4°C under permanent stirring with precooled buffers (10,000 MWCO Dialysis Cassettes; Pierce). Protein concentration was measured with the bicinchoninic acid kit (Pierce). For solubility test, protein samples were adjusted to final protein concentration of 1 mg/ml, mixed with respective CpG amounts, and analyzed by differential ultracentrifugation in a TL-100 ultracentrifuge (1 h at 100,000 x g; TLA-45 rotor). Protein concentration of the supernatant was measured with bicinchoninic acid kit and calculated in percentage of soluble control sample.
Preparation of microspheres
Microspheres were prepared using solvent/evaporation technique (43). Polymer was emulsified in methylene chloride with Ag/adjuvant solution at high speed. This primary organic phase was then mixed with an aqueous phase containing 3% polyvinyl alcohol, followed by stirring for 4–5 h, allowing the methylene chloride to evaporate. Further details are described elsewhere (42).
Preparation of DNA vaccination microspheres
The 3F4-tagged full-length murine PrP DNA sequence was obtained by subcloning from a vector construct (44) using restriction enzymes XbaI and PstI (NEB). tPrP sequence bearing 3F4 tags and flanked by signal peptides (1–231-linker-23–254) was subcloned using BamHI and XbaI (26). Both constructs were ligated into the DNA vaccination vector pVAX (Invitrogen Life Technologies) and purified with Qiagen Plasmid Giga kit before coencapsulation into microspheres with CpG-ODN (42).
Atomic force microscopy (AFM)
Protein samples for AFM were dialyzed and diluted to a concentration of 0.1 mg/ml with respective dialysis buffer. Samples were mixed with CpG to a final concentration of 100 µM and immediately placed on freshly cleaved mica attached to AFM sample discs (Ted Pella). After 3 min of adsorption at 25°C, the discs were washed five times with ddH2O before allowing to air dry. Contact mode imaging was performed on a multimode scanning probe microscope (Veeco) by using silicon nitride probes (type DNP-S20, Veeco NanoProbe Tips; spring constant K = 0.06 N/m).
Fourier transformation infrared spectroscopy (FTIR)
tPrP was dialyzed either with 10 mM NaAc (pH 4.5) or ddH2O overnight, followed by desalting with G25 MicroSpin devices (GE-Healthcare) against acetate Na/D2O (pD 4.1) or D2O, respectively. The protein concentrations were 6 mg/ml. FTIR spectra were recorded with a Jasco 810 infrared spectrometer equipped with a thermostated cell holder. Each spectrum was an overage of 20 scans with a 4 cm–1 resolution. The spectrum deconvolution was made using homemade software.
Size exclusion chromatography
Size exclusion chromatography was done with Åkta fast protein liquid chromatography equipment (GE-Healthcare) and a 7 x 600-mm gel filtration column TSK 4000SW (Interchim). Before use, the column was calibrated for molecular mass and Stokes radius with low and high molecular mass calibration kits (GE-Healthcare). Previously to each measurement, the column was equilibrated with a minimum of 4-column vol of elution buffer. Flow rate was 1 ml/min at 20°C; protein elution was monitored with UV absorption at 280 nm.
Immunization
NMRI mice were immunized with 100 µl containing 1 mg/ml freshly dialyzed protein; CpG 1826 was added directly before administration to a final concentration of 100 µM. Injections were given in two or more portions s.c. into the dorsal region. Three boost injections were administered every 3 wk. Blood samples and organs were taken 10–21 days after the final boost. For microsphere immunization, 5 mg of the respective PLGA-MS was diluted in PBS and injected s.c. into the tail base in 4-wk intervals, a total of four injections. DNA vaccination was performed by i.m. application of PBS-diluted PLGA-MS in two 50-µl portions.
ELISA
Ninety-six-well plates were coated with 150 µl of sodium-carbonate buffer (0.1 M (pH 9.5)) containing 1 µg of NaAc-dialyzed mPrP or tPrP and incubated overnight at room temperature. After washing six times with PBST (PBS, 1% Tween 20), the plates were blocked for a minimum of 2 h with 150 µl of PBST, 3% BSA, at 37°C. Sera of mice were diluted (as indicated) in PBST with 3% serum albumin, and 100 µl/well was incubated for 1 h at 37°C. After washing six times with PBST (300 µl in a Tecan plate washer), plates were incubated with 1/4000 diluted HRP-labeled anti-mouse IgG Ab (GE-Healthcare) for 1 h at 37°C. The plates were washed again and incubated with 100 µl of ABTS solution (Sigma-Aldrich). After 10 min at 25°C, the OD405 was measured (Tecan plate reader). Cutoff was defined as three times the OD405 value of 1/100 diluted preimmune serum of respective mice.
For Ab isotypes, plate preparation was the same as for ELISA. Sera were diluted 1/100 and incubated with 1/4000 dilution of biotin-labeled secondary Ab for 45 min (IgG1, IgG2a, IgG2b, IgG3, IgA, IgM). After washing, wells were incubated with HRP-labeled streptavidin with 1/1000 diluted for 30 min. The final wash was followed by 10-min incubation with ABTS solution and measurement of OD405.
Epitope mapping
For epitope mapping, a peptide bank encompassing the mature full-length murine PrP consisting of 20 residues with an overlap of 5 residues was used (PrP 23–231; depicted in Fig. 2C). Peptide 6 encompasses the 3F4 epitope; peptide 6b represents the corresponding wt murine sequence. The poly-histidine (MRG SHH HHH HGS CKK RPK PG) and linker region (DGR RSS AGA IGG AKK RPK P) of tPrP was used in addition. As full-length protein control nontagged mPrP and tPrP were used. Polyclonal Ab A7 was obtained by tPrP immunization in rabbit (26). The assay was performed as described in detail elsewhere (26). In brief, CovaLink NH microtiter plates (Nunc) were activated with the bifunctional linker disuccinimidyl suberate in carbonate buffer and incubated with peptide or recombinant protein, respectively, overnight at room temperature. The coated plates were blocked and incubated with prediluted sera, washed, and incubated with corresponding HRP-labeled conjugate. Upon washing and incubation with ABTS solution, OD405 was measured and documented by scanning or digital photography.
|
Spleens of individual mice were harvested, isolated with a 100-µm cell strainer mesh (BD Biosciences), and individually analyzed. Lysis of erythrocytes was achieved by incubation with NH4Cl (0.15 M (pH 7.4)) for 7 min at 25°C, followed by filtration with 100-µm mesh. A total of 2 x 107 cells in 2 ml of RPMI 1640 medium, containing 5% FCS and penicillin/streptomycin, was plated in 12-well plates and restimulated for 16 h either with 20 µg of freshly dialyzed tPrP or mock treated at 37°C. Control cells were treated with PMA/ionomycin (25 ng/ml, 1 µg/ml, respectively) for 6 h. A total of 2 µl of Golgi plug (BD Biosciences) was added per 3 h before addition of ethidium monoazide solution (1 µl of ethidium monoazide, 10 µl of anti-mouse CD16/CD32 (Fc-Block; BD Biosciences) per ml wash buffer (PBS with 2.5% FCS)). Incubation on ice with strong light for 20 min was followed by transfer and separation into 96-well plates. After washing three times with 300 µl wash buffer, cells were surface stained with 1/100 dilution of PE-Cy5-CD4 and PE-CD8 Abs (BD Biosciences) in wash buffer for 20 min. Washing was performed in accordance with the manufacturers instructions.
-IFN-
FITC- or FITC-labeled isotype Ab were diluted 1/500 in Permwash buffer and incubated 30 min before final washing. Samples were dissolved in PBS/1% paraformaldehyde. Samples were measured with a Beckman Coulter Epics XL flow cytometer (Beckman Coulter); resulting data were analyzed by EXPO 32 software.
Histology
The brain, spleen, mesenterial lymph nodes, and small and large intestine were explanted and immediately fixed in Roti-Histofix (Roth) and embedded in paraffin for histological examination. Sections (3–5 mm thick) were cut and stained with H&E.
| Results |
|---|
|
|
|---|
Previously, we have shown that tPrP is able to partly overcome self-tolerance to PrP in PrP wt mice (26). This feature was obviously influenced by the design and preparation of the Ag and of the adjuvant used. Of note, rmPrP was significantly less or not effective in our hands (24, 26). Therefore, we first analyzed in more detail the biochemical properties of tPrP.
tPrP was expressed in E. coli and purified under denaturing conditions, as described previously (26). Analysis of structural features was done by FTIR, because precipitated protein also can be measured. Fig. 1A shows FTIR spectra from measurements of different proteins and preparations, including deconvolution.
-Helical structure is characterized by a peak at 1651 cm–1 (in blue), and β-sheet content by a peak at 1619 cm–1 (in red); percentage rates are indicated in respective figures. rmPrP, confirming previous CD measurements, contained mainly
-helical structure after refolding (
-helix, 41%; β-sheet, 5%). In contrast, tPrP showed the formation of mainly β-sheet structure with both dialysis buffers. tPrP(NaAc) revealed a higher β-sheet content (
-helix, 15%; β-sheet, 37%) compared with tPrP(H2O) (
-helix, 18%; β-sheet, 25%).
|
70% soluble (Fig. 1B). Interestingly, mixing of refolded protein with adjuvant CpG-ODN led to very substantial and immediate precipitation. Virtually all tPrP(NaAc) precipitated after addition of 5 or 10 nmol CpG, whereas 28 and 23% tPrP(H2O) remained soluble, respectively (Fig. 1B). From these data we conclude that CpG significantly influenced the aggregation behavior of the tPrP immunogen. To gain further insight into the aggregation behavior of tPrP in the presence or absence of CpG, we applied AFM (Fig. 1C). Without CpG, AFM revealed no obvious difference between tPrP(H2O) and tPrP(NaAc) in structural appearance. Evenly widespread, unstructured proteins are visible in both preparations. In contrast, protein CpG mixtures showed the formation of two different types of aggregates. Canyon-like structures were of two-dimensional appearance and seemed to consist of smaller aggregates gathered in bigger groups. The second type of aggregate appeared as three-dimensional clumps in tPrP(NaAc) preparations, and smaller and less frequently in tPrP(H2O) samples, indicating unstructured, oligomeric aggregates (Fig. 1C).
We next used size exclusion chromatography to study the aggregation state of tPrP depending on ionic strength (Fig. 1D). Under low ionic strength buffer conditions, comparable to our dialysis buffer, tPrP was found mainly in a monomeric state. In contrast, high ionic strength, corresponding to conditions after injection into mice, showed significant oligomerization (Fig. 1D).
Taken together, the biochemical characterization clearly shows the formation of mainly β-sheets for refolded tPrP, in contrast to high
-helical content in mPrP. Furthermore, solubility and conformation of rtPrP are influenced by refolding conditions, ionic strength, and interaction with CpG adjuvant.
Ag preparation influences the induction of PrP autoantibodies
Having found that various conditions significantly affected the biochemical properties of tPrP Ag, we next asked whether this is reflected by variations in humoral immune response in vivo. Therefore, we immunized PrP wt mice with different tPrP preparations and measured Ab titers, isotypes, and linear epitopes of individual mice after final boosts. Three times OD of 1/100 diluted preimmune sera of respective mice were defined as cutoff in the recombinant PrP (recPrP)-specific endpoint titration ELISA. Coating plates with either mPrP or tPrP revealed no considerable differences in OD values (data not shown). Fig. 2A shows the endpoint titers of individual mice of three immunization groups. None of the CpG-only-treated mice displayed a titer, whereas 3 of 12 tPrP(NaAc)-treated mice clearly showed autoantibody titers. Of note, in the group of tPrP(H2O)-immunized mice, 8 of 12 mice exhibited anti-PrP Abs. However, the mean titer values between the two groups did not significantly differ.
Isotyping of titer-positive sera from our study showed no differences within the two protein preparation groups and revealed the presence of mainly IgG1, IgG2a, and IgG2b, but not of IgG3, IgA, or IgM Abs (Fig. 2B). To further analyze differences in quality of humoral response, we performed a linear epitope mapping (Fig. 2, C and D). We found reactive Abs mainly against epitope 10 and several N-terminal epitopes, confirming previous results (26). Neither reactivity against epitope 6, representing the 3F4 epitope tag sequence contained in tPrP used for immunization in this study, nor against to the respective wt epitope 6 was found. In addition, no binding to potential poly-histidine tag or linker peptide epitopes was detected. We found no differences in signal intensity of full-length rmPrP without tags and tPrP controls. The reactivity of a polyclonal antiserum obtained by immunization of rabbits with mouse tPrP is shown for comparison (A7; Fig. 2, C and D).
In summary, B cell reaction, upon immunization with different Ag preparations, mainly varies in number of responding animals, and not in the quality of response because isotypes and linear epitopes are similar. Only few mice react after administration of tPrP(NaAc), although with high titers, whereas the majority of tPrP(H2O)-treated mice react with a broad variation of individual titers.
T cell response depending on the structure of tPrP
Next, we examined whether intensity of T cell responses is influenced by Ag preparation, similarly to the differences in B cell response described above. Therefore, spleen cells of individual immunized mice were harvested and individually restimulated with tPrP or not. Cells were CD4 and CD8 surface labeled to distinguish T cell subpopulations and stained for intracellular IFN-
.
Fig. 3A shows the percentage rates of IFN-
-expressing CD4+ splenocytes from all individual mice; horizontal lines indicate the mean values of groups. Two mice of the tPrP(NaAc) group responded with a well-defined IFN-
expression, whereas the others showed only a weak increase. Overall, the results were statistically significant compared with the CpG-only control group (p = 0.0159; Mann-Whitney t test). Within the tPrP(H2O) group, four of five animals clearly responded with even higher IFN-
mean values, thus leading to very significant increase compared with the control group (p = 0.0079). From CD8+ T cell subpopulation, only one mouse of the tPrP(H2O) group responded with a slightly increased IFN-
expression, not leading to significant mean value (Fig. 3B). Exemplary results of FACS measurements are shown as dot-blot analysis of CD4+ or CD8+ T cells (Fig. 3, A and B, right panels).
|
Microsphere encapsulation drastically enhances T cell response
Encouraged by the promising T cell responses found in some individual mice, we tried to increase this effect by encapsulation of immunogens in PLGA-MS. We used two different PLGA-MS preparations for our immunization trials. PLGA-MS containing either tPrP or CpG-DNA exclusively were mixed together and applied to mice s.c. at the tail base. To the other PLGA-MS coencapsulating both protein and CpG-DNA were used. T cell responses of individual mice from both groups displayed partial IFN-
secretion upon in vitro restimulation of primed splenocytes (Fig. 4). The CD4+ T cells in the group of coencapsulated PLGA-MS revealed a clear response in seven of eight animals, including two high responders (Fig. 4A). These results were statistically significant compared with the group of mixed PLGA-MS (p = 0.0109; Mann-Whitney t test), therefore representing an ideal control group. Furthermore, CD8+ T cells of individual mice primed with coencapsulated PLGA-MS showed a consistent IFN-
secretion (Fig. 4B), leading to a statistically significant difference compared with the control group (p = 0.0186; Mann-Whitney t test).
|
DNA vaccination with encapsulated tPrP-encoding vector fails to induce a significant immune response
DNA vaccines have been shown to be very efficient when applied encapsulated in microspheres (45). Vaccination of wt mice with either pVAX (mPrP) or pVAX (tPrP) in combination with CpG, applied via microspheres i.m., resulted in very modest immune responses of only a few animals (Fig. 5). Only one mouse of eight from the pVAX (tPrP) group showed an Ab titer in an endpoint titration ELISA not higher than 1:200 (Fig. 5A). Two animals showed a slight expression of IFN-
in CD4+ splenocytes in the same group (Fig. 5B). No activation of CD8+ cells was seen (data not shown).
|
Histology and examination of side effects
Given the possibility that PrP autoimmunization might result in induction of severe side effects in reactive mice, we performed a detailed histological examination. This was even more important, because it was reported that repeated application of high CpG doses lead to major deleterious effects in immunological organs (46). Histological examinations of representative lymphatic organs displayed no severe side effects in the autoimmunization situation described in this study. Neither the application of CpG alone (Fig. 6A), nor treatment with tPrP(NaAc) (Fig. 6B) or tPrP(H2O) (Fig. 6C), both in combination with CpG, revealed signs of severe changes in the lymphatic organs in any of the analyzed mice. In the gut lumen of ileum, only small-sized or no Peyers patches were visible after treatment with tPrP(H2O); tPrP(NaAc)-treated mice showed normal- or middle-sized Peyers patches without germinal centers.
|
|
| Discussion |
|---|
|
|
|---|
In previous studies by us (26) and others (23, 24, 25, 33, 35), it became obvious that Ag design and application are major determinants in overcoming PrP self-tolerance. For example, we described the induction of comparable Ab titers and binding to the same linear epitopes after mPrP and tPrP immunization in wt mice (26). From the fact that only tPrP-induced sera led to an efficient blocking of the de novo PrPSc formation in prion-infected cell culture, we concluded the induction of Abs with conformational epitopes, eventually specific to PrPSc or a folding intermediate. In this study, we analyzed some biochemical features of the tPrP Ag in attempts to optimize conditions for improved immune responses. We also examined the T cell responses that might help to trigger B cells producing conformational Abs, and we tried improved methods for Ag administration by using PLGA-MS and DNA vaccination approaches.
The conformation of rPrPs has been extensively studied (47, 48, 49, 50, 51). Overall, full-length rPrP 23–231 showed virtually identical biochemical properties compared with cellular PrPC (52, 53, 54). Refolding conditions are a critical factor in the formation of structure and level of solubility. In contrast to rmPrP with mainly
-helical structure after NaAc dialysis, in both tPrP preparations β-sheet prevailed, leading to the conclusion that the covalent linkage of two PrP monomers intrinsically forces the protein into β-sheet conformation, with possible similarities to PrPSc structure. Moreover, physiological salt conditions caused the formation of distinct soluble oligomers, also found by others (53, 54). The partial insolubility of tPrP(H2O) might be based on a salting-out effect while dialyzed against ddH2O.
Properties described under in vitro conditions do not necessarily reflect the in vivo situation. Importantly, immunization needs application of an immunogen to a living organism, thereby significantly changing the biochemical environment and eventually also biochemical properties of the immunogen. Interestingly, mixing of highly soluble tPrP(NaAc) with negatively charged adjuvant CpG resulted in an immediate and almost complete aggregation of protein. In contrast, less soluble tPrP(H2O) was less influenced by CpG contact. It is known that mouse rPrP has a high DNA-binding capacity and can form aggregates in the presence of high m.w. DNA (55, 56, 57). In contrast, contact to DNA ODN, as is basically CpG-ODN, resulted apparently in a dimerization of mPrP (58). As DNA binds to the structured C-terminal region of PrP, modification of protein structure is likely (59).
Although we could characterize the biochemical properties of tPrP preparations under defined in vitro situations (they seemed to differ rather in solubility than in structure), predictions concerning its immunogenicity in vivo are difficult.
Abs directed against PrPC or PrPSc are in principle able to inhibit de novo synthesis of PrPSc in vivo and in vitro (13, 16, 17, 25, 26). In this study, we confirmed the previously demonstrated induction of PrP-specific Abs in PrP wt mice after immunization with tPrP in combination with CpG adjuvant (26). Thereby, tPrP(NaAc) induced high Ab titers only in a few animals, whereas in immunizations with tPrP(H2O) most mice developed titers with variations in individual endpoints.
The Ab isotyping revealed the production of mainly IgG1, IgG2a, and IgG2b PrP-specific isotypes and did not show differences between immunization using tPrP(H2O) or tPrP(NaAc).
ELISA-based epitope mapping provides a versatile tool for investigating linear epitopes, although the exposure of covalently linked peptides does not necessarily reflect the respective amino acid side chains as exposed by native authentic proteins. Epitopes were measured individually for all ELISA titer-positive sera and showed weak reactivity near cutoff against some N-terminal epitopes, i.e., epitopes 11 and 12, and the outstanding epitope 10 (Fig. 2, C and D). We can exclude the induction of Abs reactive against linear epitopes a foreign origin as contained in our protein sequence, like the 3F4 tag, the poly-histidine tag, or the linker region of tPrP. Control serum A7 of a tPrP-immunized rabbit exhibited Abs specific to the complete panel of linear epitopes, thereby demonstrating the principal binding ability to the covalently linked peptides. In particular, the differentiation between epitopes 6 and 6b, corresponding to 3F4 epitope tag or PrP wt sequence, respectively, was clearly demonstrated with this positive control experiment.
The quantity of humoral anti-PrP response as usually tested in ELISA-based formats is not an appropriate method to determine the efficacy of Ab binding to native, authentic PrPC on living cells. To date, we failed to conclusively show specific binding of autoantibodies to authentic PrPC or PrPSc in immunoprecipitation assays using brain homogenates or cell lysates, although we could measure some reactivity of final sera as compared with respective preimmune sera in PrP surface FACS-staining approaches (data not shown). We could show specific binding of some sera to PrPC in immunoblot stripe blots with brain homogenate, even under denaturing conditions (data not shown). In contrast, the absence of detectable binding in these assays may not be a negative result, because concentration of Abs in murine blood might be too small or because Abs might be able to interfere in prion conversion by other mechanisms, e.g., by reacting with putative folding intermediates. In fact, the β-sheeted tandem version of PrP as used in this study is thought to mimic such putative folding intermediates. Importantly, in our hands, sera of immunized wt mice have shown significant interference in the de novo PrPSc formation in cell culture, whereas mPrP induced similar linear Ab reactivity, but mainly failed to prevent PrPSc formation (26). In a promising in vivo pilot vaccination approach using tPrP immunogen, two of eight mice survived longer than 600 days after otherwise deadly i.p. prion challenge (24). These results are presently under further investigation (G. Kaiser-Schulz, H. Schätzl, and M. Groschup, unpublished observations).
We propose the induction of conformational Abs by specific binding of B cells to mainly soluble, β-sheeted tPrP, followed by internalization, processing, and presentation of linear tPrP epitopes via MHC II. Epitope-specific CD4+ T cells are then able to trigger B cell proliferation and Ab production. We show the induction of a tPrP-specific T cell response in PrP wt mice using a syngenic full-length rPrP immunogen with only minor changes (e.g., two amino acid exchanges in the 3F4 tag). The significant IFN-
expression of CD4+ splenocytes after restimulation with full tPrP protein together with the absence of Th2-specific IL-4 and IL-10 (data not shown) indicate the induction of a Th1 response at the final stage of immunization, presumably polarized by the adjuvant CpG-ODN (38). The CpG-ODN-based general adjuvant effect may be further enhanced by the strong binding between CpG and tPrP, because it is known for covalently linked CpG proteins that promote a strong Th1 response via concomitant uptake of Ag and adjuvant (39, 40, 60). PrP-specific T cell responses in wt mice in experimental vaccination scenarios and the protective effect of cellular-based immunity in prion infection are in part characterized (19, 20, 22, 34, 35). Peptide immunizations with adjuvant CpG demonstrated two immunogenic PrP T cell epitopes (residues 143–172; 158–187) in wt mice (C57BL/6) (20). Souan et al. (22) successfully generated a T cell response in various mouse strains by designing peptides optimized to fit into the MHC class II-binding groove of NOD mice (aa 131–150; 211–230). β-sheet-rich rPrP, like the tPrP described in this study, is known to be partially proteinase K resistant, in contrast to easily degradable
-helical mPrP (48). Potentially, this biochemical characteristic of tPrP might lead to changes in the lysosomal degradation in APCs and thereby could lead to different MHC class II peptide loading, inducing T cells with possible cryptic PrP epitopes. At least in PrP0/0 mice, such variances in T cell response were discussed for
-helical and β-sheeted rPrP (33). Besides changes of wt epitopes, xenogenic epitopes, in our case most likely against the two amino acid exchanges of the 3F4 tag in the tPrP Ag, might help to break tolerance and enhance T cell response, as it was elegantly demonstrated for Syrian hamster PrP-immunized mice (35).
With our actual findings, we are now able to give statements for the obvious lack of functional conformational Abs in mPrP vaccinations seen in our previous work. The mainly
-helical folded monomer might present only conformational epitopes similar to PrPC, thereby being recognized as self. The explanation for differences between tPrP(H2O) and tPrP(NaAc) immunization might follow similar mechanistic correlations. Both protein preparations share β-sheet conformation, whereas tPrP(H2O) was partially soluble even after contact to adjuvant CpG. Ag presentation of B cells is essential for proper humoral response, and the formation of huge aggregates in tPrP(NaAc) preparations might decrease or hinder the uptake and processing by B cells.
In summary, these data together with other reported data provide solid experimental evidence that an effective humoral antiprion response can be evoked in certain experimental autoimmunization scenarios.
To increase CD4+ T cell help, we tested a vaccination protocol in which both tPrP and CpG-DNA are coencapsulated in biodegradable microspheres (42, 43). PLGA-MS are taken up and digested mainly by macrophages and DCs after s.c. application in vivo (61). The efficacy of loaded PLGA-MS seems to be based on the enhanced concurrent uptake of proteinaceous Ag and the adjuvant into endosomes of DCs and macrophages (62). Exogenous Ags are classically internalized and processed by the MHC class II-presenting pathway, resulting in APC-activating CD4 T cell (Th cell) response (63). In addition, DCs are able to direct exogenous Ag into the MHC class I presentation pathway via cross-presentation (64). CpG-ODN motifs are known to specifically activate innate immune cells, including macrophages and DCs via targeting TLR9. Interaction between CpG-ODN and TLR9 takes place in late endosomal compartments (65), and receptor-mediated endocytosis was reported to be essential for cross-priming (40). Via PLGA-MS immunization, CpG-ODN and exogenous Ag are effectively cotranslocated into endosomal compartments, thus enhancing TLR9 interaction and subsequent cross-presentation. This might explain the very pronounced increase in the response of CD4+ cells, in the range of 1% and higher, and, interestingly, also of a CD8+ T cell response. When PLGA-MS containing either tPrP or CpG-ODN alone are mixed, no detectable immune response at all is induced. One obvious explanation is that under these conditions no cointernalization of tPrP and CpG-ODN into the same APCs is taking place. The lack of Ab titers, despite the enhanced CD4 T cell response, might be due to the encapsulation of Ag, thereby preventing direct contact to B cells. We have to evaluate the potential lytic function of induced CD8 T cells, because they might cause severe autoimmune responses.
Inducing autoimmunity bears the obvious risk of induction of severe side effects. As was the case in anti-PrP immunization approaches reported by others (18, 19, 20, 21, 23, 24, 25, 26, 27, 29), we could not find such deleterious effects in our studies. In Alzheimer phase II vaccination studies, in some cases meningoencephalitis appeared (66). To exclude similar side effects in our approach, brain sections of mice were examined for signs of inflammation. All examined brains appeared normal without any severe changes as indicative for meningoencephalitis. Recently, it was reported that repeated daily administrations of high CpG-ODN doses have deleterious effects on the morphology and function of lymphoid organs (46). Such dramatic side effects can be excluded in our studies.
In summary, we describe in this work for the first time the induction of a substantial CD8 T cell response in PrP wt mice immunized with a rPrP of the same species. We need now to test whether application of coencapsulated Ag and CpG in PLGA-MS provides a new and very promising prophylactic approach against prion diseases in animals and humans.
| Acknowledgments |
|---|
| Disclosures |
|---|
|
|
|---|
| Footnotes |
|---|
1 This work was supported by SFB-576 (Project B12), SFB-596 (Projects A8 and B14), Bundesministerium fur Bildung, Wissenschaft, Forschung und Technologie (01KO0108), and the European Union Network of Excellence Neuroprion. ![]()
2 Address correspondence and reprint requests to Dr. Hermann M. Schätzl, Institute of Virology, Prion Research Group, Technical University of Munich, Trogerstr. 30, 81675 Munich, Germany. E-mail address: schaetzl{at}lrz.tum.de ![]()
3 Abbreviations used in this paper: CJD, Creutzfeldt-Jakob disease; AFM, atomic force microscopy; DC, dendritic cell; FTIR, Fourier transformation infrared spectroscopy; mPrP, monomeric prion protein; ODN, oligonucleotide; PLGA-MS, polylactide-coglycolide microspheres; PrP, prion protein; PrP0/0, knockout mice; PrPC, cellular PrP; PrPSc, abnormal isoformprion protein; tPrP, tandem PrP; wt, wild type; rec PrP, recombinant PrP. ![]()
Received for publication February 15, 2007. Accepted for publication June 25, 2007.
| References |
|---|
|
|
|---|
-helix to β-sheet transition in the recombinant prion protein. Biochemistry 40: 6982-6987. [Medline]This article has been cited by other articles:
![]() |
A. Sacquin, A. S. Bergot, P. Aucouturier, and M. Bruley-Rosset Contribution of Antibody and T Cell-Specific Responses to the Progression of 139A-Scrapie in C57BL/6 Mice Immunized with Prion Protein Peptides J. Immunol., July 1, 2008; 181(1): 768 - 775. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |