peptides 27 (2006) 2695–2705
available at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/peptides
A single prion protein peptide can elicit a panel of isoform specific monoclonal antibodies Tanja Vranac a, Katrina Pretnar Hartman a, Mara Popovic´ b, Anja Venturini a, Eva Zˇerovnik c, Vladka Cˇurin Sˇerbec a,* Blood Transfusion Centre of Slovenia, Sˇlajmerjeva 6, SI-1000 Ljubljana, Slovenia Medical Faculty, University of Ljubljana, Institute of Pathology, Korytkova 2, SI-1000 Ljubljana, Slovenia c Department of Biochemistry and Molecular Biology, Jozˇef Stefan Institute, Jamova 39, SI-1000 Ljubljana, Slovenia a
b
article info
abstract
Article history:
The main step in the pathogenesis of transmissible spongiform encephalopathies (TSE) is
Received 22 December 2005
the conformational change of the normal cellular prion protein (PrPC) into the abnormal
Received in revised form
isoform, named prion (PrPSc). Since PrP is a highly conserved protein, the production of
26 May 2006
monoclonal antibodies (mAbs) of high specificity and affinity to PrP is a difficult task. In the
Accepted 26 May 2006
present study we show that it is possible to overcome the unresponsiveness of the immune
Published on line 21 July 2006
system by immunizing wild-type BALB/c mice with a 13 amino acid PrP peptide from the Cterminal part of PrP, bound to the keyhole limpet hemocyanin (KLH). Immunization induced
Keywords:
predominantly anti-PrPSc humoral immune response. Furthermore, we were able to obtain a
Monoclonal antibody
panel of mAbs of IgG class specific for different non-self-conformations of PrP, with anti-
Peptide antigen
PrPSc-specific mAbs being the most abundant. # 2006 Elsevier Inc. All rights reserved.
Prion protein Epitope conformation Immune tolerance Anti-TSE vaccine
1.
Introduction
Although scrapie in sheep has been known for centuries, the importance of prion diseases only became widely recognized in the last decades with the epidemics of bovine spongiform encephalopathy (BSE) and the appearance of a new variant of Creutzfeldt–Jakob disease (vCJD) in humans. Despite significant efforts, neither the etiology nor immunology of prion diseases is well understood at the moment. No significant humoral or cellular immune response has been observed in patients with sporadic or acquired prion diseases [1,26]. B- and T-cell tolerance are the main obstacles for an effective immune response to the incorrectly folded self-protein or to PrP of other mammals: sequence similarity among them is
typically higher than 85%. Hence, the development of reliable immunodiagnostic tools as well as a potential anti-TSE vaccine is a major problem. With the aim of producing anti-PrP mAbs, different strategies have been used to overcome tolerance, however with limited success. Two anti-PrP mAbs, 3F4 and 6H4, are currently widely in use for diagnosing prion diseases in humans and cattle, respectively, by immunohistochemistry (IHC) and Western blot. They were prepared by immunizing mice with PrPSc rich extracts of infected hamsters’ brains (3F4) [16] and by immunizing Prnp0/0 mice with recombinant bovine PrP (6H4) [17]. Neither of the two mAbs is able to discriminate between the two isoforms of the protein. Consequently, for the specific detection of PrPSc, proteinase K (PK) digestion is an
* Corresponding author. Tel.: +386 1 5438 165; fax: +386 1 2302 224. E-mail address:
[email protected] (V. Cˇurin Sˇerbec). 0196-9781/$ – see front matter # 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.peptides.2006.05.026
2696
peptides 27 (2006) 2695–2705
essential step in the majority of immunological methods. PK digestion completely eliminates the predominantly a-helical PrPC, while the b-sheet rich PrPSc is degraded to a PK resistant fragment, PrPres, that can be detected immunologically. In theory, Abs of predetermined specificity can be obtained by immunization with peptides selected from the primary structure of the protein and bound to a carrier molecule [20]. Peptides of 13 amino acid residues contain sufficient structural information to induce the production of protein reactive Abs and are, at the same time, short enough to assume a variety of conformations [20]. Immunization with non-infectious PrP peptides provides a safe way for production of conformation specific mAbs useful for diagnostic purposes. These studies may also lead to the development of a vaccine against TSE. It is crucial to select peptides carefully, however, because not all epitopes are exposed in the native protein and, especially in the case of self-proteins, not all peptides are immunogenic [31,32]. In previous studies, peptides selected from the primary structure of PrP of different species were bound to carriers and used for immunization of mice, rabbits and chickens [12,21,29,36]. Only a few peptides provoked a strong polyclonal humoral immune response and not all sera reacted with PrP. Consequently, only a few high-affinity mAbs were obtained in these studies. None of them could discriminate between the two isoforms of the protein for different reasons: the predetermined epitope on the PrP may not change sufficiently during the transition of PrPC to PrPSc, the chosen peptides were unable to mimic these changes in the immunized animal, or the selection of mAbs was not optimal. We proposed that, by careful selection of the peptide, it should be possible to obtain mAbs against its different conformations. By immunizing wild-type mice expressing normal PrP, we expected B-cell tolerance to restrict the humoral immune response to non-self-conformations of the chosen PrP peptide. Our group has already obtained a powerful PrPSc-specific mAb V5B2 [6]. In this paper we report for the first time that, by immunization of BALB/c mice with a 13 amino acid peptide selected from the C-terminus of the human PrP sequence, it is possible to obtain mAbs of IgG class specific for different nonself-conformations of human PrP. PrPSc-specific mAbs were the most numerous among them, demonstrating that their appearance was the rule, rather than the exception. Our results also indicate that C-terminal part of PrP is very flexible and able to assume different conformations, either during the transition of PrPC to PrPSc or in response to the changes of chemical conditions in various immunological assays.
2.
Materials and methods
2.1.
Peptides and peptide conjugates
Three peptides were chosen from the primary structure of human PrP and synthesized by Bachem: P1 (amino acid residues 214–226): CITQYERESQAYY of Mw = 1653.8 and pI = 4.53, P2 (167–179): DEYSNQNNFVHDC of Mw = 1584.6 and pI = 4.02 and P3 (139–150, with the cysteine added Nterminally): CIHFGSDYEDRYY of Mw = 1667.8 and pI = 4.54.
Peptides were covalently bound to keyhole limpet hemocyanin (KLH) via the N- or C-terminal cysteine (Bachem) for the immunization. The rationale for the selection of the three peptides was the following: the peptides were chosen from the hydrophilic regions of PrP molecule, because only these parts are usually exposed in aqueous solutions. We also tried to select peptides in the way that they comprise natural cysteines at one terminus, however to the peptide P3, cysteine was added artificially. At the time when peptides were selected it was assumed that at the transition of PrPC to PrPSc the biggest conformational changes occur around the first a-helix [26] and this was the reason for the selection of peptide P3 in this region. Since the crystal structure of PrPSc has not been determined so far, little was (and still is) known about the differences in exposure of epitopes between PrPC and PrPSc. From this point of view, we found interesting the research paper of Kaneko et al. [15], claiming selective accessibility of residues 168, 172, 215 and 219 in PrPC, but not in PrPSc. Peptide P2 comprises the first two residues (168 and 172) and peptide P1 the last two (215 and 219). Apart from this, the entire sequence of the bovine version of P1, as well as parts of bovine versions of P2 and P3 were shown to compose a part of discontinuous epitope of a mAb 15B3, which was at that time the only PrPSc-specific mAb [17]. Two other peptides were synthesized in order to test antiP1-KLH mAbs specificity: P1m (the mouse version of the peptide P1) CVTQYQKESQAYY (Pepscan Systems) of Mw = 1610.7 and pI = 5.99 and P1s (the scrambled peptide P1) RQYIEYCSTEQYA (JPT Peptide Technologies).
2.2. Immunization and preparation of monoclonal antibodies Three groups of five BALB/c mice were injected subcutaneously on day 0 with 0.2 mg of each of the three peptides, bound to KLH (P–KLH) per mouse, in Freund’s complete adjuvant (0.2 ml/mouse). On days 14 and 28, the mice were injected intraperitoneally with 0.1 mg P–KLH per mouse in Freund’s incomplete adjuvant (IFA) (0.2 ml/mouse). Blood was taken from the tail vein 10 days after the last inoculation. Antibodies against KLH, P–KLH and peptide alone were detected in sera by indirect ELISA. A final booster dose of P– KLH was injected on day 45 intravenously in physiological saline (0.1 mg/mouse in 0.1 ml) to mice with the highest titers of mAbs against each of the peptides. Mice were sacrificed on day 48 and their spleens removed. Splenocytes were isolated and fused with mouse NS1 myeloma cells using 50% PEG for 3 min, according to standard techniques. Cells were washed and resuspended in 96-well microtiter plates in DMEM (Dulbecco’s modification of Eagle’s medium, ICN Biomedical) supplemented with 13% bovine serum (Hy Clone) (subsequently designated DMEM) and with feeder cells of mouse thymocytes. The next day, DMEM supplemented with hypoxanthine-aminopterine-thymidine (HAT, Sigma) mixture was added to all the wells. The presence of specific antibodies was determined in the supernatants after 10–14 days by indirect ELISA. Hybridomas from positive wells were transferred into larger volumes of HAT DMEM and the specificity of antibodies was determined by immunohistochemistry and dot blot.
peptides 27 (2006) 2695–2705
Selected hybridomas were cloned in DMEM by the limiting dilution method and frozen in liquid nitrogen. Three mice, immunized with P1-KLH that were not sacrificed for the first fusion, were injected again intraperitoneally on day 60 with 0.1 mg P1–KLH per mouse in IFA (0.2 ml/mouse) and boosted on day 70 with an intravenous injection of P1–KLH in physiological saline (0.1 mg/mouse in 0.1 ml). Two mice were sacrificed on day 73 and the second fusion was performed in the same way as the first.
2.3.
Indirect ELISA
Microtiter plates (Nunc) were coated with 1 mg/ml of peptide, 5 mg/ml of peptide-KLH (Bachem) or KLH (Bachem, Sigma) and incubated overnight at 4 8C in carbonate-bicarbonate buffer, pH 9.6. The next day plates were washed three times in sodium phosphate buffer, containing 150 mM NaCl and 0.05% Tween 20, pH 7.2–7.4. They were subsequently incubated for 30 min at 37 8C with blocking buffer (1% BSA in washing buffer). The plates were washed again and then incubated for 1.5 h at 37 8C with polyclonal sera or mAbs diluted in blocking buffer. After washing, plates were incubated with goat anti-mouse Abs conjugated to HRP (Jackson ImmunoResearch) diluted 1:5000 in the blocking buffer, and incubated again for 1.5 h at 37 8C. Plates were washed and incubated with ABTS substrate (Sigma) diluted in citrate-phosphate buffer, pH 4.5. Plates were incubated for 15 min at 37 8C and the color was quantitated at 405 nm.
2.4.
Purification of antibodies
Monoclonal antibodies were purified from the cell culture supernatants by affinity chromatography on Protein GSepharose (Sigma), using 0.1 M glycine, pH 2.7, for elution.
2.5.
Brain tissue sample preparation
Brain tissue of a deceased sporadic CJD (sCJD) patient was homogenized (HT1000 Potter homogenizer) in ice-cold PBS containing 0.5% Nonident P-40 and 0.5% deoxycholate. Homogenates were centrifuged for 5 min at 5000 g at +4 8C. The supernatant of brain homogenate was aliquoted and stored at 80 8C. The samples were used as prepared or digested with proteinase K (50 mg/ml, 30 min at 37 8C).
2.6.
Dot blot analysis
The 5 diluted supernatants of brain homogenate were loaded on 0.2 mm nitrocellulose membrane (BioRad) using dot blot (BioRad). The membranes were blocked with 5% (w/v) non-fat dried milk (BioRad) in TTBS (Tris-buffered saline with 0.1% (v/v) Tween 20) overnight at +4 8C. The next day, membranes were incubated with primary mAbs (5 mg/ml in 1% milk in TTBS) for 1.5 h with shaking at room temperature. After washing in TTBS, membranes were incubated with HRP-labelled anti-mouse secondary antibody (Amersham Biosciences) diluted 1:1500 in 1% milk in TTBS for 1.5 h, with shaking at room temperature. Membranes were washed in TTBS and the immune reaction was detected using a chemiluminescence detection kit (ECL, Amersham Biosciences).
2.7.
2697
Western blot analysis
The fusion protein of the peptide P1 and ketosteroid isomerase (KSI) was prepared as described [6]. For Western blot analyses 3 mg of P1-KSI or KSI were loaded per lane in SDS-PAGE sample buffer (Novex), containing 10% b-mercaptoethanol. For Western blot analyses of mAb reactivity with brain tissue PrP, 5 diluted supernatants of brain homogenate were denatured with SDS-PAGE sample buffer (Novex). Proteins were separated by SDS-PAGE and then blotted on nitrocellulose membranes (BioRad) as described [6]. PrP was detected with mAbs as described for dot blot analysis.
2.8.
Immunohistochemistry (IHC)
Brain of the sCJD case was fixed in 4% buffered paraformaldehyde for two weeks. Samples of the brain were treated with 96% formic acid for 1 h before dehydration in increasing concentrations of ethanol and embedding in paraffin. The 5 mm thick sections were mounted on three DAKO ChemMateTM capillary gap microscope slides. After overnight drying at 60 8C the sections were cooled to room temperature and deparaffinized in decreasing concentrations of ethanol and xylene for 15 min, followed by two changes of absolute ethanol, washed under tap water for 3 min and rinsed in distilled water. For antigen retrieval the sections were microwaved at 95 8C in EDTA buffer (pH 8) for 10 min. Immunohistochemistry was performed in the DAKO TechMateTM 500/1000 immunostainer at room temperature. The sections were incubated for 25 min with primary mAbs (4 mg/ml) or 20–200 diluted mouse polyclonal serum and rinsed three times in DAKO ChemMateTM buffer. The procedure was continued for 25 min with DAKO ChemMateTM biotinylated secondary antibodies in buffered solution, the sections then being rinsed and incubated with endogenous peroxidase blocking solution for 7.5 min and with streptavidin-peroxidase in buffered solution for 25 min. The immunoreaction was visualized with 30 3-diaminobenzidine included in DAKO ChemMateTM detection kit. The slides were counterstained with DAKO ChemMateTM hematoxylin for 15 s. Slides were rinsed in water, dehydrated in three changes of ethanol, cleared in xylene and coverslipped in RCM 2000 coverslipper (Medite).
2.9.
Flow cytometry
Human peripheral blood mononuclear leucocytes (PBML) were isolated from freshly collected blood obtained by venipuncture from healthy adult volunteers with MM, MV and VV polymorphisms at position 129 of the PrP. Mononuclear leucocyte layer was collected by gradient centrifugation of buffy coat on Histopaque 1077 (Hybri-Max, Sigma); the cells were washed in PBS three times, then resuspended in ice-cold staining solution (2% FCS, 2% Na-azide in PBS). For each experiment 3 105 cells were first stained with primary antibody (20 mg/ml) for 30 min at +4 8C, then washed three times and incubated in the dark with secondary antibody (GAM-FITC, Becton Dickinson) for another 30 min at +4 8C. After staining, cells were again washed. PBMLs were then resuspended in 300 ml of the staining solution and immediately detected on a flow cytometer (Becton Dickinson). Forward and side scatter criteria were used to gate and define
2698
peptides 27 (2006) 2695–2705
the lymphocyte population. mAb 3F4 (Senetec) was used as a positive control at the concentration of 2 mg/ml.
2.10.
Determination of minimal epitopes by SPOT
SPOT synthesis has been described extensively by Frank [9]. In our case, six octamer peptides, shifted by one residue in such a way as to cover the whole P1 sequence, were synthesized on a cellulose membrane using Fmoc chemistry, provided in the SPOT synthesis kit (Sigma Genosys), following the manufacturer’s instructions precisely. As a positive control, the whole P1 peptide, as well as the mouse and bovine versions of P1, were also synthesized.
2.11.
CD spectroscopy
CD spectra were measured using an Aviv model 62ADS CD spectropolarimeter (Aviv Associates). The 0.1- and 1-cm cuvettes at bandwidths of 1 and 0.5 nm were used to record far- and near-UV CD spectra, respectively. Temperature was maintained at 25 8C throughout. Data were collected every 1 nm in the far- and every 0.5 nm in the near-UV. Protein concentrations for the near UV CD were 1.2 mg/ml for the antibodies and 0.6 mg/ml for the free peptide. They were diluted by half when forming the complex, i.e. 0.6 mg/ml for the antibody and 0.3 mg/ml for the peptide. For the far UV CD, all concentrations were reduced by two-fold. To obtain the difference spectra, from experimental CD spectrum of the complex (multiplied by 2 due to the effect of dilution) a sum of individual CD spectra of the respective antibody and the peptide was deduced. CD spectra were expressed in the usual units. The measured values of ellipticity in mdeg were transformed to mean residue ellipticity [Q]MRW (8 cm2/dmol) based on the equation: ½QMRW ¼
QMMRW 10cl
MMRW of 112 was taken and the amount of the complex was equated to the molarity of the antibody.
2.12. BLAST search of sequences matching hexameric parts of P1, P2 and P3 To further prove the sequence specificity of the selected peptides compared to other human proteins, we performed BLAST search of sequences matching hexameric parts of P1, P2 and P3. Minimal epitopes of antibodies are thought to comprise at least six amino acid residues [25], therefore for each peptide eight hexameric parts shifted by one residue were submitted to BLAST search. The settings of the program were as follows: BLASTp; short, nearly exact matches; organism: Homo sapiens.
3.
Results
3.1.
Polyclonal responses to peptides
The highest anti-peptide titers were observed in sera from mice immunized with P1-KLH, with values of 1:107 in some
Fig. 1 – Humoral immune responses of BALB/c mice, immunized with conjugates P1-KLH, P2-KLH and P3-KLH. Titers to the peptide, conjugate and KLH alone were determined by indirect ELISA.
animals (Fig. 1). Polyclonal sera were used for IHC on CJD brain tissue samples. Prominent staining of PrPSc plaques was only noted with anti-P1-KLH serum (Fig. 2).
3.2.
A panel of conformational mAbs against P1
From mice immunized with P1-KLH, two separate cell fusions of mouse splenocytes with NS1 myeloma cells were performed. Hybridomas, producing antibodies that reacted in ELISA with P1 and P1-KLH, but not with KLH, were selected for cloning. A panel of anti-PrP mAbs was isolated by further selection. They were classified into three groups according to their specificity towards the two PrP isoforms, as determined by IHC, although each of the selected mAbs had its own specific characteristics (Table 1). One mAb was chosen from each group for further analysis: C7/5, PrPC-specific mAb, C1/1, which reacted with both isoforms of PrP, and E9/5, mAb that was PrPSc-specific. The most potent PrPSc-specific mAb V5B2, obtained by a separate fusion, has been characterized in detail in Ref. [6]. From mice immunized with P2-KLH and P3-KLH no potent mAbs were isolated after the fusions.
3.3.
Characterization of selected mAbs
The IHC reaction of the three mAbs with sCJD brain tissue sections is shown in Fig. 3. C7/5 did not label PrPSc aggregates, in the presented case kuru plaques, but labelled the surrounding brain tissue. C1/1 reacted with kuru plaques as well as with surrounding brain tissue, while E9/5 labelled the plaques exclusively, without any PK pre-treatment of the brain section. C7/5 and C1/1 also labelled non-CJD brain tissue (Fig. 3). Western blot analysis of the three mAbs demonstrated their reactivity with monomeric P1, expressed as a fusion protein with KSI, but not with KSI alone (Fig. 4). It was shown however, that PrPSc-specific mAbs from our panel (as E9/5 and V5B2, for example) demonstrate higher affinity to di- or multimeric forms of P1 compared to monomeric P1 [30 and unpublished data]. On Western blots of CJD and non-CJD brain homogenates, the mAbs gave weak or no reaction, indicating the sensitivity of the C-terminal epitope of PrP to SDS
2699
peptides 27 (2006) 2695–2705
Fig. 2 – The IHC staining with murine polyclonal antisera: sCJD brain tissue (CJD+) and normal brain tissue (CJDS) stained with murine polyclonal sera anti-P1-KLH (diluted 200), anti-P2-KLH (diluted 20) and anti-P3-KLH (diluted 20). Arrows indicate PrPSc aggregates, in the presented case kuru plaques.
denaturation. Nevertheless, the results confirmed the reactivity of C7/5 with PrPC, predominantly with the diglycosylated form. C1/1, that labelled both isoforms on IHC, reacted with PrPC as well as with the PK resistant part of PrPSc (PrPres) on Western blot (Fig. 5A). Interestingly, neither of the two mAbs reacted with native PrPC of non-CJD brain on dot blot (Table 2). The PrPSc-specific mAb E9/5 did not react with PrP on Western blot. Apart from the improper renaturation after exposure to SDS, the reason for this might be dissociation of PrPSc aggregates to which PrPSc-specific mAbs probably bind with
the highest affinity [30]. Reaction with native, undissociated PrPSc as well as with PrPres was confirmed by dot blot of brain homogenates (Fig. 5B). In addition, none of the mAbs obtained by immunizing BALB/c mice with P1-KLH reacted with native PrPC on the surface of lymphocytes from individuals polymorphic at the position 129 of the PrP, as determined by flow cytometry analyses. No differences between the genotypes analyzed were observed, so only results with the MV genotype, which is prevalent in the Slovene population [10] are shown (Fig. 6).
Table 1 – A panel of isoform specific anti-PrP mAbs, obtained in two separate cell fusions Antibody
Type
P1 (ELISA)
P1m (ELISA)
P1-KLH (ELISA)
1. Fusion
K1/F11 K4/B3 K4/H5 K5/D5 V5/B2
M G1 G1 M G1
+ +++ ++ ++ +++
+ +++ ++ ++ +++
+ +++ ++ ++ +++
Sc Sc Sc C/Sc Sc
2. Fusion
A3/4 B8/1 B11/1 C1/1 C7/2 C7/5 D9/4 E9/5 G12/1
G3 G1 G1 G1 G3 G1 G1 G2a G1
+ +++ ++ +++ ++ + ++ +++ ++
+++ ++ +++ + + ++ + ++
+ +++ ++ +++ ++ +++ ++ +++ ++
Sc Sc Sc C/Sc Sc C Sc Sc C/Sc
" " "
P1s (ELISA)
P2 (ELISA)
P3 (ELISA)
PrP isoform specificity (IHCa)
Reactivities with peptide P1, mouse version of P1 (P1m), P1-KLH, scrambled P1 (P1s), P2 and P3 were detected by indirect ELISA: (+) represents hybridoma culture supernatant’s titer between 1:10 and 1:102; (++) titer of 1:103 to 1:104 and (+++) titer of 1:105 to 1:106. a IHC: immunohistochemistry, performed on CJD and non-CJD human brain tissue samples was used to determine specificity for PrP isoforms. The arrows accentuate the characteristics of the three selected mAbs.
2700
peptides 27 (2006) 2695–2705
Table 2 – Reactivity of mAbs with native or denatured PrP isoforms PrP C Native (dot blot, FC) C7/5 C1/1 E9/5
PrP Sc Denatured (IHC, WBa) + +
Native (dot blot)
Denatured (IHC, WBa)
+
+ + (in IHC only)
Pre-treatment procedures required in various immunological assays cause different expositions of C-terminal epitopes. FC, flow cytometry; IHC, immunohistochemistry; WB, Western blot. a The reactions with PrP in Western blot were weak, requiring prolonged exposition times.
3.4.
Determination of minimal epitopes
The SPOT method was used to determine minimal epitopes of the selected mAbs. C1/1 was the only antibody for which a minimal epitope could be determined. This was ERESQA(Y), in which first E and R can be replaced by Q and K in mouse PrP. The first C-terminal tyrosine significantly improved affinity of the Ab towards the epitope, while the second tyrosine appeared not to be important for binding. The other two antibodies were not shown to react by this method, not even with whole length SPOT-synthesized P1 (Fig. 7). No reaction was obtained when using only secondary antibodies. The presence of peptides on the cellulose membranes tested with E9/5 and C7/5 was confirmed by re-probing them with C1/1. The reactivity of E9/5 and C7/5 in the primary antibody solutions, used for testing SPOT membranes, was confirmed by testing dot-blotted P1 and/or P1-KLH in the same antibody solutions (not shown).
Fig. 4 – mAbs reactivity with monomeric P1. Peptide P1 was expressed C-terminally of ketosteroid isomerase (KSI) in the form of fusion protein. Western blotting performed with C7/5, C1/1 and E9/5 demonstrated the reactivity of all three mAbs with peptide P1 monomer.
3.5.
CD spectroscopy
In Fig. 8 difference CD spectra obtained upon formation of the complex between each of the three mAbs and the peptid P1 are shown. In the far UV there is no significant change down to 205 nm, what shows that secondary structure of the peptide does not change significantly after binding to a mAb. P1 probably binds in its initial conformation as in solution, which at pH 7.0 is extended and most likely dimeric due to the presence of the N-terminal cysteine. However, in the near UV there are significant changes in the case of mAb C1/1 and mAb E9/5, confirming strong interactions of these two mAbs with
Fig. 3 – The IHC staining with the three selected mAbs: sCJD brain tissue (CJD+) and normal brain tissue (CJDS) stained with the mAbs C7/5, C1/1, E9/5. Arrows indicate PrPSc aggregates, in the presented case kuru plaques.
peptides 27 (2006) 2695–2705
2701
Fig. 5 – (A) Western blot of human brain tissue homogenates, performed with mAbs C7/5 (10 mg/ml) and C1/1 (10 mg/ml), compared to 3F4 (0.2 mg/ml). (B) Dot blot of human brain tissue homogenates, performed with the mAb E9/5 (5 mg/ml) that did not react in Western blot assay, compared to 3F4 (0.2 mg/ml). PK+ indicates digestion of tissue with proteinase K.
the peptide. The negative peaks at wavelengths 275–280 nm could be ascribed to the interactions involving tyrosine residues. Much less interaction is seen by mAb C7/5, which is in accordance with the fact that C7/5 binds poorly to the peptide alone. A mAb with the epitope located near the first ahelix of PrP was used as a negative control producing no difference spectra.
3.6. BLAST search of sequences matching hexameric parts of P1, P2 and P3 The results of BLAST search demonstrate that there are a few matches for some hexapeptides from P1 and two matches for a hexapeptide from P3 (Table 3). All the matches for the hexapeptide TQYERE from P1, belong to members of a family of scaffolding proteins, responsible for manoeuvring MAP
Fig. 6 – mAbs reactivity with native PrPC, present on human lymphocytes. Flow cytometry was performed with mAbs C7/5, C1/1 and E9/5 (20 mg/ml). mAb 3F4 (2 mg/ml) was used as a positive control.
Fig. 7 – Mapping of minimal epitopes of mAbs C7/5, C1/1 and E9/5. On cellulose membrane overlapping hexapeptides from P1 were synthesized (spots 1–6 from left), as well as the whole P1 (spot 7), the corresponding mouse P1–P1m (spot 8), bovine P1–P1b (spot 9) and P1, in which four central residues were exchanged with alanines (spot 10).
kinases signalling pathways [BLAST at NCBI]. They are ubiquitously expressed [19], so there is no possibility for their unique localization in PrPSc plaques. Furthermore, SPOT analyses show that none of the three mAbs binds to the sequence TQYERE, therefore cross-reaction with MAP kinase scaffolding proteins with mAbs presented in this paper seems unlikely. The results obtained by SPOT indicate that the mAb C1/1 could hypothetically cross-react with a yet unisolated human protein Secernin 2, based on its cDNA sequence. However, the function of this protein, its tissue distribution and even whether the protein is actually expressed is not yet known [34].
2702
peptides 27 (2006) 2695–2705
Fig. 8 – Difference CD spectra obtained upon complex formation between the peptide P1 and each of the mAbs in the far and near UV.
4.
Discussion
In the search for efficient diagnostic tools and possible immunotherapy for TSE, the breaking of tolerance against the self-protein PrP, poses a significant problem. In recent years, Prnp0/0 mice have been predominantly used for mAb production, with recombinant PrP as an antigen. Their lack of PrP, and
hence the absence of immune tolerance to the protein, makes this task much easier. However, for studying the immune response with possible vaccine development in mind, this is obviously not the method of choice. By immunizing mice with a peptide selected from the Cterminus of human PrP, we were able to obtain a panel of unique mAbs that are specific for different isoforms of PrP, but
Table 3 – Human proteins matching hexameric parts of P1, P2 and P3 Hexapeptide parts, shifted by one residue P1
Number of matches (besides PrP) with 100% sequence similarity
CITQYE ITQYER
0 1
TQYERE
3
QYERES YERESQ ERESQA RESQAY ESQAYY
0 0 0 1 0
P2
DEYSNQ EYSNQN YSNQNN SNQNNF NQNNFV QNNFVH NNFVHD NTVHDC
0 0 0 0 0 0 0 0
P3
CIHFGS IHFGSD HFGSDY FGSDYE GSDYED
0 0 0 0 2
SDYEDR DYEDRY YEDRYY
0 0 0
Protein
Sperm associated antigen 9 JNK-associated leucine-zipper protein; MAP kinase 8 interacting protein 3; sperm associated antigen 9
Function/localization
Cell signalling in spermatogenesis/ testicles Scaffolding protein family in MAP kinases signalling pathways/many cell types
Secernin 2
Unknown
Adenylate kinase 5; Adenylate kinase 6
Metabolism of adenine nucleotides/ brain, glands, other?
BLAST search program was used to identify human proteins that share potential hexapeptide epitopes with peptides, used for immunization.
peptides 27 (2006) 2695–2705
with the predominance of PrPSc-specific mAbs. This supports our hypothesis that, when immunizing wild-type animals, immune tolerance would restrict B-cell response to non-selfconformations of an antigen. The majority of mAbs obtained showed equal affinity to human P1 and the corresponding mouse P1 (P1m) (Table 1), demonstrating that the basis for their specificity is the conformation of the epitope and not the differences in amino acid sequences between mouse and human PrP. The characterization of these conformational mAbs has proved to be difficult due to the high sensitivity of their paratopes to slight conformational changes of the Cterminal PrP region, comprising the P1 sequence, under various chemical conditions, which is a logical consequence of the conformational specificity of mAbs. Conformational and/or steric sensitivity of paratopes was confirmed also by SPOT epitope analyses of the three selected mAbs. C1/1, being the least sensitive to SDS denaturation, is the only one that reacts with the epitope, represented by peptides on SPOT membrane. The other two antibodies did not react with them. It is likely that the epitope of PrPSc-specific mAb E9/5, which reacts well with P1 on dot blot and ELISA (Fig. 5 and Table 1) but not on SPOT (Fig. 7), resides in the very C-terminal part of the peptide, comprising the tyrosine pair at positions 225 and 226 of human PrP. This part of the peptide is less flexible and less accessible, due to the C-terminal binding of peptides to the SPOT membrane. In contrast, the C-terminus was exposed in the conjugate used for immunization, in which P1 was bound to KLH by N-terminal cysteine. It was shown by Paramithiotis et al. [23] that epitopes rich in tyrosine are specifically exposed in bsheet rich PrP, although more importance was given to tyrosine pairs at positions 149–150 and 162–163 of human PrP. Also the epitope of the PrPC-specific mAb C7/5 is conformationally very sensitive. C7/5 does not react with P1 alone, but reacts strongly with P1 when bound to KLH (Table 1), when expressed as a fusion protein with KSI (Fig. 4) or as a part of PrPC (Figs. 3 and 5). It appears likely that P1 alone very rarely assumes the conformation recognized by C7/5, but that the carrier molecule stabilizes the epitope in the conformation that is also present in denatured PrPC. CD spectroscopy analyses (Fig. 8) confirmed that tyrosines are likely to be involved in the interaction between P1 and mAbs C1/1 and E9/5. However, due to a significant number of tyrosine residues in hypervariable regions of both mAbs, observed after sequencing of these regions (manuscript in preparation), it is difficult to say whether tyrosines from the peptide or from the antibody (or both) are responsible for the observed change in the near CD spectra. The final interpretation of obtained difference CD spectra will only be possible after solving the crystal structure of mAbs’ Fab-P1 complex, which is on the way. Antibodies directed against the P1 conformation resembling that in native PrPC could cause autoimmune response. However, further analyses proved that none of the obtained mAbs reacted either with recombinant human PrP in ELISA (results not shown) or with native PrPC, as determined by dot blot of non-CJD brain homogenates and flow cytometry on non-CJD lymphocytes (Fig. 6). This again confirms our expectation that, in wild-type BALB/c mice, only mAbs specific for the non-selfconformations of the peptide P1 can be obtained. However, two mAbs, C7/5 and C1/1, reacted with PrPC in immunological assays that involve denaturation. It is impor-
2703
tant to stress that the conformation of PrPC after exposure to heat and acid in IHC or SDS in Western blot, is most probably different from that in untreated tissue. It has been proved that a b-sheet rich and very stable intermediate of recombinant PrP can be obtained by exposing the recombinant prion protein to low pH [13,28] or to temperatures above 40 8C [5]. A closer look at the protein domains involved in the conformational switch at low pH confirmed moderate structural changes in the Cterminal a-helix, in which the P1 epitope resides [3,11]. It is not clear, however, if those studies can explain the conformational changes of the protein in vivo, especially in its Cterminus which is bound to the cell membrane by a glycosylphosphatidylinositol (GPI) anchor. Our results confirm that the C-terminal part of human PrP is very flexible when exposed to the denaturing conditions in different immunological assays and that it undergoes significant conformational changes during the transition of PrPC to PrPSc, as predicted also in a recent study by Dima and Thirumalai [7]. In spite of clear IHC results, weak or no reaction of mAbs in Western blot and their unreactivity with recombinant PrP might lead to the critical question: are this mAbs really specific for PrP? We performed extensive BLAST search of possible matches for hexameric parts of the peptide P1 (Table 3) that left very little space for a negative answer. Apart from this, it should be emphasized that the unreactivity of mAbs with the native PrPC itself is a proof of the conformational specificity of mAbs. Immunization of wild-type mice with conjugate P1-KLH, but not with P2-KLH and P3-KLH, led to an exceptionally good polyclonal response in BALB/c mice, reaching titers up to 1:107 (Fig. 1). High immunogenicity of the C-terminal part of PrP from different mammals has also been reported by other authors [12,36]. The flexibility demonstrated here for this region could be one of the reasons for our results. It has already been proposed by Van Regenmortel that the peptide, as well as the corresponding protein epitope, needs to possess some mobility in order to adapt to and cross-react with preexisting antibody [31]. The ability of peptides to stimulate helper T-cells might also play a role in provoking high polyclonal titers. It was shown by Souan and co-workers that peptide p211–230, comprising amino acid residues 211–230 from mouse PrP (and containing the entire P1m!) was extremely efficient in T-cell stimulation in different mouse strains, indicating that this peptide possesses a sequence that is recognized by different MHC molecules. They even reported that immunization of mice with p211–230 caused a reduction of PrPSc levels in neuroblastoma cell tumors that were induced in immunized mice. PrPC levels remained unaltered [27]. It is possible that in their experiments Abs to non-self-conformations were induced, similarly to the case we describe, and involved in the PrPSc clearance process. The aptness of anti-PrPC mAbs to prevent and reverse the formation of PrPSc deposits has been shown in vitro [8,24] and in vivo [35]. Although no obvious side effects were reported in treated animals, it is not yet clear what consequences such an autoimmune therapeutic would provoke in humans in the long term. With a vaccine that promotes anti PrPSc-specific mAbs formation this issue could be overcome. The first evidence that the binding of Abs to PrPSc (although not exclusively) might be crucial in inhibiting prion replication has already been reported [2].
2704
peptides 27 (2006) 2695–2705
The ability of peptide P1 to promote a specific B, and possibly also T [27] cell response is an important characteristic necessary for peptide vaccine development. According to our results and to other studies that included several other PrP peptides [12,27] there are only a few, if any, parts of PrP with the same attributes. There are evidences that the interaction of overlapping Cterminal epitopes with antibodies [24,27] or other molecules, e.g. quinacrine [33], can slow down PrPSc formation. It has also been reported recently that PrPSc-specific anti-(Tyr-Tyr-Arg) mAbs, possibly interacting also with the C-terminal tyrosine pair at positions 225 and 226 of PrP, could reduce PrPSc levels in infected neuroblastoma cell cultures [4]. Besides, the importance of isoform specific mAbs in diagnostics is obvious. With the emerging reports that only moderate changes in chemical conditions can alter the degree of PK cleavage significantly [18,22], the use of PK in routine tests is becoming questionable. PrPSc-specific mAbs from the panel were shown to be especially useful for the diagnosing of CJD as well as BSE by IHC [6,14]. Apart from this, a sandwich ELISA test has been successfully developed with the mAb V5B2 for diagnosing CJD [6]. Sandwich ELISA, based on the same mAb, has also been used to develop the BSE test - PrionType post-mortem that has been in the last validation process by European Commission. The other mAbs from the panel were not subjected to test development so far.
5.
[3]
[4] [5]
[6]
[7]
[8]
[9]
[10]
Conclusion
The relative conformational freedom of peptides comprising part of the primary structure of self-proteins can be used successfully for producing mAbs against non-self-conformations of these proteins, as postulated on the basis of immune tolerance. We exploited this feature to produce a panel of PrPSc-specific mAbs, as well as mAbs specific for denatured PrPC. These mAbs could be very useful for TSE diagnostics and possibly also for the treatment of the disease. Our results indicate that peptide vaccines should not be overlooked for prevention of diseases caused by protein misfolding.
[11]
[12]
[13]
[14]
Acknowledgements We thank Iva Hafner-Bratkovicˇ for fusion protein isolate, Marjana Sˇprohar for technical assistance and Dr. Bozˇidar Voljcˇ for his support. The Ministry of Education, Science and Sport of the Republic of Slovenia supported this work financially with ˇ Sˇ). The Ministry of Grants L3-3435 and L3-6006 (both to VC Education, Science and Sport of the Republic of Slovenia also financed the PhD studies of TV and AV.
[15]
[16]
[17]
references [18]
[1] Aucouturier P, Carp RI, Carnaud C, Wisniewski T. Prion diseases and the immune system. Clin Immunol 2000;96:79–85. [2] Beringue V, Vilette D, Mallinson G, Archer F, Kaisar M, Tayebi M, et al. PrPSc binding antibodies are potent
[19]
inhibitors of prion replication in cell lines. J Biol Chem 2004;279(38):39671–6. Calzolai L, Zahn R. Influence of pH on NMR structure and stability of the human prion protein globular domain. J Biol Chem 2003;278(37):35592–6. Cashman NR, Caughey B. Prion diseases—close to effective therapy? Nat Rev Drug Discov 2004;3(10):874–84. Cordeiro Y, Kraineva J, Ravindra R, Lima LM, Gomes MP, Foguel D, et al. Hydration and packing effects on prion folding and beta-sheet conversion. High pressure spectroscopy and pressure perturbation calorimetry studies. J Biol Chem 2004;279(31):32354–9. Cˇurin Sˇerbec V, Bresjanac M, Popovic´ M, Pretnar Hartman K, Galvani V, Rupreht R, et al. Monoclonal antibody against a peptide of human prion protein discriminates between Creutzfeldt–Jakob disease-affected and normal brain tissue. J Biol Chem 2004;279(5): 3694–8. Dima RI, Thirumalai D. Probing the instabilities in the dynamics of helical fragments from mouse PrPC. Proc Natl Acad Sci USA 2004;101(43):15335–40. Enari M, Flechsig E, Weissmann C. Scrapie prion protein accumulation by scrapie-infected neuroblastoma cells abrogated by exposure to a prion protein antibody. Proc Natl Acad Sci USA 2001;98(16):9295–9. Frank R. Spot-synthesis: an easy technique for the positionally addressable, parallel chemical synthesis on a membrane support. Tetrahedron 1992;48(22):9217–32. Galvani V, Rupreht RR, Cˇurin Sˇerbec V, Vidan-Jeras B. Genetic risk factors associated with Creutzfeld–Jakob disease in Slovenians and a rapid typing for PRNP codon 129 single nucleotide polymorphism. Transfus Med 2005;15(3):197–207. Gu W, Wang T, Zhu J, Shi Y, Liu H. Molecular dynamics simulation of the unfolding of the human prion protein domain under low pH and high temperature conditions. Biophys Chem 2003;104(1):79–94. Harmeyer S, Pfaff E, Groschup MH. Synthetic peptide vaccines yield monoclonal antibodies to cellular and pathological prion proteins of ruminants. J Gen Virol 1998;79:937–45. Hornemann S, Glockshuber R. A scrapie-like unfolding intermediate of the prion protein domain PrP(121–231) induced by acidic pH. Proc Natl Acad Sci USA 1998;95(11):6010–4. Juntes P, Zabavnik Piano J, Cˇurin Sˇerbec V, Kovacˇ Z, Pogacˇnik M. Neuropathological lesions in the first case of bovine spongiform encephalopathy in Slovenia. Slo Vet Res 2002;39(2):105–14 [English edition]. Kaneko K, Zulianello L, Scott M, Cooper CM, Wallace AC, James TL, et al. Evidence for protein X binding to a discontinuous epitope on the cellular prion protein during scrapie prion propagation. Proc Natl Acad Sci USA 1997;94(19):10069–74. Kascsak RJ, Rubenstein R, Merz PA, Tonna-De Masi M, Fersko R, Carp RI, et al. Mouse polyclonal and monoclonal antibody to scrapie-associated fibril proteins. J Virol 1987;61:3688–93. Korth C, Stierli B, Streit P, Moser M, Schaller O, Fischer R, et al. Prion (PrPSc)-specific epitope defined by a monoclonal antibody. Nature 1997;390:74–7. Kuczius T, Buschmann A, Zhang W, Karch H, Becker K, Peters G, et al. Cellular prion protein acquires resistance to proteolytic degradation following copper ion binding. Biol Chem 2004;385(8):739–47. Lee CM, Onesime D, Reddy CD, Dhanasekaran N, Reddy EP. JLP: a scaffolding protein that tethers JNK/p38MAPK signaling modules and transcription factors. Proc Natl Acad Sci USA 2002;99(22):14189–94.
peptides 27 (2006) 2695–2705
[20] Lerner RA. Antibodies of predetermined specificity in biology and medicine. Adv Immunol 1984;36:1–44. [21] Matsushita K, Horiuchi H, Furusawa S, Horiuchi M, Shinagawa M, Matsuda H. Chicken monoclonal antibodies against synthetic bovine prion protein peptide. J Vet Med Sci 1998;60(6):777–9. [22] Notari S, Capellari S, Giese A, Westner I, Baruzzi A, Ghetti B, et al. Effects of different experimental conditions on the PrPSc core generated by protease digestion: implications for strain typing and molecular classification of CJD. J Biol Chem 2004;279(16):16797–804. [23] Paramithiotis E, Pinard M, Lawton T, LaBoissiere S, Leathers VL, Zou WQ, et al. A prion protein epitope selective for the pathologically misfolded conformation. Nat Med 2003;9(7):893–9. [24] Peretz D, Williamson RA, Kaneko K, Vergara J, Leclerc E, Schmitt-Ulms G, et al. Antibodies inhibit prion propagation and clear cell cultures of prion infectivity. Nature 2001;412(6848):739–43. [25] Price KM. Production and characterisation of synthetic peptide-derived antibodies. In: Ritter MA, Ladyman, editors. Monoclonal antibodies: production, engineering and clinical application. Cambridge University Press; 1995. p. 60–84. [26] Prusiner SB. Prions. Proc Natl Acad Sci USA 1998;95:13363–8. [27] Souan L, Tal Y, Felling Y, Cohen IR, Taraboulos A, Mor F. Modulation of proteinase-K resistant prion protein by prion peptide immunization. Eur J Immunol 2001;31(8):2338–46. [28] Swietnicki W, Petersen R, Gambetti P, Surewicz WK. pHDependent stability and conformation of the recombinant human prion protein PrP(90–231). J Biol Chem 1997;272(44):27517–20.
2705
[29] Takahashi H, Takahashi RH, Hasegawa H, Horiuchi M, Shinagawa M, Yokoyama T, et al. Characterization of antibodies raised against bovine-PrP-peptides. J Neurovirol 1999;5(3):300–7. [30] Ulrih NP, Skrt M, Veranic P, Galvani V, Veranicˇ T, Cˇurin Sˇerbec V. Oligomeric forms of peptide fragment PrP (214–226) in solution are preferentially recognized by PrP(Sc)-specific antibody. Biochem Biophys Res Commun 2006;344(4):1320–6. [31] Van Regenmortel MHV. Antigenic cross-reactivity between proteins and peptides: new insights and applications. Trends Biochem Sci 1987;12:237–40. [32] Van Regenmortel MHV. Which structural features determine protein antigenicity? Trends Biochem Sci 1986;11:36–9. [33] Vogtherr M, Grimme S, Elshorst B, Jacobs DM, Fiebig K, Griesinger C, et al. Antimalarial drug quinacrine binds to C-terminal helix of cellular prion protein. J Med Chem 2003;46(17):3563–4. [34] Way G, Morrice N, Smythe C, O’Sullivan AJ. Purification and identification of secernin, a novel cytosolic protein that regulates exocytosis in mast cells. Mol Biol Cell 2002;13(9):3344–54. [35] White AR, Enever P, Tayebi M, Mushens R, Linehan J, Brandner S, et al. Monoclonal antibodies inhibit prion replication and delay the development of prion disease. Nature 2003;422(6927):80–3. [36] Yokoyama T, Kimura K, Tagawa Y, Yuasa N. Preparation and characterization of antibodies against mouse prion protein (PrP) peptides. Clin Diagn Lab Immunol 1995;2(2):172–6.
