University of Zurich Zurich Open Repository and Archive
Winterthurerstr. 190 CH-8057 Zurich http://www.zora.uzh.ch
Year: 2008
Novel dominant-negative prion protein mutants identified from a randomized library Ott, D; Taraborrelli, C; Aguzzi, A
Ott, D; Taraborrelli, C; Aguzzi, A (2008). Novel dominant-negative prion protein mutants identified from a randomized library. Protein Engineering Design and Selection : PEDS, 21(10):623-629. Postprint available at: http://www.zora.uzh.ch Posted at the Zurich Open Repository and Archive, University of Zurich. http://www.zora.uzh.ch Originally published at: Protein Engineering Design and Selection : PEDS 2008, 21(10):623-629.
Novel dominant-negative prion protein mutants identified from a randomized library Abstract Prion diseases are untreatable neurodegenerative disorders characterized by accumulation of PrP(Sc), an aggregated isoform of the cellular prion protein (PrP(C)). We generated a library of PrP variants with random mutations in the helix-3 domain and screened for dominant-negative mutants (DNMs) that would inhibit replication of prions (the Rocky Mountain Laboratory strain) in infected N2a cells. Two of the identified PrP mutants, Q167R and Q218K, were already known to counteract prion replication, thereby validating the effectiveness of this approach. In addition, novel DNMs were found efficiently to antagonize PrP(Sc) propagation in cells. In contrast to Q167R and Q218K, the newly identified DNMs S221P and Y217C resided on the cell surface at a substantially lower level, suggesting that robust cell surface display of DNM might not be a general prerequisite for efficient prion antagonism. The newly identified DNMs point to useful target-selective therapeutic tools for the treatment of prion diseases.
Novel dominant negative prion protein mutants identified from a randomized library
David Ott1,2,3, Cornelia Taraborrelli1,2,4 and Adriano Aguzzi1,5
1
Institute of Neuropathology, University Hospital Zurich, Schmelzbergstrasse 12, CH-8091
Zurich, Switzerland 2
These authors contributed equally
3
Present address: MorphoSys AG, Lena-Christ-Str. 48, DE-82152 Martinsried/Planegg,
Germany 4
Present address: Klinische Immunologie, University Hospital Zürich, Moussonstrasse 13,
CH-8044 Zürich 5
Corresponding author:
Adriano Aguzzi, Institute of Neuropathology, University Hospital Zürich, Schmelzbergstrasse 12, CH-8091 Zürich, Switzerland. Tel.: +41 1 255 2107, Fax.: +41 1 255 4402, Email:
[email protected]
Running head: Novel dominant-negative PrP-mutants
List of used abbreviations: DNM, dominant negative mutant; GPI, glycosylphosphatidylinositol (anchor); NI (no insert), expression vector without the PrP-ORF inserted into the multiple cloning site; ORF, open reading frame; PK, proteinase K; RML, the Rocky Mountain Laboratory prion-strain; SNP, single nucleotide polymorphism; WT, wild-type (PrP)
1
Abstract Prion diseases are untreatable neurodegenerative disorders characterized by accumulation of PrPSc, an aggregated isoform of the cellular prion protein, PrPC. We generated a library of PrP variants with random mutations in the helix-3 domain, and screened for dominantnegative mutants (DNMs) that would inhibit replication of prions (RML strain) in infected N2a cells. Two of the identified mutants, Q167R and Q218K, were already known to counteract prion replication, thereby validating the effectiveness of this approach. In addition, novel DNMs were found to efficiently antagonize PrPSc propagation in cells. In contrast to Q167R and Q218K, the newly identified S221P and Y217C mutants resided on the cell surface at a substantially lower level, suggesting that robust cell surface display of DNM might not be a general prerequisite for efficient prion antagonism. The newly identified DNMs point to useful target-selective therapeutic tools for treatment of prion diseases.
Keywords Prion / PrP / dominant-negative mutant / screening / helix-3 / library
Funding European Commission (LSHB-CT-2005-018805 and FOOD-CT-2004-023144); Swiss National Science Foundation (3100A0-111917 and NFP38+/403840-102445); NCCR on "Neural Plasticity and Repair"; Forschungskredit der Universität Zürich; Theodor and Ida Herzog-Egli Foundation.
Introduction
2
Prions are the infectious agents causing transmissible spongiform encephalopathies (TSE) such as Creutzfeldt-Jakob disease in human and BSE in cattle. The central event in the pathogenesis of prion diseases is the ordered aggregation of misfolded isoform (PrPSc) of the cellular prion protein (PrPC). The structural and mechanistic basis of the aggregation remains elusive, primarily because a high-resolution structure of PrPSc is not available. A promising approach for gaining mechanistic understanding of the conversion of PrPC into PrPSc consists in the identification of amino-acid residue substitutions which interfere with the conversion. The resulting structural perturbations might shed light onto the motifs and determinants of the transformation of the cellular protein PrPC into an infectious agent. In the present study we present an approach based on functional screening for potent dominant negative PrP mutants from a focused PrP library.
Materials and Methods Materials Reagents were obtained from Sigma, Fluka or Merck, unless stated otherwise. TBS-Tw (20 mM Tris, pH 7.5, 150 mM NaCl, 0.1% Tween-20) was used as a buffer for the blocking buffer and for the incubation with antibodies during SCA and Western blot. PAA buffer [PBS/NP-40 (2%)/Tween-20 (2%)] or RIPA buffer [PBS/NP-40 (1%)/sodium deoxycholate (0.25%)] were used for cell lysis. The generation of anti-PrP antibodies POM1, POM2 and POM3 was described elsewhere (Polymenidou et al., 2005). POM1 is directed against a discontinuous SDS-stable epitope on helix-1. POM2 and POM3 recognize, respectively, octapeptide repeats in the N-proximal region of PrP and residues 95-100 of murine PrP. Construction of the PrP mutant library Error-prone PCR by using nucleotide analogs was performed essentially as described previously (Zaccolo et al., 1996). The helix-3 library of murine PrP was generated on the DNA level by PCR (25 cycles) with primer pair CAAGGGGGAGAACTTCACC and
3
GTGCTGCTGGATCTTCTCCC (resulting in mutagenesis of residues Glu199 – Asp 226) in the presence of Taq DNA polymerase, dNTPs (0.2 mM) and dNTP analogs: 8-oxo-dGTP and dPTP (50 μM each). Full-length ORF for PrP was assembled by conventional overlapping PCR (25 cycles) with Vent DNA polymerase and dNTPs (0.2 mM). ORF encoding the PrP library or control PrP variants was cloned into expression vector pBMN-IRES-GFP (Swift et al., 1999). The resulting pBMN-PrP-IRES-GFP plasmid encodes for PrP variant and GFP (translated due to the IRES element in the mRNA transcript). Hereafter, we refer to the pBMN-IRES-GFP vector without any PrP-ORF inserted into the multiple cloning site as the ‘NI’ (no insert plasmid) control.
Scrapie cell-based assay (SCA) in 96-well format SCA used in the screening was modified, based on the original protocol initially developed for sensitive detection of scrapie prions in a cell-based bioassay (Klohn et al., 2003). Neuroblastoma N2a cells (the subclone N2aPK1 (Klohn, et al., 2003)), chronically infected with the RML scrapie prion strain, were seeded on 96-well cell culture plate in growth medium (OptiMEM with 10% FCS, Pen/Strep/GlutaMax) at density 1.2⋅104 cells per well, and transiently transfected with pBMN-PrP-IRES-GFP plasmid 2 h later. Mock-infected N2aPK1 cells (treated with brain homogenate from non-inoculated mouse) were used as a control. Per one well, we routinely used 0.2 μg plasmid DNA and 0.5 μl Lipofectamine 2000 (Invitrogen), each diluted in 25 μl of OptiMEM. The total volume of the growth medium was 200 μl per well. Each PrP variant was tested in a triplet on a cell culture 96-well plate. Four days later, the growth medium was removed, and the cells were harvested in 200 μl PBS/EDTA (2mM). One aliquot of the cell suspension (80 μl) was used for determination of cell viability by using the MTS assay (CellTiter 96 Aqueous One Solution Cell Proliferation Assay Kit; Promega; absorbance readout at 490 nm; background signal was defined as the reading of samples devoid of cells), and another aliquot was used for determination of the PrPSc accumulation in transfected cells as follows. The cell suspension (80 μl) was mixed
4
with 80 μl proteinase K solution (60 μg/ml) in the PAA buffer, and incubated for 1 hour at 37°C. All following incubation steps were carried out at room temperature. The digested cell lysate was transferred into the Immobilon-P plate (hydrophobic, high protein binding white plate; Millipore) pre-rinsed with 70% ethanol and PBS, and passed through the filter by applying vacuum. The plate was washed with PBS, and residual PK activity was stopped by incubation with PMSF (1 mM in PBS) for 10 minutes. PrPSc adherent to the filters was then denatured with guanidine thiocyanate (3 M in PBS) for 10 minutes. Wells were washed with PBS and incubated sequentially with blocking solution (Top-block) for 1 hour, the POM1 antiPrP antibody (0.4 μg/ml), and eventually with anti-murine IgG1-HRP conjugate (0.2 μg/ml; Zymed, Cat. No. 61-0120). Luminescence signal (relative luminescence units, RLU) was recorded by using “SuperSignal ELISA Pico chemiluminescent substrate” (Pierce; 100 μl per well). In a pilot experiments we found that the luminescence signal in the modified SCA reflected the percentage of the RML-infected cells in a sample containing a mixture of RMLinfected and non-infected N2a cells (data not shown). We concluded that the luminescence signal linearly reflected amount of PrPSc in our samples, and hence it was not necessary to use calibration (= serial dilution of purified PrPSc-standard) in our modified SCA. PrPSc accumulation (RLU) was normalized to the cell viability (A490 nm; see above) resulting in RLU´value for every single well. Average RLU´-value (ØRLU´) was determined from three (or more) technical replicas (=wells) on the 96-well plate. An example of raw data is given in Supplementary Figure 1. PrPSc accumulation for each tested library member was corrected for residual PrPC-signal, and then related to the corrected PrPSc -signal from the RML cells transfected with dummy vector (the NI-control) as follows: PrPSc accum.i = (ØRLU´i RML - ØRLU´ PrP/WT mock) / (ØRLU´NI RML - ØRLU´ NI mock) x 100
SCA has been repeated in at least two additional independent experiments for all library members, which caused a decrease of the PrPSc accumulation in the cells by more than 50% related to the NI control.
5
Quantification of PrPSc accumulation in transfected N2a/RML cells by Western blot The N2a cells, chronically infected with the RML prions, were seeded on a 25-cm2 cell culture T-flask, and transiently transfected as described above (a scale-up factor 30, compared to the 96-well plate). Three days later, the cells were harvested, and the pellet was resuspended in 120 μl of the RIPA buffer. The cell lysate was cleared by centrifugation at 2000 g for 3 minutes, adjusted to protein concentration 5 mg/ml (determined by BCA assay, Pierce), and incubated with PK (20 μg/ml) at 37°C for 30 minutes. Non-digested aliquot was kept on ice. The samples were then incubated with NuPAGE sample buffer (Invitrogen; supplemented with 100 mM DTT) for 5 minutes at 95°C. Proteins in the sample (25 μg per lane) were separated on a pre-cast NuPAGE 12% Bis-Tris gel, and blotted onto nitrocellulose membrane. We used anti-PrP antibody POM3 (5 μg/ml) for detection of PrPSc because: i- its epitope (residues 95-100 in murine PrP) remains intact during the PK-digest, and ii- recognition of PrP by the POM1 antibody used for the primary screening was found later to be affected in mutants containing mutation Y217C (data not shown). Sample aliquots not digested with PK were probed by Western blot by using anti-GFP antibody (100 ng/ml; Roche Applied Science) or anti-β-actin antibody (100 ng/ml; Chemicon) in order to assess transfection efficiency in N2a cells, and protein loading on the gel, respectively. Rabbit antimurine IgG1-HRP conjugate was used as a secondary antibody. Light development using “SuperSignal West Pico chemiluminescent substrate” (Pierce) was detected by a CCD camera, and intensity of protein bands was quantified by TINA software (Raytest). Quantification of PrP expression in HpL3-4 cells by Western blot Immortalized HpL3-4 cells derived from hippocampus of Prnp-/- mouse (Kuwahara et al., 1999) were seeded on 12-well cell culture plate at density 1.8⋅105 cells per well, and transiently transfected as described above (a scale-up factor 5, compared to the 96-well plate). Twenty four hours later, the cells were harvested, resuspended in 50 μl of the PAA
6
buffer supplemented with protease inhibitor cocktail (0.5%; Sigma, Cat. No. P-8340), and cleared by centrifugation at 2000 g for 3 minutes. An aliquot of the cell lysate was used for protein determination by the BCA assay. The cell lysates was mixed with the NuPAGE sample buffer, and incubated for 5 minutes at 95°C. The Western blot was performed as described above. Here, anti-PrP antibody POM2 (0.4 μg/ml) was used as a primary antibody for detection of full-length PrP because POM2 has a very high avidity due to its multiple binding of octa-repeats in the N-terminus of PrP. For quantification, signals corresponding to the amount of detected PrP and GFP (co-expressed due to the IRES-element) were normalized to the signal corresponding to β-actin, and then the PrP signal was normalized to the GFP signal (compensation for variation in transfection efficiency among the cell samples). Expression level of DNMs was related to PrP WT. Quantification of cell surface expression of PrP in HpL3-4 cells by flow cytometry HpL3-4 cells were seeded on 12-well cell culture plate, and transiently transfected, as described above. Twenty four hours after the transfection, the cells were harvested, and stained with the conjugate of anti-PrP antibody POM2 with Cy5. PBS/FCS (1%)/EDTA (10 mM)/NaN3 (0.1%) was used for the antibody dilution and subsequent washes. The incubation with the POM2-Cy5 and washes were carried out at 4°C in order to prevent internalization of PrP in the cells. Data were acquired on an instrument “BD FACS Calibur”. We gated on live and GFP-positive cells, and the mean Cy5 fluorescence was used for quantification of the PrP expression level. Signal from the cells transfected with the NI control was used for offset of the background.
7
Results Design and generation of a focused PrP library. We set out to generate a broad range of amino acid residue substitutions which might interfere with the conversion of PrP. Since random mutagenesis is an established tool for extensive exploration of the sequence space in proteins (Drummond et al., 2005, Hermes et al., 1990), we opted to produce a library of mutants randomized in the open reading frame of murine Prnp by error-prone PCR. The library was then screened for dominant negative mutants (henceforth termed DNMs) of PrP. A number of methods is available for in vitro conversion of PrPC into a PrPSc-like state (Kocisko et al., 1994, Saborio et al., 2001) yet none of these methods has been validated for the discovery of dominant-negative mutants. While more cumbersome and not easily amenable to forward genetics, cell culture assays were deemed more suitable for this purpose because of their sensitivity and precise quantitative responses (Klohn, et al., 2003). Therefore, we decided to screen a focused small-size PrP library for DNMs in a functional, albeit low-throughput screening in cells. It has been observed that N-proximal truncations of PrP (deletion of residues 23-88) did not prevent PrPSc propagation in cell culture (Rogers et al., 1993) and in mice (Fischer et al., 1996), suggesting that the N-terminus does not play a crucial role in the PrP conversion. In contrast, PrP-mutations, identified due to naturally occurring ‘prion-resistant’ polymorphism in sheep and human, were found to possess potent dominant-negative properties. In particular, the substitutions Q167R in the loop connecting β-sheet 2 and helix-2 (loop ‘L1’, residues 165-173), and Q218K in the helix-3 (residues 199-226), were found to strongly inhibit PrPSc propagation in cell culture and in transgenic mice (Kaneko et al., 1997, Perrier et al., 2002). We decided to analyze a series of random mutations within the helix-3 domain. We predicted that the replacement of many of its residues might result in disruption of stabilizing intrahelical and intramolecular hydrogen bonds and salt bridges, thereby bringing about structural perturbations in the PrP molecule that may be a prerequisite for interference with PrPSc aggregation. The length of helix-3 (28 residues) was expected to allow for satisfactory
8
residue substitution coverage by screening ~100 members of a library with a mutational load of ~1-2 mutations per clone. Hence the above strategy would result in testing ~5 substitutions of each residue within the helix-3. Noteworthy, we opted not to introduce a tag for the 3F4-antibody (Kascsak et al., 1987) (the antibody recognizes murine PrP with mutations L108M/V111M) often used to distinguish endogenous and heterologous PrP in murine neuroblastoma cells. It has been reported that the murine PrP/L108M/V111M heterologously expressed in the N2a cells is not converted per se into PrPSc, and it can antagonize accumulation of wild-type PrPSc in the N2a cells (Priola et al., 1994). Similarly, we have observed that PrP/L108M/V111M, in contrast to wildtype PrP, heterologously expressed in chronically-infected N2a cells did not contribute to the generation of PrPSc in the cells (data not shown). If the additional L108M/V111M-mutations were introduced into the sequence of the PrP/helix 3-library members, interpretation of the dominant-negative effect would be considerably compromised. The library of PrP randomized in the helix-3 has been generated on the DNA level by errorprone PCR, as described in Materials and Methods. We analyzed sequences of 40 randomly picked clones in order to characterize the quality of the library. As expected, the vast majority of nucleotide substitutions consisted of the transitions A↔G and T↔C. The rate of mutations was ~3% at the level of DNA, and the mutational load was 2±2 mutations per clone. Most of the analyzed clones (97%) encoded full-length PrP. The vast majority (~90%) of all analyzed mutations were found within the gene region that encodes the helix-3. We conclude that quality of the generated library was suitable for our screening purpose.
Screening of the PrP library by scrapie cell-based assay. The neuroblastoma N2a cells (Race et al., 1987) provide an attractive cell culture model of the chronic propagation of scrapie-derived prions in vitro, and have been exploited in many paradigms for high-throughput screens of small molecular weight anti-prion compounds (Kocisko and Caughey, 2006). Prion propagation in the infected cells is usually stable for
9
several months, and does not seem to affect cell viability. We decided to use scrapie cell assay (hereafter SCA) (Klohn, et al., 2003) suitable for handling of N2a cells in the 96-well format. N2a cells (subclone PK1) chronically infected with RML prions (hereafter N2a/RML) were transfected with plasmids encoding individual members of the PrP library, and several days later the amount of PrPSc accumulated in the cells was assessed. We modified SCA conditions in order to ensure efficient removal of PrPC which interferes with the detection of PrPSc. This is an important consideration in our set-up because the transfected N2a cells express endogenous PrP under the control of the authentic murine PrP promoter, and also a PrP variant under the control of a viral promoter present in the expression vector. In particular, we found that it was crucial to increase the PK concentration from 0.5 to 30 μg/ml (final concentration; data not shown for the optimization of SCA). Under these optimized conditions, signals yielded by N2a/RML cells transfected with the NI control plasmid were ca. 10-fold higher than those of mock-infected N2a cells transfected with plasmid encoding wild-type PrP. This indicates that the detection of PrPSc was not compromised by the overexpression of PrP in transfected cells. We screened 140 library members using the optimized SCA (raw data for a representative SCA are shown in Suppl. Figure 1). Since the screen was aimed at identifying potent DNMs which would significantly inhibit PrPSc propagation in the cells, we focused on analysis of PrP mutants which caused more than 50% decrease of PrPSc accumulation in N2a/RML cells when compared to a NI control (Figure 1). The mutations within the sequence of the helix-3 domain in these DNMs were distributed across the entire domain. In addition to the positive control (PrP/Q218K), we identified three other single mutations among the DNMs: PrP/Q167R, PrP/S221P and PrP/Y225C. Q167 does not reside in the helix-3, but in the L1 loop which had not been intentionally mutagenized. Hence the Q167R mutation was incidentally introduced during assembly of the PrP gene by PCR. All other DNMs contained multiple mutations. We introduced mutations K203R, T215A, Y225H, or Y217C which were frequently found in several multiple DNMs (Figure 1) into the PrP gene, and evaluated the effect of these single mutants in SCA. We found that PrP/Y217C is a very potent DNM, 10
whereas the inhibitory effects of the mutants PrP/Y225H and PrP/T215A were more moderate. The mutant PrP/K203R (with a basic residue substituted for another basic residue) behaved in SCA like PrP WT (Figure 1). We tested in the SCA also rabbit and canine PrP, which may be particularly resilient to conversion into PrPSc given that rabbits and dogs are resistant to various prion strains (Loftus and Rogers, 1997, Lysek et al., 2004). PrPSc propagation in N2a/RML cells transfected with these PrP variants was similar to that observed in NI control cells (85±15 % and 93±5 % for rabbit and canine PrP, respectively; N=2), suggesting that rabbit PrP and canine PrP were indeed not converted into PrPSc form, but on the other hand indicating that these PrP variants do not antagonize propagation of RML prions. In order to confirm the above results from the screening, we analyzed the effect of the most potent DNMs on clearance of RML in the cells by Western blot. In agreement with the SCA results,
transfection
of
N2a/RML
cells
with
PrP/Q167R,
PrP/Y217C,
PrP/S221P,
PrP/Y217C/Y225H/D226G, or PrP/M212L/Q218R/S221P resulted in substantial (70-90%) reduction of PrPSc, whereas transfection of the N2a/RML cells with a plasmid encoding PrP WT resulted in ~3-fold increase in PrPSc accumulation (Figure 2). Notably, the prevalence of transfected N2a/mock or N2a/RML cells was only ~30-50%. This suggests that DNM expression, in addition to clearing prions from transfected cells, interferes with prion propagation also in non-transfected N2a/RML cells by a mechanism which remains to be elucidated. Expression of selected dominant negative mutants in HpL3-4 cells. The biogenesis of the most potent DNMs was characterized in PrP-deficient HpL3-4 cells, which were originally derived from the hippocampus of a Prnp-/- mouse (Kuwahara, et al., 1999). Under these conditions, the folding and cellular trafficking of the tested PrP variants cannot be modified by the co-expression of WT PrPC, as may be the case in N2a cells. Western blot analyses by using the anti-PrP antibody POM2, which binds octa-repeats in the N-terminus of PrP (Polymenidou, et al., 2005), showed that the overall expression level and
11
expression profile (mobility of PrP glycoforms) of PrP/Q167R, PrP/Y217C, PrP/Q218K, PrP/S221P, PrP/Y217C/Y225H/D226G, and PrP/M212L/Q218R/S221P was similar to that of transfected PrP WT (Figure 3a). We detected a tendency for PrP/Q218K to be expressed at somewhat higher levels than PrP WT; however, expression levels of PrP WT and tested DNMs did not differ significantly (Figure 3b). Mature
PrP
molecules
are
normally
attached
to
the
plasma
membrane
via
glycosylphosphatidylinositol (GPI) anchor, and localized in lipid rafts (Stahl et al., 1987). We therefore wondered whether the selected DNMs stably reside on the cell surface. In order to address this question, we opted to quantify PrP localized on the cell surface by flow cytometry. HpL3-4 cells were transfected with DNM-encoding plasmids, and stained with conjugate of the anti-PrP-antibody POM2 with the Cy5 fluorophore. Staining was carried out at 4°C in order to prevent internalization of the label prior to flow cytometric analysis. The data reported in Figure 4 suggest that PrP/Q218K and PrP/Q167R reside on the cell surface at a similar level as PrP WT. In contrast, the surface presence of other DNMs (PrP/Y217C, PrP/S221P, PrP/Y217C/Y225H/D226G, and PrP/M212L/Q218R/S221P) was much lower compared to the PrP WT. PrPC is known be to constitutively internalized (Kaneko et al., 1997, Prado et al., 2004). It is plausible that the DNMs, which seem to be not abundant on the cell surface, are properly targeted to the cell surface; however, due to their destabilized structure they are internalized in a substantially faster rate than wild-type PrPC. An analogous mechanism has been previously demonstrated for several mutated G protein-coupled receptors which also reside in lipid rafts (Ott et al., 2004). Taken together, it seems that stable localization on the cell surface of the selected DNMs is not a prerequisite for the prion antagonism. In fact, fast internalization into early endosomes may enhance prion antagonism for, at least, some of the selected DNMs since it was suggested that PrP is likely to be converted under the low pH conditions found in endosomes (Borchelt et al., 1990, Kaneko, et al., 1997).
12
Discussion In this study, we screened for PrP mutants that would inhibit prion propagation in prioninfected N2a cells. Antagonistic effect of heterologous (mutated or chimeric) PrP-molecules on prion propagation in cell culture was initially reported by Priola and colleagues (Priola, et al., 1994). In following mutagenesis studies, researchers tested effects of naturally occurring protective polymorphisms (Crozet et al., 2004, Kaneko, et al., 1997, Zulianello et al., 2000), glutamine residue substitutions (Atarashi et al., 2006), or reversal of charged residues in the helix-1 (Norstrom and Mastrianni, 2006) on PrP conversion. Here, we reasoned that the structural determinants of prion propagation are very poorly understood, and therefore rational predictions about the impact of individual mutations might be treacherous if not impossible. For these reasons we opted to undertake an unbiased effort by randomizing the sequence of a defined region of PrP, helix-3 within the carboxy-proximal globular domain of PrP. We found a relatively broad distribution and various kinds of amino acid residue substitution in the selected dominant negative mutants (DNMs). Some of these individual mutations, in particular Q167R, Y217C, Q218K and S221P, had a pronounced effect on prion propagation. While PrP/Q167R and PrP/Q218K had been already identified as potent DNMs based on the natural ‘prion-resistant’ polymorphism in ovine and human PrP, respectively (Kaneko, et al., 1997), the Y217C and S221P mutations point to hitherto unreported hot spots that may be crucial for the process of PrP conversion. It is conceivable that the additional cysteine residue in the PrP/Y217C mutant, possibly affecting the formation of the conserved S-S bond during the folding of PrP, and the S221P replacement within the helix-3, which can cause a break in the helix, have effect on PrP misfolding and aggregation resulting in the interference with the PrP conversion in the presence of wild-type PrPC in the cells. How do DNMs exert their anti-prion effects? The nucleation hypothesis of prion propagation (Aguzzi and Weissmann, 1997, Jarrett and Lansbury, 1993) provides a very simple model for their action. In this framework, prions consist essentially of highly ordered, oligomeric aggregates of ca. 15-20 PrPSc molecules (Silveira et al., 2005). In many similar paradigms of 13
nucleation, such aggregates grow appositionally at one end by addition of individual monomers. Here, DNMs may yield ‘defective’ monomers that can be recruited by existing seeds, yet do not have reactive interface necessary for the recruitment of further monomers. Hence DNMs might effectively “cap” seeds and quench their infectious potential (Horiuchi et al., 2000). Testable structural models of DNMs, derived e.g. from nuclear magnetic resonance studies and/or from folding simulations, are likely to assist in the elucidation of the mechanism of the PrP conversion at the atomic level, and may help devising even more powerful DNMs based on rational predictions rather than random mutagenesis. However, several alternative models of conversion inhibition are thinkable and would be compatible with the results reported above. For example, it is possible that DNMs affect prion propagation by competing with PrPC for a putative cofactor. The extraordinarily inefficient generation of prion infectivity in vitro from chemically defined constituents (Legname et al., 2004) suggests the existence of such cofactors in cells and in tissues. Circumstantial evidence suggests that the site of PrP conversion, and possibly of any hypothetical auxiliary factor (Kaneko, et al., 1997) may be congruent with caveolae-like domains of the cell surface or endosomes (Kaneko, et al., 1997, Nunziante et al., 2003). However, other cellular localizations are by no means excluded, and the auxiliary factors may encompass also nonproteinaceous polyanionic molecules such as nucleic acids (Deleault et al., 2007). As a next step, it will be of interest to ascertain the potential of the newly discovered DNMs to inhibit PrPSc propagation in vivo. For any therapeutic application to be successful, it will be crucial that prion antagonists efficiently diffuse within the CNS and exert their effects noncell-autonomously. Encouragingly, recombinant non-glycosylated PrP/Q218K administered in a soluble form (without the GPI anchor) was found to antagonize PrPSc propagation in prion-infected N2a cells (Kishida et al., 2004). However, another study found that cellautonomous expression and tethering by a GPI anchor is important for efficient PrPSc antagonism by PrP/Q218K in prion-infected N2a cells (Zulianello, et al., 2000). We made similar observations when selected DNMs were expressed in a soluble form, without the GPI 14
anchor, in N2a/RML cells (data not shown). Maybe the GPI anchor is important for efficient targeting of DNM into endosomes, where PrPC is presumed to be converted into PrPSc. If so, this requirement could be fulfilled by the use of molecular targeting signals for transport of DNMs into endosomes. Efficient targeting signals may be, for instance, derived from the internalization domain of bacterial or plant toxins (Medina-Kauwe, 2007, Sandvig and van Deurs, 2002).
15
Acknowledgements We thank Christian Julius and Gino Miele for discussions, Jan Kranich for Cy5-labeled POM2, Garry Nolan (Stanford University; California) for the pBMN-IRES-GFP vector, and Amedeo Caflisch and François Marchand for calculations of PrP aggregation propensity. AA is supported by EU grants LSHB-CT-2005-018805 and FOOD-CT-2004-023144, the Swiss National Science Foundation, the Stammbach Foundation, the Ernst-Jung Foundation, and the NCCR on "Neural Plasticity and Repair". DO was supported by the Forschungskredit der Universität Zürich and the Theodor und Ida Herzog-Egli Foundation. The funding institutes and foundations had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
16
References Aguzzi A. and Weissmann C. (1997) Nature, 389, 795-798. Atarashi R., Sim V.L., Nishida N., Caughey B. and Katamine S. (2006) J. Virol., 80, 78547862. Borchelt D.R., Scott M., Taraboulos A., Stahl N. and Prusiner S.B. (1990) J. Cell. Biol., 110, 743-752. Crozet C., Lin Y.L., Mettling C., Mourton-Gilles C., Corbeau P., Lehmann S. and Perrier V. (2004) J. Cell. Sci., 117, 5591-5597. Deleault N.R., Harris B.T., Rees J.R. and Supattapone S. (2007) Proc. Natl. Acad. Sci. USA, 104, 9741-9746. Drummond D.A., Iverson B.L., Georgiou G. and Arnold F.H. (2005) J Mol Biol, 350, 806-816. Fischer M., Rülicke T., Raeber A., Sailer A., Moser M., Oesch B., Brandner S., Aguzzi A. and Weissmann C. (1996) EMBO J., 15, 1255-1264. Hermes J.D., Blacklow S.C. and Knowles J.R. (1990) Proc Natl Acad Sci U S A, 87, 696700. Horiuchi M., Priola S.A., Chabry J. and Caughey B. (2000) Proc Natl Acad Sci U S A, 97, 5836-5841. Jarrett J.T. and Lansbury P.T., Jr. (1993) Cell, 73, 1055-1058. Kaneko K., Vey M., Scott M., Pilkuhn S., Cohen F.E. and Prusiner S.B. (1997) Proc. Natl. Acad. Sci. USA, 94, 2333-2338. Kaneko K., Zulianello L., Scott M., Cooper C.M., Wallace A.C., James T.L., Cohen F.E. and Prusiner S.B. (1997) Proc. Natl. Acad. Sci. USA, 94, 10069-10074. Kascsak R.J., Rubenstein R., Merz P.A., Tonna-DeMasi M., Fersko R., Carp R.I., Wisniewski H.M. and Diringer H. (1987) J Virol, 61, 3688-3693. Kishida H., Sakasegawa Y., Watanabe K., Yamakawa Y., Nishijima M., Kuroiwa Y., Hachiya N.S. and Kaneko K. (2004) Amyloid, 11, 14-20. Klohn P.C., Stoltze L., Flechsig E., Enari M. and Weissmann C. (2003) Proc. Natl. Acad. Sci. USA, 100, 11666-11671. Kocisko D.A. and Caughey B. (2006) Methods Enzymol., 412, 223-234. Kocisko D.A., Come J.H., Priola S.A., Chesebro B., Raymond G.J., Lansbury P.T. and Caughey B. (1994) Nature, 370, 471-474. Kuwahara C., Takeuchi A.M., Nishimura T., Haraguchi K., Kubosaki A., Matsumoto Y., Saeki K., Matsumoto Y., Yokoyama T., Itohara S. et al. (1999) Nature, 400, 225-226.
17
Legname G., Baskakov I.V., Nguyen H.O., Riesner D., Cohen F.E., DeArmond S.J. and Prusiner S.B. (2004) Science, 305, 673-676. Loftus B. and Rogers M. (1997) Gene, 184, 215-219. Lysek D.A., Nivon L.G. and Wüthrich K. (2004) Gene, 341, 249-253. Medina-Kauwe L.K. (2007) Adv. Drug Deliv. Rev., 59, 798-809. Norstrom E.M. and Mastrianni J.A. (2006) J. Virol., 80, 8521-8529. Nunziante M., Gilch S. and Schätzl H.M. (2003) Chembiochem, 4, 1268-1284. Ott D., Frischknecht R. and Plückthun A. (2004) Protein Eng. Des. Sel., 17, 37-48. Perrier V., Kaneko K., Safar J., Vergara J., Tremblay P., DeArmond S.J., Cohen F.E., Prusiner S.B. and Wallace A.C. (2002) Proc. Natl. Acad. Sci. USA, 99, 13079-13084. Polymenidou M., Stoeck K., Glatzel M., Vey M., Bellon A. and Aguzzi A. (2005) Lancet Neurol., 4, 805-814. Prado M.A., Alves-Silva J., Magalhaes A.C., Prado V.F., Linden R., Martins V.R. and Brentani R.R. (2004) J. Neurochem., 88, 769-781. Priola S.A., Caughey B., Race R.E. and Chesebro B. (1994) J Virol, 68, 4873-4878. Race R.E., Fadness L.H. and Chesebro B. (1987) J. Gen. Virol., 68 ( Pt 5), 1391-1399. Rogers M., Yehiely F., Scott M. and Prusiner S.B. (1993) Proc. Natl. Acad. Sci. USA, 90, 3182-3186. Saborio G.P., Permanne B. and Soto C. (2001) Nature, 411, 810-813. Sandvig K. and van Deurs B. (2002) FEBS Lett., 529, 49-53. Silveira J.R., Raymond G.J., Hughson A.G., Race R.E., Sim V.L., Hayes S.F. and Caughey B. (2005) Nature, 437, 257-261. Stahl N., Borchelt D.R., Hsiao K. and Prusiner S.B. (1987) Cell, 51, 229-240. Swift S., Lorens J., Achacoso P. and Nolan G.P. (1999) Curr. Protoc. Immunol., 10.17.1410.17.29. Zaccolo M., Williams D.M., Brown D.M. and Gherardi E. (1996) J. Mol. Biol., 255, 589-603. Zulianello L., Kaneko K., Scott M., Erpel S., Han D., Cohen F.E. and Prusiner S.B. (2000) J. Virol., 74, 4351-4360.
18
Figure Legend
Figure 1. Accumulation of PrPSc determined by SCA for N2a/RML cells transfected with plasmid encoding PrP variant. N2a/RML cells were transfected with the pBMN-PrPIRES-GFP plasmid, harvested 4 days later, and treated with PK (30 μg/ml) for detection of PrPSc in SCA (see Materials and Methods). The values (average and standard deviation from N independent experiments) were calculated as a ratio of luminescence signals, normalized to the number of viable cells, for a tested PrP variant and the NI control (set as 100%). Values in the upper panel are shown only for the selected PrP library members, which caused a decrease in the PrPSc accumulation of more than 50%, related to the NI control. In each run of the SCA, we used plasmids encoding PrP WT or PrP/Q218K as controls. In the lower panel, values are shown for single PrP mutants, which were constructed after the primary screen, based on mutations frequently found in DNMs, listed in the upper panel.
Figure 2. Accumulation of PrPSc is decreased in N2a/RML cells transfected with selected DNMs, as determined by Western blot. N2a cells, chronically infected with RML (annotated as ‘RML’ cells), and mock-infected N2a cells (annotated as ‘mock’ cells) were transfected with the pBMN-PrP-IRES-GFP plasmid, harvested 3 days later, and treated with PK (20 μg/ml) for detection of PrPSc. Western blot data representative of two independent transfections are shown. Normalized value for the NIcontrol (100%) did not lie within the confidence interval determined for each of the five tested PrP-mutants (N=2). Thus the effect of each tested DNM was significantly distinct from the effect of the NI-control. The tested DNMs are annotated as follow: M1: PrP/Q218K, M2: PrP/Q167R, M3: PrP/Y217C, M4: S221P, M5: PrP/Y217C/Y225H/D226G. WT relates to wild-type PrP and NT relates to non-transfected cells.
Figure 3. Overall expression of selected DNMs in transfected HpL3-4 cells, determined by Western blot. A: The cells were transiently transfected with the pBMN-PrP-IRES-GFP 19
plasmid, and harvested 24 hours later. The tested DNMs are annotated as follow: M1: PrP/Q218K, M2: PrP/Q167R, M3: PrP/Y217C, M4: S221P, M5: PrP/Y217C/Y225H/D226G. WT relates to wild-type PrP and NT relates to non-transfected cells. B: PrP expression level (related to PrP WT) determined in 4 independent transfection experiments. The confidence interval (at 95% probability level) is defined as average value ± 1.96 standard deviation.
Figure 4. Cell surface expression of selected DNMs in transfected HpL3-4 cells, determined by flow cytometry. The cells were transiently transfected with the pBMN-PrPIRES-GFP plasmid, and harvested 24 hours later. For the analysis of PrP expression on the cell surface, gates were set on live and GFP-positive cells. Values are based on 3 independent transfection experiments. NI relates to the cells transfected with the NI control plasmid.
20
Figure 1
residue number in murine PrP helix 3 Accumulation of PrPSc: related to the NI control average (%) SD (%) N WT
Q218K
167
199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226
187
39
27
Q
E
T
D
V
K
M
M
E
R
V
V
E
Q
M
C
V
T
Q
Y
Q
K
E
S
Q
A
Y
Y
D
21 22 24 28 33 37 37 40 42 42 44 45 46 48 48 49
11 14 17 6 12 18 4 4 11 6 6 12 15 11 17 14
6 6 6 4 4 27 3 3 3 3 3 3 3 3 3 3
R Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q
E E E E E E E E E E E E E E E E
T T T T A T T T T T T T A T T T
D D D D D D D D D D D D D D G D
V V V V V V V V V V V V V A V V
K K K K K K K K R R K R K K K K
M M M M M M M M M M M M M M M M
M M M M M M M T M M M M M M M M
E E E E E E E E E E E E E E E E
R R R R R R R R R R R R R R R R
V V V V V V V V V V V V V V A V
V V V V V V V V V V V V V V V V
E E E E E E E E E E E K E E E E
Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q
M M M L M M M M V M M M M M M M
C C C C C C C C C C C C C C C C
V V V V V V V V V V V V A V V V
T T T T T T A T T T T A T T T T
Q Q Q Q Q Q Q R Q Q Q Q Q Q Q Q
Y Y C Y C Y Y Y Y Y Y Y Y Y Y Y
Q Q Q R Q K R Q Q Q Q Q Q R Q R
K K K K K K R K K E K K K K K K
E E E E E E E E E E E E E E E E
S P S P S S S S S P S S S S S S
Q Q Q Q Q Q Q Q Q Q Q Q Q Q R Q
A A A A A A A A A A A A A A A A
Y Y Y Y Y Y Y Y Y Y Y H Y Y Y Y
Y Y H Y Y Y Y H Y Y C Y Y Y Y Y
D D G D D D D D D D D N D D D D
E E E E
S S S S
Q Q Q Q
A A A A
Y Y Y Y
Y H Y Y
D D D D
single mutants constructed after the primary screen Y217C Y225H T215A K203R
25 51 79 185
14 4 22 56
4 2 2 2
Q Q Q Q
T T T T
D D D D
V V V V
K K K R
M M M M
M M M M
E E E E
R R R R
V V V V
V V V V
E E E E
Q Q Q Q
M M M M
C C C C
V V V V
T T A T
Q Q Q Q
C Y Y Y
Q Q Q Q
K K K K
Figure 2
WT
NT
M5
M4
mock
M3
M2
M1
WT
NT
-PK
RML
37 25
PrP
20
GFP
M5
M4
M3
M2
M1
WT
NT
Figure 3a
37
PrP 25
β-actin GFP
Figure 3b
expression level related to PrP WT
5
4
3
2
1
0 NT
WT
Q218K
Q167R
Y217C
S221P
Y217C/ Y225H/ D226G
M212L/ Q218R/ S221P
Figure 4
surface expression (in %) related to PrP WT
120
100
80
60
40
20
0 NI
WT
Q218K
Q167R
Y217C
S221P
Y217C/ Y225H/ D226G
M212L/ Q218R/ S221P
Supplementary Figure 1. Example of raw data from SCA. N2a cells (mock-infected or RML-infected) were plated in 96-well plates and transiently transfected with plasmid encoding PrP variants. Four days later, cells in the plates were processed and analyzed. Bars and error bars represent average values and standard deviations, respectively, of triplicate determinations. A: Cell viability determined by MTS assay. All values were corrected for the background (defined as the signal in a sample devoid of cells). B: PrPSc accumulation in cells determined after PK digestion by using antibody POM1.
1
