SEARCH

SEARCH BY CITATION

Keywords:

  • epigallocatechin;
  • green tea;
  • neurodegeneration;
  • prion;
  • scrapie;
  • stress-protection

Abstract

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

A hallmark in prion diseases is the conformational transition of the cellular prion protein (PrPC) into a pathogenic conformation, designated scrapie prion protein (PrPSc), which is the essential constituent of infectious prions. Here, we show that epigallocatechin gallate (EGCG) and gallocatechin gallate, the main polyphenols in green tea, induce the transition of mature PrPC into a detergent-insoluble conformation distinct from PrPSc. The PrP conformer induced by EGCG was rapidly internalized from the plasma membrane and degraded in lysosomal compartments. Isothermal titration calorimetry studies revealed that EGCG directly interacts with PrP leading to the destabilizing of the native conformation and the formation of random coil structures. This activity was dependent on the gallate side chain and the three hydroxyl groups of the trihydroxyphenyl side chain. In scrapie-infected cells EGCG treatment was beneficial; formation of PrPSc ceased. However, in uninfected cells EGCG interfered with the stress-protective activity of PrPC. As a consequence, EGCG-treated cells showed enhanced vulnerability to stress conditions. Our study emphasizes the important role of PrPC to protect cells from stress and indicate efficient intracellular pathways to degrade non-native conformations of PrPC.

Abbreviation used
CD

circular dichroism

ECG

epicatechin gallate

EGCG

epigallocatechin gallate

eq

molar equivalents

GCG

gallocatechin gallate

GFP

green fluorescent protein

GPI

glycosylphosphatidylinositol

HD

hydrophobic domain

PBS

phosphate-buffered saline

PK

proteinase K

PMSF

phenylmethylsulfonyl fluoride

PrPC

cellular prion protein

PrPSc

scrapie prion protein

rPrP

recombinant mouse prion protein

RT

room temperature

ScN2a

scrapie-infected N2a

SDS/PAGE

sodium dodecyl sulfate/polyacrylamide gel electrophoresis

The cellular prion protein (PrPC) is a highly conserved glycosylphosphatidylinositol (GPI)-anchored protein, mainly present at the plasma membrane of neuronal and lymphatic cells. In prion diseases of humans and mammals PrPC is converted into a detergent-insoluble and partially proteinase K (PK)-resistant isoform, designated scrapie PrP (PrPSc), which is the main component of infectious prions (reviewed in Collinge 2001; Prusiner et al. 1998; Weissmann et al. 1996; Chesebro 2003). The mechanism of PrPSc formation is still enigmatic but the conversion is thought to occur after PrPC has reached the plasma membrane or is re-internalized for degradation (Caughey and Raymond 1991; Caughey et al. 1991; Borchelt et al. 1992). In the majority of prion diseases, neurodegeneration is tightly linked to the formation of infectious prions; however, studies in transgenic animals indicated that misfolding or mistargeting of PrP can induce neuronal cell death even in the absence of prion propagation (reviewed in Winklhofer et al. 2008). How PrPSc or neurotoxic PrP mutants cause neurodegeneration remains elusive (reviewed in Hunter 2006), but different studies indicate that neuronal expression of GPI-anchored PrPC is required to mediate neurotoxic effects of prion propagation (Brandner et al. 1996; Mallucci et al. 2003; Chesebro et al. 2005). On the other hand, PrPC expression can alleviate the toxic effects of PrPΔHD, a PrP mutant devoid of the internal hydrophobic domain (HD) (Shmerling et al. 1998; Baumann et al. 2007; Li et al. 2007). This activity of PrPC could be related to its physiological function. Employing stroke models in mice and rats it was demonstrated that wt PrPC has a neuroprotective activity after an ischemic insult (McLennan et al. 2004; Shyu et al. 2005; Spudich et al. 2005; Weise et al. 2006; Mitteregger et al. 2007). This stress-protective activity was dependent on the octarepeat region located in the unstructured N-terminal domain of PrPC (Mitteregger et al. 2007). Recently, we have identified two novel domains linked to the stress-protective activity of PrP: the internal HD and the C-terminal GPI anchor. Based on a functional characterization of the HD it appeared that the stress-protective signaling of wt PrPC is linked to dimer formation with the HD as part of the dimer interface (Rambold et al. 2008). Based on these and other studies in cultured cells it seems plausible to assume that PrPC can confer enhanced survival under stress (reviewed in Flechsig and Weissmann 2004; Roucou and LeBlanc 2005; Westergard et al. 2007).

Mature PrPC can be internalized to be degraded in a lysosomal compartment or to recycle to the plasma membrane (Shyng et al. 1993). Internalization of PrPC is mediated mainly by the unstructured N-terminal domain (Shyng et al. 1995; Nunziante et al. 2003; Sunyach et al. 2003; Kiachopoulos et al. 2004); however, the precise mechanism is controversial. Two internalization pathways, via coated pits or caveolae-dependent structures, have been described (Shyng et al. 1994; Marella et al. 2002; Peters et al. 2003; Sunyach et al. 2003).

Down-regulation of cell surface localized PrPC by stimulating endocytosis could be a strategy to interfere with prion propagation. Experiments in scrapie-infected hamsters indicated that suramin, a polysulfonated naphthylurea derivative which has been used for the treatment of African trypanosomiasis (Dressel and Oesper 1961), can prolong the incubation time (Ladogana et al. 1992). Subsequent studies provided a mechanistic explanation for the anti-prion activity of suramin; it induces a conformational transition of PrPC into a detergent-insoluble conformation. This conformational transition induces rapid internalization, mediated by the unstructured N-terminal domain, and subsequent intracellular degradation of PrPC in a lysosomal compartment (Kiachopoulos et al. 2004). As a consequence, PrPSc propagation was inhibited in scrapie-infected mouse neuroblastoma (ScN2a) cells and the onset of prion disease was significantly delayed in scrapie-infected mice (Gilch et al. 2001).

Down-regulation of cell surface PrPC by stimulating endocytosis was also described for copper (Pauly and Harris 1998; Sumudhu et al. 2001). Indeed, it turned out that copper, similarly to suramin, has an impact on the conformation of PrPC at the cell surface and interferes with the propagation of PrPSc in scrapie-infected cells (Kiachopoulos et al. 2004). What might be the functional relevance of this phenomenon? Internalization of detergent-insoluble conformers of PrPC and their subsequent degradation in a lysosomal compartment might function as a cellular quality control mechanism to eliminate non-native, presumably non-functional PrPC conformers from the cell surface. The copper-induced conformational transition of PrPC might act as a stimulus for the internalization of copper-bound PrPC, suggesting a role in copper homeostasis or signal transduction. Thus, the conformational plasticity of PrPC might not only confer susceptibility to misfolding and conversion into a pathogenic conformer, it could also be an immanent feature associated with the physiological function of PrPC.

In this study, we show that epigallocatechin gallate (EGCG) and gallocatechin gallate (GCG), the major polyphenols in green tea, induce rapid transition of PrPC into a detergent-insoluble conformation. As a consequence, EGCG- or GCG-treated cells are depleted of mature PrPC localized at the cell surface; these cells are protected against PrPSc propagation, however, they are more vulnerable to stress-induced cell death.

Material and methods

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Reagents, antibodies, and plasmids

Green tea extracts (−)-GCG (G6782), (−)-EGCG (E4143), and (−)-epicatechin gallate (ECG; E3893), black tea extract (T5550) and caffeine (C0750), cupric sulfate (CuSO4) as well as NH4Cl were purchased from Sigma (St Louis, MO, USA), MG132, kainate, and bafilomycin A1 from Calbiochem (San Diego, CA, USA). For the incubation of cells, green tea extracts, CuSO4, kainate, and NH4Cl were dissolved in bidistilled H2O, bafilomycin A1, and MG132 in dimethylsulfoxide. The following antibodies were used: anti-PrP 3F4 monoclonal antibody (Kascsak et al. 1987), anti-PrP antiserum A7 (Winklhofer et al. 2003), anti-heat-shock protein 70 antibody N27 (kindly provided by William J. Welch), anti-β actin antibody (Sigma), and anti-Active Caspase 3 antibody (Promega, Madison, WI, USA). The plasmid [pcDNA3.1Zeo(+); Invitrogen, Carlsbad, CA, USA] encoding for wt PrP contains the 3F4 epitope and was described previously (Gilch et al. 2001). Notably, the 3F4 antibody recognizes endogenous human PrP (SH-SY5Y cells) and the 3F4-tagged mouse PrP (wt PrP) used for transfection. Anti-PrP antiserum A7 recognizes human and murine PrP.

Cell culture and transfection

SH-SY5Y, N2a, and ScN2a cells were cultivated as described earlier (Winklhofer and Tatzelt 2000). Cells were transfected by a liposome-mediated method using Lipofectamine Plus reagent (Invitrogen) according to the manufacturer’s instructions.

Detergent solubility assay, proteinase K digestion of cell lysates, and western blot analysis

As described previously (Tatzelt et al. 1996), cells were washed twice with phosphate-buffered saline (PBS), scraped off the plate, pelleted by centrifugation, and lysed in cold detergent buffer (0.5% Triton X-100 and 0.5% sodium deoxycholate in PBS). The lysate was centrifuged at 15 000 g for 20 min at 4°C. After boiling in Laemmli sample buffer, proteins present in the supernatant and pellet fraction were subjected to sodium dodecyl sulfate/polyacrylamide gel electrophoresis (SDS/PAGE). Following SDS/PAGE, proteins were transferred onto a nitrocellulose membrane and analyzed by immunoblotting as previously described (Tatzelt et al. 1996). For PK digestion, PK (20 μg/mL) was added to the supernatant and the resuspended pellet. After 40 min at 25°C, the digestion was terminated by the addition of phenylmethylsulfonyl f (PMSF) and residual PrP detected by western blotting. Quantifications were performed using aida quantification software (Raytest, Straubenhardt, Germany) and were based on at least three independent experiments. Data were expressed as mean ± SEM. Statistical analysis was performed using Student’s t-test.

Indirect immunofluorescence assay

Indirect immunofluorescence assays were performed as described (Rambold et al. 2006). In brief, N2a cells were grown on glass coverslips and fixed with 3% para-formaldehyde for 20 min. To detect specifically cell surface PrP, antibody incubation was performed directly after fixation without permeabilization. To detect total PrP, cells were permeabilized with 0.2% Triton X-100 for 10 min at 25°C prior to antibody incubation. The primary antibody was incubated for 45 min at 37°C in PBS containing 1% bovine serum albumin. After extensive washing with cold PBS, incubation with the Cy3-conjugated secondary antibody followed at 37°C for 30 min. Cells were mounted onto glass slides and examined by fluorescence microscopy (Zeiss LSM 510 Zeiss Axioscope2 plus; Zeiss, Thornwood, NY, USA).

Apoptosis assay

Activation of caspase 3 was determined as described previously (Rambold et al. 2006). Briefly, SH-SY5Y cells were grown on glass coverslips; 24 h after transfection, cells were incubated with kainate (500 μM) for 4 h. N2a cells were cultivated on coverslips for 3 days and treated with copper (500 μM) for 5 h. The cells were then fixed with 3% para-formaldehyde for 20 min, permeabilized with 0.2% Triton X-100 for 10 min at 25°C, and blocked with 1% bovine serum albumin in PBS for 1 h at 25°C. Fixed cells were incubated with anti-active caspase 3 antibody for 16 h at 4°C, washed, and incubated with Alexa 555-conjugated secondary antibody for 1 h at 25°C. After extensive washing, cells were mounted onto glass slides and examined by fluorescence microscopy using a Zeiss Axioscope 2 plus microscope. To detect cells undergoing apoptosis, the number of activated caspase 3-positive cells out of at least 300 transfected cells was determined. Quantifications were based on at least three independent experiments. Data were expressed as mean ± SEM.

Expression, purification, and refolding of recombinant PrP(89–232)

N-terminally truncated recombinant mouse prion protein (rPrP) was cloned into the pTXB3 vector and expressed in Escherichia  coli BL21(DE3) RIL (Stratagene, La Jolla, CA, USA) as a fusion protein, N-terminally attached to GyrA intein from Mycobacterium Xenopii and a chitin-binding domain separated by a His tag. Protein expression, purification, and cleavage of the fusion proteins was carried out as described before (Olschewski et al. 2007). The denatured protein was diluted to a final concentration of 0.1 mg/mL into a buffer containing 0.6 M l-arginine, 50 mM Tris–HCl, pH 8.6, 5 mM GSH, and 0.5 mM GSSG and kept for 12 h at 4°C. To remove l-arginine and the redox reagent GSSG, the solution was dialyzed against a buffer containing 50 mM Tris–HCl at pH 7.5. The protein solution was concentrated by ultrafiltration to yield solutions with concentrations of rPrP between 27 and 50 μM.

CD spectroscopy

Far-UV circular dichroism (CD) spectra (200–250 nm) were recorded on a Jasco J-715 spectropolarimeter (Jasco, Easton, MD, USA) in a 0.1 cm cuvette. For measurements, the concentration of rPrP was adjusted to 0.2 mg/mL (12.5 μM) in 50 mM Tris–HCl (pH 7.5).

In vitro aggregation assay and proteinase K digestion

Folded rPrP (50 μM) was dialyzed against urea-buffer consisting of 3 M urea, 1 M Gdn-HCl, 150 mM NaCl, and 20 mM Na-phosphate buffer at pH 6.8. Samples were incubated at 37°C for 3 days. Aggregation was tested by PK digestion. Digestion was carried out using an rPrP to enzyme ratio of 50 : 1 at 37°C in 100 mM Tris–HCl, pH 7.5. After 1 h, the reaction was stopped by adding PMSF. The efficiency of proteolysis was analyzed by SDS/PAGE and western blotting.

ITC measurements

Isothermal titration calorimetry was carried out on an automated VP-ITC device (MicroCal, Northampton, MA, USA). The ITC cell was filled with 1.8 mL of a 27 μM rPrP solution and titrated in step of 8 μL with a 0.5 mM solution of inhibitor at various temperatures. The time interval between injections was 1400 s. Control experiments were carried out using 27 μM rPrP in the cell and buffer in the syringe and 0.5 M EGCG in the syringe and buffer in the cell. The data sets were evaluated using the software package Origin-ITC (MicroCal, Northampton, MA, USA).

Results

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Epigallocatechin gallate depletes cells of PrPC and interferes with the formation of PrPSc

The present study was designed to identify new small molecules that would interfere with the propagation of PrPSc. The screen was based on a collection of natural compounds with a documented activity to interfere with amyloid fibril formation in vitro (Porat et al. 2006). We used scrapie-infected N2a (ScN2a) cells as a model to study propagation of PK-resistant PrPSc and infectious scrapie-prions in rodent cells (Butler et al. 1988; Caughey and Raymond 1991). A fast and robust method to determine the relative amount of PrPSc is the detergent solubility assay. It relies on the fact that upon lysis of ScN2a cells in detergent buffer the pathogenic, partially PK-resistant isoform PrPSc partitions into the detergent-insoluble phase (P), while PrPC is found in the detergent-soluble fraction (S) (Fig. 1a) (Tatzelt et al. 1996; Winklhofer and Tatzelt 2000). Thus, the relative amount of PrPSc and PrPC present in cells can be analyzed in parallel. With this approach we identified EGCG, the main polyphenol in green tea, and its stereoisomer GCG as potent drugs to interfere with the formation of PrPSc in ScN2a cells. Caffeine or black tea extracts had no impact on PrPSc formation. The relative amounts of other cellular proteins such as heat-shock protein 70 were not reduced by EGCG or GCG treatment (Fig. 1b, ScN2a). However, the reduction of PrPSc was paralleled by a decrease in PrPC present in the soluble (S) fraction. This result indicates that the anti-PrPSc activity of EGCG and GCG is most likely because of a depletion of PrPC, the precursor of PrPSc. To test this possibility, we included uninfected N2a cells in our analysis. Indeed, incubation with EGCG and GCG led to a decrease in the level of endogenous PrPC (Fig. 1b, N2a).

image

Figure 1.  Epigallocatechin gallate and gallocatechin gallate derived from green tea interfere with the formation of PrPSc. (a) PK-resistant PrP partitions in the detergent-insoluble fraction. ScN2a cells were lysed in detergent buffer and fractionated by centrifugation. The fractions were treated with PK (20 μg/mL, 40 min, RT) or left untreated (PK -) and PrP present in the detergent-soluble (S) and -insoluble fraction (P) was analyzed by western blotting using the anti-PrP antiserum A7. (b) N2a and ScN2a cells were grown for 3 days. Gallocatechin gallate (GCG), epigallocatechin gallate (EGCG), epicatechin gallate (ECG), black tea extract, or caffeine (100 μM each) dissolved in water were added to the cell culture medium and cells were cultivated for additional 48 h. Cells were lysed in detergent buffer and fractionated by centrifugation. PrP present in the detergent soluble (S) and insoluble fraction (P) was analyzed by western blotting using the anti-PrP antiserum A7. As a control, heat-shock protein (HSP70) present in the supernatant was detected by the mAb N27. (c) Molecular structure of the tested components from green tea extract. GCG: (−)-gallocatechin gallate, EGCG: (−)-epigallocatechin gallate (EGCG) and ECG: (−)-epicatechin gallate. The hydroxyl groups (bold) and the gallate side chain (boxed) required for the anti-PrPSc activity are marked. (d) N2a and ScN2a cells were subjected to increasing concentrations of ECG and EGCG, as indicated. PrP was analyzed as described under (b). As a control, HSP70 and actin present in the supernatant were analyzed by western blotting.

Download figure to PowerPoint

The activity of EGCG and GCG is dependent on three hydroxyl groups of the trihydroxyphenyl side chain

The initial results indicated a lower activity of ECG, a compound closely related to EGCG and GCG (Fig. 1c). To investigate this phenomenon in more detail, we tested the activity of EGCG and ECG in a concentration-dependent manner. While EGCG interfered with PrPSc accumulation efficiently at 50 μM, a comparable effect was seen for ECG only at 200 μM (Fig. 1d). These findings indicate that the additional hydroxyl group of the trihydroxyphenyl side chain, and possibly its presence at the meta position, increases the activity of EGCG to induce cellular depletion of PrPC. In addition, the gallate side chain (Fig. 1c, boxed) is essential. Epicatechin and gallocatechin, which both lack the gallate side chain, did not interfere with PrPSc propagation (data not shown), corroborating previous results (Kocisko et al. 2003).

EGCG induces a transition of PrPC into a detergent-insoluble conformation followed by lysosomal degradation

The apparent decrease in mature PrPC indicated that EGCG either interferes with biogenesis of PrP, or induces its degradation. To gain insight into the mechanism, we first analyzed transiently transfected cells over-expressing 3F4-tagged PrP (Gilch et al. 2001). Similarly to endogenous PrPC, ectopically expressed PrP was depleted from the detergent-soluble fraction upon EGCG treatment. However, the over-expressed PrP now appeared in the detergent-insoluble fraction of EGCG-treated cells (Fig. 2a). Obviously, EGCG induced a conformational transition of PrP into a detergent-insoluble conformation. To identify the cellular localization of the misfolded PrP in EGCG-treated cells, we performed indirect immunofluorescence experiments. In non-permeabilized EGCG-treated cells, PrP was not detectable, indicating that the detergent-insoluble PrP detected in the pellet fraction by western blotting was present intracellularly. To obtain further experimental evidence, cells were permeabilized prior to antibody incubation; PrP was now found intracellularly in a pattern reminiscent of vesicular staining (Fig. 2b). Based on our previous studies employing different anti-prion compounds, it seemed plausible that intracellular PrP in EGCG-treated cells is present in lysosomal vesicles (Gilch et al. 2001; Kiachopoulos et al. 2004). We therefore treated untransfected N2a cells with EGCG and the lysosomal inhibitors bafilomycin or NH4Cl. Under these conditions, PrP was detectable in the EGCG-treated N2a cells (Fig. 2c, Bafilomycin, NH4Cl). A possible explanation for the effect that detergent-insoluble PrP is only detectable in EGCG-treated cells transfected with PrP is that the proteolytic capacity of the lysosomal compartment is overburden by the high over-expression of PrP. Notably, inhibition of the proteasome by MG132 did not stabilize PrP in EGCG-treated cells (Fig. 2c, MG132). These data indicate that EGCG induces a conformational transition of PrPC at the plasma membrane followed by internalization and subsequent lysosomal degradation.

image

Figure 2.  EGCG treatment leads to lysosomal degradation of PrPC. (a and b) N2a cells were transiently transfected with 3F4-tagged PrP. (a) Cells were treated with increasing concentrations of EGCG and PrP present in the detergent soluble (S) and insoluble (P) fractions was detected by western blotting (left panel) using the mAb 3F4. Quantified data of at least three independent experiments shows the percentage of PrP present in the detergent-soluble (S) and -insoluble (P) fractions (right panel). Plotted is the mean ± SEM; **< 0.001. (b) Cellular localization of PrP after incubation with 100 μM EGCG and ECG for 3 h was visualized by indirect immmunofluorescence in permeabilized (perm) and non-permeabilized (non-perm) cells using the mAb 3F4. Scale bar represents 35 μm. (c) Uninfected N2a cells were grown for 5 days and treated for 16 h with EGCG (100 μM) or H2O (100 μM) and the inhibitors MG132 (5 μM), bafilomycin (100 nM), NH4Cl (30 mM), or dimethylsulfoxide (DMSO) as control. Endogenous PrP present in the detergent-soluble (S) and -insoluble (P) fractions was detected by western blotting using the anti-PrP pAb A7 (left panel). The relative amount of detergent-insoluble PrP in EGCG-treated cells was quantified from at least three independent experiments (right panel). The amount in DMSO-treated cells was set as 1. Plotted is the mean ± SEM; *< 0.05.

Download figure to PowerPoint

EGCG binds to natively folded PrP and induces the formation of random coil structure

To test for a direct interaction of PrP with EGCG, isothermal titration calorimetry experiments were performed using purified PrP (rPrP90–232) expressed in and purified from E. coli (Olschewski et al. 2007) (Fig. 3a). The long delay of 1400 s between injections was required to achieve full thermodynamic equilibrium of the measured system. Such long equilibration times are characteristic of a binding event associated with concomitantly occurring conformational changes. The binding curve for EGCG to rPrP90–232 can be fitted by assuming a model with two binding sites with different affinities. In the first step one molecule of EGCG binds to one molecule of rPrP with a KD of 130 nM and a ΔH of −43 kJ (Fig. 3a). The second phase shows a much smaller enthalpy change. The real nature of this biphasic behavior is not clear. It could be either explained by conformational transitions of rPrP triggered by higher concentrations of the inhibitor or by multiple binding sites for EGCG. The analysis of full-length rPrP23–232 did not reveal additional binding sites (see Supporting information). It should be noted that aggregation processes can be excluded. Based on gel filtration analysis only monomeric rPrP90–232 could be detected even at high EGCG concentrations. Enthalpy changes of the latter phase encompass about 1/3 of that of the initial binding event. Contrary to the results obtained for EGCG, no changes in enthalpies for the titration with ECG were observed (Fig 3a, inset). To exclude any interference of heat capacity effects that could prevent the detection of changes in enthalpy, the measurements were also carried out at 25°C and 15°C (data not shown). At all temperatures, no binding of ECG to rPrP90–232 was observed confirming the lack of interaction of ECG with rPrP.

image

Figure 3.  Binding of EGCG to PrP induces unfolding and the formation of PK-sensitive aggregates. (a) ITC curves of rPrP90–232 with GCG and ECG (inset) at 37°C. The concentration in the syringe was 0.5 mM of EGCG and ECG, respectively. The substances were added in 8 μL steps to a 27 μM solution of rPrP. The delay between injections was set to 1400 s to allow full equilibration of the system. (b) Far-UV CD spectra of rPrP90–232 in the absence (dashed line), in presence of 5 eq of EGCG (solid line) or 5 eq ECG (triangles) in 20 mM Tris–HCl, pH 7.5. Far-UV CD spectra of EGCG (135 μM) in 20 mM Tris–HCl, pH 7.5 (dots). (c) Far-UV CD spectra of rPrP90–232 without EGCG (black line) and after incubation with 5 eq of EGCG for 5 min (red line), 15 min (dashed black line), and 1 h (dashed red line). (d) Analytical size exclusion chromatography of rPrP90–232 after 2 h incubation with 5 molar equivalents of EGCG (solid line). The dashed line shows a molecular weight standard (HMW; Biorad, Hercules, CA, USA). (e) rPrP was incubated in Tris buffer (top panel) or in aggregation buffer (3 M urea and 1 M Gdn-HCl) (in vitro aggregation) for 3 days. In parallel samples in vitro aggregation was carried out in the presence of EGCG or ECG. The samples were centrifuged and the supernatant (S) and the pellet (P) was treated with PK or left untreated (PK−). Residual PrP was detected by western blotting using the anti-PrP antiserum A7.

Download figure to PowerPoint

The effect of EGCG binding on the secondary structure of rPrP90–232 was analyzed by recording Far-UV CD spectra. Figure 3b displays the CD spectrum of rPrP taken after 2 h incubation in the presence and absence of 5 molar equivalents (eq) EGCG as well as in the presence of 5 eq ECG. In the absence of EGCG the typical CD spectrum of rPrP90–232 showing a mainly α-helical tertiary structure is observed (Fig. 3b, dashed line) (Nandi et al. 2002). However, in the presence of EGCG the recorded CD spectrum displays a curve with a maximum at ∼220 nm, which can be interpreted as a random coil structure (Fig. 3b, solid line). The CD spectrum of only EGCG and ECG, respectively did not show a significant signal between 210 and 250 nm at the concentration used here. However, below 210 nm strong absorption of both molecules led to saturation of the detector and prevented measurements in this wavelength range.

To resolve the structural transition in a time-dependent manner CD spectra of rPrP90–232 incubated with EGCG were recorded at different time points. Before addition of EGCG freshly prepared rPrP90–232 (27 μM) displays a predominantly α-helical structure with two characteristic minima at 222 nm and 208 nm identical to published spectra (Fig. 3c). The addition of 5 eq of EGCG leads to an initial increase in β-sheet structures as indicated by the appearance of a minimum at 218 nm after 5 min of incubation. This intermediate undergoes further structural transitions over time, which are characterized by a decrease of the minimum at 218 nm after 15 min and 1h (Fig. 3c) and the appearance of a maximum at ∼220 nm after 2 h (Fig. 3b). These changes can be interpreted as a loss of secondary structure elements. Similar results were observed for rPrP23–232 (Supporting information). Even though a slight change in the CD spectrum of rPrP is also induced by the addition of ECG (Fig. 3b, triangles), a random coil structure is not observed.

The time course of this transition was followed by the change of ellipticity at 222 nm (data not shown). This data indicate that the major conformational change occurs in the first 20 min after addition of EGCG, whereas for ECG almost no changes were observed over time. These observations are in good agreement with the above described measurements after 2 h incubation with EGCG and ECG, respectively. Aggregation and/or oligomerization of rPrP had not occurred at this stage as proven by size exclusion chromatography (Fig. 3d). In this experiment rPrP90–232, which had been incubated for 2 h with EGCG was analyzed by analytical gel filtration using a Sephadex S-200 (Amersham Biosciences, Piscataway, NY, USA) column. The resulting chromatogram clearly showed that EGCG-treated rPrP90–232 remained monomeric with an apparent molecular weight of ∼14 kDa (Fig. 3d).

EGCG interferes with the formation of proteinase K-resistant aggregates

The influence of EGCG and ECG on the in vitro aggregation of rPrP was tested with a PK resistance assay. The in vitro aggregation of rPrP90–232 was accelerated by transferring the protein into a buffer containing 3 M urea, 1 M Gdn-HCl, 150 mM NaCl, and 20 mM NaPi at pH 6.8 (Bocharova et al. 2005) and was followed by testing the sensitivity of rPrP to PK. To determine the effect of EGCG and ECG on the formation of PK-resistant aggregates 5 eq of each compound were added. All samples were incubated under constant shaking at 600 rpm at 37°C for 3 days. (Fig. 3e). After 3 days all samples were centrifuged and separated into supernatant (S) and pellet (P). PK was added to all fractions for 1 h and the digestion was quenched by addition of 2 mM PMSF prior to western blot analysis (Fig. 3e). With the exception of untreated rPrP (Fig. 3e, top panel), which did not precipitate, in all other samples rPrP90–232 was solely found in the pellet. However, the resistance of these aggregates against a limited proteolytic digestion differed depending on the presence or absence of EGCG. Untreated rPrP90–232 was almost quantitatively converted into PK-resistant material by this procedure (Fig. 3e, second panel). In contrast, the rPrP90–232 aggregates formed in the presence of 5 eq of EGCG were not PK-resistant (Fig. 3e, third panel). The effect of ECG on the formation of PK-sensitive rPrP90–232 aggregates was not as pronounced as for EGCG, even though a reduction in the amount of PK-resistant aggregates could also be observed (Fig. 3e, fourth panel). These results suggest structural differences of rPrP90–232 aggregates formed in vitro in the absence or presence of EGCG. Indeed, preliminary results indicate that EGCG interferes with the formation of ordered fibrils; instead amorphous aggregates are formed (data not shown).

EGCG decreases the stress-protective activity of PrPC

Studies in cultured cells, mice and rats revealed that PrPC expression confers increased viability to neuronal cells under stress conditions (reviewed in Westergard et al. 2007). To monitor the protective activity of PrPC, we exposed human neuroblastoma SH-SY5Y cells to the excitotoxin kainate, an established inducer of apoptosis. SH-SY5Y cells have been previously used to monitor apoptotic cell death (Rambold et al. 2006, 2008) and express extremely low levels of endogenous PrPC. This phenomenon is illustrated in Fig. 4a. First, we established that kainate does not interfere with the EGCG-induced conformational transition of PrP into a detergent-insoluble conformation. As expected, detergent-soluble PrPC was depleted from the EGCG-treated cells, instead PrP was now found in the detergent-insoluble fraction. Note that GPI-anchored green fluorescent protein (GFP) (GFP-GPI) remained in the detergent-soluble phase (Fig. 4b). To monitor apoptosis transiently transfected SH-SY5Y cells expressing either wt PrP or GPI-anchored GFP were exposed to kainate for 4 h and apoptotic cell death was analyzed by indirect immunofluorescence using an anti-active caspase 3 antibody (Fig. 5b). Indeed, expression of wt PrP significantly decreased apoptotic cell death induced by kainate (Fig. 5a, kainic acid). However, the protective effect of PrPC over-expression against kainate-induced apoptosis was significantly reduced in cells exposed to EGCG: over-expressing PrPC cells, pre-treated with EGCG, were as sensitive to kainate as transfected cells expressing GFP-GPI (Fig. 5a, kainic acid + EGCG).

image

Figure 4.  (a) SH-SY5Y cells express low levels of endogenous PrPC. Post-nuclear supernatants of untransfected or wt PrP transfected SH-SY5Y and N2a cells were analyzed by western blotting using the mAb 3F4 or the pAb A7. A7 detects both endogenous mouse PrP and transfected 3F4-tagged PrP, while mAb 3F4 does not recognize endogenous mouse PrP in untransfected N2a cells. Because of low expression levels the 3F4 or the A7 antibody could not detect endogenous PrPC in SH-SY5Y cells. (b) Kainic acid does not interfere with the formation of detergent-insoluble PrP in EGCG-treated cells. SH-SY5Y cells were transiently co-transfected with EYFP and GFP-GPI or wt PrP. 24 h after transfection cells were pre-incubated with EGCG (50 μM, +EGCG) or water (−EGCG) for 1 h prior to exposure to kainic acid (500 μM, 4 h). Cells were lysed in detergent-buffer and GFP-GPI and PrP present in the detergent-soluble (S) and insoluble (P) fraction were analyzed by western blotting.

Download figure to PowerPoint

image

Figure 5.  EGCG interferes with the stress-protective activity of PrPC. (a and b) SH-SY5Y cells were transiently co-transfected with EYFP and GFP-GPI or wt PrP; 24 h after transfection cells were pre-incubated with EGCG (10 μM, +EGCG) or water (−EGCG) for 2 h prior to exposure to kainic acid (KA, 500 μM, 3 h). Induction of apoptosis was visualized by indirect immunofluorescence using an anti-active caspase 3 antibody (b) (scale bar represents 50 μm) and the percentage of apoptotic cells was quantified (a). (c) N2a cells were grown for 3 days and treated with EGCG (100 μM for 2 days) or EGCG (100 μM for 2 days) and CuSO4 (500 μM for 5 h). Apoptosis was analyzed as described in (a). Endogenous PrP was visualized by western blot using the pAb A7 (right panel). Data represent at least three independent experiments. Shown is the mean ± SEM; *< 0.01; **< 0.001.

Download figure to PowerPoint

To test whether down-regulation of endogenous PrPC increases susceptibility to stress we turned to N2a cells, which express high levels of PrPC. It has been reported that PrPC protects primary neurons and N2a cells against copper-induced toxicity (Brown et al. 1998; Vassallo et al. 2005). Thus, we employed copper-induced toxicity as a stress paradigm to monitor if N2a cells exposed to EGCG are more sensitive to stress. Cultivation in the presences of CuSO4 (500 μM) for 5 h had no adverse effect on the viability of N2a cells (Fig. 5c, left panel, CuSO4). However, pre-treatment with EGCG let to a depletion of endogenous PrPC (Fig. 5c, right panel) and sensitized N2a cells to copper-induced apoptosis (Fig. 5c, left panel, EGCG + CuSO4).

Discussion

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

In this study, we show that the major polyphenols of green tea, EGCG and GCG, interfere with cell surface expression of mature PrPC by converting natively folded PrPC into detergent-insoluble conformers, which are rapidly degraded intracellularly. As a consequence, cells are protected against PrPSc propagation; however, they are more vulnerable to stress-induced apoptosis.

PrPC as a target to interfere with the formation of PrPSc

The most potent way to prevent prion diseases is to ablate PrPC (Büeler et al. 1993). Our study indicated that the anti-PrPSc activity of EGCG and GCG is based on a similar mechanism. In ScN2a as well as in uninfected N2a cells, the relative amount of mature PrPC was observably reduced after incubation with EGCG. The effect occurred fast, 16 h after onset of treatment mature PrPC was hardly detectable by western blotting (Fig. 1b). On the other hand, the effect was only transient. As soon as EGCG was omitted from the cell culture medium, mature PrPC re-populated the plasma membrane (data not shown). Based on our results with PrPC-over-expressing cells and with inhibitors of lysosomal or proteasomal degradation, we would like to propose a model in which EGCG and GCG induce a conformational transition of mature, natively folded PrPC into detergent-insoluble conformers, which are rapidly degraded in lysosomal compartments (Fig. 2b and c). This observation further supports the notion that a conformational change of cell surface PrPC can result in rapid internalization (Pauly and Harris 1998; Gilch et al. 2001; Quaglio et al. 2001; Sumudhu et al. 2001; Kiachopoulos et al. 2004). EGCG-treated cells are depleted from mature PrPC, consequently, the precursor for the formation of PrPSc is missing and propagation of PrPSc ceases.

In vitro experiments with purified components support our model. We could show that EGCG directly binds to recombinantly expressed and refolded PrP (rPrP). Binding of EGCG destabilized the natively folded rPrP and promoted the formation of random coil structure. The random coil conformers formed in the presence of EGCG were initially monomeric and in a second step formed PK-sensitive aggregates.

Further studies are now aimed at identifying the domain(s) of PrP that interact with EGCG. Obviously, the unstructured N-terminal domain of PrP is dispensable, because the rPrP used in our study lacks amino acid 23–90. To test for a possible therapeutic value of EGCG, bioassays in scrapie-infected mice or hamsters have to be performed. Fortunately, the low toxicity of EGCG will allow prolonged treatment at relatively high concentrations. In addition, the pharmacokinetic and pharmacodynamic properties of EGCG are promising, in particular as it has been shown to cross the blood–brain barrier (Chow et al. 2003; Lin et al. 2007). Furthermore, a protective activity of EGCG for other protein misfolding disorders has been demonstrated in animal models of Alzheimer’s and Huntington’s disease (Rezai-Zadeh et al. 2005; Ehrnhoefer et al. 2006).

Loss of mature PrPC is accompanied by enhanced vulnerability to stress

Based on its inhibitory effect on PrPSc propagation, EGCG seems to be a molecule with beneficial activities. However, our studies also revealed that the effect of EGCG on the conformation of PrPC could have adverse effects. This became apparent when we analyzed the viability of cells exposed to stress. Several studies in transgenic mice and cell culture models indicated that PrPC protects cells from stress-induced apoptosis (reviewed in Flechsig and Weissmann 2004; Roucou and LeBlanc 2005; Westergard et al. 2007). Using kainate- and copper-induced apoptosis as a stress paradigm, we corroborated these findings and showed that expression of wt PrPC decreased apoptotic cell death of SH-SY5Y cells exposed to kainate. This protective effect of PrPC expression was significantly reduced by pre-treatment with EGCG, indicating that the detergent-insoluble conformers of PrP found in EGCG-treated cells lack this physiological activity of mature PrPC. Similarly, down-regulation of endogenous wt PrPC increased cell death of EGCG-treated N2a cells exposed to copper. It will now be interesting to see if the internalization of detergent-insoluble PrP conformers, similarly to those induced by suramin, copper or EGCG, might be implicated in the regulation of the physiological function of PrPC.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

We thank Veronika Müller for experimental assistance. This work was supported by grants from the Deutsche Forschungsgemeinschaft (SFB 596), from the Bayerische Staatsminister für Wissenschaft, Forschung und Kunst (for Prion, MPI3), the Bundesministerium für Bildung und Forschung (01KO0110, BioDisc, DIP), NeuroPrion and the Max Planck Society.

References

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Fig. S1 Binding of EGCG to full-length rPrP.

Appendix S1 Binding of EGCG to full-length rPrP.

Please note: Blackwell Publishing are not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.

FilenameFormatSizeDescription
JNC_5611_sm_appendixS1.doc72KSupporting info item
JNC_5611_sm_Suppfig1.tif6253KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.