International Journal of Peptide Research and Therapeutics, Vol. 13, No. 3, September 2007 (Ó 2007) pp. 413–421 DOI: 10.1007/s10989-006-9055-y

Characterization of a Peptide that Specifically Blocks the Ras Binding Domain of p75 Silvia Egert,1,2 Heike Piechura,1 Nina Hambruch,1 Martin Feigel,2 and Andrea Blo¨chl1,3 (Accepted November 8, 2006; Online publication February 7, 2007)

The neurotrophin receptor p75 interacts with the GTPase Ras. Unstimulated it inactivates Ras while ligand binding induces Ras activation. We developed an inhibitory peptide (ip75RBD) which interferes with the binding domain of Ras of the intracellular domain of p75. ip75RBD inhibits the binding of Ras to the receptor in vitro. It is membrane-permeable and inhibits ligand-induced Ras activation via p75 in vivo but does not influence Ras activation by the stimulated receptor tyrosine kinases Trk and the epidermal growth factor receptor EGFR. The activation of the neutral sphingomyelinase by stimulated p75 is slightly delayed but not inhibited by the peptide. p75-mediated neuronal death induced by NGF or aggregated beta-amyloid1–42 is reduced. We conclude that ip75RBD specifically blocks the Ras binding site of p75 and can be used to analyze p75-induced Ras signaling.

KEY WORDS: p75; Ras; inhibition.

discussion (reviews: Gentry et al., 2004; Bronfman and Fainzilber, 2004; Yamashita et al., 2005). We recently demonstrated an activation of the Ras/Erk pathway upon stimulation of p75 with NGF (Blo¨chl et al., 2004) or aggregated Ab (Susen and Blo¨chl, 2005) and could show that Ras directly interacts with the receptor’s intracellular domain. The unstimulated receptor promotes the inactivation of Ras and lowers the level of Ras-GTP within the cell whereas receptor stimulation elevates the level of Ras-GTP (Blo¨chl et al., 2004). p75-mediated effects range from neurotrophic to apoptotic events, depending on cell type and cell condition. Ras activation has been linked to trophic effects but may also contribute to the induction of apoptosis (via Rac). A specific inhibitor of Ras activation by p75 would be an appropriate tool to investigate the role of Ras in p75 signaling and to find out how Ras participates in trophic and possibly in apoptotic effects of p75. As peptides are widely used to inhibit receptors or enzymes we used this approach to construct a suitable inhibitor. For p75 peptides have been

INTRODUCTION Neurotrophins like nerve growth factor (NGF) are essential for development and synaptic plasticity of the nervous system of vertebrates. Their action is mediated by two different receptor types: the Trk receptor family of tyrosine kinase receptors which bind their ligands with high affinity and selectivity, and p75 which belongs to the superfamily of tumor necrosis factor receptors and binds all neurotrophins and also other ligands like aggregated beta-amyloid1–42 (Ab) and prion Prp62. While Trk signaling is well characterized (reviews: Huang and Reichardt, 2003; Teng and Hempstead, 2004), signaling via p75 is still under 1

2

3

Faculty of Chemistry, Biochemistry II, Ruhr-University Bochum, NC7/132, 44780, Bochum, Germany. Naturstoffchemie, Ruhr-University Bochum, NC7/132, 44780, Bochum, Germany. Correspondence should be addressed to: Andrea Blo¨chl, Faculty of Chemistry, Biochemistry II, Ruhr-University Bochum, NC7/ 132, 44780, Bochum, Germany. Tel.: +49-234-3226208; e-mail: [email protected]

413 1573-3149/07/0900–0413/0 Ó 2007 Springer Science+Business Media, LLC

414 designed that bind to the intracellular domain of the receptor. Rudert et al. (1998) and Ilag et al. (1999) used the selective phage infection method to create a peptide library. One of the peptides found by Ilag and coauthors interferes with the interaction of p75 and Rho (Yamashita et al., 2003). For our purpose, we created a peptide library for the interaction of p75 and Ras. The starting point was a mastoparanlike site within the intracellular domain of p75 (Feinstein and Larhammar, 1990; Chapman, 1995; Dostaler et al., 1998). Mastoparan is known to interact with G-proteins and small GTPases (Weingarten et al., 1990; Koch et al., 1991; Klinker et al., 1996; Kusunoki et al., 1998) and to influence nucleotide exchange (Daniel et al., 2002) and velocity of GTP hydrolysis (Kusunoki et al., 1998) by GTPases. The binding of mastoparan to the trimeric Gi1a protein has been mapped to the C-terminal fifth helix of Gi1a (Tanaka et al., 1998). Our investigation yielded a peptide with high affinity to p75 which reduces the interaction of p75 with Ras in vivo and in vitro and blocks the p75/Ras-mediated outgrowth of neurites.

Egert et al. (1:200): Molecular Probes (Leiden, Netherlands). All other agents were obtained from Sigma (Munich, Germany) if not indicated else.

Constructs and Protein Purification DNA sequences of p75 intracellular domain (p75icd) were amplified by polymerase chain reaction (PCR) using the following primers: p75icd (GAGGTGGAACAGCTGCAAAC) together with p75C (GTTCACACTGGGGATG TGGCA) subcloned into the SmaI-site of pGEX-2T (Amersham Corp., Bucks, UK) to obtain p75icd. The p75 mutants p75icd (intracellular domain of p75), p75icdY337F (intracellular domain with tyrosine Y337 exchanged for phenylalanine), p75icdY366F (intracellular domain with tyrosine Y366 exchanged for phenylalanine) and p75icdFF (intracellular domain with both tyrosines exchanged for phenylalanine) were subcloned into the BamH1-EcoR1-sites of the vector pGex-2T and used as described previously (Blo¨chl et al., 2004). All constructs were verified by sequencing. The pGEXRas-binding domain (RBD) was a kind gift of C. Herrmann. Glutathione S transferase (GST) fusion proteins were affinitypurified with glutathione-sepharose 4B (Amersham Corp., Buckinghamshire, Great Britain) according to the provider in buffer A (50 mM Tris, pH7.4, 2 mM MgCl2, 100 mM NaCl, 2 mM DTT) and further purified by gel-filtration using a Superdex75HR column (Amersham Corp., Buckinghamshire, Great Britain); for Src kinase assays, the GST part was cleaved off with thrombin prior to gel-filtration.

MATERIALS AND METHODS

Peptide Library, Peptide Synthesis and Screening

Materials

Based on the known interface of Gi1a and mastoparan we found by molecular modelling a sequence potentially aiming at p75-Ras interaction. Starting from this sequence, a peptide library has been established using spot-synthesis on cellulose membranes (according to Blankemeyer-Menge and Frank, 1988; Frank, 1992; Eichler et al., 1989) using Fmoc-protected amino acids. This library was screened for best binding to the intracellular domain of p75 (p75icd) using p75icd as bait, and p75LNTR antibody (Santa Cruz) for detection of bound p75-protein (see Fig. 1). The peptide with highest binding was synthesized on a preparative scale by use of semi-automatic solid-phase Fmoc strategy, using Fmoc-Ala-Wang-Resin for the solid phase. Prior to the first acylation, the solid phase was swollen in dichlormethan for 1 h and the Fmoc protection group was cleaved of using 20% piperidine in dimetyhlformamide.

Dulbecco’s modified Eagle’s medium, fetal calf serum, glutamine, penicillin/streptomycin, and trypsin-EDTA were obtained from PAA Laboratories (Co¨lbe, Germany), 2.5S NGF from Alomone (ICS, Munich, Germany), Ab from Bachem (Weil, Germany), active c-Src from Biosource (Camerillo, USA) and Fmocand Boc-Amino acids, HBTU, HoBt, PyBOP from Nova Biochem (Darmstadt, Germany). Antibodies, used in the indicated dilutions, were obtained from the following sources: anti-p75 MC192 (800 ng/ml): Roche Diagnostics (Mannheim, Germany); anti-p75 (polyclonal, 1:400): Santa Cruz (Heidelberg, Germany); anti-pan Ras AB3 (1:1000): Chemicon (Hofheim, Germany); anti-phosphoERK (1:2000): New England Biolabs (Ipswich, USA) anti-mouse horse radish peroxidase (HRP; 1:10000) and anti-rabbit HRP (1:8000): Dianova (Hamburg, Germany), and phalloidin-Alexa 488

Fig. 1. Filter screening of the peptide library. Filter membranes with different peptides were incubated with p75icd protein. Bound protein was detected using p75 antibody and anti-rabbit antibody coupled to alkaline phosphatase. Several peptides gave positive signals but the most prominent signal could be detected with peptide E2 (arrow).

Inhibition of p75/Ras interaction Acylations were performed with 4 equiv (relative to the solid phase) of the Fmoc-protected amino acids, 3,8 equiv of PyBOP, 4 equiv of HOBt, and 8 equiv of DIPEA. Each acylation was monitored by the ninhydrin test. The peptide was cleaved from the resin by dichlormethan/trifluoro acetic acid (TFA) (50:50). After hydrogen fluoride removal, the peptide were dissolved in TFA and precipitated with ether. The peptide was purified with an acetonitrile/water (0.01% TFA) gradient on a reverse-phase semipreparative nucleosil 120-5-C18 RP HPLC column. Peptide purity and identity were confirmed by analytical HPLC, ESI and MALDI-spectrometry and 1H- and 13CNMR spectroscopy.

Cell Culture, Stimulation of Cells, Ras Pulldown and Western Blot We used PCNA cells which stably express p75 under the control of the CMV promoter but do not express any other neurotrophin receptor. The cells were grown in a monolayer in DMEM, 10% FCS, 2 mM glutamine. For experiments, cells were seeded on dishes (TPP, Trasadingen, Switzerland) at a density of 15000–20000 cells per cm2 and used for experiments on the second day in culture. 8 h prior to experiments, the medium was exchanged for serum-free confined medium (B18; Brewer et al., 1993) without antibiotics. Cells were stimulated according to Susen et al. (1999) by adding NGF (final concentration: 100 ng/ml, dissolved in medium) or aggregated Ab1–42 (final concentration: 25 nM) to the cells without changing the medium to minimize stress. Ab1–42 was aggregated as described previously (Susen and Blo¨chl, 2005). After stimulation cells were placed on ice, washed with ice-cold PBS and lysed in Boehringer lysis buffer (50 mM Tris–HCl at pH 7.4, 150 mM NaCl, 40 mM NaF, 5 mM EDTA, 1 mM Na3VO4, 1% (v/v) Nonidet P40, 0.1% (w/v) sodium deoxycholate, 0.1% (w/v) sodium dodecyl sulfate (SDS), 1 mM phenylmethylsulfonyl fluoride) for 10 min (4°C) on a shaker. Primary cerebellar neuronal cultures were obtained from 2 days old rats as described previously (Susen et al., 1999); at this age these neurons express p75 and TrkB but not TrkA, the specific tyrosine kinase receptor for NGF. These cells, too, were used on the second day in culture. Ras pulldown experiments were performed as described in detail by De Rooij and Bos (1997). Briefly, 800 lg of protein were incubated with 10 lg of Ras binding domain of Raf fused to glutathione S-transferase and pulldown was performed with 20 ll of glutathione-sepharose (Amersham Corp., Buckinghamshire, Great Britain). After washing of the precipitates five times in the lysis buffer (100 mM NaCl, 50 mM Tris–HCl, 2 mM MgCl2, 1 mM DTT, 10% Glycerol, 1% Nonidet-p40) the precipitates were dissolved in Laemmli buffer, boiled for 5 min and loaded on a 12% SDS-PAGE. Activated Erk1/2 was analyzed by Western blot using equal amounts of protein (15 lg/slot). The nitrocellulose membranes were blotted with anti-phospho Erk1/2 antibodies, and—after stripping off the antibodies—reblotted with antibodies raised against total Erk1/2. Band intensity was analyzed by densitometric measurement (using the program TINA 2.09, Raytest, Germany) as described before (Susen et al., 1999). Interaction of Ras with p75icd was measured like Ras pulldown described above, using p75icd fused to glutathione S-transferase as bait.

415 The activity of sphingomyelinase was measured by enzymatic cleavage of NBD-C6-sphingomyelin. 25 lg of cell lysates were incubated for 1 h with 3 lM NBD-C6-sphingomyelin in a buffer containing 25 mM Tris, pH7.4 and 10 mM MgCl2. Lipids were extracted with a CHCl3:CH3OH (2:1) mixture and separated via thin layer chromatography using a CHCl3:CH3OH:NH3 mixture of 70:30:3. Quantification of the spots was performed with the programm TINA 2.09 (Raytest, Sprockho¨vel, Germany). For the in vitro phosphorylation of p75 with the Src kinase c-Src, purified p75 protein (10 lg each) was incubated with c-Src (9  10)3 U per reaction) at 30°C for 30 min in a buffer containing 20 mM Tris pH 7.4, 10 mM MgCl2, 5 mM MnCl2, 1 mM DTT, 1 mM EGTA, 500 lM pervanadate and 100 lM ATP. The reaction was stopped by addition of Laemmli sample buffer. Immunostaining of the cytoskeleton to evaluate neurite outgrowth was performed as described previously (Blo¨chl et al., 2004) by using Alexa-488-coupled phalloidin. Terminal deoxynucleotidyl transferase-mediated UTP nick and labelling (TUNEL) assays were carried out for neurons in paraformaldehyde fixed using the detection kit from Boehringer (Mannheim, Germany) according to the manufacturer’s description.

RESULTS AND DISCUSSION Peptide Design and Screening The Ras/Erk pathway is an essential component of p75 signaling, and we intended to design an inhibitory peptide that specifically blocks its induction. The starting sequence of the peptide library (NH2–A1A2I3I4V5D6T7V8A9D10F11V12F13V14A15A16 –COOH) created to this purpose was based on the interaction of mastoparan and Gi1a described by Tanaka et al. (1998). We knew that Ras directly interacts with the p75 receptor and that phosphorylation of the receptor at tyrosine residue Y366 is necessary for Ras activation while phosphorylation of Y337 regulates the phosphorylation of Y366 (Blo¨chl et al., 2004; see below). Direct interaction with Ras is not known for receptor tyrosine kinases like the Trk receptors. We presumed that a possible binding of p75 to Ras would take place at the fifth helix which resembles mastoparan (Feinstein and Larhammar, 1990; Chapman, 1995; Dostaler et al., 1998). Previous studies showed that this helix is essential for the interaction of Ras with the receptor (Blo¨chl et al., 2004). Blocking this site might interrupt the signaling of activated p75 to Ras and Erk. Computer simulations hinted at a salt bridge between D6 and D10 of the initial peptide sequence with R403, R402 and R407 of p75icd, which would be in agreement with the hypothesized interaction of trimeric Gi1a and mastoparan (Tanaka et al., 1998).

416 We optimized the binding of the peptide to p75 using p75icd as bait; best sequences were further analyzed in Ras pulldown experiments (see below). Screening of the peptide library revealed the necessity of a core sequence (V5D6T7V8A9D10F11) for p75 binding (Fig. 1). Exchange of A9 for D and F11 for A increased the binding affinity. The resulting peptide VDTVDDA, which was named ip75RBD, also showed optimal binding (Fig. 1). The peptide was then synthesized in high quantities, purified by HPLC and characterized by 1D–1H–NMR analysis. The analysis of the peptide by ROESY-spectrum suggests that part of the peptide might temporarily adopt an a-helical conformation, as some NOEs of the type dNN(i, i + 1) are observed (Fig. 2). A weak NOE signal between the amide NH of D6 and the NH of D10 is perhaps the result of an intermolecular aggregation. ip75RBD Inhibits the Interaction of Ras and p75icd in vitro If the designed peptide binds to p75 at the same site as Ras it should inhibit the interaction of Ras and p75. In a pulldown assay, we analyzed the binding of the GST-tagged p75icd or (as a positive control) GSTtagged Ras binding domain of Raf to H-RasGppNHP or H-Ras-GDP. The two proteins were

Fig. 2. Characterization of the ip75RBD sequence by NH/NH ROESY correlation. The sequence NOE signals between NHi and NHi + 1 are clearly visible indicating a partial helical conformation.

Egert et al.

incubated in equimolar concentrations (100 nM) for 30 min. Prior to the incubation with Ras-GDP, p75icd was incubated with ip75RBD in concentrations ranging from 1 nM to 100 nM. As shown previously (Blo¨chl et al., 2004) p75 preferentially interacts with the GDP-bound form of Ras. This interaction is significantly inhibited by the peptide in all used concentrations suggesting a high binding affinity of the peptide to p75icd. For unknown reasons, this inhibition was lower for a concentration of 10 nM than for 1 nM and 100 nM (Fig. 3). One hypothesis concerning the interaction of p75icd and the peptide is that one or more of three arginines protruding from p75 (Liepinsh et al., 1997) might constitute the binding site of both Ras and the peptide. Arginines play a major role in the binding of Ras to other proteins. For instance, the interaction of Ras with Raf-kinase, Ras-GAP or Ras-GEF depends on arginines (Browbridge et al., 1993; Fabian et al., 1994; Scheffzek et al., 1997); mutations of these arginine residues cause the loss of Ras binding. In Cell Cultures ip75RBD Inhibits the p75-Mediated Ras/Erk Pathway but not the EGF-Induced Ras/Erk Pathway or the p75-Mediated Activation of nSMase The next step was the analysis if the peptide is capable of interfering with p75-induced signaling in cell cultures. First, the uptake of the peptide into cells was verified by coupling ip75RBD with fluorescein and visualizing intracellular peptide by confocal microscopy. The peptide seems to be taken up by the cell and partly co-localizes with the receptor p75 (data not shown).

Fig. 3. ip75RBD inhibits binding of Ras-GDP to p75icd. Binding of Ras-GDP (100 nM) and p75icd (100 nM) was detected in vitro using p75icd fused to glutathione S-transferase as a bait. Rasbinding domain (RBD) from Raf (100 nM) was used as a positive control. While Raf interacts with Ras-GTP, p75icd interacts with Ras-GDP. This interaction is reduced by ip75RBD already at a concentration of 1 nM.

Inhibition of p75/Ras interaction The data obtained in vitro suggest a selective inhibition of the Ras/Erk pathway. To verify this interference, we pre-incubated cells for 2 h with the peptide (1 nM–100 nM) and performed Ras pulldown experiments with the lysates of unstimulated cells and cells stimulated for 2.5 min with 100 ng/ml NGF. At this time maximal Ras activation has been observed in PCNA cells (Susen and Blo¨chl, 2005). The peptide blocked p75-induced Ras activation significantly (Fig. 4a). The activation of Erk1/2 induced by stimulation of p75 with 100 ng/ml NGF was also significantly reduced by pretreatment of the cells for 2 h with 10 nM ip75RBD; the inhibitory effect of the peptide was stable over 8 h (Fig. 4b).

417 We also analyzed a possible influence of the peptide on the signaling of other receptors that activate Ras. The peptide did not affect the activation of Ras or Erk1/2 via the epidermal growth factor receptor (and receptor tyrosine kinase) EGFR (Fig. 5) or via stimulation of the neurotrophin TrkB receptor (see next paragraph). p75 also activates the neutral sphingomyelinase (nSMase) and causes hydrolysis of sphingomyelin to ceramide. To verify that the peptide probably interferes only with the Ras binding domain of p75 and not with the juxtamembrane domain, we analyzed the hydrolysis of sphingomyelin in PCNA cells stimulated with NGF (100 ng/ml) for 2–30 min. NGF induced a time-dependent increase in ceramide production peaking at 2–5 min, which was slightly delayed but not diminished by pretreatment with the peptide (Fig. 6). We therefore presume that signaling pathways of p75 that are mediated by the juxtamembrane domain essentially are not affected by ip75RBD. We conclude that ip75RBD specifically inhibits activation and inactivation of Ras by p75 but does not interfere with Ras activation by other receptors and only slightly modifies another basic signaling pathway of p75 (the activation of nSMase). ip75RBD Does not Inhibit Src kinase-Mediated Phosphorylation of p75icd

Fig. 4. Ras/Erk activation upon NGF stimulation of p75 is blocked by ip75RBD in vivo. (a) PCNA-cells (expressing p75 but not TrkA) were stimulated with NGF (100 ng/ml) for 2 min. Cultures were pretreated 2 h prior to stimulation with ip75RBD in the given concentrations. The amount of activated Ras-GTP was measured in the lysates by pulldown experiments using the Ras binding domain of Raf as bait. The histogram represents a quantification of Ras-GTP levels from 5 experiments given in percent of the unstimulated control without peptide and NGF. (b) Activation of Erk1/2 was measured by Western blot in PCNA cells stimulated for 7 min with NGF. ip75RBD (10 nM) was given up to 8 h prior to stimulation. The histogram represents phosphorylated to unphosphorylated Erk1/2 from 5 experiments given as ratios of stimulation-induced activity to basic activity in unstimulated cells (in percent).

If ip75RBD indeed interacts with the RBD of p75 then the inactivation of Ras by the unstimulated receptor should be inhibited as well, and indeed we observed a significant increase of activated Ras in unstimulated treated cells pretreated with ip75RBD (Fig. 4a). Inactivation of Ras by the non-stimulated receptor takes place through replacing GTP by GDP due to a slightly higher affinity of the receptor to the Ras-GDP form (Blo¨chl et al., 2004). The tyrosine residue Y337 of p75ICD is a key element for both inactivation of Ras by the unstimulated receptor and activation of Ras by the stimulated receptor. The observed suppression of Ras inactivation might therefore signify that ip75RBD blocks Y337 instead of the RBD of p75. The phosphorylation of Y337 upon p75 stimulation controls the phosphorylation of Y366 which is necessary for Ras activation (Blo¨chl et al. 2004). If ip75RBD really blocked Y337 then it should prevent phosphorylation of Y337 and Y366 in vitro. To investigate this, we first identified the kinase that phosphorylates the two tyrosine residues of p75ICD.

418

Egert et al.

Fig. 5. Ras/Erk activation after stimulation of the receptor tyrosine kinase EGFR is not reduced by ip75RBD. (a) THUVEC cells were stimulated with EGF for 2 min; ip75RBD was given at different concentrations 30 min prior to stimulation. Level of Ras-GTP was measured by pulldown experiments with the Ras binding domain of Raf as bait. The histogram represents Ras-GTP levels from 5 experiments given as ratios of stimulation-induced activity to basic activity in unstimulated cells (in percent). (b) Erk1/2 phosphorylation was measured in THUVEC cells after stimulation with EGF for 7 min. ip75RBD (10 nM) was given up to 24 h prior to stimulation. The histogram represents phosphorylated to unphosphorylated Erk1/2 from 5 experiments given as ratios of stimulation-induced activity to basic activity in unstimulated cells (in percent).

Fig. 6. p75-induced activation of neutral sphingomyelinase is not blocked by ip75RBD. PCNA-cells (expressing p75 but no TrkAreceptor) were stimulated with NGF (100 ng/ml) for up to 30 min. Cultures were pretreated 2 h prior to stimulation with ip75RBD in the given concentrations. Activity of neutral sphingomyelinase was measured by the amount of hydrolyzed NBD-C6 sphingomyelin to NBD-C6 ceramide. The histogram represents sphingomyelinase activities from 5 experiments given as ratios of stimulation-induced sphingomyelinase activity to basal sphingomyelinase activity in unstimulated cells (in percent).

Since p75 does not express an intrinsic tyrosine kinase, we hypothesized that a member of the family of Src kinases might phosphorylate the two tyrosines within the intracellular domain of the receptor after ligand binding. Preincubation of PCNA cells for 30 min with the specific Src kinase inhibitor PP2 (Hanke et al., 1996) but not with its inactive analogue PP3 abolished Ras activation induced by stimulation of the cells with either NGF (100 ng/ml) or aggregated Ab1–42 (Fig. 7a) for up to 5 min. p75icd, but not a protein consisting of the death domain of p75 (p75dd), could be phosphorylated in vitro by purified c-Src (Fig. 7b). There is no unspecific phosphorylation if both tyrosines of the intracellular domain have been exchanged for phenylalanine (p75FF). c-Src could not phosphorylate Y366 if Y337 had been mutated to phenylalanine (p75Y337F) whereas Y337 was also phosphorylated if Y336 had been mutated to phenylalanine . ip75RBD added in great excess of p75icd (up to 100 fold) did not inhibit

Inhibition of p75/Ras interaction

419

Fig. 7. Y337 is phosphorylated by Src kinases and regulates Ras activation via p75. (a) PCNA cells were stimulated with NGF (100 ng/ml) or aggregated Ab(25 nM) as indicated, and activated Ras was precipitated with RBD-GST as bait (PD: RBD). The precipitates and total lysates were analysed in SDS-PAGE and Western Blot using anti-Ras antibodies for detection. Stimulation-induced Ras activation was clearly reduced by application of PP2 but not by its inactive analogue PP3. The histogram shows the average of up to six densitometrically analyzed experiments. (b) Equal amounts (10 lM) of purified p75icd and its mutants p75Y337F, p75Y366F and p75FF were incubated with active c-Src kinase, and phosphorylation of p75 was detected in Western blots using anti-phospho-tyrosine antibodies. p75icd was only phosphorylated when Y337 was present. (c) p75icd at a concentration of 10 lM was mixed with the indicated amounts of ip75RBD and the mixture was incubated for 30 min at 25°C. Active c-Src kinase was added and phosphorylation of p75icd was detected in Western blots using anti-phosphotyrosine antibodies. Even a 100fold excess of the peptide did not reduce the amount of ip75icd phosphorylation by c-Src kinase.

the phosphorylation of p75icd (Fig. 7c) but even slightly enhance it. Therefore ip75RBD does not block Y337. These data, together with the pulldown experiments of p75icd and Ras in vitro and our data obtained from cell cultures, make it very probable that ip75RBD binds to the RBD of p75. ip75RBD Blocks p75-Induced but not TrkB-Induced Neurite Outgrowth and Inhibits p75-Mediated Apoptosis NGF induces neurite outgrowth in cerebellar neurons of newborn rats through activated p75 (Blo¨chl et al., 2004; Susen and Blo¨chl, 2005). This outgrowth is at least partly due to the receptor’s ability to activate Ras. NGF stimulation of neuronal cultures pretreated with ip75RBD no longer caused

the formation of new primary and secondary neurites. In addition, growth cone formation was reduced (Fig. 8). Like NGF, BDNF also induces the outgrowth of primary and secondary neurites in cerebellar neurons but this neurite formation largely depends on TrkB stimulation since inhibition of TrkB by K252A largely suppresses this outgrowth (data not shown). ip75RBD did not inhibit BDNF-induced neurite outgrowth; consequently, ip75RBD does not block TrkB-induced Ras activation (Fig. 8). These results clearly demonstrate the dependence of p75induced neurite formation on Ras activation. In addition we performed TUNEL assays in cerebellar neurons using NGF and aggregated Ab1–42 for the induction of apoptosis. Both p75 ligands induce apoptosis after prolonged stimulation (>48 h) but not in cells pretreated with ip75RBD (Fig. 8).

420

Egert et al.

Fig. 8. NGF-mediated neurite outgrowth in cerebellar neurons from newborn rats is blocked by ip75RBD. Neurons were stimulated for 2 min and 24 h with NGF (100 ng/ml); for a control BDNF (10 ng/ml), which stimulates the receptor tyrosine kinase TrkB was given for 24 h. The number of short (<2 lm) and long primary and secondary processes was counted. In addition, the number of filopodia per growth cone was determined. The data are given as percentage of the according numbers in the control groups without inhibitor and neurotrophins. Effects of NGF treatment are inhibited by pretreatment of the cells with ip75RBD (10 nM, added 2 h prior to neurotrophins) while BDNF effects are not reduced or even increased. The TUNNEL assay was performed 48 h after treatment with NGF (100 ng/ml) or Ab1–42 (100 nM). The inhibitor was given to the neurons 2 h prior to stimulation.

CONCLUDING REMARKS The peptide ip75RBD inhibits in vivo specifically Ras inactivation by the unstimulated neurotrophin receptor p75 and Ras activation by the stimulated p75, and abolishes in vitro the interaction of Ras and p75. As the peptide does not block the tyrosine residue Y337 of p75, which regulates inactivation and activation of Ras by p75, it can be assumed that it binds to the Ras binding site of p75 as it was designed to do. It probably does not interfere with other pathways of p75 or with Ras activation through other receptors and should therefore be an appropriate tool to analyze the role of Ras in p75 signaling, e.g. in different cell types. For instance, using this inhibitor we could show that Ras activation is indeed respon-

sible for p75-induced neurite outgrowth and growth cone formation during short stimulation with NGF and also for p75-mediated neuronal cell death after prolonged stimulation with NGF or Ab. The latter result indicates that the novel inhibitor could be useful in the research of neurodegenerative diseases especially as p75 appears to play an important role in Alzheimer’s disease (Schor, 2005).

REFERENCES Blo¨chl, A., Blumenstein, L. and Ahmadian, M. R.: 2004, Europ. J. Neurosci. 20, 2321–2335. Blankemeyer-Menge, B. and Frank, R.: 1988, Tetrahedron 44, 6031–6040.

Inhibition of p75/Ras interaction Brewer, G. J., Torricelli, J. R., Evege, E. K. and Price, P. J.: 1993, J. Neurosci. Res. 35, 567–576. Bronfman, F. C. and Fainzilber, M.: 2004, EMBO Rep. 5, 867–871. Brownbridge, G. G., Lowe, P. N., Moore, K. J., Skinner, R. H. and Webb, M. R.: 1993, J. Biol. Chem. 268, 10914–10919. Chapman, B. S.: 1995, FEBS Lett. 374, 216–220. Daniel, S., Noda, M., Cerione, R. A. and Sharp, G. W.: 2002, Biochemistry 41, 9663–9671. Dostaler, S. M., Ross, G. M., Myers, S. M., Weaver, D. F., Ananthanarayanan, V. and Riopelle, R. J.: 1998, Eur. J. Neurosci. 8, 870–879. Eichler, J., Beyermann, M. and Bienert, M.: 1989, Collect. Czech. Chem. Commun. 54, 1746–1752. Fabian, J. R., Vojtek, A. B., Cooper, J. A. and Morrison, D. K.: 1994, Proc. Natl. Acad. Sci. USA 91, 5982–5986. Feinstein, D. L. and Larhammar, D.: 1990, FEBS Lett. 272, 7–11. Frank, R.: 1992, Tetrahedron 48, 9217–9232. Gentry, J. J., Barker, P. A. and Carter, B. D.: 2004, Prog. Brain Res. 146, 25–39. Hanke, J. H., Gardner, J. P., Dow, R. L., et al.: 1996, J. Biol.Chem 271, 695–701. Huang, E. J. and Reichardt, L. F.: 2003, Annu. Rev. Biochem. 72, 609–642. Ilag, L. L., Rottenberger, C., Liepinsh, E., Wellnhofer, G., Rudert, F. and Otting, G.: 1999, Biochem. Biophys. Res. Commun. 255, 104–109.

421 Koch, G., Haberman, B., Mohr, C., Just, I. and Aktories, K.: 1991, FEBS Lett. 291, 336–340. Klinker, J. F., Laugwitz, K. L., Hageluken, A. and Seifert, R.: 1996, Biochem. Pharmacol. 51, 217–223. Kusunoki, H., Wakamatsu, K., Sato, K., Miyazawa, T. and Kohno, T.: 1998, Biochem. 37, 4782–4790. Liepinsh, E., Ilag, L. L., Otting, G. and Ibanez, C. F.: 1997, EMBO J. 16, 4999–5005. De Rooij, J. and Bos, J. L.: 1997, Oncogene 14, 623–625. Rudert, F., Woltering, C., Frisch, C., Rottenberger, C. and Ilag, L. L.: 1998, FEBS Lett. 440, 135–140. Scheffzek, K., Ahmadian, M. R., Kabsch, W., et al.: 1997, Science 277, 333–338. Schor, N. F.: 2005, Prog. Neurobiol. 77, 201–214. Susen, K., Heumann, R. and Blo¨chl, A.: 1999, FEBS Lett. 463, 231–234. Susen, K. and Blo¨chl, A.: 2005, J. Mol. Med. 83, 720–735. Tanaka, T., Kohno, T., Kinoshita, S., et al.: 1998, J. Biol. Chem. 273, 3247–3252. Teng, K. K. and Hempstead, B. L.: 2004, Cell. Mol. Life Sci. 61, 35–48. Weingarten, R., Ransnas, L., Mueller, H., Sklar, L. A. and Bokoch, G. M.: 1990, J. Biol. Chem. 265, 11044–11049. Yamashita, T. and Tohyama, M.: 2003, Nat. Neurosci. 6, 461–467. Yamashita, T., Fujitani, M., Hata, K., Mimura, F. and Yamagishi, S.: 2005, Anat. Sci. Int. 80, 37–41.