Cell-permeable peptide system for treating diseases caused by glutamate excitotoxicity

Abstract

Peptide system including at least one peptide blocking the presynaptic release of glutamate. The peptide has the sequence SEQ ID NO 5: GRKKRRQRRRPPIEQSIEQEEGLNRS and/or sequence SEQ ID NO 8: GRKKRRQRRRPPMSEYNATQSDYRER for use in treating pathologies associated with glutamate excitotoxicity.

Claims

1. A method for blocking N-Methyl D-Aspartate (NMDA)-induced presynaptic release of glutamate in a subject in need thereof, comprising administering to the subject an effective amount of a peptide system comprising at least one peptide blocking the presynaptic release of glutamate, wherein the peptide blocking the presynaptic release of glutamate comprises an amino acid sequence selected from SEQ ID NOs: 5 and 8.

2. The method according to claim 1, wherein the amino acids in the peptide are L-amino acids or D-amino acids.

3. The method according to claim 1, wherein the peptide system is administered as a medicament comprising the peptide system.

4. The method according to claim 1, wherein the peptide further comprises a tag sequence selected from the group consisting of HA, Glutathione S-Transferase (GST), Myc, and enhanced green fluorescent protein (EGFP).

5. The method according to claim 1, wherein the peptide system is administered as a pharmaceutical composition comprising the peptide system.

6. The method according to claim 1, wherein the peptide system is administered as a pharmaceutical composition comprising the peptide system, and wherein the pharmaceutical composition is in a form selected from the group consisting of solution, suspension, ointment, patch, tablet, and capsule.

Description

BRIEF DESCRIPTION OF THE FIGURES

(1) FIG. 1 shows the effect that the inhibition of the JNK protein has on the presynaptic release of glutamate, and it also shows the synaptic activation of the JNK protein under the action of different stimuli. A) shows the results regarding the synaptosomes prepared from the cortex of Wild-type mice, which were pre-loaded with the radioactive neurotransmitter and then incubated with different stimuli as indicated. The results are expressed as percentage of the induced release. The data is expressed as meanSEM (standard error of the mean) obtained from 3 experiments reproduced in triplicate (Three perfusion chambers for each experimental condition). **p<0.01 vs. basal release. B) Synaptosomes prepared from the cortex of Wild-type mice were pre-loaded with the radioactive neurotransmitter and incubated with L-JNKi1 (2 M) for 30 min before the stimuli as indicated. The results are expressed as percentage of the induced release. The data is expressed as meanSEM (standard error of the mean) obtained from 3 experiments reproduced in triplicate (Three perfusion chambers for each experimental condition). L-JNKi1 prevented the release evoked by the NMDA stimulation. **p<0.01 vs. 100% of the release evoked by NMDA. C-D-E) representative Western blots with relative quantifications that show the p-JNK level reached after 90 sec of stimulation of KCl (C), 10 min of NMDA (D), 10 min of AMPA (E) and the treatment of the synaptosomes with L-JNKi1. The results are expressed as percentage of increase of the p-JNK/JNK ratio on the control (placed at 100%). The data represents the meanSEM of 4 experiments for each condition. The treatment with NMDA induced an increase of the phosphorylation of JNK **p<0.01 vs. 100% of the control) while JNKi1 does not reduce the effect of the NMDA (*p<0.05 vs. 100% of the control).

(2) FIG. 2 shows the functional role of JNK in the release of glutamate dependent on NMDA. A) depicts the histograms (meanS.E.M) of the paired-pulse ratio (PPR) of evoked EPSCs that are registered by 7 neurons from slices of mouse of the strain C57BL6/J and by 10 neurons of slices of mouse C57BL6/J pre-incubated with L-JNKi1 (2 M). On the left, the representative traces were obtained from the same neurons, in control condition, during the administration of D-AP5 (50 M) and after the washing of D-AP5. B) shows the cumulative distribution of the amplitude of the mEPSC (left) and of the inter-event intervals (right) recorded from single neurons of C57BL6/J, or of neurons of C57BL6/J pre-incubated with L-JNKi1 in response to the D-AP5. The top traces were obtained from the same neurons, in control conditions, during the administration of D-AP5 and subsequent washing. C) depicts the histograms (meanS.E.M) of the paired-pulse ratio (PPR) of evoked EPSCs which are recorded from 11 neurons of slices of mouse of the strain C57BL6/J and from 10 neurons of slices of mouse C57BL6/J pre-incubated with L-JNKi1 (2 M) in response to the D-AP5. **p<0.01 (t-test).

(3) FIG. 3 shows that the JNK controls the phosphorylation of Syntaxin 1a and that JNK2/3 co-immunoprecipitates together with Syntaxin 1a. A) shows the representative Western blot results with respective quantifications of the phosphorylation of STX1a during the treatment as shown. The results are expressed as a percentage and are normalized with the control. The data is the meanSEM of 5 experiments in which 10 min of NMDA+L-JNKi1 reduced the relation pSTX1a/STX1a (**p<0.01 vs. 100% of the control) as did L-JNKi1 on its own (**p<0.01 vs. 100% of the control). B) shows the representative Western blot results with the respective quantifications that show the levels of the phosphorylation of JNK during the treatment with D-AP5. The results are expressed as percentage with respect to the control (100%). The data is the meanSEM of 3 experiments. C) shows the representative Western blot results with respective quantifications that show the levels of the phosphorylation of STX1a during the treatment with D-AP5. The results are expressed as percentage and are normalized with the control. The data represents the meanSEM of 3 experiments. D) shows the results of the experiments of co-immunoprecipitation (Co-IP) with relative quantifications. Cortex synaptosomes of wild-type mice were stimulated as indicated in the figure and the immunoprecipitation executed with STX1a antibody. Representative Western blot results show Co-Ip of p-JNK and JNK after stripping procedure. The membranes were blotted for STX1a (as control of the immunoprecipitation) and for SYT1 (as control of the specificity). The results are expressed as percentage of increase of the relation p-JNK/STX1a or JNK/STX1a with respect to the control condition. The data is the meanSEM of 4 experiments for each treatment (*p<0.05, **p<0.01 vs. 100% of the control).

(4) FIG. 4 shows how the NMDA-evoked release of glutamate and the NMDA-induced phosphorylation of JNK are inhibited in the JNK2 KO mice.

(5) A) shows the results of the glutamate release induced by NMDA stimulation in synaptosomes from the cortex of wild-type and JNK-KO mice pre-loaded with radioactive tracer. The results are expressed as a percentage of induced release. The data is the meanSEM of 4 experiments conducted in triplicate (3 perfusion chambers for each experimental condition). **p<0.01 vs. WT; p<0.05 vs. JNK1-KO or JNK3-KO. B) shows the representative Western blot results with quantifications that show the ratio pJNK/JNK in the JNK2-KO or JNK3-KO mice during the treatments. The results are expressed as percentages with respect to the control (100%). The data represents the meanSEM of 3 experiments (**p<0.01 vs. 100% of control).

(6) FIG. 5 shows how the isoform JNK2 is involved in the NMDA-mediated release of glutamate. A) represents the cumulative distribution of the amplitude of the mEPSC (left) and the inter-event intervals (right) recorded from single neurons of isoform JNK KO mice in response to D-AP5 (50 M). The top lines are obtained from the same neurons, in control conditions, during the administration of D-AP5 and subsequent washing. B) represents the histograms (meanS.E.M) of the means of the amplitude values of mEPSC (left) and inter-event intervals (right) with n=8 JNK1 KO, n=7 JNK2 KO and n=6 JNK3 KO neurons in response to the D-AP5. **p<0.01 t-test.

(7) FIG. 6 shows the model of molecular docking of the protein interaction JNK-STX1a. A,B,C) represent the three best docking possibilities obtained from the study of the interaction between STX1a and JNK2 respectively identified as models 1, 2 and 3. The JNK2 protein is represented by the group 100, while the N-terminal portion of STX1a is represented by the group 200. D) shows the structure of the N-terminal portion of STX1a: the sections 300, 400, and 500 are shown which represent the portions that, with highest probability, interact with JNK2. The rendering of all the images was carried out by means of the use of VMD software.

(8) FIG. 7 shows that the treatment with L-JNKi1 does not induce changes in the NMDA and AMPA receptor subunits (A), nor does it induce changes on the proteins of the release mechanism (B), neither during NMDA stimulation conditions nor if applied on its own.

(9) FIG. 8 shows the effect of the two peptides JGRi1 and JGRi2 on the release of glutamate in wild-type mouse cortex slices in electrophysiological patch-clamp experiments. More in detail, the figure in question shows the histograms (meanS.E.M) of the means of the amplitude values of the inter-event intervals with n=9 neurons where JGRi1 and JGRi2 were applied together or n=4 neurons for JGRi1 and n=6 neurons for JGRi2. **p<0.01 t-test.

(10) FIG. 9 shows an example of a solid-phase chemical synthesis process for peptides. The symbols indicate: little circle, group of protection for the amino acid chain; R, protection for the amino group; X, Y, Z, L, functional groups.

(11) The process is cyclically repeated until the peptide is attained.

(12) FIG. 10 shows the chromatogram obtained from the HPLC analysis of the peptide JGRi1 having sequence SEQ ID NO. 5: GRKKRRQRRRPPIEQSIEQEEGLNRS.

(13) FIG. 11 shows the chromatogram obtained from the HPLC analysis of the peptide JGRi2 having sequence SEQ ID NO. 8: GRKKRRQRRRPPMSEYNATQSDYRER.

(14) FIG. 12 shows the mass spectrum of the peptide JGRi1 having sequence SEQ ID NO. 5: GRKKRRQRRRPPIEQSIEQEEGLNRS.

(15) FIG. 13 shows the mass spectrum of the peptide JGRi2 having sequence SEQ ID NO. 8: GRKKRRQRRRPPMSEYNATQSDYRER.

DETAILED DESCRIPTION OF THE INVENTION

(16) The present invention is based on the demonstration of the presence of the JNK isoforms (JNK1, JNK2, JNK3) at the STX1a presynaptic level. Taking this protein interaction as the starting point, two peptides were designed: cell-permeable JGRi1 and JGRi2, which are capable of blocking this JNK2-STX1a interaction and reducing the release of glutamate induced by the stimulation of the NMDA receptor.

(17) Experimental Data

(18) c-Jun N-Terminal Kinase (JNK) Regulates the Presynaptic Release of Glutamate

(19) The physiological presence of JNK at the presynaptic level was demonstrated by executing the glutamate release experiments. Due to this experiment type, it was possible to study the role of the JNK kinases in the presynaptic compartment. Isolated cortical terminals (synaptosomes) of mouse strain C57/BL6 were used. These synaptosomes pre-loaded with tritiated D-aspartate ([.sup.3H] D-Asp) were subjected to glutamate release induced by different stimuli. The synaptosomes were stimulated with a method that has been used for years, namely superfusion, in which the synaptosomes are continuously washed by a continuous current solution, and at a certain point they are stimulated. The tritiated glutamate that is liberated in this procedure is exclusively released due to presynaptic intracellular events and once collected is distributed into the beta-emitting radiation counters. The transient exposure of the synaptosomes (90 s) to a slight depolarization stimulation (8 mM KCl) induced a significant exocytotic release of [.sup.3H] D-Asp (15837% vs. control **P<0.01), which is comparable to that obtained from 10 minutes of exposure to 100 M NMDA (20221% vs. control. **p<0.01). The NMDA stimulation was applied in a medium lacking Mg.sup.2+ in order to ensure the effectiveness of the presynaptic NMDA receptor. In the same manner, a stimulation of 10 min of 50 M (S)-AMPA was capable of giving a significant increase of release of [.sup.3H] D-Asp (9811% vs. control. **P<0.01) (FIG. 1A). All the applied stimuli are known for triggering the release of neurotransmitter in a Ca.sup.2+-dependent manner. Then, in order to study possible presynaptic roles of JNK, a specific inhibitor (L-JNKi1) was used which blocks the JNK signaling cascade, preventing JNK interaction with the target proteins that bind the JNK binding domain (JBD). L-JNKi1 is a commercial cell-permeable peptide with which cortical synaptosomes were then treated for 30 minutes at 2 M concentration. It is interesting to observe that only the release of NMDAstimulated glutamate was strongly inhibited (6010% vs. control=100%. **P<0.01) by the L-JNKi1 (FIG. 1B), indicating that there is a specificity of the JNK signaling cascade in the modulation exclusively of the NMDA-evoked release of glutamate.

(20) The Activation of JNK at the Presynaptic Button Level

(21) The above-reported results indicate for the first time the presence and hence the functionality of JNK in the presynaptic button. In addition, it is clear that the JNK are capable of regulating the release of glutamate induced by the activation of the NMDA receptor. In order to confirm the presynaptic position of the JNK and their possible activation, biochemical experiments were carried out on cortical synaptosomes of mice, measuring the phosphorylation induced by the above-listed stimuli (FIG. 1C-E). The phosphorylation of JNK was increased following 10 minutes of exposure to 100 M NMDA in Mg.sup.2+-free medium (+6318% vs. control=100%. **P<0.01). Since L-JNKi1 does not affect the phosphorylation of JNK, even when it has been pre-incubated, the applied stimulus of NMDA continues to produce an increase of the phosphorylated JNK (+4910% vs. control=100%. *P<0.05). Conversely, the JNK basal phosphorylation remained unchanged when L-JNKi1 was applied on its own, indicating that this substance has no effect in a basal system (FIG. 1D). The other two depolarizing stimuli, AMPA and KCl, are instead unable to increase the phosphorylation of JNK, indicating that it is the NMDA stimulus that specifically modulates the activity of JNK (FIG. 1C-E).

(22) JNK Mediates the Presynaptic Release of NMDA-Dependent Glutamate

(23) The presynaptic NMDA receptors on the glutamatergic terminals are activated by the spontaneously-released glutamate and, by means of a positive feedback mechanism, determine a tonic liberation of glutamate. Electrophysiological recordings were made in order to investigate the functional role of JNK in the release of glutamate. For this reason, a parameter termed paired-pulse facilitation (PPF) was measured from pyramidal neurons of cerebral slices of entorhinal cortex (EC)obtained from adult mice C57BL6/J incubated with or without the inhibitor of JNK, L-JNKi1 (2 M for 30 min). Such paired-pulse facilitation (PPF) is a paradigm studied by means of the electrophysiological recording technique in patch-clamp configuration.

(24) In this case, the perfusion with the antagonist of the NMDA receptor, D-AP5 (50 M), irreversibly reduced the paired-pulse (PPR) relation obtained at 50 ms interval (**p<0.01, Student t-test ctrl vs. D-AP5 n=7; **p<0.01, Student t-test D-AP5 vs. washing n=7; FIG. 2A). When the slices were pre-incubated with L-JNKi1, the perfusion with D-AP5 did not change the PPR (p>0.05, Student t-test, ctrl vs. D-AP5, n=10; FIG. 2A). This data indicates that L-JNKi1 eliminates the probability of NMDA-dependent presynaptic release and is in line with the data of the release by synaptosomes, even in a more integral system like the cerebral slices.

(25) Subsequently, the effect of L-JNKi1 was studied on the NMDA-dependent presynaptic glutamate release by measuring the excitatory currents of miniature type (mEPSCs), according to a protocol that is already commonly used. When the D-AP5 was applied in the presence of extracellular tetrodotoxin (1 mM) and picrotoxin (100 M) together with MK-801 (10 mM), inhibitor of NMDA in the patch pipette (in order to block the postsynaptic NMDA channels in the cell being recorded), a reversible increase was seen of the distribution of the intervals between cumulative events (IEI) (***p<0.001, KS test, C57BL6/J vs. D-AP5, n=11; FIG. 2B) and of the mean values thereof (**p<0.01, Student t-test, ctrl vs. D-AP5, n=11; FIG. 2C); in addition, neither the cumulative amplitude nor the mean of the same events were affected (p>0.05, KS test, C57BL6/J vs. D-AP5, n=11; FIG. 2B, p>0.05, Student t-test, ctrl vs. D-AP5, n=11; FIG. 2C). The reduction of the frequency by D-AP5 is due to the block of the presynaptic NMDA receptors, which facilitates the release of tonic glutamate. In slices pre-incubated with L-JNKi1, the D-AP5 effect is lost. Indeed, it was discovered that the perfusion with D-AP5 did not lead to significant changes of inter-event interval (p>0.05, KS test, C57BL6/J+L-JNKi1 vs. D-AP5, n=10; FIG. 2B; p>0.05, t-test of Student, ctrl vs. D-AP5, n=10; FIG. 2C), nor of the amplitude thereof (p>0.05, KS test, C57BL6/J+L-JNKi1 vs. D-AP5, n=10; FIG. 2B; p>0.05, Student t-test, ctrl vs. D-AP5, n=10; FIG. 2C). These results suggest that the inhibition of JNK reduces the release of glutamate mediated by the presynaptic NMDA receptors.

(26) Overall, our electrophysiological data indicates that JNK performs a functional role in the NMDA-dependent release of glutamate in cerebral slices acutely treated with NMDA.

(27) JNK Regulates the Phosphorylation of STX1a

(28) Syntaxin 1a is an essential protein for the assembly of the SNARE complex (Soluble NSF Attachment Protein REceptor), which is an essential step for the release of neurotransmitter. The role of the phosphorylation of STX1a in ser14 (pSTX1a) in the neurotransmitter release mechanism is still a controversial subject. Some have reported that the reduction of pSTX1a has as direct consequence the diminution of the neurotransmitter release, while others demonstrated that the diminution of pSTX1a leads to facilitating the release of neurotransmitters.

(29) Following an exposure of NMDA for 10 min, there is an increase of the JNK phosphorylation (FIG. 1D), which causes a slight but non-significant decrease in the phosphorylation of STX1a (STX1a (S14)) (FIG. 3A). It is interesting to observe that a pre-treatment with L-JNKi1 for 30 minutes followed by a stimulation of NMDA causes a significant reduction of pSTX1a (406.4% vs. control=100%. **P<0.01). The administration of L-JNKi1 on its own, even if it had no effect on the phosphorylation of JNK (FIG. 1D), surprisingly causes a strong reduction of the phosphorylation of STX1a (456.4% vs. control=100%. **P<0.01). In order to explain this phenomenon, we can assume that, since the synaptosomes are held in batches with a medium lacking Mg.sup.2+, the neurotransmitters, including glutamate, could be accumulated in the extracellular compartment, and are thus able to activate the receptors even when L-JNKi1 is applied on its own. The NMDA antagonist D-AP5 was then applied for estimating possible changes in the pJNK (+108.9% vs. Ctrl) (FIG. 3B) or pSTX1a (107.2% vs. Ctrl) (FIG. 3C) with respect to the control conditions (100%), therefore excluding the possibility that such changes occurred. Since other proteins are known to be part of the neurotransmitter release system, the possible interaction of JNK with the phosphorylation of synapsin was analyzed.

(30) The phosphorylation of synapsin (SYN (S9)) was not affected by the stimuli caused in all experimental conditions (FIG. 3A). Overall, these results allow us to assume that the inhibition of the JNK activity specifically reduces the phosphorylation of STX1a.

(31) JNK Interacts with STX1a

(32) In order to better understand the synergistic interaction between JNK and STX1a in the NMDA-induced release of glutamate, co-immunoprecipitation (Co-IP) experiments were conducted (FIG. 3D). The proteins extracted from the synaptosomes were immunoprecipitated with an antibody specific for STX1a and then by means of Western blot pJNK and JNK were detected with specific antibodies. The control (non-immunoprecipitated sample) gave the typical JNK signal with a lower band (JNK1, 46 kDa) and an upper band (JNK2/3, 54 kDa). Surprisingly, STX1a immunoprecipitates only the JNK 2 and 3 isoforms, both as JNK and in the phosphorylated pJNK form; indeed, the antibody detected the upper band. This result clearly prevents the protein STX1a from bonding to JNK1, and suggests that the formation of the STX1a and JNK2/3 interaction is already verified in basal conditions (Ctrl). In addition, both the immunoreactivity of JNK and pJNK were strongly reduced with respect to control conditions when L-JNKi1 was applied with NMDA (fig. ratio 3D JNK/STX1a:422% vs. control=100%; **p<0.01 and ratio pJNK/STX1a:393% vs. control=100%; *p<0.05), suggesting that the protein-protein interaction is perturbed by L-JNKi1. It is interesting to observe that the STX1a-JNK2/3 interaction was reduced in non-stimulated conditions by applying L-JNKi1 (FIG. 3D JNK/STX1a:444% vs. control=100%; **p<0.01 and ratio pJNK/STX1a:468% vs. control=100%; **p<0.01), showing that the interaction is constitutively present. In order to evaluate the specificity of the immunoprecipitate, the anti-synaptotagmin 1 antibody (Syt1) was also used, which is not immunoprecipitated, indicating that the interaction STX1a-JNK2/3 is specific.

(33) In JNK2 Knock-Out Mice, the Release of Glutamate and the Phosphorylation of JNK are No Longer Induced by NMDA

(34) In order to identify the specific isoform of JNK that is activated by the NMDA stimulation in the presynaptic compartment, glutamate release experiments were conducted on cortical synaptosomes obtained from JNK knock-out (KO) mice. The release of glutamate, measured as outflow of [.sup.3H] D-Asp, was induced by NMDA applied for 10 min. It is interesting to observe that NMDA induces a release that is significantly lower, with respect to the basal (2819% with respect to the overflow basal), only in JNK2-KO mice with respect to JNK1, JNK3-KO mice and wild-type mice (WT) (FIG. 4A). This result suggests that the presynaptic JNK2 isoform in particular controls the NMDAevoked release.

(35) In order to confirm the release experiments, it was studied, on synaptosomes stimulated with NMDA, if the absence of JNK2 reduced the phosphorylation of JNK. As expected, in the JNK2-KO mice the phosphorylation of JNK was totally eliminated with respect to the control. Conversely, it is still maintained in the JNK3-KO mice where NMDA induces an increase of pJNK (+7610% with respect to the control=100%; **p<0.01) (FIG. 4B) fully comparable with that obtained in WT mice (FIG. 1D). In both cases, L-JNKi1 on its own is not able to affect the phosphorylation of JNK.

(36) JNK2 Modulates the NMDA-Dependent Release of Glutamate in Patch-Clamp Experiments

(37) Electrophysiological recordings were carried out by recording the mEPSC, using the same above-described protocol from JNK1 KO, JNK2-KO and JNK3-KO mice. The perfusion of D-AP5 induces an increase of the cumulative distribution of the inter-event intervals in JNK1-KO (***p<0.001, KS test, JNK1 KO vs. D-AP5, FIG. 5A) and JNK3 KO (***p<0.001, KS test, JNK3-KO vs. D-AP5, FIG. 5A) and an increase of the mean values (**p<0.01, Student t-test, control vs. JNK1 or JNK3 KO; FIG. 5B). In both genotypes, the amplitude of the mEPSC was never affected by D-AP5 (p>0.05, KS test, JNK1 or JNK3-KO vs. D-AP5, FIG. 5A; P>0.05, Student t-test, control vs. JNK1 or JNK3 KO; FIG. 5B), while the diminution of the glutamate release due to the application of D-AP5 has been specifically reduced in JNK2-KO mice. Indeed, D-AP5 produced no significant changes in mEPSC (p>0.05, KS test, JNK2-KO vs. D-P5; FIG. 5A; p>0.05, Student t-test, control vs. JNK2 KO, FIG. 5B), or in amplitude (p>0.05, KS test, JNK2-KO vs. D-P5, FIG. 5A; p>0.05, Student t-test, control vs. JNK2 KO, FIG. 5B). Overall, this data suggests that only the JNK2 isoform mediates the presynaptic release of glutamate in our experimental conditions.

(38) Docking Results for the JNK Variants

(39) By comparing the different sequences of the three JNK, it was observed that JNK1 provides 18 characteristic residues, JNK2 45, while JNK3 only 11. Characteristic residues are defined as those residues that are specific variants of only one isoform, being differentiated with respect to the other two JNK.

(40) The characteristic residues are important in the protein-protein interactions, and specifically if Syntaxin 1a specifically interacts with only one variant of JNK, the characteristic residues of such variant should be different with respect to the other JNKs. The fact that JNK2 is significantly different from JNK1 and JNK3 could be a first indication of its specific interaction with Syntaxin.

(41) Subsequently, simulations of protein-protein docking were executed for modeling the complex between Syntaxin 1a and JNK2 using the Rosetta software.

(42) It was observed that the three main structures deriving from the analysis are the three best-classified structures based on the interface score. A rendering image of each of these is reported in FIG. 6, in which JNK2 is represented by the group 100 and the N-terminal of Syntaxin 1a by the group (200).

(43) In order to better evaluate these structures, the hidden surface areas were calculated. In models 1 (FIG. 6a) and 2 (FIG. 6b), there is a hidden surface area of about 1700 A2, while model 3 (FIG. 6c) presents a surface area of 2680 A2. The interface residues were then analyzed and compared with the characteristic residues of JNK2. This test demonstrated that in model 3, the N-terminal of Syntaxin 1a interacts with eight characteristic residues, in model 2 with five such residues and in model 1 with only two such residues. These observations led us to interpret model 3 as the best interaction conformation obtained.

(44) A rendering image was then obtained that indicates the possible contact sites of the N-terminal portion of Syntaxin 1a (sections 300 and 400).

(45) The Inhibition of JNK does not Induce Changes in the Machinery of the Vesicular Proteins and of the Subunit Levels of the NMDA or AMPA Receptor.

(46) The expression levels were then monitored of several presynaptic proteins such as Syntaxin 1a (STX1a), synapsin (SYN), Munc18-1 and synaptotagmin 1 (Syt1), since their alteration compromises the release mechanism of the neurotransmitter. In our experimental conditions, L-JNKi1, if applied on its own or in combination with NMDA, did not modify the presence of STX1a, SYN, Munc18-1 and Syt1 in synaptosomes (FIG. 7A). In the same manner, the expression of the subunits of the NMDA and AMPA receptor (FIG. 7B) remained unchanged, thus ensuring that the NMDA-mediated effect on presynaptic JNK does not depend on specific modifications of their expression levels.

(47) Analysis and Construction of the Peptides Inhibiting the Release of Glutamate

(48) Analyzing the amino acid sequence of the N-terminal portion of Syntaxin 1a, 3 different residues can be identified which form the characteristic 3 alpha-helices; these are Ha, Hb and Hc. Hereinbelow, the amino acid sequence is shown (288 amino acids total):

(49) TABLE-US-00001 SEQIDNO.1: MKDRTQELRTAKDSDDDDDVTVTVDRDR(28)FMDEFFEQVEEIRGFID KIAENVEEVKRKHSAIL(62)ASPNPDEKT(71)KEELEELMSDIKKTA NKVRSKLKSIEQSIEQEE(104)GLNRSSA(111)DLRIRKTQHSTLSR KFVEVMSEYNATQSDYRER(144)CKGRIQRQLEITGRTTTSEELEDML ESGNPAIFASGIIMDSSISKQALSEIETRHSEIIKLETSIRELHDMFMD MAMLVESQGEMIDRIEYNVEHAVDYVERAVSDTKKAVKYQSKARRKKIM IIICCVILGIIIASTIGGIFG

(50) Ha goes from amino acid 28 to 62

(51) Hb goes from amino acid 71 to 104

(52) Hc goes from amino acid 111 to 144

(53) The two sequences that have been identified as point of interaction of JNK2 on Syntaxin 1a are the following:

(54) TABLE-US-00002 1) SEQIDNO.2: IEQEEGLNRS 2) SEQIDNO.3: MSEYNATQSDYRER

(55) These two sequences were then taken under consideration for the construction of two cell-permeable peptides JGRi1 and JGRi2.

(56) The functionality of the peptides is based on the principle of interruption of the interaction of the JNK2 protein with the N-terminal portion of Syntaxin 1a. Hence, if the cell is provided with an amino acid sequence of Syntaxin 1a of different length, that varies from the sites of minimum interaction between the two proteins or up to the entire portion of the N-terminal of Syntaxin 1a, these peptides will competitively participate in the interaction with JNK2, thus inhibiting the bond with endogenous Syntaxin 1a. This event is capable of reducing the release of glutamate. The peptides that can be considered effective vary in length starting from the entire N-terminal portion of Syntaxin 1a, 288 amino acids, to a minimum of 10 amino acids corresponding to the sequence SEQ ID NO. 2: IEQEEGLNRS or of 14 amino acids corresponding to the sequence SEQ ID NO. 3: MSEYNATQSDYRER.

(57) In order to render the peptides cell-permeable, the sequence of the Tat protein (48-59) of the HIV virus was inserted before the N-terminal portion of the present sequences, hence adding another 12 amino acids.

(58) Tat (48-59) has the following sequence of 12 amino acids: GRKKRRQRRRPP.

(59) This sequence is very effective, whether it is added in N- or C-terminal position of each of the possible synthesized peptides. One example of peptides are those synthesized in a manner so as to have a total length of 26 amino acids.

(60) The sequences of the two peptides are the following: JGRi1 (SEQ ID NO. 5): GRKKRRQRRRPPIEQSIEQEEGLNRS or (SEQ ID NO. 6) IEQSIEQEEGLNRSGRKKRRQRRRPP or even the respective mirrored sequences with regard to the tat sequence e.g. (SEQ ID NO. 7) PPRRRQRRKKRGIEQSIEQEEGLNRS JGRi2 (SEQ ID NO. 8): GRKKRRQRRRPPMSEYNATQSDYRER or (SEQ ID NO. 9) MSEYNATQSDYRERGRKKRRQRRRPP or even the respective mirrored sequences with regard to the tat sequence e.g. (SEQ ID NO. 10) PPRRRQRRKKRGMSEYNATQSDYRER

(61) The two peptides were synthesized with amino acids having L enantiomer structure, but also synthesized as D enantiomers in order to improve the stability thereof in a living organism.

(62) The same sequences can be inserted, as corresponding DNA sequences, also in different vectors for expressions in cell lines, bacteria and also in vectors for virus production.

(63) The insertion in the expression vectors can be of whole sequences, and hence as JGRi1 (SEQ ID NO. 5): GRKKRRQRRRPPIEQSIEQEEGLNRS or (SEQ ID NO. 6) IEQSIEQEEGLNRSGRKKRRQRRRPP or also the respective mirrored sequences with regard to the tat sequence.

(64) For example (SEQ ID NO. 7) PPRRRQRRKKRGIEQSIEQEEGLNRS JGRi2 (SEQ ID NO. 8): GRKKRRQRRRPPMSEYNATQSDYRER or (SEQ ID NO. 9) MSEYNATQSDYRERGRKKRRQRRRPP or also the respective mirrored sequences with regard to the tat sequence e.g. (SEQ ID NO. 5) PPRRRQRRKKRGMSEYNATQSDYRER or by adding the different domains, hence first the sequence of TAT and then that (SEQ ID NO. 2) IEQEEGLNRS or (SEQ ID NO. 3) MSEYNATQSDYRER or vice versa. In addition, in order to render the final protein recognizable in cellular systems, small pieces of DNA may be added which encode small (tag) peptides, such as HA, GST, Myc, EGFP etc.

(65) For example: pGEX-4, pTAT, p-EGFP-C1, pc-DNA, 3 generation lentiviral vectors, commercial or not also used for gene therapy, vectors for Sindbis virus, adenovirus for infection (transduction) of neurons.

(66) The two peptides were tested in glutamate release experiments (as described above) and in electrophysiology experiments (patch clamp, as described above) in order to test their effectiveness in reducing the NMD-evoked release of neurotransmitter.

(67) Both peptides resulted active in decreasing the NMDA-dependent release of glutamate in patch-clamp experiments. Mouse cortex slices were treated for 30 minutes before the recording of the cumulative distribution of the inter-events of the mEPSC with the peptides in various condition. The two peptides (JGRi1 and JGRi2), whether added together at the concentration of 5 M or separately at the concentration of 10 M, reduced the functionality of the inhibitor D-AP5 in blocking the NMDA presynaptic receptor (FIG. 8). This data indicates that the NMDA receptor is already effectively blocked by the two peptides in the intracellular pathway and hence the receptor inhibitor, which acts on the extracellular portion of NMDA, no longer has effect.

(68) The two peptides demonstrated effectiveness both on the synaptosomes and on cerebral slices, indicating that, given that they were administered in the extracellular environment, they are capable of penetrating into the cells, also diffusing into integral tissue. It is also expected that the peptides are capable of being entirely diffused in an organism. In order to reduce their degradation, and hence increase their hemi-life, the present peptide system was synthesized by using amino acids belonging to the D stereochemical series. D-JGRi1: dG-dR-dK-dK-dR-dR-dQ-dR-dR-dR-dP-dP-dI-dE-dQ-dS-dI-dE-dQ-dE-dE-dG-dL-dN-dR-dS D-JGRi2: d-G-dR-dK-dK-dR-dR-dQ-dR-dR-dR-dP-dP-dM-dS-dE-dY-dN-dA-dT-dQ-dS-dD-dY-dR-dE-dR.

(69) The possible administration routes of these cell-permeable peptides comprise all the known combinations, which range from those with immediate bioavailability to those with release modified over time.

(70) We include in the administration modes those via intraperitoneal, intramuscular, subcutaneous and intravenous injection (physiological solutions of the peptide system), transdermal absorption pathways (plaster, ointments, creams, oils, etc.), sublingual (pasty or solid solutions), intranasal (various physiological solutions) and oral (oral preparations such as suspensions, tablets, capsules, etc.) routes.

(71) In addition, all the nanotechnological formulations that produce a modified release, whether transdermal or systemic via injection or oral administration.

(72) In addition, a new administration route is considered: the production of neuronal stem cells, engineered such that they produce the peptide system in cell-permeable form (expressing the sequence Tat) or non-cell-permeable form. These cells, once injected and once they have reached the cerebral zone hit by degenerative pathology, can proliferate and liberate the peptide system in situ.

(73) Process of Preparing the Present Peptide System

(74) The synthesis of these peptides was carried out by means of chemical practice often used today, which allows obtaining amino acid chains up to even 70 residues. A solid-phase chemical synthesizer was used in which the first amino acid is anchored to a solid base (resin) and then by means of isopeptide reactions between the amino acids, the chain of the peptide is elongated via repeated cycles (FIG. 9; modified by Current Protocols in Molecular Biology (2002) 11.15.1-11.15.9).

(75) After synthesis, an HPLC (High Performance Liquid Chromatography) analysis is used for purifying the peptide from contaminants, which is the reaction waste. In addition, a cycle of Mass spectrometry is carried out in order to verify the identity of the peptide. The peptides are considered good when the obtained yield (with regard to the peptides with 26 amino acids) exceeds 95%.

(76) The peptide chain is formed using amino acids in L stereochemical form, if the final peptide is in L form, as well as those in D form, if the final peptide is in D form.

(77) Plasmid Vector Synthesis.

(78) With regard to the plasmid systems, the nucleotide sequence of the peptideswhether they all comprise the N-terminal portion of Syntaxin 1a or even smaller peptides up to a length of 10 amino acids corresponding to the sequence (SEQ ID NO. 2) IEQEEGLNRS and of 14 amino acids corresponding to the sequence (SEQ ID NO. 3) MSEYNATQSDYRER and for all possible lengths, also with the addition at the front (N-terminal) or at the back (C-terminal) of the sequence Tat=(SEQ ID NO. 4) GRKKRRQRRRPPare inserted in the plasmid expression vectors.

(79) For example, the nucleotide sequences are inserted corresponding to the amino acid sequences (SEQ ID NO. 2) IEQEEGLNRS, or (SEQ ID NO. 11) IEQSIEQEEGLNRS and (SEQ ID NO. 3) MSEYNATQSDYRER and also the entire N-terminal portion of Syntaxin 1a is inserted in the bacterial expression vector pTAT. The particular nature of this vector lies in the fact that the fusion protein that the bacteria synthesize contains different tags which, starting from the N-terminal portion, are: 6His, the sequence Tat=(SEQ ID NO. 4) GRKKRRQRRRPP and the tag Human influenza hemagglutinin (HA). Except for the sequence Tat that serves for rendering the synthesized protein cell permeable, the other tags are useful for the recognition of the exogenous protein by means of specific antibodies.

(80) Preparation of Plasmids for Cellular Expression and Protein Purification from Bacterial Synthesis in Plasmid

(81) Hereinbelow, several examples are described relative to the preparation of bacterial strains containing the plasmid vectors comprising the peptide system that is the object of the present description. The quantities of the substances, indicated hereinbelow, are only reported as a non-limiting example. In the course of the present description, it is therefore assumed that the average man skilled in the art understands the possibility to use even different quantities, so as long as the ratios between the reagents are respected as in the following examples.

(82) The lentiviral plasmidpGEX-4 or pTAT or p-EGFP-C1 or pc-DNAcontaining the above-described sequences is to be inserted in E. coli bacteria, e.g. strain DH5 alpha or BL21 and chemically competent derivatives (rosetta). Then, 5 g of Plasmid DNA (p-EGFP-C1 or pc-DNA) is added to 20 l of KCM 5 (containing 0.5 M KCl; 0.15M CaCl2; 0.25 M MgCl2), 5 l MnCl2 and H2O as needed up to 100 l of solution. At this point, an equal volume of chemically competent DH5 alpha bacterial cells are added. They are left several minutes in ice (4 C.). A thermal shock is caused by placing the entire solution at room temperature for several minutes.

(83) Then, a volume quantity is added equal to 600 l of Luria Broth medium (LB) (10 g/1 tryptone, 5 g/1 yeast extract, 10 g/1 NaCl, 15 g/1 agar) heated at 37 C. containing Ampicillin 100 g/ml, since the plasmids contain the resistance to this antibiotic. Then the solution is plated on agar with the same concentration of Ampicillin. Then, proceed with the purification of the Plasmid DNA by means of MAXI-prep commercial kits, hence used for cellular transfections or for synthesis of lentiviral particles with classical methods from the literature.

(84) The transformation of the bacteria BL21 and derivatives is carried out with a different protocol.

(85) 5 g of vector (pGEX-4 or pTAT) is added to 100 l of bacteria, and then this is incubated for 30 min at 4 C. Then, a thermal shock is induced by placing the solution at 42 C. for 45 seconds. In this manner, the bacteria incorporate the plasmid. The solution is then left for 5 minutes in ice and then a quantity in volume equal to 4 times (400 l) of LB is added. The entire solution is plated on agar containing Ampicillin and the colonies are left to grow for 12 hours at approximately 37 C.

(86) MINI-PREP, Cut with Restriction Enzymes

(87) Since the sequences of interest were inserted between the restriction enzymes, the bacterial growth will be tested so as to control the actual incorporation of the plasmid within the bacteria.

(88) Then the grown colonies are extracted with sterile points and placed in 5 ml of LB with Ampicillin. After a bacterial growth at 37 C. for 5-6 h, a miniprep is prepared with the commercial kits. Then, an aliquot of DNA solution obtained from each colony is subjected to enzymatic cutting, with restriction enzymes used for subcloning (37 C. for 60 minutes).

(89) After this, the DNA is made to run on agarose gel and the cut sequence is controlled at the photometer.

(90) Several l of the colony that expresses more plasmid is made to grow, in order to then prepare the glycerol stocks that are frozen at 80 C.

(91) Growth and Induction:

(92) From the glycerol stocks, the bacteria, coming from a single colony, are made to grow at 37 C. under stirring (250 rpm) as pre-inoculum in LB+Ampicillin (100 g/ml). 1/50 of the pre-inoculum is made to grow in fresh LB+Amp (the quantity depends on the intended size of the production) at 37 C. (250 rpm) up to obtaining OD600 of 0.8-1.

(93) 0.5 mM IPTG is then added in order to induce the production of the exogenous protein.

(94) 1 ml of pre-inoculum is extracted and centrifuged at 5000 rpm for 5 minutes. The pellet is then stored at 20 C. in order to then be analyzed on SDS-PAGE gel.

(95) The induction lasts 4-5 h at 37 C. at a speed of 250 rpm.

(96) Then, the bacteria are centrifuged at 8000 rpm for 20 minutes at 4 C.

(97) The supernatant is decanted and the pellet is weighted and stored at 20 C.

(98) Lysis of the Bacteria Pellet

(99) A Lysis Buffer is added to the pellet with the following composition (0.1M TRIS-HCl, 1 mM EDTA pH 7.5) with a proportion of 2 ml of Lysis Buffer per gram of pellet.

(100) The pellet is then resuspended and 1.5 mg/g of Lisozyme is added (from a 10 mg/ml stock) in order to lyse the bacteria. The solution is incubated for 2 h at room temperature and stirred at 200 rpm.

(101) The lysate is incubated for 10 minutes on ice and then sonicated with 4 pulses each of 50 second duration. Between one pulse and the next, the solution is left for 1 minute in ice.

(102) Then, a centrifuge is executed at 12000 rpm for 20 minutes at 4 C.

(103) The supernatant is stored and a flake of pellet is removed; such flake is broken up, making it soft.

(104) Isolation of the Inclusion Bodies

(105) Added to the lysate volume (after sonication) is a volume equal to of Triton Buffer. Then, this is incubated for 30 minutes at room temperature with a stirring of 200 rpm. Centrifuge then for 20 minutes at 4 C. at a speed of 12000 rpm. The pellet is resuspended in Lysis Buffer. Then, Triton buffer is added in a volume equal to half the volume that is in the tube. This is then incubated at room temperature at 200 rpm. Centrifuge at 12000 rpm for 20 minutes at 4 C.

(106) The pellet is resuspended in washing buffer and centrifuged at 12000 rpm for 20 minutes at 4 C. Wash the pellet again, another 3 or 4 times, always with washing buffer. The pellet is then stored at 20 C.

(107) Solubilization and Dialysis

(108) The pellet is solubilized with 5 ml/g of solubilization buffer containing 6M Guanidinium and 1 mM DTT.

(109) After 2-3 h, the pH is adjusted to 3-4 with HCl. Centrifuge then for 20 minutes at 4 C. at 12000 rpm.

(110) The supernatant is then dialyzed with 6M Guanidinium at pH 3-4.

(111) The dialysis solution is changed 3 times: each washing is carried out after 3 hours and the third washing is kept the entire night at 4 C.

(112) The dialysis solution is then changed with a sodium phosphate buffer **50 mM at pH 7 with the same mode used for the dialysis with guanidinium.

(113) Then, at the end, an aliquot of dialyzed substance is used on acrylamide gel, with coomassie staining in which one must see the band corresponding to the synthesized protein.

(114) The rest of the dialyzed substance is purified on an ion-exchange column.

(115) **Preparation of the Phosphate Buffer:

(116) Prepare a 1 M NaH2PO4 solution and a 1 M Na2HPO4 solution. The two solutions are added together and pH 7 is verified.

(117) Purification on Ion-Exchange Column.

(118) There are three solutions used: 50 mM pH 7 phosphate buffer; 50 mM pH 7 phosphate buffer+1M NaCl; 20% Ethanol+20% sodium acetate in H2O.

(119) The solutions are filtered, degassed and stored at +4 C. in order to be cold at the time of use.