PDZ Binding peptides Characterisation of Neuroprotective PDZ binding

PDZ Binding peptides Characterisation of Neuroprotective PDZ binding peptides by Jamie Al-Nasir Project supervisor: Prof. Brian Austen Presented as part of M. Pharm project at: Kingston university and St. Georges Hospital medical school, 24/01/2011

Introduction What are PDZ domains? ◦ Recall that domains are regions of a protein with a particular sequence ◦ PDZ domains are important regions of intracellular signaling proteins ◦ They function as modular protein-protein signalling complexes in Eukaryotic cells ◦ PDZ acquires its name from the first proteins found to contain these regions PSD-95 (Post-Synaptic-Density, 95 k. Da), DLG (Drasophila melangolaster Discs Large Protein) and ZO-1 (Zonula occludens 1) ◦ Some proteins contain more than one PDZ region ◦ PDZ domains have been found in over 250 proteins across a wide range of organisms from Drasophila to Humans ◦ PDZs bind to complementary sequences at C-termini ◦ Implicated in a wide range of cellular processes ◦ – therefore an extremely useful drug target ◦

PDZs in Neuroscience PDZ domains and Neuroprotective binding ◦ PDZ Domains are found in the NMDA receptor complex where they function as signaling scaffolds that mediate protein-protein interactions (1 and 2, figure below) ◦ In normal physiological function Glutamate is a neurotransmitter that acts on NMDA receptors resulting in the activation of intracellular processes ◦ Excesses of Glutamate however can cause downstream toxicity, termed Excitotoxicity - a mechanism implicated in a variety of neurologically degenerative disorders such as Alzheimers, Parkinsons, epilepsy and stroke. This pathological process results in damage to the neurons! ◦ PSD-95 is a PDZ containing protein which forms part of the NMDA receptor complex disruption of which can mitigate and reduce glutamate mediated neuro-degeneration. ◦ Prof. Austen and his team at St. Georges Academic Neuroscience department are working to develop PDZ binding ligands to achieve this therapeutic objective.

Mechanism & Evidence of Approach The NMDA receptor complex 1) Glutamate binds on NR 2 A sub-unit 2) Ca 2+ pumped into intracellular space 3) Intracellular signaling pathways are activated 4) This is coupled by PSD-95 (pink) containing two PDZ domains (1 and 2). 5) Normal response OR Pathological response if prolonged activation (in presence of excess glutamate) Evidence favoring this approach ◦ Immunopreciptation studies by Prof. Austen et al confirm that the NR 2 A subunit interacts with PMCA 2 b via PSD-95. ◦ Spaller, Pizerchio et al have demonstrated that disrupting this interaction using peptides generated from PDZ binding proteins decreases the vulnerability of neurons to excitotoxicity

Ligand design Objectives ◦ “To alter the function of the NMDA receptor complex by targeting PDZ 1 domain of PSD-95 thereby disrupting it’s interaction with PMCA-2 b ◦ …to develop a PDZ binding ligand with improved binding characteristics over an existing PDZ binding peptide ligand, R 2… ◦ …the putative ligand should therefore possess a smaller Kd dissociation constant binding at nano-molar concentrations – R 2 currently binds at micromolar concentrations. ” Structure and Function ◦ Pizerchio et al proposed a cyclic peptide structure based on the straight chain C-Termini of PMCA 2 b but with a cyclic linker – This led to development of R 2 ligand by St. Georges, Royal Holloway, university of Otago, New Zealand uni. of London - R 2 presented at Association International Conference on Alzheimer's Disease, 2010 ◦ Later studies by Li, Spaller et al have yielded important thermodynamic data relating to the effects of alterations in the ring length and residues on binding stability

Ligand design & Molecular modeling New Ligand J 1 (Jamie-1) based on R 2 (Ruth-2) ◦ Incorporates Hexa-Arginine sequence – Cell penetration enhancer Backbone - based on C-Terminus of PMCA 2 b Cyclic ring - with β-Aspartyl linker (previously β-Alanine) Biotin Marker - to aid in microscopy and assess how much drug penetrates neurones ◦ J 1 modeled In Silico Left : J 1 shown without Hexa-Arg or Biotin Above : J 1, note shorthand Hexa-Arginine on left of structure

Ligand design & Molecular modeling Molecular docking ◦ A methodology by which a putative ligand, stored as a 3 d data model is docked to a target protein’s active site (also stored as a 3 d model). ◦ The software re-arranges the atoms to compute the lowest overall energy by means of energy minimisation algorithms such as GROMOS or docking scores such as PMF Argus Lab: PSD-95 protein with J 1 Ligand (yellow) Scigress: PSD-95 protein with J 1 Ligand (red/blue/grey)

Ligand design & Molecular modeling Molecular docking ◦ β-Aspartyl was chosen as a linker as we want to facilitate the bonding between the β-Asp negative COO- side-chain and a nearby Lysine-98 which has a positive NH 3+ side-chain. ◦ This bonding will be ionic providing extra stability to the bound complex Zeus-PDB: PSD-95 protein with J 1 Ligand (green) Deep. View: PSD-95 protein with J 1 Ligand (red/blue/white) Lysine 98 (green) – (Zoomed in)

In Silico results I simulatated the docking of a selection of L- and D- linker ligands in both alpha and beta- configurations, the results determined what was synthesised L- and D- amino acids in alpha and beta config Some standard amino acids in L-a- config

Synthesis of J 1 ligand SPPS, Solid Phase Peptide Synthesis ◦ ◦ Original method pioneered by Robert Bruce Merrifield Essentially synthesis of peptide from C-terminus to the N-terminus, stepwise First residue is bound to solid beads and is protected Beads are washed with a solution of the subsequent residue, which is then washed again to deprotect it. The cycle is repeated with each further residue and finally washed with TFA (Tri. Fluoracetic Acid to cleave the peptide from the beads). ◦ Straight chain of J 1 can be automated using St. George’s Peptide synthesiser ◦ Linker (ring with β-Asp) will be added through a separate reaction to connect the side-chains (Lysine-2 and Glutamate-4 to the linker, β-Aspartyl). Synthesis of J 1 ◦ Synthesis will be on a Leu Peg-PS resin using temporary α-amino Fmoc protection and orthogonally protected residues ◦ Residues will be deprotected and reacted with the bridging residue before finally being deprotected. ◦ The synthesised peptide will be purified using HPLC ◦ The product is then verified using mass spectrometry (MS) and MS 2 by Maldi

Orthogonal Protecting group strategy (Figure 30) - Schematic of the J-1 (Jamie-1) backbone – precursor to cyclic J-1 based on R-2 (Ruth-2). (From right to left) blue region shows the initial Leucine, Leu-1 bound to the Resin (Fmoc-Leu-PEG-PS) followed by Orthogonally protected groups (shown in red) on Lys-2(Iv. Dde) and Glu-4(Phi. Pr) which are separated by Thr-3. Two conventionally Fmoc coupled residues Leu-5 and Ser-6 precede Biotinylated Lysine, Lys-7 (shown in Green). Gly-8 and Gly-9 are proceeded by Pbf protected Hexa-Arginine unit (Purple with Pbf shown in Magenta) which has been acetylated. Two t-Butyl groups are shown in Orange and are used to protect the hydroxyl of Threonine and Serine.

FMoc deprotection and cleavage Fmoc protection strategy is stable in acidic conditions and therefore required Basic condition for cleavage of temporary protection (Piperidine 20% in DMF)

HATU activation of acyl groups during peptide construction

Characterisation of synthesised peptides HPLC (High performance Liquid Chromatography) MALDI mass spectrometry

Testing J 1: Binding assays Determine the Kd binding constant GST (Glutathione-S-Transferase) Pull-down assay used to attain Kd dissociation constant GST is a marker-tag that can be attached to proteins under study If binding occurs then the complex will contain marked protein+bound protein In the presence of J 1 we can quantify the potency of binding between J 1 and PSD 95 -PDZ 1 Quantifying the effect of J 1 on altering excitotoxicity ◦ Utilises a cell line, SHSY-5 Y cells derived from 30 year old blastoma stored at St. Georges, refrigerated at a specific temp. & under CO 2 to prevent alkalinisation. ◦ J 1 will be added to the SHSY-5 Y neuronal cells in presence of excess glutamate (2 mmol) and quantifying the amount of cells that survive. Immunohistological analysis ◦ Used to study the extent to which J 1 internalises within SHSY-5 Y cells and to PSD-95 ◦ Makes use of the integrated biotin marker within the structure as a target for anti-biotin antibody - visualised by microscopy Biotin+Ab. marked neuronal cells

Cell culturing: SHSY-5 Y line SHSY-5 Y cells cultured in DMEM-F 12 buffer

Immunohistochemistry Fluorophores (three Flurophores were used) ◦ DAPI – to label the A-T rich regions of DNA in the Nuclei (for contrasting and locating cells) FITC-Avidin conjugate to label Biotin incorporated in PDZ peptides (labels the peptide) TRITC-anti-rabbit antibody for PSD-95 (labels the PSD-95 target on the membrane) Jamie-3 -Asn Slide# Jamie-3 -Glu 1 A 2 A 3 A 4 A 1 B 2 B 3 B 4 B 2. 7 u. M PDZ peptide X X X X DAPI (Blue) X X X X FITCAvidin(Green) X X X Anti-PSD 95 X X X TRITC Antirabbit (Red) X X X

Immunohistochemistry DAPI / FITC / TRITC stained SHSY-5 Y cells, Jamie-3 peptide (x 40) Jamie 3 -Glu (DAPI+FITC) ( Slide #1 B) Jamie 3 -Glu (DAPI+FITC) Jamie 3 -Glu (DAPI+TRITC) Negative control slides ( lacking either FITC {peptide} or TRITC {psd 95 target} )

Immunohistochemistry DAPI / FITC / TRITC stained SHSY-5 Y cells, Jamie-3 -Asn (x 63) Jamie 3 -Asn (DAPI+FITC+TRITC) Top left = DAPI Top right = FITC (peptide) Bottom left = TRITC Bottom right = Composite image (DAPI+FITC+TRITC) showing good co-localisation (orange and yellow) of Jamie-3 -Asn (green) and target PSD-95 (red)

Immunohistochemistry FITC / TRITC stained SHSY-5 Y cells, Jamie-3 Asn (x 63) Composite image of Jamie 3 -Asn FITC+TRITC but excluding DAPI. It shows co-localisation (orange and yellow) of the Jamie-3 -Asn peptide with its PSD-95 target. Jamie-3 -Asn was stained with FITC and the PSD-95 target with Anti-PSD 95 TRITC the co-localisation is the combination of the two colours Slide #3 A

Immunohistochemistry FITC / TRITC stained SHSY-5 Y cells, Jamie-3 Asn (x 63) Composite image of Jamie 3 Asn FITC+TRITC+DAPI It shows co-localisation (orange and yellow) of the Jamie-3 -Asn peptide with its PSD-95 target. Jamie-3 -Asn was stained with FITC and the PSD-95 target with Anti-PSD 95 TRITC the colocalisation is the combination of the two colours Slide #3 A

Conclusions PDZ domains ◦ PDZ domains are important protein-protein cell signaling mediators ◦ Implicated in a variety of cellular processes not just neurological ◦ Important drug target for novel drugs Molecular modeling and In Silico methods ◦ Molecular modeling and docking is extremely useful method of simulating thermodynamic stability and binding between ligands and proteins ◦ Data can be used to select a putative ligand prior synthesis vs the traditional combinatorial approach Immunohistochemical analysis ◦ Jamie-3 -Asn pdz binding peptide showed good co-localisation of peptide with PSD-95 ◦ Jamie-3 -Glu showed some co-localisation but was less consistent across repeated imaging ◦ Results consistent with MTT assay, in which Jamie-3 -Asn pdz binding peptide performed better at mitigating excitotoxic damage than Jamie-3 -Glu Implications ◦ More favorable binding characteristics than R 2 - further development may yield a useful pharmacological agent that could be administered to slow down neurological degeneration in diseases such as Alzeimers, Parkinsons, epilepsy and stroke.

References used 1. Harris BZ, Lim WA. Mechanism and role of PDZ domains in signaling complex assembly. Journal of Cell Science, 2010; 114: 3219 -3231 2. Cui H, Hayashi A, Sun HS, Belmares MP, Cobey C, Phan T et al. PDZ Protein Interactions Underlying NMDA Receptor-Mediated Excitotoxicity and Neuroprotection by PSD-95 Inhibitors. The Journal of Neuroscience, 2007; 27(37): 9901 -9915 3. Li T, Saro D, Spaller MR. Thermodynamic profiling of conformationally constrained cyclic ligands for the PDZ domain. Bioorganic & Medicinal Chemistry Letters, 2004; 14(6): 1385 -1388 4. Austen BM, Duberley K, Turner P, Empson R. Cyclic hexa-arg PDZ-binding peptides that bind PSD 95 inhibit glutamate-mediated toxicity. Alzheimer's Association International Conference on Alzheimer's Disease 2010, 6(4): Supp. 562 Software used (in order of images shown) 1. 2. 3. 4. Argus Lab by Mark Thompson Scigress explorer by Fujitsu Siemens Zeus PDB viewer by Jamie al-nasir Deep. View/Swiss-PDB viewer