Probing the catalytic mechanism of an antifibrotic copper metallodrug 4 December 2017 Lizelle Lubbe Supervisor: Prof Ed Sturrock
Transition metals offer benefit of designing compounds with complex architectures, chemical diversity & novel MOA Metal chelation and linkage to a target recognition domain allows ROS production at specific sites Desired protein target Off-target protein Metal-binding Target recognition Advantages of catalytic metallodrugs Effective at sub-stoichiometric drug concentrations metallodrug recycled following irreversible inactivation of target lower & less frequent dose
Renin angiotensin system Kallikrein-kinin system Angiotensinogen Kininogen D R V Y I H P F H L V I H G M I S L M K R P P G F S P F R S S R I G E Renin Kallikrein Angiotensin I Bradykinin D R V Y I H P F H L R P P G F S P F R Angiotensin II ACE Blood Pressure ACE D R V Y I H P F Inactive peptide R P P G F S P B2 receptor AT 1 receptor Vasoconstriction Vasodilation
Two catalytically active domains (N- and C-domain) AcSDKP N 60% sequence identity, 90% active site similarity Display diverse functions Inactive fragments AcSDKP Blood pressure mainly controlled by C-dom Acetyl-Ser-Asp-Lys-Pro (AcSDKP) peptide cleaved by N-dom Bradykinin Inactive peptides C AngII AngI Prevent hydrolysis of AcSDKP by selective N-dom inactivation plasma concentration fibroblast proliferation collagen deposition No influence on blood pressure - Potential for treatment of fibrosis
N-selective inactivation Aim: Determine mechanism of N-selective catalytic inactivation
kda kda 100 70 55 100 70 55 Inhibitor binding affinities (K i ) K i = 0,86nM K i = 0,087nM K i = 6,41nM K i = 9,99nM K i = 44,94nM K i = 15,57nM Lisinopril GGH-Lisinopril Cu-GGH-Lisinopril
no = no co-reactant A = ascorbic acid H = H 2 O 2 AH = ascorbic acid + H 2 O 2 Ndom 100 100 % R em aining activity 80 60 40 20 0 0 50 100 150 200 250 300 350 400 450 500 EB + A EB + H EB + HA + CuGGHLis % R em aining activity 80 60 Full inactivation 40 E I + no E I + A 20 E I + H E I + H A 0 0 50 100 150 200 250 300 350 400 450 500 Tim e (m in) Tim e Cdom 100 100 % R em aining activity 80 60 40 20 0 0 50 100 150 200 250 300 350 400 450 500 EB + A EB + H EB + HA + CuGGHLis % R em aining activity 80 Non-zero plateau 60 40 E I + no E I + A 20 E I + H E I + H A 0 0 50 100 150 200 250 300 350 400 450 500 Tim e (m in) Tim e
Protein cleavage can occur when OH abstracts H from C α or Pro, Glu / Asp sidechains are oxidized 7 hour inactivation of protein with/-out co-reactants and inhibitor (@ IC 20 ) No cleavage observed suggesting amino acid sidechain oxidation Ndom Incubation Cu-GGHLis + Cu-GGHLis Cdom Incubation Cu-GGHLis + Cu-GGHLis - - A H AH - A H AH - - A H AH - A H AH 150 100 80 60 50 25 100 80 60 50 25 no = no co-reactant A = ascorbic acid H = H 2 O 2 AH = ascorbic acid + H 2 O 2
ACE inactivation by ROS - OH from PMA-activated neutrophils oxidize ACE in endothelial cells (Chen et al, 1992-3) - g OH inactivate Cdom>Ndom (42% vs 66% remaining activity) (Michel et al, 2001) - H 2 O 2 + ascorbate incubation inactivates Cdom>Ndom (43% vs 65% remaining activity) (current work) ACE protection - 8-ANS binding to a hydrophobic site protects against γ-radiation (Cdom>Ndom) (Voronov et al, 2002-3)
Neutrophil activation ACE Inactivation Formyl-MLF Bradykinin Neutrophil recruitment Vasodilation Macrophage activation Degranulation Substance P Pain response Nitric Oxide Species Reactive Oxygen Species Inactivation Inactive peptides Elucidate mechanism of ACE oxidation by diffuse ROS design ACE-protecting agent based on 8-ANS
Inter-subdomain glycan interactions allow Ndom active site closure and ROS protection Cdom is more susceptible to diffuse radical oxidation due to low occupancy and lacking glycan interactions - 8-ANS binding to site on Cdom surface stabilizes protein and shields against ROS 8-ANS site Perform molecular dynamics simulations (Amber) to study effect of glycan motions on ROS shielding Integrate with kinetics and oxidized sites identified by mass spectrometry
Attached bisialylated fucosylated complex type glycan to N-X-T/S (X P) motifs - 9 Ndom and 6 Cdom sites Zn site forcefield parameters from Brás N.F et al (2014) and hybrid bonded/non-bonded model 1. Solvation and neutralization in TIP3P box using tleap in AMBER 2. 5 000 steps solvent minimization followed by 20 000 steps full system minimization 3. Heating to 300K over 600ps 4. 0.8ns solvent equilibration followed by 1ns full system equilibration (CPU) 5. Additional 2.5ns full equilibration using PMEMD (GPU) 6. 30ns production dynamics using PMEMD (GPU) Ndom: 188 042 atoms Cdom: 103 072 atoms
N- and Cdom backbone and Zn site were stable while glycans were highly flexible Average Zn site geometry Ndom + glycans Ndom backbone Cdom + glycans Cdom backbone His His Zn Glu H 2 O His Glu His Zn H 2 O
Glycan Occupancy VolMaps Glycoprotein principal component analysis revealed long-range intersubdomain glycan motions driving Ndom active site cleft closure Ndom PC1 Unique glycans Glycan interactions and large range of motion creates a shield around the Ndom (3 unique glycans) Cdom PC1
357-381 425-472 490-518 Overall the backbone flexibility profile is similar between N- and C-domain Zn
Sub-domain I N-ter Hinges Ndom Subdomain interaction Stable active site Cdom Subdomain repulsion Flexible active site W137 L143 I146 L129 V127 L156 F158 M155 Unique Ndom glycan Subdomain interface Unique glycan ANS binding: slight ROS protection Residues with RMSF Cdom > Ndom Lacking glycan ANS binding: Stabilization, active site closure & ROS protection Sub-domain II Proposed 8-ANS binding site C-ter Unique Ndom glycans
Investigated whether glycans restrict ROS access to the active site by affecting tunnel dynamics (29-30ns) Sub-domain I Sub-domain I Zn Zn Zn Zn Sub-domain II Active site cleft Prime subsite tunnels shielded Sub-domain II Prime subsite tunnels exposed Active site cleft
A. Confirm the 8-ANS binding site by co-crystallization B. Collaboration with Prof RK Acharya (University of Bath) to solve Cu-GGHLis co-crystal structures Proximity to substrate binding/hydrolysing residues C-domain Cu-GGH proximity to oxidizable N/C residues: - Understand inactivation in presence of metallodrug - Suboptimal orientation for ROS production in Cdom? Cu-GGH-Lisinopril B. Mass spectrometric analysis of inactivated samples Identify sites of selective metal-catalysed/diffuse oxidation Hocharoen et al (2013) docking into Cdom
Cu-catalysed oxidation of substrate binding and hydrolysing residues led to rapid N-selective inactivation X-ray crystallography and mass spectrometry will offer more details w.r.t. the mechanism The C-domain s lower carbohydrate content and subdomain repulsion leads to increased active site mobility, tunnel dynamics and ultimately susceptibility to diffuse radical oxidation 8-ANS binding at the proposed site could stabilize the C-domain to enable active site closure and decrease exposure to reactive oxygen species Elucidating the mechanisms of ACE oxidation will aid the future design of selective antifibrotic drugs and further our understanding of ACE s role in inflammation
Prof Ed Sturrock (supervisor) & ACE lab members Collaborators Prof James A Cowan Financial assistance Prof Trevor Sewell (co-supervisor) Prof Ravi K Acharya