Post on 11-Jun-2015
description
Collaborative MedicinalCollaborative Medicinal Chemistry Research:Building More Porous
OResearch OrganisationsThe Academic-Industrial Interface in 21st Century Drug Discovery
Tuesday 24 June 2014
David Andrews*, Andy Merritt, Martin Swarbrick
Outline of Talk
• Introduction to AstraZeneca’s Open Innovation Efforts• Examples of compound collection collaborations
C t d ll b ti i l ti D i M k T t A l• Case study : collaborating in real time : Design–Make–Test–Analyze- Opportunities - IssuesIssues- Solutions
• Open Innovation Platform• Future Outlook
2 David M Andrews | 24 June 2014 R & D | Oncology
AstraZeneca is driving science through collaborations…
…more than 90 partnerships in the last three years
Three examples of how we are helping to drive Open Innovation across our industry: Leveraging compound collections to share to maximise value
TB D A l tTB Drug Accelerator
Delivering support for neglected disease strategy and sharing libraries to find
di i f lif
Delivering reciprocal access to high quality chemical start points with libraries
th $MM
Delivering early access to new target ideas and partnerships with
d i LG t new medicines for life-threatening neglected diseases
worth $MMs academic LG centres, with first right of refusal on targets at LO
P id ll i ( ) Hi F dProvide collection(s) → Hits → Freedom-to-use‘Clean’; with a clear exit strategy
Medicinal chemistry within different collaborative discovery model settingsy gModel Description Advantages ChallengesStrategic long-term, shared risks motivation and engagement, ‘true- role definition, manage-Model Description Advantages ChallengesStrategic long-term, shared risks motivation and engagement, ‘true- role definition, manage-S a eg cAlliance*
o g te , s a ed s sand incentives
o a o a d e gage e , uetype’ collaboration, learning/expertise, cost effective
o e de o , a agement, processes, accountability, IP
Industry- risks & rights at clear roles IP and decisions utilizing full potential of
S a eg cAlliance*
o g te , s a ed s sand incentives
o a o a d e gage e , uetype’ collaboration, learning/expertise, cost effective
o e de o , a agement, processes, accountability, IP
Industry- risks & rights at clear roles IP and decisions utilizing full potential ofIndustry-sponsored
risks & rights at industry sponsor
clear roles, IP and decisions,speed, flexible
utilizing full potential of the team, costs
Government/ Charity
research project grant approval
neglected diseases, diverse groups and skills longer term planning
bureaucracy, IP, management
Industry-sponsored
risks & rights at industry sponsor
clear roles, IP and decisions,speed, flexible
utilizing full potential of the team, costs
Government/ Charity
research project grant approval
neglected diseases, diverse groups and skills longer term planning
bureaucracy, IP, managementCharity-
funded**approval and skills, longer term planning,
cost effective management
Crowd- use of the entire Easy to accommodate, low cost, IP, management of
Charity-funded**
approval and skills, longer term planning, cost effective
management
Crowd- use of the entire Easy to accommodate, low cost, IP, management of sourcing MedChem community powerful in idea generation design ideasInnovation incubator
on-campus model training, tool compounds, line of sight
IP, limited to early discovery phases
sourcing MedChem community powerful in idea generation design ideasInnovation incubator
on-campus model training, tool compounds, line of sight
IP, limited to early discovery phases
Precompeti-tive consortia
common interest in developing tools
cost effective, learning/expertise management, IP, limited to early discov.
H Wild et al Angew Chem Int Ed 2013 52 2684
Precompeti-tive consortia
common interest in developing tools
cost effective, learning/expertise management, IP, limited to early discov.
H. Wild et al., Angew. Chem. Int. Ed. 2013, 52, 2684.* R. Wellenreuther et al., Drug Discov. Today 2012, 17, 1242.* R. Williams et al., Drug Discov. Today 2012, 15, 1359.* D. Andrews et al., Drug Discov. Today 2014, 19, 496. ** A.L. Hopkins et al. Nature 449, 166
Setting Up the CollaborationsThe Initial Model
Shared Series ofInitial HTS Hit Shared Series of Interest
The Problems• What happens to the hits that don’t go anywhere?
• Led to a reluctance to unblind structuresLed to a reluctance to unblind structures
• ‘Two countries divided by a common language’ : we used slightly different terminology and acronyms for the same things
The Solution• Create an agreement that gives the chemists the maximum freedom to work innovatively and
i ti llsynergistically
• Control the risk of inadvertent reach-through into the broader proprietary information or the parent organizations
• Agree common terminology, common ground rules
6 David M Andrews | 24 June 2014 R & D | Oncology
Opportunities in Compound Collaboration
• Ownership of compound series rests with the originator until initial liabilities are mitigated• Prevents the non-originating parties collection becoming populated by compounds that can’t
progress• Incentivizes teams to overcome initial liabilities• ‘Productive SAR’ triggers shared ownership and a fully collaborative research optimization
programprogram
7 David M Andrews | 24 June 2014 R & D | Oncology
Additional Impacts of Clearer ground RulesAllows testing of newer compounds
16
18
20
12
14
16
Years
6
8
10Years
2
4
6
0 10 20 30 40 50 60 70 80 90 1000
C l ti t8
Cumulative percentDavid M Andrews | 24 June 2014 R & D | Oncology
Additional Impacts of Clearer ground RulesAllows testing of quality compounds
12
0
1
34 6
Calculated logD
distribution-2-1 4 Rotatable
bonds2
12
810
-4 -32 5
67
12
34
4
57
8Rings
≤25
6
…and expansion into full deck screening
Number of Acceptors
23
4
9
10 11 screening9
2 11
David M Andrews | 24 June 2014 R & D | Oncology
Issues to Overcome….….and solutions
Design Make Test Analyze
What should Who?External
database?Preferred
we make?Priority?
Who?Route?
database?Post data between
partners?
workflows / analysis tools?
10 David M Andrews | 24 June 2014 R & D | Oncology
Collaboration ToolsCh T X• Capture of design ideas and outcomes (knowledge management)
ChemTraXp g ( g g )
• Platform for real time collaboration between partners and service providers
• Easy visibility of on going chemistry within a project and planned next rounds of chemistry, ensuring optimal deployment of resources
• Built to support today’s ways of working with partners and CROs where information visibility and user functionality is easily controlled to fit all modes ofinformation visibility and user functionality is easily controlled to fit all modes of operation
f f• Information access is set at the project level, enabling easy set up of multiple projects to work with multiple organizations
11 David M Andrews | 24 June 2014 R & D | Oncology
Overview of FeaturesChemTraX Tracking Board
Process• Steps a design set follows from conceptionOverview of Features • Steps a design set follows from conception
through to completion
Color Design Set• Multi parameter way of visualizing information
• Here we see color by organization that is
assigned the design set for synthesis
Design Set• A collection of chemical structures
Swim Lanes• Multi parameter way of separating the design sets
• This example shows split by priority of design set
12 David M Andrews | 24 June 2014 R & D | Oncology
designed to address a specific project issue
(potency, solubility etc)
• This example shows split by priority of design set
Design SetsSharing Ideas and Compounds to SynthesizeSharing Ideas and Compounds to Synthesize
Collaborative sharing of:• Design hypotheses
C d t id f th i• Compounds to consider for synthesis• Status of individual compounds (in
synthesis, complete etc)
R & D | Oncology
y , p )• Design set outcomes
Data Sharing
Partner 2Partner 1
Visualisation Visualisation
Query & retrieval
Query & retrievalExport and
transfer
Corporate Database Corporate Database
transfer
p p
Input Input
14
Data generation Data generation
PIP5K and PI4K – Complex BiologyA ideal target area for risk-sharing collaboration
N
O
H
O
O
PI4K
O
O
LY294002
O
OO
OOO
Wortmannin
Pharmacological manipulation of
cellular PI4P LY294002 Wortmannin
pIC50
PI4Kα <4.3
pIC50
PI4Kα 5.9
levels A challenge due to
non-specificity and lack of
PIP5KPI4Kβ 4.4
PI3Kα 6.2
PI4Kβ 5.8
PI3Kα 8.1
and lack of potency of PI 4-kinase inhibitors
• AZ/CRT team identified potent and selective small molecule inhibitors:
• Of both type III PI 4-kinase isoforms
• Cross-subtype selective inhibitors of PIP5KPhosphoinositides in cell regulation and membrane dynamicsPhosphoinositides in cell regulation and membrane dynamicsNature 443, 651-657 (12 October 2006) | doi:10.1038/nature05185
Author | 00 Month Year15 R & D | Oncology
PI4Kβ Lead GenerationStarting with a potent non-selective hit
1 2
HTS Hit – 1 2PI4Kα pIC50 8.0 5.3PI4Kβ pIC50 8.3 7.2PI3Kα pIC50 8.5 6.1PIP5Kγ pIC50 6.2 6.2
Potent, selective small molecule inhibitors of type III phosphatidylinositol-4-kinase-α…Ch C 2014 50 5388 5390 htt //d d i /10 1039/C3CC48391F
16 David M Andrews | 24 June 2014 R & D | Oncology
Chem. Commun., 2014, 50, 5388-5390 http://dx.doi.org/10.1039/C3CC48391F
PI4Kβ Series – Amide SAR
pIC50
PI4Kα 5.3
pIC50
PI4Kα 5.0
pIC50
PI4Kα 5.5
pIC50
PI4Kα 6.0
pIC50
PI4Kα 5.0
PI4Kβ 7.2
PI3Kα 6.2
PIP5K 4 9
PI4Kβ 7.7
PI3Kα 4.8
PIP5K 4 0
PI4Kβ 8.0
PI3Kα 5.4
PIP5K 4 4
PI4Kβ 8.2
PI3Kα 5.9
PIP5K 4 7
PI4Kβ 8.0
PI3Kα 4.5
PIP5K 4 9PIP5Kγ 4.9
LogD 2.5
PIP5Kγ <4.0
LogD -
PIP5Kγ <4.4
LogD 3.6
PIP5Kγ 4.7
LogD 1.6
PIP5Kγ 4.9
LogD 3.0
PI4KβPI3Kα
17 David M Andrews | 24 June 2014 R & D | Oncology
Kinase selectivity of inhibitors
N
O
N H 2
S
N
@ 1
0μM
60
70
80
90
100
60
70
80
90
100
@ 1
0μM
FGR 98%
ZIPK 72%
STK17A 68%
inhi
bitio
n @
20
30
40
50
20
30
40
50
inhi
bitio
n @
%
Millipore 125 kinase panel
0
10
0 50 100 150 200 250 3000
10
0 20 40 60 80 100 120 140
%
Millipore 259 kinase panel
18 David M Andrews | 24 June 2014 R & D | Oncology
Live cell imaging• The PH domain of PLCδ1 binds specifically to PI(4,5)P2
• U2OS cells overexpressing PH-PLCδ1 pre-incubated with inhibitors for 60 min atp g p37ºC before reading fluorescence
• In this system, Wortmannin and the PI4Kα inhibitor modulate PI(4,5)P2 levels,the PI4Kβ inhibitor is inactivethe PI4Kβ inhibitor is inactive
19 David M Andrews | 24 June 2014 R & D | Oncology
Open Innovation – Industry Perspective• Stefan Lindegaard survey – 2010• http://www.15inno.com/2010/03/29/oibigpharma/• Quick and dirty survey 10 largest pharma + ‘Open Innovation’• Quick and dirty survey – 10 largest pharma + Open Innovation• GSK – ‘Innovation at GSK’ – the only well-developed web site
• Four years on….
20 David M Andrews | 24 June 2014
Open Innovation offerings across all stages of hresearch
R & D | Oncologyhttp://openinnovation.astrazeneca.com
Target InnovationHow Does it Work process flow diagram
Proposals can seek
How Does it Work – process flow diagram
Continuous call for proposalsContinuous call for proposals ①Seed funding (up to $100K) to strengthen hypothesis ②Request an AZ compound library for them to screen
Continuous call for proposalsContinuous call for proposals
High th h t
High th h tAZAZSeed fundingSeed funding screen③Request to run a full HTS at AZ facility
Tools provided to help investigators:
throughput screen in AZ facility
throughput screen in AZ facility
AZ compound
library
AZ compound
library
Seed funding for Target validation
Seed funding for Target validation
• “Instructions to Authors”• AZ interests and proposal scoring criteria• Review feedback
AZ scientific ReviewAZ scientific Review
Typical arrangement is risk/reward sharing:• AZ provides compound supply or seed funding or
screening capability
‘Full Project Proposal’ under CDA‘Full Project Proposal’ under CDA
screening capability• PI has obtained funding via grant (unless grant
awarded by AZ)• Rewards include: publication(s) background info
AZ scientific ReviewAZ scientific Review
Project ExecutedProject Executed Rewards include: publication(s), background info. for follow-on studies, royalties (if successful)
Project ExecutedProject Executed
David M Andrews | 24 June 2014 R & D | Oncology
New Molecule ProfilingHow Does it Work process flow diagram
Two step process:
How Does it Work – process flow diagram
New molecules submitted New molecules submitted • Cheminformatics evaluation• Screening evaluation
More details:
to external cheminformatics service providerto external cheminformatics service provider
Cheminformatics evaluationCheminformatics evaluationMore details:• Molecules are submitted securely to an external
cheminformatics service provider so that AZ does not see the structures
Report of property calculations and novelty checks sent to AZ/submitter
Report of property calculations and novelty checks sent to AZ/submitter
• Physicochemical and biological properties are calculated and molecules are checked for novelty against the AZ and public collectionsAZ d b itt i t f h i f ti
AZ scientific reviewAZ scientific review
• AZ and submitter receive a report of cheminformaticsevaluation results
• AZ reviews report and accepts/rejects compounds into the HTS screening collection
MTAMTA
g• MTA between AZ and submitter and samples added
to the HTS screening collection• HTS screening report generated yearly and sent to
b itt
Samples added to Screening collection Samples added to Screening collection
submitter• If ‘screening hit’ then AZ/submitter
negotiate/collaborate.
Negotiate/collaborate if ‘screening hit’Negotiate/collaborate if ‘screening hit’
The Future?
M t li d / il• More streamlined / agile start-up?
• Further vendor tools to facilitate
• Remote working?g• E.g. virtual whiteboards : • http://www.chemaxon.com/wp-content/uploads/2012/10/Patcore.pdf• ‘Skype for chemists’• Refinement of interaction models
Addi i l ll b i d l24 David M Andrews | 24 June 2014 R & D | Oncology
• Additional collaboration models
What you may hear about collaborative MedChem…
We can’t l h IP
Remember the Boeing
We are
control the IP risks!
Dreamliner project!
giving away our crown
jewels!
This is too complicated
and can neverand can never work!
cf. H. Wild et al., Angew. Chem. Int. Ed. 2013, 52, 2684.
Summary• Collaborative MedChem in our hands has been
versatile and successful, projects have advanced fast d t hi h d i litand at high design quality
• An incentive to invest in novel therapeutic approaches over the longer term – e.g. AZ/CRT Cancer metabolism Alliance
• Many ways of ‘constructing’ the collaborative DMTA teams (project dependent)(p j p )
• Pursuit of more than one project in partnership brings many synergies
• We have found ways to incentivize teams to overcome• We have found ways to incentivize teams to overcome initial liabilities with novel chemical series / hits
• Future areas for focusTechnology sharing / access• Technology sharing / access
• Derisking area of biology• Exploring new target classes
• Should be tried more often• Should be tried more often
Acknowledgements
• A large number of bench scientists at AstraZeneca, Cancer Research UK and MRCT
• ….in particular – Mike Waring
• Jörg Holenz AZNeuro
• Phil Spencer AZ Discovery Sciencep y
• Peter Simpson AZ Discovery Science
• David Hollinshead• Martin Harrison• Paul Faulder• Andrew Griffin
The ChemTraX team at Elixir
• Andrew Griffin
http://www elixirsoftware co uk/chemTrax htmlhttp://www.elixirsoftware.co.uk/chemTrax.html
27 David M Andrews | 24 June 2014 R & D | Oncology
Confidentiality NoticeConfidentiality Notice This file is private and may contain confidential and proprietary information. If you have received this file in error, please notify us and remove it from your system and note that you must not copy, distribute or take any action in reliance on it. Any unauthorized use or disclosure of the contents of this file is not permitted and may be unlawful. AstraZeneca PLC, 2 Kingdom Street, London, W2 6BD, UK, T: +44(0)20 7604 8000, F: +44 (0)20 7604 8151, www.astrazeneca.com
28