gabriele cruciani, laura goracci · efsa parma, 13-15 june 2017 gabriele cruciani washing twice...
TRANSCRIPT
Gabriele CrucianiEFSA Parma, 13-15 June 2017
Gabriele Cruciani, Laura GoracciDepartment of Chemistry, Biology and Biotechnology, University of Perugia, Italy
& Molecular Discovery, London
Photos from Perugia, Umbria, Italy
Lipidomics and food metabolites
Parma Summer School 2017 In silico methods for food safety EFSA Parma, 13-15 June 2017
Gabriele CrucianiEFSA Parma, 13-15 June 2017
The limits of QSAR (QSPR)
1. Local models (plastics and drugs cannot be located in
the same model)
2. Activity (Property) must be generated by the same
mechanism
3. Similarity methods depend on descriptors
4. Chemical descriptors must be relevant to P
5. Biological descriptors must be relevant to P
6. Pruning descriptors very dangerous !!
7. Combination of methods better than consensus methods
8. AI << IA (interpretation …interpretation …interpretation)
Gabriele CrucianiBrazMedChem 2014
DRUG-impurities
Foodingredients-
toxins
Detergents
WANTED EFFECTS
Chemical description01100011100010101
Tox
xenobiotics
UNWANTED EFFECTS
METABOLITES
QSPR
Struture-Property Relationships
Gabriele CrucianiBrazMedChem 2014
DRUG-impurities
Foodingredients-
toxins
Detergents
WANTED EFFECTS
Chemical description01100011100010101
Tox
xenobiotics
UNWANTED EFFECTS
METABOLITES
QSPR
Struture-Property Relationships
Phase I metabolites – Phase II metabolites
Gabriele CrucianiBrazMedChem 2014
DRUG-impurities
Foodingredients-
toxins
Detergents
WANTED EFFECTS
NO metabolomics !
QSPR
Tox
Chemical description01100011100010101
xenobiotics
METABOLITES
UNWANTED EFFECTS
Struture-Toxicity Relationships
Gabriele CrucianiBrazMedChem 2014
Identification of metabolites
chemical structure, abundance, kinetics
QSPR
Toxtime
abu
nd
ance
Gabriele CrucianiBrazMedChem 2014
Identification of metabolites
chemical structure, abundance, kinetics
time
abu
nd
ance
QSPR
No Tox
Gabriele CrucianiBrazMedChem 2014
BioMatrix analysis
DHep
HHep
MHep
MsHep
RHep0
20000
40000
60000
80000
100000
120000
140000
160000
180000
DHep
HHep
MHep
MsHep
RHep
Hepatocytes across species Analysis
humandog
Gabriele CrucianiBrazMedChem 2014
25%
7%
68%
toxicityQSPR
No Tox
DHep
MsHep0
50000
100000
150000
200000
M1
2 -
17
8
M1
4 -
16
4
M1
6 +
14
8
M1
7 +
16
2
M2
0 +
16
2
M2
1 +
14
8
M2
5 -
14
M6
-2
M7
-2
Sub
stra
te +
0
DHep
HHep
MHep
MsHep
RHep
Metabolic patways
Take home messages
Gabriele CrucianiEFSA Parma, 13-15 June 2017
GENOMICS
PROTEOMICS
METABOLOMICS
LIPIDOMICS
IMMUNOMICS TRANSCRIPTOMICS
TRANSPORTOMICSSECRETOMICS
THE «OMICS» TREE1980
2000
2010
1990
2017
According to SciFinder only 18 publications are reported in Untargeted Lipidomics till today (2017)
Targeted Untargeted
Gabriele CrucianiEFSA Parma, 13-15 June 2017
‘STARTING POINT’
‘END POINTS =
EFFECTS’
Any change/modification of a cellular function induced by drugs will be reflected in several modifications of lipid content, composition and distribution.
Why
untargeted
Lipidomics ?
Gabriele CrucianiEFSA Parma, 13-15 June 2017
Several thousands of lipids must be identifyed (up to 1000000)
Control Treated
Time zero Time x
Normal Disease
Gabriele CrucianiEFSA Parma, 13-15 June 2017
Several thousands of lipids must be identifyed (up to 1000000)
An MS/MS spectra of the ion m/z=878.816 representing characteristic
fragment ions to support for the identification of TAG 52:1
3 energies
2 polarities
10 adducts
isotopes
dimers
water-adducts
Gabriele CrucianiEFSA Parma, 13-15 June 2017
Several thousands of lipids must be identifyed (up to 1000000)
Identification of 10 lipids may require a days of workFor 2000 lipids one need 6-12 months time !!!
Proposed solution: database of commercial lipids
However: 1000 lipids availables = 0.1% potential lipids
very high cost (1 mg – 250 USD)
low stability (oxydation also at -80°C)
Gabriele CrucianiEFSA Parma, 13-15 June 2017
Several thousands of lipids must be identifyed (up to 1000000)
Identification of 10 lipids may require a days of workFor 2000 lipids one need 6-12 months time !!!
Proposed solution:
LipoStar is doing in less than a minute
Gabriele CrucianiEFSA Parma, 13-15 June 2017
❑ Amiodarone (1, 4, 8, 12 µM)❑ Cimetidine (4, 20, 40 µM)❑ CTR-CTRV
AMIO= 1 µM
AMIO= 8 µM
AMIO= 12 µM
AMIO= 4 µM
Lipids modification in liver Amiodarone (Cholestasis)
Experiments show the tox-dose level
In collaboration with
Lipidomic research in public and pharmaceutical industries
Gabriele CrucianiEFSA Parma, 13-15 June 2017
«Campioni anomali»: Vit_D_ 63 e Vit_D_3
In collaboration with
Steatosis: blood samples
Gabriele CrucianiEFSA Parma, 13-15 June 2017
In collaboration with
Ovarian cancer – serum samples
Gabriele CrucianiEFSA Parma, 13-15 June 2017
RAT BRAIN SLICES
Anticancer compound - Ceramide Sinthase (CS) inhibitor
D7-SPHCS
D7-SPH-1P D7-Cer
D7-Cer-1P
D7-SM
D7-SULFATIDES(representingglycosphingolipids)
- Rat brain slices were treated with deuterated sphingosine (D7-SPH), aslabelling agent
- After a certain labelling time, some samples were treated also with a CS inhibitor (CS-INH)
- The aim was to compare CS-INH treatedsamples with controls
D7-hexadecenal
D7-palmitoylCoA
D7-PC, PE, PI, PS…
(labelling agent)
CS
INH
Gabriele CrucianiEFSA Parma, 13-15 June 2017
D7-SPHCS
D7-SPH-1P D7-Cer
D7-Cer-1P
D7-SM
D7-hexadecenal
D7-palmitoylCoA
D7-PC, PE, PI, PS…
(labelling agent)
D7-SPH
CTRL CS-INHSHORT
CS-INHLONG
CTRL= controlsCS-INH SHORT= treated with CS-INH for a short timeCS-INH LONG = treated with CS-INH for a long time
D7-SULFATIDES(representingglycosphingolipids)
D7-Cer
D7-SM
CTRL CS-INHSHORT
CS-INHLONG
CTRL CS-INHSHORT
CS-INHLONG
The effect of the inhibitor on the CS is confirmed
CS
INH
Gabriele CrucianiEFSA Parma, 13-15 June 2017
LipoTox Technology Platform
Lipidomics in 3D-Human
microtissues
(hepatocytes)
Biomarker responses to drugs
are stored in the database
>300 drugs
Informatics tools are used to
predict clinical outcomes
High Throughput Human Lipidomics
LipoStar®
Fingerprinting System
Reference
(DILI) Profile DatabasePredictive
Cheminformatic Tool
Gabriele CrucianiEFSA Parma, 13-15 June 2017
Portfolio of assay-ready microtissues
Gabriele CrucianiEFSA Parma, 13-15 June 2017
Washing twice with PBS
Centrifugation Lipid extraction with MMC*
(Methanol:MTBE:Chloroform) method on cells pellet
Adding of 150 µL of MMC
mixture
MMC extraction method
* Anal Bioanal Chem. 2014 Dec;406(30), Pellegrino RM, Di Veroli A, Valeri A, Goracci L, Cruciani G.
Vortex and incubation
Centrifugation Separation of supernatant
LC/MS Q TOFAnalysis
Microtissue retrival
One pot Lipid extraction from 3D microtissues
MetID
Adding DCP
Gabriele CrucianiEFSA Parma, 13-15 June 2017
Why Lipidomics ?
Peak-detectionSignal/noise data pretreatmentSignal alignmentStructural assignment
Gabriele CrucianiEFSA Parma, 13-15 June 2017
Identification of 10 lipids may require days of workFor 2000 lipids one need 6-12 months time !!!
LipoStar is doing in less than a minute
Identification, structural assignment and matrix building
Gabriele CrucianiEFSA Parma, 13-15 June 2017
Gabriele CrucianiEFSA Parma, 13-15 June 2017
LipoTox Technology Platform
Similarity Analysis of Lipids Profiles
Highly correlated → Similar
Low correlation → Not similar
Medium correlated → Almost similar
Gabriele CrucianiEFSA Parma, 13-15 June 2017
LipoTox Technology Platform
Similarity Analysis of Profiles
Lipid profiles – drug 01/02Lipid profiles – drug 01- different doses
Lip
id p
rofi
les –
dru
g 0
1 &
02
Gabriele CrucianiEFSA Parma, 13-15 June 2017
phenformin
tolcapone
Triacylglycerols profiles for Drug 1 and Drug 2 (negative values are differences from control)
Gabriele CrucianiEFSA Parma, 13-15 June 2017
LipoTox Technology Platform
Tamoxifen DILI+(Steatohepatitis)
Gabriele CrucianiEFSA Parma, 13-15 June 2017
2 days4 days6 days8 days11
days
14 days
LipoTox Technology Platform
Gabriele CrucianiEFSA Parma, 13-15 June 2017
Tamoxifen DILI+(Steatohepatitis)
Toxicity increases
Gabriele CrucianiEFSA Parma, 13-15 June 2017
2 days
14
11
9
7
4
Tamoxifen in membranes
Drug in membranes
Gabriele CrucianiEFSA Parma, 13-15 June 2017
358,2353 in medium
564,2558 in medium
563,2183 in medium
374,2368 in medium
Tamoxifen in medium
Drug in medium
Gabriele CrucianiEFSA Parma, 13-15 June 2017
564,2558
Medium
563,2183
Medium
358,2353
Medium
& lipids
374,2368
Medium
NOT
observed
374,2368
Medium
lipids
Ultra fast automatic MetID
Gabriele CrucianiEFSA Parma, 13-15 June 2017
• Analyze overall toxicity profiles• Profile characteristics
• Unsupervised and supervised approaches to compare profiles
• Focus on individual endpoints• Correlate to external data
• Build an understanding of clinical mechanisms
36
What Can We Do With LipoTox Profile Data?
Gabriele CrucianiEFSA Parma, 13-15 June 2017
• Compound characterization
• ADME Property profiles
• Pathways, possible clinical indications
• Mechanism of action
• Unexpected off-targets (toxicity)
• Support therapeutic hypotheses
• Compare to competitor molecules, clinical standards of care
• Identify translational biomarkers
37
Applications
Gabriele CrucianiEFSA Parma, 13-15 June 2017
• Challenges for studying drug combinations:
• System may include more drugs
• Suitably robust to capture combination effects
38
Drug Combinations
Gabriele CrucianiEFSA Parma, 13-15 June 2017
• Lipid profiling in human (or animal) 3D-microtissues generates property profiles (close to disease process) that can be used to:
• Group chemicals into bio-activity classes
• Generate mechanicistic hypotheses
• Identify properties that may correlate with in vivo outcomes
• High throughput in vitro data is most informative when combined with available external information
• Known drugs as reference
• In vivo properties
LipoTox gives biological insight that may be crucial to improve drugs and to reduce time to market substantially.
Summary
Gabriele CrucianiEFSA Parma, 13-15 June 2017
Alessandra Di Veroli
Aurora Valeri
Fabien Fontain
Laura Goracci
Massimo Baroni
Paolo Benedetti
Paolo Tiberi
Roberto Pellegrino
Sara Tortorella
Stefano di Bona
Robert Stanton
Simone Sciabola
Molecular Discovery
Thanks to …
Gabriele CrucianiEFSA Parma, 13-15 June 2017
Rosso di sera .. bel tempo si spera !!
Red sky at night, shepherd’s delight
Thanks all …