Transcript Folie 1 - Chilealimentos

Max Planck Institute
of Molecular Plant Physiology
http://www.mpimp-golm.mpg.de
Nature 433, 6th January 2005, p.12
.. journey southwest
from Berlin to Golm,
a small village near
Potsdam, is a 90minute train trip to
the end of the world …
Thank You
………. outside Potsdam
the only view from the
window is farmland
stretching to the
horizon, until an ultramodern glass building
looms out of the fog.
Structure of MPI-MP
3 departments
comprising 12 departmental research groups
mostly led by young scientists,
all now qualified as ‚Research Group Leader‘
2 independent junior research groups, funded by the MPG
3 central infrastructure groups (expression profiling, metabolite profiling, bioinformatics)
3 service units (plant growth, microscopy, biophysics)
2 university guest groups + 2 GoFORSYS guest groups
Horizontal structure with independence for groups led by (time-limited) young scientists
Personnel 2010
Admin.
service
Women in science:
Directors &
Group Leaders
- 52 % of the PhD students
- 34 % of the post-docs
5% 5%
- 18
% of
the group leaders
Directors and
Group
Leaders
Scientific
service
19%
Postdocs
33%
Internationality:
PostPhD Students
Docs
Students - 39 % of the PhD students
- 65 % of the post-docs
Students 12%
Scientific Service
-masters
-working
- 24 % of the group leaders
Administrative Service
26%
PhD
Students
Internationality
Internationality
is atmosphere
and environment,
Internationality:
- 39 % of the PhD students
- 65 % of the post-docs
- 24 % of the group leaders
not just counting heads
- 33% of the Directors
32 different countries …
English is the official institute language
Poland (20)
- all talks and seminars
- all information, operating instructions
safety and other papers
- the new institute intranet / sharepoint
- work contracts
South America (12),
Temporary accommodation
on Campus, or in Golm
Help where possible with the authorities
esp. Brazil, Argentina, Chile, Mexico
China (11)
India (7), Nepal (4)
Australia (3), New Zealand (2)
Macedonia (3)
The main biological question
What determines plant growth and composition?
Our entry point
Flowers
Seeds
Light
CO2
Nutrients
Water
Young
leaves
Metabolism
A complex integrated process
Many entry points for research
Developmental regulation
- Meristem activity
- Cell cycle
- Cell growth
Roots
and the biogenesis and use of
machinery that is needed to
turn metabolites into biomass
e.g. chloroplasts, ribosomes, cellulose synthase
….. using systems approaches
Biophysics
- Water movement
- Cell expansion
Whole plant allocation
- leaf area /unit biomass
- root volume /unit biomass
Evolution of Research Strategies
Time series
Multiple steady states
Bioinformatics
Huge amounts of
data
Environmental
conditions
Many different sorts
of data
Genetically differing plants
- Natural Diversity
- Gene Technology
nuclear
plastome
Multilevel
Analysis
Molecular traits
Metabolic traits
Physiological traits
Integrative traits
Basic understanding, Biotechnology, Breeding, Biomarkers
Systems
Biology
Rigorous conceptual
analysis
Metabolomics
• The metabolome comprises all small
molecules present in a given biological
system
• Metabolomics aims at the quantitative
determination of all small molecules
• The metabolome contains molecules
hugely varying in three parameters :
concentration, structure and chemical
behaviour
Data acquisition
DTD – GC/TOF: high throughput, high quality
~ 200 primary metabolites
robotic derivatisation & full extract injection:
no fractionation, no cross contamination, reproducibility <10%RSD
„What Is This Peak?“
Acquired on 09-Dec-1998 at 12: 59: 56
8343AO01
Scan EI+
T IC
1.20e8
100
%
0
7.500
10.000
12.500
15.000
17.500
20.000
22.500
25.000
27.500
30.000
32.500
35.000
37.500
40.000
42.500
rt
45.000
LTQ FT Ultra
•
•
•
Resolution
– > 1 000 000
Mass Range
– m/z 50-2000
Dynamic Range
– 1 000
Mass Accuracy
102 ppm
- 0.017
5.93 ppm
- 0.00103
Phenylalanine [M+H+]+
0.26 ppm
+ 0.00004
Formula calculation is depending on mass accuracy and resolution
[M+H+]+ Mass 181.07066
allowed
chemical elements:
C = 30
H = 50
N= 5
O = 10
P= 5
S= 5
500 ppm = 0.09085 Da Error  268 predicted Formulas
100 ppm = 0.01817 Da Error  54 predicted Formulas
10 ppm = 0.00181 Da Error 
6 predicted Formulas
1 ppm = 0.00018 Da Error 
1 predicted Formula 
C6H12O6
All-in-One-Extraction Procedure combined
with isotope labelling
Derivatisation
(prim. Metabolites)
13CO
2
GC-MS Analysis
20%
Aequous 80%
UPLC (C18) -MS
phase
(sec. Metabolites)
neg. mode
Organic
phase
neg. mode
pos. mode
MeOH:MTBE:H2O
15N
UPLC (C8) -MS
(Lipids)
pos. mode
Ammonium nitrate
Thermo Exactive
(Orbitrap)
plant tissue
Isotope labeling — Annotation
NL: 3.73E7
12c_s_pos_1#2315 RT:
9.20 AV: 1 T: FTMS {1,1}
+ p ESI Full ms
[100.00-1500.00]
756.55457
100
12C
50
Labelled Sample
554.51471 591.49866 628.50470
0
100
814.60767
716.56818
877.62915
NL: 2.02E7
13c_s_pos_1#2605 RT:
9.20 AV: 1 T: FTMS {1,1}
+ p ESI Full ms
[100.00-1500.00]
798.69470
13C
Labelled Sample
50
0
100
519.43634
15N
587.62531
628.62268
716.56122
756.70197
757.55157
819.67371 856.74835
NL: 6.57E7
15n_s_pos_1#2643 RT:
9.19 AV: 1 T: FTMS {1,1}
+ p ESI Full ms
[100.00-1500.00]
Labelled Sample
50
550
600
m/z 756.55457
C
0
100
H
0
200
O
0
20
N
0
10
S
0
10
P
0
10
779.53314 816.60785
717.56573
591.49915
0
500
650
700
m/z
235 hits
C42
750
800
878.62500
850
C42H84ON2P4
C H O NP
8 hits 42 79 8
C42H72O6N6
C42H82O2N3S3
C42H84O3N2P2S
C42H74O3N7S
C42H84O3P2S2
C42H80O5N2S2
N1
1 hit C42H79O8NP
Metabolic Profiling
• allows a rapid and simple discrimination
between genotypes, developmental and
environmental stages ( fingerprinting)
• allows functional analysis of genes with
respect to their influence on metabolic
composition
• is an indispensible level in systems
approaches
• allows identification of biomarkers
Why (Metabol)Omics?
DNA
RNA
Genomics
Protein
Metabolite
Proteomics
Metabolomics
Transcriptomics
Complex phenotype
Why (Metabol)Omics?
• Growth and performance of any biological
system are to a large extent (if not totally)
driven by its metabolic activity
• Metabolites are the last level of the
realization of genetic information
• This level is most near to the complex
phenotype ( in a linear thinking which is of
course not correct)
Why (Metabol)Omics?
• Biosynthesis (and degradation) of
metabolites is characterized by multiple
chemical transformations
• A + B -> C + D
• This means that metabolism by principle
represents a network
Why (Metabol)Omics?
• Most phenotypes are due to both linear
and epistatic contributions
• Genetic markers at first approximation can
only represent linear contributions
• Metabolites due to their inherent network
characteristic represent the action and
interaction of many gene products
• They should thus have the potential to
mirror also epistatic interactions
Metabolomics and Diagnostics
• Prediction of plant biomass
• Discrimination of cancer vs non-cancer tissues
• Patient categorization (personalized medicine)
• Prediction of type 2 diabetes
• Metabolomics as a measure in wine production
Biomass prediction of field
grown corn plants
• 300 corn inbred lines representing 3 maturity
groups were grown on two field sites
• Leaves from 10 plants of each genotype and
plot were harvested four weeks after
germination ( leaf size : approx. 5 cm)
• Metabolic profiles were run for each
genotype
• Biomass and flowering time were modeled
using the metabolite data following a random
forest approach
Discrimination of maturity groups in
corn
Metabolic profiles have a high
diagnostic power for complex traits
such as flowering time and early and
late biomass
Division of the entire data set into a training set and a test set
allows the prediction of biomass
Metabolomics and Diagnostics
• Prediction of plant biomass
• Discrimination of cancer vs non-cancer tissues
• Patient categorization (personalized medicine)
• Prediction of type 2 diabetes
• Metabolomics as a measure in wine production
Metabolic profiling allows assigment of
normal and clear cell kidney carcinoma
class
assignment
s
Control
(normal)
Clear-cell
RCC
RCC
(other)
Control
(normal)
66
0
0
Clear-cell
RCC
1
60
0
RCC
(other)
2
2
1
Metabolic Signature in CSF of Depressive Patients Accurately
Predicts Antidepressant Treatment Response
Tanja Gärtner, Joachim Selbig, Abdelhalim Larhlimi, Patrick Giavalisco,
Gareth Catchpole, Christian Namendorf, Lothar Willmitzer, Manfred Uhr,
Florian Holsboer
• Classification performance in strict cross
validation
•
•
•
•
Accuracy 88 %
Sensitivity 82 %
Specificity 93 %
False-discovery rate 8 %
Predicting Fasting Plasma Glucose Level Development
using Human Metabolic Profiles
Manuela Hische, Abdelhalim Larhlimi, Gareth S Catchpole, Andreas FH Pfeiffer, Lothar
Willmitzer, Joachim Selbig & Joachim Spranger
Charite Berlin, Berlin,
University of Potsdam,
Max-Planck-Institute for Molecular Plant Physiology, Potsdam, German Institute of Human Nutrition,
Potsdam
Results of Classification: Established risk markers are: gender, waist circumference,
BMI, age and baseline fasting glucose levels.
Variables
Specificity
Sensitivity
Accuracy
Metabolites
0.70
0.67
0.68
Established markers
0.56
0.50
0.53
Metabolites +
Established markers
0.70
0.67
0.68
Metabolomics and Diagnostics
• Prediction of plant biomass
• Discrimination of cancer vs non-cancer tissues
• Patient categorization (personalized medicine)
• Prediction of type 2 diabetes
• Metabolomics as a measure in wine production
Problems
 The wine market is highly fragmented
with respect to quality criteria and
assurance
 This leads to uncertainty concerning
quality, origin, year and variety
(authenticity)
 The problem is rooted in the absence
of any objective technology
32
Some first principles :
The sum of its compounds reflects history
and determines taste and quality of each
wine
The sum of its compounds is equivalent to its
metabolic composition
33
The workflow
RT: 0.00 - 15.00
UPLC/LTQ-Orbitrap
6.89
80
6.00
60
5.32
4.83
3.50
40
7.91
6.08
4.61
20
Relative Abundance
NL:
9.47E6
Base Peak
MS
R_A_Carm_
05_4_neg
3.30
100
1.15
1.33
1.93 2.66
0
100
7.82
7.26
8.51
8.87
10.20
11.09 11.78 12.45
14.14
NL:
1.10E7
Base Peak
MS
r_ct_05_2_n
eg
6.91
80
3.30
60
40
20
1.24 1.32
1.69
3.10
3.49
4.19
0
100
5.33
6.02
4.83
6.10
7.72
8.50
8.88
10.20
11.06
12.31
12.73
14.43
NL:
1.08E7
Base Peak
MS
r_a_syrah_0
5_4_neg
4.61
80
60
6.01
3.32
40
20
1.15
1.33
1.71
3.50
4.52
5.32
8.50
9.51
3.11
0
100
9.70
10.89 11.32 12.21
12.51
14.47
NL:
7.98E6
Base Peak
MS
v_ta_m_06_
2_neg
6.90
80
5.32
5.99
60
3.33
40
Sample analysis
7.92
7.08
6.89
1.26 1.33
20
0.84
0
0
1
2.09
2
3.09
3
6.09
6.45
4.61
7.93
8.51
8.87 9.08 10.20 11.11 12.04 12.69
4.35
4
5
6
7
8
Time (min)
9
10
11
Metabolic
Mass spectrum
12
13
14.35
14
Data Analysis & Interpretation
Sample preparation
Wine samples
• Relative quantitative sample comparisons
• Determine statistical significance
• Identify metabolite(s) of interest
5
Biomarkers
Data Analysis & Interpretation/own system
• Identification of masses
• Potential BioMarkers for discrimination of wine samples
34
Cultivar Identification
a
CM
C
ME
CM
SY
S
b
i
ii
CS
ME
SY
i
ii
iii
iii
iv
iv
v
v
vi
vi
35
Vineyard Discrimination
Vineyard CT
Vineyard Vasco
Vineyard VSP
From the same country
36
Vintage ( production year) discrimination
Year 2004
Year 2005
Year 2006
37
Quality Discrimination
LDA of Wines (Negative Mod)
LDA of Wines (Positive Mod)
38
Metabolomics and Diagnostics
• Prediction of plant biomass
• Discrimination of cancer vs non-cancer tissues
• Patient categorization (personalized medicine)
• Prediction of type 2 diabetes
• Metabolomics as a measure in wine production