Transcript Presentation - University of Warwick

Life or Cell Death:
Deciphering c-Myc Regulated Gene
Networks In Two Distinct Tissues
Sam Robson
MOAC DTC, Coventry House, University of Warwick,
Gibbet Hill Road, Coventry, CV4 7AL
[email protected]
Mco
A
Talk Outline
1.
2.
3.
4.
5.
6.
7.
Introduction
Skin vs. Pancreas
Methods
Quality Control
Results
Linear Models
Conclusions
Project Aims
•
•
•
•
Analyse differences in gene-expression
between two models of c-Myc activation with
distinctly opposing phenotypes
Identify c-Myc targets that promote cell
replication/survival and apoptotic cell death to
help understand dual potential of c-Myc
To improve understanding of the complex
activity of c-Myc in diseases such as cancer
To understand how c-Myc regulates vastly
different paradoxical phenotypes in vivo
1: Introduction
• Transcription factor- wide range of
cellular functions
• “Master regulator of genes”
• Deregulated in majority of human
cancers
• Possible therapeutic target?
• We know WHAT c-Myc does, but we
want to know WHY it does it
c-Myc Regulated Processes
Angiogenesis
External
Signals
Growth
c-Myc
Loss of Differentiation
(eg. mitogens,
survival factors)
Apoptosis
Proliferation
2: Skin vs. Pancreas
• Controlled activation of c-Myc in
target cells of adult mice in vivo
• Targetted to pancreatic islet β-cells
(insulin promoter) and skin suprabasal keratinocytes (involucrin
promoter)
• Opposing phenotypic outcomes
TAM
c-MycER
Activation
Inactive
Skin
Suprabasal
layer
Pancreas
Pelengaris et al. (1999), Molecular Cell, Vol. 3(5), 565-577
Pelengaris et al. (2002), Cell, Vol. 109(3), 321-334
Active
Suprabasal
layer
TAM
c-MycER
Activation
• Skin
Unchecked proliferation, no
apoptosis - Replication
• Pancreas
Synchronous cell cycle entry
and apoptosis – Death
• Myc activation regulates
two opposing phenotypes
Pelengaris et al. (1999), Molecular Cell, Vol. 3(5), 565-577
Pelengaris et al. (2002), Cell, Vol. 109(3), 321-334
3: Methods
1: Treatment of
Transgenics
2: Extraction of
Tissue
Controlled activation of
c-Myc in two diverse
tissues
6: Microarray
Hybridisation
3: Laser
Capture
Microdissection
Excision of target tissue
QC
5: 2-Cycle IVT
Preparation of cRNA for
microarray hybridisation
Hybridise fragmented,
labelled cRNA to
microarrays
Isolation of homogenous
tissue
QC
4: RNA
Extraction
Isolate total RNA from
target cells
QC
7: Microarray
Data Analysis
Analysis of microarray
data
QC
8: Validation
Studies
9: Functional
Validation
Validation studies to
confirm results
Linking results to the
biology of the system
Experimental Setup
Myc OFF
Myc ON
Gene
Expression
Skin
Tissue
Time course n=3
Gene
Expression
Time course n=3
4hr
8hr
16hr
4hr
32hr
16hr
32hr
Time course n=3
Gene
Expression
Pancreas
Tissue
Gene
Expression
Time course n=3
8hr
4hr
8hr
16hr
32hr
4hr
8hr
16hr
32hr
Laser Capture Microdissection
Laser
LCM
Eppendorf
Tube
Membrane
Slide
Tissue
Glass Slide
Support
• β-cells make up only ~2% of
pancreas
• Interesting changes masked
by changes in non-Myc
encoding cells
• LCM allows isolation of
homogenous cell populations
• Nikon SL Microcut LCM
system
Laser Capture Microdissection
1: Find Islet
2: Cut Islet
3: Lift Islet
4: Extracted Islet
Laser Capture Microdissection
1: Find Islet
2: Cut Islet
3: Lift Islet
4: Extracted Islet
Laser Capture Microdissection
1: Find Islet
2: Cut Islet
3: Lift Islet
4: Extracted Islet
LCM Optimisation - Pancreas
Control
section
18s and
28s peaks
nonexistent
18s and
28s peaks
nonexistent
18s and
28s peaks
nonexistent
• Pancreas has large number of RNases
• RNA Integrity decreases rapidly using standard LCM
protocols (Arcturus Pixcell, PALM MicroBeam, etc.)
• RNA fully degraded before reaching laser capture platform
LCM Optimisation - Pancreas
• All slides washed and baked before use
• 100% ethanol used at all times
• All surfaces cleaned thoroughly, gloves
worn at all times, etc.
• Don’t let tissue thaw (fix in ice-cold 100%
ethanol)
• Very quick staining (10 secs with 1%
toluidine blue in 100% ethanol)
• Air-dry sections in dessicator
• Limit time on LCM platform (10 mins max)
LCM Optimisation - Skin
•
•
•
•
Difficult to cut – Strong cell-cell bonds
Slow laser burns and damages cells
Difficult to lift from surrounding tissue
Area of captured tissue much smaller
than for islets
• Difficult to obtain sufficient RNA yield
• LCM not essential for skin compared to
pancreas – decided to stick to whole
sections
4: Quality Control
• Quality control at every stage:
I. RNA extraction
II. IVT
III.Microarray hybridisation
IV.Probe-level
V. Post-normalisation
Minimising Error
• Technical Error
– Same person performs all procedures
– Meticulous planning
– Standardise all protocols used
– Randomisation of batches
• Biological Error
– Inbred transgenic mice used
– All male
– All same age
– All culled at same time of day
Quality Control
• Quality control at every stage:
I. RNA extraction
II. IVT
III.Microarray hybridisation
IV.Probe-level
V. Post-normalisation
+/+ 4 (1)
+/+ 4 (2)
+/+ 4 (3)
+/+ 8 (1)
+/+ 8 (2)
+/+ 8 (3)
+/+ 16 (1)
+/+ 16 (2)
+/+ 16 (3)
+/+ 32 (1)
+/+ 32 (2)
+/+ 32 (3)
+/+ 72 (1)
+/+ 72 (2)
+/+ 72 (3)
+/- 4 (1)
+/- 4 (2)
+/- 4 (3)
+/- 8 (1)
+/- 8 (2)
+/- 8 (3)
+/- 16 (1)
+/- 16 (2)
+/- 16 (3)
+/- 32 (1)
+/- 32 (2)
+/- 32 (3)
+/- 72 (1)
+/- 72 (2)
+/- 72 (3)
-/+ 4 (1)
-/+ 4 (2)
-/+ 4 (3)
-/+ 8 (1)
-/+ 8 (2)
-/+ 8 (3)
-/+ 16 (1)
-/+ 16 (2)
-/+ 16 (3)
-/+ 32 (1)
-/+ 32 (2)
-/+ 32 (3)
-/+ 72 (1)
-/+ 72 (2)
-/+ 72 (3)
-/- 4 (1)
-/- 4 (2)
-/- 4 (3)
-/- 8 (1)
-/- 8 (2)
-/- 8 (3)
-/- 16 (1)
-/- 16 (2)
-/- 16 (3)
-/- 32 (1)
-/- 32 (2)
-/- 32 (3)
-/- 72 (1)
-/- 72 (2)
-/- 72 (3)
4h 1 T
4h 2 T
4h 3 T
4h 1 U
4h 2 U
4h 3 U
8h 1 T
8h 2 T
8h 3 T
8h 1 U
8h 2 U
8h 3 U
16h 1 T
16h 2 T
16h 3 T
16h 1 U
16h 2 U
16h 3 U
32h 1 T
32h 2 T
32h 3 T
32h 1 U
32h 2 U
32h 3 U
RIN
RNA Integrity
10.0
9.0
8.0
7.0
6.0
5.0
4.0
3.0
2.0
1.0
0.0
RNA quality lower than threshold (RIN < 5)
Quality Control
• Quality control at every stage:
I. RNA extraction
II. IVT
III.Microarray hybridisation
IV.Probe-level
V. Post-normalisation
nd
2
Round cRNA Yield
2nd round cRNA yield
ug
80
70
60
50
Standard Protocol
40
Double Volume Protocol
30
20
10
0
P60 P58 S9 S6 P41 S1 S10 S2 P42 S15 P44 P12 P28 S17 S16 P52 P25 P2 P27 P11 S11 P16 P6
Sample name
• Standard protocol results in low cRNA
yield – Possible contaminant?
• Using double volume vastly improves yield
– Contaminant diluted?
+/+ 4 (1)
+/+ 4 (2)
+/+ 4 (3)
+/+ 8 (1)
+/+ 8 (2)
+/+ 8 (3)
+/+ 16 (1)
+/+ 16 (2)
+/+ 16 (3)
+/+ 32 (1)
+/+ 32 (2)
+/+ 32 (3)
+/+ 72 (1)
+/+ 72 (2)
+/+ 72 (3)
+/- 4 (1)
+/- 4 (2)
+/- 4 (3)
+/- 8 (1)
+/- 8 (2)
+/- 8 (3)
+/- 16 (1)
+/- 16 (2)
+/- 16 (3)
+/- 32 (1)
+/- 32 (2)
+/- 32 (3)
+/- 72 (1)
+/- 72 (2)
+/- 72 (3)
-/+ 4 (1)
-/+ 4 (2)
-/+ 4 (3)
-/+ 8 (1)
-/+ 8 (2)
-/+ 8 (3)
-/+ 16 (1)
-/+ 16 (2)
-/+ 16 (3)
-/+ 32 (1)
-/+ 32 (2)
-/+ 32 (3)
-/+ 72 (1)
-/+ 72 (2)
-/+ 72 (3)
-/- 4 (1)
-/- 4 (2)
-/- 4 (3)
-/- 8 (1)
-/- 8 (2)
-/- 8 (3)
-/- 16 (1)
-/- 16 (2)
-/- 16 (3)
-/- 32 (1)
-/- 32 (2)
-/- 32 (3)
-/- 72 (1)
-/- 72 (2)
-/- 72 (3)
4h 1 T
4h 2 T
4h 3 T
4h 1 U
4h 2 U
4h 3 U
8h 1 T
8h 2 T
8h 3 T
8h 1 U
8h 2 U
8h 3 U
16h 1 T
16h 2 T
16h 3 T
16h 1 U
16h 2 U
16h 3 U
32h 1 T
32h 2 T
32h 3 T
32h 1 U
32h 2 U
32h 3 U
2nd round cRNA yield
ug
nd
2
Round cRNA Yield
100
90
80
70
60
50
40
30
20
10
0
Yield lower than recommended cutoff of 10 µg
Effect of RNA Quality on Yield
2nd round cRNA yield
ug
100
90
80
70
60
50
40
30
20
10
0
3
4
5
6
7
8
9
RIN
• General trend between RNA quality and 2nd round cRNA
yield – Weakly correlated
• Low RIN does not necessarily mean poor yield
• High RIN samples can still give low yield
• RIN cannot accurately predict yield
10
Effect of RNA Quality on Yield
• Skin samples have higher RNA quality and yield than
pancreas samples
• Many differences between skin and pancreas
– Greater ribonuclease activity in pancreas
– More intense processing for pancreas tissue RNA compared to skin
Effect of RNA Extraction Batch
Effect of 2-cycle IVT Batch
Quality Control
• Quality control at every stage:
I. RNA extraction
II. IVT
III.Microarray hybridisation
IV.Probe-level
V. Post-normalisation
Percent Present
Low 2nd round cRNA yield (< 10 µg)
Percent Present vs RNA Quality
Percent Present vs RNA Yield
+/+ 4 (1)
+/+ 4 (2)
+/+ 4 (3)
+/+ 8 (1)
+/+ 8 (2)
+/+ 8 (3)
+/+ 16 (1)
+/+ 16 (2)
+/+ 16 (3)
+/+ 32 (1)
+/+ 32 (2)
+/+ 32 (3)
+/+ 72 (1)
+/+ 72 (2)
+/+ 72 (3)
+/- 4 (1)
+/- 4 (2)
+/- 4 (3)
+/- 8 (1)
+/- 8 (2)
+/- 8 (3)
+/- 16 (1)
+/- 16 (2)
+/- 16 (3)
+/- 32 (1)
+/- 32 (2)
+/- 32 (3)
+/- 72 (1)
+/- 72 (2)
+/- 72 (3)
-/+ 4 (1)
-/+ 4 (2)
-/+ 4 (3)
-/+ 8 (1)
-/+ 8 (2)
-/+ 8 (3)
-/+ 16 (1)
-/+ 16 (2)
-/+ 16 (3)
-/+ 32 (1)
-/+ 32 (2)
-/+ 32 (3)
-/+ 72 (1)
-/+ 72 (2)
-/+ 72 (3)
-/- 4 (1)
-/- 4 (2)
-/- 4 (3)
-/- 8 (1)
-/- 8 (2)
-/- 8 (3)
-/- 16 (1)
-/- 16 (2)
-/- 16 (3)
-/- 32 (1)
-/- 32 (2)
-/- 32 (3)
-/- 72 (1)
-/- 72 (2)
-/- 72 (3)
4h 1 T
4h 2 T
4h 3 T
4h 1 U
4h 2 U
4h 3 U
8h 1 T
8h 2 T
8h 3 T
8h 1 U
8h 2 U
8h 3 U
16h 1 T
16h 2 T
16h 3 T
16h 1 U
16h 2 U
16h 3 U
32h 1 T
32h 2 T
32h 3 T
32h 1 U
32h 2 U
32h 3 U
Scale Factor
Scale Factor
50
45
40
35
30
25
20
15
10
5
0
Low 2nd round cRNA yield (< 10 µg)
Quality Control
• Quality control at every stage:
I. RNA extraction
II. IVT
III.Microarray hybridisation
IV.Probe-level
V. Post-normalisation
Probe Level Models
Example 1
Example 2
Example 3
.CEL File
PLM Residuals
Image
• Pseudo-images of PLM summary - accounts for
strong probe effects
• Can see artefacts that may be otherwise hidden
• Package affyPLM in Bioconductor in R
Data Distribution – Pre-Normalised
Pancreas
Myc OFF
Myc ON
Skin
Quality Control
• Quality control at every stage:
I. RNA extraction
II. IVT
III.Microarray hybridisation
IV.Probe-level
V. Post-normalisation
Skin vs Pancreas
Skin
Pancreas
• Clustering – Group
similar samples
together
• Branching tree like
structure – samples
on the same branch
most similar
• Data cluster nicely on
tissue (some outliers)
• Given the protocol,
the data looks great!
Outliers
Data Distribution – Post-Normalised
Pancreas
Myc OFF
Myc ON
Skin
Removal of Outliers
Outliers to be removed
• Outliers tended to be poor across all QC tests
• Good pancreas samples as outliers?
• Remove or keep?
Effect of RNA Quality on Outliers
Sample QC Penalty
25
20
15
Pancreas
10
Skin
5
0
0
1
2
3
4
5
6
7
8
9
10
RIN
• Outliers have wide range of RINs
• Only one of four RIN < 5 samples classed as an outlier
• Low RIN samples can produce good reproducible data
Effect of RNA Yield on Outliers
Sample QC Penalty
25
20
15
Pancreas
10
Skin
5
0
0
10
20
30
40
50
60
70
80
90
100
2nd Round cRNA Yield
ug
• Outliers have wide range of 2nd round cRNA yields
• Only 4 of the outliers had low cRNA yields
• Good quality data with less than 4 µg cRNA!
5: Results
• Early time point analysis – looking
for direct effects of c-Myc activation
• Untreated versus Treated
• Combined 4 and 8 hour time points
• Analysis in Genespring GX
• Two-way ANOVA on tissue type
and treatment, p = 0.05
5: Results
Gene
Pancreas
Skin
Comments
Insulin
↓ 4-fold
nc
Involucrin
nc
↓ 2-fold
Tissue-specific
differentiation markers
down
Cyclin D2
nc
↑ 2-fold
CDK4
nc
↑ 4-fold
Cyclin E
↑ 4-fold
nc
p27KIP1
↓ 2-fold
↓ 4-fold
p19ARF
↑ 2-fold
nc
ODC
↑ 2-fold
nc
Fas receptor
↑ 6-fold
nc
Cell cycle progressors
up & repressors down
(some anomalies)
Apoptotic markers up in
pancreas
6: Linear Models
• Close collaboration with Agilent Technologies for
linear modelling of microarray data
• Part of multidisciplinary PhD - MOAC
• Bioconductor package in R, GUI in Tcl/Tk
• Implementation in GeneSpring GX (Agilent)
• Simple to use for non-statisticians
• Work in Progress - currently testing the program
on a number of diverse data sets
Linear Models
• Can be used in the following ways:
1.To ensure superfluous parameters have
minimal effect on gene expression (eg
batching effects)
2.To find interesting parameters
3.To find genes that change based on
interesting parameters whilst taking other
parameters and interactions into account
(eg clinical data)
Linear Models vs ANOVA
Parameter 1
Interaction
Parameter 2
LM
ANOVA
• LM with 2 factors equivalent to 2-way ANOVA
• Comparable results to ANOVA
• Found a few genes (163) more than ANOVA for
parameter 2 – borderline p-values
7: Conclusions - What have I learnt?
• Talk to people (Affymetrix, UKAffy, GeneSpring
Users Group, R/Bioconductor community,
conferences, colleagues, etc.)
• Randomise everything
• Keep things really, really clean (be paranoid!)
• Plan everything with military precision
• Minimum of four replicates if possible
• You may still be able to get good results from poor
quality RNA
• Using double volume of reagents in 1st cycle of
IVT reaction can increase overall cRNA yield
Further Work
• Analysis of microarray data
• Use of LM tool and comparison of results
with standard methods (ANOVA)
• Validation of results –
Immunohistochemistry, quantitative real
time PCR, western blots, etc.
• Functional validation – siRNA, ChIP-onchip, etc.
• Thesis…
Acknowledgements
Project Supervisors:
Michael Khan
David Epstein
Stella Pelengaris
Special Thanks:
Sheena Lee
Paul Heath
Geoff Scopes
Giorgia Riboldi-Tunnicliffe
Collaborators:
Helen Brown
Lesley Ward
Sue Davis
Heather Turner
Ewan Hunter
Sponsors:
EPSRC, BBSRC, AICR, Eli Lilly and Amylin Pharmaceuticals Inc.
Mco
A
Acknowledgements
Outline of Talk
1.
2.
3.
Introduction
Skin vs. pancreas
Methods
i.
ii.
iii.
iv.
v.
4.
5.
Overview and technical considerations
Laser Capture Microscopy
mRNA Isolation
Microarray Hybridisation
Microarray Analysis
Results
General linear models
RNA Integrity
Poor quality:
Okay quality:
Majority of peaks
at lower levels
18S and 28S peaks
more prominent, but
many peaks at
lower levels
Good quality:
Excellent quality:
Fewer peaks at
lower levels
18S and 28S peaks
clear with almost no
peaks at lower levels
Cell-Cycle Progression
Ub
MYC MAX
CCND2
CDK4
Cyclin D2 CDK4
CACGTG
Proteosome
E-Box sequence in
promoter sequence
of target gene
CAK
p27KIP1
Cyclin E
CDK2
Active
Inactive
P
MIZ-1
MYC MAX
Sp1/Sp3
CUL1
CKS
MYC
p15Ink4b (CDKN2B)
p21Waf1 (CDKN1A)
Cell-Cycle
Entry
Apoptosis – Cell Death
FAS Ligand
FAS “Death
Receptor”
Death Induced Signalling Complex (DISC)
FADD
BID
BCL-2
Apoptosome
Procaspase 8
Caspase
Cascade
Effector
ARF
caspases
FLIP
BAX/
BAK
tBID
APAF-1
MOMP
c-Myc
Procaspase 9
Cytochrome c
Smac
DIABLO
Mitochondrion
ATP
IAPs
BIM
IAPs
PUMA
Apoptosis
p53
AIF
NOXA
Endo G
Cellular
targets
Omi/
Htra2
Effector caspases
Scatterplots
Example 1
R2
R1
R3
R3
Example 2
R2
R1
R1
R3
R2
R1
R2
R3
Skin vs Pancreas
Skin
Pancreas
• Clustering – Group
similar samples
together
• Branching tree like
structure – samples
on the same branch
most similar
• Data cluster nicely on
tissue (some outliers)
• Given the protocol,
the data looks great!
Tissue-Specific Differentiation Markers
Insulin
~4-fold down
in pancreas
Involucrin
~2-fold down
in skin
Cell-Cycle Progression
Cyclin D2
~2-fold up E
Cyclin
in skin
~4-fold up
in pancreas
• Ccnd2 and CDK4 upregulated in skin –
Indicates G1/S cell cycle progression
• No change in pancreas – Odd
• CDK inhibitor p27 downregulated in both
• Cyclin E upregulated in pancreas and
not skin – Again, very odd
CDK4
~4-fold up
in skin
p27KIP1
~2-fold down in pancreas
~4-fold down in skin
Apoptosis
p19ARF
~2-fold up
in pancreas
•
Increase in p19 – Oncogenic stress (p53
dependent pathway)
•
No change in p53 at transcriptional level –
Changes may occur at protein level
•
Massive increase in Fas receptor
expression – Extrinsic pathway
•
Myc seems to drive apotosis through
extrinsic and intrinsic pathways
Fas Receptor
~6-fold up
in pancreas
p53
No change
Linear Models
• Unsupervised linear regressive technique.
• Model gene-expression data as a linear
combination of parameter variables:
y  b1 x1  b2 x2  ...  b p x p  
y = (y1,…,yn)T is the response variable (gene-expression) for each sample
xi = (x1,…,xn)T are the explanatory variables (1 ≤ i ≤ p) for each sample
bi is the model coefficient for explanatory variable xi
n is the number of samples, p is the number of parameters
ε is some error term
c-Myc Regulated Processes
Angiogenesis
External
Signals
Growth
Loss of
Differentiation
c-Myc
(eg. mitogens,
survival factors)
Proliferation
Apoptosis
Effect of RNA Quality on Outliers
Sample QC Penalty
25
20
15
Pancreas
10
Skin
5
0
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
10.0
RIN
• Outliers have wide range of RINs
• Only one low RIN sample (RIN < 5) classed as an outlier
• Low RIN samples can produce reproducible data
Effect of RNA Yield on Outliers
Sample QC Penalty
25
20
15
Pancreas
10
Skin
5
0
0
10
20
30
40
50
60
70
80
90
100
2nd Round cRNA Yield
ug
• Outliers have wide range of 2nd round cRNA yields
• Only 4 of the outliers had low 2nd round cRNA
• Good quality data with less than 4 µg 2nd round cRNA!
Laser Capture Microdissection
• Heterogeneity of tissue may
cause problems
• β-cells make up only ~2% of
pancreas
• Interesting changes masked
by changes in non-Myc
encoding cells
• LCM allows isolation of
homogenous cell populations
Laser Capture Microdissection
Laser
LCM
Eppendorf
Tube
Membrane
Slide
Tissue
Glass Slide
Support
• Heterogeneity of tissue may
cause problems
• β-cells make up only ~2% of
pancreas
• Interesting changes masked
by changes in non-Myc
encoding cells
• LCM allows isolation of
homogenous cell populations