Transcript presentation

Computational biology of
cancer cell pathways
Modelling of cancer cell function
and response to therapy
Signalling pathways
Signal from
outside cell
Source:
Biocarta
database
Signal from
inside cell
Epidermal growth factor (EGF) signalling
Signal from
outside cell
Gene
expression
Signalling pathways
Signal from outside or
Receptors
inside cell
E.g., kinases
Signal transmitted to
genome
Changes in gene
expression = cell
response to signal
Information about
cell’s environment
and internal state is
coupled to gene
expression
Transcription
factors
Signalling pathways in cancer cells
Mutations in cancer
• Point mutations - changes
in protein sequence or
control sequences
• Genome instability: Loss or
gain of single genes or
chromosome portions
(many genes)
Abnormal amounts of
proteins
Abnormal function, e.g
always ‘switched on’ or
inactivated
Signalling pathways in cancer cells
Mutations in cancer
• Point mutations - changes
in protein sequence or
control sequences
• Genome instability: Loss or
gain of single genes or
chromosome portions
(many genes)
Abnormal amounts of
proteins
Abnormal function, e.g
always ‘switched on’ or
inactivated
Changes in information
processing underpin hallmarks of
cancer
Signalling pathways
in cancer cells
Epidermal growth factor receptor (EGFR)
• Overexpression of EGFR is common in many solid
tumours
• Correlates with increased metastasis, decreased
survival and a poor prognosis
• Protects malignant tumour cells from the cytotoxic
effects of chemotherapy and radiotherapy, making these
treatments less effective
EGFR is the target for several new anticancer therapies
EGFR-targeted therapy
cell surface portion
binds epidermal growth factor
CELL MEMBRANE
intracellular tyrosine kinase
transmit signal by phosphorylation
EGFR-targeted therapy
Therapeutic
antibody: Cetuximab
(colorectal cancer)
Small molecule
inhibitors
Inhibition of EGFR
• Both types of inhibitors
block signalling from the
EGF receptor
• Inhibition limits tumour
growth, dissemination,
angiogenesis
• Reduces resistance to
chemotherapy and
radiotherapy
• Aids the induction of cell
death (apoptosis)
Not a linear pathway, but a complex network
Genome
Not a linear pathway, but a complex network
Growth factor (EGF)
Receptor tyrosine kinase
PLC
ERK
Ras
PI3K
PKC
MAPK
PKB/Akt
TFs
Functional targets
CELL GROWTH AND PROLIFERATION
Cross-activation
by other
pathways
Not a linear pathway, but a complex network
Signal processing by the entire network and
mutations or expression changes in signal proteins
can limit response to therapy or cause side effects.
Complexity needs to be modelled in the
computer
Computer models of pathways
need biochemical kinetic data for
every connection.
Enable one to simulate
– the change in concentration
or
– activation (eg. phosphorylation)
of the proteins of the pathway
over time.
Modelling gives information on
how signals are processed.
The effect of the number of active EGFR
molecules on ERK activation
EGFR
PLC
PKC
ERK TFs
Ras
MAPK
PI3K
PKB/Akt
Functional targets
CELL GROWTH AND PROLIFERATION
The effect of the number of active EGFR
molecules on ERK activation
EGFR
PLC
PKC
ERK TFs
Ras
MAPK
500,000 active
receptors
PI3K
PKB/Akt
Functional targets
CELL GROWTH AND PROLIFERATION
50,000 active
receptors =
Inhibition by one order
of magnitude
Schoeberl et al., 2002, Nat. Biotech. 20: 370
The effect of active EGFR number on ERK
activation
500,000 active
receptors
50,000 active
receptors
Can this be achieved
by receptor
inactivation alone?
The effect of active EGFR number on ERK
activation
… Or, what might
happen if
• ERK is
overexpressed?
• Several proteins
in the pathway
are abnormally
expressed?
The effect of active EGFR number on ERK
activation
50,000 active
receptors
with normal levels of
ERK
or
ERK overexpression
and cross-activation
Computer models of pathways
• Integration of complex knowledge
– Biological processes are mediated by pathways in health
and disease
– Modelling aids integrative understanding of relationships
between physiology and clinical observations and the
molecular level
• Analysis and simulation of signal processing in cancer cells
– Effects of mutations and abnormal gene expression
– Discovery of new targets for therapy: key modulators of
pathway function
– Effects of therapeutic inhibitors
– Possible side effects
Effects of abnormal gene expression
• Roughly 90% of human cancers are epithelial in origin and
exhibit a large number of changes in the structure and function
of the genome.
• Abnormal expression levels can be observed for a a large
number of genes.
• This complexity might be the reason for the clinical diversity of
tumours (even with similar histology).
• A comprehensive analysis of the multiple genetic alterations
present is required for an understanding of abnormal signal
processing in cancer and differences between tumours.
Gene expression analysis
• The use of expression microarrays enables the largescale analysis of mRNA expression (expression profiling)
in tumour samples.
• Expression profiling can be used to simultaneously
assess the expression of the entire human genome.
• mRNA concentration is used as a surrogate for protein
conc. - protein concentrations may be hypothetically
inferred.
Example: Gene expression profiles from
breast cancer patient samples
ER-positive breast
tumour subtype has a
distinct microarray
expression profile
ER = oestrogen
receptor
Role of ER and EGFR in anti-oestrogen
therapy
• Response to tamoxifen is dependent on ER expression.
• Overexpression of EGFR is associated with tamoxifen
resistance - EGFR as a target for therapy.
EGFR as a target for therapy
ER-positive tumours
genes
EGFR
• Differences in expression of EGFR and other proteins in the
network between patients?
• May account for different responses to therapy with EGFR
inhibitors.
Modelling of individual response
Gene expression of EGFR network genes in
each individual tumour
• Input in computer model
Approximation of
relative changes in
protein expression
Caveat: may be not
directly comparable –
direct measurement of
protein concentration
• Model signal processing in different
tumours in response to EGFR
inhibition
• Hypothesis generation re. response
• Validate with clinical response
Summary
Molecular changes in cancer are highly complex.
They affect signal processing in pathways and networks.
Changes in signal processing underpin hallmarks of
cancer.
Computer modelling gives information on how signals
are processed.
Modelling aids fundamental understanding of cancer,
discovery of new targets for therapy, prediction of effects
of therapy and possible side effects.