Transcript Introduction 2

Robert A. Weinberg
The Biology of Cancer
First Edition
Chapter 12:
Maintenance of Genomic Integrity
and the Development of Cancer
Copyright © Garland Science 2007
“The capacity to blunder slightly is the real marvel of DNA. Without this special
attribute we would still be anaerobic bacteria and there would be no music”.
Lewis Thomas
Over time DNA sequences are the the most fixed, unchangeable components of the cell;
most of its other parts are in constant flux.
It is the stability of DNA that underpins the most robust defenses against cancer. There
are so many defenses against cancer each requiring a rare mutational event to
overcome and involving several layers of cells that it is surprising the incidence of
cancer is as high as it is. Suggests that populations on the way to transformation
undergo a disproportionate increase in mutation rate-so called “mutator phenotype” or
they gain genetic instability.
How are tissues organized to minimize the progressive accumulation of
mutations?
Stem cell compartments- 0.1% to 1% of total cell mass.
Figure 12.3a The Biology of Cancer (© Garland Science 2007)
Figure 12.1 The Biology of Cancer (© Garland Science 2007)
Figure 12.2b The Biology of Cancer (© Garland Science 2007)
Figure 12.2a The Biology of Cancer (© Garland Science 2007)
Figure 12.2c The Biology of Cancer (© Garland Science 2007)
Figure 12.3b The Biology of Cancer (© Garland Science 2007)
Figure 12.3c The Biology of Cancer (© Garland Science 2007)
Figure 12.4 The Biology of Cancer (© Garland Science 2007)
Figure 12.5a The Biology of Cancer (© Garland Science 2007)
Figure 12.5b The Biology of Cancer (© Garland Science 2007)
Figure 12.5c The Biology of Cancer (© Garland Science 2007)
Major attraction of cancer stem cell theory is it provided an explanation for the failure of clinical therapy.
CSC renew themselves, are resistant to chemotherapy and radiotherapy, divide infrequently and therefore
survive long periods of dormancy, have the potential to metastasize and colonize distant sites.
Raises the prospect of more effective therapy designed to eradicate the “Achilles” heel of the tumor.
This may have led to greater enthusiasm for the theory than the data warranted.
First demonstration human stem cell was in leukemia:
Lapidot T. et al. A cell initiating human acute myeloid leukemia after transplantation
into nude mice. Nature 367: 645-648 (1994)
Most human AML cells fail to grow in culture or as xenografts suggesting that they
have a very limited proliferative potential. Dick et al. showed that AML subtypes could
be engrafted into nude mice, but only from CD34+ CD38- fractions. Cells homed to
bone marrow proliferated extensively in response to in vivo cytokine treatment.
Frequency of initiating cell was 1 per 106 tumor cells.
Al-Hajj et al. Prospective Identification of tumorigenic breast cancer cells. Proc. Natl. Acad. Sci. USA 100:
3983-3988 (2003)
As few as 100 CD44+ CD24- cells were tumorigenic whereas tens of thousands of cells of alternative phenotype
were not. Tumorigenic subpopulations could be serially passaged generating both CD44+ CD24- and more
differentiated progeny.
Identification of Cancer Stem Cell in Human Brain Tumors:
Sigh et al. Cancer Research 63: 5821-5828 (2003).
Brain tumor stem cells isolated from CD133+ cell fraction:
Markedly increased capacity for proliferation, self-renewal and differentiation.
Self-renewal capacity highest from more aggressive–medulloblastomas compared to low-grade
gliomas.
CD133+ cells could differentiate in culture into tumor cells that resembled those in the patient.
Physically separated two populations of cells
that differ in their cell surface antigens and
capacity to seed new tumors in-vivo. After
implantation CSC enriched populations
generate tumors that are no longer enriched
in CSC’s indicating CSC both renew and
differentiate into non-CSC progeny. Cancer
cells within a tumor exist in multiple states of
differentiation that show distinct tumorseeding properties. Whereas normal stem
cells are oligopotent CSC may not be.
Spheres differentiate to express immunophenotypes similar to the primary tumor
Efficient tumor formation by single human melanoma cells. Quintana E et al. Nature 456: 593-598 (2004)
Heterogeneity in Cancer: Cancer Stem Cells versus Clonal
Evolution. Shackelton M et al. Cell: 138: 822-829 (2009)
25% of cells in human melanomas have tumorigenic
potential and no markers found to identify that subset.
Melanoma may therefore not follow the CSC model.
Cancer Stem Cells: mirage or reality? Nature Medicine 15: 1010-1012 (2009)
Challenges to theory: reports that up to 25% of cells are CSC’s (Quintana Nature 456: 593-598
(2008). Kelly Science 317: 337 (2007). Cells fully capable of reconstituting a tumor are therefore not
rare and should not be considered stem cells.
CSC’s defined by ability to seed tumors in animal hosts, self-renew AND generate differentiated
progeny. CSC’s quantitated as numbers of cells required at limiting dilution to seed new tumors.
Initial reports stem cells a very minor component of total cancer cell population. However, nothing in
early reports precluded possibility that some tumors might contain much higher CSCs.
CSC representation-function of:
cell type, stromal microenvironment, accumulated somatic mutations and stage of malignant
progression. CSC in leukemias observed to vary 500-fold between patient samples. Undifferentiated
tumors higher CSC’s than differentiated. CSC differ considerably between cancer subtypes arising
from a single tissue.
Early studies demonstrate cancer cells exist in at least two alternative phenotypic states that show
very different tumor-seeding potential but do not provide evidence for their ratios.
PROBLEM: CSC evaluated in-vivo but hosts vary tremendously in ability to support individual tumors
(vascularization, ECM, growth factors, host immuno-competence). Any measure of CSC therefore
poorly quantitative and relative to animal model. Studies showing high CSC’s co-inoculation with
ECM, late-stage patient samples, genetic mutations that give rise to mouse tumors lacking
heterogeneity of human cancers.
Proportions of human CSC’s in human populations often underestimated because of residual
immuno-competence.
More de-differentiated tumors higher CSC’s-regulators of differentiation strong determinants of CSC
biology.
EMT in transformed mammary epithelial cells –highly enriched in CSC as determined by tumorseeding ability, mammosphere formation, cell surface markers. Similarly EMT induces changes in
immortalized, non-tumorigenic mammary epithelial cells.
Fractionation of naturally existing normal and neoplastic epithelial cells –cells with CSC surface
markers show multiple attributes of mesenchymal transdifferentiation and expression EMT
transcription factors.
Carcinoma cells at invasive edge of tumors undergo EMT under influence of contextual signals from
adjacent stroma. Reminiscent of embryogenesis TGF-beta and Wnt ligands induce cells to undergo
EMT.
Reversibility: CSC’s are therefore at variance with models of normal stem cells. Phenotypic plasticity.
Much greater than observed in normal stem cell populations. However, that plasticity might be
possible in normal tissues under certain extreme conditions.
Tumors likely to contain dynamic equilibrium between CSC and non-CSC and that equilibrium would
be regulated by contextual signals. Unlikely the balance is very dynamic. All evidence indicates that
CSC’s exist in a meta-stable state.
Interconvertibility of CSC’s by environmental signals may vary greatly between tumors. In part
answers paradox that in order for tumors to progress they must have mutation rates at least twice
above that of stem cells. Treatment implications: CSC arising by EMT are more resistant to therapy.
Current therapies target non-stem cells. This should limit progression to a nonmutable population but
opposite observed.
Perspective on cancer cell metastasis (Chaffer and Weinberg. Science 331: 15591564 2011)
Mutational Drivers and clonal dynamics
Darwinian Evolution- purposeless genetic variation of reproductive individuals united by common
descent with natural selection of the fittest variants.
Clones arise through selectively advantageous driver mutations with selectively neutral ‘passenger’
lesions and deleterious lesions. ‘Mutator’ lesions increase rate of other genetic changes. Microenvironmental changes alter the the fitness effects of those lesions.
The Cancer Ecosystem
Tissue ecosystems provide the adaptive landscapes for fitness selection. Complex dynamic states
with multiple components that influence cancer clone evolution. TGF-b is a cancer ecosystem
regulatory molecule. Cancer cells interact reciprocally with tissue habitat. Cancer cells remodel tissue
microenvironments and niches to their competitive advantage. Cancer clone expansion is controlled
by architectural constraints - sequestration of stem cells in crypts and need for external signals for
proliferation and survival.
Cancer genomics
2nd generation whole genome sequencing. Individual cancers contain hundreds to tens of thousands
of mutations and chromosomal alterations (largely neutral mutations arising from genetic instability).
Evolutionarily neutral alterations register in screens because they hitchhike on clonal expansions
driven by selectively advantageous alterations or by drift. Each cancer has an individually unique
genomic profile. Almost infinite variety of evolutionary trajectories and multiple combinations of driver
mutations.
Genome profiles underestimate complexity. Individual metastases are clonal in origin and can be
traced to the primary tumor.
Mutational Drivers and clonal dynamics
Darwinian Evolution- purposeless genetic variation of reproductive individuals united by common
descent with natural selection of the fittest variants.
Clones arise through selectively advantageous driver mutations with selectively neutral ‘passenger’
lesions and deleterious lesions. ‘Mutator’ lesions increase rate of other genetic changes. Microenvironmental changes alter the the fitness effects of those lesions.
The Cancer Ecosystem
Tissue ecosystems provide the adaptive landscapes for fitness selection. Complex dynamic states
with multiple components that influence cancer clone evolution. TGF-b is a cancer ecosystem
regulatory molecule. Cancer cells interact reciprocally with tissue habitat. Cancer cells remodel tissue
microenvironments and niches to their competitive advantage. Cancer clone expansion is controlled
by architectural constraints - sequestration of stem cells in crypts and need for external signals for
proliferation and survival.
Cancer genomics
2nd generation whole genome sequencing. Individual cancers contain hundreds to tens of thousands
of mutations and chromosomal alterations (largely neutral mutations arising from genetic instability).
Evolutionarily neutral alterations register in screens because they hitchhike on clonal expansions
driven by selectively advantageous alterations or by drift. Each cancer has an individually unique
genomic profile. Almost infinite variety of evolutionary trajectories and multiple combinations of driver
mutations.
Genome profiles underestimate complexity. Individual metastases are clonal in origin and can be
traced to the primary tumor.
Subclonal segregation of mutations and clonal architecture
Sequential acquisition of mutations and with successive sub-clonal dominance or sweeps.
Multiplexed single cell mutational (serial) analysis is the most appropriate way to examine clonal
architecture. Evidence shows that evolutionary trajectories are complex and branching. Selection to
fit different and separate natural habitats (Darwins finches).
Subclones may be mixed in the primary tissue but also occupy distinct territories. Level of diversity
within the subclonal structure can be measured and is a robust biomarker for predicting progression.
Units of selection and cancer stem cells
For natural selection to work a the level of the cell that cell has to have extensive replication potential.
Cancer stem cells have an extensive potential for self-renewal. Cancer stem cell theory predicts that
stem cells should evolve their genotype and phenotype concomitant with that of the cancer (following
therapy). Cancer progression should be accompanied by slective pressure for cells with the most
extensive self renewing capacity and at the expense of cells with the ability to differentiate. Loss of
TP53 DNA damage checkpoint seems to release stem cell like transcriptional signatures that lead to
enhahnced self renewal in mammosphere cultures. Cancer stem cells should be prone to genetic
variation. Cancer stem-cell restraint or elimination should be the goal of therapy but if cancer stem
cells are as genetically and epigenetically diverse as evolutionary considerations and experiments
indicate this would account for therapeutic failure. Intrinsically lowered susceptibility to irradiation and
chemotherapy coupled with genetic diversity.
Subclonal segregation of mutations and clonal architecture
Sequential acquisition of mutations and with successive sub-clonal dominance or sweeps.
Multiplexed single cell mutational (serial) analysis is the most appropriate way to examine clonal
architecture. Evidence shows that evolutionary trajectories are complex and branching. Selection to
fit different and separate natural habitats (Darwin’s finches).
Sub-clones may be mixed in the primary tissue but also occupy distinct territories. Level of diversity
within the sub-clonal structure can be measured and is a robust biomarker for predicting progression.
Units of selection and cancer stem cells
For natural selection to work a the level of the cell that cell has to have extensive replication potential.
Cancer stem cells have an extensive potential for self-renewal. Cancer stem cell theory predicts that
stem cells should evolve their genotype and phenotype concomitant with that of the cancer (following
therapy). Cancer progression should be accompanied by selective pressure for cells with the most
extensive self renewing capacity and at the expense of cells with the ability to differentiate. Loss of
TP53 DNA damage checkpoint seems to release stem cell like transcriptional signatures that lead to
enhanced self-renewal in mammosphere cultures. Cancer stem cells should be prone to genetic
variation. Cancer stem-cell restraint or elimination should be the goal of therapy but if cancer stem
cells are as genetically and epigenetically diverse as evolutionary considerations and experiments
indicate this would account for therapeutic failure. Intrinsically lowered susceptibility to irradiation and
chemotherapy coupled with genetic diversity.
Clonal Evolution in cancer
Tissue ecosystem - optimize multicellular function - restrain unregulated expansion.
Has to be mechanisms to permit regulated maintenance and restoration:
Cellular self-renewal
Stabilization of telomeres
Angiogenesis
Cell migration
Invasion
Micro-environmental mechanisms to constrain or allow these events.
Cancer represents the sequential and random searches for phenotypic solutions to over-come
intrinsic and environmental constraints.
Limited resources, architecture, and other constraints limit size of solid tumors at every stage.
Natural selection as in organisms takes place through competition for space and resources.
Therapeutic modalities that are genotoxic can lead to enhanced fitness of surviving cells.